**Meet the editors**

Er. Kumar has been working in the field of Environmental Engineering and Management in general and Solid and Hazardous Waste Management in particular at National Environmental Engineering Research Institute (NEERI), Council of Scientific and Industrial Research, Government of India. He has published many research papers in journal of international repute with high im-

pact factor covered referred journals and many papers are under review/ submission stage in various international journals. He is the Editorial Board member of International Journal of Environmental Monitoring and Assessment besides working as Guest Editors for many international journals in the domain of Environmental Engineering in general and solid waste management in particular.

Dr. Rakesh Kumar is a Chief Scientist and Head at Mumbai Zonal Centre of Council of Scientific and Industrial Research-National Environmental Engineering Research Institute (CSIR-NEERI). He has been closely associated with many prestigious studies carried out by different funding agencies in the domain of air quality monitoring and modeling which could lead toward poli-

cy formulation in this sector in India. He has numerous publications in the field of air pollution engineering. He is a member of various committees relating to air quality.

Contents

**Preface IX** 

Chapter 4 **Environmental Parameter** 

and Mark Ojogba Okolo

Chapter 1 **Relationship Between Fungal Contamination of Indoor** 

**Air and Health Problems of Some Residents in Jos 1**  Grace Mebi Ayanbimpe, Wapwera Samuel Danjuma

Chapter 2 **A Qualitative Study of Air Pollutants from Road Traffic 19** 

Kaushik K. Shandilya, Mukesh Khare and A. B. Gupta

**Effects on the Fate of a Chemical Slick 59** 

Chapter 5 **Indoor Air Quality and Thermal Comfort in Naturally Ventilated Low-Energy Residential Houses 79** 

Chapter 7 **A Methodology of Estimation on Air Pollution and** 

Shogo Sakamoto and Shigeru Kashima

Chapter 8 **Evaluation of Dry Atmospheric Deposition in** 

**Its Health Effects in Large Japanese Cities 145** 

**Two Sites in the Vicinity of Fuel Oil-Fired Power** 

Cerón-Bretón Rosa María,Cerón-Bretón Julia Griselda, Cárdenas-González Beatriz, Ortínez-Álvarez José Abraham, Carballo-Pat Carmen Guadalupe, Díaz-Morales Berenice

Gabriela Raducan and Ioan Stefanescu

Chapter 3 **Estimation of Ambient Air Quality in Delhi 37** 

Stéphane Le Floch, Mélanie Fuhrer, Pierre Slangen and Laurent Aprin

M. Krzaczek and J. Tejchman

Hermann Fromme

Chapter 6 **Particles in the Indoor Environment 117** 

Keiko Hirota, Satoshi Shibuya,

**Plants in Mexico 167** 

and Muriel-García Manuel

### Contents

#### **Preface XI**



### Preface

There is a great challenge for monitoring of Air quality specially for formulating abatement plans which will decrease the burden of environmental pollution. The management of polluted and fungal contaminated air is required to be dealt with different science and technological applications to reduce its impact on the environment.

There are many publications dealing with air quality monitoring and its health impact but still very few documents are available which could reveal the information relationship between fungal contamination of indoor air and health problems of some residuals. Microbial pollution is a key element of indoor air pollution. It is caused by hundreds of species of bacteria and fungi, in particular filamentous fungi (mould), growing indoors when sufficient moisture is available. In indoor environments, the air and surface fungal contaminations are influenced not only by temperature, relative humidity and air flux/drafting but also by the weather conditions, hygiene standards and human activity.

In the present book, chapters covering a methodology for estimation of air pollution and its health effects in large Asian cities, a qualitative study of air pollutants from road traffic and air pollution, air quality in low-energy buildings, particulates in indoor environment, evaluation of dry atmospheric deposition at sites in the vicinity of fuel oil fired power, PM 10 and its chemical composition, etc. are discussed in detail. Each chapter of the book carries valuable and updated information from basics to apex which help readers to understand more precisely.

#### **Sunil Kumar**

Scientist Council of Scientific and Industrial Research-National Environmental Engineering Research Institute (CSIR-NEERI), India

#### **Rakesh Kumar**

Chief Scientist and Head CSIR-NEERI Mumbai Zonal Laboratory India

**1** 

*Nigeria* 

**Relationship Between Fungal**

Grace Mebi Ayanbimpe1,

*3Department of Microbiology, Jos University Teaching Hospital Jos,* 

**Contamination of Indoor Air and Health** 

The problem of indoor air quality has in recent times attracted more concern than was the case several decades ago. Morey, (1999) presented some probable reasons why this is so. Construction materials have markedly changed from stones, wood and other 'natural' products to synthetic, pressed wood and amorphous cellulose products, some of which may be less resistant to microorganisms than the earlier ones. The indoor environment is invariably the closest to humans in terms of daily interaction, as a large proportion of man's time is spent indoors, especially in the modern cities and urban societies. There are indications however, that in many parts of the world, our homes, schools and offices are heavily contaminated with airborne molds and other biological contaminants. Biological contaminants of the indoor air include fungi, bacteria, viruses, pollen etc.; and many species of fungi are able to grow wherever there is moisture and an organic substrate. Building materials such as ceiling tiles, wood, paint, carpet rugs, etc., present very good environment for growth of fungi (Robin et al, 2007; Morey, 1999; Heseltine and Rosen, 2009). Water accumulation and moist dirty surfaces also encourage growth of fungi within a building, and contribute to human health problems. Fungi are very successful organisms capable of survival in diverse environments. This is possible because of their physiological versatility and genetic plasticity. Their spores are produced in large quantities and easily spread over a wide area. These spores can remain dormant for a long period of time during unfavorable conditions. Fungal spores and products of their metabolism are able to trigger allergic reactions which include hyper-sensitivity pneumonitis, allergic rhinitis and some types of asthma. Infections such as influenza, respiratory fungal infections, mycotoxicoses, and eye irritation by fungal volatile substances, among others, are health risks and challenges to susceptible individuals. Children, the elderly and immuno-compromised persons are at greater risks (Simoni

**1. Introduction** 

et al, 2003).

**Problems of Some Residents in Jos** 

Wapwera Samuel Danjuma2 and Mark Ojogba Okolo3 *1Department of Medical Microbiology University of Jos Nigeria, 2Department of Geography and Planning University of Jos,*

### **Relationship Between Fungal Contamination of Indoor Air and Health Problems of Some Residents in Jos**

Grace Mebi Ayanbimpe1,

Wapwera Samuel Danjuma2 and Mark Ojogba Okolo3 *1Department of Medical Microbiology University of Jos Nigeria, 2Department of Geography and Planning University of Jos, 3Department of Microbiology, Jos University Teaching Hospital Jos, Nigeria* 

#### **1. Introduction**

The problem of indoor air quality has in recent times attracted more concern than was the case several decades ago. Morey, (1999) presented some probable reasons why this is so. Construction materials have markedly changed from stones, wood and other 'natural' products to synthetic, pressed wood and amorphous cellulose products, some of which may be less resistant to microorganisms than the earlier ones. The indoor environment is invariably the closest to humans in terms of daily interaction, as a large proportion of man's time is spent indoors, especially in the modern cities and urban societies. There are indications however, that in many parts of the world, our homes, schools and offices are heavily contaminated with airborne molds and other biological contaminants. Biological contaminants of the indoor air include fungi, bacteria, viruses, pollen etc.; and many species of fungi are able to grow wherever there is moisture and an organic substrate. Building materials such as ceiling tiles, wood, paint, carpet rugs, etc., present very good environment for growth of fungi (Robin et al, 2007; Morey, 1999; Heseltine and Rosen, 2009). Water accumulation and moist dirty surfaces also encourage growth of fungi within a building, and contribute to human health problems. Fungi are very successful organisms capable of survival in diverse environments. This is possible because of their physiological versatility and genetic plasticity. Their spores are produced in large quantities and easily spread over a wide area. These spores can remain dormant for a long period of time during unfavorable conditions. Fungal spores and products of their metabolism are able to trigger allergic reactions which include hyper-sensitivity pneumonitis, allergic rhinitis and some types of asthma. Infections such as influenza, respiratory fungal infections, mycotoxicoses, and eye irritation by fungal volatile substances, among others, are health risks and challenges to susceptible individuals. Children, the elderly and immuno-compromised persons are at greater risks (Simoni et al, 2003).

Relationship Between Fungal Contamination of

Flannigan, 1997; Dubey et al, 2011).

differences influence the seasonal patterns of outdoor fungi.

Indoor Air and Health Problems of Some Residents in Jos 3

**1.2.1 Common fungal contaminants and factors responsible for indoor contamination**  Several fungi have been isolated from the indoor environment in various parts of the world and there seems to be relatedness in the types found from different places. Common airborne fungi in indoor environments include: *Alternaria sp, Aspergillus glaucus group, Aspergillus versicolor, Aspergillus fumigatus, A sydowii, Penicilium aurantiogriseum, Penicillium chrysogenum*, *Stachybotrys chartarum, Chaetomium globosum, Cladosporium cladosporoides, Ulocladium, Aspergillus nidulans,* and *Alternaria alternata*. Others are *Mortierella sp., Trichoderma sp.Penicilium lividum, P. arenicola, P. verrucosum P. citrinum, P. bilai, P. cyaneum, P. granulatum, P. sublateritium, Cladosporium sphaerospermum, C. herbarum, Fusarium sp. Trichoderma sp. Memnoniella sp. Epicoccum nigrum, Scopulariopsis fusca and Chrysosporium queenslandicum*. There seems to be some variation in the abundance of fungi at different seasons of the year (Shelton et al., 2002; Mirabelli et al, 2006; Sautour, 2009) while others occurred irrespective of the season (Horner et al, 2004). *Cladosporium species* are abundant in the summer, but *Penicillium* concentrations are higher in the wetter months (Verhoeff, 1993;

Mold types and concentrations indoors are primarily a function of outdoor fungi (Shelton et al., 2002). They presented an exhaustive study of the mycoflora of outdoor and indoor atmospheres in all USA regions. More than 12,000 samplings were carried out, both in outdoor and indoor atmospheres in more than 1,700 buildings. The moisture content of the substrate (related to indoor humidity level) has also been considered. Higher concentrations of outdoor molds and other fungi occur where trees, shrubs and landscape irrigation occur close to exterior building walls (McNeel and Krautzer, 1996). Humidity is the most important factor determining fungal growth in indoor environments (Got et al, 2003). Atmospheric relative humidity influences directly the release of conidia from conidiophores, and concomitantly, the concentration of spores in the atmosphere. The pattern may differ for different types of fungi. Pasanen et al., 1991, showed that spore release by *Cladosporium* was favored by low humidity unlike in *Penicilium* where the reverse was the case. These

In general, the types and concentrations of molds that affect indoor air quality are similar to those found in outdoor air and molds may occur in homes without dampness problems (Gots et al, 2003; Horner et al, 2004; O'connor et al, 2004; Codina et al, 2008). However, background mold numbers may shift whenever water accumulates in buildings. Damage caused by floods, plumbing leaks, poor understanding of moisture dynamics and careless building design and construction lead to structures that are more susceptible to water intrusion. Also, lack of good maintenance practices in some buildings can lead to moisture buildups that, when left alone, can result in microbial contamination and higher levels of bioaerosols ( BCBS, 2007; Stetzenbach et al, 2004). Dampness has usually been associated with moldiness indoor and its attendant ill health conditions (Dales et al, 1997; O'connor et al, 2004; BC Non-profit Housing Association,2007; Vocaturo et al, 2008). Damp buildings often have a moldy smell or obvious mold growth; some molds are human pathogens (Kuhn et al, 2003). Fungal colonization of buildings is especially common in low income communities in developing countries, where buildings are situated indiscriminately without consideration for environmental sanitation, urban planning and building regulations (Ahiamba et al, 2008; Ayanbimpe et al, 2010). Among factors responsible for indoor air contamination in developing countries are poor locations of refuse dumps, standing water

#### **1.1 Medically important fungi**

Molds are part of the kingdom Fungi. Fungi are a diverse group of organisms within a wide range of species that include mushrooms, bracket fungi, molds and yeasts. Fungi, are eukaryotic, heterotrophic organisms which may be filamentous, i.e. multicellular (molds) or single celled (yeasts). They are found everywhere around the human environment and more than 1, 500,000 species of molds exist in the world (Hawksworth, 1991; Cannon & Hawksworth, 1995; Hawksworth, 1999). Fungi are not photosynthetic and hence obtain nutrients by absorption. Molds are composed of linear chains of cells (hyphae) that branch and intertwine to form the fungus body called mycelium. All fungal cell walls contain (1-3) beta-D-glucan, chitin and mannan (Hawksworth, 1998). They exist as saprobes, mutualists or parasites in their relationships with other organisms. They reproduce by forming sexual or asexual spores which are in most instances, their dispersal propagules. Fungi grow very easily in damp and wet conditions; especially on materials and products containing moisture and an organic substrate. When these conditions are met, molds will grow and reproduce by forming spores that are released into the air. Molds are very adaptable and can grow even on damp inorganic materials such as glass, metal, concrete or painted surfaces if a microscopic layer of organic nutrients is available. Such nutrients can be found on household dust and soil particles (Robins and Morell, 2007). Fungi are capable of extracting their food from the organic materials they grow on and the ability to reproduce by way of minute spores makes them ubiquitous (DeHoog et al., 2007). Fungi are a part of nature's recycling system and play an important role in breaking down materials such as plants, leaves, wood and other natural matter (Robins and Morell, 2007). Because various genera grow and reproduce at different substrate water concentrations and temperatures, molds occur in a wide range of habitats. Microbial growth may result in greater numbers of spores, cell fragments, allergens, mycotoxins, endotoxins, β-glucans and volatile organic compounds in indoor air (Nielson, 2003). Persistent dampness and microbial growth on interior surfaces and in building structures should be avoided or minimized, as they may lead to adverse health effects. When mold growth is observed remediation should be given top priority (Zaslow, 1993; Douwes, 2009).

#### **1.2 Fungi and the Indoor environment**

Fungi grow on any material that provides a carbon source and adequate moisture. Fungi can grow on building and other materials, including: ceiling tiles; wood products; paint; wallpaper; carpeting; some furnishings; books/papers; the paper on gypsum wallboard (drywall); clothes; and other fabrics (Robins and Morrell, 2007). Mildew most often appears on natural fibers, such as cotton, linen, silk, and wool. It can actually rot the fabric. Mold can also grow on moist, dirty surfaces such as concrete, fiberglass insulation, and ceramic tiles (North Carolina Cooperative Extension Services 2009; Robins and Morell, 2007). Fungal growth in building materials is more dependent on the moisture content of the substrate than on atmospheric relative humidity. The minimum moisture content of building materials allowing fungal growth is near 76 %. Wood, wood composites (plywood, chipboard), and materials with a high starch content are capable of supporting fungal growth, at the lowest substrate moisture content. Plasterboard reinforced with cardboard and paper fibers, or inorganic materials coated with paint or treated with additives that offer an easily-degradable carbon source, are excellent substrates for fungal growth when substrate moisture content reaches 85-90 % (Pasanen et al, 1992; Dubey et al, 2011).

Molds are part of the kingdom Fungi. Fungi are a diverse group of organisms within a wide range of species that include mushrooms, bracket fungi, molds and yeasts. Fungi, are eukaryotic, heterotrophic organisms which may be filamentous, i.e. multicellular (molds) or single celled (yeasts). They are found everywhere around the human environment and more than 1, 500,000 species of molds exist in the world (Hawksworth, 1991; Cannon & Hawksworth, 1995; Hawksworth, 1999). Fungi are not photosynthetic and hence obtain nutrients by absorption. Molds are composed of linear chains of cells (hyphae) that branch and intertwine to form the fungus body called mycelium. All fungal cell walls contain (1-3) beta-D-glucan, chitin and mannan (Hawksworth, 1998). They exist as saprobes, mutualists or parasites in their relationships with other organisms. They reproduce by forming sexual or asexual spores which are in most instances, their dispersal propagules. Fungi grow very easily in damp and wet conditions; especially on materials and products containing moisture and an organic substrate. When these conditions are met, molds will grow and reproduce by forming spores that are released into the air. Molds are very adaptable and can grow even on damp inorganic materials such as glass, metal, concrete or painted surfaces if a microscopic layer of organic nutrients is available. Such nutrients can be found on household dust and soil particles (Robins and Morell, 2007). Fungi are capable of extracting their food from the organic materials they grow on and the ability to reproduce by way of minute spores makes them ubiquitous (DeHoog et al., 2007). Fungi are a part of nature's recycling system and play an important role in breaking down materials such as plants, leaves, wood and other natural matter (Robins and Morell, 2007). Because various genera grow and reproduce at different substrate water concentrations and temperatures, molds occur in a wide range of habitats. Microbial growth may result in greater numbers of spores, cell fragments, allergens, mycotoxins, endotoxins, β-glucans and volatile organic compounds in indoor air (Nielson, 2003). Persistent dampness and microbial growth on interior surfaces and in building structures should be avoided or minimized, as they may lead to adverse health effects. When mold growth is observed remediation should be given

Fungi grow on any material that provides a carbon source and adequate moisture. Fungi can grow on building and other materials, including: ceiling tiles; wood products; paint; wallpaper; carpeting; some furnishings; books/papers; the paper on gypsum wallboard (drywall); clothes; and other fabrics (Robins and Morrell, 2007). Mildew most often appears on natural fibers, such as cotton, linen, silk, and wool. It can actually rot the fabric. Mold can also grow on moist, dirty surfaces such as concrete, fiberglass insulation, and ceramic tiles (North Carolina Cooperative Extension Services 2009; Robins and Morell, 2007). Fungal growth in building materials is more dependent on the moisture content of the substrate than on atmospheric relative humidity. The minimum moisture content of building materials allowing fungal growth is near 76 %. Wood, wood composites (plywood, chipboard), and materials with a high starch content are capable of supporting fungal growth, at the lowest substrate moisture content. Plasterboard reinforced with cardboard and paper fibers, or inorganic materials coated with paint or treated with additives that offer an easily-degradable carbon source, are excellent substrates for fungal growth when

substrate moisture content reaches 85-90 % (Pasanen et al, 1992; Dubey et al, 2011).

**1.1 Medically important fungi** 

top priority (Zaslow, 1993; Douwes, 2009).

**1.2 Fungi and the Indoor environment** 

#### **1.2.1 Common fungal contaminants and factors responsible for indoor contamination**

Several fungi have been isolated from the indoor environment in various parts of the world and there seems to be relatedness in the types found from different places. Common airborne fungi in indoor environments include: *Alternaria sp, Aspergillus glaucus group, Aspergillus versicolor, Aspergillus fumigatus, A sydowii, Penicilium aurantiogriseum, Penicillium chrysogenum*, *Stachybotrys chartarum, Chaetomium globosum, Cladosporium cladosporoides, Ulocladium, Aspergillus nidulans,* and *Alternaria alternata*. Others are *Mortierella sp., Trichoderma sp.Penicilium lividum, P. arenicola, P. verrucosum P. citrinum, P. bilai, P. cyaneum, P. granulatum, P. sublateritium, Cladosporium sphaerospermum, C. herbarum, Fusarium sp. Trichoderma sp. Memnoniella sp. Epicoccum nigrum, Scopulariopsis fusca and Chrysosporium queenslandicum*. There seems to be some variation in the abundance of fungi at different seasons of the year (Shelton et al., 2002; Mirabelli et al, 2006; Sautour, 2009) while others occurred irrespective of the season (Horner et al, 2004). *Cladosporium species* are abundant in the summer, but *Penicillium* concentrations are higher in the wetter months (Verhoeff, 1993; Flannigan, 1997; Dubey et al, 2011).

Mold types and concentrations indoors are primarily a function of outdoor fungi (Shelton et al., 2002). They presented an exhaustive study of the mycoflora of outdoor and indoor atmospheres in all USA regions. More than 12,000 samplings were carried out, both in outdoor and indoor atmospheres in more than 1,700 buildings. The moisture content of the substrate (related to indoor humidity level) has also been considered. Higher concentrations of outdoor molds and other fungi occur where trees, shrubs and landscape irrigation occur close to exterior building walls (McNeel and Krautzer, 1996). Humidity is the most important factor determining fungal growth in indoor environments (Got et al, 2003). Atmospheric relative humidity influences directly the release of conidia from conidiophores, and concomitantly, the concentration of spores in the atmosphere. The pattern may differ for different types of fungi. Pasanen et al., 1991, showed that spore release by *Cladosporium* was favored by low humidity unlike in *Penicilium* where the reverse was the case. These differences influence the seasonal patterns of outdoor fungi.

In general, the types and concentrations of molds that affect indoor air quality are similar to those found in outdoor air and molds may occur in homes without dampness problems (Gots et al, 2003; Horner et al, 2004; O'connor et al, 2004; Codina et al, 2008). However, background mold numbers may shift whenever water accumulates in buildings. Damage caused by floods, plumbing leaks, poor understanding of moisture dynamics and careless building design and construction lead to structures that are more susceptible to water intrusion. Also, lack of good maintenance practices in some buildings can lead to moisture buildups that, when left alone, can result in microbial contamination and higher levels of bioaerosols ( BCBS, 2007; Stetzenbach et al, 2004). Dampness has usually been associated with moldiness indoor and its attendant ill health conditions (Dales et al, 1997; O'connor et al, 2004; BC Non-profit Housing Association,2007; Vocaturo et al, 2008). Damp buildings often have a moldy smell or obvious mold growth; some molds are human pathogens (Kuhn et al, 2003). Fungal colonization of buildings is especially common in low income communities in developing countries, where buildings are situated indiscriminately without consideration for environmental sanitation, urban planning and building regulations (Ahiamba et al, 2008; Ayanbimpe et al, 2010). Among factors responsible for indoor air contamination in developing countries are poor locations of refuse dumps, standing water

Relationship Between Fungal Contamination of

Indoor Air and Health Problems of Some Residents in Jos 5

reactions in which mild and occasional, to severe and chronic symptoms caused by most species of fungi may develop in susceptible individuals who may present with hay fever, runny nose, breathing problems, and sinusitis (Bryant, 2002; Moloughney, 2004, Wu et al, 2007; Kuhn and Ghannoum,, 2004; Mirabelli et al, 2006). Some people could develop skin diseases and skin rashes as an allergic reaction to toxic mold. (Gots et al, 2003; http://www.cleanwaterpartners.org/mold/black-mold.html). Most people can recognize and react to the moldy smell negatively with symptoms like headaches, vomiting, nausea, blocked noses, and asthma. Volatile products of indoor molds have also been suggested to cause toxic illnesses or mycotoxicoses (Abbott, 2002; Straus and Wilson, 2006). Mycotoxins can be inhaled by people through the air, ingested through food, and can also come in contact with skin, resulting in many illnesses. Different molds produce different mycotoxins, which also depends on what kind of material the toxic mold is growing on. Among the illnesses caused by mycotoxins are gastrointestinal, respiratory, and reproductive disorders (O'connor et al, 2004; Mirabelli et al, 2006). Also reported are eye irritations, poor taste sensations, sore throat and tiredness (U.S. Environmental Protection Agency, 2003; Worksafe Bulletin, 1999). Even severe neurologic conditions such as depression have been linked to indoor mold colonization (Shenassa et al, 2007). Under certain metabolic conditions, many fungi produce mycotoxins, natural organic compounds that initiate a toxic response in vertebrates. Molds that are important potential producers of toxins indoors are certain species of *Fusarium*, *Penicillium*, and *Aspergillus*. In water-damaged buildings *Stachybotrys chartarum* and *Aspergillus versicolor* may also produce toxic metabolites. A large body of information is available on the human and animal health effects from ingestion of certain mycotoxins (Sorenson, 1989; Smith and Henderson, 1991), Two classes of mycotoxins have been isolated from house dust samples: aflatoxins from some strains of *Aspergillus flavus* and trichothecenes from some species and strains of *Fusarium, Cephalosporium, Stachybotrys* and *Trichoderma* (Kuhn and Ghanoum, 2004). The fungus cell wall component, (1-3)-beta-D-glucan, a medically significant glucose polymer that has immunosuppressive, mitogenic (i.e. causing mitosis or cell transformation) and inflammatory properties. This mold cell wall component also appears to act synergistically with bacterial endotoxins to

produce airway inflammation following inhalation exposure in guinea pigs.

The problem of environmental pollution is enormous and various attempts have been made to establish facilities for its control and regulation in various parts of the world. In Nigeria, the Federal Military Government set up the Federal Environmental Protection Agency (FEPA) Act 1988, established under Decree 58 of 30 December 1988 with statutory responsibility for overall protection of the environment and setting and enforcing ambient and emission standards for air, water and noise pollution (Federal Government of Nigeria (FGN), 1988). The Environmental Impact Assessment (EIA) Act 1992 reaffirmed the powers of the FEPA and defined the minimum requirements for an EIA (Environmental Policy, 1999). The democratic government created a Ministry of Environment in 1999, thus bringing agencies such as the FEPA and the Environmental Health Unit of the Ministry of Health under one administrative system. Further effort to ensure the maintenance of a healthy environment was made by instituting the National Environmental Standards and Regulations Enforcement Agency (NESREA) charged with the responsibility for protection, and development of the Nigerian environment (Federal Republic of Nigeria, 2007). Among

**1.5 Environmental pollution and regulations** 

bodies, water damaged materials and absence of proper urban planning. According to Heseltine &Rosen, (2009) indicators of dampness and microbial growth include the presence of condensation on surfaces or in structures, visible mould, perceived mould odour and a history of water damage, leakage or penetration. Thorough inspection and, if necessary, appropriate measurements can be used to confirm indoor moisture and microbial growth. Similarly, contaminated air handling systems can become breeding grounds for molds and mildews and also distribute spores all over the home. Excessive humidity and high water content of building materials, lack of thermal insulations and incorrect behaviours of residents may also contribute (Ahiamba et al, 2008).

#### **1.3 Indoor molds and health**

A healthy indoor environment is a necessity for every human being. A number of studies have examined the health risks related to the presence of fungi indoors and shown positive associations between fungal levels and health outcomes. Evidences abound on the problems of dampness which consequently contribute to mold colonization of the indoor environment (European Environment and Health Information System, 2007). To control diseases caused by fungi the primary avenues of their introduction into the indoor environment should be attended to and knowledge of these avenues is required. Identifying conditions which contribute to fungal presence in the indoor environment and correction of such is a necessary step to preventing ill-health and resultant losses. Various suggestions have been made to this effect (Buck, 2002).

Some reports have highlighted the importance of housing conditions as determinants of mental and physical health and the importance of housing as a medium for promoting universal health efforts (BC Non-profit Housing Association, 2007; Shenasa et al, 2007; BCBS, 2007; Canadian Center for Occupational Health and Safety, 2008). Surveys of indoor environments have shown that symptoms of ill-health associated with air-borne fungi and their products are higher in households with molds and mildews and that are characterized by moist or damp surfaces, damp walls, wet surfaces, damaged materials, standing water and refuse dumps (Shelton et al, 2002; O'connor et al, 2004; Ahiamba et all, 2008; Ayanbimpe et al, 2010;). Pathogenic infections may be elicited by some molds which are primarily found in soil or on vegetation but are carried into homes by human foot wears or animals. When spores of some air-borne fungi are inhaled by immunocompromised persons, they may develop disease processes, for example Aspergillosis, and histoplasmosis. Respiratory illnesses are the major reported problems among persons living in homes colonized by molds (Dales et al., 1991). Persons at greater risk of effects of indoor air fungal colonization are the young, elderly, immunocompromised individuals, persons undergoing intensive surgical procedures like transplant patients and those on immunosuppressive therapy (Simoni et al, 2003; Ritz et al, 2009;). Fungal components implicated in pathogenesis include spores; β-1,3-D-glucan; Hydrophobins; Volatile organic compounds (VOCs) and Dihdroxynaphthalene (DHN) Melanin.

#### **1.4 Symptoms of ill health associated with poor indoor air quality**

There are suggestions that even very low levels of some common air pollutants can gradually cause health problems. Such condition have been reported to include; allergic

bodies, water damaged materials and absence of proper urban planning. According to Heseltine &Rosen, (2009) indicators of dampness and microbial growth include the presence of condensation on surfaces or in structures, visible mould, perceived mould odour and a history of water damage, leakage or penetration. Thorough inspection and, if necessary, appropriate measurements can be used to confirm indoor moisture and microbial growth. Similarly, contaminated air handling systems can become breeding grounds for molds and mildews and also distribute spores all over the home. Excessive humidity and high water content of building materials, lack of thermal insulations and incorrect behaviours of

A healthy indoor environment is a necessity for every human being. A number of studies have examined the health risks related to the presence of fungi indoors and shown positive associations between fungal levels and health outcomes. Evidences abound on the problems of dampness which consequently contribute to mold colonization of the indoor environment (European Environment and Health Information System, 2007). To control diseases caused by fungi the primary avenues of their introduction into the indoor environment should be attended to and knowledge of these avenues is required. Identifying conditions which contribute to fungal presence in the indoor environment and correction of such is a necessary step to preventing ill-health and resultant losses. Various suggestions have been

Some reports have highlighted the importance of housing conditions as determinants of mental and physical health and the importance of housing as a medium for promoting universal health efforts (BC Non-profit Housing Association, 2007; Shenasa et al, 2007; BCBS, 2007; Canadian Center for Occupational Health and Safety, 2008). Surveys of indoor environments have shown that symptoms of ill-health associated with air-borne fungi and their products are higher in households with molds and mildews and that are characterized by moist or damp surfaces, damp walls, wet surfaces, damaged materials, standing water and refuse dumps (Shelton et al, 2002; O'connor et al, 2004; Ahiamba et all, 2008; Ayanbimpe et al, 2010;). Pathogenic infections may be elicited by some molds which are primarily found in soil or on vegetation but are carried into homes by human foot wears or animals. When spores of some air-borne fungi are inhaled by immunocompromised persons, they may develop disease processes, for example Aspergillosis, and histoplasmosis. Respiratory illnesses are the major reported problems among persons living in homes colonized by molds (Dales et al., 1991). Persons at greater risk of effects of indoor air fungal colonization are the young, elderly, immunocompromised individuals, persons undergoing intensive surgical procedures like transplant patients and those on immunosuppressive therapy (Simoni et al, 2003; Ritz et al, 2009;). Fungal components implicated in pathogenesis include spores; β-1,3-D-glucan; Hydrophobins; Volatile organic compounds (VOCs) and

residents may also contribute (Ahiamba et al, 2008).

**1.3 Indoor molds and health** 

made to this effect (Buck, 2002).

Dihdroxynaphthalene (DHN) Melanin.

**1.4 Symptoms of ill health associated with poor indoor air quality** 

There are suggestions that even very low levels of some common air pollutants can gradually cause health problems. Such condition have been reported to include; allergic

reactions in which mild and occasional, to severe and chronic symptoms caused by most species of fungi may develop in susceptible individuals who may present with hay fever, runny nose, breathing problems, and sinusitis (Bryant, 2002; Moloughney, 2004, Wu et al, 2007; Kuhn and Ghannoum,, 2004; Mirabelli et al, 2006). Some people could develop skin diseases and skin rashes as an allergic reaction to toxic mold. (Gots et al, 2003; http://www.cleanwaterpartners.org/mold/black-mold.html). Most people can recognize and react to the moldy smell negatively with symptoms like headaches, vomiting, nausea, blocked noses, and asthma. Volatile products of indoor molds have also been suggested to cause toxic illnesses or mycotoxicoses (Abbott, 2002; Straus and Wilson, 2006). Mycotoxins can be inhaled by people through the air, ingested through food, and can also come in contact with skin, resulting in many illnesses. Different molds produce different mycotoxins, which also depends on what kind of material the toxic mold is growing on. Among the illnesses caused by mycotoxins are gastrointestinal, respiratory, and reproductive disorders (O'connor et al, 2004; Mirabelli et al, 2006). Also reported are eye irritations, poor taste sensations, sore throat and tiredness (U.S. Environmental Protection Agency, 2003; Worksafe Bulletin, 1999). Even severe neurologic conditions such as depression have been linked to indoor mold colonization (Shenassa et al, 2007). Under certain metabolic conditions, many fungi produce mycotoxins, natural organic compounds that initiate a toxic response in vertebrates. Molds that are important potential producers of toxins indoors are certain species of *Fusarium*, *Penicillium*, and *Aspergillus*. In water-damaged buildings *Stachybotrys chartarum* and *Aspergillus versicolor* may also produce toxic metabolites. A large body of information is available on the human and animal health effects from ingestion of certain mycotoxins (Sorenson, 1989; Smith and Henderson, 1991), Two classes of mycotoxins have been isolated from house dust samples: aflatoxins from some strains of *Aspergillus flavus* and trichothecenes from some species and strains of *Fusarium, Cephalosporium, Stachybotrys* and *Trichoderma* (Kuhn and Ghanoum, 2004). The fungus cell wall component, (1-3)-beta-D-glucan, a medically significant glucose polymer that has immunosuppressive, mitogenic (i.e. causing mitosis or cell transformation) and inflammatory properties. This mold cell wall component also appears to act synergistically with bacterial endotoxins to produce airway inflammation following inhalation exposure in guinea pigs.

#### **1.5 Environmental pollution and regulations**

The problem of environmental pollution is enormous and various attempts have been made to establish facilities for its control and regulation in various parts of the world. In Nigeria, the Federal Military Government set up the Federal Environmental Protection Agency (FEPA) Act 1988, established under Decree 58 of 30 December 1988 with statutory responsibility for overall protection of the environment and setting and enforcing ambient and emission standards for air, water and noise pollution (Federal Government of Nigeria (FGN), 1988). The Environmental Impact Assessment (EIA) Act 1992 reaffirmed the powers of the FEPA and defined the minimum requirements for an EIA (Environmental Policy, 1999). The democratic government created a Ministry of Environment in 1999, thus bringing agencies such as the FEPA and the Environmental Health Unit of the Ministry of Health under one administrative system. Further effort to ensure the maintenance of a healthy environment was made by instituting the National Environmental Standards and Regulations Enforcement Agency (NESREA) charged with the responsibility for protection, and development of the Nigerian environment (Federal Republic of Nigeria, 2007). Among

Relationship Between Fungal Contamination of

Indoor Air and Health Problems of Some Residents in Jos 7

Mycotoxicosis may also be acquired through exposure to toxins produced by some fungi as a result of their secondary metabolism, elaborated through various metabolic pathways (Wang et al, 2002). Mycotoxins have been detected from crude building materials and include Acetyl-T-2 toxin; Aflatoxin A; DON, HT-2 toxin; OchratoxinA; citrinine; Nivalenol, indoor environments (Nielson 2003). These mycotoxins may target various organs of the human body such as the liver, lungs, kidney, nervous system, immune system and the endocrine system. The severity of the mycotoxicoses depends on the target organ, nutritional and health status of the individual, type of toxin, and the synergistic effect of the

Indoor air pollution has been identified as one of the biggest health threats facing Americans in recent times (US Environmental Protection Agency, 2003). This may be even more in the developing countries but are underreported because of the near absence of research into this aspect. Knowledge of indoor air quality, its health significance and the factors that cause poor quality is key to enabling action by relevant stakeholders, including building owners, developers, users and occupants (Heseltine &Rosen, 2009). Maintaining a clean indoor air involves every one of the above. It is clear that elucidating the actual state of indoor air and its relationship to health of residents is an important step that would enable proper action, by regulation and policy to be taken by appropriate authorities. The building users on their part would also play a role in this very important issue. There are very few reported investigations in the study area on the health effects of indoor air fungi on the residents and

This household survey and examination of health data of residents of houses with indoor air fungal contamination was carried out to establish the association between microbial contaminations of indoor air and the health of residents. Some aspect of this was highlighted in a previous report by us (Ayanbimpe et al, 2010), thus the present study is a follow up on subjects previously identified in order to establish a relationship, if any, between the indoor air fungal colonization and the health of residents. The criteria considered for determination of a relationship between ill-health symptoms in residents and

1. The presence of fungus growth indoor and isolation of fungi from indoor air of the

2. Dampness of the indoor environment and presence of other parameters considered as

3. The experience of symptoms among residents while in the specific indoor environment.

The Jos metropolis is the capital city of Plateau State of Nigeria located at 1900m above sea level in the North Central part of the country. The Jos Plateau is bounded approximately by latitudes 8° 0'55''N and 10° 0''N, and longitude 8 0'S 22'' E and 9 0' 30''E. Tin mining which

mycotoxins with other chemicals or mycotoxins in the target organ.

**1.8 The general challenge of indoor air quality** 

because of this, little or no data is available.

**2. Materials and methods** 

indoor air were as follows:

risk factors (see Table 3).

**2.1 Description of the study area** 

4. The reduction of symptoms following mold remediation.

residence.

numerous initiatives advanced to attain the United Nations Millennium Development Goal is the relationship between the physical environment and human health, and this is a top priority of the WHO. Here the need to ensure that people, especially children, live in environments with clean air was emphasized. The WHO recognized healthy indoor air as a basic right because people spend a large part of their time each day indoors; in homes, offices, schools, health care facilities, or other private or public buildings, therefore quality of the air they breathe in those buildings is an important determinant of their health and well-being (World Health Organization, 2009). Regulations by relevant authorities to control the environmental air pollution have focused on industrial emissions and chemical contaminants but less on microbiological contaminants, especially of the indoor air.

#### **1.6 Relationship between indoor air fungi and health of residents**

Indoor air pollution, such as from dampness and mould, chemicals and other biological agents, is a major cause of morbidity and mortality worldwide (Environmental Protection Agency, 2010). Biological pollutants, volatile organic compounds and indoor molds have been associated with symptoms such as rhinitis, wheezing, conjunctiva irritation, skin rashes, tightness in the chest, difficulty in breathing, runny nose, fever, muscle aches, headaches, diarrhea, nausea and nose bleeding. Aggravation of respiratory symptoms has been linked to increased concentration of airborne particles (McNeel and Kreutzer, 1996; Wong et al, 1999; Petrescul, et al, 2011). Interventions to reduce the colonization of indoor environment by contaminants which may serve as triggers to respiratory illnesses, especially asthma in various places yielded positive results. Follow-up from one of such interventions (Morgan et al. 2004), showed that the subjects experienced fewer symptom days for 2 years, and there were significant reductions in dust mite and cockroach allergens, also for 2 years. Another study showed the rate of doctor-diagnosed asthma was higher (25%) in children residing in deteriorated public housing compared with only 8% in those in other houses. A similar observation was made by Howell et al. (2005), on the self-reported health status of some individuals. Petrescu1 et al., (2011) related increase in chronic respiratory symptoms to increased concentrations of particles in the air. Shenasa et al, (2009) reported a strong association between depression and indoor mold. In their study, 9% of the respondents reported 3 or 4 depressive symptoms and older women and unemployed persons were more prone depressive symptoms. This was irrespective of dampness of the indoor environment.

#### **1.7 Health effects of mycotoxins and volatile organic compounds**

Molds produce acute health effects through toxin-induced inflammation, allergy, or infection. Some reports have associated overgrowths of trichothecene-producing fungi with human health effects such as cold and flu-like symptoms, sore throats, headache and general malaise (Croft et al., 1986; Johanning et al., 1993; Pasanen et al., 1996). Molds also produce a large number of volatile organic compounds (VOCs) which are responsible for the musty odors produced by growing molds. These VOCs have not concretely been associated with disease, although ethanol, the most common VOC, synergizes many fungal toxins. When fungal spores are inhaled, they may reach the lung alveoli and induce an inflammatory reaction, creating toxic pneumonitis. Severe toxic pneumonitis can cause fever, flu-like symptoms and fatigue (organic toxic dust syndrome). Hypersensitivity pneumonitis, a particular form of granulomatous lung disease, is a syndrome caused by inhalation of large concentrations of dust containing organic material including fungal spores.

numerous initiatives advanced to attain the United Nations Millennium Development Goal is the relationship between the physical environment and human health, and this is a top priority of the WHO. Here the need to ensure that people, especially children, live in environments with clean air was emphasized. The WHO recognized healthy indoor air as a basic right because people spend a large part of their time each day indoors; in homes, offices, schools, health care facilities, or other private or public buildings, therefore quality of the air they breathe in those buildings is an important determinant of their health and well-being (World Health Organization, 2009). Regulations by relevant authorities to control the environmental air pollution have focused on industrial emissions and chemical

contaminants but less on microbiological contaminants, especially of the indoor air.

Indoor air pollution, such as from dampness and mould, chemicals and other biological agents, is a major cause of morbidity and mortality worldwide (Environmental Protection Agency, 2010). Biological pollutants, volatile organic compounds and indoor molds have been associated with symptoms such as rhinitis, wheezing, conjunctiva irritation, skin rashes, tightness in the chest, difficulty in breathing, runny nose, fever, muscle aches, headaches, diarrhea, nausea and nose bleeding. Aggravation of respiratory symptoms has been linked to increased concentration of airborne particles (McNeel and Kreutzer, 1996; Wong et al, 1999; Petrescul, et al, 2011). Interventions to reduce the colonization of indoor environment by contaminants which may serve as triggers to respiratory illnesses, especially asthma in various places yielded positive results. Follow-up from one of such interventions (Morgan et al. 2004), showed that the subjects experienced fewer symptom days for 2 years, and there were significant reductions in dust mite and cockroach allergens, also for 2 years. Another study showed the rate of doctor-diagnosed asthma was higher (25%) in children residing in deteriorated public housing compared with only 8% in those in other houses. A similar observation was made by Howell et al. (2005), on the self-reported health status of some individuals. Petrescu1 et al., (2011) related increase in chronic respiratory symptoms to increased concentrations of particles in the air. Shenasa et al, (2009) reported a strong association between depression and indoor mold. In their study, 9% of the respondents reported 3 or 4 depressive symptoms and older women and unemployed persons were more prone depressive symptoms. This was irrespective of dampness of the indoor environment.

**1.6 Relationship between indoor air fungi and health of residents** 

**1.7 Health effects of mycotoxins and volatile organic compounds** 

dust containing organic material including fungal spores.

Molds produce acute health effects through toxin-induced inflammation, allergy, or infection. Some reports have associated overgrowths of trichothecene-producing fungi with human health effects such as cold and flu-like symptoms, sore throats, headache and general malaise (Croft et al., 1986; Johanning et al., 1993; Pasanen et al., 1996). Molds also produce a large number of volatile organic compounds (VOCs) which are responsible for the musty odors produced by growing molds. These VOCs have not concretely been associated with disease, although ethanol, the most common VOC, synergizes many fungal toxins. When fungal spores are inhaled, they may reach the lung alveoli and induce an inflammatory reaction, creating toxic pneumonitis. Severe toxic pneumonitis can cause fever, flu-like symptoms and fatigue (organic toxic dust syndrome). Hypersensitivity pneumonitis, a particular form of granulomatous lung disease, is a syndrome caused by inhalation of large concentrations of Mycotoxicosis may also be acquired through exposure to toxins produced by some fungi as a result of their secondary metabolism, elaborated through various metabolic pathways (Wang et al, 2002). Mycotoxins have been detected from crude building materials and include Acetyl-T-2 toxin; Aflatoxin A; DON, HT-2 toxin; OchratoxinA; citrinine; Nivalenol, indoor environments (Nielson 2003). These mycotoxins may target various organs of the human body such as the liver, lungs, kidney, nervous system, immune system and the endocrine system. The severity of the mycotoxicoses depends on the target organ, nutritional and health status of the individual, type of toxin, and the synergistic effect of the mycotoxins with other chemicals or mycotoxins in the target organ.

#### **1.8 The general challenge of indoor air quality**

Indoor air pollution has been identified as one of the biggest health threats facing Americans in recent times (US Environmental Protection Agency, 2003). This may be even more in the developing countries but are underreported because of the near absence of research into this aspect. Knowledge of indoor air quality, its health significance and the factors that cause poor quality is key to enabling action by relevant stakeholders, including building owners, developers, users and occupants (Heseltine &Rosen, 2009). Maintaining a clean indoor air involves every one of the above. It is clear that elucidating the actual state of indoor air and its relationship to health of residents is an important step that would enable proper action, by regulation and policy to be taken by appropriate authorities. The building users on their part would also play a role in this very important issue. There are very few reported investigations in the study area on the health effects of indoor air fungi on the residents and because of this, little or no data is available.

#### **2. Materials and methods**

This household survey and examination of health data of residents of houses with indoor air fungal contamination was carried out to establish the association between microbial contaminations of indoor air and the health of residents. Some aspect of this was highlighted in a previous report by us (Ayanbimpe et al, 2010), thus the present study is a follow up on subjects previously identified in order to establish a relationship, if any, between the indoor air fungal colonization and the health of residents. The criteria considered for determination of a relationship between ill-health symptoms in residents and indoor air were as follows:


#### **2.1 Description of the study area**

The Jos metropolis is the capital city of Plateau State of Nigeria located at 1900m above sea level in the North Central part of the country. The Jos Plateau is bounded approximately by latitudes 8° 0'55''N and 10° 0''N, and longitude 8 0'S 22'' E and 9 0' 30''E. Tin mining which

Relationship Between Fungal Contamination of

2010)

remediation

Indoor Air and Health Problems of Some Residents in Jos 9

**Fungus Average number percent**  *Chaetomium globosum* 64 16.8 *Stachybotrys alterans* 53 13.9 *Aspergillus versicolor* 53 13.9 *Alternaria alternata* 53 13.9 *Aspergillus fumigatus* 37 13.9 *Xylohypha bantianum* 29 7.6 *Penicilium sp* 16 4.2 *Rhizopus sp* 28 7.8 *Stemphylum sp* 24 6.3 *Sepedonium sp* 13 3.4 *Phoma sp* 10 2.6 Total 380 100 Table 1. Fungi isolated from indoor air in the study area (Adopted from Ayanbimpe et al.,

**Symptoms 0-10 11-20 21-30 31-40 41-50 51-60 61-70 71-80 Total**  Cough, 7 8 1 0 10 4 0 1 31 Nocturnal cough 2 2 0 0 1 1 0 1 7 Dyspnea 0 1 0 0 8 0 0 0 9 Sore throat 1 12 2 0 8 2 0 0 25 Hoarseness 1 4 2 0 3 0 0 1 26 Rhinitis 7 6 3 0 7 2 0 1 26 Nasal bleeding 1 4 0 0 0 0 0 0 5 Sinusitis 0 0 0 0 1 1 0 0 2 Skin rash 2 4 1 0 3 0 0 1 11 Headaches 3 9 3 0 9 2 0 1 27 Nausea 2 8 0 0 8 0 0 0 18 Diarrhea 4 7 1 0 4 0 0 0 16 Tightness in the chest 2 4 0 0 3 0 0 0 9 Fever 7 8 2 0 4 2 0 0 23 Muscle aches 2 8 1 0 12 1 0 1 25 Eye irritation 4 9 2 0 4 4 0 1 24

Table 2. Symptoms reported among residents of various age groups before mold

began around 1904 was a major occupation of people of this area. This activity has led to creation of ponds; gullies and water logged ditches (some of which are used as dumping sites by residents) around residential areas that may contribute to dampness. The rock materials are used extensively for various forms of building construction in and around Jos (Solomon, et al 2002). The study area is characterized by several unplanned settlements, with haphazardly located buildings.

#### **2.2 Evaluation of indoor air quality in the study area**

A previous study had highlighted the poor indoor air quality of the study area (Ayanbimpe et al, 2010). The following fungi were isolated: *Aspergillus versicolor, Eurotium sp, Stachybotrys chartarum, Chaetomium globosum, Stachybotrys alternans, Aspergillus versicolor, Alternaria alternata, Aspergillus fumigatus, Xylohypha bantianum, Rhizopus sp, Stemphylium sp, Penicillium sp, Sepedonium sp* and *Phoma sp*. (see Table 1). More than 75% of the population in that report had complaints of respiratory symptoms; frequent headache, eye irritation, skin rash etc. (see Table 2). 17 households out of the 150 had molds growing in one or more parts of the house. The relationship between poor housing conditions and indoor air contamination was also established.

#### **2.3 Investigation of factors in the study area**

Using structured questionnaires and one on one interviews the conditions of the buildings considered as risk factors were determined. This was done by obtaining responses for the following questions:


The following problems were identified as risk factors for the different households in the study area:


Residents with health problems related to airborne fungi had been identified and specific symptoms recorded. The present study assumed that the presence of moldy surfaces indoors contributed to the concentration of airborne spores. By a random selection, 8 of the 17 households with indoor mold colonization and high concentration of airborne fungi were visited for follow-up. Each of the 8 households revisited underwent mold clean up and remediation or correction of conditions / factors supposedly responsible for indoor mold colonization.

began around 1904 was a major occupation of people of this area. This activity has led to creation of ponds; gullies and water logged ditches (some of which are used as dumping sites by residents) around residential areas that may contribute to dampness. The rock materials are used extensively for various forms of building construction in and around Jos (Solomon, et al 2002). The study area is characterized by several unplanned settlements,

A previous study had highlighted the poor indoor air quality of the study area (Ayanbimpe et al, 2010). The following fungi were isolated: *Aspergillus versicolor, Eurotium sp, Stachybotrys chartarum, Chaetomium globosum, Stachybotrys alternans, Aspergillus versicolor, Alternaria alternata, Aspergillus fumigatus, Xylohypha bantianum, Rhizopus sp, Stemphylium sp, Penicillium sp, Sepedonium sp* and *Phoma sp*. (see Table 1). More than 75% of the population in that report had complaints of respiratory symptoms; frequent headache, eye irritation, skin rash etc. (see Table 2). 17 households out of the 150 had molds growing in one or more parts of the house. The relationship between poor housing conditions and indoor air

Using structured questionnaires and one on one interviews the conditions of the buildings considered as risk factors were determined. This was done by obtaining responses for the

e. Did you experience flooding at any time or are there leaking areas in the house?

The following problems were identified as risk factors for the different households in the

Residents with health problems related to airborne fungi had been identified and specific symptoms recorded. The present study assumed that the presence of moldy surfaces indoors contributed to the concentration of airborne spores. By a random selection, 8 of the 17 households with indoor mold colonization and high concentration of airborne fungi were visited for follow-up. Each of the 8 households revisited underwent mold clean up and remediation or correction of conditions / factors supposedly responsible for indoor mold

g. Do you have moist walls and moldy spots/areas within the house?



with haphazardly located buildings.

contamination was also established.

following questions:

study area: - Flooding

colonization.



a. How old is the building?

b. What type of accommodation is it? c. How many rooms are there in the house? d. How many persons live in a room?

**2.3 Investigation of factors in the study area** 

f. How many households share the building?


**2.2 Evaluation of indoor air quality in the study area** 


Table 1. Fungi isolated from indoor air in the study area (Adopted from Ayanbimpe et al., 2010)


Table 2. Symptoms reported among residents of various age groups before mold remediation

Relationship Between Fungal Contamination of

i. Frequency of symptoms ii. Type of complaint iii. Moldy odor in the home?

experienced by residents.

**3. Results** 

1 Building above 25 years old was considered old

Indoor Air and Health Problems of Some Residents in Jos 11

The symptoms included cough, nocturnal cough, dyspnea, sore throat, hoarseness, rhinitis, nasal bleeding, sinusitis, skin rashes, headaches, nausea, diarrhea, tightness in the chest, fever and muscle aches. The questionnaire responses were analyzed to establish the state of health of subjects post remediation. The Pearson correlation test was applied to determine significant association between mold presence or absence in homes and ill health symptoms

Table 3 shows outcome of questionnaire responses with respect to the eight houses revisited. Mold growth, Arthropod infestation and presence of suggestive symptoms were reported in all (100%) of the houses before remediation. All the households inspected, post remediation, had complete absence of moldy spots. There was significant correlation between reduction in frequency of symptoms of some health problems among residents and mold remediation (r = 0.577; P>0.05). Significant reduction of the following symptoms,

**Features of responses Yes (%) No (%)**  Location unplanned 72 28 Males 100 0 Females 100 0 Age of building (old)1 62.5 37.5 Leaking 50 50 Flooding 12.5 87.5 Dampness indoors 75 25 Mold growth indoor 100 0 Musty odor indoors 55 45 Arthropods infestation 100 0 Suggestive symptoms 100 0

Table 3. Features of Questionnaire responses with respect to the eight residences

cough (83.3%), dyspnea (100%), nasal bleeding (96.9%), skin rash (85.1%), tightness in the chest (83.3%), rhinitis (85.1), muscle aches (92.3%) and eye irritation (92.3%), were observed (Table 4). Reduction in frequency of occurrence of symptoms with respect to age and gender is shown in table 3. 72.3% of individuals experienced significant reduction or total

#### **2.4 Fungal clean up and remediation**

Because of the seriousness of the problem of indoor air contamination, the WHO guideline recommended that remediation of the conditions that lead to adverse exposure should be given priority to prevent an additional contribution to poor health in populations who are already living with an increased burden of disease (WHO Regional Office for Europe, 2006). The various conditions that supposedly contributed to mold colonization were physically assessed and the necessary remediation procedure determined. Methods proposed in standard manuals and guidelines (Kennedy; Buck, 2002; New York City Department of Health and Mental Hygiene; 2008; North Carolina University Cooperative Extension Services, 2007) were followed, with slight modifications, under close expert monitoring. The steps adopted for each house depended on the extent of mold contamination and the part of the building affected. Two of the households completely relocated to new accommodations. The process of remediation essentially involved:


#### **2.5 Questionnaires to determine link between indoor air fungi and health symptoms**

From these households, 47 residents, made up of 20 males and 27 females, ages between 0 and 80 years old, with one or more complaints of respiratory and other symptoms before remediation, were closely monitored after remediation for a period of 3 to 6 months to determine their state of health with respect to the previously observed symptoms. This was done with the aid of structured questionnaires designed to obtain personal information on age, sex, location before and after remediation and other salient features considered in determining the state of health of the residents post remediation. The following questions were asked:


The symptoms included cough, nocturnal cough, dyspnea, sore throat, hoarseness, rhinitis, nasal bleeding, sinusitis, skin rashes, headaches, nausea, diarrhea, tightness in the chest, fever and muscle aches. The questionnaire responses were analyzed to establish the state of health of subjects post remediation. The Pearson correlation test was applied to determine significant association between mold presence or absence in homes and ill health symptoms experienced by residents.

#### **3. Results**

10 Air Quality – Monitoring and Modeling

Because of the seriousness of the problem of indoor air contamination, the WHO guideline recommended that remediation of the conditions that lead to adverse exposure should be given priority to prevent an additional contribution to poor health in populations who are already living with an increased burden of disease (WHO Regional Office for Europe, 2006). The various conditions that supposedly contributed to mold colonization were physically assessed and the necessary remediation procedure determined. Methods proposed in standard manuals and guidelines (Kennedy; Buck, 2002; New York City Department of Health and Mental Hygiene; 2008; North Carolina University Cooperative Extension Services, 2007) were followed, with slight modifications, under close expert monitoring. The steps adopted for each house depended on the extent of mold contamination and the part of the building affected. Two of the households completely relocated to new accommodations.

1. Flooded apartment – Only one of the households experienced flooding which affected two of the rooms. Here the rugs were removed, washed with detergent solution and rinsed with water and Sodium hypochlorite disinfectant solution (diluted according to Manufacturer's instructions). These were dried in the sun. The rooms were also

2. Bad Roofs – Leaky roofs were repaired by replacing punctured sheets and boards. This

3. Fungal growth on asbestos and wooden ceiling sheets – These were removed using absorbent cotton pads soaked in disinfectant, held in place by a wooden handle over a disposable plastic collector. The spot was then scrubbed with detergent solution and dabbed dry, after which paint was applied. Adequate protective measures were observed with every procedure. A physical inspection of each household was done to

**2.5 Questionnaires to determine link between indoor air fungi and health symptoms**  From these households, 47 residents, made up of 20 males and 27 females, ages between 0 and 80 years old, with one or more complaints of respiratory and other symptoms before remediation, were closely monitored after remediation for a period of 3 to 6 months to determine their state of health with respect to the previously observed symptoms. This was done with the aid of structured questionnaires designed to obtain personal information on age, sex, location before and after remediation and other salient features considered in determining the state of health of the residents post remediation. The following questions

a. What symptom(s) or complaint(s) have you experienced and when did it begin?

g. After the repairs and mold clean-up of your home did you notice a difference?

b. Do symptoms occur only when you are in the home, or at other places?

e. Does any other person in your home experience similar symptoms?

c. Do symptoms disappear when you are out of the house?

f. What steps did you take to solve the problem?

d. How frequent was the symptom

washed, disinfected and allowed to dry before re-occupation.

ascertain the absence of moldy areas following remediation.

**2.4 Fungal clean up and remediation** 

The process of remediation essentially involved:

was done by a carpenter.

were asked:

Table 3 shows outcome of questionnaire responses with respect to the eight houses revisited. Mold growth, Arthropod infestation and presence of suggestive symptoms were reported in all (100%) of the houses before remediation. All the households inspected, post remediation, had complete absence of moldy spots. There was significant correlation between reduction in frequency of symptoms of some health problems among residents and mold remediation (r = 0.577; P>0.05). Significant reduction of the following symptoms,


Table 3. Features of Questionnaire responses with respect to the eight residences

cough (83.3%), dyspnea (100%), nasal bleeding (96.9%), skin rash (85.1%), tightness in the chest (83.3%), rhinitis (85.1), muscle aches (92.3%) and eye irritation (92.3%), were observed (Table 4). Reduction in frequency of occurrence of symptoms with respect to age and gender is shown in table 3. 72.3% of individuals experienced significant reduction or total

<sup>1</sup> Building above 25 years old was considered old

Relationship Between Fungal Contamination of

**Age group (Years)**

**4. Discussion** 

2 Number of persons in each group

3 Number of persons with significant reduction of symptoms

Indoor Air and Health Problems of Some Residents in Jos 13

0 – 10 4 2(50.0) 3 3(100.0) 7 5(71.4) 11 – 20 6 3(50.0) 14 11(78.6) 20 14(70.0) 21 – 30 2 2(100.0) 1 1(100.0) 3 3(100.0)

41 – 50 4 4(100.0) 8 5(62.5) 12 9(75.0) 51 – 60 4 2(50.0) 0 0 4 2(50.0)

71 – 80 0 0 1 1(100.0) 1 1(100.0) **Total 20 13(65.0) 27 21(77.8) 47 34(72.3)** 

Indoor air fungi may be responsible for some of the health problems experienced by residents of contaminated homes. There was persistent cough, headache, skin rashes and sneezing in some of the residents before mold cleanup. This did not reoccur after relocation of cleanup of molds. The major factors suggested to contribute to indoor air fungi were observed in a significant percentage of the households. Mold growth was seen in all the houses, at one point of the other with varying abundance and locations. There was significant reduction in frequency of symptoms of some health problems among residents post remediation. This implies that there may be a link between indoor air contamination by fungi and occurrence of such health problems. Jacobs et al., (2007) reported that housing interventions and remediation effected reduction of symptoms of asthma among subjects. Dales et al, (1999) and Shenasa et al, (2007) had, independently, related some ill-health conditions to indoor mold contamination. Some health conditions however barely changed, suggesting that sources other than fungi may be responsible. Symptoms like headaches and fever are also indicators of several other disease entities such as malaria, typhoid and other endemic diseases in the study area. Others such as sinusitis may have assumed a chronic status and hence the little or no reduction recorded. Frequencies of occurrence of symptoms among residents recorded before and after remediation showed remarkable difference between males and females. Shenasa et al, (2007) made a similar observation persons with symptoms of depression. This may reflect the level of exposure to sources of fungi outside the home, and the difference in reaction to triggers of such symptoms in relation to proximity to source of fungi. A similar observation was made by O'connor, (2004) and Mirabell et al, (2006) relating outdoor exposure with asthma in children. Males are more

31 – 40 0 0 0 0 0 0

61 – 70 0 0 0 0 0 0

Table 5. Reduction of symptoms among residents with respect to age and sex

**Males Female Total** 

**N2 R3 (%) N1 R2 (%) N1 R2 (%)**

disappearance of symptoms of ill-health. More females (77.8%) experienced reduction in frequency of symptoms than males. Among the age groups examined, there was no significant difference in the rate of reduction of symptoms. There was no obvious change in the condition of some residents even six months beyond remediation and clear absence of moldy spots in their homes. Some symptoms ranked very low in reduction and were not significantly affected by mold remediation. These included sinusitis (Percentrank =0), nocturnal cough (Percentrank = 0.066), and fever (Percentrank = 0.133). Headache and nausea reduced appreciably but there was no strong association between mold absence and their reduction. Residents also reported relief from some of the symptoms like sneezing, cough and headaches when they left the home but reoccurred on their return. Exposure to moldy environment was influenced by location of building, dampness of the indoor environment and age of building (Student ttest =0.5; P<0.05).


Table 4. Frequency of symptoms of ill-health experienced by residents before and after mold remediation


Table 5. Reduction of symptoms among residents with respect to age and sex

#### **4. Discussion**

12 Air Quality – Monitoring and Modeling

disappearance of symptoms of ill-health. More females (77.8%) experienced reduction in frequency of symptoms than males. Among the age groups examined, there was no significant difference in the rate of reduction of symptoms. There was no obvious change in the condition of some residents even six months beyond remediation and clear absence of moldy spots in their homes. Some symptoms ranked very low in reduction and were not significantly affected by mold remediation. These included sinusitis (Percentrank =0), nocturnal cough (Percentrank = 0.066), and fever (Percentrank = 0.133). Headache and nausea reduced appreciably but there was no strong association between mold absence and their reduction. Residents also reported relief from some of the symptoms like sneezing, cough and headaches when they left the home but reoccurred on their return. Exposure to moldy environment was influenced by location of building, dampness of the indoor

> **Frequency after (days/year)**

Cough, 90 15 75 83.3 0.533

cough 60 35 25 41.7 0.066

Dyspnea 33 0 33 100 0.933 Sore throat 90 20 70 77.8 0.333 Hoarseness 66 12 54 81.8 0.466 Rhinitis 87 13 74 85.1 0.666

bleeding 65 2 63 96.9 0.866

Sinusitis 94 94 0 0 0 Skin rash 67 10 57 85.1 0.666 Headaches 94 40 54 57.4 0.2 Nausea 45 0 45 100 0.933 Diarrhea 52 15 37 71.2 0.266

the chest 30 5 25 83.3 0.533

Fever 87 45 42 48.3 0.133

Table 4. Frequency of symptoms of ill-health experienced by residents before and after mold

Muscle aches 90 20 70 77.8 0.333 Eye irritation 65 5 60 92.3 0.8

**Difference (days/year)**

**Actual** 

**reduction (%) Percentrank** 

environment and age of building (Student ttest =0.5; P<0.05).

**Frequency before (days/year)** 

**Symptoms of ill-health** 

Nocturnal

Nasal

Tightness in

remediation

Indoor air fungi may be responsible for some of the health problems experienced by residents of contaminated homes. There was persistent cough, headache, skin rashes and sneezing in some of the residents before mold cleanup. This did not reoccur after relocation of cleanup of molds. The major factors suggested to contribute to indoor air fungi were observed in a significant percentage of the households. Mold growth was seen in all the houses, at one point of the other with varying abundance and locations. There was significant reduction in frequency of symptoms of some health problems among residents post remediation. This implies that there may be a link between indoor air contamination by fungi and occurrence of such health problems. Jacobs et al., (2007) reported that housing interventions and remediation effected reduction of symptoms of asthma among subjects. Dales et al, (1999) and Shenasa et al, (2007) had, independently, related some ill-health conditions to indoor mold contamination. Some health conditions however barely changed, suggesting that sources other than fungi may be responsible. Symptoms like headaches and fever are also indicators of several other disease entities such as malaria, typhoid and other endemic diseases in the study area. Others such as sinusitis may have assumed a chronic status and hence the little or no reduction recorded. Frequencies of occurrence of symptoms among residents recorded before and after remediation showed remarkable difference between males and females. Shenasa et al, (2007) made a similar observation persons with symptoms of depression. This may reflect the level of exposure to sources of fungi outside the home, and the difference in reaction to triggers of such symptoms in relation to proximity to source of fungi. A similar observation was made by O'connor, (2004) and Mirabell et al, (2006) relating outdoor exposure with asthma in children. Males are more

<sup>2</sup> Number of persons in each group

<sup>3</sup> Number of persons with significant reduction of symptoms

Relationship Between Fungal Contamination of

3, No. 4, pp. 14-24.

May 2007. Accessed January 2009.

Occupational Health & Safety (2006).

172-176.

**7. References** 

Indoor Air and Health Problems of Some Residents in Jos 15

Abbot, S. P. (2002). Mycotoxins and indoor moulds. *Indoor Environment CONNECTIONS* Vol.

Ahiamba J.E.; Dimuna, K.O. & Okogun, G.R.A. (2008). Built Environment Decay and Urban Health in Nigeria. *Journal of Human Ecology*, Vol. 23 No. 3 pp. 259-265. Ayanbimpe, G. M.; Wapwera, S.D.& Kuchin D. (2010). Indoor air mycoflora of residential

British Columbia Non-Profit Housing Association (2007). Determining Good Health as Part

Canadian Center for Occupational Health & Safety (2006). Canadian Centre for

Canadian Center for Occupational Health and Safety (2009). Biological Hazards. Indoor Aie

Cannon, P. F. and Hawksworth, D. L. (1995)The Diversity of Fungi Associated with

Codina, R.; Fox, R. W.; Lockey, R. F.; Demarco, P. & Bagg, A. (2008). Typical levels of

Croft W. A., Jarvis B. B. and Yarawaya C. S. (1986). Airborne outbreak of Trichothecene

Dales, R. E.; Zwanenburg H.; Burnett R. & Franklin C. A. (1991). Respiratory health effects of

Dales, R. E.; Miller, D. & Mc Mullen, E. D. (1997). Indoor Air Quality and Health: Validity

DeHoog, G. S.; Guarro, J.; Gene, J. & Figueras M. J. (2004). Atlas of Clinical Fungi. Atlas

Douwes, J. (2009). Building dampness and its effect on indoor exposure to biological and

Dubey S.; Lanjewar S.; Sahu, M.; Pandey, K. and Kutti, U. (2011). The Monitoring of

Olowoporoku D. (2007). Air Quality Management in Lagos. Air Quality Management

BCBS, (2007). Key Questions: Environment and Housing Quality. Version 1.1, Accessed July

Environmental Protection Agency (2010). Indoor Air Pollution: An introduction for Health Professionals. Available from: http://epa.gov/iaq/pubs/hpguide.html

Buck, K. M. (2002). Fighting Fungi Environmental Contaminants in Healthcare.

Quality Molds and Fungi. Available from: www.ccohs.ca

toxicosis. Atmosphere and Environment vol. 20. Pp. 549-552.

*of Epidemiology* Vol. 26, No. 1, July 1996, pp. 120-125.

Advances in Plant Science vol. 11, pp. 277-302.

*Immunology* Vol. 18, No. 3, pp. 156-162.

Epidemiology Vol. 134, pp. 196-203.

ISSN 978-92-890-4168-3, Copenhagen.

Resource Centre, UWE, Bristol 9 May 2007.

2009, Available from www.designforhealth.net

Wurzburg Germany.

3, No. 4, pp. 13-14.

dwellings in Jos metropolis. *African Health Sciences* Vol. 10, No. 2, (June 2010), pp.

of Housing Solution inBritish Columbia. In: *Housing Affects Health Affects Housing.* 

Vascular Plants: the known, the unknown and the need to bridge knowledge gap.

airborne fungal spores in houses without obvious moisture problems during a rainy season in Florida, USA. *Journal of Investigative Allergology and Clinical* 

home dampness and mold among Canadian xhildren. American Journal of

and Determinants of Reported Home Dampness and Moulds. *International Journal* 

Version 2004.1, CD realization by Weniger, T. Computer Science ii University of

non- biological pollutant, In: Dampness and Mould, vol. 7, pp.7-29; WHO Europe,

Filamentous Fungi in the Indoor , Air Quality and Health. Journal of Phytology vol.

involved in outdoor activities either for relaxation or work. There was no significant difference in the rate of reduction of symptom among the age groups. Of note, however is the only elderly female subject, who suffered prolonged respiratory symptoms (cough and runny nose) prior to remediation, which resolved barely three weeks after remediation. It was clear that the source of the symptoms was indoor mold since the old woman spent most of her days indoors.

Residents are gradually taking interest in the relationship between their residential environment and recurring health problems. Other researchers have suggested that effective communication with building occupants is an important component of all remedial efforts. Awareness of the causes and sources of agent of ill health would enable individuals who believe they have mold-related health problems to seek medical attention.

The health symptoms reduced in frequency after leaky roofs and other risk factors that promote mold growth were attended to. This agrees with the numerous reports which emphasize that the health risks of biological contaminants of indoor air could thus be addressed by considering dampness as the risk indicator. Several widely acknowledged global trends contribute to the conditions associated with increased exposure to dampness and mould (Canadian Center for Occupational Health and Safety, 2009). Conditions that favoured presence of indoor molds was well established in the residences studied and may have contributed to the high occurrence of symptoms. The high relative humidity of the study area is another factor that may encourage growth of fungi. Mshelgaru and Olonitola (2010) in considering the post occupancy contamination of timber buildings in Nigeria, found that fungi were predominant in the rain forest belts which included Plateau State. He found that 64% of the workers on such sites suffered symptoms of the sick building syndrome.

To our understanding, most people in the study area, both in administration and planning positions, as well as residents who are the direct victims of a moldy indoor environment, are still very much ignorant of the enormity of the problem. There is a serious need for more research, enlightenment and implementation of research findings by all stakeholders if a better indoor air quality is to be achieved.

#### **5. Conclusions**

Dampness in indoor environment encourages mold colonization and indoor air contamination. Molds and moisture indoor were associated with symptoms of ill health among residents. Some adverse health effects of indoor fungi on residents were established in the study area. The young children and the elderly were particular prone to mold-related symptoms and remediation provided relief. Reduction of symptoms was easily noticeable among these individuals when molds were removed from the home. The need for collaborative efforts to tackling the problem of indoor contamination is further highlighted. All stakeholders should participate in engaging resources at our disposal to prevention and control of contamination of the indoor environment with pathogenic fungi.

#### **6. Acknowledgement**

The authors wish to acknowledge the following: All member of the households who agreed to be enlisted in this work, the data analyst and family members.

#### **7. References**

14 Air Quality – Monitoring and Modeling

involved in outdoor activities either for relaxation or work. There was no significant difference in the rate of reduction of symptom among the age groups. Of note, however is the only elderly female subject, who suffered prolonged respiratory symptoms (cough and runny nose) prior to remediation, which resolved barely three weeks after remediation. It was clear that the source of the symptoms was indoor mold since the old woman spent most

Residents are gradually taking interest in the relationship between their residential environment and recurring health problems. Other researchers have suggested that effective communication with building occupants is an important component of all remedial efforts. Awareness of the causes and sources of agent of ill health would enable individuals who

The health symptoms reduced in frequency after leaky roofs and other risk factors that promote mold growth were attended to. This agrees with the numerous reports which emphasize that the health risks of biological contaminants of indoor air could thus be addressed by considering dampness as the risk indicator. Several widely acknowledged global trends contribute to the conditions associated with increased exposure to dampness and mould (Canadian Center for Occupational Health and Safety, 2009). Conditions that favoured presence of indoor molds was well established in the residences studied and may have contributed to the high occurrence of symptoms. The high relative humidity of the study area is another factor that may encourage growth of fungi. Mshelgaru and Olonitola (2010) in considering the post occupancy contamination of timber buildings in Nigeria, found that fungi were predominant in the rain forest belts which included Plateau State. He found that 64% of

To our understanding, most people in the study area, both in administration and planning positions, as well as residents who are the direct victims of a moldy indoor environment, are still very much ignorant of the enormity of the problem. There is a serious need for more research, enlightenment and implementation of research findings by all stakeholders if a

Dampness in indoor environment encourages mold colonization and indoor air contamination. Molds and moisture indoor were associated with symptoms of ill health among residents. Some adverse health effects of indoor fungi on residents were established in the study area. The young children and the elderly were particular prone to mold-related symptoms and remediation provided relief. Reduction of symptoms was easily noticeable among these individuals when molds were removed from the home. The need for collaborative efforts to tackling the problem of indoor contamination is further highlighted. All stakeholders should participate in engaging resources at our disposal to prevention and

The authors wish to acknowledge the following: All member of the households who agreed

believe they have mold-related health problems to seek medical attention.

the workers on such sites suffered symptoms of the sick building syndrome.

control of contamination of the indoor environment with pathogenic fungi.

to be enlisted in this work, the data analyst and family members.

better indoor air quality is to be achieved.

**5. Conclusions** 

**6. Acknowledgement** 

of her days indoors.


Relationship Between Fungal Contamination of

Palm Beach, Florida.

http://www.ces.ncsu.edu

vol. 24, No. 7, pp. 703-712.

vol. 39, pp. 103-117.

Available from: www.nyc.gov/health

Environment Digest vol. 10, No. 2, pp. 9-12.

and Technology vol. 4, No. 1, pp. 013-020.

Available from: www.nyc.gov/health

Indoor Air and Health Problems of Some Residents in Jos 17

McNeel, S. V.; and Kreutzer R.A. (1996). Fungi and Indoor Air Quality. Health and

Mirabell, M. C.; Wing, S.; Marshall, S. W. & Wilkosky, T. C. (2006). Asthma Symptoms

Swine Feeding Operations. *Paediatrics* Vol. 118, No. 1, July 2006, pp. e66-e75. Moloughney, B. (2004). Housing and Population Health: The State of Current research

Morey P. R. (1999). Indoor Air Quality, Fungi, and Removal of Fungal Colonization.

Morgan W. J.; Crain, E.F.; Gruchalla, R.S.; O'Connor, G.T. Kattan, M.; Evans, R. III et al.

New York City Department of Health and Mental Hygiene (2009). Guidelines on

North Carolina Extension Service (2009). Mildews Prevention in the Home. Housing and Home Furnishing Specialists. Revised by Kirby S. D. Available from:

O'Connor, G. T.; Walter, M.; Mitchell, H.; Kattan, M.; Morgan, W. J.; Gruchalla R. S.;

Pasanen A.L., Lapalainen S., and Pasanen P. (1996). Volatile organic Metabolites Associated with some toxic fungi and their mycotoxins. Analyst vol. 12, pp. 1949-1953. Pasanen A. L., Korip, A., Kasanen, J.P. and Pasanen P. (1996). Critical aspects on significant

Petrescu, C.; Suciu, O.; Lonovici R.; Herbarth, O.;Franck U. and Schlink U. (2011).

New York City Department of Health and Mental Hygiene (2008). Preventing and Cleaning

Nielson, K. F. (2003). Mycotoxin Production by Indoor Molds. Fungal Genetics and Biology

Robins C. & Morell J. (2007). Mold Housing and Wood. In: Western Wood Products Association 2007. Available from: © 2007 Western Wood Products Association.

*Immunology* Vol. 114, No. 3, September 2004, pp. 599 - 606.

with asthma. *New England Journal of Medicine* Vol. 351, pp.1068–1080. Mshelgaru I. H. and Olonitola O.S. (2010). Health and Safety Conditions of Buikding

Knowledge. Ottawa: Canadian Institute for Health Information.

Among Adolescents Who Attend Public Schools That Are Located Near Confined

Presented at: Mealy Sick Building Litigation Conference November 1999. West

(2004). Results of a home-based environmental intervention among urban children

Maintenance Sites in Nigeria: Evaluating the Post-Occupancy Contamination of Timber Buildings buy Microorganisms. African Journal of Environmental Science

Assessment and Remediation of Fungi in Indoor Environments November 2008.

Pongracic, J. A.; Smartt, E.; Stout, J. W.; Evans, R.; Crain, E. F. & Burge, H. A. (2004). Airborne fungi in homes of children with asthma in low-income urban communities: The inner-City asthma study. *Journal of Allergy and Clinical* 

of microbial volatile metabolite as indoor air pollutants. Environment International

Respiratory Health Effectof Air Pollution and Climate Parameters in the Population of Drobeta Turnu-Severin, Romania. In: Air Pollution-New Developments, Moldoveanu A. M. (Ed.), ISBN: 978-953-307-527-3, InTech, Available from: http://www.intechopen.com/article/show/title/respiratory-health-effects-of-airpollution-and-climate-parameters-in-the-population-of-population-of-drobeta-turn

Mold Growth, Fact Sheet for Building Owners and Managers. Accessed July 2011.


http://www.communities.gov.uk/index.asp?id=1152820


European Environment and Health Information System (2007). Children Living in Homes with Problems of Damp. Fact Sheet No. 3.5- May 2007- Code RPG3\_Hous\_Ex2. Enhis (2008). Children living in homes with problems of dampness. In Enhis Version 1.8, 21

Federal Republic of Nigeria (2007). National Environmental Standard and Regulation

FEPA, (1999). National Master Plan for Public Awareness on environment and Natural

FGN (1988). Federal Environmental Protection Agency Decree 58, 1988,Federal Ministry of

Gots, R. E.; Layton, N. J. (2003). Indoor Health: Background Levels of Fungi. Journal of

Hawksworth, D. L. (1991). The fungal dimension of biodiversity: magnitude, significance,

Hawksworth, D. L. (1998). Kingdom Fungi: Fungal Phylogeny and Systematics. In: Collier

Heseltine E. and Rosen, J. (2009). WHO Guidelines for Indoor Air Quality: Dampness and

Housing Health and Safety Rating Systems London Communities and Local Government

Horner, W. E.; Worthan, A. G. & Morey P. R. (2004). Air- and dustborne mycoflora in houses

Howell E.; Harris, L. J. & Popkin S. J. (2005). The health status of Hope VI public housing residents. *Journal of Health Care Poor & Underserved* Vol. 16, pp. 273–285.

Husman T. (1996). Health effects of indoor-air microorganisms. *Scand J Work Environ Health* 

Jacobs, E. D.; Kelly, T. & Sobolewsky, J. (2007). Linking Public Health, Housing, and Indoor

Johaninning E., Biagini R., Hull D., Morey P. R., Jarvis, B., and Landsbergis, P. (1996). Health

Kennedy, R. (n.d). Strategies for Preventing or Removing Mold Growth After

Kuhn, D. M. & Ghannoum, M. A. (2003). Indoor mold, toxigenic fungi, and Stachybotrys

L., Balows, A., Sussman, M. (eds.). Topley and Wilson's Microbiology and

Mold. WHO Europe Damness and Molds. Available from: www.euro.who.int

free of water damage and fungal growth. *Applied Environmental Microbiology*. Vol.

Environmental Policy: Successes and Challenges at Local and Federal Agencies in the United States. *Environmental Health Perspectives* Vol. 115, No. 6, June 2007,

and Immunology Studies following exposure to toxigenic fungi (Starchybotrys chartarum) in a water-damaged office environment. Int. Arch. Occup. Environ

Contamination, Handout 4. In: US Department of Agriculture,University of

chartarum: infectious disease perspective. *Clinic Microbiology Reviews* Vol. 16, No.1,

Resource conservation in Nigeria, FEPA Garki, Abuja.

Occupational Health and Hygiene vol. 64, No. 4, 427-438.

and conservation. *Mycological Reasearch* Vol. 95, pp. 641-655.

Microbial Infections , 9th edition. Arnold, London. 1988; pp. 43-54.

Information and Culture, Lagos, Nigeria.

(2007). Accesses July 2009. A vailable from: http://www.communities.gov.uk/index.asp?id=1152820

70, No. 11, November 2004, pp. 6394-6400.

http://www.epa.gov/mold/moldresources.html

Health. Vol. 68, pp.207-218.

Akansas, Cooperative and Extension Services.

Vol. 223, pp. 5-13.

pp.976-982.

144-172.

Agency (Establishment) Act, 2007. Federal Republic of Nigeria Official Gazette vol.

October 2008.

94, No. 92., 31 July 2007.

ISBN: 978 92 870 4168 3.


**2** 

*Romania* 

**A Qualitative Study of** 

*Isotopic Technologies, Rm. Valcea,* 

Gabriela Raducan\* and Ioan Stefanescu

**Air Pollutants from Road Traffic** 

*The National Research and Development Institute for Cryogenics and* 

Air quality in urban areas continues to represent a major concern, taking into account the impact of pollution on environment and human health. Air quality study involves researchers of several scientific disciplines and requires the development of theoretical, practical and methodological concepts, always new, for understanding, controlling and combating urban air pollution. Together with monitoring, modeling air pollution is a very important tool in the management of air quality. It enables efficient verification in order to reduce emissions and make prediction of pollution levels associated with changes in urban

In recent decades many experimental and modeling studies have been conducted, in order to obtain information on the dispersion of pollutants and transformations suffered by them in the street canyon. Depending on the purpose, different techniques were used for monitoring and modeling. Some of these studies are purely experimental, which means they rely only on small-scale measurements. Others are purely theoretical, concentrating on investigating the various schemes of atmospheric fluid flows and dispersion of pollutants, using mathematical models. The most common are those that combine mathematical

This work presents a series of studies for the correlation of emissions from road traffic, meteorology and concentrations of pollutants at a local scale. It is a qualitative analysis meaning that the study compares the concentration of fictitious values. The aim of this study has been to enrich and understand how to make an analysis of the spatial and

The most important urban air pollutants in terms of emissions and the impact on the environment and human health which could be taken into account are: carbon monoxide (CO), nitrogen oxides (NO and NO2), volatile organic compounds (VOCs), ozone (O3) and

modeling with field measurements, as the models for street canyon.

temporal variability of air pollution in urban areas.

aerosol particles with a diameter less than 2.5μm (PM2,5).

**2. Urban air pollutants** 

Corresponding Author

 \*

**1. Introduction** 

infrastructure.


## **A Qualitative Study of Air Pollutants from Road Traffic**

Gabriela Raducan\* and Ioan Stefanescu *The National Research and Development Institute for Cryogenics and Isotopic Technologies, Rm. Valcea, Romania* 

#### **1. Introduction**

18 Air Quality – Monitoring and Modeling

Sautour, M.; Sixt, N.; Dalle, F.; L'Ollivier, C.; Fourquenet, V.; Calinon, C.; Paul, K.; Valvin, S.;

Shelton, B. G.; Kirkland K. H.; Flanders W. D. & Morris, G. K. (2002). Profiles of airborne

Solomon A. O.; Ike E. E.; Ashaano E. C. and Jwanbot D. N. (2002). Natural Background

Stetzenbach, L. D.; Amman, H.; Johanning, E.; King, G & Shaughnessy, R. J. (2004).

Straus, D. C. & Wilson, S. C. (2006). Reespiratory Trichothecene Mycotoxins can be

Verhoeff, A. P. & Burge, H. A., (1997). Health risk assessment of fungi in home

Vocaturo, E.; Kunseler, E.; Slovakova G.; Ruut, J.; Cavoura O. & Otorepec P. (2008). Children

Workplace Health, Safety and Compensation Commission of New Brunswick, (2000):

Wu, F.; Biksey, T. & Karol, M. H. (2007). Can Mold Contamination of Homes Be Regulated?

Zaslow S. A.; and Genter M. B. (1993). Mold, Dust Mites, Fungi, Sporesand Pollen:

Microbiology (2004), Indoor Air Quality, December, 2004, pp. 1-20.

buildings. *Journal of Allergy and Clinical Immunology* Vol. 118, pp. 760. U.S. Environmental Protection Agency (2003). An Introduction to Indoor Air Quality (IAQ).

Environmental Microbiology, Vol. 68, No. 4, April 2002, pp. 1743-1753. Shenassa, E. D.; Liebhaber, A.; Daskalakis, C.; Braubach, M. & Brown M. (2007). Dampness

vol. 407, No. 12, pp.3766-3771.

*Journal of Public Health* Vol.97, No.10, 1893-1899.

AfricanJournal of Natural Sciences vol. 5.

Biological Pollutants. Accessed, April 2011.

Home\ Environment and health issues\ Housing

Worksafe Bulletin (1999). Building Related Moulds. Bulletin No. 194.

*Technology*,Vol.41, No. 14, pp. 4861-4867.

5. Available from: www.ces.ncsu.edu

Maurel, A.; Aho, S.; Couillault, G.; Cachia, C.; Vagner, O.; Cuisenier, B.; Caillot, D. & Bonnin, A. (2009). Profiles and seasonal distribution of airborne fungi in indoor and outdoor environments at a French hospital.Science of the Total Environment,

fungi in buildings and.outdoor environments in the United States. Applied

and Mold in the Home and Depression: An Examination of Mold-Related Illness and Perceived Control of One's Home as Possible Depression Pathways. *American* 

Radiation on the Jos Plateau and use of the Rock for House Construction.

Microorganisms, Mold, and Indoor Air Quality. In: American Society for

Demonstrated in the air of Stachybotrys chartarum contaminated buildings

environments. *Annals of Allergy*, *Asthma and Immunology.* Vol. 78, No. 6, pp.544-556.

living in homes with problems of dampness. In: *ENHIS, World Health Organisation*, Version 3.5, Menu. Bilthoven, accessed 22 October 2008. Available from: RIVM,

Microbials and indoor air quality Moulds and Bacteria. Published: December, 2000

Lessons Learned from Radon and Lead Policies. *Environmental Science* 

Bioaerosols in the human Environment. North Carolina Extension Service FCS-360-

Air quality in urban areas continues to represent a major concern, taking into account the impact of pollution on environment and human health. Air quality study involves researchers of several scientific disciplines and requires the development of theoretical, practical and methodological concepts, always new, for understanding, controlling and combating urban air pollution. Together with monitoring, modeling air pollution is a very important tool in the management of air quality. It enables efficient verification in order to reduce emissions and make prediction of pollution levels associated with changes in urban infrastructure.

In recent decades many experimental and modeling studies have been conducted, in order to obtain information on the dispersion of pollutants and transformations suffered by them in the street canyon. Depending on the purpose, different techniques were used for monitoring and modeling. Some of these studies are purely experimental, which means they rely only on small-scale measurements. Others are purely theoretical, concentrating on investigating the various schemes of atmospheric fluid flows and dispersion of pollutants, using mathematical models. The most common are those that combine mathematical modeling with field measurements, as the models for street canyon.

This work presents a series of studies for the correlation of emissions from road traffic, meteorology and concentrations of pollutants at a local scale. It is a qualitative analysis meaning that the study compares the concentration of fictitious values. The aim of this study has been to enrich and understand how to make an analysis of the spatial and temporal variability of air pollution in urban areas.

#### **2. Urban air pollutants**

The most important urban air pollutants in terms of emissions and the impact on the environment and human health which could be taken into account are: carbon monoxide (CO), nitrogen oxides (NO and NO2), volatile organic compounds (VOCs), ozone (O3) and aerosol particles with a diameter less than 2.5μm (PM2,5).

<sup>\*</sup> Corresponding Author

A Qualitative Study of Air Pollutants from Road Traffic 21

Aerosol particles, both those produced naturally, as well as those produced by anthropogenic sources are involved in the extinction of solar radiation, can spread the light and reduce visibility and, therefore, can influence, both direct and indirect (cloud formation), the global or regional climate. Inhalation by humans and animals may have adverse effects on health. The small particles, with the diameter less than 1µm, behave like gases; they are involved in the Brownian movement, follow the lines of fluid current near the obstacles, are able to coagulate and are deposited on the surface of the earth very hard. Particles larger than 10µm are strongly affected by gravity, being involved in processes of

The preliminary assessment of atmospheric pollution in urban area is a very important step in finding the locations where to deploy fixed monitoring stations. Such locations are selected according to information which includes: distribution of pollutants sources in the urban area, the expected maximum concentration points (hot points), the meteorological and topographical conditions, the model applications and other suitable features. The monitoring sites have to be representative of a sufficiently large area in the vicinity, so that the sampling station can be considered representative of a larger area or representative of

The examination of modern methods used for air monitoring permits to answer the question: how air monitoring has to be performed. The variety of analytical methods offers a wide selection of procedures, which can be carried out by means of static, mechanized or automatic devices. The choice will depend upon the use of the monitoring data and the aim

Static devices, such as the ones employed to measure the amount of deposited particles (dust fall) are used for mapping, for definition of special problem areas and for general survey.

Mechanized bubbler devices are used to collect nitrogen dioxide, mercury, and other gases and vapors. These samplers, although typically designed for collecting 24 h integrated samples, can be modified to collect 1 or 2 h samples in sequence, and thus allow definition

In automatic sampler—analyzers, collection and analysis are combined in a single device. These automatic instruments produce continuous analysis, with the output in a machine-

The approach of the various instruments described to monitor a pollutant either by laboratory analysis or by analyzers is the same. These instruments can be called 'point' sensors as they measure the concentration of the given pollutant at a single point. Another approach for monitoring is the 'remote sensing'. This term indicates the use of instruments which can provide the average concentration of a pollutant in a certain area either by looking at the emissions as they exit at the stack output or by sampling an optical volume at a point within the plume and conducting a spatially integrated measurement across the diameter of the plume. Remote sensing can be performed also by means of a 'long-path sensor'; this term indicates any device which permits one to measure extended or diffuse

readable format or in a suitable form to a central data-acquisition facility.

sources, such as oil refineries and chemical complexes between two points.

dry and wet settling.

which has to be reached.

of diurnal variations.

**3. Methods of monitoring atmospheric pollutants** 

sites characterized by similar environmental conditions.

Carbon monoxide is found in abundance in urban air pollution in quantities that no other toxic gases are. Carbon monoxide is formed, usually by one of the following three processes: incomplete combustion of fuel containing carbon, reactions of carbon dioxide and materials containing carbon at high temperatures and dissociation of carbon dioxide at high temperatures.

The main nitrogen oxides are nitrogen monoxide (NO) and nitrogen dioxide (NO2), and together are called noxious NOx. About the effects on health, NO are considered harmless, at least at concentrations that usually exist in the street. Instead, NO2 can cause serious problems. Only a small fraction of the gases emitted by engines is represented by NO2, the principal amount being NO. This NO2 in ambient air is mainly due to further oxidation NO. Chemistry of nitrogen oxides is quite complex, but due to the small time the pollutants reside in the street canyon, the reactions of practical interest are: oxidation of nitrogen monoxide to nitrogen dioxide, nitrogen dioxide decomposition in nitrogen monoxide and oxygen and the forming of ozone resulting from the reaction between the oxygen molecule and the oxygen root.

In the category of volatile organic compounds (VOCs) there are a large number of air pollutants, emitted from industrial sources and urban traffic, as well as from other sources. In terms of chemistry, volatile organic compounds include aliphatic and aromatic hydrocarbons, halogenated hydrocarbons, some alcohols, esters and aldehydes. The importance of these compounds is obvious, considering the fact that the agency EPA (Environmental Protection Agency) has designated them as belonging to the six types of critical air pollutants.

Both natural sources and anthropogenic sources contribute to VOC emissions. Natural sources include oil, forest fires and some chemical reactions to produce these compounds. The main anthropogenic sources include fossil fuel combustion at high temperature, the obtaining of oil refined and unrefined, incineration, burning of crops and other debris after collection.

Ozone is formed and destroyed chemically in the atmosphere, the stratosphere or the low troposphere. The complicated processes of training and distribution of ozone in the stratosphere are of great importance for the amount of ultraviolet radiation reaching the earth's surface and for the energy balance of earth-atmosphere system. Ozone is formed under the influence of direct solar radiation from volatile organic compounds and nitrogen oxides. Both the training and the destruction of ozone are photochemical reactions. Ozone concentration in the troposphere has increased lately, especially in the northern hemisphere, due to increasing anthropogenic emissions. Ozone is a strong oxidizer and high concentrations at this level can cause strong negative effects on the environment and human health.

Aerosol particles cover a wide dimensional field, from the cluster of a few molecules to particles greater than 100 µm in diameter. Dimensional distribution of aerosol particles presents a certain regularity that is the result of production of nuclei from different sources and loss through different processes. Each type of particles, taken individually, has its own characteristics, related to its size, geometry, concentration, chemical composition and physical properties.

The particles with less than 2.5µm diameter are a real concern because they remain suspended in the atmosphere, where, depending on size, can be settled relatively difficult.

Carbon monoxide is found in abundance in urban air pollution in quantities that no other toxic gases are. Carbon monoxide is formed, usually by one of the following three processes: incomplete combustion of fuel containing carbon, reactions of carbon dioxide and materials containing carbon at high temperatures and dissociation of carbon dioxide at high

The main nitrogen oxides are nitrogen monoxide (NO) and nitrogen dioxide (NO2), and together are called noxious NOx. About the effects on health, NO are considered harmless, at least at concentrations that usually exist in the street. Instead, NO2 can cause serious problems. Only a small fraction of the gases emitted by engines is represented by NO2, the principal amount being NO. This NO2 in ambient air is mainly due to further oxidation NO. Chemistry of nitrogen oxides is quite complex, but due to the small time the pollutants reside in the street canyon, the reactions of practical interest are: oxidation of nitrogen monoxide to nitrogen dioxide, nitrogen dioxide decomposition in nitrogen monoxide and oxygen and the forming of ozone resulting from the reaction between the oxygen molecule

In the category of volatile organic compounds (VOCs) there are a large number of air pollutants, emitted from industrial sources and urban traffic, as well as from other sources. In terms of chemistry, volatile organic compounds include aliphatic and aromatic hydrocarbons, halogenated hydrocarbons, some alcohols, esters and aldehydes. The importance of these compounds is obvious, considering the fact that the agency EPA (Environmental Protection Agency) has designated them as belonging to the six types of

Both natural sources and anthropogenic sources contribute to VOC emissions. Natural sources include oil, forest fires and some chemical reactions to produce these compounds. The main anthropogenic sources include fossil fuel combustion at high temperature, the obtaining of oil

Ozone is formed and destroyed chemically in the atmosphere, the stratosphere or the low troposphere. The complicated processes of training and distribution of ozone in the stratosphere are of great importance for the amount of ultraviolet radiation reaching the earth's surface and for the energy balance of earth-atmosphere system. Ozone is formed under the influence of direct solar radiation from volatile organic compounds and nitrogen oxides. Both the training and the destruction of ozone are photochemical reactions. Ozone concentration in the troposphere has increased lately, especially in the northern hemisphere, due to increasing anthropogenic emissions. Ozone is a strong oxidizer and high concentrations

Aerosol particles cover a wide dimensional field, from the cluster of a few molecules to particles greater than 100 µm in diameter. Dimensional distribution of aerosol particles presents a certain regularity that is the result of production of nuclei from different sources and loss through different processes. Each type of particles, taken individually, has its own characteristics, related to its size, geometry, concentration, chemical composition and

The particles with less than 2.5µm diameter are a real concern because they remain suspended in the atmosphere, where, depending on size, can be settled relatively difficult.

refined and unrefined, incineration, burning of crops and other debris after collection.

at this level can cause strong negative effects on the environment and human health.

temperatures.

and the oxygen root.

critical air pollutants.

physical properties.

Aerosol particles, both those produced naturally, as well as those produced by anthropogenic sources are involved in the extinction of solar radiation, can spread the light and reduce visibility and, therefore, can influence, both direct and indirect (cloud formation), the global or regional climate. Inhalation by humans and animals may have adverse effects on health. The small particles, with the diameter less than 1µm, behave like gases; they are involved in the Brownian movement, follow the lines of fluid current near the obstacles, are able to coagulate and are deposited on the surface of the earth very hard. Particles larger than 10µm are strongly affected by gravity, being involved in processes of dry and wet settling.

#### **3. Methods of monitoring atmospheric pollutants**

The preliminary assessment of atmospheric pollution in urban area is a very important step in finding the locations where to deploy fixed monitoring stations. Such locations are selected according to information which includes: distribution of pollutants sources in the urban area, the expected maximum concentration points (hot points), the meteorological and topographical conditions, the model applications and other suitable features. The monitoring sites have to be representative of a sufficiently large area in the vicinity, so that the sampling station can be considered representative of a larger area or representative of sites characterized by similar environmental conditions.

The examination of modern methods used for air monitoring permits to answer the question: how air monitoring has to be performed. The variety of analytical methods offers a wide selection of procedures, which can be carried out by means of static, mechanized or automatic devices. The choice will depend upon the use of the monitoring data and the aim which has to be reached.

Static devices, such as the ones employed to measure the amount of deposited particles (dust fall) are used for mapping, for definition of special problem areas and for general survey.

Mechanized bubbler devices are used to collect nitrogen dioxide, mercury, and other gases and vapors. These samplers, although typically designed for collecting 24 h integrated samples, can be modified to collect 1 or 2 h samples in sequence, and thus allow definition of diurnal variations.

In automatic sampler—analyzers, collection and analysis are combined in a single device. These automatic instruments produce continuous analysis, with the output in a machinereadable format or in a suitable form to a central data-acquisition facility.

The approach of the various instruments described to monitor a pollutant either by laboratory analysis or by analyzers is the same. These instruments can be called 'point' sensors as they measure the concentration of the given pollutant at a single point. Another approach for monitoring is the 'remote sensing'. This term indicates the use of instruments which can provide the average concentration of a pollutant in a certain area either by looking at the emissions as they exit at the stack output or by sampling an optical volume at a point within the plume and conducting a spatially integrated measurement across the diameter of the plume. Remote sensing can be performed also by means of a 'long-path sensor'; this term indicates any device which permits one to measure extended or diffuse sources, such as oil refineries and chemical complexes between two points.

A Qualitative Study of Air Pollutants from Road Traffic 23

Analyzer for nitrogen oxides (NO, NO2,) works on the principle of chemiluminescence. Gas sample is first filtered through a filter that can have a Teflon membrane. Nitric oxide present in the sample reacts with ozone generated within the analyzer, producing nitrogen dioxide in the excited state. It emits electromagnetic radiation that is detected and measured by photomultiplier tube, being proportional to the concentration of NO in the sample. A catalytic converter that reduces NO2 to NO is used to measure NO2, allowing measurement of total amount of nitrogen oxides (NOx). The measuring cycle of the NO and NO2

Low levels of O3 could be measured using a spectrophotometer. Ozone absorbs UV radiation of 254 nm wavelength, according to Beer-Lambert law (Schneider et al, 1997). The analyzer determines the concentration of ozone in the sample by measuring the attenuation of UV radiation due to ozone absorption wavelength above mentioned. The gas sample to be measured is led into an analyzer using a diaphragm pump. It passes first through a filter for aerosol particles or by a catalytic converter. The converter selectively changes the ozone sample in O2, thus generating a reference gas to contain no ozone. When the reference gas passes through the cell is set "0" of light intensity. This value is kept as a reference, by the microprocessor. When the sample gas replaces the reference gas the luminous intensity is measured and then the system calculates the difference which is proportional to the mass of ozone contained in the cell. To obtain a relationship of proportionality between light intensity and amount of light absorbed, the ozone analyzer was equipped with a temperature and pressure compensator which includes a temperature converter and a

To measure the particulate matters (PM), the element oscillating microbalance (Tapered Element Oscillating Microbalance- TEOM) could be used or a device based on beta

TEOM is a device used to measure real-time mass concentration of particles with sizes smaller than 2,5 m (or 10 m), indoors and outdoors. It records: mass concentration, mass

Beta radiation attenuation monitor (BAM-Beta Attenuation Monitor): the beta radiation flow, emitted by a radioactive source located inside the monitor, is directed to the filter containing deposited particles and is attenuated in proportion to the mass of collected particles. This system also allows continuous hourly (or 30 minutes) measurements of

In the initial dispersion of pollutants from the emission sources, they are dispersed into the environment by movements that depend on weather and surveying conditions. In the dispersion process, significant weather phenomena are related to wind speed and direction,

The wind speed plays an important role in transport and dilution of pollutants. If the emission speed is relatively constant, a doubling of wind speed will result in halving the

amount of pollutants, so the concentration is inversely proportional to wind speed.

concentrations is very important for the NOx analyzer.

radiation. Both methods are based on basic physical principles.

ratio and total mass accumulated in the filter cartridge.

aerosol particles of different sizes.

turbulence and atmospheric stability.

**4. The dispersion process** 

pressure converter.

For some pollutants however, there are no methods of measurement based on sensors and monitors, and in that case the system requires a periodic transfer of this data.

Choosing compounds to be measured is related to the type of monitoring stations. These, according to the standards, may be:


In recent years the focus has been on the use of modern techniques, easy to use, inexpensive and having a good accuracy. Thus, modern information systems for monitoring air quality were made, which enable integrated approach and direct and rapid access to data. They can be used for the purpose of evaluating and planning future actions. Such a surveillance system could be consisted of:


*Automatic analyzers* can provide continuous measurements of concentrations of pollutants without further laboratory analysis. Monitoring procedures are fully automated and use analysis techniques such as chemiluminescence or infrared absorption. Concentration determined by the analyzer is recorded in a data storage device that can be internal or external. The monitoring stations usually operate with multiple analyzers (eg CO, NO, NO2, NOx, O3), because the price is much smaller than the individual analyzers, taken together. A typical monitoring station, along with the analyzer, implies the existence of a room weatherproof and highly secure, air-conditioning system, sample collection lines, a powerful vacuum pump to absorb quantities of sample in ambient air, calibration gas, a data storage system and another data transfer.

There are several types of measurement devices with various features. It is very important to choose them properly.

To measure CO and CO2 concentrations devices operating on the principle of non-dispersive IR analysis could be used. The IR radiation is absorbed by the CO and CO2 at specific wavelengths. Gas sample found in the room where measurements are made, is exposed to IR radiation coming from a powerful source. Absorbed energy is measured. The camera is filled with sample gas and clean air, in turn, and the difference of energy absorbed is calculated.

For some pollutants however, there are no methods of measurement based on sensors and

Choosing compounds to be measured is related to the type of monitoring stations. These,

1. Urban Background Station (for nitrogen oxides, ozone, aerosol particles and sulfur

2. Industrial / Residential Station (for nitrogen oxides, sulfur dioxide, aerosol particles and possibly other compound-specific ammonia or volatile organic compounds); 3. Traffic Station (for nitrogen oxides, carbon monoxide, aerosol particles, volatile organic

4. Regional Background Station (for nitrogen oxides, sulfur dioxide, ozone, aerosol

In recent years the focus has been on the use of modern techniques, easy to use, inexpensive and having a good accuracy. Thus, modern information systems for monitoring air quality were made, which enable integrated approach and direct and rapid access to data. They can be used for the purpose of evaluating and planning future actions. Such a surveillance

Data package transfer systems and programs related to quality assurance and control

Numerical or statistical models (including models of dispersion and meteorological

 Distribution systems and communication of results through network in order to inform users and the public or for preparing strategies to reduce pollution by the authorities. *Automatic analyzers* can provide continuous measurements of concentrations of pollutants without further laboratory analysis. Monitoring procedures are fully automated and use analysis techniques such as chemiluminescence or infrared absorption. Concentration determined by the analyzer is recorded in a data storage device that can be internal or external. The monitoring stations usually operate with multiple analyzers (eg CO, NO, NO2, NOx, O3), because the price is much smaller than the individual analyzers, taken together. A typical monitoring station, along with the analyzer, implies the existence of a room weatherproof and highly secure, air-conditioning system, sample collection lines, a powerful vacuum pump to absorb quantities of sample in ambient air, calibration gas, a

There are several types of measurement devices with various features. It is very important

To measure CO and CO2 concentrations devices operating on the principle of non-dispersive IR analysis could be used. The IR radiation is absorbed by the CO and CO2 at specific wavelengths. Gas sample found in the room where measurements are made, is exposed to IR radiation coming from a powerful source. Absorbed energy is measured. The camera is filled with sample gas and clean air, in turn, and the difference of energy absorbed is calculated.

Sampling systems, sensors and monitors installed in special monitoring stations;

Database for the air quality, weather and the emission of pollutants;

Modules for graphic presentation of the results of tests;

data storage system and another data transfer.

monitors, and in that case the system requires a periodic transfer of this data.

compounds (benzene) and sulfur dioxide (optional));

particles, volatile organic compounds).

according to the standards, may be:

dioxide (optional));

system could be consisted of:

forecasting models);

to choose them properly.

data;

Analyzer for nitrogen oxides (NO, NO2,) works on the principle of chemiluminescence. Gas sample is first filtered through a filter that can have a Teflon membrane. Nitric oxide present in the sample reacts with ozone generated within the analyzer, producing nitrogen dioxide in the excited state. It emits electromagnetic radiation that is detected and measured by photomultiplier tube, being proportional to the concentration of NO in the sample. A catalytic converter that reduces NO2 to NO is used to measure NO2, allowing measurement of total amount of nitrogen oxides (NOx). The measuring cycle of the NO and NO2 concentrations is very important for the NOx analyzer.

Low levels of O3 could be measured using a spectrophotometer. Ozone absorbs UV radiation of 254 nm wavelength, according to Beer-Lambert law (Schneider et al, 1997). The analyzer determines the concentration of ozone in the sample by measuring the attenuation of UV radiation due to ozone absorption wavelength above mentioned. The gas sample to be measured is led into an analyzer using a diaphragm pump. It passes first through a filter for aerosol particles or by a catalytic converter. The converter selectively changes the ozone sample in O2, thus generating a reference gas to contain no ozone. When the reference gas passes through the cell is set "0" of light intensity. This value is kept as a reference, by the microprocessor. When the sample gas replaces the reference gas the luminous intensity is measured and then the system calculates the difference which is proportional to the mass of ozone contained in the cell. To obtain a relationship of proportionality between light intensity and amount of light absorbed, the ozone analyzer was equipped with a temperature and pressure compensator which includes a temperature converter and a pressure converter.

To measure the particulate matters (PM), the element oscillating microbalance (Tapered Element Oscillating Microbalance- TEOM) could be used or a device based on beta radiation. Both methods are based on basic physical principles.

TEOM is a device used to measure real-time mass concentration of particles with sizes smaller than 2,5 m (or 10 m), indoors and outdoors. It records: mass concentration, mass ratio and total mass accumulated in the filter cartridge.

Beta radiation attenuation monitor (BAM-Beta Attenuation Monitor): the beta radiation flow, emitted by a radioactive source located inside the monitor, is directed to the filter containing deposited particles and is attenuated in proportion to the mass of collected particles. This system also allows continuous hourly (or 30 minutes) measurements of aerosol particles of different sizes.

#### **4. The dispersion process**

In the initial dispersion of pollutants from the emission sources, they are dispersed into the environment by movements that depend on weather and surveying conditions. In the dispersion process, significant weather phenomena are related to wind speed and direction, turbulence and atmospheric stability.

The wind speed plays an important role in transport and dilution of pollutants. If the emission speed is relatively constant, a doubling of wind speed will result in halving the amount of pollutants, so the concentration is inversely proportional to wind speed.

A Qualitative Study of Air Pollutants from Road Traffic 25

weather conditions, it is dominant in windy, in neutral atmospheric conditions. Turbulence has the effect of increasing the dispersion process, even when there is mechanical turbulence, on in certain areas, high concentrations of pollutants could be recorded,

In large urban centres there are a multitude of fixed and mobile pollution sources. This urban community forms a heated and polluted island which is one of the biggest problems of the contemporary world. In an air quality analysis the ventilation rate of the entire region must be taken into account, because the dispersion can be influenced by both atmospheric mesoscale and macroscale movements. The atmospheric conditions as the anticyclones or thermal inversions slow the dispersion, resulting in high concentrations of pollutants. On the other hand, high weather variability such as moving cyclones, associated with strong winds, will enhance the dispersion of pollutants in urban areas. The pollution can move on, over very large areas, affecting air quality within hundreds of kilometers distances

Dispersion of car emissions is a complex process that acts at different scales of time and space and is strongly influenced by weather variability and air dynamics around vehicles and buildings. The amount of the exhaust gas varies over time, and it moves in time and space. Immediately after being discharged, the gases are dispersed due to turbulence that occurs around the vehicle, a phenomenon influenced by the non-stationary flow of air masses. A few seconds later, the dispersion is further accelerated by turbulence from other

In parallel with the mixture due to movement of vehicles, emissions are dispersed by air masses that have a dynamic movement because of the weather and topographical

Movement of vehicles creates uneven air movements (Plate, 1998) which mix the pollutants emitted from the exhaust pipe. After moving, the vehicle vortices arise. Basically, there isn't a database of local-scale air movements produced by motor vehicles in urban areas

The influence of driving mode on vehicle emissions is of particular importance in such a study as higher emission rates are associated with certain traffic characteristics. (Tate, 2004) The urban environment is characterized by different widths streets and buildings with different geometries. Dispersion of emissions from vehicles in narrow streets from urban areas has become a very intensely studied research theme (Chan et al, 2002, Xie et al, 2003) because it is known that very high levels of pollutants can be found within these streets,

The term "street canyon" refers to a relatively narrow street with buildings built by both sides, continuously (Nicholson, 1975). Dimensions of the street canyons are expressed usually by the characteristic ratio H / W, where H is the average height of a building located on both sides of the street, and W is the width of the street. A street canyon can be considered regular if the characteristic ratio is approximately equal to 1 and there are no major gaps in buildings on both sides of the canyon, which are like walls. If the wind blows obliquely to the direction of street-canyon, air flow reflection (wind in the roof) in the

which means, most often, very high levels of pollution to which citizens are exposed.

conditions (the buildings, trees, open areas, etc.) (Dobre et al, 2005).

(Godish, 1997).

(Briggs, 1969).

(Tate, 2004).

vehicles, moving within the street area.

The speed of the horizontal wind is affected by friction, which is proportional to surface roughness. The last one depends on local natural features: mountains, valleys, rivers, lakes, forests, cultivated fields and buildings. Wind speed over soft surfaces (cultivated areas or lakes, etc.) is on average higher than the wind speed over rough surfaces (mountains, buildings, etc.). The effect of surface roughness on wind speed is a function of height above the earth's surface: the wind speed is lower near the surface.

Dispersion of pollutants is also affected by wind direction variability. If wind speed is relatively constant, the same area will be exposed to high levels of pollutant concentrations. On the other hand, if the wind direction will change constantly, it will disperse pollutants over a larger area, and concentrations in any of the exposed areas will be much lower. Wind direction and wind speed can be represented by a type of graph named wind rose. The spokes' length indicates wind direction frequency. This kind of graph can be used to predict dispersion from a point source or surface.

Fig. 1. Wind Rose

The air flow is not uniform near the surface of the earth, but presents a series of movements called vortices. The vortices are produced by two specific processes: thermal turbulence (turbulence caused by the atmospheric heat) and mechanical turbulence (resulting from the movement of air near various obstacles). Thermal turbulence is dominant in bright and sunny days with weak winds. Although mechanical turbulence is produced in various

The speed of the horizontal wind is affected by friction, which is proportional to surface roughness. The last one depends on local natural features: mountains, valleys, rivers, lakes, forests, cultivated fields and buildings. Wind speed over soft surfaces (cultivated areas or lakes, etc.) is on average higher than the wind speed over rough surfaces (mountains, buildings, etc.). The effect of surface roughness on wind speed is a function of height above

Dispersion of pollutants is also affected by wind direction variability. If wind speed is relatively constant, the same area will be exposed to high levels of pollutant concentrations. On the other hand, if the wind direction will change constantly, it will disperse pollutants over a larger area, and concentrations in any of the exposed areas will be much lower. Wind direction and wind speed can be represented by a type of graph named wind rose. The spokes' length indicates wind direction frequency. This kind of graph can be used to predict

The air flow is not uniform near the surface of the earth, but presents a series of movements called vortices. The vortices are produced by two specific processes: thermal turbulence (turbulence caused by the atmospheric heat) and mechanical turbulence (resulting from the movement of air near various obstacles). Thermal turbulence is dominant in bright and sunny days with weak winds. Although mechanical turbulence is produced in various

the earth's surface: the wind speed is lower near the surface.

dispersion from a point source or surface.

Fig. 1. Wind Rose

weather conditions, it is dominant in windy, in neutral atmospheric conditions. Turbulence has the effect of increasing the dispersion process, even when there is mechanical turbulence, on in certain areas, high concentrations of pollutants could be recorded, (Godish, 1997).

In large urban centres there are a multitude of fixed and mobile pollution sources. This urban community forms a heated and polluted island which is one of the biggest problems of the contemporary world. In an air quality analysis the ventilation rate of the entire region must be taken into account, because the dispersion can be influenced by both atmospheric mesoscale and macroscale movements. The atmospheric conditions as the anticyclones or thermal inversions slow the dispersion, resulting in high concentrations of pollutants. On the other hand, high weather variability such as moving cyclones, associated with strong winds, will enhance the dispersion of pollutants in urban areas. The pollution can move on, over very large areas, affecting air quality within hundreds of kilometers distances (Briggs, 1969).

Dispersion of car emissions is a complex process that acts at different scales of time and space and is strongly influenced by weather variability and air dynamics around vehicles and buildings. The amount of the exhaust gas varies over time, and it moves in time and space. Immediately after being discharged, the gases are dispersed due to turbulence that occurs around the vehicle, a phenomenon influenced by the non-stationary flow of air masses. A few seconds later, the dispersion is further accelerated by turbulence from other vehicles, moving within the street area.

In parallel with the mixture due to movement of vehicles, emissions are dispersed by air masses that have a dynamic movement because of the weather and topographical conditions (the buildings, trees, open areas, etc.) (Dobre et al, 2005).

Movement of vehicles creates uneven air movements (Plate, 1998) which mix the pollutants emitted from the exhaust pipe. After moving, the vehicle vortices arise. Basically, there isn't a database of local-scale air movements produced by motor vehicles in urban areas (Tate, 2004).

The influence of driving mode on vehicle emissions is of particular importance in such a study as higher emission rates are associated with certain traffic characteristics. (Tate, 2004)

The urban environment is characterized by different widths streets and buildings with different geometries. Dispersion of emissions from vehicles in narrow streets from urban areas has become a very intensely studied research theme (Chan et al, 2002, Xie et al, 2003) because it is known that very high levels of pollutants can be found within these streets, which means, most often, very high levels of pollution to which citizens are exposed.

The term "street canyon" refers to a relatively narrow street with buildings built by both sides, continuously (Nicholson, 1975). Dimensions of the street canyons are expressed usually by the characteristic ratio H / W, where H is the average height of a building located on both sides of the street, and W is the width of the street. A street canyon can be considered regular if the characteristic ratio is approximately equal to 1 and there are no major gaps in buildings on both sides of the canyon, which are like walls. If the wind blows obliquely to the direction of street-canyon, air flow reflection (wind in the roof) in the

A Qualitative Study of Air Pollutants from Road Traffic 27

**Traffic flow and primary pollutants: NO and PM2.5**

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

**CO concentrations (ppm)**

0:00:00

2:00:00

4:00:00

set lower, records higher concentration values.

6:00:00

8:00:00

10:00:00

12:00:00

**Time (h)**

Fig. 4. The device 2 was closer by the street and it recorded higher level of pollution

14:00:00

When there are many devices which measure the same pollutant, installed at different heights, they must record the data as it can be seen in the graph below: the device which is

16:00:00

18:00:00

20:00:00

22:00:00

**Traffic flow (vehicles/h)**

trend

0:00

2:00

4:00

points or the more distant points from the street.

6:00

8:00

pollutants, which demonstrates that their main source is the traffic.

10:00

12:00

**Time (h)**

Fig. 3. Traffic flow - NO and PM 2.5 concentrations (primarily pollutants) – almost the same

The graphs show a relatively large amounts of traffic flow and concentration of pollutants around two moments of the day: in the morning, when people go to work and in the afternoon, when they come back, and a good correlation between the above mentioned

Also, relatively high concentrations are recorded near the street, compared with the higher

**Two monitoring devices located at different height**

14:00

16:00

18:00

20:00

22:00

PM2.5 NO

**NO concentrations (ppb)** 

Traffic flow (vehicles/h)

**and PM2.5 (µg/m3)**

Device 1 Device 2

canyon wall which is exposed to the wind flow, induces a spiral flow along the canyon (Johnson and Hunter, 1999).

Atmospheric fluid dynamics inside the street canyon have been extensively studied by all types of experiments. Historically, most studies were based on idealization of the canyon with buildings placed evenly on both sides of the canyon, considering a wind perpendicular to the street (Berkowicz 1997, Chan et al, 2002). In reality, the geometry and weather conditions are statistically very unlikely.

When analyzing measurements of air pollution in street canyon it is important to have basic knowledge about the dynamics of air and wind effects on dispersion. Also, the moving vehicles generate turbulence which is preferable to be taken into account, if possible. A serious limitation of many studies of urban pollution is the lack of good data set to describe the synoptic weather conditions. Wherever possible, these critical parameters (wind speed, direction, and turbulent flows) must be measured in the area.

#### **5. A study of spatial and temporal variability of air pollution in urban area**

The purpose of this work is to understand how the examination of spatial and temporal variations of the pollutants' concentration in a street canyon could be done. It has to be linked with the traffic data, with emissions and weather conditions. To do this, the frequency of road traffic, the generation of emissions and the pollutants dispersion phenomenon must be taken into account.

A long term database is preferred, to also make an analysis of the pollutants trend. The database must be created using the raw measurements and eliminate negative values or strange data which could appear because of the device malfunctioning. If the measurement period is long, the devices must be calibrated from time to time (e.g. one time a month) and a few data could be lost. Time series plots must be done, pollutants together with traffic flow and meteorological data, to make complex analysis of the monitored data. The graphs must be analyzed, to find the correlations between different pollutants and the traffic flow. So, carbon monoxide, nitrogen oxide and particulate matter are emitted by vehicle engines during usage – primarily pollutants (Raducan and Stefan, 2009).

Fig. 2. Traffic flow - CO concentrations (primarily pollutant) – almost the same trend

canyon wall which is exposed to the wind flow, induces a spiral flow along the canyon

Atmospheric fluid dynamics inside the street canyon have been extensively studied by all types of experiments. Historically, most studies were based on idealization of the canyon with buildings placed evenly on both sides of the canyon, considering a wind perpendicular to the street (Berkowicz 1997, Chan et al, 2002). In reality, the geometry and weather

When analyzing measurements of air pollution in street canyon it is important to have basic knowledge about the dynamics of air and wind effects on dispersion. Also, the moving vehicles generate turbulence which is preferable to be taken into account, if possible. A serious limitation of many studies of urban pollution is the lack of good data set to describe the synoptic weather conditions. Wherever possible, these critical parameters (wind speed,

**5. A study of spatial and temporal variability of air pollution in urban area** 

The purpose of this work is to understand how the examination of spatial and temporal variations of the pollutants' concentration in a street canyon could be done. It has to be linked with the traffic data, with emissions and weather conditions. To do this, the frequency of road traffic, the generation of emissions and the pollutants dispersion

A long term database is preferred, to also make an analysis of the pollutants trend. The database must be created using the raw measurements and eliminate negative values or strange data which could appear because of the device malfunctioning. If the measurement period is long, the devices must be calibrated from time to time (e.g. one time a month) and a few data could be lost. Time series plots must be done, pollutants together with traffic flow and meteorological data, to make complex analysis of the monitored data. The graphs must be analyzed, to find the correlations between different pollutants and the traffic flow. So, carbon monoxide, nitrogen oxide and particulate matter are emitted by vehicle engines

**Traffic flow and CO concentrations**

(Johnson and Hunter, 1999).

conditions are statistically very unlikely.

phenomenon must be taken into account.

**Traffic flow** 

**(vehicles/h)**

1:00:00

3:00:00

5:00:00

7:00:00

9:00:00

11:00:00

13:00:00

**Time (h)**

Fig. 2. Traffic flow - CO concentrations (primarily pollutant) – almost the same trend

15:00:00

17:00:00

19:00:00

21:00:00

23:00:00

Traffic flow

0 0.2 0.4 0.6 0.8 1 1.2 1.4

**CO concentrations** 

CO concentrations (ppm)

**(ppm)**

direction, and turbulent flows) must be measured in the area.

during usage – primarily pollutants (Raducan and Stefan, 2009).

Fig. 3. Traffic flow - NO and PM 2.5 concentrations (primarily pollutants) – almost the same trend

The graphs show a relatively large amounts of traffic flow and concentration of pollutants around two moments of the day: in the morning, when people go to work and in the afternoon, when they come back, and a good correlation between the above mentioned pollutants, which demonstrates that their main source is the traffic.

Also, relatively high concentrations are recorded near the street, compared with the higher points or the more distant points from the street.

Fig. 4. The device 2 was closer by the street and it recorded higher level of pollution

When there are many devices which measure the same pollutant, installed at different heights, they must record the data as it can be seen in the graph below: the device which is set lower, records higher concentration values.

A Qualitative Study of Air Pollutants from Road Traffic 29

Mathematical modeling is one of the most powerful modern tools used in most studies of air pollution. Since the monitoring urban pollutants is very expensive, the modeling of the pollutants dispersion is an alternative that offers good results in the study of air quality.

A simple definition of the concept of "model" may be as follows: the model is a mathematical

The models are used mainly for: generalization and interpretation of measurement, the pollution prediction and the simulation of the pollution scenarios (answers to questions:

Within the operational plans, the models are now widely used, the main focus being

 The establishment of legislative measures to control emission of pollutants into the atmosphere, by determining the maximum emission of pollutants that can take place

 The assessment of the proposed emission control techniques (air cleaning technologies); The selection of positions for the future pollution sources, in order to minimize

 The control of pollution accidents by defining the measures and strategies to be taken immediately after the accidental pollution episodes (warning systems and real-time

The assumption of the responsibilities, for higher levels of pollution, by establishing the

Physical models are small-scale representations of "real world", e.g. representations and reproductions of the phenomena of nature in the laboratory. An example of such common representations is the wind tunnel or water where they can reproduce the dispersion experiments in the presence of miniature buildings, relief, vegetation - forests, etc. Another example is the fog chambers where chemical reactions can be studied under the same atmospheric pressure, temperature, humidity and solar radiation. There are several limitations in this small-scale reproduction of the phenomena from the nature: the conditions of turbulence in wind tunnels or water can not fully simulate the real conditions,

Physical models have shown very interesting results, revealing some mechanisms and

Empirical models generalize the relationships between different parameters, experimentally

representation of facts, factors and interpretation of data quantity or situations.

**6. Air quality modeling** 

what will happen if ...?)

environmental impact;

strategy to reduce emissions);

source-receptor relationship.

without exceeding air quality standards;

Models currently used can be classified as follows:

many chemical reactions occur in fog room at their walls, etc.

providing validation data, then used in developing mathematical models.

(Zanetti, 1993):

*Types of models* 

determined.

 Physical models; Empirical models; Mathematical models;

Fig. 5. The devices 6 and 7 ware situated in the same point near the street. Device 6 was located above

Meteorology has a key role; the results clearly show the importance of wind direction and speed: increasing wind speed cause a good dispersion of pollutants, so the registration of low concentration of pollutants. On the contrary, stable atmospheric conditions cause an accumulation of pollutants in the area of the street, considerably increasing concentrations.

Fig. 6. When the wind speed is lower, the CO concentrations are higher

When the wind direction is perpendicular to the street, a vortex is created in the street, carrying pollutants towards the wind. If the wind direction is parallel to the street, the atmospheric fluid transports the pollutants along the street increasing the rate of concentrations in this direction. When the direction of the wind is skew toward the street the atmospheric fluid is moving helicoidally, carrying the pollutants both to the other side of the road, and along the street.

#### **6. Air quality modeling**

28 Air Quality – Monitoring and Modeling

**Two monitoring devices located at different distances from the street** 

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8

located above

concentrations.

**Carbon monoxide concentrations (ppm)**

**Wind speed (m/s)**

0 0.5 1 1.5 2 2.5 3 3.5 4

road, and along the street.

**CO concentrations (ppm)**

0:00:00

2:00:00

4:00:00

6:00:00

8:00:00

10:00:00

12:00:00

**Time (h)**

Meteorology has a key role; the results clearly show the importance of wind direction and speed: increasing wind speed cause a good dispersion of pollutants, so the registration of low concentration of pollutants. On the contrary, stable atmospheric conditions cause an accumulation of pollutants in the area of the street, considerably increasing

**The influence of wind speed to the CO concentrations**

**Timp(h)**

When the wind direction is perpendicular to the street, a vortex is created in the street, carrying pollutants towards the wind. If the wind direction is parallel to the street, the atmospheric fluid transports the pollutants along the street increasing the rate of concentrations in this direction. When the direction of the wind is skew toward the street the atmospheric fluid is moving helicoidally, carrying the pollutants both to the other side of the

Fig. 6. When the wind speed is lower, the CO concentrations are higher

Fig. 5. The devices 6 and 7 ware situated in the same point near the street. Device 6 was

14:00:00

16:00:00

18:00:00

20:00:00

22:00:00

CO concentrations Wind speed

Device 6 Device 7

> Mathematical modeling is one of the most powerful modern tools used in most studies of air pollution. Since the monitoring urban pollutants is very expensive, the modeling of the pollutants dispersion is an alternative that offers good results in the study of air quality.

> A simple definition of the concept of "model" may be as follows: the model is a mathematical representation of facts, factors and interpretation of data quantity or situations.

> The models are used mainly for: generalization and interpretation of measurement, the pollution prediction and the simulation of the pollution scenarios (answers to questions: what will happen if ...?)

> Within the operational plans, the models are now widely used, the main focus being (Zanetti, 1993):


#### *Types of models*

Models currently used can be classified as follows:


Physical models are small-scale representations of "real world", e.g. representations and reproductions of the phenomena of nature in the laboratory. An example of such common representations is the wind tunnel or water where they can reproduce the dispersion experiments in the presence of miniature buildings, relief, vegetation - forests, etc. Another example is the fog chambers where chemical reactions can be studied under the same atmospheric pressure, temperature, humidity and solar radiation. There are several limitations in this small-scale reproduction of the phenomena from the nature: the conditions of turbulence in wind tunnels or water can not fully simulate the real conditions, many chemical reactions occur in fog room at their walls, etc.

Physical models have shown very interesting results, revealing some mechanisms and providing validation data, then used in developing mathematical models.

Empirical models generalize the relationships between different parameters, experimentally determined.

A Qualitative Study of Air Pollutants from Road Traffic 31

Eulerian approach, although capable to include complex physical phenomena in the diffusion equation (dry deposition, wet deposition, topographic effects and building, chemical reactions), has serious mathematical problems when the equations must be solved,

Lagrangian formalism, by the statistical description of particle displacement along current lines, is mathematically easier to approach, but the results are limited by the impossibility of defining a complete statistic for a system of particles. Furthermore, Lagrangian formalism can't include a number of important physical phenomena such as nonlinear chemical

Operational Street Pollution Model (OSPM) has been tested many times by comparison with data obtained in different measurement campaigns (Hertel and Berkowicz, 1989a, b, c, Berkowicz, 2000a, b). Test results have helped to improve the model parameterization.

Quality of results obtained with the model depends on the quality of input data. Thus, studies by Schaedler et al. (1996, 1999) and Roeckle and Judges (1995) examined the influence of quality input data on modeled concentrations. As with any model of urban pollution, the most important parameters of the model used as input data are meteorological and background data, traffic data and emission factors that are needed to

Pollution generated by road traffic in a street canyon is characterized by a strong temporal but also spatial variability, both horizontally and vertically, which not only depends on diurnal variation of traffic, but also on weather conditions. Very little things are known about the vertical distribution of traffic pollution. Some of them are obtained from experiments in wind tunnels (Pavageau et al., 1999) or from direct measurements (Vachon et al., 1999). The analysis of measurements of turbulence in the street is shown in Vachon et al. (2001) and Louka et al. (2001). The dependence of urban pollutant concentrations on traffic and thermal effects, and comparison of experimental results with those modeled by OSPM

OSPM is a semiempiric, parameterized model, where the concentrations of exhaust gas are calculated using a combination of a plume model for direct contribution and a box model for the recirculated part of the pollutants from the street. OSPM uses a very simplified parameterization of atmospheric fluid dynamics and dispersion of pollutants in street canyon. This parameterization was deducted from the analysis of experimental data and testing of the model. The results of these tests were used to improve the model's performance, especially in connection with various configurations of the street and different weather conditions. On the other hand, OSPM is based on many empirical assumptions that may not be applicable in all urban environments. The comparison data measured in the street with the OSPM model proves that a model with a simple parameterization like this can be applied to predict pollution from traffic. An important feature is the ability of the model to determine the concentrations of pollutants on a very weak wind, when it records

estimate emissions and topographical conditions (Pavageau et al., 1997).

using different parameterizations for this purpose.

**7. OSPM – A street canyon model** 

was realized in Berkowicz (2000a,b).

data of severe pollution.

reactions.

Mathematical models, the most complex models currently developed, can be divided into:


The models of type receiver could be good examples of statistical models. Receptor type models use measured data for different concentrations of pollutants, for a long time. They can identify the impact of the contribution from different sources using statistical analysis of the measured concentrations at a certain point (receiver). Receptor type models are used mainly for the interpretation of air quality measurements.

An example of a deterministic model could be a diffusion model in which the output data generated by the model (pollutant concentrations) is calculated based on mathematical computations performed on the input data (emissions, atmospheric parameters, etc.).

Deterministic models are most suitable for practical applications, because if they are correctly calibrated and used, they generate very accurate and consistent results regarding the source-receiver relationship. Establishing accurate source-receptor relationships is the target of any study which is made to improve the air quality or to keep the current levels of concentration for future development of industrial or urban areas. In other words, only deterministic models can generate clear assessment of the responsibility of each source of pollution from each receiver, this thing being absolutely necessary in the design and implementation of emission control strategies.

In terms of mathematical models, the modeling techniques used in the deterministic models must be applied to all aspects and phenomena related to air pollution. The main phenomena which are mathematically modeled:


Turbulent diffusion theory is focused on the behavior of pollutant particles which follow trajectories that allegedly are caused by the wind field. Due to the randomness of atmospheric dynamics it is impossible to determine the exact distribution of pollutants concentration in the atmosphere due to the pollution sources. Although the basic equations describing turbulent diffusion can be written without difficulty, there is no single mathematical model that can be used to calculate concentrations field in all possible states of the atmosphere.

There are currently two theories developed that address turbulent diffusion: the Eulerian approach and the Lagrangian approach.

The Eulerian method is based on solving the equation of mass conservation in a fixed region of space, while the Lagrangian method is based on tracking the behavior of pollutant particles, carried along the flow lines. Each approach has a number of shortcomings that make it virtually impossible to determine the exact diffusion problems. For example, the

Mathematical models, the most complex models currently developed, can be divided into: Statistical models based on semi-empirical statistical relationships between various data

 Deterministic models based on mathematical description of atmospheric processes, where the effects (e.g. air pollution) are generated by causes (e.g. pollutant emissions). The models of type receiver could be good examples of statistical models. Receptor type models use measured data for different concentrations of pollutants, for a long time. They can identify the impact of the contribution from different sources using statistical analysis of the measured concentrations at a certain point (receiver). Receptor type models are used

An example of a deterministic model could be a diffusion model in which the output data generated by the model (pollutant concentrations) is calculated based on mathematical computations performed on the input data (emissions, atmospheric parameters, etc.).

Deterministic models are most suitable for practical applications, because if they are correctly calibrated and used, they generate very accurate and consistent results regarding the source-receiver relationship. Establishing accurate source-receptor relationships is the target of any study which is made to improve the air quality or to keep the current levels of concentration for future development of industrial or urban areas. In other words, only deterministic models can generate clear assessment of the responsibility of each source of pollution from each receiver, this thing being absolutely necessary in the design and

In terms of mathematical models, the modeling techniques used in the deterministic models must be applied to all aspects and phenomena related to air pollution. The main phenomena

Turbulent diffusion theory is focused on the behavior of pollutant particles which follow trajectories that allegedly are caused by the wind field. Due to the randomness of atmospheric dynamics it is impossible to determine the exact distribution of pollutants concentration in the atmosphere due to the pollution sources. Although the basic equations describing turbulent diffusion can be written without difficulty, there is no single mathematical model that can be used to calculate concentrations field in all possible states of

There are currently two theories developed that address turbulent diffusion: the Eulerian

The Eulerian method is based on solving the equation of mass conservation in a fixed region of space, while the Lagrangian method is based on tracking the behavior of pollutant particles, carried along the flow lines. Each approach has a number of shortcomings that make it virtually impossible to determine the exact diffusion problems. For example, the

and available measurements.

mainly for the interpretation of air quality measurements.

implementation of emission control strategies.

which are mathematically modeled:

deposition to the soil surface.

turbulent diffusion in the atmosphere,

approach and the Lagrangian approach.

atmospheric photochemical and chemical reactions,

atmospheric transport,

the atmosphere.

Eulerian approach, although capable to include complex physical phenomena in the diffusion equation (dry deposition, wet deposition, topographic effects and building, chemical reactions), has serious mathematical problems when the equations must be solved, using different parameterizations for this purpose.

Lagrangian formalism, by the statistical description of particle displacement along current lines, is mathematically easier to approach, but the results are limited by the impossibility of defining a complete statistic for a system of particles. Furthermore, Lagrangian formalism can't include a number of important physical phenomena such as nonlinear chemical reactions.

#### **7. OSPM – A street canyon model**

Operational Street Pollution Model (OSPM) has been tested many times by comparison with data obtained in different measurement campaigns (Hertel and Berkowicz, 1989a, b, c, Berkowicz, 2000a, b). Test results have helped to improve the model parameterization.

Quality of results obtained with the model depends on the quality of input data. Thus, studies by Schaedler et al. (1996, 1999) and Roeckle and Judges (1995) examined the influence of quality input data on modeled concentrations. As with any model of urban pollution, the most important parameters of the model used as input data are meteorological and background data, traffic data and emission factors that are needed to estimate emissions and topographical conditions (Pavageau et al., 1997).

Pollution generated by road traffic in a street canyon is characterized by a strong temporal but also spatial variability, both horizontally and vertically, which not only depends on diurnal variation of traffic, but also on weather conditions. Very little things are known about the vertical distribution of traffic pollution. Some of them are obtained from experiments in wind tunnels (Pavageau et al., 1999) or from direct measurements (Vachon et al., 1999). The analysis of measurements of turbulence in the street is shown in Vachon et al. (2001) and Louka et al. (2001). The dependence of urban pollutant concentrations on traffic and thermal effects, and comparison of experimental results with those modeled by OSPM was realized in Berkowicz (2000a,b).

OSPM is a semiempiric, parameterized model, where the concentrations of exhaust gas are calculated using a combination of a plume model for direct contribution and a box model for the recirculated part of the pollutants from the street. OSPM uses a very simplified parameterization of atmospheric fluid dynamics and dispersion of pollutants in street canyon. This parameterization was deducted from the analysis of experimental data and testing of the model. The results of these tests were used to improve the model's performance, especially in connection with various configurations of the street and different weather conditions. On the other hand, OSPM is based on many empirical assumptions that may not be applicable in all urban environments. The comparison data measured in the street with the OSPM model proves that a model with a simple parameterization like this can be applied to predict pollution from traffic. An important feature is the ability of the model to determine the concentrations of pollutants on a very weak wind, when it records data of severe pollution.

A Qualitative Study of Air Pollutants from Road Traffic 33

**Statistical parameters used to analyze the quality of model output data** Bias FB

**Device 1 Device 2 Device 3 Device 4 Device 5 Device 6 Device 7**

**Monitoring devices**

Fig. 8. The statistical parameters: the mean error (bias), the normalized residue (FB)

Fig. 9. A good correlation between simulated and measured data (R2> 0,5)

Fig. 10. A weak correlation between simulated and measured data (R2< 0,5)

**-0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4**

**Statistical parameters** 

**values**

The results obtained with WinOSPM can be saved in a file that can be text, Excel or Access. The easiest to use is the .xls file, because it allows an easily analysis of the correlation between modeled and measured data. The modeled data is compared with those measured by calculating a set of statistical indicators used to validate the model. By this method, the quality of the output data can be determined, and so the model performance.

This data set includes the averages, the medians and the standard deviations for measured and simulated values, the mean error (bias), the normalized residue (FB), the normalized square error (NMSE), the correlation coefficient (R2) and the fraction of predicting the measured values up to a factor 2 (FA2) (Raducan, 2008b).

Fig. 7. The statistical parameters: the averages, the medians calculated for measured and modeled data

If the values of the averages and medians for both sets of data (modeled and measured) are close, it means that there are not very large extreme values which could be wrong.

The standard deviation shows how much variation or "dispersion" there is from the average. A low standard deviation indicates that the data points tend to be very close to the mean, whereas high standard deviation indicates that the data points are spread out over a large range of values.

If Bias <0 the model underestimates. Otherwise, it overestimates. Also, if FB>0, the model underestimates. Otherwise, it overestimates. So, figure 8 shows that the model underestimates.

When the normalized square error (NMSE) is low, it indicates that the model correctly describes the processes in time and space.

Analyzing the correlation coefficients between the simulated and measured data (Raducan, 2008a) there is good correlation if R2> 0,5 and a weak correlation if R2< 0,5.

The results obtained with WinOSPM can be saved in a file that can be text, Excel or Access. The easiest to use is the .xls file, because it allows an easily analysis of the correlation between modeled and measured data. The modeled data is compared with those measured by calculating a set of statistical indicators used to validate the model. By this method, the

This data set includes the averages, the medians and the standard deviations for measured and simulated values, the mean error (bias), the normalized residue (FB), the normalized square error (NMSE), the correlation coefficient (R2) and the fraction of predicting the

> **Statistical parameters used to analyze the quality of model output data**

Modeled average Measured average Modeled median Measured median

**Device 1 Device 2 Device 3 Device 4 Device 5 Device 6 Device 7 Monitoring devices**

Fig. 7. The statistical parameters: the averages, the medians calculated for measured and

close, it means that there are not very large extreme values which could be wrong.

If the values of the averages and medians for both sets of data (modeled and measured) are

The standard deviation shows how much variation or "dispersion" there is from the average. A low standard deviation indicates that the data points tend to be very close to the mean, whereas high standard deviation indicates that the data points are spread out over a large

If Bias <0 the model underestimates. Otherwise, it overestimates. Also, if FB>0, the model underestimates. Otherwise, it overestimates. So, figure 8 shows that the model underestimates. When the normalized square error (NMSE) is low, it indicates that the model correctly

Analyzing the correlation coefficients between the simulated and measured data (Raducan,

2008a) there is good correlation if R2> 0,5 and a weak correlation if R2< 0,5.

quality of the output data can be determined, and so the model performance.

measured values up to a factor 2 (FA2) (Raducan, 2008b).

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

modeled data

range of values.

describes the processes in time and space.

**Statistical parameters values**

Fig. 8. The statistical parameters: the mean error (bias), the normalized residue (FB)

Fig. 9. A good correlation between simulated and measured data (R2> 0,5)

Fig. 10. A weak correlation between simulated and measured data (R2< 0,5)

A Qualitative Study of Air Pollutants from Road Traffic 35

A complex analysis of urban air quality must contain an observational part that examines the concentrations of pollutants recorded in the street and a dispersion modeling part.

First, the monitored pollutants must be established. Also, the monitoring devices must be chosen and the location in the urban area. Then, from the multitude of pollutants dispersion models, the most appropriate must be chosen (could be more then one model in order to

Data analysis is done taking into account both the measured concentrations in the street, and emissions, traffic, weather conditions and topography, all being related to one another. The best way to compare data is to use some types of graphs to highlight the features that interest us.

The results should be stated clearly, visibly and after careful and thorough analysis. It is preferable to highlight both the results and shortcomings after a practical study. The latter could become research topics for those who made that study but also for other researchers.

[1] Berkowicz, R., 1997, Modelling Street Canyon Pollution: Model Requirements and Expectations, International Journal of Environment and Pollution 8 (3–6), 609–619. [2] Berkowicz, R., 2000a, OSPM – A Parameterized Street Pollution Model, Environment

[3] Berkowicz, R., 2000b, A simple model for urban background pollution. Environmental

[5] Chan, T.L., Dong, G., Leung, C.W., Cheung, C.S., Hung, W.T., 2002, Validation of a Two-

[6] Dobre A., Arnold S.J., Smalley R.J., Boddy J.W.D. Barlow J.F., Tomlin A.S., Belcher S.E.,

[7] Godish, T., 1997, Motor Vehicle Emissions Control. Air Quality. Boca Raton: CRC Press. [8] Hertel, O. and Berkowicz, R., 1989a, Modelling Pollution from Traffic in a Street Canyon. Evaluation of Data and Model Development, DMU Luft A-129, 77p. [9] Hertel, O. and Berkowicz, R., 1989b, Modelling NO2 Concentrations in a Street Canyon,

[10] Hertel, O. and Berkowicz, R., 1989c, Operational Street Pollution Model (OSPM). Evaluation of the Model on Data from St. Olavs Street in Oslo, DMU Luft A-135. [11] Johnson, G.T., Hunter, L.J., 1999, Some Insights into Typical Urban Canyon Air Flows.

[12] Louka, P., Vachon, G., Sini, J.F., Mestayer, P.G., Rosant, J.M., 2001, Thermal Effects on the

[13] Nicholson, S.E., 1975, A Pollution Model for Street-Level Air, Atmospheric

[14] Palmgren, F., Berkowicz, R., Ziv, A., Hertel, O., 1999, Actual Car Fleet Emissions

Air Flow in a Street Canyon—Nantes'99 Experimental Results and Model Simulations, Third International Conference on Urban Air Quality, Loutraki, Greece.

Estimated from Urban Air Quality Measurements and Street Pollution Models.

Dimensional Pollutant Dispersion Model in an Isolated Street Canyon.

2005, Flow Field Measurements in the Proximity of an Urban Intersection in

[4] Briggs G.A., 1969, Plume rise, ABC Critical Review Series, Washington DC

London, UK, Atmospheric Environment 39, 4647-4657.

compare the models output data).

Monitoring Assessment, 65, 323–331.

Monitoring and Assessment 65, 259–267.

Atmospheric Environment 36, 861–872.

Atmospheric Environment 33, 3991–3999.

Science of the Total Environment 235, 101–109.

DMU Luft A-131, 31p.

Environment 9, 19–31.

**10. References** 

The values of the fractions of predicting (FA2) must be between 0,5 and 2.

#### **8. The model's parameterization**

The correlation between the simulated and measured data could be improved if the errors are corrected. The input data are very important for the quality of output data. In the input data there are a few parameterization mathematical relationships which could be improved. The model must be run with the new inputs. The results must be compared as it can be seen in the graph below (Raducan, 2008a)

If the values of the correlation coefficients are higher, the parameterization has improved the model outputs and it is validated for that set of data. This parameterization must be tested for many others sets of data in the same location and if it works properly, than it can be used within that street canyon. Also, the parameterization must be tested from time to time, using real data.

Fig. 11. A comparison between two sets of model outputs, before and after a parameterization; the model was run in 7 points in the street canyon

In many studies with OSPM, the results are compared with those obtained with other models to improve physical concepts of the model parameterization. Yet, the numerical models do not describe well enough some important phenomena and they require various parameterization relationships. These relationships are very important in the case when the wind is weak and the turbulence produced by traffic has a great influence.

It is noted that OSPM has been successfully applied to analysis of data obtained from routine monitoring in order to predict trends in traffic emissions (Palmgren et al., 1999).

Because OSPM, as software, does not require significant resources of the computer system, several years' data were processed within a few minutes.

#### **9. Conclusions**

The aim of the study has been enriching and understanding how to make an analysis of the urban air pollution.

A complex analysis of urban air quality must contain an observational part that examines the concentrations of pollutants recorded in the street and a dispersion modeling part.

First, the monitored pollutants must be established. Also, the monitoring devices must be chosen and the location in the urban area. Then, from the multitude of pollutants dispersion models, the most appropriate must be chosen (could be more then one model in order to compare the models output data).

Data analysis is done taking into account both the measured concentrations in the street, and emissions, traffic, weather conditions and topography, all being related to one another. The best way to compare data is to use some types of graphs to highlight the features that interest us.

The results should be stated clearly, visibly and after careful and thorough analysis. It is preferable to highlight both the results and shortcomings after a practical study. The latter could become research topics for those who made that study but also for other researchers.

#### **10. References**

34 Air Quality – Monitoring and Modeling

The correlation between the simulated and measured data could be improved if the errors are corrected. The input data are very important for the quality of output data. In the input data there are a few parameterization mathematical relationships which could be improved. The model must be run with the new inputs. The results must be compared as it can be seen

If the values of the correlation coefficients are higher, the parameterization has improved the model outputs and it is validated for that set of data. This parameterization must be tested for many others sets of data in the same location and if it works properly, than it can be used within that street canyon. Also, the parameterization must be tested from time to

Correlation coefficient R2 before

Correlation coefficient R2 after

parameterization

parameterization

**Device 1 Device 2 Device 3 Device 4 Device 5 Device 6 Device 7 Monitoring devices**

In many studies with OSPM, the results are compared with those obtained with other models to improve physical concepts of the model parameterization. Yet, the numerical models do not describe well enough some important phenomena and they require various parameterization relationships. These relationships are very important in the case when the

It is noted that OSPM has been successfully applied to analysis of data obtained from routine monitoring in order to predict trends in traffic emissions (Palmgren et al., 1999).

Because OSPM, as software, does not require significant resources of the computer system,

The aim of the study has been enriching and understanding how to make an analysis of the

The values of the fractions of predicting (FA2) must be between 0,5 and 2.

**Correlation coefficients before and after parameterization**

Fig. 11. A comparison between two sets of model outputs, before and after a

wind is weak and the turbulence produced by traffic has a great influence.

several years' data were processed within a few minutes.

parameterization; the model was run in 7 points in the street canyon

**8. The model's parameterization** 

in the graph below (Raducan, 2008a)

time, using real data.

**0.4 0.45 0.5 0.55 0.6 0.65 0.7**

**9. Conclusions** 

urban air pollution.


**3** 

*1USA 2,3India* 

**Estimation of Ambient Air Quality in Delhi** 

*3Civil Engineering Department, Malaviya National Institute of Technology, Jaipur,* 

For decades, it has been well known that air pollution causes adverse health effects on humans (Thurston et al., 1989, Shandilya and Khare, 2012). Since the late 1980s, epidemiological studies were published relating the mortality and morbidity to the ambient levels of fine particles without showing the expected results (Borja-Aburto et al., 1997). Most emphasis was then given to the particles with an aerodynamic diameter smaller than 10µm (PM10) particularly to the class of the fine particulates. A special study carried out by United Nations Sources (UNECE, 1979) shows that the frequency of finding a particle of diameter range 2-4m is 27.5%, which is higher in an overall range of 1-12m, so the danger of the fine particulates of diameter ranges 2-4m is higher than the danger by any particle which may lie in the range of 4-12m. Therefore, there is a rising awareness about further fine fraction i.e., particles with an aerodynamic diameter smaller than 2.5µm (PM2.5). Aerosols, which originate directly from the sources, are termed as "Primary aerosols". There may be the conversion of gaseous matter into the particulate due to many Physico-Chemical reactions. Such types of particulates are known as "Secondary Aerosols" (Perkins, 1974,

Delhi, the Capital and third largest city of India, ranks third in population among other Indian cities, estimated among 8.5 million (Crowther et al., 1990). Rapid urbanization and the unprecedented industrial and economic development during the last three decades have increased the vehicular population of Delhi by several folds. Consequently, harmful emission such as Pb, PAH, SO2, NOx, particulate matter and carbon monoxide have also registered a sharp increase in these intervening years. As per the WHO report, Delhi is the fourth most populated city in the world. A 1997 study by the Center for Science and Environment (CSE) revealed that at least one person died prematurely every hour in Delhi in 1995 strictly because of Suspended Particulate Matter (SPM). A 1998 study by All India Institute of Medical Sciences (AIIMS), New Delhi, showed how emergency visits and deaths due to respiratory and heart problems are the highest when particulate levels peak during the winter (Down to Earth, 2000). It is estimated that in Delhi, the incidence of chronic

**1. Introduction** 

Shandilya et al., 2009).

\* Corresponding Author

Kaushik K. Shandilya1,\*, Mukesh Khare2 and A. B. Gupta3 *1Civil Engineering Department, The University of Toledo, Toledo, OH, 2Civil Engineering Department, Indian Institute of Technology, Delhi,* 


## **Estimation of Ambient Air Quality in Delhi**

Kaushik K. Shandilya1,\*, Mukesh Khare2 and A. B. Gupta3

*1Civil Engineering Department, The University of Toledo, Toledo, OH, 2Civil Engineering Department, Indian Institute of Technology, Delhi, 3Civil Engineering Department, Malaviya National Institute of Technology, Jaipur, 1USA 2,3India* 

#### **1. Introduction**

36 Air Quality – Monitoring and Modeling

[15] Pavageau, M., Rafailidis, S. and Schatzmann, M., 1997, A Comprehensive Experimental

[16] Pavageau, M., Schatzmann, M., 1999, Wind Tunnel Measurements of Concentration Fluctuations in an Urban Street Canyon. Atmospheric Environment 33, 3961–3971. [17] Plate, E.J. and Kastner-Klein, P.,1998, Wind Tunnel Modelling of Traffic Induced

[19] Raducan, G., 2008b, Pollutants Dispersion Modeling with OSPM in a Street Canyon from Bucharest, Romania Reports in Physics, Romania, no.4, 1099-1114. [20] Raducan, G., Stefan, S., 2009, Characterization of Traffic-Generated Pollutants in

[21] Roeckle, R., C.J. Judges, 1995, Ermittlung des Strömungs- und Konzentrationsfeldesim

[22] Schaedler, G., W. Baechlin, A. Lohmeyer, T. Van Wees, 1996, Vergleich und Bewertung

Umweltforschung Baden-Württemberg (FZKA-PEF 138), (English Version),

[23] Schaedler, G., W. Baechlin, A. Lohmeyer, 1999, Immissionsprognosen mit

[24] Schneider, T., Koning, H. W. and Brassed, L. J., 1997, Air Pollution Reference

[25] Tate J.E., 2004, A Study of Vehicular Emissions and Ambient Air Quality at the Local-

[26] Vachon, G., Rosant, J.-M., Mestayer, P.G. and Sini, J.F.,1999, Measurements of Dynamic

[27] Vachon, G., Louka, P., Rosant, J.M., Mestayer, P., Sini, J.F., 2001, Measurements of

[28] Xie, S., Zhang, Y., Qi, L. and Tang, X., 2003, Spatial Distribution of Traffic-Related Pollutant Concentrations in Street Canyons. Atmos. Env., 37, 3213-3224. [29] Zanetti, P., 1993, Air Pollution Modelling: Theories, Computational Methods, and Available Software. Computational Mechanics Publications, Southampton, UK.

Regulatory Purposes, 11–14 October, Rouen, France, Paper 124.

Förderkennzeichen PEF 2 96 004. FZKA-BWPLUS 14, (English Version),

Nahfeld typischer Gebäudekonfigurationen.- Modellrechnungen-. Abschlußbericht

derzeit verfügbarer Strömungs- und Ausbreitungsmodelle. In: Berichte

meikroscaligen Modellen. Vergleich von berechneten und gemessenen Größen.

Measurement Methods and Systems, Studies in Environmental Science 2. Elsevier,

and Thermal Field in a Street Canyon, URBCAP Nantes'99', Proceedings of the 6th Int. Conf. on Harmonisation within Atmospheric Dispersion Modelling for

Traffic-Induced Turbulence within a Street Canyon during the Nantes '99 Experiment. Third International Conference on Urban Air Quality, Loutraki,

[18] Raducan, G., 2008a, Study of the air pollution regional variability, PhD Thesis.

PEF 92/007/02, Forschungszentrum Karlsruhe, English Version,

Environ. Pollut. 8, 738–746.

http://bwplus.fzk.de

http://bwplus.fzk.de

http://bwplus.fzk.de

New York, 1998.

Scale, PhD Thesis

Greece.

for Urban Safety and Security in March 1998.

Bucharest, Atmosfera, Mexic, Vol. 22, No.1, 99-110.

Databank for the Verification of Urban Car Emission Dispersion Models, Int. J.

Pollution in Cities, Proceedings of the 2nd German-Japanese meeting "Klimaanalyse für die Stadtplanung", 25-27.9.97, to appear in the series Report of Research Center

> For decades, it has been well known that air pollution causes adverse health effects on humans (Thurston et al., 1989, Shandilya and Khare, 2012). Since the late 1980s, epidemiological studies were published relating the mortality and morbidity to the ambient levels of fine particles without showing the expected results (Borja-Aburto et al., 1997). Most emphasis was then given to the particles with an aerodynamic diameter smaller than 10µm (PM10) particularly to the class of the fine particulates. A special study carried out by United Nations Sources (UNECE, 1979) shows that the frequency of finding a particle of diameter range 2-4m is 27.5%, which is higher in an overall range of 1-12m, so the danger of the fine particulates of diameter ranges 2-4m is higher than the danger by any particle which may lie in the range of 4-12m. Therefore, there is a rising awareness about further fine fraction i.e., particles with an aerodynamic diameter smaller than 2.5µm (PM2.5). Aerosols, which originate directly from the sources, are termed as "Primary aerosols". There may be the conversion of gaseous matter into the particulate due to many Physico-Chemical reactions. Such types of particulates are known as "Secondary Aerosols" (Perkins, 1974, Shandilya et al., 2009).

> Delhi, the Capital and third largest city of India, ranks third in population among other Indian cities, estimated among 8.5 million (Crowther et al., 1990). Rapid urbanization and the unprecedented industrial and economic development during the last three decades have increased the vehicular population of Delhi by several folds. Consequently, harmful emission such as Pb, PAH, SO2, NOx, particulate matter and carbon monoxide have also registered a sharp increase in these intervening years. As per the WHO report, Delhi is the fourth most populated city in the world. A 1997 study by the Center for Science and Environment (CSE) revealed that at least one person died prematurely every hour in Delhi in 1995 strictly because of Suspended Particulate Matter (SPM). A 1998 study by All India Institute of Medical Sciences (AIIMS), New Delhi, showed how emergency visits and deaths due to respiratory and heart problems are the highest when particulate levels peak during the winter (Down to Earth, 2000). It is estimated that in Delhi, the incidence of chronic

<sup>\*</sup> Corresponding Author

Estimation of Ambient Air Quality in Delhi 39

location was mainly a residential area. Except for Shakarpur, where the Indraprastha power

SPM was collected on a filter paper weighed before and after exposure. It was dried in an oven at 80oC for 45 minutes before each weighing. The size classification of PM was achieved through a cyclone installed in a Hi-Vol sampler, which separates the respirable PM10 and non-respirable fraction in the case of the Hi-Vol sampler. The particles sized <10μm were collected on the filter, and >10μm were collected in a separate sampling cup, also known as a dust collector. Flow rates were monitored before and after every sampling period with manometer, which was factory calibrated with an accuracy of 1.3% full scale. The flow rate checks were made before and after each sampling event to ensure that filter

These gases were monitored with the help of APM411 assembly attached with APM460DX (Envirotech Instruments Pvt. Ltd.) using the impingers. The modified Jacob-Hochheiser method (or arsenate method) have determined nitrogen dioxide (IS-5182, 1975). For the 24 hour field sampling, it is preferable to use this method. The West–Gaeke spectrophotometric method is the standard method for SO2 monitoring of 0.0005-5 ppm SO2 in the ambient air.

Based on the monitoring data of various pollutants at different sites on different days, an index called "Air Quality Index" was calculated using the following equation, thereby assessing the present air quality status of Delhi. As per the EPA guidelines, pollutants such

( ) *Hi Lo*

In the present study, the pollutants used for calculating the AQI were NO2, SO2, and RSP. AQI was calculated for individual pollutants simultaneously at a place, and the maximum value corresponding to one particular pollutant was reported as the AQI of that area. The same procedure was adopted at all sites on each day, with AQI finally reported along with

For all sites, we have calculated the mean, standard deviation, and maximum, and we also calculated the geometric means. The geometric mean is the basis of preparing the air quality standards. The mean, standard deviation, maximum value, and geometric mean values are

*Lo Lo*

(1)

as O3, CO, SO2, NO2, PM10, and PM2.5 are used for calculating the AQI for an area.

*Hi Lo l l lp Cp BP l BP BP*

station is situated nearby, the sources of the local emissions were considered.

holders were not leaking and to determine that fitter clogging had not occurred.

**2.2 Air quality index determination** 

lp=the index for pollutant P

the responsible pollutants.

listed as below:

**3. Results and discussion** 

Cp=the rounded concentration of pollutant P BHHi=the breakpoint greater than or equal to Cp LHi=the AQI value corresponding to BPHi lLo=the AQI value corresponding to BPLo BPLo=the breakpoint less than or equal to Cp

bronchitis is 6-14 times more than the average elsewhere in India because of the thermal power plants and the vehicular pollution (Rao & Rao, 1989).

The objective of this study was to monitor NOx, SO2, PM10, PM2.5, and Non Respirable Suspended Particulate (NRSP) levels in South Delhi at different land use patterns; to compare it with the levels found in the crowded residential site of East Delhi; and to compare the observed SPM,NOx, and SO2 levels with the prescribed CPCB limits (except PM2.5 compared with USEPA primary standards), which serves as the guidelines for the environmental planning of the region. Another objective was to find out the spatial and diurnal variation of NRSP, the Organic Matter level in NRSP, and the 12 hour variation of NOx and SO2. The final objective was the determination of Air Pollution Index at different places.

#### **2. Materials and methods**

A background to the Study Area is outlined in detail in Shandilya et al. (2007). In order to assess ambient pollution level, particulate samples were taken with the help of respirable dust sampler and fine particulate sampler. For the high volume Measurement of RSP, air was drawn through a 20.3X25.4cm Whatman GF/A glass fiber filter at a flow rate which was kept around 1-1.1m3/min. Detailed Procedures of Equipment (APM 460DX) are specified in operating manuals, available from equipment manufacturers (Envirotech Instruments Private Limited). The gases were determined with APM 411. The entire weighing procedure was done under controlled conditions (room temperature) in a balance room at Envirotech Instruments Pvt. Ltd., New Delhi using a Dhona-100DS microbalance. For PM2.5 monitoring, ambient air enters APM550 system through an omni directional inlet designed to provide a clean aerodynamic cut-point for particles >10μm.

#### **2.1 Sampling sites and details**

The air quality monitoring in Delhi for present investigation was carried out from October 19, 2000 to November 29, 2000, during which PM10, PM2.5, and non-respirable suspended particulate concentrations were measured at four monitoring sites. In order to obtain a clear picture of a source-based pollution, it was required that the monitoring must be carried out at different places of different characteristics in New Delhi. It was important to consider residential, industrial, and commercial areas for carrying out such source differentiation. Based on the above considerations, four sites were chosen for setting up the sampling sites, namely Envirotech Instruments Pvt. Ltd, Okhla, II Floor Roof (Industrial Area); Indian Institute of Technology, Delhi, V Block IV Floor Roof (Residential Area); Vasant Vihar Bus Depot II Floor Roof (Commercial Area); Shakarpur a Residence, III Floor Roof (Residential Area). More details and figures of sampling locations can be obtained from Shandilya et al., (2007, 2012).

Out of the four sites chosen, two were in residential areas, one each in industrial and commercial areas. The criterion that was taken into account was the presence of local sources. The Shakarpur site was of particular importance, since the results obtained were used to compare the SPM level in South Delhi. Simultaneously, when we analyzed the respirable suspended particulates in respect of heavy metals, it was interesting to know the concentration near prominent sources like the Indraprastha power station. The site was located just on the bank of the Yamuna, when on the other bank was the Indraprastha power station. But the site

bronchitis is 6-14 times more than the average elsewhere in India because of the thermal

The objective of this study was to monitor NOx, SO2, PM10, PM2.5, and Non Respirable Suspended Particulate (NRSP) levels in South Delhi at different land use patterns; to compare it with the levels found in the crowded residential site of East Delhi; and to compare the observed SPM,NOx, and SO2 levels with the prescribed CPCB limits (except PM2.5 compared with USEPA primary standards), which serves as the guidelines for the environmental planning of the region. Another objective was to find out the spatial and diurnal variation of NRSP, the Organic Matter level in NRSP, and the 12 hour variation of NOx and SO2. The final

A background to the Study Area is outlined in detail in Shandilya et al. (2007). In order to assess ambient pollution level, particulate samples were taken with the help of respirable dust sampler and fine particulate sampler. For the high volume Measurement of RSP, air was drawn through a 20.3X25.4cm Whatman GF/A glass fiber filter at a flow rate which was kept around 1-1.1m3/min. Detailed Procedures of Equipment (APM 460DX) are specified in operating manuals, available from equipment manufacturers (Envirotech Instruments Private Limited). The gases were determined with APM 411. The entire weighing procedure was done under controlled conditions (room temperature) in a balance room at Envirotech Instruments Pvt. Ltd., New Delhi using a Dhona-100DS microbalance. For PM2.5 monitoring, ambient air enters APM550 system through an omni directional inlet

The air quality monitoring in Delhi for present investigation was carried out from October 19, 2000 to November 29, 2000, during which PM10, PM2.5, and non-respirable suspended particulate concentrations were measured at four monitoring sites. In order to obtain a clear picture of a source-based pollution, it was required that the monitoring must be carried out at different places of different characteristics in New Delhi. It was important to consider residential, industrial, and commercial areas for carrying out such source differentiation. Based on the above considerations, four sites were chosen for setting up the sampling sites, namely Envirotech Instruments Pvt. Ltd, Okhla, II Floor Roof (Industrial Area); Indian Institute of Technology, Delhi, V Block IV Floor Roof (Residential Area); Vasant Vihar Bus Depot II Floor Roof (Commercial Area); Shakarpur a Residence, III Floor Roof (Residential Area). More details and figures of sampling locations can be obtained from

Out of the four sites chosen, two were in residential areas, one each in industrial and commercial areas. The criterion that was taken into account was the presence of local sources. The Shakarpur site was of particular importance, since the results obtained were used to compare the SPM level in South Delhi. Simultaneously, when we analyzed the respirable suspended particulates in respect of heavy metals, it was interesting to know the concentration near prominent sources like the Indraprastha power station. The site was located just on the bank of the Yamuna, when on the other bank was the Indraprastha power station. But the site

power plants and the vehicular pollution (Rao & Rao, 1989).

**2. Materials and methods** 

**2.1 Sampling sites and details** 

Shandilya et al., (2007, 2012).

objective was the determination of Air Pollution Index at different places.

designed to provide a clean aerodynamic cut-point for particles >10μm.

location was mainly a residential area. Except for Shakarpur, where the Indraprastha power station is situated nearby, the sources of the local emissions were considered.

SPM was collected on a filter paper weighed before and after exposure. It was dried in an oven at 80oC for 45 minutes before each weighing. The size classification of PM was achieved through a cyclone installed in a Hi-Vol sampler, which separates the respirable PM10 and non-respirable fraction in the case of the Hi-Vol sampler. The particles sized <10μm were collected on the filter, and >10μm were collected in a separate sampling cup, also known as a dust collector. Flow rates were monitored before and after every sampling period with manometer, which was factory calibrated with an accuracy of 1.3% full scale. The flow rate checks were made before and after each sampling event to ensure that filter holders were not leaking and to determine that fitter clogging had not occurred.

These gases were monitored with the help of APM411 assembly attached with APM460DX (Envirotech Instruments Pvt. Ltd.) using the impingers. The modified Jacob-Hochheiser method (or arsenate method) have determined nitrogen dioxide (IS-5182, 1975). For the 24 hour field sampling, it is preferable to use this method. The West–Gaeke spectrophotometric method is the standard method for SO2 monitoring of 0.0005-5 ppm SO2 in the ambient air.

#### **2.2 Air quality index determination**

Based on the monitoring data of various pollutants at different sites on different days, an index called "Air Quality Index" was calculated using the following equation, thereby assessing the present air quality status of Delhi. As per the EPA guidelines, pollutants such as O3, CO, SO2, NO2, PM10, and PM2.5 are used for calculating the AQI for an area.

$$llp = \frac{l\_{Hi} - l\_{Lo}}{BP\_{Hi} - BP\_{Lo}} (Cp - BP\_{Lo}) + l\_{Lo} \tag{1}$$

lp=the index for pollutant P Cp=the rounded concentration of pollutant P BHHi=the breakpoint greater than or equal to Cp LHi=the AQI value corresponding to BPHi lLo=the AQI value corresponding to BPLo BPLo=the breakpoint less than or equal to Cp

In the present study, the pollutants used for calculating the AQI were NO2, SO2, and RSP. AQI was calculated for individual pollutants simultaneously at a place, and the maximum value corresponding to one particular pollutant was reported as the AQI of that area. The same procedure was adopted at all sites on each day, with AQI finally reported along with the responsible pollutants.

#### **3. Results and discussion**

For all sites, we have calculated the mean, standard deviation, and maximum, and we also calculated the geometric means. The geometric mean is the basis of preparing the air quality standards. The mean, standard deviation, maximum value, and geometric mean values are listed as below:

Estimation of Ambient Air Quality in Delhi 41

agriculture; smoke from residential wood combustion or prescribed burning; directly emitted exhaust from motor vehicle engines; and secondary sulfates, nitrates, and organics formed from gaseous ammonia, sulfur dioxide, nitrogen oxides, and reactive organic gas emissions (Sharma et al., 2004, Shandilya et al., 2012). The emissions inventories for these "nontraditional" types of sources were found to be much less accurate than those obtained for "traditional" ducted emissions from the industrial sources. The coarse fraction was mainly composed of the re-suspended dust. Farming operations (e.g., agriculture tillage, harvesting, and travel on unpaved roads) account for large portions of the primary PM10 in many emissions inventories. PM10 non-attainment areas are suspected of having RWC as a major contributor. The prescribed burning of agricultural fields and forest slash is suspected of being a major contributor to PM10 in other areas. PM10 is considered more a local and urban scale problem, whereas PM2.5 has a much longer residence time and is considered a regional or trans-boundary problem. The spatial and temporal distribution of fine particle

Natural background levels, excluding all anthropogenic sources in the US and elsewhere, were 4-11-µgm-3 and 1-5-µgm-3 for PM10 and PM2.5, respectively. PM10 mean concentrations were typically from 10-80-µgm-3 around the world. 10-20-µgm-3 is found at remote sites with no local sources (i.e., even in clean air). In urban areas, 60–220-μg/m3 is typical, and in a heavily polluted area, levels may approach 2000-μg/m3. More than 50-µgm-3 was found at heavily polluted sites with high traffic density and/or local sources in the surroundings, where heavily polluted urban areas may reach 100-µgm-3 on average. In the heavily polluted cities of southern Europe, Latin America, and Asia, average PM10 concentrations have reached 100 µgm-3, and more. PM2.5 mean concentrations were between 7-80-µgm-3. PM10 trends were decreasing in the last decade, for emission and ambient levels, in the developed and developing world. The ratio of PM2.5 to PM10 at the urban sites usually averaged 0.55-0.6.

Many studies indicated that TSP and PM10 concentration in the ambient air would be affected by various meteorological factors such as wind speed, wind direction, solar radiation, relative humidity, rainfall, and source conditions. Apart from the emission factors of the vehicles, SPM concentrations on the street level would be mainly affected by the mechanical turbulence (created by moving vehicles and wind) and by thermal turbulence

24-hour average concentrations of the FSP, RSP, and TSP were calculated (See Methodology for a sample calculation) from samples collected at four stations at Southern Delhi and East Delhi and from samples taken from Satna. Our study is also not limited by the fact that, during winter, mixing depth in the early morning is lower than in the summer, thereby concentrating primary crustal aerosol emissions (e.g., from road dust). Since our study did not take place in the summer or winter, this condition is not applicable and the purpose of

The sources prevalent at VV are Diesel Buses, CNG Buses, Traffic, and Mechanical Disturbances, while at IIT the sources are Traffic, Vegetative Burning, construction work, and residential sources like cooking. At Okhla, the sources are Vegetative Burning, Diesel Generators, Mechanical Disturbances, Traffic, and Industrial Activities. Concentrations of

produced by the hot vehicle exhaust gas (Shandilya and Kumar, 2009) .

levels may vary substantially.

our study is not forfeited.

**3.2 Comparison of SPM level in Delhi** 


Table 1. Statistical Analysis of Results.

#### **3.1 General discussion**

In general, for PM2.5 the secondary formation is more relevant than the primary emissions, but Wilson and Suh (1997) stated that some natural particles also occur in size range <2.5µm. A typical study (Pekkanen et al., 1997) reports that a considerable fraction of the road dust consisted of particles<2.5μm in diameter. The primary particles are emitted by a source as particles and are dispersed in the atmosphere without any major chemical transformation (Shandilya and Kumar, 2010a, 2010b, 2011).

PM10 mainly consisted of three distinct components: Primary vehicular emissions, secondary pollutants, and windblown dust. Specifically, major contributors to excessive PM10 levels were combinations of area-wide emitters such as fugitive dust from roads, construction, and

**GEO MEAN** 268.6 126.71 405.35 687.71 0.47 0.39 **MEAN** 271.11 129.7 425.2 696.3 0.49 0.40 **STD DEV.** 39.1 28.6 139.1 117.4 0.11 0.11 **MAXIMUM** 320.3 157.3 599.2 827.6 0.56 0.53

**GEO MEAN** 202.1 139.02 289.0 492.1 0.69 0.41 **MEAN** 204.3 144.1 290.7 495.0 0.71 0.41 **STD DEV.** 35.4 38.6 32.9 59.3 0.20 0.04 **MAXIMUM** 274.7 170.6 326.7 601.3 0.91 0.46

**GEO MEAN** 256.9 152.58 251.3 514.6 0.59 0.50 **MEAN** 260.7 159.5 260.3 521.1 0.65 0.50 **STD DEV.** 49.7 45.0 77.9 97.3 0.26 0.09 **MAXIMUM** 322.6 198.1 394.5 717.2 0.94 0.64

**GEO MEAN** 226.8 174.05 192.3 427.6 0.77 0.53 **MEAN** 233.0 175.0 199.4 432.4 0.78 0.54 **STD DEV.** 64.5 21.2 67.1 73.7 0.18 0.11 **MAXIMUM** 321.2 198.4 299.7 497.4 0.95 0.66

**GEO MEAN** 238.1 144.8 283.3 532.0 0.61 0.45 **MEAN** 243.1 150.0 302.5 545.6 0.65 0.46 **STD DEV.** 51.1 37.0 117.6 130.3 0.21 0.10 **MAXIMUM** 322.6 198.4 599.2 827.6 0.95 0.66

**GEO MEAN** 245.65 371.35 628.99 0.39 **MEAN** 257.62 384.62 642.24 0.40 **STD DEV.** 86.28 101.04 130.81 0.10 **MAXIMUM** 586.00 659.00 984.00 0.63

In general, for PM2.5 the secondary formation is more relevant than the primary emissions, but Wilson and Suh (1997) stated that some natural particles also occur in size range <2.5µm. A typical study (Pekkanen et al., 1997) reports that a considerable fraction of the road dust consisted of particles<2.5μm in diameter. The primary particles are emitted by a source as particles and are dispersed in the atmosphere without any major chemical

PM10 mainly consisted of three distinct components: Primary vehicular emissions, secondary pollutants, and windblown dust. Specifically, major contributors to excessive PM10 levels were combinations of area-wide emitters such as fugitive dust from roads, construction, and

**OKHLA** 

**IIT** 

**VV** 

**SKP** 

**Delhi** 

**ITO Delhi** 

Table 1. Statistical Analysis of Results.

transformation (Shandilya and Kumar, 2010a, 2010b, 2011).

**3.1 General discussion** 

**RSP FSP NRSP TSP PM2.5/PM10 RSP/TSP** 

agriculture; smoke from residential wood combustion or prescribed burning; directly emitted exhaust from motor vehicle engines; and secondary sulfates, nitrates, and organics formed from gaseous ammonia, sulfur dioxide, nitrogen oxides, and reactive organic gas emissions (Sharma et al., 2004, Shandilya et al., 2012). The emissions inventories for these "nontraditional" types of sources were found to be much less accurate than those obtained for "traditional" ducted emissions from the industrial sources. The coarse fraction was mainly composed of the re-suspended dust. Farming operations (e.g., agriculture tillage, harvesting, and travel on unpaved roads) account for large portions of the primary PM10 in many emissions inventories. PM10 non-attainment areas are suspected of having RWC as a major contributor. The prescribed burning of agricultural fields and forest slash is suspected of being a major contributor to PM10 in other areas. PM10 is considered more a local and urban scale problem, whereas PM2.5 has a much longer residence time and is considered a regional or trans-boundary problem. The spatial and temporal distribution of fine particle levels may vary substantially.

Natural background levels, excluding all anthropogenic sources in the US and elsewhere, were 4-11-µgm-3 and 1-5-µgm-3 for PM10 and PM2.5, respectively. PM10 mean concentrations were typically from 10-80-µgm-3 around the world. 10-20-µgm-3 is found at remote sites with no local sources (i.e., even in clean air). In urban areas, 60–220-μg/m3 is typical, and in a heavily polluted area, levels may approach 2000-μg/m3. More than 50-µgm-3 was found at heavily polluted sites with high traffic density and/or local sources in the surroundings, where heavily polluted urban areas may reach 100-µgm-3 on average. In the heavily polluted cities of southern Europe, Latin America, and Asia, average PM10 concentrations have reached 100 µgm-3, and more. PM2.5 mean concentrations were between 7-80-µgm-3. PM10 trends were decreasing in the last decade, for emission and ambient levels, in the developed and developing world. The ratio of PM2.5 to PM10 at the urban sites usually averaged 0.55-0.6.

Many studies indicated that TSP and PM10 concentration in the ambient air would be affected by various meteorological factors such as wind speed, wind direction, solar radiation, relative humidity, rainfall, and source conditions. Apart from the emission factors of the vehicles, SPM concentrations on the street level would be mainly affected by the mechanical turbulence (created by moving vehicles and wind) and by thermal turbulence produced by the hot vehicle exhaust gas (Shandilya and Kumar, 2009) .

24-hour average concentrations of the FSP, RSP, and TSP were calculated (See Methodology for a sample calculation) from samples collected at four stations at Southern Delhi and East Delhi and from samples taken from Satna. Our study is also not limited by the fact that, during winter, mixing depth in the early morning is lower than in the summer, thereby concentrating primary crustal aerosol emissions (e.g., from road dust). Since our study did not take place in the summer or winter, this condition is not applicable and the purpose of our study is not forfeited.

#### **3.2 Comparison of SPM level in Delhi**

The sources prevalent at VV are Diesel Buses, CNG Buses, Traffic, and Mechanical Disturbances, while at IIT the sources are Traffic, Vegetative Burning, construction work, and residential sources like cooking. At Okhla, the sources are Vegetative Burning, Diesel Generators, Mechanical Disturbances, Traffic, and Industrial Activities. Concentrations of

Estimation of Ambient Air Quality in Delhi 43

contribution to PM10 is 47%, which increases to 59% at VV, further increases to 69% at IIT and achieves its maximum (77%) at SKP. PM2.5/PM10 ratio standard deviation is highest (0.26) at the VV site, suggesting the heterogeneous contribution of secondary particle formation sources. It has been observed that a part of PM10 is PM2.5 but not as much as in the case of IIT in South Delhi. The standard deviation (PM2.5/RSP) low value at Okhla (0.11) shows that the contribution of PM2.5/RSP is more consistent compared to other sites, and PM2.5 sources contributing to RSP remains the same. The ratios of PM2.5/PM10 are generally between 20 and 30%, with abnormal values from industrial particulate (both high-50% in Cynon- and low-17% in Port Talbot). The highest contribution of PM2.5/PM10 strongly shows that industrial sources are predominant in the SKP area. Comparison with higher ratios generally found in large urban areas (Birmingham-QUARG report) indicates that the finer traffic particulate stick together or deposit on surfaces less rapidly than the coarse fraction of PM10, so they tend to build up in larger urban areas or still conditions. It has been observed that a part of PM10 is PM2.5, but not as much as in the case of SKP. It shows that the condensation products of the chemical reactions primarily contribute PM10 fraction that originated in this area. Data (PM2.5/PM10) shows that PM10 fraction that originated in industrial areas contains the

condensation products of the chemical reactions and some primary particulate.

November 26, 2000, which may have been due to the wash out caused by the rain.

With a minimum of one daily sample at a site, 24-hour TSP concentration at the urbanindustrial Delhi site varies from 553.4-827.6-µgm-3. 24-hr TSP concentration at the urbancommercial Delhi site varies from 453.2-717.2-µgm-3 and 24-hr TSP concentration at the urban-residential Delhi sites varies from 345.2-601.3-µgm-3. The TSP concentration value was always greater than the 24-hr NAAQS of TSP i.e., 200-μg/m3 in the residential area. The TSP concentration value was always greater than the NAAQS of TSP i.e., 500-μg/m3 in the industrial and commercial area while it crossed the limit (NAAQS, 500-μg/m3) only once on

The maximum PM10 concentration in Delhi, South Delhi and VV 322.6-μg/m3 was on November 17, 2000 during our study; the lowest PM10 of 174.5-μg/m3 was observed at Shakarpur. This can be due to some change of wind pattern since there is a reduction in PM2.5 concentration (171.8-μg/m3), which eradicates the possibility of contribution from secondary particles and is further confirmed by an increase in NRSP fraction. Maximum PM10 concentration at Okhla, 320.3-μg/m3, was on October 29, 2000 next to Deepawali. This high concentration can be attributed to crackers; this is well confirmed by the reduction of PM2.5 concentration (85.1-μg/m3), which is primarily caused by vehicles and is further confirmed by decline in the contribution of PM2.5 to PM10 suggested by the value of PM2.5/PM10 (0.27, second lowest in all). Maximum PM10 concentration at IIT, 274.7-μg/m3, occurred on November 3, 2000, the day when a celebration was held, so there was a great rush of vehicles in IIT. The increase in PM10 was due to the resuspension of dust and vehicle exhaust. It was further confirmed by the increase in NRSP fraction on that day and an increase in PM2.5. The PM2.5 value on the second day remained on the same level (168.7-μgm-3), showing that the residence time PM2.5 was much higher, which was not the case with PM10 and NRSP. Maximum PM10 concentration at SKP, 321.2-μg/m3, took place on

**3.2.5 Peak day composition** 

**3.2.6 Variation of TSP** 

November 17, 2000 at the VV commercial area.

fine particles are close to main roads. Road transport and diesel exhaust are major sources of fine particles in urban areas (Shandilya and Kumar, 2009, Kumar et al., 2011). PM10 and PM2.5 concentrations are elevated close to main roads compared to background sites. More monitoring details can be found in Shandilya et al (2006, 2011).

#### **3.2.1 Variation of PM2.5**

With a minimum of one daily sample at a site, 24-hour PM2.5 concentration at the urbanindustrial Delhi site varies from 85.1-157.3µgm-3. 24-hour PM2.5 concentration at the urbancommercial Delhi site varies from 78.6-198.1µgm-3, and 24-hour PM2.5 concentration at the urban-residential Delhi sites varies from 84.6-198.4µgm-3. The value at Delhi is well above 24-hour PM2.5 USNAAQS (65µgm-3). More details of this monitoring can be found in Shandilya et al., (2011). Variation of PM2.5 value is highest on SKP, which suggests that the contribution from different sources varies to a great extent. The Standard Deviation values suggest that PM2.5 source emission rate at VV varies greater than the emission rate at any place in South Delhi. The large variation is most likely due to the differences in air mass, emission characteristics, and meteorology.

#### **3.2.2 Peak day composition**

The maximum PM2.5 concentration in Delhi, 198.4-μg/m3, was at SKP on November 26, 2000, which may be due to a wash out caused by the rain. The lowest PM2.5 value of 78.6 μg/m3 was observed at VV bus depot (November, 24, 2000). When there was a strike and there was less traffic on the road, there was a significant decrease in PM2.5 fraction. It shows an important relationship between PM2.5 fractions with the traffic sources. Maximum PM2.5 concentration in South Delhi, 198.1-μg/m3, was at VV on November 22, 2000. Maximum PM2.5 concentration at Okhla, 157.3-μg/m3, was on October 21, 2000. Maximum PM2.5 concentration at IIT, 170.6-μg/m3, was on November 6, 2000.

#### **3.2.3 Variation of PM10**

With a minimum of one daily sample at a site, 24-hour PM10 concentration at the urbanindustrial Delhi site varies from 205.0-320.3-µgm-3. 24-hour PM10 concentration at the urbancommercial Delhi site varies from 210.0-322.6-µgm-3, and 24-hour PM10 concentration at the urban-residential Delhi sites varies from 174.5-321.2-µgm-3. The PM10 concentration value was always greater than the 24-hour NAAQS (150-μg/m3 in industrial and commercial areas and 100-μg/m3 in residential areas).

#### **3.2.4 PM2.5/PM10 ratio**

The most common ratio of PM2.5/PM10 is 0.60, but the prevalence of the local sources, with either a dominating fine or coarse fraction, may shift this ratio. PM2.5/PM10 concentration at the urban-industrial Delhi site varies from 0.27-0.56 with an average±standard deviation of 0.47±0.11. PM2.5/RSP concentration at the urban-commercial Delhi site varies from 0.25-0.94 with an average±standard deviation of 0.59±0.26. PM2.5/RSP concentration at the urbanresidential South Delhi sites varies from 0.43-0.91 with an average±standard deviation of 0.69±0.20. PM2.5/RSP concentration at the urban-residential East Delhi sites varies from 0.62- 0.95 with an average±standard deviation of 0.77±0.18. This shows that at Okhla, PM2.5

fine particles are close to main roads. Road transport and diesel exhaust are major sources of fine particles in urban areas (Shandilya and Kumar, 2009, Kumar et al., 2011). PM10 and PM2.5 concentrations are elevated close to main roads compared to background sites. More

With a minimum of one daily sample at a site, 24-hour PM2.5 concentration at the urbanindustrial Delhi site varies from 85.1-157.3µgm-3. 24-hour PM2.5 concentration at the urbancommercial Delhi site varies from 78.6-198.1µgm-3, and 24-hour PM2.5 concentration at the urban-residential Delhi sites varies from 84.6-198.4µgm-3. The value at Delhi is well above 24-hour PM2.5 USNAAQS (65µgm-3). More details of this monitoring can be found in Shandilya et al., (2011). Variation of PM2.5 value is highest on SKP, which suggests that the contribution from different sources varies to a great extent. The Standard Deviation values suggest that PM2.5 source emission rate at VV varies greater than the emission rate at any place in South Delhi. The large variation is most likely due to the differences in air mass,

The maximum PM2.5 concentration in Delhi, 198.4-μg/m3, was at SKP on November 26, 2000, which may be due to a wash out caused by the rain. The lowest PM2.5 value of 78.6 μg/m3 was observed at VV bus depot (November, 24, 2000). When there was a strike and there was less traffic on the road, there was a significant decrease in PM2.5 fraction. It shows an important relationship between PM2.5 fractions with the traffic sources. Maximum PM2.5 concentration in South Delhi, 198.1-μg/m3, was at VV on November 22, 2000. Maximum PM2.5 concentration at Okhla, 157.3-μg/m3, was on October 21, 2000. Maximum PM2.5

With a minimum of one daily sample at a site, 24-hour PM10 concentration at the urbanindustrial Delhi site varies from 205.0-320.3-µgm-3. 24-hour PM10 concentration at the urbancommercial Delhi site varies from 210.0-322.6-µgm-3, and 24-hour PM10 concentration at the urban-residential Delhi sites varies from 174.5-321.2-µgm-3. The PM10 concentration value was always greater than the 24-hour NAAQS (150-μg/m3 in industrial and commercial

The most common ratio of PM2.5/PM10 is 0.60, but the prevalence of the local sources, with either a dominating fine or coarse fraction, may shift this ratio. PM2.5/PM10 concentration at the urban-industrial Delhi site varies from 0.27-0.56 with an average±standard deviation of 0.47±0.11. PM2.5/RSP concentration at the urban-commercial Delhi site varies from 0.25-0.94 with an average±standard deviation of 0.59±0.26. PM2.5/RSP concentration at the urbanresidential South Delhi sites varies from 0.43-0.91 with an average±standard deviation of 0.69±0.20. PM2.5/RSP concentration at the urban-residential East Delhi sites varies from 0.62- 0.95 with an average±standard deviation of 0.77±0.18. This shows that at Okhla, PM2.5

monitoring details can be found in Shandilya et al (2006, 2011).

concentration at IIT, 170.6-μg/m3, was on November 6, 2000.

**3.2.1 Variation of PM2.5** 

emission characteristics, and meteorology.

areas and 100-μg/m3 in residential areas).

**3.2.2 Peak day composition** 

**3.2.3 Variation of PM10** 

**3.2.4 PM2.5/PM10 ratio** 

contribution to PM10 is 47%, which increases to 59% at VV, further increases to 69% at IIT and achieves its maximum (77%) at SKP. PM2.5/PM10 ratio standard deviation is highest (0.26) at the VV site, suggesting the heterogeneous contribution of secondary particle formation sources. It has been observed that a part of PM10 is PM2.5 but not as much as in the case of IIT in South Delhi. The standard deviation (PM2.5/RSP) low value at Okhla (0.11) shows that the contribution of PM2.5/RSP is more consistent compared to other sites, and PM2.5 sources contributing to RSP remains the same. The ratios of PM2.5/PM10 are generally between 20 and 30%, with abnormal values from industrial particulate (both high-50% in Cynon- and low-17% in Port Talbot). The highest contribution of PM2.5/PM10 strongly shows that industrial sources are predominant in the SKP area. Comparison with higher ratios generally found in large urban areas (Birmingham-QUARG report) indicates that the finer traffic particulate stick together or deposit on surfaces less rapidly than the coarse fraction of PM10, so they tend to build up in larger urban areas or still conditions. It has been observed that a part of PM10 is PM2.5, but not as much as in the case of SKP. It shows that the condensation products of the chemical reactions primarily contribute PM10 fraction that originated in this area. Data (PM2.5/PM10) shows that PM10 fraction that originated in industrial areas contains the condensation products of the chemical reactions and some primary particulate.

#### **3.2.5 Peak day composition**

The maximum PM10 concentration in Delhi, South Delhi and VV 322.6-μg/m3 was on November 17, 2000 during our study; the lowest PM10 of 174.5-μg/m3 was observed at Shakarpur. This can be due to some change of wind pattern since there is a reduction in PM2.5 concentration (171.8-μg/m3), which eradicates the possibility of contribution from secondary particles and is further confirmed by an increase in NRSP fraction. Maximum PM10 concentration at Okhla, 320.3-μg/m3, was on October 29, 2000 next to Deepawali. This high concentration can be attributed to crackers; this is well confirmed by the reduction of PM2.5 concentration (85.1-μg/m3), which is primarily caused by vehicles and is further confirmed by decline in the contribution of PM2.5 to PM10 suggested by the value of PM2.5/PM10 (0.27, second lowest in all). Maximum PM10 concentration at IIT, 274.7-μg/m3, occurred on November 3, 2000, the day when a celebration was held, so there was a great rush of vehicles in IIT. The increase in PM10 was due to the resuspension of dust and vehicle exhaust. It was further confirmed by the increase in NRSP fraction on that day and an increase in PM2.5. The PM2.5 value on the second day remained on the same level (168.7-μgm-3), showing that the residence time PM2.5 was much higher, which was not the case with PM10 and NRSP. Maximum PM10 concentration at SKP, 321.2-μg/m3, took place on November 26, 2000, which may have been due to the wash out caused by the rain.

#### **3.2.6 Variation of TSP**

With a minimum of one daily sample at a site, 24-hour TSP concentration at the urbanindustrial Delhi site varies from 553.4-827.6-µgm-3. 24-hr TSP concentration at the urbancommercial Delhi site varies from 453.2-717.2-µgm-3 and 24-hr TSP concentration at the urban-residential Delhi sites varies from 345.2-601.3-µgm-3. The TSP concentration value was always greater than the 24-hr NAAQS of TSP i.e., 200-μg/m3 in the residential area. The TSP concentration value was always greater than the NAAQS of TSP i.e., 500-μg/m3 in the industrial and commercial area while it crossed the limit (NAAQS, 500-μg/m3) only once on November 17, 2000 at the VV commercial area.

Estimation of Ambient Air Quality in Delhi 45

Our results could also be compared with the results of Jaipur, a study carried out by Das et al. (1998). They found at the industrial site that TSP levels were 384.61-μg/m3 and, in another study by Gajraj et al. (submitted) in the Vishwakarma industrial area, mean PM10 concentration during May 1999 was 242-μg/m3; during June 1999, the mean concentration was PM10 (208.5-μg/m3). Our TSP levels at Okhla are certainly higher than TSP levels in

The terminal settling velocity is an important factor for the spatial variation of SPM (Willeke & Baron, 1993). For very fine particles, such as a particle of 0.1-µm dae setting velocity, is 8.65x10-5-cms-1 for a particle of 1-µm 3.48x10-3 and for a particle of 10-µm in size 3.06x10-1 cms-1. Primary particles (Aitken particles <0.1-µm) and large particles (e.g.>10-µm) are expected to have a bigger spatial variability than the particles in the accumulation mode (0.1–1-µm). Coagulation (for very fine particles <0.01-µm) and gravitational settling (for particles>1-µm) are the underlying mechanisms, which cause spatial heterogeneity. The distribution of PM2.5 might be uniform in a situation where the secondary formation is important; in cities, however, with large emissions from heavy duties (diesel exhaust), PM2.5 can also exhibit significant spatial variation. Secondary aerosols that contribute to the accumulation mode (0.1-2-µm) show quite homogeneous spatial distribution. In general, small-scale spatial variations of PM2.5 were described to be smaller than the spatial variations of PM10 (Tuch et al., 1997). The spatial variation was small for PM2.5 but larger for PM10. Our results are in agreement with Monn (2001). With respect to the small-scale spatial variation in the urban areas, the largest variations occurred in the ultrafine (0.1-µm) and coarse mode (PM10-2.5) (Burton et al., 1996). Ultrafine particle number counts have large spatial variations, and they are not well correlated to mass data, so spatial variation is much

The spatial variation was small for PM2.5 but larger for PM10. Spatial correlations for PM2.5 are found to be lower than 0.9 and around 0.6 for PM10. The contribution of coarse particle fraction (PM10-2.5) on PM10 can be important. Conclusively, we can say that SPM is complex with respect to its sources in Delhi. With the exception of the geological material, PM2.5 source contribution estimates are not similar to those for PM10. A deep insight in the source contribution could be obtained if we consider emission factors of SPM from different

In order to understand the sources, behavior, and mechanism of particle formation in the atmosphere, it is important to measure the precursors like SOx and NOx; this primarily contributes to the formation of the secondary particulate, which again constitutes a big fraction of total particulate matter in the atmosphere. This type of precursor characterization for the continental, marine background, and urban influenced aerosols has been reported from various sites throughout the world (Garg et al., 2001). 24-hour average concentrations of SO2 and NO2 (See Methodology) from the samples collected at four stations at Southern Delhi and East Delhi are given in the table below; they were calculated by taking the

Jaipur, as obtained by Das et al. (1998).

**3.2.9 Explanation of spatial variation** 

more reflected in PM10 than in PM2.5.

**3.3 Oxides of sulfur and nitrogen** 

average of the day and night value.

sources.

#### **3.2.7 RSP/TSP ratio**

RSP/TSP at the urban-industrial Delhi site varies from 0.27-0.56 with an average±standard deviation of 0.39±0.11. RSP/TSP concentration at the urban-commercial Delhi site varies from 0.42-0.64µgm-3 with an average±standard deviation of 0.50±0.09. RSP/TSP concentration at the urban-residential South Delhi sites varies from 0.37-0.46 with an average±standard deviation of 0.41±0.04. RSP/TSP concentration at the urban-residential East Delhi sites varies from 0.40-0.66 with an average±standard deviation of 0.53±0.11. This shows that at Okhla, RSP contribution to TSP is 39%, increasing to 50% at VV, decreasing to 41% at IIT, and again achieving its highest of 53% at SKP. The standard deviation (RSP/TSP) low value at IIT (0.04) shows that the contribution of RSP to TSP is consistent, while it varies equally at SKP and at Okhla. Therefore, the RSP emitting sources in the SKP area were contributing equally throughout the study, while these sources' emission pattern was changed at the other places.

#### **3.2.8 Peak day composition**

The maximum TSP concentration in Delhi, South Delhi and at Okhla 827.6-μg/m3 occurred on October 23, 2000. The maximum TSP concentration at IIT, 601.3-μg/m3, occurred on November 3, 2000. It is further confirmed by an increase in NRSP fraction on that day. The maximum TSP concentration at VV, 717.2-μg/m3, occurred on November 17, 2000, which may be due to some change of wind pattern since there is a reduction in PM2.5 concentration (171.8-μg/m3) while there is an increase in NRSP fraction. The maximum TSP concentration at SKP, 489.4-μg/m3, was on November 26, 2000, which may be due to the wash out caused by the rain. In short, TSP concentrations at IIT and at SKP sites are lower than the other values observed in Delhi, but PM2.5 concentrations are higher than those in the Okhla industrial location.

Thus, industrial emissions have considerable impact on the ambient non-respirable particulate matter concentrations, while residential emissions have considerable impact on PM2.5 concentration. Similarly, PM10 concentrations at the Okhla site are higher than the values found at any other site. This suggests that the industrial emissions contribute more to PM2.5-10 fraction.

Our observations at Okhla in South Delhi showed that PM2.5, RSP, and TSP concentrations were 126.71±28.6, 268.6±39.1, and 687.71±117.4-µgm-3, respectively; IIT showed that PM2.5, RSP, and TSP concentrations were 139.02±38.6, 202.1±35.4, and 492.1±59.3-µgm-3, respectively; and VV showed that PM2.5, RSP, and TSP concentrations were 152.58±45.0, 251.3±49.7, and 514.6±97.3-µgm-3, respectively. As stated before, SKP in East Delhi showed that PM2.5, RSP and TSP concentrations were 174.05.0±21.1, 226.8±64.5, and 427.6±73.6-µgm-3, respectively.

Our results could be compared with the results of Kanpur, a study carried out by Sharma et al. (2003). They found at the background site, IIT Kanpur that PM10 levels were 62±11.22 μg/m3 and at Agricultural University, the residential area, PM10 levels were 71.46-335.83 µg/m3. At the commercial area Naveen Market, PM10 levels were 114.22-634.68-μg/m3, and at the industrial area, Lajpat Nagar, PM10 levels were 221.77-650.73-µg/m3. On comparison, it can be concluded that PM10 levels in all areas (industrial, commercial, residential) found the same range at both places (Delhi and Kanpur).

RSP/TSP at the urban-industrial Delhi site varies from 0.27-0.56 with an average±standard deviation of 0.39±0.11. RSP/TSP concentration at the urban-commercial Delhi site varies from 0.42-0.64µgm-3 with an average±standard deviation of 0.50±0.09. RSP/TSP concentration at the urban-residential South Delhi sites varies from 0.37-0.46 with an average±standard deviation of 0.41±0.04. RSP/TSP concentration at the urban-residential East Delhi sites varies from 0.40-0.66 with an average±standard deviation of 0.53±0.11. This shows that at Okhla, RSP contribution to TSP is 39%, increasing to 50% at VV, decreasing to 41% at IIT, and again achieving its highest of 53% at SKP. The standard deviation (RSP/TSP) low value at IIT (0.04) shows that the contribution of RSP to TSP is consistent, while it varies equally at SKP and at Okhla. Therefore, the RSP emitting sources in the SKP area were contributing equally throughout the study, while these sources' emission pattern was

The maximum TSP concentration in Delhi, South Delhi and at Okhla 827.6-μg/m3 occurred on October 23, 2000. The maximum TSP concentration at IIT, 601.3-μg/m3, occurred on November 3, 2000. It is further confirmed by an increase in NRSP fraction on that day. The maximum TSP concentration at VV, 717.2-μg/m3, occurred on November 17, 2000, which may be due to some change of wind pattern since there is a reduction in PM2.5 concentration (171.8-μg/m3) while there is an increase in NRSP fraction. The maximum TSP concentration at SKP, 489.4-μg/m3, was on November 26, 2000, which may be due to the wash out caused by the rain. In short, TSP concentrations at IIT and at SKP sites are lower than the other values observed in Delhi, but PM2.5 concentrations are higher than those in the Okhla

Thus, industrial emissions have considerable impact on the ambient non-respirable particulate matter concentrations, while residential emissions have considerable impact on PM2.5 concentration. Similarly, PM10 concentrations at the Okhla site are higher than the values found at any other site. This suggests that the industrial emissions contribute more to

Our observations at Okhla in South Delhi showed that PM2.5, RSP, and TSP concentrations were 126.71±28.6, 268.6±39.1, and 687.71±117.4-µgm-3, respectively; IIT showed that PM2.5, RSP, and TSP concentrations were 139.02±38.6, 202.1±35.4, and 492.1±59.3-µgm-3, respectively; and VV showed that PM2.5, RSP, and TSP concentrations were 152.58±45.0, 251.3±49.7, and 514.6±97.3-µgm-3, respectively. As stated before, SKP in East Delhi showed that PM2.5, RSP and

Our results could be compared with the results of Kanpur, a study carried out by Sharma et al. (2003). They found at the background site, IIT Kanpur that PM10 levels were 62±11.22 μg/m3 and at Agricultural University, the residential area, PM10 levels were 71.46-335.83 µg/m3. At the commercial area Naveen Market, PM10 levels were 114.22-634.68-μg/m3, and at the industrial area, Lajpat Nagar, PM10 levels were 221.77-650.73-µg/m3. On comparison, it can be concluded that PM10 levels in all areas (industrial, commercial, residential) found

TSP concentrations were 174.05.0±21.1, 226.8±64.5, and 427.6±73.6-µgm-3, respectively.

the same range at both places (Delhi and Kanpur).

**3.2.7 RSP/TSP ratio** 

changed at the other places.

**3.2.8 Peak day composition** 

industrial location.

PM2.5-10 fraction.

Our results could also be compared with the results of Jaipur, a study carried out by Das et al. (1998). They found at the industrial site that TSP levels were 384.61-μg/m3 and, in another study by Gajraj et al. (submitted) in the Vishwakarma industrial area, mean PM10 concentration during May 1999 was 242-μg/m3; during June 1999, the mean concentration was PM10 (208.5-μg/m3). Our TSP levels at Okhla are certainly higher than TSP levels in Jaipur, as obtained by Das et al. (1998).

#### **3.2.9 Explanation of spatial variation**

The terminal settling velocity is an important factor for the spatial variation of SPM (Willeke & Baron, 1993). For very fine particles, such as a particle of 0.1-µm dae setting velocity, is 8.65x10-5-cms-1 for a particle of 1-µm 3.48x10-3 and for a particle of 10-µm in size 3.06x10-1 cms-1. Primary particles (Aitken particles <0.1-µm) and large particles (e.g.>10-µm) are expected to have a bigger spatial variability than the particles in the accumulation mode (0.1–1-µm). Coagulation (for very fine particles <0.01-µm) and gravitational settling (for particles>1-µm) are the underlying mechanisms, which cause spatial heterogeneity. The distribution of PM2.5 might be uniform in a situation where the secondary formation is important; in cities, however, with large emissions from heavy duties (diesel exhaust), PM2.5 can also exhibit significant spatial variation. Secondary aerosols that contribute to the accumulation mode (0.1-2-µm) show quite homogeneous spatial distribution. In general, small-scale spatial variations of PM2.5 were described to be smaller than the spatial variations of PM10 (Tuch et al., 1997). The spatial variation was small for PM2.5 but larger for PM10. Our results are in agreement with Monn (2001). With respect to the small-scale spatial variation in the urban areas, the largest variations occurred in the ultrafine (0.1-µm) and coarse mode (PM10-2.5) (Burton et al., 1996). Ultrafine particle number counts have large spatial variations, and they are not well correlated to mass data, so spatial variation is much more reflected in PM10 than in PM2.5.

The spatial variation was small for PM2.5 but larger for PM10. Spatial correlations for PM2.5 are found to be lower than 0.9 and around 0.6 for PM10. The contribution of coarse particle fraction (PM10-2.5) on PM10 can be important. Conclusively, we can say that SPM is complex with respect to its sources in Delhi. With the exception of the geological material, PM2.5 source contribution estimates are not similar to those for PM10. A deep insight in the source contribution could be obtained if we consider emission factors of SPM from different sources.

#### **3.3 Oxides of sulfur and nitrogen**

In order to understand the sources, behavior, and mechanism of particle formation in the atmosphere, it is important to measure the precursors like SOx and NOx; this primarily contributes to the formation of the secondary particulate, which again constitutes a big fraction of total particulate matter in the atmosphere. This type of precursor characterization for the continental, marine background, and urban influenced aerosols has been reported from various sites throughout the world (Garg et al., 2001). 24-hour average concentrations of SO2 and NO2 (See Methodology) from the samples collected at four stations at Southern Delhi and East Delhi are given in the table below; they were calculated by taking the average of the day and night value.

Estimation of Ambient Air Quality in Delhi 47

The natural sources of nitrogen oxides are the geothermal activity and bacterial action. Natural sources include the soil microbes, vegetation, biomass, slum, burning, and lightning. Natural sources include the biological emissions (e.g. vegetation and soil microbes) and abiotic processes (e.g., geogenic, lightning, and biomass burning). Globally, lightning strikes occur about 100-times/second (Borucki and Chameides, 1984). Lighting strikes are characterized by a highly ionized and high temperature channel, in which molecular nitrogen is dissociated and reacts with oxygen to form NO (Guenther et al, 2000). The primary sources of NOx are the motor vehicles, electric utilities, and other industrial, commercial, and residential sources that burn fuels. Other sources include the chemical and nitric acid manufacture, the detonation of nitrate-containing explosives, and electric sources. Man made sources of nitrogen oxide include the automobile, railways (coal fired and diesel locos), airways (landing and takeoff operations), industrial processes like chemicals (nitric acid plants, processes where nitrogenous compounds are used), metal processing, fertilizer (nitrogenous fertilizers), and other miscellaneous activity like domestic combustion, forest fires and managed fires, industrial boilers, furnaces, and commercial combustion. Although most NO is produced by the natural bacterial action, it must be emphasized that the rate of emission from man-related sources has, by estimate, increased by one-fourth between 1966 and 1968. Certainly, man's greatest contribution to the family of nitrogen oxides arises from the combustion of fuels. The contribution to the total NOx can be categorized as Utilities 27%; Motor Vehicles 49%; Other Sources 5%; and Industrial/ Commercial/Residential 19%. Nearly 2/3 of NOx emissions originate at the Earth's surface from fossil fuel and biomass combustion. Vehicle exhausts are known to increase the inputs of NO2 to the atmosphere (see, e.g., Hargreaves et al., 2000). The burning of natural gas (which produces negligible

SOx, CO, and HC) does produce a very significant quantity of oxides of nitrogen.

oxidation) all make substantial contributions to the total emissions.

55 in 1995. The largest NOx emission chunk in Delhi is from the oil sources.

Man caused nitrogen emissions as equivalent NO2 are estimated as 53x106 TPA over the world, 95% of which are emitted in the Northern Hemisphere. Half of these are from the combustion of coal. Natural sources of NOx are estimated as between 15 and 17-times pollutant emissions. (Sawyer et al, 1998). The major source of NOx is the combustion when the fixation of atmospheric N occurs at the high flame temperature. The oxides are emitted mainly as NO, which is normally rapidly oxidized to NO2 by atmospheric O2 and O3. Motor vehicle exhaust contributes a sizable fraction of the total emissions of NOx. As well as the stationary combustion sources, the manufacture of HNO3 and nitrate fertilizer is the source of NO & NO2. On a global basis, natural emissions of oxides of nitrogen (such as releases from soil and the ocean, lightning strikes, oxidation of ammonia, and stratospheric

Garg et al (2001) has reported the sectoral shares of NOx emissions for India in 1995. According to them, a major part of NOx emissions is contributed from 28% by the nonenergy sources, 28% by electric power generation, 2% by road transport, 19% by Biomass burning, 4% by other transports, and 19% by industries. Delhi tops the table of top 10 NOx emitting districts in the terms of total emissions by emitting total NOx 66-Gg in 1990 and 82- Gg in 1995; in the terms of emissions per unit area, NOx/area (tons km-2) was 44 in 1990 and

**3.3.1 Sources of sulfur oxides and nitrogen oxides** 

For all sites, we have calculated the mean, standard deviation, and minimum and maximum values. We have also calculated the geometric means to show a homogeneous distribution of the data. Geometric mean concentrations are better representations of the data population as the elements are log-normally distributed. The geometric mean, standard deviation, maximum and minimum values, and geometric mean values are listed as below:


Table 2. Statistical Analysis of the Result Obtained

For all sites, we have calculated the mean, standard deviation, and minimum and maximum values. We have also calculated the geometric means to show a homogeneous distribution of the data. Geometric mean concentrations are better representations of the data population as the elements are log-normally distributed. The geometric mean, standard deviation,

> **Night Value**

**Geo.Mean** 5.86 25.47 15.76 32.48 107.85 70.68 0.22 0.18 0.24 **Mean** 6.13 25.55 15.84 32.58 110.57 71.58 0.23 0.19 0.24 **Std Dev** 1.89 2.32 1.76 2.73 24.20 11.68 0.06 0.06 0.08 **Maximum** 8.42 30.12 18.45 36.89 136.62 84.87 0.35 0.28 0.42 **Minimum** 3.50 23.58 13.54 28.20 64.55 50.72 0.18 0.11 0.19

**Geo.Mean** 4.30 14.45 9.40 24.56 57.29 40.95 0.23 0.18 0.25 **Mean** 4.37 14.49 9.43 24.72 57.83 41.27 0.23 0.18 0.25 **Std Dev** 0.83 1.21 0.81 3.06 8.67 5.67 0.02 0.03 0.03 **Maximum** 5.78 16.32 10.48 29.85 70.37 50.11 0.26 0.22 0.29 **Minimum** 3.21 12.96 8.45 21.00 47.13 35.12 0.20 0.14 0.23

**Geo.Mean** 5.48 26.67 16.12 37.86 97.48 68.36 0.24 0.14 0.27 **Mean** 5.51 26.97 16.24 38.21 99.98 69.09 0.24 0.15 0.28 **Std Dev** 0.64 4.38 2.15 5.53 24.71 11.07 0.04 0.02 0.06 **Maximum** 6.32 33.18 19.12 44.60 139.19 85.49 0.30 0.16 0.38 **Minimum** 4.90 20.65 12.81 31.05 68.45 54.55 0.20 0.11 0.22

**Geo.Mean** 5.32 29.66 17.49 42.88 133.91 88.45 0.20 0.12 0.22 **Mean** 5.40 30.00 17.70 42.99 134.91 88.95 0.20 0.12 0.22 **Std Dev** 1.07 5.26 3.13 3.47 19.36 11.10 0.02 0.01 0.02 **Maximum** 6.56 36.84 21.70 46.92 162.02 104.47 0.21 0.14 0.24 **Minimum** 4.46 24.52 14.52 39.45 116.09 78.43 0.17 0.11 0.19

**Geo.Mean** 5.22 22.83 14.11 32.98 92.48 63.19 0.22 0.16 0.25 **Mean** 5.38 23.81 14.59 33.81 98.28 66.04 0.23 0.16 0.25 **Std Dev** 1.36 6.65 3.69 7.51 33.25 19.08 0.04 0.04 0.06 **Maximum** 8.42 36.84 21.70 46.92 162.02 104.47 0.35 0.28 0.42 **Minimum** 3.21 12.96 8.45 21.00 47.13 35.12 0.17 0.11 0.19

**Geo.Mean** 17.30 78.42 0.22 **Mean** 17.81 80.57 0.22 **Std Dev** 4.40 18.83 0.04 **Maximum** 28.00 126.00 0.30 **Minimum** 11.00 42.00 0.10

Table 2. Statistical Analysis of the Result Obtained

**NOx 24-hr** **Sox/NOx 24-hr**

**Sox/NOx Day**

**Sox/NOx Night** 

maximum and minimum values, and geometric mean values are listed as below:

**Day Value**

**Okhla Day**

**IIT** 

**VV** 

**SKP** 

**Delhi** 

**ITO** 

**Value**

**Night Value**

**Sox 24-hr**

#### **3.3.1 Sources of sulfur oxides and nitrogen oxides**

The natural sources of nitrogen oxides are the geothermal activity and bacterial action. Natural sources include the soil microbes, vegetation, biomass, slum, burning, and lightning. Natural sources include the biological emissions (e.g. vegetation and soil microbes) and abiotic processes (e.g., geogenic, lightning, and biomass burning). Globally, lightning strikes occur about 100-times/second (Borucki and Chameides, 1984). Lighting strikes are characterized by a highly ionized and high temperature channel, in which molecular nitrogen is dissociated and reacts with oxygen to form NO (Guenther et al, 2000).

The primary sources of NOx are the motor vehicles, electric utilities, and other industrial, commercial, and residential sources that burn fuels. Other sources include the chemical and nitric acid manufacture, the detonation of nitrate-containing explosives, and electric sources. Man made sources of nitrogen oxide include the automobile, railways (coal fired and diesel locos), airways (landing and takeoff operations), industrial processes like chemicals (nitric acid plants, processes where nitrogenous compounds are used), metal processing, fertilizer (nitrogenous fertilizers), and other miscellaneous activity like domestic combustion, forest fires and managed fires, industrial boilers, furnaces, and commercial combustion. Although most NO is produced by the natural bacterial action, it must be emphasized that the rate of emission from man-related sources has, by estimate, increased by one-fourth between 1966 and 1968. Certainly, man's greatest contribution to the family of nitrogen oxides arises from the combustion of fuels. The contribution to the total NOx can be categorized as Utilities 27%; Motor Vehicles 49%; Other Sources 5%; and Industrial/ Commercial/Residential 19%. Nearly 2/3 of NOx emissions originate at the Earth's surface from fossil fuel and biomass combustion. Vehicle exhausts are known to increase the inputs of NO2 to the atmosphere (see, e.g., Hargreaves et al., 2000). The burning of natural gas (which produces negligible SOx, CO, and HC) does produce a very significant quantity of oxides of nitrogen.

Man caused nitrogen emissions as equivalent NO2 are estimated as 53x106 TPA over the world, 95% of which are emitted in the Northern Hemisphere. Half of these are from the combustion of coal. Natural sources of NOx are estimated as between 15 and 17-times pollutant emissions. (Sawyer et al, 1998). The major source of NOx is the combustion when the fixation of atmospheric N occurs at the high flame temperature. The oxides are emitted mainly as NO, which is normally rapidly oxidized to NO2 by atmospheric O2 and O3. Motor vehicle exhaust contributes a sizable fraction of the total emissions of NOx. As well as the stationary combustion sources, the manufacture of HNO3 and nitrate fertilizer is the source of NO & NO2. On a global basis, natural emissions of oxides of nitrogen (such as releases from soil and the ocean, lightning strikes, oxidation of ammonia, and stratospheric oxidation) all make substantial contributions to the total emissions.

Garg et al (2001) has reported the sectoral shares of NOx emissions for India in 1995. According to them, a major part of NOx emissions is contributed from 28% by the nonenergy sources, 28% by electric power generation, 2% by road transport, 19% by Biomass burning, 4% by other transports, and 19% by industries. Delhi tops the table of top 10 NOx emitting districts in the terms of total emissions by emitting total NOx 66-Gg in 1990 and 82- Gg in 1995; in the terms of emissions per unit area, NOx/area (tons km-2) was 44 in 1990 and 55 in 1995. The largest NOx emission chunk in Delhi is from the oil sources.

Estimation of Ambient Air Quality in Delhi 49

sources, 36% by electric power generation, 8% by road transport, 6% by biomass burning, 3% by other transports, and 1% by industries. The details of industrial SO2 emissions shows that 40%, 27%, 14%, 10%, 7%, and 2% are shared by non-energy industries, steel industry,

In Delhi, SO2/area (tkm-2) in 1990 was 47.9, in 1995 it was 46.7, and the average annual concentrations (μg/m3) were 16 in 1990 and 24 in 1995. The largest contribution to SO2 levels is by the transport sector. Our study results should be considered by keeping the points of the study done by Kamyotra et al (2000) and Jain (2001). These studies have pointed out two important points: by increasing the sampling period, the concentration shown becomes less (Kamyotra et al., 2000) and the diurnal pattern over 24 hours changes with time, local

The nighttime high values can be attributed to the phenomenon of the inversion. According to Kamyotra et al (2000), the lower values found during the day can be ascribed for the evaporation of the liquid, but this evaporation is absent during the night time which, in turn, increases the night time values, so high values obtained can be attributed to the

With a minimum of one 12-hour daily sample at a site, 12-hour a day SO2 concentration at the urban-industrial Delhi site varies from 3.50-8.42-μgm-3, 12-hour a night SO2 concentration varies from 23.58-30.12-μgm-3, and the average 24-hour SO2 concentration varies from 13.54-18.45-μgm-3. With a minimum of one 12-hour daily sample at a site, 12 hour a day SO2 concentration at the urban-commercial Delhi site varies from 4.90-6.32-μgm-3, 12-hour a night SO2 concentration varies from 20.65-33.18-μgm-3, and average 24-hour SO2 concentration varies from 12.81-19.12-μgm-3. With a minimum of one 12-hour daily sample at two sites, 12-hour a day SO2 concentration at the urban-residential Delhi site varies from 3.21-6.56-μgm-3, 12-hour a night SO2 concentration varies from 12.96-36.84-μgm-3, and

On Okhla, the highest day SO2 concentration variation (1.89-μgm-3) suggests that the source emission rate variation at Okhla is higher comparatively. On the VV site, day SO2 concentration variation was lowest in all (0.64-μgm-3), suggesting that the source diversity and source emission rate variation are lower comparatively. On the IIT site, day SO2 concentration variation is 0.83-μgm-3, and on the SKP site day SO2 concentration variation is

On SKP, Night SO2 concentration variation was the highest in all (5.26-μgm-3), suggesting that the source emission rate variation at SKP is higher comparatively. On the IIT site, night-SO2 concentration variation was the lowest in all (1.21-μgm-3), suggesting that the source diversity and source emission rate variation are lower comparatively. On the VV site, night SO2 concentration variation is 4.38-μgm-3, and on the Okhla site night SO2 concentration variation is 2.32-μgm-3. In all of Delhi, night SO2 concentration varies from 12.96-36.84-μgm-3. On SKP, 24-hour SO2 concentration variation was highest in all (3.13-μgm-3), suggesting that the source emission rate variation at SKP is higher comparatively. On the IIT site, 24 hour SO2 concentration variation was the lowest in all (0.81-μgm-3), suggesting that the source diversity and source emission rate variation are lower comparatively. On the VV site,

fertilizer, cement, other industries, and refineries, respectively.

average 24- hour SO2 concentration varies from 8.45-21.70-μgm-3.

1.07-μgm-3. In all of Delhi, day SO2 concentration varies from 3.21-8.42-μgm-3.

conditions, and activities (Jain, 2001).

combined effect of these facts.

**3.3.2 Variation in SO2** 

The major atmospheric man made sources of sulphur dioxide (SO2) are the burning of fossil fuel in stationary combustion and industrial process, viz. petroleum, chemical, metallurgical and mineral industries, domestic emission and mineral industries, domestic emission, and fires, while the natural source are the volcanoes (wherever these are present, but in India there are none), geothermal activity, the bacterial decomposition of organic matter, forest fires, and managed fires. Motor vehicles are a relatively minor source of SO2, since refined motor fuel normally has low sulfur content.

Ambient SO2 results largely from the stationary sources (that burn the coal and oil, refineries, pulp, and paper mills) and from the nonferrous metal smelters. The sources of the pollution include the emissions from the on-road vehicles, non-road vehicles (like planes, ships, trains and industries), and small businesses and households where polluting products are used.

Nitrogen oxides (NOx) are produced when fossil fuels are burned in the motor vehicles, power plants, furnaces, and turbines. Other gasoline powered mobile sources (motorcycles, recreational vehicles, lawn, garden, and utility equipment) have high emissions on per quantity of the fuel consumed basis, but their contribution to the total emissions is small. Heavy-duty diesel vehicles are the dominant mobile source of the Oxides of nitrogen. Oxides of nitrogen emissions on a fuel-consumed basis are much greater from the diesel mobile sources than from the gas.

Another common source of SO2 in the atmosphere is the metallurgical operations. Many ores, like Zn, Cu, and Pb are primarily sulfides. During the smelting of these ores, SO2 is evolved in the stack concentrations of 5-10% (SO2). Among miscellaneous operations releasing SO2 into the atmosphere are the sulphuric acid plant, paper-manufacturing plants, open burning of the refuse, and municipal incinerators which contribute some amount of SO2 to the atmosphere (Peavy et al, 1985).

In the entire sulphur entering atmosphere, 1/3rd is by the anthropogenic activities, and 2/3rd come from natural sources such as Hydrogen Sulfide (H2S) or SOx. The problem with the manmade sources is the distribution, rather than the amount. Delhi, in this instance, is being recognized as a city with a potentially high traffic load on its roads. Therefore, the combustion of fossil fuels like diesel and petrol can be considered as a major source of the SO2 in Delhi. Besides this thermal power plant, which is located in the Southeast about 1-1.5 km far from the SKP station, it is also contributing SO2 in the atmosphere because the coal is used as a main fuel for the heat and the power generation. However, SO2 concentration in the air also depends upon the sulfur content of fuels that vary from less than 1% of good quality anthracite to over 4% for bituminous coal.

Most crude petroleum products contain less than 1% sulfur; a few contain up to 5%. In the coming years, Delhi is supposed to have less concentration of SO2 in its atmosphere because the fuel with less S content has already been introduced into the market. Another common source of the sulfur dioxide in Delhi is metallurgical operation. Among the miscellaneous operations releasing releasing SO2 into the atmosphere are sulfuric plants and paper manufacturing plants. The open burning of refuse and municipal incinerator also contribute some amount of SO2.

Garg et al (2001) also reported the sectoral shares of SO2 emissions for India in 1995. According to them, a major part of SO2 emissions is contributed to 46% by non-energy

The major atmospheric man made sources of sulphur dioxide (SO2) are the burning of fossil fuel in stationary combustion and industrial process, viz. petroleum, chemical, metallurgical and mineral industries, domestic emission and mineral industries, domestic emission, and fires, while the natural source are the volcanoes (wherever these are present, but in India there are none), geothermal activity, the bacterial decomposition of organic matter, forest fires, and managed fires. Motor vehicles are a relatively minor source of SO2, since refined

Ambient SO2 results largely from the stationary sources (that burn the coal and oil, refineries, pulp, and paper mills) and from the nonferrous metal smelters. The sources of the pollution include the emissions from the on-road vehicles, non-road vehicles (like planes, ships, trains and industries), and small businesses and households where polluting products are used.

Nitrogen oxides (NOx) are produced when fossil fuels are burned in the motor vehicles, power plants, furnaces, and turbines. Other gasoline powered mobile sources (motorcycles, recreational vehicles, lawn, garden, and utility equipment) have high emissions on per quantity of the fuel consumed basis, but their contribution to the total emissions is small. Heavy-duty diesel vehicles are the dominant mobile source of the Oxides of nitrogen. Oxides of nitrogen emissions on a fuel-consumed basis are much greater from the diesel

Another common source of SO2 in the atmosphere is the metallurgical operations. Many ores, like Zn, Cu, and Pb are primarily sulfides. During the smelting of these ores, SO2 is evolved in the stack concentrations of 5-10% (SO2). Among miscellaneous operations releasing SO2 into the atmosphere are the sulphuric acid plant, paper-manufacturing plants, open burning of the refuse, and municipal incinerators which contribute some amount of

In the entire sulphur entering atmosphere, 1/3rd is by the anthropogenic activities, and 2/3rd come from natural sources such as Hydrogen Sulfide (H2S) or SOx. The problem with the manmade sources is the distribution, rather than the amount. Delhi, in this instance, is being recognized as a city with a potentially high traffic load on its roads. Therefore, the combustion of fossil fuels like diesel and petrol can be considered as a major source of the SO2 in Delhi. Besides this thermal power plant, which is located in the Southeast about 1-1.5 km far from the SKP station, it is also contributing SO2 in the atmosphere because the coal is used as a main fuel for the heat and the power generation. However, SO2 concentration in the air also depends upon the sulfur content of fuels that vary from less than 1% of good

Most crude petroleum products contain less than 1% sulfur; a few contain up to 5%. In the coming years, Delhi is supposed to have less concentration of SO2 in its atmosphere because the fuel with less S content has already been introduced into the market. Another common source of the sulfur dioxide in Delhi is metallurgical operation. Among the miscellaneous operations releasing releasing SO2 into the atmosphere are sulfuric plants and paper manufacturing plants. The open burning of refuse and municipal incinerator also contribute

Garg et al (2001) also reported the sectoral shares of SO2 emissions for India in 1995. According to them, a major part of SO2 emissions is contributed to 46% by non-energy

motor fuel normally has low sulfur content.

mobile sources than from the gas.

SO2 to the atmosphere (Peavy et al, 1985).

quality anthracite to over 4% for bituminous coal.

some amount of SO2.

sources, 36% by electric power generation, 8% by road transport, 6% by biomass burning, 3% by other transports, and 1% by industries. The details of industrial SO2 emissions shows that 40%, 27%, 14%, 10%, 7%, and 2% are shared by non-energy industries, steel industry, fertilizer, cement, other industries, and refineries, respectively.

In Delhi, SO2/area (tkm-2) in 1990 was 47.9, in 1995 it was 46.7, and the average annual concentrations (μg/m3) were 16 in 1990 and 24 in 1995. The largest contribution to SO2 levels is by the transport sector. Our study results should be considered by keeping the points of the study done by Kamyotra et al (2000) and Jain (2001). These studies have pointed out two important points: by increasing the sampling period, the concentration shown becomes less (Kamyotra et al., 2000) and the diurnal pattern over 24 hours changes with time, local conditions, and activities (Jain, 2001).

The nighttime high values can be attributed to the phenomenon of the inversion. According to Kamyotra et al (2000), the lower values found during the day can be ascribed for the evaporation of the liquid, but this evaporation is absent during the night time which, in turn, increases the night time values, so high values obtained can be attributed to the combined effect of these facts.

#### **3.3.2 Variation in SO2**

With a minimum of one 12-hour daily sample at a site, 12-hour a day SO2 concentration at the urban-industrial Delhi site varies from 3.50-8.42-μgm-3, 12-hour a night SO2 concentration varies from 23.58-30.12-μgm-3, and the average 24-hour SO2 concentration varies from 13.54-18.45-μgm-3. With a minimum of one 12-hour daily sample at a site, 12 hour a day SO2 concentration at the urban-commercial Delhi site varies from 4.90-6.32-μgm-3, 12-hour a night SO2 concentration varies from 20.65-33.18-μgm-3, and average 24-hour SO2 concentration varies from 12.81-19.12-μgm-3. With a minimum of one 12-hour daily sample at two sites, 12-hour a day SO2 concentration at the urban-residential Delhi site varies from 3.21-6.56-μgm-3, 12-hour a night SO2 concentration varies from 12.96-36.84-μgm-3, and average 24- hour SO2 concentration varies from 8.45-21.70-μgm-3.

On Okhla, the highest day SO2 concentration variation (1.89-μgm-3) suggests that the source emission rate variation at Okhla is higher comparatively. On the VV site, day SO2 concentration variation was lowest in all (0.64-μgm-3), suggesting that the source diversity and source emission rate variation are lower comparatively. On the IIT site, day SO2 concentration variation is 0.83-μgm-3, and on the SKP site day SO2 concentration variation is 1.07-μgm-3. In all of Delhi, day SO2 concentration varies from 3.21-8.42-μgm-3.

On SKP, Night SO2 concentration variation was the highest in all (5.26-μgm-3), suggesting that the source emission rate variation at SKP is higher comparatively. On the IIT site, night-SO2 concentration variation was the lowest in all (1.21-μgm-3), suggesting that the source diversity and source emission rate variation are lower comparatively. On the VV site, night SO2 concentration variation is 4.38-μgm-3, and on the Okhla site night SO2 concentration variation is 2.32-μgm-3. In all of Delhi, night SO2 concentration varies from 12.96-36.84-μgm-3. On SKP, 24-hour SO2 concentration variation was highest in all (3.13-μgm-3), suggesting that the source emission rate variation at SKP is higher comparatively. On the IIT site, 24 hour SO2 concentration variation was the lowest in all (0.81-μgm-3), suggesting that the source diversity and source emission rate variation are lower comparatively. On the VV site,

Estimation of Ambient Air Quality in Delhi 51

On the VV site, night-NO2 concentration variation was the highest in all (24.71-µg/m3), which suggests that the source emission rate variation at SKP is higher comparatively. On the IIT site, night NO2 concentration variation was the lowest in all (8.67-μgm-3), which suggests that the source diversity and source emission rate variation is lower comparatively. On the Okhla site night NO2 concentration variation is 24.20-μgm-3, and on the SKP site night NO2 concentration variation is 19.36-μgm-3. In all of Delhi, night NO2 concentration

On Okhla, 24-hour NO2 concentration variation was highest in all (11.68-μgm-3), which suggests that the source emission rate variation at Okhla is higher comparatively. On the IIT site, 24-hour NO2 concentration variation was the lowest in all (5.67-μgm-3), which suggests that the source diversity and source emission rate variation are lower comparatively. On the VV site, 24-hour NO2 concentration variation is 11.07-μgm-3, and on the SKP site, 24-hour NO2 concentration variation is 11.10-μgm-3. In all of Delhi, 24-hour NO2 concentration varies

The maximum day NOx concentration in Delhi, 46.92-μg/m3, was at SKP on November 27, 2000. The lowest day NOx value, 21.00-µg/m3, was observed at IIT on November 4, 2000. The maximum day NOx concentration in Okhla, 36.89-μg/m3, was on October 25, 2000. The maximum day NOx concentration at VV, 44.60-μg/m3, was on November 15, 2000. The maximum day NOx concentration at IIT, 29.85-μg/m3, was on November 6, 2000.

The maximum night NOx concentration in Delhi, 162.02-μg/m3, was at SKP on November 27, 2000. The lowest night NOx value, 47.13-μgm-3, was observed at IIT on November 2, 2000. The maximum night NOx concentration in Okhla, 136.62-μg/m3, was on October 23, 2000. The maximum night NOx concentration at VV, 139.19-μg/m3, was on November 23, 2000.

The maximum 24-hour NOx concentration in Delhi, 104.47-μg/m3, was at SKP on November 27, 2000. The lowest 24-hour NOx value, 35.12-μgm-3, was observed at IIT on November 2, 2000. The maximum 24-hour NOx concentration in Okhla, 84.87-μg/m3, was on October 23, 2000. The maximum 24-hour NOx concentration at VV, 85.49-μg/m3, was on November 23,

Our observations at Okhla in South Delhi showed that day NOx, night NOx, and average 24- hour concentrations were 32.48±2.73, 107.85±24.20, and 70.68±11.68-μgm-3, respectively. Observations at IIT showed that day NOx, night NOx, and average 24-hour concentrations were 24.56±3.06, 57.29±8.67, and 40.95±5.67-μgm-3, respectively. VV showed that day NOx, night NOx, and average 24-hour concentrations were 37.86±5.53, 97.48±24.71, and 68.36±11.07-μgm-3, respectively. SKP in East Delhi showed that day NOx, night NOx, and average 24-hour concentrations were 42.88±3.47, 133.91±19.36, and 88.45±11.10-μgm-3, respectively, and the overall concentrations in Delhi were 32.98±7.51, 92.48±33.25, and

As evident from the table 2, the concentration of NOx was the highest at SKP, followed by Okhla, VV, and IIT. The trend is attributed to the prevailing activities at the sites. 24-hour average NOx concentration is mainly contributed by the nighttime values. These nighttime higher NOx concentrations at the different sites also reflect the local source conditions.

The maximum night NOx concentration at IIT, 70.37-μg/m3, was on November 6, 2000.

2000. The maximum NOx concentration at IIT, 50.11-μg/m3, was on November 6, 2000.

varies from 47.13-162.02-μgm-3.

from 35.12-104.47-μgm-3.

63.19±19.08-μgm-3.

**3.3.5 Peak day composition** 

24-hour SO2 concentration variation is 2.15-μgm-3, and on the Okhla site 24-hour SO2 concentration variation is 1.76-μgm-3. In all of Delhi, 24-hour SO2 concentration varies from 8.45-21.70-μgm-3.

#### **3.3.3 Peak day composition**

The maximum day SOx concentration in Delhi, 8.42-μg/m3, was at Okhla on October 25, 2000. The lowest day SOx value 3.21-μgm-3 was observed at IIT on November 2, 2000. The maximum day SOx concentration in SKP, 6.56-μg/m3, was on November 27, 2000. The maximum day SOx concentration at VV, 6.32-μg/m3, was on November 22, 2000. The maximum night SOx concentration in Delhi, 36.84-μg/m3, was at SKP on November 27, 2000. The lowest night SOx value 12.96-μgm-3 was observed at IIT on November 5, 2000. The maximum night SOx concentration in Okhla, 30.12-μg/m3, was on October 23, 2000. The maximum night SOx concentration at VV, 33.18-μg/m3, was on November 23, 2000.

The maximum 24-hour SOx concentration in Delhi, 21.70-μg/m3, was at SKP on November 27, 2000. The lowest 24-hour SOx value, 8.45μg/m3, was observed at IIT on November 2, 2000. The maximum 24-hr SOx concentration in Okhla, 18.45-μg/m3, was on October 23, 2000. The maximum 24-hour SOx concentration at VV, 19.12-μg/m3, was on November 23, 2000. Summarizing our observations at Okhla in South Delhi showed that day SOx, night SOx, and average 24-hour concentrations were 5.86±1.89, 25.47±2.32, and 15.76±1.76-μgm-3, respectively. IIT showed that day SOx, night SOx, and average 24-hour concentrations were 4.30±0.83, 14.45±1.21, and 9.40±0.81-μgm-3, respectively. VV showed that day SOx, night SOx, and average 24-hour concentrations were 5.48±0.64, 26.67±4.38, and 16.12±2.15-μgm-3, respectively. SKP in East Delhi showed that day SOx, night SOx, and average 24-hour concentrations were 5.32±1.07, 29.66±5.26, and 17.49±3.13-μgm-3, respectively, and overall in Delhi were 5.22±1.36, 22.83±6.65, and 14.11±3.69-μgm-3.

#### **3.3.4 Variation in NO2**

With a minimum of one 12-hour daily sample at a site, 12-hour per day NO2 concentration at the urban-industrial Delhi site varies from 28.20-36.89-μgm-3, 12-hour per night NO2 concentration varies from 64.55-136.62-μgm-3 and average 24-hour NO2 concentration varies from 50.72-84.87-μgm-3. With a minimum of one 12-hour daily sample at a site, 12-hour per day NO2 concentration at the urban-commercial Delhi site varies from 31.05-44.60-μgm-3, 12 hour per night NO2 concentration varies from 68.45-139.19-μgm-3, and average 24-hour NO2 concentration varies from 54.55-85.49-μgm-3. With a minimum of one 12-hour daily sample at two sites, 12-hour per day NO2 concentration at the urban-residential Delhi site varies from 21.00-46.92-μgm-3, 12-hour per night NO2 concentration varies from 47.13-162.02-μgm-3 and average 24-hour NO2 concentration varies from 35.12-104.47-μgm-3.

On VV, day NO2 concentration variation was highest in all (5.53-μgm-3), which suggests that the source emission rate variation at VV is higher comparatively. On the Okhla site, day NO2 concentration variation was the lowest in all (2.73-μgm-3), which suggests that the source diversity and source emission rate variation is lower comparatively. On the IIT site, day-NO2 concentration variation is 3.06-μgm-3, and on the SKP site, the day NO2 concentration variation is 3.47-μgm-3. In all of Delhi, the day NO2 concentration varies from 21.00-46.92-μgm-3.

24-hour SO2 concentration variation is 2.15-μgm-3, and on the Okhla site 24-hour SO2 concentration variation is 1.76-μgm-3. In all of Delhi, 24-hour SO2 concentration varies from

The maximum day SOx concentration in Delhi, 8.42-μg/m3, was at Okhla on October 25, 2000. The lowest day SOx value 3.21-μgm-3 was observed at IIT on November 2, 2000. The maximum day SOx concentration in SKP, 6.56-μg/m3, was on November 27, 2000. The maximum day SOx concentration at VV, 6.32-μg/m3, was on November 22, 2000. The maximum night SOx concentration in Delhi, 36.84-μg/m3, was at SKP on November 27, 2000. The lowest night SOx value 12.96-μgm-3 was observed at IIT on November 5, 2000. The maximum night SOx concentration in Okhla, 30.12-μg/m3, was on October 23, 2000. The

The maximum 24-hour SOx concentration in Delhi, 21.70-μg/m3, was at SKP on November 27, 2000. The lowest 24-hour SOx value, 8.45μg/m3, was observed at IIT on November 2, 2000. The maximum 24-hr SOx concentration in Okhla, 18.45-μg/m3, was on October 23, 2000. The maximum 24-hour SOx concentration at VV, 19.12-μg/m3, was on November 23, 2000. Summarizing our observations at Okhla in South Delhi showed that day SOx, night SOx, and average 24-hour concentrations were 5.86±1.89, 25.47±2.32, and 15.76±1.76-μgm-3, respectively. IIT showed that day SOx, night SOx, and average 24-hour concentrations were 4.30±0.83, 14.45±1.21, and 9.40±0.81-μgm-3, respectively. VV showed that day SOx, night SOx, and average 24-hour concentrations were 5.48±0.64, 26.67±4.38, and 16.12±2.15-μgm-3, respectively. SKP in East Delhi showed that day SOx, night SOx, and average 24-hour concentrations were 5.32±1.07, 29.66±5.26, and 17.49±3.13-μgm-3, respectively, and overall in

With a minimum of one 12-hour daily sample at a site, 12-hour per day NO2 concentration at the urban-industrial Delhi site varies from 28.20-36.89-μgm-3, 12-hour per night NO2 concentration varies from 64.55-136.62-μgm-3 and average 24-hour NO2 concentration varies from 50.72-84.87-μgm-3. With a minimum of one 12-hour daily sample at a site, 12-hour per day NO2 concentration at the urban-commercial Delhi site varies from 31.05-44.60-μgm-3, 12 hour per night NO2 concentration varies from 68.45-139.19-μgm-3, and average 24-hour NO2 concentration varies from 54.55-85.49-μgm-3. With a minimum of one 12-hour daily sample at two sites, 12-hour per day NO2 concentration at the urban-residential Delhi site varies from 21.00-46.92-μgm-3, 12-hour per night NO2 concentration varies from 47.13-162.02-μgm-3

On VV, day NO2 concentration variation was highest in all (5.53-μgm-3), which suggests that the source emission rate variation at VV is higher comparatively. On the Okhla site, day NO2 concentration variation was the lowest in all (2.73-μgm-3), which suggests that the source diversity and source emission rate variation is lower comparatively. On the IIT site, day-NO2 concentration variation is 3.06-μgm-3, and on the SKP site, the day NO2 concentration variation is 3.47-μgm-3. In all of Delhi, the day NO2 concentration varies

maximum night SOx concentration at VV, 33.18-μg/m3, was on November 23, 2000.

Delhi were 5.22±1.36, 22.83±6.65, and 14.11±3.69-μgm-3.

and average 24-hour NO2 concentration varies from 35.12-104.47-μgm-3.

8.45-21.70-μgm-3.

**3.3.3 Peak day composition** 

**3.3.4 Variation in NO2** 

from 21.00-46.92-μgm-3.

On the VV site, night-NO2 concentration variation was the highest in all (24.71-µg/m3), which suggests that the source emission rate variation at SKP is higher comparatively. On the IIT site, night NO2 concentration variation was the lowest in all (8.67-μgm-3), which suggests that the source diversity and source emission rate variation is lower comparatively. On the Okhla site night NO2 concentration variation is 24.20-μgm-3, and on the SKP site night NO2 concentration variation is 19.36-μgm-3. In all of Delhi, night NO2 concentration varies from 47.13-162.02-μgm-3.

On Okhla, 24-hour NO2 concentration variation was highest in all (11.68-μgm-3), which suggests that the source emission rate variation at Okhla is higher comparatively. On the IIT site, 24-hour NO2 concentration variation was the lowest in all (5.67-μgm-3), which suggests that the source diversity and source emission rate variation are lower comparatively. On the VV site, 24-hour NO2 concentration variation is 11.07-μgm-3, and on the SKP site, 24-hour NO2 concentration variation is 11.10-μgm-3. In all of Delhi, 24-hour NO2 concentration varies from 35.12-104.47-μgm-3.

#### **3.3.5 Peak day composition**

The maximum day NOx concentration in Delhi, 46.92-μg/m3, was at SKP on November 27, 2000. The lowest day NOx value, 21.00-µg/m3, was observed at IIT on November 4, 2000. The maximum day NOx concentration in Okhla, 36.89-μg/m3, was on October 25, 2000. The maximum day NOx concentration at VV, 44.60-μg/m3, was on November 15, 2000. The maximum day NOx concentration at IIT, 29.85-μg/m3, was on November 6, 2000.

The maximum night NOx concentration in Delhi, 162.02-μg/m3, was at SKP on November 27, 2000. The lowest night NOx value, 47.13-μgm-3, was observed at IIT on November 2, 2000. The maximum night NOx concentration in Okhla, 136.62-μg/m3, was on October 23, 2000. The maximum night NOx concentration at VV, 139.19-μg/m3, was on November 23, 2000. The maximum night NOx concentration at IIT, 70.37-μg/m3, was on November 6, 2000.

The maximum 24-hour NOx concentration in Delhi, 104.47-μg/m3, was at SKP on November 27, 2000. The lowest 24-hour NOx value, 35.12-μgm-3, was observed at IIT on November 2, 2000. The maximum 24-hour NOx concentration in Okhla, 84.87-μg/m3, was on October 23, 2000. The maximum 24-hour NOx concentration at VV, 85.49-μg/m3, was on November 23, 2000. The maximum NOx concentration at IIT, 50.11-μg/m3, was on November 6, 2000.

Our observations at Okhla in South Delhi showed that day NOx, night NOx, and average 24- hour concentrations were 32.48±2.73, 107.85±24.20, and 70.68±11.68-μgm-3, respectively. Observations at IIT showed that day NOx, night NOx, and average 24-hour concentrations were 24.56±3.06, 57.29±8.67, and 40.95±5.67-μgm-3, respectively. VV showed that day NOx, night NOx, and average 24-hour concentrations were 37.86±5.53, 97.48±24.71, and 68.36±11.07-μgm-3, respectively. SKP in East Delhi showed that day NOx, night NOx, and average 24-hour concentrations were 42.88±3.47, 133.91±19.36, and 88.45±11.10-μgm-3, respectively, and the overall concentrations in Delhi were 32.98±7.51, 92.48±33.25, and 63.19±19.08-μgm-3.

As evident from the table 2, the concentration of NOx was the highest at SKP, followed by Okhla, VV, and IIT. The trend is attributed to the prevailing activities at the sites. 24-hour average NOx concentration is mainly contributed by the nighttime values. These nighttime higher NOx concentrations at the different sites also reflect the local source conditions.

Estimation of Ambient Air Quality in Delhi 53

hour NAAQS of NOx i.e. 80-μg/m3 in the residential area and 120-μg/m3 in the industrial area, except on November 23, 2000 at VV and on November 26, 27, and 28, 2000 at

The ITO site can be considered as a commercial site in East Delhi, so we will compare our results of VV with it. We will also compare our results of SKP with it. The levels obtained at ITO in East Delhi showed that NOx, SOx concentrations and SOx/NOx ratio were 78.42±18.83, 17.30±4.40-μgm-3, and 0.22±0.04, while at VV levels were 68.36±11.07, 16.12±2.15-μgm-3, and 0.24±0.04. Our data also did not show any weekend or holiday effect. The variation is most

Das et al (1997) reported in NO2, SO2, and NO2/SO2 levels in Jaipur for the sampling time of two-hours in the evening peak traffic hours at 49-stations. At residential areas, NO2 and SO2 values varied between 4.61-332.89-μg/m3 and 115.84-296.74-μg/m3; at the commercial areas, between 1.07-974.44-μg/m3 and 136.96-912.64-μg/m3; and at the industrial areas, between 26.16-78.81-μg/m3 and 141.44-262.02-μg/m3. Gajraj et al (2000) reported for May-June 1999 SO2 and NOx levels during the morning and the evening for the rush hours in Jaipur as 56.7- 184.4-μg/m3 and 70.8-139-μg/m3. Sharma et al. (2003) reported SO2 levels in Kanpur for 24 hours at five stations. They found at IIT Kanpur SO2, the levels were 5.44±1.58-μg/m3, at GT road were 28.6±9.9-μg/m3, at Agricultural University were 11.0±2.7-μg/m3, at Naveen

CPCB (http://www.envfor.nic.in/divisions/cpoll/delpoll.html) have reported an increase in the annual SO2 level over the years (1989-1997), such as 8.7, 10.2, 13.3, 18.4, 18.5, 19.5, 19.0, 19.0, and 16.2-μg/m3, and in annual NO2 level such as 18.5, 22.5, 27.2, 30.4, 33.2, 33.0, 34.1, 33.7, and 33.0-μg/m3. Though the annual mean value of SO2 (15-26-µg/m3) and NOx (28-46-µg/m3) remain within the prescribed limits of 60-80-µg/m3, there is a rising trend. Compared to 1989, SO2 atmospheric concentrations in 1996 have registered a 109% rise, and NOx an 82% rise. Glikson et al (1995) reported in Brisbane, Australia for the period of the study that the maximum 1-hour average for NO2 was 6.7-ppm. Cheng and Lam (2000) reported in Hong Kong the seasonal daily mean SO2 concentrations for the entire period (1983-1992) were 28.2, 38.3, 20.0, and 23.5-µgm-3 for spring, summer, autumn, and winter, respectively; their standard deviations were 30.8, 42.7, 16.9, and 23.6-µgm-3. The corresponding figures of the overall mean NOx concentrations were 129.7, 110.4, 121.0, and 148.6-µgm-3 with standard deviations of 54.2, 49.7, 40.1, and 74.5-µgm-3. Mondal et al (2000) reported the ground level concentration of NOx at 19-important traffic intersection points in Calcutta ranging from 55-222-μgm-3. Lebret et al (2000) reported small area variations in ambient NO2 concentrations in four European areas, i.e., in Amsterdam (Netherlands), Huddersfield (UK), Poznan (Poland), and Prague (Czech Republic) found NO2 concentrations as 24-72, 10-79, 12-48, and 9.6-18.4-μgm-3. In NOx emission Delhi holds second position, and in SO2 Delhi is not in the top 10-districts of India (Garg et al., 2001). India is the second largest contributor of SO2 emissions in Asia, after China. For India, the estimates of national SO2 emissions (Gg/yr.) in the period 1985-97 shows a rising trend as the emissions were 3402.0, 3609.1, 3829.9, 4085.2, 4256.1, 4437.2, 4684.3, 4863.1, 5039.6, 5289.3, 5609.5, 6106.3, and 6276.6-Gg/yr in the years 1985, 1986, 1987, 1988, 1989, 1990, 1991, 1992, 1993, 1994, 1995, 1996, and 1997, respectively (Streets et al., 2000). Our study is in good

likely due to differences in the air mass, emission characteristics, and the meteorology.

Market were 14.6±3.8-μg/m3, and at Lajpat Nagar were 17.73±4.38-μg/m3.

Shakarpur (80-μg/m3).

**3.3.9 Comparison with other studies** 

agreement with all of these studies.

#### **3.3.6 SOx to NOx day ratio**

The day ratio at an urban-industrial Delhi site varies from 0.11-0.28 and at the urbancommercial Delhi site varies from 0.11-0.16; at two urban-residential Delhi sites, it varies from 0.11-0.22. At SKP, variation of the ratio (0.01) is lowest, while it is highest at the Okhla (0.06). The variation of ratio at VV was 0.02, and at IIT it was 0.03, and in Delhi the ratio ranged from 0.11-0.28. The variation in Delhi (4%) shows that the sources liberating NOx are also releasing SOx. Summarizing our observations at Okhla in South Delhi showed that the day ratio was 0.18±0.06. IIT in South Delhi showed that the ratio was 0.18±0.03. VV in South Delhi showed that the ratio was 0.14±0.02. SKP in East Delhi showed that the ratio was 0.12±0.01, and overall the Delhi ratio was 0.16±0.04.

#### **3.3.7 SOx to NOx night ratio**

The night ratio at the urban-industrial Delhi site varies from 0.19-0.42, and at the urbancommercial Delhi site varies from 0.22-0.38, and at two urban-residential Delhi sites varies from 0.19-0.29. At SKP, the variation of the ratio (0.02) is the lowest, while it remains the highest at Okhla (0.08). The variation of ratio at VV was 0.06, and at IIT it was 0.03, and in Delhi the ratio ranges from 0.19-0.42. The variation in Delhi (6%) shows that sources liberating NOx are also releasing SOx. Summarizing our observations at Okhla in South Delhi showed that the night ratio was 0.24±0.08. IIT in South Delhi showed that the ratio was 0.25±0.03. VV in South Delhi showed that the ratio was 0.27±0.06. SKP in East Delhi showed that the ratio was 0.22±0.02, and overall in Delhi the ratio was 0.25±0.06.

#### **3.3.8 Sox to NOx 24-hr ratio**

The 24-hour ratio at the urban-industrial Delhi site varies from 0.18-0.35, and at the urbancommercial Delhi site varies from 0.20-0.30, and at two urban-residential Delhi sites varies from 0.17-0.26. The SKP and IIT variation of the ratio (0.02) is the lowest, while the highest is at Okhla (0.06). The variation of ratio at VV was 0.04, and overall in Delhi the ratio ranges from 0.17-0.35. The variation in Delhi (4%) shows that the sources liberating NOx are also releasing SOx.

Summarizing our observations at Okhla in South Delhi showed that the 24-hour ratio was 0.22±0.06. IIT in South Delhi showed that the ratio was 0.23±0.02. VV in South Delhi showed that the ratio was 0.24±0.04. SKP in East Delhi showed that the ratio was 0.20±0.02, and overall in Delhi the ratio was 0.22±0.04.

Many studies indicated that NOx and SOx concentration in ambient air would be affected by the various meteorological factors such as the wind speed, wind direction, solar radiation, relative humidity, rainfall, and source conditions. Our results could also be compared with the values obtained by the CPCB personnel at ITO during the study. These were taken from their site (http://www.envfor.nic.in/). Our values are in good agreement with the CPCB values. During comparison, it was kept in mind that these values are of those samples, which were taken for eight-hour duration as per CPCB rules. But these values are normalized for 24-hours to represent the daily mean value (as per CPCB personnel).

The SOx levels were always lesser than the 24-hour NAAQS of SOx, i.e., 80-μg/m3 in the residential area and 120-μg/m3 in the industrial area. The NOx levels were lesser than 24-

The day ratio at an urban-industrial Delhi site varies from 0.11-0.28 and at the urbancommercial Delhi site varies from 0.11-0.16; at two urban-residential Delhi sites, it varies from 0.11-0.22. At SKP, variation of the ratio (0.01) is lowest, while it is highest at the Okhla (0.06). The variation of ratio at VV was 0.02, and at IIT it was 0.03, and in Delhi the ratio ranged from 0.11-0.28. The variation in Delhi (4%) shows that the sources liberating NOx are also releasing SOx. Summarizing our observations at Okhla in South Delhi showed that the day ratio was 0.18±0.06. IIT in South Delhi showed that the ratio was 0.18±0.03. VV in South Delhi showed that the ratio was 0.14±0.02. SKP in East Delhi showed that the ratio was

The night ratio at the urban-industrial Delhi site varies from 0.19-0.42, and at the urbancommercial Delhi site varies from 0.22-0.38, and at two urban-residential Delhi sites varies from 0.19-0.29. At SKP, the variation of the ratio (0.02) is the lowest, while it remains the highest at Okhla (0.08). The variation of ratio at VV was 0.06, and at IIT it was 0.03, and in Delhi the ratio ranges from 0.19-0.42. The variation in Delhi (6%) shows that sources liberating NOx are also releasing SOx. Summarizing our observations at Okhla in South Delhi showed that the night ratio was 0.24±0.08. IIT in South Delhi showed that the ratio was 0.25±0.03. VV in South Delhi showed that the ratio was 0.27±0.06. SKP in East Delhi

The 24-hour ratio at the urban-industrial Delhi site varies from 0.18-0.35, and at the urbancommercial Delhi site varies from 0.20-0.30, and at two urban-residential Delhi sites varies from 0.17-0.26. The SKP and IIT variation of the ratio (0.02) is the lowest, while the highest is at Okhla (0.06). The variation of ratio at VV was 0.04, and overall in Delhi the ratio ranges from 0.17-0.35. The variation in Delhi (4%) shows that the sources liberating NOx are also

Summarizing our observations at Okhla in South Delhi showed that the 24-hour ratio was 0.22±0.06. IIT in South Delhi showed that the ratio was 0.23±0.02. VV in South Delhi showed that the ratio was 0.24±0.04. SKP in East Delhi showed that the ratio was 0.20±0.02, and

Many studies indicated that NOx and SOx concentration in ambient air would be affected by the various meteorological factors such as the wind speed, wind direction, solar radiation, relative humidity, rainfall, and source conditions. Our results could also be compared with the values obtained by the CPCB personnel at ITO during the study. These were taken from their site (http://www.envfor.nic.in/). Our values are in good agreement with the CPCB values. During comparison, it was kept in mind that these values are of those samples, which were taken for eight-hour duration as per CPCB rules. But these values are

The SOx levels were always lesser than the 24-hour NAAQS of SOx, i.e., 80-μg/m3 in the residential area and 120-μg/m3 in the industrial area. The NOx levels were lesser than 24-

normalized for 24-hours to represent the daily mean value (as per CPCB personnel).

showed that the ratio was 0.22±0.02, and overall in Delhi the ratio was 0.25±0.06.

**3.3.6 SOx to NOx day ratio** 

**3.3.7 SOx to NOx night ratio** 

**3.3.8 Sox to NOx 24-hr ratio** 

overall in Delhi the ratio was 0.22±0.04.

releasing SOx.

0.12±0.01, and overall the Delhi ratio was 0.16±0.04.

hour NAAQS of NOx i.e. 80-μg/m3 in the residential area and 120-μg/m3 in the industrial area, except on November 23, 2000 at VV and on November 26, 27, and 28, 2000 at Shakarpur (80-μg/m3).

The ITO site can be considered as a commercial site in East Delhi, so we will compare our results of VV with it. We will also compare our results of SKP with it. The levels obtained at ITO in East Delhi showed that NOx, SOx concentrations and SOx/NOx ratio were 78.42±18.83, 17.30±4.40-μgm-3, and 0.22±0.04, while at VV levels were 68.36±11.07, 16.12±2.15-μgm-3, and 0.24±0.04. Our data also did not show any weekend or holiday effect. The variation is most likely due to differences in the air mass, emission characteristics, and the meteorology.

#### **3.3.9 Comparison with other studies**

Das et al (1997) reported in NO2, SO2, and NO2/SO2 levels in Jaipur for the sampling time of two-hours in the evening peak traffic hours at 49-stations. At residential areas, NO2 and SO2 values varied between 4.61-332.89-μg/m3 and 115.84-296.74-μg/m3; at the commercial areas, between 1.07-974.44-μg/m3 and 136.96-912.64-μg/m3; and at the industrial areas, between 26.16-78.81-μg/m3 and 141.44-262.02-μg/m3. Gajraj et al (2000) reported for May-June 1999 SO2 and NOx levels during the morning and the evening for the rush hours in Jaipur as 56.7- 184.4-μg/m3 and 70.8-139-μg/m3. Sharma et al. (2003) reported SO2 levels in Kanpur for 24 hours at five stations. They found at IIT Kanpur SO2, the levels were 5.44±1.58-μg/m3, at GT road were 28.6±9.9-μg/m3, at Agricultural University were 11.0±2.7-μg/m3, at Naveen Market were 14.6±3.8-μg/m3, and at Lajpat Nagar were 17.73±4.38-μg/m3.

CPCB (http://www.envfor.nic.in/divisions/cpoll/delpoll.html) have reported an increase in the annual SO2 level over the years (1989-1997), such as 8.7, 10.2, 13.3, 18.4, 18.5, 19.5, 19.0, 19.0, and 16.2-μg/m3, and in annual NO2 level such as 18.5, 22.5, 27.2, 30.4, 33.2, 33.0, 34.1, 33.7, and 33.0-μg/m3. Though the annual mean value of SO2 (15-26-µg/m3) and NOx (28-46-µg/m3) remain within the prescribed limits of 60-80-µg/m3, there is a rising trend. Compared to 1989, SO2 atmospheric concentrations in 1996 have registered a 109% rise, and NOx an 82% rise. Glikson et al (1995) reported in Brisbane, Australia for the period of the study that the maximum 1-hour average for NO2 was 6.7-ppm. Cheng and Lam (2000) reported in Hong Kong the seasonal daily mean SO2 concentrations for the entire period (1983-1992) were 28.2, 38.3, 20.0, and 23.5-µgm-3 for spring, summer, autumn, and winter, respectively; their standard deviations were 30.8, 42.7, 16.9, and 23.6-µgm-3. The corresponding figures of the overall mean NOx concentrations were 129.7, 110.4, 121.0, and 148.6-µgm-3 with standard deviations of 54.2, 49.7, 40.1, and 74.5-µgm-3. Mondal et al (2000) reported the ground level concentration of NOx at 19-important traffic intersection points in Calcutta ranging from 55-222-μgm-3. Lebret et al (2000) reported small area variations in ambient NO2 concentrations in four European areas, i.e., in Amsterdam (Netherlands), Huddersfield (UK), Poznan (Poland), and Prague (Czech Republic) found NO2 concentrations as 24-72, 10-79, 12-48, and 9.6-18.4-μgm-3. In NOx emission Delhi holds second position, and in SO2 Delhi is not in the top 10-districts of India (Garg et al., 2001). India is the second largest contributor of SO2 emissions in Asia, after China. For India, the estimates of national SO2 emissions (Gg/yr.) in the period 1985-97 shows a rising trend as the emissions were 3402.0, 3609.1, 3829.9, 4085.2, 4256.1, 4437.2, 4684.3, 4863.1, 5039.6, 5289.3, 5609.5, 6106.3, and 6276.6-Gg/yr in the years 1985, 1986, 1987, 1988, 1989, 1990, 1991, 1992, 1993, 1994, 1995, 1996, and 1997, respectively (Streets et al., 2000). Our study is in good agreement with all of these studies.

Estimation of Ambient Air Quality in Delhi 55

based on AQI for RSPM at VV was unhealthy on November 16, 17, and 24, 2000, while it was unhealthy for sensitive people on November 15, 22, and 23, 2000. The category of air quality based on AQI for RSPM at SKP was always unhealthy for sensitive people except on one day i.e. November 26, 2000, when it was unhealthy for the general public as well. The category of

It can be concluded that PM is the critical parameter for the deterioration of ambient air

In the industrial area of Jaipur, Gajraj et al (2000) found 24 observations of RSPM, where the air quality was unhealthy for 14 days and unhealthy for sensitive groups for the remaining days. For 32 observations of RSPM in Jaipur, Jain (2001) found that the air quality was good on six days, moderate on 20 days, and unhealthy for sensitive people on six days. She also

The PM2.5 value at Delhi is well above 24-hour PM2.5 USNAAQS (65-µgm-3). The variation of PM2.5 value is highest on SKP, which suggests that the contribution from different sources is varying to a great extent. The PM10 concentration value was always greater than the 24-hour NAAQS (150-μg/m3 in industrial and commercial areas and 100-μg/m3 in residential areas). The study shows that, at Okhla, PM2.5 contribution to PM10 is 47%; it increases to 59% at VV, further increases to 69% at IIT, and achieves its maximum (77%) at SKP. The highest contribution of PM2.5/PM10 strongly shows that industrial sources are predominant in the SKP area. This shows that at Okhla RSP contribution to TSP is 39%; it increases to 50% at VV and decreases to 41% at IIT and again achieves its highest of 53% at SKP. The above discussion put forward that the secondary formation and diesel exhaust, which are prime important mechanisms of the formation of PM2.5, are predominant at IIT but the primary particulate sources also seem like prime important reasons for the formation of RSP. In all of Delhi, 24 hour SO2 concentration varies from 8.45-21.70-μgm-3 and 24-hour NO2 concentration varies from 35.12-104.47-μgm-3. The variation in Delhi shows that sources liberating NOx are also releasing SOx. The SOx levels were always lesser than the 24-hour NAAQS of SOx i.e. 80 μg/m3 in the residential area and 120-μg/m3 in the industrial area. The NOx levels were lesser than the 24-hour NAAQS of NOx i.e. 80-μg/m3 in the residential area and 120-μg/m3 in the industrial area. The SO2 and NOx data do not show any weekend or holiday effect. The variations observed are most likely due to differences in the air mass, emission characteristics, and the meteorology. The category of air quality based on AQI for SOx and NOx at all sites was always good for the public. The category of air quality based on AQI for PM at all sites was always unhealthy for the public. It can be concluded that PM is the critical parameter for

The authors are thankful to Envirotech Instruments Private Limited, Delhi for their sponsorship of the study. Thanks were due to all the persons, organizations, institutions, and companies who provided a secure location for the sampling equipment used in this

air quality based on AQI for SOx and NOx at all sites was always good for the public.

found for 24 observations of SO2 that the air quality was good on all days.

quality in Delhi.

**4. Conclusions** 

**3.4.1 Comparison with other studies** 

the deterioration of ambient air quality in Delhi.

**5. Acknowledgement** 

#### **3.4 Air quality index**

In order to assess the degree of air pollution in Delhi, an index called air quality index (AQI) has been used as recommended by USEPA (EPA-454/R-99-010). More details are given in the literature review section, and the calculation method adopted is given in the methodology section. Calculated AQI has been presented in the table below.


Table 3. Daily AQI along with Respective Parameter Concentrations

The category of air quality based on AQI for RSPM at Okhla was always unhealthy except on one day i.e. November 9, when it was unhealthy for sensitive people. The category of air quality based on AQI for RSPM at IIT was always unhealthy for sensitive people except on one day i.e. November 3, when it was unhealthy for the general public. The category of air quality

In order to assess the degree of air pollution in Delhi, an index called air quality index (AQI) has been used as recommended by USEPA (EPA-454/R-99-010). More details are given in the literature review section, and the calculation method adopted is given in the

**DATE RSP SOx NOx AQI RSP AQI SOx AQI NOx** 

**20/10** 259.06 14.41 77.48 153.03 20.59 70.10 **21/10** 297.51 15.66 71.03 172.04 23.03 67.20 **23/10** 273.07 18.45 84.87 159.96 27.13 73.42 **24/10** 271.74 17.74 50.72 159.27 26.10 58.07 **29/10** 320.31 16.07 62.90 183.32 23.63 63.54 **9-Nov** 204.99 13.54 72.67 125.75 19.91 67.93

**2-Nov** 179.09 8.45 35.12 112.93 12.43 51.05 **3-Nov** 274.65 9.09 42.62 160.75 13.37 54.53 **4-Nov** 189.11 9.57 36.17 117.88 14.07 51.53 **5-Nov** 196.80 8.76 38.94 121.69 12.88 52.77 **6-Nov** 200.86 10.22 50.11 123.72 15.03 57.79 **7-Nov** 185.06 10.48 44.69 115.90 15.41 55.36

**15/11** 245.62 12.81 63.90 145.84 18.84 63.99 **16/11** 258.31 17.68 62.80 152.63 26.00 63.50 **17/11** 322.63 15.26 77.01 184.46 22.44 69.89 **22/11** 210.57 16.34 70.80 128.52 24.03 67.09 **23/11** 209.97 19.12 85.49 128.23 28.12 73.70 **24/11** 317.32 16.24 54.55 181.84 23.88 59.79

**25/11** 174.54 16.08 78.43 110.65 23.65 70.52 **26/11** 321.24 18.48 88.07 183.77 27.18 74.86 **27/11** 238.72 21.70 104.47 142.43 31.91 82.23 **28/11** 197.64 14.52 84.82 122.09 21.35 73.40

The category of air quality based on AQI for RSPM at Okhla was always unhealthy except on one day i.e. November 9, when it was unhealthy for sensitive people. The category of air quality based on AQI for RSPM at IIT was always unhealthy for sensitive people except on one day i.e. November 3, when it was unhealthy for the general public. The category of air quality

Table 3. Daily AQI along with Respective Parameter Concentrations

methodology section. Calculated AQI has been presented in the table below.

**3.4 Air quality index** 

**Okhla** 

**IIT** 

**VV** 

**SKP** 

based on AQI for RSPM at VV was unhealthy on November 16, 17, and 24, 2000, while it was unhealthy for sensitive people on November 15, 22, and 23, 2000. The category of air quality based on AQI for RSPM at SKP was always unhealthy for sensitive people except on one day i.e. November 26, 2000, when it was unhealthy for the general public as well. The category of air quality based on AQI for SOx and NOx at all sites was always good for the public.

It can be concluded that PM is the critical parameter for the deterioration of ambient air quality in Delhi.

#### **3.4.1 Comparison with other studies**

In the industrial area of Jaipur, Gajraj et al (2000) found 24 observations of RSPM, where the air quality was unhealthy for 14 days and unhealthy for sensitive groups for the remaining days. For 32 observations of RSPM in Jaipur, Jain (2001) found that the air quality was good on six days, moderate on 20 days, and unhealthy for sensitive people on six days. She also found for 24 observations of SO2 that the air quality was good on all days.

#### **4. Conclusions**

The PM2.5 value at Delhi is well above 24-hour PM2.5 USNAAQS (65-µgm-3). The variation of PM2.5 value is highest on SKP, which suggests that the contribution from different sources is varying to a great extent. The PM10 concentration value was always greater than the 24-hour NAAQS (150-μg/m3 in industrial and commercial areas and 100-μg/m3 in residential areas). The study shows that, at Okhla, PM2.5 contribution to PM10 is 47%; it increases to 59% at VV, further increases to 69% at IIT, and achieves its maximum (77%) at SKP. The highest contribution of PM2.5/PM10 strongly shows that industrial sources are predominant in the SKP area. This shows that at Okhla RSP contribution to TSP is 39%; it increases to 50% at VV and decreases to 41% at IIT and again achieves its highest of 53% at SKP. The above discussion put forward that the secondary formation and diesel exhaust, which are prime important mechanisms of the formation of PM2.5, are predominant at IIT but the primary particulate sources also seem like prime important reasons for the formation of RSP. In all of Delhi, 24 hour SO2 concentration varies from 8.45-21.70-μgm-3 and 24-hour NO2 concentration varies from 35.12-104.47-μgm-3. The variation in Delhi shows that sources liberating NOx are also releasing SOx. The SOx levels were always lesser than the 24-hour NAAQS of SOx i.e. 80 μg/m3 in the residential area and 120-μg/m3 in the industrial area. The NOx levels were lesser than the 24-hour NAAQS of NOx i.e. 80-μg/m3 in the residential area and 120-μg/m3 in the industrial area. The SO2 and NOx data do not show any weekend or holiday effect. The variations observed are most likely due to differences in the air mass, emission characteristics, and the meteorology. The category of air quality based on AQI for SOx and NOx at all sites was always good for the public. The category of air quality based on AQI for PM at all sites was always unhealthy for the public. It can be concluded that PM is the critical parameter for the deterioration of ambient air quality in Delhi.

#### **5. Acknowledgement**

The authors are thankful to Envirotech Instruments Private Limited, Delhi for their sponsorship of the study. Thanks were due to all the persons, organizations, institutions, and companies who provided a secure location for the sampling equipment used in this

Estimation of Ambient Air Quality in Delhi 57

Mondal, R., Sen, G.K., Chatterjee, M., Sen, B.K., Sen, S. (2000). Short Communication:

Intersection Points In Calcutta. *Atmospheric Environment,* 34, pp. (629-633). Monn, C. (2001). Exposure Assessment of Air Pollutants: A Review on Spatial Heterogeneity

Pekkanen, J., Timonen, K.L., Ruuskanen, J., Reponen, A., Mirme, A. (1997). Effects of

Rao, M.N. & Rao, H.V.N. (1989). *Air Pollution,* Tata–McGraw Hill publishing Company

Sawyer, R.F., Hardely, R.A., Cadle, S.H., Norbeck, J.M., Slott, R., Bravo, H.A. (1998). Mobile

Shandilya, K.K., Khare, M., Gupta, A.B. (2006). Suspended Particulate Matter concentrations

Shandilya, K.K., Khare, M., Gupta, A.B. (2007). Suspended Particulate Matter concentrations

Shandilya, K.K., Gupta, A.B., Khare, M. (2009). Defining Aerosols by physical and chemical characteristics, *Indian Journal of Air Pollution Control,* Vol. IX, No. 1, pp. (107-126). Shandilya, K.K., Kumar, A. (2009). Analysis of Research Studies on Exhaust Emission from

Shandilya, K.K., Kumar, A. (2010a). Qualitative Evaluation of Particulate Matter inside

Shandilya, KK., Kumar, A., (2010b). Morphology of Single Inhalable Particle inside Public Transit Biodiesel Fueled Bus. *Journal of Environmental Sciences,* 22(2), pp. (263–270). Shandilya, K.K., Kumar, A. (2011). Physical Characterization of Fine Particulate Matter

Shandilya, K.K., Khare, M., Gupta, A.B. (2011). Particulate Matter Concentrations in Delhi

Shandilya, K.K., Gupta, A.B., Khare, M. (2012). Formation of Atmospheric Nitrate under

Nitrogen Dioxide and Ozone. *Atmospheric Environment,* 35, pp. (1-32). Peavy, H.S., Rowe, D.R., Tchobanoglous, G. (1985). Environment Engineering, International

with asthmatic symptoms. *Environmental Research*, 74, pp. (24-33).

Perkins, H.C. (1974). *Air Pollution,* McGraw -Hill Kogakusha Ltd., Tokyo.

*monitoring and assessment,* 128(1-3), pp. (431-45).

Publishers, Inc., ISBN: 978-1-60741-492-6, NewYork..

*Hazardous Materials,* Volume 190, Issues 1-3, pp. (508-514).

Ed., McGraw-Hill Book Company.

Limited.

pp. (2161-2181).

*Protection,* 26(8), pp. (705-717).

*Journal*, 3, pp. (13-20).

*Control,* Paper in Press.

*Management,* Paper in Press.

Ground-Level Concentration Of Nitrogen Oxides (NOx) At Some Traffic

and Indoor/Outdoor/Personal Exposure to Suspended Particulate Matter,

ultrafine and fine particles in urban air on peak expiratory flow among children

Sources Critical Review: 1998 NARSTO Assessment, *Atmospheric Environment,* 34,

in Commercial East Delhi and South Delhi, India. *Indian Journal of Environmental* 

in Rural-Industrial Satna and in Urban-Industrial South Delhi. *Environmental* 

the Heavy Duty Diesel Engine Fueled by Biodiesel. In: *Handbook of Environmental Research,* Editors: Aurel Edelstein and Dagmar Bär, pp. (101-151), Nova Science

Public Transit Buses Operated by Biodiesel. *The Open Environmental Engineering* 

inside the Public Transit Buses Fueled by Biodiesel in Toledo, Ohio. *Journal of* 

before Changing to Compressed Natural Gas (CNG). *Indian Journal of Air Pollution* 

High Particulate Matter Concentration. *World Review Of Science, Technology And Sustainable Development, (WRSTSD), Special Issue: Waste And Environment* 

study, as well as those who provided supplemental information. Authors were thankful to Mr. Ronald Zallocco for correcting the English.

#### **6. References**


study, as well as those who provided supplemental information. Authors were thankful to

Borja-Aburto, V., Loomis, D., Bangdiwala, S., Shy, C., Rascon-Pacheco, R. (1997). Ozone,

Burton, R., Suh, H., Koutrakis, P. (1996). Spatial Variation in Particulate Concentrations

Cheng, S., Lam, K.C. (2000). Synoptic Typing And Its Application To The Assessment Of

Crowther, G., Finlay, H., Raj P.A., Wheeler, T. (1990). India: *A travel survival kit,* (IV ed.),

Das, D.B., Gupta, A.B., Bhargava, A., Kushwaha, Pandit, M.K. (1998). Quarrying Induced

Gajraj V., Jain R., Gupta T.P., Gupta, A.B. (2000). A study of air quality of Vishwakarma

Guenther, A., Geron, C., Pierce, T., Lamb, B., Harley, P., Fall, R. (2000). Natural Emissions Of

and its control, Feb. 18-19, 2000, BHUIT, Varanasi, India, pp. (10-19). Garg, A., Shukla, P.R., Bhattacharya, S., Dadhwal, V.K. (2001). Sub-Region (District) And

http://www.envfor.nic.in/divisions/cpoll/delpoll.html, Date of access: July, 28, 2009. Jain, R. (2001). Dispersion Modeling of Gaseous Pollutants in Jaipur City and Assessment of

Kamyotra & Ashish. (2001). Characterization of different methods used for analysis of

Kumar, A., Kadiyala, A., Somuri, D., Shandilya, K.K., Velagapudi, S., Nerella, V.K.V., (2011).

Mitigation Flexibility, *Atmospheric Environment,* 35, pp. (703-713).

Hong Kong, *Atmospheric Environment,* 34, pp. (585-594).

E.P.A, 1999, *Guideline for reporting daily air quality,* EPA-454/R-99-010 July 1999

Lonely Planet Publications, pp. (140-169).

Down To Earth. (2000). *Killing Pollutants,* Vol 9, No.10, pp. (23).

http://www.envfor.nic.in/, Date of access: September, 9, 2010.

College, University of Rajasthan.

Delhi.

660-3.

Suspended Particulates, and Daily Mortality in Mexico City. *American Journal of* 

within Metropolitan Philadelphia. *Environmental Science and Technology,* 30, pp.

Climatic Impact On Concentrations Of Sulfur Dioxide And Nitrogen Oxides In

Particulate Air Pollution In Jaipur City, Western India. *Journal Of Nepal Geological* 

Industrial Area, Jaipur, *Proc. IPC 2000,* National seminar on industrial pollution

Sector Level SO2 And NOx Emissions For India: Assessment Of Inventories And

Non-Methane Volatile Organic Compounds, Carbon Monoxide, And Oxides Of Nitrogen From North America. *Atmospheric Environment,* 34, pp. (2205-2230). Hargreaves, P.R., Leidi, L.A., Grubb, H.J., Howe, M.I., Muggle Stone, M.A. (2000). Local &

Seasonal Variations In Atmospheric Nitrogen Dioxide Levels At Both Amsted, UK & Relations With Metrological Conditions. *Atmospheric Environment,* 843-853.

Their Health Effects on Residents Ph.D. Thesis, Malviya Regional Engineering

sulfur-di-oxide, M. Tech. Thesis, Indraprastha College, Delhi University, New

Characterization of Emissions and Indoor Air Quality of Public Transport Buses Using Alternative Fuels. In: *Biodiesel: Blends, Properties and Applications,* Editors: Marchetti, J.M., Fang, Z., In Press, Nova Science Publishers, Inc., ISBN 978-1-61324-

Mr. Ronald Zallocco for correcting the English.

*Epidemiology,* 145, pp. (250-257).

*Society,* 18, pp. (369-378).

**6. References** 

(392-399).


**4** 

*France* 

**Environmental Parameter** 

Stéphane Le Floch1, Mélanie Fuhrer1,2, Pierre Slangen2 and Laurent Aprin2

*2ISR, LGEI, Ecole des Mines d'Alès, Ales* 

*1CEDRE - Centre de Documentation, de Recherche et* 

**Effects on the Fate of a Chemical Slick** 

*D'Expérimentations sur les Pollutions Accidentelles des Eaux, Brest* 

Recent data highlights the growing trend in the transport of dangerous substances and the consequent evolution in legislation concerning such substances, whether on a European level (evolution of the Standard European Behaviour Classification, or SEBC code) or on a worldwide scale (new MARPOL-Annex II classification which entered into force on 1st January 2007, the OPRC-HNS Protocol, as well as the HNS Convention). This evolution is however based exclusively on data from the literature which all too often cannot be applied to spills, as it does not take into account the influence of factors in the marine environment

Yet, following a spill at sea of a bulk liquid chemical, the response authorities must immediately take measures in order to reduce the risks of exposure of the surrounding population as well as to prevent contamination of the marine ecosystem as a whole and, more widely, to protect all life forms. To succeed in this effort that forms a part of emergency response, it is vital to know which compartment of the ecosystem will be affected (atmosphere, water surface, water column, sediments…) and to monitor the

The accident of the chemical tanker *Ievoli Sun* which sank in the English Channel in 2000 releasing approximately 1,000 tonnes of styrene into the sea (Law et al., 2003) has raised concerns about the fate of this compound at sea. In the aquatic environment, styrene is often reported to be of low risk for chronic effects due to its volatility and low bioaccumulation potential. Hence, the potential for styrene to produce long term adverse environmental impacts appears to be negligible. However, most of these statements are based on tests performed in fresh water environments whilst very few data exist concerning the marine environment. Also, the value of some of these tests can on some occasions be questionable as they are reliant on a static mode of exposure and measurements of the actual exposure concentrations are not performed during the tests. Hence, although styrene's volatility and biodegradability would in most cases significantly reduce the levels in water, the behaviour and bioavailability of styrene in saline water is not very well known. Thereby, a better

on the physico-chemical characteristics of the product spilt.

evolution of concentrations in order to define the environmental risks.

**1. Introduction** 


## **Environmental Parameter Effects on the Fate of a Chemical Slick**

Stéphane Le Floch1, Mélanie Fuhrer1,2, Pierre Slangen2 and Laurent Aprin2 *1CEDRE - Centre de Documentation, de Recherche et D'Expérimentations sur les Pollutions Accidentelles des Eaux, Brest 2ISR, LGEI, Ecole des Mines d'Alès, Ales France* 

#### **1. Introduction**

58 Air Quality – Monitoring and Modeling

Shandilya, K.K., Khare, M. (2012). Particulate Matter: Sources, Emission Rates and Health

Sharma, M., Kiran, Y.N.V.M., Shandilya, K.K. (2003). Investigations into Formation of

Sharma, M., Kiran, Y.N.V.M., Shandilya, K.K. (2004). Subset Statistical Modeling of Atmospheric Sulfate. *Clean Air and Environmental Quality,* 38, No. 4, pp. (37-41). Streets, D.G., Tsai, N.Y., Akimoto, H., Oka, K. (2000). Sulfur Dioxide Emissions In Asia In

Thurston, G.D., Ito, K., Lippmann, M., Hayes, C. (1989). Reexamination of London, England,

Tuch, T., Brand, P., Wichmann, H.E., Heyder, J. (1997). Variation of Particle Number and

WHO. (1979). *Fine Particulate Pollution: A Report of the United Nations Economic Commission for* 

Wilson, W.E., Suh, H.H. (1997). Fine particles and coarse particles: Concentration

The Period 1985-1997, *Atmospheric Environment,* 34, pp. (4413-4424).

In Press, NOVA publishers, ISBN: 978-1-62100-837-8.

*Environmental Health Perspectives,* 79, pp. (73-82).

*Europe (UNECE),* Pargamon Press sec. I, A, 2.

*Management Association*, 47, pp. (1238-1249).

Germany, *Atmospheric Environment,* 31, pp. (4193- 4197).

pp. (2005-2013).

Effects, In: Advances in Environmental Research Volume 23, Editor: Daniels, J.A.,

Atmospheric Sulfate under High PM10 Concentration, *Atmospheric Environment,* 37,

mortality in relation to exposure to acidic aerosols during 1963-1972 winters.

Mass Concentration in Various Size Ranges of Ambient Aerosols in Eastern

relationships relevant to epidemiologic studies. *Journal of the Air & Waste* 

Recent data highlights the growing trend in the transport of dangerous substances and the consequent evolution in legislation concerning such substances, whether on a European level (evolution of the Standard European Behaviour Classification, or SEBC code) or on a worldwide scale (new MARPOL-Annex II classification which entered into force on 1st January 2007, the OPRC-HNS Protocol, as well as the HNS Convention). This evolution is however based exclusively on data from the literature which all too often cannot be applied to spills, as it does not take into account the influence of factors in the marine environment on the physico-chemical characteristics of the product spilt.

Yet, following a spill at sea of a bulk liquid chemical, the response authorities must immediately take measures in order to reduce the risks of exposure of the surrounding population as well as to prevent contamination of the marine ecosystem as a whole and, more widely, to protect all life forms. To succeed in this effort that forms a part of emergency response, it is vital to know which compartment of the ecosystem will be affected (atmosphere, water surface, water column, sediments…) and to monitor the evolution of concentrations in order to define the environmental risks.

The accident of the chemical tanker *Ievoli Sun* which sank in the English Channel in 2000 releasing approximately 1,000 tonnes of styrene into the sea (Law et al., 2003) has raised concerns about the fate of this compound at sea. In the aquatic environment, styrene is often reported to be of low risk for chronic effects due to its volatility and low bioaccumulation potential. Hence, the potential for styrene to produce long term adverse environmental impacts appears to be negligible. However, most of these statements are based on tests performed in fresh water environments whilst very few data exist concerning the marine environment. Also, the value of some of these tests can on some occasions be questionable as they are reliant on a static mode of exposure and measurements of the actual exposure concentrations are not performed during the tests. Hence, although styrene's volatility and biodegradability would in most cases significantly reduce the levels in water, the behaviour and bioavailability of styrene in saline water is not very well known. Thereby, a better

Environmental Parameter Effects on the Fate of a Chemical Slick 61

experimental open cells were set up in Brest Bay, each of them made of a 9 square metre surface surrounded with a 3 metre high nylon skirt set on a metal structure. As the cell sides were flexible, the contained water was subject to the influence of swell and waves. As the cells were not covered, the sea surface was subject to atmospheric influences (wind, sunshine and rain). After releasing the products at the surface, weathering processes were studied by measuring the viscosity of the slick, solubility in the water column and gas cloud formation.

Several international, regional and national authorities, aware of the risk connected with the transport of chemicals by sea, have published operational guides to describe the possible response options. For instance, IMO has published several manuals (IMO, 1987, 1992, 2006) and REMPEC has done the same for the Mediterranean area (REMPEC, 1996, 1999, 2004). We can also mention the Helsinki Baltic Sea Convention manuals (HELCOM, 1991, 2002, 2007) and the North Sea Bonn Agreement manuals (Bonn Agreement, 1985, 1994). At a national level in France, *Cedre* has produced response guides, each of which is specifically

Alongside this information, which ranges from general (i.e. in the case of a guide on response at sea to hazardous substances) to specific (i.e. guides that characterize the behaviour and identify the risks related to a given product), the Standard European Behaviour Classification system or SEBC code determines the short term behaviour of any chemical as long as its physico-chemical characteristics are known. This tool is of the utmost importance for operational staff in charge of response, as it can be used in an emergency to indicate the main components of the ecosystem that will be affected. However, experience has made it clear that this code can be tricky to use as it can be difficult to understand how environmental factors (wind, sunshine, water and air temperature, etc.) will affect the chemicals' categorization. Indeed these parameters have a very significant influence on the short-term fate of chemicals in the marine environment, in particular on their evaporation kinetics (transfer from the water surface to the atmosphere), their dissolution kinetics (transfer from the water surface to the water column) and their transport and transformation

In the case of a spill, response usually follows three generally accepted scenarios: i) response is not possible because the spill occurred in a geographical environment that is incompatible with reasonable response times; ii) response is not possible due to the reactivity of the

If, in the latter scenario, the substance spilled is not particularly reactive, it is important to bear in mind that the slick formed will evolve. Immediately after its accidental release into the marine environment, the substance will tend to move into different components of the ecosystem, i.e. the atmosphere, the sea surface, the water column and the seafloor. These different forms of distribution are gathered under the umbrella term of "fate", for which we generally distinguish short term behaviour (a few hours) and longer term behaviour (up to

**2. Fate of chemicals in the environment after accidental release** 

dedicated to a given chemical.

processes in the aqueous phase.

several years).

**2.1 Fate of substances with low reactivity** 

substances (major, imminent danger); iii) response is possible.

evaluation of the actual toxic effects of this compound in marine aquatic organisms is needed. In the recent example of the *Ievoli Sun* vessel, a relatively large amount of styrene was discharged in a relatively short time. Yet, studies showed that only low levels of styrene contamination were measured in edible tissues of crabs caught in the immediate vicinity of the wreck site (Law et al., 2003). Bioaccumulation does not necessarily represent the real uptake of a contaminant as many aquatic organisms including fish and crustaceans are able to efficiently biotransform the mother compounds into more polar metabolites, thereby considerably reducing the bioaccumulation factor of the mother compound. Even at low concentrations, it is known that some substances can disturb important biochemical pathways which can impair biochemical functioning of cells and lead to deleterious effects in tissue, like cancer development, in the longer term (Alexander, 1997; Vaghef & Hellman, 1998; Moller et al., 2000). Furthermore, the styrene released from the wreck lying on the floor of the Channel, whose density is lower than that of seawater, rose through the water column to form slicks at the sea surface. These slicks drifted under the combined influence of the wind and marine currents, without ever reaching the coast. Under the effect of the wind and sunshine, the styrene evaporated to form an invisible cloud whose fate was then dependent, both in terms of drift and dilution on aerial currents alone. Following this succession of physico-chemical events, the response authorities had to deal with contamination of the atmosphere, in addition to water pollution, by a product recognised as being particularly neurotoxic (Laffon et al., 2002). The French Customs operators conducting aerial surveillance above the spill area were not prepared for this new risk and were exposed to styrene vapours, resulting in disabling headaches. Just as concerning, on the inhabited island of Aurigny, naval firefighters sporadically yet repeatedly detected the presence of styrene in the air. The inability to predict, i.e. to anticipate, arrivals of styrene gas on the coast rendered contingency plans ineffective for the protection of the population. While no consequences on the health of neighbouring populations were reported in this incident, as the concentrations always remained below the ADI for styrene, things could have been quite different with other more toxic substances.

While classifications can be used to predict the theoretical fate of a product in the event of a spill in the environment, they do not take into account the specificities of this environment and should therefore be used cautiously. For instance, wind and sunshine can significantly affect chemicals' evaporation. A similar observation can be made for the salinity and temperature of seawater and the chemicals' hydrosolubility. Furthermore, while the prediction of a substance's fate in the event of a spill has been facilitated by the emergence of forecast software (CLARA, CHEMMAP…), these predictions must nevertheless be validated by field measurements. Forecasting predicts behaviour, i.e. the fate of products in the environment, from the physico-chemical characteristics of products as determined in laboratory conditions, to which mathematical equations are applied which often underestimate certain environmental parameters such as the influence of surface agitation on the spreading of the slick or sunshine on evaporation. This software also has difficulty in predicting the interaction of chemicals with the sea water matrix in the case of a leaking wreck, due to a lack of information on the effects of pressure, temperature and salinity on dissolution processes (Le Floch et al., 2007; 2009).

This work aimed to obtain experimental data on the behaviour of 3 liquid chemicals when released at sea and to characterise them through *in situ* experiments. For this purpose,

evaluation of the actual toxic effects of this compound in marine aquatic organisms is needed. In the recent example of the *Ievoli Sun* vessel, a relatively large amount of styrene was discharged in a relatively short time. Yet, studies showed that only low levels of styrene contamination were measured in edible tissues of crabs caught in the immediate vicinity of the wreck site (Law et al., 2003). Bioaccumulation does not necessarily represent the real uptake of a contaminant as many aquatic organisms including fish and crustaceans are able to efficiently biotransform the mother compounds into more polar metabolites, thereby considerably reducing the bioaccumulation factor of the mother compound. Even at low concentrations, it is known that some substances can disturb important biochemical pathways which can impair biochemical functioning of cells and lead to deleterious effects in tissue, like cancer development, in the longer term (Alexander, 1997; Vaghef & Hellman, 1998; Moller et al., 2000). Furthermore, the styrene released from the wreck lying on the floor of the Channel, whose density is lower than that of seawater, rose through the water column to form slicks at the sea surface. These slicks drifted under the combined influence of the wind and marine currents, without ever reaching the coast. Under the effect of the wind and sunshine, the styrene evaporated to form an invisible cloud whose fate was then dependent, both in terms of drift and dilution on aerial currents alone. Following this succession of physico-chemical events, the response authorities had to deal with contamination of the atmosphere, in addition to water pollution, by a product recognised as being particularly neurotoxic (Laffon et al., 2002). The French Customs operators conducting aerial surveillance above the spill area were not prepared for this new risk and were exposed to styrene vapours, resulting in disabling headaches. Just as concerning, on the inhabited island of Aurigny, naval firefighters sporadically yet repeatedly detected the presence of styrene in the air. The inability to predict, i.e. to anticipate, arrivals of styrene gas on the coast rendered contingency plans ineffective for the protection of the population. While no consequences on the health of neighbouring populations were reported in this incident, as the concentrations always remained below the ADI for styrene, things could

While classifications can be used to predict the theoretical fate of a product in the event of a spill in the environment, they do not take into account the specificities of this environment and should therefore be used cautiously. For instance, wind and sunshine can significantly affect chemicals' evaporation. A similar observation can be made for the salinity and temperature of seawater and the chemicals' hydrosolubility. Furthermore, while the prediction of a substance's fate in the event of a spill has been facilitated by the emergence of forecast software (CLARA, CHEMMAP…), these predictions must nevertheless be validated by field measurements. Forecasting predicts behaviour, i.e. the fate of products in the environment, from the physico-chemical characteristics of products as determined in laboratory conditions, to which mathematical equations are applied which often underestimate certain environmental parameters such as the influence of surface agitation on the spreading of the slick or sunshine on evaporation. This software also has difficulty in predicting the interaction of chemicals with the sea water matrix in the case of a leaking wreck, due to a lack of information on the effects of pressure, temperature and salinity on

This work aimed to obtain experimental data on the behaviour of 3 liquid chemicals when released at sea and to characterise them through *in situ* experiments. For this purpose,

have been quite different with other more toxic substances.

dissolution processes (Le Floch et al., 2007; 2009).

experimental open cells were set up in Brest Bay, each of them made of a 9 square metre surface surrounded with a 3 metre high nylon skirt set on a metal structure. As the cell sides were flexible, the contained water was subject to the influence of swell and waves. As the cells were not covered, the sea surface was subject to atmospheric influences (wind, sunshine and rain). After releasing the products at the surface, weathering processes were studied by measuring the viscosity of the slick, solubility in the water column and gas cloud formation.

### **2. Fate of chemicals in the environment after accidental release**

Several international, regional and national authorities, aware of the risk connected with the transport of chemicals by sea, have published operational guides to describe the possible response options. For instance, IMO has published several manuals (IMO, 1987, 1992, 2006) and REMPEC has done the same for the Mediterranean area (REMPEC, 1996, 1999, 2004). We can also mention the Helsinki Baltic Sea Convention manuals (HELCOM, 1991, 2002, 2007) and the North Sea Bonn Agreement manuals (Bonn Agreement, 1985, 1994). At a national level in France, *Cedre* has produced response guides, each of which is specifically dedicated to a given chemical.

Alongside this information, which ranges from general (i.e. in the case of a guide on response at sea to hazardous substances) to specific (i.e. guides that characterize the behaviour and identify the risks related to a given product), the Standard European Behaviour Classification system or SEBC code determines the short term behaviour of any chemical as long as its physico-chemical characteristics are known. This tool is of the utmost importance for operational staff in charge of response, as it can be used in an emergency to indicate the main components of the ecosystem that will be affected. However, experience has made it clear that this code can be tricky to use as it can be difficult to understand how environmental factors (wind, sunshine, water and air temperature, etc.) will affect the chemicals' categorization. Indeed these parameters have a very significant influence on the short-term fate of chemicals in the marine environment, in particular on their evaporation kinetics (transfer from the water surface to the atmosphere), their dissolution kinetics (transfer from the water surface to the water column) and their transport and transformation processes in the aqueous phase.

#### **2.1 Fate of substances with low reactivity**

In the case of a spill, response usually follows three generally accepted scenarios: i) response is not possible because the spill occurred in a geographical environment that is incompatible with reasonable response times; ii) response is not possible due to the reactivity of the substances (major, imminent danger); iii) response is possible.

If, in the latter scenario, the substance spilled is not particularly reactive, it is important to bear in mind that the slick formed will evolve. Immediately after its accidental release into the marine environment, the substance will tend to move into different components of the ecosystem, i.e. the atmosphere, the sea surface, the water column and the seafloor. These different forms of distribution are gathered under the umbrella term of "fate", for which we generally distinguish short term behaviour (a few hours) and longer term behaviour (up to several years).

Environmental Parameter Effects on the Fate of a Chemical Slick 63

FED Floater that evaporates and dissolves

In terms of response, it is also useful to be aware of longer term behaviour as it provides information on the fate of a pollutant that cannot be removed from the environment through response operations. The residual chemical will spread through the atmosphere and/or the water column where it may become associated with sediment particles in suspension before settling. In water and air, the ultimate fate of chemicals will depend on the processes of dispersion, breakdown and other physical, chemical or biological transformations. These processes can extend over periods ranging from a few months to several years when the

In terms of fate, all these phenomena will contribute, in the best case scenario, to the disappearance of the product and, in the worst case scenario, to its accumulation or even its

Spill response fundamentally requires rapid identification of the compartment(s) of the ecosystem in which the pollutant will accumulate so as to implement the appropriate means to reduce, or even eliminate, its impact. While the SEBC code provides initial elements of response, the operational personnel in charge of response must interpret this classification critically so as to assess whether or not the specificities of the environment in which the incident has occurred will alter the results. This critical interpretation may go as far as altering the categorization of a substance as in the case of the industrial incident at the Jilin

On 13 November 2005, an explosion occurred in a petrochemical plant resulting in a spill of around one hundred tonnes of various substances into the Songhua River. Analyses conducted by the Chinese Research Academy of Environmental Science showed that the water contained benzene and nitrobenzene concentrations 30 to 100 times higher than the accepted standards, of 0.01 mg.L-1 and 0.017 mg.L-1 respectively (Ambrose, 2006; UNEP, 2006). The pollution was carried by the current, and concentrations 34 times higher than standard levels were found in Harbin, 380 km downstream of Jilin, and 10 times higher than standard levels in Jiamusi, 550 km downstream. This decrease can be explained both by the

**Main group Subgroup** 

E Evaporator ED Evaporator that dissolves FE Floater that evaporates F Floater FD Floater that dissolves

D Dissolver DE Dissolver that evaporates S Sinker SD Sinker that dissolves

Table 1. The 12 Behaviour Groups according to the Standard European Behaviour

G Gas GD Gas that dissolves

Classification System

**2.3 Long term fate** 

plant in China.

substance remains in the environment.

transfer into another part of the ecosystem.

**2.4 Influence of environmental factors on fate** 

#### **2.2 Short term fate**

Short term behaviour covers a duration ranging from a few minutes to a few hours, and it must be known in order to implement response measures and techniques. It was in light of this observation that the Standard European Behaviour Classification, or SEBC code, was developed (Cedre, 1988; Bonn Agreement, 1994; GESAMP, 2002).

This code classifies the behaviour of chemical substances according to their state and some of their physical properties, i.e. density, vapour pressure and solubility. The state of the substance refers to whether it is in the form of a gas, a liquid or a solid at 20°C. The density, defined as the mass per unit volume compared to that of seawater (1.03 g.cm-3, at 20°C), determines whether or not a substance will float. The vapour pressure is defined as the partial vapour pressure of a compound in equilibrium with its pure condensed phase (liquid or solid). It is generally accepted that a floating substance will not evaporate if its vapour pressure is lower than 0.3 kPa and it will evaporate rapidly if its vapour pressure is higher than 3 kPa. For dissolved substances, evaporation occurs when the vapour pressure is greater than 10 kPa. Solubility is defined as the maximum abundance of a compound per unit volume in the aqueous phase when the solution is in equilibrium with the pure compound in its initial state. It is usually measured at 20°C, at a pressure of 1 atmosphere, in distilled water (zero salinity). The criteria used are different according to the physical state of the substance. A substance is considered insoluble if its solubility is less than 0.1% for liquids and less than 10% for solids. At the other end of the scale, the phenomenon of dilution will take place if the solubility is greater than 5% for liquids and greater than 100% for solids.

These limits are presented in Figure 1 and can be used to identify 12 types of behaviour, which are summarized in Table 1.


Fig. 1. SEBC Classification of Chemical Products according to their Physical State and Physical Properties


Table 1. The 12 Behaviour Groups according to the Standard European Behaviour Classification System

#### **2.3 Long term fate**

62 Air Quality – Monitoring and Modeling

Short term behaviour covers a duration ranging from a few minutes to a few hours, and it must be known in order to implement response measures and techniques. It was in light of this observation that the Standard European Behaviour Classification, or SEBC code, was

This code classifies the behaviour of chemical substances according to their state and some of their physical properties, i.e. density, vapour pressure and solubility. The state of the substance refers to whether it is in the form of a gas, a liquid or a solid at 20°C. The density, defined as the mass per unit volume compared to that of seawater (1.03 g.cm-3, at 20°C), determines whether or not a substance will float. The vapour pressure is defined as the partial vapour pressure of a compound in equilibrium with its pure condensed phase (liquid or solid). It is generally accepted that a floating substance will not evaporate if its vapour pressure is lower than 0.3 kPa and it will evaporate rapidly if its vapour pressure is higher than 3 kPa. For dissolved substances, evaporation occurs when the vapour pressure is greater than 10 kPa. Solubility is defined as the maximum abundance of a compound per unit volume in the aqueous phase when the solution is in equilibrium with the pure compound in its initial state. It is usually measured at 20°C, at a pressure of 1 atmosphere, in distilled water (zero salinity). The criteria used are different according to the physical state of the substance. A substance is considered insoluble if its solubility is less than 0.1% for liquids and less than 10% for solids. At the other end of the scale, the phenomenon of dilution will take place if the solubility is

These limits are presented in Figure 1 and can be used to identify 12 types of behaviour,

Fig. 1. SEBC Classification of Chemical Products according to their Physical State and

developed (Cedre, 1988; Bonn Agreement, 1994; GESAMP, 2002).

greater than 5% for liquids and greater than 100% for solids.

which are summarized in Table 1.

Physical Properties

**2.2 Short term fate** 

In terms of response, it is also useful to be aware of longer term behaviour as it provides information on the fate of a pollutant that cannot be removed from the environment through response operations. The residual chemical will spread through the atmosphere and/or the water column where it may become associated with sediment particles in suspension before settling. In water and air, the ultimate fate of chemicals will depend on the processes of dispersion, breakdown and other physical, chemical or biological transformations. These processes can extend over periods ranging from a few months to several years when the substance remains in the environment.

In terms of fate, all these phenomena will contribute, in the best case scenario, to the disappearance of the product and, in the worst case scenario, to its accumulation or even its transfer into another part of the ecosystem.

#### **2.4 Influence of environmental factors on fate**

Spill response fundamentally requires rapid identification of the compartment(s) of the ecosystem in which the pollutant will accumulate so as to implement the appropriate means to reduce, or even eliminate, its impact. While the SEBC code provides initial elements of response, the operational personnel in charge of response must interpret this classification critically so as to assess whether or not the specificities of the environment in which the incident has occurred will alter the results. This critical interpretation may go as far as altering the categorization of a substance as in the case of the industrial incident at the Jilin plant in China.

On 13 November 2005, an explosion occurred in a petrochemical plant resulting in a spill of around one hundred tonnes of various substances into the Songhua River. Analyses conducted by the Chinese Research Academy of Environmental Science showed that the water contained benzene and nitrobenzene concentrations 30 to 100 times higher than the accepted standards, of 0.01 mg.L-1 and 0.017 mg.L-1 respectively (Ambrose, 2006; UNEP, 2006). The pollution was carried by the current, and concentrations 34 times higher than standard levels were found in Harbin, 380 km downstream of Jilin, and 10 times higher than standard levels in Jiamusi, 550 km downstream. This decrease can be explained both by the

Environmental Parameter Effects on the Fate of a Chemical Slick 65

Xylene is a volatile aromatic hydrocarbon produced through petrochemistry. It is in liquid form at room temperature and is often a mixture of isomers (ortho, para and meta – often the most dominant) containing variable proportions of ethylbenzene, although always below <15%. In France, xylene is the most commonly used hydrocarbonated solvent, with 34,000 tonnes in 2004, in particular in the paint, varnish, glue and printing ink industries. The main physico-chemical characteristics of this substance indicate that it will, in theory, behave as a floater/evaporator with very low solubility (Vp = 8.9hPa at 20°C, density of 0.88g.cm-3, and solubility between 175 and 200 mg per litre of seawater; Cedre, 2007 and

Methyl methacrylate (MMA) is described as a transparent volatile liquid with a characteristic smell that can be detected at very low concentrations (< 1 ppm in air). It is mainly used in the polymer and copolymer industry (plastic sheets such as Plexiglas®, Perspex® and Lucite…). Its vapour pressure is 3.9 kPa at 20°C, its solubility

Methyl ethyl ketone (MEK) is a sweet-smelling colourless liquid. It is a volatile organic compound, that is readily flammable and is of little harm to the environment. MEK is an eye and respiratory irritant as well as a central nervous system depressant in humans. It is mainly used as a solvent in various coatings including vinyl, nitrocellulose and acrylic coatings. It is also used as an extraction agent in certain oils as well as in the processing of products and food ingredients. Its vapour pressure is 10.5 kPa at 20°C, its solubility 158 g.L-1 and its density 0.805 g.cm-3. It is therefore a highly evaporative and soluble product, with

The test, replicated for each product, required 25 litres to be released at the sea surface of each floating cell. To monitor physical oceanographic conditions that may influence the behavioural processes of products, weather conditions, prevailing light, wind speed and

Slicks were sampled daily, with a funnel equipped with a tap, in order to characterise the emulsification processes. Samples were transferred into 0.5 L amber bottles and brought

Before the release of xylene, the internal standard pentacosane, a non-evaporating and insoluble substance, was introduced. By measuring the pentacosane/xylene ratio throughout the trial, it was possible to monitor the disappearance kinetics of xylene at the

The natural dispersion of the products in the water column was monitored by performing *in situ* fluorescence measurements at 2 positions in each floating cell at 3 depths (0.5, 1 and 1.50 m). In addition, for each *in situ* fluorescence measurement (at a given position and a given depth), a sample of seawater (1 L) was taken. All samples were placed in an amber

back to the laboratory for viscosity measurement and water content determination.

Lyman et al., 1996).

**3.3 Experimental protocol** 

**3.3.1 Slick sampling** 

water surface.

**3.3.2 Water column sampling** 

15 g.L-1 and its density 0.944 g.cm-3 (Cedre, 2008).

particularly low persistence at the water surface (Cedre, 2009).

temperature conditions were recorded, for both air and sea.

processes of dilution and adsorption of the chemicals by organic matter present in the water, either in suspension, or on the banks (Levshina et al., 2009). These unexpectedly high concentrations can be explained by the particularly harsh weather conditions on site at this time of year (negative temperatures [< -10 °C] and river partially frozen) which restricted evaporation, promoting dilution.

With this as a background, *Cedre* developed an experimental system designed to study the fate of chemicals when released into the marine environment in field conditions, i.e. as close as possible to those encountered in the event of a spill.

### **3. Materials and methods**

The aim of this experimental investigation was to characterise the fate of liquid chemicals released at the sea surface.

#### **3.1 Floating cell enclosures**

Six floating cells were constructed by *Cedre* and moored in Brest Bay, France (48°22'32N and 4°29'32W) to study the influence of natural conditions (i.e., prevailing air and sea temperatures, wind, radiant energy) on the weathering of products spilled onto the sea surface (Figure 2). Each cell consisted of a 3 x 3 m rigid aluminium framework surrounded by a nylon skirt which extended approximately 2 m below and 0.6 m above the sea surface. As cells were not covered and were open to the sea at the base, evaporation and dissolution processes were unimpaired but lateral dispersion at the sea surface was restricted. Observations have shown that while wind-induced capillary waves were reduced by the presence of the skirt, the majority of the mixing action of the sea passed into the cell almost unimpeded. The volume of seawater inside the skirt of each floating cell was approximately 18 m3. As the floating cell enclosures were independent of each other, replica experiments could be conducted simultaneously. During this research programme only 3 of the 6 cells were used.

Fig. 2. Floating cell enclosure

#### **3.2 Chemical products tested**

Following an investigation into the chemicals transported most frequently and in the greatest quantities, three products were selected: xylene, methyl methacrylate and methyl ethyl ketone.

processes of dilution and adsorption of the chemicals by organic matter present in the water, either in suspension, or on the banks (Levshina et al., 2009). These unexpectedly high concentrations can be explained by the particularly harsh weather conditions on site at this time of year (negative temperatures [< -10 °C] and river partially frozen) which restricted

With this as a background, *Cedre* developed an experimental system designed to study the fate of chemicals when released into the marine environment in field conditions, i.e. as close

The aim of this experimental investigation was to characterise the fate of liquid chemicals

Six floating cells were constructed by *Cedre* and moored in Brest Bay, France (48°22'32N and 4°29'32W) to study the influence of natural conditions (i.e., prevailing air and sea temperatures, wind, radiant energy) on the weathering of products spilled onto the sea surface (Figure 2). Each cell consisted of a 3 x 3 m rigid aluminium framework surrounded by a nylon skirt which extended approximately 2 m below and 0.6 m above the sea surface. As cells were not covered and were open to the sea at the base, evaporation and dissolution processes were unimpaired but lateral dispersion at the sea surface was restricted. Observations have shown that while wind-induced capillary waves were reduced by the presence of the skirt, the majority of the mixing action of the sea passed into the cell almost unimpeded. The volume of seawater inside the skirt of each floating cell was approximately 18 m3. As the floating cell enclosures were independent of each other, replica experiments could be conducted

Following an investigation into the chemicals transported most frequently and in the greatest quantities, three products were selected: xylene, methyl methacrylate and methyl

simultaneously. During this research programme only 3 of the 6 cells were used.

evaporation, promoting dilution.

**3. Materials and methods** 

released at the sea surface.

**3.1 Floating cell enclosures** 

Fig. 2. Floating cell enclosure

**3.2 Chemical products tested** 

ethyl ketone.

as possible to those encountered in the event of a spill.

Xylene is a volatile aromatic hydrocarbon produced through petrochemistry. It is in liquid form at room temperature and is often a mixture of isomers (ortho, para and meta – often the most dominant) containing variable proportions of ethylbenzene, although always below <15%. In France, xylene is the most commonly used hydrocarbonated solvent, with 34,000 tonnes in 2004, in particular in the paint, varnish, glue and printing ink industries. The main physico-chemical characteristics of this substance indicate that it will, in theory, behave as a floater/evaporator with very low solubility (Vp = 8.9hPa at 20°C, density of 0.88g.cm-3, and solubility between 175 and 200 mg per litre of seawater; Cedre, 2007 and Lyman et al., 1996).

Methyl methacrylate (MMA) is described as a transparent volatile liquid with a characteristic smell that can be detected at very low concentrations (< 1 ppm in air). It is mainly used in the polymer and copolymer industry (plastic sheets such as Plexiglas®, Perspex® and Lucite…). Its vapour pressure is 3.9 kPa at 20°C, its solubility 15 g.L-1 and its density 0.944 g.cm-3 (Cedre, 2008).

Methyl ethyl ketone (MEK) is a sweet-smelling colourless liquid. It is a volatile organic compound, that is readily flammable and is of little harm to the environment. MEK is an eye and respiratory irritant as well as a central nervous system depressant in humans. It is mainly used as a solvent in various coatings including vinyl, nitrocellulose and acrylic coatings. It is also used as an extraction agent in certain oils as well as in the processing of products and food ingredients. Its vapour pressure is 10.5 kPa at 20°C, its solubility 158 g.L-1 and its density 0.805 g.cm-3. It is therefore a highly evaporative and soluble product, with particularly low persistence at the water surface (Cedre, 2009).

#### **3.3 Experimental protocol**

The test, replicated for each product, required 25 litres to be released at the sea surface of each floating cell. To monitor physical oceanographic conditions that may influence the behavioural processes of products, weather conditions, prevailing light, wind speed and temperature conditions were recorded, for both air and sea.

#### **3.3.1 Slick sampling**

Slicks were sampled daily, with a funnel equipped with a tap, in order to characterise the emulsification processes. Samples were transferred into 0.5 L amber bottles and brought back to the laboratory for viscosity measurement and water content determination.

Before the release of xylene, the internal standard pentacosane, a non-evaporating and insoluble substance, was introduced. By measuring the pentacosane/xylene ratio throughout the trial, it was possible to monitor the disappearance kinetics of xylene at the water surface.

#### **3.3.2 Water column sampling**

The natural dispersion of the products in the water column was monitored by performing *in situ* fluorescence measurements at 2 positions in each floating cell at 3 depths (0.5, 1 and 1.50 m). In addition, for each *in situ* fluorescence measurement (at a given position and a given depth), a sample of seawater (1 L) was taken. All samples were placed in an amber

Environmental Parameter Effects on the Fate of a Chemical Slick 67

Figure 3 shows the persistence of the xylene slicks at the water surface. The results are expressed as a percentage of the quantity spilt at T0. After 2 hours, over 50% of the slick had disappeared, and after 5 hours, less than 30% of the initial quantity remained. The quantities lost were either transferred from the surface to the air (evaporation) or from the surface to the water column (natural dispersion). After day 2, no traces of the slick remained visible at

> T0 T+2h T+5h **Time (Hours)**

Fig. 3. Evolution over time of the slicks remaining at the surface (average of results obtained

The emulsification kinetics could only be monitored during the first day of the experiment

the surface. The time during which the slick was present was estimated at 34 hours.

Photo 2. Xylene slick at T+5h

0

using samples taken from cells 1 and 2).

**4.1.4 Monitoring of emulsification** 

(Figure 4).

20

40

60

**Percentage of residual Xylene**

80

100

**4.1.3 Slick persistence** 

bottle and were stored in a refrigerator before analysis, performed the following day. These samples, which were analysed by GC-MS after liquid-liquid extraction, were used to correct data obtained by fluorimetry measurements (Katz, 1987).

#### **3.3.3 Air monitoring**

During the tests, evaporation was monitored by measuring the Volatile Organic Carbon using two Photo Ionization Detectors (PID MiniRAE 2000, RAE systems). A stationary PID was placed in a corner of the floating cell, while the second was placed on a wind vane so as to monitor evaporation independently of the wind direction.

During these trials, an infrared hyperspectral SPIM camera was deployed to assess its capacity to monitor the formation and movement of a gas cloud formed from slicks of MMA.

#### **4. Results and discussion**

#### **4.1 Xylene**

To study the fate of xylene, two 25 litre releases were performed.

#### **4.1.1 Field observations**

Photo 1, taken at T0, illustrates the chemical release. One hour after the release, the slick began to emulsify and, two hours later, the emulsion covered the entire surface of the floating cell. After 5 hours, the slick began to fragment and several small slicks scattered across the entire cell appeared (Photo 2).

#### **4.1.2 Prevailing sea and weather conditions**

The week of the trials was marked by relatively harsh weather conditions, with a slight sea according to the Douglas sea scale (sea 3), a north-easterly wind at force 3 on the Beaufort scale (12 to 19 km.h-1) and low temperatures (between 3 and 5°C). Several snow showers occurred during this week.

Photo 1. Xylene release (T0)

Photo 2. Xylene slick at T+5h

#### **4.1.3 Slick persistence**

66 Air Quality – Monitoring and Modeling

bottle and were stored in a refrigerator before analysis, performed the following day. These samples, which were analysed by GC-MS after liquid-liquid extraction, were used to correct

During the tests, evaporation was monitored by measuring the Volatile Organic Carbon using two Photo Ionization Detectors (PID MiniRAE 2000, RAE systems). A stationary PID was placed in a corner of the floating cell, while the second was placed on a wind vane so as

During these trials, an infrared hyperspectral SPIM camera was deployed to assess its capacity

Photo 1, taken at T0, illustrates the chemical release. One hour after the release, the slick began to emulsify and, two hours later, the emulsion covered the entire surface of the floating cell. After 5 hours, the slick began to fragment and several small slicks scattered

The week of the trials was marked by relatively harsh weather conditions, with a slight sea according to the Douglas sea scale (sea 3), a north-easterly wind at force 3 on the Beaufort scale (12 to 19 km.h-1) and low temperatures (between 3 and 5°C). Several snow showers

to monitor the formation and movement of a gas cloud formed from slicks of MMA.

data obtained by fluorimetry measurements (Katz, 1987).

to monitor evaporation independently of the wind direction.

To study the fate of xylene, two 25 litre releases were performed.

**3.3.3 Air monitoring** 

**4. Results and discussion** 

**4.1.1 Field observations** 

occurred during this week.

Photo 1. Xylene release (T0)

across the entire cell appeared (Photo 2).

**4.1.2 Prevailing sea and weather conditions** 

**4.1 Xylene** 

Figure 3 shows the persistence of the xylene slicks at the water surface. The results are expressed as a percentage of the quantity spilt at T0. After 2 hours, over 50% of the slick had disappeared, and after 5 hours, less than 30% of the initial quantity remained. The quantities lost were either transferred from the surface to the air (evaporation) or from the surface to the water column (natural dispersion). After day 2, no traces of the slick remained visible at the surface. The time during which the slick was present was estimated at 34 hours.

Fig. 3. Evolution over time of the slicks remaining at the surface (average of results obtained using samples taken from cells 1 and 2).

#### **4.1.4 Monitoring of emulsification**

The emulsification kinetics could only be monitored during the first day of the experiment (Figure 4).

Environmental Parameter Effects on the Fate of a Chemical Slick 69

T+1h

T+2h

T+5h

T+7h

T+25h

T+29h

0 2 4 6 810 12 **Concentration (mg.L-1)**

Fig. 6. Evolution over time of xylene concentrations in the water column in cell 2.

Beyond this sample time, traces of xylene were no longer detected in the water column.

expected: 100 times higher than the theoretical solubility found in the literature.

In the two cells, the maximum concentrations were obtained at T+2h (10.8 and 11.7 mg.L-1, respectively in cells 1 and 2), and represented a quantity of around 200 g of dispersed xylenes. These concentrations then gradually decreased over time, and dropped below 1 mg.L-1 at T+29h.

During these *in situ* trials, the natural dispersion of xylenes in the water column (dissolution + emulsification) showed rapid kinetics, and the concentrations measured were higher than

These trials demonstrated that the persistence of a slick of xylenes at the water surface is relatively short-lived (around 2 days in these trial conditions, not very conducive to evaporation) and that this disappearance could be explained by dissolution processes which were higher than expected (in the literature this chemical is identified as having low solubility). In terms of the transfer of the chemical into the atmosphere, evaporation kinetics could not be measured due to the particularly harsh weather conditions (the snow interfered with the photoionisator). However, by comparing it to styrene (this chemical was previously tested and these two products have an equivalent vapour pressure and are classified as FE according to the SEBC code), it can be presumed that xylenes evaporate less quickly than theory predicts: after 5 hours, the entire styrene slick had disappeared, while this took 34 hours with xylenes. These results can be explained by emulsification processes: the xylene slicks rapidly emulsified to a great extent during the first day (90% at T+7h), a process which does not take place with styrene. This raises the question of surface tension which is not taken into account in the SEBC classification; yet it is this factor that defines a product's capacity to emulsify, and emulsification promotes natural dispersion in the water column.

0

0.5

**Depth (m)**

1

1.5

**4.1.6 Fate of xylenes** 

Fig. 4. Emulsification kinetics of xylene slicks within cells 1 and 2.

Xylenes have a high water sorption capacity and rapidly form emulsions. At T+7h the water content of slicks in cells 1 and 2 were respectively 65% and 90%. However, these emulsions are unstable as they are non-persistent: from day 2, it was no longer possible to sample the surface slick due to insufficient quantities, and on day 3, no pollution remained visible.

#### **4.1.5 Monitoring of dispersion in the water column**

Figures 5 and 6 show the xylene concentrations in the water column, respectively in cells 1 and 2, at the three sampled depths (0.5, 1 and 1.5 m) over time. After 29 hours, concentrations of the substance were no longer detected.

Fig. 5. Evolution over time of xylene concentrations in the water column in cell 1.

T0 T+1h T+2h T+5h T+6h T+7h **Time (Hours)**

Floating Cell n°1 Floating Cell n°2

Xylenes have a high water sorption capacity and rapidly form emulsions. At T+7h the water content of slicks in cells 1 and 2 were respectively 65% and 90%. However, these emulsions are unstable as they are non-persistent: from day 2, it was no longer possible to sample the surface slick due to insufficient quantities, and on day 3, no pollution remained visible.

Figures 5 and 6 show the xylene concentrations in the water column, respectively in cells 1 and 2, at the three sampled depths (0.5, 1 and 1.5 m) over time. After 29 hours,

**)**

T+1h

T+2h

T+5h

T+7h

T+25h

T+29h

0 2 4 6 8 10 12 **Concentration (mg.L-1**

Fig. 5. Evolution over time of xylene concentrations in the water column in cell 1.

Fig. 4. Emulsification kinetics of xylene slicks within cells 1 and 2.

**4.1.5 Monitoring of dispersion in the water column** 

concentrations of the substance were no longer detected.

0

0

0.5

**Depth (m)** 

1

1.5

20

40

60

**Percentage of seawater into the slick**

80

100

Fig. 6. Evolution over time of xylene concentrations in the water column in cell 2.

In the two cells, the maximum concentrations were obtained at T+2h (10.8 and 11.7 mg.L-1, respectively in cells 1 and 2), and represented a quantity of around 200 g of dispersed xylenes. These concentrations then gradually decreased over time, and dropped below 1 mg.L-1 at T+29h. Beyond this sample time, traces of xylene were no longer detected in the water column.

During these *in situ* trials, the natural dispersion of xylenes in the water column (dissolution + emulsification) showed rapid kinetics, and the concentrations measured were higher than expected: 100 times higher than the theoretical solubility found in the literature.

#### **4.1.6 Fate of xylenes**

These trials demonstrated that the persistence of a slick of xylenes at the water surface is relatively short-lived (around 2 days in these trial conditions, not very conducive to evaporation) and that this disappearance could be explained by dissolution processes which were higher than expected (in the literature this chemical is identified as having low solubility). In terms of the transfer of the chemical into the atmosphere, evaporation kinetics could not be measured due to the particularly harsh weather conditions (the snow interfered with the photoionisator). However, by comparing it to styrene (this chemical was previously tested and these two products have an equivalent vapour pressure and are classified as FE according to the SEBC code), it can be presumed that xylenes evaporate less quickly than theory predicts: after 5 hours, the entire styrene slick had disappeared, while this took 34 hours with xylenes. These results can be explained by emulsification processes: the xylene slicks rapidly emulsified to a great extent during the first day (90% at T+7h), a process which does not take place with styrene. This raises the question of surface tension which is not taken into account in the SEBC classification; yet it is this factor that defines a product's capacity to emulsify, and emulsification promotes natural dispersion in the water column.

Environmental Parameter Effects on the Fate of a Chemical Slick 71

The MMA evaporated very rapidly: it was released at 11:18 and the PIDs began to detect significant values from 11:25 (1 ppm at 11:24, then 53 ppm at 11:25). A peak in concentration was reached at 11:27 with 130 ppm, the time at which a local maximum was recorded for light intensity (130 mW/cm²), while the wind varied little throughout the duration of measurements (5 to 6 m/s on average with peaks at 7 m/s and minima at 4 m/s). These results, illustrated in Figure 6, show the importance in evaporation processes of sunshine, which had a greater influence than the wind: positive correlation between solar peaks and

Date 25/06 25/06 25/06 Time 11:00-12:00 12:00-13:00 13:00-14:00

Air T (°C) 16.4 16.7 17.1 Wind speed (m/s) 5.5 5.6 5.6 Solar intensity (mW/cm²) 61.5 71.2 83.4

Table 2. Evolution of weather conditions during the release of MMA

evaporation peaks (T+9, T+14, T+35, T+49, T+56, T+74 min).

Concentration en MAM dans l'air

Concentration en MAM dans l'air (en

ppm)

(en ppm)

11:18

11:18

11:25

11:33

11:40

vane) and variations in wind and sun intensity throughout the experiment.

11:48

11:55

Fig. 7. MMA concentrations in the air measured using PIDs (stationary and on the wind

12:03

PID Fixe PID girouette intensité lumineuse

12:10

12:18

12:25

12:33

Radiation solaire (en mW/cm²)

11:25

11:33

11:40

11:48

11:55

12:03

PID Fixe PID girouette moy vit vent

12:10

12:18

12:25

12:33

Vélocité du vent (en m.s-1)

**4.2.3 Evaporation** 

The results obtained *in situ* therefore directly contradict the literature which mainly describes the behaviour of a xylene slick in terms of transfer into the atmosphere. It is clear that the slick will initially emulsify, causing evaporation kinetics to slow down and promoting transfer into the water column by natural dispersion processes (dissolution and emulsification). It is important, however, to remember that the sea and weather conditions on site should be taken into account when considering these results, as, in these trials, the low temperatures curbed evaporation processes and surface agitation promoted emulsification. This fate is in accordance with what was observed during the benzene spill in Jilin (higher dissolution than predicted and low evaporation processes).

In terms of response, this means that all the issues related to pollution by soluble products (contamination of the water column and problem of water intakes, impact on benthic flora and fauna…) must be taken into account even if the product is classified as a floater and evaporator. Nevertheless, the gas cloud must not be overlooked, as it presents a risk for responders both in terms of intoxication and as an explosion hazard. In the case of xylene, vapours are heavier than air and will move around just above the water surface, pushed by the wind. This danger is even greater if the spill occurs in a coastal area, or even in a port or harbour, where the population density is liable to be high.

#### **4.2 Methyl methacrylate**

Methyl methacrylate (or MMA) is classed ED according to the SEBC classification, and its main behaviour is evaporation according to its Henry's law constant.

#### **4.2.1 Observations**

The MMA spread out across the water surface forming a heterogeneous slick: various sized clusters of floating droplets were observed (Photo 3). This phenomenon was only temporary, as in less than two hours all of the substance had disappeared.

Photo 3. Appearance of the MMA slick immediately after the spill

#### **4.2.2 Prevailing sea and weather conditions**

The prevailing weather conditions onsite at the time of the trial are presented in Table 2. The key point is the presence of wind.


Table 2. Evolution of weather conditions during the release of MMA

#### **4.2.3 Evaporation**

70 Air Quality – Monitoring and Modeling

The results obtained *in situ* therefore directly contradict the literature which mainly describes the behaviour of a xylene slick in terms of transfer into the atmosphere. It is clear that the slick will initially emulsify, causing evaporation kinetics to slow down and promoting transfer into the water column by natural dispersion processes (dissolution and emulsification). It is important, however, to remember that the sea and weather conditions on site should be taken into account when considering these results, as, in these trials, the low temperatures curbed evaporation processes and surface agitation promoted emulsification. This fate is in accordance with what was observed during the benzene spill

In terms of response, this means that all the issues related to pollution by soluble products (contamination of the water column and problem of water intakes, impact on benthic flora and fauna…) must be taken into account even if the product is classified as a floater and evaporator. Nevertheless, the gas cloud must not be overlooked, as it presents a risk for responders both in terms of intoxication and as an explosion hazard. In the case of xylene, vapours are heavier than air and will move around just above the water surface, pushed by the wind. This danger is even greater if the spill occurs in a coastal area, or even in a port or

Methyl methacrylate (or MMA) is classed ED according to the SEBC classification, and its

The MMA spread out across the water surface forming a heterogeneous slick: various sized clusters of floating droplets were observed (Photo 3). This phenomenon was only

The prevailing weather conditions onsite at the time of the trial are presented in Table 2. The

in Jilin (higher dissolution than predicted and low evaporation processes).

harbour, where the population density is liable to be high.

main behaviour is evaporation according to its Henry's law constant.

temporary, as in less than two hours all of the substance had disappeared.

Photo 3. Appearance of the MMA slick immediately after the spill

**4.2.2 Prevailing sea and weather conditions** 

key point is the presence of wind.

**4.2 Methyl methacrylate** 

**4.2.1 Observations** 

The MMA evaporated very rapidly: it was released at 11:18 and the PIDs began to detect significant values from 11:25 (1 ppm at 11:24, then 53 ppm at 11:25). A peak in concentration was reached at 11:27 with 130 ppm, the time at which a local maximum was recorded for light intensity (130 mW/cm²), while the wind varied little throughout the duration of measurements (5 to 6 m/s on average with peaks at 7 m/s and minima at 4 m/s). These results, illustrated in Figure 6, show the importance in evaporation processes of sunshine, which had a greater influence than the wind: positive correlation between solar peaks and evaporation peaks (T+9, T+14, T+35, T+49, T+56, T+74 min).

Fig. 7. MMA concentrations in the air measured using PIDs (stationary and on the wind vane) and variations in wind and sun intensity throughout the experiment.

Environmental Parameter Effects on the Fate of a Chemical Slick 73

Immediately after the spill, the MEK spread across the entire water surface within the floating cell to form a homogeneous, colourless slick that was difficult to detect. This observation is consistent with the results obtained in the laboratory on MEK-water interfacial tension which predict complete spreading, or even infinite spreading in the absence of containment. This high degree of spreading promoted the transfer of the product into the atmosphere and the water column, resulting in complete disappearance of the slick in less than 1 hour and the end

The prevailing weather conditions on site at the time of the experiment are presented in Table 3, and can be summarised as alternating sunny spells and threatening clouds due to

Time 10:25-11:00 11:00-11:30 11:30 -12:00

Note that the mean solar intensity on this day was higher than that measured during the

Air T (°C) 15.6 15.7 15.7 Wind speed (m/s) 5.1 5.1 5.2 Solar intensity (mW/cm²) 93.1 106.1 104.6

Date 25/06 25/06 25/06

Photo 5. Infrared image of the MMA gas cloud near the floating cells

of sampling and PID measurements at 12:00 (no signal from PID).

Table 3. Evolution of weather conditions during the MEK release

**4.3.2 Prevailing sea and weather conditions** 

**4.3.1 Observations** 

the constant wind.

MMA trial.

The advantage of using the wind vane to monitor the concentration of the product in the atmosphere can be observed in Figure 7: while the shapes of the two curves (stationary PID and PID on wind vane) match, the concentration peaks are more intense for the PID on the wind vane and, in general, its curve shows slightly higher values, indicating better quantification of the vapours emitted by the slick. This result was obtained despite a relatively constant wind direction throughout the experiment (250 – 270°), therefore suggesting that the gain would be even greater in the case of quantification with a swirling wind.

#### **4.2.4 Dissolution**

Despite its solubility (10 g.L-1), the MMA did not dissolve at all in these experimental conditions: the chemical was not detected at any of the three depths sampled (0.5, 1 and 1.5 m).

#### **4.2.5 Fate of methyl methacrylate**

MMA is listed as ED according to the SEBC code, i.e. it is first and foremost an evaporator, but also dissolves. In the prevailing weather conditions during the trial, this product only evaporated, and it did so with rapid kinetics: in less than 90 minutes, no traces of MMA were visible at the water surface or detected by the PIDs. Given its negligible dissolution, an atmospheric transfer coefficient can be estimated at around 4.1 litres per square metre per hour (4.1 L.m-2.h-1), in these experimental conditions.

In the event of a spill, response should therefore focus on the gas cloud as it will be explosive (MMA can form flammable vapours with air) and toxic (TEEL1 of 400 ppm). It is therefore essential to have operational tools to visualise the formation and movement of a gas cloud as well as a computer forecast model.

Due to its marked evaporation, MMA was selected to test the SPIM camera. Using this camera, three periods were identified:


These results, although partial, are very promising in order to monitor the formation and movement of an invisible gas cloud. This technology, already employed for the detection of gas leaks at industrial facilities, could be put to good use in the field of pollution response.

#### **4.3 Methyl ethyl ketone**

Methyl ethly ketone (MEK), like MMA, is classed ED.

<sup>1</sup> TEEL = TEmporary Exposure Limit

Photo 5. Infrared image of the MMA gas cloud near the floating cells

#### **4.3.1 Observations**

72 Air Quality – Monitoring and Modeling

The advantage of using the wind vane to monitor the concentration of the product in the atmosphere can be observed in Figure 7: while the shapes of the two curves (stationary PID and PID on wind vane) match, the concentration peaks are more intense for the PID on the wind vane and, in general, its curve shows slightly higher values, indicating better quantification of the vapours emitted by the slick. This result was obtained despite a relatively constant wind direction throughout the experiment (250 – 270°), therefore suggesting that the gain would be even greater in the case of quantification with a swirling

Despite its solubility (10 g.L-1), the MMA did not dissolve at all in these experimental conditions: the chemical was not detected at any of the three depths sampled (0.5, 1 and

MMA is listed as ED according to the SEBC code, i.e. it is first and foremost an evaporator, but also dissolves. In the prevailing weather conditions during the trial, this product only evaporated, and it did so with rapid kinetics: in less than 90 minutes, no traces of MMA were visible at the water surface or detected by the PIDs. Given its negligible dissolution, an atmospheric transfer coefficient can be estimated at around 4.1 litres per square metre per

In the event of a spill, response should therefore focus on the gas cloud as it will be explosive (MMA can form flammable vapours with air) and toxic (TEEL1 of 400 ppm). It is therefore essential to have operational tools to visualise the formation and movement of a

Due to its marked evaporation, MMA was selected to test the SPIM camera. Using this




These results, although partial, are very promising in order to monitor the formation and movement of an invisible gas cloud. This technology, already employed for the detection of gas leaks at industrial facilities, could be put to good use in the field of pollution response.

wind.

1.5 m).

**4.2.4 Dissolution** 

**4.2.5 Fate of methyl methacrylate** 

hour (4.1 L.m-2.h-1), in these experimental conditions.

gas cloud as well as a computer forecast model.

camera, three periods were identified:

**4.3 Methyl ethyl ketone** 

1 TEEL = TEmporary Exposure Limit

denser around the outside (Photo)

while remaining detectable by the SPIM

dilution so that it completely disappears.

Methyl ethly ketone (MEK), like MMA, is classed ED.

Immediately after the spill, the MEK spread across the entire water surface within the floating cell to form a homogeneous, colourless slick that was difficult to detect. This observation is consistent with the results obtained in the laboratory on MEK-water interfacial tension which predict complete spreading, or even infinite spreading in the absence of containment. This high degree of spreading promoted the transfer of the product into the atmosphere and the water column, resulting in complete disappearance of the slick in less than 1 hour and the end of sampling and PID measurements at 12:00 (no signal from PID).

#### **4.3.2 Prevailing sea and weather conditions**

The prevailing weather conditions on site at the time of the experiment are presented in Table 3, and can be summarised as alternating sunny spells and threatening clouds due to the constant wind.


Table 3. Evolution of weather conditions during the MEK release

Note that the mean solar intensity on this day was higher than that measured during the MMA trial.

Environmental Parameter Effects on the Fate of a Chemical Slick 75

and to form a homogeneous slick. The coefficient is the same as for MMA, i.e. 4.1 L.m-2.h-1, bearing in mind that for MEK it is not constant throughout evaporation. It is estimated that 80% of the slick evaporated in 15 minutes, giving a coefficient of 8.9 L.m-2.h-1, while the remaining 20% took around 75 minutes, giving a considerably lower transfer coefficient of

In the event of a spill, response should therefore focus on the gas cloud which will form

In the event of a chemical spill at sea, it is crucial to be aware of the behaviour of the substances involved to determine the most appropriate response strategy. While classifications exist and are valuable tools, they are based on laboratory data for which the influence of environmental parameters is not taken into consideration, i.e. the effects of wind and sunshine on evaporation and of surface agitation on dissolution are not taken into account. Moreover, they consider the physico-chemical properties of substances independently of each other, while in the event of a spill, dissolution and evaporation processes will occur simultaneously. Therefore, if evaporation kinetics are rapid, dissolution processes will not have time to occur and a product, even if soluble, could evaporate completely into the atmosphere. With this in mind, *Cedre* developed an experimental tool, floating cells, which can be used to characterise the main processes governing a substance's

This paper presents the results obtained during a series of trials in these floating cells, which aimed to characterise the behaviour of three chemicals: xylene, methyl methacrylate and

These trials provided two vital pieces of information on these chemicals, to add to the information provided by the SEBC classification. The first relates to xylene which is listed as FE. It became apparent that, in certain weather conditions, xylene will indeed behave as a Floater/Evaporator, however if the temperature drops and especially if the surface agitation and wind are sufficient (wind speed greater than 3 m.s-1), it will form an emulsion and will mainly dissolve, thus exhibiting FDE behaviour. The two other products (MMA and MEK) are classed ED while, in these experimental conditions, they

These results highlight the need to study *in situ* the behaviour of chemicals to add to the theoretical classification based only on laboratory data, especially as it is clear that factors such as surface agitation, wind, sunshine and outside temperature significantly affect the transfer of products from the surface to another environmental compartment. The floating cells, developed by *Cedre*, adequately fulfil this purpose, and, what's more, are an innovative tool which, in addition to helping to define a response strategy in the event of a

Concretely, such experiments on the behaviour of HNS releases in sea water provide:

i. further information in addition to the chemical emergency response guides. For example, on the one hand, xylene, which is considered as a Floater/Evaporator, can

spill, can be used routinely as realistic input for forecast software.

around 0.45 L.m-2.h-1.

**5. Conclusion** 

fate after its release at sea.

behaved only as Evaporators.

methyl ethyl ketone.

rapidly and will be flammable and toxic.

#### **4.3.3 Evaporation**

The MEK evaporated very swiftly and to a great extent: less than 5 minutes after the spill, a peak in concentration was measured both by the stationary PID and the PID on the wind vane (350 ppm). Concentrations then gradually decreased and levelled out at around 20 ppm at T+30 min before disappearing completely at T+90 min (Figure 8). The evaporation of MEK was so intense that it was unaffected by the wind or sun; no peaks could be correlated with a gust of wind or spell of sunshine.

Fig. 8. MEK concentrations in the air measured by the PIDs (stationary and wind vane) during the trial.

#### **4.3.4 Dissolution**

Despite its high solubility (≈ 300 g.L-1 i.e. 30 times higher than MMA), MEK did not dissolve at all in the conditions of this experiment: the chemical was not detected at any of the three depths sampled (0.5, 1 and 1.5 m).

#### **4.3.5 Fate of methyl ethyl ketone**

MEK is listed as ED according to the SEBC code, i.e. it is first and foremost an evaporator, but also dissolves. In the weather conditions during this trial, the product only evaporated, and did so with intense, rapid kinetics: in less than 5 minutes, an evaporation peak was observed and thereafter concentrations in the air continually decreased. This behaviour was similar to MMA with, however, even faster evaporation kinetics, unaffected by the wind or sun. This trend can be explained by MEK's capacity to spread widely at the water surface and to form a homogeneous slick. The coefficient is the same as for MMA, i.e. 4.1 L.m-2.h-1, bearing in mind that for MEK it is not constant throughout evaporation. It is estimated that 80% of the slick evaporated in 15 minutes, giving a coefficient of 8.9 L.m-2.h-1, while the remaining 20% took around 75 minutes, giving a considerably lower transfer coefficient of around 0.45 L.m-2.h-1.

In the event of a spill, response should therefore focus on the gas cloud which will form rapidly and will be flammable and toxic.

### **5. Conclusion**

74 Air Quality – Monitoring and Modeling

The MEK evaporated very swiftly and to a great extent: less than 5 minutes after the spill, a peak in concentration was measured both by the stationary PID and the PID on the wind vane (350 ppm). Concentrations then gradually decreased and levelled out at around 20 ppm at T+30 min before disappearing completely at T+90 min (Figure 8). The evaporation of MEK was so intense that it was unaffected by the wind or sun; no peaks could be correlated

**4.3.3 Evaporation** 

with a gust of wind or spell of sunshine.

0

10:19

depths sampled (0.5, 1 and 1.5 m).

**4.3.5 Fate of methyl ethyl ketone** 

10:33

10:48

11:02

Fig. 8. MEK concentrations in the air measured by the PIDs (stationary and wind vane)

Despite its high solubility (≈ 300 g.L-1 i.e. 30 times higher than MMA), MEK did not dissolve at all in the conditions of this experiment: the chemical was not detected at any of the three

MEK is listed as ED according to the SEBC code, i.e. it is first and foremost an evaporator, but also dissolves. In the weather conditions during this trial, the product only evaporated, and did so with intense, rapid kinetics: in less than 5 minutes, an evaporation peak was observed and thereafter concentrations in the air continually decreased. This behaviour was similar to MMA with, however, even faster evaporation kinetics, unaffected by the wind or sun. This trend can be explained by MEK's capacity to spread widely at the water surface

11:16

Heure

PID fixe girouette

11:31

11:45

12:00

100

200

Concentration en MEC dans l'air

during the trial.

**4.3.4 Dissolution** 

(en ppm)

300

400

In the event of a chemical spill at sea, it is crucial to be aware of the behaviour of the substances involved to determine the most appropriate response strategy. While classifications exist and are valuable tools, they are based on laboratory data for which the influence of environmental parameters is not taken into consideration, i.e. the effects of wind and sunshine on evaporation and of surface agitation on dissolution are not taken into account. Moreover, they consider the physico-chemical properties of substances independently of each other, while in the event of a spill, dissolution and evaporation processes will occur simultaneously. Therefore, if evaporation kinetics are rapid, dissolution processes will not have time to occur and a product, even if soluble, could evaporate completely into the atmosphere. With this in mind, *Cedre* developed an experimental tool, floating cells, which can be used to characterise the main processes governing a substance's fate after its release at sea.

This paper presents the results obtained during a series of trials in these floating cells, which aimed to characterise the behaviour of three chemicals: xylene, methyl methacrylate and methyl ethyl ketone.

These trials provided two vital pieces of information on these chemicals, to add to the information provided by the SEBC classification. The first relates to xylene which is listed as FE. It became apparent that, in certain weather conditions, xylene will indeed behave as a Floater/Evaporator, however if the temperature drops and especially if the surface agitation and wind are sufficient (wind speed greater than 3 m.s-1), it will form an emulsion and will mainly dissolve, thus exhibiting FDE behaviour. The two other products (MMA and MEK) are classed ED while, in these experimental conditions, they behaved only as Evaporators.

These results highlight the need to study *in situ* the behaviour of chemicals to add to the theoretical classification based only on laboratory data, especially as it is clear that factors such as surface agitation, wind, sunshine and outside temperature significantly affect the transfer of products from the surface to another environmental compartment. The floating cells, developed by *Cedre*, adequately fulfil this purpose, and, what's more, are an innovative tool which, in addition to helping to define a response strategy in the event of a spill, can be used routinely as realistic input for forecast software.

Concretely, such experiments on the behaviour of HNS releases in sea water provide:

i. further information in addition to the chemical emergency response guides. For example, on the one hand, xylene, which is considered as a Floater/Evaporator, can

Environmental Parameter Effects on the Fate of a Chemical Slick 77

HELCOM. (2002). HELCOM Response Manual, Volume 2, ANNEX 3, A3- Case histories of

HELCOM. (2007). Report on shipping accidents in the Baltic Sea area for the year 2007. http://www.helcom.fi/stc/files/shipping/shipping\_accidents\_2007.pdf IMO. (1987). *Manual on Chemical Pollution*, Section 1: Evaluation and intervention, IMO,

IMO. (1992). Manual on Chemical Pollution, Section 2: Search and Recovery of Merchandise

IMO. (2006). International Convention for the Prevention of Pollution from Ships, 1973, as modified by the Protocol of 1978 relating thereto (MARPOL). London. Katz, E. (1987). Quantitative analysis using chromatographic techniques. *Separation science* 

Laffon, B., Pasaro, E., Mendez, J., 2002. Evaluation of genotoxic effects in a group of workers

Law, R.J., Kelly, C., Matthiessen, P., Aldridge, J., 2003. The loss of the chemical

Le Floch S., Dumont J., Jaffrennou C., Olier R. (2007). Accidental spills: the C.E.C., a new

Le Floch S., Benbouzid H., Olier R. (2009). Operational Device and Procedure to Test the

Levshina S.L., N.N. Efimov and V.N. Bazarkin. "Assessment of the Amur River Ecosystem

Lyman, W.J., Reehl, W.F., Rosenblatt, D.H. (1996). Handbook of chemical property

Moller, P., Knudsen, L.E., Loft, S., Wallin, H., 2000. The comet assay as a rapid test in

REMPEC. (1996). Compendium of notes on preparedness and response to maritime pollution emergencies involving hazardous substances. REMPEC, Malta

UNEP. (2006). The Songhua River spill. China. in *Field Mission Report.* UNEP, 26p.,

confounding factors. Cancer Epidem. Biomar. 9, 1005-1015.

REMPEC. (1999). Practical Guide for Marine Chemical Spills. REMPEC, Malta

tanker Ievoli Sun in the English Channel, October 2000. Mar. Pollut. Bull. 46,

device to test initial dissolution rate of chemicals in seawater column. In *7th International Symposium on Maritime Safety, Security and Environmental Protection*, 20-

Initial Dissolution Rate of Chemicals after Ship Accidents: The Cedre Experimental Column. In *The Open Environmental Pollution & Toxicology Journal.*

Pollution with Benzene and its Derivatives Caused by an Accident at the Chemical Plant in Jilin City, China". *Bull Environ Contam Toxicol*, 83: 776-779,

estimation method: Environmental behavior of organic compounds. *American* 

biomonitoring occupational exposure to DNA-damaging agents and effect of

*series*. John Wiley & Sons, Chichester, New York, Brisbane, 427p.

exposed to low levels of styrene. Toxicology. 171, 175-186.

involving chemicals

London, p 82.

254-257.

Vol. 1, 1-10.

2009.

December

marine chemical accidents

and Packages Lost at Sea, IMO, London, p 41.

21 September 2007, Athènes, Grèce.

*Chemical Society*, 4ème Ed., Whashington.

REMPEC. (2004). Oil Accidents Report. REMPEC, Malta

Baltic Sea Area, 1974 (Helsinki Convention), vol. III Response to incidents

behave as a FDE; in the case of emergency response in the same weather conditions as the trials, the attention of the intervention team must also be focused on the water column. On the other hand for both other substances (MMA and MEK), the atmospheric compartment has to be especially considered by the intervention team which has, firstly, to protect itself against the gas cloud.


#### **6. Acknowledgments**

This study was financially supported by the French Navy which is in charge of the response at sea following an incident and the French Ministère de l'Ecologie, du Développement durable, des Transports et du Logement (MEEDDM).

#### **7. References**


ii. experimental data for the implementation of existing software. Forecast software is a useful tool but few programs are validated by field experiments. Here, results are used to validate the Clara software which is fully dedicated to predicting the behaviour of

iii. the Authorities in charge of the response with operational information. They require

This study was financially supported by the French Navy which is in charge of the response at sea following an incident and the French Ministère de l'Ecologie, du Développement

Alexander, M., 1997. Environmental fate and effects of styrene. Critical Reviews in Environ.

Ambrose, P. (2006). China Involved in Massive Chemical Spill". In *Marine Pollution Bulletin,*

Bonn Agreement. (1985). *Bonn Agreement: Counter-Pollution Manual*, Bonn Agreement,

Bonn Agreement. (1994). European classification system. Bonn Agreement: *Counter-Pollution* 

Cedre. (1988). Système européen de classification pour les produits chimiques déversés en

Cedre. (2007). Xylènes. *Guide d'intervention chimique.* ISBN 978-2-87893-085-6. ISSN 1950-

Cedre. (2008). Méthacrylate de méthyle stabilisé. *Guide d'intervention chimique.* ISBN 978-2-

Cedre. (2009). Méthyléthylcétone. *Guide d'intervention chimique.* ISBN 978-2-87893-095-5.

Gesamp. (2002). (IMO/FAO/UNESCO-IOC/WMO/WHO/IAEA/UN/UNEP Joint Group

HELCOM. (1991). Manual on Co-operation in Combating Marine Pollution within the

of Experts on the Scientific Aspects of Marine Environmental Protection). *Revised GESAMP Hazard Evaluation Procedure for Chemical Substances Carried by Ships.* Rep.

framework of the Convention on the Protection of the Marine Environment of the

*pollution*. BAWG OTOPSA 18/13/3-F, 3-6 mai 1988, 8p

mer". *18E réunion du groupe de travail de l'Accord de Bonn sur les questions opérationnelles, techniques et scientifiques concernant les activités de lutte contre la* 

which has, firstly, to protect itself against the gas cloud.

durable, des Transports et du Logement (MEEDDM).

87893-091-7. ISSN 1950-0556. p 72.

Stud. Gesamp No. 64, IMO, London, 126pp.

ISSN 1950-0556. p 70.

Sci. Technol. 27, 383-410.

accurate data in order to organize the most appropriate response.

HNS at sea.

**6. Acknowledgments** 

52, news.

London

0556. p 69.

*Manual.* London.

**7. References** 

behave as a FDE; in the case of emergency response in the same weather conditions as the trials, the attention of the intervention team must also be focused on the water column. On the other hand for both other substances (MMA and MEK), the atmospheric compartment has to be especially considered by the intervention team Baltic Sea Area, 1974 (Helsinki Convention), vol. III Response to incidents involving chemicals


REMPEC. (1999). Practical Guide for Marine Chemical Spills. REMPEC, Malta


**5** 

*Poland* 

**Indoor Air Quality and Thermal Comfort in Naturally Ventilated** 

**Low-Energy Residential Houses** 

*Faculty of Civil and Environmental Engineering, Gdańsk,* 

The indoor environmental quality (IEQ) and occupant comfort are closely related. The current indoor environmental assessment includes four aspects, namely thermal comfort (TC), indoor air quality (IAQ), visual comfort (VC) and aural comfort (AC). IAQ, as the nature of air in an indoor environment with relation to the occupant health and comfort is not an easily defined concept. In a broad context, it is the result of complex interactions between building, building systems and people. Comparative risk studies performed by the United States Environmental Protection Agency (USEPA) ranked IAQ as one of the 5 top environmental risks to the public health. Over the past decades, exposure to indoor air pollutants increased due to a variety of factors including: construction of tightly sealed buildings, reduction of ventilation rates (for energy saving) and use of synthetic building materials and furnishings as well as chemically formulated personal care products, pesticides and household cleaners. The effect of chemical pollutants on the perceived IAQ was investigated in several studies. The volatile organic compounds (VOCs) were suspected to cause ''sick-building'' symptoms, like headache, eye and mucous membrane irritation, fatigue and asthmatic symptoms (Redlich et al. 1997). The WHO air quality guidelines exist for major ambient air pollutants such as nitrogen dioxide and ozone, a few organic pollutants including mainly chlorinated and aromatic hydrocarbons (World Health Organization, 2000). The International Agency of Cancer Research recently upgraded formaldehyde to the group 1, known human carcinogen (IARC, 2004). However, there are still inadequate data about health effects of other VOCs. The total amount of VOCs and TVOC was not proven to correlate with symptoms. Investigations of all types of indoor air pollutants for the general air quality monitoring and assessment are complicated. In many studies, it was suggested that the measurement and analysis of the indoor carbon dioxide (CO2) concentration could be useful for understanding the Indoor Air Quality (IAQ) and ventilation effectiveness. Healthy people can tolerate the CO2 level up to 10,000 ppm without serious health effects. An acceptable indoor CO2 level should be kept below 1000 ppm or 650 ppm above the ambient level in order to prevent any accumulation of associated human body odors. The indoor carbon dioxide is relatively easy to measure and its low level

in the indoor air usually corresponds to a low level of VOCs.

**1. Introduction** 

M. Krzaczek and J. Tejchman *Gdańsk University of Technology,* 

Vaghef, H., Hellman, B., 1998. Detection of styrene and styrene oxide-induced DNA damage in various organs of mice using the comet assay. Pharmacol. Toxicol. 83, 69-74.

### **Indoor Air Quality and Thermal Comfort in Naturally Ventilated Low-Energy Residential Houses**

M. Krzaczek and J. Tejchman *Gdańsk University of Technology, Faculty of Civil and Environmental Engineering, Gdańsk, Poland* 

#### **1. Introduction**

78 Air Quality – Monitoring and Modeling

Vaghef, H., Hellman, B., 1998. Detection of styrene and styrene oxide-induced DNA

69-74.

damage in various organs of mice using the comet assay. Pharmacol. Toxicol. 83,

The indoor environmental quality (IEQ) and occupant comfort are closely related. The current indoor environmental assessment includes four aspects, namely thermal comfort (TC), indoor air quality (IAQ), visual comfort (VC) and aural comfort (AC). IAQ, as the nature of air in an indoor environment with relation to the occupant health and comfort is not an easily defined concept. In a broad context, it is the result of complex interactions between building, building systems and people. Comparative risk studies performed by the United States Environmental Protection Agency (USEPA) ranked IAQ as one of the 5 top environmental risks to the public health. Over the past decades, exposure to indoor air pollutants increased due to a variety of factors including: construction of tightly sealed buildings, reduction of ventilation rates (for energy saving) and use of synthetic building materials and furnishings as well as chemically formulated personal care products, pesticides and household cleaners. The effect of chemical pollutants on the perceived IAQ was investigated in several studies. The volatile organic compounds (VOCs) were suspected to cause ''sick-building'' symptoms, like headache, eye and mucous membrane irritation, fatigue and asthmatic symptoms (Redlich et al. 1997). The WHO air quality guidelines exist for major ambient air pollutants such as nitrogen dioxide and ozone, a few organic pollutants including mainly chlorinated and aromatic hydrocarbons (World Health Organization, 2000). The International Agency of Cancer Research recently upgraded formaldehyde to the group 1, known human carcinogen (IARC, 2004). However, there are still inadequate data about health effects of other VOCs. The total amount of VOCs and TVOC was not proven to correlate with symptoms. Investigations of all types of indoor air pollutants for the general air quality monitoring and assessment are complicated. In many studies, it was suggested that the measurement and analysis of the indoor carbon dioxide (CO2) concentration could be useful for understanding the Indoor Air Quality (IAQ) and ventilation effectiveness. Healthy people can tolerate the CO2 level up to 10,000 ppm without serious health effects. An acceptable indoor CO2 level should be kept below 1000 ppm or 650 ppm above the ambient level in order to prevent any accumulation of associated human body odors. The indoor carbon dioxide is relatively easy to measure and its low level in the indoor air usually corresponds to a low level of VOCs.

Indoor Air Quality and Thermal Comfort in Naturally Ventilated Low-Energy Residential Houses 81

c. buildings with a low heat consumption consume heat by 1/3 less than conventional

d. buildings equipped with ventilation systems do not reach a theoretically calculated

e. consumption of the electric energy by ventilators and water pumps increases the

f. when installing a ventilation system, the demand for the ventilation heat per year can

Low-energy houses that use very small amounts of energy for heating purposes do not need any conventional heating system. If windows have low transmission losses, the risk of drafts is minimized and the heating appliances (i.e. radiators) are redundant (Karlsson & Moshfegh 2006). When a house can be designed to require less than 10 W/m2 of the heating capacity to maintain 20°C by −10 °C ambient conditions, a conventional heating system (i.e. a gas fired furnace, circulation pipes and radiators) can be omitted and the total energy consumption drops to a small fraction of normal levels. The result is a drastic reduction in both operating costs and environmental impacts. Hastings (2004) studied five building projects as example of successful solutions. He reported that the estimated space heating demand varied from 12.6 kWh/(m2 a) up to 22.4 kWh/(m2 a), while the thickness of the insulation layers varied between 30-50 cm. In apartment projects, the ventilation system typically ran 24 h/day and 365 days/year. He concluded that high performance housing depended on keeping the heat in. To achieve the required *U*-values, the walls had to be too thick. The superior glazing (*U*=0*.*5 W*/*m2 K) together with highly insulating frame constructions are already on the market today, but they are still too expensive and are one of several reasons for a lager cost of the high performance housing. The ventilation is ambiguously related to the energy saving rationale originating from the global warming problem and non-renewable energy sources saving. Since it makes up for about half of the energy consumption in well-insulated buildings, it is an

However, simply reducing ventilation rates has unwanted repercussions on the indoor air quality. Advances in several disciplines of the knowledge such as the growing understanding of the global warming (IPCC, 2007) and its effects on our environment, the increasing evidence of the limited nature of our major energy supply and large costs, both economical and human, of the air pollution related illnesses are dramatically altering the goals of innovations in the building technology. The focus is shifted towards low-energy, 'green' or sustainable buildings, seeking concepts that allow to maintain or even further increase the comfort level that we are accustomed to, while significantly to reduce the associated energy used in every aspect of the human life. Consequently, this field represents a massive gross energy saving potential. Simply reducing ventilation rates, however, deteriorate the indoor air quality and therefore cause unwanted effects such as an increase in the incidence of the respiratory illness and loss of productivity. The purpose of each ventilation system is to provide an acceptable micro-climate (in this context, the microclimate refers to the thermal environment and air quality). The process to create an acceptable interior micro-climate can be conducted basically by two sides: ventilation to achieve a good air quality and heating or cooling inside air to have a thermal comfort. It is relatively easy to control mechanical ventilation systems to maintain the best air exchange rate (Wong |& Huang 2004, Ho et al. 2011, Paul et al. 2010, Laverge et al. 2011). However, the mechanical ventilation and air-conditioning of buildings are responsible for a lot of non-

consumption of the electric energy in buildings up to 50%,

be reduced by 5%, and in the case of heat exchangers even by 20%.

houses,

level of the heat consumption,

attractive target for energy saving measures.

The chapter focuses on the concentration of CO2 in the indoor air. The term comfort is not commonly used in relation to the indoor air quality and it is mainly linked with the lack of the discomfort due to odor and sensory irritations. The acceptable air quality is defined as "air in which there are not known contaminants at harmful concentrations as determined by cognizant authorities and with which, a substantial majority (80% or more) of people do not express a dissatisfaction". Consequently most standards including requirements for the indoor air quality define conditions by providing the minimum percentage of persons who are dissatisfied with the air quality. They are mainly based on the discomfort and annoyance caused for visitors in indoor spaces. Recently, some standards also deal with requirements for occupants.

A problem of the thermal and indoor air quality rises in low-energy buildings. The lowenergy buildings are tight and impermeable. A relative contribution of the heat consumption due to the ventilation in the total heat consumption in low-energy buildings is very high. Thus, the ventilation system significantly affects the thermal comfort and indoor air quality in low-energy buildings. However, the mechanical ventilation and airconditioning of buildings are responsible for a lot of the non-renewable fossil-based energy consumption in the world. Therefore, the natural ventilation can be used as a cheap airexchange system for buildings. The natural ventilation effectiveness depends on many factors such as: size and location of openings (Hummelgaard et al. 2007, Stavrakakis et al. 2010), location of inlet gaps, ambient climate conditions, etc. During the natural ventilation, the air exchange is caused by temperature gradients, but can be enforced by fans located in the exhaust duct inlets (outlets of the indoor zone). The air quality and thermal comfort in buildings equipped with the natural ventilation depend strongly on air flow patterns which are influenced by a type of the space heating system.

In the chapter, the indoor air quality and thermal comfort in the naturally ventilated lowenergy buildings are discussed. The investigations are limited to residential houses equipped with the natural ventilation system with and without exhaust ventilators only. Two locations of the fresh air inlets are considered: above and below the window. The CFD simulations were carried out for two heating systems: a radiator heating system and a floor heating system. The results of numerical investigations are compared with the corresponding experimental measurements.

#### **2. Low-energy buildings**

The term 'low-energy building' is not well defined. The definition is different in several countries and in national and international standards and regulations. However, modern residential buildings can be characterized by the following features with respect to ventilation systems (Meyer 1993, Reinmuth 1994):


The chapter focuses on the concentration of CO2 in the indoor air. The term comfort is not commonly used in relation to the indoor air quality and it is mainly linked with the lack of the discomfort due to odor and sensory irritations. The acceptable air quality is defined as "air in which there are not known contaminants at harmful concentrations as determined by cognizant authorities and with which, a substantial majority (80% or more) of people do not express a dissatisfaction". Consequently most standards including requirements for the indoor air quality define conditions by providing the minimum percentage of persons who are dissatisfied with the air quality. They are mainly based on the discomfort and annoyance caused for visitors in indoor spaces. Recently, some standards also deal with requirements

A problem of the thermal and indoor air quality rises in low-energy buildings. The lowenergy buildings are tight and impermeable. A relative contribution of the heat consumption due to the ventilation in the total heat consumption in low-energy buildings is very high. Thus, the ventilation system significantly affects the thermal comfort and indoor air quality in low-energy buildings. However, the mechanical ventilation and airconditioning of buildings are responsible for a lot of the non-renewable fossil-based energy consumption in the world. Therefore, the natural ventilation can be used as a cheap airexchange system for buildings. The natural ventilation effectiveness depends on many factors such as: size and location of openings (Hummelgaard et al. 2007, Stavrakakis et al. 2010), location of inlet gaps, ambient climate conditions, etc. During the natural ventilation, the air exchange is caused by temperature gradients, but can be enforced by fans located in the exhaust duct inlets (outlets of the indoor zone). The air quality and thermal comfort in buildings equipped with the natural ventilation depend strongly on air flow patterns which

In the chapter, the indoor air quality and thermal comfort in the naturally ventilated lowenergy buildings are discussed. The investigations are limited to residential houses equipped with the natural ventilation system with and without exhaust ventilators only. Two locations of the fresh air inlets are considered: above and below the window. The CFD simulations were carried out for two heating systems: a radiator heating system and a floor heating system. The results of numerical investigations are compared with the

The term 'low-energy building' is not well defined. The definition is different in several countries and in national and international standards and regulations. However, modern residential buildings can be characterized by the following features with respect to

a. relative contribution of the heat consumption due to ventilation in the total heat consumption can be even higher than 50% (thus, the influence of the mechanical

b. buildings are tight and impermeable; however, they must be ventilated due to the minimum air exchange rate related to hygienic requirements in order to reach the appropriate air quality in rooms and good comfort perception of residents; thus, the

type of ventilation system has a great meaning for a comfort perception,

are influenced by a type of the space heating system.

corresponding experimental measurements.

ventilation systems (Meyer 1993, Reinmuth 1994):

ventilation on the heat consumption is very high),

**2. Low-energy buildings** 

for occupants.


Low-energy houses that use very small amounts of energy for heating purposes do not need any conventional heating system. If windows have low transmission losses, the risk of drafts is minimized and the heating appliances (i.e. radiators) are redundant (Karlsson & Moshfegh 2006). When a house can be designed to require less than 10 W/m2 of the heating capacity to maintain 20°C by −10 °C ambient conditions, a conventional heating system (i.e. a gas fired furnace, circulation pipes and radiators) can be omitted and the total energy consumption drops to a small fraction of normal levels. The result is a drastic reduction in both operating costs and environmental impacts. Hastings (2004) studied five building projects as example of successful solutions. He reported that the estimated space heating demand varied from 12.6 kWh/(m2 a) up to 22.4 kWh/(m2 a), while the thickness of the insulation layers varied between 30-50 cm. In apartment projects, the ventilation system typically ran 24 h/day and 365 days/year. He concluded that high performance housing depended on keeping the heat in. To achieve the required *U*-values, the walls had to be too thick. The superior glazing (*U*=0*.*5 W*/*m2 K) together with highly insulating frame constructions are already on the market today, but they are still too expensive and are one of several reasons for a lager cost of the high performance housing. The ventilation is ambiguously related to the energy saving rationale originating from the global warming problem and non-renewable energy sources saving. Since it makes up for about half of the energy consumption in well-insulated buildings, it is an attractive target for energy saving measures.

However, simply reducing ventilation rates has unwanted repercussions on the indoor air quality. Advances in several disciplines of the knowledge such as the growing understanding of the global warming (IPCC, 2007) and its effects on our environment, the increasing evidence of the limited nature of our major energy supply and large costs, both economical and human, of the air pollution related illnesses are dramatically altering the goals of innovations in the building technology. The focus is shifted towards low-energy, 'green' or sustainable buildings, seeking concepts that allow to maintain or even further increase the comfort level that we are accustomed to, while significantly to reduce the associated energy used in every aspect of the human life. Consequently, this field represents a massive gross energy saving potential. Simply reducing ventilation rates, however, deteriorate the indoor air quality and therefore cause unwanted effects such as an increase in the incidence of the respiratory illness and loss of productivity. The purpose of each ventilation system is to provide an acceptable micro-climate (in this context, the microclimate refers to the thermal environment and air quality). The process to create an acceptable interior micro-climate can be conducted basically by two sides: ventilation to achieve a good air quality and heating or cooling inside air to have a thermal comfort. It is relatively easy to control mechanical ventilation systems to maintain the best air exchange rate (Wong |& Huang 2004, Ho et al. 2011, Paul et al. 2010, Laverge et al. 2011). However, the mechanical ventilation and air-conditioning of buildings are responsible for a lot of non-

Indoor Air Quality and Thermal Comfort in Naturally Ventilated Low-Energy Residential Houses 83

Fig. 2. Natural ventilation without mechanical exhaust ventilators (Maier et al. 2009)

Fig. 3. Natural ventilation with mechanical exhaust ventilators (Maier et al. 2009)

renewable fossil-based energy consumptions in the world. Therefore, the natural ventilation can be used as a cheap air-exchange system for buildings.

The comfort and health factors are the major elements for designing proper indoor environments. However, they often act opposite to each other (Behne 1999). This opposing behavior of design elements is often not taken into account. In contrast, modern designers make an imaginative use of glass and space to create well-lit and attractive interiors. These buildings are perfectly isolated from outdoors, and often create unusual conditions leading to a violation of traditional ''rules of thumb'' for the thermal comfort and hygiene (Hastings 2004). Poorly designed naturally ventilated buildings are uncomfortable to live and work in and lead to a reduced quality of the life and loss of productivity. The natural ventilation is an energy-efficient alternative for reducing the energy use in buildings and, with a proper design, it is able to create the thermal comfort and healthy indoor conditions. Typically, the energy cost of naturally ventilated building is 40% less than that of air-conditioned buildings (Energy Consumption Guide 1993)

#### **3. Reference measurements**

#### **3.1 Reference buildings**

To investigate the indoor air quality and thermal comfort in the naturally ventilated lowenergy buildings, the colony of identical residential houses located in Germany by Leipzig was chosen for experiments to obtain reliable experimental results (Maier et al. 2009) and to compare them to the results of our CFD simulations. From the total number of 149 houses, 22 houses standing in one row were chosen for measurements (Fig.1). They were situated in the central part of the entire complex. The residential area of the medium size house was 114 m2, where 102 m2 was the heating area (the ratio of the area to the volume was *A/V*=0.54). Each measured object had two external walls exposed to the air influence. The residential rooms with a large window area were oriented towards the south. The transmission heat coefficient *U* was 0.35 W/(m2 K) (ground floor), 0.22 W/(m2 K) (external walls), 1.30 W/(m2 K) (windows) and 0.22 W/(m2 K) (roof). The theoretical heat demand per year in the houses was 36.8 kWh/(m2 a). For further investigations, only two residential houses were considered:


Fig. 1. The reference house for experiments (Maier et al. 2009)

renewable fossil-based energy consumptions in the world. Therefore, the natural ventilation

The comfort and health factors are the major elements for designing proper indoor environments. However, they often act opposite to each other (Behne 1999). This opposing behavior of design elements is often not taken into account. In contrast, modern designers make an imaginative use of glass and space to create well-lit and attractive interiors. These buildings are perfectly isolated from outdoors, and often create unusual conditions leading to a violation of traditional ''rules of thumb'' for the thermal comfort and hygiene (Hastings 2004). Poorly designed naturally ventilated buildings are uncomfortable to live and work in and lead to a reduced quality of the life and loss of productivity. The natural ventilation is an energy-efficient alternative for reducing the energy use in buildings and, with a proper design, it is able to create the thermal comfort and healthy indoor conditions. Typically, the energy cost of naturally ventilated building is 40% less than that of air-conditioned

To investigate the indoor air quality and thermal comfort in the naturally ventilated lowenergy buildings, the colony of identical residential houses located in Germany by Leipzig was chosen for experiments to obtain reliable experimental results (Maier et al. 2009) and to compare them to the results of our CFD simulations. From the total number of 149 houses, 22 houses standing in one row were chosen for measurements (Fig.1). They were situated in the central part of the entire complex. The residential area of the medium size house was 114 m2, where 102 m2 was the heating area (the ratio of the area to the volume was *A/V*=0.54). Each measured object had two external walls exposed to the air influence. The residential rooms with a large window area were oriented towards the south. The transmission heat coefficient *U* was 0.35 W/(m2 K) (ground floor), 0.22 W/(m2 K) (external walls), 1.30 W/(m2 K) (windows) and 0.22 W/(m2 K) (roof). The theoretical heat demand per year in the houses was 36.8

kWh/(m2 a). For further investigations, only two residential houses were considered:

b. equipped with natural ventilation enforced by single exhaust ventilators (Fig.3).

a. equipped with natural gravitational ventilation (Fig.2),

Fig. 1. The reference house for experiments (Maier et al. 2009)

can be used as a cheap air-exchange system for buildings.

buildings (Energy Consumption Guide 1993)

**3. Reference measurements** 

**3.1 Reference buildings** 

Fig. 2. Natural ventilation without mechanical exhaust ventilators (Maier et al. 2009)

Fig. 3. Natural ventilation with mechanical exhaust ventilators (Maier et al. 2009)

Indoor Air Quality and Thermal Comfort in Naturally Ventilated Low-Energy Residential Houses 85

Fig. 4. Location of measuring sensors in residential houses (Maier et al. 2009)

Fig. 5. The relationship between number of residents and air quality (Maier et al. 2009)

In the case of the natural ventilation with exhaust ventilators, the area of the air supply included children rooms, living rooms and bedrooms. However, a bathroom, toilet and rooms with the internal building installation were associated to the area of the air outlet. The single air ventilators were used, and each of them operated separately. The output of the stream in the mode of the basic ventilation was around 0.32 Wh/m3. The value of 0.25 Wh/m3 for the ventilation output friendly for users was unfortunately exceeded. The following parameters were monitored in houses by sensors: air humidity, air quality (expressed by the emission of both CO2 and malodors), temperature, electricity consumption, gas consumption, heat consumption, working time of the window ventilation (windows totally and slightly open), working time of the mechanical ventilation and number of residents. The low-temperature radiation was not measured due to high costs of transducers. In addition, sun radiation, wind power, wind direction, air humidity, air quality and the air temperature were registered outside houses. Figure 4 shows the location of measuring sensors in residential houses.

#### **3.2 Weather conditions**

The weather conditions during both heating periods were similar (Maier et al. 2009). The time interval between consecutive measurements was 1 hour. According to the occurrence frequency concept (Stavrakakis et al. 2010), the prevailing climatic data were determined as recorded-range weighted averages for the time period between 8:00 and 18:00, i.e. the weights used were the occurrence frequencies of the weather data ranges. Consequently, the mean values of temperature, total solar radiation and wind speed were 4.0 °C, 20.0 W/m2 and 1.6 m/s respectively. Referring to the associated wind direction, south-west (SW) wind directions were dominant, as they corresponded to the occurrence frequency of approximately 39%.

#### **3.3 Experimental results**

The experimental results covered a long time period of 2 years. The weather conditions during both heating periods were similar. Thus, the relationships were established only between the mean values for the sake of readability (maximum and minimum values were not shown). To establish relationships between experimental results, a linear regression method was used. All ventilation systems showed the lower heat consumption than the theoretically calculated one. The average year heat consumption for all systems (22 houses) was 3063 kWh: for the gravitational ventilation 3168 kWh and natural ventilation with exhaust ventilators 3586 kWh. Any relation between the ventilation time and the parameters of the indoor climate was not found. The average concentration of CO2 for each system was: for the gravitational ventilation 1266 ppm and for the natural ventilation with exhaust ventilators 647 ppm. The house with a gravitational ventilation system had the worst air quality, but it had also the highest number of residents (5 persons) what was the second reason for a poor air quality. However, the air quality in all houses was in a permissible comfort range (<1500 ppm) (ISO 2008). In the case of the relative air humidity and air quality (Fig.5), the number of residents realistically depicted a distribution of their values.

A strong relationship between the concentration of CO2 per resident and opening time of windows is visible in Fig.6 (including 22 houses with different ventilation systems).

In the case of the natural ventilation with exhaust ventilators, the area of the air supply included children rooms, living rooms and bedrooms. However, a bathroom, toilet and rooms with the internal building installation were associated to the area of the air outlet. The single air ventilators were used, and each of them operated separately. The output of the stream in the mode of the basic ventilation was around 0.32 Wh/m3. The value of 0.25 Wh/m3 for the ventilation output friendly for users was unfortunately exceeded. The following parameters were monitored in houses by sensors: air humidity, air quality (expressed by the emission of both CO2 and malodors), temperature, electricity consumption, gas consumption, heat consumption, working time of the window ventilation (windows totally and slightly open), working time of the mechanical ventilation and number of residents. The low-temperature radiation was not measured due to high costs of transducers. In addition, sun radiation, wind power, wind direction, air humidity, air quality and the air temperature were registered outside houses. Figure 4 shows the location

The weather conditions during both heating periods were similar (Maier et al. 2009). The time interval between consecutive measurements was 1 hour. According to the occurrence frequency concept (Stavrakakis et al. 2010), the prevailing climatic data were determined as recorded-range weighted averages for the time period between 8:00 and 18:00, i.e. the weights used were the occurrence frequencies of the weather data ranges. Consequently, the mean values of temperature, total solar radiation and wind speed were 4.0 °C, 20.0 W/m2 and 1.6 m/s respectively. Referring to the associated wind direction, south-west (SW) wind directions were dominant, as they corresponded to the occurrence frequency of

The experimental results covered a long time period of 2 years. The weather conditions during both heating periods were similar. Thus, the relationships were established only between the mean values for the sake of readability (maximum and minimum values were not shown). To establish relationships between experimental results, a linear regression method was used. All ventilation systems showed the lower heat consumption than the theoretically calculated one. The average year heat consumption for all systems (22 houses) was 3063 kWh: for the gravitational ventilation 3168 kWh and natural ventilation with exhaust ventilators 3586 kWh. Any relation between the ventilation time and the parameters of the indoor climate was not found. The average concentration of CO2 for each system was: for the gravitational ventilation 1266 ppm and for the natural ventilation with exhaust ventilators 647 ppm. The house with a gravitational ventilation system had the worst air quality, but it had also the highest number of residents (5 persons) what was the second reason for a poor air quality. However, the air quality in all houses was in a permissible comfort range (<1500 ppm) (ISO 2008). In the case of the relative air humidity and air quality (Fig.5), the number of residents realistically depicted a distribution of their values.

A strong relationship between the concentration of CO2 per resident and opening time of

windows is visible in Fig.6 (including 22 houses with different ventilation systems).

of measuring sensors in residential houses.

**3.2 Weather conditions** 

approximately 39%.

**3.3 Experimental results** 

Fig. 4. Location of measuring sensors in residential houses (Maier et al. 2009)

Fig. 5. The relationship between number of residents and air quality (Maier et al. 2009)

Indoor Air Quality and Thermal Comfort in Naturally Ventilated Low-Energy Residential Houses 87

Fig. 7. Acceptance level of separate ventilation systems in residential houses (on the basis of

To investigate the fluid flow patterns and to track CO2 concentration in the indoor air for two inlet gap configurations and two natural ventilation systems, the reference house was considered as one indoor zone which represented the dining and living room (Figs.2 and 3). The indoor zone was modeled together with exhaust ducts and created one fluid flow domain (Fig.8). The exhaust duct was added to the model to make it more realistic, especially in the case of the natural ventilation without fans, induced by the mixture temperature differences.

the questionnaire) (Maier et al. 2009)

**4. Physical model of ventilated zone** 

Fig. 8. Geometry of ventilated zone and fluid domain

Fig. 6. Results of experimental measurements: time of windows opening and CO2 concentration/residents number (Maier et al. 2009)

#### **3.4 Analysis of questionnaire results**

After the measurements program was finished, a detailed questionnaire program was performed (Maier et al. 2009). The results obtained from questionnaires performed in 21 houses equipped with different ventilation systems were compared (Fig.7). In the case of the natural ventilation with exhaust ventilators, the new designed in-take ventilator elements did not disappoint expectations. No problems took place in reference to the air draught. The operational elements were estimated positively. Considering the air quality in rooms and installation regulation, the residents saw the necessity for introducing several improvements (however, from the technical point of view it was impossible). Any negative opinion of the comfort perception or ventilation system was not found. The air quality and comfort perception obtained the notes "good" and "very good" for all ventilation systems and no evident differences of the level of the residents' satisfaction with the ventilation system were detected. In spite of "good" and "very good" notes in questionnaires, in all cases occupants adapted the thermal environment to improve their comfort perception by opening operable windows. It was an unexpected observation because notes in questionnaires were high and the measured air quality was in the comfort range.

#### **4. Physical model of ventilated zone**

86 Air Quality – Monitoring and Modeling

Fig. 6. Results of experimental measurements: time of windows opening and CO2

After the measurements program was finished, a detailed questionnaire program was performed (Maier et al. 2009). The results obtained from questionnaires performed in 21 houses equipped with different ventilation systems were compared (Fig.7). In the case of the natural ventilation with exhaust ventilators, the new designed in-take ventilator elements did not disappoint expectations. No problems took place in reference to the air draught. The operational elements were estimated positively. Considering the air quality in rooms and installation regulation, the residents saw the necessity for introducing several improvements (however, from the technical point of view it was impossible). Any negative opinion of the comfort perception or ventilation system was not found. The air quality and comfort perception obtained the notes "good" and "very good" for all ventilation systems and no evident differences of the level of the residents' satisfaction with the ventilation system were detected. In spite of "good" and "very good" notes in questionnaires, in all cases occupants adapted the thermal environment to improve their comfort perception by opening operable windows. It was an unexpected observation because notes in questionnaires were high and

concentration/residents number (Maier et al. 2009)

the measured air quality was in the comfort range.

**3.4 Analysis of questionnaire results** 

To investigate the fluid flow patterns and to track CO2 concentration in the indoor air for two inlet gap configurations and two natural ventilation systems, the reference house was considered as one indoor zone which represented the dining and living room (Figs.2 and 3). The indoor zone was modeled together with exhaust ducts and created one fluid flow domain (Fig.8). The exhaust duct was added to the model to make it more realistic, especially in the case of the natural ventilation without fans, induced by the mixture temperature differences.

Fig. 8. Geometry of ventilated zone and fluid domain

$$\frac{\partial p}{\partial t} + \frac{\partial (\rho v\_{\chi})}{\partial \chi} + \frac{\partial \{\rho v\_{\chi}\}}{\partial \chi} = 0 \tag{1}$$

$$T\_{\chi} = \frac{\partial}{\partial \chi} \left( \mu \frac{\partial v\_{\chi}}{\partial \chi} \right) + \frac{\partial}{\partial \chi} \left( \mu \frac{\partial v\_{\chi}}{\partial \chi} \right) \tag{2}$$

$$T\_{\mathcal{Y}} = \frac{\partial}{\partial \boldsymbol{\chi}} \left( \mu \frac{\partial \boldsymbol{v}\_{\mathcal{X}}}{\partial \boldsymbol{\chi}} \right) + \frac{\partial}{\partial \boldsymbol{\chi}} \left( \mu \frac{\partial \boldsymbol{v}\_{\mathcal{Y}}}{\partial \boldsymbol{\chi}} \right). \tag{3}$$

$$\begin{split} \frac{\partial}{\partial t} \{\rho c\_p T\_0\} + \frac{\partial}{\partial \mathbf{x}} \{\rho v\_x c\_p T\_0\} + \frac{\partial}{\partial \mathbf{y}} \{\rho v\_y c\_p T\_0\} \\ = \frac{\partial}{\partial \mathbf{x}} \{\lambda \frac{\partial T\_0}{\partial \mathbf{x}}\} + \frac{\partial}{\partial \mathbf{y}} \{\lambda \frac{\partial T\_0}{\partial \mathbf{y}}\} + W^V + E^k + Q\_V + \Phi + \frac{\partial P}{\partial t'} \end{split} \tag{4}$$

$$\frac{\partial(\rho Y\_{alr})}{\partial t} + \nabla \cdot (\rho Y\_{alr}\nu) - \nabla \cdot (\rho D\_{ma\,air} \nabla Y\_{alr}) = 0 \tag{5}$$

$$Y\_{CO2} = 1 - Y\_{air} \tag{6}$$

$$\rho = \frac{P}{RT \sum\_{1}^{2} \frac{Y\_{l}}{M\_{l}}} \tag{7}$$

$$
\upsilon\_{\chi} = \upsilon\_{\text{inlet}\prime} \quad \upsilon\_{\chi} = 0 \tag{8}
$$

$$\frac{\partial v}{\partial n} = 0,\tag{9}$$

$$\frac{\partial v}{\partial n} = 0\tag{10}$$

$$T = T\_{\text{intlet}} \tag{11}$$

$$\frac{\partial T}{\partial n} = 0\tag{12}$$

$$\boldsymbol{\upsilon}\_{inlet} = \boldsymbol{\upsilon}\_{ref} \left(\frac{\mathbf{z}}{h}\right)^{a} \tag{13}$$

$$T\_{e-sol} = \ ^\circ T\_e + \frac{al}{h\_e} \tag{14}$$

$$h\_{\mathbf{e}}(t) = \max\left[5, \frac{8.6 \,\,\nu(t)^{0.6}}{l^{0.4}}\right] \tag{15}$$

$$
\upsilon\_{\chi} = \upsilon\_{fan\prime} \quad \upsilon\_{\chi} = 0 \tag{16}
$$

Indoor Air Quality and Thermal Comfort in Naturally Ventilated Low-Energy Residential Houses 93

where *he* is the convective heat transfer coefficient (ISO 2007) on the internal surface of the external wall or window. For the above assumptions, the convective heat transfer coefficient *he eqv*=0.222 W/m2 K and 1.370 W/ m2 K, for the external wall and window, respectively. Opposite to surfaces in a contact with the external wall and window, other surfaces of the fluid domain are considered to be adiabatic. To simplify an analysis of flow patterns in the indoor zone, it is assumed that the only heat source in the indoor zone is a resident. The human body is concentrated in one heat source which is distributed over the edges of the rectangle (0.3 m × 1.6 m). Figure 12 presents the location of the heat source in the fluid domain. Murakami el at. (2000) studied a problem of airflow, thermal radiation and moisture transport for predicting a heat release from a human body. They reported that a heat release from a human body (in a standing position) to the surrounding environment by convection was 29.14 W/m2 and the mean value of the convective heat transfer coefficient was 4.95 W/m2 K. Following the above conclusions and considering the skin surface area of 1.5696 m2 (Murakami et al. 2000), the

�� ����

(21)

�� = 0 (22)

of 0.14×10-3 m3/s. According to

ℎ����� <sup>=</sup> �������� ������������ �

constant heat flux of 13.26 W/m2 along the human body's rectangle edges is assumed.

��

resident exhales CO2 into the indoor zone with a rate ��

**6. Numerical input data** 

In the indoor zone model, the main CO2 sources are residents. To track the CO2 concentration, two residents occupying the indoor zone are assumed. The respiration process of human beings was investigated by several researchers (Lawrence & Braun 2007, Hyun & Kleinstreuer 2001). They concluded that the frequency of respiration under light physical work is 17 times per minute with the time-mean rate of 8.4 l/min. Following their conclusions, a steady inhalation process is assumed. Consequently, it is chosen that a

research results of Yanes et al., the average value of the CO2 concentration in exhaled air by human beings is 55.100 ppm. Hence, the CO2 source is modeled as the source generating the mixture of 55,100 ppm CO2 concentration with the velocity �� = ���� and �� = 0 m/s.

The Ansys FLOTRAN software package (2007) is used to solve Eqs.1-5. In numerical simulations, the two fluid domain configurations are considered: the inlet gap located above the window (Fig.13a) and the inlet gap located below the window (Fig.13b). A fine mesh generation requires the mesh to be well suited to the boundary conditions, heat and CO2 sources, etc. Thus, the fluid domain is simulated as a multi-area region and a variable mesh density is applied (Fig.14). To compute the IAQ indices, the average breathing zone is

modeled according to the international guideline EN ISO (2007) (Fig.15).

In a composite gas analysis, the species concentration depends strongly on boundary and initial conditions. The boundary conditions are defined by the CO2 concentration over the fluid domain boundaries and the initial conditions are defined by the initial CO2 distribution over the fluid domain. According to the research results of Li et al. (2010), it is assumed that the ambient CO2 concentration is approximately 700 mg/m3 (389 ppm). Consequently, the CO2 concentration at the inlet gap *Cinl* =389 ppm. The same assumption is taken for initial conditions over the fluid domain. For any other boundaries of the fluid domain, one took:

and for the temperature are �� �� = 0. The '*x'* component of the velocity at the inlet to the exhaust duct is computed according to the assumption of the air exchange rate in the ventilated zone of 0.5 1/h. Hence, ���� is 0.23 m/s. The boundary conditions for the velocity components on the surfaces in a contact with internal walls, floor, ceiling and the exhaust duct are

$$
\boldsymbol{\upsilon}\_{\boldsymbol{\chi}} = \boldsymbol{0} \quad \text{and} \quad \boldsymbol{\upsilon}\_{\boldsymbol{\chi}} = \boldsymbol{0} \tag{17}
$$

On the boundary with the external wall and window, the equivalent convection is assumed (Fig.11):

$$q = h\_{e\
eq\nu} (T\_{Fl} - T\_e) \tag{18}$$

where *TFi* is the temperature of the internal surface of the wall or window [°C], *Te* is the ambient air temperature [°C] and *he eqv* is the equivalent convective heat transfer coefficient [W/(m K)]. The convection heat flux *q* is characterized by the equivalent convective heat transfer coefficient *he eqv*. For steady-state heat transfer problems, the convective heat transfer coefficient *he eqv* is independent of ambient and indoor climate conditions. The conduction heat flux *qcond* in a steady-state heat transfer for the external wall and window is:

$$q\_{cond} = U(T\_l - T\_\text{e}) \tag{19}$$

and the equivalent convection heat flux *qe eqv* is

$$q\_{e\
eq\upsilon} = h\_{e\
eq\upsilon}(T\_{Fl} - T\_e),\tag{20}$$

where *U* is the coefficient of the heat transmission [W/(m2 K)], *Te* is the ambient air temperature [°C] and *Ti* is the indoor AIR-CO2 mixture temperature [°C].

Fig. 11. The equivalent convection heat flux at boundary with external wall and window

According to the assumption ����� equals to �� ���, the convective heat transfer coefficient *he eqv* is

$$h\_{\text{e }eq\nu} = \frac{\upsilon(\tau\_{l} - \tau\_{e})}{\left(\tau\_{l} - \upsilon(\tau\_{l} - \tau\_{e})\frac{1}{h\_{\text{e}}} - \tau\_{e}\right)}\tag{21}$$

where *he* is the convective heat transfer coefficient (ISO 2007) on the internal surface of the external wall or window. For the above assumptions, the convective heat transfer coefficient *he eqv*=0.222 W/m2 K and 1.370 W/ m2 K, for the external wall and window, respectively. Opposite to surfaces in a contact with the external wall and window, other surfaces of the fluid domain are considered to be adiabatic. To simplify an analysis of flow patterns in the indoor zone, it is assumed that the only heat source in the indoor zone is a resident. The human body is concentrated in one heat source which is distributed over the edges of the rectangle (0.3 m × 1.6 m). Figure 12 presents the location of the heat source in the fluid domain. Murakami el at. (2000) studied a problem of airflow, thermal radiation and moisture transport for predicting a heat release from a human body. They reported that a heat release from a human body (in a standing position) to the surrounding environment by convection was 29.14 W/m2 and the mean value of the convective heat transfer coefficient was 4.95 W/m2 K. Following the above conclusions and considering the skin surface area of 1.5696 m2 (Murakami et al. 2000), the constant heat flux of 13.26 W/m2 along the human body's rectangle edges is assumed.

In a composite gas analysis, the species concentration depends strongly on boundary and initial conditions. The boundary conditions are defined by the CO2 concentration over the fluid domain boundaries and the initial conditions are defined by the initial CO2 distribution over the fluid domain. According to the research results of Li et al. (2010), it is assumed that the ambient CO2 concentration is approximately 700 mg/m3 (389 ppm). Consequently, the CO2 concentration at the inlet gap *Cinl* =389 ppm. The same assumption is taken for initial conditions over the fluid domain. For any other boundaries of the fluid domain, one took:

$$\frac{\partial \mathcal{C}}{\partial n} = 0 \tag{22}$$

In the indoor zone model, the main CO2 sources are residents. To track the CO2 concentration, two residents occupying the indoor zone are assumed. The respiration process of human beings was investigated by several researchers (Lawrence & Braun 2007, Hyun & Kleinstreuer 2001). They concluded that the frequency of respiration under light physical work is 17 times per minute with the time-mean rate of 8.4 l/min. Following their conclusions, a steady inhalation process is assumed. Consequently, it is chosen that a resident exhales CO2 into the indoor zone with a rate �� of 0.14×10-3 m3/s. According to research results of Yanes et al., the average value of the CO2 concentration in exhaled air by human beings is 55.100 ppm. Hence, the CO2 source is modeled as the source generating the mixture of 55,100 ppm CO2 concentration with the velocity �� = ���� and �� = 0 m/s.

#### **6. Numerical input data**

92 Air Quality – Monitoring and Modeling

exhaust duct is computed according to the assumption of the air exchange rate in the ventilated zone of 0.5 1/h. Hence, ���� is 0.23 m/s. The boundary conditions for the velocity components on the surfaces in a contact with internal walls, floor, ceiling and the exhaust

On the boundary with the external wall and window, the equivalent convection is assumed

where *TFi* is the temperature of the internal surface of the wall or window [°C], *Te* is the ambient air temperature [°C] and *he eqv* is the equivalent convective heat transfer coefficient [W/(m K)]. The convection heat flux *q* is characterized by the equivalent convective heat transfer coefficient *he eqv*. For steady-state heat transfer problems, the convective heat transfer coefficient *he eqv* is independent of ambient and indoor climate conditions. The conduction

where *U* is the coefficient of the heat transmission [W/(m2 K)], *Te* is the ambient air

Fig. 11. The equivalent convection heat flux at boundary with external wall and window

According to the assumption ����� equals to �� ���, the convective heat transfer coefficient *he* 

heat flux *qcond* in a steady-state heat transfer for the external wall and window is:

temperature [°C] and *Ti* is the indoor AIR-CO2 mixture temperature [°C].

and the equivalent convection heat flux *qe eqv* is

�� = 0. The '*x'* component of the velocity at the inlet to the

�� = 0 and �� = 0 (17)

�=ℎ� ������� � ��� (18)

����� = ���� � ��� (19)

�� ��� = ℎ� ������� � ���, (20)

and for the temperature are ��

duct are

(Fig.11):

*eqv* is

The Ansys FLOTRAN software package (2007) is used to solve Eqs.1-5. In numerical simulations, the two fluid domain configurations are considered: the inlet gap located above the window (Fig.13a) and the inlet gap located below the window (Fig.13b). A fine mesh generation requires the mesh to be well suited to the boundary conditions, heat and CO2 sources, etc. Thus, the fluid domain is simulated as a multi-area region and a variable mesh density is applied (Fig.14). To compute the IAQ indices, the average breathing zone is modeled according to the international guideline EN ISO (2007) (Fig.15).

$$\text{PMV}\_{l} = [0.303e^{(-0.036\text{M})} + 0.028]TL \tag{23}$$

$$\text{PMV}\_{\text{avg}} = \frac{\sum\_{k=1}^{n} \text{PMV}\_{k} \cdot A\_{k}}{\sum\_{k=1}^{n} A\_{k}} \tag{24}$$

$$PD = 100 - 94 \cdot \exp\{-1.387 + 0.118 \cdot T\_f - 0.0025 \cdot T\_f^2\},\tag{25}$$

$$\varepsilon\_t = \frac{T\_{outlet} - T\_{inlet}}{T\_{average} - T\_{inlet}} \tag{26}$$

$$T\_{average} = \frac{\int\_{A\_x} T \, dA}{A\_z} \tag{27}$$

$$T\_{outlet} = \frac{\int\_{S\_r} T \, d\mathcal{S}}{S\_r} \tag{28}$$

$$T\_{inlet} = \frac{\int\_{\mathcal{S}\_s} T \, d\mathcal{S}}{\mathcal{S}\_{\mathcal{S}}} \tag{29}$$

$$
\eta\_V = \frac{\mathcal{C}\_r - \mathcal{C}\_s}{\mathcal{C}\_z - \mathcal{C}\_s} \tag{30}
$$

$$\mathcal{C}\_{\mathbf{z}} = \frac{\int\_{A\_{\mathbf{z}}} \mathcal{C} \, dA}{A\_{\mathbf{z}}},\tag{31}$$

$$\mathcal{C}\_r = \frac{\int\_{S\_r} \mathcal{C} \, dS}{S\_r} \tag{32}$$

$$\mathcal{L}\_s = \frac{\int\_{S\_s} \mathcal{C} \, dS}{S\_s} \tag{33}$$

Indoor Air Quality and Thermal Comfort in Naturally Ventilated Low-Energy Residential Houses 99

Case 1 1015 0,81 0,1917 6,0 1,0139 Case 2 1007 1,22 0,3300 6,1 0,8265 Case 3 729 -1,88 0,0500 10,6 2,3943 Case 4 1243 0,92 0,3244 6,4 0,9408 Case 5 991 1,07 0,3455 6,0 0,8278 Case 6 1776 -0,44 0,2206 8,0 1,6427 Case 7 1016 0,77 0,1914 13,8 0,9668 Case 8 1008 1,15 0,3344 13,7 0,9068 Case 9 1266 -3.00 0,1676 13,5 1,7275 Case 10 1243 0,93 0,3244 13,9 0,9783 Case 11 991 1,18 0,3455 13,7 0,9058 Case 12 1776 -1,01 0,1846 13,5 1,6528 Case 13 743 0,77 0,2571 6,6 0,9581 Case 14 688 1,24 0,0010 7,0 0,8974 Case 15 645 -1,34 0,4805 8,5 1,9367 Case 16 904 1,11 0,2738 6,9 0,9330 Case 17 688 1,05 0,0010 7,0 0,9004 Case 18 967 -1,49 0,3443 7,7 1,6136 Case 19 743 0,52 0,2571 13,8 0,9666 Case 20 688 1,07 0,0010 13,9 0,8913 Case 21 957 -0,53 0,2165 13,7 1,5133 Case 22 904 -2,78 0,2738 13,9 0,9483 Case 23 688 1,06 0,0010 13,8 0,8913 Case 24 962 -1,30 0,3473 13,5 1,5942

Figure 17a shows the distributions of the air velocity for the air inlet located below the window (case 4). The indoor space is heated by radiators without exhaust ventilators. Unlikely the air inlet located above the window, in the case of windy weather, the downward flow is still dominated by the forced convection from the imposed inlet velocity. However, main flow moves forward, parallel to the floor, just to the resident body with the velocity of about 0.63 m/s (Fig.17a). The air mixture reaches the resident body at the height of 0.6 m and turns down along the body, slowing down to the velocity of about 0.33 m/s. At the bottom of the indoor zone, the main flow turns back to the external wall and circulates in the closed cell. The cell is bounded by the floor, horizontal surface located at the height of 0.6 m, human body and external wall. The height of the cell is about 0.6 m. This flow pattern results in a poor air zone located just above the floor (Fig.17b). Unlikely the air inlet located above window, the fresh air zone is moved up and out of the resident face. A resident is not

ࣁ [-]

 [%] തࢿ [-]

*PMVavg*  [-]

**Case** *Cz* 

Table 2. The CFD simulation results

in a contact with the fresh air zone.

[ppm]

human's body and the external wall. The cell height is about 1.8 m. This flow pattern results in a fresh air zone located at the front of the resident's face (Fig.16b). The average CO2 concentration in a cellular zone (diameter of 1.0 m) at the front of the resident's head is 735 ppm, while over the breathing zone 1015 ppm.


Table 1. Specification of the simulation cases


Table 2. The CFD simulation results

human's body and the external wall. The cell height is about 1.8 m. This flow pattern results in a fresh air zone located at the front of the resident's face (Fig.16b). The average CO2 concentration in a cellular zone (diameter of 1.0 m) at the front of the resident's head is 735

Case 2 UP Calm NO NO Radiator Case 3 UP Windy YES NO Radiator

Case 5 DOWN Calm NO NO Radiator Case 6 DOWN Windy YES NO Radiator

Case 8 UP Calm NO NO Floor Case 9 UP Windy YES NO Floor

Case 11 DOWN Calm NO NO Floor Case 12 DOWN Windy YES NO Floor

Case 14 UP Calm NO YES Radiator Case 15 UP Windy YES YES Radiator

Case 17 DOWN Calm NO YES Radiator Case 18 DOWN Windy YES YES Radiator

Case 20 UP Calm NO YES Floor Case 21 UP Windy YES YES Floor

Case 23 DOWN Calm NO YES Floor Case 24 DOWN Windy YES YES Floor

location Weather Open

window

DOWN Windy NO NO Radiator

UP Windy NO NO Floor

DOWN Windy NO NO Floor

UP Windy NO YES Radiator

DOWN Windy NO YES Radiator

UP Windy NO YES Floor

DOWN Windy NO YES Floor

UP Windy NO NO Radiator

Exhaust fan

Heating system

ppm, while over the breathing zone 1015 ppm.

1

2

Table 1. Specification of the simulation cases

Gravitational

Mechanical exhaust

Case 1

Case 4

Case 7

Case 10

Case 13

Case 16

Case 19

Case 22

Case name House Group Subgroup Inlet

a

b

c

d

i

ii

i

ii

i

ii

i

ii

Figure 17a shows the distributions of the air velocity for the air inlet located below the window (case 4). The indoor space is heated by radiators without exhaust ventilators. Unlikely the air inlet located above the window, in the case of windy weather, the downward flow is still dominated by the forced convection from the imposed inlet velocity. However, main flow moves forward, parallel to the floor, just to the resident body with the velocity of about 0.63 m/s (Fig.17a). The air mixture reaches the resident body at the height of 0.6 m and turns down along the body, slowing down to the velocity of about 0.33 m/s. At the bottom of the indoor zone, the main flow turns back to the external wall and circulates in the closed cell. The cell is bounded by the floor, horizontal surface located at the height of 0.6 m, human body and external wall. The height of the cell is about 0.6 m. This flow pattern results in a poor air zone located just above the floor (Fig.17b). Unlikely the air inlet located above window, the fresh air zone is moved up and out of the resident face. A resident is not in a contact with the fresh air zone.

Indoor Air Quality and Thermal Comfort in Naturally Ventilated Low-Energy Residential Houses 101

a) b) Fig. 17. The air-CO2 mixture local flow characteristics (case 4): a) velocity vector field, b) CO2

In the case of the calm weather, the location of the air inlet below the window is better. The air-CO2 mixture flows along the resident body downward to the floor with the velocity of 0.27 m/s (Fig.21a). Next, it turns back to the external wall and slows down to the velocity of 0.13 m/s. This flow pattern and flow velocities are enough to create the fresh air zone at the front of the resident head. The CO2 concentration of 1352 ppm in the fresh air zone is still high (Fig.21b), but significantly lower than at the location of the air inlet above the window.

The exchange of the heating system into the floor heating system can improve a little the air quality in the fresh air zone located at the front of the resident's head during the calm weather. Opposite to the radiator heating system, the floor heating system decreases slightly the flow intensity, but surprisingly it results in a creation of the cellular "fresh" air zone at the front of the resident's head (Fig.22a). Even though the "fresh" air zone is created, the average CO2 concentration is high and equals to 1425 ppm (Fig.22b), while the average CO2 concentration in the breathing zone is 1008 ppm. It can results in the resident dissatisfaction.

a) b) Fig. 18. The air-CO2 mixture local flow characteristics (case 7): a) velocity vector field, b) CO2

The average CO2 concentration in the breathing zone is 991 ppm.

concentration distribution

concentration distribution

In the case of the floor heating system (case 7), the flow pattern is alike the radiator heating system. However, main flow bends down to the resident head and reaches the velocity of about 0.51 m/s (Fig.18a). It is much higher than in the case of the radiator heating system (0.36 m/s). The air mixture, enforced by the breathing affect, flows down along a body of the resident reaching the flow velocity of 0.57 m/s and is smaller than in the case of the radiator heating system (Fig.18b). Even though, the CO2 concentrations in the breathing zone for the radiator and floor heating system are almost the same and equal to 1015 ppm and 1016 ppm, respectively, the fresh air zone at the front of the resident is bigger for the radiator heating system. The average CO2 concentration in a cellular zone (diameter of 1.0 m) at the front of the resident head is 628 ppm.

The exhaust ventilators installation does not change the flow pattern significantly. In case of the radiator heating system applied and the air inlet located above the window, for the windy weather, the main flow bends down to the resident head with a slightly higher velocity of 0.43 m/s (Fig.19a). The air-CO2 mixture flows down and back to the external wall, creating a cellular flow pattern in front of the resident. The average flow velocity in the cell is 0.24 m/s and the average CO2 concentration in a cellular zone (diameter of 1.0 m) at the front of the resident head is 453 ppm (Fig.19b), while the average CO2 concentration in the breathing zone is 743 ppm.

The calm weather is more unfavorable for the natural ventilation. In the case of the calm weather, when the gravitational ventilation and the radiator heating system are applied, for the air inlet located above the window there is no fresh air zone at the front of the resident (Fig.20a). There can be observed the mixture flow along the resident body downward to the floor with the velocity of 0.4 m/s and the mixture flow upward and along the radiator, and the window with the velocity of 0.082 m/s. These two mixture streams are not able to create a clear cellular flow of the fresh air. Thus, at the front of the resident head, the CO2 concentration is high, about 1900 ppm (Fig.20b), while the average CO2 concentration is 1002 ppm in the breathing zone.

Fig. 16. The air-CO2 mixture local flow characteristics (case 1): a) velocity vector field, b) CO2 concentration distribution

In the case of the floor heating system (case 7), the flow pattern is alike the radiator heating system. However, main flow bends down to the resident head and reaches the velocity of about 0.51 m/s (Fig.18a). It is much higher than in the case of the radiator heating system (0.36 m/s). The air mixture, enforced by the breathing affect, flows down along a body of the resident reaching the flow velocity of 0.57 m/s and is smaller than in the case of the radiator heating system (Fig.18b). Even though, the CO2 concentrations in the breathing zone for the radiator and floor heating system are almost the same and equal to 1015 ppm and 1016 ppm, respectively, the fresh air zone at the front of the resident is bigger for the radiator heating system. The average CO2 concentration in a cellular zone (diameter of 1.0

The exhaust ventilators installation does not change the flow pattern significantly. In case of the radiator heating system applied and the air inlet located above the window, for the windy weather, the main flow bends down to the resident head with a slightly higher velocity of 0.43 m/s (Fig.19a). The air-CO2 mixture flows down and back to the external wall, creating a cellular flow pattern in front of the resident. The average flow velocity in the cell is 0.24 m/s and the average CO2 concentration in a cellular zone (diameter of 1.0 m) at the front of the resident head is 453 ppm (Fig.19b), while the average CO2 concentration in

The calm weather is more unfavorable for the natural ventilation. In the case of the calm weather, when the gravitational ventilation and the radiator heating system are applied, for the air inlet located above the window there is no fresh air zone at the front of the resident (Fig.20a). There can be observed the mixture flow along the resident body downward to the floor with the velocity of 0.4 m/s and the mixture flow upward and along the radiator, and the window with the velocity of 0.082 m/s. These two mixture streams are not able to create a clear cellular flow of the fresh air. Thus, at the front of the resident head, the CO2 concentration is high, about 1900 ppm (Fig.20b), while the average CO2 concentration is 1002

a) b) Fig. 16. The air-CO2 mixture local flow characteristics (case 1): a) velocity vector field, b) CO2

m) at the front of the resident head is 628 ppm.

the breathing zone is 743 ppm.

ppm in the breathing zone.

concentration distribution

Fig. 17. The air-CO2 mixture local flow characteristics (case 4): a) velocity vector field, b) CO2 concentration distribution

In the case of the calm weather, the location of the air inlet below the window is better. The air-CO2 mixture flows along the resident body downward to the floor with the velocity of 0.27 m/s (Fig.21a). Next, it turns back to the external wall and slows down to the velocity of 0.13 m/s. This flow pattern and flow velocities are enough to create the fresh air zone at the front of the resident head. The CO2 concentration of 1352 ppm in the fresh air zone is still high (Fig.21b), but significantly lower than at the location of the air inlet above the window. The average CO2 concentration in the breathing zone is 991 ppm.

The exchange of the heating system into the floor heating system can improve a little the air quality in the fresh air zone located at the front of the resident's head during the calm weather. Opposite to the radiator heating system, the floor heating system decreases slightly the flow intensity, but surprisingly it results in a creation of the cellular "fresh" air zone at the front of the resident's head (Fig.22a). Even though the "fresh" air zone is created, the average CO2 concentration is high and equals to 1425 ppm (Fig.22b), while the average CO2 concentration in the breathing zone is 1008 ppm. It can results in the resident dissatisfaction.

Fig. 18. The air-CO2 mixture local flow characteristics (case 7): a) velocity vector field, b) CO2 concentration distribution

Indoor Air Quality and Thermal Comfort in Naturally Ventilated Low-Energy Residential Houses 103

a) b) Fig. 20. The air-CO2 mixture local flow characteristics (case 2): a) velocity vector field, b) CO2

a) b) Fig. 21. The air-CO2 mixture local flow characteristics (case 5): a) velocity vector field, b) CO2

a) b) Fig. 22. The air-CO2 mixture local flow characteristics (case 8): a) velocity vector field, b) CO2

concentration distribution

concentration distribution

concentration distribution

Fig. 19. The air-CO2 mixture local flow characteristics (case 13): a) velocity vector field, b) CO2 concentration distribution

Moving down the air inlet location below the window, the air-CO2 mixture flow pattern is not changed (Fig.23a) and the average value of the CO2 concentration in the "fresh" air zone is reduced down to 1390 ppm (Fig.23b), while the average CO2 concentration in the breathing zone is 991 ppm. However, this still can result in the resident's dissatisfaction.

In the case of the calm weather, the exhaust ventilators slightly improve the indoor air quality. In the following discussion, only the fresh air inlet located above the window is considered and two simulation cases are discussed: radiator heating system and floor heating system. For the radiator heating system installed, there is not a clear cellular fresh air zone (Fig.24a). However, the average CO2 concentration at the front of the resident's head is 986 ppm (Fig.24b) and the value is still in the comfort range, while the average CO2 concentration in the breathing zone is 688 ppm.

The heating system exchange into the floor heating system improves significantly the air quality near the resident head, however there is not a clear cellular fresh air zone (Fig.25a). In contrast to the radiator heating system, when the floor system is installed, the zone of the fresh air is significantly larger and the average value of the CO2 concentration near the resident's head is about 583 ppm (Fig.25b). It means that the air quality near the resident head is excellent, while the average CO2 concentration in the breathing zone is 622 ppm.

Analyzing the FE results, it can be concluded that the local air quality in the zone located at the front of the resident head can be significantly different than the one predicted by indices concerning the breathing zone. It can result in the unpredicted resident dissatisfaction. This conclusion is confirmed by experiments (Section 3.3). The momentary dissatisfaction of the local air quality results in actions aiming at the comfort level recovery (e.g. windows opening), while the overall indoor air quality indices indicate IAQ in a comfort range.

a) b)

Moving down the air inlet location below the window, the air-CO2 mixture flow pattern is not changed (Fig.23a) and the average value of the CO2 concentration in the "fresh" air zone is reduced down to 1390 ppm (Fig.23b), while the average CO2 concentration in the breathing zone is 991 ppm. However, this still can result in the resident's dissatisfaction.

In the case of the calm weather, the exhaust ventilators slightly improve the indoor air quality. In the following discussion, only the fresh air inlet located above the window is considered and two simulation cases are discussed: radiator heating system and floor heating system. For the radiator heating system installed, there is not a clear cellular fresh air zone (Fig.24a). However, the average CO2 concentration at the front of the resident's head is 986 ppm (Fig.24b) and the value is still in the comfort range, while the average CO2

The heating system exchange into the floor heating system improves significantly the air quality near the resident head, however there is not a clear cellular fresh air zone (Fig.25a). In contrast to the radiator heating system, when the floor system is installed, the zone of the fresh air is significantly larger and the average value of the CO2 concentration near the resident's head is about 583 ppm (Fig.25b). It means that the air quality near the resident head is excellent, while the average CO2 concentration in the breathing zone is

Analyzing the FE results, it can be concluded that the local air quality in the zone located at the front of the resident head can be significantly different than the one predicted by indices concerning the breathing zone. It can result in the unpredicted resident dissatisfaction. This conclusion is confirmed by experiments (Section 3.3). The momentary dissatisfaction of the local air quality results in actions aiming at the comfort level recovery (e.g. windows opening), while the overall indoor air quality indices indicate IAQ

Fig. 19. The air-CO2 mixture local flow characteristics (case 13): a) velocity vector field, b)

CO2 concentration distribution

622 ppm.

in a comfort range.

concentration in the breathing zone is 688 ppm.

Fig. 20. The air-CO2 mixture local flow characteristics (case 2): a) velocity vector field, b) CO2 concentration distribution

Fig. 21. The air-CO2 mixture local flow characteristics (case 5): a) velocity vector field, b) CO2 concentration distribution

Fig. 22. The air-CO2 mixture local flow characteristics (case 8): a) velocity vector field, b) CO2 concentration distribution

Indoor Air Quality and Thermal Comfort in Naturally Ventilated Low-Energy Residential Houses 105

The indoor air quality is usually assessed by indices averaged over the average breathing zone according to the international guideline EN ISO (2004): the average CO2 concentration *Cz* and the ventilation effectiveness ߟ) which is a measure of how well the supply airflow mixes with the breathing zone for a removal of CO2). In this study, the average breathing zone is modeled as a separate area (Fig.15). The average CO2 concentration over the breathing zone is computed according to Eq.31. The results of simulations for the house equipped with the

a)

b)

c)

d)

Fig. 26. The average CO2 concentration *Cz* in breathing zone for house equipped with

natural ventilation without exhaust ventilators

natural ventilation without exhaust ventilators (gravitational) are presented in Fig.26.

**8.2 Overall indoor air quality** 

Fig. 23. The air-CO2 mixture local flow characteristics (case 11): a) velocity vector field, b) CO2 concentration distribution

Fig. 24. The air-CO2 mixture local flow characteristics (case 14): a) velocity vector field, b) CO2 concentration distribution

Fig. 25. The air-CO2 mixture local flow characteristics (case 20): a) velocity vector field, b) CO2 concentration distribution

#### **8.2 Overall indoor air quality**

104 Air Quality – Monitoring and Modeling

a) b) Fig. 23. The air-CO2 mixture local flow characteristics (case 11): a) velocity vector field, b)

a) b) Fig. 24. The air-CO2 mixture local flow characteristics (case 14): a) velocity vector field, b)

a) b) Fig. 25. The air-CO2 mixture local flow characteristics (case 20): a) velocity vector field, b)

CO2 concentration distribution

CO2 concentration distribution

CO2 concentration distribution

The indoor air quality is usually assessed by indices averaged over the average breathing zone according to the international guideline EN ISO (2004): the average CO2 concentration *Cz* and the ventilation effectiveness ߟ) which is a measure of how well the supply airflow mixes with the breathing zone for a removal of CO2). In this study, the average breathing zone is modeled as a separate area (Fig.15). The average CO2 concentration over the breathing zone is computed according to Eq.31. The results of simulations for the house equipped with the natural ventilation without exhaust ventilators (gravitational) are presented in Fig.26.

Fig. 26. The average CO2 concentration *Cz* in breathing zone for house equipped with natural ventilation without exhaust ventilators

Indoor Air Quality and Thermal Comfort in Naturally Ventilated Low-Energy Residential Houses 107

is passing by and over the resident. Between the resident and the fresh air stream, there is a a relatively large zone of the poor quality air. The poor quality air zone is large enough to increase significantly the value of the average CO2 concentration in the breathing zone. The best air quality (729 ppm) is reached by window opening in the case of the air inlet located above the window and the radiator heating system. The poorest air quality (2104 ppm) is obtained at the calm weather with the radiator heating system. The results of numerical simulations are confirmed by the experimental measurements (Section 3.3). The simulations show that the average CO2 concentration in the breathing zone varies in the range of (729 – 2104) ppm with the mean value of 1352 ppm and is very close to the measured long term

Contrary to the ventilation system without exhaust ventilators, the average CO2 concentration *Cz* in the breathing zone varies very slightly in the range of (645 – 967) ppm (Fig.28). The worse results are obtained for the floor heating system rather than for the radiator heating system. The simulations for the radiator heating system show that the average CO2 concentration in the breathing zone varies in the range of (645 – 967) ppm with the mean value of 772 ppm and is very close to the measured long term average value of 647 ppm. The indoor air quality depends strongly on the ventilation system effectiveness

The simulation results show that the ventilation system effectiveness does not depend on the type of the heating system. Surprisingly, the higher effectiveness is obtained with the

(Fig.29), which is expressed in terms of the ߟ index (Eq. 30).

gravitational ventilation system without exhaust ventilators.

Fig. 28. The average CO2 concentration *Cz* in the breathing zone

average value of 1266 ppm.

Fig. 27. The average CO2 concentration *Cz* in breathing zone for house equipped with natural ventilation with exhaust ventilators

The results of the simulations for the house equipped with the natural ventilation with exhaust ventilators are presented in Fig.27.

In the house equipped with the natural ventilation system without exhaust ventilators (gravitational ventilation), the average CO2 concentration varies significantly under the weather conditions depending upon the location of the fresh air inlet and the type of the heating system (Fig.28). Surprisingly, the very high CO2 concentration (1776 ppm) can be observed for the windy weather and the open window (cases 6 and 12). Studying Figs.26b and 26d, it can be noticed that a stream of the fresh air incoming through the open window

a)

b)

c)

d)

The results of the simulations for the house equipped with the natural ventilation with

In the house equipped with the natural ventilation system without exhaust ventilators (gravitational ventilation), the average CO2 concentration varies significantly under the weather conditions depending upon the location of the fresh air inlet and the type of the heating system (Fig.28). Surprisingly, the very high CO2 concentration (1776 ppm) can be observed for the windy weather and the open window (cases 6 and 12). Studying Figs.26b and 26d, it can be noticed that a stream of the fresh air incoming through the open window

Fig. 27. The average CO2 concentration *Cz* in breathing zone for house equipped with

natural ventilation with exhaust ventilators

exhaust ventilators are presented in Fig.27.

is passing by and over the resident. Between the resident and the fresh air stream, there is a a relatively large zone of the poor quality air. The poor quality air zone is large enough to increase significantly the value of the average CO2 concentration in the breathing zone. The best air quality (729 ppm) is reached by window opening in the case of the air inlet located above the window and the radiator heating system. The poorest air quality (2104 ppm) is obtained at the calm weather with the radiator heating system. The results of numerical simulations are confirmed by the experimental measurements (Section 3.3). The simulations show that the average CO2 concentration in the breathing zone varies in the range of (729 – 2104) ppm with the mean value of 1352 ppm and is very close to the measured long term average value of 1266 ppm.

Contrary to the ventilation system without exhaust ventilators, the average CO2 concentration *Cz* in the breathing zone varies very slightly in the range of (645 – 967) ppm (Fig.28). The worse results are obtained for the floor heating system rather than for the radiator heating system. The simulations for the radiator heating system show that the average CO2 concentration in the breathing zone varies in the range of (645 – 967) ppm with the mean value of 772 ppm and is very close to the measured long term average value of 647 ppm. The indoor air quality depends strongly on the ventilation system effectiveness (Fig.29), which is expressed in terms of the ߟ index (Eq. 30).

The simulation results show that the ventilation system effectiveness does not depend on the type of the heating system. Surprisingly, the higher effectiveness is obtained with the gravitational ventilation system without exhaust ventilators.

Fig. 28. The average CO2 concentration *Cz* in the breathing zone

Indoor Air Quality and Thermal Comfort in Naturally Ventilated Low-Energy Residential Houses 109

Fig. 30. The average CO2 concentration *Cz* in function of fresh air inlet velocity

Fig. 31. The ventilation effectiveness ߟ in function of fresh air inlet velocity

Fig. 29. The ventilation effectiveness in terms of ߟ index

#### **8.3 Optimal fresh air inlet velocity for natural ventilation**

According to the simulations results, the indoor air quality in houses with natural ventilation systems depends strongly on the fresh air inlet velocity. To find the optimum inlet velocity, some simulations were carried out (Figs.30-33). The investigations were limited to the case of the natural ventilation system without exhaust ventilators, fresh air inlet location above the window and radiator heating system. The simulations considered inlet velocities in the range of (0.00 – 0.98) m/s.

The optimum range of the fresh air inlet velocities varies between 0.2 m/s and 0.4 m/s. This range of velocities satisfies the indices of the indoor air quality and the thermal comfort.

#### **8.4 Overall thermal comfort**

Concerning the problem of IAQ in residential houses it is reasonable to take into account the thermal comfort indices. As the IAQ indices, the thermal comfort indices influence the Indoor Environment Quality. All of them depend on the air-CO2 mixture flow patterns and, in consequence, on the temperature distribution in the indoor zone. Thus, in this section the main overall thermal comfort indices are considered.

Figure 34 presents the average PMV value (Eq.24) for each simulation, where the gravitational ventilation is installed. In turn, Fig.35 presents the average PMV value (Eq.24) for each simulation with the gravitational ventilation and exhaust ventilators.

According to the simulations results, the indoor air quality in houses with natural ventilation systems depends strongly on the fresh air inlet velocity. To find the optimum inlet velocity, some simulations were carried out (Figs.30-33). The investigations were limited to the case of the natural ventilation system without exhaust ventilators, fresh air inlet location above the window and radiator heating system. The simulations considered

The optimum range of the fresh air inlet velocities varies between 0.2 m/s and 0.4 m/s. This range of velocities satisfies the indices of the indoor air quality and the thermal comfort.

Concerning the problem of IAQ in residential houses it is reasonable to take into account the thermal comfort indices. As the IAQ indices, the thermal comfort indices influence the Indoor Environment Quality. All of them depend on the air-CO2 mixture flow patterns and, in consequence, on the temperature distribution in the indoor zone. Thus, in this section the

Figure 34 presents the average PMV value (Eq.24) for each simulation, where the gravitational ventilation is installed. In turn, Fig.35 presents the average PMV value (Eq.24)

for each simulation with the gravitational ventilation and exhaust ventilators.

Fig. 29. The ventilation effectiveness in terms of ߟ index

inlet velocities in the range of (0.00 – 0.98) m/s.

main overall thermal comfort indices are considered.

**8.4 Overall thermal comfort** 

**8.3 Optimal fresh air inlet velocity for natural ventilation** 

Fig. 30. The average CO2 concentration *Cz* in function of fresh air inlet velocity

Fig. 31. The ventilation effectiveness ߟ in function of fresh air inlet velocity

Indoor Air Quality and Thermal Comfort in Naturally Ventilated Low-Energy Residential Houses 111

a)

b

c)

d)

Fig. 34. The PMV index distribution for houses with natural ventilation and without exhaust

ventilators

Fig. 32. The percentage dissatisfied *PD* in function of fresh air inlet velocity

Fig. 33. The overall ventilation effectiveness for temperature distribution �̅ � in function of fresh air inlet velocity

Fig. 32. The percentage dissatisfied *PD* in function of fresh air inlet velocity

Fig. 33. The overall ventilation effectiveness for temperature distribution �̅

fresh air inlet velocity

� in function of

Fig. 34. The PMV index distribution for houses with natural ventilation and without exhaust ventilators

Indoor Air Quality and Thermal Comfort in Naturally Ventilated Low-Energy Residential Houses 113

inlet location, while for the floor heating system, the thermal comfort strongly varies and depends on weather conditions and the air inlet location. Surprisingly, the best comfort level is reached in the house equipped with the radiator heating system and air inlet located below the window (*PMVavg*=1.11), while the worst result is obtained in the house with the

According to the *PMVavg* index, it can be concluded that the best comfort level is observed in the house with the natural ventilation and without exhaust ventilators. More precise conclusions can be derived by studying the simulations results of the PD index (Fig. 37). It can be noted that the percentage of dissatisfied ones (PD index) depends only on the type of the heating system installed. The biggest values of the PD index of 13.9 % are obtained for

The overall ventilation effectiveness for a temperature distribution is provided by the

range of (0.8265 – 1.0139)). It means that a temperature homogeneity level is high enough.

2.3943. However, the open window is an abnormal situation and the corresponding values

results, it can be seen that the natural ventilation without exhaust ventilators can maintain the satisfying thermal comfort level. It can be even higher by installing the radiator heating system. The location of the air inlet gap does not influence the thermal comfort when using

� index are artificial and thus can't be taken into account. Analyzing the above

� related to the way in which the heat is distributed over the indoor zone

� index is almost the same (varies in the

� index reaches

floor heating system and air inlet located below the window (*PMVavg*=-2.78).

The only exception is observed for the cases with open windows; the �̅

the natural ventilation system with or without exhaust ventilators.

Fig. 36. The average *PMVavg* index for houses with natural ventilation

the floor heating system.

(Fig.38). However, for all investigated cases, the �̅

quantitative index �̅

of the �̅

Fig. 35. The PMV index distribution for houses with natural ventilation and exhaust ventilators

In the case of houses with the gravitational ventilation without exhaust ventilators, a good thermal comfort is maintained independently of the heating system (Fig.36), air inlet location and weather conditions. It is not surprising that the thermal comfort is better during the calm weather (*PMVavg*=1.22) than during the windy weather (*PMVavg*=0.81). For houses with natural ventilation and exhaust ventilators, the results are significantly different (Fig.36). The house with the radiator heating system maintains the good thermal comfort level (*PMVavg* is in range of (0.77–1.11)) independently of weather conditions and the air

a)

b)

c)

d)

In the case of houses with the gravitational ventilation without exhaust ventilators, a good thermal comfort is maintained independently of the heating system (Fig.36), air inlet location and weather conditions. It is not surprising that the thermal comfort is better during the calm weather (*PMVavg*=1.22) than during the windy weather (*PMVavg*=0.81). For houses with natural ventilation and exhaust ventilators, the results are significantly different (Fig.36). The house with the radiator heating system maintains the good thermal comfort level (*PMVavg* is in range of (0.77–1.11)) independently of weather conditions and the air

Fig. 35. The PMV index distribution for houses with natural ventilation and exhaust

ventilators

inlet location, while for the floor heating system, the thermal comfort strongly varies and depends on weather conditions and the air inlet location. Surprisingly, the best comfort level is reached in the house equipped with the radiator heating system and air inlet located below the window (*PMVavg*=1.11), while the worst result is obtained in the house with the floor heating system and air inlet located below the window (*PMVavg*=-2.78).

According to the *PMVavg* index, it can be concluded that the best comfort level is observed in the house with the natural ventilation and without exhaust ventilators. More precise conclusions can be derived by studying the simulations results of the PD index (Fig. 37). It can be noted that the percentage of dissatisfied ones (PD index) depends only on the type of the heating system installed. The biggest values of the PD index of 13.9 % are obtained for the floor heating system.

The overall ventilation effectiveness for a temperature distribution is provided by the quantitative index �̅ � related to the way in which the heat is distributed over the indoor zone (Fig.38). However, for all investigated cases, the �̅ � index is almost the same (varies in the range of (0.8265 – 1.0139)). It means that a temperature homogeneity level is high enough. The only exception is observed for the cases with open windows; the �̅ � index reaches 2.3943. However, the open window is an abnormal situation and the corresponding values of the �̅ � index are artificial and thus can't be taken into account. Analyzing the above results, it can be seen that the natural ventilation without exhaust ventilators can maintain the satisfying thermal comfort level. It can be even higher by installing the radiator heating system. The location of the air inlet gap does not influence the thermal comfort when using the natural ventilation system with or without exhaust ventilators.

Fig. 36. The average *PMVavg* index for houses with natural ventilation

Indoor Air Quality and Thermal Comfort in Naturally Ventilated Low-Energy Residential Houses 115

The fresh air inlet location above the window opening leads to a better indoor air quality,

A type of the heating system does not influence the indoor air quality, while strongly affects the thermal comfort. From the thermal comfort point of view, the radiator heating system is

The installation of exhaust ventilators significantly improves the indoor air quality, however, slightly deteriorates the thermal comfort. The exhaust ventilators installed in houses with the fresh air inlet located below the window deteriorate the thermal comfort

The *CZ* index of the average CO2 concentration in the breathing zone as the assessment method is not well suited for the natural ventilation in residential houses located in coldclimate regions. The results indicate significant differences between the *CZ* index and the average CO2 concentration in the local fresh air zone located at the front of the resident's head. It can be a main reason of the dissatisfaction, leading to the unpredicted resident's

The natural ventilation without exhaust ventilators can maintain the good thermal comfort and acceptable indoor air quality, but the fresh air inlet has to be installed above the window and the radiator heating system should be applied. However, this type of ventilation is very sensitive to ambient conditions, thus the fresh air inlets should be

Our further research will focus on experimental investigations of the local fresh air zone and

Behne, M. (1999). Indoor air quality in rooms with cooled ceilings. Mixing ventilation or rather displacement ventilation? *Energy and Buildings*, 30(2), pp.155–66 Duffie, J. A. & Beckman, W. A. (1991). Solar Engineering of Thermal Processes, New York:

Energy Consumption Guide 19, Energy efficiency in offices, London: Energy Efficiency

Hastings, S. R. (2004). Breaking the "heating barrier". Learning from the first houses without

Ho, S. H., Rosario, L. & Rahman, R. R. (2011). Comparison of underfloor and overhead air

Hummelgaard, J., Juhl, P., Sebjornsson, K. O., Clausen G., Toftum, J. & Langkilde, G. (2007).

distribution systems in an office environment. *Building and Environment,* 46,

Indoor air quality and occupant satisfaction in five mechanically and four naturally ventilated open-plan office buildings. *Building and Environment*, 42, pp.4051–4058 Hyun, S. & Kleinstreuer, C. (2001). Numerical simulation of mixed convection heat and mass

transfer in a human inhalation test chamber. *International Journal of Heat and Mass* 

conventional heating. *Energy and Buildings*, 26, pp.373-380

however decreases the thermal comfort.

even stronger.

**10. References** 

significantly better (based on the PD index) than the floor one.

actions aimed at recovering the satisfaction, e.g. by windows opening.

controlled in order to maintain a constant optimum air velocity.

Ansys Inc., ANSYS Release 11.0 Documentation for ANSYS, 2007. Awbi, H. (2003). Ventilation of building, E & FN Spon, 2003.

its influence on the subjective satisfaction sensation.

John Wiley and Sons

Office/HMSO, 1993.

*Transfer*, 44, pp.2247–60

pp.1415-1427

Fig. 37. The average *PD* index for houses with natural ventilation

Fig. 38. The average ventilation effectiveness for temperature distribution �̅ � in houses with natural ventilation

#### **9. Conclusions**

The CFD simulations were carried out to study the indoor air quality and the thermal comfort in residential houses with the natural ventilation and with or without exhaust ventilators. The average results were compared with the experimental measurements and a good correlation was found between them.

The fresh air inlet location above the window opening leads to a better indoor air quality, however decreases the thermal comfort.

A type of the heating system does not influence the indoor air quality, while strongly affects the thermal comfort. From the thermal comfort point of view, the radiator heating system is significantly better (based on the PD index) than the floor one.

The installation of exhaust ventilators significantly improves the indoor air quality, however, slightly deteriorates the thermal comfort. The exhaust ventilators installed in houses with the fresh air inlet located below the window deteriorate the thermal comfort even stronger.

The *CZ* index of the average CO2 concentration in the breathing zone as the assessment method is not well suited for the natural ventilation in residential houses located in coldclimate regions. The results indicate significant differences between the *CZ* index and the average CO2 concentration in the local fresh air zone located at the front of the resident's head. It can be a main reason of the dissatisfaction, leading to the unpredicted resident's actions aimed at recovering the satisfaction, e.g. by windows opening.

The natural ventilation without exhaust ventilators can maintain the good thermal comfort and acceptable indoor air quality, but the fresh air inlet has to be installed above the window and the radiator heating system should be applied. However, this type of ventilation is very sensitive to ambient conditions, thus the fresh air inlets should be controlled in order to maintain a constant optimum air velocity.

Our further research will focus on experimental investigations of the local fresh air zone and its influence on the subjective satisfaction sensation.

#### **10. References**

114 Air Quality – Monitoring and Modeling

Fig. 37. The average *PD* index for houses with natural ventilation

Fig. 38. The average ventilation effectiveness for temperature distribution �̅

The CFD simulations were carried out to study the indoor air quality and the thermal comfort in residential houses with the natural ventilation and with or without exhaust ventilators. The average results were compared with the experimental measurements and a

natural ventilation

**9. Conclusions** 

good correlation was found between them.

� in houses with

Ansys Inc., ANSYS Release 11.0 Documentation for ANSYS, 2007.

Awbi, H. (2003). Ventilation of building, E & FN Spon, 2003.


**6** 

Hermann Fromme

 *Germany* 

*Bavarian Health and Food Safety Authority, Dep. of Chemical Safety and Toxicology* 

**Particles in the Indoor Environment** 

As a result of a change in living and work habits, we now stay in industrial countries every day more than 90% of the time inside buildings. Against this backdrop findings about the exposure of users are relevant. Given their heterogeneity, very complex exposure patterns exist in the indoor environment in respect of which not only input from the outdoor air but also important sources inside the rooms themselves have to be taken into account. At any rate, different indoor environments have to be identified (e.g. living, bed, handicraft, leisure and cellar rooms, working rooms and workplaces in buildings, public buildings, restaurants and inns, community facilities such as schools and kindergartens as well as spaces in motor vehicles and other public transportation systems). Furthermore, it has been shown that the amount of airborne particle content in indoor environments can be highly variable in terms of space but also in terms of time. Apart from the conditions prevailing in the outdoor air close to the indoor environment (e.g. location close to a heavily trafficked street or in a rural region) and the current climatic conditions, the structural conditions of the building and the ventilation conditions are important. Furthermore, activities in indoor environments, such as the deposition and resuspension of house dust, cooking and cleaning activities or

smoking can make a considerable contribution to the respective pollution situation.

Particulates (particulate matter, PM) which are dispersedly distributed in the air form colloidal systems with the gases which are also referred to as aerosols. Overall, the composition of aerosols strongly depend on the specific sources. The particles of the fine fraction develop primarily through transformation processes from gases or within the framework of combustion processes. They are typically composed of nitrates, sulphates, ammonium, elementary carbon, a large number of organic compounds and trace elements. By contrast, the particles in the coarse fraction develop largely mechanically following the disintegration of larger solid particles and consist typically of whirled up dust from industrial processes and biological material such as pollen and bacteria and

PM in indoor environments consist of very different particles which are considerably varying in terms of size, form and chemical composition. Whereas the larger particles determine primarily the mass of the environmental aerosol, the particle number concentration (PNC) and the particle surface are dominated almost exclusively by the ultra

**1. Introduction** 

their fragments.

fine particles (<100 nm).

ISO 16000-1, (2004). E, Indoor air - part 1: General aspects of sampling strategy


### **Particles in the Indoor Environment**

#### Hermann Fromme

*Bavarian Health and Food Safety Authority, Dep. of Chemical Safety and Toxicology Germany* 

#### **1. Introduction**

116 Air Quality – Monitoring and Modeling

ISO 16814 (2008), Building environment design - Indoor air quality - Methods of expressing

ISO 6946:2007, Building components and building elements. Thermal resistance and thermal

ISO 7730:2005, Ergonomics of the thermal environment. Analytical determination and

Karlsson, J. F. & Moshfegh, B. (2006). Energy demand and indoor climate in a low energy

Launder, B. E. & Spalding, D. B. (1974). The Numerical Computation of Turbulent Flows. *Computer Methods in Applied Mechanics and Engineering,* Vol. 3, pp 269-289 Laverge, J., Van Den Bossche, N., Heijmans, N. & Janssens, A. (2011). Energy saving

Lawrence, T. M. & Braun, J. E. (2007). A methodology for estimating occupant CO2 source

Li, J., Way, O. W. H., Li,Y. S., Zhan, J., Ho, Y. A., Li, J. & Lam, E. (2010). Effect of green roof on ambient CO2 concentration. *Building and Environment,* 45, pp.2644-2651 Maier, T., Krzaczek, M. & Tejchman, J. (2009). Comparison of physical performances of the

Meyer, F. (1993). Niedrigenergiehäuser Heidenheim: Hauskonzepte und erste

Murakami, S., Kato, S. & Zeng J. (2000). Combined simulation of airflow, radiation and

Paul, T., Sree, D. & Aglan, H. (2010). Effect of mechanically induced ventilation on the indoor air quality of building envelopes. *Energy and Buildings*, 42, pp. 326–332 Redlich, C. A., Sparer, J. & Cullen M. R. (1997). *Sick-building syndrome*. Lancet, 349 (9057),

Reinmuth, F. (1994). Energieeinsparung in der Gebäudetechnik: Baukörper und technische

Stavrakakis, G. M., Zervas, P. L., Sarimveis, H. & Markatos, N. C. (2010) Development of a

naturally ventilated rural houses. *Building and Environment*, 45(1), pp.65-80 Viollet, P. L. (1987). The Modelling of Turbulent Re-circulating flows for the purpose of reactor thermal-hydraulic analysis. *Nuclear Engineering and Design*, 99, pp.365-377 Wong, N. H. & Huang, B. (2004). Comparative study of the indoor air quality of naturally

computational tool to quantify architectural design effects on thermal comfort in

ventilated and air-conditioned bedrooms of residential buildings in Singapore.

Systeme der Energieverwendung, Vogel Verlag, Würzburg

*Building and Environment*, 39, pp.1115-1123

ventilation strategies. *Building and Environment*, 46, pp.1497-1503

interpretation of thermal comfort using calculation of the PMV and PPD indices

building – changed control strategies and boundary condition. *Energy and Building*,

potential and repercussions on indoor air quality of demand controlled residential

generation rates from measurements in small commercial buildings. *Building and* 

ventilation systems in low-energy residential houses. *Energy and Building*, 41,

Meßergebnisse, Eggenstein-Leopoldshafen: Fachinformationszentrum, Bine Projekt

moisture transport for heat release from a human body. *Building and Environment*,

ISO 16000-1, (2004). E, Indoor air - part 1: General aspects of sampling strategy

the quality of indoor air for human occupancy

and local thermal comfort criteria, Annex D, 2005.

transmittance. Calculation method, 2007.

38, pp.315-326

pp.337-353

35, pp.489–500

pp.1013–6

*Environment*, 42, pp.623-639

Info-Service Nr. 9, Karlsruhe

As a result of a change in living and work habits, we now stay in industrial countries every day more than 90% of the time inside buildings. Against this backdrop findings about the exposure of users are relevant. Given their heterogeneity, very complex exposure patterns exist in the indoor environment in respect of which not only input from the outdoor air but also important sources inside the rooms themselves have to be taken into account. At any rate, different indoor environments have to be identified (e.g. living, bed, handicraft, leisure and cellar rooms, working rooms and workplaces in buildings, public buildings, restaurants and inns, community facilities such as schools and kindergartens as well as spaces in motor vehicles and other public transportation systems). Furthermore, it has been shown that the amount of airborne particle content in indoor environments can be highly variable in terms of space but also in terms of time. Apart from the conditions prevailing in the outdoor air close to the indoor environment (e.g. location close to a heavily trafficked street or in a rural region) and the current climatic conditions, the structural conditions of the building and the ventilation conditions are important. Furthermore, activities in indoor environments, such as the deposition and resuspension of house dust, cooking and cleaning activities or smoking can make a considerable contribution to the respective pollution situation.

Particulates (particulate matter, PM) which are dispersedly distributed in the air form colloidal systems with the gases which are also referred to as aerosols. Overall, the composition of aerosols strongly depend on the specific sources. The particles of the fine fraction develop primarily through transformation processes from gases or within the framework of combustion processes. They are typically composed of nitrates, sulphates, ammonium, elementary carbon, a large number of organic compounds and trace elements. By contrast, the particles in the coarse fraction develop largely mechanically following the disintegration of larger solid particles and consist typically of whirled up dust from industrial processes and biological material such as pollen and bacteria and their fragments.

PM in indoor environments consist of very different particles which are considerably varying in terms of size, form and chemical composition. Whereas the larger particles determine primarily the mass of the environmental aerosol, the particle number concentration (PNC) and the particle surface are dominated almost exclusively by the ultra fine particles (<100 nm).

Particles in the Indoor Environment 119

crack diameters and geometry, the largest penetration factor seems to exist for the particles

**Size Interval (µm)**

Fig. 2. Penetration efficiencies (P) and deposition rates (k) (all homes nightly averaged data

The deposition of particles on surfaces is based on different physical mechanisms such as gravitation and diffusion. Apart from the deposition speed, this process is described by the so-called deposition rate (k) (example, Fig. 2). This process is strongly dependent on the particle diameter and reaches a minimum for particles with an aerodynamic diameter of approximately 0.4 µm. However, there is a considerable variation range (Morawska & Salthammer, 2003; Miguel et al., 2005; Hussein et al., 2009). The particle deposition, in particular of coarse particles, increases with a rising draught in the room and an increasing

Particles deposited on the surfaces of the room can become resuspended in the indoor air in particular through activities in the indoor environment (Thatcher & Layton, 1995; Hussein et al., 2006). Hu et al. (2005) state that essentially three parameters like mechanical vibration, aerodynamic as well as electrostatic forces can achieve a stronger effect than gravitation and

**0.4-0.5**

**0.7-1**

**1-2**

**2-3**

**3-4**

**4-5**

**5-6**

**all homes summer deposition**

**0.3-0.4**

with diameter between approximately >0.05 and < 1 µm (see Fig. 2).

**0.02-0.03**

**2.2 Deposition** 

**2.3 Resuspension** 

**0**

**0.1**

**0.2**

**0.3**

**0.4**

**0.5**

**0.6**

**0.7**

**0.8**

**0.9**

**1.0**

**1.1**

**Penetration / deposition efficiency**

**0.03-0.04**

**0.04-0.06**

**0.06-0.08**

**0.8-0.1**

**0.1-0.15**

**0.15-0.2**

from n=98-106 and in summer from n=8) (modified from Long et al., 2001)

room area and also varies depending on the degree of interior decoration.

**0.2-0.3**

Concerning the measurement of PM in air, different sampling conventions have established themselves, often using the aerodynamic diameter of the particles. In order to better reflect the human respiratory characteristics, conventions such as PM10 (Particulate Matter) or PM2.5 were introduced by the US Environmental Protection Agency and European authorities. PM2.5 is, for instance, the particle fraction which passes through a size-selective air inlet which has a separation efficiency of 50% for an aerodynamic diameter of 2.5 µm. Depending on the specific context, other definitions may be applied, for example in indoor working environments.

#### **2. Behaviour, transport, and fate of particles in the indoor environment**

The transport and fate of particles in indoor environments are fundamentally influenced by a series of physical and chemical processes (Fig. 1). This can lead to considerable changes in terms of their chemical composition, their physical characteristics, their distribution patterns and finally the measurable contents (Thatcher et al., 2001; Morawska & Salthammer, 2003; Nazaroff, 2004).

Fig. 1. Transport and transformation processes with impact on the indoor concentration of particulate matter (modified from Thatcher et al., 2001)

#### **2.1 Infiltration/ penetration**

The dimensionless penetration factor (P) is defined as the share of the particle fraction with a specific diameter which reaches the indoor environment through the inflow of outdoor air. In the scientific literature there are results of different studies which are based on the observation of the indoor to outdoor ratio of the particles, manipulations of the external building envelope, experimental simulations in the laboratory or mathematical modelling (e.g. Long et al., 2000; Vette et al., 2001; Riley et al., 2002; Riley et al., 2002; Liu & Nazaroff, 2003; Chen & Zhao, 2011). The results show that for different types of buildings and gap /

Concerning the measurement of PM in air, different sampling conventions have established themselves, often using the aerodynamic diameter of the particles. In order to better reflect the human respiratory characteristics, conventions such as PM10 (Particulate Matter) or PM2.5 were introduced by the US Environmental Protection Agency and European authorities. PM2.5 is, for instance, the particle fraction which passes through a size-selective air inlet which has a separation efficiency of 50% for an aerodynamic diameter of 2.5 µm. Depending on the specific context, other definitions may be applied, for example in indoor

**2. Behaviour, transport, and fate of particles in the indoor environment** 

The transport and fate of particles in indoor environments are fundamentally influenced by a series of physical and chemical processes (Fig. 1). This can lead to considerable changes in terms of their chemical composition, their physical characteristics, their distribution patterns and finally the measurable contents (Thatcher et al., 2001; Morawska & Salthammer, 2003;

Fig. 1. Transport and transformation processes with impact on the indoor concentration of

The dimensionless penetration factor (P) is defined as the share of the particle fraction with a specific diameter which reaches the indoor environment through the inflow of outdoor air. In the scientific literature there are results of different studies which are based on the observation of the indoor to outdoor ratio of the particles, manipulations of the external building envelope, experimental simulations in the laboratory or mathematical modelling (e.g. Long et al., 2000; Vette et al., 2001; Riley et al., 2002; Riley et al., 2002; Liu & Nazaroff, 2003; Chen & Zhao, 2011). The results show that for different types of buildings and gap /

particulate matter (modified from Thatcher et al., 2001)

**2.1 Infiltration/ penetration** 

working environments.

Nazaroff, 2004).

crack diameters and geometry, the largest penetration factor seems to exist for the particles with diameter between approximately >0.05 and < 1 µm (see Fig. 2).

Fig. 2. Penetration efficiencies (P) and deposition rates (k) (all homes nightly averaged data from n=98-106 and in summer from n=8) (modified from Long et al., 2001)

#### **2.2 Deposition**

The deposition of particles on surfaces is based on different physical mechanisms such as gravitation and diffusion. Apart from the deposition speed, this process is described by the so-called deposition rate (k) (example, Fig. 2). This process is strongly dependent on the particle diameter and reaches a minimum for particles with an aerodynamic diameter of approximately 0.4 µm. However, there is a considerable variation range (Morawska & Salthammer, 2003; Miguel et al., 2005; Hussein et al., 2009). The particle deposition, in particular of coarse particles, increases with a rising draught in the room and an increasing room area and also varies depending on the degree of interior decoration.

#### **2.3 Resuspension**

Particles deposited on the surfaces of the room can become resuspended in the indoor air in particular through activities in the indoor environment (Thatcher & Layton, 1995; Hussein et al., 2006). Hu et al. (2005) state that essentially three parameters like mechanical vibration, aerodynamic as well as electrostatic forces can achieve a stronger effect than gravitation and

Particles in the Indoor Environment 121

Tobacco smoking constitutes an essential particle source in indoor environments which results in an increase in the particle mass as well as the ultrafine particles. In the Harvard Six City Study, for instance, the annual mean values in smoker households were higher by approximately factor 3 compared to non smoker households (Neas et al., 1994). Fig. 4 shows the increase in indoor pollution depending on the number of cigarettes smoked. In the same way the particle number increases considerably during cigarette smoking, partly to values up to 213,000 particles/cm³ (He et al., 2004; Afshari et al., 2005; Hussein et al.,

When burning candles or oil lamps in indoor environments, an increase in ultrafine particles was likewise observed (Fine et al., 1999; Hussein et al., 2006; Wallace & Ott, 2011). This involved significantly higher concentrations when extinguishing candles compared to the burning itself (Hussein et al., 2006). During the burning of incense sticks it is also possible to detect high particle contents, in particular in the range from 0.06 to 2.5 µm, in indoor air

**3. Sources of particles indoors** 

(Chao et al., 1998; Jetter et al., 2002).

**3.1 Burning processes** 

2006).

**Respirable particulate matter (µg/m³)**

**0**

**Never and former**

the home (modified from Neas et al., 1994)

**Changed status**

**Consistently smoking**

**0.5-1 pack**

**>1 pack**

**< 0.5 pack**

Fig. 4. Distribution percentiles for the annual average concentration of indoor particulate matter by household smoking status and the estimated number of cigarette packs smoked in

**20**

**40**

**60**

**80**

**100**

**120**

hence influence the resuspension of particles. In different field studies it was shown that activities in the indoor environment (e.g. running, playing kids) resulted in a significant increase in PM contents, whereby essentially coarse particles were whirled up (Thatcher & Layton, 1995; Long et al., 2000; Miguel et al., 2005) (Fig. 3). Moreover it could be shown that the resuspension in rooms with wall to wall carpet was significantly higher compared to rooms with a smooth flooring (Long et al., 2000).

Fig. 3. The ratio of the suspended particle concentration after a resuspension activity to the indoor concentration before that activity, by particle size (modified from Thatcher & Layton, 1995)

#### **2.4 Particle formation**

Within the framework of chemical processes in the indoor environment particles can be newly formed or there can be a growth in particle size. The coagulation of particles in the indoor environment is based on the fact that e.g. depending on the particle number they come together with a certain probability and then tend to agglomerate. This process is, for instance, relevant for ultra fine particles in indoor environments, since the latter exist e.g. in high number concentrations when for example burning candles. They then agglomerate over time; this can be observed through a shift in the peak value of particle distribution (Dennekamp et al., 2001). The phenomenon of phase transition, too, describes an "ageing process" during which a growth of the particles is observed through the adsorption of organic substances or water.

#### **3. Sources of particles indoors**

#### **3.1 Burning processes**

120 Air Quality – Monitoring and Modeling

hence influence the resuspension of particles. In different field studies it was shown that activities in the indoor environment (e.g. running, playing kids) resulted in a significant increase in PM contents, whereby essentially coarse particles were whirled up (Thatcher & Layton, 1995; Long et al., 2000; Miguel et al., 2005) (Fig. 3). Moreover it could be shown that the resuspension in rooms with wall to wall carpet was significantly higher compared to

**11.4 29.5 9.2**

**Particle diameter (µm)**

Fig. 3. The ratio of the suspended particle concentration after a resuspension activity to the indoor concentration before that activity, by particle size (modified from Thatcher & Layton,

Within the framework of chemical processes in the indoor environment particles can be newly formed or there can be a growth in particle size. The coagulation of particles in the indoor environment is based on the fact that e.g. depending on the particle number they come together with a certain probability and then tend to agglomerate. This process is, for instance, relevant for ultra fine particles in indoor environments, since the latter exist e.g. in high number concentrations when for example burning candles. They then agglomerate over time; this can be observed through a shift in the peak value of particle distribution (Dennekamp et al., 2001). The phenomenon of phase transition, too, describes an "ageing process" during which a growth of the particles is observed through the adsorption of

**0.5-1 >1-5 >5-10 >10-25**

rooms with a smooth flooring (Long et al., 2000).

2 Min.walk/sit 4 people, 5 minutes 4 people; 30 minutes

Walk in

Cleaning and vacuuming

**Ratio after / before activity**

**0**

**2.4 Particle formation** 

organic substances or water.

1995)

**1**

**2**

**3**

**4**

**5**

**6**

**7**

Tobacco smoking constitutes an essential particle source in indoor environments which results in an increase in the particle mass as well as the ultrafine particles. In the Harvard Six City Study, for instance, the annual mean values in smoker households were higher by approximately factor 3 compared to non smoker households (Neas et al., 1994). Fig. 4 shows the increase in indoor pollution depending on the number of cigarettes smoked. In the same way the particle number increases considerably during cigarette smoking, partly to values up to 213,000 particles/cm³ (He et al., 2004; Afshari et al., 2005; Hussein et al., 2006).

When burning candles or oil lamps in indoor environments, an increase in ultrafine particles was likewise observed (Fine et al., 1999; Hussein et al., 2006; Wallace & Ott, 2011). This involved significantly higher concentrations when extinguishing candles compared to the burning itself (Hussein et al., 2006). During the burning of incense sticks it is also possible to detect high particle contents, in particular in the range from 0.06 to 2.5 µm, in indoor air (Chao et al., 1998; Jetter et al., 2002).

Fig. 4. Distribution percentiles for the annual average concentration of indoor particulate matter by household smoking status and the estimated number of cigarette packs smoked in the home (modified from Neas et al., 1994)

Particles in the Indoor Environment 123

During cleaning and in particular vacuum cleaning, an increase in coarse particles and hence in particular the particle mass is observed in indoor air (Abt et al., 2000). See also figure 5. In two US studies the contribution of cleaning activities to the PM2.5 indoor pollution was estimated at 23-32 µg/m³ (Long et al., 2000; Ferro et al., 2004). Afshari et al. (2004) describe, by contrast, merely an insignificant minor increase for ultrafine particles.

Long et al. (2000) investigated the influence of the use of commercial cleaning agents on a pine oil basis on the exposure in a living room. During the activities the PNCs rose from initially 2,000 particles/cm³ to a maximum of 190,000 particles/cm³ and the PM2.5 contents increased from 5 to 38 µg/m³. This phenomenon was explained by referring to the new particle formation and / or particle growth through oxidative processes in the indoor environment. Other working groups, too, were able to detect in test chambers in the presence of ozone and the simultaneous application of terpene-containing cleaning agents a significant increase in particle number concentrations and the particle mass (Sarwar et al.,

Following chemical reactions of the gas and aerosol phase, so-called secondary organic aerosols (SOAs) are newly formed in indoor environments (Weschler et al. 2006). The formation of SOAs through the reaction of ozone with terpenes and other unsaturated organic compounds was demonstrated and confirmed in many test chamber experiments (e.g. Wainman et al., 2000; Fan et al., 2003; Sarwar et al., 2004; Liu et al., 2004b; Vartiainen et al., 2006; Destaillats et al., 2006; Aoki & Tanabe, 2007). In two office rooms, for instance, there was an increase in the particle mass and the PNC (Fig. 6) with realistic ozone and limonene contents (Weschler et al., 2003). Ozone was in these experiments the limiting factor

The contribution of outdoor air to the amount of PM concentration in indoor air depends, in addition to the particle fraction, in particular on the ventilation behaviour of the room user, the tightness of the building envelope, the dust deposition rates indoors, the resuspension effects in the room and the coagulation behaviour of the particles. The ventilation behaviour itself is naturally dependent to a large extent on the season and the meteorology (Nazaroff, 2004). Through the windows and doors but also through leakages of the building envelope there is an exchange of air between the indoor air and the outdoor air. This results in a highly variable share of outdoor air in the amount of particle concentration in the indoor air. Other factors such as the building geometry (e.g. floor height) and location (e.g. close to a heavily trafficked road) can have a significant influence on the exposure situation. Cyrys et al. (2004) report in respect of the examination of two model rooms without an indoor activity that 75% of the indoor air contents of PM2.5 but only 43% of the PNCs can be explained by corresponding outdoor air contents. During the parallel measurements of particle distribution in rooms without indoor source and outdoor air there were in the event of closed windows and doors in the indoor environments significantly lower contents in the particle size classes than outdoors (Franck et al., 2003). Fig. 7 shows results which represent the ventilation-related influencing of

PM from outside to residential indoor environments (Riley et al., 2002).

**3.3 Cleaning activities** 

2004; Singer et al., 2006; Destaillats et al., 2006).

**3.4 Secondary organic aerosols (SOA)** 

in the formation of SOAs.

**3.5 Outdoor air as a source** 

#### **3.2 Cooking activities**

During cooking, too, fine and ultrafine particles are released. Different working groups were able to detect very high peak pollutions during cooking with electric stoves and in particular gas stoves of 100,000 to 560,000 particles /cm³ (Morawska et al., 2003; Dennekamp et al., 2001; He et al., 2004; Afshari et al., 2005; Ogulei et al., 2006; Hussein et al., 2006). The large concentration range is attributable to the different cooking activities (eg baking, roasting, frying, toasting), the use of energy, the respective cooking goods, the ventilation conditions and the room geometry. Dennekamp et al. (2001) describe PNCs of up to 110,000 or 150,000 particles/cm³ when using four electric or gas rings. Peak values of up to 590,000 ultra fine particles/cm³ were reached at the frying of bacon on a gas stove. After a short period of time the particles grew up in the indoor air and a displacement towards larger diameters. (Abt et al., 2000; Dennekamp et al., 2001; Hussein et al., 2006). After the end of the cooking activity the concentration rapidly decreases (Fig. 5). Referred to the particle mass, these activities likewise constitute a certain source. In the American PTEAM Study it was determined by means of a regression model that cooking increased the basic load of PM10 in the indoor environment by approximately 12 - 26 µg/m³ (PM2.5: approximately 13 µg/m³) (Wallace et al., 2003). Extremely high pollutions are to be expected when cooking on open fireplaces as, for instance, in third world countries (e.g. Naeher et al., 2000).

**time (minutes after start of experiment)**

Fig. 5. Particle number concentration (PNC) and PM10 in the kitchen baking with an electric oven or vacuuming

#### **3.3 Cleaning activities**

122 Air Quality – Monitoring and Modeling

During cooking, too, fine and ultrafine particles are released. Different working groups were able to detect very high peak pollutions during cooking with electric stoves and in particular gas stoves of 100,000 to 560,000 particles /cm³ (Morawska et al., 2003; Dennekamp et al., 2001; He et al., 2004; Afshari et al., 2005; Ogulei et al., 2006; Hussein et al., 2006). The large concentration range is attributable to the different cooking activities (eg baking, roasting, frying, toasting), the use of energy, the respective cooking goods, the ventilation conditions and the room geometry. Dennekamp et al. (2001) describe PNCs of up to 110,000 or 150,000 particles/cm³ when using four electric or gas rings. Peak values of up to 590,000 ultra fine particles/cm³ were reached at the frying of bacon on a gas stove. After a short period of time the particles grew up in the indoor air and a displacement towards larger diameters. (Abt et al., 2000; Dennekamp et al., 2001; Hussein et al., 2006). After the end of the cooking activity the concentration rapidly decreases (Fig. 5). Referred to the particle mass, these activities likewise constitute a certain source. In the American PTEAM Study it was determined by means of a regression model that cooking increased the basic load of PM10 in the indoor environment by approximately 12 - 26 µg/m³ (PM2.5: approximately 13 µg/m³) (Wallace et al., 2003). Extremely high pollutions are to be expected when cooking on open fireplaces as,

PNC (Number x1000/cm³)

**Baking cakes**

)

PM10 (µg/m<sup>3</sup>

**time (minutes after start of experiment)**

Fig. 5. Particle number concentration (PNC) and PM10 in the kitchen baking with an electric

**0 20 40 60 80 100 120 140 160 180 200 220 240 260**

for instance, in third world countries (e.g. Naeher et al., 2000).

**3.2 Cooking activities** 

**vacuum cleaning**

oven or vacuuming

During cleaning and in particular vacuum cleaning, an increase in coarse particles and hence in particular the particle mass is observed in indoor air (Abt et al., 2000). See also figure 5. In two US studies the contribution of cleaning activities to the PM2.5 indoor pollution was estimated at 23-32 µg/m³ (Long et al., 2000; Ferro et al., 2004). Afshari et al. (2004) describe, by contrast, merely an insignificant minor increase for ultrafine particles.

Long et al. (2000) investigated the influence of the use of commercial cleaning agents on a pine oil basis on the exposure in a living room. During the activities the PNCs rose from initially 2,000 particles/cm³ to a maximum of 190,000 particles/cm³ and the PM2.5 contents increased from 5 to 38 µg/m³. This phenomenon was explained by referring to the new particle formation and / or particle growth through oxidative processes in the indoor environment. Other working groups, too, were able to detect in test chambers in the presence of ozone and the simultaneous application of terpene-containing cleaning agents a significant increase in particle number concentrations and the particle mass (Sarwar et al., 2004; Singer et al., 2006; Destaillats et al., 2006).

#### **3.4 Secondary organic aerosols (SOA)**

Following chemical reactions of the gas and aerosol phase, so-called secondary organic aerosols (SOAs) are newly formed in indoor environments (Weschler et al. 2006). The formation of SOAs through the reaction of ozone with terpenes and other unsaturated organic compounds was demonstrated and confirmed in many test chamber experiments (e.g. Wainman et al., 2000; Fan et al., 2003; Sarwar et al., 2004; Liu et al., 2004b; Vartiainen et al., 2006; Destaillats et al., 2006; Aoki & Tanabe, 2007). In two office rooms, for instance, there was an increase in the particle mass and the PNC (Fig. 6) with realistic ozone and limonene contents (Weschler et al., 2003). Ozone was in these experiments the limiting factor in the formation of SOAs.

#### **3.5 Outdoor air as a source**

The contribution of outdoor air to the amount of PM concentration in indoor air depends, in addition to the particle fraction, in particular on the ventilation behaviour of the room user, the tightness of the building envelope, the dust deposition rates indoors, the resuspension effects in the room and the coagulation behaviour of the particles. The ventilation behaviour itself is naturally dependent to a large extent on the season and the meteorology (Nazaroff, 2004). Through the windows and doors but also through leakages of the building envelope there is an exchange of air between the indoor air and the outdoor air. This results in a highly variable share of outdoor air in the amount of particle concentration in the indoor air. Other factors such as the building geometry (e.g. floor height) and location (e.g. close to a heavily trafficked road) can have a significant influence on the exposure situation. Cyrys et al. (2004) report in respect of the examination of two model rooms without an indoor activity that 75% of the indoor air contents of PM2.5 but only 43% of the PNCs can be explained by corresponding outdoor air contents. During the parallel measurements of particle distribution in rooms without indoor source and outdoor air there were in the event of closed windows and doors in the indoor environments significantly lower contents in the particle size classes than outdoors (Franck et al., 2003). Fig. 7 shows results which represent the ventilation-related influencing of PM from outside to residential indoor environments (Riley et al., 2002).

Particles in the Indoor Environment 125

In the scientific literature a large number of measurements of particle mass concentrations in indoor air are described. Table 1 shows the results for the mass-related measurements in residential indoor environments. It must be taken into account that due to different sampling and measurement methods the results can only be compared to a limited extent.

A: Athens, B: Basel, H: Helsinki, P:

Prague; n: 186; 1996-2000

SU: 2000

13 Denmark; n: 389; 1999-2002

**Reference Concentration Description** 

Fromme et al., 2005\* 30 (Wi), 27 (Su) Berlin, Germany; n: 62; WI: 1997/98;

(H), 36 (P)

Lai et al., 2004 10 UK; n: 42; 1998-2000

Wallace et al., 2003+ 28 USA; n: 294; 7 cities Meng et al., 2005 14 USA; n: 212; 1999-2001

Baxter et al., 2007 17 USA; n: 43; 2003-2005 Héroux et al., 2010 6 Canada; n: 96; 2007 Jung et al., 2010 14 USA; n: 286; 2005-2010

Lim et al., 2011 48 Korea; n: 60; 2008

Wi: winter; Su: summer; S: smoker; NS: non smoker; \*: PM4; +: mean

Simons et al., 2007 35 (a), 10 (b) USA; n: 100 city (a), 20 suburban (b)

Li & Lin, 2003+ 39 (Wi), 37 (Su) Taiwan; urban; n: 10; 1999-2000 Chao & Wong, 2002+ 45 Hong Kong; n: 34; 1999-2000

Table 1. Median concentrations of PM2.5 in the indoor air of residences in µg/m³

Breysse et al., 2005+ 26 USA; n: 90

Link et al., 2004 19 Germany; n: 126; 2001-2002 Franck et al., 2011 32 Germany; n: 129; 2001/2002 Stranger et al., 2007+ 36 Belgium; n: 19; 2002-2003 Wichmann et al., 2010 10 Sweden; n: 29; 2003/2004 Osman et al., 2007 18 Scotland; n: 75; 2004/2005 Santen et al., 2009 3-15 Germany; n: 50; 2007 Cattaneo et al., 2011 23 Italy; n: 107; 2007/2008

**4. Occurrence of particles in indoor spaces** 

Hänninen et al., 2004+ 31 (A), 26 (B), 13

**4.1 Particles in residences** 

Raaschou-Nielsen et al.,

**America, Australia** 

**Europe** 

2011

**Asia** 

Fig. 6. Comparison between the concentrations of particles (left: number, PNC; right: mass) in an office with a limonene source and one without (modified from Weschler et al., 2003)

Fig. 7. Predicted proportion of outdoor particles in three urban residential scenarios (modified from Riley et al., 2002)

**Particle mass (µg/m³)**

**0**

Fig. 6. Comparison between the concentrations of particles (left: number, PNC; right: mass) in an office with a limonene source and one without (modified from Weschler et al., 2003)

Fig. 7. Predicted proportion of outdoor particles in three urban residential scenarios

**2**

**4**

**6**

**8**

**10**

**12**

**14**

without Limonene with Limonene

> **particle diameter [µm] 0.1-0.2 0.2-0.3 0.3-0.4 0.4-0.5**

**particle diameter [µm] 0.1-0.2 0.2-0.3 0.3-0.4 0.4-0.5**

**Particle number concentration (N/cm³)**

**0**

(modified from Riley et al., 2002)

**1000**

**2000**

**3000**

**4000**

**5000**

**6000**

#### **4. Occurrence of particles in indoor spaces**

#### **4.1 Particles in residences**

In the scientific literature a large number of measurements of particle mass concentrations in indoor air are described. Table 1 shows the results for the mass-related measurements in residential indoor environments. It must be taken into account that due to different sampling and measurement methods the results can only be compared to a limited extent.


Wi: winter; Su: summer; S: smoker; NS: non smoker; \*: PM4; +: mean

Table 1. Median concentrations of PM2.5 in the indoor air of residences in µg/m³

Particles in the Indoor Environment 127

**Wichmann et al. 2010 Almeida et al. 2010 Gabrio et al. 2007 Stranger et al. 2007 Fromme et al. 2007 (Su) Fromme et al. 2007 (Wi) Link et al. 2004 Molnar et al. 2007\* Shaughnessy et al. 2002\* Janssen et al. 2001 Fromme et al. 2005 Lahrz et al. 2003 Yang et al. 2009 Almeida et al. 2010 Fromme et al. 2007 (Su) Fromme et al. 2007 (Wi) Diapouli et al. 2007\* Son et al. 2005\* Roorda-Knape et al. 1998 Janssen et al. 1997**

Su: summer; wi: winter

determined (Oeder et al., 2011).

**µg/m³ 0 50 100 150 200 250 300 350 400 600**

Fig. 8. Particulate matter in the indoor air of schools (minimum, median, maximum); \*: mean;

Different studies showed for PM2.5 a ratio of indoor to outdoor air was in the range of 1 and a strong dependence on the outdoor air contents (Diapouli et al., 2007; Fromme et al., 2007; Wichmann et al., 2010; Guo et al., 2010). The situation is different if coarse particle fractions are considered. In a German study it was observed that 90% of the variability of the daily indoor medians of PM1 were attributable to differences between the schools and / or the days but in this way only 45% of the PM10 variants could be explained (Fromme et al., 2007). Indoor sources themselves seem to be highly significant in this connection. In order to be able to assess the contribution of different sources and outdoor air, different studies also determined the elementary and/or ionic composition of PM (Diapouli et al., 2007; Molnar et al., 2007; Fromme et al., 2008). It turned out that in particular the coarse particles did not originate from the outdoor air but that the source was the classroom itself. Examinations of the filters by EDX (energy dispersive X-ray spectroscopy) suggested that the PM10 contents were mainly composed of floor particles and other mineral substances, attrition of building materials and chalk dust (Fromme et al., 2008). A further study revealed significantly more silicate particles (36% of all particles), organic particles (29%, probably from human skin) and Ca carbonate particles (12%, probably from paper) indoors, whereas in the corresponding outdoor filters in particular Ca sulphate containing particles (38%) were

The physical activity of the pupils and the associated whirling up of suspended particles from the floor seems to be the main reason for the high PM10 concentrations in classrooms

(Fromme et al., 2007, 2008; Almeida et al., 2010; Guo et al., 2010; Oeder et al., 2011).

**PM2.5**

**PM4**

**PM10 <sup>12</sup>**

**40 3**

**N**

**79 7**

**3 74**

**33 55**

**24 73**

**10 15**

**12 3**

In different studies it could be shown that smoking is the most important influencing factor for the PM contents (e.g. Özkaynak et al., 1995; Wallace & Howard-Reed, 2002; Lai et al., 2004; Fromme et al., 2005; Breysse et al., 2005; Héroux et al., 2010; Franck et al., 2011). In Germany, the mean PM4 concentrations in smoker households amounted in winter and summer, for instance, to 109 µg/m³ and 59 µg/m³ respectively, and in non smoker households they amounted during the two seasons only to 28 µg/m³ (Fromme et al., 2005). Other important influencing factors for the indoor air contents are the season, the outdoor air, the ventilation behaviour, the age and the location of the buildings and indoor activities such as cooking, the use of ovens and the burning of incense sticks (Mönkkönen et al., 2005; Martuzevicius et al., 2008; Santen et al., 2009; Rodes et al., 2010; Héroux et al., 2010; Byun et al., 2010; Raaschou-Nielsen et al., 2011).

Studies on the ultrafine particles (as particle number concentration, PNC) in residences resulted in Germany in cities on average in 20,400 particles/cm³ (Link et al., (2004) or to between 4,000 and 25,000 particles/cm³ in a monthly median (n: 50) (Santen et al., 2009) and in an epidemiological study in 59 residences in the median of 9,000 particles/cm³ (Franck et al., 2011). McLaughlin et al. (2005) report in seven Irish residences about mean PNCs between 4,900 and 105,200 particles/cm³ with a maximum value of up to 485,300 particles/cm³. In a Swedish study three residences were investigated with mean daily values between approximately 1,800 and 8,300 particles/cm³ (Matson, 2005). The proportion of indoor to outdoor ranged between 0.7 and 2.5.

In the USA in an apartment in Boston mean PNCs of 16,000 particles/cm³ (Levy et al., 2002) and in seven Californian homes values of 9,200 to 35,000 particles/cm³ were measured (Bhangar et al., 2011). With indoor sources a mean value of 18,700 particles /cm³ (maximum: 300,000 particles/cm³) was found in a house; without indoor sources it only amounted to 2,400 particles/cm³ (maximum: 58,000 particles/cm³) (Wallace & Howard-Reed, 2002). In 36 houses in Canada mean contents of 21,600 particles/cm³ were determined during the afternoon whereas during the night the average contents were only at 6,700 particles/cm³ (Weichenthal et al., 2007). In another Canadian study median PNCs of 2,700 particles/cm³ (summer) were determined in 94 flats, 3,700 particles/cm³ (winter) and 2,600 particles/cm³ (summer) (Kearney et al., 2011).

In Australia Morawska et al. (2003) measured mean PNCs of 18,200 particles/cm³ (during indoor activities) and 12,400 particles/cm³ (without corresponding activities) when examining kitchens in 15 flats in 1999.

#### **4.2 Particles in schools**

Figure 8 shows some examples of results from schools. In most of the studies the PM2.5 contents ranged on average between 8 and 20 µg/m³. Merely in a study in 27 Belgian schools 61 µg/m³ were described, i.e. comparatively high concentrations (Stranger et al., 2007). By contrast, the PM10 contents at schools were highly variable with medians in the range of 50 - 100 µg/m³. Significantly higher contents were determined in a Greek study in which there was, however, also a high outdoor air pollution (Diapouli et al., 2007). In a European survey of 45 schools contents between 14 and 260 µg/m³ (PM10) were measured (HESE, 2006).

In different studies it could be shown that smoking is the most important influencing factor for the PM contents (e.g. Özkaynak et al., 1995; Wallace & Howard-Reed, 2002; Lai et al., 2004; Fromme et al., 2005; Breysse et al., 2005; Héroux et al., 2010; Franck et al., 2011). In Germany, the mean PM4 concentrations in smoker households amounted in winter and summer, for instance, to 109 µg/m³ and 59 µg/m³ respectively, and in non smoker households they amounted during the two seasons only to 28 µg/m³ (Fromme et al., 2005). Other important influencing factors for the indoor air contents are the season, the outdoor air, the ventilation behaviour, the age and the location of the buildings and indoor activities such as cooking, the use of ovens and the burning of incense sticks (Mönkkönen et al., 2005; Martuzevicius et al., 2008; Santen et al., 2009; Rodes et al., 2010; Héroux et al., 2010; Byun et

Studies on the ultrafine particles (as particle number concentration, PNC) in residences resulted in Germany in cities on average in 20,400 particles/cm³ (Link et al., (2004) or to between 4,000 and 25,000 particles/cm³ in a monthly median (n: 50) (Santen et al., 2009) and in an epidemiological study in 59 residences in the median of 9,000 particles/cm³ (Franck et al., 2011). McLaughlin et al. (2005) report in seven Irish residences about mean PNCs between 4,900 and 105,200 particles/cm³ with a maximum value of up to 485,300 particles/cm³. In a Swedish study three residences were investigated with mean daily values between approximately 1,800 and 8,300 particles/cm³ (Matson, 2005). The proportion

In the USA in an apartment in Boston mean PNCs of 16,000 particles/cm³ (Levy et al., 2002) and in seven Californian homes values of 9,200 to 35,000 particles/cm³ were measured (Bhangar et al., 2011). With indoor sources a mean value of 18,700 particles /cm³ (maximum: 300,000 particles/cm³) was found in a house; without indoor sources it only amounted to 2,400 particles/cm³ (maximum: 58,000 particles/cm³) (Wallace & Howard-Reed, 2002). In 36 houses in Canada mean contents of 21,600 particles/cm³ were determined during the afternoon whereas during the night the average contents were only at 6,700 particles/cm³ (Weichenthal et al., 2007). In another Canadian study median PNCs of 2,700 particles/cm³ (summer) were determined in 94 flats, 3,700 particles/cm³ (winter) and 2,600 particles/cm³

In Australia Morawska et al. (2003) measured mean PNCs of 18,200 particles/cm³ (during indoor activities) and 12,400 particles/cm³ (without corresponding activities) when

Figure 8 shows some examples of results from schools. In most of the studies the PM2.5 contents ranged on average between 8 and 20 µg/m³. Merely in a study in 27 Belgian schools 61 µg/m³ were described, i.e. comparatively high concentrations (Stranger et al., 2007). By contrast, the PM10 contents at schools were highly variable with medians in the range of 50 - 100 µg/m³. Significantly higher contents were determined in a Greek study in which there was, however, also a high outdoor air pollution (Diapouli et al., 2007). In a European survey of 45 schools contents between 14 and 260 µg/m³ (PM10) were measured

al., 2010; Raaschou-Nielsen et al., 2011).

of indoor to outdoor ranged between 0.7 and 2.5.

(summer) (Kearney et al., 2011).

**4.2 Particles in schools** 

(HESE, 2006).

examining kitchens in 15 flats in 1999.

Fig. 8. Particulate matter in the indoor air of schools (minimum, median, maximum); \*: mean; Su: summer; wi: winter

Different studies showed for PM2.5 a ratio of indoor to outdoor air was in the range of 1 and a strong dependence on the outdoor air contents (Diapouli et al., 2007; Fromme et al., 2007; Wichmann et al., 2010; Guo et al., 2010). The situation is different if coarse particle fractions are considered. In a German study it was observed that 90% of the variability of the daily indoor medians of PM1 were attributable to differences between the schools and / or the days but in this way only 45% of the PM10 variants could be explained (Fromme et al., 2007). Indoor sources themselves seem to be highly significant in this connection. In order to be able to assess the contribution of different sources and outdoor air, different studies also determined the elementary and/or ionic composition of PM (Diapouli et al., 2007; Molnar et al., 2007; Fromme et al., 2008). It turned out that in particular the coarse particles did not originate from the outdoor air but that the source was the classroom itself. Examinations of the filters by EDX (energy dispersive X-ray spectroscopy) suggested that the PM10 contents were mainly composed of floor particles and other mineral substances, attrition of building materials and chalk dust (Fromme et al., 2008). A further study revealed significantly more silicate particles (36% of all particles), organic particles (29%, probably from human skin) and Ca carbonate particles (12%, probably from paper) indoors, whereas in the corresponding outdoor filters in particular Ca sulphate containing particles (38%) were determined (Oeder et al., 2011).

The physical activity of the pupils and the associated whirling up of suspended particles from the floor seems to be the main reason for the high PM10 concentrations in classrooms (Fromme et al., 2007, 2008; Almeida et al., 2010; Guo et al., 2010; Oeder et al., 2011).

Particles in the Indoor Environment 129

offices ranged between 7 – 51 µg/m³ and 30 - 63 µg/m³, respectively. Noticeably low values resulted from the most extensive examination in 100 buildings with air conditioning systems in the USA (Burton et al., 2000). By contrast, particularly high concentrations were observed in Greek offices (Gemenetzis et al., 2006). These are attributed to the high outdoor air

Concerning the ultra fine particles, higher PNCs were observed in offices exposed to tobacco smoke than in outdoor air whereby they ranged between approximately 1,000 and 13,000 particles/cm³ in offices with air conditioning (Matson, 2005). In an Australian study a mean concentration of 6,500 particles/cm³ was measured during and 1,200 particles/cm³ after working hours (He et al., 2007) in an open plan office with ventilation and air conditioning system and smoking ban. The highest measured concentration amounted to 38,000

An overview of the exposure in pubs, restaurants and similar venues is provided by Table 3. The worldwide studies all reach the conclusion that in venues in which smoking is permitted very high concentrations have to be expected. A German study in discos (n = 10) resulted, for instance, for PM10 in a median of 1,014 µg/m³ and for PM2.5 of 869 µg/m³ (Bolte et al., 2008). In pubs (n = 18) the same working group measured medians of 210 µg/m³ (PM10) and 195 µg/m³ (PM2.5). Figure 9 shows, for instance, the PM2.5 concentration time course in three venues which were examined during the above

Results about the development of indoor air pollution in bars, restaurants and similar venues after the introduction of smoking bans are available so far to a larger extent from the USA, Italy, Ireland, Scotland and Norway (summary in Fromme et al., 2009). Overall, it turned out that a considerable reduction of the PM2.5 contents between 70 and 97%, mostly above 90%, can be achieved through the implementation of a consistent smoking ban in

On the other hand, the published results proved that through spatially not completely separated smoking areas in pubs and with ventilation systems no or only a low decrease in particle pollution is achieved. This is confirmed in a position paper by the American Society of Heating, Refrigerating and Air Conditioning Engineers which does not see ventilation systems as a useful instrument to protect from passive smoking in these venues (ASHRAE,

So far there are hardly any study results on the number of ultrafine particles. Milz et al., (2007) investigated 2 restaurants in two American cities. Whereas in non-smoker restaurants the mean contents amounted to ca. 15,000 particles/cm³, 82,000 particles/cm³ and ca. 106,000 particles/cm³ were observed in smoker rooms. Concerning ultrafine particles, smoker rooms result in a pollution of areas nearby in which smoking is banned. In Germany very high median PNCs of 221,100 particles/cm³ were measured in 4 cafés/ restaurants, 119,100 particles/cm³ in 2 bars and 289,900 particles/cm³ in 7 discos (Bolte et

concentrations and the presence of smokers.

particles/cm³ in this study.

mentioned study.

these venues alone.

2005).

al., 2008).

**4.4 Particles in hospitality venues** 

Measurements of the particle number concentration (PNC) have only been carried out so far in some cases at schools. In 36 German classrooms PNCs of 2,600 to 12,100 particles/cm³ were measured (Fromme et al., 2007). In another German study the contents ranged between 2,400 and 75,500 particles/cm³ (city) or 1,720 and 47,100 particles/cm³ (rural area) (Link et al., 2004). In a study in Greece mean PNCs of 24,000 particles/cm³ were determined during the class time at seven primary schools in Athens (Greece) which correlated well with the outdoor air contents (32,000 particles/cm³) (Diapouli et al., 2007). In an Australian study 3,100 particles/ cm³ were determined as mean value which increased within the framework of indoor activities such as cooking or cleaning of the floor surfaces to a maximum of 100,000 particles/cm³ (Guo et al., 2010). Since in school classrooms these classical sources for ultra fine particles are as a rule missing, the exposure of pupils during class time is essentially determined by the pollution of the outdoor air.

#### **4.3 Particles in offices**


Table 2 represents the mass related contents in indoor air of office buildings. The study results are difficult to compare with one another, since it was partly not mentioned whether smoking was allowed in the rooms. The median PM2.5 and PM10 values in non-smoker

\*: Mean; +: geometric mean; #: geometric mean per building; DL: detection limit; S: smoker; NS: non smoker

Table 2. Concentrations of particulate matter in the indoor air in office buildings in µg/m³

Measurements of the particle number concentration (PNC) have only been carried out so far in some cases at schools. In 36 German classrooms PNCs of 2,600 to 12,100 particles/cm³ were measured (Fromme et al., 2007). In another German study the contents ranged between 2,400 and 75,500 particles/cm³ (city) or 1,720 and 47,100 particles/cm³ (rural area) (Link et al., 2004). In a study in Greece mean PNCs of 24,000 particles/cm³ were determined during the class time at seven primary schools in Athens (Greece) which correlated well with the outdoor air contents (32,000 particles/cm³) (Diapouli et al., 2007). In an Australian study 3,100 particles/ cm³ were determined as mean value which increased within the framework of indoor activities such as cooking or cleaning of the floor surfaces to a maximum of 100,000 particles/cm³ (Guo et al., 2010). Since in school classrooms these classical sources for ultra fine particles are as a rule missing, the exposure of pupils during

Table 2 represents the mass related contents in indoor air of office buildings. The study results are difficult to compare with one another, since it was partly not mentioned whether smoking was allowed in the rooms. The median PM2.5 and PM10 values in non-smoker

Phillips et al., 1998 53 (NS) \*; 63 (S) \* France; n: 222 personal monitoring; 1995 Gemenetzis et al., 2006 103 (25- 370) Greece; 40 rooms in 2 buildings; natural

Lahrz et al., 2002 29 (5- 120) (NS) Germany; n: 25; natural ventilation; 2001 Gemenetzis et al., 2006 77 (11- 250) Greece; 40 rooms in 2 buildings; natural

Horemans et al., 2007 11 (5- 28) Belgium; n:9; natural ventilation; 2007 Burton et al., 2000 7 (1- 25) + USA ; n: 100; with AC; 1994-1998 Liu et al., 2004a 28 (3- 103) China, Peking; n: 11; 2002/2003 \*: Mean; +: geometric mean; #: geometric mean per building; DL: detection limit; S: smoker;

Table 2. Concentrations of particulate matter in the indoor air in office buildings in µg/m³

ventilated

28 (S); 1992

ventilated

**Description** 

USA, New Jersey, Pennsylvania; n: 52 (NS) and

Greece; n: 6; natural ventilation; 2006

class time is essentially determined by the pollution of the outdoor air.

**(Min-Max)** 

67 (18- 217) (S) \*

Mosqueron et al., 2005 26 (5- 265) France; n: 55; 1999/2000

107 (39- 63) (S)

Burton et al., 2000 11 (3- 35) + USA; n: 100; with AC; 1994-1998 Reynolds et al., 2001 14 to 36# USA; n: 6; with AC; 1996/1997 Liu et al., 2004a 63 (14- 166) China, Peking; n: 11; 2002/2003

**4.3 Particles in offices** 

**PM10**

**PM2.5**

NS: non smoker

**Reference Median** 

Heavner et al., 1996 30 (<DL- 98)(NS) \*

Vardavas et al., 2007 51 (39- 63)\* (NS)

offices ranged between 7 – 51 µg/m³ and 30 - 63 µg/m³, respectively. Noticeably low values resulted from the most extensive examination in 100 buildings with air conditioning systems in the USA (Burton et al., 2000). By contrast, particularly high concentrations were observed in Greek offices (Gemenetzis et al., 2006). These are attributed to the high outdoor air concentrations and the presence of smokers.

Concerning the ultra fine particles, higher PNCs were observed in offices exposed to tobacco smoke than in outdoor air whereby they ranged between approximately 1,000 and 13,000 particles/cm³ in offices with air conditioning (Matson, 2005). In an Australian study a mean concentration of 6,500 particles/cm³ was measured during and 1,200 particles/cm³ after working hours (He et al., 2007) in an open plan office with ventilation and air conditioning system and smoking ban. The highest measured concentration amounted to 38,000 particles/cm³ in this study.

#### **4.4 Particles in hospitality venues**

An overview of the exposure in pubs, restaurants and similar venues is provided by Table 3. The worldwide studies all reach the conclusion that in venues in which smoking is permitted very high concentrations have to be expected. A German study in discos (n = 10) resulted, for instance, for PM10 in a median of 1,014 µg/m³ and for PM2.5 of 869 µg/m³ (Bolte et al., 2008). In pubs (n = 18) the same working group measured medians of 210 µg/m³ (PM10) and 195 µg/m³ (PM2.5). Figure 9 shows, for instance, the PM2.5 concentration time course in three venues which were examined during the above mentioned study.

Results about the development of indoor air pollution in bars, restaurants and similar venues after the introduction of smoking bans are available so far to a larger extent from the USA, Italy, Ireland, Scotland and Norway (summary in Fromme et al., 2009). Overall, it turned out that a considerable reduction of the PM2.5 contents between 70 and 97%, mostly above 90%, can be achieved through the implementation of a consistent smoking ban in these venues alone.

On the other hand, the published results proved that through spatially not completely separated smoking areas in pubs and with ventilation systems no or only a low decrease in particle pollution is achieved. This is confirmed in a position paper by the American Society of Heating, Refrigerating and Air Conditioning Engineers which does not see ventilation systems as a useful instrument to protect from passive smoking in these venues (ASHRAE, 2005).

So far there are hardly any study results on the number of ultrafine particles. Milz et al., (2007) investigated 2 restaurants in two American cities. Whereas in non-smoker restaurants the mean contents amounted to ca. 15,000 particles/cm³, 82,000 particles/cm³ and ca. 106,000 particles/cm³ were observed in smoker rooms. Concerning ultrafine particles, smoker rooms result in a pollution of areas nearby in which smoking is banned. In Germany very high median PNCs of 221,100 particles/cm³ were measured in 4 cafés/ restaurants, 119,100 particles/cm³ in 2 bars and 289,900 particles/cm³ in 7 discos (Bolte et al., 2008).

Particles in the Indoor Environment 131

**PM10 (µg/m³)**

**0**

week (e.g. Lee et al., 2010).

**4.5 Particles in transportation systems** 

**4.5.1 PM in aboveground transportation systems** 

**200**

al., 2008)

al., 2005).

**400**

**600**

**800**

**1000**

**1200**

**1400**

**1600**

**1800**

Discotheque

Pub Café

**2000**

**time (minutes after start of experiment)** 

Fig. 9. Time course of PM2.5 in three hospitality venues in Germany (modified from Bolte et

The contents of PM2.5 in above ground buses and cars are shown in Figure 10. The highest contents in cars and buses were observed in Asian cities; merely in one study in Mexico City, in Boston (Levy et al., 2002) and in Peru (Han et al., 2005) similar high concentrations were described. The other studies, in particular in Europe and Australia, refer, by contrast, to a mean exposure level for PM2.5 of approximately 10 - 40 µg/m³; as a rule the concentrations are significantly higher indoors than in the ambient air. There was a dependency of the indoor air contents on the outdoor levels, the time of day and the day of

Table 4 shows the results of the measurements of ultrafine particles in cars and buses. The mean PNC ranges between 10,000 and 50,000 particles/cm³. By contrast, very high contents were described by Kaur et al. (2005) in the City of London which ranged on average between 90,000 and 100,000 particles/cm³. During a drive on the freeway with an open window high concentrations were likewise determined (Eiguren-Fernandez et al., 2005). Under special conditions, eg a diesel truck ahead, short term peak concentrations of up to 500,000 particles/cm³ were observed (Abraham et al., 2002; Eiguren-Fernandez et

**0 20 40 60 80 100 120 140 160 180 200 220 240**


\*: mean; +: PM4 or respirable particulate matter (RPM)

Table 3. Concentrations of particulate matter (PM2.5) in indoor air of hospitality venues in µg/m³

75 (0- 277) + Restaurants; Schwitzerland

**Reference Median (Min-Max) Description**

201 (62- 391) + Restaurants; UK

Bolte et al., 2008 164 (55- 570) 11 restaurants, cafes; Germany; 2005/2006 203 (103- 1250) 7 pubs and bars; Germany; 2005/2006 869 (291- 4475) 10 discotheques; Germany; 2005/2006 Vardavas et al., 2007 268 (19- 612)\* 31 bars, pubs, cafes, clubs; Greece; 2006

Daly et al., 2011 83 (51- 108)\* 70 bars, cafes, restaurants; Switzerland; 2008

82 (0- 768) + Bars; USA; 1996/1997

(47- 253) 4 bars; Canada

Bohanon et al., 2003 194 (56- 312)+ Restaurants; France

Gee et al., 2006 94\*+ 59 pubs; England; 2001

Semple et al., 2010 197 (8- 902) 42 bars; Scotland; 2006 92 (5- 1005) 52 bars; England; 2007 184 (16- 872) 12 bars; Wales; 2007 Rosen et al., 2007 465 (66- 862)+ 6 bars, pubs; Israel; 2007 52 (18- 557)+ 8 cafes; Israel; 2007

Maskarinec et al., 2000 66 (0- 233) + Restaurants; USA; 1996/1997

Baek et al., 1997 159 (33 - 475)+ 6 restaurants; Korea; 1994/1995 Lee et al., 1999 400 - 1760 3 restaurants; China; 1996/1997

Lee et al., 2010 92 (17- 565)\* 55 restaurants; 7 countries; 2008/2009 114 (14- 565)\* 35 cafes; 7 countries; 2008/2009 191 (33- 748)\* 34 bars, clubs; 7 countries; 2008/2009

169 (4- 881)\* 44 entertainment venues; 7 countries; 2008/2009

Table 3. Concentrations of particulate matter (PM2.5) in indoor air of hospitality venues in

Brauer et al., 2000 (11- 163) 11 restaurants; Canada

Repace et al., 2006 178 (43- 323) 6 pubs; USA; 2003 Connolly et al., 2005 206 (23- 727) 28 locations, USA; 2005 Brennan et al., 2010 61 (6- 338) 19 pubs; Australia; 2007 Jiang et al., 2011 63 (18- 183) 36 casinos, USA; 2008

Bohanon et al., 2003 194 (0- 611) + Restaurants; Japan 107 (54- 172) + Restaurants; Korea

\*: mean; +: PM4 or respirable particulate matter (RPM)

Edwards et al., 167 (54- 1395) 33 pubs, with cooking; UK; 2004 2006a 217 (15- 1227) 31 Pubs; no cooking; UK; 2004 Goodman et al., 2007 35,5 \* 42 pubs; Ireland; 2004/2005 Valente et al., 2007 119 40 locations; Italy; 2005 Schneider et al., 2008 173 (22- 831) 38 restaurants; Germany; 2005 131 (24- 1029) 20 cafes; Germany; 2005 378 (144- 2022) 11 bars; Germany; 2005

**Europe** 

**America, Australia**

**Asia** 

µg/m³

Fig. 9. Time course of PM2.5 in three hospitality venues in Germany (modified from Bolte et al., 2008)

#### **4.5 Particles in transportation systems**

#### **4.5.1 PM in aboveground transportation systems**

The contents of PM2.5 in above ground buses and cars are shown in Figure 10. The highest contents in cars and buses were observed in Asian cities; merely in one study in Mexico City, in Boston (Levy et al., 2002) and in Peru (Han et al., 2005) similar high concentrations were described. The other studies, in particular in Europe and Australia, refer, by contrast, to a mean exposure level for PM2.5 of approximately 10 - 40 µg/m³; as a rule the concentrations are significantly higher indoors than in the ambient air. There was a dependency of the indoor air contents on the outdoor levels, the time of day and the day of week (e.g. Lee et al., 2010).

Table 4 shows the results of the measurements of ultrafine particles in cars and buses. The mean PNC ranges between 10,000 and 50,000 particles/cm³. By contrast, very high contents were described by Kaur et al. (2005) in the City of London which ranged on average between 90,000 and 100,000 particles/cm³. During a drive on the freeway with an open window high concentrations were likewise determined (Eiguren-Fernandez et al., 2005). Under special conditions, eg a diesel truck ahead, short term peak concentrations of up to 500,000 particles/cm³ were observed (Abraham et al., 2002; Eiguren-Fernandez et al., 2005).

Particles in the Indoor Environment 133

bus car


bus car taxi

bus car

bus car

Leeds, UK

London, UK; 2003

Boston, USA; 2000

car Los Angeles, USA, car; AC: air

small streets; Y: freeway

condition, nAC: windows open; X:

Sydney; Australia; 40 trips; 2004

**Reference Mean (Min-Max) Description** 

Diapouli et al., 2007 94 (25-217) car Athens, Greece; through city

Abraham et al., 2002 30 b (4-190) New York, USA; 3 city routes

Rim et al., 2008 6-35 bus Austin; USA; 6 busses; 2006 Wallace & Ott, 2011 29-34 (-) car USA; 17 trips; 2005-2009

Zhang & Zhu, 2010 7.3-34 bus Texas; USA; school buses; 2008

Table 4. Particle number concentrations (PNC) in indoor air of transportation systems (10³ of

Studies in this micro environment show that the exposure is significantly above the values measured in above ground transportation systems. In the Berlin underground mean PM4 contents of 141 (124-169 µg/m³) were measured in winter and 153 µg/m³ (121-176 µg/m³) in summer (Fromme et al., 1998). Similar results were obtained in the London underground with mean PM2.5 concentrations of 247 µg/m³ (105 – 371 µg/m³) (summer) and 157 µg/m³ (12 – 263 µg/m³) (winter) (Adams et al., 2001). In a more recent study conducted in London mean values of 180 - 200 µg/m³ (PM2.5) were measured (Hurley et al., 2004). By contrast, significantly lower mean pollutions were observed in Boston (70 µg/m³ PM2.5), Los Angeles (13.7 µg/m³) and Helsinki (21 µg/m³) (Levy et al., 2002; Aarnio et al., 2005; Kam et al., 2011). In the underground of Mexico City the mean values amounted to 61 µg/m³ (31 - 99 µg/m³)

Geiss et al., 2010 16 (8-30) car Italy; 18 cars; 2009

Dennekamp et al., 2002 53 (-) a bus Aberdeen, UK; n: 11

58 (8-282)

100 (37-152) 88 (52-114)

~39 (11-83)

25 (X, AC) 55 (X, nAC) 69 (Y, AC) 246 (Y,nAC)

9

**4.5.2 PM in underground transportation systems** 

**Europe** 

2005

2005

Mackay, 2004 44 (10-143)

Kaur et al., 2005 101 (65-159)

Levy et al., 2002 ~32 (12-80)

Krausse & Mardaljevic,

**North America, Australia** 

Eiguren-Fernandez et al.,

Knibbs & de Dear, 2010 11

a: median; b: mean of three cycles

particles/cm³)

Fig. 10. Concentrations of PM2.5 in cars and buses (\*: PM4) (minimum, median, maximum)

Overall, the exposure in the indoor environment of transportation systems is influenced by many different factors such as ventilation, quality of driving, traffic volume and traffic composition, built-up area and meteorology. Car passengers seem to be exposed to a slightly higher PM2.5 and PNCs than cyclists (Adams et al., 2001; Kaur et al., 2005), whereas they are stated as the same or only slightly different in other studies (Gulliver & Briggs, 2004; Boogaard et al., 2009; Zuurbier et al., 2010; Int Panis et al., 2010). In this connection it must, however, be taken into account that due to the approximately 2 – 4.5 fold higher breathing volume cyclists have a significantly higher inhaled dose compared to car and bus users (Zuurbier et al., 2010; Int Panis et al., 2010).

In addition, air conditioning and filter systems, which can have a major influence on the contents in the indoor environment of transportation systems depending on the quality and the separation level, must be taken into account (Rim et al., 2008). Studies in an urban environment also showed a higher exposure for pedestrians than during car use or compared to the general outdoor air pollution (Kaur et al., 2005).

A special pollution situation results from passive smoking exposure. During a drive in Wellington with a fully opened side window mean PM2.5 contents of 169 µg/m³ (maximum: 217 µg/m³) were measured while smoking a cigarette (Edwards et al., 2006b). With a closed window the mean values were 2,962 µg/m³ (maximum: 3,645 µg/m³) and in a Canadian study mean PM2.5 contents between 790 and 4,626 µg/m³ (maximum: 7,635 µg/m³) were determined (Sendzik et al., 2006). Rees & Connolly (2006) determined in 45 measurements during smoking with closed windows mean PM2.5 values of 271 µg/m³ (maximum: approximately 500 µg/m³) and with opened side windows approximately 50 µg/m³ (maximum: approximately 100 µg/m³). Liu & Zhu (2010) observed inside cars the tenfold PNC and 120-fold PM2.5 contents compared to outdoor air.

**µg/m³ 0 20 40 60 80 100 120 140 160**

Fig. 10. Concentrations of PM2.5 in cars and buses (\*: PM4) (minimum, median, maximum)

Overall, the exposure in the indoor environment of transportation systems is influenced by many different factors such as ventilation, quality of driving, traffic volume and traffic composition, built-up area and meteorology. Car passengers seem to be exposed to a slightly higher PM2.5 and PNCs than cyclists (Adams et al., 2001; Kaur et al., 2005), whereas they are stated as the same or only slightly different in other studies (Gulliver & Briggs, 2004; Boogaard et al., 2009; Zuurbier et al., 2010; Int Panis et al., 2010). In this connection it must, however, be taken into account that due to the approximately 2 – 4.5 fold higher breathing volume cyclists have a significantly higher inhaled dose compared to car and bus

In addition, air conditioning and filter systems, which can have a major influence on the contents in the indoor environment of transportation systems depending on the quality and the separation level, must be taken into account (Rim et al., 2008). Studies in an urban environment also showed a higher exposure for pedestrians than during car use or

A special pollution situation results from passive smoking exposure. During a drive in Wellington with a fully opened side window mean PM2.5 contents of 169 µg/m³ (maximum: 217 µg/m³) were measured while smoking a cigarette (Edwards et al., 2006b). With a closed window the mean values were 2,962 µg/m³ (maximum: 3,645 µg/m³) and in a Canadian study mean PM2.5 contents between 790 and 4,626 µg/m³ (maximum: 7,635 µg/m³) were determined (Sendzik et al., 2006). Rees & Connolly (2006) determined in 45 measurements during smoking with closed windows mean PM2.5 values of 271 µg/m³ (maximum: approximately 500 µg/m³) and with opened side windows approximately 50 µg/m³ (maximum: approximately 100 µg/m³). Liu & Zhu (2010) observed inside cars the tenfold

**Car**

**USA**

**USA UK UK UK Germany**

**Italy Netherland Australia Australia UK**

**Bus**

**Taiwan Finland Australia USA China Mexico USA China UK UK**

**Fromme et al. 1998\* Gee et al. 1999 Adams et al. 2001 Kaur et al. 2005 Levy et al. 2002 Riediker et al. 2004 Chertok et al. 2004 Kaur et al. 2005 Knibbs & de Dear 2010 Boogaard et al. 2009 Geiss et al. 2010 Adams et al. 2001 Kaur et al. 2005 Chan et al. 2002 Sabin et al. 2004 Gomez-Perales et al. 2004 Chan et al. 2002 Rim et al. 2008 Knibbs & de Dear 2010 Asmi et al. 2009 Hsu & Huang 2009**

users (Zuurbier et al., 2010; Int Panis et al., 2010).

compared to the general outdoor air pollution (Kaur et al., 2005).

PNC and 120-fold PM2.5 contents compared to outdoor air.


a: median; b: mean of three cycles

Table 4. Particle number concentrations (PNC) in indoor air of transportation systems (10³ of particles/cm³)

#### **4.5.2 PM in underground transportation systems**

Studies in this micro environment show that the exposure is significantly above the values measured in above ground transportation systems. In the Berlin underground mean PM4 contents of 141 (124-169 µg/m³) were measured in winter and 153 µg/m³ (121-176 µg/m³) in summer (Fromme et al., 1998). Similar results were obtained in the London underground with mean PM2.5 concentrations of 247 µg/m³ (105 – 371 µg/m³) (summer) and 157 µg/m³ (12 – 263 µg/m³) (winter) (Adams et al., 2001). In a more recent study conducted in London mean values of 180 - 200 µg/m³ (PM2.5) were measured (Hurley et al., 2004). By contrast, significantly lower mean pollutions were observed in Boston (70 µg/m³ PM2.5), Los Angeles (13.7 µg/m³) and Helsinki (21 µg/m³) (Levy et al., 2002; Aarnio et al., 2005; Kam et al., 2011). In the underground of Mexico City the mean values amounted to 61 µg/m³ (31 - 99 µg/m³)

Particles in the Indoor Environment 135

Asmi, E.; Antola, M.; Yli-Tuomi, T.; Jantunen, M.; Aarnio, P.; Mäkelä, P.; Hillamo, R. &

Baxter, L.K.; Clougherty, J.E.; Laden, F. & Levy, J.I. (2007). Predictors of concentrations of

cocioeconimic status urban homes. *J Expos Sci Environ Epidemiol*, 17, 433-444. Birmili, W.; Kinnersley, R.P. & Baker, J. (2003). Factors influencing human exoposure to fine

Boogaard, H.; Borgman, F.; Kamminga, J. & Hoek, G. (2009). Exposure to ultrafine and fine

Bohanon, H.R.; Piadé, J.J.; Schorp, M.K. & Saint-Jalm, Y. (2003). An international survey of

Bolte, G.; Heitmann, D.; Kiranoglu, M.; Schierl, R.; Diemer, J.; Körner, W. & Fromme, H.

Brauer, M.; Hirtle, R.; Lang, B. & Ott, W. (2000). Assessment of indoor fine aerosol

Brennan, E.; Cameron, M.; Warne, C.; Durkin, S.; Borland, R.; Travers, M.J.; Hyland, A. &

Breysse, P.N.; Buckley, T.J.; Williams, D.A.; Beck, C.M.; Jo, S.J.; Merriman, B.;

Burton, L.E.; Girman, J.G. & Womble, S.E. (2000). Airborne particulate matter within 100

Byun, H.; Bae, H.; Kim, D.; Shin, H. & Yoon, C. (2010). Effects of socioeconomic factors and

Cattaneo, A.; Peruzzo, C.; Garramone, G.; Urso, P.; Ruggeri, R.; Carrer, P. & Cavallo, D.M.

Chan, L.Y.; Lau, W.L.; Lee, S.C. & Chan, C.Y. (2002). Commuter exposure to particulate

Chao, Y.H.; Tung, T.C.W. & Burnett, J. (1998). Influence of different indoor activities on the indoor particulate levels in residential buildings. *Indoor Built Environ*, 7, 110-121. Chao, C.Y. & Wong, K.K. (2002). Residential indoor PM10 and PM2.5 in Hong Kong and the

trams in Helsinki, Finland. *Sci Total Environ*, 2860–2867.

Aerosol Conference.

*Tobacco Res*, 12, 271–277.

*Healthy Buildings*, 1, 157-162.

*Arch Occup Environ Health*, 83, 867–878.

*Expo Anal Environ Epidemiol*, 13, 378-392.

discotheques. *J Expo Sci Environ Epidemiol*, 18, 262-271.

nephelometer. *J Expo Anal Environ Epidemiol*, 10, 136-144.

children in inner-city Baltimore. *Environ Res*, 98, 167-176.

structures in Lodi Province, Italy. *Indoor Air*, accepted.

elemental composition. *Atmos Environ*, 36, 265-277.

4234–4242.

Hämeri, K. (2009). Driver and passenger exposure to aerosol particles in buses and

nitrogen dioxide, fine particulate matter, and particle constituents inside lower

and ultrafine particles in a city centre office. Abstract presented at the European

particles and noise during cycling and driving in 11 Dutch cities. *Atmos Environ*, 43,

indoor air quality, ventilation, and smoking activity in restaurants: a pilot study. *J* 

(2008). Exposure to environmental tobacco smoke in German restaurants, pubs and

contributions from environmental tobacco smoke and cooking with a portable

Wakefield, M.A. (2010) Secondhand smoke drift: Examining the influence of indoor smoking bans on indoor and outdoor air quality at pubs and bars. *Nicotine &* 

Kanchanaraksa, S.; Swartz, L.J.; Callahan, K.A.; Butz, A.M.; Rand, C.S.; Diette, G.B.; Krishnan, J.A.; Moseley, A.M.; Curtin-Brosnan, J.; Durkin, N.B. & Eggleston, P.A. (2005). Indoor exposures to air pollutants and allergens in the homes of asthmatic

randomily selected office buildings in the united states (BASE), *Proceedings of* 

human activities on children's PM10 exposure in inner-city households in Korea. *Int* 

(2011) Airborne particulate matter and gaseous air pollutants in residential

matter in public transportation modes in Hong Kong. *Atmos Environ*, 36, 3363-3373.

(Gómez-Perales et al., 2004). Measurements in the Seoul underground and in a Chinese city, resulted in concentrations of 148 µg/m³ (Sohn et al., 2005) and 67 µg/m³ (26-123 µg/m³), respectively (Chan et al., 2002).

In the Berlin underground stations the PM4 contents ranged between 128 and 311 µg/m³ during operation (Fromme et al., 1998), and in London the average PM2.5 contents were 270 - 480 µg/m³ (Hurley et al., 2004). In Boston 130 µg/m³ were measured at the underground station (Levy et al., 2002). On the other hand the contents in Taipei were on average only 25- 40 µg/m³ (Cheng & Yan, 2011) and in Paris (Raut et al., 2009) the contents were 61 µg/m³ (normal hours) and 93 µg/m³ (rush hours) correspondingly lower.

At present only a few measurements on ultrafine particles are available. In three underground lines in London 17,000 – 23,000 particles/cm³ (>50 nm) were measured on average whereas on the platforms of the three underground stations the average contents determined were 14,000 – 29,000 particles/cm³ (Hurley et al., 2004). Similar results were reached by measurements at the underground of Boston with mean values of approximately 21,000 particles/cm³ (Levy et al., 2002) and in Helsinki, with 27,000 particles/cm³ (14,000- 50,000 Pt./cm³) (Aarnio et al., 2005).

#### **5. References**


(Gómez-Perales et al., 2004). Measurements in the Seoul underground and in a Chinese city, resulted in concentrations of 148 µg/m³ (Sohn et al., 2005) and 67 µg/m³ (26-123 µg/m³),

In the Berlin underground stations the PM4 contents ranged between 128 and 311 µg/m³ during operation (Fromme et al., 1998), and in London the average PM2.5 contents were 270 - 480 µg/m³ (Hurley et al., 2004). In Boston 130 µg/m³ were measured at the underground station (Levy et al., 2002). On the other hand the contents in Taipei were on average only 25- 40 µg/m³ (Cheng & Yan, 2011) and in Paris (Raut et al., 2009) the contents were 61 µg/m³

At present only a few measurements on ultrafine particles are available. In three underground lines in London 17,000 – 23,000 particles/cm³ (>50 nm) were measured on average whereas on the platforms of the three underground stations the average contents determined were 14,000 – 29,000 particles/cm³ (Hurley et al., 2004). Similar results were reached by measurements at the underground of Boston with mean values of approximately 21,000 particles/cm³ (Levy et al., 2002) and in Helsinki, with 27,000 particles/cm³ (14,000-

Aarnio, P.; Yli-Tuomi, T.; Kousa, A.; Mäakelä, T.; Hirsikko, A.; Hämeri, K.; Räisänen, M.;

Abraham, J.L.; Siwinski, G. & Hunt, A. (2002). Ultrafine particulate exposures in indoor,

Abt, E.; Suh, H.H.; Allen, G. & Koutrakis, P. (2000). Characterization of indoor particle

Adams, H.S.; Nieuwenhuijsen, M.J.; Colvile, R.N.; McMullen, M.A.S. & Khandelwal, P.

Afshari, A.; Matson, U. & Ekberg, LE. (2005). Characterization of indoor sources of fine and

Almeida, S.M.; Canha, N.; Silva, A.; do Carmo Freitas, M.; Pegas, P.; Alves, C.; Evtyugina,

Aoki, T. & Tanabe, S. (2007). Generation of sub-micron particles and secondary pollutants from building materials by ozone reaction. *Atmos Environ*, 41, 3139-3150. ASHRAE (American Society of Heating, Refrigerating and Air-Conditioning Engineers)

microenvironments, London, UK. *Sci Total Environ*, 279, 29-44.

Board of Directors. June 30, 2005, online: www.ashrae.org.

Hillamo, R.; Koskentalo, T. & Jantunen, M. (2005). The concentrations and composition of and exposure to fine particles (PM2.5) in the Helsinki subway

outdoor, personal and mobile environments: effects of diesel, traffic, pottery kiln, cooking and HEPA filtration on micro-environmental particle number

sources: a study conducted in the metropolitan Boston area. *Environ Health Perspect*,

(2001). Fine particle (PM2.5) personal exposure levels in transport

ultrafine particles: a study conducted in a full-scale chamber. *Indoor Air*, 15, 141-

M. & Pio, C.A. (2010). Children exposure to atmospheric particles in indoor of Lisbon primary schools. *Atmos Environ*, corrected proof,

(2005). Environmental Tobacco Smoke -Position Document Approved by ASHRAE

(normal hours) and 93 µg/m³ (rush hours) correspondingly lower.

respectively (Chan et al., 2002).

50,000 Pt./cm³) (Aarnio et al., 2005).

108, 35-44.

150.

system. *Atmos Environ*, 39, 5059–5066.

doi:10.1016/j.atmosenv.2010.11.052.

concentration. *Ann Occup Hyg*, 46 (suppl 1), 406-411.

**5. References** 


Particles in the Indoor Environment 137

Ferro, A.R.; Kopperud, R.J. & Hildemann, L.M. (2004). Elevated personal exposure to

Fine, P.M.; Cass, G.R. & Simoneit, B.R.T. (1999). Characterization of fine particle emissions

Franck, U.; Herbarth, O.; Wehner, B.; Wiedensohler, A. & Manjarrez, M. (2003). How do the

Franck, U.; Herbarth, O.; Röder, S.; Schlink, U.; Borte, M.; Diez, U.; Krämer, U. & Lehmann,

Fromme, H.; Oddoy, A.; Lahrz, T.; Krause, M. & Piloty, M. (1998). Polycyclic aromatic

Fromme, H.; Lahrz, T.; Hainsch, A.; Oddoy, A.; Piloty, M. & Rüden, H. (2005). Elemental

nursery schools and outdoor air in Berlin (Germany). *Indoor Air*, 15, 335-341. Fromme, H.; Twardella, D.; Dietrich, S.; Heitmann, D.; Schierl, R.; Liebl, B. & Rüden, H.

Fromme, H.; Diemer, J.; Dietrich, S.; Cyrys, J.; Heinrich, J.; Lang, W.; Kiranoglu, M. &

Fromme, H.; Kuhn, J. & Bolte, G. (2009). Secondhand smoke in hospitality venues. Exposure,

Gabrio, T.; Volland, G.; Baumeister, I.; Bendak, J.; Flicker-Klein, A.; Gickeleiter, M.; Kersting,

Gee, I.L. & Raper, D.W. (1999). Commuter exposure to respirable particles inside buses and

Gee, I.L; Watson, A.F.; Carrington, J.; Edwards, P.R.; van Tongeren, M.; McElduff, P. &

Public Health White Paper proposals go far enough? *J Public Health*, 28, 17-23. Geiss, O.; Barrero-Moreno, J.; Tirendi, S. & Kotzias, D. (2010). Exposure to particulate matter

Gemenetzis, P.; Moussas, P. & Arditsoglou, (2006). A. Mass concentration and elemental

Gomez-Perales, J.E.; Colvile, R.N.; Nieuwenhuijsen, M.J.; Fernandez-Bremauntz, A.;

in vehicle abins of private cars. *Aerosol Air Qual Res*, 10, 581–588.

from burning church candles. *Environ Sci Technol*, 33, 2352-2362.

dependent. *Sci Total Environ*, 409, 1621–1631.

and a subway train. *Sci Total Environ*, 217, 165-173.

Munich and surrounding. *Atmos Environ*, 41, 854-866.

*Gesundheitswesen*, 71, 242-257 [in German].

*Gefahrstoffe - Reinhaltung der Luft*, 67, 96-102.

northern Greece. *Atmos Environ*, 40, 3195-3206.

Mexico City. *Atmos Environ*, 38, 1219-1229.

by bicycle. *Sci Total Environ*, 235, 403-405.

*Epidemiol*, 14, S34-S40.

13, 174-181.

particulate matter from human activities in a residence. *J Expo Anal Environ* 

indoor size distributions of airborne submicron and ultrafine particles in the absence of significant indoor sources depend on outdoor distributions? *Indoor Air*,

I. (2011). Respiratory effects of indoor particles in young children are size

hydrocarbons (PAH) and diesel engine emission (elemental carbon) inside a car

carbon and respirable particulate matter in the indoor air of apartments and

(2007). Particulate matter in the indoor air of classrooms – exploratory results from

Twardella, D. (2008). Chemical and morphological properties of particulate matter (PM10, PM2.5) from indoor of schools and outdoor air. *Atmos Environ*, 42, 6597–6605.

body burden, economic and health aspects in conjugation with smoking bans. *Das* 

G.; Maisner, V. & Zöllner, I. (2007). Messung von Feinstäuben in Innenräumen.

Edwards, R.E. (2006). Second-hand smoke levels in UK pubs and bars: do the English

composition of indoor PM2,5 and PM10 in University rooms in Thessaloniki,

Gutierrez-Avedoy, V.J.; Paramo-Figueroa, V.H.; Blanco-Jimenez, S.; Bueno-Lopez, E.; Mandujano, F.; Bernabe-Cabanillas, R. & Ortiz-Segovia, E. (2004). Commuters` exposure to PM2.5, CO, and benzene in public transport in the metropolitan area of


Chen, C. & Zhao, B. (2011). Review of relationship between indoor and outdoor particles: I/O ratio, infiltration factor and penetration factor. *Atmos Environ*, 45, 275-288. Cheng, YH. & Yan, J.W. (2011). Comparisons of particulate matter, CO, and CO2 levels in

Chertok, M.; Voukelatos, A.; Sheppeard, V. & Rissel, C. (2004). Comparison of personal

Connolly, G.N.; Carpenter, C.; Alpert, H.R.; Skeer, M. & Travers, M. (2005). Evaluation of the

Cyrys, J.; Pitz, M.; Bischof, W.; Wichmann, H.E. & Heinrich, J. (2004). Relationship between

Daly, B.J.; Schmid, K. & Riediker, M. (2011). Contribution of fine particulate matter sources to indoor exposure in bars, restaurants, and cafes. *Indoor Air*, 20, 204–212. Dennekamp, M.; Howarth, S.; Dick, C.A.J.; Cherrie, W.; Donaldson, K. & Seaton, A. (2001).

Dennekamp, M.; Mehenni, O.; Cherrie, J. & Seaton, A. (2002). Exposure to ultrafine particles and PM2.5 in different microenvironments. *Ann Occup Hyg*, 46 (suppl. 1), 412-414. Dermentzoglou, M.; Manoli, E.; Voutas, D. & Samara, C. (2003). Sources and patterns of

Destaillats, H.; Lunden, M.M.; Singer, B.C.; Coleman, B.K.; Hodgson, A.T.; Weschler, C.J. &

Diapouli, E.; Chaloulakou, A.; Mihalopoulos, N. & Spyrellis, N. (2008). Indoor and outdoor

Edwards, R.; Hasselholdt, C.P.; Hargreaves, K.; Probert, C.; Holford, R.; Hart, J.; Van

Edwards, R.; Wilson, N. & Pierse, N. (2006b). Highly hazardous air quality associated with smoking in cars: New Zealand pilot study. NZMJ, 119, 27 October 2006. Eiguren-Fernandez, A.; Miguel, A.H.; Zhu, Y.F. & Hering, S.V. (2005). In-cabin passenger

Fan, Z.H.; Lioy, P.; Weschler, C.J.; Fiedler, N.; Kipen, H. & Zhang, J.F. (2003). Ozone-

matter of Greek houses. *Fresenius Environ Bull*, 12, 1511-1519.

prepared for NSW Department of Health, Australia.

*Atmos Environ*, 45, 4882-4891.

*Occup Environ Med*, 58, 511-516.

*Technol*, 40, 4421-4428.

West England. *BMC Public Health*, 6, 42

indoor conditions. *Environ Sci Technol*, 37, 1811-1821.

*Assess*, 136, 13-20.

1763-1767. Peking.

14, 275-284.

underground and ground-level stations in the Taipei mass rapid transit system.

exposures to air pollutants by commuting mode in Sydney. BTEX & NO2. Report

Massachusetts smoke-free workplace law. A preliminary report of the Division of Public Health Practice, Harvard School of Public Health, Tobacco Research Program. Online: http://www.hsph.harvard.edu/php/pri/tcrtp/Smoke-free\_Workplace.pdf.

indoor and outdoor levels of fine particle mass, particle number concentrations and black smoke under different ventilation conditions. *J Expo Anal Environ Epidemiol*,

Ultrafine particles and nitrogen oxides generated by gas and electric cooking.

polycyclic aromatic hydrocarbons and heavy metals in fine indoor particulate

Nazaroff, W.W. (2006). Indoor secondary pollutants from household product emissions in the presence of ozone: a bench-scale chamber study. *Environ Sci* 

PM mass and number concentrations at schools in the Athens area. *Environ Monit* 

Tongeren, M. & Watson, A.F.R. (2006a). Levels of second hand smoke in pubs and bars by deprivation and food-serving status: a cross-sectional study from North

exposure to ultrafine and nano-particles during daily commute in Los Angeles roads and freeways: evaluation of a HEPA filtration system. *Proceedings: Indoor Air*,

initiated reactions with mixtures of volatile orghanic compounds under simulated


Particles in the Indoor Environment 139

Hussein, T.; Hruska, A.; Dohanyosová, P.; Dzumbová, L.; Hemerka, J.; Kulmala, M.; Smolík,

Int Panis, L.; de Geus, B.; Vandenbulcke, G.; Willems, H.; Degraeuwe, B.; Bleux, N.; Mishra,

Jetter, J.J.; Guo, Z.; McBrian, J.A. & Flynn, M.R. (2002). Characterization of emissions from

Jiang, R.T.; Cheng, K.C.; Acevedo-Bolton; V.; Klepeis, N.E.; Repace, J.L.; Ott, W.R. &

Jung, K.H.; Patel, M.M.; Moors, K.; Kinney, P.L.; Chillrud, S.N.; Whyatt, R.; Hoepner, L.;

Kam, W.; Cheung, K.; Daher, N. & Sioutas, C. (2011). Particulate matter (PM) concentrations

Kaur, S.; Nieuwenhuijsen, M. & Colvile, R. (2005). Personal exposure of street canyon

Kearney, J.; Wallace, L.; MacNeill, M.;Xuc, X.; Van Ryswyk, K.; Youa, H.; Kulka, R. &

Knibbs, L.D. & de Dear, R.J. (2011). Exposure to ultrafine particles and PM2.5 in four Sydney

Krausse, B. & Mardaljevic, J. (2005). Patterns of driver´s exposure to particulate matter. In:

Lahrz, T.; Piloty, M.; Pfeiler, P. & Honigmann, I. (2002). Messungen von Schadstoffen an

Lai, H.K.; Kendall, M.; Ferrier, H.; Lindup, I.; Alm. S.; Hänninen, O.; Jantunen, M.; Mathys,

und Tierseuchen, Fachbereich Umwelt- und Gesundheitsschutz, Berlin. Lahrz, T.; Piloty, M.; Oddoy, A. & Fromme, H. (2003). Gesundheitlich bedenkliche Substanzen

Tierseuchen, Fachbereich Umwelt- und Gesundheitsschutz. Berlin.

comparison of cyclists and car passengers. *Atmos Environ*, 44, 2263-2270. Janssen, N.A.H.; van Vliet, P.H.N.; Aaarts, F.; Harssema, H. & Brunekreef, B. (2001).

inside a test chamber. *Atmos Environ*, 43, 905-914.

near motorways. *Atmos Environ*, 35, 3875-3884.

burning incense. *Sci Total Environ*, 295, 51-67.

London, UK. *Atmos Environ*, 39, 3629-3641.

transport modes. *Atmos Environ*, 44, 3224-3227.

Oxford, UK. *Atmos Environ*, 38, 6399-6410.

Ontario. *Atmos Environ*, 45, 7583-7593.

31–41.

*Environ*, 44, 4545-4552,

*Environ*, 45, 1506-1516.

Aldershot, UK.

J. (2009). Deposition rates on smooth surfaces and coagulation of aerosol particles

V.; Thomas, I. & Meeusen, R. (2010). Exposure to particulate matter in traffic: A

Assessment of exposure to traffic related air pollution of children attending schools

Hildemann, L.M. (2011). Measurement of fine particles and smoking activity in a statewide survey of 36 California Indian casinos. *J Expo Sci Environ Epidemiol*, 21,

Garfinkel, R.; Yan, B.; Ross, J.; Camann, D.; Perera, F.P. & Miller, R.L. (2010). Effects of heating season on residential indoor and outdoor polycyclic aromatic hydrocarbons, black carbon, and particulate matter in an urban birth cohort. *Atmos* 

in underground and ground-level rail systems of the Los Angeles Metro. *Atmos* 

intersection users to PM2.5, ultrafine particle counts and carbon monoxide in central

Wheeler, A.J. (2010). Residential indoor and outdoor ultrafine particles in Windsor,

Williams K (Ed.). Spatial planning, urban form and sustainable transport. Ashgate,

Berliner Büroarbeitsplätzen, Bericht des Institutes für Lebensmittel, Arzneimittel

in öffentlichen Einrichtungen in Berlin. Untersuchungen zur Innenraumluftqualität in Berliner Schulen. Bericht des Instituts für Lebensmittel, Arzneimittel und

P.; Colvile, R.; Ashmore, M.R.; Cullinan, P. & Nieuwenhuijsen, M.J. (2004). Personal exposures and microenvironment concentrations of PM2.5, VOC, NO2 and CO in


Goodman, P.; Agnew, M.; McCaffrey, M.; Paul, G. & Clancy, L. (2007). Effects of the Irish

Gulliver, J. & Briggs, D.J. (2004). Personal exposure to particulate air pollution in transport

Guo, H.; Morawska, L.; He, C.; Zhang, Y.L.; Ayoko, G. & Cao, M. (2010). Characterization of

Han, X.; Aguilar-Villalobos, M.; Allen, J.; Charlton, C.S.; Ronbinson, R.; Bayer, C. & Naeher,

Hänninen, O.O.; Lebret, E.; Ilacqua, V.; Katsouyanni, K.; Künzli, N.; Sráme, R.J. & Jantunen,

He, C.; Morawska, L. & Tablin, L. (2007). Particle emission characteristics of office printers.

Heavner, D.L.; Morgan, W.T. & Ogden, M.W. (1996). Determination of volatile organic

Héroux, M.E.; Clark, N.; Van Ryswyk, K.; Mallick, R.; Gilbert, N.L.; Harrison, I.; Rispler, K.;

HESE (Health Effects of School Environment) (2006). Final Scientific Report prepared for the

Horemans, B.; Worobiec, A.; Buczynska, A.; Van Meel, K. & Van Grieken, R. (2008).

Hsu, D.J. & Huang, H.L. (2009). Concentrations of volatile organic compounds, carbon

Hu, B.; Freihaut, J.D.; Bahnfleth, W.; Gomes, C.A.S. & Brandolyn. T. (2005). Literatur review

Hurley, F.; Cherrie, J.; Donaldson, K.; Seaton, A. & Tran, L. (2004). Assessment of health

underground. Universitiy of Aberdeen. Research report TM/02/04. Hussein, T.; Glytsos, T.; Ondrácek, J.; Dohányosová, P.; Zdimal, V.; Hämeri, K.; Lazaridis,

rates during indoor activities in a home. *Atmos Environ*, 40, 4285-4307.

Pennsylvania homes and workplaces, *Environ Int*, 22, 159-183.

PM2.5 in residences of four European cities. *Atmos Environ*, 38, 6411-6423. He, C., Morawska, L.; Hitchins, J. & Gilbert, D. (2004). Contribution from indoor sources to

pollution on indoor air. *Environ Sci Pollut Res*, 17, 1268–1278.

*Am J Respir Crit Care Med*, 175, 840-845.

microenvironments. *Atmos Environ*, 38, 1-8.

Peru. *Int J Occup Environ Health*, 11, 276-288.

*Environ Sci Technol*, 41, 6039-6045.

*J Environ Res Public Health*, 7, 3080-3099.

*Atmos Environ*, 43, 5723–5730.

*Indoor Air*, 1541-1545.

3405-3415.

04.pdf

867–876.

smoking ban on respiratory health of bar workers and air quality in Dublin pubs.

particle number concentrations and PM2.5 in a school: influence of outdoor air

L. (2005). Traffic-related occupational exposures to PM2.5, CO and VOCs in Trujillo,

M. (2004). Infiltration of ambient PM2.5 and levels of indoor generated non-ETS

particle number and mass concentrations in residential houses. *Atmos Environ*, 38,

compounds and respirable suspended particulate matter in New Jersey an

Wang, D.; Anastassopoulos, A.; Guay, M.; MacNeill, M. & Wheeler, A.J. (2010). Predictors of indoor air concentrations in smoking and non-smoking residences. *Int* 

Health & Consumer Protection Directorate General. Siena, Italy. Online: http://ec.europa.eu/health/ph\_projects/2002/pollution/fp\_pollution\_2002\_frep\_

Airborne particulate matter and BTEX in office environments. *J Environ Monit*, 10,

monoxide, carbon dioxide and particulate matter in buses on highways in Taiwan.

and parametric study: indoor particle resuspension by human activity. *Proceedings:* 

effects of long-term occupational exposure to tunnel dust in the London

M.; Smolik, J. & Kulmala, M. (2006). Particle size characterization and emission


Particles in the Indoor Environment 141

Miguel, A.F.; Aydin, M. & Reis, A.H. (2005). Indoor deposition and forced re-suspension of

Milz, S.; Akbar-Khanzadeh, F.; Ames, A.; Spino, S.; Tex, C. & Lanza, K. (2007). Indoor air

Mönkkönen, P.; Pai, P.; Maynard, A.; Lehtinen, K.E.J.; Hämeri, K.; Rechkemmer, P.;

Molnár, P.; Bellander, T.; Sällsten, G. & Boman, J. (2007). Indoor and outdoor concentrations

Morawska, L.; He, C.; Hitchins, J.; Mengersen, K. & Gilbert, D. (2003). Characteristics of

Morawska, L. & Salthammer, T. (2003). Fundamentals of indoor particles and settled dust.

Mosqueron, L.; Momas, I. & Moullec, Y. (2002). Personal exposure of Paris office workers to nitogen, dioxide and fine particeles. *Occup Environ Med*, 59, 550-556. Naeher, L.P.; Smith, K.R.; Leaderer, B.P.; Mage, D. & Grajeda, R. (2000). Indoor and outdoor

Neas, L.M.; Dockery, D.W., Ware, J.H.; Spengler, J.D.; Ferris, B.G. & Speizer, F.E. (1994).

Oeder, S.; Weichenmeier, I.; Dietrich, S.; Schober, W.; Pusch, G.; Jörres, R.A.; Schierl, R.;

Ogulei, D.; Hopke, P.K. & Wallace, L.A. (2006). Analysis of indoor particle size distributions in an occupied townhouse using positive matrix factorization. *Indoor Air*, 16, 204-215. Osman, L.M.; Douglas, J.G.; Garden, C.; Reglitz, K.; Lyon, J.; Gordon, S. & Ayres, J.G. (2007).

Özkaynak, H.; Xue, J., Weker, R.; Butler, D.; Koutrakis, P. & Spengler, J.D. (1995). The

Phillips, K.; Bentley, M.C.; Howard, D.A. & Alvan, G. (1998). Assesment of air quality in

and environmental tobacco smoke. *Environ Int*, 24, 405-425.

Analyses of RIOPA data. *J Expo Anal Environ Epidemiol*, 15, 17-28.

respirable particles. *Indoor Built Environ*, 14, 391-396.

*Hyg*, 4, 246-252.

*Environ Monit*, 9, 348-57.

*Epidemiol*, 10, 544-551.

Australia. *Atmos Environ*, 37, 4195-4203.

children. *Am J Epidemiol*, 139, 1088-1099.

10.1111/j.1600-0668.2011.00743.x.

*Am J Respir Crit Care Med*, 176, 465-472.

Environmental Protection Agency.

settled dust. Wiley-VCH Verlag Weinheim, Germany.

Nazaroff, W.W. (2004). Indoor particle dynamics. *Indoor Air*, 4 (Suppl. 7), 175-183.

131-147.

ambient (outdoor) sources on residential indoor and personal PM2.5 concentrations:

quality in restaurants with and without designeted smoking rooms. *J Occup Environ* 

Ramachandran, G.; Prasad, B. & Kulmala. M. (2005). Fine particle number and mass concentration measurements in urban Indian households. *Sci Total Environ*, 347,

of PM2.5 trace elements at homes, preschools and schools in Stockholm, Sweden. *J* 

particle number and mass concentrations in residential houses in Brisbane,

In: Morawska L, Salthammer T. (Eds.) Indoor environment. Airborne particles and

PM2.5 and CO in high- and low-density Guatemalan villages. *J Expo Anal Environ* 

Concentration of indoor particulate matter as a determinant of respiratory health in

Nowak, D.; Fromme, H.; Behrendt, H. & Buters, J. (2011). Toxicity and elemental composition of particulate matter from outdoor and indoor air of elementary schools in Munich, Germany. *Indoor Air* Accepted manuscript online: DOI:

Indoor air quality in homes of patients with chronic obstructive pulmonary disease.

Particle TEAM (PTEAM) Study: analysis of the data. Final report. Vol. III, US -

Paris by personal monitoring of non-smokers for respirable suspended particles


Lee, J.; Lim, S.; Lee, K.; Guo, X.; Kamath, R.; Yamato H.; Abas, L.E.; Nandasena, S.; Nafees,

Li, S.C. & Lin, C.H. (2003). Carbon profile of residential indoor PM1 and PM2.5 in the

Link, B.; Gabrio, T.; Zöllner, I.; Schwenk, M.; Siegel, D.; Schultz, E.; Scharring, S. & Borm, P.

Lim, J.M.; Jeong, J.H.; Lee, J.H.; Moon, J.H.; Chung, Y.S. & Kim, K.H. (2011). The analysis of

Liu, D.L. & Nazaroff, W.W. (2003). Particle penetration through building cracks. *Aerosol Sci* 

Liu, Y.; Chen, R.; Shen, X. & Mao, X. (2004a). Wintertime indoor air levels of PM10, PM2.5 and PM1 at public places and their contributions to TSP. *Environ Int*, 30, 189-197. Liu, X.Y.; Mason, M.; Krebs, K. & Sparks, L. (2004b). Full-scale chamber investigation and

Liu, S. & Zhu, Y. (2010). A case study of exposure to ultrafine particles from secondhand

Long, C.M.; Suh, H.H. & Koutrakis, P. (2000). Characterization of indoor particle sources using continuous mass and size monitors. *J Air & Waste Manage Assoc*, 50, 1236-1250. Long, C.M.; Suh, H.H.; Catalano, P.J. & Koutrakis, P. (2001). Using time- and size-resolved

Mackay, E. (2004). An investigation of the variation in personal exposure to carbon

Martuzevicius, D.; Grinshpun, S.A.; Lee, T.; Hu, S.; Biswas, P.; Reponen, T. & LeMasters, G.

Matson, U. (2005). Indoor and outdoor concentrations of ultrafine particles in some

Matson, U. & Ekberg, L.E. (2005). Prediction of ultrafine particle concentrations in various

McLaughlin, J.; Hogg, C. & Guo, L.Y. (2005). Ultrafine and coarse mode aerosol measurements in selected dwellings in Ireland. *Proceedings Indoor Air*, 698-701. Meng, Q.J.; Turpin, B.J.; Korn. L.; Weisel, C.P.; Morandi, M.; Colome, S.; Zhang, J.; Stock, T.;

Scandinavian rural and urban areas. *Sci Total Environ*, 343, 169-176.

Massachusetts. *J Expo Anal Environ Epidemiol*, 12, 104-14.

Bericht des Landesgesundheitsamtes Baden-Württemberg.

tobacco smoke in an automobile. *Indoor Air*, 20, 412-423.

US city. *J Expo Anal Environ Epidemiol*, 10, 36-49.

indoor enviroments. *Proceedings of Indoor Air*, 1581-1585.

subtropical region. *Atmos Environ*, 37, 881-888.

urban area. *Indoor Air*, 21, 145–155.

*Technol*, 37, 565-573.

*Sci Technol*, 35, 2089-2099.

38, 2802-2812.

A.A. & Sathiakumar, N. (2010). Secondhand smoke exposures in indoor public places in seven Asian countries. International *J Hyg Environ Health*, 213, 348–351. Levy, J.I.; Dumyahn, T. & Sprengler, J.D. (2002). Particulate matter and polycyclic

hydrocarbon concentrations in indoor and outdoor microenvironments in Boston,

(2004). Feinstaubbelastung und deren gesundheitliche Wirkungen bei Kindern.

PM2.5 and associated elements and their indoor/outdoor pollution status in an

simulation of air freshener emissions in the presence of ozone. *Environ Sci Technol*,

particulate data to quantify indoor penetration and deposition behavior. *Environ* 

monoxide and particulates on the A660 in Leeds. M.Sc. thesis, University of Leeds.

(2008). Traffic-related PM2.5 aerosol in residential houses located near major highways: indoor versus outdoor concentrations. *Atmos Environ*, 42, 6575-6585. Maskarinec, M.P.; Jenkins, R.A.; Counts, R.W. & Dindal, A.B. (2000). Determination of

exposure to environmental tobacco smoke in restaurant and tavern workers in one

Spektor, D.; Winer, A.; Zhang, L.; Lee, J.H.; Giovanetti, R.; Cui, W.; Kwon, J.; Alimokhtari, S.; Shendell, D.; Jones, J.; Farrar, C. & Maberti, S. (2005). Influence of ambient (outdoor) sources on residential indoor and personal PM2.5 concentrations: Analyses of RIOPA data. *J Expo Anal Environ Epidemiol*, 15, 17-28.


Particles in the Indoor Environment 143

Sendzik, T.; Fong, G.; Travers, M. & Hyland, A. (2006). The hazard of tobacco smoke

Shaughnessy, R.J.; Turk, B.; Evans, S.; Fowler, F.; Casteel, S. & Louie, S. (2002). Preliminary

Simons, E.; Curtin-Brosnan, J.; Buckley, T.; Breysse, P. & Eggleston, PA. (2007). Indoor

Singer, B.C.; Coleman, B.K.; Destaillats, H.; Hodgson, A.T.; Lunden, M.M.; Weschler, C.J. &

Son, B.S.; Song, M.R. & Yang, W.H. (2005). A study on PM10 and VOCs concentrations of

Stranger, M.; Potgieter-Vermaak, S.S. & Van Grieken, R. (2007). Comparative overview of

Thatcher, T.L. & Layton, D. (1995). Deposition, Resuspension, and Penetration of Particles

Thatcher, T.L.; McKone, T.E.; Fisk, W.J.; Sohn, M.D.; Delp, W.W.; Riley, W.J. & Sextro, R.G.

Valente, P.; Forastiere, F., Bacosi, A.; Cattani, G., Di Carlo, S.; Ferri, M.; Figà-Talamanca, I.;

Vardavas, C.I.; Kondilis, B.; Travers, M.J.; Petsetaki, E.; Tountas, Y. & Kafatos, A.G. (2007).

Vartiainen, E.; Kulmala, M.; Ruuskanen, T.M.; Taipale, R.; Rinne, J. & Vehkamäki, H. (2006).

Vette, A.F.; Rea, A.W.; Law, P.A.; Rodes, C.E.; Evans, G.; Highsmith, V.R. & Sheldon, S.

during the Fresno PM Exposure Studies. *Aerosol Sci Technol*, 34, 118 – 126. Wainman, T.; Zhang, J.; Weschler, C.J. & Lioy, P.J. (2000). Ozone and limonene in indoor air: a source of submicron particle exposure. *Environ Health Perspect*, 108, 1139-1145. Wallace, L. & Howard-Reed, C. (2002). Continuous monitoring of ultrafine, fine and coarse

Wallace, L.A.; Mitchell, H.; O´Connor, G.T.; Neas, L.; Lippmann, M.; Kattan, M.; Koeng, J.,

freshner use in the presence of ozone. *Atmos Environ*, 40, 6696-6710.

indoor air quality in Antwerp, Belgium. *Environ Int*, 33, 789–797.

Laboratory (LBNL). Report under contract No. DW-89938748.

within a Residence, *Atmos Environ*, 29, 1487-1497.

smoking ban, Italy 2005. *Tobacco Control*, 16, 312-317.

oxidation. *Atmos. Environ*, 40, 7882-7892.

Conference on Tobacco and Health, Washington DC.

asthma. *J Urban Health*, 84, 577-590.

*Indoor Air*, 827-832.

7, 302.

828-844.

uncarpeted classrooms. *Proceedings of Indoor Air*, 974-979.

pollution in cars: evidence from an air quality monitoring study. 13th World

study of flooring in school in the U.S.: airborne particulate exposure in carpeted vs.

environmental differences between inner city and suburban homes of children with

Nazaroff, W.W. (2006). Indoor secondary pollutants from cleaning products and air

indoor environment in school and recognition of indoor air quality. *Proceedings:* 

(2001). Factors affecting the concentration of outdoor particles indoors (COPI): identification of data needs and existing data. Lawrence Berkeley National

Marconi, A., Paoletti, L., Perucci, C. & Zuccaio, P. (2007). Exposure to fine and ultrafine particles from secondhand smoke in public placet bifore and after the

Environmental tobacco smoke in hospitality venues in Greece. *BMC Public Health*,

Formation and growth of indoor air aerosol particles as a result of d-limonene

(2001). Characterization of indoor-outdoor aerosol concentration relationships

particles in a residence for 18 months in 1999-2000. *J Air Waste Manage Assoc*, 52,

Stout, J.W., Vaughn, B.J.; Wallace, D., Walter, M.; Adams, K. & Liu, L.-J.S. (2003). Particle concentrations in inner-city homes of children with asthma: the effect of smoking, cooking, and outdoor pollution. *Environ Health Perspect*, 111, 1265-1272.


Raaschou- Nielsen, O.; Sørensen, M.; Hertel, O.; Chawes, B.L.K.; Vissing, N.; Bønnelykke, K.

Raut, J.C.; Chazette, P. & Fortain, A. (2009). Link between aerosol optical, microphysical and

Rees, V.W. & Connolly, G.N. (2006). Measuring air quality to protect children from second

Repace, J.; Hyde, J.N. & Brugge, D. (2006). Air pollution in Boston bars before and after a

Reynolds, S.J.; Bleck, D.W.; Borin, S.S.; Breuer, G.; Burmeister, L.F.; Fuortes, L.J.; Smith, T.F.;

Riediker, M.; Cascio, W.E.; Griggs, T.R.; Herbst, M.C.; Bromberg, P.A.; Neas, L.; Williams, R.W.

Rim, D.; Siegel, J.; Spinhirne, J.; Webb, A. & McDonald-Buller, E. (2008). Characteristics of cabin air quality in school buses in Central Texas. *Atmos Environ*, 42, 6453–6464. Rodes, C.E.; Lawless, P.A.; Thornburg, J.W.; Williams, R.W. & Croghan, C.W. (2010). DEARS

Rojas-Bracho, L.; Suh, H.H. & Koutrakis, P. (2000). Relationships among personal, indoor,

Rosen, L.J.; Zucker, D.M.; Rosen, B.J. & Connolly, G.N. (2010). Second-hand smoke levels in

Sabin, L.D.; Behrentz, E.; Winer, A.M.; Jeong, S.; Fitz, D.R.; Pankratz, D.V.; Colome, S.D. &

during school bus commutes. *J Expo Anal Environ Epidemiol*, 15, 377–387. Santen, M.; Wesselmann, M.; Fittschen, U.; Cremer, R.; Braun, P.; Lüdecke, A. & Moriske,

living rooms. Gefahrstoffe-Reinhaltung der Luft, 69, 63-70 [in German]. Sarwar, G.; Olson, D.A.; Corsi, R.L. & Weschler, C.J. (2004). Indoor fine particles: the role of terpene emissions from consumer products. *J Air & Waste Manage Assoc*, 54, 367-377. Schneider, S.; Seibold, B.; Schunk, S.; Jentzsch, E.; Dresler, C.; Travers, M.J.; Hyland, A. &

Semple, S.; Van Tongeren, M.; Galea, K.S.; Maccalman, L.; Gee, I.; Parry, O.; Naji, A. &

bars in Scotland, England, and Wales. *Ann Occup Hyg*, 54, 272–280.

*Technol*, 36, 200-207. Erratum in: *Environ Sci Technol*, 36, 1868.

settings for a general population, *Atmos. Environ.*, 44, 1386-1399.

bedrooms in Denmark. *Environ Res*, 111, 87–93.

hand smoke in cars. *Am J Prev Med*, 31, 363-368.

States. *App Occup Environ Hyg*, 16, 1065-1077.

*Expo Anal Environ Epidemiol*, 10, 294-306.

*Nicotine & Tobacco Res*, 10, 547–555.

legislation. *European J Public Health*, 21, 15–20.

smoking ban. *BMC Public Health*, 6, 266.

43, 860–868.

& Bisgaard, H. (2011). Predictors of indoor fine particulate matter in infants'

chemical measurements in an underground railway station in Paris. *Atmos Environ*,

Stein, M.A., Subramanian, P.; Thorne, P.S. & Whitten, P. (2001). Indoor Enviromental Quality in Six Commercial Office Buildings in the Midwest United

& Devlin, R.B. (2004). Particulate matter exposure in cars is associated with cardiovascular effects in healthy young men. *Am J Respir Crit Care Med*, 169, 934-940. Riley, W.J.; McKone, T.E.; Lai, A.C. & Nazaroff, W.W. (2002). Indoor particulate matter of

outdoor origin: importance of size-dependent removal mechanisms. *Environ Sci* 

particulate matter relationships for personal, indoor, outdoor, and central site

and outdoor fine and coarse particle concentrations for individuals with COPD. *J* 

Israeli bars, pubs and cafes before and after implementation of smoke-free

Fruin, S.A. (2005). Characterizing the range of children's air pollutant exposure

H.J. (2009). Measurements of fine and ultrafine particles in indoor environment of

Pötschke-Langer, M. (2008). Exposure to secondhand smoke in Germany: Air contamination due to smoking in German restaurants, bars, and other venues.

Ayres, J.G. (2010). UK smoke-free legislation: changes in PM2.5 concentrations in


**7** 

*Japan* 

Keiko Hirota1, Satoshi Shibuya2,

*1Japan Automobile Research Institute* 

Shogo Sakamoto2 and Shigeru Kashima2

*2Faculty of Science and Engineering, Chuo University* 

**A Methodology of Estimation on Air Pollution** 

**and Its Health Effects in Large Japanese Cities** 

The correlation between air pollution and health effects in large Japanese cities presents a great challenge owing to the limited availability of data on the exposure to pollution, health effects and the uncertainty of mixed causes. A methodology for quantitative relationships (between the emission volume and air quality, and the air quality and health effects) is analysed with a statistical method in this chapter; the correlation of air pollution reduction

This chapter discusses a step-by-step methodology of determining the direct correlation between emission volumes, air quality, and health effects. Figure 1 shows the Japanese compensation system with two hypotheses in order to clarify the correlation. Hypothesis 1 states that the total emission volume affects air quality. Hypothesis 2 states that the air

Fig. 1. The relationship between emissions, air quality and the number of certified patients

**1. Introduction** 

policy in Japan from 1974 to 2007.

quality influences the number of certified patients.

(above) and flow of compensation (below)


## **A Methodology of Estimation on Air Pollution and Its Health Effects in Large Japanese Cities**

Keiko Hirota1, Satoshi Shibuya2, Shogo Sakamoto2 and Shigeru Kashima2 *1Japan Automobile Research Institute 2Faculty of Science and Engineering, Chuo University Japan* 

### **1. Introduction**

144 Air Quality – Monitoring and Modeling

Wallace, L. & Ott, W. (2011). Personal exposure to ultrafine particles. *J Expos Sci Environ* 

Weichenthal, S.; Dufresne, A.; Infante-Rivard, C. & Joseph, L. (2007). Indoor ultrafine

homes during the winter months. *J Expo Sci Environ Epidemiol*, 17, 288-297. Weschler, C.J. (2003). Indoor chemistry as a source of particles. In: Morawska, L. &

Weschler, C.J.; Wells, J.R.; Poppendieck, D.; Hubbard, H. & Pearce, T.A. (2006). Workgroup report: indoor air chemistry and health. *Environ Health Perspect*, 114, 442-446.. Wichmann, J.; Lind, T.; Nilsson, M.A.M. & Belland, T. (2010). PM2.5, soot and NO2 indoor

Yang, W.; Sohn, J.; Kim, J.; Son, B. & Park, J. (2009). Indoor air quality investigation according to age of the school buildings in Korea. *J Environ Manage*, 90, 348e-354e. Zhang, Q. & Zhu, Y. (2010). Measurements of ultrafine particles and other vehicular pollutants inside school buses in South Texas. *Atmos Environ*, 44, 253-261. Zuurbier, M.; Hoek, G.; Oldenwening, M.; Lenters, V.; Meliefste, K.; van den Hazel, P. &

particle exposure and home heating systems: A cross-sectional survey of Canadian

Salthammer, T. (Eds.) Indoor environment. Airborne particles and settled dust.

outdoor relationships at homes, pre-schools and schools in Stockholm, Sweden.

Brunekreef, B. (2010). Commuters' Exposure to Particulate Matter Air Pollution Is Affected by Mode of Transport, Fuel Type, and Route. *Environ Health Perspect*, 118,

*Epidemiol*, 21, 20–30.

Wiley-VCH Verlag Weinheim, 2003.

*Atmos Environ*, 44, 4536-4544.

783–789.

The correlation between air pollution and health effects in large Japanese cities presents a great challenge owing to the limited availability of data on the exposure to pollution, health effects and the uncertainty of mixed causes. A methodology for quantitative relationships (between the emission volume and air quality, and the air quality and health effects) is analysed with a statistical method in this chapter; the correlation of air pollution reduction policy in Japan from 1974 to 2007.

This chapter discusses a step-by-step methodology of determining the direct correlation between emission volumes, air quality, and health effects. Figure 1 shows the Japanese compensation system with two hypotheses in order to clarify the correlation. Hypothesis 1 states that the total emission volume affects air quality. Hypothesis 2 states that the air quality influences the number of certified patients.

Fig. 1. The relationship between emissions, air quality and the number of certified patients (above) and flow of compensation (below)

A Methodology of Estimation on Air Pollution and Its Health Effects in Large Japanese Cities 147

areas "Class 1", where health damage has occurred. The compensation law also classifies "certifying patients" who are affected by one of the four pollution-related respiratory and circulatory diseases (pulmonary emphysema, asthmatic bronchitis, bronchial asthma, and chronic bronchitis). The patients with one of these four diseases are considered as certified patients if they satisfy certain other conditions (such as extended residence in the certified area, non-smoker status, or illness diagnosed by a medical doctor) in the certified areas [ERCA Homepage http://www.erca.go.jp]. The compensation system collects emission charge from 9,000 factories and companies on the basis of their SO2 emission volumes. SO2 emission charge from polluting industries (80%) and a vehicle tax (20%) covered the welfare

Fig. 2. Motorization and emission reduction measures from the automobile sector in Japan

The industrial sector promoted the government to amend the compensation law in the beginning of 1980s. The committee of air pollution reduction measures reported that certain areas could have had higher rates of morbidity and higher prevalence rates of chronic obstructive pulmonary disease (Abbreviation: COPD) caused by air pollution in 1950s and 1960s. However, it is not the same situation as those of 1950s and 1960s. In September 1987, amendments were made to the compensation system based on the report of the committee. In March 1988, the Association of Environmental Restoration and Conservation Agency of Japan started an additional project to prevent health damage caused by air pollution on the basis of the revision of the compensation law. This new project intended to assist local public organizations with health consultations, health examinations, and functional training

services for the certified patients.

The relationship between total emissions (NOx, PM) (from both stationary and mobile sources) and air quality (NO2, SPM) was found to be significant, which supports hypothesis 1. The correlation analysis of emission volume, and air quality suggests that NOx and PM levels worsen according to increases in NO2.

When the correlation between the air pollutant and the type of health effects (certified, mortality, recovery, and newly registered) was examined according to the certified area, an inverse relationship was observed. The relationship between air quality (NO2) and health effect was found to be significant, which support hypothesis 2. When NO2 worsens, certified patients, mortality rates and newly certified patients increase, according to the data from 1989 to 2007 with dummy variable analysis.

#### **2. Methodology**

#### **2.1 Pollution-related health damage compensation system and the law concerning special measures to reduce the total amount of nitrogen oxides and particulate matter emitted from motor vehicles in specified areas**

Mie prefecture rapidly deteriorated as a result of rapid economic growth in the early 1960s, the air quality in the industrial sector of Yokkaichi. Residents had higher rates of morbidity and higher prevalence rates of respiratory and circulatory diseases in Yokkaichi than in other areas.

Environmental laws and standards were established to address those issues by the government in order to mitigate this deterioration of air quality in the late 1960s to early 1970s. The Basic Law of Pollution Prevention was established in 1967. This law determined the permissible limits of each contaminant in the air (SO2 in 1969, CO in 1970, suspended particulate matter (SPM) in 1972, and NO2 in 1973). The policies initially focused on stationary sources such as industries for reducing air pollution.

The health effects caused by air pollution became a social issue. A major issue was the apportioning of the compensation among the potential polluters. There was strong opposition from the industrial sector when discussions began among the polluters [Matsuura, 1994 a, 1994 b, 1994 c]. The affected patients sought compensation from the Japanese government and the industrial sector.

To address the social issues associated with the health effects observed since the early 1960s, the Pollution-related Health Damage Compensation Law (Abbreviation: the compensation law) was enacted in 1973 to support the affected patients by providing prompt and fair compensation to cover damages and by implementing the necessary programs for the welfare of these patients. The Pollution-related Health Damage Compensation and Prevention System (Abbreviation: the compensation system) was established in June 1974 in accordance with the compensation law. The certified regions and diseases for the compensation system were determined by the government in 1974 [Amagasaki City Government 1994, Kawasaki City Government 1994, Kita Kyushu City Government 1994 & Osaka City Government 1974] (See (6) of Figure 2).

The compensation system implements welfare services to facilitate the recovery of patients with health impairments based on the compensation law. The compensation law is responsible for determining the groups that should bear the expenses in the designated

The relationship between total emissions (NOx, PM) (from both stationary and mobile sources) and air quality (NO2, SPM) was found to be significant, which supports hypothesis 1. The correlation analysis of emission volume, and air quality suggests that NOx and PM

When the correlation between the air pollutant and the type of health effects (certified, mortality, recovery, and newly registered) was examined according to the certified area, an inverse relationship was observed. The relationship between air quality (NO2) and health effect was found to be significant, which support hypothesis 2. When NO2 worsens, certified patients, mortality rates and newly certified patients increase, according to the data

**2.1 Pollution-related health damage compensation system and the law concerning special measures to reduce the total amount of nitrogen oxides and particulate matter** 

Mie prefecture rapidly deteriorated as a result of rapid economic growth in the early 1960s, the air quality in the industrial sector of Yokkaichi. Residents had higher rates of morbidity and higher prevalence rates of respiratory and circulatory diseases in Yokkaichi than in

Environmental laws and standards were established to address those issues by the government in order to mitigate this deterioration of air quality in the late 1960s to early 1970s. The Basic Law of Pollution Prevention was established in 1967. This law determined the permissible limits of each contaminant in the air (SO2 in 1969, CO in 1970, suspended particulate matter (SPM) in 1972, and NO2 in 1973). The policies initially focused on

The health effects caused by air pollution became a social issue. A major issue was the apportioning of the compensation among the potential polluters. There was strong opposition from the industrial sector when discussions began among the polluters [Matsuura, 1994 a, 1994 b, 1994 c]. The affected patients sought compensation from the

To address the social issues associated with the health effects observed since the early 1960s, the Pollution-related Health Damage Compensation Law (Abbreviation: the compensation law) was enacted in 1973 to support the affected patients by providing prompt and fair compensation to cover damages and by implementing the necessary programs for the welfare of these patients. The Pollution-related Health Damage Compensation and Prevention System (Abbreviation: the compensation system) was established in June 1974 in accordance with the compensation law. The certified regions and diseases for the compensation system were determined by the government in 1974 [Amagasaki City Government 1994, Kawasaki City Government 1994, Kita Kyushu City Government 1994 &

The compensation system implements welfare services to facilitate the recovery of patients with health impairments based on the compensation law. The compensation law is responsible for determining the groups that should bear the expenses in the designated

levels worsen according to increases in NO2.

from 1989 to 2007 with dummy variable analysis.

**emitted from motor vehicles in specified areas** 

Japanese government and the industrial sector.

Osaka City Government 1974] (See (6) of Figure 2).

stationary sources such as industries for reducing air pollution.

**2. Methodology** 

other areas.

areas "Class 1", where health damage has occurred. The compensation law also classifies "certifying patients" who are affected by one of the four pollution-related respiratory and circulatory diseases (pulmonary emphysema, asthmatic bronchitis, bronchial asthma, and chronic bronchitis). The patients with one of these four diseases are considered as certified patients if they satisfy certain other conditions (such as extended residence in the certified area, non-smoker status, or illness diagnosed by a medical doctor) in the certified areas [ERCA Homepage http://www.erca.go.jp]. The compensation system collects emission charge from 9,000 factories and companies on the basis of their SO2 emission volumes. SO2 emission charge from polluting industries (80%) and a vehicle tax (20%) covered the welfare services for the certified patients.

Fig. 2. Motorization and emission reduction measures from the automobile sector in Japan

The industrial sector promoted the government to amend the compensation law in the beginning of 1980s. The committee of air pollution reduction measures reported that certain areas could have had higher rates of morbidity and higher prevalence rates of chronic obstructive pulmonary disease (Abbreviation: COPD) caused by air pollution in 1950s and 1960s. However, it is not the same situation as those of 1950s and 1960s. In September 1987, amendments were made to the compensation system based on the report of the committee.

In March 1988, the Association of Environmental Restoration and Conservation Agency of Japan started an additional project to prevent health damage caused by air pollution on the basis of the revision of the compensation law. This new project intended to assist local public organizations with health consultations, health examinations, and functional training

A Methodology of Estimation on Air Pollution and Its Health Effects in Large Japanese Cities 149

These results suggest that more variables, such as lifestyle, age, and sex, need to be included for further research. The data set of the Japanese compensation system is employed because

The compensation system covers 25 certified areas (Figure 3). The NOx-PM law designated a total of 276 communities in the Tokyo, Saitama, Kanagawa, Chiba, Aichi, Mie, Osaka, and Hyogo Prefectures as certified areas with significant air pollution due to nitrogen oxides

Five certified areas including 19 wards in Tokyo , Kawasaki (Saiwai, Kawasaki), the whole of Osaka city, Amagasaki (East region and South region), and Kita Kyushu (Wakamatsu, Tobata, Yawata, Kokura, Kokusetsu-Kita Kyushu, Nijima, Kurosaki, and Higashi-Kokura)

Fig. 3. Areas of certified (Class 1) under the Pollution-Related Health Damage Prevention system (the research areas assessed in this paper are underlined) and the NOx-PM area

Figure 4 shows the air pollutants and their affected diseases. USEPA and WHO considered respiratory and cardiovascular diseases caused by air pollution. PM is a representative air pollutant for analysis according to USEPA and WHO. [UPEPA 1999, WHO 2002, UK DH

this system certified patients suffering from diseases caused by air pollution.

**2.3.1 Certified areas of the compensation system and the NOx-PM law** 

were selected for the analysis of Hypothesis 1 & 2.

emitted from motor vehicles. Kita Kyushu area is not the NOx-PM certified area.

**2.3 Precondition** 

(Highlighted in yellow)

2006].

**2.3.2 Air pollutants and health effects** 

to carry out research and to distribute knowledge leading to health maintenance in local communities. The Law was revised in 1988 to remove the Class 1 specification (compensation disbursement was continued) and to start health damage prevention projects [ERCA Homepage http://www.erca.go.jp].

The focus of emission reduction policies shifted from the industrial sector to non-industrial sources, namely, mobile sources in 1990s due to motorization. The Ministry of the Environment adopted a total emission volume control measure by the "Law Concerning Special Measures to Reduce the Total Amount of Nitrogen Oxides Emitted from Motor Vehicles in Specified Areas", called in short "The Motor Vehicle NOx Law" in 1992. The law copes with NOx pollution problems from existing vehicle fleets in highly populated metropolitan areas. Under the Motor Vehicle NOx Law, several measures had to be taken to control NOx from in-use vehicles, including enforcing emission standards for specified vehicle categories.

The regulation was amended in June 2001 to tighten the existing NOx requirements and to add PM control provisions. The amended rule is called the "Law Concerning Special Measures to Reduce the Total Amount of Nitrogen Oxides and Particulate Matter Emitted from Motor Vehicles in Specified Areas", or in short "the Motor Vehicle NOx and PM Law". The amended regulation became effective in October 2002.

The law concerning Special Measures for Total Emission Reduction of Nitrogen Oxides and Particulate Matter (Abbreviation: The NOx-PM Law) introduces emission standards for specified categories of in-use highway vehicles including commercial goods (cargo) vehicles such as trucks and vans, buses, and special purpose motor vehicles, irrespective of the fuel type. The regulation also applies to diesel powered passenger cars.

#### **2.2 Literature survey**

This section explains why this research focuses on correlation analysis by comparing some previous studies. Iwai K, Mizuno S, Miyasaka Y, and Mori T [2005] analysed the odds ratio from the 2001 data for particulate matters PM2.5 converted from suspended particulate matter (SPM) for the whole of Japan. The authors tried to find out the direct correlation between respiratory diseases and air quality. Their contribution showed a strong correlation between the classification of diseases and air quality. Following this methodology of the direct correlation analysis, this chapter focuses on the diseases caused by air pollution and air quality.

Makino [1996] focused on 23 Tokyo wards from 1958 to 1989, analysing SOx, NOx, and dust and comparing those to the total registered vehicles. In this research, a time lag of 15 to 20 years from the time of exposure to the appearance of the health effect was found. This demonstrates the importance of long-term analysis.

Sunyer J et al. [2006] focused their research on 21 European cities from 1991 to 1993. The author analysed the correlation (cross section) between the prevalence ratio of chronic bronchitis and air quality. The prevalence ratio was adjusted for attributes such as sex, place of residence, smoking habits, income level, and social status based on a questionnaire (n = 6824). Higher NO2 levels resulted in a higher prevalence ratio of female patients.

These results suggest that more variables, such as lifestyle, age, and sex, need to be included for further research. The data set of the Japanese compensation system is employed because this system certified patients suffering from diseases caused by air pollution.

#### **2.3 Precondition**

148 Air Quality – Monitoring and Modeling

to carry out research and to distribute knowledge leading to health maintenance in local communities. The Law was revised in 1988 to remove the Class 1 specification (compensation disbursement was continued) and to start health damage prevention projects

The focus of emission reduction policies shifted from the industrial sector to non-industrial sources, namely, mobile sources in 1990s due to motorization. The Ministry of the Environment adopted a total emission volume control measure by the "Law Concerning Special Measures to Reduce the Total Amount of Nitrogen Oxides Emitted from Motor Vehicles in Specified Areas", called in short "The Motor Vehicle NOx Law" in 1992. The law copes with NOx pollution problems from existing vehicle fleets in highly populated metropolitan areas. Under the Motor Vehicle NOx Law, several measures had to be taken to control NOx from in-use vehicles, including enforcing emission standards for specified

The regulation was amended in June 2001 to tighten the existing NOx requirements and to add PM control provisions. The amended rule is called the "Law Concerning Special Measures to Reduce the Total Amount of Nitrogen Oxides and Particulate Matter Emitted from Motor Vehicles in Specified Areas", or in short "the Motor Vehicle NOx and PM Law".

The law concerning Special Measures for Total Emission Reduction of Nitrogen Oxides and Particulate Matter (Abbreviation: The NOx-PM Law) introduces emission standards for specified categories of in-use highway vehicles including commercial goods (cargo) vehicles such as trucks and vans, buses, and special purpose motor vehicles, irrespective of the fuel

This section explains why this research focuses on correlation analysis by comparing some previous studies. Iwai K, Mizuno S, Miyasaka Y, and Mori T [2005] analysed the odds ratio from the 2001 data for particulate matters PM2.5 converted from suspended particulate matter (SPM) for the whole of Japan. The authors tried to find out the direct correlation between respiratory diseases and air quality. Their contribution showed a strong correlation between the classification of diseases and air quality. Following this methodology of the direct correlation analysis, this chapter focuses on the diseases caused by air pollution and

Makino [1996] focused on 23 Tokyo wards from 1958 to 1989, analysing SOx, NOx, and dust and comparing those to the total registered vehicles. In this research, a time lag of 15 to 20 years from the time of exposure to the appearance of the health effect was found. This

Sunyer J et al. [2006] focused their research on 21 European cities from 1991 to 1993. The author analysed the correlation (cross section) between the prevalence ratio of chronic bronchitis and air quality. The prevalence ratio was adjusted for attributes such as sex, place of residence, smoking habits, income level, and social status based on a questionnaire (n =

6824). Higher NO2 levels resulted in a higher prevalence ratio of female patients.

[ERCA Homepage http://www.erca.go.jp].

The amended regulation became effective in October 2002.

demonstrates the importance of long-term analysis.

type. The regulation also applies to diesel powered passenger cars.

vehicle categories.

**2.2 Literature survey** 

air quality.

#### **2.3.1 Certified areas of the compensation system and the NOx-PM law**

The compensation system covers 25 certified areas (Figure 3). The NOx-PM law designated a total of 276 communities in the Tokyo, Saitama, Kanagawa, Chiba, Aichi, Mie, Osaka, and Hyogo Prefectures as certified areas with significant air pollution due to nitrogen oxides emitted from motor vehicles. Kita Kyushu area is not the NOx-PM certified area.

Five certified areas including 19 wards in Tokyo , Kawasaki (Saiwai, Kawasaki), the whole of Osaka city, Amagasaki (East region and South region), and Kita Kyushu (Wakamatsu, Tobata, Yawata, Kokura, Kokusetsu-Kita Kyushu, Nijima, Kurosaki, and Higashi-Kokura) were selected for the analysis of Hypothesis 1 & 2.

Fig. 3. Areas of certified (Class 1) under the Pollution-Related Health Damage Prevention system (the research areas assessed in this paper are underlined) and the NOx-PM area (Highlighted in yellow)

#### **2.3.2 Air pollutants and health effects**

Figure 4 shows the air pollutants and their affected diseases. USEPA and WHO considered respiratory and cardiovascular diseases caused by air pollution. PM is a representative air pollutant for analysis according to USEPA and WHO. [UPEPA 1999, WHO 2002, UK DH 2006].

A Methodology of Estimation on Air Pollution and Its Health Effects in Large Japanese Cities 151

certified patients reduced because no new patients could be registered after 1989. Two time series will be considered in order to present the correlation. One is for the whole time period

②

Fig. 5. Number of certificated patient in the pollution-related health damage compensation

These two hypotheses are expressed by the following 5 equations. The annual average value NO2 of residential areas depends on emission volumes including those from stationary and mobile sources (hypothesis 1) in Eq. 1. The health effects (certified patients, mortality rates, recovery, and newly registered patients) depend on NO2 level (hypothesis 2) from Eq. 2 to

The emission volume is divided by the surface area in order to compare the health effects. The number of certified patients is divided by the total population because this group becomes patient group from normal population group. Certified patient includes newly registered and re-registered patients. The numbers of deceased and recovered are divided by the number of certified patients because these groups change status from certified patients. The number of newly registered patients is divided by the number of population

> , , *Mtc tc*

because this group becomes patient group from normal population group.

X: Emission volume of NOx (stationary + mobile sources) (t/year/km2)

Methodology of hypothesis 1 (Emission-air quality)

M: Annual average value of NO2 in residential area (ppm)

①1974

②1988

The Pollution-related Health Damage Compensation and

Prevention System

ln *X c* (1)

Revision of the Pollutionrelated Health Damage Compensation Law

(1974–2007), and the other is for two separate time series (1974–1988 and 1989–2007).

①

and prevention system (compensation system)

**2.3.4 Estimation models** 

Eq. 5.

The Japanese definition of disease caused by air pollution is narrower compared to those of USEPA and WHO. NO2 and SO2 affect the four certified respiratory diseases in the compensation system in Japan. Emission fees are determined by the volume of SO2 emissions in the compensation system because SO2 measurement method was already established. NO2 measurement did not seem to be well-established in the early 1970s. The relationship between NO2 and health effects has been a representative indicator for the evaluation of health effects in epidemiological research in Japan.

First, we used both SPM and NO2, but we found that the analysis with NO2 has more significant results than SPM. Second, the Air Quality Index (abbreviation: AQI) data were also revised in order to include all the pollutants. The results of the AQI data confirm that PM and NOx are the main indicators for the assessment of the health effects of air pollution. Finally, NOx was selected for the analysis based on previous literature and on its strong relationship to health effects.

Fig. 4. Air pollutants and their affected diseases

#### **2.3.3 Health effects caused by air pollution**

The data of the compensation system has two benefits. First, local governments issue white paper. The white paper shows the number of certified, mortality, newly registered and recovered patients in public. These consistency data is useful to check if it is reliable by calculation. Second, a certified patient is required to be non-smoker as a rule. The data is classified patients of certified 4 diseases and non-smokers in the compensation system.

There are two different trends according to statistical data of certified patients in the compensation system; one before 1988 and one after 1989 (Figure 5). The air quality dramatically diminished and the number of patients increased before 1988. The number of certified patients reduced because no new patients could be registered after 1989. Two time series will be considered in order to present the correlation. One is for the whole time period (1974–2007), and the other is for two separate time series (1974–1988 and 1989–2007).

Fig. 5. Number of certificated patient in the pollution-related health damage compensation and prevention system (compensation system)

#### **2.3.4 Estimation models**

150 Air Quality – Monitoring and Modeling

The Japanese definition of disease caused by air pollution is narrower compared to those of USEPA and WHO. NO2 and SO2 affect the four certified respiratory diseases in the compensation system in Japan. Emission fees are determined by the volume of SO2 emissions in the compensation system because SO2 measurement method was already established. NO2 measurement did not seem to be well-established in the early 1970s. The relationship between NO2 and health effects has been a representative indicator for the

First, we used both SPM and NO2, but we found that the analysis with NO2 has more significant results than SPM. Second, the Air Quality Index (abbreviation: AQI) data were also revised in order to include all the pollutants. The results of the AQI data confirm that PM and NOx are the main indicators for the assessment of the health effects of air pollution. Finally, NOx was selected for the analysis based on previous literature and on its strong

Respiratory

Bronchial asthma Asthmatic bronchitis Chronic bronchitis(CB) Pulmonary [pulmonic]

Pneumonia Lung cancer Cardiovascular

The data of the compensation system has two benefits. First, local governments issue white paper. The white paper shows the number of certified, mortality, newly registered and recovered patients in public. These consistency data is useful to check if it is reliable by calculation. Second, a certified patient is required to be non-smoker as a rule. The data is classified patients of certified 4 diseases and non-smokers in the compensation system.

There are two different trends according to statistical data of certified patients in the compensation system; one before 1988 and one after 1989 (Figure 5). The air quality dramatically diminished and the number of patients increased before 1988. The number of

Irritation skin and eyes

emphysema (PE)

WHO definition

Data source:USEPA,HEI,WHO

evaluation of health effects in epidemiological research in Japan.

All within this square

Certified disease by Japanese

compensation system

Fig. 4. Air pollutants and their affected diseases

**2.3.3 Health effects caused by air pollution** 

relationship to health effects.

SPM

SO2

NO2

HC

CO

These two hypotheses are expressed by the following 5 equations. The annual average value NO2 of residential areas depends on emission volumes including those from stationary and mobile sources (hypothesis 1) in Eq. 1. The health effects (certified patients, mortality rates, recovery, and newly registered patients) depend on NO2 level (hypothesis 2) from Eq. 2 to Eq. 5.

The emission volume is divided by the surface area in order to compare the health effects. The number of certified patients is divided by the total population because this group becomes patient group from normal population group. Certified patient includes newly registered and re-registered patients. The numbers of deceased and recovered are divided by the number of certified patients because these groups change status from certified patients. The number of newly registered patients is divided by the number of population because this group becomes patient group from normal population group.

Methodology of hypothesis 1 (Emission-air quality)

$$M\_{t,c} = \mathcal{a} \ln X\_{t,c} + \mathcal{c} \tag{1}$$

M: Annual average value of NO2 in residential area (ppm)

X: Emission volume of NOx (stationary + mobile sources) (t/year/km2)

A Methodology of Estimation on Air Pollution and Its Health Effects in Large Japanese Cities 153

government. The outlier for Osaka was 1972. It is also the highest level of emission volume

These data were checked by the Tokyo metropolitan government and the Osaka city local government for verification. There are multiple data sources from different research projects for the same year. The criterion for selection is that the estimation must represent the actual

Data on total NOx emission volumes (stationary and mobile sources) in Tokyo (23 wards) [Environmental Bureau of Tokyo Metropolitan Government, 1971, 1973, 1984, 1994, 1996, 2000a, 2000b, 2004, 2005, 2009, Tokyo Metropolitan Government, 1983, 1995, 1998], Kawasaki [Kawasaki City Government 2008, 2009], and Osaka [Osaka City Government, 1975, 1976, 1991, 2001, 2004, 2007] from 1972 to 2007 are available from various annual

> Air quality (NO2,SPM, CO,HC, SO2)

Certified

(Whole city) ●●●●●

(Whole city) ●●●●●

● ●

The annual average values of the pollutants (NO2) in residential areas in Tokyo (All 23 wards), [National Institute for Environmental Studies. http://www-gis.nies.go.jp/], Kawasaki [Kawasaki City Government 2008 a, 2008 b, 2009, Kanagawa Prefecture Government 1983], and Osaka [Osaka City Government, 1975, 1976, 1991, 2001, 2004, 2007]

Patient Mortality Recovery

●●●●●

●●●●●

Newly registered patient

situation without any policy implementation assumption.

environmental reports written by local governments.

Emission Volume (NOx, PM)

● (Tokyo 23 wards)

●

●

● (Hyogo prefecture)

Data source

Environmental Bureau of Tokyo Metropolitan Government

Kawasaki City Government, Kanagawa Prefecture Government

Government

Kita Kyushu Government, Fukuoka Prefecture Government

from 1960 to 2007 were used.

Emission volume = stationary + mobile sources Table 1. Data sources of each item (1974–2007)

Amagasaki City Government. Environment in Amagasaki, Hyougo Prefecture Government

Tokyo (19 wards)

Kawasaki

Amagasaki

**3.2 Air quality** 

Kita Kyushu

Osaka Osaka City

among the Osaka data.

t: year (t = 33)

```
Tokyo: 1974–1976,1980,1985,1990,1995,1997 
    Kawasaki:1985,1993,1995,1997,2000,2005 
    Osaka:1972,1978,1980,1988,1990,1992,1994,1996,1998,2000,2002–2006 
c: certified area (n = 1–3)
```
Methodology of hypothesis 2 (Air quality and health effects)

*QDRN* ,,, , are estimated from the coefficients and NO2

$$Q\_{t,r} = a\_1 M\_{t,r} + b\_1 \tag{2}$$

$$D\_{t,r} = a\_2 M\_{t,r} + b\_2 \tag{3}$$

$$R\_{t,r} = a\_3 M\_{t,r} + b\_3 \tag{4}$$

$$N\_{t,r} = a\_4 M\_{t,r} + b\_4 \tag{5}$$

Q: certified patients per 1000 population of certified areas D: mortality per 1000 certified patients of certified areas R: recovery per 1000 certified patients of certified areas N: newly certified patients per 1000 population of certified areas M: annual average value of NO2 in residential areas t: year (t = 1–33) a: coefficient (n = 1–4) b: constant (n = 1–4) c: certified area (n = 1–5)

#### **3. Data description**

#### **3.1 Emission volume**

The NOx, PM, SO2, and CO emission volumes from stationary sources have been measured since the early 1970s in Tokyo, Kawasaki, Osaka, Amagasaki, and Kita Kyushu. The data was in public by environmental white paper of each local government. The NOx and PM emission volumes from mobile sources have been measured since the 1970s primarily to evaluate policy implementation of the NOx-PM law.

The total emission data (stationary and mobile sources) were estimated for certain years in Tokyo, Kawasaki, and Osaka (Table 1). They were not continuous data of every year. The Amagasaki data were included in the total emissions from Hyogo prefecture according to the Hyogo prefecture environment white paper [Hyogo Prefecture Government 2003]. There was no separate data set available for only Amagasaki. There was no data set of emissions from mobile sources in Kita Kyushu because Kita Kyushu is not certified of the NOx-PM law [Kita Kyushu City Government, 1982]. Only the data set for emission volumes from stationary sources was available in Kita Kyushu.

There appear to be some outliers in the correlation analysis. There are four estimated values of emission volumes for 1985 in four different reports of the Tokyo metropolitan government. The outlier for Osaka was 1972. It is also the highest level of emission volume among the Osaka data.

These data were checked by the Tokyo metropolitan government and the Osaka city local government for verification. There are multiple data sources from different research projects for the same year. The criterion for selection is that the estimation must represent the actual situation without any policy implementation assumption.

Data on total NOx emission volumes (stationary and mobile sources) in Tokyo (23 wards) [Environmental Bureau of Tokyo Metropolitan Government, 1971, 1973, 1984, 1994, 1996, 2000a, 2000b, 2004, 2005, 2009, Tokyo Metropolitan Government, 1983, 1995, 1998], Kawasaki [Kawasaki City Government 2008, 2009], and Osaka [Osaka City Government, 1975, 1976, 1991, 2001, 2004, 2007] from 1972 to 2007 are available from various annual environmental reports written by local governments.


Emission volume = stationary + mobile sources

Table 1. Data sources of each item (1974–2007)

#### **3.2 Air quality**

152 Air Quality – Monitoring and Modeling

The NOx, PM, SO2, and CO emission volumes from stationary sources have been measured since the early 1970s in Tokyo, Kawasaki, Osaka, Amagasaki, and Kita Kyushu. The data was in public by environmental white paper of each local government. The NOx and PM emission volumes from mobile sources have been measured since the 1970s primarily to

The total emission data (stationary and mobile sources) were estimated for certain years in Tokyo, Kawasaki, and Osaka (Table 1). They were not continuous data of every year. The Amagasaki data were included in the total emissions from Hyogo prefecture according to the Hyogo prefecture environment white paper [Hyogo Prefecture Government 2003]. There was no separate data set available for only Amagasaki. There was no data set of emissions from mobile sources in Kita Kyushu because Kita Kyushu is not certified of the NOx-PM law [Kita Kyushu City Government, 1982]. Only the data set for emission volumes from

There appear to be some outliers in the correlation analysis. There are four estimated values of emission volumes for 1985 in four different reports of the Tokyo metropolitan

*Q aM b tr tr* , 1, 1 (2)

*D aM b tr tr* , 2, 2 (3)

*R aM b tr tr* , 3, 3 (4)

*N aM b tr tr* , 4, 4 (5)

Osaka:1972,1978,1980,1988,1990,1992,1994,1996,1998,2000,2002–2006

t: year (t = 33)

t: year (t = 1–33) a: coefficient (n = 1–4) b: constant (n = 1–4) c: certified area (n = 1–5)

**3. Data description 3.1 Emission volume** 

c: certified area (n = 1–3)

Tokyo: 1974–1976,1980,1985,1990,1995,1997 Kawasaki:1985,1993,1995,1997,2000,2005

Methodology of hypothesis 2 (Air quality and health effects)

*QDRN* ,,, , are estimated from the coefficients and NO2

Q: certified patients per 1000 population of certified areas D: mortality per 1000 certified patients of certified areas R: recovery per 1000 certified patients of certified areas

M: annual average value of NO2 in residential areas

evaluate policy implementation of the NOx-PM law.

stationary sources was available in Kita Kyushu.

N: newly certified patients per 1000 population of certified areas

The annual average values of the pollutants (NO2) in residential areas in Tokyo (All 23 wards), [National Institute for Environmental Studies. http://www-gis.nies.go.jp/], Kawasaki [Kawasaki City Government 2008 a, 2008 b, 2009, Kanagawa Prefecture Government 1983], and Osaka [Osaka City Government, 1975, 1976, 1991, 2001, 2004, 2007] from 1960 to 2007 were used.

A Methodology of Estimation on Air Pollution and Its Health Effects in Large Japanese Cities 155

Figure 7 shows the correlation between total PM emission (stationary and mobile sources) and SPM in Tokyo, Kawasaki and Osaka. The data for Tokyo (B) were estimated in 1975. The data in (A) were estimated in 1990. Both PM emission and SPM have improved from (B) to (A). The t-value of PM emission volume is 2.225, which is sufficiently high. The Osaka

Tokyo (23 wards) Kawasaki (whole city)

B

Y = 0.009 ln(X) +0.019 (2.225) (2.134) Adjusted R2 =0.300

A

0 5 10 15 20

PM emission volume (stationary+mobile sources) (t/year/km2)

Fig. 7. Correlation between total PM emissions and SPM in Tokyo (23 wards), Kawasaki-

The second hypothesis focuses on the correlation between air quality and health effects (number of certified patients, recovered patients, new patients, and total mortality figures) in the areas that were certified as Class 1 under the compensation law. Figure 8 shows NO2 and certified patients in the 5 Japanese cities. The range of NO2 is from 0.02 ppm to 0.06 ppm for the Tokyo data. The Ohta ward (1974), the Itabashi ward (1987, 1988, 1991–1995, 2001), and the Kita ward(1975, 1991, 1992, 1995) have higher annual averages compared to the rest of the Tokyo wards, Kawasaki, Osaka, Amagasaki, and Kita Kyushu. The highest number of certified patients per 1000 people was in the Ohta ward (26 persons, 1974). The

Some values are outside the general trend because of the data instabilities caused by the introduction of the compensation system and air quality measurements. In fact, many experts have pointed out that the methodology for air quality measurements was not wellestablished around 1974, when the permissible limits were introduced. Owing to the variability of the measurements, different results were recorded in different governmental documents. Further, the data for certified patients, mortality, recovery, and new patients for that year were not stable as a result of the introduction of the compensation system in 1974.

C

data of 1998 (C) seem to be lower than those of Tokyo.

data from 1974 and 1975 are outliers; in the early 1970s.

Osaka (whole city)

0.02

0.03

0.04

SPM (mg/m3)

city, and Osaka-city

**4.2 Hypothesis 2** 

0.05

0.06

#### **3.3 Certified, mortality, recovered and newly registered patients**

We used the records of certified, deceased, recovered patients, and newly registered for correlation analysis. As certified areas, the data from Tokyo (19 wards), Kawasaki (Saiwai, Kawasaki wards), Osaka city, Amagasaki city (East region and South regions), and Kita Kyushu city (Wakamatsu, Tobata, Yawata, Nijima, Kurosaki, and Higashi-Kokura) were used.

Some data regarding mortality rates, newly registered patients, and recovered patients in the 19 wards of Tokyo are missing. The term 'recovered patients' includes actual recovery, non-renewal of registration, and cancelation of registration.

#### **4. Correlation analysis**

#### **4.1 Hypothesis 1**

To support hypothesis 1, the NOx and PM emission volume and their air quality data are analyzed to determine a direct relationship.

Figure 6 shows the correlation between the total NOx emission (stationary and mobile sources) and NOx air quality (residential area, annual average) in Tokyo, Kawasaki, and Osaka. The data for emission volumes (stationary and mobile sources) were obtained from the respective local governments. A log-linear regression model is assumed. The t-value of NOx emission volume is 5.660, which is sufficiently high.

Fig. 6. Correlation between total NOx emissions and NO2 in Tokyo (23 wards), Kawasakicity, and Osaka-city

Figure 7 shows the correlation between total PM emission (stationary and mobile sources) and SPM in Tokyo, Kawasaki and Osaka. The data for Tokyo (B) were estimated in 1975. The data in (A) were estimated in 1990. Both PM emission and SPM have improved from (B) to (A). The t-value of PM emission volume is 2.225, which is sufficiently high. The Osaka data of 1998 (C) seem to be lower than those of Tokyo.

Fig. 7. Correlation between total PM emissions and SPM in Tokyo (23 wards), Kawasakicity, and Osaka-city

#### **4.2 Hypothesis 2**

154 Air Quality – Monitoring and Modeling

We used the records of certified, deceased, recovered patients, and newly registered for correlation analysis. As certified areas, the data from Tokyo (19 wards), Kawasaki (Saiwai, Kawasaki wards), Osaka city, Amagasaki city (East region and South regions), and Kita Kyushu city (Wakamatsu, Tobata, Yawata, Nijima, Kurosaki, and Higashi-Kokura) were

Some data regarding mortality rates, newly registered patients, and recovered patients in the 19 wards of Tokyo are missing. The term 'recovered patients' includes actual recovery,

To support hypothesis 1, the NOx and PM emission volume and their air quality data are

Figure 6 shows the correlation between the total NOx emission (stationary and mobile sources) and NOx air quality (residential area, annual average) in Tokyo, Kawasaki, and Osaka. The data for emission volumes (stationary and mobile sources) were obtained from the respective local governments. A log-linear regression model is assumed. The t-value of

No.1

No.3

y = 0.005ln(x) + 0.009 (5.660) (2.190) Adjusted R² = 0.525

No.2

Tokyo (23 wards) Kawasaki (whole city)

0 50 100 150 200 250 300

NOx emission volume (stationary+mobile sources) (t/year/km2)

Fig. 6. Correlation between total NOx emissions and NO2 in Tokyo (23 wards), Kawasaki-

**3.3 Certified, mortality, recovered and newly registered patients** 

non-renewal of registration, and cancelation of registration.

NOx emission volume is 5.660, which is sufficiently high.

Osaka (whole city)

used.

**4. Correlation analysis** 

0.02

0.03

0.04

NO2 (ppm)

city, and Osaka-city

0.05

analyzed to determine a direct relationship.

**4.1 Hypothesis 1** 

The second hypothesis focuses on the correlation between air quality and health effects (number of certified patients, recovered patients, new patients, and total mortality figures) in the areas that were certified as Class 1 under the compensation law. Figure 8 shows NO2 and certified patients in the 5 Japanese cities. The range of NO2 is from 0.02 ppm to 0.06 ppm for the Tokyo data. The Ohta ward (1974), the Itabashi ward (1987, 1988, 1991–1995, 2001), and the Kita ward(1975, 1991, 1992, 1995) have higher annual averages compared to the rest of the Tokyo wards, Kawasaki, Osaka, Amagasaki, and Kita Kyushu. The highest number of certified patients per 1000 people was in the Ohta ward (26 persons, 1974). The data from 1974 and 1975 are outliers; in the early 1970s.

Some values are outside the general trend because of the data instabilities caused by the introduction of the compensation system and air quality measurements. In fact, many experts have pointed out that the methodology for air quality measurements was not wellestablished around 1974, when the permissible limits were introduced. Owing to the variability of the measurements, different results were recorded in different governmental documents. Further, the data for certified patients, mortality, recovery, and new patients for that year were not stable as a result of the introduction of the compensation system in 1974.

A Methodology of Estimation on Air Pollution and Its Health Effects in Large Japanese Cities 157

Kyushu. A dummy variable was introduced when the recovery figure was more than 40 and the level of NO2 was more than 0.035 ppm (Osaka, 1979, 1980, 1981, and 1982; Kita Kyushu, 1977, 1980, and 1983). The R2 and t-values improved when the dummy variable was applied.

Kawasaki Osaka Amagasaki Kita Kyusyu

1974-1988 y=299.900x+11.781 (2.582) (3.363) Adjusted R2 = 0.098

1974-1988 with dummy y=537.922x+15.707d-10.063 (5.520) (6.049) (-2.235)

Adjusted R2 = 0.468

1974-1988 with dummy y=-622.152x -29.399d+66.369 (-1.679) (-6.342) (5.133)

1974-1988

y=614.598x+10.524 (1.484) (0.845) Adjusted R2=0.021

Adjusted R2 = 0.433

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07

**NO2 (ppm)**

Kawasaki Osaka Amagasaki Kita Kyusyu

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07

**NO2 (ppm)**

Fig. 10. Correlation: NO2 and recovery in Kawasaki‐city, Osaka‐city, Amagasaki‐city and

Fig. 9. Correlation: NOx and mortality in Kawasaki‐city, Osaka‐city, Amagasaki‐city and

0

Kita Kyushu-city (1974–2007)

50

100

**Number of mortality per** 

**1000 certified patients of 4 certificated** 

**areas (person)**

Kita Kyushu-city

**Number of recovery value per 1000** 

**certified patients of 4 certificated areas** 

**(person)**

An NO2 dummy variable is applied when NO2 exceeds 0.05 ppm and when the number of certified patients exceeds 5 persons per 1000 population. When the dummy variable is applied, the inverse relationship and R2 improve.

Fig. 8. Correlation: NO2 and certified patients in Tokyo (19 wards), Kawasaki-city, Osaka-city, Amagasaki-city and Kita Kyushu-city (1974-2007)

Figure 9 shows NOx and mortality values of certified patients in Kawasaki, Osaka, Amagasaki, and Kita Kyushu. The air quality is between 0.02 ppm and 0.04 ppm. The range of annual mortality numbers per 1000 certified patients is 4–32 in Kawasaki. There were 4 per 1000 certified patients in 1974 in Osaka. The number increased to 11 per 1000 certified patients in 1975. The mortality rate became stable at approximately 30 per 1000 certified patients after 1976. Osaka has three outlier years: 1974, 1975, and 1976. The Amagasaki data from 1983 to 1999 are missing. The mortality figure was 33 per 1000 certified patients in 1974. The number stabilised to around 20 from 1975 to 1982. The range of mortality increased from 27 to 33 between 2001 and 2007. The mortality figures were between 13 and 24 in Kita Kyushu. When the dummy variables were introduced for the three Osaka outliers from 1974 to 1988, the R2 and t-values improved and the correlation estimation reached a significant level.

Figure 10 shows NOx and recovery values of certified patients in Kawasaki, Osaka, Amagasaki, and Kita Kyushu. The recovery rates increased from 2 people per 1000 certified patients in 1975 to 47 people per 1000 certified patients in 1976 in Kawasaki. The number of recovery cases decreased, reaching 6 recoveries in 1997. The number of recoveries went up (to 22 recoveries) and down (to 2 recoveries) after 1998. There were 29 recovery cases in 1974 in Osaka. The number decreased to 6 in 1975. The number increased to 80 recovery cases per 1000 patients in 1980. There is no 'recovery' data in Amagasaki. It includes only non-renewal of registration and retirement cases. Recovery decreased from 30 people in 1970 to 0 in 2002. The data exhibit a significant reduction—from 74 people in 1974 to 10 people in 2004 in Kita

An NO2 dummy variable is applied when NO2 exceeds 0.05 ppm and when the number of certified patients exceeds 5 persons per 1000 population. When the dummy variable is

Kita Kyushu 1974-1988

Y= 141.887x+7.543 (-0.702) (3.927) Adjusted R2 = -0.007

1974-1988 with dummy Y= 192.182x-7.790d+1.266 (2.820) (-5.130) (0.621)

Adjusted R2 = 0.274

Tokyo (19 wards) Kawasaki Osaka Amagasaki

0.00 0.01 0.02 0.03 0.04 0.05 0.06

Fig. 8. Correlation: NO2 and certified patients in Tokyo (19 wards), Kawasaki-city,

Osaka-city, Amagasaki-city and Kita Kyushu-city (1974-2007)

**NO2 (ppm)**

Figure 9 shows NOx and mortality values of certified patients in Kawasaki, Osaka, Amagasaki, and Kita Kyushu. The air quality is between 0.02 ppm and 0.04 ppm. The range of annual mortality numbers per 1000 certified patients is 4–32 in Kawasaki. There were 4 per 1000 certified patients in 1974 in Osaka. The number increased to 11 per 1000 certified patients in 1975. The mortality rate became stable at approximately 30 per 1000 certified patients after 1976. Osaka has three outlier years: 1974, 1975, and 1976. The Amagasaki data from 1983 to 1999 are missing. The mortality figure was 33 per 1000 certified patients in 1974. The number stabilised to around 20 from 1975 to 1982. The range of mortality increased from 27 to 33 between 2001 and 2007. The mortality figures were between 13 and 24 in Kita Kyushu. When the dummy variables were introduced for the three Osaka outliers from 1974 to 1988, the R2 and t-values improved and the correlation estimation reached a

Figure 10 shows NOx and recovery values of certified patients in Kawasaki, Osaka, Amagasaki, and Kita Kyushu. The recovery rates increased from 2 people per 1000 certified patients in 1975 to 47 people per 1000 certified patients in 1976 in Kawasaki. The number of recovery cases decreased, reaching 6 recoveries in 1997. The number of recoveries went up (to 22 recoveries) and down (to 2 recoveries) after 1998. There were 29 recovery cases in 1974 in Osaka. The number decreased to 6 in 1975. The number increased to 80 recovery cases per 1000 patients in 1980. There is no 'recovery' data in Amagasaki. It includes only non-renewal of registration and retirement cases. Recovery decreased from 30 people in 1970 to 0 in 2002. The data exhibit a significant reduction—from 74 people in 1974 to 10 people in 2004 in Kita

applied, the inverse relationship and R2 improve.

0

5

10

**Number of certified patient per 1000** 

**population of 5 certified areas** 

**(person)**

significant level.

15

Kyushu. A dummy variable was introduced when the recovery figure was more than 40 and the level of NO2 was more than 0.035 ppm (Osaka, 1979, 1980, 1981, and 1982; Kita Kyushu, 1977, 1980, and 1983). The R2 and t-values improved when the dummy variable was applied.

Fig. 9. Correlation: NOx and mortality in Kawasaki‐city, Osaka‐city, Amagasaki‐city and Kita Kyushu-city

Fig. 10. Correlation: NO2 and recovery in Kawasaki‐city, Osaka‐city, Amagasaki‐city and Kita Kyushu-city (1974–2007)

A Methodology of Estimation on Air Pollution and Its Health Effects in Large Japanese Cities 159

number of certified patients is not estimated to be significant. There is an inverse relationship between recovery rates and air quality owing to the large variance. The number of newly

Table 4 represents the results from Eq. 2 to Eq. 5 for the period before and the period after the system amendment (1974–1988 and 1989–2007, respectively). When the data are separated, the relationship between NO2 and the number of certified patients is indicated clearly, even though the t-value is still low. The number of recovered patients may affect the number of certified patients. Mortality rates and the numbers of newly registered patients

Table 5 represents the result from Eq 4 to Eq 7 with a dummy variable in separate periods

Newly

Constant 5.426 11.763 13.819 NO2 -10.115 386.387 242.825 t value -0.652 4.439\*\*\* 0.889 Adjusted R2 -0.001 0.138 -0.002

Constant 6.600 -42.715 11.781 10.524 NO2 -14.396 5396.263 299.900 614.598 t value -0.371 2.768\*\*\* 2.582\*\*\* 1.484\* Adjusted R2 -0.007 0.105 0.098 0.021

Constant 4.417 10.011 19.177 NO2 5.926 517.227 -150.001 t value 0.484 4.862\*\*\* -0.553 Adjusted R2 -0.002 0.261 -0.010

Newly

Constant 1.317 1.115 -10.063 66.369 NO2 68.730 17.384 537.922 -622.152 t value 1.466\* 2.083\*\* 5.520\*\*\* -1.679\* Dummy 3.071 -1.010 15.707 -29.399 t value 2.981\*\*\* -4.335\*\*\* 6.049\*\*\* -6.342\*\*\* Adjusted R2 0.052 0.274 0.468 0.433

Certified patients Newly registered Mortality Recovery

registered Mortality Recovery

registered Mortality Recovery

registered patients increased before the amendment of the compensation law in 1988.

(1974-1988). All explanatory variables are estimated significantly.

Table 4. Results of correlation 1974–1988 and 1989–2007 (Eq 2-Eq5)

patients

Table 5. Results of correlation 1974–1988 with dummy variable (Eq 2-Eq5)

patients

Certified

Table 3. Results of correlation 1974–2007 (Eq 2-Eq5)

are estimated to be significant.

Significance at \* 20%, \*\*5%, \*\*\*1%

Significance at \* 20%, \*\*5%, \*\*\*1%

Significance at \* 20%, \*\*5%, \*\*\*1%

Certified

1974–2007

1974–1988

1988–2007

1974–1988

Figure 11 shows NO2 levels and the number of new patients. The number of newly certified patients tended to increase before the compensation law was amended in 1989. In Kawasaki, Osaka, and Amagasaki, new patients decreased from 1974 to 1986. There were sharp increases in new patients in 1987 and 1988. A dummy value is applied for Kawasaki 1975, Osaka 1976, Osaka 1989 and Amagasaki 1975. R2 and t-values improve when the dummy value is applied.

Fig. 11. Correlation: NO2 and newly certified patients in Kawasaki‐city, Osaka‐city, Amagasaki‐city and Kita Kyushu-city (1974–1988)

#### **5. Results**

#### **5.1. Hypothesis 1 (Correlation between emission volume and air quality)**

Table 2 shows the results from Eq 1. Without adjustment by dummy, NOx concentrations related to NOx emission are estimated to be significant. Without adjustment by dummy, PM concentrations related to PM emission are estimated to be significant.


Significance at \* 20%, \*\*5%, \*\*\*1%

Table 2. Result of correlation (Eq 1)

#### **5.2 Hypothesis 2 (Correlation between air quality and health effects)**

Table 3 shows the results from Eq. 2 to Eq. 5 for the whole period 1974-2007. R2 tends to be low without adjustment with the dummy value. The correlation between the air quality and the number of certified patients is not estimated to be significant. There is an inverse relationship between recovery rates and air quality owing to the large variance. The number of newly registered patients increased before the amendment of the compensation law in 1988.

Table 4 represents the results from Eq. 2 to Eq. 5 for the period before and the period after the system amendment (1974–1988 and 1989–2007, respectively). When the data are separated, the relationship between NO2 and the number of certified patients is indicated clearly, even though the t-value is still low. The number of recovered patients may affect the number of certified patients. Mortality rates and the numbers of newly registered patients are estimated to be significant.

Table 5 represents the result from Eq 4 to Eq 7 with a dummy variable in separate periods (1974-1988). All explanatory variables are estimated significantly.


Significance at \* 20%, \*\*5%, \*\*\*1%

158 Air Quality – Monitoring and Modeling

Figure 11 shows NO2 levels and the number of new patients. The number of newly certified patients tended to increase before the compensation law was amended in 1989. In Kawasaki, Osaka, and Amagasaki, new patients decreased from 1974 to 1986. There were sharp increases in new patients in 1987 and 1988. A dummy value is applied for Kawasaki 1975, Osaka 1976, Osaka 1989 and Amagasaki 1975. R2 and t-values improve when the dummy

Kawasaki Osaka Amagasaki Kita Kyusyu

1974-1988 y=19.509x+0.108 (2.048) (0.363) Adjusted R2 = 0.051

1974-1988 with dummy y=17.384x-1.010d+1.115 (2.083) (-4.335) (3.195)

Adjusted R2 = 0.274

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07

NO2 (ppm)

Table 2 shows the results from Eq 1. Without adjustment by dummy, NOx concentrations related to NOx emission are estimated to be significant. Without adjustment by dummy, PM

NOx PM

Table 3 shows the results from Eq. 2 to Eq. 5 for the whole period 1974-2007. R2 tends to be low without adjustment with the dummy value. The correlation between the air quality and the

Constant 0.009 0.001 NO2/PM pollutant 0.005 0.009

t value 5.660\*\*\* 2.225\*\* Adjusted R2 0.525 0.300

Fig. 11. Correlation: NO2 and newly certified patients in Kawasaki‐city, Osaka‐city,

**5.1. Hypothesis 1 (Correlation between emission volume and air quality)** 

concentrations related to PM emission are estimated to be significant.

**5.2 Hypothesis 2 (Correlation between air quality and health effects)** 

value is applied.

Number of newly registered patient

per 1000 population of 4 certificated

**5. Results** 

1974–2007

Significance at \* 20%, \*\*5%, \*\*\*1% Table 2. Result of correlation (Eq 1)

areas (person)

0

Amagasaki‐city and Kita Kyushu-city (1974–1988)

1

2

3

0.5

1.5

2.5

Table 3. Results of correlation 1974–2007 (Eq 2-Eq5)


Significance at \* 20%, \*\*5%, \*\*\*1%

Table 4. Results of correlation 1974–1988 and 1989–2007 (Eq 2-Eq5)


Significance at \* 20%, \*\*5%, \*\*\*1%

Table 5. Results of correlation 1974–1988 with dummy variable (Eq 2-Eq5)

A Methodology of Estimation on Air Pollution and Its Health Effects in Large Japanese Cities 161

An amendment in the compensation system (1988) clearly affects the data trends, so this period is divided into two stages: 1974–1988 (before the amendment) and 1989–2007 (after the amendment). To increase accuracy, the coefficients were re-examined by separating the

The correlation is estimated by using a simple linear model to support hypothesis 2. Most of the coefficients would be expected to be positive; however, there were some inverse relationships. Estimates based only on air quality data may lead to overestimation or underestimation of the number of certified, of mortality, recovered patients, or newly

The relationship between certified patients and NO2 was estimated to be significant. The relationship between NO2 and the number of death was also estimated to be significant. When the air quality worsens, mortality increases. When the three outliers of Osaka were excluded, R2 and the t-value improved. The recovery data exhibited a large variance among the four areas; therefore, the relationship between NO2 and recovery rate was not significant. When the outliers of Osaka and Kita Kyushu from the end of the 1970s and the beginning of the 1980s were excluded, the estimate was significant. The recovery data may have affected the number of certified patients. The relationship between NO2 and the

The relationship between air quality (NO2) and health effects was found to be significant, which support hypothesis 2. When NO2 worsens, certified patients, mortality rates and newly certified patients increase, according to the data from 1989 to 2007 with dummy

The discussion focuses on how to choose the methodology for correlation analysis of the area, time period, pollutants, air quality data, emission volume, and health effects in the large Japanese cities. In order to capture the data related to health problems caused by air pollution, we focused on the data from the Japanese compensation system because these

This chapter establishes a methodology by paying attention to historical events of the compensation system. The previous studies suggest that the data set of respiratory disease caused by air pollution were not separated from the other risks. This research collected existent data of health effects caused by air pollution. The methodology is designed for

This research also indicates some improvement points. The further issues are the followings.

 Air Quality Index (AQI) is an index for reporting air quality with regard to public health [USEPA http://www.airnow.gov/]. A lower AQI indicates better air quality.

For the correlation analysis, it must be required adjustment on health effect data by age,

 Only outliner was detected for revision of data. It also needs to check all the data. The area used for analysis should cover all the compensation areas in Japan. Increasing

the analysis area can also raise the accuracy of the coefficients.

PM and NO2 values were converted to AQI values in all five regions.

period into two stages, namely, 1974–1988 and 1989–2007.

number of new certified patients was estimated to be significant.

data represent respiratory diseases caused by air pollution.

registered patient rates.

variable analysis.

**6.3 Conclusion and further issues** 

application to large Asian cities.

gender and population.

### **6. Resume of the chapter**

The correlation between air pollution and health effects in the large Japanese cities presents a great challenge owing to the limited availability of data on the exposure to air pollution, health effects and the uncertainty of mixed causes. In this paper, methodology for quantitative relationships (between emission volume and air quality, and air quality and health effects) is analyzed with a statistical method: the correlation of air pollution reduction policy in Japan from 1974 to 1988.

#### **6.1 Hypothesis 1 (Correlation between emission volume and air quality)**

Hypothesis 1 states that the total volume of emissions affects the air quality. Emissions volume data from stationary and mobile sources were collected only from the 23 Tokyo wards, Kawasaki (2 wards), and Osaka whole city. Amagasaki were not available. Kita Kyushu was not covered by the NOx-PM certified area. The emission volumes data from mobile sources were not available in Kita Kyushu.

For air quality data, the annual averages were collected from environmental white papers of the areas of hypothesis 1. The health effect data—Tokyo (23 wards), Kawasaki (2 wards), Osaka, Amagasaki, and Kita Kyushu—was collected.

The relationship between emission volumes and air quality was clarified using data on emission volumes and air quality in Tokyo, Kawasaki, and Osaka. The air quality (NOx, SPM) tended to worsen with increases in emission volumes (from both stationary and mobile sources). The relationship between total emissions and air quality was found to be significant, which supports hypothesis 1. The correlation analysis of emission volume, and air quality suggests that NOx levels worsen according to increases in NO2.

#### **6.2 Hypothesis 2 (Correlation between air quality and health effects)**

Hypothesis 2 states that the air quality influences the number of certified patients. The health effect data are adopted from the compensation system for two reasons. First, local governments issue environmental white papers which indicate the number of certified, deceased, newly registered, and recovered patients in the compensation system. These consistent data are useful for checking the reliability of the calculations. Second, to be a certified patient, a patient must be a non-smoker. The data include classified patients with one of the four certified diseases. In addition, the data include classified patients affected by diseases caused by air pollution.

In the compensation system, patients who suffer from one of the four certified respiratory diseases in a certified area can be registered with the local government. For air quality data, the annual averages were collected from environmental white papers of each local government. The health effect data—Tokyo (19 wards), Kawasaki (2 wards), Osaka whole city, Amagasaki, and Kita Kyushu—was collected. The annual averages were collected from environmental white papers of the areas of hypothesis 2 for air quality data.

Based on previous literature and on its strong relationship to health effects, NOx was selected for the analysis. In order to include all the pollutants, the Air Quality Index (AQI) data were also revised. The results of the AQI data confirm that PM and NOx are the main indicators for the assessment of the health effects of air pollution.

An amendment in the compensation system (1988) clearly affects the data trends, so this period is divided into two stages: 1974–1988 (before the amendment) and 1989–2007 (after the amendment). To increase accuracy, the coefficients were re-examined by separating the period into two stages, namely, 1974–1988 and 1989–2007.

The correlation is estimated by using a simple linear model to support hypothesis 2. Most of the coefficients would be expected to be positive; however, there were some inverse relationships. Estimates based only on air quality data may lead to overestimation or underestimation of the number of certified, of mortality, recovered patients, or newly registered patient rates.

The relationship between certified patients and NO2 was estimated to be significant. The relationship between NO2 and the number of death was also estimated to be significant. When the air quality worsens, mortality increases. When the three outliers of Osaka were excluded, R2 and the t-value improved. The recovery data exhibited a large variance among the four areas; therefore, the relationship between NO2 and recovery rate was not significant. When the outliers of Osaka and Kita Kyushu from the end of the 1970s and the beginning of the 1980s were excluded, the estimate was significant. The recovery data may have affected the number of certified patients. The relationship between NO2 and the number of new certified patients was estimated to be significant.

The relationship between air quality (NO2) and health effects was found to be significant, which support hypothesis 2. When NO2 worsens, certified patients, mortality rates and newly certified patients increase, according to the data from 1989 to 2007 with dummy variable analysis.

#### **6.3 Conclusion and further issues**

160 Air Quality – Monitoring and Modeling

The correlation between air pollution and health effects in the large Japanese cities presents a great challenge owing to the limited availability of data on the exposure to air pollution, health effects and the uncertainty of mixed causes. In this paper, methodology for quantitative relationships (between emission volume and air quality, and air quality and health effects) is analyzed with a statistical method: the correlation of air pollution reduction

Hypothesis 1 states that the total volume of emissions affects the air quality. Emissions volume data from stationary and mobile sources were collected only from the 23 Tokyo wards, Kawasaki (2 wards), and Osaka whole city. Amagasaki were not available. Kita Kyushu was not covered by the NOx-PM certified area. The emission volumes data from

For air quality data, the annual averages were collected from environmental white papers of the areas of hypothesis 1. The health effect data—Tokyo (23 wards), Kawasaki (2 wards),

The relationship between emission volumes and air quality was clarified using data on emission volumes and air quality in Tokyo, Kawasaki, and Osaka. The air quality (NOx, SPM) tended to worsen with increases in emission volumes (from both stationary and mobile sources). The relationship between total emissions and air quality was found to be significant, which supports hypothesis 1. The correlation analysis of emission volume, and

Hypothesis 2 states that the air quality influences the number of certified patients. The health effect data are adopted from the compensation system for two reasons. First, local governments issue environmental white papers which indicate the number of certified, deceased, newly registered, and recovered patients in the compensation system. These consistent data are useful for checking the reliability of the calculations. Second, to be a certified patient, a patient must be a non-smoker. The data include classified patients with one of the four certified diseases. In addition, the data include classified patients affected by

In the compensation system, patients who suffer from one of the four certified respiratory diseases in a certified area can be registered with the local government. For air quality data, the annual averages were collected from environmental white papers of each local government. The health effect data—Tokyo (19 wards), Kawasaki (2 wards), Osaka whole city, Amagasaki, and Kita Kyushu—was collected. The annual averages were collected from

Based on previous literature and on its strong relationship to health effects, NOx was selected for the analysis. In order to include all the pollutants, the Air Quality Index (AQI) data were also revised. The results of the AQI data confirm that PM and NOx are the main

**6.1 Hypothesis 1 (Correlation between emission volume and air quality)** 

air quality suggests that NOx levels worsen according to increases in NO2.

**6.2 Hypothesis 2 (Correlation between air quality and health effects)** 

environmental white papers of the areas of hypothesis 2 for air quality data.

indicators for the assessment of the health effects of air pollution.

**6. Resume of the chapter** 

policy in Japan from 1974 to 1988.

diseases caused by air pollution.

mobile sources were not available in Kita Kyushu.

Osaka, Amagasaki, and Kita Kyushu—was collected.

The discussion focuses on how to choose the methodology for correlation analysis of the area, time period, pollutants, air quality data, emission volume, and health effects in the large Japanese cities. In order to capture the data related to health problems caused by air pollution, we focused on the data from the Japanese compensation system because these data represent respiratory diseases caused by air pollution.

This chapter establishes a methodology by paying attention to historical events of the compensation system. The previous studies suggest that the data set of respiratory disease caused by air pollution were not separated from the other risks. This research collected existent data of health effects caused by air pollution. The methodology is designed for application to large Asian cities.

This research also indicates some improvement points. The further issues are the followings.


A Methodology of Estimation on Air Pollution and Its Health Effects in Large Japanese Cities 163

Environmental Bureau of Tokyo Metropolitan Government. Total Automotive Emission

Environmental Bureau of Tokyo Metropolitan Government. Total Emission Reduction Plan

Environmental Bureau of Tokyo Metropolitan Government. Total Emission Reduction Plan

Environmental Bureau of Tokyo Metropolitan Government. Total Emission Reduction Plan

Environmental Bureau of Tokyo Metropolitan Government. Total Emission Reduction Plan

Environmental Bureau of Tokyo Metropolitan Government. Total Emission Reduction Plan

Environmental Bureau of Tokyo Metropolitan Government. Total Emission Reduction Plan

Environmental Bureau of Tokyo Metropolitan Government. Total PM Emission Survey 2003

Environmental Protection Bureau of Tokyo Metropolitan Government. Trend of Pollution from Automobile (in Japanese). Tokyo Metropolitan Government, Tokyo, 1984. 6. Environmental Restoration and Conservation Agency of Japan (ERCA). The Pollution-

 <http://www.erca.go.jp/english/activities/ac\_02.html> (Accessed 23 July 2008). ERCA. Pollution-related Health Damage Compensation. ERCA annual report. Chapter 3. Fukuoka Prefecture Government. Fukuoka Environmental White Paper (in Japanese).

Fukuoka Prefecture Government. Fukuoka Environmental White Paper (in Japanese).

Fukuoka Prefecture Government. Fukuoka Environmental White Paper (in Japanese).

Hyougo Prefecture Government (2003). Automobile NOx and PM Emission Reduction Plan.

Iwai K, Mizuno S, Miyasaka Y and Mori T, Correlation between suspended particles in the

Kanagawa Prefecture Government. Kanagawa Environmental White Paper (in Japanese).

Kanagawa Prefecture Government. Kanagawa Environmental White Paper (in Japanese).

environmental air and causes of disease among inhabitants: Cross-sectional studies using the vital statistics and air pollution data in Japan, Environmental Research,

(in Japanese). Tokyo Metropolitan Government, Tokyo, 2000, 6-12.

related Health Damage Prevention Program. Available at

Fukuoka Prefecture, Fukuoka, 1970, 15.

Fukuoka Prefecture, Fukuoka, 1974, 26.

Fukuoka Prefecture, Fukuoka, 1978, 17.

Hyougo Prefecture Government. Hyougo, 1.

Kanagawa Prefecture Government, Kanagawa, 1973, 24.

Kanagawa Prefecture Government, Kanagawa, 1982, 209.

Tokyo, 2000.

Tokyo, 1994, 129.

Tokyo, 1996, 128.

Tokyo, 2000, 10.

Tokyo, 2004, 6.

Tokyo, 2005, 8.

Tokyo, 2009, 10.

99, 2005, 106-117.

Reduction Projection in the Future (in Japanese). Tokyo Metropolitan Government,

Promotion Management Survey (in Japanese). Tokyo Metropolitan Government,

Promotion Management Survey (in Japanese). Tokyo Metropolitan Government,

Promotion Management Survey (in Japanese). Tokyo Metropolitan Government,

Promotion Management Survey (in Japanese). Tokyo Metropolitan Government,

Promotion Management Survey (in Japanese). Tokyo Metropolitan Government,

Promotion Management Survey (in Japanese). Tokyo Metropolitan Government,

 Inverse relationships are observed in the data from Japan, but the replacement of outliers by dummy variables yields significant estimation results. These results suggest the possibility of introducing other variables—such as weather, lifestyle, and age—for further analysis. The area used for analysis should cover all the compensation areas in Japan. Increasing the analysis area can also raise the accuracy of the coefficients for large Asian cities.

#### **7. Acknowledgment**

This research was conducted as a part of the research for "Economic analysis for emission reduction measures from automobile sector in large Asian cities". I am grateful to Mr. Hirokazu Sugizaki (ERCA), Mr. Toshiro Kokaji (Safety and Environmental Bureau, Osaka Prefecture Government), Mr. Yamada (Environmental Protection Division, Kita Kyushu City Government), Mr. Satoshi Kakurai (NEXCO), Mr. Michiya Kousaka (Library of Health and Safety Center, Tokyo Metropolitan Government) for data, librarians Tokyo Metropolitan Institute for Environmental Protection for advice of data collection, many colleagues of Kashima Seminar (Chuo University), Dr. Tazuko Marikawa (JARI), Dr. Naomi Tsukue (JARI), Dr. Koichiro Ishii (Tokyo Metropolitan Institute for Environmental Protection) for discussion. I am solely responsible for all the errors in this paper.

#### **8. References**


 http://www.city.amagasaki.hyogo.jp/sogo\_annai/toukei/033amakan.html> (Accessed 20 June 2009).


 Inverse relationships are observed in the data from Japan, but the replacement of outliers by dummy variables yields significant estimation results. These results suggest the possibility of introducing other variables—such as weather, lifestyle, and age—for further analysis. The area used for analysis should cover all the compensation areas in Japan. Increasing the analysis area can also raise the accuracy of the coefficients for

This research was conducted as a part of the research for "Economic analysis for emission reduction measures from automobile sector in large Asian cities". I am grateful to Mr. Hirokazu Sugizaki (ERCA), Mr. Toshiro Kokaji (Safety and Environmental Bureau, Osaka Prefecture Government), Mr. Yamada (Environmental Protection Division, Kita Kyushu City Government), Mr. Satoshi Kakurai (NEXCO), Mr. Michiya Kousaka (Library of Health and Safety Center, Tokyo Metropolitan Government) for data, librarians Tokyo Metropolitan Institute for Environmental Protection for advice of data collection, many colleagues of Kashima Seminar (Chuo University), Dr. Tazuko Marikawa (JARI), Dr. Naomi Tsukue (JARI), Dr. Koichiro Ishii (Tokyo Metropolitan Institute for Environmental

Air Pollution Division Pollution Regulation Bureau Tokyo Metropolitan Government. Air

Amagasaki City Government. Amagasaki Statistical Yearbook (in Japanese). Amagasaki

Amagasaki City Government. Environment in Amagasaki (in Japanese). Amagasaki City

Amagasaki City Government. Environment in Amagasaki. (in Japanese). Amagasaki City

Amagasaki City Government. Environment in Amagasaki. (in Japanese). Amagasaki City

Amagasaki City Government. Environment in Amagasaki. (in Japanese). Amagasaki City

Amagasaki City Government. Pollution in Amagasaki (Amagasaki Environmental White Paper) (in Japanese), Amagasaki City Government, Amagasaki, 1974, 192-195. Amagasaki City Government. Pollution in Amagasaki. (in Japanese). Amagasaki City

Amagasaki City Government. Pollution in Amagasaki. (in Japanese). Amagasaki City

Amagasaki City Government. Pollution in Amagasaki. (in Japanese). Amagasaki City

http://www.city.amagasaki.hyogo.jp/sogo\_annai/toukei/033amakan.html>

Pollution Emission Survey (in Japanese). Tokyo Metropolitan Government, Tokyo,

Protection) for discussion. I am solely responsible for all the errors in this paper.

City Government, Amagasaki, 2008, 129-130.

Government, Amagasaki, 1998, 125-128.

Government, Amagasaki, 2006, 129-130.

Government, Amagasaki, 1970, 221-228.

Government, Amagasaki, 1985, 164-165.

Government, Amagasaki, 1976, 9.

Government, Amagasaki, 2001, 197.

Government, Amagasaki, 2008, 23. Available at <

large Asian cities.

**7. Acknowledgment** 

**8. References** 

1977, 41-44.

(Accessed 20 June 2009).


<http://www.erca.go.jp/english/activities/ac\_02.html> (Accessed 23 July 2008).


A Methodology of Estimation on Air Pollution and Its Health Effects in Large Japanese Cities 165

Matsuura, I. Process of establishment of Pollution-related Health Damage Compensation

National Institute for Environmental Studies. Available at < http://www-gis.nies.go.jp/ >

Osaka City Government. Osaka City Statistical Yearbook (in Japanese). Osaka, Osaka City

Osaka City Government. Osaka City White Paper (in Japanese). Osaka, Osaka City

Osaka City Government. Osaka City White Paper (in Japanese). Osaka, Osaka City

Osaka City Government. Osaka City White Paper (in Japanese). Osaka, Osaka City

Osaka City Government. Osaka City White Paper (in Japanese). Osaka, Osaka City

Osaka City Government. Osaka City White Paper (in Japanese). Osaka, Osaka City

Osaka City Government. Osaka City White Paper (in Japanese). Osaka, Osaka City

Osaka City Government. Pollution in Osaka (Osaka Environmental White Paper) (in

Osaka City Government. Statistical data of certificated Patients and cancelation. (Personal

Pollution Bureau of Tokyo Metropolitan Government. Air Pollutants Emission Estimation Survey (in Japanese). Tokyo Metropolitan Government, March, 1971, 54-55. Pollution Regulation Bureau of Tokyo Metropolitan Government. Air Pollutant Emission

Sunya J, D Jarvis and T Gotschi, et al. Chronic bronchitis and urban air pollution in an international study. Occupational Environmental Medicine, 2006, 63, 836-843. Tokyo Metropolitan Government. Number of Certificated Patients and Mortality, Tokyo

Tokyo Metropolitan Government. Number of Certificated Patients and Mortality. Tokyo

Tokyo Metropolitan Government. Number of Certificated Patients and Mortality. Tokyo

Tokyo Metropolitan government. Pollution Prevention Plan in Tokyo Region (in Japanese),

Tokyo Metropolitan Government. Tokyo Environmental White Paper (in Japanese). Tokyo

Tokyo Metropolitan Government. Tokyo Environmental White Paper (in Japanese). Tokyo

Tokyo Metropolitan Government. Tokyo Environmental White Paper (in Japanese). Tokyo

Tokyo Metropolitan Government, Tokyo, 1978.

Metropolitan Government, Tokyo, 1983, 79.

Metropolitan Government, Tokyo, 1995, 75.

Metropolitan Government, Tokyo, 1998, 32.

Factor Survey (in Japanese). Tokyo Metropolitan Government, Tokyo,July 1973,

Health Annual Report. Tokyo Metropolitan Government, Tokyo, 1985, No.37, 247-

Health Annual Report. Tokyo Metropolitan Government, Tokyo, 1995, No.47, 276-

Health Annual Report. Tokyo Metropolitan Government, Tokyo, 1999, No.51, 272-

Japanese), Osaka City Government, Osaka, 1974, 169.

Law, Jurist, No.824, 91-97.

(Accessed 27 August 2008).

Government, 1975, 48-49.

Government, 1976, 76.

Government, 1991, 25.

Government, 2001, 36.

Government, 2004, 31.

Government, 2007, 36.

contact: 04 July 2009).

149.

262.

277.

273.

Government, Chapter 20, Table 25.


Kanagawa Prefecture Government. Kanagawa Environmental White Paper (in Japanese).

Kanagawa Prefecture Government. Kanagawa Environmental White Paper (in Japanese).

Kanagawa Prefecture Government. Kanagawa Environmental White Paper (in Japanese).

Kanagawa Prefecture Government. Kanagawa Environmental White Paper (in Japanese).

Kawasaki City Government. Kawasaki City Environmental White Paper (in Japanese).

Kawasaki City Government. Pollution in Kawasaki (Kawasaki Environmental White Paper)

Kawasaki City Government. Pollution in Kawasaki (Kawasaki Environmental White Paper)

Kawasaki City Government. Statistics in Kawasaki (in Japanese). Kawasaki City

<http://www.city.kawasaki.jp/20/20tokei/home/tokeisyo/tokeisyo20/html/mo

Kita Kyushu City Government. Environment in Kita Kyushu (in Japanese). Appendix, Kita

Kita Kyushu City Government. Environment in Kita Kyushu (in Japanese). Kita Kyushu

Kita Kyushu City Government. Kita Kyushu Air Pollution Prevention Plan (in Japanese).

<http://www.city.kitakyushu.jp/pcp\_portal/PortalServlet?DISPLAY\_ID=DIREC

Kita Kyushu City Government. Pollution in Kita Kyushu (in Japanese). Appendix, Kita

Kita Kyushu City Government. Pollution in Kita Kyushu (Kita Kyushu Environmental

Kita Kyushu City Government. Pollution in Kita Kyushu. Kita Kyushu City Government,

Kita Kyushu City Government. Pollution in Kita-Kyushu (in Japanese). Kita Kyushu City

Kita Kyushu City Government. Pollutiont in Kita-Kyushu (in Japanese). Kita Kyushu City

Makino,K. Mortality from Lung Cancer and Air pollution in Tokyo. Environmental Science,

Matsuura, I. Process of establishment of Pollution-related Health Damage Compensation

Matsuura, I. Process of establishment of Pollution-related Health Damage Compensation

T&NEXT\_DISPLAY\_ID=U000004&CONTENTS\_ID=1964> (Accessed 20 June

White Paper) (in Japanese), Kita Kyushu City Government, Kita Kyushu, 1974, 246-

(in Japanese), Kawasaki City Government, Kawasaki, 1994, 103.

(in Japanese). Kawasaki City Government, Kawasaki, 1993, 26.

Kanagawa Prefecture Government, Kanagawa, 1983, 398.

Kanagawa Prefecture Government, Kanagawa, 1985, 187.

Kanagawa Prefecture Government, Kanagawa, 1988, 105.

Kanagawa Prefecture Government, Kanagawa, 1994,129. Kawasaki City Government. Environment in Kawasaki (in Japanese). 2008, 12.

Kawasaki City Government, Kawasaki, 2009, 12.

Government, Kawasaki, 2008 Available at

Kyushu City Government, Kita Kyushu, 2008, 11.

Kyushu City Government, Kita Kyushu, 1987, 215.

Government, Kita Kyushu, 1981, Appendix 243.

Government, Kita Kyushu, Appendix 1974, 87.

City Government, Kita Kyushu, 2008, Appendix 11-73.

kuji.htm> (Accessed 2 May 2009).

Kita Kyushu, 1982, Appendix 19.

Law, Jurist, 1984, No.821, 29-35.

Law, Jurist, No.822, 80-85.

2007. Available at

4,1, 1996, 025-036.

2009).

247.


**8** 

*México* 

**Evaluation of Dry Atmospheric** 

Cerón-Bretón Rosa María1,

*Botanical Garden, Carmen City, Campeche,* 

*National Institute of Ecology, México City,* 

**Deposition in Two Sites in the Vicinity** 

Cerón-Bretón Julia Griselda1, Cárdenas-González Beatriz2,

Díaz-Morales Berenice1 and Muriel-García Manuel3

*2National Center for Training and Environmental Research,* 

*3Mexican Institute of Petroleum, Carmen City, Campeche,* 

**of Fuel Oil-Fired Power Plants in Mexico** 

Ortínez-Álvarez José Abraham2, Carballo-Pat Carmen Guadalupe1,

*1Research Center on Environmental Sciences, Autonomous University of Carmen,* 

Airborne pollutants are involved in a complex cycle in the atmosphere that begins with their release and finishes with their removal and deposition on terrestrial and aquatic ecosystems. Some pollutants as dust or crustal particles remain without changes during this cycle; however, gaseous components undergo physical and chemical changes. This is the case of combustion gases such as SO2 and NOx. It is well known that atmospheric particles and gases including those pollutants emitted from natural (biogenic, volcanic activity, marine aerosol, and so on) and anthropogenic sources (vehicular and industrial emissions) are removed by wet and dry deposition processes. The phenomenon of acidic atmospheric deposition has been largely studied in different sites and locations around the world during the last years (Tsitouridou & Anatolaki, 2007), and it has been found that episodes of acid precipitation are directly related to industrial emissions, this is the case of coal and fuel

In Mexico, 72 % of generated energy comes from burning of fossil fuels ("combustoleo"1, natural gas and coal). Particularly, in the last decade, "combustoleo" has been the fuel more used in the main productive sectors. Currently, 66.8% of energy is generated from power plants burning "combustoleo", resulting in fly ash emission, and large amounts of gases considered as acid precursors (such as SO2 and NOx) that contribute to the acid precipitation in the vicinity of the plants (Zuk et al., 2006). Considering that the residence times in the atmosphere for both acid precursors are different, their deposition rates are different too. Thus,

1 It is a blend of residua and gas oil cutter stock and meets the viscosity requirements of ASTM D396. It

is commonly used in power plants in large steam boilers, drying kilns and ovens.

**1. Introduction** 

burning power plants (Flues et al., 2002).


World Health Organization. The World Health Report. WHO, Geneva

### **Evaluation of Dry Atmospheric Deposition in Two Sites in the Vicinity of Fuel Oil-Fired Power Plants in Mexico**

Cerón-Bretón Rosa María1,

Cerón-Bretón Julia Griselda1, Cárdenas-González Beatriz2, Ortínez-Álvarez José Abraham2, Carballo-Pat Carmen Guadalupe1, Díaz-Morales Berenice1 and Muriel-García Manuel3 *1Research Center on Environmental Sciences, Autonomous University of Carmen, Botanical Garden, Carmen City, Campeche, 2National Center for Training and Environmental Research, National Institute of Ecology, México City, 3Mexican Institute of Petroleum, Carmen City, Campeche, México* 

#### **1. Introduction**

166 Air Quality – Monitoring and Modeling

Tokyo Metropolitan Government. Tokyo Regional Pollution Prevention Plan (in Japanese).

Tokyo Metropolitan Government. Tokyo Statistical Yearbook. Tokyo Metropolitan

UK Department of Health (UK DH). Cardiovascular Disease and Air Pollution. A report by

http://www.airnow.gov/index.cfm?action=aqibasics.aqi > (Accessed 30 June

USEPA. The Benefits and Costs of the Clean Air Act 1990 to 2010. USEPA, Washington

the Committee on the Medical Effects of Air Pollutants. UK DH, London, 2006, 21-

Tokyo Metropolitan Government, Tokyo, 1978, 6.

US EPA, Air Quality Index, USEPA, Washington D.C. Available at <

World Health Organization. The World Health Report. WHO, Geneva

Government, 1998, 32.

137.

2009).

D.C.1999, D-58.

Airborne pollutants are involved in a complex cycle in the atmosphere that begins with their release and finishes with their removal and deposition on terrestrial and aquatic ecosystems. Some pollutants as dust or crustal particles remain without changes during this cycle; however, gaseous components undergo physical and chemical changes. This is the case of combustion gases such as SO2 and NOx. It is well known that atmospheric particles and gases including those pollutants emitted from natural (biogenic, volcanic activity, marine aerosol, and so on) and anthropogenic sources (vehicular and industrial emissions) are removed by wet and dry deposition processes. The phenomenon of acidic atmospheric deposition has been largely studied in different sites and locations around the world during the last years (Tsitouridou & Anatolaki, 2007), and it has been found that episodes of acid precipitation are directly related to industrial emissions, this is the case of coal and fuel burning power plants (Flues et al., 2002).

In Mexico, 72 % of generated energy comes from burning of fossil fuels ("combustoleo"1, natural gas and coal). Particularly, in the last decade, "combustoleo" has been the fuel more used in the main productive sectors. Currently, 66.8% of energy is generated from power plants burning "combustoleo", resulting in fly ash emission, and large amounts of gases considered as acid precursors (such as SO2 and NOx) that contribute to the acid precipitation in the vicinity of the plants (Zuk et al., 2006). Considering that the residence times in the atmosphere for both acid precursors are different, their deposition rates are different too. Thus,

<sup>1</sup> It is a blend of residua and gas oil cutter stock and meets the viscosity requirements of ASTM D396. It is commonly used in power plants in large steam boilers, drying kilns and ovens.

Evaluation of Dry Atmospheric Deposition in Two

**2.1 Los Petenes** 

and in risk of extinction species.

Sites in the Vicinity of Fuel Oil-Fired Power Plants in Mexico 169

**Tula de Allende, Hidalgo** 

**Gulf of Mexico** **Los Petenes Biosphere Reserve, Campeche** 

Fig. 1. Location of the sampling sites: Los Petenes Biosphere Reserve and Tula de Allende.

Los Petenes site is a natural protected area in a large and narrow coastline. This area covers an extension of 282,857 ha, and it is characterized by the presence of complexes habitats similar to islands with availability of sweet water during all year, called "Petenes" (Figure 2). These are individual areas of tropical forest that reach a diameter ranging from a few meters to kilometers surrounded by wetlands, with a great diversity of flora and fauna. These habitats are only found in Cuba, Florida and Mexico, generally having a "Cenote" in its central part, originating vegetation that flourishes in rings of similar centers (Figure 2). Los Petenes is located within an eco-region that includes Ria Celestún Biosphere Reserve and the Natural Protected Area El Palmar in Yucatan State. Los Petenes is a Ramsar Site and constitutes a representative bio-geographic area of an ecosystem not significantly perturbed, in which representative species of national biodiversity live, including endemic, threatened

The climate is warm sub-humid (Aw) with summer rains with the presence of the mid-summer drought, and forest fires occurring frequently. Annual rainfall ranges from 729 to 1049 mm and temperature ranges from 26.1 and 28.8 °C. Los Petenes are characterized by four types of soils: a) Sandy soils at the coast (calcareous regosols), b) holomorphic soils in marshes (histosols), c) shallow and rocky soils (rendzina type) and clay soils (gleysols). Specific sampling site was located within the Center for Conservation of Wildfire Life of Autonomous

NOx are removed in an efficient way by wet deposition, being deposited around the emission source. On the other hand, since SO2 has a residence time of 13 days, undergoes long-range transport, being deposited in distant places from the emission point. For this reason NOx are referred as local pollutants whereas SO2 is considered as a regional pollutant.

Qualitative and quantitative assessment of atmospheric deposition in regions probably impacted by anthropogenic sources is essential for understanding regional variations, to determine if critical loads are exceeded, to assess time-trends in a long-term, and to relate sources contributing to the acidity with local and regional meteorology. Long-term changes in air quality and atmospheric deposition are not obvious, in some cases due to daily and seasonal variations in parameters such as wind, temperature, precipitation and atmospheric circulation patterns which influence on dispersion, transport and deposition of pollutants. For this reason, it is important to measure wet and dry deposition fluxes at various sites all over the country during long-time periods to provide an outline of the main chemical characteristics of the deposition and to assess their trends along time. Concerning monitoring networks, air quality data are useful to determine if the current regulations are effective in the improvement of the air quality, or if it is necessary to promulgate new ones.

In 2009, National Institute of Ecology [INE] sponsored the design and operation (during two years) in a preliminary phase of the Mexican Atmospheric Deposition Network, integrated by four stations (considering both polluted and natural sites). In addition, in 2010, the Mexican Environmental Agency [SEMARNAT] supported a research project focused on the study of atmospheric deposition in five sites located in the surroundings of power plants burning "combustoleo". The aim of this work was to assess the chemical composition of dry deposition in two of these sites: Los Petenes and Tula de Allende, both sites are located in the vicinity of power plants. The first one is a Ramsar site and a Biosphere Reserve located in Yucatan Peninsula at the southeast of the country, and the second one is a polluted site located in an important industrial region of Central Mexico, in whose vicinity there are important historic monuments.

#### **2. Sampling sites description**

Dry deposition samples were collected during 2009 and 2010 at two sampling sites: Los Petenes Biosphere Reserve, Campeche and Tula de Allende, Hidalgo.


Table 1. Main features of the sampling sites.

Sampling sites were selected considering proper conditions of accessibility and safety. The specific location and the main characteristics of these sampling sites are shown in Figure 1 and Table 1.

Fig. 1. Location of the sampling sites: Los Petenes Biosphere Reserve and Tula de Allende.

#### **2.1 Los Petenes**

168 Air Quality – Monitoring and Modeling

NOx are removed in an efficient way by wet deposition, being deposited around the emission source. On the other hand, since SO2 has a residence time of 13 days, undergoes long-range transport, being deposited in distant places from the emission point. For this reason NOx are

Qualitative and quantitative assessment of atmospheric deposition in regions probably impacted by anthropogenic sources is essential for understanding regional variations, to determine if critical loads are exceeded, to assess time-trends in a long-term, and to relate sources contributing to the acidity with local and regional meteorology. Long-term changes in air quality and atmospheric deposition are not obvious, in some cases due to daily and seasonal variations in parameters such as wind, temperature, precipitation and atmospheric circulation patterns which influence on dispersion, transport and deposition of pollutants. For this reason, it is important to measure wet and dry deposition fluxes at various sites all over the country during long-time periods to provide an outline of the main chemical characteristics of the deposition and to assess their trends along time. Concerning monitoring networks, air quality data are useful to determine if the current regulations are effective in the improvement of the air quality, or if it is necessary to promulgate new ones. In 2009, National Institute of Ecology [INE] sponsored the design and operation (during two years) in a preliminary phase of the Mexican Atmospheric Deposition Network, integrated by four stations (considering both polluted and natural sites). In addition, in 2010, the Mexican Environmental Agency [SEMARNAT] supported a research project focused on the study of atmospheric deposition in five sites located in the surroundings of power plants burning "combustoleo". The aim of this work was to assess the chemical composition of dry deposition in two of these sites: Los Petenes and Tula de Allende, both sites are located in the vicinity of power plants. The first one is a Ramsar site and a Biosphere Reserve located in Yucatan Peninsula at the southeast of the country, and the second one is a polluted site located in an important industrial region of Central Mexico, in whose vicinity there are

Dry deposition samples were collected during 2009 and 2010 at two sampling sites: Los

Sampling sites were selected considering proper conditions of accessibility and safety. The specific location and the main characteristics of these sampling sites are shown in Figure 1

Biosphere Reserve 4 20° 51´ 30" N 90° 45´15" W

(m asl) Latitude Longitude

2020 20° 05´08" N 99° 33´05" W

Petenes Biosphere Reserve, Campeche and Tula de Allende, Hidalgo.

Sites Classification area Altitude

Ramsar Site

Urban and Industrial, with the presence of historic monuments in the vicinity

Table 1. Main features of the sampling sites.

referred as local pollutants whereas SO2 is considered as a regional pollutant.

important historic monuments.

Los Petenes, Campeche

> Tula de Allende, Hidalgo

and Table 1.

**2. Sampling sites description** 

Los Petenes site is a natural protected area in a large and narrow coastline. This area covers an extension of 282,857 ha, and it is characterized by the presence of complexes habitats similar to islands with availability of sweet water during all year, called "Petenes" (Figure 2). These are individual areas of tropical forest that reach a diameter ranging from a few meters to kilometers surrounded by wetlands, with a great diversity of flora and fauna. These habitats are only found in Cuba, Florida and Mexico, generally having a "Cenote" in its central part, originating vegetation that flourishes in rings of similar centers (Figure 2). Los Petenes is located within an eco-region that includes Ria Celestún Biosphere Reserve and the Natural Protected Area El Palmar in Yucatan State. Los Petenes is a Ramsar Site and constitutes a representative bio-geographic area of an ecosystem not significantly perturbed, in which representative species of national biodiversity live, including endemic, threatened and in risk of extinction species.

The climate is warm sub-humid (Aw) with summer rains with the presence of the mid-summer drought, and forest fires occurring frequently. Annual rainfall ranges from 729 to 1049 mm and temperature ranges from 26.1 and 28.8 °C. Los Petenes are characterized by four types of soils: a) Sandy soils at the coast (calcareous regosols), b) holomorphic soils in marshes (histosols), c) shallow and rocky soils (rendzina type) and clay soils (gleysols). Specific sampling site was located within the Center for Conservation of Wildfire Life of Autonomous

Evaluation of Dry Atmospheric Deposition in Two

**2.2 Tula de Allende** 

Allende.

Sites in the Vicinity of Fuel Oil-Fired Power Plants in Mexico 171

The second study site is Tula de Allende, Hidalgo located at the Central Region of Mexico, at 20°03' N and 99° 20' W with an altitude of 2020 m asl, at 70 km from Mexico City and 3 km from the fuel oil-fired power plant "Francisco Pérez Ríos" located along the Tula-Vito-Apaxco industrial complex. Due to large volumes of atmospheric pollutants generated by the Tula-Vito-Apaxco corridor, this region has been classified as a critical zone by the Mexican environmental authorities. Industries located in this area such as an oil refinery, a plant energy (which is one of biggest power plants in Latin America), and two important

Tula

Vito

Apaxco

cement plants, contribute to the development of the region (Figure 3).

Fig. 3. Location of the Tula-Vito-Apaxco Industrial Complex in the vicinity of Tula de

According to the latest information from the environmental authorities, 323, 000 tons per year of SO2 and 44, 000 tons per year of NOx are released in this region. The main sources are the "Miguel Hidalgo" Oil Refinery and the "Francisco Pérez Ríos" Power Plant. "Miguel Hidalgo" Refinery processes 296,000 barrels per day of crude oil, representing 20% of the total refining capacity in the country. Other industries such as cement plants, open-sky mines and agricultural activities are also responsible of emission of important quantities of particulate matter. Therefore, the coexistence in the area of these industries could produce secondary compounds that can be potentially dangerous and risky for public health, ecosystems, and historic monuments. Nearby to this site, the ruins of the ancient capital city

University of Campeche at 15 km from Campeche City and at 10 km from the fuel oil-fired power plant "Lerma". Characteristics and capacity of this power plant are shown in Table 2.

Fig. 2. Panoramic view of Los Petenes Biosphere Reserve.


Source: Zuk et al., 2006.

Table 2. Main characteristics of the power plants located in the vicinity of the sampling sites considered in this study.

#### **2.2 Tula de Allende**

170 Air Quality – Monitoring and Modeling

University of Campeche at 15 km from Campeche City and at 10 km from the fuel oil-fired power plant "Lerma". Characteristics and capacity of this power plant are shown in Table 2.

Fig. 2. Panoramic view of Los Petenes Biosphere Reserve.

Lerma;

Francisco Pérez Ríos; Tula, Hidalgo

Source: Zuk et al., 2006.

considered in this study.

Power Plant, Location Units Operation starting

9 ( steam and combined

Campeche, Mexico 4 (combined cycle) September 9, 1976 150

Table 2. Main characteristics of the power plants located in the vicinity of the sampling sites

date

cycle technologies) June 30, 1991 2000

Installed effective capacity (MW)

The second study site is Tula de Allende, Hidalgo located at the Central Region of Mexico, at 20°03' N and 99° 20' W with an altitude of 2020 m asl, at 70 km from Mexico City and 3 km from the fuel oil-fired power plant "Francisco Pérez Ríos" located along the Tula-Vito-Apaxco industrial complex. Due to large volumes of atmospheric pollutants generated by the Tula-Vito-Apaxco corridor, this region has been classified as a critical zone by the Mexican environmental authorities. Industries located in this area such as an oil refinery, a plant energy (which is one of biggest power plants in Latin America), and two important cement plants, contribute to the development of the region (Figure 3).

Fig. 3. Location of the Tula-Vito-Apaxco Industrial Complex in the vicinity of Tula de Allende.

According to the latest information from the environmental authorities, 323, 000 tons per year of SO2 and 44, 000 tons per year of NOx are released in this region. The main sources are the "Miguel Hidalgo" Oil Refinery and the "Francisco Pérez Ríos" Power Plant. "Miguel Hidalgo" Refinery processes 296,000 barrels per day of crude oil, representing 20% of the total refining capacity in the country. Other industries such as cement plants, open-sky mines and agricultural activities are also responsible of emission of important quantities of particulate matter. Therefore, the coexistence in the area of these industries could produce secondary compounds that can be potentially dangerous and risky for public health, ecosystems, and historic monuments. Nearby to this site, the ruins of the ancient capital city

Evaluation of Dry Atmospheric Deposition in Two

were probably present but were not measured.

**3.3 Quality assurance** 

Allende.

0

collected at Tula de Allende.

compounds were not specifically analyzed in this work.

100

200

300

400

Σ Anions (µEq l

 -1)

500

600

700

Sites in the Vicinity of Fuel Oil-Fired Power Plants in Mexico 173

were prepared by dilution of certified standards (J.T. Baker). The detection limits were

Repeatability was guaranteed by 3 replicate measurements for each sample. Results showed a coefficient of variation < 5% for all elements measured. To verify whether major components had been measured in the analysis, two kinds of quality control were applied. If linear regression analysis between measured anions and cations gives a regression coefficient minor than 0.9, it indicates that anions and organic acids (formate and acetate)

On the other hand, if linear regression analysis between the measured and the calculated conductivity gives a regression coefficient major than 0.9, it indicates that practically the majority of the ions were analyzed. Balance cation-anion obtained for Tula de Allende showed differences less than 10%, therefore, we considered that all the major ions were analyzed. In addition, measured conductivities in Tula de Allende were in agreement with calculated conductivities. Figure 4 shows the quality of the analytical data in Tula de

> 0 200 400 600 800 1000 Σ Cations (µEq l -1)

Fig. 4. Quality control (ionic balance) of chemical analysis for dry deposition samples

In the case of Los Petenes Biosphere Reserve, conductivity differences were greater than 10%, and ionic balance suggests an ion deficit, probably due to the missing HCO3- and the presence of organic acids that are commonly found in natural sites. These short chain

calculated as three times the standard deviation of six blank samples.

of Toltecs, also known as "Tula" or as "Tollan", are located. This site is characterized by a semi-arid climate (BSh) with an annual average rainfall of 600 mm, and an average temperature ranging between 16 and 17°C.

#### **3. Experimental**

#### **3.1 Sampling**

Dry deposition samples were collected on a weekly basis for a period of two years (from August 2009 to December 2010). Procedures to collect wet deposition are available anywhere since several years (Galloway & Likens, 1976), nevertheless, standard procedures to collect dry deposition are not available. Quantifying dry atmospheric input to natural surfaces is a difficult task due to the complex chemical processes involved (Balestrini et al., 2000). However, investigations using plastic materials, Teflon® and glass surfaces to collect dry deposition have demonstrated the importance of the physical, chemical and geometrical features of the surrogate surface in selecting the depositing material (Dasch, 1985). In this study, dry deposition was assessed using a surrogate surface according to the methodology used by Alonso et al (2005). Sampling was performed using automated wet-dry deposition collectors. Samplers were equipped with two polyethylene buckets and a lid controlled by a humid sensor, which moves depending on the beginning and the end of the rain event. Three nylon filters (Nylasorb, Pall Corporation, Gelman Lab, Michigan, USA; 47 mm diameter, 1 m pore-size) per site were used as surrogate surfaces. Filters were located on a support within the dry deposition bucket and exposed in horizontal position at about 1 m height above ground. Dry deposition fluxes were calculated based on surface area and time exposure.

#### **3.2 Analysis**

After exposition, filters were stored at -18 °C until analysis. Three nylon filters were extracted with 80 ml of deionized-distilled water for 15 min in an ultrasonic bath, the first one was used to determine cations, and the second filter was used to analyze anions. Ammonium, pH and conductivity were determined on extracted solution of the third filter. pH and conductivity measurements were obtained by using precision pH meter (TERMO ORION 290) and a conductivity meter (CL 135). All plastic ware and glassware used to prepare standard solutions, for digestion of samples and for chemical analysis were rigorously washed, brushed and rinsed with distilled water. Plastic ware used for digestion of samples for analysis of cations by atomic absorption spectrophotometer was completely immersed during 24 h in a 20% ultrapure nitric acid bath (J.T. Baker, AA Grade), then rinsed several times with deionized water type I (Hycel) and sealed into double plastic bags. Before using, all material was again rinsed with deionized water. Extract from the first filter was digested in Teflon® closed flasks (Cole-Parmer) of 100 ml, using an autoclave equipment as energy source; subsequently, Na+, K+, Ca2+ and Mg2+ were analyzed using an atomic absorption spectrophotometer (Thermoscientific ICE 3300). Extract from the second filter was analyzed for SO42-, Cl and NO3- by ion chromatography (Agilent 1100) with 200 µl sample loop. Finally, extracted solution of the third filter was used to determine NH4+ colorimetrically by means of a spectrophotometer UV (HACH DR 2800). Standard solutions were prepared by dilution of certified standards (J.T. Baker). The detection limits were calculated as three times the standard deviation of six blank samples.

#### **3.3 Quality assurance**

172 Air Quality – Monitoring and Modeling

of Toltecs, also known as "Tula" or as "Tollan", are located. This site is characterized by a semi-arid climate (BSh) with an annual average rainfall of 600 mm, and an average

Dry deposition samples were collected on a weekly basis for a period of two years (from August 2009 to December 2010). Procedures to collect wet deposition are available anywhere since several years (Galloway & Likens, 1976), nevertheless, standard procedures to collect dry deposition are not available. Quantifying dry atmospheric input to natural surfaces is a difficult task due to the complex chemical processes involved (Balestrini et al., 2000). However, investigations using plastic materials, Teflon® and glass surfaces to collect dry deposition have demonstrated the importance of the physical, chemical and geometrical features of the surrogate surface in selecting the depositing material (Dasch, 1985). In this study, dry deposition was assessed using a surrogate surface according to the methodology used by Alonso et al (2005). Sampling was performed using automated wet-dry deposition collectors. Samplers were equipped with two polyethylene buckets and a lid controlled by a humid sensor, which moves depending on the beginning and the end of the rain event. Three nylon filters (Nylasorb, Pall Corporation, Gelman Lab, Michigan, USA; 47 mm diameter, 1 m pore-size) per site were used as surrogate surfaces. Filters were located on a support within the dry deposition bucket and exposed in horizontal position at about 1 m height above ground. Dry deposition fluxes were calculated based on surface area and time

After exposition, filters were stored at -18 °C until analysis. Three nylon filters were extracted with 80 ml of deionized-distilled water for 15 min in an ultrasonic bath, the first one was used to determine cations, and the second filter was used to analyze anions. Ammonium, pH and conductivity were determined on extracted solution of the third filter. pH and conductivity measurements were obtained by using precision pH meter (TERMO ORION 290) and a conductivity meter (CL 135). All plastic ware and glassware used to prepare standard solutions, for digestion of samples and for chemical analysis were rigorously washed, brushed and rinsed with distilled water. Plastic ware used for digestion of samples for analysis of cations by atomic absorption spectrophotometer was completely immersed during 24 h in a 20% ultrapure nitric acid bath (J.T. Baker, AA Grade), then rinsed several times with deionized water type I (Hycel) and sealed into double plastic bags. Before using, all material was again rinsed with deionized water. Extract from the first filter was digested in Teflon® closed flasks (Cole-Parmer) of 100 ml, using an autoclave equipment as energy source; subsequently, Na+, K+, Ca2+ and Mg2+ were analyzed using an atomic absorption spectrophotometer (Thermoscientific ICE 3300). Extract from the second filter

sample loop. Finally, extracted solution of the third filter was used to determine NH4+ colorimetrically by means of a spectrophotometer UV (HACH DR 2800). Standard solutions

and NO3- by ion chromatography (Agilent 1100) with 200 µl

temperature ranging between 16 and 17°C.

**3. Experimental** 

**3.1 Sampling** 

exposure.

**3.2 Analysis** 

was analyzed for SO4

2-, Cl-

Repeatability was guaranteed by 3 replicate measurements for each sample. Results showed a coefficient of variation < 5% for all elements measured. To verify whether major components had been measured in the analysis, two kinds of quality control were applied. If linear regression analysis between measured anions and cations gives a regression coefficient minor than 0.9, it indicates that anions and organic acids (formate and acetate) were probably present but were not measured.

On the other hand, if linear regression analysis between the measured and the calculated conductivity gives a regression coefficient major than 0.9, it indicates that practically the majority of the ions were analyzed. Balance cation-anion obtained for Tula de Allende showed differences less than 10%, therefore, we considered that all the major ions were analyzed. In addition, measured conductivities in Tula de Allende were in agreement with calculated conductivities. Figure 4 shows the quality of the analytical data in Tula de Allende.

Fig. 4. Quality control (ionic balance) of chemical analysis for dry deposition samples collected at Tula de Allende.

In the case of Los Petenes Biosphere Reserve, conductivity differences were greater than 10%, and ionic balance suggests an ion deficit, probably due to the missing HCO3- and the presence of organic acids that are commonly found in natural sites. These short chain compounds were not specifically analyzed in this work.

Evaluation of Dry Atmospheric Deposition in Two

NORTH

SOUTH

Reserve during the study period.

WEST EAST 3% 6% 9% 12% 15%

study period.

Sites in the Vicinity of Fuel Oil-Fired Power Plants in Mexico 175

Fig. 7. Typical wind rose and backward air mass trajectory for Tula de Allende during the

WIND SPEED (m/s) >= 11.1 8.8 - 11.1 5.7 - 8.8 3.6 - 5.7 2.1 - 3.6 0.5 - 2.1 Calms: 75.00%

Fig. 8. Typical wind rose and backward air mass trajectory for Los Petenes Biosphere

#### **3.4 Meteorological data**

Two portable meteorological stations (Davies Inc, n.d.) that were operating during the whole study period in both sampling sites provided needed meteorological information. From these meteorological data, wind roses were constructed for each site by means of WRPLOT VIEW 6.5.2 (Lakes Environmental Inc, 2011).

To trace the origin of air masses for the whole study period, air-masses backward trajectories were calculated (48 h before) for both sampling sites by means of HYSPLIT-Hybrid Single Particle Lagrangian Integrated Trajectory Model (NOAA, n.d.). From calculated backward trajectories, it was observed that prevailing winds came from N-NE for Tula de Allende (Figures 5, 7) and from E for Los Petenes Biosphere Reserve (Figures 6,8).

Fig. 5. Wind direction frequency distribution for Tula de Allende during the study period.

Fig. 6. Wind direction frequency distribution for Los Petenes Biosphere Reserve during the study period.

Two portable meteorological stations (Davies Inc, n.d.) that were operating during the whole study period in both sampling sites provided needed meteorological information. From these meteorological data, wind roses were constructed for each site by means of

To trace the origin of air masses for the whole study period, air-masses backward trajectories were calculated (48 h before) for both sampling sites by means of HYSPLIT-Hybrid Single Particle Lagrangian Integrated Trajectory Model (NOAA, n.d.). From calculated backward trajectories, it was observed that prevailing winds came from N-NE for Tula de Allende (Figures 5, 7) and from E for Los Petenes Biosphere Reserve (Figures 6,8).

N NE E SE S SW W NW

N NE E SE S SW W NW

Fig. 6. Wind direction frequency distribution for Los Petenes Biosphere Reserve during the

Fig. 5. Wind direction frequency distribution for Tula de Allende during the study period.

**3.4 Meteorological data** 

WRPLOT VIEW 6.5.2 (Lakes Environmental Inc, 2011).

0

0

5

10

**Frequency**

study period.

15

20

5

10

**Frequency**

15

20

Fig. 7. Typical wind rose and backward air mass trajectory for Tula de Allende during the study period.

Fig. 8. Typical wind rose and backward air mass trajectory for Los Petenes Biosphere Reserve during the study period.

Evaluation of Dry Atmospheric Deposition in Two

0

2

4

**Relative Frequency**

6

8

10

temporal sources.

Biosphere Reserve.

**4.1.2 Tula de Allende** 

any sample was in the acid range (pH<5.0).

process.

Sites in the Vicinity of Fuel Oil-Fired Power Plants in Mexico 177

plant located at 10 km from the sampling site. Considering the residence times for SO2 (13 days) and NO2 (1 day), it can be assumed that Los Petenes site was not subjected to the input of long-range transported pollutants and probably only had the influence of local and

5.0-5.5 5.5-6.0 6.0-6.5 >6.5

**pH**

pH values found in dry deposition samples showed a great variability, 22% of the samples were below of 5.6 (with a mean value of 5.9), and a negligible percentage (5.5%) was in the acid range (pH<5.0) (Table 3, Figure 9). In spite of that nitrate levels exceeded background hemispheric levels, pH values suggest that this site does not have acidity problems, owing to some cations probably played an important role in the neutralization process. This behavior is in agreement with the calcareous soils commonly found in Yucatan Peninsula which is constituted by alkaline particles participating in active way in the neutralization

Sulfate was the most abundant ion in Tula de Allende, probably attributed to anthropogenic sources. Tula de Allende is not a coastal site; therefore, it can be assumed that the total content of sulfate and nitrate is in excess (anthropogenic origin). Sulfate excess, [SO42-]xs) levels found (68.78 Eq l-1) exceeded almost seven times the background hemispheric levels reported by Galloway et al (1982) for remote sites (10 Eq l-1). In addition, nitrate levels (56.37 Eq l-1) also exceeded the background hemispheric levels reported for remote sites (2.8 Eq l-1). Both, nitrate and sulfate contributed to the acidity. Considering the residence times for SO2 (13 days) and NO2 (1 day), it can be assumed that this site had the influence of local sources (anthropogenic sources such as a local power plant and oil refinery) as well as the input of pollutants subjected to long-range transport process. From Table 3 and Figure 10, it can be observed that pH values found in dry deposition samples ranged from 5.3 to 6.7, with a mean value of 5.77. A great percentage of the samples (59%) was below of 5.6, but

Fig. 9. pH frequency distribution for dry deposition samples collected at Los Petenes

#### **3.5 Statistical analysis**

Pearson´s correlation analysis was applied to test the relationship among the total trace element concentrations for each sampling site. ANOVA (GLM) was performed to test the differences between each element.

Means comparison was also performed using the average concentration values for trace elements as well as pH values. The calculations were carried out using SAS 8.0 software (SAS Inc, 1998).

#### **4. Results and discussion**

#### **4.1 Chemical composition**

Mean concentration values and ionic abundance of principal ions in dry deposition at Los Petenes Biosphere Reserve and Tula de Allende are shown in Table 3.


Table 3. Mean values of pH and concentration (Eq l-1) of dry deposition samples.

#### **4.1.1 Los Petenes Biosphere Reserve**

Ammonium was the most abundant ion in Los Petenes, probably due to agriculture practices and swine husbandry in the zone and biological decay. On the other hand, nitrate was second more abundant ion, probably attributed to biomass burning activities and emissions from a local power plant. The influence of forest fires on nitrate and ammonium levels has been reported in other works (Hegg et al, 1988; Galbally & Gillel, 1988). It is well known that forest fires produce great amounts of nitrogen oxides, which are rapidly converted to nitric acid in the atmosphere. Since marine aerosol does not contribute to the nitrate levels, it can be assumed that total NO3- content present in samples was in excess, and that it had an anthropogenic origin. In spite of Los Petenes is an important natural protected area, it cannot be considered a remote site, for this reason, nitrate levels (63.38 Eq l-1) exceeded the background hemispheric levels reported for remote sites (2.8 Eq l-1).

On the other hand, since Los Petenes can be considered as a coastal site (8 km from the coastline), it was required to estimate marine aerosol contribution to the sulfate levels, in order to determine the marine and anthropogenic fractions (commonly known as sulfate excess, [SO42-]xs). Sulfate excess levels found (12.66 Eq l-1) were in agreement with the background hemispheric levels reported by Galloway et al (1982) for remote sites (10 Eq l-1), suggesting that there was a minimal influence of local sources that contributed to this ion; this is in agreement to the residence time for SO2 in the atmosphere. However, (SO42- )xs/NO3 - was lower than other coastal sites (0.031), indicating a significant enrichment of nitrates. Clearly, there was a local source influencing on nitrate levels, probably forest fires occurring as a result of the mid-summer drought, and emissions coming from the power

Pearson´s correlation analysis was applied to test the relationship among the total trace element concentrations for each sampling site. ANOVA (GLM) was performed to test the

Means comparison was also performed using the average concentration values for trace elements as well as pH values. The calculations were carried out using SAS 8.0 software

Mean concentration values and ionic abundance of principal ions in dry deposition at Los

Site pH Na+ K+ Ca2+ Mg2+ NH4+ Cl- SO42- NO3-

+>NO3-

Ammonium was the most abundant ion in Los Petenes, probably due to agriculture practices and swine husbandry in the zone and biological decay. On the other hand, nitrate was second more abundant ion, probably attributed to biomass burning activities and emissions from a local power plant. The influence of forest fires on nitrate and ammonium levels has been reported in other works (Hegg et al, 1988; Galbally & Gillel, 1988). It is well known that forest fires produce great amounts of nitrogen oxides, which are rapidly converted to nitric acid in the atmosphere. Since marine aerosol does not contribute to the

and that it had an anthropogenic origin. In spite of Los Petenes is an important natural protected area, it cannot be considered a remote site, for this reason, nitrate levels (63.38 Eq l-1) exceeded the background hemispheric levels reported for remote sites (2.8 Eq l-1).

On the other hand, since Los Petenes can be considered as a coastal site (8 km from the coastline), it was required to estimate marine aerosol contribution to the sulfate levels, in order to determine the marine and anthropogenic fractions (commonly known as sulfate excess, [SO42-]xs). Sulfate excess levels found (12.66 Eq l-1) were in agreement with the background hemispheric levels reported by Galloway et al (1982) for remote sites (10 Eq l-1), suggesting that there was a minimal influence of local sources that contributed to this ion; this is in agreement to the residence time for SO2 in the atmosphere. However, (SO42- )xs/NO3- was lower than other coastal sites (0.031), indicating a significant enrichment of nitrates. Clearly, there was a local source influencing on nitrate levels, probably forest fires occurring as a result of the mid-summer drought, and emissions coming from the power

Ionic Abundance: Na+> Ca2+> Cl->SO42-> NO3-

Table 3. Mean values of pH and concentration (Eq l-1) of dry deposition samples.

5.9 2.00 5.79 70.51 16.15 132.75 60.49 14.61 90.63

5.77 153.6 4.30 109.50 26.20 20.50 103.50 83.30 56.40

> Ca2+> Cl-> SO42-> Mg2+> K+>Na+

> Mg2+>NH4

content present in samples was in excess,

+> K+

Petenes Biosphere Reserve and Tula de Allende are shown in Table 3.

Ionic Abundance: NH4

**3.5 Statistical analysis** 

(SAS Inc, 1998).

Los Petenes

Tula de Allende

differences between each element.

**4. Results and discussion 4.1 Chemical composition** 

**4.1.1 Los Petenes Biosphere Reserve** 

nitrate levels, it can be assumed that total NO3-

plant located at 10 km from the sampling site. Considering the residence times for SO2 (13 days) and NO2 (1 day), it can be assumed that Los Petenes site was not subjected to the input of long-range transported pollutants and probably only had the influence of local and temporal sources.

Fig. 9. pH frequency distribution for dry deposition samples collected at Los Petenes Biosphere Reserve.

pH values found in dry deposition samples showed a great variability, 22% of the samples were below of 5.6 (with a mean value of 5.9), and a negligible percentage (5.5%) was in the acid range (pH<5.0) (Table 3, Figure 9). In spite of that nitrate levels exceeded background hemispheric levels, pH values suggest that this site does not have acidity problems, owing to some cations probably played an important role in the neutralization process. This behavior is in agreement with the calcareous soils commonly found in Yucatan Peninsula which is constituted by alkaline particles participating in active way in the neutralization process.

#### **4.1.2 Tula de Allende**

Sulfate was the most abundant ion in Tula de Allende, probably attributed to anthropogenic sources. Tula de Allende is not a coastal site; therefore, it can be assumed that the total content of sulfate and nitrate is in excess (anthropogenic origin). Sulfate excess, [SO42-]xs) levels found (68.78 Eq l-1) exceeded almost seven times the background hemispheric levels reported by Galloway et al (1982) for remote sites (10 Eq l-1). In addition, nitrate levels (56.37 Eq l-1) also exceeded the background hemispheric levels reported for remote sites (2.8 Eq l-1). Both, nitrate and sulfate contributed to the acidity. Considering the residence times for SO2 (13 days) and NO2 (1 day), it can be assumed that this site had the influence of local sources (anthropogenic sources such as a local power plant and oil refinery) as well as the input of pollutants subjected to long-range transport process. From Table 3 and Figure 10, it can be observed that pH values found in dry deposition samples ranged from 5.3 to 6.7, with a mean value of 5.77. A great percentage of the samples (59%) was below of 5.6, but any sample was in the acid range (pH<5.0).

Evaluation of Dry Atmospheric Deposition in Two

sources (power plant and biomass burning).

NH4 (µEq l-1)

SO4 (µEq l

Na+ 1

**4.2.2 Tula de Allende** 

NO3-

K+ 0.41 1

Ca2+ 0.52 0.41 1

Mg2+ 0.02 0.09 0.65 1

NH4+ 0.26 0.21 0.26 0.14 1

at p<0.001) and Ca2+ (r=0.74 at p<0.001); whereas SO4


)

M01

M02

Fig. 11. Time-trends of ionic concentrations in dry deposition in Los Petenes.

0.51 0.27 0.89 0.54 0.18 1

M03

M04

Na+ K+ Ca2+ Mg2+ NH4+ NO3- Cl- SO42-

Cl- 0.78 0.67 0.77 0.77 0.33 0.71 1 SO42- 0.51 0.37 0.55 0.04 0.26 0.63 0.21 1 Table 4. Pearson´s correlation matrix among trace elements for Los Petenes Biosphere Reserve.

Pearson matrix (Table 5) indicates that nitrate ion correlated significantly with SO42- (r=0.95

M05

M06

M07

M01

M02

M03

M04

SO4 NO3

M05

M06

M07

NO3(µEq l-1)

NO3(µEq l-1)

2- correlated with Ca2+ (r=0.79 at

Sites in the Vicinity of Fuel Oil-Fired Power Plants in Mexico 179

had a common origin, and being evident an enrichment of nitrates derived from local

NH4 NO3

Fig. 10. pH frequency distribution for dry deposition samples collected at Tula de Allende.

In despite of nitrate levels exceeded the background hemispheric levels, it can be concluded that this site has enough buffer capacity to neutralize the acidity, probably due to Ca2+ (attributed to Portland cement plants located near this sampling site) playing an important role in the neutralization process.

#### **4.2 Correlation analysis and time trends**

The origin of elements in atmospheric deposition samples can be inferred by correlations between elements and let us to identify common sources, similar removal processes or a strong acid-base relationship.

#### **4.2.1 Los Petenes Biosphere Reserve**

From Pearson matrix (Table 4), it can be observed that nitrate ion showed a good correlation with Ca2+ (r=0.89 at p<0.001) and Mg2+ (r=0.54 at p<0.001), being evident a strong acid-base relation. In addition, Ca2+ was the third more abundant ion, indicating that this alkaline element, derived from crustal, was the main responsible of the neutralization process, resulting in pH values above of 5.6. NO3- - SO42- correlated significantly (r=0.63 at p<0.001), suggesting that both ions had a common source.

Figure 11 shows the variations of ionic concentrations in dry deposition with time at Los Petenes Biosphere Reserve. Backward air masses trajectories analysis carried out using the HYSPLIT Model from NOAA demonstrates that winds during the sampling period had an East component. Therefore, it can be inferred that a specific source located in this direction (the power plant near to this site) possibly contributed to the significant levels of nitrate and sulfate.

Some temporal emissions as forest fires occurring as a result of the mid-summer drought could also contribute to these high levels of nitrate and sulfate. In Yucatan Peninsula, the mid-summer drought is associated with warm temperatures and the lack of humidity in soils favoring the conditions for forest fires. It can be observed that NO3- and NH4+ showed the same behavior, it suggests that both ions had a common source (forest fires). In addition, it can be observed that SO42- and NO3- showed the same pattern, suggesting that both ions

5.0-5.5 5.5-6.0 6.0-6.5 >6.5

**pH**

Fig. 10. pH frequency distribution for dry deposition samples collected at Tula de Allende.

In despite of nitrate levels exceeded the background hemispheric levels, it can be concluded that this site has enough buffer capacity to neutralize the acidity, probably due to Ca2+ (attributed to Portland cement plants located near this sampling site) playing an important

The origin of elements in atmospheric deposition samples can be inferred by correlations between elements and let us to identify common sources, similar removal processes or a

From Pearson matrix (Table 4), it can be observed that nitrate ion showed a good correlation with Ca2+ (r=0.89 at p<0.001) and Mg2+ (r=0.54 at p<0.001), being evident a strong acid-base relation. In addition, Ca2+ was the third more abundant ion, indicating that this alkaline element, derived from crustal, was the main responsible of the neutralization process, resulting in pH values above of 5.6. NO3- - SO42- correlated significantly (r=0.63 at p<0.001),

Figure 11 shows the variations of ionic concentrations in dry deposition with time at Los Petenes Biosphere Reserve. Backward air masses trajectories analysis carried out using the HYSPLIT Model from NOAA demonstrates that winds during the sampling period had an East component. Therefore, it can be inferred that a specific source located in this direction (the power plant near to this site) possibly contributed to the significant levels of nitrate and sulfate. Some temporal emissions as forest fires occurring as a result of the mid-summer drought could also contribute to these high levels of nitrate and sulfate. In Yucatan Peninsula, the mid-summer drought is associated with warm temperatures and the lack of humidity in

the same behavior, it suggests that both ions had a common source (forest fires). In addition, it can be observed that SO42- and NO3- showed the same pattern, suggesting that both ions

and NH4+ showed

soils favoring the conditions for forest fires. It can be observed that NO3-

**Relative Frequency**

role in the neutralization process.

strong acid-base relationship.

**4.2 Correlation analysis and time trends** 

**4.2.1 Los Petenes Biosphere Reserve** 

suggesting that both ions had a common source.

had a common origin, and being evident an enrichment of nitrates derived from local sources (power plant and biomass burning).

Fig. 11. Time-trends of ionic concentrations in dry deposition in Los Petenes.


Table 4. Pearson´s correlation matrix among trace elements for Los Petenes Biosphere Reserve.

#### **4.2.2 Tula de Allende**

Pearson matrix (Table 5) indicates that nitrate ion correlated significantly with SO42- (r=0.95 at p<0.001) and Ca2+ (r=0.74 at p<0.001); whereas SO42- correlated with Ca2+ (r=0.79 at

Evaluation of Dry Atmospheric Deposition in Two

Tula de Allende.

common source.

Petenes.

**4.3 Annual deposition fluxes** 

**4.3.1 Los Petenes Biosphere Reserve** 

**Dry deposition fluxes (Kg ha -1 yr -1)**

**4.3.2 Tula de Allende** 

Sites in the Vicinity of Fuel Oil-Fired Power Plants in Mexico 181

Backward air masses trajectories analysis demonstrated that winds during the sampling period had a N-NE component. These results imply that a specific source located in this direction (the power plant near to this site) contributed to the significant levels of nitrate and sulfate. Figure 12 shows the variations of ionic concentrations in dry deposition with time at

Tula de Allende sampling site constitutes a complex mosaic of rural, agricultural and industrialized areas. At the edges of the City, agriculture activities prevail, making evident the influence of ammonium on neutralization process. Furthermore, it can be observed that SO42- and NO3- showed the same behavior, suggesting that both ions possibly had a

The average dry deposition fluxes (in Kg ha-1 yr-1) ranged between 0.39 and 7.19 for Los

Figure 13 illustrates that dry deposition fluxes for ammonium and nitrate were the highest. During dry periods, the reduced mixing layer concentrates only ionic species of local origin; it supports the fact of high levels of nitrate and ammonium had a local origin, in this case, forest fires occurring during the mid-summer drought and emissions from "Lerma" Power Plant.

Ca2+ K+ Mg2+ Na+ NH4+ NO3- Cl- SO42-

The average dry deposition fluxes (in Kg ha-1 yr-1) ranged between 0.34 and 8.04 for Tula de Allende. It can be observed, from Figure 14, that dry deposition fluxes for calcium, sulfate and nitrate were the highest. In this case, anthropogenic emissions from Tula-Vito-Apaxco complex contributed to high levels of calcium, sulfate and nitrate, having in mind that, during

dry periods, the reduced mixing layer only concentrates ionic species of local origin.

Fig. 13. Dry deposition fluxes for Los Petenes Biosphere Reserve.


p<0.001). It was completely evident a strong relationship acid-base between these ion pairs, suggesting that these ions were deposited as calcium nitrate and sulfate. These results support the fact that the co-existence of cement plant and power plant in the same location resulted in pH values near to neutrality value reported for natural rainwater.

Table 5. Pearson´s correlation matrix among trace elements for Tula de Allende.

Fig. 12. Time-trends of ionic concentrations in dry deposition in Tula de Allende.

Backward air masses trajectories analysis demonstrated that winds during the sampling period had a N-NE component. These results imply that a specific source located in this direction (the power plant near to this site) contributed to the significant levels of nitrate and sulfate. Figure 12 shows the variations of ionic concentrations in dry deposition with time at Tula de Allende.

Tula de Allende sampling site constitutes a complex mosaic of rural, agricultural and industrialized areas. At the edges of the City, agriculture activities prevail, making evident the influence of ammonium on neutralization process. Furthermore, it can be observed that SO4 2- and NO3 - showed the same behavior, suggesting that both ions possibly had a common source.

#### **4.3 Annual deposition fluxes**

180 Air Quality – Monitoring and Modeling

p<0.001). It was completely evident a strong relationship acid-base between these ion pairs, suggesting that these ions were deposited as calcium nitrate and sulfate. These results support the fact that the co-existence of cement plant and power plant in the same location

SO4

2-

resulted in pH values near to neutrality value reported for natural rainwater.

0.14 0.44 0.74 0.51 0.10 1

Table 5. Pearson´s correlation matrix among trace elements for Tula de Allende.

Na+ 1

NO3-

SO4

K+ 0.29 1

Ca2+ 0.10 0.74 1

Mg2+ 0.20 0.77 0.91 1

0

30

M01

M03

M05

M07

M09

M11

Fig. 12. Time-trends of ionic concentrations in dry deposition in Tula de Allende.

M13

M15

M17

M19

M21

M23

M25

M27

M29

M31

NO3 - (µEq l-1)

55

80

105

130

155

M01

M03

M05

M07

M09

M11

M13

M15

M17

M19

M21

SO4 NO3

M23

M25

M27

M29

M31

NO3 - (µEq l-1)

25

50

NH4 + (µEq l-1)

SO 4

2-

(µEq l


)

75

100

NH4+ 0.38 0.06 0.18 0.11 1

Na+ K+ Ca2+ Mg2+ NH4+ NO3- Cl-

Cl- 0.59 0.27 0.10 0.20 0.41 0.13 1

2- 0.13 0.47 0.79 0.60 0.13 0.95 0.13 1

NH4 NO3

#### **4.3.1 Los Petenes Biosphere Reserve**

The average dry deposition fluxes (in Kg ha-1 yr-1) ranged between 0.39 and 7.19 for Los Petenes.

Figure 13 illustrates that dry deposition fluxes for ammonium and nitrate were the highest. During dry periods, the reduced mixing layer concentrates only ionic species of local origin; it supports the fact of high levels of nitrate and ammonium had a local origin, in this case, forest fires occurring during the mid-summer drought and emissions from "Lerma" Power Plant.

Fig. 13. Dry deposition fluxes for Los Petenes Biosphere Reserve.

#### **4.3.2 Tula de Allende**

The average dry deposition fluxes (in Kg ha-1 yr-1) ranged between 0.34 and 8.04 for Tula de Allende. It can be observed, from Figure 14, that dry deposition fluxes for calcium, sulfate and nitrate were the highest. In this case, anthropogenic emissions from Tula-Vito-Apaxco complex contributed to high levels of calcium, sulfate and nitrate, having in mind that, during dry periods, the reduced mixing layer only concentrates ionic species of local origin.

Evaluation of Dry Atmospheric Deposition in Two

located in the vicinity of this place.

ISSN 1352-2310.

1470, ISSN 1352-2310.

pp. 241-258, ISSN 0049-6979.

August 1985, pp. 721-725, ISSN 0013-936X.

**6. Acknowledgement** 

this work.

**7. References** 

**5. Conclusion** 

Sites in the Vicinity of Fuel Oil-Fired Power Plants in Mexico 183

It can be concluded that in Los Petenes site, alkaline particles from the calcareous soil played an important role in the acidity neutralization process derived from high emissions of NO3 from the power plant and due to biomass burning in the vicinity of this sampling site. N input in Los Petenes exceeded slightly the critical values reported for other sites of the world. Considering the ecological importance of this site, local strategic policies must be

On the other hand, in spite of high sulfate and nitrate levels found in Tula de Allende (attributed to a power plant), acidity neutralization process was improved by the presence of cement plants near to this sampling site. However, S input exceeded critical values reported for several sites around the world. In spite of Tula de Allende is not an eco-region, since is already an industrialized site, S deposition could be a threat to historic wealth

This work was financially co-supported by Sectorial Fund CONACYT-SEMARNAT (project 107948) and by National Institute of Ecology (projects INE/A1-047/2008; INE/A1-007/2009; INE/A1-031/2010). Authors gratefully acknowledge their help to students that supported

Alonso, R., Bytnerowicz, A., & Boarman, W.I. (2005). Atmospheric dry deposition in the

Balestrini, R., Galli, L., & Tartari, G. (2000). Wet and dry atmospheric deposition at prealpine

Dasch, M.J. (1985). Direct measurement of dry deposition to a polyethylene bucket and

Davies Inc (n.d.). Davies Instruments. Vantage Pro2. Weather Link. Data Logger & Software. Flues,M., Hama, P., Lemes, M.J.L., Dantas, E.S.K., & Fornaro, A. (2002). Evaluation of the

Galbally, I.E. & Gillel, R.W. (1988). Acidification in Tropical Countries. Chapter 3. In:

Herrera. Pp. 1-405, John Wiley & Sons, ISBN 047 91870 9, Great Britain. Galloway, J.N. & Likens, G.E. (1976). Calibration of collection procedures for the

vicinity of the Salton Sea, California – I: Air pollution and deposition in a desert environment. *Atmospheric Environment,*Vol39, No. 26, August 2005, pp. 4671-4679,

and alpine sites in northern Italy. *Atmospheric Environment*, Vol. 34, No. 9, pp. 1455-

various surrogate surfaces. *Environmental, Science and Technology,* Vol. 19, No. 8,

rainwater acidity of a rural region due to a coal-fired power plant in Brazil. *Atmospheric Environment*, Vol. 36, No. 14, May 2002, pp. 2397-2404, ISSN 1352-2310.

*Processes regulating nitrogen compounds in the tropical atmosphere.* H. Rodhe & R.

determination of precipitation chemistry. *Water, Air & Soil Pollution*, Vol. 6, No. 2-4,

applied focused on protecting the wildfire life in this Biosphere Reserve.

Fig. 14. Dry deposition fluxes for Tula de Allende.

#### **4.4 Deposition fluxes and critical loads**

A critical load of 10 Kg N ha-1 yr-1 has been proposed in the United Kingdom (Jones, 2005), however, given that the sensitivity varies from one specie to other, Power et al (2006) proposed a value of 8 Kg N ha-1 yr-1. In addition, it has been proposed a value of 5 Kg N ha-1 yr -1 for sensitive ecosystems. For the impacted ecosystems in U.S, the critical load reported ranges between 3 and 8 Kg N ha-1 yr-1. In the case of S, for natural forests, some authors propose values between 2 and 5 Kg S ha-1 yr-1 (Nilsson & Grennfelt, 1988).

Data on dry deposition inputs of N and S in Mexican sites are scarce, taking in account the critical values reported for Europe and United States of America, we can observe that in Los Petenes Biosphere Reserve, N input (11.99 Kg ha-1 yr-1) is slightly above the critical loads, whereas S input (1.4 Kg ha-1 yr-1) did not exceeded the critical values, this supports the proposal that this site is only subject to local emissions. These results indicate that critical loads of nitrogen can increase in a short time period if strategic measures are not taken. Our results are sub estimated considering that we are not reporting data for wet deposition, since this site is a natural reserve, it can constitute a serious threat to flora and fauna species that live in this ecosystem.

On the other hand, N input (7.6 Kg ha-1 yr-1) for Tula de Allende, exceeds proposed value for sensitive ecosystems, but it is below of critical load reported for American and European sites. However, in the case of S, we can observe that S input (8.04 Kg ha-1 yr-1) exceeded the critical value for various eco-regions in the world. It is clear that emissions released by sources in Tula-Vita-Apaxco industrial complex contributed in a significant way to this input. Alkaline particles emitted from cement plants contributed to neutralize the acidity, therefore, pH values resulting from this neutralization process were not in the acid range. However, pH itself it is not a conservative tracer of ecological effects of acid deposition; for that reason, we propose to improve the current environmental policies and the promulgation of new regulations based on a conservative tracer and focused to decrease sulfate levels in this region.

### **5. Conclusion**

182 Air Quality – Monitoring and Modeling

Ca2+ K+ Mg2+ Na+ NH4+ NO3- Cl- SO42-

A critical load of 10 Kg N ha-1 yr-1 has been proposed in the United Kingdom (Jones, 2005), however, given that the sensitivity varies from one specie to other, Power et al (2006) proposed a value of 8 Kg N ha-1 yr-1. In addition, it has been proposed a value of 5 Kg N ha-1 yr -1 for sensitive ecosystems. For the impacted ecosystems in U.S, the critical load reported ranges between 3 and 8 Kg N ha-1 yr-1. In the case of S, for natural forests, some authors

Data on dry deposition inputs of N and S in Mexican sites are scarce, taking in account the critical values reported for Europe and United States of America, we can observe that in Los Petenes Biosphere Reserve, N input (11.99 Kg ha-1 yr-1) is slightly above the critical loads, whereas S input (1.4 Kg ha-1 yr-1) did not exceeded the critical values, this supports the proposal that this site is only subject to local emissions. These results indicate that critical loads of nitrogen can increase in a short time period if strategic measures are not taken. Our results are sub estimated considering that we are not reporting data for wet deposition, since this site is a natural reserve, it can constitute a serious threat to flora and fauna species

On the other hand, N input (7.6 Kg ha-1 yr-1) for Tula de Allende, exceeds proposed value for sensitive ecosystems, but it is below of critical load reported for American and European sites. However, in the case of S, we can observe that S input (8.04 Kg ha-1 yr-1) exceeded the critical value for various eco-regions in the world. It is clear that emissions released by sources in Tula-Vita-Apaxco industrial complex contributed in a significant way to this input. Alkaline particles emitted from cement plants contributed to neutralize the acidity, therefore, pH values resulting from this neutralization process were not in the acid range. However, pH itself it is not a conservative tracer of ecological effects of acid deposition; for that reason, we propose to improve the current environmental policies and the promulgation of new regulations based on a conservative tracer and focused to decrease

propose values between 2 and 5 Kg S ha-1 yr-1 (Nilsson & Grennfelt, 1988).

**4.4 Deposition fluxes and critical loads** 

that live in this ecosystem.

sulfate levels in this region.

Fig. 14. Dry deposition fluxes for Tula de Allende.

**Dry deposition fluxes (Kg ha -1 yr -1)**

It can be concluded that in Los Petenes site, alkaline particles from the calcareous soil played an important role in the acidity neutralization process derived from high emissions of NO3 from the power plant and due to biomass burning in the vicinity of this sampling site. N input in Los Petenes exceeded slightly the critical values reported for other sites of the world. Considering the ecological importance of this site, local strategic policies must be applied focused on protecting the wildfire life in this Biosphere Reserve.

On the other hand, in spite of high sulfate and nitrate levels found in Tula de Allende (attributed to a power plant), acidity neutralization process was improved by the presence of cement plants near to this sampling site. However, S input exceeded critical values reported for several sites around the world. In spite of Tula de Allende is not an eco-region, since is already an industrialized site, S deposition could be a threat to historic wealth located in the vicinity of this place.

#### **6. Acknowledgement**

This work was financially co-supported by Sectorial Fund CONACYT-SEMARNAT (project 107948) and by National Institute of Ecology (projects INE/A1-047/2008; INE/A1-007/2009; INE/A1-031/2010). Authors gratefully acknowledge their help to students that supported this work.

#### **7. References**


**0**

**9**

A. Omrane\*

*French Guiana*

(1.1)

**Pollution Problems**

*Route de Baduel BP 792, 97337 Cayenne Cedex*

**Some Aspects of the Sentinel Method for**

*Laboratoire CEREGMIA, EA 2440, I.E.S Guyane, Campus Universitaire Trou-Biran,*

Modelling environmental problems leads to mathematical systems with missing data: Weather problems have generally missing initial conditions. This chapter is concerned with identifying pollution terms arising in the state equation of some dissipative system with

To this aim the so-called sentinel method is used. We explain how the problemn of determining a sentinel is equivalent to a null-controllability problem for which Carleman

In a second part of the chapter, we use the same techniques to discuss of how to get instantaneous information (at fixed *t* = *T* ∈ [0, +∞[) on pollution terms in distributed systems

We consiter a fixed final time *T* > 0, and Ω an open subset of **R***<sup>d</sup>* of smooth boundary *∂*Ω, and we denote by *Q* = Ω×]0, *T*[ the space-time cylinder. We are intersted with systems partially

> *y*(0) = *y*<sup>0</sup> + *τy*ˆ<sup>0</sup> in Ω, *y* = 0 on Σ1,

*∂ν* <sup>=</sup> 0 on <sup>Σ</sup>2,

where *y* = *y*(*x*, *t*; *λ*, *τ*), and where Σ<sup>1</sup> is a piece of the boundary Σ = *∂*Ω×]0, *T*[ and Σ<sup>2</sup> = <sup>Σ</sup>\Σ1. We assume here that *<sup>f</sup>* : **<sup>R</sup>** <sup>→</sup> **<sup>R</sup>** is of class *<sup>C</sup>*1, the functions *<sup>ξ</sup>* and *<sup>y</sup>*<sup>0</sup> are known with

\*This projet is supported by the CPER AGROECOTROP 2007-2013, Région Guadeloupe-État-Europe

*ξ* in *Q*,

*ξ* (so-called pollution term) and *τy*ˆ<sup>0</sup> (so-called

*ξ* and *y*ˆ<sup>0</sup> are renormalized and represent the size of pollution

<sup>0</sup>�*L*<sup>2</sup>(Ω) <sup>≤</sup> 1, so that the reals *<sup>λ</sup>*, *<sup>τ</sup>* are small enough.

*<sup>y</sup>*� <sup>−</sup> <sup>Δ</sup>*<sup>y</sup>* <sup>+</sup> *<sup>f</sup>*(*y*) = *<sup>ξ</sup>* <sup>+</sup> *<sup>λ</sup>* <sup>ˆ</sup>

*∂y*

of incomplete data in some ecology and/or meteorology problems.

⎧ ⎪⎪⎪⎨

⎪⎪⎪⎩

known; we consider here the state equation :

*<sup>ξ</sup>* <sup>∈</sup> *<sup>L</sup>*2(*Q*) and *<sup>y</sup>*<sup>0</sup> <sup>∈</sup> *<sup>L</sup>*2(Ω). But, the terms : *<sup>λ</sup>* <sup>ˆ</sup>

perturbation term) are unknown, ˆ

*ξ*�*L*<sup>2</sup>(*Q*) ≤ 1, �*y*ˆ

and perturbation

� ˆ

**1. Introduction**

incomplete initial condition.

inequalities are revisited.


### **Some Aspects of the Sentinel Method for Pollution Problems**

A. Omrane\*

*Laboratoire CEREGMIA, EA 2440, I.E.S Guyane, Campus Universitaire Trou-Biran, Route de Baduel BP 792, 97337 Cayenne Cedex French Guiana*

#### **1. Introduction**

184 Air Quality – Monitoring and Modeling

Galloway, J.N., Thornton, J.D., Norton, S.A., Volchok, H.L., & McLean, R.A.N. (1982)Trace

Hegg, D.A., Radke, L.F., Hobbs, P.V., & Riggon, P.J. (1988). Ammonia emissions from biomass burning. *Geophys. Res. Letter,* Vol. 15, No. 4, pp. 335-337, ISSN 0094-8276. Jones, M.L.M. (2005). Nitrogen deposition in upland grasslands: Critical loads, management

Lakes Environmental Inc. (2011). Lakes Environmental Software. WRPLOT View v.6.5.2.

Nilsson, J. & Grennfelt, P. (1988). Critical loads for sulphur and nitrogen. Report from a

Ministers. Report 1988: 15, Copenhagen, Denmark. ISBN 87-7303-248-4. NOAA (n.d.). National Oceanic and Atmospheric Administration. U.S. Department of

Model, In: *Air Resources Laboratory http://ready.arl.noaa.gov/HYSPLIT.php.*  Power, S.A., Green, E.R., Barker, C.G., Bell, J.N.B., & Ashmore, M.R. (2006). Ecosystem

SAS Inc. (1998). SAS/STAT, Guide for personal computers. Version 8. United States of

Tsitouridou, R. & Anatolaki, Ch. (2007). On the wet and dry deposition of ionic species in

Zuk, M., Garibay, V., Iniestra, R., López, M.T., Rojas, L., & Laguna, I. (2006). *Introducción a la* 

*Biology,* Vol. 12, No. 7, July 2006, pp. 1241-1252, ISSN 1354-1013.

workshop held at Skokloster, Sweden. March 19-24. The Nordic Council of

Commerce. HYSPLIT –Hybrid Single Particle Lagrangian Integrated Trajectory

recovery: heathland response to a reduction in nitrogen deposition. *Global Change* 

the vicinity of coal-fired power plants, northwestern Greece. *Atmospheric Research*,

*evaluación de los impactos de las termoeléctricas de México (Primera Edición Octubre 2006).* Secretaría del Medio Ambiente y Recursos Naturales, Instituto Nacional de

*Environment*, Vol. 16, No. 7, pp. 1677-1700, ISSN 1352-2310.

and recovery. Thesis. University of Sheffield. U.K.

Freeware Wind Rose Plots for Meteorological Data.

Ecología, ISBN 968-817-804-7, México D.F, México.

America.

Vol 83, pp. 93-105, ISSN 0169-8095.

metals in atmospheric deposition: a review and assessment. *Atmospheric* 

Modelling environmental problems leads to mathematical systems with missing data: Weather problems have generally missing initial conditions. This chapter is concerned with identifying pollution terms arising in the state equation of some dissipative system with incomplete initial condition.

To this aim the so-called sentinel method is used. We explain how the problemn of determining a sentinel is equivalent to a null-controllability problem for which Carleman inequalities are revisited.

In a second part of the chapter, we use the same techniques to discuss of how to get instantaneous information (at fixed *t* = *T* ∈ [0, +∞[) on pollution terms in distributed systems of incomplete data in some ecology and/or meteorology problems.

We consiter a fixed final time *T* > 0, and Ω an open subset of **R***<sup>d</sup>* of smooth boundary *∂*Ω, and we denote by *Q* = Ω×]0, *T*[ the space-time cylinder. We are intersted with systems partially known; we consider here the state equation :

$$\begin{cases} y' - \Delta y + f(y) = \xi + \lambda \hat{\xi} & \text{in} \quad Q\_{\prime} \\ y(0) &= y^0 + \tau y^0 \text{ in} \quad \Omega\_{\prime} \\ y &= 0 & \text{on} \quad \Sigma\_{1\prime} \\ \frac{\partial y}{\partial \nu} &= 0 & \text{on} \quad \Sigma\_{2\prime} \end{cases} \tag{1.1}$$

where *y* = *y*(*x*, *t*; *λ*, *τ*), and where Σ<sup>1</sup> is a piece of the boundary Σ = *∂*Ω×]0, *T*[ and Σ<sup>2</sup> = <sup>Σ</sup>\Σ1. We assume here that *<sup>f</sup>* : **<sup>R</sup>** <sup>→</sup> **<sup>R</sup>** is of class *<sup>C</sup>*1, the functions *<sup>ξ</sup>* and *<sup>y</sup>*<sup>0</sup> are known with *<sup>ξ</sup>* <sup>∈</sup> *<sup>L</sup>*2(*Q*) and *<sup>y</sup>*<sup>0</sup> <sup>∈</sup> *<sup>L</sup>*2(Ω). But, the terms : *<sup>λ</sup>* <sup>ˆ</sup> *ξ* (so-called pollution term) and *τy*ˆ<sup>0</sup> (so-called perturbation term) are unknown, ˆ *ξ* and *y*ˆ<sup>0</sup> are renormalized and represent the size of pollution and perturbation

� ˆ *ξ*�*L*<sup>2</sup>(*Q*) ≤ 1, �*y*ˆ <sup>0</sup>�*L*<sup>2</sup>(Ω) <sup>≤</sup> 1, so that the reals *<sup>λ</sup>*, *<sup>τ</sup>* are small enough.

<sup>\*</sup>This projet is supported by the CPER AGROECOTROP 2007-2013, Région Guadeloupe-État-Europe

**1.1 The informations given by the sentinel**

In (1.4), *S*(*λ*, *τ*) is observed and using (1.2),

*<sup>λ</sup> <sup>∂</sup><sup>S</sup>*

*∂λ* (0, 0) �

*∂S*

⎧ ⎪⎪⎪⎨ *y*�

⎪⎪⎪⎩

� *Q* �

function *q* = *q*(*x*, *t*) as the solution of the backward problem :

The insensibility condition (1.5) holds if and only if

The derivative *<sup>y</sup><sup>λ</sup>* = (*∂y*/*∂λ*)(0, 0) only depends on <sup>ˆ</sup>

the estimates (1.6) contains the informations on *λ* ˆ

transforming the insensibility condition (1.5).

**2. Null-controllability problem**

Then the function *y<sup>τ</sup>* is solution of

*<sup>S</sup>*(*λ*, *<sup>τ</sup>*) � *<sup>S</sup>*(0, 0) + *<sup>λ</sup> <sup>∂</sup><sup>S</sup>*

*<sup>S</sup>*(*λ*, *<sup>τ</sup>*) = �

*∂λ* (0, 0) = �

� *Q* *Q*

Some Aspects of the Sentinel Method for Pollution Problems 187

*Q*

The existence of a sentinel is equivalent to a null-controllability property. Indeed, we begin by

*<sup>d</sup><sup>τ</sup> <sup>y</sup>*(*λ*, *<sup>τ</sup>*)

where *y*<sup>0</sup> = *y*(0, 0). Problem (2.1) is linear and has a unique solution *y<sup>τ</sup>* under mild

�

We can transform (2.2) by introducing the classical adjoint state. More precisely, we define the

� � � *λ*=*τ*=0 .

(*y*0)*y<sup>τ</sup>* = 0 in *Q*, *yτ*(0) = *y*ˆ<sup>0</sup> in Ω, *y<sup>τ</sup>* = 0 on Σ1,

*∂ν* <sup>=</sup> 0 on <sup>Σ</sup>2,

where here *χ<sup>O</sup>* and *χω* denote the characteristic functions of *O* and *ω* respectively.

*<sup>y</sup><sup>τ</sup>* <sup>=</sup> *<sup>d</sup>*

*<sup>τ</sup>* − Δ*y<sup>τ</sup>* + *f* �

*∂y<sup>τ</sup>*

*h*0*χ<sup>O</sup>* + w*χω*

*∂λ* (0, 0), for *<sup>λ</sup>*, *<sup>τ</sup>* small.

(*h*0*χ<sup>O</sup>* + w*χω*)(*mo* − *y*0) *dxdt*, (1.6)

*ξ* (see for details remark 1 below).

*ξ* and other known data. Consequently,

*y<sup>τ</sup> dxdt* = 0. (2.2)

(2.1)

(*h*0*χ<sup>O</sup>* + w*χω*)*mo dxdt*

(*h*0*χ<sup>O</sup>* + w*χω*)*y<sup>λ</sup> dxdt*,

Because of (1.5) we can write

so that (1.5) becomes

with

Denote by

assumptions on *f* .

Some non empty open subset *O* ⊂ Ω, is called observatory set. The observation is *y* on *O*, for the time *T*. We denote by *yobs* this observation

$$y\_{obs} = m\_0 \in L^2(\mathbb{O} \times (0, T)). \tag{1.2}$$

We suppose that (1.1) has a unique solution denoted by *y*(*λ*, *τ*) := *y*(*x*, *t*; *λ*, *τ*) in some relevant space. The question is

> 

how to calculate the pollution term *λ* ˆ *ξ* in the state equation, independently from the variation *τy*ˆ<sup>0</sup> around the initial data ? (1.3)

**Least squares.** Question (1.3) is natural and leds to some developments; some answer is given by the least squares method. The method consists in considering the unknowns {*<sup>λ</sup>* <sup>ˆ</sup> *<sup>ξ</sup>*, *<sup>τ</sup>y*ˆ0} <sup>=</sup> {*v*, *w*} as control variables, then the state *y*(*x*, *t*; *v*, *w*) has to be driven as close as possible to *mo*.

This comes to some optimal control problem. By this way we look for the pair (*v*, *w*), there is then no real possibility to find *v* or *w* independently.

**Sentinels.** The sentinel method of Lions Lions J.-L. (1992) is a particular least squares method which is adapted to the identification of parameters in ecosystems with incomplete data; many models can be found in litterature. The sentinel concept relies on the following three objects : some state equation (for instance (1.1)), some observation function (1.2), and some control function w to be determined.

Many papers use the definition of Lions in the theoretical aspect (see for example Bodart Bodart O. (1994), Bodart-Fabre Bodart O. & Fabre C. (1993)Bodart O. & Fabre C. (1995)), as well as in the numerical one (see Bodart-Demeestere Bodart O. & Demeestere P. (1997), Demeestere Demeestere P. (1997) and Kernevez Kernevez J.-P. (1997)).

We use the techniques in Miloudi Y. et al. (2007) to give an answer to the question (1.3). Let *<sup>h</sup>*<sup>0</sup> be some function in *<sup>L</sup>*2(*<sup>O</sup>* <sup>×</sup> (0, *<sup>T</sup>*)). Let on the other hand *<sup>ω</sup>* be some open and non empty subset of <sup>Ω</sup>. For a control function <sup>w</sup> <sup>∈</sup> *<sup>L</sup>*2(*<sup>ω</sup>* <sup>×</sup> (0, *<sup>T</sup>*)), we define the functional

$$S(\lambda, \tau) = \int\_0^T \int\_{\mathcal{O}} h\_0 y(\lambda, \tau) \, d\mathbf{x} dt + \int\_0^T \int\_{\omega} \mathbf{w} y(\lambda, \tau) \, d\mathbf{x} dt. \tag{1.4}$$

We say that *S* defines a sentinel for the problem (1.1) if there exists w such that *S* is insensitive (at first order) with respect the to missing terms *τy*ˆ0, which means

$$\frac{\partial \mathcal{S}}{\partial \tau}(0,0) = 0\tag{1.5}$$

for any *y*ˆ<sup>0</sup> where here (0, 0) corresponds to *λ* = *τ* = 0, and if w minimizes the norm �v�*L*<sup>2</sup>(*ω*×(0,*T*)).

**Important Remark.** The Lions sentinels *S* assume *ω* = *O*. In this case, the observation and the control share the same support, and the solution w = −*h*<sup>0</sup> is trivial.

The definition (1.4) extends the one by Lions to the case where the observation and the control have different supports. This point of view (where *ω* �= *O*) has been considered for the first time in Nakoulima O. (2004) and Miloudi Y. et al. (2007).

#### **1.1 The informations given by the sentinel**

Because of (1.5) we can write

$$\mathcal{S}(\lambda, \tau) \simeq \mathcal{S}(0, 0) + \lambda \frac{\partial \mathcal{S}}{\partial \lambda}(0, 0), \qquad \text{for} \quad \lambda, \tau \quad \text{small}.$$

In (1.4), *S*(*λ*, *τ*) is observed and using (1.2),

$$S(\lambda, \tau) = \int\_{Q} (h\_0 \chi\_O + \mathbf{w} \chi\_{\omega}) m\_o \, d\mathbf{x} dt$$

so that (1.5) becomes

$$
\lambda \frac{\partial \mathcal{S}}{\partial \lambda}(0,0) \simeq \int\_{Q} (h\_0 \chi\_O + \mathfrak{w} \chi\_{\omega})(m\_o - y\_0) \, d\mathbf{x} dt,\tag{1.6}
$$

with

2 Will-be-set-by-IN-TECH

Some non empty open subset *O* ⊂ Ω, is called observatory set. The observation is *y* on *O*, for

We suppose that (1.1) has a unique solution denoted by *y*(*λ*, *τ*) := *y*(*x*, *t*; *λ*, *τ*) in some relevant

the state equation, independently from the variation *τy*ˆ<sup>0</sup> around the initial data ?

**Least squares.** Question (1.3) is natural and leds to some developments; some answer is given

{*v*, *w*} as control variables, then the state *y*(*x*, *t*; *v*, *w*) has to be driven as close as possible to

This comes to some optimal control problem. By this way we look for the pair (*v*, *w*), there is

**Sentinels.** The sentinel method of Lions Lions J.-L. (1992) is a particular least squares method which is adapted to the identification of parameters in ecosystems with incomplete data; many models can be found in litterature. The sentinel concept relies on the following three objects : some state equation (for instance (1.1)), some observation function (1.2), and some control

Many papers use the definition of Lions in the theoretical aspect (see for example Bodart Bodart O. (1994), Bodart-Fabre Bodart O. & Fabre C. (1993)Bodart O. & Fabre C. (1995)), as well as in the numerical one (see Bodart-Demeestere Bodart O. & Demeestere P. (1997), Demeestere

We use the techniques in Miloudi Y. et al. (2007) to give an answer to the question (1.3). Let *<sup>h</sup>*<sup>0</sup> be some function in *<sup>L</sup>*2(*<sup>O</sup>* <sup>×</sup> (0, *<sup>T</sup>*)). Let on the other hand *<sup>ω</sup>* be some open and non empty

We say that *S* defines a sentinel for the problem (1.1) if there exists w such that *S* is insensitive

for any *y*ˆ<sup>0</sup> where here (0, 0) corresponds to *λ* = *τ* = 0, and if w minimizes the norm

**Important Remark.** The Lions sentinels *S* assume *ω* = *O*. In this case, the observation and

The definition (1.4) extends the one by Lions to the case where the observation and the control have different supports. This point of view (where *ω* �= *O*) has been considered for the first

 *T* 0 *ω*

w *y*(*λ*, *τ*) *dxdt*. (1.4)

*∂τ* (0, 0) = <sup>0</sup> (1.5)

subset of <sup>Ω</sup>. For a control function <sup>w</sup> <sup>∈</sup> *<sup>L</sup>*2(*<sup>ω</sup>* <sup>×</sup> (0, *<sup>T</sup>*)), we define the functional

*h*<sup>0</sup> *y*(*λ*, *τ*) *dxdt* +

*∂S*

by the least squares method. The method consists in considering the unknowns {*<sup>λ</sup>* <sup>ˆ</sup>

how to calculate the pollution term *λ* ˆ

*yobs* <sup>=</sup> *mo* <sup>∈</sup> *<sup>L</sup>*2(*<sup>O</sup>* <sup>×</sup> (0, *<sup>T</sup>*)). (1.2)

*ξ* in

(1.3)

*<sup>ξ</sup>*, *<sup>τ</sup>y*ˆ0} <sup>=</sup>

the time *T*. We denote by *yobs* this observation

 

then no real possibility to find *v* or *w* independently.

Demeestere P. (1997) and Kernevez Kernevez J.-P. (1997)).

 *T* 0 *O*

(at first order) with respect the to missing terms *τy*ˆ0, which means

the control share the same support, and the solution w = −*h*<sup>0</sup> is trivial.

time in Nakoulima O. (2004) and Miloudi Y. et al. (2007).

*S*(*λ*, *τ*) =

space. The question is

function w to be determined.

�v�*L*<sup>2</sup>(*ω*×(0,*T*)).

*mo*.

$$\frac{\partial S}{\partial \lambda}(0,0) = \int\_{\mathcal{Q}} (h\_0 \chi\_{\mathcal{O}} + \mathbf{w} \chi\_{\omega}) y\_{\lambda} \, d\mathbf{x} dt\_{\lambda}$$

where here *χ<sup>O</sup>* and *χω* denote the characteristic functions of *O* and *ω* respectively.

The derivative *<sup>y</sup><sup>λ</sup>* = (*∂y*/*∂λ*)(0, 0) only depends on <sup>ˆ</sup> *ξ* and other known data. Consequently, the estimates (1.6) contains the informations on *λ* ˆ *ξ* (see for details remark 1 below).

#### **2. Null-controllability problem**

The existence of a sentinel is equivalent to a null-controllability property. Indeed, we begin by transforming the insensibility condition (1.5).

Denote by

$$y\_{\tau} = \frac{d}{d\tau}y(\lambda, \tau)\Big|\_{\lambda=\tau=0}$$

Then the function *y<sup>τ</sup>* is solution of

$$\begin{cases} y\_\tau' - \Delta y\_\tau + f'(y\_0)y\_\tau = 0 & \text{in } Q\_\prime \\ y\_\tau(0) &= \hat{y}^0 \text{ in } \Omega\_\prime \\ y\_\tau &= 0 \quad \text{on } \Sigma\_{1\prime} \\ \frac{\partial y\_\tau}{\partial \nu} &= 0 \quad \text{on } \Sigma\_{2\prime} \end{cases} \tag{2.1}$$

.

where *y*<sup>0</sup> = *y*(0, 0). Problem (2.1) is linear and has a unique solution *y<sup>τ</sup>* under mild assumptions on *f* .

The insensibility condition (1.5) holds if and only if

$$\int\_{Q} \left( h\_{0} \chi\_{O} + \mathfrak{w} \chi\_{\omega} \right) \mathcal{y}\_{\mathbb{T}} \, d\mathbf{x} dt \, = \,\, \mathbf{0}. \tag{2.2}$$

We can transform (2.2) by introducing the classical adjoint state. More precisely, we define the function *q* = *q*(*x*, *t*) as the solution of the backward problem :

**Theorem 1.** *Let be u* ∈ V*, then there exists a positive constant C* = *C*(Ω, *ω*,*O*, *T*, *f* �

*Q*

*v* �→ �*v*�*<sup>V</sup>* =

*<sup>v</sup>* <sup>∈</sup> *<sup>L</sup>*2(*Q*) *such that :*

Cauchy-Schwartz inequality, we deduce that the linear form defined on *V* by

*v* �→ *Q*

*Q*

*Q* 1 *<sup>θ</sup>*<sup>2</sup> |*v*|

then, *V* is a Hilbert space for the scalar product *a*(*v*, *v*ˆ) and the associated norm.

*<sup>Q</sup>* <sup>|</sup>*L u*<sup>|</sup>

Some Aspects of the Sentinel Method for Pollution Problems 189

According to the RHS of (3.2), we consider the space V endowed with the bilinear form *a*(., .)

*Lu Lv dxdt* +

**Remark 2.** *We can precise the structure of the elements of V. Indeed, let H<sup>θ</sup>* (*Q*) *be the weigthed*

<sup>2</sup> *dxdt* +

 *T* 0 *ω*

*a*(*v*, *v*),

*Q* 1 *<sup>θ</sup>*<sup>2</sup> <sup>|</sup>*v*<sup>|</sup>

<sup>2</sup> *dxdt*) 1

*h*0*χ<sup>O</sup> v dxdt*

*h*0*χ<sup>O</sup> v dxdt* ∀*v* ∈ *V*.

*<sup>θ</sup>* (*Q*)*, and let u be the unique solution of (3.3). We set*

is continuous. Therefore, from the Lax-Milgram theorem there exits a unique *u* in *V* solution

w = −*uχω*

*q* = *Lu*.

*Then, the pair* (w, *q*) *is such that (2.3)-(2.4) hold (i.e there is some insensitive sentinel defined by*

*u v dxdt*.

<sup>2</sup> *dxdt* < ∞

 ,

<sup>2</sup> *. This shows that V is imbedded*

*<sup>θ</sup>* (*Q*)), then from (3.2) and the

 *T* 0 *<sup>ω</sup>* <sup>|</sup>*u*<sup>|</sup>

<sup>2</sup> *dxdt*

<sup>2</sup> *dxdt* <sup>≤</sup> *<sup>C</sup>*

*<sup>θ</sup> bounded.*

*<sup>a</sup>*(*u*, *<sup>v</sup>*) =

Now if *<sup>h</sup>*<sup>0</sup> <sup>∈</sup> *<sup>L</sup>*2(*Q*) and *<sup>θ</sup>h*<sup>0</sup> <sup>∈</sup> *<sup>L</sup>*2(*Q*) (i.e. : *<sup>h</sup>*<sup>0</sup> <sup>∈</sup> *<sup>L</sup>*<sup>2</sup>

*<sup>a</sup>*(*u*, *<sup>v</sup>*) =

Let *V* be the completion of V with respect to the norm

*<sup>H</sup><sup>θ</sup>* (*Q*) =

*endowed with the natural norm* �*v*�*<sup>θ</sup>* = (

 *Q* 1 *<sup>θ</sup>*<sup>2</sup> <sup>|</sup>*u*<sup>|</sup>

*where <sup>θ</sup>* ∈ C2(*Q*) *positive with* <sup>1</sup>

defined by :

*Hilbert space defined by*

*continuously* �*v*�*<sup>θ</sup>* ≤ *C* �*v*�*V.*

of the variational equation :

**Theorem 2.** *Assume that h*<sup>0</sup> <sup>∈</sup> *<sup>L</sup>*<sup>2</sup>

*and*

*(1.4)-(1.5)).*

(*yo*)) *such that*

, (3.2)

$$\begin{cases} -q' - \Delta q + f'(y\_0)q = h\_0 \chi\_O + \mathbf{w} \chi\_{\omega} \text{ in } Q\_{\prime} \\ q(T) &= 0 \\ q &= 0 \\ \frac{\partial q}{\partial \nu} &= 0 \end{cases} \quad \text{in } \Omega\_{1\prime} \tag{2.3}$$

As for the problem (2.1), the problem (2.3) has a unique solution *q* (under mild assumptions on *f* � (*y*0)). The function *q* depends on the control w that we shall determine : Indeed, if we multiply the first equation in (2.3) by *yτ*, and we integrate by parts over *Q*, we obtain

$$\int\_{Q} \left( h\_{0} \chi\_{O} + \mathfrak{w} \chi\_{\omega} \right) y\_{\tau} \, d\mathbf{x} dt = \int\_{\Omega} q(0) \mathfrak{J}^{0} d\mathbf{x} \quad \forall \mathfrak{J}^{0}, \quad \|\mathfrak{J}^{0}\|\_{L^{2}(\Omega)} \le 1.$$

So, the condition (1.5) (or (2.2)) is equivalent to

$$q(0) = 0.\tag{2.4}$$

This is a null-controllability problem.

**Remark 1.** *The knowledge of the optimal control* w *provides informations about the pollution term λ* ˆ *ξ. Indeed, denote by*

$$L = \frac{\partial}{\partial t} - \Delta + f'(y\_0) \, I\_d$$

*and let y<sup>λ</sup>* <sup>=</sup> *<sup>∂</sup><sup>y</sup> ∂λ* (0, 0) *be the solution of*

$$\begin{cases} \begin{aligned} \,^L y\_\lambda &= \dot{\xi} \text{ in } \Omega\_{\dot{\nu}}\\ y\_\lambda(0) &= 0 \text{ in } \Omega\_{\dot{\nu}}\\ \,^{y\_\lambda} &= 0 \text{ on } \Sigma\_{1\nu} \end{aligned} \end{cases}$$

$$\begin{aligned} \begin{aligned} \frac{\partial y\_\lambda}{\partial \nu} &= 0 \text{ on } \Sigma\_2. \end{aligned} \end{aligned}$$

*Integrating by parts, we then obtain*

$$\int\_{Q} y\_{\lambda} \, L^\* q \, d\mathfrak{x} dt = \int\_{Q} q \hat{\xi} \, d\mathfrak{x} dt$$

*with*

$$L^\* = -\frac{\partial}{\partial t} - \Delta + f'(y\_0) \, I\_d.$$

*So that from (2.3) and (1.6) we deduce*

$$\int\_{Q} \lambda \xi \, d\mathbf{x} dt = \int\_{Q} (h\_0 \chi\_O + \mathbf{w} \chi\_{\omega}) (m\_o - y\_0) \, d\mathbf{x} dt.$$

#### **3. Existence of a sentinel**

We begin with some observability inequality, which will be proved in detail in the last section. Denote by

$$\mathcal{V} = \left\{ v \in \mathcal{C}^{\infty}(\overline{\mathcal{Q}}) \text{ such that } \left. \left. v \right|\_{\mathbb{E}\_1} = \frac{\partial v}{\partial t} \right|\_{\mathbb{E}\_1} = 0 \text{ and } \left. \frac{\partial v}{\partial v} \right|\_{\mathbb{E}\_2} = 0 \right\}. \tag{3.1}$$

Then we have :

**Theorem 1.** *Let be u* ∈ V*, then there exists a positive constant C* = *C*(Ω, *ω*,*O*, *T*, *f* � (*yo*)) *such that*

$$\int\_{Q} \frac{1}{\theta^2} |u|^2 \, d\mathbf{x} dt \le \mathbb{C} \left[ \int\_{Q} |\mathbf{L} \, u|^2 \, d\mathbf{x} dt + \int\_{0}^{T} \int\_{\omega} |u|^2 \, d\mathbf{x} dt \right],\tag{3.2}$$

*where <sup>θ</sup>* ∈ C2(*Q*) *positive with* <sup>1</sup> *<sup>θ</sup> bounded.*

4 Will-be-set-by-IN-TECH

As for the problem (2.1), the problem (2.3) has a unique solution *q* (under mild assumptions

Indeed, if we multiply the first equation in (2.3) by *yτ*, and we integrate by parts over *Q*, we

**Remark 1.** *The knowledge of the optimal control* w *provides informations about the pollution term*

− Δ + *f* �

*yλ*(0) = 0 *in* Ω, *y<sup>λ</sup>* = 0 *on* Σ1,

*∂ν* <sup>=</sup> <sup>0</sup> *on* <sup>Σ</sup>2.

− Δ + *f* �

We begin with some observability inequality, which will be proved in detail in the last section.

� *Q q* ˆ *ξ dxdt*

(*y*0) *Id*.

(*h*0*χ<sup>O</sup>* + w*χω*)(*mo* − *y*0) *dxdt*.

<sup>=</sup> *<sup>∂</sup><sup>v</sup> ∂t* � � Σ1 <sup>=</sup> 0 and *<sup>∂</sup><sup>v</sup>*

*∂ν* � � Σ2 = 0 �

. (3.1)

*L y<sup>λ</sup>* = <sup>ˆ</sup>

� Ω *q*(0)*y*ˆ

(*y*0)). The function *q* depends on the control w that we shall determine :

(*y*0)*q* = *h*0*χ<sup>O</sup>* + w*χω* in *Q*, *q*(*T*) = 0 in Ω, *q* = 0 on Σ1,

<sup>0</sup>*dx* <sup>∀</sup>*y*<sup>ˆ</sup>

(*y*0) *Id*

*ξ in Q*,

0, �*y*<sup>ˆ</sup>

*q*(0) = 0. (2.4)

<sup>0</sup>�*L*<sup>2</sup>(Ω) <sup>≤</sup> 1.

(2.3)

*∂ν* <sup>=</sup> <sup>0</sup> on <sup>Σ</sup>2.

⎧ ⎪⎪⎪⎨ −*q*� − Δ*q* + *f* �

�

*∂q*

*y<sup>τ</sup> dxdt* =

*<sup>L</sup>* <sup>=</sup> *<sup>∂</sup> ∂t*

> ⎧ ⎪⎪⎪⎨

> ⎪⎪⎪⎩

*<sup>L</sup>*<sup>∗</sup> <sup>=</sup> <sup>−</sup> *<sup>∂</sup> ∂t*

> � *Q*

*<sup>v</sup>* ∈ C∞(*Q*) such that : *<sup>v</sup>*|<sup>Σ</sup><sup>1</sup>

� *Q* *∂y<sup>λ</sup>*

*y<sup>λ</sup> L*∗*q dxdt* =

⎪⎪⎪⎩

*h*0*χ<sup>O</sup>* + w*χω*

So, the condition (1.5) (or (2.2)) is equivalent to

*∂λ* (0, 0) *be the solution of*

This is a null-controllability problem.

*Integrating by parts, we then obtain*

*So that from (2.3) and (1.6) we deduce*

**3. Existence of a sentinel**

V = � � *Q λ* ˆ *ξ dxdt* =

on *f* �

obtain

*λ* ˆ

*with*

Denote by

Then we have :

� *Q* �

*ξ. Indeed, denote by*

*and let y<sup>λ</sup>* <sup>=</sup> *<sup>∂</sup><sup>y</sup>*

According to the RHS of (3.2), we consider the space V endowed with the bilinear form *a*(., .) defined by :

$$a(u,v) \quad = \int\_{Q} Lu \, \mathcal{L}v \, d\mathbf{x}dt + \int\_{0}^{T} \int\_{\omega} u \, v \, d\mathbf{x}dt.$$

Let *V* be the completion of V with respect to the norm

$$v \mapsto \|v\|\_V = \sqrt{a(v, v)}\_\prime$$

then, *V* is a Hilbert space for the scalar product *a*(*v*, *v*ˆ) and the associated norm.

**Remark 2.** *We can precise the structure of the elements of V. Indeed, let H<sup>θ</sup>* (*Q*) *be the weigthed Hilbert space defined by*

$$H\_{\theta}(Q) = \left\{ v \in L^{2}(Q) \quad \text{such that} \, : \quad \int\_{Q} \frac{1}{\theta^{2}} \left| v \right|^{2} \, d\mathbf{x} dt < \infty \right\},$$

*endowed with the natural norm* �*v*�*<sup>θ</sup>* = ( *Q* 1 *<sup>θ</sup>*<sup>2</sup> |*v*| <sup>2</sup> *dxdt*) 1 <sup>2</sup> *. This shows that V is imbedded continuously* �*v*�*<sup>θ</sup>* ≤ *C* �*v*�*V.*

Now if *<sup>h</sup>*<sup>0</sup> <sup>∈</sup> *<sup>L</sup>*2(*Q*) and *<sup>θ</sup>h*<sup>0</sup> <sup>∈</sup> *<sup>L</sup>*2(*Q*) (i.e. : *<sup>h</sup>*<sup>0</sup> <sup>∈</sup> *<sup>L</sup>*<sup>2</sup> *<sup>θ</sup>* (*Q*)), then from (3.2) and the Cauchy-Schwartz inequality, we deduce that the linear form defined on *V* by

$$v \mapsto \int\_{Q} h\_{0} \chi\_{O} \, v \, dxdt$$

is continuous. Therefore, from the Lax-Milgram theorem there exits a unique *u* in *V* solution of the variational equation :

$$a(\mu, v) = \int\_{Q} h\_0 \chi\_{\mathcal{O}} v \, d\mathfrak{x} dt \qquad \forall v \in V.$$

**Theorem 2.** *Assume that h*<sup>0</sup> <sup>∈</sup> *<sup>L</sup>*<sup>2</sup> *<sup>θ</sup>* (*Q*)*, and let u be the unique solution of (3.3). We set*

$$
\Psi = -u\chi\_{\omega}
$$

*and*

$$q = Lu.$$

*Then, the pair* (w, *q*) *is such that (2.3)-(2.4) hold (i.e there is some insensitive sentinel defined by (1.4)-(1.5)).*

We use the method by Fursikov and Imanuvilov A. Fursikov & O. Yu. Imanuvilov (1996), Imanuvilov O. Yu. Imanuvilov (1995), Lebeau and Robbiano Lebeau G. & Robbiano L. (1995),

Some Aspects of the Sentinel Method for Pollution Problems 191

−*sη*(*x*,*t*)

*w* (*x*, 0) = *w* (*x*, *T*) = 0.

*P*1*w* + *P*2*w* = *gs*,

*<sup>∂</sup><sup>t</sup>* <sup>+</sup> <sup>2</sup>*sλϕ* <sup>∇</sup>*<sup>ψ</sup>* <sup>∇</sup>*<sup>w</sup>* <sup>+</sup> <sup>2</sup>*sλ*2*ϕ*|∇*ψ*<sup>|</sup>

 *Q*

*P*1*wP*2*w dxdt*. This will give 9 terms *Ik*,*l*.

In order to organize our calculus, we denote by *A* and *B* the quantities such that *A* contains

*ϕ* |∇*w*|

 *Q*

<sup>3</sup>*λ*<sup>3</sup> *Q <sup>ϕ</sup>*<sup>3</sup> <sup>|</sup>*w*<sup>|</sup>

We denote by *ν* the outer normal on Γ. Note down that *ψ* cancels on Γ. We then have the

Δ*w dxdt*

*∂ν <sup>d</sup>γdt* <sup>+</sup> 0,

<sup>2</sup>*w* + *s*

*∂η ∂t w*,

<sup>2</sup>*<sup>w</sup>* <sup>−</sup> *<sup>s</sup>λϕ* <sup>Δ</sup>*<sup>ψ</sup> <sup>w</sup>*.

*P*1*wP*2*w dxdt* =

<sup>2</sup> *dxdt*,

<sup>2</sup> *dxdt*.

<sup>2</sup>*w dxdt* = *B*,

<sup>2</sup>*λ*2*ϕ*2|∇*ψ*<sup>|</sup>

<sup>−</sup>*s<sup>η</sup> <sup>g</sup>* <sup>+</sup> *<sup>s</sup>λ*2*ϕ*|∇*ψ*<sup>|</sup>

<sup>2</sup> *dxdt* + 2

*u* (*x*, *t*).

<sup>2</sup>*w*,

 *<sup>Q</sup>* <sup>|</sup>*gs*<sup>|</sup>

<sup>2</sup> *dxdt*.

*w* (*x*, *t*) = *e*

and Puel Puel J.-P. (2001) (case Σ<sup>2</sup> = ∅ and Σ<sup>1</sup> = Σ).

Calculating *<sup>g</sup>* = (*∂<sup>t</sup>* <sup>−</sup> <sup>Δ</sup>)(*esηw*), with notation (4), we get

*<sup>P</sup>*1*<sup>w</sup>* <sup>=</sup> *<sup>∂</sup><sup>w</sup>*

*gs* = *e*

*<sup>Q</sup>* <sup>|</sup>*P*2*w*<sup>|</sup>

*c sλ* + *λ*<sup>2</sup>

*c s* <sup>2</sup>*λ*<sup>4</sup> + *s*

*I*1,1 = −

= − Σ *∂w ∂t ∂w*

*I*1,2 = −*s*

 *Q ∂w ∂t*

<sup>2</sup>*λ*<sup>2</sup> *Q ∂w <sup>∂</sup><sup>t</sup> <sup>ϕ</sup>*2|∇*ψ*<sup>|</sup>

<sup>2</sup> *dxdt* +

*Q*

all the terms which can be upper bounded by

and by *B* all those which can be bounded by

*P*2*w* = −Δ*w* − *s*

For *s* ≥ *s*<sup>0</sup> and *λ* ≥ *λ*0, we define

We easily notice that

Taking the *L*<sup>2</sup> norm we get : 

We shall now calculate

following results :

*<sup>Q</sup>* <sup>|</sup>*P*1*w*<sup>|</sup>

where

#### **4. Proof of theorem 1**

The proof for the observability inequality in theorem 1 will hold from Carleman estimates that we carefully show in the following results.

**Lemma 1.** *Let be <sup>ω</sup>*<sup>0</sup> *an open set such that <sup>ω</sup>*<sup>0</sup> <sup>⊂</sup> *<sup>ω</sup>. Then there is <sup>ψ</sup>* <sup>∈</sup> *<sup>C</sup>*2(Ω) *such that*

$$\begin{cases} \quad \psi(\mathbf{x}) & > 0 \,\,\forall \,\mathbf{x} \in \Omega, \\\quad \psi(\mathbf{x}) & = 0 \,\,\forall \,\mathbf{x} \in \Gamma, \\\, |\nabla \psi(\mathbf{x})| \neq 0 \,\,\forall \,\mathbf{x} \in \overline{\Omega - \omega\_{0}}. \end{cases}$$

See Imanuvilov O. Yu. Imanuvilov (1995)

We now use a function *ψ* as given by the previous lemma, to define convenient *weight* functions. For *λ* > 0, we set

$$\varphi(x,t) = \frac{e^{\lambda\psi(x)}}{t(T-t)}\text{ },$$

and

$$\eta(x,t) = \frac{e^{2\lambda|\psi|\_{\infty}} - e^{\lambda\psi(x)}}{t(T-t)}.$$

Then

$$
\nabla \varphi = \lambda \varrho \nabla \psi \,\,\, \nabla \eta = -\lambda \varrho \nabla \psi.
$$

We also notice the following properties :

$$\left|\frac{\partial\varphi}{\partial t}\right| \le T\varphi^2, \left|\frac{\partial^2\varphi}{\partial t^2}\right| \le T^2\varphi^3,\tag{4.1}$$

$$\left|\frac{\partial\eta}{\partial t}\right| \le T\rho^2, \left|\frac{\partial^2\eta}{\partial t^2}\right| \le T^2\rho^3. \tag{4.2}$$

**Remark 3.** *Note that η increases to* +∞ *when t* → *T or t* → 0*, but η is uniformly bounded on* Ω × [*δ*, *T* − *δ*] *for any δ* > 0*.*

*On the other hand, for fixed s* <sup>&</sup>gt; <sup>0</sup> *the function e*−*sη*(*x*,*t*) *goes to* <sup>0</sup> *when t* <sup>→</sup> *T or t* <sup>→</sup> <sup>0</sup>*.*

The following theorem states the Carleman inequalities concerning (3.1) :

� � � �

**Proposition 1.** *There exist constants s*<sup>0</sup> > 0*, λ*<sup>0</sup> > 0 *and C* > 0 *depending on* Ω*, ω, ψ, and T, such that for all s* ≥ *s*0*, λ* ≥ *λ*0*, and for any function u* ∈ V *given by (3.1), we have*

$$\begin{split} 2\mathbf{s}^{3}\lambda^{4}\int\_{Q}\rho^{3}e^{-2s\eta}\left|u\right|^{2}dxdt + 4\mathbf{s}^{2}\lambda\int\_{\Sigma\_{2}}\frac{\partial\eta}{\partial t}\frac{\partial\overline{\psi}}{\partial\upsilon}e^{-2s\eta}\left|u\right|^{2}d\gamma dt \\ -4\mathbf{s}^{3}\lambda^{3}\int\_{\Sigma\_{2}}\rho^{3}|\nabla\psi|^{2}\frac{\partial\overline{\psi}}{\partial\upsilon}e^{-2s\eta}|u|^{2}d\gamma dt - 4\mathbf{s}^{2}\lambda^{3}\int\_{\Sigma\_{2}}\rho^{2}|\nabla\psi|^{2}\frac{\partial\overline{\psi}}{\partial\upsilon}e^{-2s\eta}|u|^{2}d\gamma dt \\ -2s\lambda\int\_{\Sigma\_{2}}\rho\frac{\partial\overline{\psi}}{\partial\upsilon}e^{-2s\eta}\frac{\partial u}{\partial t}u\,d\gamma dt - 4s\lambda\int\_{\Sigma\_{1}}\rho\nabla\Psi e^{-2s\eta}\nabla u\,\frac{\partial u}{\partial \upsilon}d\gamma dt \\ +2s\lambda\int\_{\Sigma}\rho\frac{\partial\overline{\psi}}{\partial\upsilon}e^{-2s\eta}|\nabla u|^{2}\,d\gamma dt \\ \leq \quad\tag{4.3} \\ \leq \quad \mathsf{C}\left(\int\_{Q}e^{-2s\eta}\left|\frac{\partial u}{\partial t} - \Delta u\right|^{2}dxdt + s^{3}\lambda^{4}\int\_{0}^{T}\int\_{\omega}\rho^{3}e^{-2s\eta}|u|^{2}\,d\mathbf{x}dt\right). \end{split} \tag{4.4}$$

We use the method by Fursikov and Imanuvilov A. Fursikov & O. Yu. Imanuvilov (1996), Imanuvilov O. Yu. Imanuvilov (1995), Lebeau and Robbiano Lebeau G. & Robbiano L. (1995), and Puel Puel J.-P. (2001) (case Σ<sup>2</sup> = ∅ and Σ<sup>1</sup> = Σ). For *s* ≥ *s*<sup>0</sup> and *λ* ≥ *λ*0, we define

$$w\left(\mathbf{x},t\right) = e^{-s\eta\left(\mathbf{x},t\right)}u\left(\mathbf{x},t\right).$$

We easily notice that

$$w\left(\mathbf{x},0\right) = w\left(\mathbf{x},T\right) = 0.$$

Calculating *<sup>g</sup>* = (*∂<sup>t</sup>* <sup>−</sup> <sup>Δ</sup>)(*esηw*), with notation (4), we get

$$P\_1 w + P\_2 w = g\_{s\lambda}$$

where

6 Will-be-set-by-IN-TECH

The proof for the observability inequality in theorem 1 will hold from Carleman estimates that

*ψ*(*x*) > 0 ∀ *x* ∈ Ω, *ψ*(*x*) = 0 ∀ *x* ∈ Γ, |∇*ψ*(*x*)| �= 0 ∀ *x* ∈ Ω − *ω*0.

We now use a function *ψ* as given by the previous lemma, to define convenient *weight*

*<sup>ϕ</sup>*(*x*, *<sup>t</sup>*) = *<sup>e</sup>λψ*(*x*)

*<sup>η</sup>*(*x*, *<sup>t</sup>*) = *<sup>e</sup>*2*λ*|*ψ*|<sup>∞</sup> <sup>−</sup> *<sup>e</sup>λψ*(*x*)

∇*ϕ* = *λϕ*∇*ψ*, ∇*η* = −*λϕ*∇*ψ*.

� � � � *∂*2*ϕ ∂t*<sup>2</sup> � � � �

� � � � *∂*2*η ∂t*<sup>2</sup> � � � �

**Remark 3.** *Note that η increases to* +∞ *when t* → *T or t* → 0*, but η is uniformly bounded on*

**Proposition 1.** *There exist constants s*<sup>0</sup> > 0*, λ*<sup>0</sup> > 0 *and C* > 0 *depending on* Ω*, ω, ψ, and T, such*

<sup>2</sup> *dxdt* + 4*s*

<sup>2</sup>*λ*<sup>3</sup> � Σ2

*u dγdt* − 4*sλ*

*dxdt* + *s*

2*λ* � Σ2 *<sup>ϕ</sup> ∂η ∂t ∂ψ ∂ν e*

> � Σ1

+2*sλ* � Σ *<sup>ϕ</sup> ∂ψ ∂ν e*

<sup>3</sup>*λ*<sup>4</sup> � *T* 0 � *ω ϕ*3*e*

*<sup>ϕ</sup>*2|∇*ψ*<sup>|</sup>

*ϕ*∇*ψe*

<sup>2</sup> *∂ψ ∂ν e*

<sup>≤</sup> *<sup>T</sup>ϕ*2,

<sup>≤</sup> *<sup>T</sup>ϕ*2,

*On the other hand, for fixed s* <sup>&</sup>gt; <sup>0</sup> *the function e*−*sη*(*x*,*t*) *goes to* <sup>0</sup> *when t* <sup>→</sup> *T or t* <sup>→</sup> <sup>0</sup>*.*

The following theorem states the Carleman inequalities concerning (3.1) :

*that for all s* ≥ *s*0*, λ* ≥ *λ*0*, and for any function u* ∈ V *given by (3.1), we have*

<sup>−</sup>2*s<sup>η</sup>* <sup>|</sup>*u*<sup>|</sup>

<sup>−</sup>2*s<sup>η</sup>* <sup>|</sup>*u*<sup>|</sup>

<sup>−</sup>2*s<sup>η</sup> <sup>∂</sup><sup>u</sup> ∂t*

<sup>2</sup> *<sup>d</sup>γdt* <sup>−</sup> <sup>4</sup>*<sup>s</sup>*

� � � � 2

2*s* <sup>3</sup>*λ*<sup>4</sup> � *Q ϕ*3*e*

<sup>2</sup> *∂ψ ∂ν e*

*<sup>ϕ</sup>*3|∇*ψ*<sup>|</sup>

−2*sλ* � Σ2 *<sup>ϕ</sup> ∂ψ ∂ν e*

> � � *Q e* −2*sη* � � � � *∂u <sup>∂</sup><sup>t</sup>* <sup>−</sup> <sup>Δ</sup>*<sup>u</sup>*

≤ *C*

*t*(*T* − *t*)

*<sup>t</sup>*(*<sup>T</sup>* <sup>−</sup> *<sup>t</sup>*) .

,

<sup>≤</sup> *<sup>T</sup>*2*ϕ*3, (4.1)

<sup>≤</sup> *<sup>T</sup>*2*ϕ*3. (4.2)

<sup>−</sup>2*sη*|*u*<sup>|</sup>

<sup>−</sup>2*sη*|*u*<sup>|</sup>

*∂u ∂ν <sup>d</sup>γdt*

<sup>−</sup>2*sη*∇*<sup>u</sup>*

<sup>−</sup>2*s<sup>η</sup>* |∇*u*<sup>|</sup>

<sup>−</sup>2*s<sup>η</sup>* <sup>|</sup>*u*<sup>|</sup>

<sup>2</sup> *dγdt*

<sup>2</sup> *dγdt*

<sup>2</sup> *dγdt*

(4.3)

<sup>2</sup> *dxdt* � .

**Lemma 1.** *Let be <sup>ω</sup>*<sup>0</sup> *an open set such that <sup>ω</sup>*<sup>0</sup> <sup>⊂</sup> *<sup>ω</sup>. Then there is <sup>ψ</sup>* <sup>∈</sup> *<sup>C</sup>*2(Ω) *such that*

⎧ ⎨ ⎩

> � � � � *∂ϕ ∂t* � � � �

� � � � *∂η ∂t* � � � �

**4. Proof of theorem 1**

we carefully show in the following results.

See Imanuvilov O. Yu. Imanuvilov (1995)

We also notice the following properties :

functions. For *λ* > 0, we set

Ω × [*δ*, *T* − *δ*] *for any δ* > 0*.*

<sup>−</sup>4*s*3*λ*<sup>3</sup>

� Σ2

and

Then

$$P\_1 w = \frac{\partial w}{\partial t} + 2s\lambda \varrho |\nabla \psi \nabla w + 2s\lambda^2 \varrho |\nabla \psi|^2 w\_\prime$$

$$P\_2 w = -\Delta w - s^2 \lambda^2 \varrho^2 |\nabla \psi|^2 w + s \frac{\partial \eta}{\partial t} w\_\prime$$

$$g\_s = e^{-s\eta} g + s\lambda^2 \varrho |\nabla \psi|^2 w - s\lambda \varrho \Delta \psi w.$$

Taking the *L*<sup>2</sup> norm we get :

$$\int\_{Q} |P\_1 w|^2 \, d\mathbf{x} dt + \int\_{Q} |P\_2 w|^2 \, d\mathbf{x} dt + 2 \int\_{Q} P\_1 w P\_2 w \, d\mathbf{x} dt = \int\_{Q} |\mathbf{g}\_s|^2 \, d\mathbf{x} dt.$$

We shall now calculate *Q P*1*wP*2*w dxdt*. This will give 9 terms *Ik*,*l*.

In order to organize our calculus, we denote by *A* and *B* the quantities such that *A* contains all the terms which can be upper bounded by

$$c\left(s\lambda + \lambda^2\right) \int\_Q \varrho \left|\nabla w\right|^2 \,d\mathbf{x}dt\_\lambda$$

and by *B* all those which can be bounded by

$$c\left(s^2\lambda^4 + s^3\lambda^3\right)\int\_{\mathcal{Q}} \rho^3 \left|w\right|^2 \,d\mathbf{x}dt.$$

We denote by *ν* the outer normal on Γ. Note down that *ψ* cancels on Γ. We then have the following results :

$$\begin{aligned} I\_{1,1} &= -\int\_{Q} \frac{\partial w}{\partial t} \Delta w \, dxdt \\ &= -\int\_{\Sigma} \frac{\partial w}{\partial t} \frac{\partial w}{\partial \nu} \, d\gamma dt + 0, \\ I\_{1,2} &= -s^2 \lambda^2 \int\_{Q} \frac{\partial w}{\partial t} \varphi^2 |\nabla \psi|^2 w \, dxdt = B\_{\nu} \end{aligned}$$

But |∇*ψ*| �= 0 on Ω − *ω*0, hence there is *δ* > 0 such that

 *<sup>Q</sup>* <sup>|</sup>*gs*<sup>|</sup>

<sup>2</sup> *dxdt* +

*<sup>Q</sup>* <sup>|</sup>*P*1*w*<sup>|</sup>

+2*sλ* Σ *<sup>ϕ</sup> ∂ψ ∂ν* |∇*w*<sup>|</sup>

+2*s* <sup>3</sup>*λ*4*δ*<sup>4</sup> *Q <sup>ϕ</sup>*<sup>3</sup> <sup>|</sup>*w*<sup>|</sup>

*<sup>Q</sup>* <sup>|</sup>*P*1*w*<sup>|</sup>

+2*s* 2*λ* Σ *<sup>ϕ</sup> ∂η ∂t ∂ψ ∂ν* <sup>|</sup>*w*<sup>|</sup>

*ϕP*2*wθ*2*wdxdt* +

*ϕ* |∇*w*|

<sup>2</sup> *dxdt* + *B* + 2*sλ*2*δ*<sup>2</sup>

On the other hand

Consequently :

≤ *Q e* <sup>−</sup>2*s<sup>η</sup>* <sup>|</sup>*g*<sup>|</sup>

> ≤ *C Q*

2*sλ*2*δ*<sup>2</sup>

 *T* 0 *ω*<sup>0</sup>

so that


Some Aspects of the Sentinel Method for Pollution Problems 193

<sup>2</sup> *dxdt* + 2

≤ *Q e* <sup>−</sup>2*s<sup>η</sup>* <sup>|</sup>*g*<sup>|</sup>

 Σ *∂w ∂t ∂w*

*<sup>Q</sup>* <sup>|</sup>*P*2*w*<sup>|</sup>

<sup>2</sup> *<sup>d</sup>γdt* <sup>−</sup> <sup>2</sup>*<sup>s</sup>*

<sup>2</sup> *<sup>d</sup>γdt* <sup>−</sup> <sup>4</sup>*sλ*<sup>2</sup>

*ϕ* |∇*w*|

*ϕθ*<sup>2</sup> |∇*w*<sup>|</sup>

<sup>2</sup> *<sup>θ</sup>*<sup>2</sup> |∇*w*<sup>|</sup> *<sup>ϕ</sup>*<sup>1</sup>

<sup>Ω</sup> <sup>|</sup>*P*2*w*<sup>|</sup>

<sup>−</sup>2*s<sup>η</sup>* <sup>|</sup>*u*<sup>|</sup> 2.

<sup>2</sup> *dxdt* + *B*,

*P*1*wP*2*w dxdt*

<sup>2</sup> *dxdt* + *B*.

 Σ

*<sup>ϕ</sup>*3|∇*ψ*<sup>|</sup>

<sup>2</sup> *∂w*

 *T* 0 *ω*<sup>0</sup>

 *Q*

*ϕ*∇*ψ*.∇*w*

<sup>2</sup> *∂ψ ∂ν* <sup>|</sup>*w*<sup>|</sup>

*∂ν* .*w dγdt* <sup>+</sup> *<sup>A</sup>* <sup>+</sup> *<sup>B</sup>*

*<sup>ϕ</sup>*<sup>3</sup> <sup>|</sup>*w*<sup>|</sup>

*ϕ*3*w*2*dxdt*

<sup>2</sup> *dxdt*,

*ϕ* |∇*w*|

*∂w ∂ν <sup>d</sup>γdt*

<sup>2</sup> *dxdt*

<sup>2</sup> *dγdt*

<sup>2</sup> *dxdt*.

,

<sup>2</sup> *dxdt* + 2*sλ*2*δ*<sup>2</sup>

*∂ν <sup>d</sup>γdt* <sup>−</sup> <sup>4</sup>*s<sup>λ</sup>*

 Σ *ϕ*|∇*ψ*|

<sup>2</sup> *dxdt* + 2*s*

<sup>2</sup> *dxdt*

<sup>2</sup> *w dxdt* + *s*

<sup>2</sup> *dxdt* + *cs*3*λ*<sup>4</sup>

<sup>3</sup>*λ*<sup>3</sup> Σ

<sup>3</sup>*λ*4*δ*<sup>4</sup>

<sup>2</sup>*λ*<sup>2</sup> *T* 0 *ω*

> *T* 0 *ω <sup>ϕ</sup>*<sup>3</sup> <sup>|</sup>*w*<sup>|</sup>

 *Q*

 *Q e* <sup>−</sup>2*s<sup>η</sup>* <sup>|</sup>*g*<sup>|</sup>

<sup>2</sup> *dxdt* <sup>≤</sup>

*<sup>Q</sup>* <sup>|</sup>*P*2*w*<sup>|</sup>

<sup>2</sup> *dxdt* +

<sup>2</sup> *dxdt* <sup>−</sup> <sup>2</sup>

 *T* 0 *ω*<sup>0</sup>

 *T* 0 *ω*

> 1 2 *T* 0

Now, we should write the inequality below in terms of the solution *u*, since


 *T* 0 *ω ϕ*1

<sup>2</sup> *dxdt* <sup>≤</sup>

for *θ* ∈ D(*ω*) such that 0 ≤ *θ* ≤ 1 and *θ*(*x*) = 1 on *ω*<sup>0</sup> gives

We can eliminate *A* and *B* by choosing *s* and *λ* large enough. And we observe that :

$$\begin{split} I\_{1,3} &= s \int\_{Q} \frac{\partial w}{\partial t} \frac{\partial \eta}{\partial t} w \, d\mathbf{x} dt = \frac{s}{2} \int\_{Q} \frac{\partial \eta}{\partial t} \frac{\partial}{\partial t} \left( |w|^2 \right) \, d\mathbf{x} dt \\ &= -\frac{s}{2} \int\_{Q} \frac{\partial^2 \eta}{\partial t^2} |w|^2 \, d\mathbf{x} dt = B. \end{split}$$

And,

$$\begin{split} I\_{2,1} &= -2s\lambda \int\_{Q} \oint \boldsymbol{\varrho} \nabla \boldsymbol{\psi} \nabla \boldsymbol{w} \Delta \boldsymbol{w} \, d\mathbf{x} dt \\ &= -2s\lambda \int\_{\Sigma} \boldsymbol{\varrho} \nabla \boldsymbol{\psi} \nabla \boldsymbol{w} \frac{\partial \boldsymbol{w}}{\partial \boldsymbol{\nu}} \, d\boldsymbol{\gamma} dt + 2s\lambda^{2} \int\_{Q} \boldsymbol{\varrho} |\nabla \boldsymbol{\psi} \nabla \boldsymbol{w}|^{2} \, d\mathbf{x} dt \\ &+ s\lambda \int\_{\Sigma} \boldsymbol{\varrho} \frac{\partial \boldsymbol{\Psi}}{\partial \boldsymbol{\nu}} |\nabla \boldsymbol{w}|^{2} \, d\boldsymbol{\gamma} dt - s\lambda^{2} \int\_{Q} \boldsymbol{\varrho} |\nabla \boldsymbol{\psi}|^{2} |\nabla \boldsymbol{w}|^{2} \, d\mathbf{x} dt + A\_{\nu} \end{split}$$

$$\begin{split} I\_{2,2} &= -2s^3 \lambda^3 \int\_Q \varphi^3 |\nabla \psi|^2 \nabla \psi \nabla w.w = -s^3 \lambda^3 \int\_Q \varphi^3 |\nabla \psi|^2 \nabla \psi \nabla |w|^2 \\ &= -s^3 \lambda^3 \int\_\Sigma \varphi^3 |\nabla \psi|^2 \frac{\partial \psi}{\partial \nu} |w|^2 + 3s^3 \lambda^4 \int\_Q \varphi^3 |\nabla \psi|^4 |w|^2 + B\_\nu \end{split}$$

$$\begin{aligned} I\_{2,3} &= 2s^2 \lambda \int\_Q \varrho \nabla \psi \, \nabla w \, \frac{\partial \eta}{\partial t} w \, dxdt \\ &= s^2 \lambda \int\_\Sigma \varrho \frac{\partial \eta}{\partial t} \frac{\partial \psi}{\partial \nu} |w|^2 \, dxdt + B. \end{aligned}$$

Finally :

$$\begin{aligned} I\_{3,1} &= -2s\lambda^2 \int\_Q \varrho |\nabla \psi|^2 \Delta w.w \, dxdt \\ &= -2s\lambda^2 \int\_\Sigma \varrho \, |\nabla \psi|^2 \frac{\partial w}{\partial \nu}.w + 2s\lambda^2 \int\_Q \varrho |\nabla \psi|^2 |\nabla w|^2 + A + B. \end{aligned}$$

$$I\_{3,2} = -2s^3 \lambda^4 \int\_Q \varrho^3 |\nabla \psi|^4 |w|^2 \, dxdt,$$

$$I\_{3,3} = 2s^2 \lambda^2 \int\_Q \varrho \frac{\partial \eta}{\partial t} |\nabla \psi|^2 |w|^2 \, dxdt = B.$$

Summing all the terms, it follows :

$$\begin{array}{ll} 2\int\_{Q} P\_{1}w P\_{2}w \,d\boldsymbol{x}dt = \boldsymbol{A} + \boldsymbol{B} + 2s\boldsymbol{\lambda}^{2} \int\_{Q} \boldsymbol{\varrho} |\nabla \boldsymbol{\psi}|^{2} |\nabla w|^{2} \,d\boldsymbol{x}dt \\ & + 2s^{3}\boldsymbol{\lambda}^{4} \int\_{Q} \boldsymbol{\varrho}^{3} |\nabla \boldsymbol{\psi}|^{4} \,|w|^{2} \,d\boldsymbol{x}dt + 4s\boldsymbol{\lambda}^{2} \int\_{Q} \boldsymbol{\varrho} |\nabla \boldsymbol{\psi} \cdot \nabla w|^{2} \,d\boldsymbol{x}dt \\ & - 2\int\_{\Sigma} \frac{\partial \boldsymbol{w}}{\partial t} \frac{\partial \boldsymbol{w}}{\partial \boldsymbol{\varrho}} \,d\boldsymbol{\gamma} dt - 4s\boldsymbol{\lambda} \int\_{\Sigma} \boldsymbol{\varrho} \nabla \boldsymbol{\psi} \cdot \nabla \boldsymbol{w} \frac{\partial \boldsymbol{w}}{\partial \boldsymbol{\varrho}} \,d\boldsymbol{\gamma} dt \\ & + 2s\boldsymbol{\lambda} \int\_{\Sigma} \boldsymbol{\varrho} \frac{\partial \boldsymbol{\uprho}}{\partial \boldsymbol{\varrho}} |\nabla \boldsymbol{w}|^{2} \,d\boldsymbol{\gamma} dt - 2s^{3}\boldsymbol{\lambda}^{3} \int\_{\Sigma} \boldsymbol{\varrho}^{3} |\nabla \boldsymbol{\psi}|^{2} \frac{\partial \boldsymbol{\uprho}}{\partial \boldsymbol{\varrho}} |\boldsymbol{w}|^{2} \,d\boldsymbol{\uprho} dt \\ & + 2s^{2}\boldsymbol{\lambda} \int\_{\Sigma} \boldsymbol{\varrho} \frac{\partial \boldsymbol{\uprho}}{\partial t} \frac{\partial \boldsymbol{\uprho}}{\partial \boldsymbol{\varrho}} |\boldsymbol{w}|^{2} \,d\boldsymbol{\uprho} dt - 4s\boldsymbol{\lambda}^{2} \int\_{\Sigma} \boldsymbol{\varrho} \,|\nabla \boldsymbol{\psi}|^{2} \frac{\partial \boldsymbol{w}$$

But |∇*ψ*| �= 0 on Ω − *ω*0, hence there is *δ* > 0 such that

$$|\nabla \psi| \ge \delta \quad \text{on} \quad \overline{\Omega - \omega\_0}.$$

On the other hand

$$\int\_{Q} \left| g\_{s} \right|^{2} d\mathfrak{x} dt \le \int\_{Q} e^{-2s\eta} \left| g \right|^{2} d\mathfrak{x} dt + B\_{\eta}$$

so that

8 Will-be-set-by-IN-TECH

2 *Q ∂η ∂t ∂ ∂t* |*w*| 2 *dxdt*

<sup>2</sup> *dxdt* = *B*.

*∂ν <sup>d</sup>γdt* <sup>+</sup> <sup>2</sup>*sλ*<sup>2</sup>

<sup>2</sup> *<sup>d</sup>γdt* <sup>−</sup> *<sup>s</sup>λ*<sup>2</sup>

<sup>2</sup>∇*ψ*∇*w*.*<sup>w</sup>* <sup>=</sup> <sup>−</sup>*<sup>s</sup>*

<sup>2</sup> + 3*s* <sup>3</sup>*λ*<sup>4</sup> *Q*

 *Q*

<sup>3</sup>*λ*<sup>3</sup> *Q*

 *Q ϕ*|∇*ψ*|

*ϕ*|∇*ψ*.∇*w*|

<sup>2</sup>|∇*w*<sup>|</sup>

*<sup>ϕ</sup>*3|∇*ψ*<sup>|</sup>

<sup>2</sup>|∇*w*<sup>|</sup>

<sup>4</sup> <sup>|</sup>*w*<sup>|</sup>

*<sup>ϕ</sup>*<sup>3</sup> |∇*ψ*<sup>|</sup>

<sup>2</sup> *dxdt*

<sup>2</sup> *dxdt* + *A*,

<sup>2</sup>∇*ψ*∇|*w*<sup>|</sup>

<sup>2</sup> + *B*,

<sup>2</sup> + *A* + *B*,

*ϕ*|∇*ψ*.∇*w*|

<sup>2</sup> *∂ψ ∂ν* <sup>|</sup>*w*<sup>|</sup>

<sup>2</sup> *∂w*

<sup>2</sup> *dxdt*

<sup>2</sup> *dγdt*

*∂ν* .*w dγdt*.

2

*w dxdt* <sup>=</sup> *<sup>s</sup>*

*∂*2*η <sup>∂</sup>t*<sup>2</sup> <sup>|</sup>*w*<sup>|</sup>

*ϕ*∇*ψ*∇*w*Δ*w dxdt*

*∂w*

<sup>2</sup> *∂ψ ∂ν* <sup>|</sup>*w*<sup>|</sup>

*∂t*

*w dxdt*

<sup>2</sup> *dxdt* + *B*.

<sup>2</sup>Δ*w*.*w dxdt*

*∂ν* .*<sup>w</sup>* <sup>+</sup> <sup>2</sup>*sλ*<sup>2</sup>

<sup>2</sup> *dxdt*,

<sup>2</sup> *dxdt* = *B*.

<sup>4</sup> <sup>|</sup>*w*<sup>|</sup>

<sup>2</sup>|∇*w*<sup>|</sup>

 Σ

<sup>2</sup> *<sup>d</sup>γdt* <sup>−</sup> <sup>4</sup>*sλ*<sup>2</sup>

<sup>2</sup> *<sup>d</sup>γdt* <sup>−</sup> <sup>2</sup>*<sup>s</sup>*

<sup>2</sup> *dxdt*

*ϕ*∇*ψ*.∇*w*

 Σ

<sup>3</sup>*λ*<sup>3</sup> Σ  *Q*

*∂w ∂ν <sup>d</sup>γdt*

*<sup>ϕ</sup>*3|∇*ψ*<sup>|</sup>

*ϕ* |∇*ψ*|

<sup>2</sup> *dxdt* + 4*sλ*<sup>2</sup>

 *Q ϕ*|∇*ψ*|

<sup>2</sup> *∂w*

<sup>4</sup>|*w*<sup>|</sup>

 *Q ϕ*|∇*ψ*|

*<sup>ϕ</sup>*3|∇*ψ*<sup>|</sup>

*∂ν <sup>d</sup>γdt* <sup>−</sup> <sup>4</sup>*s<sup>λ</sup>*

*ϕ*∇*ψ*.∇*w*

*<sup>ϕ</sup>*3|∇*ψ*<sup>|</sup>

*<sup>ϕ</sup>*<sup>3</sup> |∇*ψ*<sup>|</sup>

*<sup>ϕ</sup>*∇*<sup>ψ</sup>* <sup>∇</sup>*<sup>w</sup> ∂η*

*ϕ* |∇*ψ*|

*<sup>ϕ</sup>*3|∇*ψ*<sup>|</sup>

*I*1,3 = *s*

*I*2,1 = −2*sλ*

*I*2,2 = −2*s*

*I*2,3 = 2*s*

= *s* 2*λ* Σ *<sup>ϕ</sup> ∂η ∂t ∂ψ ∂ν* <sup>|</sup>*w*<sup>|</sup>

*<sup>I</sup>*3,1 <sup>=</sup> <sup>−</sup>2*sλ*<sup>2</sup>

*I*3,2 = −2*s*

*I*3,3 = 2*s*

Summing all the terms, it follows :

2 *Q* <sup>=</sup> <sup>−</sup>2*sλ*<sup>2</sup>

 *Q ϕ*|∇*ψ*|

 Σ

<sup>3</sup>*λ*<sup>4</sup> *Q*

<sup>2</sup>*λ*<sup>2</sup> *Q <sup>ϕ</sup> ∂η ∂t* |∇*ψ*| <sup>2</sup>|*w*<sup>|</sup>

*P*1*wP*2*w dxdt*=*A* + *B* + 2*sλ*<sup>2</sup>

+2*s*3*λ*<sup>4</sup>

−2 Σ *∂w ∂t ∂w*

+2*sλ* Σ *<sup>ϕ</sup> ∂ψ ∂ν* |∇*w*<sup>|</sup>

+2*s*2*λ* Σ *<sup>ϕ</sup> ∂η ∂t ∂ψ ∂ν* <sup>|</sup>*w*<sup>|</sup>

 *Q*

= −*s* <sup>3</sup>*λ*<sup>3</sup> Σ

> 2*λ Q*

= −2*sλ*

+*sλ* Σ *<sup>ϕ</sup> ∂ψ ∂ν* |∇*w*<sup>|</sup>

> <sup>3</sup>*λ*<sup>3</sup> *Q*

And,

Finally :

 *Q ∂w ∂t ∂η ∂t*

<sup>=</sup> <sup>−</sup> *<sup>s</sup>* 2 *Q*

> *Q*

 Σ

$$\begin{aligned} \int\_{Q} |P\_1 w|^2 \, dxdt + \int\_{Q} |P\_2 w|^2 \, dxdt + 2 \int\_{Q} P\_1 w P\_2 w \, dxdt \\ \leq \int\_{Q} e^{-2s\eta} \, |g|^2 \, dxdt + B. \end{aligned}$$

Consequently :

≤

 *<sup>Q</sup>* <sup>|</sup>*P*1*w*<sup>|</sup> <sup>2</sup> *dxdt* + *<sup>Q</sup>* <sup>|</sup>*P*2*w*<sup>|</sup> <sup>2</sup> *dxdt* + 2*sλ*2*δ*<sup>2</sup> *Q ϕ* |∇*w*| <sup>2</sup> *dxdt* +2*s* <sup>3</sup>*λ*4*δ*<sup>4</sup> *Q <sup>ϕ</sup>*<sup>3</sup> <sup>|</sup>*w*<sup>|</sup> <sup>2</sup> *dxdt* <sup>−</sup> <sup>2</sup> Σ *∂w ∂t ∂w ∂ν <sup>d</sup>γdt* <sup>−</sup> <sup>4</sup>*s<sup>λ</sup>* Σ *ϕ*∇*ψ*.∇*w ∂w ∂ν <sup>d</sup>γdt* +2*sλ* Σ *<sup>ϕ</sup> ∂ψ ∂ν* |∇*w*<sup>|</sup> <sup>2</sup> *<sup>d</sup>γdt* <sup>−</sup> <sup>2</sup>*<sup>s</sup>* <sup>3</sup>*λ*<sup>3</sup> Σ *<sup>ϕ</sup>*3|∇*ψ*<sup>|</sup> <sup>2</sup> *∂ψ ∂ν* <sup>|</sup>*w*<sup>|</sup> <sup>2</sup> *dγdt* +2*s* 2*λ* Σ *<sup>ϕ</sup> ∂η ∂t ∂ψ ∂ν* <sup>|</sup>*w*<sup>|</sup> <sup>2</sup> *<sup>d</sup>γdt* <sup>−</sup> <sup>4</sup>*sλ*<sup>2</sup> Σ *ϕ*|∇*ψ*| <sup>2</sup> *∂w ∂ν* .*w dγdt* <sup>+</sup> *<sup>A</sup>* <sup>+</sup> *<sup>B</sup> Q e* <sup>−</sup>2*s<sup>η</sup>* <sup>|</sup>*g*<sup>|</sup> <sup>2</sup> *dxdt* + *B* + 2*sλ*2*δ*<sup>2</sup> *T* 0 *ω*<sup>0</sup> *ϕ* |∇*w*| <sup>2</sup> *dxdt* + 2*s* <sup>3</sup>*λ*4*δ*<sup>4</sup> *T* 0 *ω*<sup>0</sup> *<sup>ϕ</sup>*<sup>3</sup> <sup>|</sup>*w*<sup>|</sup> <sup>2</sup> *dxdt*.

We can eliminate *A* and *B* by choosing *s* and *λ* large enough. And we observe that :

$$\int\_0^T \int\_{\omega} \varrho \theta^2 \left| \nabla w \right|^2 dxdt$$

$$0 \le \mathcal{C} \left( \int\_Q \varrho P\_2 w \theta^2 w dxdt + \int\_0^T \int\_{\omega} \varrho^{\frac{1}{2}} \theta^2 \left| \nabla w \right| \varrho^{\frac{1}{2}} w \, dxdt + s^2 \lambda^2 \int\_0^T \int\_{\omega} \varrho^3 w^2 dxdt \right).$$

for *θ* ∈ D(*ω*) such that 0 ≤ *θ* ≤ 1 and *θ*(*x*) = 1 on *ω*<sup>0</sup> gives

$$2s\lambda^2 \delta^2 \int\_0^T \int\_{\omega\_0} \varrho \left| \nabla w \right|^2 dxdt \le \frac{1}{2} \int\_0^T \int\_{\Omega} \left| P\_2 w \right|^2 dxdt + cs^3 \lambda^4 \int\_0^T \int\_{\omega} \varrho^3 \left| w \right|^2 dxdt, \lambda$$

Now, we should write the inequality below in terms of the solution *u*, since

$$|w|^2 = e^{-2s\eta}|u|^2.$$

and

Then

where

we still have the properties (4.1) and (4.2).

For *s* ≥ *s*<sup>0</sup> and *λ* ≥ *λ*0, we define

Calculating *<sup>P</sup><sup>w</sup>* <sup>=</sup> *<sup>e</sup>*−*s<sup>η</sup><sup>g</sup>* <sup>=</sup> *<sup>e</sup>*−*s<sup>η</sup>*

We easily notice that

+4*s* <sup>3</sup>*λ*<sup>3</sup> Σ2

*where* <sup>1</sup>

*<sup>θ</sup>*<sup>2</sup> <sup>=</sup> *<sup>ϕ</sup>*3*<sup>e</sup>*

*<sup>η</sup>*(*x*, *<sup>t</sup>*) = *<sup>e</sup>*2*λ*|*ψ*|<sup>∞</sup> <sup>−</sup> *<sup>e</sup>*−*λψ*(*x*)

Some Aspects of the Sentinel Method for Pollution Problems 195

<sup>∇</sup>*<sup>ϕ</sup>* <sup>=</sup> <sup>−</sup>*λϕ*<sup>∇</sup>*ψ*, <sup>∇</sup>*<sup>η</sup>* <sup>=</sup> *λϕ*<sup>∇</sup>*ψ*;

<sup>−</sup>*s<sup>η</sup>*(*x*,*t*)

 , we set

*<sup>∂</sup><sup>t</sup>* <sup>−</sup> <sup>2</sup>*sλϕ* <sup>∇</sup>*<sup>ψ</sup>* <sup>∇</sup>*<sup>w</sup>* <sup>+</sup> <sup>2</sup>*sλ*2*<sup>ϕ</sup>*|∇*ψ*<sup>|</sup>

<sup>2</sup>*<sup>w</sup>* <sup>+</sup> *<sup>s</sup>*

2*λ* Σ2 *ϕ∂η ∂t ∂ψ ∂ν e*

> Σ1

−2*sλ* Σ *ϕ∂ψ ∂ν e*

<sup>3</sup>*λ*<sup>4</sup> *T* 0 *ω ϕ*3*e*

*∂η ∂t w*,

*<sup>ϕ</sup>*<sup>2</sup>|∇*ψ*<sup>|</sup>

*<sup>ϕ</sup>*<sup>∇</sup>*ψ<sup>e</sup>*

<sup>2</sup> *∂ψ ∂ν e*

<sup>2</sup>*<sup>w</sup>* <sup>+</sup> *<sup>s</sup>λϕ*<sup>Δ</sup>*<sup>ψ</sup> <sup>w</sup>*.

<sup>2</sup>*λ*2*<sup>ϕ</sup>*<sup>2</sup>|∇*ψ*<sup>|</sup>

**Proposition 2.** *There exists s*<sup>0</sup> > 0*, λ*<sup>0</sup> > 0 *and C a positive constant depending on* Ω*, ω, ψ, and T,*

<sup>2</sup> *dxdt* <sup>−</sup> <sup>4</sup>*<sup>s</sup>*

<sup>2</sup>*λ*<sup>3</sup> Σ2

*u dγdt* + 4*sλ*

*dxdt* + *s*

 2

*dxdt* +

 *T* 0 *<sup>ω</sup>* <sup>|</sup>*u*<sup>|</sup>

The proof is similar to the one of Proposition 1, we let it to the reader. We obtain from the

<sup>−</sup>*s<sup>η</sup><sup>g</sup>* <sup>+</sup> *<sup>s</sup>λ*2*<sup>ϕ</sup>*|∇*ψ*<sup>|</sup>

<sup>−</sup>2*s<sup>η</sup>* <sup>|</sup>*u*<sup>|</sup>

<sup>−</sup>2*s<sup>η</sup> <sup>∂</sup><sup>u</sup> ∂t*

<sup>2</sup> *dγdt* + 4*s*

 2

*<sup>w</sup>* (*x*, *<sup>t</sup>*) <sup>=</sup> *<sup>e</sup>*

(*∂<sup>t</sup>* <sup>−</sup> <sup>Δ</sup>)(*es<sup>η</sup><sup>w</sup>*)

*P*

*P*

2*s* <sup>3</sup>*λ*<sup>4</sup> *Q ϕ*3*e*

<sup>2</sup> *∂ψ ∂ν e*

*<sup>ϕ</sup>*<sup>3</sup>|∇*ψ*<sup>|</sup>

≤ *C*

 *Q* 1 *<sup>θ</sup>*<sup>2</sup> <sup>|</sup>*u*<sup>|</sup>

<sup>−</sup>2*s<sup>η</sup>* <sup>+</sup> *<sup>ϕ</sup>*<sup>3</sup>*<sup>e</sup>*

+2*sλ* Σ2 *ϕ∂ψ ∂ν e*

*Q e* <sup>−</sup>2*s<sup>η</sup> ∂u <sup>∂</sup><sup>t</sup>* <sup>−</sup> <sup>Δ</sup>*<sup>u</sup>*

We easily deduce the following result :

<sup>1</sup>*<sup>w</sup>* <sup>=</sup> *<sup>∂</sup><sup>w</sup>*

*<sup>g</sup><sup>s</sup>* <sup>=</sup> *<sup>e</sup>*

*<sup>P</sup>*<sup>2</sup>*<sup>w</sup>* <sup>=</sup> <sup>−</sup>Δ*<sup>w</sup>* <sup>−</sup> *<sup>s</sup>*

*such that for all s* ≥ *s*0*, λ* ≥ *λ*0*, and for any fuction u of (3.1), we have*

<sup>−</sup>2*s<sup>η</sup>*<sup>|</sup>*u*<sup>|</sup>

above Propositions 1 and 2 the following observability inequality :

<sup>2</sup> *dxdt* <sup>≤</sup> *<sup>C</sup>*

**Corollary 1.** *There is a positive constant C* = *C*(Ω, *ω*, *ψ*, *T*) *such that we have*

 *Q ∂u <sup>∂</sup><sup>t</sup>* <sup>−</sup> <sup>Δ</sup>*<sup>u</sup>*

<sup>−</sup>2*s<sup>η</sup> is a bounded weight function.*

*<sup>t</sup>*(*<sup>T</sup>* <sup>−</sup> *<sup>t</sup>*) .

*u* (*x*, *t*). (4.4)

<sup>−</sup>2*s<sup>η</sup>*<sup>|</sup>*u*<sup>|</sup>

<sup>−</sup>2*s<sup>η</sup>*<sup>|</sup>*u*<sup>|</sup>

*∂u ∂ν <sup>d</sup>γdt*

<sup>−</sup>2*s<sup>η</sup>*<sup>∇</sup>*<sup>u</sup>*

<sup>−</sup>2*s<sup>η</sup>* <sup>|</sup>*u*<sup>|</sup>

<sup>−</sup>2*s<sup>η</sup>*|∇*u*<sup>|</sup>

<sup>2</sup> *dxdt*

<sup>2</sup> *dγdt*

<sup>2</sup> *dγdt*

<sup>2</sup> *dγdt*

, (4.8)

(4.7)

<sup>2</sup> *dxdt* .

*<sup>w</sup>* (*x*, 0) <sup>=</sup> *<sup>w</sup>* (*x*, *<sup>T</sup>*) <sup>=</sup> 0. (4.5)

<sup>1</sup>*<sup>w</sup>* <sup>+</sup> *<sup>P</sup>*<sup>2</sup>*<sup>w</sup>* <sup>=</sup> *<sup>g</sup><sup>s</sup>*, (4.6)

2*w*,

So,

$$\begin{split} \frac{1}{2} \int\_{Q} |P\_{1}w|^{2} \, d\mathbf{x}dt + \frac{1}{2} \int\_{Q} |P\_{2}w|^{2} \, d\mathbf{x}dt \\ + 2s\lambda^{2} \int\_{Q} \varrho \, |\nabla w|^{2} \, d\mathbf{x}dt + 2s^{3}\lambda^{4} \int\_{Q} \varrho^{3}e^{-2s\eta} \, |u|^{2} \, d\mathbf{x}dt \\ + 2s^{2}\lambda \int\_{\Sigma} \varrho \, \frac{\partial \eta}{\partial t} \, \frac{\partial \Psi}{\partial \nu} e^{-2s\eta} |u|^{2} \, d\gamma \, d\mathbf{x}dt - 2s^{3}\lambda^{3} \int\_{\Sigma} \varrho^{3} |\nabla \psi|^{2} \frac{\partial \Psi}{\partial \nu} e^{-2s\eta} |u|^{2} \, d\gamma dt \\ - 2 \int\_{\Sigma} \frac{\partial w}{\partial t} \, \frac{\partial w}{\partial \nu} \, d\gamma dt - 4s\lambda \int\_{\Sigma} \varrho \, \nabla \psi \, \nabla w \frac{\partial w}{\partial \nu} \, d\gamma dt \\ + 2s\lambda \int\_{\Sigma} \varrho \frac{\partial \Psi}{\partial \nu} |\nabla w|^{2} \, d\gamma dt - 4s\lambda^{2} \int\_{\Sigma} \varrho |\nabla \psi|^{2} \frac{\partial w}{\partial \nu} \, w \, d\gamma dt \\ \leq \mathsf{C} \left( \int\_{Q} e^{-2s\eta} \, |g|^{2} \, dxdt + s^{3}\lambda^{4} \int\_{0}^{T} \int\_{\omega} \varrho^{3} e^{-2s\eta} |u|^{2} \, dxdt \right). \end{split}$$

Now from

$$\nabla \mu = e^{s\eta} \left( \nabla w - s\lambda \rho \nabla \psi w \right) \mu$$

we deduce

$$\int\_{Q} \oint\_{Q} \varrho \varepsilon^{-2s\eta} |\nabla u|^2 \, dxdt \le \mathbb{C} \left( \int\_{Q} \varrho |\nabla w|^2 \, dxdt + s^2 \lambda^2 \int\_{Q} \varrho^3 |w|^2 \, dxdt \right).$$

We then use the explicit form of *P*1*w* and *P*2*w*, and get

$$\frac{1}{s} \int\_{Q} \frac{1}{\varrho} \left| \frac{\partial w}{\partial t} \right|^2 d\mathbf{x} dt \le \mathbb{C} \left( \int\_{Q} e^{-2s\eta} |\mathbf{g}|^2 d\mathbf{x} dt + \mathbf{s}^3 \lambda^4 \int\_0^T \int\_{\omega} \varrho^3 |w|^2 d\mathbf{x} dt \right) \lambda$$

and

$$\frac{1}{s} \int\_{Q} \frac{1}{q} \left| \Delta w \right|^{2} dxdt \le \mathbb{C} \left( \int\_{Q} e^{-2s\eta} |g|^{2} dxdt + s^{3} \lambda^{4} \int\_{0}^{T} \int\_{\omega} \rho^{3} |w|^{2} dxdt \right).$$

We sum up to finally have

$$\begin{split} 2s^{3}\lambda^{4}\int\_{Q}\varphi^{3}e^{-2s\eta}\left|u\right|^{2}dxdt+4s^{2}\lambda\int\_{\Sigma}\varphi\frac{\partial\eta}{\partial t}\frac{\partial\psi}{\partial t}e^{-2s\eta}\left|u\right|^{2}d\gamma dt \\ -4s^{3}\lambda^{3}\int\_{\Sigma}\varphi^{3}|\nabla\psi|^{2}\frac{\partial\psi}{\partial\nu}e^{-2s\eta}|u|^{2}d\gamma dt-4s^{2}\lambda^{3}\int\_{\Sigma}\varphi^{2}|\nabla\psi|^{2}\frac{\partial\psi}{\partial\nu}e^{-2s\eta}\left|u\right|^{2}d\gamma dt \\ +\int\_{\Sigma}\left(2s\frac{\partial\eta}{\partial t}-4s\lambda^{2}\varphi\vert\nabla\psi\vert^{2}-4s^{2}\lambda^{2}\varphi^{2}\vert\nabla\psi\vert^{2}\right)e^{-2s\eta}\frac{\partial u}{\partial\nu}du \\ -2s\lambda\int\_{\Sigma}\varphi\frac{\partial\psi}{\partial\nu}e^{-2s\eta}\frac{\partial u}{\partial t}u\,d\gamma dt-4s\lambda\int\_{\Sigma}\varphi\nabla\psi e^{-2s\eta}\nabla u\frac{\partial u}{\partial\nu}\,d\gamma dt \\ +2s\lambda\int\_{\Sigma}\varphi\frac{\partial\psi}{\partial\nu}e^{-2s\eta}|\nabla u|^{2}\,d\gamma dt - 2\int\_{\Sigma}e^{-2s\eta}\frac{\partial u}{\partial t}\frac{\partial u}{\partial\nu}\,d\gamma dt \\ \leq \,\,\,\mathsf{C}\left(\int\_{\mathcal{Q}}e^{-2s\eta}\left|\mathcal{S}\right|^{2}d\mathrm{x}dt + s^{3}\lambda^{4}\int\_{0}^{T}\int\_{\omega}g^{3}e^{-2s\eta}\left|u\right|^{2}d\mathrm{x}dt\right). \end{split}$$

Using the fact that *<sup>u</sup>* <sup>=</sup> *<sup>∂</sup><sup>u</sup> ∂t* Σ1 <sup>=</sup> 0, and *<sup>∂</sup><sup>u</sup> ∂ν* Σ2 = 0, we obtain (4.3).

Now we proceed as the following. We define

$$
\tilde{\varphi}(\mathbf{x}, t) = \frac{e^{-\lambda\psi(x)}}{t(T - t)}.
$$

and

10 Will-be-set-by-IN-TECH

1 2 

<sup>2</sup> *dxdt* + *s*

*ϕ* |∇*w*|

*ϕ* |∇*w*|

<sup>2</sup> *<sup>d</sup>γdt* <sup>−</sup> <sup>2</sup>*<sup>s</sup>*

+2*sλ*<sup>2</sup> *Q*

<sup>−</sup>2*s<sup>η</sup>* <sup>|</sup>*u*<sup>|</sup>

+2*sλ* Σ *<sup>ϕ</sup> ∂ψ ∂ν* |∇*w*<sup>|</sup>

≤ *C Q e* <sup>−</sup>2*s<sup>η</sup>* <sup>|</sup>*g*<sup>|</sup>

<sup>−</sup>2*s<sup>η</sup>* |∇*u*<sup>|</sup>

We then use the explicit form of *P*1*w* and *P*2*w*, and get

−2 Σ *∂w ∂t ∂w*

∇*u* = *e*

 *Q e* <sup>−</sup>2*s<sup>η</sup>* <sup>|</sup>*g*<sup>|</sup>

 *Q e* <sup>−</sup>2*s<sup>η</sup>* <sup>|</sup>*g*<sup>|</sup>

<sup>−</sup>2*sη*|*u*<sup>|</sup>

<sup>−</sup>2*s<sup>η</sup>* <sup>|</sup>*u*<sup>|</sup>

<sup>−</sup>2*s<sup>η</sup> <sup>∂</sup><sup>u</sup> ∂t*

> *∂ν* Σ2

*<sup>ϕ</sup>*(*x*, *<sup>t</sup>*) = *<sup>e</sup>*−*λψ*(*x*)

*t*(*T* − *t*)

<sup>2</sup> *<sup>d</sup>γdt* <sup>−</sup> <sup>4</sup>*<sup>s</sup>*

*<sup>∂</sup><sup>t</sup>* <sup>−</sup> <sup>4</sup>*sλ*2*ϕ*|∇*ψ*<sup>|</sup>

 *Q*

<sup>2</sup> *dxdt* <sup>≤</sup> *<sup>C</sup>*

*dxdt* ≤ *C*

<sup>2</sup> *dxdt* <sup>≤</sup> *<sup>C</sup>*

<sup>2</sup> *∂ψ ∂ν e*

+ Σ 2*s ∂η*

≤ *C*

+2*sλ* Σ *<sup>ϕ</sup> ∂ψ ∂ν e*

 *Q e* <sup>−</sup>2*s<sup>η</sup>* <sup>|</sup>*g*<sup>|</sup>

<sup>=</sup> 0, and *<sup>∂</sup><sup>u</sup>*

2*s* <sup>3</sup>*λ*<sup>4</sup> *Q ϕ*3*e*

−2*sλ* Σ *<sup>ϕ</sup> ∂ψ ∂ν e*

*∂t* Σ1

Now we proceed as the following. We define

*<sup>ϕ</sup>*3|∇*ψ*<sup>|</sup>

*<sup>Q</sup>* <sup>|</sup>*P*1*w*<sup>|</sup>

<sup>2</sup> *dxdt* + 2*s*

*∂ν <sup>d</sup>γdt* <sup>−</sup> <sup>4</sup>*s<sup>λ</sup>*

<sup>2</sup> *<sup>d</sup>γdt* <sup>−</sup> <sup>4</sup>*sλ*<sup>2</sup>

<sup>3</sup>*λ*<sup>4</sup> *T* 0 *ω ϕ*3*e*

*<sup>s</sup><sup>η</sup>* (∇*<sup>w</sup>* <sup>−</sup> *<sup>s</sup>λϕ*∇*ψw*),

<sup>2</sup> *dxdt* + *s*

<sup>2</sup>*dxdt* + *s*

<sup>2</sup>*dxdt* + *s*

<sup>2</sup> *dxdt* + 4*s*

<sup>2</sup>*λ*<sup>3</sup> Σ

*u dγdt* − 4*sλ*

<sup>−</sup>2*s<sup>η</sup>* |∇*u*<sup>|</sup>

<sup>2</sup> *dxdt* + *s*

<sup>2</sup>*λ*<sup>2</sup> *Q <sup>ϕ</sup>*<sup>3</sup> <sup>|</sup>*w*<sup>|</sup>

<sup>3</sup>*λ*<sup>4</sup> *T* 0 *ω <sup>ϕ</sup>*3|*w*<sup>|</sup>

<sup>3</sup>*λ*<sup>4</sup> *T* 0 *ω <sup>ϕ</sup>*3|*w*<sup>|</sup>

2*λ* Σ *<sup>ϕ</sup> ∂η ∂t ∂ψ ∂ν e*

<sup>2</sup> <sup>−</sup> <sup>4</sup>*<sup>s</sup>*

<sup>3</sup>*λ*<sup>4</sup> *T* 0 *ω ϕ*3*e*

= 0, we obtain (4.3).

,

 Σ *ϕ*∇*ψe*

<sup>2</sup> *<sup>d</sup>γdt* <sup>−</sup> <sup>2</sup>

*<sup>ϕ</sup>*2|∇*ψ*<sup>|</sup>

<sup>2</sup> *∂ψ ∂ν e*

<sup>2</sup>*λ*2*ϕ*2|∇*ψ*<sup>|</sup>

 Σ *e* <sup>−</sup>2*s<sup>η</sup> <sup>∂</sup><sup>u</sup> ∂t ∂u ∂ν <sup>d</sup>γdt*

<sup>3</sup>*λ*<sup>3</sup> Σ <sup>2</sup> *dxdt* +

<sup>3</sup>*λ*<sup>4</sup> *Q ϕ*3*e*

*<sup>ϕ</sup>*3|∇*ψ*<sup>|</sup>

 Σ

 Σ *ϕ*|∇*ψ*|

1 2 

<sup>2</sup> *∂ψ ∂ν e*

*ϕ*∇*ψ*.∇*w*

<sup>−</sup>2*s<sup>η</sup>* <sup>|</sup>*u*<sup>|</sup>

*<sup>Q</sup>* <sup>|</sup>*P*2*w*<sup>|</sup>

<sup>−</sup>2*s<sup>η</sup>* <sup>|</sup>*u*<sup>|</sup>

<sup>−</sup>2*s<sup>η</sup>* <sup>|</sup>*u*<sup>|</sup>

<sup>2</sup> *∂w*

*∂w ∂ν <sup>d</sup>γdt*

*∂ν* .*w dγdt*

<sup>2</sup> *dxdt* .

<sup>2</sup> *dxdt* .

> <sup>2</sup>*dxdt* ,

<sup>2</sup>*dxdt* .

<sup>−</sup>2*sη*|*u*<sup>|</sup>

<sup>−</sup>2*sη*|*u*<sup>|</sup>

*∂u ∂ν <sup>d</sup>γdt*

<sup>2</sup> *dxdt* .

2 *e* <sup>−</sup>2*s<sup>η</sup> <sup>∂</sup><sup>u</sup> ∂ν <sup>u</sup>*

<sup>−</sup>2*sη*∇*<sup>u</sup>*

<sup>−</sup>2*s<sup>η</sup>* <sup>|</sup>*u*<sup>|</sup>

<sup>2</sup> *dγdt*

<sup>2</sup> *dγdt*

<sup>2</sup> *dxdt*

<sup>2</sup> *dxdt*

<sup>2</sup> *dγdt*

So,

Now from

we deduce

and

+2*s* 2*λ* Σ *<sup>ϕ</sup> ∂η ∂t ∂ψ ∂ν e*

 *Q ϕe*

1 *s Q* 1 *ϕ ∂w ∂t* 2

1 *s Q* 1 *<sup>ϕ</sup>* <sup>|</sup>Δ*w*<sup>|</sup>

We sum up to finally have

−4*s* <sup>3</sup>*λ*<sup>3</sup> Σ

Using the fact that *<sup>u</sup>* <sup>=</sup> *<sup>∂</sup><sup>u</sup>*

$$
\tilde{\eta}(x,t) = \frac{e^{2\lambda |\psi|\_{\infty}} - e^{-\lambda \psi(x)}}{t(T-t)}.
$$

Then

$$
\nabla \tilde{\varphi} = -\lambda \tilde{\varphi} \nabla \psi \,\,\, \nabla \tilde{\eta} = \lambda \tilde{\varphi} \nabla \psi \,\, \,.
$$

we still have the properties (4.1) and (4.2).

For *s* ≥ *s*<sup>0</sup> and *λ* ≥ *λ*0, we define

$$
\tilde{w}\left(\mathbf{x},t\right) = e^{-s\eta\left(\mathbf{x},t\right)}u\left(\mathbf{x},t\right).\tag{4.4}
$$

We easily notice that

$$
\tilde{w}\left(\mathbf{x},\mathbf{0}\right) = \tilde{w}\left(\mathbf{x},T\right) = \mathbf{0}.\tag{4.5}
$$

Calculating *<sup>P</sup><sup>w</sup>* <sup>=</sup> *<sup>e</sup>*−*s<sup>η</sup><sup>g</sup>* <sup>=</sup> *<sup>e</sup>*−*s<sup>η</sup>* (*∂<sup>t</sup>* <sup>−</sup> <sup>Δ</sup>)(*es<sup>η</sup><sup>w</sup>*) , we set

$$
\bar{P}\_1 \tilde{w} + \bar{P}\_2 \tilde{w} = \tilde{\mathcal{g}}\_{\text{s}} \tag{4.6}
$$

where

$$
\begin{split}
\tilde{P}\_1 \tilde{w} &= \frac{\partial \tilde{w}}{\partial t} - 2s\lambda \tilde{\varphi} \, \nabla \psi \, \nabla \tilde{w} + 2s\lambda^2 \tilde{\varphi} |\nabla \psi|^2 \tilde{w}, \\
\tilde{P}\_2 \tilde{w} &= -\Delta \tilde{w} - s^2 \lambda^2 \tilde{\varphi}^2 |\nabla \psi|^2 \tilde{w} + s \frac{\partial \tilde{\eta}}{\partial t} \tilde{w}, \\
\tilde{g}\_s &= e^{-s\tilde{\eta}} g + s \lambda^2 \tilde{\varphi} |\nabla \psi|^2 \tilde{w} + s \lambda \tilde{\varphi} \, \Delta \psi \, \tilde{w}.
\end{split}
$$

We easily deduce the following result :

**Proposition 2.** *There exists s*<sup>0</sup> > 0*, λ*<sup>0</sup> > 0 *and C a positive constant depending on* Ω*, ω, ψ, and T, such that for all s* ≥ *s*0*, λ* ≥ *λ*0*, and for any fuction u of (3.1), we have*

$$\begin{split} 2\mathbf{s}^{3}\lambda^{4}\int\_{Q}\tilde{\varphi}^{3}e^{-2s\tilde{\eta}}|u|^{2}\,dxdt - 4\mathbf{s}^{2}\lambda\int\_{\Sigma\_{2}}\tilde{\phi}\frac{\partial\tilde{\eta}}{\partial t}\frac{\partial\tilde{\eta}}{\partial\nu}e^{-2s\tilde{\eta}}|u|^{2}\,d\gamma dt \\ + 4\mathbf{s}^{3}\lambda^{3}\int\_{\Sigma\_{2}}\tilde{\varphi}^{3}|\nabla\psi|^{2}\frac{\partial\tilde{\eta}}{\partial\nu}e^{-2s\tilde{\eta}}|u|^{2}\,d\gamma dt + 4\mathbf{s}^{2}\lambda^{3}\int\_{\Sigma\_{2}}\tilde{\phi}^{2}|\nabla\psi|^{2}\frac{\partial\tilde{\eta}}{\partial\nu}e^{-2s\tilde{\eta}}|u|^{2}\,d\gamma dt \\ + 2s\lambda\int\_{\Sigma\_{2}}\tilde{\phi}\frac{\partial\Psi}{\partial\nu}e^{-2s\tilde{\eta}}\frac{\partial u}{\partial t}\,u\,d\gamma dt + 4s\lambda\int\_{\Sigma\_{1}}\tilde{\phi}\nabla\Psi e^{-2s\tilde{\eta}}|\nabla u\,\frac{\partial u}{\partial \nu}\,d\gamma dt \\ - 2s\lambda\int\_{\Sigma}\tilde{\varphi}\frac{\partial\Psi}{\partial\nu}e^{-2s\tilde{\eta}}|\nabla u|^{2}\,d\gamma dt \\ \leq \mathsf{C}\left(\int\_{Q}e^{-2s\tilde{\eta}}\left|\frac{\partial u}{\partial t}-\Delta u\right|^{2}\,dxdt + s^{3}\lambda^{4}\int\_{0}^{T}\int\_{\omega}\tilde{\phi}^{3}e^{-2s\eta}|u|^{2}\,d\mathbf{x}dt\right). \end{split} \tag{4.7}$$

The proof is similar to the one of Proposition 1, we let it to the reader. We obtain from the above Propositions 1 and 2 the following observability inequality :

**Corollary 1.** *There is a positive constant C* = *C*(Ω, *ω*, *ψ*, *T*) *such that we have*

$$\int\_{Q} \frac{1}{\theta^2} \left| \boldsymbol{u} \right|^2 d\mathbf{x} dt \le \mathbb{C} \left[ \int\_{Q} \left| \frac{\partial \boldsymbol{u}}{\partial t} - \Delta \boldsymbol{u} \right|^2 d\mathbf{x} dt + \int\_{0}^{T} \int\_{\omega} \left| \boldsymbol{u} \right|^2 d\mathbf{x} dt \right],\tag{4.8}$$

*where* <sup>1</sup> *<sup>θ</sup>*<sup>2</sup> <sup>=</sup> *<sup>ϕ</sup>*3*<sup>e</sup>* <sup>−</sup>2*s<sup>η</sup>* <sup>+</sup> *<sup>ϕ</sup>*<sup>3</sup>*<sup>e</sup>* <sup>−</sup>2*s<sup>η</sup> is a bounded weight function.*

where �*y*ˆ*o*� ≤ 1, and where *<sup>τ</sup>y*ˆ*<sup>o</sup>* is the missing data. To the boundary condition *<sup>ξ</sup>* is added

Some Aspects of the Sentinel Method for Pollution Problems 197

Still, one can use the least square method : In the context of the above problem, it consists in

is then *y*(*T*, *x*; *v*, *w*), and we want this solution to be *as close as possible* to some measurement *mo*. We consider then the distance *J*(*v*, *w*) = �*y*(*v*, *w*) − *mo*� in an appropriate norm and we

where *v* and *w* are arbitrary. As we know, with this method, we can not seperate *v* and *w*. The more suitable is then the Sentinel method (see Lions J.-L. (1992), Miloudi Y. et al. (2009),

*J*(*v*, *w*)

inf {*v*,*w*}

where *ω* ⊂ Ω the open set of controls. We search for w, such that the mean function

*hoχ*<sup>O</sup> + <sup>w</sup>*χω*

*∂*S

The integral (5.7) is called *sentinel*, following the definition of Lions Lions J.-L. (1992).

*ho y*(*T*, *x*; *λ*, *τ*) *dx* +

(here *χ<sup>O</sup>* and *χω* denote the characteristic functions of O and *ω* respectively), is insensitive to

*∂τ* (0, 0) = 0,

The pollution sources in (5.3) can be considered as functionals whose position and nature are known, but their amplitudes are however unknown. The goal is to estimate these amplitudes;

 *ω*

Σ<sup>0</sup> = Γ<sup>0</sup> ⊂ Σ of the total boundary Σ =]0, *T*[×Γ, as we have :

The question is to obtain information on the pollution *λ* ˆ

Kernevez J.-P. (1997) and the references therein).

<sup>S</sup>(*λ*, *<sup>τ</sup>*) =

= Ω 

O

the other missing terms (in the initial data (5.2)) do not interest us.

The state solution *y*(*λ*, *τ*) := *y*(*t*, *x*; *λ* ˆ

the missing data at the first order; i.e.

for *λ* = *τ* = 0, and such that

Ω, but here we observe *y* at fixed *instant T* :

*y* =

*ξ* + *λ* ˆ

*ξ* which is unknown, and which *we want* to identify. It appears on a part

0 on <sup>Σ</sup>\Σ0. (5.3)

*<sup>ξ</sup>*, *<sup>τ</sup>y*ˆ*o*} <sup>=</sup> {*v*, *<sup>w</sup>*} as two control variables. At time *<sup>T</sup>*, the state

*<sup>ξ</sup>*, *<sup>τ</sup>y*ˆ*o*). The observation is given in an observatory O ⊂

*y*obs = *mo*. (5.4)

*ho* <sup>∈</sup> *<sup>L</sup>*2(O) (5.5)

<sup>w</sup> <sup>∈</sup> *<sup>L</sup>*2(*ω*) (5.6)

*y*(*T*, *x*; *λ*, *τ*) *dx* (5.7)

w *y*(*T*, *x*; *λ*, *τ*) *dx*

�w�*L*<sup>2</sup>(*ω*) is of minimal norm. (5.8)

*ξ*, not affected by the missing term

*ξ* on Σ0,

a pollution term *λ* ˆ

*τy*ˆ*<sup>o</sup>* of the initial data.

search for

Now, we are given

and

considering the unknowns {*<sup>λ</sup>* <sup>ˆ</sup>

Summing the terms in (4.3) and (4.7) we get the following :

2*s* <sup>3</sup>*λ*<sup>4</sup> *Q ϕ*3*e* <sup>−</sup>2*s<sup>η</sup>* <sup>+</sup> *<sup>ϕ</sup>*<sup>3</sup>*<sup>e</sup>* <sup>−</sup>2*s<sup>η</sup>* |*u*| <sup>2</sup> *dxdt* −4*s* 2*λ* Σ2 *<sup>ϕ</sup> ∂η ∂t e* <sup>−</sup>2*s<sup>η</sup>* <sup>−</sup> *<sup>ϕ</sup>∂η ∂t e* <sup>−</sup>2*s<sup>η</sup> ∂ψ ∂ν* <sup>|</sup>*u*<sup>|</sup> <sup>2</sup> *dγdt* <sup>−</sup>4*s*3*λ*<sup>3</sup> Σ2 *ϕ*3*e* <sup>−</sup>2*s<sup>η</sup>* <sup>−</sup> *<sup>ϕ</sup>*<sup>3</sup>*<sup>e</sup>* <sup>−</sup>2*s<sup>η</sup>* |∇*ψ*| <sup>2</sup> *∂ψ ∂ν* <sup>|</sup>*u*<sup>|</sup> <sup>2</sup> *dγdt* <sup>−</sup>4*s*2*λ*<sup>3</sup> Σ2 *ϕ*2*e* <sup>−</sup>2*s<sup>η</sup>* <sup>−</sup> *<sup>ϕ</sup>*<sup>2</sup>*<sup>e</sup>* <sup>−</sup>2*s<sup>η</sup>* |∇*ψ*| <sup>2</sup> *∂ψ ∂ν* <sup>|</sup>*u*<sup>|</sup> <sup>2</sup> *dγdt* +2*sλ* Σ2 *ϕe* <sup>−</sup>2*s<sup>η</sup>* <sup>−</sup> *<sup>ϕ</sup><sup>e</sup>* <sup>−</sup>2*s<sup>η</sup> ∂ψ ∂ν ∂u ∂t u dγdt* −4*sλ* Σ1 *ϕe* <sup>−</sup>2*s<sup>η</sup>* <sup>−</sup> *<sup>ϕ</sup><sup>e</sup>* <sup>−</sup>2*s<sup>η</sup>* ∇*ψ*∇*u ∂u ∂ν <sup>d</sup>γdt* +2*sλ* Σ *ϕe* <sup>−</sup>2*s<sup>η</sup>* <sup>−</sup> *<sup>ϕ</sup><sup>e</sup>* <sup>−</sup>2*s<sup>η</sup> ∂ψ ∂ν* |∇*u*<sup>|</sup> <sup>2</sup> *dγdt* ≤ *C Q e* <sup>−</sup>2*s<sup>η</sup>* + *e* <sup>−</sup>2*s<sup>η</sup> ∂u <sup>∂</sup><sup>t</sup>* <sup>−</sup> <sup>Δ</sup>*<sup>u</sup>* 2 *dxdt* +*s*3*λ*<sup>4</sup> *T* 0 *ω ϕ*3*e* <sup>−</sup>2*s<sup>η</sup>* <sup>+</sup> *<sup>ϕ</sup>*<sup>3</sup>*<sup>e</sup>* <sup>−</sup>2*s<sup>η</sup>* |*u*| <sup>2</sup> *dxdt* .

Now, it suffices to notice that *<sup>ϕ</sup>* <sup>=</sup> *<sup>ϕ</sup>* and *<sup>η</sup>* <sup>=</sup> *<sup>η</sup>* on <sup>Σ</sup>.

#### **5. Instantaneous sentinels**

In this part, we discuss of how to get instantaneous information (at fixed *t* = *T* ∈ [0, +∞[) on pollution terms in systems of incomplete data in ecology and/or meteorology problems.

Here, the ecological system is affected by pollution in the boundary of the domain (a border of an air pollution cloud for example). We verify that if the initial data is completely unknown, the sentinel is nul. If there is some information on the initial data, the *instantaneous* sentinel naturally exists if the control set is bigger than the one where the observation may be defined.

Finally, we give the characterization of the instantaneous sentinel with some remarks.

Using the techniques of the sections below and those in Miloudi Y. et al. (2009). We prove the existence and characterization of a sentinel, which permits to identify the pollution parameters at fixed time *T*. The problem we consider now is the following :

$$\frac{\partial y}{\partial t} + Ay + f(y) = 0 \qquad \text{in } Q := ]0,\\ T[ \times \Omega,\tag{5.1}$$

where <sup>Ω</sup> <sup>⊂</sup> **<sup>R</sup>***<sup>d</sup>* is an open domain of regular boundary <sup>Γ</sup> <sup>=</sup> *<sup>∂</sup>*<sup>Ω</sup> for instance, *<sup>A</sup>* represents a second order elliptic operator such that *akl* <sup>=</sup> *alk*, *akl* ∈ C2(Ω¯ ), and <sup>∑</sup>*<sup>d</sup> <sup>i</sup>*,*j*=<sup>1</sup> *aijηiη<sup>j</sup>* <sup>≥</sup> *<sup>c</sup>*2|*η*<sup>|</sup> 2, and *<sup>f</sup>* : **<sup>R</sup>** <sup>→</sup> **<sup>R</sup>** is a (nonlinear) <sup>C</sup><sup>1</sup> function.

To (5.1) we should add initial and boundary conditions. It is in these conditions that we can meet incomplete data. We assume that we know that *<sup>y</sup>*|*t*=<sup>0</sup> is in a ball of *<sup>L</sup>*2(Ω) of center *<sup>y</sup>o*, say �*y*(0, .) <sup>−</sup> *<sup>y</sup>o*(.)�*L*<sup>2</sup>(Ω) <sup>≤</sup> *<sup>τ</sup>*, that we should write :

$$y(0) = y^\rho + \tau \mathfrak{H}^\rho,\tag{5.2}$$

where �*y*ˆ*o*� ≤ 1, and where *<sup>τ</sup>y*ˆ*<sup>o</sup>* is the missing data. To the boundary condition *<sup>ξ</sup>* is added a pollution term *λ* ˆ *ξ* which is unknown, and which *we want* to identify. It appears on a part Σ<sup>0</sup> = Γ<sup>0</sup> ⊂ Σ of the total boundary Σ =]0, *T*[×Γ, as we have :

$$y = \begin{cases} \tilde{\xi} + \lambda \, \hat{\xi} & \text{on } \Sigma\_{0\prime} \\ 0 & \text{on } \Sigma \backslash \Sigma\_0. \end{cases} \tag{5.3}$$

The question is to obtain information on the pollution *λ* ˆ *ξ*, not affected by the missing term *τy*ˆ*<sup>o</sup>* of the initial data.

Still, one can use the least square method : In the context of the above problem, it consists in considering the unknowns {*<sup>λ</sup>* <sup>ˆ</sup> *<sup>ξ</sup>*, *<sup>τ</sup>y*ˆ*o*} <sup>=</sup> {*v*, *<sup>w</sup>*} as two control variables. At time *<sup>T</sup>*, the state is then *y*(*T*, *x*; *v*, *w*), and we want this solution to be *as close as possible* to some measurement *mo*. We consider then the distance *J*(*v*, *w*) = �*y*(*v*, *w*) − *mo*� in an appropriate norm and we search for

$$\inf\_{\{v,w\}} J(v,w)$$

where *v* and *w* are arbitrary. As we know, with this method, we can not seperate *v* and *w*. The more suitable is then the Sentinel method (see Lions J.-L. (1992), Miloudi Y. et al. (2009), Kernevez J.-P. (1997) and the references therein).

The state solution *y*(*λ*, *τ*) := *y*(*t*, *x*; *λ* ˆ *<sup>ξ</sup>*, *<sup>τ</sup>y*ˆ*o*). The observation is given in an observatory O ⊂ Ω, but here we observe *y* at fixed *instant T* :

$$
\mathcal{Y}\_{\text{obs}} = m\_0. \tag{5.4}
$$

Now, we are given

$$h\_{\theta} \in L^{2}(\mathcal{O})\tag{5.5}$$

and

12 Will-be-set-by-IN-TECH

<sup>−</sup>2*s<sup>η</sup>* <sup>+</sup> *<sup>ϕ</sup>*<sup>3</sup>*<sup>e</sup>*

<sup>−</sup>2*s<sup>η</sup>* <sup>−</sup> *<sup>ϕ</sup>∂η ∂t e* <sup>−</sup>2*s<sup>η</sup>*

> <sup>−</sup>2*s<sup>η</sup>* |∇*ψ*|

> <sup>−</sup>2*s<sup>η</sup>* |∇*ψ*|

<sup>−</sup>2*s<sup>η</sup> ∂ψ ∂ν ∂u ∂t*

<sup>−</sup>2*s<sup>η</sup>* 

<sup>−</sup>2*s<sup>η</sup> ∂ψ ∂ν* |∇*u*<sup>|</sup>

<sup>−</sup>2*s<sup>η</sup> ∂u <sup>∂</sup><sup>t</sup>* <sup>−</sup> <sup>Δ</sup>*<sup>u</sup>*

<sup>−</sup>2*s<sup>η</sup>* <sup>+</sup> *<sup>ϕ</sup>*<sup>3</sup>*<sup>e</sup>*

In this part, we discuss of how to get instantaneous information (at fixed *t* = *T* ∈ [0, +∞[) on pollution terms in systems of incomplete data in ecology and/or meteorology problems.

Here, the ecological system is affected by pollution in the boundary of the domain (a border of an air pollution cloud for example). We verify that if the initial data is completely unknown, the sentinel is nul. If there is some information on the initial data, the *instantaneous* sentinel naturally exists if the control set is bigger than the one where the observation may be defined.

Using the techniques of the sections below and those in Miloudi Y. et al. (2009). We prove the existence and characterization of a sentinel, which permits to identify the pollution

where <sup>Ω</sup> <sup>⊂</sup> **<sup>R</sup>***<sup>d</sup>* is an open domain of regular boundary <sup>Γ</sup> <sup>=</sup> *<sup>∂</sup>*<sup>Ω</sup> for instance, *<sup>A</sup>* represents a

To (5.1) we should add initial and boundary conditions. It is in these conditions that we can meet incomplete data. We assume that we know that *<sup>y</sup>*|*t*=<sup>0</sup> is in a ball of *<sup>L</sup>*2(Ω) of center *<sup>y</sup>o*,

*y*(0) = *y<sup>o</sup>* + *τy*ˆ

Finally, we give the characterization of the instantaneous sentinel with some remarks.

parameters at fixed time *T*. The problem we consider now is the following :

second order elliptic operator such that *akl* <sup>=</sup> *alk*, *akl* ∈ C2(Ω¯ ), and <sup>∑</sup>*<sup>d</sup>*

<sup>−</sup>2*s<sup>η</sup>* <sup>−</sup> *<sup>ϕ</sup>*<sup>3</sup>*<sup>e</sup>*

<sup>−</sup>2*s<sup>η</sup>* <sup>−</sup> *<sup>ϕ</sup>*<sup>2</sup>*<sup>e</sup>*

<sup>−</sup>2*s<sup>η</sup>* <sup>−</sup> *<sup>ϕ</sup><sup>e</sup>*

<sup>−</sup>2*s<sup>η</sup>* <sup>−</sup> *<sup>ϕ</sup><sup>e</sup>*

<sup>−</sup>2*s<sup>η</sup>* <sup>−</sup> *<sup>ϕ</sup><sup>e</sup>*

<sup>−</sup>2*s<sup>η</sup>* |*u*| <sup>2</sup> *dxdt*

> *∂ψ ∂ν* <sup>|</sup>*u*<sup>|</sup>

> > <sup>2</sup> *∂ψ ∂ν* <sup>|</sup>*u*<sup>|</sup>

> > <sup>2</sup> *∂ψ ∂ν* <sup>|</sup>*u*<sup>|</sup>

∇*ψ*∇*u*

<sup>−</sup>2*s<sup>η</sup>* |*u*| <sup>2</sup> *dxdt* .

*u dγdt*

<sup>2</sup> *dγdt*

*∂u ∂ν <sup>d</sup>γdt*

 2 *dxdt*

*<sup>∂</sup><sup>t</sup>* <sup>+</sup> *Ay* <sup>+</sup> *<sup>f</sup>*(*y*) = 0 in *<sup>Q</sup>* :=]0, *<sup>T</sup>*[×Ω, (5.1)

*<sup>i</sup>*,*j*=<sup>1</sup> *aijηiη<sup>j</sup>* <sup>≥</sup> *<sup>c</sup>*2|*η*<sup>|</sup>

*<sup>o</sup>*, (5.2)

2, and

<sup>2</sup> *dγdt*

<sup>2</sup> *dγdt*

<sup>2</sup> *dγdt*

Summing the terms in (4.3) and (4.7) we get the following :

2*s* <sup>3</sup>*λ*<sup>4</sup> *Q ϕ*3*e*

> Σ2 *ϕ*3*e*

> Σ2 *ϕ*2*e*

+2*sλ* Σ2 *ϕe*

−4*sλ* Σ1 *ϕe*

+2*sλ* Σ *ϕe*

≤ *C Q e* <sup>−</sup>2*s<sup>η</sup>* + *e*

+*s*3*λ*<sup>4</sup>

Now, it suffices to notice that *<sup>ϕ</sup>* <sup>=</sup> *<sup>ϕ</sup>* and *<sup>η</sup>* <sup>=</sup> *<sup>η</sup>* on <sup>Σ</sup>.

*∂y*

say �*y*(0, .) <sup>−</sup> *<sup>y</sup>o*(.)�*L*<sup>2</sup>(Ω) <sup>≤</sup> *<sup>τ</sup>*, that we should write :

*<sup>f</sup>* : **<sup>R</sup>** <sup>→</sup> **<sup>R</sup>** is a (nonlinear) <sup>C</sup><sup>1</sup> function.

**5. Instantaneous sentinels**

 *T* 0 *ω ϕ*3*e*

−4*s* 2*λ* Σ2 *<sup>ϕ</sup> ∂η ∂t e*

<sup>−</sup>4*s*3*λ*<sup>3</sup>

<sup>−</sup>4*s*2*λ*<sup>3</sup>

$$\mathbf{w} \in L^2(\omega) \tag{5.6}$$

where *ω* ⊂ Ω the open set of controls. We search for w, such that the mean function

$$\mathcal{S}(\lambda, \tau) = \int\_{\mathcal{O}} h\_{\mathcal{O}} \, y(T, x; \lambda, \tau) \, dx + \int\_{\omega} \mathbf{w} \, y(T, x; \lambda, \tau) \, dx$$

$$= \int\_{\Omega} \left( h\_{\mathcal{O}} \chi\_{\mathcal{O}} + \mathbf{w} \chi\_{\omega} \right) y(T, x; \lambda, \tau) \, dx \tag{5.7}$$

(here *χ<sup>O</sup>* and *χω* denote the characteristic functions of O and *ω* respectively), is insensitive to the missing data at the first order; i.e.

$$\frac{\partial \mathcal{S}}{\partial \tau}(0,0) = 0.$$

for *λ* = *τ* = 0, and such that

$$\|\|\mathbf{w}\|\|\_{L^2(\omega)}\quad\text{is of minimal norm.}\tag{5.8}$$

The integral (5.7) is called *sentinel*, following the definition of Lions Lions J.-L. (1992). The pollution sources in (5.3) can be considered as functionals whose position and nature are known, but their amplitudes are however unknown. The goal is to estimate these amplitudes; the other missing terms (in the initial data (5.2)) do not interest us.

**Proof -** Multiplying (6.1) by *y<sup>τ</sup>* and integrating by parts we obtain :

*q*(0)*yτ*(0) *dx* −

*hoχ*<sup>O</sup> + <sup>w</sup>*χω*

 *Q*

 Σ *∂q ∂ν <sup>y</sup><sup>τ</sup> <sup>d</sup><sup>σ</sup>* <sup>+</sup>

*q* L*y<sup>τ</sup> dxdt* −

Some Aspects of the Sentinel Method for Pollution Problems 199

*yτ*(*T*) *dx* +

*<sup>o</sup> dx* = 0 for every *y*ˆ

**Corollary 2.** *If there is no information on the missing data, then the instantaneous sentinel is nul.*

**Proof -** The proof is easy and lies on the backward uniqueness property (see Lions-Malgrange Lions J.-L. & Malgrange B. (1960)) : If an instantaneous sentinel exists, then from the above

*q* ≡ 0 in *Q*. Thus, in particular for *<sup>t</sup>* = *<sup>T</sup>*, we have *<sup>q</sup>*(*T*) = 0. So that *hoχ*<sup>O</sup> + <sup>w</sup>*χω* = 0 in <sup>Ω</sup>, and hence -as

We briefly show how (5) and (5.8) are sufficient to get information on the pollution term *λ* ˆ

*y<sup>λ</sup>* = (*∂y*/*∂λ*)(0, 0)

**Remark 6.** *From the previous corollary, we see that in order to get a non nul sentinel, we need more information (on the structure) of the missing initial data. We will need here the assumption (7) below.*

*ξ*. Indeed, *y<sup>λ</sup>* is solution to the well-posed problem <sup>L</sup> *<sup>y</sup><sup>λ</sup>* <sup>=</sup> 0 in *<sup>Q</sup>*, *<sup>y</sup>λ*(0) = 0 in <sup>Ω</sup>, and *<sup>y</sup><sup>λ</sup>* <sup>=</sup> <sup>ˆ</sup>

*∂λ* (0, 0), for *<sup>λ</sup>*, *<sup>τ</sup>* small.

*ξ* and other known data. Consequently, the estimates (6.6) contain the

(*h*0*χ<sup>O</sup>* + w*χω*)(*mo* − *y*0) *dxdt*.

(*h*0*χ<sup>O</sup>* + w*χω*)(*mo* − *y*0) *dxdt*, (6.6)

*ξ* on Σ. (6.7)

 *Q*  Ω

 Σ *∂y<sup>τ</sup>*

 Ω *q*(0) *y*ˆ

*q*(*T*)*yτ*(*T*) *dx*

*<sup>o</sup> dx* = 0.

*<sup>o</sup>* <sup>∈</sup> *<sup>L</sup>*2(Ω),

*∂ν q d<sup>σ</sup>* <sup>=</sup> 0. (6.5)

*ξ*.

(L∗*q*) *y<sup>τ</sup> dxdt* =

But, *y<sup>τ</sup>* is solution to (6), and *q* verifies (6.2) and (6.3). Hence

and consequently *q*(0) = 0 in Ω. The converse is obvious.

Lemma we have (6.1), (6.3) and (6.4). We deduce that

*<sup>S</sup>*(*λ*, *<sup>τ</sup>*) � *<sup>S</sup>*(0, 0) + *<sup>λ</sup> <sup>∂</sup><sup>S</sup>*

(*h*0*χ<sup>O</sup>* + w*χω*)*y<sup>λ</sup> dxdt* �

 *Q*

in Lions Lions J.-L. (1992)- the sentinel is nul.

6.0.0.1 Information given by the sentinel

We write

Using (5.4), we have

where the derivative

only depends on ˆ

information on *λ* ˆ

*<sup>λ</sup> <sup>∂</sup><sup>S</sup>*

*∂λ* (0, 0) =

*Q*

Multiplying by *q* and integrating by parts, we obtain *Q λ* ˆ *ξ dxdt* =

 *Q*

> + Ω

− Ω 

If a sentinel exists, then we have (5). So it remains : Ω *q*(0) *y*ˆ

**Remark 4.** *In Lions Lions J.-L. (1992), ω* = O*, so there exists always a sentinel (at least the one where* w = −*ho). Then the problem is only to find* w *solution to (5.8). At last, we have to be sure that* w �= −*ho.*

Below, we use the sentinel method as formulated in Miloudi Y. et al. (2007) and Miloudi Y. et al. (2009) and Nakoulima O. (2004), where the observation and the control have their support in two different open sets. Indeed, one can observe somewhere and control elsewhere !

#### **6. Instantaneous sentinel. Case of no information on the missing data**

In this section, we extend the method of sentinels to the case of observation and control having their supports in two different open sets. Moreover, we want information at precise time *T* which is a difficult problem.

Denote by *y*(*t*, *x*; *λ*, *τ*) := *y*(*λ*, *τ*), the state solution of (5.1)-(5.3). We begin by noticing that the solution *y* = *y*(*λ*, *τ*) satisfies the system :

$$\begin{cases} \mathcal{L}y\_\tau = 0 \quad \text{in } Q\_\prime\\ y\_\tau(0) = \mathcal{Y}^\flat \text{ in } \Omega\_\prime\\ y\_\tau = 0 \quad \text{on } \Sigma\_\prime \end{cases}$$

where *y<sup>τ</sup>* denotes the derivative

$$y\_{\tau} = \frac{\partial y}{\partial \tau}(0,0)$$

and where L is the linear differential operator defined by

$$\mathcal{L} = \frac{\partial}{\partial t} + A + f'(y\_o) \, I\_d$$

with *yo* = *y*(*t*, *x*; 0, 0) for *λ* = *τ* = 0. Then

$$\frac{\partial \mathcal{S}}{\partial \tau}(0,0) = \int\_{\Omega} \left( h\_{\boldsymbol{\sigma}} \chi\_{\mathcal{O}} + \mathsf{w} \chi\_{\boldsymbol{\omega}} \right) \boldsymbol{y}\_{\tau}(T) \, d\boldsymbol{x}\_{\tau}$$

**Remark 5.** *From above, the insensitive criterion (5) to the missing term τy*ˆ*<sup>o</sup> is given by :*

$$\int\_{\Omega} \left( h\_{\boldsymbol{\alpha}} \chi\_{\mathcal{O}} + \mathsf{w} \chi\_{\boldsymbol{\omega}} \right) y\_{\boldsymbol{\pi}}(T) \, d\boldsymbol{x} = 0.$$

**Lemma 2.** *Let q be the solution to the following well-posed backward problem :*

$$\mathcal{L}^\* q = 0 \qquad\qquad\text{im}\quad Q\_\prime \tag{6.1}$$

$$q(T) = h\_o \chi\_{\mathcal{O}} + \mathbf{w} \chi\_{\omega} \quad \text{in} \quad \Omega\_{\prime} \tag{6.2}$$

$$
\mathfrak{q} = \mathbf{0} \qquad\qquad\text{or}\quad \Sigma\_{\prime} \tag{6.3}
$$

*where* <sup>L</sup><sup>∗</sup> <sup>=</sup> <sup>−</sup> *<sup>∂</sup> <sup>∂</sup><sup>t</sup>* + *A*<sup>∗</sup> + *f* � (*yo*) *Id is the adjoint operator. Then, the existence of an instantaneous sentinel insensitive to the missing data (i.e. such that (5) holds), is equivalent to the null-controllability problem (6.1)-(6.3) with*

$$q(0) = 0 \qquad \qquad \text{in} \quad \Omega. \tag{6.4}$$

14 Will-be-set-by-IN-TECH

**Remark 4.** *In Lions Lions J.-L. (1992), ω* = O*, so there exists always a sentinel (at least the one where* w = −*ho). Then the problem is only to find* w *solution to (5.8). At last, we have to be sure that*

Below, we use the sentinel method as formulated in Miloudi Y. et al. (2007) and Miloudi Y. et al. (2009) and Nakoulima O. (2004), where the observation and the control have their support in two different open sets. Indeed, one can observe somewhere and control elsewhere !

In this section, we extend the method of sentinels to the case of observation and control having their supports in two different open sets. Moreover, we want information at precise time *T*

Denote by *y*(*t*, *x*; *λ*, *τ*) := *y*(*λ*, *τ*), the state solution of (5.1)-(5.3). We begin by noticing that

L *y<sup>τ</sup>* = 0 in *Q*, *yτ*(0) = *y*ˆ

*y<sup>τ</sup>* = 0 on Σ,

*∂τ* (0, 0)

+ *A* + *f* �

*hoχ*<sup>O</sup> + <sup>w</sup>*χω*

�

*sentinel insensitive to the missing data (i.e. such that (5) holds), is equivalent to the null-controllability*

*<sup>y</sup><sup>τ</sup>* <sup>=</sup> *<sup>∂</sup><sup>y</sup>*

*<sup>o</sup>* in Ω,

(*yo*) *Id*

�

*yτ*(*T*) *dx* = 0.

*yτ*(*T*) *dx*,

L∗*q* = 0 *in Q*, (6.1) *<sup>q</sup>*(*T*) = *hoχ*<sup>O</sup> + <sup>w</sup>*χω in* <sup>Ω</sup>, (6.2) *q* = 0 *on* Σ, (6.3)

(*yo*) *Id is the adjoint operator. Then, the existence of an instantaneous*

*q*(0) = 0 *in* Ω. (6.4)

**6. Instantaneous sentinel. Case of no information on the missing data**

⎧ ⎨ ⎩

<sup>L</sup> <sup>=</sup> *<sup>∂</sup> ∂t*

> � Ω �

**Remark 5.** *From above, the insensitive criterion (5) to the missing term τy*ˆ*<sup>o</sup> is given by :*

*hoχ*<sup>O</sup> + <sup>w</sup>*χω*

**Lemma 2.** *Let q be the solution to the following well-posed backward problem :*

w �= −*ho.*

which is a difficult problem.

where *y<sup>τ</sup>* denotes the derivative

*where* <sup>L</sup><sup>∗</sup> <sup>=</sup> <sup>−</sup> *<sup>∂</sup>*

*problem (6.1)-(6.3) with*

with *yo* = *y*(*t*, *x*; 0, 0) for *λ* = *τ* = 0. Then

*<sup>∂</sup><sup>t</sup>* + *A*<sup>∗</sup> + *f* �

the solution *y* = *y*(*λ*, *τ*) satisfies the system :

and where L is the linear differential operator defined by

*∂*S *∂τ* (0, 0) =

> � Ω �

**Proof -** Multiplying (6.1) by *y<sup>τ</sup>* and integrating by parts we obtain :

$$\int\_{Q} \left( \mathcal{L}^\* q \right) y\_\tau \, d\mathbf{x} dt = \int\_{Q} q \, \mathcal{L} y\_\tau \, d\mathbf{x} dt - \int\_{\Omega} q(T) y\_\tau(T) \, d\mathbf{x}$$

$$+ \int\_{\Omega} q(0) y\_\tau(0) \, d\mathbf{x} - \int\_{\Sigma} \frac{\partial q}{\partial \nu} y\_\tau \, d\sigma + \int\_{\Sigma} \frac{\partial y\_\tau}{\partial \nu} q \, d\sigma = 0. \tag{6.5}$$

But, *y<sup>τ</sup>* is solution to (6), and *q* verifies (6.2) and (6.3). Hence

$$-\int\_{\Omega} \left( h\_o \chi\_{\mathcal{O}} + \mathfrak{w} \chi\_{\omega} \right) y\_\tau(T) \, d\mathfrak{x} + \int\_{\Omega} q(0) \, \mathfrak{z}^o \, d\mathfrak{x} = 0.1$$

If a sentinel exists, then we have (5). So it remains :

$$\int\_{\Omega} q(0) \, \mathfrak{J}^o \, d\mathfrak{x} = 0 \qquad \text{for every} \quad \mathfrak{J}^o \in L^2(\Omega).$$

and consequently *q*(0) = 0 in Ω. The converse is obvious.

**Corollary 2.** *If there is no information on the missing data, then the instantaneous sentinel is nul.*

**Proof -** The proof is easy and lies on the backward uniqueness property (see Lions-Malgrange Lions J.-L. & Malgrange B. (1960)) : If an instantaneous sentinel exists, then from the above Lemma we have (6.1), (6.3) and (6.4). We deduce that

$$q \equiv 0 \qquad \text{in} \quad Q.$$

Thus, in particular for *<sup>t</sup>* = *<sup>T</sup>*, we have *<sup>q</sup>*(*T*) = 0. So that *hoχ*<sup>O</sup> + <sup>w</sup>*χω* = 0 in <sup>Ω</sup>, and hence -as in Lions Lions J.-L. (1992)- the sentinel is nul.

6.0.0.1 Information given by the sentinel

We briefly show how (5) and (5.8) are sufficient to get information on the pollution term *λ* ˆ *ξ*. We write

$$S(\lambda, \tau) \simeq S(0, 0) + \lambda \frac{\partial S}{\partial \lambda}(0, 0), \qquad \text{for} \quad \lambda, \tau \quad \text{small} \dots$$

Using (5.4), we have

$$
\lambda \frac{\partial \mathbf{S}}{\partial \lambda}(\mathbf{0}, \mathbf{0}) = \int\_{Q} (h\_{0} \chi\_{O} + \mathbf{w} \chi\_{\omega}) y\_{\lambda} \, d\mathbf{x} dt \simeq \int\_{Q} (h\_{0} \chi\_{O} + \mathbf{w} \chi\_{\omega}) (m\_{o} - y\_{0}) \, d\mathbf{x} dt,\tag{6.6}
$$

where the derivative

$$y\_{\lambda} = \left(\partial y / \partial \lambda\right)(0,0)$$

only depends on ˆ *ξ* and other known data. Consequently, the estimates (6.6) contain the information on *λ* ˆ *ξ*. Indeed, *y<sup>λ</sup>* is solution to the well-posed problem

$$
\mathcal{L}\,y\_{\lambda} = 0 \text{ in } Q, \quad y\_{\lambda}(0) = 0 \text{ in } \Omega, \quad \text{and } y\_{\lambda} = \oint \text{ on } \Sigma. \tag{6.7}
$$

Multiplying by *q* and integrating by parts, we obtain

$$\int\_{Q} \lambda \widehat{\xi} \, d\mathbf{x} dt = \int\_{Q} (h\_0 \chi\_O + \mathbf{w} \chi\_{\omega}) (m\_o - y\_0) \, d\mathbf{x} dt.$$

**Remark 6.** *From the previous corollary, we see that in order to get a non nul sentinel, we need more information (on the structure) of the missing initial data. We will need here the assumption (7) below.*

**Remark 7.** *The controllability problem (7.3)-(7.4) has at least a solution: Given ho* <sup>∈</sup> *<sup>L</sup>*2(O)*, the*

Some Aspects of the Sentinel Method for Pollution Problems 201

<sup>−</sup>*ho in* O ∩ *<sup>ω</sup>*,

(w,*z*)∈A �w�*L*<sup>2</sup>(*ω*)

**Proof -** The domain A is no empty and is closed. The mapping : w −→ �w�*L*<sup>2</sup>(*ω*) is continuous, coercitive and strictly convex. We deduce that there exists a unique solution to

�<sup>w</sup>�*L*<sup>2</sup>(*ω*) ≤ �w�*L*<sup>2</sup>(*ω*) <sup>∀</sup> (w, *<sup>q</sup>*) ∈ A.

We now use the penalization method in order to characterize the optimal control (<sup>w</sup>, *<sup>q</sup>*). Let

*<sup>L</sup>*<sup>2</sup>(*ω*) +

(w,*q*)∈ U

The following proposition gives the existence of a solution to the penalized problem (P*ε*).

**Proof -** We have A⊂U. Moreover, A is nonempty by the previous theorem. Consequently, U is nonempty and closed. The cost function *J<sup>ε</sup>* is continuous, coercitive and strictly convex.

1 2*ε*

�L∗*q*�<sup>2</sup>

*Jε*(w, *q*)

<sup>L</sup>∗*<sup>q</sup>* <sup>∈</sup> *<sup>L</sup>*2(*Q*) in *<sup>Q</sup>*, *<sup>q</sup>*(0) <sup>⊥</sup> *<sup>y</sup><sup>o</sup>*

*<sup>q</sup>*(*T*) = *hoχ*O∩*<sup>ω</sup>* + <sup>w</sup>*χω* in <sup>Ω</sup>, *<sup>q</sup>* = 0 on <sup>Σ</sup>

*<sup>L</sup>*<sup>2</sup>(*Q*),

<sup>L</sup>∗*<sup>q</sup>* <sup>=</sup> 0 in *<sup>Q</sup>*, *<sup>q</sup>*(0) <sup>⊥</sup> *<sup>y</sup><sup>o</sup>*

*<sup>q</sup>*(*T*) = *hoχ*O∩*<sup>ω</sup>* + <sup>w</sup>*χω* in <sup>Ω</sup>, *<sup>q</sup>* = 0 on <sup>Σ</sup>

Here we are interested in the problem (5.8). We consider the optimization problem :

(P) min

<sup>0</sup> *in <sup>ω</sup>* \ (O ∩ *<sup>ω</sup>*), (7.5)

*<sup>i</sup>* 1 ≤ *i* ≤ *N*,

*<sup>i</sup>* 1 ≤ *i* ≤ *N*,

 .

 .

w =

*and so, the set of solutions to (7.3)-(7.4) is not empty.*

(w,*q*) such that

(P) denoted by (<sup>w</sup>, *<sup>q</sup>*) ∈ A, which satisfies

and we consider the problem (P*ε*) given by

(w,*q*) such that

Hence, problem (P*ε*) has a unique solution.

be *ε* > 0, we introduce the function

 

**Theorem 3.** *There is a unique pair* (<sup>w</sup>, *<sup>q</sup>*) *solution to the problem* (P)*.*

*<sup>J</sup>ε*(w, *<sup>q</sup>*) = <sup>1</sup>

 

**Proposition 3.** *The problem* (P*ε*) *has a unique solution denoted by* (w*ε*, *qε*).

2 �w�<sup>2</sup>

(P*ε*) min

*control 'solution'* w *is given by*

**8. Penalization**

A = 

with

with

U = 

#### **7. Instantaneous sentinel. Case of partial information on the missing data**

In this section, we consider the case of the *incomplete* initial data :

$$y(0) = y^o + \sum\_{i=1}^{N} \tau\_i \hat{y}\_i^o \tag{7.1}$$

where *y*ˆ*<sup>o</sup> <sup>i</sup>* , 1 <sup>≤</sup> *<sup>i</sup>* <sup>≤</sup> *<sup>N</sup>* are linearly independent in *<sup>L</sup>*2(Ω), and belong to a vector subspace of *<sup>N</sup>* dimension, which we denote by *<sup>Y</sup>* <sup>=</sup> �*y*ˆ*<sup>o</sup>* <sup>1</sup>, *<sup>y</sup>*ˆ*<sup>o</sup>* <sup>2</sup>, ··· , *<sup>y</sup>*ˆ*<sup>o</sup> <sup>N</sup>*�. The parameters *τ<sup>i</sup>* are *unknown* and are supposed small.

If we denote by *<sup>y</sup>τ<sup>i</sup>* <sup>=</sup> *<sup>∂</sup><sup>y</sup> ∂τ<sup>i</sup>* (0, 0) for *λ* = *τ<sup>i</sup>* = 0, 1 ≤ *i* ≤ *N*, then *yτ<sup>i</sup>* is solution with *yτi*(0) = *y*ˆ *o i* . Now, we assume the following :

$$
\omega \cap \mathcal{O} \neq \mathcal{O}
$$

and we define the instantaneous sentinel by :

$$\mathcal{S}(\lambda, \tau) = \int\_{\Omega} \left( h\_o \chi\_{\circlearrowright \omega} + \mathsf{w} \chi\_{\omega} \right) y(T; \lambda, \tau) \, d\mathfrak{x} = 0$$

with *τ* = (*τ*1, ··· , *τN*). Hence, looking for w such that the sentinel S is insensitive to the missing terms *τiy*ˆ*<sup>o</sup> <sup>i</sup>* , is finding w such that :

$$\frac{\partial \mathcal{S}}{\partial \boldsymbol{\pi}\_{l}}(\mathbf{0}, \boldsymbol{\tau}\_{l} = \mathbf{0}) = \int\_{\Omega} \left( h\_{0} \boldsymbol{\chi}\_{\mathcal{O} \cap \omega} + \mathbf{w} \boldsymbol{\chi}\_{\omega} \right) \boldsymbol{y}\_{\boldsymbol{\pi}\_{l}}(T) \, d\mathbf{x} = \mathbf{0}, \quad 1 \le i \le N. \tag{7.2}$$

**Lemma 3.** *The existence of an instantaneous sentinel insensitive to the missing terms is equivalent to the existence of a unique pair* (w, *q*) *such that we have :*

$$\begin{cases} \mathcal{L}^\* q = 0 & \text{in } \mathcal{Q}\_{\prime} \\ q(T) = h\_{\partial \chi\_{\mathcal{O} \cap \omega}} + \mathfrak{w} \chi\_{\omega} & \text{in } \bar{\Omega}\_{\prime} \\ q = 0 & \text{on } \Sigma. \end{cases} \tag{7.3}$$

*and such that*

$$q(0) \in Y^{\perp}.\tag{7.4}$$

**Proof -** Multiplying the first equation in (7.3) by *yτ<sup>i</sup>* and integrating by parts, we find :

$$-\int\_{\Omega} q(T) \, y\_{\tau\_l}(T) \, d\mathbf{x} + \int\_{\Omega} q(0) \hat{y}\_l^\rho \, d\mathbf{x} = \mathbf{0}\_{\lambda}$$

since *yτ<sup>i</sup>* is solution of (6). And so,

$$-\int\_{\Omega} \left( h\_{\boldsymbol{\vartheta}} \chi\_{\boldsymbol{\vartheta} \cap \boldsymbol{\omega}} + \mathbf{w} \chi\_{\boldsymbol{\omega}} \right) \mathcal{Y}\_{\mathbb{T}\_i}(T) \, d\boldsymbol{x} + \int\_{\Omega} q(0) \hat{y}\_i^{\boldsymbol{\vartheta}} \, d\boldsymbol{x} = 0, \qquad 1 \le i \le N.$$

But the sentinel should satisfy (7.2). Hence,

$$\int\_{\Omega} q(0) \mathfrak{J}\_i^{\rho} \, d\mathfrak{x} = 0, \quad \text{for all} \quad 1 \le i \le N.$$

So that

$$q(0) \perp \mathcal{Y}\_{i'}^{\rho} \qquad 1 \le i \le N \qquad \text{if and only if} \qquad q(0) \in Y^{\perp},$$
 where  $Y^{\perp}$  is the orthogonal of  $Y$  in  $L^2(\Omega)$ .

**Remark 7.** *The controllability problem (7.3)-(7.4) has at least a solution: Given ho* <sup>∈</sup> *<sup>L</sup>*2(O)*, the control 'solution'* w *is given by*

$$\mathcal{W} = \left\{ \begin{array}{l} -\text{h}\_{0} \text{ in } \mathcal{O} \cap \omega, \\ 0 \quad \text{in } \omega \mid (\mathcal{O} \cap \omega), \end{array} \right. \tag{7.5}$$

*and so, the set of solutions to (7.3)-(7.4) is not empty.*

#### **8. Penalization**

Here we are interested in the problem (5.8). We consider the optimization problem :

$$(\mathcal{P}) \qquad \min\_{(\mathbf{w}, z) \in \mathcal{A}} \|\mathbf{w}\|\_{L^2(\omega)}$$

with

16 Will-be-set-by-IN-TECH

*N* ∑ *i*=1 *τ<sup>i</sup> y*ˆ *o*

*<sup>i</sup>* , 1 <sup>≤</sup> *<sup>i</sup>* <sup>≤</sup> *<sup>N</sup>* are linearly independent in *<sup>L</sup>*2(Ω), and belong to a vector subspace of *<sup>N</sup>*

�

(0, 0) for *λ* = *τ<sup>i</sup>* = 0, 1 ≤ *i* ≤ *N*, then *yτ<sup>i</sup>* is solution with *yτi*(0) = *y*ˆ

*y*(*T*; *λ*, *τ*) *dx* = 0

<sup>2</sup>, ··· , *<sup>y</sup>*ˆ*<sup>o</sup>*

*ω* ∩ O �= ∅

*hoχ*O∩*<sup>ω</sup>* + w*χω*

*hoχ*O∩*<sup>ω</sup>* + w*χω*

**Proof -** Multiplying the first equation in (7.3) by *yτ<sup>i</sup>* and integrating by parts, we find :

*q*(*T*) *yτi*(*T*) *dx* +

*yτi*(*T*) *dx* +

�

with *τ* = (*τ*1, ··· , *τN*). Hence, looking for w such that the sentinel S is insensitive to the

**Lemma 3.** *The existence of an instantaneous sentinel insensitive to the missing terms is equivalent to*

L∗*q* = 0 *in Q*, *q*(*T*) = *hoχ*O∩*<sup>ω</sup>* + w*χω in* Ω, *q* = 0 *on* Σ.

> � Ω *q*(0)*y*ˆ *o <sup>i</sup> dx* = 0,

� Ω *q*(0)*y*ˆ *o*

*<sup>i</sup> dx* = 0, for all 1 ≤ *i* ≤ *N*.

*<sup>i</sup>* , 1 ≤ *i* ≤ *N* if and only if *q*(0) ∈ *Y*⊥,

�

*<sup>i</sup>* (7.1)

*o i* .

(7.3)

*<sup>N</sup>*�. The parameters *τ<sup>i</sup>* are *unknown* and are

*yτi*(*T*) *dx* = 0, 1 ≤ *i* ≤ *N*. (7.2)

*q*(0) ∈ *Y*⊥. (7.4)

*<sup>i</sup> dx* = 0, 1 ≤ *i* ≤ *N*.

**7. Instantaneous sentinel. Case of partial information on the missing data**

*y*(0) = *y<sup>o</sup>* +

<sup>1</sup>, *<sup>y</sup>*ˆ*<sup>o</sup>*

In this section, we consider the case of the *incomplete* initial data :

dimension, which we denote by *<sup>Y</sup>* <sup>=</sup> �*y*ˆ*<sup>o</sup>*

*∂τ<sup>i</sup>*

and we define the instantaneous sentinel by :

(0, *τ<sup>i</sup>* = 0) =

*the existence of a unique pair* (w, *q*) *such that we have :*

− � Ω

*hoχ*O∩*<sup>ω</sup>* + w*χω*

� Ω *q*(0)*y*ˆ *o*

*o*

But the sentinel should satisfy (7.2). Hence,

*q*(0) ⊥ *y*ˆ

where *Y*<sup>⊥</sup> is the orthogonal of *Y* in *L*2(Ω).

S(*λ*, *τ*) =

*<sup>i</sup>* , is finding w such that :

� Ω �

⎧ ⎨ ⎩ � Ω �

where *y*ˆ*<sup>o</sup>*

supposed small.

missing terms *τiy*ˆ*<sup>o</sup>*

*and such that*

So that

If we denote by *<sup>y</sup>τ<sup>i</sup>* <sup>=</sup> *<sup>∂</sup><sup>y</sup>*

Now, we assume the following :

*∂*S *∂τ<sup>i</sup>*

since *yτ<sup>i</sup>* is solution of (6). And so,

− � Ω �

$$\mathcal{A} = \left\{ (\mathsf{w}\boldsymbol{\mathcal{q}}) \text{ such that} \, \middle| \begin{aligned} \mathcal{L}^\* \boldsymbol{\mathcal{q}} &= \boldsymbol{0} & \text{in } Q, \boldsymbol{q}(\boldsymbol{0}) \perp \boldsymbol{\mathcal{Y}}\_l^0 \, \middle| \, 1 \le i \le N, \\ \boldsymbol{\mathcal{q}}(T) &= \boldsymbol{h}\_{\boldsymbol{0}} \boldsymbol{\chi}\_{\mathcal{O}\cap\mathcal{U}^\omega} + \mathsf{w} \boldsymbol{\chi}\_{\omega} \text{ in } \Omega, \quad \boldsymbol{q} = \boldsymbol{0} & \text{on } \Sigma \end{aligned} \right\}.$$

**Theorem 3.** *There is a unique pair* (<sup>w</sup>, *<sup>q</sup>*) *solution to the problem* (P)*.*

**Proof -** The domain A is no empty and is closed. The mapping : w −→ �w�*L*<sup>2</sup>(*ω*) is continuous, coercitive and strictly convex. We deduce that there exists a unique solution to (P) denoted by (<sup>w</sup>, *<sup>q</sup>*) ∈ A, which satisfies

$$\|\|\widehat{\mathbf{w}}\|\|\_{L^{2}(\omega)} \leq \|\mathbf{w}\|\|\_{L^{2}(\omega)} \qquad \forall \ (\mathbf{w}, q) \in \mathcal{A}.$$

We now use the penalization method in order to characterize the optimal control (<sup>w</sup>, *<sup>q</sup>*). Let be *ε* > 0, we introduce the function

$$J\_{\varepsilon}(\mathfrak{w}, q) = \frac{1}{2} \|\mathfrak{w}\|\_{L^2(\omega)}^2 + \frac{1}{2\varepsilon} \|\mathcal{L}^\* q\|\_{L^2(Q)'}^2$$

and we consider the problem (P*ε*) given by

$$(\mathcal{P}\_{\varepsilon}) \qquad \min\_{(\mathbf{w}, q) \in \mathcal{U}} J\_{\varepsilon}(\mathbf{w}, q)$$

with

$$\mathcal{U} = \left\{ (\mathsf{w}, q) \text{ such that} \begin{aligned} \mathcal{L}^\* q &\in L^2(\mathbb{Q}) \quad \text{in } Q, \, q(0) \perp \widehat{\mathcal{Y}}\_i^0 \, 1 \le i \le N\_\prime \\\, \mathsf{h}\_\prime \mathsf{X} \boldsymbol{\mathcal{O}} \, \mathsf{m} + \mathsf{w} \boldsymbol{\chi}\_\prime \, \text{in } \Omega, \quad q = 0 \qquad \text{on } \Sigma \end{aligned} \right\}.$$

The following proposition gives the existence of a solution to the penalized problem (P*ε*).

**Proposition 3.** *The problem* (P*ε*) *has a unique solution denoted by* (w*ε*, *qε*).

**Proof -** We have A⊂U. Moreover, A is nonempty by the previous theorem. Consequently, U is nonempty and closed. The cost function *J<sup>ε</sup>* is continuous, coercitive and strictly convex. Hence, problem (P*ε*) has a unique solution.

does not depend on the observation and then makes the method efficient in case of several measures. We refer to Ainseba B.E. et al. (1994a)Ainseba B.-E. et al. (1994b) and to the book of

Some Aspects of the Sentinel Method for Pollution Problems 203

As we have seen, the existence of an instantaneous sentinel is equivalent to controllability problems. In the case of no supplementary information on the missing data (that we do not

In the case where we have some more information on the structure of the missing data (which becomes incomplete data), we proved that the existence of the instantaneous sentinel is equivalent to a nontrivial controllability problem, which has a solution under the assumption

Ainseba B.-E., Kernevez J.-P., Luce R. *Identification de pollution dans une rivière*,

Ainseba B.-E., Kernevez J.-P., Luce R. *Identification de paramètres dans les problèmes non linéaires*

Bodart O., Application de la méthode des sentinelles à l'identification de sources de

Bodart O., Demeestere P., Contrôle frontière de l'explosion en temps fini de la solution d'une équation de combustion. C.R. Acad. Sci., Série I, 325(8), 1997, pp 841 − 845. Bodart O., Fabre C., Contrôle insensibilisant la norme de la solution d'une équation de la chaleur semi-linéaire. C.R. Acad. Sci., Série I, 316(8), 1993, pp 789 − 794. Bodart O., Fabre C., Controls insensitizing the norm of a semilinear heat equation. Journ. of

Brezis H., Analyse fonctionnelle, Collection mathématiques appliquées pour la maîtrise,

Demeestere P., Méthode des sentinelles : étude comparative et application à l'identification

A. Fursikov, O. Yu. Imanuvilov, Controllability of evolution equations. Lecture Notes, Research Institute of Mathematics, Seoul National University, Korea (1996). O. Yu. Imanuvilov Controllability of parabolic equations. Sbornik Mathematics 186 : 6, 1995,

Kernevez J.-P., The Sentinel Method and its Application to Environmental Pollution Problems,

Lebeau G., Robbiano L., Contrôle exacte de l'équation de la chaleur, Comm. P.D.E., 20, 1995,

Lions J.-L., Contrôle optimal de systèmes gouvernés par des équations aux dérivées

Lions J.-L. Sentinelles pour les systèmes distribués à données incomplètes. Masson, Paris,

Lions J.-L., Malgrange B. *Sur l'unicité rétrograde dans les problèmes mixtes paraboliques*, Math.

d'une frontière Contrôle du temps d'explosion d'une équation de combustion. Thèse

Math. Analysis and Applications, 195, 1995, pp 658 − 683.

de Doctorat, Université de Technologie de Compiègne, 1997.

ESAIM:Mathematical Modeling and Numerical Analysis M2AN, 1994, pp 819 − 823.

*à données manquantes*, ESAIM:Mathematical Modeling and Numerical Analysis

pollution dans un système distribué. Thèse de Doctorat, Université de Technologie

J.-P. Kernevez Kernevez J.-P. (1997) for further information and numerical details.

want to identify), the sentinel is nul as in Lions Lions J.-L. (1992).

M2AN, 1994, pp 313 − 328.

de Compiègne, 1994.

Masson, 1984.

pp 879 − 900.

pp 335 − 356.

1992.

CRC. Press, Boca Raton, FL, 1997.

Scand., 8, pp 277-286 (1960).

partielles. Dunod, gauthier-Villars, Paris 1968.

O ∩ *ω* �= ∅.

**10. References**

We now prove the following proposition.

**Proposition 4.** *Let be* (w*ε*, *qε*) *the unique solution of* (P*ε*)*, then for ε* −→ 0 *we obtain*

$$\begin{cases} \mathsf{w}\_{\varepsilon} \rightharpoonup \widehat{\mathsf{w}} & \text{weakly in} \quad L^2(\omega)\_{\prime} \\ \eta\_{\varepsilon} \rightharpoonup \widehat{q} & \text{weakly in} \quad \mathcal{W}(0,T)\_{\prime} \end{cases}$$

*where* <sup>W</sup>(0, *<sup>T</sup>*) = *<sup>q</sup>* ∈ C(0, *<sup>T</sup>*); *<sup>q</sup>*(*t*) <sup>∈</sup> *<sup>L</sup>*2(Ω) *.*

**Proof -** Since (w*ε*, *qε*) is solution to (P*ε*) then

$$J\_{\varepsilon}(\mathsf{w}\_{\varepsilon}, q\_{\varepsilon}) \le J\_{\varepsilon}(\mathsf{w}, q) \,\,\forall (\mathsf{w}, q) \in \mathcal{A}.$$

In particular (<sup>w</sup>, *<sup>q</sup>*) ∈A⊂U is solution to (P*ε*). Then :

$$\|\mathfrak{w}\_{\varepsilon}\|\_{L^{2}(\omega)} \leq \|\widehat{w}\|\_{L^{2}(\omega)} \leq \mathbb{C} \qquad \text{and so} \qquad \|\mathcal{L}^{\*}q\_{\varepsilon}\|\_{L^{2}(Q)} \leq \mathbb{C}\sqrt{\varepsilon}\rho$$

where *C* is a positive constant which is not the same at each time.

Knowing that (w*ε*, *qε*) ∈ U, we deduce

$$\|q\_{\varepsilon}\|\_{H^{2,1}(Q)} \le \mathbb{C}.$$

Then, there is a subsequent (w*ε*, *<sup>q</sup>ε*), and two functions <sup>w</sup><sup>0</sup> <sup>∈</sup> *<sup>L</sup>*2(*ω*) and *<sup>q</sup>*<sup>0</sup> <sup>∈</sup> *<sup>H</sup>*2,1(*Q*) such that

$$ \begin{cases} \mathsf{w}\_{\varepsilon} \rightharpoonup \mathsf{w}\_{0} & \text{in } L^{2}(\omega), \\ q\_{\varepsilon} \rightharpoonup q\_{0} & \text{in } H^{2,1}(Q). \end{cases} $$

Now as *<sup>H</sup>*2,1(*Q*) <sup>→</sup> *<sup>L</sup>*2(*Q*) with compact injection, the pair (w0, *<sup>q</sup>*0) satisfies the following :

$$
\mathcal{L}^\* q\_0 = h + \mathbf{k}\_0 \chi\_\omega \text{ in } Q, \quad q\_0(T) = q\_0 = 0 \text{ in } \Omega, \quad q\_0 = 0 \text{ in } \Sigma. \tag{8.1}
$$

But,

$$\frac{1}{2} \|\mathfrak{w}\_0\|\_{L^2(\omega)}^2 \le \liminf\_{\varepsilon \to 0} J\_\varepsilon(\mathfrak{w}\_{\varepsilon\prime} q\_\varepsilon) \le J\_\varepsilon(\widehat{w}, \widehat{q}) \le \frac{1}{2} \||\widehat{w}\||\_{L^2(\omega)}.$$

Since (<sup>w</sup>, *<sup>q</sup>*) is the unique solution of (P), then <sup>w</sup> <sup>=</sup> <sup>w</sup>0. Finally, as *<sup>q</sup>*<sup>0</sup> satisfies (8.1), we obtain *<sup>q</sup>* <sup>=</sup> *<sup>q</sup>*0.

#### **9. Concluding remarks**

From the numerical aspect, there where some results in the litterature. In B.-E. Ainseba et al. Ainseba B.E. et al. (1994a)Ainseba B.-E. et al. (1994b), the authors compare numerically the least square method and the sentinel one during time [0, *T*[ and find that they are equivalent in the linear case *f*(*y*) = *C y*. However, the sentinel approach has an advantage in case of several measures -as it is really the case in general-; Indeed, it suffices to calculate a simple integral to identify the parameter each time, since one has to minimize a quadratic cost functional for the least square approach in order to determine both the parameter and the missing terms.

In the nonlinear case and when the observation is noisy, the least square method costs a lot numerically and can fail after a large number of iterations, while the sentinel method is relatively robust face to the perturbations of the observation. Moreover, the calculus of w does not depend on the observation and then makes the method efficient in case of several measures. We refer to Ainseba B.E. et al. (1994a)Ainseba B.-E. et al. (1994b) and to the book of J.-P. Kernevez Kernevez J.-P. (1997) for further information and numerical details.

As we have seen, the existence of an instantaneous sentinel is equivalent to controllability problems. In the case of no supplementary information on the missing data (that we do not want to identify), the sentinel is nul as in Lions Lions J.-L. (1992).

In the case where we have some more information on the structure of the missing data (which becomes incomplete data), we proved that the existence of the instantaneous sentinel is equivalent to a nontrivial controllability problem, which has a solution under the assumption O ∩ *ω* �= ∅.

#### **10. References**

18 Will-be-set-by-IN-TECH

 <sup>w</sup>*<sup>ε</sup>* <sup>w</sup> *weakly in L*<sup>2</sup> (*ω*), *<sup>q</sup><sup>ε</sup> q weakly in* <sup>W</sup> (0, *<sup>T</sup>*),

> *.*

*Jε*(w*ε*, *qε*) ≤ *Jε*(w, *q*) ∀(w, *q*) ∈ A.

√*ε*,

�w*ε*�*L*<sup>2</sup>(*ω*) ≤ �*<sup>w</sup>*�*L*<sup>2</sup>(*ω*) <sup>≤</sup> *<sup>C</sup>* and so �L∗*qε*�*L*<sup>2</sup>(*Q*) <sup>≤</sup> *<sup>C</sup>*

�*qε*�*H*2,1(*Q*) ≤ *C*.

Then, there is a subsequent (w*ε*, *<sup>q</sup>ε*), and two functions <sup>w</sup><sup>0</sup> <sup>∈</sup> *<sup>L</sup>*2(*ω*) and *<sup>q</sup>*<sup>0</sup> <sup>∈</sup> *<sup>H</sup>*2,1(*Q*) such

 w*<sup>ε</sup>* w<sup>0</sup> in *L*2(*ω*), *q<sup>ε</sup> q*<sup>0</sup> in *H*2,1(*Q*).

Now as *<sup>H</sup>*2,1(*Q*) <sup>→</sup> *<sup>L</sup>*2(*Q*) with compact injection, the pair (w0, *<sup>q</sup>*0) satisfies the following :

L∗*q*<sup>0</sup> = *h* + k0*χω* in *Q*, *q*0(*T*) = *q*<sup>0</sup> = 0 in Ω, *q*<sup>0</sup> = 0 in Σ. (8.1)

1 2

�*<sup>w</sup>*�*L*<sup>2</sup>(*ω*).

inf *<sup>J</sup>ε*(w*ε*, *<sup>q</sup>ε*) <sup>≤</sup> *<sup>J</sup>ε*(*<sup>w</sup>*, *<sup>q</sup>*) <sup>≤</sup>

Since (<sup>w</sup>, *<sup>q</sup>*) is the unique solution of (P), then <sup>w</sup> <sup>=</sup> <sup>w</sup>0. Finally, as *<sup>q</sup>*<sup>0</sup> satisfies (8.1), we obtain

From the numerical aspect, there where some results in the litterature. In B.-E. Ainseba et al. Ainseba B.E. et al. (1994a)Ainseba B.-E. et al. (1994b), the authors compare numerically the least square method and the sentinel one during time [0, *T*[ and find that they are equivalent in the linear case *f*(*y*) = *C y*. However, the sentinel approach has an advantage in case of several measures -as it is really the case in general-; Indeed, it suffices to calculate a simple integral to identify the parameter each time, since one has to minimize a quadratic cost functional for the least square approach in order to determine both the parameter and the missing terms.

In the nonlinear case and when the observation is noisy, the least square method costs a lot numerically and can fail after a large number of iterations, while the sentinel method is relatively robust face to the perturbations of the observation. Moreover, the calculus of w

**Proposition 4.** *Let be* (w*ε*, *qε*) *the unique solution of* (P*ε*)*, then for ε* −→ 0 *we obtain*

*<sup>q</sup>* ∈ C(0, *<sup>T</sup>*); *<sup>q</sup>*(*t*) <sup>∈</sup> *<sup>L</sup>*2(Ω)

where *C* is a positive constant which is not the same at each time.

*<sup>L</sup>*<sup>2</sup>(*ω*) ≤ lim *ε*→0

We now prove the following proposition.

Knowing that (w*ε*, *qε*) ∈ U, we deduce

1 2 �w0�<sup>2</sup>

**9. Concluding remarks**

**Proof -** Since (w*ε*, *qε*) is solution to (P*ε*) then

In particular (<sup>w</sup>, *<sup>q</sup>*) ∈A⊂U is solution to (P*ε*). Then :

*where* W(0, *T*) =

that

But,

*<sup>q</sup>* <sup>=</sup> *<sup>q</sup>*0.


**10** 

*Thailand* 

Somporn Chantara *Chiang Mai University* 

**PM10 and Its Chemical Composition:** 

**A Case Study in Chiang Mai, Thailand** 

This chapter describes the chemical composition and spatial and temporal variations of airborne particulate matter that have a diameter of less than 10 µm (PM10), as well as its possible health effects. A case study of air pollution in the city of Chiang Mai, Thailand was conducted due to its geographical features and meteorological conditions; i.e. temperature inversion, wind velocity and precipitation levels. Chiang Mai, with an altitude of approximately 310 m above sea level, is situated approximately 700 km north of Bangkok. The city covers an area of approximately 20,107 km2 and is the country's second largest province. The city is situated in the Chiang Mai-Lamphun Basin and is surrounded by mountains. In the dry season, there is a low level of precipitation, and there are calm winds and a vertical temperature inversion, while air pollutants are generated from various sources of mostly anthropogenic activities and accumulate in the lower atmosphere and definitely have an effect on human health and the environment. PM10 is considered the most significant air pollutant that contributes to serious air pollution in the dry season of Northern Thailand. Its major sources are open burning and internal combustion exhaust from traffic. However, traffic density seems to be constant for the whole year, while open burning is mostly performed in the dry season, which coincides with the peak of the annual haze episode in the upper region of Northern Thailand. Large scale open burning in this region consists of forest fires and the burning of agricultural waste. These activities

definitely emit a variety of air pollutants in the forms of both particulates and gases.

Air pollution is quite common in most countries with big cities. Atmospheric aerosol particles originate from a wide variety of natural and anthropogenic sources. Primary particles are directly emitted from sources, such as biomass burning, incomplete combustion of fossil fuels and traffic-related suspension of road, soil, dust, sea salt and biological materials. Secondary particles are formed by gas-to-particle conversion in the atmosphere (Pöschl, 2005). Concentration, composition and size distribution of atmospheric aerosol particles are temporally and spatially highly variable. The predominant particle components of air particulate matter (PM) are sulfate, nitrate, ammonium, sea salt, mineral dust, organic compounds and black or elemental carbon (Pöschl, 2005). The effects of aerosols to the atmosphere, climate and public health are among the central topics in the current environmental research. An airborne particulate matter consists of several of inorganic and organic species, many of which can adversely affect human health (Pöschl, 2005; Mauderly

**1. Introduction** 


### **PM10 and Its Chemical Composition: A Case Study in Chiang Mai, Thailand**

Somporn Chantara *Chiang Mai University Thailand* 

#### **1. Introduction**

20 Will-be-set-by-IN-TECH

204 Air Quality – Monitoring and Modeling

Miloudi Y., Nakoulima O., Omrane A., A method of detection of pollution in dissipative

Miloudi Y., Nakoulima O., Omrane A., Instantaneous sentinels for distributed systems of

Nakoulima O., Contrôlabilité à zéro avec contraintes sur le contrôle. C. R. Acad. Sci. Paris,

Puel J.-P., Contrôlablité approchée et contrôlabilité exacte. Notes de cours de D.E.A.

ESAIM:Proc Vol. 17, 2007, pp 67 − 79.

Université Pierre et Marie Curie, Paris, 2001.

Ser. I 339, 2004, pp 405 − 410.

451 − 459.

systems of incomplete data, Proc. of the European Society of Appl. and Indust. Math.,

incomplete data, Inverse Problems in Science and Engineering, Vol. 17, 4, 2009, pp

This chapter describes the chemical composition and spatial and temporal variations of airborne particulate matter that have a diameter of less than 10 µm (PM10), as well as its possible health effects. A case study of air pollution in the city of Chiang Mai, Thailand was conducted due to its geographical features and meteorological conditions; i.e. temperature inversion, wind velocity and precipitation levels. Chiang Mai, with an altitude of approximately 310 m above sea level, is situated approximately 700 km north of Bangkok. The city covers an area of approximately 20,107 km2 and is the country's second largest province. The city is situated in the Chiang Mai-Lamphun Basin and is surrounded by mountains. In the dry season, there is a low level of precipitation, and there are calm winds and a vertical temperature inversion, while air pollutants are generated from various sources of mostly anthropogenic activities and accumulate in the lower atmosphere and definitely have an effect on human health and the environment. PM10 is considered the most significant air pollutant that contributes to serious air pollution in the dry season of Northern Thailand. Its major sources are open burning and internal combustion exhaust from traffic. However, traffic density seems to be constant for the whole year, while open burning is mostly performed in the dry season, which coincides with the peak of the annual haze episode in the upper region of Northern Thailand. Large scale open burning in this region consists of forest fires and the burning of agricultural waste. These activities definitely emit a variety of air pollutants in the forms of both particulates and gases.

Air pollution is quite common in most countries with big cities. Atmospheric aerosol particles originate from a wide variety of natural and anthropogenic sources. Primary particles are directly emitted from sources, such as biomass burning, incomplete combustion of fossil fuels and traffic-related suspension of road, soil, dust, sea salt and biological materials. Secondary particles are formed by gas-to-particle conversion in the atmosphere (Pöschl, 2005). Concentration, composition and size distribution of atmospheric aerosol particles are temporally and spatially highly variable. The predominant particle components of air particulate matter (PM) are sulfate, nitrate, ammonium, sea salt, mineral dust, organic compounds and black or elemental carbon (Pöschl, 2005). The effects of aerosols to the atmosphere, climate and public health are among the central topics in the current environmental research. An airborne particulate matter consists of several of inorganic and organic species, many of which can adversely affect human health (Pöschl, 2005; Mauderly

PM10 and Its Chemical Composition: A Case Study in Chiang Mai, Thailand 207

metals present and their concentrations and the presence of soluble ligands in the surrounding fluids (Valerie et al., 2003). The metals that are of high environmental priority,,due to their toxicity from char and ash samples of biomass, could be released to the atmosphere when biomass is burnt. The elemental compositions of biomass are complex. They involve 6 major elements (C, Cl, H, N, O and S) in the organic phase and at least 10 other elements (Al, Ca, Fe, K, Mg, Na, P, Si, Sr and Ti) in the inorganic phase called ash elements, which is important to ash characterization. Heavy metals (As, Cd, Co, Cr, Hg, Ni,

In attempt to identify the major emission sources responsible for adverse health effects, several comprehensive surveys of atmospheric contaminants from a variety of sources have been performed worldwide (Furton & Pentzke, 1998; Pengchai, et al., 2009). In Thailand, only a few studies have been conducted to determine atmospheric PAHs in big cities, such as Bangkok and Chiang Mai. In particular, the northern part of Thailand has been facing air pollution in the dry season almost every year. It is important to know the levels of air pollutants, including PAHs and their variations, throughout the year. Moreover, sources of

Ambient PM10 concentrations have been automatically measured at the Air Quality Monitoring (AQM) station set up by the Pollution Control Department (PCD) of Thailand. They were measured with the tapered element oscillating microbalance (TEOM) Series 1400a (Rupprecht & Patashnick, USA). TEOM mass detectors or microbalances utilize an inertial mass weighing principle. A TEOM detector consists of a substrate (usually a filter cartridge) placed on the end of a hollow tapered tube. The tube with the filter on the free end was oscillated in a clamped-free mode at its resonant frequency. This frequency depends on the physical characteristics of the tube and the mass on its free end. A particle laden air steam was drawn through the filter where the particles were deposited and then through the hollow tube. The frequency of oscillation was measured and recorded by the microprocessor; the change in frequency was used to calculate the mass of particulate matter deposited on the filter (Teflon coated with glass fiber filter surface). The air-flow rate at

was sampled through the sampling head and divided between the filter flow

mass transducer, which contained the analyte fraction of the particulate. The inlet was heated to 50°C prior to the particles being deposited onto the filter in order to eliminate the effect of condensation or evaporation of particle water (Washington State Department of

Twenty-four hour PM10 concentrations from 2005-2009 monitored at Chiang Mai AQM station located at the city center were plotted and are illustrated in Fig.1. The trend of PM10 concentrations are almost the same every year. They were high in the dry season especially in February and March. The level decreased in the middle of April due to rain precipitation. During the first 2 weeks of March 2007, air pollution levels in Chiang Mai and the surrounding provinces rose steadily above the safety limit and produced haze that significantly cut visibility down to less than 1 km. PM10 reached their peak concentration on

). The filter flow was sent to the instrument's

Pb, Sb and Se) in both phases of biomass are trace levels (Demirbas et al., 2008).

air pollutants are also needed to be identified.

16.67 L min-

1

(3 L min-

1

Ecology, 2004; Patashnick et al., 2002).

**2. PM10 situation in the city of Chiang Mai** 

) and an auxiliary flow (13.67 L min-<sup>1</sup>

& Chow, 2008). Of these constituents, the classes of polycyclic aromatic hydrocarbons (PAHs) are widespread environmental pollutants which are formed during the incomplete combustion process of organic material emitted from a large variety of industrial processes, motor vehicles, domestic waste burning and regenerated burning in agriculture (Lee, et al., 1995). PAHs have been measured in many media (air, water, food, and soil samples) because some of them are known to be carcinogenic. Specifically prepared foods, such as charcoal broiled or smoked meat, contain higher concentrations of PAHs than city air. Nevertheless, the daily air intake of humans (10–25 m3) is of comparable mass to the daily intake of food and water (2–4 kg) (Furton & Pentzke, 1998). Therefore, the air polluted by airborne PAHs is of significant concern.

Numerous particulates and gaseous compounds (e.g. CO and volatile organic compounds) that come from biomass burning are known to be hazardous to human health (Torigoe et al., 2000). Mass concentrations of CO, NO2, PM2.5, organic carbon (OC) and elemental carbon (EC) in PM2.5 were significantly associated with emergency department visits at hospitals due to cardiovascular diseases (Metzger et al., 2004). The most important gasses which have affected the acidic dry deposition are HNO3, HCl and H2SO4. Those gasses can be transformed into aerosols by neutralization reactions (Seinfeld & Pandis, 1998). Semivolatile NH4NO3 and NH4Cl are formed via reversible phase equilibrium with NH3, HNO3 and HCl (Pio & Harrison, 1987). Understanding the formation of secondary particulate species SO42-, NO3- and NH4+ formation, requires investigation of the atmospheric concentration levels for HNO3, HCl and SO2 and related aerosol ions. The inorganic component is made up of some insoluble dust and ash material, and soluble salts, in which potassium, ammonium, sulfate and nitrate are the most important species (Fuzzi et al., 2007).

Toxic metals in air are considered a global problem and a growing threat to the environment. Heavy metals are part of a large group of air pollutants called air toxins, which upon inhalation or ingestion can be responsible for a range of health effects such as cancer, neurotoxicity, immunotoxicity, cardiotoxicity, reproductive toxicity, teratogenesis and genotoxicity. For most metals, the bound or complex species are less toxic than the elemental forms, hence there is a loss in specificity when they are broadly classified as metals and their compounds. When inhaled, very small particles containing metals or their compounds deposit beyond the bronchial regions of the lungs into the alveoli region. Epidemiological studies have established relationships between inhaled PM and morbidity or mortality in populations. Studies in occupational or community settings have established the health effects of exposure to heavy metals, such as lead and their compounds. The localised release of some heavy metals from inhaled PM have been hypothesised to be responsible for lung tissue damage (U.S. EPA, 1999c). Metals occur in air in different phases as solids, gases or adsorbed to particles having aerodynamic sizes ranging from below 0.01 to 100 µm and larger. Two major categories of PM are fine particles (PM2.5) and coarse particles (PM10). Heavy metals such as As, Ga, Eu, W, Au, Cd, Ga, Mo, Pb, Sb, W, Zn, Fe, V, Cr, Co, Ni, Mn, Cu and Se exist in both coarse and fine fractions in ambient air. Ca, Al, Ti, Mg, Sc, La, Hf and Th exist predominantly in the coarse fraction. Metals such as Ba, Cs and Se enrich the fine fraction of PM. The degree to which the metals are absorbed depends on the properties of the metals concerned, i.e. valence, radius, degree of hydration and coordination with oxygen, physicochemical environment, nature of the adsorbent, other

& Chow, 2008). Of these constituents, the classes of polycyclic aromatic hydrocarbons (PAHs) are widespread environmental pollutants which are formed during the incomplete combustion process of organic material emitted from a large variety of industrial processes, motor vehicles, domestic waste burning and regenerated burning in agriculture (Lee, et al., 1995). PAHs have been measured in many media (air, water, food, and soil samples) because some of them are known to be carcinogenic. Specifically prepared foods, such as charcoal broiled or smoked meat, contain higher concentrations of PAHs than city air. Nevertheless, the daily air intake of humans (10–25 m3) is of comparable mass to the daily intake of food and water (2–4 kg) (Furton & Pentzke, 1998). Therefore, the air polluted by

Numerous particulates and gaseous compounds (e.g. CO and volatile organic compounds) that come from biomass burning are known to be hazardous to human health (Torigoe et al., 2000). Mass concentrations of CO, NO2, PM2.5, organic carbon (OC) and elemental carbon (EC) in PM2.5 were significantly associated with emergency department visits at hospitals due to cardiovascular diseases (Metzger et al., 2004). The most important gasses which have affected the acidic dry deposition are HNO3, HCl and H2SO4. Those gasses can be transformed into aerosols by neutralization reactions (Seinfeld & Pandis, 1998). Semivolatile NH4NO3 and NH4Cl are formed via reversible phase equilibrium with NH3, HNO3 and HCl (Pio & Harrison, 1987). Understanding the formation of secondary particulate species SO42-, NO3- and NH4+ formation, requires investigation of the atmospheric concentration levels for HNO3, HCl and SO2 and related aerosol ions. The inorganic component is made up of some insoluble dust and ash material, and soluble salts, in which potassium, ammonium, sulfate and nitrate are the most important species (Fuzzi et al.,

Toxic metals in air are considered a global problem and a growing threat to the environment. Heavy metals are part of a large group of air pollutants called air toxins, which upon inhalation or ingestion can be responsible for a range of health effects such as cancer, neurotoxicity, immunotoxicity, cardiotoxicity, reproductive toxicity, teratogenesis and genotoxicity. For most metals, the bound or complex species are less toxic than the elemental forms, hence there is a loss in specificity when they are broadly classified as metals and their compounds. When inhaled, very small particles containing metals or their compounds deposit beyond the bronchial regions of the lungs into the alveoli region. Epidemiological studies have established relationships between inhaled PM and morbidity or mortality in populations. Studies in occupational or community settings have established the health effects of exposure to heavy metals, such as lead and their compounds. The localised release of some heavy metals from inhaled PM have been hypothesised to be responsible for lung tissue damage (U.S. EPA, 1999c). Metals occur in air in different phases as solids, gases or adsorbed to particles having aerodynamic sizes ranging from below 0.01 to 100 µm and larger. Two major categories of PM are fine particles (PM2.5) and coarse particles (PM10). Heavy metals such as As, Ga, Eu, W, Au, Cd, Ga, Mo, Pb, Sb, W, Zn, Fe, V, Cr, Co, Ni, Mn, Cu and Se exist in both coarse and fine fractions in ambient air. Ca, Al, Ti, Mg, Sc, La, Hf and Th exist predominantly in the coarse fraction. Metals such as Ba, Cs and Se enrich the fine fraction of PM. The degree to which the metals are absorbed depends on the properties of the metals concerned, i.e. valence, radius, degree of hydration and coordination with oxygen, physicochemical environment, nature of the adsorbent, other

airborne PAHs is of significant concern.

2007).

metals present and their concentrations and the presence of soluble ligands in the surrounding fluids (Valerie et al., 2003). The metals that are of high environmental priority,,due to their toxicity from char and ash samples of biomass, could be released to the atmosphere when biomass is burnt. The elemental compositions of biomass are complex. They involve 6 major elements (C, Cl, H, N, O and S) in the organic phase and at least 10 other elements (Al, Ca, Fe, K, Mg, Na, P, Si, Sr and Ti) in the inorganic phase called ash elements, which is important to ash characterization. Heavy metals (As, Cd, Co, Cr, Hg, Ni, Pb, Sb and Se) in both phases of biomass are trace levels (Demirbas et al., 2008).

In attempt to identify the major emission sources responsible for adverse health effects, several comprehensive surveys of atmospheric contaminants from a variety of sources have been performed worldwide (Furton & Pentzke, 1998; Pengchai, et al., 2009). In Thailand, only a few studies have been conducted to determine atmospheric PAHs in big cities, such as Bangkok and Chiang Mai. In particular, the northern part of Thailand has been facing air pollution in the dry season almost every year. It is important to know the levels of air pollutants, including PAHs and their variations, throughout the year. Moreover, sources of air pollutants are also needed to be identified.

#### **2. PM10 situation in the city of Chiang Mai**

Ambient PM10 concentrations have been automatically measured at the Air Quality Monitoring (AQM) station set up by the Pollution Control Department (PCD) of Thailand. They were measured with the tapered element oscillating microbalance (TEOM) Series 1400a (Rupprecht & Patashnick, USA). TEOM mass detectors or microbalances utilize an inertial mass weighing principle. A TEOM detector consists of a substrate (usually a filter cartridge) placed on the end of a hollow tapered tube. The tube with the filter on the free end was oscillated in a clamped-free mode at its resonant frequency. This frequency depends on the physical characteristics of the tube and the mass on its free end. A particle laden air steam was drawn through the filter where the particles were deposited and then through the hollow tube. The frequency of oscillation was measured and recorded by the microprocessor; the change in frequency was used to calculate the mass of particulate matter deposited on the filter (Teflon coated with glass fiber filter surface). The air-flow rate at 16.67 L min-<sup>1</sup> was sampled through the sampling head and divided between the filter flow (3 L min-1 ) and an auxiliary flow (13.67 L min-<sup>1</sup> ). The filter flow was sent to the instrument's mass transducer, which contained the analyte fraction of the particulate. The inlet was heated to 50°C prior to the particles being deposited onto the filter in order to eliminate the effect of condensation or evaporation of particle water (Washington State Department of Ecology, 2004; Patashnick et al., 2002).

Twenty-four hour PM10 concentrations from 2005-2009 monitored at Chiang Mai AQM station located at the city center were plotted and are illustrated in Fig.1. The trend of PM10 concentrations are almost the same every year. They were high in the dry season especially in February and March. The level decreased in the middle of April due to rain precipitation. During the first 2 weeks of March 2007, air pollution levels in Chiang Mai and the surrounding provinces rose steadily above the safety limit and produced haze that significantly cut visibility down to less than 1 km. PM10 reached their peak concentration on

PM10 and Its Chemical Composition: A Case Study in Chiang Mai, Thailand 209

other anthropogenic sources, such as traffic, etc. Due to the fact that the station at CMU was on the rooftop of high building, it had a minimized influence from traffic emissions. Therefore, during the non-burning season PM10 concentrations at this sampling site were

Dry season Rainy season Dry season

Fig. 2. PM10 concentrations (µg/m3) at Chiang Mai University (CMU), Provincial Hall (PH)

PH

Preci

pitation (mm)

This study aims to investigate PM10 concentrations and PM10 chemical composition in the ambient air of the cities located in the Chiang Mai-Lamphun Basin in order to assess the air quality and predict possible health effects. Three study sites were located in Chiang Mai Province: Yuparaj Witayalai School (YP), Municipality Hospital (HP), Sarapee District (SP) and one site was located in Muang District, Lamphun Province (LP). A map of the sampling sites is illustrated in Fig. 3. YP site is located at the centre of Chiang Mai City. It is a residential area surrounded by government buildings, schools and temples with quite high traffic density. HP site is a densely populated commercial area. It has been reported to possess particulate concentrations exceeding the standard throughout the year (Pengchai, et al., 2009 as cited in Vinitketkumnuen et al., 2002). SP site is surrounded by government buildings and is considered an area experiencing dense traffic within the countryside. There have been significant reports of people suffering from respiratory disease (Pengchai, et al., 2009 as cited in Vinitketkumnuen et al., 2002). LP site is located in a residential area at the

and Yupparaj Wittayalai School (YP) and precipitation (mm) from 2010-2011

**3. Sampling site selection for the study of PM10 and its chemical** 

**composition** 

PM10 (µ

g/m3)

centre of Lamphun with light traffic density.

lower than the other two sites, where increased human activities existed.

March 14th 2007at 383 μg/m3 three times higher than 24 hrs PM10 standard of Thailand (120 μg/m3). About 500,000 people were reported by the Public Health Ministry to be affected by the heavy air pollution. Hospitals and clinics across the affected area reported a surge in the number of patients with respiratory problems in March 2007, approximately a 20% increase compared to the same period in 2006 (Pengchai, et al., 2009).

Fig. 1. PM10 concentrations in Chiang Mai's ambient air from 2005-2009.

PM10 concentrations and rain precipitation from late February 2010 to the beginning of March 2011 are illustrated in Fig. 2. PM10 samples were collected by mini volume air samplers (Airmetric, USA) at the rooftop of a 9-storey building (ScB1), Faculty of Science, Chiang Mai University (CMU). This station was set up to monitor ambient PM10 variation by avoiding the direct effects of traffic emissions. Another two PM10 data series were obtained from the AQM Station belonging to PCD located at Provincial Hall (PH) and Yupparaj Wittayalai School (YP). Regional area air quality data was downloaded from the PCD website (Pollution Control Department, 2011). All mentioned sampling sites are shown in Fig. 3. Data of rain precipitation was obtained from website of Chiang Mai meteorological station (http://www.cmmet.tmd.go.th/index1.php).

In the dry season of the year 2010, PM10 concentrations obtained from those 3 stations were much closer to each other. This is probably due to the effect from the large scale of open burning, which is mostly performed in the dry season. In the rainy season of the year 2010 and at the beginning of the dry season in the year 2011, PM10 in the air was generated from

March 14th 2007at 383 μg/m3 three times higher than 24 hrs PM10 standard of Thailand (120 μg/m3). About 500,000 people were reported by the Public Health Ministry to be affected by the heavy air pollution. Hospitals and clinics across the affected area reported a surge in the number of patients with respiratory problems in March 2007, approximately a

20% increase compared to the same period in 2006 (Pengchai, et al., 2009).

Fig. 1. PM10 concentrations in Chiang Mai's ambient air from 2005-2009.

station (http://www.cmmet.tmd.go.th/index1.php).

PM10 concentrations and rain precipitation from late February 2010 to the beginning of March 2011 are illustrated in Fig. 2. PM10 samples were collected by mini volume air samplers (Airmetric, USA) at the rooftop of a 9-storey building (ScB1), Faculty of Science, Chiang Mai University (CMU). This station was set up to monitor ambient PM10 variation by avoiding the direct effects of traffic emissions. Another two PM10 data series were obtained from the AQM Station belonging to PCD located at Provincial Hall (PH) and Yupparaj Wittayalai School (YP). Regional area air quality data was downloaded from the PCD website (Pollution Control Department, 2011). All mentioned sampling sites are shown in Fig. 3. Data of rain precipitation was obtained from website of Chiang Mai meteorological

In the dry season of the year 2010, PM10 concentrations obtained from those 3 stations were much closer to each other. This is probably due to the effect from the large scale of open burning, which is mostly performed in the dry season. In the rainy season of the year 2010 and at the beginning of the dry season in the year 2011, PM10 in the air was generated from other anthropogenic sources, such as traffic, etc. Due to the fact that the station at CMU was on the rooftop of high building, it had a minimized influence from traffic emissions. Therefore, during the non-burning season PM10 concentrations at this sampling site were lower than the other two sites, where increased human activities existed.

Fig. 2. PM10 concentrations (µg/m3) at Chiang Mai University (CMU), Provincial Hall (PH) and Yupparaj Wittayalai School (YP) and precipitation (mm) from 2010-2011

#### **3. Sampling site selection for the study of PM10 and its chemical composition**

This study aims to investigate PM10 concentrations and PM10 chemical composition in the ambient air of the cities located in the Chiang Mai-Lamphun Basin in order to assess the air quality and predict possible health effects. Three study sites were located in Chiang Mai Province: Yuparaj Witayalai School (YP), Municipality Hospital (HP), Sarapee District (SP) and one site was located in Muang District, Lamphun Province (LP). A map of the sampling sites is illustrated in Fig. 3. YP site is located at the centre of Chiang Mai City. It is a residential area surrounded by government buildings, schools and temples with quite high traffic density. HP site is a densely populated commercial area. It has been reported to possess particulate concentrations exceeding the standard throughout the year (Pengchai, et al., 2009 as cited in Vinitketkumnuen et al., 2002). SP site is surrounded by government buildings and is considered an area experiencing dense traffic within the countryside. There have been significant reports of people suffering from respiratory disease (Pengchai, et al., 2009 as cited in Vinitketkumnuen et al., 2002). LP site is located in a residential area at the centre of Lamphun with light traffic density.

PM10 and Its Chemical Composition: A Case Study in Chiang Mai, Thailand 211

mean concentration of the Sarapee (SP) site was higher than those of the other sites

Fig. 4. PM10 concentrations of all sampling stations from June 2005 – June 2006

concentration

b Groups of number in *italic* and/or **bold** are not significant different (*p* > 0.05)

Dry: Dec – Mar (n = 160) 73.64 25.34 /7 LP (n=127) 46.53 29.30 /1

Trans II: Oct – Nov (n = 80) **47.19 12.36 /2** YP (n=127) **50.30 23.26 /6**

The post-weighed quartz fiber paper (8 x 10 inches) of the PM10 sample was equally divided into eight pieces (2.5 x 4 inches) by a stainless steel roller blade cutter. Two of the eight pieces were grouped for the analysis of carbon, PAHs, soluble ions and elements. All fiber samples, except the samples for carbon analysis, were cut into small pieces by stainless

Table 1. Mean PM10 concentrations (g/m3) SD for 4 periods and 4 sampling sites

Sampling sitea,b PM10 concentration

 *25.33 /2* SP (n=127) **55.94 36.64 /2**

 *21.39 /1* HP (n=127) 44.08 20.60 /3

Seasona,b PM10

Trans I: Apr – May (n = 84) *40.44* 

Rainy: Jun – Sep (n = 184) *33.17* 

a/n number of non detected or incomplete data

steel scissors prior to analysis.

**4. Analysis of PM10 chemical composition** 

(Chantara et al., 2009).

Fig. 3. Map of sampling sites

#### **3.1 Ambient PM10 concentrations**

Twenty-four hour PM10 samples were collected on quartz fiber filters every 3 days at each sampling site using a High Volume PM10 Air Sampler (Wedding & Associates Inc., USA) at the flow rate of 1,130 L/min. Sampling had been done for 1 year from June 2005 to June 2006 to cover 3 seasonal periods, which were the wet season (June - September), the dry season (December – March) and the transition periods (October – November and April - May). PM10 samples were weighed using the five digit-scale balance (Mettler Toledo AG285, USA) covered by the temperature and humidity controlling cabinet (DE-300).

PM10 concentrations of all sampling stations from June 2005 – June 2006 are illustrated in Fig. 4. Their concentrations for all seasons were almost the same. They were lower than 80 µg/m3 in the rainy season and getting higher in the dry season. The highest concentration was found in March (150 µg/m3) at all three sampling sites except at the LP site, which was uncommonly high in April. This was probabaly due to local activities, such as open burning of community garbage.

Seasonal and spatial variations of mean PM10 concentrations and their standard deviation (SD) are illustrated in Table 1. It was found that concentrations of PM10 collected in the dry season (Dec – Mar) were significantly higher than those of the other seasons, while the lowest concentration was found in the rainy season (Jun – Sep). Average PM10 concentrations in each sampling site were found to be very much similar. However, the

Twenty-four hour PM10 samples were collected on quartz fiber filters every 3 days at each sampling site using a High Volume PM10 Air Sampler (Wedding & Associates Inc., USA) at the flow rate of 1,130 L/min. Sampling had been done for 1 year from June 2005 to June 2006 to cover 3 seasonal periods, which were the wet season (June - September), the dry season (December – March) and the transition periods (October – November and April - May). PM10 samples were weighed using the five digit-scale balance (Mettler Toledo AG285, USA)

PM10 concentrations of all sampling stations from June 2005 – June 2006 are illustrated in Fig. 4. Their concentrations for all seasons were almost the same. They were lower than 80 µg/m3 in the rainy season and getting higher in the dry season. The highest concentration was found in March (150 µg/m3) at all three sampling sites except at the LP site, which was uncommonly high in April. This was probabaly due to local activities, such as open burning

Seasonal and spatial variations of mean PM10 concentrations and their standard deviation (SD) are illustrated in Table 1. It was found that concentrations of PM10 collected in the dry season (Dec – Mar) were significantly higher than those of the other seasons, while the lowest concentration was found in the rainy season (Jun – Sep). Average PM10 concentrations in each sampling site were found to be very much similar. However, the

covered by the temperature and humidity controlling cabinet (DE-300).

Fig. 3. Map of sampling sites

of community garbage.

**3.1 Ambient PM10 concentrations** 

mean concentration of the Sarapee (SP) site was higher than those of the other sites (Chantara et al., 2009).

Fig. 4. PM10 concentrations of all sampling stations from June 2005 – June 2006


a/n number of non detected or incomplete data

b Groups of number in *italic* and/or **bold** are not significant different (*p* > 0.05)

Table 1. Mean PM10 concentrations (g/m3) SD for 4 periods and 4 sampling sites

#### **4. Analysis of PM10 chemical composition**

The post-weighed quartz fiber paper (8 x 10 inches) of the PM10 sample was equally divided into eight pieces (2.5 x 4 inches) by a stainless steel roller blade cutter. Two of the eight pieces were grouped for the analysis of carbon, PAHs, soluble ions and elements. All fiber samples, except the samples for carbon analysis, were cut into small pieces by stainless steel scissors prior to analysis.

PM10 and Its Chemical Composition: A Case Study in Chiang Mai, Thailand 213

than those found in the PM10 samples in this study. This is probably due to a much higher

LP 0.861 Dry (Dec-Mar) 0.849 SP 0.682 Transition (Apr-May) 0.836 HP 0.784 Rainy (Jun-Sep) 0.208 YP 0.798 Transition (Oct-Nov) 0.445

Gas Chromatography – Mass Spectrometry (GC-MS) (Hewlett Packard, USA) was used for investigation of PAHs in this study. One micro liter of a mixed solution of 16-PAH standards (2,000 μg/ml in methylenechloride, Restek, USA) and internal standards (D10- ACE, D10-PHE and D12-PYR from Supelco, USA) was injected in splitless mode onto an HP-5MS GC capillary column of 30 m length, 0.25 mm diameter and 0.25 μm thickness internally coated with 5% phenyl methyl polysiloxane stationary phase. Standard solutions of the 16 PAHs with concentrations ranging from 0.005 to 0.500 μg/ml were combined with 0.10 μg/ml internal standards and were injected onto the same GC capillary column of the GC-MS employed under an optimum condition in order to obtain a suitable calibration curve in which peak area ratios were plotted against concentrations. The GC condition was as follows. The injector temperature was 275°C. The GC oven was programmed with an initial temperature of 70°C, held for 2 min, then increased to 290°C at 8°C/min. The MS was operated in SIM mode. Characteristic ions of selected PAHs were used for quantification, which was based on the peak area of the PAH standard relative to the deuterated internal standard closest in molecular weight to the analyte. Detection limits of GC-MS for PAHs and recoveries of PAHs from extraction are illustrated in Table 3. Detection limit of GC-MS for an individual PAH was calculated as three times of the standard deviation of blank concentration (Miller & Miller, 1993). Detection limits for 16-PAHs obtained were ranged from 0.003-0.007 ng/m3. Recoveries of PAHs from the extraction of spiked filters ranged from 58-120%. Of all the 16 PAHs investigated, 12 PAHs were found to yield the recovery

Optimization for the extraction procedure of PAHs in PM 10 samples was conducted using ultrasonication to obtain an appropriate extraction volume of acetonitrile (HPLC grade, Lab-Scan Analytical Science) and evaporation conditions. The quartz filter was cut into small pieces using a pair of stainless steel scissors and the cut filter was put into a 60-ml amberbottle wrapped with aluminium foil and mounted by paraffin film. The 0.3 μg/ml of mixed 16-PAH standards was spiked onto the paper, covered and left for 10 min. The spiked samples were extracted in 35 ml acetonitrile by ultrasonicator (T710DH, Elma, Germany) at 100% ultrasound power for 30 min under controlled temperature (approximately 20°C). The

**Sampling site r Season r** 

number of carbon sources, especially in terms of the traffic in Bangkok.

Table 2. Correlation of carbon (µg/m3) and PM10 weight (mg)

**4.2.1 Validation of analysis method for PM10 bound PAHs using GC-MS** 

**4.2 PM10-bound PAHs** 

higher than 70%.

#### **4.1 Analysis of total carbon in PM10 samples**

#### **4.1.1 Analysis method of PM10 carbon content by CHN-S/O analyzer**

A tin capsule (1 x 1 cm) as a sample container was weighed by MX5 Automated-S Microbalance (Mettler Toledo). The quartz filter contains PM10 was punched by a puncher with a diameter of 0.55 cm. The punched filter was put into a prepared capsule, wrapped and weighed. The sample was then analysed for total carbon (TC) by Elemental Analyzer (PE 2400 series II CHN-S/O analyzer, Perkin Elmer Cooperation) (Perkin Elmer, 1991) using L-cystine and BS Slag 2 as standards and an unused quartz filter paper as a blank.

#### **4.1.2 Carbon content of PM10 samples**

Spatial and temporal variations of mean carbon concentrations are shown in Fig. 5. It was found that the distribution pattern of carbon contents in all stations were almost the same. However, the PM10 samples collected at the SP station had higher carbon content than the others. A high concentration of carbon was found in the dry season, especially in March, while in the rainy season the carbon content was low. A correlation of carbon content with PM10 weight (Table 2) revealed that they were strongly correlated in the dry season and the transition period (Apr-May) for all sampling sites. It can be concluded that carbon content was positively correlated to PM10 weight. However, this consumption relationship can not be applied to the rainy season, in which the correlation was low. It can be explained that low concentrations of PM10 and carbon in the rainy season gave a relatively high variation of values. Moreover, ratios of carbon and PM10 (µg/mg PM10) at all sites were higher in the rainy season than the other seasons. Therefore a correlation was not obviously observed.

Fig. 5. Spatial and temporal variations of mean carbon concentrations

Carbon concentrations found in this study were compared with the study in Bangkok (Chuersuwan et. al., 2008), in which carbon content in PM2.5 was determined. The carbon concentrations were 38.5±19.3, 21.7±12.7, 21.9±13.3 and 17.6±11.0 µg/m3 in Din Daeng, Chankasaem, Ban Somdej and Bangna, respectively. The contents were 4-8 times higher


than those found in the PM10 samples in this study. This is probably due to a much higher number of carbon sources, especially in terms of the traffic in Bangkok.

Table 2. Correlation of carbon (µg/m3) and PM10 weight (mg)

#### **4.2 PM10-bound PAHs**

212 Air Quality – Monitoring and Modeling

A tin capsule (1 x 1 cm) as a sample container was weighed by MX5 Automated-S Microbalance (Mettler Toledo). The quartz filter contains PM10 was punched by a puncher with a diameter of 0.55 cm. The punched filter was put into a prepared capsule, wrapped and weighed. The sample was then analysed for total carbon (TC) by Elemental Analyzer (PE 2400 series II CHN-S/O analyzer, Perkin Elmer Cooperation) (Perkin Elmer, 1991) using

Spatial and temporal variations of mean carbon concentrations are shown in Fig. 5. It was found that the distribution pattern of carbon contents in all stations were almost the same. However, the PM10 samples collected at the SP station had higher carbon content than the others. A high concentration of carbon was found in the dry season, especially in March, while in the rainy season the carbon content was low. A correlation of carbon content with PM10 weight (Table 2) revealed that they were strongly correlated in the dry season and the transition period (Apr-May) for all sampling sites. It can be concluded that carbon content was positively correlated to PM10 weight. However, this consumption relationship can not be applied to the rainy season, in which the correlation was low. It can be explained that low concentrations of PM10 and carbon in the rainy season gave a relatively high variation of values. Moreover, ratios of carbon and PM10 (µg/mg PM10) at all sites were higher in the rainy season than the other seasons. Therefore a correlation

**4.1.1 Analysis method of PM10 carbon content by CHN-S/O analyzer** 

Fig. 5. Spatial and temporal variations of mean carbon concentrations

Carbon concentrations found in this study were compared with the study in Bangkok (Chuersuwan et. al., 2008), in which carbon content in PM2.5 was determined. The carbon concentrations were 38.5±19.3, 21.7±12.7, 21.9±13.3 and 17.6±11.0 µg/m3 in Din Daeng, Chankasaem, Ban Somdej and Bangna, respectively. The contents were 4-8 times higher

L-cystine and BS Slag 2 as standards and an unused quartz filter paper as a blank.

**4.1 Analysis of total carbon in PM10 samples** 

**4.1.2 Carbon content of PM10 samples** 

was not obviously observed.

#### **4.2.1 Validation of analysis method for PM10 bound PAHs using GC-MS**

Gas Chromatography – Mass Spectrometry (GC-MS) (Hewlett Packard, USA) was used for investigation of PAHs in this study. One micro liter of a mixed solution of 16-PAH standards (2,000 μg/ml in methylenechloride, Restek, USA) and internal standards (D10- ACE, D10-PHE and D12-PYR from Supelco, USA) was injected in splitless mode onto an HP-5MS GC capillary column of 30 m length, 0.25 mm diameter and 0.25 μm thickness internally coated with 5% phenyl methyl polysiloxane stationary phase. Standard solutions of the 16 PAHs with concentrations ranging from 0.005 to 0.500 μg/ml were combined with 0.10 μg/ml internal standards and were injected onto the same GC capillary column of the GC-MS employed under an optimum condition in order to obtain a suitable calibration curve in which peak area ratios were plotted against concentrations. The GC condition was as follows. The injector temperature was 275°C. The GC oven was programmed with an initial temperature of 70°C, held for 2 min, then increased to 290°C at 8°C/min. The MS was operated in SIM mode. Characteristic ions of selected PAHs were used for quantification, which was based on the peak area of the PAH standard relative to the deuterated internal standard closest in molecular weight to the analyte. Detection limits of GC-MS for PAHs and recoveries of PAHs from extraction are illustrated in Table 3. Detection limit of GC-MS for an individual PAH was calculated as three times of the standard deviation of blank concentration (Miller & Miller, 1993). Detection limits for 16-PAHs obtained were ranged from 0.003-0.007 ng/m3. Recoveries of PAHs from the extraction of spiked filters ranged from 58-120%. Of all the 16 PAHs investigated, 12 PAHs were found to yield the recovery higher than 70%.

Optimization for the extraction procedure of PAHs in PM 10 samples was conducted using ultrasonication to obtain an appropriate extraction volume of acetonitrile (HPLC grade, Lab-Scan Analytical Science) and evaporation conditions. The quartz filter was cut into small pieces using a pair of stainless steel scissors and the cut filter was put into a 60-ml amberbottle wrapped with aluminium foil and mounted by paraffin film. The 0.3 μg/ml of mixed 16-PAH standards was spiked onto the paper, covered and left for 10 min. The spiked samples were extracted in 35 ml acetonitrile by ultrasonicator (T710DH, Elma, Germany) at 100% ultrasound power for 30 min under controlled temperature (approximately 20°C). The

0

PAHs\*†

0.05

0.1

0.15

**Bar (ng/m3**

**)**

0.2

0.25

0.3

PM10 and Its Chemical Composition: A Case Study in Chiang Mai, Thailand 215

 (n = 46) (n = 19) (n = 40) (n = 21) (n = 46) (n = 19) (n = 39) (n = 21) (n = 45) (n = 20) (n = 38) (n = 21) (n = 46) (n = 20) (n = 36) (n = 19) Wet Trans I Dry Trans II Wet Trans I Dry Trans II Wet Trans I Dry Trans II Wet Trans I Dry Trans II LP SP HP YP

Seasonal variation of PM10-bound PAHs can be observed in Table 4 . Mean concentrations of various groups of PAHs including those with 2-3 rings, 4-6 rings, carcinogenic and noncarcinogenic PAHs and sums of 16 PAHs were statistically tested based on the season basis. The results revealed that all PAH groups presented a similar pattern. The highest concentration was found in the dry season (Dec-Mar) and were significantly higher (*p*<0.05) than the other seasons, in which PAHs concentrations were not significantly different. In terms of the sampling site (Table 5), high molecular weight (4-6 rings) and carcinogenic PAHs were significantly higher at the SP site than the other sites (Chantara et al., 2010).

> Apr – May (n = 84)

2-3 rings **0.15 0.13 /1** 0.10 0.16 /2 0.07 0.04 /5 0.08 0.06 /4

Non carcinogenic **0.36 0.20 /1** 0.20 0.18 /2 0.13 0.06 /1 0.18 0.09 /2

Jun – Sep (n = 184)

 *2.96 /1 3.59*

 *2.42 /1 2.73*

 *2.97 /1 3.67* 

Fig. 6. Spatial and temporal variations of PAHs concentrations

Dec – Mar (n = 160)

4-6 rings **6.58 3.20 /1** 2.76 2.05 /2 *4.06* 

Carcinogenic‡ **5.29 2.72 /1** 2.05 1.65 /2 *3.15* 

Total 16 PAHs **6.72 3.20 /1** 2.87 2.08 /2 *4.13* 

‡ Carcinogenic PAHs are BaA, CHR, BbF, BkF, BaP, DBA and IND

† Groups of number in *italic*, underline and/or **bold** are not significant different (*p* > 0.05)

Table 4. Mean concentrations (ng/m3) SD of PM10-bound PAHs in the four seasons

\* /n number of non detected or incomplete data

0

Oct – Nov (n = 80)

> *1.69 /2*

> *1.41 /2*

> *1.71 /2*

0.5

1

1.5

**Line (ng/m3**

**)**

NAP ACY ACE FLU PHE ANT FLA PYR BaA CHR BbF BkF BaP IND DbA BPER

2

2.5

solution was filtered through a 0.45 μm nylon filter and evaporated by low-pressure evaporator at 30°C until it became nearly dry. The pellet was re-extracted in 30 ml acetronitrile by repeating the previous step. The solution was transferred into a 1 ml volumetric flask and a mixed solution of internal standards was added, with the final volume adjusted to 1 ml with acetronitrile. The solution was then analyzed for 16 PAHs by GC-MS. Percent recoveries of individual PAHs were obtained.


\* carcinogen suspects

Table 3. Detection limits (DL) of GC-MS for PAHs and recoveries of PAHs from extraction

#### **4.2.2 Analysis of 16 PAHs in PM10 samples by GC-MS**

After sampling, the weighed quartz filter was cut into small pieces by a pair of stainless steel scissors and put into a 60-ml amber-bottle wrapped with aluminium foil and mounted with paraffin film. The samples were extracted using the same conditions as the standards described above. Each of the sample solutions was then analyzed for 16 PAHs by GC-MS.

Fig. 6 illustrates spatial and temporal variations of PAHs concentrations (Chantara et al., 2010). The highest concentrations of an individual and total PAHs, as well as carcinogenic PAHs, were found in the dry season and were higher than those found in other seasons.

214 Air Quality – Monitoring and Modeling

solution was filtered through a 0.45 μm nylon filter and evaporated by low-pressure evaporator at 30°C until it became nearly dry. The pellet was re-extracted in 30 ml acetronitrile by repeating the previous step. The solution was transferred into a 1 ml volumetric flask and a mixed solution of internal standards was added, with the final volume adjusted to 1 ml with acetronitrile. The solution was then analyzed for 16 PAHs by

Naphthalene (NAP) 2 0.006 64±5 Acenaphthylene (ACY) 3 0.005 76±6 Acenaphthene (ACE) 3 0.006 66±6 Fluorene (FLU) 3 0.004 80±7 Phenanthrene (PHE) 3 0.006 60±1 Anthracene (ANT) 3 0.004 77±2 Fluoranthene (FLA) 4 0.003 76±2 Pyrene (PYR) 4 0.003 72±1

Benzo(a)anthracene\* (BaA) 4 0.007 120±4

Benzo(b)fluoranthene\* (BbF) 5 0.006 112±7 Benzo(k)fluoranthene\* (BkF) 5 0.007 84±4 Benzo(a)pyrene\* (BaP) 5 0.005 81±12 Indeno(1,2,3-cd)pyrene\* (IND) 5 0.004 73±14 Dibenzo(a,h)anthracene\* (DBA) 6 0.005 77±16 Benzo(g,h,i)perylene (BPER) 6 0.007 58±15

Chrysene\* (CHR) 4 0.005 78±2

Table 3. Detection limits (DL) of GC-MS for PAHs and recoveries of PAHs from extraction

After sampling, the weighed quartz filter was cut into small pieces by a pair of stainless steel scissors and put into a 60-ml amber-bottle wrapped with aluminium foil and mounted with paraffin film. The samples were extracted using the same conditions as the standards described above. Each of the sample solutions was then analyzed for 16 PAHs by GC-MS.

Fig. 6 illustrates spatial and temporal variations of PAHs concentrations (Chantara et al., 2010). The highest concentrations of an individual and total PAHs, as well as carcinogenic PAHs, were found in the dry season and were higher than those found in other seasons.

of ring DL (ng/m3) Recovery±SD (%)

GC-MS. Percent recoveries of individual PAHs were obtained.

Compound Number

**4.2.2 Analysis of 16 PAHs in PM10 samples by GC-MS** 

\* carcinogen suspects

Fig. 6. Spatial and temporal variations of PAHs concentrations

Seasonal variation of PM10-bound PAHs can be observed in Table 4 . Mean concentrations of various groups of PAHs including those with 2-3 rings, 4-6 rings, carcinogenic and noncarcinogenic PAHs and sums of 16 PAHs were statistically tested based on the season basis. The results revealed that all PAH groups presented a similar pattern. The highest concentration was found in the dry season (Dec-Mar) and were significantly higher (*p*<0.05) than the other seasons, in which PAHs concentrations were not significantly different. In terms of the sampling site (Table 5), high molecular weight (4-6 rings) and carcinogenic PAHs were significantly higher at the SP site than the other sites (Chantara et al., 2010).


\* /n number of non detected or incomplete data

† Groups of number in *italic*, underline and/or **bold** are not significant different (*p* > 0.05)

‡ Carcinogenic PAHs are BaA, CHR, BbF, BkF, BaP, DBA and IND

Table 4. Mean concentrations (ng/m3) SD of PM10-bound PAHs in the four seasons

PM10 and Its Chemical Composition: A Case Study in Chiang Mai, Thailand 217

The carcinogenic ratios of all sites found in this study were higher than those found in roadsides of the inner part of Bangkok from November 02 to April 03 (Norramit et al., 2005) (Table 7). This reveals that our study sites presented higher concentrations of carcinogenic PAHs but lower PM10 contents in comparison to the Bangkok area. This should be noted

Chiang Mai Jun 05 – Jun 06 Roadside area 1.3-12.2 This study Lamphun Jun 05 – Jun 06 Roadside area 1.7-7.4 This study

> Community area Rural area

Roadside area

General area

Jun 1996 – Apr 1997 Road site 17.4 Kim Oanh et al.,

Bangkok Nov 02 – Apr 03 Roadsides 12.590.94 Norramit et al.,

a Sum of concentrations of 18 PAHs: PHE, ANT, 1-methyl PHE, 2-methyl PHE, 3-methyl PHE, 9-methyl PHE, FLA, PYR, BaA, CHR, BbF, BjF, BkF, BeP, BaP, IND, BPER and coronene. b Sum of concentrations of 11 PAHs: ANT, PHE, PYR, BaA, CHR, BbF, BkF, BaP, IND, DBA and BPER

Table 7. Comparison of mean PAH concentrations in Chiang Mai, Lamphun and Bangkok,

Correlations between PM10 content and total PAH concentrations for each sampling site in each season are shown in Table 8. Correlations between PM10 content and PAH concentrations for all seasons were found to be in the range 0.618 – 0.731. The highest correlation was found in the LP site, while the lowest one was in the YP site. The correlations for all sites were relatively high in the dry season and low in the wet season, respectively. The correlations for the SP and HP sites were high throughout the year; they were also high in the transition periods and were obviously higher than those for the LP and YP sites. This is probably due to local sources, which distributed PM10-bound PAHs into the atmosphere during the sampling time period. In terms of PAH source, both SP and HP sampling sites are surrounded by markets with high traffic volume and human activities. The difference between these two places is that the SP site is a place where small scale factories are situated. The fuels used in such factories could have contributed to the level of

(ng/m3)

7.6-16.6 3.9-9.1 2.7-8.4

7.2-10.0 10.1-28.1

> 50-53 11-12

Reference

Chantara & Sangchan, 2009

Boonyatumanond et al., 2007

2005

Chetwittayachan et al., 2002

2000

and of significant concern for air quality management.

Chiang Mai Jun – Nov 04 Traffic area

Bangkok a Nov 03 – Jan 04 Rural area

Bangkok b Mar 01 Roadside area

Pathumthani (40 km north of Bangkok)

Thailand

**4.2.3 Correlation of PM10 and PAHs** 

pollutants emitted to the air.

City Study period Location Mean


\* /n number of non detected or incomplete data

† Groups of number in *italic* and/or **bold** are not significant different (*p* > 0.05)

‡ Carcinogenic PAHs are BaA, CHR, BbF, BkF, BaP, DBA and IND

Table 5. Mean concentrations (ng/m3) SD of PM10-bound PAHs from four sampling sites

Table 6 illustrates the ratios of non-carcinogenic and carcinogenic PAHs found in PM10 samples from each sampling site, which were 1.0:3.7 (LP), 1.0:3.8 (SP), 1.0:2.7 (YP) and 1.0:2.8 (HP). SP and LP ratios revealed that carcinogenic PAHs were obviously higher than noncarcinogenic ones. The highest carcinogenic PAHs content was found at the SP site and was significantly higher than other sites (Table 5). The carcinogenic ratios of all sites found in this study were higher than those found along the roadsides of the inner part of Bangkok (1.0:2.5) from November 02 to April 03 (Norramit et al., 2005). This reveals that our study sites presented higher concentrations of carcinogenic PAHs but lower PM10 content in comparison to the Bangkok area. This fact should be noted and brought into the attention of relevant authorities for subsequent air quality management.


Table 6. Ratios of non-carcinogenic to carcinogenic PM10-bound PAHs

Mean total PAHs concentrations detected in this study were almost the same as values found in the community area of the previous study (Chantara & Sangchan, 2009) carried out in 2004 (Table 6). However, the concentrations found in this study were obviously lower than those found in Bangkok and Pathumthanee, even under the same category (roadside area) (Boonyatumanond et al., 2007; Norramit et al., 2005; Chetwittayachan et al., 2002; Kim Oanh et al., 2000), but the values were similar to the rural area (Boonyatumanond et al., 2007) and the general area (Chetwittayachan et al., 2002) as is shown in Table 7. This is attributed to the transport pattern and the traffic volume that exist near the sampling site.

Based on the ratios of non-carcinogenic and carcinogenic PAHs found in PM10 samples from each sampling site (Table 6), the highest carcinogenic PAHs content was found at the SP site.

Non carcinogenic 0.17 0.11 /1 *0.23*

† Groups of number in *italic* and/or **bold** are not significant different (*p* > 0.05)

‡ Carcinogenic PAHs are BaA, CHR, BbF, BkF, BaP, DBA and IND

relevant authorities for subsequent air quality management.

Sampling site % Non–carcinogenic

Bangkok

\* /n number of non detected or incomplete data

PAHs\*† LP (n=127) SP (n=127) HP (n=127) YP (n=127) 2-3 rings 0.07 0.06 /5 0.10 0.09 /3 0.10 0.08 /2 **0.14 0.17 /2** 4-6 rings 3.97 2.53 /1 **6.93 4.16 /2** 3.58 2.03 /1 3.82 1.81 /2 Carcinogenic‡ 3.18 2.12 /1 **5.57 3.51 /2** 2.71 1.61 /1 2.88 1.46 /2

Total 16 PAHs 4.04 2.55 /1 **7.03 4.18 /2** 3.68 2.06 /1 3.96 1.84 /2

Table 5. Mean concentrations (ng/m3) SD of PM10-bound PAHs from four sampling sites

Table 6 illustrates the ratios of non-carcinogenic and carcinogenic PAHs found in PM10 samples from each sampling site, which were 1.0:3.7 (LP), 1.0:3.8 (SP), 1.0:2.7 (YP) and 1.0:2.8 (HP). SP and LP ratios revealed that carcinogenic PAHs were obviously higher than noncarcinogenic ones. The highest carcinogenic PAHs content was found at the SP site and was significantly higher than other sites (Table 5). The carcinogenic ratios of all sites found in this study were higher than those found along the roadsides of the inner part of Bangkok (1.0:2.5) from November 02 to April 03 (Norramit et al., 2005). This reveals that our study sites presented higher concentrations of carcinogenic PAHs but lower PM10 content in comparison to the Bangkok area. This fact should be noted and brought into the attention of

> LP 21.3 78.7 1.0 : 3.7 SP 20.8 79.2 1.0 : 3.8 HP 26.4 73.6 1.0 : 2.8 YP 27.3 72.7 1.0 : 2.7

(Norramit et al., 2005) 28.6 71.4 1.0 : 2.5

Mean total PAHs concentrations detected in this study were almost the same as values found in the community area of the previous study (Chantara & Sangchan, 2009) carried out in 2004 (Table 6). However, the concentrations found in this study were obviously lower than those found in Bangkok and Pathumthanee, even under the same category (roadside area) (Boonyatumanond et al., 2007; Norramit et al., 2005; Chetwittayachan et al., 2002; Kim Oanh et al., 2000), but the values were similar to the rural area (Boonyatumanond et al., 2007) and the general area (Chetwittayachan et al., 2002) as is shown in Table 7. This is attributed to the transport pattern and the traffic volume that exist near the sampling site.

Based on the ratios of non-carcinogenic and carcinogenic PAHs found in PM10 samples from each sampling site (Table 6), the highest carcinogenic PAHs content was found at the SP site.

Table 6. Ratios of non-carcinogenic to carcinogenic PM10-bound PAHs

 *0.16 /2* **0.22 0.15 /1** *0.28* 

PAHs % Carcinogenic PAHs non-carcinogenic :

 *0.22 /2*

carcinogenic PAHs

The carcinogenic ratios of all sites found in this study were higher than those found in roadsides of the inner part of Bangkok from November 02 to April 03 (Norramit et al., 2005) (Table 7). This reveals that our study sites presented higher concentrations of carcinogenic PAHs but lower PM10 contents in comparison to the Bangkok area. This should be noted and of significant concern for air quality management.


a Sum of concentrations of 18 PAHs: PHE, ANT, 1-methyl PHE, 2-methyl PHE, 3-methyl PHE, 9-methyl PHE, FLA, PYR, BaA, CHR, BbF, BjF, BkF, BeP, BaP, IND, BPER and coronene. b Sum of concentrations of 11 PAHs: ANT, PHE, PYR, BaA, CHR, BbF, BkF, BaP, IND, DBA and BPER

Table 7. Comparison of mean PAH concentrations in Chiang Mai, Lamphun and Bangkok, Thailand

#### **4.2.3 Correlation of PM10 and PAHs**

Correlations between PM10 content and total PAH concentrations for each sampling site in each season are shown in Table 8. Correlations between PM10 content and PAH concentrations for all seasons were found to be in the range 0.618 – 0.731. The highest correlation was found in the LP site, while the lowest one was in the YP site. The correlations for all sites were relatively high in the dry season and low in the wet season, respectively. The correlations for the SP and HP sites were high throughout the year; they were also high in the transition periods and were obviously higher than those for the LP and YP sites. This is probably due to local sources, which distributed PM10-bound PAHs into the atmosphere during the sampling time period. In terms of PAH source, both SP and HP sampling sites are surrounded by markets with high traffic volume and human activities. The difference between these two places is that the SP site is a place where small scale factories are situated. The fuels used in such factories could have contributed to the level of pollutants emitted to the air.

PM10 and Its Chemical Composition: A Case Study in Chiang Mai, Thailand 219

Fig. 7. Monthly mean BaPE and PM10 concentrations from all sampling sites

Dianostic ratio can be used for the identification of possible emission sources. The diagnostic ratios of this study were calculated by separating sampling sites and months as shown in Table 9 (Chantara et al., 2010). The IND/(IND+BPER) ratio can be used to identify traffic sources. The value in a range from 0.35-0.70 indicates that the PAHs were emitted from diesel engines (Lui et. al, 2007 and Kavouras et. al, 2001). In this study, the average ratios of all sites were 0.50-0.56. Therefore, the PAHs found in this study could have originated from diesel engine emissions. The BaP/BPER ratio was also used for characterization of the PAHs sources. The low BaP/BPER ratio (<0.60) is an evidence of greater emission of BPER from traffic sources (Pandey et. al., 1999). The mean ratios of the 4 sampling sites were 0.56-0.80. The mean ratios of HP and YP sites were fitted with the low ratio value, therefore traffic might be the main source for these two sites. However, the ratio values in some months, especially in the dry season were greater than 0.60, which indicated some other PAH sources. The SP and LP sites had a higher mean ratio value (0.80 and 0.64, respectively) as well as monthly values, depicted as *italic* values in Table 9. This could indicate that the PAHs in these two sites were not only from the traffic but were also from some other sources on a similar basis described for the BaPE above. Apart from these two ratios, the BaP/(BaP + CHR) ratio was also considered. The ratio values of 0.57 and 0.89 could be related to diesel and gasoline exhaust, respectively (Khalili et al., 1995). The BaP/(BaP + CHR) ratio values obtained in this work were 0.68-0.77, which could be reckoned as somewhere between 0.57 and 0.89. Based on this ratio alone, it could not be concluded what the major source of PAHs was in the areas studied in this work. However, the IND/(IND+BPER) ratio of 0.50-0.56 obtained in this study fits within the range of 0.35-0.70 which suggests that the diesel engine source plays a more important role than the gasoline

**4.2.5 Diagnostic ratios** 

source.


Table 8. Pearson correlation of PM10 and PAHs

In comparison to previous studies, the correlation found in this study is almost the same as that reported by Chantara & Sangchan (2009) in which the mass of PM10 collected by a Minivol air sampler in traffic areas was found to be slightly correlated (R2 = 0.45) with the total PAH concentration of the solution obtained from extraction of the collected particulate. The afore-mentioned correlation was higher than the value in the study by Koyano et al. (1998) in which airborne particulate samples were collected from commercial and residential sites of Chiang Mai from July to December 1989 and their results revealed that there was not a large differnce in PAHs concentrations between different sampling sites. However, the airborne particulates and large amounts of rough particulates encountered in their study prompted them to conclude that there was wide spread air pollution throughout the city.

#### **4.2.4 BaP and BaP-equivalent carcinogenic power (BaPE)**

The limit value of BaP recommended by WHO is 1 ng/m3 due to its high carcinogenic property. Other PAHs such as BaA, BbF, BkF and DBA also have carcinogenic potential. To combine these potential carcinogenic compounds, the benzo(a)pyrene-equivalent carcinogenic power (BaPE) is calculated. The BaPE is an index that has been introduced for better denoting aerosol carcinogenicity related to the whole PAH fraction instead of the BaP (Liu et al., 2007). The BaPE was calculated in this study using the following equation:

$$\text{BaPE} = \left(\text{BaA} \times 0.06\right) + \left(\left(\text{BbF} + \text{BkF}\right) \times 0.07\right) + \text{BaP} + \left(\text{DBA} \times 0.6\right) + \left(\text{IND} \times 0.08\right) \tag{1}$$

Values of monthly mean BaPE and PM10 concentrations for each sampling site are illustrated in Fig. 7. It can be obviously seen that the BaPE values of the SP site are the highest, while those of the other sites are approximately two times lower. It can therefore be concluded that the air quality of the SP site was worse than the other sampling sites.

Patterns of the BaPE values from June 05 to June 06 were found to be similar for all sites. Fig. 7 shows two peaks of BaPE values. The first peak was obtained in the rainy season (Aug 05) while the second one was obtained in the dry season (Jan-Mar 06). Based on the fact that traffic is the main PAHs source in the urban area, it presents almost the same level of pollutants to the air throughout the year. Therefore, the BaPE values contributed from the traffic should be a background value for all sites. It is quite clear that in the dry season, open burning, i.e. burning of agricultural waste and community garbage, as well as forest fires, mainly contributed pollutants to the atmosphere. In the rainy season, the high BaPE values might have come from other sources. The air pollutants in this season could be from various local activities apart from open burning. One of the possibilities is from small scale factories in the area, whereby seasonal agricultural products are processed.

Fig. 7. Monthly mean BaPE and PM10 concentrations from all sampling sites

#### **4.2.5 Diagnostic ratios**

218 Air Quality – Monitoring and Modeling

**Season LP SP HP YP**  Wet (Jun-Sep) 0.530 0.549 0.703 0.595 Dry (Dec- Mar) 0.704 0.572 0.601 0.590

(Apr-May, Oct-Nov) 0.257 0.650 0.690 0.365

All Season 0.731 0.686 0.659 0.618

In comparison to previous studies, the correlation found in this study is almost the same as that reported by Chantara & Sangchan (2009) in which the mass of PM10 collected by a Minivol air sampler in traffic areas was found to be slightly correlated (R2 = 0.45) with the total PAH concentration of the solution obtained from extraction of the collected particulate. The afore-mentioned correlation was higher than the value in the study by Koyano et al. (1998) in which airborne particulate samples were collected from commercial and residential sites of Chiang Mai from July to December 1989 and their results revealed that there was not a large differnce in PAHs concentrations between different sampling sites. However, the airborne particulates and large amounts of rough particulates encountered in their study prompted them to conclude that there was wide spread air pollution throughout the city.

The limit value of BaP recommended by WHO is 1 ng/m3 due to its high carcinogenic property. Other PAHs such as BaA, BbF, BkF and DBA also have carcinogenic potential. To combine these potential carcinogenic compounds, the benzo(a)pyrene-equivalent carcinogenic power (BaPE) is calculated. The BaPE is an index that has been introduced for better denoting aerosol carcinogenicity related to the whole PAH fraction instead of the BaP

 BaPE = (BaA x 0.06) + ((BbF + BkF) x 0.07) + BaP + (DBA x 0.6) + (IND x 0.08) (1) Values of monthly mean BaPE and PM10 concentrations for each sampling site are illustrated in Fig. 7. It can be obviously seen that the BaPE values of the SP site are the highest, while those of the other sites are approximately two times lower. It can therefore be concluded

Patterns of the BaPE values from June 05 to June 06 were found to be similar for all sites. Fig. 7 shows two peaks of BaPE values. The first peak was obtained in the rainy season (Aug 05) while the second one was obtained in the dry season (Jan-Mar 06). Based on the fact that traffic is the main PAHs source in the urban area, it presents almost the same level of pollutants to the air throughout the year. Therefore, the BaPE values contributed from the traffic should be a background value for all sites. It is quite clear that in the dry season, open burning, i.e. burning of agricultural waste and community garbage, as well as forest fires, mainly contributed pollutants to the atmosphere. In the rainy season, the high BaPE values might have come from other sources. The air pollutants in this season could be from various local activities apart from open burning. One of the possibilities is from small scale factories

(Liu et al., 2007). The BaPE was calculated in this study using the following equation:

that the air quality of the SP site was worse than the other sampling sites.

in the area, whereby seasonal agricultural products are processed.

Transition period

Table 8. Pearson correlation of PM10 and PAHs

**4.2.4 BaP and BaP-equivalent carcinogenic power (BaPE)** 

Dianostic ratio can be used for the identification of possible emission sources. The diagnostic ratios of this study were calculated by separating sampling sites and months as shown in Table 9 (Chantara et al., 2010). The IND/(IND+BPER) ratio can be used to identify traffic sources. The value in a range from 0.35-0.70 indicates that the PAHs were emitted from diesel engines (Lui et. al, 2007 and Kavouras et. al, 2001). In this study, the average ratios of all sites were 0.50-0.56. Therefore, the PAHs found in this study could have originated from diesel engine emissions. The BaP/BPER ratio was also used for characterization of the PAHs sources. The low BaP/BPER ratio (<0.60) is an evidence of greater emission of BPER from traffic sources (Pandey et. al., 1999). The mean ratios of the 4 sampling sites were 0.56-0.80. The mean ratios of HP and YP sites were fitted with the low ratio value, therefore traffic might be the main source for these two sites. However, the ratio values in some months, especially in the dry season were greater than 0.60, which indicated some other PAH sources. The SP and LP sites had a higher mean ratio value (0.80 and 0.64, respectively) as well as monthly values, depicted as *italic* values in Table 9. This could indicate that the PAHs in these two sites were not only from the traffic but were also from some other sources on a similar basis described for the BaPE above. Apart from these two ratios, the BaP/(BaP + CHR) ratio was also considered. The ratio values of 0.57 and 0.89 could be related to diesel and gasoline exhaust, respectively (Khalili et al., 1995). The BaP/(BaP + CHR) ratio values obtained in this work were 0.68-0.77, which could be reckoned as somewhere between 0.57 and 0.89. Based on this ratio alone, it could not be concluded what the major source of PAHs was in the areas studied in this work. However, the IND/(IND+BPER) ratio of 0.50-0.56 obtained in this study fits within the range of 0.35-0.70 which suggests that the diesel engine source plays a more important role than the gasoline source.

PM10 and Its Chemical Composition: A Case Study in Chiang Mai, Thailand 221

Average (SD) concentrations (ng/m3) of PM10-bound ions in each season and sampling site are shown in Tables 10-11 and Fig. 9. Both anions (SO42- , NO3-, and Cl-) and cations (NH4+, Na+, K+, Mg2+ and Ca2+) were significantly higher in the dry period (Dec-Mar) and the transition period I (Oct-Nov) than those found in other seasons (Table 10). The dominant anion and cation were SO42- and NH4+, respectively. There was a strong correlation between

This suggested that the main acidity of dry deposition in this region was due to H2SO4, which was neutralized by NH4+ (Chantara & Chunsuk, 2009). Noticeably, K+ (biomarker of vegetative burning) was significantly 2-3 times higher in the dry season. Therefore, it can be

> **Apr – May (n = 84)**

Mg2+ **0.09 0.17 /3** 0.03 0.02 /2 0.04 0.04 /2 **0.09 0.04 /2** Ca2+ **1.29 0.52** 0.92 0.39 /2 0.79 0.29 /1 **1.17 0.43 /2 Total cations 4.38 1.89** 2.41 1.55 /2 1.36 0.48 /1 **3.11 1.31 /2 Total ions 10.59 4.62** 5.76 3.87 /2 3.14 1.49 /1 **10.07 6.56 /2**

*1.92 /2 1.20* 

*0.61 /2 0.46* 

*0.10 /2 0.15* 

**Jun – Sep (n = 184)**

*0.50 /2* 0.24 0.13 /1 *0.55* 

*0.78 /17* 0.15 0.16 /7 **1.17 1.06 /4**

*0.09 /2* **0.15 0.11 /4 0.16 0.11 /2**

(r = 0.606).

**Oct – Nov (n = 80)** 

> *0.34 /2*

> *0.46 /2*

*0.97 /2* **6.04 5.35 /2**

*0.11 /1* 0.20 0.15 /2

*1.10 /1* **6.95 5.35 /2**

*0.24 /7 0.71* 

NH4+ and SO42- (r = 0.953), followed by Na+ and SO42- (r = 0.651) and K+ and Cl-

Fig. 8. pH distribution of PM10 extractions from four sampling stations

**Ion\*† Dec – Mar**

**(n = 160)**

SO42- **4.02 2.39** *2.56* 

NO3- **1.84 1.31** *0.68* 

Cl- **0.33 0.33** *0.11* 

NH4+ **1.45 1.22** *0.96* 

Na+ **0.17 0.09 /3** *0.11* 

K+ **1.38 0.65** *0.56* 

\* /n number of non detected or incomplete data

**Total anions 6.21 3.09** 3.35 2.44 /2 *1.79* 

† Groups of number in *italic* and/or **bold** are not significant different (p > 0.05)

Table 10. Mean concentrations (ng/m3) SD of PM10-bound ions in each season


Table 9. Diagnostic ratios of PAHs for four sampling sites in different months

#### **5. Analysis of dissolved ions in PM10 samples by Ion Chromatography**

#### **5.1 Ion analysis method by Ion Chromatography**

Dissolved ions including Na+, NH4+ , K+, Ca2+, Mg2+, Cl-, NO3- and SO4 2- were analyzed following a method developed from the Technical Document for Filter Pack Method (Acid Deposition Monitoring Network in East Asia, 2003) using Ion Chromatograph (IC). Samples were put into 100 ml beakers followed by 50 ml of deionized water. The beakers were then covered by paraffin film and ultrasonicated at 100% ultrasound power (T490DH, Elma, Germany) for 30 min. The solution was filtered through a cellulose acetate filter (pore size 45 μm, diameter 13 mm) into a plastic bottle for further analysis by IC (Metrohm, Switzerland). Analytical columns were Metrosep A Supp4 (4 x 250 mm) and Metrosep C2 (4 x 100 mm) for anions and cations, respectively.

#### **5.2 pH and ion contents of PM10**

The extracted samples were divided into 2 parts for pH measurement and ion analysis. The pH distribution for all sampling stations was plotted as shown in Fig. 8. More than 50% of the samples from all sampling sites except the YP station (45%) had pH value less than 5.6. Specifically, half of all the samples from the SP station was in the lower pH range value (<5.0). The seasonal variation of the pH values illustrated that low pH values were inspected in the dry season, particularly in March and April.

Jun-05 0.54 0.57 0.50 0.48 *0.61 0.87 0.63* 0.50 0.66 0.57 0.58 0.60 Jul-05 0.53 0.55 0.49 0.48 *0.70 1.06 0.66 0.61* 0.86 0.79 0.83 0.83 Aug-05 0.61 0.62 0.56 0.55 *0.61 0.95 0.70 0.67* 0.90 0.83 0.86 0.86 Sep-05 0.57 0.60 0.52 0.52 0.46 0.59 0.48 0.46 0.77 0.71 0.71 0.72 Oct-05 0.60 0.58 0.52 0.52 0.48 *0.62* 0.51 0.45 0.83 0.78 0.79 0.80 Nov-05 0.56 0.55 0.50 0.50 *0.81 0.94 0.69 0.61* 0.75 0.77 0.83 0.86 Dec-05 0.52 0.54 0.47 0.48 *0.83 1.03 0.64 0.63* 0.68 0.66 0.73 0.76 Jan-06 0.52 0.56 0.45 0.46 *0.86 0.86 0.66 0.64* 0.67 0.56 0.81 0.78 Feb-06 0.57 0.58 0.53 0.52 *0.89 1.01 0.71 0.63* 0.65 0.62 0.81 0.78 Mar-06 0.55 0.55 0.54 0.54 *0.65 0.79* 0.59 0.54 0.68 0.55 0.67 0.67 Apr-06 0.52 0.55 0.48 0.48 0.53 *0.66* 0.50 0.48 0.72 0.49 0.65 0.64 May-06 0.52 0.53 0.45 0.43 0.40 0.50 0.45 0.38 0.79 0.73 0.81 0.77 Jun-06 0.52 0.52 0.47 0.54 0.44 0.56 0.47 *0.61* 0.88 0.80 0.88 0.80 **Avg 0.55 0.56 0.50 0.50 0.64 0.80 0.59 0.56 0.76 0.68 0.77 0.76**  SD 0.03 0.03 0.03 0.03 0.17 0.20 0.10 0.09 0.09 0.11 0.09 0.08

Table 9. Diagnostic ratios of PAHs for four sampling sites in different months

**5.1 Ion analysis method by Ion Chromatography** 

in the dry season, particularly in March and April.

anions and cations, respectively.

**5.2 pH and ion contents of PM10** 

**5. Analysis of dissolved ions in PM10 samples by Ion Chromatography** 

Dissolved ions including Na+, NH4+ , K+, Ca2+, Mg2+, Cl-, NO3- and SO42- were analyzed following a method developed from the Technical Document for Filter Pack Method (Acid Deposition Monitoring Network in East Asia, 2003) using Ion Chromatograph (IC). Samples were put into 100 ml beakers followed by 50 ml of deionized water. The beakers were then covered by paraffin film and ultrasonicated at 100% ultrasound power (T490DH, Elma, Germany) for 30 min. The solution was filtered through a cellulose acetate filter (pore size 45 μm, diameter 13 mm) into a plastic bottle for further analysis by IC (Metrohm, Switzerland). Analytical columns were Metrosep A Supp4 (4 x 250 mm) and Metrosep C2 (4 x 100 mm) for

The extracted samples were divided into 2 parts for pH measurement and ion analysis. The pH distribution for all sampling stations was plotted as shown in Fig. 8. More than 50% of the samples from all sampling sites except the YP station (45%) had pH value less than 5.6. Specifically, half of all the samples from the SP station was in the lower pH range value (<5.0). The seasonal variation of the pH values illustrated that low pH values were inspected

IND/(IND+BPER) BaP/BPER BaP/(BaP+CHR) LP SP HP YP LP SP HP YP LP SP HP YP

Fig. 8. pH distribution of PM10 extractions from four sampling stations

Average (SD) concentrations (ng/m3) of PM10-bound ions in each season and sampling site are shown in Tables 10-11 and Fig. 9. Both anions (SO4 2- , NO3-, and Cl- ) and cations (NH4 +, Na+, K+, Mg2+ and Ca2+) were significantly higher in the dry period (Dec-Mar) and the transition period I (Oct-Nov) than those found in other seasons (Table 10). The dominant anion and cation were SO42- and NH4+, respectively. There was a strong correlation between NH4+ and SO42- (r = 0.953), followed by Na+ and SO42- (r = 0.651) and K+ and Cl- (r = 0.606). This suggested that the main acidity of dry deposition in this region was due to H2SO4, which was neutralized by NH4+ (Chantara & Chunsuk, 2009). Noticeably, K+ (biomarker of vegetative burning) was significantly 2-3 times higher in the dry season. Therefore, it can be


\* /n number of non detected or incomplete data

† Groups of number in *italic* and/or **bold** are not significant different (p > 0.05)

Table 10. Mean concentrations (ng/m3) SD of PM10-bound ions in each season

PM10 and Its Chemical Composition: A Case Study in Chiang Mai, Thailand 223

Correlations of ion concentrations (µg/m3) and PM10 weight (mg) were analysed among all sampling stations (Fig. 10a) and among all seasons (Fig. 10b). It was found that almost all ions, except Na+, Mg2+, were well correlated with PM10 (*p< 0.001*). K+ concentrations and had a very strong correlation with PM10 in every sampling site in the dry season (December – May) and the transition period (April – March). As mentioned above, K+ is often used as a tracer for biomass burning. This figure obviously showed the possibility of a high frequency and huge area of open burning, including forest fires and agricultural burning in Northern

Fig. 10. Correlations between ion and PM10 concentrations; (a) sampling sites and (b)

from Feb 02 – Jan 03. It was found that concentrations of SO4

Table 12 illustrates the major ion composition of PM10 samples in this study in comparison with other studies, which were 1) ion composition of dry deposition collected by 4-stages filter pack sampler of the Mae Hia sampling site (approximately 7 km southwest from the city center of Chiang Mai) from Oct 05 – Sep 06 and 2) PM10-bound ions in the Bangkok area

detected in this study were approximately 2 times higher than those in Mae Hia. This is

2- , NO3

(a)

(b)


+ and Ca2+

Thailand and neighboring countries.

seasons

concluded that biomass burning took place in the area during the dry season. Total ion concentrations including total anions and cations were not significantly different among all sites (Table 11). Standing out was the SP site with significantly higher concentrations of some major ions such as NO3- and K+. This is probably due to a higher frequency or larger scale of open burning happening in the area.


\* /n number of non detected or incomplete data

† Groups of number in *italic* and/or **bold** are not significant different (p > 0.05)

Table 11. Mean concentrations (ng/m3) SD of PM10-bound ions in each sampling site

Fig. 9. Spatial and temporal variations of ion concentrations (µg/m3)

concluded that biomass burning took place in the area during the dry season. Total ion concentrations including total anions and cations were not significantly different among all sites (Table 11). Standing out was the SP site with significantly higher concentrations of some major ions such as NO3- and K+. This is probably due to a higher frequency or larger

**Ion\*† LP (n=127) SP (n=127) HP (n=127) YP (n=127)**  SO42- 2.67 2.62 /2 3.39 3.40 /1 2.86 3.16 /1 3.37 3.45 /2

Cl- **0.27 0.28 /1 0.31 0.30 /1** 0.13 0.11 /1 0.13 0.10 /2 **Total anions** 3.98 3.42 /1 4.84 3.98 /1 3.76 3.44 /1 4.42 3.79 /2 NH4+ 0.76 0.92 /6 1.00 1.09 /8 0.79 1.00 /7 0.89 1.13 /7

 *0.62 /1* **0.84 0.73 /1** *0.59* 

**Total cations** 2.56 1.77 /1 3.02 1.91 /1 2.55 1.69 /1 2.93 2.03 /2 **Total ions** 6.54 4.99 /1 7.85 5.64 /1 6.31 4.83 /1 7.35 5.51 /2

Table 11. Mean concentrations (ng/m3) SD of PM10-bound ions in each sampling site

Mg2+ 0.07 0.11 /2 0.07 0.08 /2 0.06 0.08 /3 0.07 0.14 /2

 *1.21 /4* **1.16 1.09 /3** 0.78 0.78 /2 *0.91* 

 *0.43 /1 1.02* 

 *0.08 /3 0.14* 

 *0.62 /1 0.68*

 *0.39 /1 1.19* 

 *0.87 /2*

 *0.09 /4*

 *0.67 /2*

 *0.52 /2*

scale of open burning happening in the area.

Na+ **0.17 0.12 /1 0.16 0.10 /3** *0.12* 

 *0.48 /1 1.00* 

† Groups of number in *italic* and/or **bold** are not significant different (p > 0.05)

Fig. 9. Spatial and temporal variations of ion concentrations (µg/m3)

NO3- *1.07* 

K+ *0.69* 

Ca2+ *0.91* 

\* /n number of non detected or incomplete data

Correlations of ion concentrations (µg/m3) and PM10 weight (mg) were analysed among all sampling stations (Fig. 10a) and among all seasons (Fig. 10b). It was found that almost all ions, except Na+, Mg2+, were well correlated with PM10 (*p< 0.001*). K+ concentrations and had a very strong correlation with PM10 in every sampling site in the dry season (December – May) and the transition period (April – March). As mentioned above, K+ is often used as a tracer for biomass burning. This figure obviously showed the possibility of a high frequency and huge area of open burning, including forest fires and agricultural burning in Northern Thailand and neighboring countries.

Fig. 10. Correlations between ion and PM10 concentrations; (a) sampling sites and (b) seasons

Table 12 illustrates the major ion composition of PM10 samples in this study in comparison with other studies, which were 1) ion composition of dry deposition collected by 4-stages filter pack sampler of the Mae Hia sampling site (approximately 7 km southwest from the city center of Chiang Mai) from Oct 05 – Sep 06 and 2) PM10-bound ions in the Bangkok area from Feb 02 – Jan 03. It was found that concentrations of SO4 2- , NO3 -, NH4 + and Ca2+ detected in this study were approximately 2 times higher than those in Mae Hia. This is

PM10 and Its Chemical Composition: A Case Study in Chiang Mai, Thailand 225

(157%). The instrument detection limit (IDL) ranged from between 0.01 ng/m3 (Sr) to

Fig. 11 shows the spatial and temporal variation of elements in all sampling sites. Mean concentrations of elements and metals were high in the dry season, except for zinc, for which its concentrations were quite high in the rainy season in every sampling site. Calcium concentration was high all year round in all sampling sites. It was generated from soil resuspension, which was related to high amounts of particulates being suspended in the air during the dry season. Potassium and silicon concentrations were high in the dry season in every sampling site, especially at the SP site. Both of them were ash elements that were emitted from biomass burning (Demirbas et al., 2008). The result was very well in agreement

Toxic metals including lead (Pb), mercury (Hg), cadmium (Cd) and arsenic (As), were detected only in some samples in tracer levels in the dry season as well as in the transition period (Oct-Nov). There was no significant difference of element concentrations in terms of

Element concentrations found in this study were compared with the study in the Bangkok area, in which 14 elements were determined from 4 areas including Din Daeng, Chankasaem, Ban Somdej and Bangna, from Feb 02 – Jan 03 (Chuersuwan et al., 2008) (Table 13). It was found that the mean concentration of almost all elements found in the Bangkok area was higher than those in this study. This is probably because sources of elements and heavy metals, especially traffic emissions in Bangkok, were higher both in terms of

5.47 ng/m3 (K).

the sampling site.

variations and number.

**6.2 PM10 bound elements** 

with the dissolved ion content mentioned earlier.

Fig. 11. Spatial and temporal variations of PM10 bound elements

because Mae Hia is located in a sub-urban area with lower human activities than the stations of this study. When comparing the results of this study with Bangkok, it reveals that NO3 and NH4 + concentrations were almost the same, while SO4 2- of this study was slightly higher. Unsurprisingly, Cl- content in the Bangkok area was 7-10 times higher than that which was found in this study. This must be related with the distance from sea.


\* Mean values from filter pack samples; dry deposition

Table 12. Mean concentrations (µg/m3) SD of PM10-bound ions in Chiang Mai and Bangkok, Thailand

#### **6. Analysis of PM10 bound elements by Inductively Coupled Plasma – Optical Emission Spectroscopy (ICP-OES)**

#### **6.1 Method efficiency for analysis of PM10 bound elements by ICP-OES**

Filter samples were extracted based on Compendium Method IO-3.1 (US. EPA, 1999a) for 20 elements including Ca, Al, Si, Fe, Mg, K, Zn, Ti, P, Pb, Ba, Sr, Mn, Ni, Cu, V, Cr, Cd, Hg and As. They were analysed by Inductively Coupled Plasma – Optical Emission Spectroscopy (ICP-OES) (PerkinElmer Optima 3000 operated with WinLab32TM software) based on Compendium Method IO-3.4 (US. EPA, 1999b). Samples were put into Erlenmayer flasks followed by the addition of 40 ml of mixed acid solution (5.55% HNO3/16.75% HCl) and refluxed on hot plates for 3 hours. They were left until cool and transferred into 50 ml volumetric flasks. Prior to analysis by ICP, they were filtered through a 0.45 μm syringe nylon filter into plastic bottles.

Percent recoveries of elements were tested by the spiking of 2.50 µg/ml of mixed standards onto the quartz filter paper and followed the analysis method described above. The recoveries ranged between 76.0% (Mg) to 97.2% (Cr), except for Fe which was quite high (157%). The instrument detection limit (IDL) ranged from between 0.01 ng/m3 (Sr) to 5.47 ng/m3 (K).

#### **6.2 PM10 bound elements**

224 Air Quality – Monitoring and Modeling

because Mae Hia is located in a sub-urban area with lower human activities than the stations of this study. When comparing the results of this study with Bangkok, it reveals that NO3 and NH4+ concentrations were almost the same, while SO42- of this study was slightly higher. Unsurprisingly, Cl- content in the Bangkok area was 7-10 times higher than that

> Chiang Mai (Oct 05 – Sep 06)\*

MinSD MaxSD MinSD MaxSD

SO42- 2.67±2.62 3.37±3.42 1.55±1.52 1.79±0.59 2.37±0.72

NO3- 0.73±0.65 1.13±1.04 0.69±1.04 1.16±0.55 1.45±0.66

Cl- 0.13±0.11 0.31±0.30 0.29±0.76 1.10±0.56 2.00±0.91

NH4+ 0.76±0.92 1.00±1.09 0.45±0.50 0.62±0.31 0.94±0.62

Na+ 0.12±0.08 0.17±0.12 0.19±0.30 - -

K+ 0.56±0.57 0.83±0.71 0.86±2.07 - -

Mg2+ 0.06±0.08 0.07±0.14 0.16±0.34 - -

Ca2+ 0.91±0.48 1.18±0.53 0.69±1.06 - -

Table 12. Mean concentrations (µg/m3) SD of PM10-bound ions in Chiang Mai and

**6.1 Method efficiency for analysis of PM10 bound elements by ICP-OES** 

**6. Analysis of PM10 bound elements by Inductively Coupled Plasma – Optical** 

Filter samples were extracted based on Compendium Method IO-3.1 (US. EPA, 1999a) for 20 elements including Ca, Al, Si, Fe, Mg, K, Zn, Ti, P, Pb, Ba, Sr, Mn, Ni, Cu, V, Cr, Cd, Hg and As. They were analysed by Inductively Coupled Plasma – Optical Emission Spectroscopy (ICP-OES) (PerkinElmer Optima 3000 operated with WinLab32TM software) based on Compendium Method IO-3.4 (US. EPA, 1999b). Samples were put into Erlenmayer flasks followed by the addition of 40 ml of mixed acid solution (5.55% HNO3/16.75% HCl) and refluxed on hot plates for 3 hours. They were left until cool and transferred into 50 ml volumetric flasks. Prior to analysis by ICP, they were filtered through a 0.45 μm syringe

Percent recoveries of elements were tested by the spiking of 2.50 µg/ml of mixed standards onto the quartz filter paper and followed the analysis method described above. The recoveries ranged between 76.0% (Mg) to 97.2% (Cr), except for Fe which was quite high

Bangkok (Feb 02 – Jan 03) (Chuersuwan, et al., 2008)

which was found in this study. This must be related with the distance from sea.

This research (Jun 05 – Jun 06) Mae Hia,

\* Mean values from filter pack samples; dry deposition

**Emission Spectroscopy (ICP-OES)** 

nylon filter into plastic bottles.

Ion

Bangkok, Thailand

Fig. 11 shows the spatial and temporal variation of elements in all sampling sites. Mean concentrations of elements and metals were high in the dry season, except for zinc, for which its concentrations were quite high in the rainy season in every sampling site. Calcium concentration was high all year round in all sampling sites. It was generated from soil resuspension, which was related to high amounts of particulates being suspended in the air during the dry season. Potassium and silicon concentrations were high in the dry season in every sampling site, especially at the SP site. Both of them were ash elements that were emitted from biomass burning (Demirbas et al., 2008). The result was very well in agreement with the dissolved ion content mentioned earlier.

Toxic metals including lead (Pb), mercury (Hg), cadmium (Cd) and arsenic (As), were detected only in some samples in tracer levels in the dry season as well as in the transition period (Oct-Nov). There was no significant difference of element concentrations in terms of the sampling site.

Fig. 11. Spatial and temporal variations of PM10 bound elements

Element concentrations found in this study were compared with the study in the Bangkok area, in which 14 elements were determined from 4 areas including Din Daeng, Chankasaem, Ban Somdej and Bangna, from Feb 02 – Jan 03 (Chuersuwan et al., 2008) (Table 13). It was found that the mean concentration of almost all elements found in the Bangkok area was higher than those in this study. This is probably because sources of elements and heavy metals, especially traffic emissions in Bangkok, were higher both in terms of variations and number.

PM10 and Its Chemical Composition: A Case Study in Chiang Mai, Thailand 227

especially in the way of forest fires and agricultural burning in Northern Thailand and

The PM10 chemical composition, including both organic parts (carbon and PAHs) and inorganic parts (dissolved ions and elements & heavy metals) revealed that their concentrations were obviously high in the dry season in all sampling stations. They were well correlated with PM10 concentrations, especially in the dry season. The dominant organic compounds found in the PM10 samples were 4-6 rings PAHs. The ratios of carcinogenic PAHs over non-carcinogenic ones were approximately 3:1-4:1, which was about 1.0-1.5 times higher than those that were reported from the Bangkok area. The BaPE values representing carcinogenic power were highest in the samples taken from the SP site, which is located in the eastern part of the city, and this was due to the higher number of sources, i.e. open burning, street food cooking and small-scale factories. The dominant

fossil fuel combustion. PM10 major elemental composition was Ca, K and Si, which were reported to be emitted from biomass burning. To reduce air pollution in this area, their sources must be reduced. Open burning from both forest fires and agricultural burning need to be managed. Biomass residues have to be removed or reused for other purposes, such as organic fertilizer, green or bio-fuels for cooking in households, etc., to avoid the practice of large-scale burning. Better management of forest areas in the dry season with plenty of leaf litter is also needed. Apart from that, potential sources of air pollution in the urban area also need to be considered. Mobile sources; traffic volume, number and types of vehicles and fuel used, as well as the point source, such as small scale factories, should also be considered. Air pollution management campaigns are only going to be successful and sustained when the issue is brought, as a priority, into the area of public concern. Moreover, increasing the number of air quality monitoring stations is necessary to cover and represent a wider area. This will, in turn, lead to higher precision and accuracy in recording and

Financial support from the Thailand Research Fund (TRF) and research facilities from the Chemistry Department, Faculty of Science, Chiang Mai University are gratefully acknowledged. Special thanks are extended to the Air Project Team, Chemistry Dept., CMU, particularly Assoc. Prof. Dr. Mongkon Rayanakorn, Asst. Prof. Dr. Sunanta Wangkarn and Dr. Urai Tengjaroenkul. Thanks to all member of the Environmental Chemistry Research

Acid Deposition Monitoring Network in East Asia (EANET). (2003). *Technical Document for* 

Boonyatumanond, R., Murakami, M., Wattayakorn, G., Togo, A., & Takada, H. (2007).

Meeting of Acid Deposition Monitoring Network in East Asia.

*Filter Pack Method in East Asia*, The Third Interim Scientific Advisory Group

Sources of polycyclic aromatic hydrocarbons (PAHs) in street dust in a tropical Asian mega-city, Bangkok, Thailand. *Science of the Total Environment*, 384, pp. 420-

2- and NH4+, which showed high amounts of

neighbouring countries.

**8. Acknowledgment** 

**9. References** 

water-soluble inorganic substances were SO4

interpreting the data for solid air quality management.

Laboratory, CMU for their great help and support.

432. ISSN 0048-9697.


Table 13. Min and max concentrations (µg/m3) SD of mean PM10 and its elemental composition found in this study and in Bangkok, Thailand

### **7. Conclusion**

Sampling and analysis of PM10 and its chemical composition in the Chiang Mai – Lamphun Basin, Thailand revealed interesting patterns of ambient PM10 concentrations in the urban areas. The background level of PM10 was mainly generated from traffic volume, which was quite constant throughout the year. Seasonal variation of PM10 levels was quite distinct between the dry and wet seasons. Rain precipitation was a main factor that reduced concentrations of air pollutants, as both particulate and gas phases. In the wet season, a high amount of precipitation scavenges contaminated atmosphere, and the sky becomes clear. In the dry season, there is a low amount of rain and high air pressure leading to air pollutant accumulation. The geographical characteristics of the basin are that it surrounded by a wall of high mountains. Temperature inversion occurs in the dry season, especially in February and March, acting as a lid covering the basin. Heavy haze occurs almost every year in the basin, especially in the past ten years. It was proof that the main sources are open burning, especially in the way of forest fires and agricultural burning in Northern Thailand and neighbouring countries.

The PM10 chemical composition, including both organic parts (carbon and PAHs) and inorganic parts (dissolved ions and elements & heavy metals) revealed that their concentrations were obviously high in the dry season in all sampling stations. They were well correlated with PM10 concentrations, especially in the dry season. The dominant organic compounds found in the PM10 samples were 4-6 rings PAHs. The ratios of carcinogenic PAHs over non-carcinogenic ones were approximately 3:1-4:1, which was about 1.0-1.5 times higher than those that were reported from the Bangkok area. The BaPE values representing carcinogenic power were highest in the samples taken from the SP site, which is located in the eastern part of the city, and this was due to the higher number of sources, i.e. open burning, street food cooking and small-scale factories. The dominant water-soluble inorganic substances were SO42- and NH4+, which showed high amounts of fossil fuel combustion. PM10 major elemental composition was Ca, K and Si, which were reported to be emitted from biomass burning. To reduce air pollution in this area, their sources must be reduced. Open burning from both forest fires and agricultural burning need to be managed. Biomass residues have to be removed or reused for other purposes, such as organic fertilizer, green or bio-fuels for cooking in households, etc., to avoid the practice of large-scale burning. Better management of forest areas in the dry season with plenty of leaf litter is also needed. Apart from that, potential sources of air pollution in the urban area also need to be considered. Mobile sources; traffic volume, number and types of vehicles and fuel used, as well as the point source, such as small scale factories, should also be considered. Air pollution management campaigns are only going to be successful and sustained when the issue is brought, as a priority, into the area of public concern. Moreover, increasing the number of air quality monitoring stations is necessary to cover and represent a wider area. This will, in turn, lead to higher precision and accuracy in recording and interpreting the data for solid air quality management.

#### **8. Acknowledgment**

226 Air Quality – Monitoring and Modeling

Min ± SD Max ± SD Min ± SD Max ± SD

(Chuersuwan, et al., 2008)

Element This research (Jun 05 – Jun 06) Bangkok (Feb 02 – Jan 03)

PM10 33.2±21.4 73.6±25.3 57.6±23.9 108.1±35.5 Al 0.27±0.19 0.31±0.43 1.32±0.78 3.02±2.73 As 0.01±0.00 0.02±0.01 0.18±0.13 0.32±0.14 Ca 1.13±1.18 1.67±1.21 1.89±1.50 4.79±2.38 Cr 0.01±0.02 0.02±0.02 0.10±0.05 0.14±0.09 Cu 0.01±0.01 0.02±0.03 0.05±0.02 0.09±0.04 Fe 0.29±0.24 0.37±0.38 1.36±0.69 2.13±0.92 Hg 0.02±0.02 0.03±0.02 - - K 0.67±0.75 0.92±0.98 0.55±0.33 1.21±0.72 Mg 0.11±0.09 0.14±0.10 0.49±0.27 0.83±0.32 Mn 0.02±0.01 0.02±0.01 0.07±0.04 0.09±0.09 Ni 0.01±0.01 0.01±0.02 0.03±0.20 0.24±0.31 P 0.01±0.01 0.03±0.03 - - Pb 0.04±0.03 0.05±0.08 0.13±0.12 0.22±0.15 Si 0.90±0.68 1.03±0.75 - - Sr 0.003±0.003 0.003±0.003 - - Ti 0.01±0.01 0.01±0.01 - - V 0.01±0.01 0.01±0.01 0.80±0.38 1.13±0.51 Zn 0.07±0.07 0.201±1.023 0.65±0.35 0.84±0.47

Table 13. Min and max concentrations (µg/m3) SD of mean PM10 and its elemental

Sampling and analysis of PM10 and its chemical composition in the Chiang Mai – Lamphun Basin, Thailand revealed interesting patterns of ambient PM10 concentrations in the urban areas. The background level of PM10 was mainly generated from traffic volume, which was quite constant throughout the year. Seasonal variation of PM10 levels was quite distinct between the dry and wet seasons. Rain precipitation was a main factor that reduced concentrations of air pollutants, as both particulate and gas phases. In the wet season, a high amount of precipitation scavenges contaminated atmosphere, and the sky becomes clear. In the dry season, there is a low amount of rain and high air pressure leading to air pollutant accumulation. The geographical characteristics of the basin are that it surrounded by a wall of high mountains. Temperature inversion occurs in the dry season, especially in February and March, acting as a lid covering the basin. Heavy haze occurs almost every year in the basin, especially in the past ten years. It was proof that the main sources are open burning,

composition found in this study and in Bangkok, Thailand

**7. Conclusion** 

Financial support from the Thailand Research Fund (TRF) and research facilities from the Chemistry Department, Faculty of Science, Chiang Mai University are gratefully acknowledged. Special thanks are extended to the Air Project Team, Chemistry Dept., CMU, particularly Assoc. Prof. Dr. Mongkon Rayanakorn, Asst. Prof. Dr. Sunanta Wangkarn and Dr. Urai Tengjaroenkul. Thanks to all member of the Environmental Chemistry Research Laboratory, CMU for their great help and support.

#### **9. References**


PM10 and Its Chemical Composition: A Case Study in Chiang Mai, Thailand 229

Koyano, M., Endo, O., Goto, S., Tanabe, K., Koottatep, S. & Matsushita, H., (1998).

Liu, L.B., Hashi, Y., Liu, M., Wei, Y. & Lin, J.M. (2007). Determination of particle-associated

Mauderly, J.L. & Chow, J.C. (2008). Health effects of organic aerosols. *Inhalation Toxicol*ogy,

Metzger, K.B., Tolbert, P.E., Klein, M., Peel, J.L., Flanders, W.D., Todd, K., Mulholland, J.A.,

emergency department visits. *Epidemiology* 15, pp. 46-56, ISSN 1044-3983. Miller J.C. & Miller J.N. (1993). *Statistics for Analytical Chemistry*, 3rd edition, Ellis Horwood,

Norramit, P., Cheevaporn, V., Itoh, N. & Tanaka, K. (2005). Characterization and

Pandey, P.K., Patel, K.S. & Lenicek, J. (1999). Polycyclic aromatic hydrocarbons: Need for

Patashnick, H., Meyer, M. & Rogers, B. (2002). Tapered element oscillating microbalance

Pengchai, P., Chantara, S., Sopajaree, K., Wangkarn, S., Tengcharoenkul, U. & Rayanakorn,

*Environmental Monitoring and Assessment*, 154, pp. 197-218. ISSN 1573-2959. Perkin Elmer Cooperation. (1991). *Users Manual PE 2400 Series II CHNS/O Analyser*, Perkin

Pio, C.A. & Harrison, R.M. (1987). Vapor-pressure of ammonium–chloride aerosol - effect of

Pollution Control Department (2011). Regional area air quality data, 2 August 2011,

 <http://www.pcd.go.th/AirQuality/Regional/QueryAirThai.cfm?task=default> Pöschl, U. (2005). Atmospheric aerosols: composition, transformation, climate and health effects. *Angewandte Chemie International Edition* , 44, pp. 7520-7540, ISSN 1521-3773. Seinfeld, J.H. & Pandis, S.N. (1998). *Atmospheric Chemistry and Physics: From Air Pollution to Climate Change*, John Wiley and Sons Inc., ISBN 0-471-17815-2, New York. Torigoe, K., Hasegawa, S., Numata, O., Yazaki, S., Matsunaga, M., Boku, N., Hiura, M. &

in children. *Pediatrics International,* 42, pp. 143-150, ISSN 1328-8067.

1352-2310.

ISSN 0013-273X.

*Science*, 23, pp. 667-671, ISSN 0910-6340.

ISBN 0130309907/0-13-030990-7, New York.

*Science*, 51, pp. 437-446. ISSN 1347-5207.

Kingston, Ontario, Canada, 8-12 June 2002.

pp. 287–319. ISSN 1573-2959.

Elmer, Connecticut.

Available from:

2310.

20, pp. 257-288, ISSN 0895-8378.

km north of Bangkok, Thailand. *Atmospheric Environment*, 34, pp. 4557-4563, ISSN

Carcinogenic polynuclear aromatic hydrocarbons in the atmosphere in Chiang Mai, Thailand, *Japanese Journal of Toxicology and Environmental Health*, 44(3), pp. 214–225,

polycyclic aromatic hydrocabens in urban air of Beijing by GC/MS. *Analytical* 

Ryan, P.B., & Frumkin, H. (2004). Ambient air pollution and cardiovascular

carcinogenic risk assessment of polycyclic aromatic hydrocarbons in the respirable fraction of airborne particles in the Bangkok metropolitan area. *Journal of Health* 

assessment of health risks in India? *Environmental Monitoring and Assessment*, 59,

technology, *Proceedings of the North American/Ninth U.S. Mine Ventilation Symposium*,

M. (2009). Seasonal variation, risk assessment and source estimation of PM10 and PM10-Bound PAHs in the ambient air of Chiang Mai and Lamphun, Thailand.

temperature and humidity. *Atmospheric Environment* 21, pp. 2711-2715. ISSN 1352-

Ino, H. (2000). Influence of emission from rice straw burning on bronchial asthma


Chantara, S., & Chunsuk, N. (2008). Comparison of wet-only and bulk deposition at Chiang

Chantara, S., & Sangchan, W. (2009). Sensitive analytical method for particle-bound

Chantara S., Wangkarn S., Sangchan W. & Rayanakorn M. (2010). Spatial and Temporal

Chantara, S., Wangkarn, S., Tengcharoenkul, U., Sangchan, W. & Rayanakorn, M. (2009).

Chetwittayachan, T., Shimazaki, D., & Yamamoto, K. (2002) A comparison of temporal

Chuersuwan, N., Nimrat, S., Lekphet, S., & Kerdkumrai, T. (2008). Levels and major

Demirbas, A. (2008). Hazardous Emissions from Combustion of Biomass. *Energy Sources,* 

Furton, K.G., & Pentzke, G. (1998). Polycyclic Aromatic Hydrocarbon, In: *Chromatographic* 

Fuzzi, S., Decesari, S., Facchini, M.C., Cavalli, F., Emblico, L., Mircea, M., Andreae, M.O.,

Kavouras, I.G., Koutrakis, P., Tsapakis, M., Lagoudaki, E., Stephanou, E., Von Baer, D. &

*Environmental Science and Technology*, 25, pp. 2288–2294, ISSN 1735 1472. Khalili, N.R., Scheff, P.A., & Holsen, T.M. (1995). PAH source fingerprints for coke ovens,

Kim Oanh, N.T., Bætz Reutergårdh, L., Dung, N.T., Yu, M.H., Yao, W.X. & Co, H.X. (2000).

Thailand. *Atmospheric Environment.* 36, pp. 2027-2037. ISSN 1352-2310. Chiang Mai Meteorological Station. Statistic of Chiang Mai climate. 2 August 2011.

Available from <http://www.cmmet.tmd.go.th/index1.php>

112(D1), D01201, doi:10.1029/2005JD006741. ISSN 0148-0227.

*Atmospheric Environment*, 29, pp. 533–542, ISSN 1352-2310.

*International*. 34, pp. 671-677. ISSN 0160-4120.

*Part A*. 30, pp. 170–178. ISSN 1556-7036.

ISBN 0824701453, New York.

42, pp. 5511-5518. ISSN 1352-2310.

17-25. ISSN 1944-3994.

ISSN 0125-2526.

*ScienceAsia.* 35, pp. 42-48. ISSN 0303-8122.

Mai (Thailand) based on rainwater chemical composition. *Atmospheric Environment.* 

polycyclic aromatic hydrocarbons; A case study in Chiang Mai, Thailand.

Variations of Ambient PM10-bound Polycyclic Aromatic Hydrocarbons in Chiang Mai and Lamphun Provinces, Thailand. *Desalination and Water Treatment*. 19, pp.

Chemical Analysis of Airborne Particulates for Air Pollutants in Chiang Mai and Lamphun Provinces, Thailand. *Chiang Mai Journal of Science*, 36(2), pp. 001-013.

variation of particle-bound polycyclic aromatic hydrocarbons (pPAHs) concentration in different urban environments: Tokyo, Japan, and Bangkok,

sources of PM2.5 and PM10 in Bangkok Metropolitan Region. *Environmental* 

*Analysis of Environmental Food Toxicants*, Shibamoto, T. pp. 1-30. Marcel Dekker Inc.,

Trebs, I., Hoffer, A., Guyon, P., Artaxo, P., Rizzo, L., Lara, L., Pauliquevis, T., Maenhaut, W., Raes Chi, N.X., Mayol-Bracero, O., Soto-Garc´a, L.L., Claeys, M., Kourtchev, I., Rissler, J., Swietlicki, E., Tagliavini, E., Schkolnik, G., Falkovich, A., Rudich, Y., Fisch, G., & Gatti L. (2007). Overview of the inorganic and organic composition of size-segregated aerosol in Rondonia, Brazil, from the biomass burning period to the onset of the wet season. *Journal of Geophysical Research*,

Oyola, P. (2001). Source apportionment of urban particulate aliphatic and polynuclear aromatic hydrocarbons (PAHs) using multivariate methods,

diesel and gasoline engines, highway tunnels, and wood combustion emissions,

Polycyclic aromatic hydrocarbons in the airborne particulate matter at a location 40

km north of Bangkok, Thailand. *Atmospheric Environment*, 34, pp. 4557-4563, ISSN 1352-2310.


<http://www.pcd.go.th/AirQuality/Regional/QueryAirThai.cfm?task=default>


U.S. EPA (Environmental Protection Agency): Center for Environmental Research

U.S. EPA (Environmental Protection Agency):Center for Environmental Research

U.S. EPA (Environmental Protection Agency). (1999c). *National Air Toxics Program: The* 

Valerie, S., Clare, S. & Aoife, D. (2003). Microwave digestion of sediment, soils and urban

Washington State Department of Ecology, (2004). PM10 Tapered Element Oscillating

Available from: <www.ecy.wa.gov ›Ecology home › Publications & Forms>

<www.epa.gov/ttnamti1/files/ambient/inorganic/mthd-3-1.pdf>

coupled plasma (ICP) spectrometry, Retrieved from <www.epa.gov/nrmrl/pubs/625r96010/iocompen.pdf>

Register Vol. 64, No. 137,1999, Retrieved from

<www.epa.gov/ttnatw01/urban/fr19jy99.pdf>

Information, Office of Research and Development. (1999a). *Compendium of methods for the determination of inorganic compounds in ambient air.* Compendium Method IO-3.1., Selection, preparation, and extraction of filter material, Retrieved from

Information, Office of Research and Development. (1999b). *Compendium of methods for the Determination of inorganic compounds in ambient air.* Compendium Method IO-3.4., Determination of metals in ambient particulate matter using inductively

*Integrated Urban Strategy*, United States Environmental Protection Agency's Federal

particulate matter for trace metal analysis. *Talanta.* 60, pp. 715-723, ISSN 0039-9140.

Microbalance Operating Procedure, In: *Air Quality Program*, 1 September 2011,

### *Edited by Sunil Kumar and Rakesh Kumar*

The book reports research on relationship between fungal contamination and its health effects in large Asian cities, estimation of ambient air quality in Delhi, a qualitative study of air pollutants from road traffic, air quality in low-energy buildings, some aspects of the Sentinel method for pollution problem, evaluation of dry atmospheric deposition at sites in the vicinity of fuel oil fired power, particles especially PM 10 in the indoor environment, etc.

Air Quality - Monitoring and Modeling

Air Quality

Monitoring and Modeling

*Edited by Sunil Kumar and Rakesh Kumar*

Photo by leskas / iStock