**Energy Efficiency Analysis in Agricultural Productions: Parametric and Non-Parametric Approaches**

S. H. Mousavi Avval, Sh. Rafiee and A. Keyhani *Department of Agricultural Machinery Engineering Faculty of Agricultural Engineering and Technology University of Tehran Karaj, Iran* 

#### **1. Introduction**

134 Energy Efficiency – A Bridge to Low Carbon Economy

Eckerle (2007) Hydraulic division. EIPS 2 Internal gear pumps.[Online]. [Accessed 20 April

2008], available at: http: //www.eckerle.com.

China.

IEEE Vehicle Power and Propulsion Conference (VPPC), September 3–5, Harbin,

The relation between agriculture and energy is very close. Agricultural sector itself is an energy user and energy supplier in the form of bio-energy (Alam et al., 2005). It uses large quantities of locally available non-commercial energies, such as seed, farmyard manure and animate energy, and commercial energies directly and indirectly in the form of electricity, diesel fuel, chemical fertilizers, plant protections, irrigation water and farm machinery (Kizilaslan, 2009).

Nowadays, energy usage in agricultural activities has been intensified in response to continued growth of human population, tendency for an overall improved standard of living and limited supply of arable land. Consequently, additional use of energy causes problems threatening public health and environment (Rafiee et al., 2010). However, increased energy use in order to obtain maximum yields may not bring maximum profits due to increasing production costs. In addition, both the natural resources are rapidly decreasing and the amount of contaminants on the environment is considerably increasing (Esengun et al., 2007).

Efficiency is defined as the ability to produce the outputs with a minimum resource level required (Sherman, 1988). In production, efficiency is a normative measure and is defined as the ratio of weighted sum of outputs to inputs or as the actual output to the optimal output ratio. The weights for inputs and outputs are estimated to the best advantage for each unit so as to maximize its relative efficiency. In order to measure the optimal input or output, it is necessary to first specify the production frontier (Mukherjee, 2008).

Efficient use of energy resources in agriculture is one of the principal requirements for sustainable agricultural productions; it provides financial savings, fossil resources preservation and air pollution reduction; for enhancing the energy efficiency it must be attempted to increase the production yield or to conserve the energy input without affecting the output (Singh et al., 2004). Therefore, energy saving has been a crucial issue for

Energy Efficiency Analysis in Agricultural Productions:

of the observations (Zhou et al., 2008).

between the farms.

energies; while, the soybean yield was the single output.

Parametric and Non-Parametric Approaches 137

Mousavi-Avval et al. (2011b) analyzed the energy efficiency of soybean producers using parametric approach. In this study the Cobb-Douglas production function was applied to develop an aconometric model between inputs and output. The inputs were human labor, machinery, diesel fuel, chemicals, fertilizers, water for irrigation, electricity and seed

Data Envelopment Analysis (DEA) is a non-parametric linear programming (LP) based technique of frontier estimation for measuring the relative efficiency of a number of decision making units (DMUs) on the basis of multiple inputs and outputs (Zhang et al., 2009). The main advantage of non-parametric method of DEA compared to parametric approaches is that it does not require any prior assumption on the underlying functional relationship between inputs and outputs. It is, therefore, a non-parametric approach. In addition, DEA is a data-driven frontier analysis technique that floats a piecewise linear surface to rest on top

Due to the high advantages of DEA, there are a large number of its applications for evaluating the performances of DMUs in different issues. Also, it currently has been employed in some agricultural enterprises. In an earlier and related study, DEA was utilized to evaluate the technical efficiency of irrigated dairy farms in Australia. The results from this study proposed that DEA was a useful tool in helping to benchmark the dairy industry, which is continually striving to improve the productive efficiency of farms (Fraser and Cordina, 1999). Also, Dawson et al. (2000) presented technical and overall economic efficiencies for 22 rice farms in Philippines, using a frontier production function approach. The results of their study showed that the overall efficiency was changed from 84% to 95%

In another study, DEA was applied to investigate the efficiency of individual farmers and to identify the efficient units for citrus farming in Spain (Reig-Martínez and Picazo-Tadeo, 2004); also, Barnes (2006) identified the technical efficiency scores of Scottish dairy farms by applying the DEA approach. Malana and Malano (2006) employed the DEA technique to

In another study by Nassiri and Singh (2009), the DEA technique was subjected to the data of energy use from different inputs by individual paddy producers and the technical, pure technical and scale efficiencies of farmers were estimated. The results showed that, there was high correlation between technical efficiency and energy-ratio, however comparison between correlation coefficient of farmers in different farm categories and different zones showed that energy-ratio and specific energy were not enhanced indices for explaining of all kinds of the technical, pure technical and scale efficiency of farmers. Omid et al. (2010) employed this technique to analyze technical and scale efficiencies of cucumber producers. Oilseed sunflower (*Helianthus annuus L.*) is one of the most widely cultivated and important oilseed crops in the world (Latif and Anwar, 2009). Sunflower seeds are mainly used for the production of oil for human consumptions. They contain a high amount of oil (26%) which is an important source of polyunsaturated fatty acid (linoleic acid) of potential health benefits (Pimentel and Patzek, 2005). The seeds are also used as a protein source for nonruminant and ruminant animals. Iran produced more than 43000 tones of sunflower seeds from about 67000 ha harvested land area, in 2008 (FAO, 2008). In Iran, sunflower is mainly

benchmark the productive efficiency of irrigated wheat area in Pakistan and India.

grown in Golestan province in the north-east of the country (Anonymous, 2010).

sustainable development in agricultural systems. Development of energy efficient agricultural systems with low input energy compared to the output of food can reduces the greenhouse gas emissions from agricultural production systems.

Improvements in the efficiency of resource use in agriculture require not only the definition of spatial and temporal use of current resources but also the development of tightly defined and broadly applicable indices (Topp et al., 2007).

In some studies the indicators of output energy to input energy ratio and energy productivity (i.e. yield to input energy ratio) in crop production systems have been used to evaluate the performance of farmers (Mohammadi et al., 2010; Unakitan et al., 2010). Energy productivity is an important indicator for more efficient use of energy although higher energy productivity does not mean in general, more economic feasibility (Mohammadi et al., 2010). The energy input-output analysis is usually made to measure the energy efficiency and environmental aspects. This analysis will determine how efficient the energy is used. In current years, several researches have been conducted on energy use for production of different agricultural crops (Jianbo, 2006; Meul et al., 2007; Kizilaslan, 2009).

Moreover, in some studies the parametric and non-parametric approaches have been used to analyze the efficiency of farmers in agricultural productions. In parametric approach, an econometric model is used to identify the relationship between energy inputs and yield values of crop productions. In this method, the parameters of the production or cost functions are estimated statistically.

Establishing the functional forms between energy inputs and output for agricultural crops are very useful in terms of determining elasticity of different energy inputs on yield (Turhan et al., 2008). Development of a model consists of several logical steps; one of them is the sensitivity analysis to ascertain how a given model depends on its input factors (Hamby, 1994). Sensitivity analysis quantifies the sensitivity of a model's state variables to the parameters defining the model. It refers to changes in the response of each of the state variables which result from small changes in the parameter values. Sensitivity analysis for the parameters of a developed model is valuable because it identifies those parameters which have most influence on the response of the model (Chalabi and Bailey, 1991). The sensitivity analysis of energy inputs on crop production is essential because it revealed what changes in energy inputs cause greater impacts on the output. Furthermore, it is of especial importance for the policy-makers to frame suitable policies for improving energy use efficiency (Lamoureux et al., 2006).

In recent years, many researchers have developed econometric models between energy inputs and output for different agricultural crops (Banaeian et al., 2010; Mohammadi and Omid, 2010). Singh et al. (2004) and Rafiee et al. (2010) investigated energy inputs and crop yield relationship to develop an econometric model for wheat and apple productions, respectively. Moreover, they applied the marginal physical productivity (MPP) technique to analyze the sensitivity of energy inputs on yield. Kulekci (2010) applied the stochastic frontier analysis technique in the Cobb-Douglas form to determine the technical efficiency for a sample of 117 randomly selected sunflower farms in Erzurum, Turkey. This method is parametric and uses statistical techniques to estimate the parameters of the function; however, this approach requires a pre-specification of the functional form and an explicit distributional assumption for the technical inefficiency term.

sustainable development in agricultural systems. Development of energy efficient agricultural systems with low input energy compared to the output of food can reduces the

Improvements in the efficiency of resource use in agriculture require not only the definition of spatial and temporal use of current resources but also the development of tightly defined

In some studies the indicators of output energy to input energy ratio and energy productivity (i.e. yield to input energy ratio) in crop production systems have been used to evaluate the performance of farmers (Mohammadi et al., 2010; Unakitan et al., 2010). Energy productivity is an important indicator for more efficient use of energy although higher energy productivity does not mean in general, more economic feasibility (Mohammadi et al., 2010). The energy input-output analysis is usually made to measure the energy efficiency and environmental aspects. This analysis will determine how efficient the energy is used. In current years, several researches have been conducted on energy use for production of

Moreover, in some studies the parametric and non-parametric approaches have been used to analyze the efficiency of farmers in agricultural productions. In parametric approach, an econometric model is used to identify the relationship between energy inputs and yield values of crop productions. In this method, the parameters of the production or cost

Establishing the functional forms between energy inputs and output for agricultural crops are very useful in terms of determining elasticity of different energy inputs on yield (Turhan et al., 2008). Development of a model consists of several logical steps; one of them is the sensitivity analysis to ascertain how a given model depends on its input factors (Hamby, 1994). Sensitivity analysis quantifies the sensitivity of a model's state variables to the parameters defining the model. It refers to changes in the response of each of the state variables which result from small changes in the parameter values. Sensitivity analysis for the parameters of a developed model is valuable because it identifies those parameters which have most influence on the response of the model (Chalabi and Bailey, 1991). The sensitivity analysis of energy inputs on crop production is essential because it revealed what changes in energy inputs cause greater impacts on the output. Furthermore, it is of especial importance for the policy-makers to frame suitable policies for improving energy use

In recent years, many researchers have developed econometric models between energy inputs and output for different agricultural crops (Banaeian et al., 2010; Mohammadi and Omid, 2010). Singh et al. (2004) and Rafiee et al. (2010) investigated energy inputs and crop yield relationship to develop an econometric model for wheat and apple productions, respectively. Moreover, they applied the marginal physical productivity (MPP) technique to analyze the sensitivity of energy inputs on yield. Kulekci (2010) applied the stochastic frontier analysis technique in the Cobb-Douglas form to determine the technical efficiency for a sample of 117 randomly selected sunflower farms in Erzurum, Turkey. This method is parametric and uses statistical techniques to estimate the parameters of the function; however, this approach requires a pre-specification of the functional form and an explicit

different agricultural crops (Jianbo, 2006; Meul et al., 2007; Kizilaslan, 2009).

greenhouse gas emissions from agricultural production systems.

and broadly applicable indices (Topp et al., 2007).

functions are estimated statistically.

efficiency (Lamoureux et al., 2006).

distributional assumption for the technical inefficiency term.

Mousavi-Avval et al. (2011b) analyzed the energy efficiency of soybean producers using parametric approach. In this study the Cobb-Douglas production function was applied to develop an aconometric model between inputs and output. The inputs were human labor, machinery, diesel fuel, chemicals, fertilizers, water for irrigation, electricity and seed energies; while, the soybean yield was the single output.

Data Envelopment Analysis (DEA) is a non-parametric linear programming (LP) based technique of frontier estimation for measuring the relative efficiency of a number of decision making units (DMUs) on the basis of multiple inputs and outputs (Zhang et al., 2009). The main advantage of non-parametric method of DEA compared to parametric approaches is that it does not require any prior assumption on the underlying functional relationship between inputs and outputs. It is, therefore, a non-parametric approach. In addition, DEA is a data-driven frontier analysis technique that floats a piecewise linear surface to rest on top of the observations (Zhou et al., 2008).

Due to the high advantages of DEA, there are a large number of its applications for evaluating the performances of DMUs in different issues. Also, it currently has been employed in some agricultural enterprises. In an earlier and related study, DEA was utilized to evaluate the technical efficiency of irrigated dairy farms in Australia. The results from this study proposed that DEA was a useful tool in helping to benchmark the dairy industry, which is continually striving to improve the productive efficiency of farms (Fraser and Cordina, 1999). Also, Dawson et al. (2000) presented technical and overall economic efficiencies for 22 rice farms in Philippines, using a frontier production function approach. The results of their study showed that the overall efficiency was changed from 84% to 95% between the farms.

In another study, DEA was applied to investigate the efficiency of individual farmers and to identify the efficient units for citrus farming in Spain (Reig-Martínez and Picazo-Tadeo, 2004); also, Barnes (2006) identified the technical efficiency scores of Scottish dairy farms by applying the DEA approach. Malana and Malano (2006) employed the DEA technique to benchmark the productive efficiency of irrigated wheat area in Pakistan and India.

In another study by Nassiri and Singh (2009), the DEA technique was subjected to the data of energy use from different inputs by individual paddy producers and the technical, pure technical and scale efficiencies of farmers were estimated. The results showed that, there was high correlation between technical efficiency and energy-ratio, however comparison between correlation coefficient of farmers in different farm categories and different zones showed that energy-ratio and specific energy were not enhanced indices for explaining of all kinds of the technical, pure technical and scale efficiency of farmers. Omid et al. (2010) employed this technique to analyze technical and scale efficiencies of cucumber producers.

Oilseed sunflower (*Helianthus annuus L.*) is one of the most widely cultivated and important oilseed crops in the world (Latif and Anwar, 2009). Sunflower seeds are mainly used for the production of oil for human consumptions. They contain a high amount of oil (26%) which is an important source of polyunsaturated fatty acid (linoleic acid) of potential health benefits (Pimentel and Patzek, 2005). The seeds are also used as a protein source for nonruminant and ruminant animals. Iran produced more than 43000 tones of sunflower seeds from about 67000 ha harvested land area, in 2008 (FAO, 2008). In Iran, sunflower is mainly grown in Golestan province in the north-east of the country (Anonymous, 2010).

Energy Efficiency Analysis in Agricultural Productions:

2011):

machine (h).

Parametric and Non-Parametric Approaches 139

power used in different field operations of crop production. Total energy embodied in machinery included energy for raw materials, manufacturing, repairs and maintenance, and energy for transportation. Taking into account the total weight and the life of machinery as used in practice, the energy required for each operation was calculated assuming that the embodied energy of tractors and agricultural machinery be depreciated during their economical life time (Beheshti Tabar et al., 2010); so, the embodied energy in machinery was calculated by multiplying the depreciated weight of machinery (kg ha-1) with their energy equivalents (MJ kg-1). Also, the weight of machinery depreciated per hectare of sunflower production during the production period was calculated as follows (Mousavi Avval et al.,

> *G Wh TW T*

where *TW* is the depreciated machinery weight (kg ha-1); *G* is the total machine weight (kg); *Wh* is the time that machine used per unit area (h ha-1) and *T* is the economical life time of

Depending on the context, manure may be considered either a valuable source of nutrients replacing synthetic fertilizers, a waste product from livestock production, or a potential energy source, e.g. for biogas production. In this study, we regarded manure as a source of nutrients, and the substitution method was used to calculate the energy input of animal manure, which equates the energy equivalent of farmyard manure with that of mineral

The energy equivalent of water for irrigation input means indirect energy of irrigation consist of the energy consumed for manufacturing the materials for the dams, canals, pipes, pumps, and equipments as well as the energy for constructing the works and building the

Following the calculation of energy input and output equivalents, the indices of energy ratio (energy use efficiency), energy productivity, specific energy (energy intensity) and net

*Energy output MJ ha Energy ratio Energy input MJ ha*

*Sunflower yield kg ha Energy productivity Energy input MJ ha*

*Energy input MJ ha Specific energy Sunflower yield kg ha*

Energy ratio index is the ratio between the caloric heat of the output products and the total sequestered energy in the production factors. This index allows us to know the influence of inputs expressed in energy units in obtaining output energy. To improve energy ratio in a

fertilizer equivalents corresponding to its fertilization effect (Liu et al., 2010).

on-farm irrigation systems (Khan et al., 2009).

energy were calculated as follow (Rafiee et al., 2010):

(2)

1 1 ( ) ( )

> 1 1 ( ) ( )

(5)

 (4)

> 1 1 ( ) ( )

1 1 *Net energy Energy output MJ ha Energy input MJ ha* () () (6)

 (3)

This study focuses on the capability of parametric and non-parametric approaches in energy efficiency analysis for agricultural crop productions. For this purpose the energy balance for sunflower production was investigated and the efficiency of energy use was analyzed using the parametric Cobb-Douglass production function and non-parametric data envelopment analysis techniques.

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

#### **2.1 Data collection**

In this study Golestan province was chosen as a representative of the Iranian sunflower production enterprises since it is the main center of oilseed productions, especially sunflower crop, in the country, mainly due to the very favorable ecological conditions. Other oilseed crops cultivated in this province are soybean and canola.

Golestan province is located in the north-east of Iran, within 36° 30' and 38° 08' north latitude and 53° 57' and 56° 22' east longitude. A survey approach was used to collect the quantitative information on energy inputs used for the production of sunflower in the production period of 2009/10. The required sample size was determined using simple random sampling method as below (Cochran, 1977):

$$m = \frac{N \times s^2 \times t^2}{(N-1)d^2 + (s^2 \times t^2)}\tag{1}$$

where *n* is the required sample size, *N* is the number of oilseed sunflower producers in target population, *s* is the standard deviation, *t* is the *t*-value at 95% confidence limit (1.96), and *d* is the acceptable error. The permissible error in the sample size was defined to be 5% for 95% confidence. Thus the sample size was found to be 95, and then, 95 farmers from the population were randomly selected. Before collecting data, the survey form was pre-tested by a group of randomly selected farmers and these pre-tested surveys were not included in the final data set. The reliability of the questionnaires was tested using *Cronbach's alpha*. So, the *Cronbach's alpha* level of 0.7 demonstrated adequate construct reliability.

#### **2.2 Energy balance analysis method**

A standard procedure was used to convert each agricultural input and output into energy equivalents. The inputs used in sunflower production were in the form of chemical fertilizers (nitrogen, phosphate, potassium and sulfur), chemical biocides (herbicides, fungicides and insecticides), diesel fuel, electricity, farmyard manure, water for irrigation, human labor and machine power. Also, the grain yield was considered as output.

The energy equivalent may thus be defined as the energy input taking into account all forms of energy in agricultural productions. The energy equivalents were computed for all inputs and outputs using the conversion factors indicated in Table 1. Multiplying the quantity of the inputs used per hectare with their conversion factors gave the energy equivalents.

The energy equivalent associated with labor vary considerably, depending on the approach chosen; it must be adapted to the actual living conditions in the target region (Moore, 2010). In this study the energy coefficient of 1.96 MJ h-1 was applied. It means only the muscle

This study focuses on the capability of parametric and non-parametric approaches in energy efficiency analysis for agricultural crop productions. For this purpose the energy balance for sunflower production was investigated and the efficiency of energy use was analyzed using the parametric Cobb-Douglass production function and non-parametric data envelopment

In this study Golestan province was chosen as a representative of the Iranian sunflower production enterprises since it is the main center of oilseed productions, especially sunflower crop, in the country, mainly due to the very favorable ecological conditions.

Golestan province is located in the north-east of Iran, within 36° 30' and 38° 08' north latitude and 53° 57' and 56° 22' east longitude. A survey approach was used to collect the quantitative information on energy inputs used for the production of sunflower in the production period of 2009/10. The required sample size was determined using simple

> *2 2 2 22 Ns t*

(1)

*(N 1)d (s t )* 

where *n* is the required sample size, *N* is the number of oilseed sunflower producers in target population, *s* is the standard deviation, *t* is the *t*-value at 95% confidence limit (1.96), and *d* is the acceptable error. The permissible error in the sample size was defined to be 5% for 95% confidence. Thus the sample size was found to be 95, and then, 95 farmers from the population were randomly selected. Before collecting data, the survey form was pre-tested by a group of randomly selected farmers and these pre-tested surveys were not included in the final data set. The reliability of the questionnaires was tested using *Cronbach's alpha*. So,

A standard procedure was used to convert each agricultural input and output into energy equivalents. The inputs used in sunflower production were in the form of chemical fertilizers (nitrogen, phosphate, potassium and sulfur), chemical biocides (herbicides, fungicides and insecticides), diesel fuel, electricity, farmyard manure, water for irrigation,

The energy equivalent may thus be defined as the energy input taking into account all forms of energy in agricultural productions. The energy equivalents were computed for all inputs and outputs using the conversion factors indicated in Table 1. Multiplying the quantity of the inputs used per hectare with their conversion factors gave the energy equivalents.

The energy equivalent associated with labor vary considerably, depending on the approach chosen; it must be adapted to the actual living conditions in the target region (Moore, 2010). In this study the energy coefficient of 1.96 MJ h-1 was applied. It means only the muscle

Other oilseed crops cultivated in this province are soybean and canola.

*n*

the *Cronbach's alpha* level of 0.7 demonstrated adequate construct reliability.

human labor and machine power. Also, the grain yield was considered as output.

random sampling method as below (Cochran, 1977):

**2.2 Energy balance analysis method** 

analysis techniques.

**2.1 Data collection** 

**2. Materials and methods** 

power used in different field operations of crop production. Total energy embodied in machinery included energy for raw materials, manufacturing, repairs and maintenance, and energy for transportation. Taking into account the total weight and the life of machinery as used in practice, the energy required for each operation was calculated assuming that the embodied energy of tractors and agricultural machinery be depreciated during their economical life time (Beheshti Tabar et al., 2010); so, the embodied energy in machinery was calculated by multiplying the depreciated weight of machinery (kg ha-1) with their energy equivalents (MJ kg-1). Also, the weight of machinery depreciated per hectare of sunflower production during the production period was calculated as follows (Mousavi Avval et al., 2011):

$$T\mathcal{W} = \frac{G \times \mathcal{W}\_{\text{fr}}}{T} \tag{2}$$

where *TW* is the depreciated machinery weight (kg ha-1); *G* is the total machine weight (kg); *Wh* is the time that machine used per unit area (h ha-1) and *T* is the economical life time of machine (h).

Depending on the context, manure may be considered either a valuable source of nutrients replacing synthetic fertilizers, a waste product from livestock production, or a potential energy source, e.g. for biogas production. In this study, we regarded manure as a source of nutrients, and the substitution method was used to calculate the energy input of animal manure, which equates the energy equivalent of farmyard manure with that of mineral fertilizer equivalents corresponding to its fertilization effect (Liu et al., 2010).

The energy equivalent of water for irrigation input means indirect energy of irrigation consist of the energy consumed for manufacturing the materials for the dams, canals, pipes, pumps, and equipments as well as the energy for constructing the works and building the on-farm irrigation systems (Khan et al., 2009).

Following the calculation of energy input and output equivalents, the indices of energy ratio (energy use efficiency), energy productivity, specific energy (energy intensity) and net energy were calculated as follow (Rafiee et al., 2010):

$$Energy\ ratio = \frac{Energy\ output\ \text{(MJ/ha}^{-1}\text{)}}{Energy\ input\ \text{(MJ/ha}^{-1}\text{)}}\tag{3}$$

$$Energy\,\,\text{probability} = \frac{\text{Sumflower\,\,yield} \,\text{(kg\,ha}^{-1}\text{)}}{\text{Energy\,\,input\,\,[M]\,\,ha}^{-1}}\tag{4}$$

$$\text{Specific energy} = \frac{\text{Energy input(M/ha}^{-1})}{\text{Sunflow yield(kg/ha}^{-1})} \tag{5}$$

$$\text{Net energy} = \text{Energy output} \left( \text{M} \,\text{h}\,\text{h}^{-1} \right) - \text{Energy input} \left( \text{M} \,\text{h}\,\text{h}^{-1} \right) \tag{6}$$

Energy ratio index is the ratio between the caloric heat of the output products and the total sequestered energy in the production factors. This index allows us to know the influence of inputs expressed in energy units in obtaining output energy. To improve energy ratio in a

Energy Efficiency Analysis in Agricultural Productions:

difficult to separate inefficiency from random error.

of statistical significance and expected signs of parameters.

0 1

which is estimated from the model and *ui*, is the error term.

reduces to (Mousavi-Avval et al., 2011b; Samavatean et al., 2011):

*i*

a 1% increase in the *j*th input.

seed (*X8*) in MJ per hectare unit.

specified as the following forms:

*i ij j*

It can be specified in a mathematical form as follows (Singh et al., 2004):

*<sup>k</sup> <sup>α</sup> <sup>u</sup>*

Using a linear presentation, the function to be estimated could be written as:

ln ln *k i j ij i j Y α α X u* 

productions (Singh et al., 2004; Mobtaker et al., 2010).

Parametric and Non-Parametric Approaches 141

parameters of the production or cost functions statistically; however, in this approach it is

In this study following the estimation of energy balance, the relation between energy inputs and output was investigated using a prior mathematical function relation. In specifying a fit relation, the Cobb-Douglass production function was selected as the best function in terms

The Cobb-Douglass function is a power relation has been used by several authors to investigate the relationship between energy inputs and output in agricultural crop

*Y α X e ( i , ,...,n; j , ,...,*

<sup>0</sup> 1

where: *Yi*, denotes the yield of the *ith* farmer, *Xij*, is the *j*th input used by the *i*th farmer for the cultivation of crop, *α*0, is a constant term, *αj*, represent the regression coefficients of *j*th input,

In this functional form the parameters to be estimated *αj* represent the elasticity of output with respect to each input *j* which implies the percent change in output augmentation from

Assuming that when the energy input is zero, the crop production is also zero, Eq. (8)

1 ln ln *k*

6 7 78 8 ln ln ln ln ln ln ln ln ln

*j*

energy on sunflower yield, the Eq. (9) can be expanded in the following form;

1 12 23 34 45 5

*i j ij i*

*Y α X u*

Assuming that yield is a function of energy inputs, for investigating the impact of each input

*Y α X α X α X α X α X α X α X α X u* 

 where *Xj* (*i=1, 2, …, 8*) stand for energy inputs of human labor (*X1*), machinery (*X2*),diesel fuel (*X3*), chemicals (*X4*), total fertilizers (*X5*), water for irrigation (*X6*), electricity (*X7*) and

With respect to this pattern, the impacts of DE and IDE, and the effect of RE and NRE forms on the production yield were investigated. So, the Cobb-Douglass function was selected and

12 12 ) *<sup>j</sup> <sup>i</sup>*

**<sup>k</sup>** (7)

(8)

(9)

*i*

(10)


Table 1. Energy equivalent of inputs and output in sunflower production.

process, it is possible either to reduce the energy sequestered in the inputs by optimization of energy use or to increase the yield of product by reducing the losses (Kitani, 1999). Energy productivity is the measure of the amount of a product obtained per unit of input energy. Also, net energy gain is the difference between the gross energy output produced and the total energy required for obtaining it.

The energy associated with inputs comes from different sources which classified as renewable and non-renewable energy forms. Renewable energy (RE) consists of water for irrigation, human labor and seeds, whereas non-renewable energy (NRE) includes machinery, diesel fuel, electricity, fertilizers and chemicals energy inputs. On the other hand, energy demand in agricultural productions can be classified in two main groups including direct and indirect forms. Direct energy (DE) form covers human labor, diesel fuel, water for irrigation and electrical energy; while, indirect energy (IDE) includes energy embodied in fertilizers, chemicals, seeds and machinery (Mobtaker et al., 2010).

#### **2.3 Parametric approach**

#### **2.3.1 Model development**

There are several parametric and non-parametric techniques to measure productive efficiency. Parametric methods assume a particular functional form (e. g., Cobb-Douglas production function or a Translog function) between inputs and output and estimate the

1. Human labor h 1.96 ( Rafiee et al., 2010)

a. Tractor 93.61 ( Canakci et al., 2005) b. Self propelled combine 87.63 ( Canakci et al., 2005) c. Other machinery 62.70 ( Canakci et al., 2005) 3. Diesel fuel L 47.80 ( Canakci et al., 2005)

a. Herbicides 238.00 ( Erdal et al., 2007) b. Fungicides 216.00 ( Erdal et al., 2007) c. Insecticides 101.20 ( Erdal et al., 2007)

a. Nitrogen 66.14 ( Rafiee et al., 2010) b. Phosphate (P2O5) 12.44 ( Rafiee et al., 2010) c. Potassium (K2O) 11.15 ( Rafiee et al., 2010) d. Sulfur (S) 1.12 ( Rafiee et al., 2010) e. Farmyard manure 0.30 ( Rafiee et al., 2010) 6. Water for irrigation m3 1.02 ( Rafiee et al., 2010)

7. Seed kg 3.60 (Beheshti Tabar et al., 2010) 8. Electricity kWh 11.93 (Mobtaker et al., 2010)

1. Sunflower kg 25.00 (Beheshti Tabar et al., 2010)

process, it is possible either to reduce the energy sequestered in the inputs by optimization of energy use or to increase the yield of product by reducing the losses (Kitani, 1999). Energy productivity is the measure of the amount of a product obtained per unit of input energy. Also, net energy gain is the difference between the gross energy output produced

The energy associated with inputs comes from different sources which classified as renewable and non-renewable energy forms. Renewable energy (RE) consists of water for irrigation, human labor and seeds, whereas non-renewable energy (NRE) includes machinery, diesel fuel, electricity, fertilizers and chemicals energy inputs. On the other hand, energy demand in agricultural productions can be classified in two main groups including direct and indirect forms. Direct energy (DE) form covers human labor, diesel fuel, water for irrigation and electrical energy; while, indirect energy (IDE) includes energy

There are several parametric and non-parametric techniques to measure productive efficiency. Parametric methods assume a particular functional form (e. g., Cobb-Douglas production function or a Translog function) between inputs and output and estimate the

embodied in fertilizers, chemicals, seeds and machinery (Mobtaker et al., 2010).

Table 1. Energy equivalent of inputs and output in sunflower production.

(MJ unit-1) Reference

Inputs Unit Energy equivalent

2. Machinery kg

4. Chemicals kg

5. Fertilizer kg

and the total energy required for obtaining it.

**2.3 Parametric approach 2.3.1 Model development** 

A. Inputs

B. Output

parameters of the production or cost functions statistically; however, in this approach it is difficult to separate inefficiency from random error.

In this study following the estimation of energy balance, the relation between energy inputs and output was investigated using a prior mathematical function relation. In specifying a fit relation, the Cobb-Douglass production function was selected as the best function in terms of statistical significance and expected signs of parameters.

The Cobb-Douglass function is a power relation has been used by several authors to investigate the relationship between energy inputs and output in agricultural crop productions (Singh et al., 2004; Mobtaker et al., 2010).

It can be specified in a mathematical form as follows (Singh et al., 2004):

$$Y\_i = a\_0 \prod\_{j=1}^k X\_{ij}^{a\_j} \ e^{u\_i} \qquad \text{ (i = 1, 2, \dots, n; j = 1, 2, \dots, k)}\tag{7}$$

Using a linear presentation, the function to be estimated could be written as:

$$\ln Y\_i = a\_0 + \sum\_{j=1}^{k} a\_j \ln \left( X\_{ij} \right) + u\_i \tag{8}$$

where: *Yi*, denotes the yield of the *ith* farmer, *Xij*, is the *j*th input used by the *i*th farmer for the cultivation of crop, *α*0, is a constant term, *αj*, represent the regression coefficients of *j*th input, which is estimated from the model and *ui*, is the error term.

In this functional form the parameters to be estimated *αj* represent the elasticity of output with respect to each input *j* which implies the percent change in output augmentation from a 1% increase in the *j*th input.

Assuming that when the energy input is zero, the crop production is also zero, Eq. (8) reduces to (Mousavi-Avval et al., 2011b; Samavatean et al., 2011):

$$\ln Y\_i = \sum\_{j=1}^{k} a\_j \ln \left( X\_{ij} \right) + u\_i \tag{9}$$

Assuming that yield is a function of energy inputs, for investigating the impact of each input energy on sunflower yield, the Eq. (9) can be expanded in the following form;

$$\begin{aligned} \ln Y\_i &= a\_1 \ln X\_1 + a\_2 \ln X\_2 + a\_3 \ln X\_3 + a\_4 \ln X\_4 + a\_5 \ln X\_5 \\ &+ a\_6 \ln X + a\_7 \ln X\_7 + a\_8 \ln X\_8 + u\_i \end{aligned} \tag{10}$$

 where *Xj* (*i=1, 2, …, 8*) stand for energy inputs of human labor (*X1*), machinery (*X2*),diesel fuel (*X3*), chemicals (*X4*), total fertilizers (*X5*), water for irrigation (*X6*), electricity (*X7*) and seed (*X8*) in MJ per hectare unit.

With respect to this pattern, the impacts of DE and IDE, and the effect of RE and NRE forms on the production yield were investigated. So, the Cobb-Douglass function was selected and specified as the following forms:

Energy Efficiency Analysis in Agricultural Productions:

those with the highest level of productive efficiency.

inputs consumed in the production process.

 *Max:* <sup>1</sup>

**2.4.1 Technical efficiency (TE)** 

analyzed.

**2.4 Non-parametric approach** 

also, each farmer called a DMU.

Parametric and Non-Parametric Approaches 143

Apart from the parametric approach, in this study a non-parametric method of DEA was employed to evaluate the technical, pure technical and scale efficiencies of individual farmers which use similar inputs, produce the same product (sunflower) and operate in a relatively homogenous region (e. g., topography, soil type, climatic conditions, etc.). So, energy consumptions from different inputs including human labor, machinery, diesel fuel, chemicals, total fertilizers, water for irrigation, electricity and seed energies in terms of MJ ha-1 were considered as inputs; while, the grain yield of sunflower was the single output;

DEA technique builds a linear piece-wise function from empirical observations of inputs and outputs. DEA is a nonparametric approach for estimating productive efficiency based on mathematical linear programming techniques. Unlike parametric methods, DEA does not require a function to relate inputs and outputs. The DEA envelops the data in such a way that all observed data points lie on or below the efficient frontier (Coelli, 1996). The efficient frontier is established by efficient units from a group of observed units. Efficient units are

In DEA an inefficient DMU can be made efficient either by minimizing the input levels while maintaining the same level of outputs (input oriented), or, symmetrically, by maximizing the output levels while holding the inputs constant (output oriented). Sunflower production similar to wheat (Malana and Malano, 2006), paddy (Chauhan et al., 2006) and greenhouse cucumber (Omid et al., 2010) productions relies on finite and scarce resources; therefore the use of input-oriented DEA models is more appropriate to reduce

In this study, the technical, pure technical and scale efficiencies of farmers were

TE can be defined as the ability of a DMU (e.g. a farm) to produce maximum output given a set of inputs and technology level. The TE score (*θ*) in the presence of multiple-input and output factor can be calculated by the ratio of sum of weighted outputs to the sum of

> 11 22 1 11 22 <sup>1</sup>

Let the DMU*j* to be evaluated on any trial be designated as DMU*o* (*o* = 1, 2, . . ., *n*). To measure the relative efficiency of a DMU*o* based on a series of *n* DMUs, the model is

*u y u y u y u y vx vx v x v x*

*j j s sj r rj r <sup>m</sup> j j m mj i ij <sup>i</sup>*

1

*u y v x*

*s r ro r m i io i*

*s*

(14)

(15)

... ...

structured as a fractional programming problem as follows (Cooper et al., 2006):

weighted inputs or in a mathematical expression as follows (Cooper et al., 2004):

$$
\ln Y\_i = \beta\_1 \ln DE + \beta\_2 \ln IDE + u\_i \tag{11}
$$

$$
\ln Y\_i = \chi\_1 \ln RE + \chi\_2 \ln NRE + \mu\_\mathbf{i} \tag{12}
$$

where *Yi* is the *i*th farmer's yield (kg ha-1), *β* and *γ* are coefficient of exogenous variables. Eqns. (10) to (12) were estimated using Ordinary Least Square (OLS) technique.

#### **2.3.2 Returns to scale**

Following the estimation of econometric model, in order to describe the changes in output subsequent to a proportional change in all the inputs (when all inputs change by a constant factor) the return to scale index was investigated (Rafiee et al., 2010). In a Cobb-Douglas production function, the sum of elasticity values derived in the form of regression coefficients represent the degree of returns to scale. If the sum of coefficients is more than, equal to, or less than one, implying that there is increasing, constant, or decreasing returns to scale, respectively (Rafiee et al., 2010). Increasing (decreasing) returns to scale indicate that an increase in the input resources produces more (less) than proportionate increase in outputs. In this study, an increasing, constant and decreasing return to scale indicate that when all of the energy inputs are increased by *X* value, then the sunflower yield increases by more than, exactly and less than *X* value, respectively.

#### **2.3.3 Sensitivity analysis**

Since several parameters affect the model output, the sensitivity analysis of energy inputs on yield was used to identify which factors had a greater effect on the production yield. For this purpose, the marginal physical productivity (MPP) method, based on the response coefficients of the inputs was utilized. The MPP of a factor implies the change in the total output with a unit change in the factor input, assuming all other factors are fixed at their geometric mean level.

A positive value of MPP of any input variable identifies that the total output is increasing with an increase in input; so, one should not stop increasing the use of variable inputs so long as the fixed resource is not fully utilized. A negative value of MPP of any variable input indicates that every additional unit of input starts to diminish the total output of previous units; therefore, it is better to keep the variable resource in surplus rather than utilizing it as a fixed resource.

The MPP of the various inputs was calculated using the *αj* of the various energy inputs as follows (Singh et al., 2004; Rafiee et al., 2010):

$$MPP\_{\mathbf{x}\mathbf{j}} = \frac{GM(\mathbf{Y})}{GM(\mathbf{X}\_{\mathbf{j}})} \times a\_{\mathbf{j}} \tag{13}$$

where *MPPxj* is the marginal physical productivity of *j*th input, *αj*, the regression coefficient of *j*th input, *GM(Y)*, geometric mean of yield, and *GM (Xj*), the geometric mean of *j*th input energy on per hectare basis.

#### **2.4 Non-parametric approach**

142 Energy Efficiency – A Bridge to Low Carbon Economy

where *Yi* is the *i*th farmer's yield (kg ha-1), *β* and *γ* are coefficient of exogenous variables.

Following the estimation of econometric model, in order to describe the changes in output subsequent to a proportional change in all the inputs (when all inputs change by a constant factor) the return to scale index was investigated (Rafiee et al., 2010). In a Cobb-Douglas production function, the sum of elasticity values derived in the form of regression coefficients represent the degree of returns to scale. If the sum of coefficients is more than, equal to, or less than one, implying that there is increasing, constant, or decreasing returns to scale, respectively (Rafiee et al., 2010). Increasing (decreasing) returns to scale indicate that an increase in the input resources produces more (less) than proportionate increase in outputs. In this study, an increasing, constant and decreasing return to scale indicate that when all of the energy inputs are increased by *X* value, then the sunflower yield increases

Since several parameters affect the model output, the sensitivity analysis of energy inputs on yield was used to identify which factors had a greater effect on the production yield. For this purpose, the marginal physical productivity (MPP) method, based on the response coefficients of the inputs was utilized. The MPP of a factor implies the change in the total output with a unit change in the factor input, assuming all other factors are fixed at their

A positive value of MPP of any input variable identifies that the total output is increasing with an increase in input; so, one should not stop increasing the use of variable inputs so long as the fixed resource is not fully utilized. A negative value of MPP of any variable input indicates that every additional unit of input starts to diminish the total output of previous units; therefore, it is better to keep the variable resource in surplus rather than utilizing it as

The MPP of the various inputs was calculated using the *αj* of the various energy inputs as

*GM(Y) MPP <sup>α</sup>*

where *MPPxj* is the marginal physical productivity of *j*th input, *αj*, the regression coefficient of *j*th input, *GM(Y)*, geometric mean of yield, and *GM (Xj*), the geometric mean of *j*th input

*xj j j*

*GM(X )* (13)

Eqns. (10) to (12) were estimated using Ordinary Least Square (OLS) technique.

by more than, exactly and less than *X* value, respectively.

**2.3.2 Returns to scale** 

**2.3.3 Sensitivity analysis** 

geometric mean level.

a fixed resource.

energy on per hectare basis.

follows (Singh et al., 2004; Rafiee et al., 2010):

1 2 ln ln ln *Yi β DE β IDE u***i** (11)

1 2 ln ln ln *Yi γ RE γ NRE u***i** (12)

Apart from the parametric approach, in this study a non-parametric method of DEA was employed to evaluate the technical, pure technical and scale efficiencies of individual farmers which use similar inputs, produce the same product (sunflower) and operate in a relatively homogenous region (e. g., topography, soil type, climatic conditions, etc.). So, energy consumptions from different inputs including human labor, machinery, diesel fuel, chemicals, total fertilizers, water for irrigation, electricity and seed energies in terms of MJ ha-1 were considered as inputs; while, the grain yield of sunflower was the single output; also, each farmer called a DMU.

DEA technique builds a linear piece-wise function from empirical observations of inputs and outputs. DEA is a nonparametric approach for estimating productive efficiency based on mathematical linear programming techniques. Unlike parametric methods, DEA does not require a function to relate inputs and outputs. The DEA envelops the data in such a way that all observed data points lie on or below the efficient frontier (Coelli, 1996). The efficient frontier is established by efficient units from a group of observed units. Efficient units are those with the highest level of productive efficiency.

In DEA an inefficient DMU can be made efficient either by minimizing the input levels while maintaining the same level of outputs (input oriented), or, symmetrically, by maximizing the output levels while holding the inputs constant (output oriented). Sunflower production similar to wheat (Malana and Malano, 2006), paddy (Chauhan et al., 2006) and greenhouse cucumber (Omid et al., 2010) productions relies on finite and scarce resources; therefore the use of input-oriented DEA models is more appropriate to reduce inputs consumed in the production process.

In this study, the technical, pure technical and scale efficiencies of farmers were analyzed.

#### **2.4.1 Technical efficiency (TE)**

TE can be defined as the ability of a DMU (e.g. a farm) to produce maximum output given a set of inputs and technology level. The TE score (*θ*) in the presence of multiple-input and output factor can be calculated by the ratio of sum of weighted outputs to the sum of weighted inputs or in a mathematical expression as follows (Cooper et al., 2004):

$$\theta = \frac{\mu\_1 y\_{1j} + \mu\_2 y\_{2j} + \dots + \mu\_s y\_{sj}}{\upsilon\_1 \mathbf{x}\_{1j} + \upsilon\_2 \mathbf{x}\_{2j} + \dots + \upsilon\_m \mathbf{x}\_{mj}} = \frac{\sum\_{r=1}^s \mu\_r y\_{rj}}{\sum\_{i=1}^m \upsilon\_i \mathbf{x}\_{ij}} \tag{14}$$

Let the DMU*j* to be evaluated on any trial be designated as DMU*o* (*o* = 1, 2, . . ., *n*). To measure the relative efficiency of a DMU*o* based on a series of *n* DMUs, the model is structured as a fractional programming problem as follows (Cooper et al., 2006):

$$\text{Max: } \theta = \frac{\sum\_{r=1}^{s} \mu\_r y\_{ro}}{\sum\_{i=1}^{m} v\_i x\_{io}} \tag{15}$$

Energy Efficiency Analysis in Agricultural Productions:

**2.4.2 Pure technical efficiency (PTE)** 

disproportionate change in outputs.

frequently in the referent sets, is highly ranked.

formula is as follows (Hu and Kao, 2007):

**2.4.3 Scale efficiency (SE)** 

Parametric and Non-Parametric Approaches 145

The TE derived from CCR model, comprehend both the technical and scale efficiencies. So, Banker et al. (1984) developed a model in DEA, which was called BCC model to calculate the PTE of DMUs. The BCC model is provided by adding a restriction on *λ* (*λ* =1) in the model (17), resulted to no condition on the allowable returns to scale. This model assums variable returns to scale (VRS), indicating that a change in inputs is expected to result in a

SE relates to the most efficient scale of operations in the sense of maximizing the average productivity. An scale efficient farmer has the same level of technical and pure technical

> *TE SE PTE*

SE gives the quantitative information of scale characteristics. It is the potential productivity gained from achieving optimum size of a DMU. However, scale inefficiency can be due to the existence of either IRS or DRS. A shortcoming of the SE score is that it does not indicate if a DMU is operating under IRS or DRS conditions. This problem is resolvable by solving a non-increasing returns of scale (NIRS) DEA model, which is obtained by substituting the VRS constraint of *λ* =1 in the BCC model with *λ* ≤ 1 (Scheel, 2000). IRS and DRS can be determined by comparing the efficiency scores obtained by the BCC and NIRS models; so that, if the two efficiency scores are equal, then DRS apply, else IRS prevail (Omid et al., 2010). The information on whether a farmer operates at IRS, CRS or DRS status is particularly helpful in indicating the potential redistribution of resources between the farmers, and thus, enables them to achieve to the higher output (Chauhan et al., 2006).

The results of standard DEA models divide the DMUs into two sets of efficient and inefficient units. The inefficient units can be ranked according to their efficiency scores; while, DEA lacks the capacity to discriminate between efficient units. A number of methods are in use to enhance the discriminating capacity of DEA (Adler et al., 2002). In this study, the bencmarking method was applied to overcome this problem. In this method, an efficient unit which is chosen as the useful target for many inefficient DMUs and so appears

In the analysis of efficient and inefficient DMUs, the energy saving target ratio (ESTR) was used to specify the inefficiency level of energy usage for the DMUs under consideration. The

where energy saving target is the total reducing amount of energy inputs which could be saved without reducing the output level. A higher ESTR percentage implies higher energy use inefficiency, and thus, a higher energy saving amount (Hu and Kao, 2007). In this study, the Microsoft Excel spreadsheet, SPSS 17.0 software and the DEA software Efficiency Measurement Systems (EMS), Version 1.3 (Scheel, 2000) were employed to analyze the data.

*Energy savingt et ESTR*

( arg ) (%) <sup>100</sup> ( )

*Actual energy input* (19)

(18)

efficiency scores. It can be calculated as below (Nassiri and Singh, 2009):

S.t:

$$\begin{aligned} \frac{\sum\_{r=1}^s \boldsymbol{\mu}\_r \boldsymbol{\mathcal{y}}\_{r\bar{j}}}{\sum\_{i=1}^m \boldsymbol{\upsilon}\_i \boldsymbol{\mathcal{x}}\_{i\bar{j}}} & \leq 1 & \quad j = 1, 2, \dots, m \\\\ \boldsymbol{\mu}\_r & \geq 0 & \quad \boldsymbol{\upsilon}\_i \geq 0 \end{aligned}$$

where *n* is the number of DMUs in the comparison, *s* the number of outputs, *m* the number of inputs, *ur* (*r* = 1, 2, …, *s*) the weighting of output *yr* in the comparison, *vi* (*i* = 1, 2, …, *m*) the weighting of input *xi*, and *yrj* and *xij* represent the values of the outputs and inputs *yj* and *xi* for DMU*j*, respectively. Eq. (15) can equivalently be written as a linear programming (LP) problem as follows (Cooper et al., 2006):

$$\text{Maxx: } \theta = \sum\_{r=1}^{s} u\_r y\_{ro} \tag{16}$$

S.t:

$$
\sum\_{r=1}^{s} \mu\_r y\_{rj} - \sum\_{i=1}^{m} \upsilon\_i x\_{ij} \le 0 \quad \text{j} = 1, 2, \dots, \text{n}
$$

$$
\sum\_{i=1}^{m} \upsilon\_i x\_{i0} = 1
$$

$$
\mu\_r \ge 0 \quad \upsilon\_i \ge 0
$$

The dual linear programming (DLP) problem is simpler to solve than Eq. (16) due to fewer constraints. Mathematically, the DLP problem is written in vector–matrix notation as follows (Cooper et al., 2006):

$$\text{Min } \mathfrak{i}\theta \tag{17}$$

S.t:

where *yo* is the 1 *s* vector of the value of original outputs produced and *xo* is the 1 *m* vector of the value of original inputs used by the *o*th DMU. *Y* is the *s n* matrix of outputs and *X* is the *m n* matrix of inputs of all *n* units included in the sample. *λ* is a 1 *n* vector of weights and *θ* is a scalar with boundaries of one and zero which determines the technical efficiency score of each DMU. Model (17) is known as the input-oriented CCR DEA model. It assumes constant returns to scale (CRS), implying that a given increase in inputs would result in a proportionate increase in outputs.

#### **2.4.2 Pure technical efficiency (PTE)**

The TE derived from CCR model, comprehend both the technical and scale efficiencies. So, Banker et al. (1984) developed a model in DEA, which was called BCC model to calculate the PTE of DMUs. The BCC model is provided by adding a restriction on *λ* (*λ* =1) in the model (17), resulted to no condition on the allowable returns to scale. This model assums variable returns to scale (VRS), indicating that a change in inputs is expected to result in a disproportionate change in outputs.

#### **2.4.3 Scale efficiency (SE)**

144 Energy Efficiency – A Bridge to Low Carbon Economy

*<sup>j</sup>*= 1, 2, . . .,*<sup>n</sup>*

<sup>0</sup> *ur* , <sup>0</sup> *<sup>i</sup> <sup>v</sup>*

where *n* is the number of DMUs in the comparison, *s* the number of outputs, *m* the number of inputs, *ur* (*r* = 1, 2, …, *s*) the weighting of output *yr* in the comparison, *vi* (*i* = 1, 2, …, *m*) the weighting of input *xi*, and *yrj* and *xij* represent the values of the outputs and inputs *yj* and *xi* for DMU*j*, respectively. Eq. (15) can equivalently be written as a linear programming (LP)

1

*r u y* 

*r rj i ij r i uy vx* j = 1, 2, . . ., n

<sup>1</sup> <sup>1</sup> *<sup>m</sup> i io <sup>i</sup> v x*

0 *ur* , 0 *<sup>i</sup> v*

The dual linear programming (DLP) problem is simpler to solve than Eq. (16) due to fewer constraints. Mathematically, the DLP problem is written in vector–matrix notation as

 *Min :θ* (17)

*λ* ≥ 0 where *yo* is the 1 *s* vector of the value of original outputs produced and *xo* is the 1 *m* vector of the value of original inputs used by the *o*th DMU. *Y* is the *s n* matrix of outputs and *X* is the *m n* matrix of inputs of all *n* units included in the sample. *λ* is a 1 *n* vector of weights and *θ* is a scalar with boundaries of one and zero which determines the technical efficiency score of each DMU. Model (17) is known as the input-oriented CCR DEA model. It assumes constant returns to scale (CRS), implying that a given increase in inputs would

1 1 <sup>0</sup> *s m*

*r ro*

(16)

*s*

1

1 1

*u y v x* 

*s r rj r m i ij i*

problem as follows (Cooper et al., 2006):

follows (Cooper et al., 2006):

Yλ ≥ yo

 *Xλ- θxo* ≤ 0

result in a proportionate increase in outputs.

Max:

S.t:

S.t:

S.t:

SE relates to the most efficient scale of operations in the sense of maximizing the average productivity. An scale efficient farmer has the same level of technical and pure technical efficiency scores. It can be calculated as below (Nassiri and Singh, 2009):

$$SE = \frac{TE}{PTE} \tag{18}$$

SE gives the quantitative information of scale characteristics. It is the potential productivity gained from achieving optimum size of a DMU. However, scale inefficiency can be due to the existence of either IRS or DRS. A shortcoming of the SE score is that it does not indicate if a DMU is operating under IRS or DRS conditions. This problem is resolvable by solving a non-increasing returns of scale (NIRS) DEA model, which is obtained by substituting the VRS constraint of *λ* =1 in the BCC model with *λ* ≤ 1 (Scheel, 2000). IRS and DRS can be determined by comparing the efficiency scores obtained by the BCC and NIRS models; so that, if the two efficiency scores are equal, then DRS apply, else IRS prevail (Omid et al., 2010). The information on whether a farmer operates at IRS, CRS or DRS status is particularly helpful in indicating the potential redistribution of resources between the farmers, and thus, enables them to achieve to the higher output (Chauhan et al., 2006).

The results of standard DEA models divide the DMUs into two sets of efficient and inefficient units. The inefficient units can be ranked according to their efficiency scores; while, DEA lacks the capacity to discriminate between efficient units. A number of methods are in use to enhance the discriminating capacity of DEA (Adler et al., 2002). In this study, the bencmarking method was applied to overcome this problem. In this method, an efficient unit which is chosen as the useful target for many inefficient DMUs and so appears frequently in the referent sets, is highly ranked.

In the analysis of efficient and inefficient DMUs, the energy saving target ratio (ESTR) was used to specify the inefficiency level of energy usage for the DMUs under consideration. The formula is as follows (Hu and Kao, 2007):

$$ESTR\left(\%\right) = \frac{\left(Energy\,saving\,\,\text{arg}\,\,et\right)}{\left(Actual\,energy\,\,input\right)} \times 100\tag{19}$$

where energy saving target is the total reducing amount of energy inputs which could be saved without reducing the output level. A higher ESTR percentage implies higher energy use inefficiency, and thus, a higher energy saving amount (Hu and Kao, 2007). In this study, the Microsoft Excel spreadsheet, SPSS 17.0 software and the DEA software Efficiency Measurement Systems (EMS), Version 1.3 (Scheel, 2000) were employed to analyze the data.

Energy Efficiency Analysis in Agricultural Productions:

to reduce its environmental footprints.

Fig. 1. The shares of energy inputs for sunflower production.

The energy indices including energy output to input ratio, energy productivity, specific energy and net energy gain are presented in Table 3. Energy ratio in sunflower production was found to be 4.24; showing that output energy of sunflower is obtained 4.24 times greater than total input energy. Also, specific energy was accounted as 5.90 MJ kg-1. Energy output to input ratio and specific energy are integrative indices indicating the potential environmental impacts associated with the production of crops (Khan et al., 2009); also, these parameters can be used to determine the optimum intensity of land and crop management from an environmental point of view. Energy ratio in some agricultural crop productions were reported as 1.5 for sesame, 2.8 for wheat, 3.8 for maize, 4.8 for cotton (Canakci et al., 2005) and 4.68 for canola production (Unakitan et al., 2010). The average energy productivity of sunflower production was 0.17 kg MJ-1. This means that 0.17 unit (kg) output is obtained per unit energy (MJ). Calculation of energy productivity for other

Parametric and Non-Parametric Approaches 147

The percentage distribution of the energy associated with the inputs is seen in Fig. 1. It is evident that, the greatest part of total energy input was consumed by diesel fuel (35.8%); followed by total fertilizer (29%). The distribution of total fertilizers energy input was 68.7% nitrogen, 7.7% phosphate, 1.3% potassium, 0.1% sulfur and 22.2% farmyard manure. Similar studies have also reported that diesel fuel and fertilizer were the most intensive energy inputs (Erdal et al., 2007; Kizilaslan, 2009; Mobtaker et al., 2010); Kallivroussis et al. (2002) reported that the main energy consuming inputs for sunflower production in Greece were nitrogen fertilizer (42.4%) and diesel fuel (33.9%). Excessive use of chemical fertilizers energy input in agriculture may create serious environmental consequences such as nitrogen loading in the environment and receiving waters, poor water quality, carbon emissions and contamination of the food chain (Khan et al., 2009). Integrating a legume into the crop rotation, application of composts, chopped residues or other soil amendments may increases soil fertility in the medium term and so reduces the need for chemical fertilizer energy inputs. Moreover, applying a better machinery management technique, employing the conservation tillage methods or technological upgrade to substitute fossil fuels with renewable energy resources may be the pathways to minimize the fossil fuel usage and thus

#### **3. Results and discussions**

#### **3.1 Energy balance in sunflower production**

Amount of inputs and output in sunflower production are given in Table 2. Based on the evaluation of collected data, average human labor required in the study area was 131.7 h ha-1. Approximately 37% of total human labor was used in harvesting, 30% in weeding and 13% in irrigation operations. Also, machine power mainly was used in harvesting and tillage operations. The use of diesel fuel for operating tractors, combine harvesters and water pumping systems was calculated as 72 L ha-1. Moreover, sunflower production used 1 kg of chemicals, 28 kg of nitrogen and 16.6 kg of phosphate per hectare. As tabulated in the third column of Table 2, the total energy consumption during the production period of sunflower was found to be 9600 MJ ha-1. In some related studies total energy input has been reported as 10491 MJ ha-1 for sunflower production in Greece (Kallivroussis et al., 2002), 18297.61 MJ ha-1 for canola (Unakitan et al., 2010), 14348.9 MJ ha-1 for cotton, 11366.2 MJ ha-1 for maize and 18680.8 MJ ha-1 for wheat production (Canakci et al., 2005). The average yield value of sunflower seed was found to be 1626.5 kg ha-1; accordingly, the total output energy was calculated as 40663 MJ ha-1.


Table 2. Amounts of inputs, output and their energy equivalents for sunflower production.

Amount of inputs and output in sunflower production are given in Table 2. Based on the evaluation of collected data, average human labor required in the study area was 131.7 h ha-1. Approximately 37% of total human labor was used in harvesting, 30% in weeding and 13% in irrigation operations. Also, machine power mainly was used in harvesting and tillage operations. The use of diesel fuel for operating tractors, combine harvesters and water pumping systems was calculated as 72 L ha-1. Moreover, sunflower production used 1 kg of chemicals, 28 kg of nitrogen and 16.6 kg of phosphate per hectare. As tabulated in the third column of Table 2, the total energy consumption during the production period of sunflower was found to be 9600 MJ ha-1. In some related studies total energy input has been reported as 10491 MJ ha-1 for sunflower production in Greece (Kallivroussis et al., 2002), 18297.61 MJ ha-1 for canola (Unakitan et al., 2010), 14348.9 MJ ha-1 for cotton, 11366.2 MJ ha-1 for maize and 18680.8 MJ ha-1 for wheat production (Canakci et al., 2005). The average yield value of sunflower seed was found to be 1626.5 kg ha-1; accordingly, the total output energy

Inputs Quantity per unit area (ha) Total energy equivalent

1. Human labor (h) 131.7 258.1 2. Machinery (kg) 8.1 676.4 a. Tractor 2.6 239.6 b. Self propelled combine 3.7 323.6 c. Other machinery 1.8 113.2 3. Diesel fuel (L) 72.0 3440.4 4. Chemicals (kg) 1.0 205.0 a. Herbicides 0.7 176.6 b. Fungicides 0.1 10.2 c. Insecticides 0.2 18.1 5. Total fertilizer (kg) 2048.8 2696.7 a. Nitrogen 28.0 1852.3 b. Phosphate (P2O5) 16.6 205.9 c. Potassium (K2O) 3.2 36.0 d. Sulfur (S) 2.7 3.1 e. Farmyard manure 1998.3 599.5 6. Water for irrigation ( m3) 644.8 657.7 7. Electricity (kWh) 137.0 1634.2 8. Seed (kg) 8.8 31.6 Total energy input 9600

1. Sunflower seed (kg) 1626.5 40663 Total energy output 40663

Table 2. Amounts of inputs, output and their energy equivalents for sunflower production.

(MJ ha-1)

**3. Results and discussions** 

was calculated as 40663 MJ ha-1.

A. Inputs

B. Output

**3.1 Energy balance in sunflower production** 

The percentage distribution of the energy associated with the inputs is seen in Fig. 1. It is evident that, the greatest part of total energy input was consumed by diesel fuel (35.8%); followed by total fertilizer (29%). The distribution of total fertilizers energy input was 68.7% nitrogen, 7.7% phosphate, 1.3% potassium, 0.1% sulfur and 22.2% farmyard manure. Similar studies have also reported that diesel fuel and fertilizer were the most intensive energy inputs (Erdal et al., 2007; Kizilaslan, 2009; Mobtaker et al., 2010); Kallivroussis et al. (2002) reported that the main energy consuming inputs for sunflower production in Greece were nitrogen fertilizer (42.4%) and diesel fuel (33.9%). Excessive use of chemical fertilizers energy input in agriculture may create serious environmental consequences such as nitrogen loading in the environment and receiving waters, poor water quality, carbon emissions and contamination of the food chain (Khan et al., 2009). Integrating a legume into the crop rotation, application of composts, chopped residues or other soil amendments may increases soil fertility in the medium term and so reduces the need for chemical fertilizer energy inputs. Moreover, applying a better machinery management technique, employing the conservation tillage methods or technological upgrade to substitute fossil fuels with renewable energy resources may be the pathways to minimize the fossil fuel usage and thus to reduce its environmental footprints.

Fig. 1. The shares of energy inputs for sunflower production.

The energy indices including energy output to input ratio, energy productivity, specific energy and net energy gain are presented in Table 3. Energy ratio in sunflower production was found to be 4.24; showing that output energy of sunflower is obtained 4.24 times greater than total input energy. Also, specific energy was accounted as 5.90 MJ kg-1. Energy output to input ratio and specific energy are integrative indices indicating the potential environmental impacts associated with the production of crops (Khan et al., 2009); also, these parameters can be used to determine the optimum intensity of land and crop management from an environmental point of view. Energy ratio in some agricultural crop productions were reported as 1.5 for sesame, 2.8 for wheat, 3.8 for maize, 4.8 for cotton (Canakci et al., 2005) and 4.68 for canola production (Unakitan et al., 2010). The average energy productivity of sunflower production was 0.17 kg MJ-1. This means that 0.17 unit (kg) output is obtained per unit energy (MJ). Calculation of energy productivity for other

Energy Efficiency Analysis in Agricultural Productions:

negative relationship with output.

Durbin-Watson 1.66 *R*2 0.98 Return to scale 1.42

Table 4. Econometric model estimation for sunflower production.

Endogenous variable: yield

a Indicates significance at 1% level. b Indicates significance at 5% level. c Indicates significance at 10% level.

with output of wheat production.

production.

Parametric and Non-Parametric Approaches 149

there was no autocorrelation in the estimated model. The *R2* (coefficient of determination) was as 0.98 for this linear regression model. This implies that all the explanatory variables

The estimated regression coefficients for the model 1 are presented in the second column of Table 4. The results revealed that, human labor machinery, diesel fuel, chemicals and seed energy inputs were the most important inputs, significantly contributed to yield. Also, all of the statistically significant inputs showed the positive relationships with output. Moreover, diesel fuel energy input had the highest elasticity on output (0.95). The second and third important energy inputs were machinery and seed with the elasticity values of 0.25 and 0.14, respectively. With respect to the obtained results, increasing 10% in the consumed energy from diesel fuel, machinery and seed energies, would led to 9.5%, 2.5% and 1.4%, increase in sunflower seed yield, respectively. On the other hand, the impacts of electricity, total fertilizer and water for irrigation energies on yield were estimated statistically insignificant and in the cases of total fertilizer and water for irrigation, the coefficients showed the

Exogenous variables Coefficient *t*-ratio MPP Model 1: *i1 12 23 34 4 5 56 67 7 8 8i lnY α lnX α lnX α lnX α lnX α lnX α lnX α lnX α lnX e* Human labor 0.06c 1.59 0.52 Machinery 0.25b 2.18 0.69 Diesel fuel 0.95a 5.58 0.47 Chemicals 0.02b 1.98 0.12 Total fertilizer -0.01 -0.11 -0.01 Water for irrigation -0.01 -0.71 -0.01 Electricity 0.02 1.42 0.00 Seed 0.14a 3.01 7.06

Mobtaker et al. (2010) developed an econometric model for barley production in Hamedan province of Iran. They reported that human labor, total fertilizer, machinery, diesel fuel, electricity and water for irrigation energies were the important inputs, significantly contributed to yield and machinery energy had highest elasticity. Singh et al. (2004) found that the use of electricity and fertilizers energy inputs in zone 4 of Punjab was inconsistent

The degree of returns to scale for the model (1) was calculated by summation of the regression coefficients as 1.42. The value of return to scale more than unity implies increasing return to scale for sunflower production in the region. These results indicate that 1% increase in all the energy inputs would result by 1.42% increase in the sunflower

included in the regression equation had contributed to the yield by 98%.

oilseed crops has been reported as 0.17 kg MJ-1 for canola (Unakitan et al., 2010) and 0.18 kg MJ-1 for soybean production (De et al., 2001).


Table 3. Some energy indices in sunflower production.

The input energy classification used for sunflower production according to direct, indirect, renewable and non-renewable energy forms are presented in Table 3 and Fig. 2. It is evident that, the ratios of direct and indirect energy forms are nearly the same; but the ratios of renewable and non-renewable energy forms are fairly different from each other (Fig. 2). The ratio of non-renewable energy is very high (83.9%), indicating that sunflower production in the region depends mainly on fossil fuels. Several researchers have founded the ratio of DE higher than that of IDE, and the rate of NRE much greater than that of RE in production of different agricultural crops (Erdal et al., 2007; Mobtaker et al., 2010).

Fig. 2. Distribution of energy forms in sunflower production.

#### **3.2 Econometric model estimation for sunflower production**

The results of econometric model estimation of sunflower production are show in Table 4. For the data used in this study presence of autocorrelation in the residuals from the regression analysis was tested using the Durbin–Watson statistical test (Rafiee et al., 2010). The test result revealed that Durbin–Watson value was as 1.66 for Eq. (10); indicating that

oilseed crops has been reported as 0.17 kg MJ-1 for canola (Unakitan et al., 2010) and 0.18 kg

The input energy classification used for sunflower production according to direct, indirect, renewable and non-renewable energy forms are presented in Table 3 and Fig. 2. It is evident that, the ratios of direct and indirect energy forms are nearly the same; but the ratios of renewable and non-renewable energy forms are fairly different from each other (Fig. 2). The ratio of non-renewable energy is very high (83.9%), indicating that sunflower production in the region depends mainly on fossil fuels. Several researchers have founded the ratio of DE higher than that of IDE, and the rate of NRE much greater than that of RE in production of

Item Unit Quantity Energy ratio - 4.24 Energy productivity kg MJ-1 0.17 Specific energy MJ kg-1 5.90 Net energy MJ ha-1 31062.7 Direct energy MJ ha-1 5990.4 Indirect energy MJ ha-1 3609.6 Renewable energy MJ ha-1 1546.8 Non-renewable energy MJ ha-1 8053.3 Total energy input MJ ha-1 9600

MJ-1 for soybean production (De et al., 2001).

Table 3. Some energy indices in sunflower production.

different agricultural crops (Erdal et al., 2007; Mobtaker et al., 2010).

Fig. 2. Distribution of energy forms in sunflower production.

**3.2 Econometric model estimation for sunflower production** 

The results of econometric model estimation of sunflower production are show in Table 4. For the data used in this study presence of autocorrelation in the residuals from the regression analysis was tested using the Durbin–Watson statistical test (Rafiee et al., 2010). The test result revealed that Durbin–Watson value was as 1.66 for Eq. (10); indicating that there was no autocorrelation in the estimated model. The *R2* (coefficient of determination) was as 0.98 for this linear regression model. This implies that all the explanatory variables included in the regression equation had contributed to the yield by 98%.

The estimated regression coefficients for the model 1 are presented in the second column of Table 4. The results revealed that, human labor machinery, diesel fuel, chemicals and seed energy inputs were the most important inputs, significantly contributed to yield. Also, all of the statistically significant inputs showed the positive relationships with output. Moreover, diesel fuel energy input had the highest elasticity on output (0.95). The second and third important energy inputs were machinery and seed with the elasticity values of 0.25 and 0.14, respectively. With respect to the obtained results, increasing 10% in the consumed energy from diesel fuel, machinery and seed energies, would led to 9.5%, 2.5% and 1.4%, increase in sunflower seed yield, respectively. On the other hand, the impacts of electricity, total fertilizer and water for irrigation energies on yield were estimated statistically insignificant and in the cases of total fertilizer and water for irrigation, the coefficients showed the negative relationship with output.


a Indicates significance at 1% level.

b Indicates significance at 5% level.

c Indicates significance at 10% level.

Table 4. Econometric model estimation for sunflower production.

Mobtaker et al. (2010) developed an econometric model for barley production in Hamedan province of Iran. They reported that human labor, total fertilizer, machinery, diesel fuel, electricity and water for irrigation energies were the important inputs, significantly contributed to yield and machinery energy had highest elasticity. Singh et al. (2004) found that the use of electricity and fertilizers energy inputs in zone 4 of Punjab was inconsistent with output of wheat production.

The degree of returns to scale for the model (1) was calculated by summation of the regression coefficients as 1.42. The value of return to scale more than unity implies increasing return to scale for sunflower production in the region. These results indicate that 1% increase in all the energy inputs would result by 1.42% increase in the sunflower production.

Energy Efficiency Analysis in Agricultural Productions:

Durbin-Watson 1.31 R2 0.98

Durbin-Watson 1.15 R2 0.98

Endogenous variable: yield

Model 2: *i1 2 i lnY β lnDE β lnIDE e*

Model 3: *i1 2 i lnY γ lnRE γ lnNRE e*

energy compared to renewable energy forms.

a Indicates significant at 1% level.

environmental footprints.

scale of their practices.

**3.4 Measuring the efficiency of farmers** 

Parametric and Non-Parametric Approaches 151

Exogenous variables Coefficient *t*-ratio MPP

Direct energy 0.84a 3.01 0.29 Indirect energy 0.03 0.37 0.02

Renewable energy 0.02 0.36 0.04 Non-renewable energy 0.85a 3.52 0.22

direct energy was higher than that of indirect energy; also it was higher for non-renewable

These results may be due to the fact that renewable energy forms such as human labor and farmyard manure were used partially by only some of the farmers and its share was very low; while, non-renewable energy forms especially diesel fuel and machinery were used intensively by majority of the farmers. Additional use of non-renewable energy sources to boost agricultural productions in developing countries with low levels of technological knowledge not only results in environmental deterioration, but also confronts us with the dilemma of a rapid rate of depletion of energetic resources; while, renewable energy sources can be used indefinitely with minimal environmental impacts associated with their production and use (Fadai, 2007). Development of renewable energy usage technologies such as farm machinery or water pumping systems using biodiesel or solar power, employing integrated pest management technique and utilization of alternative sources of energy such as organic fertilizers (compost, manure, etc.) may be the pathways to substitute the non-renewable energy forms with renewable resources and to reduce their

The results of distribution of farmers based on the efficiency score obtained by the application of CCR and BCC DEA models are shown in Fig. 3. As it is evident, about 33% (31 farmers) and 54% (51 farmers) from total farmers were recognized as the efficient farmers under constant and variable returns to scale assumptions, respectively. Moreover, 33% and 32%, with respect, had their technical and pure technical efficiency scores between 0.8 and 1 range. Also, when the BCC model is assumed, only 1% had an efficiency score of less than 0.6; whereas, when the CCR model is applied, 11% had the efficiency scores of less than 0.6. The results of returns to scale estimation indicated that all of the technically efficient farmers (based on the CCR model) were operating at CRS, showing the optimum

Table 5. Econometric model estimation of energy forms in sunflower production.

#### **3.3 Sensitivity analysis of energy inputs on sunflower yield**

The sensitivity of energy inputs was analyzed using the MPP method and partial regression coefficients on output level. The results are presented in the last column of Table 4. As it is seen, the major MPP value was drown by seed (7.06) and it followed by machinery (0.69) and human labor energies (0.52). This implies that an additional use of 1 MJ ha-1 from each of the seed, machinery and human labor energies would lead to an additional increase in yield value of sunflower by 7.06, 0.69 and 0.52 kg ha-1, respectively. In other words, there is a higher potential for increasing output by additional use of these inputs for sunflower production in the surveyed region. On the other hand, the MPP value of total fertilizer and water for irrigation energies were found negative, indicating that use of these inputs is high for sunflower production, resulting in energy dissipation as well as imposing negative effects to environment and human health. The results of sensitivity analysis indicate that which variables should be identified and measured most carefully to assess the state of the environmental system, and which environmental factors should be managed preferentially (Drechsler, 1998). Within this framework, sensitivity analysis of energy inputs is important for improving energy use efficiency and lowering the environmental footprints of energy consumption.

For investigating the relationship between energy forms (i. e. direct, indirect, renewable and non-renewable) and the yield value of sunflower the models (2) and (3) were estimated using Eqs. (11) and (12), respectively. For these models the estimated coefficients, *t*-values, MPP values and validation statistical parameters are presented in Table 5. Durbin–Watson statistical test revealed that Durbin–Watson values were 1.31 and 1.15 for the models (2) and (3), respectively; indicating that there is no autocorrelation in the estimated models. The R2 values were 0.98 for both the estimated models.

The results of model development between direct and indirect energies showed that both the forms of energy had the expected sign and the impact of direct energy was statistically significant, with an elasticity value of 0.84; while indirect energy form had no statistically significant impact on yield. These imply that a 10% increase in direct energy inputs would led to 8.4% increase in yield.

Looking at the Table 5 it also can be seen that, the regression coefficients of non-renewable energy forms was positive and significant at 1% level, while the impact of renewable energies was insignificant; also, the elasticity of non-renewable energy was higher than that of renewable energy (0.85 versus 0.02), implying that a 10% increase in non-renewable energy inputs would led to 8.4% increase in yield, while 10% increase in renewable energy resources increases the output by only 0.2%.

In the literature, similar results have been reported. For example, the impact of direct energy was more than the impact of indirect energy on yield (Hatirli et al., 2005), and the impact of non-renewable energy was higher than that of renewable energy (Mousavi-Avval et al., 2011a).

As can be seen from Table 5, the MPP values of direct and indirect energy forms were 0.29 and 0.02, respectively. Moreover, the sensitivity analysis of renewable and non-renewable energy forms showed that additional use of 1 MJ in non-renewable energies would lead to an additional increase in yield by 0.22 kg; while in the case of renewable energy forms only 0.04 kg is obtained by additional use of 1 MJ. Rafiee et al. (2010) reported that sensitivity of



a Indicates significant at 1% level.

150 Energy Efficiency – A Bridge to Low Carbon Economy

The sensitivity of energy inputs was analyzed using the MPP method and partial regression coefficients on output level. The results are presented in the last column of Table 4. As it is seen, the major MPP value was drown by seed (7.06) and it followed by machinery (0.69) and human labor energies (0.52). This implies that an additional use of 1 MJ ha-1 from each of the seed, machinery and human labor energies would lead to an additional increase in yield value of sunflower by 7.06, 0.69 and 0.52 kg ha-1, respectively. In other words, there is a higher potential for increasing output by additional use of these inputs for sunflower production in the surveyed region. On the other hand, the MPP value of total fertilizer and water for irrigation energies were found negative, indicating that use of these inputs is high for sunflower production, resulting in energy dissipation as well as imposing negative effects to environment and human health. The results of sensitivity analysis indicate that which variables should be identified and measured most carefully to assess the state of the environmental system, and which environmental factors should be managed preferentially (Drechsler, 1998). Within this framework, sensitivity analysis of energy inputs is important for improving energy use efficiency and lowering the environmental footprints of energy

For investigating the relationship between energy forms (i. e. direct, indirect, renewable and non-renewable) and the yield value of sunflower the models (2) and (3) were estimated using Eqs. (11) and (12), respectively. For these models the estimated coefficients, *t*-values, MPP values and validation statistical parameters are presented in Table 5. Durbin–Watson statistical test revealed that Durbin–Watson values were 1.31 and 1.15 for the models (2) and (3), respectively; indicating that there is no autocorrelation in the estimated models. The R2

The results of model development between direct and indirect energies showed that both the forms of energy had the expected sign and the impact of direct energy was statistically significant, with an elasticity value of 0.84; while indirect energy form had no statistically significant impact on yield. These imply that a 10% increase in direct energy inputs would

Looking at the Table 5 it also can be seen that, the regression coefficients of non-renewable energy forms was positive and significant at 1% level, while the impact of renewable energies was insignificant; also, the elasticity of non-renewable energy was higher than that of renewable energy (0.85 versus 0.02), implying that a 10% increase in non-renewable energy inputs would led to 8.4% increase in yield, while 10% increase in renewable energy

In the literature, similar results have been reported. For example, the impact of direct energy was more than the impact of indirect energy on yield (Hatirli et al., 2005), and the impact of non-renewable energy was higher than that of renewable energy (Mousavi-Avval et al.,

As can be seen from Table 5, the MPP values of direct and indirect energy forms were 0.29 and 0.02, respectively. Moreover, the sensitivity analysis of renewable and non-renewable energy forms showed that additional use of 1 MJ in non-renewable energies would lead to an additional increase in yield by 0.22 kg; while in the case of renewable energy forms only 0.04 kg is obtained by additional use of 1 MJ. Rafiee et al. (2010) reported that sensitivity of

**3.3 Sensitivity analysis of energy inputs on sunflower yield** 

consumption.

values were 0.98 for both the estimated models.

resources increases the output by only 0.2%.

led to 8.4% increase in yield.

2011a).

Table 5. Econometric model estimation of energy forms in sunflower production.

direct energy was higher than that of indirect energy; also it was higher for non-renewable energy compared to renewable energy forms.

These results may be due to the fact that renewable energy forms such as human labor and farmyard manure were used partially by only some of the farmers and its share was very low; while, non-renewable energy forms especially diesel fuel and machinery were used intensively by majority of the farmers. Additional use of non-renewable energy sources to boost agricultural productions in developing countries with low levels of technological knowledge not only results in environmental deterioration, but also confronts us with the dilemma of a rapid rate of depletion of energetic resources; while, renewable energy sources can be used indefinitely with minimal environmental impacts associated with their production and use (Fadai, 2007). Development of renewable energy usage technologies such as farm machinery or water pumping systems using biodiesel or solar power, employing integrated pest management technique and utilization of alternative sources of energy such as organic fertilizers (compost, manure, etc.) may be the pathways to substitute the non-renewable energy forms with renewable resources and to reduce their environmental footprints.

#### **3.4 Measuring the efficiency of farmers**

The results of distribution of farmers based on the efficiency score obtained by the application of CCR and BCC DEA models are shown in Fig. 3. As it is evident, about 33% (31 farmers) and 54% (51 farmers) from total farmers were recognized as the efficient farmers under constant and variable returns to scale assumptions, respectively. Moreover, 33% and 32%, with respect, had their technical and pure technical efficiency scores between 0.8 and 1 range. Also, when the BCC model is assumed, only 1% had an efficiency score of less than 0.6; whereas, when the CCR model is applied, 11% had the efficiency scores of less than 0.6. The results of returns to scale estimation indicated that all of the technically efficient farmers (based on the CCR model) were operating at CRS, showing the optimum scale of their practices.

Energy Efficiency Analysis in Agricultural Productions:

Input Target use

by him.

Parametric and Non-Parametric Approaches 153

The results of ESTR calculations showed that if all farmers operated efficiently, the reduction of electricity, water for irrigation and total fertilizer energy inputs, with respect, by 39.6%, 23.8% and 20.5% would have been possible without affecting the yield level. These energy inputs had the highest inefficiency which was owing mainly to the excess use of water and also electricity in water pumping systems. High percentage of fertilizer energy inputs can also be interpreted by the low prices and freely availability of these inputs in surveyed region. Accurate fertilizer management by increasing its profitability with the crops and reducing losses by improving management practices can improve energy use. These results are consistent with the results of energy efficiency analysis in parametric approach, in which, the use of these inputs was inconsistent with output (Section 3.3). On the other hand, the ESTR for human labor and seeds energy inputs was found to be 3.9% and 0.5%, respectively; indicating that these inputs were mainly used efficiently by the farmers in the region. This is consistent with the previouse results, in which, the seed and human labor energies had the relatively high MPP values on output of sunflower (Section

3.3). Similar results also had been reported in the literature by Omid et al. (2010).

Moreover, the results revealed that, the ESTR percentage for total energy input was 16.4%, indicating that by adopting the recommendations resulted from this study, on average, about 16.4% (1571.6 MJ ha-1) from total input energy in sunflower production could be saved without affecting the yield level. Singh et al. (2004) reported that 15.9% (11305 MJ ha-1) from total energy input for wheat production could be saved without affecting the yield level. Also, Mousavi-Avval et al. (2011d) found that about 20% of overall resources in soybean production could be reduced if all of the farmers operate efficiently. Using the information of Table 7, it is possible to advise the inefficient farmers regarding the better operating practices followed by his peers in order to reduce the input energy levels to the target values indicated in the analysis while achieving the output level presently achieved

(MJ ha-1)

1. Human labor 248.1 10.0 3.9 2. Machinery 627.9 48.5 7.2 3. Diesel fuel 3300.5 139.9 4.1 4. Chemicals 187.2 17.8 8.7 5. Total fertilizer 2145.1 551.6 20.5 6. Water for irrigation 500.9 156.7 23.8 7. Electricity 987.3 646.9 39.6 8. Seeds 31.4 0.2 0.5 Total energy 8028.4 1571.6 16.4

Table 7. Optimum energy requirement and saving energy for sunflower production.

Fig. 4 shows the distribution of saving energy from different sources for sunflower production. It is evident that the maximum contribution to the total saving energy is 41.2% from electricity. Also, electricity, total fertilizer, water for irrigation and diesel fuel energy inputs contributed to the total saving energy by about 95%. This is consistent with the results of previous studies that diesel fuel and electricity had the highest potential for improving energy productivity in the production of different agricultural crops (Chauhan et

Saving energy (MJ ha-1)

ESTR (%)

Fig. 3. Distribution of sunflower producers based on efficiency scores

The summarized statistics for the three estimated measures of efficiency are presented in Table 6. The results revealed that the average values of technical and pure technical efficiency scores were 0.83 and 0.93, respectively. Also, the technical efficiency varied from 0.33 to 1 range. The wide variation in the technical efficiency implies that all the farmers were not fully aware of the right production techniques or did not apply them properly. Based on the literature, the technical efficiency scores of 0.77 for paddy production (Chauhan et al., 2006), 0.75 for tomato, 0.81 for asparagus production (Iráizoz et al., 2003) and 0.74 for canola production (Mousavi-Avval et al., 2011c) have been reported.

The average scale efficiency score was relatively low as 0.89, showing the disadvantagiouse conditions of scale size. This indicates that if all of the inefficient farmers operated at the most productive scale size, about 11% savings in energy use from different sources would be possible without affecting the yield level.


Table 6. Average efficiencies of farmers for sunflower production in Golestan, Iran

#### **3.5 Setting realistic input levels for inefficient farmers**

A pure technical efficiency score of less than one for a farmer indicates that, at present conditions, he is using higher values of energy than required. Therefore, it is desired to suggest realistic levels of energy to be used from each source for every inefficient farmer in order to avert wastage of energy. The summarized information for setting realistic input levels are given in Table 7. It gives the average energy usage in target conditions (MJ ha-1), possible energy savings and ESTR percentage for different energy sources. It is evident that, total energy input could be reduced to 8028.4 MJ ha-1; while, maintaining the current production level and also assuming no other constraining factors. Diesel fuel, total fertilizer and electricity energies were required as 3311.5, 2145.1 and 987.3 MJ ha-1, respectively. Moreover, machinery, water for irrigation, human labor, chemicals and seeds energy inputs were required as 627.9, 500.9, 248.1, 187.2 and 31.4 MJ ha-1, respectively.

The summarized statistics for the three estimated measures of efficiency are presented in Table 6. The results revealed that the average values of technical and pure technical efficiency scores were 0.83 and 0.93, respectively. Also, the technical efficiency varied from 0.33 to 1 range. The wide variation in the technical efficiency implies that all the farmers were not fully aware of the right production techniques or did not apply them properly. Based on the literature, the technical efficiency scores of 0.77 for paddy production (Chauhan et al., 2006), 0.75 for tomato, 0.81 for asparagus production (Iráizoz et al., 2003)

The average scale efficiency score was relatively low as 0.89, showing the disadvantagiouse conditions of scale size. This indicates that if all of the inefficient farmers operated at the most productive scale size, about 11% savings in energy use from different sources would

Particular Average SD Min Max Technical efficiency 0.83 0.17 0.33 1 Pure technical efficiency 0.93 0.10 0.54 1 Scale efficiency 0.89 0.14 0.43 1

A pure technical efficiency score of less than one for a farmer indicates that, at present conditions, he is using higher values of energy than required. Therefore, it is desired to suggest realistic levels of energy to be used from each source for every inefficient farmer in order to avert wastage of energy. The summarized information for setting realistic input levels are given in Table 7. It gives the average energy usage in target conditions (MJ ha-1), possible energy savings and ESTR percentage for different energy sources. It is evident that, total energy input could be reduced to 8028.4 MJ ha-1; while, maintaining the current production level and also assuming no other constraining factors. Diesel fuel, total fertilizer and electricity energies were required as 3311.5, 2145.1 and 987.3 MJ ha-1, respectively. Moreover, machinery, water for irrigation, human labor, chemicals and seeds energy inputs

Table 6. Average efficiencies of farmers for sunflower production in Golestan, Iran

were required as 627.9, 500.9, 248.1, 187.2 and 31.4 MJ ha-1, respectively.

and 0.74 for canola production (Mousavi-Avval et al., 2011c) have been reported.

Fig. 3. Distribution of sunflower producers based on efficiency scores

be possible without affecting the yield level.

**3.5 Setting realistic input levels for inefficient farmers** 

The results of ESTR calculations showed that if all farmers operated efficiently, the reduction of electricity, water for irrigation and total fertilizer energy inputs, with respect, by 39.6%, 23.8% and 20.5% would have been possible without affecting the yield level. These energy inputs had the highest inefficiency which was owing mainly to the excess use of water and also electricity in water pumping systems. High percentage of fertilizer energy inputs can also be interpreted by the low prices and freely availability of these inputs in surveyed region. Accurate fertilizer management by increasing its profitability with the crops and reducing losses by improving management practices can improve energy use. These results are consistent with the results of energy efficiency analysis in parametric approach, in which, the use of these inputs was inconsistent with output (Section 3.3). On the other hand, the ESTR for human labor and seeds energy inputs was found to be 3.9% and 0.5%, respectively; indicating that these inputs were mainly used efficiently by the farmers in the region. This is consistent with the previouse results, in which, the seed and human labor energies had the relatively high MPP values on output of sunflower (Section 3.3). Similar results also had been reported in the literature by Omid et al. (2010).

Moreover, the results revealed that, the ESTR percentage for total energy input was 16.4%, indicating that by adopting the recommendations resulted from this study, on average, about 16.4% (1571.6 MJ ha-1) from total input energy in sunflower production could be saved without affecting the yield level. Singh et al. (2004) reported that 15.9% (11305 MJ ha-1) from total energy input for wheat production could be saved without affecting the yield level. Also, Mousavi-Avval et al. (2011d) found that about 20% of overall resources in soybean production could be reduced if all of the farmers operate efficiently. Using the information of Table 7, it is possible to advise the inefficient farmers regarding the better operating practices followed by his peers in order to reduce the input energy levels to the target values indicated in the analysis while achieving the output level presently achieved by him.


Table 7. Optimum energy requirement and saving energy for sunflower production.

Fig. 4 shows the distribution of saving energy from different sources for sunflower production. It is evident that the maximum contribution to the total saving energy is 41.2% from electricity. Also, electricity, total fertilizer, water for irrigation and diesel fuel energy inputs contributed to the total saving energy by about 95%. This is consistent with the results of previous studies that diesel fuel and electricity had the highest potential for improving energy productivity in the production of different agricultural crops (Chauhan et

Energy Efficiency Analysis in Agricultural Productions:

healthy sunflower production systems in the region.

of energy inputs and to obtain sustainable food production systems.

**4. Conclusion** 

Parametric and Non-Parametric Approaches 155

To sum it up, oilseed sunflower is a crop with relatively high requirements on nonrenewable energy resources; its fertilizer and electrical energy requirements are high and it needs a high amount of diesel fuel consumption. The farmers mainly don't have enough knowledge on more efficient input use and there is a common belief that increased use of energy resources will increase the yield. The methodologies presented in this study demonstrate how energy use efficiency in sunflower production may improve by applying the operational management tools to assess the performance of farmers. On an average, considerable savings in energy inputs may be obtained by adopting the best practices of high-performing ones in crop production process. Adoption of more energy-efficient cultivation systems would help in energy conservation and better resource allocation.

Some strategies such as providing better extension and training programs for farmers and use of advanced technologies should be developed in order to increase the energy efficiency of agricultural crop productions in the region. The farmers should be trained with regard to the optimal use of inputs, especially, fertilizers and water for irrigation as well as employing the new production technologies. The local agricultural institutes in the region have an important role in these cases to establish the more energy efficient and environmentally

The study describes the application of parametric and non-parametric approaches to analyze energy efficiency in agricultural production. Therefore, energy use pattern for oilseed sunflower production in Iran was investigated and the parametric method of Cob-Douglas production function and non-parametric method of data envelopment analysis were applied to analyze efficiency of farmers. These methologies helped to identify the impact of energy use from different inputs on output, measure efficiency scores of farmers, segregate efficient farmers from inefficient farmers and to find the wasteful uses of energy by inefficient farmers. The results from both the parametric and non-parametric techniques revealed that the use of machinery, human labor and seed energies had higher impacts on output. In other words, there is higher potential for increasing output by additional use of these inputs for sunflower production. On the other hand, the use of fertilizer, water for irrigation and electrical energy was inconsistent with output, indicating that use of these inputs is high, resulting in energy dissipation as well as imposing negative effects to environment and human health. The results of DEA application further indicated that there are substantial production inefficiencies for farmers; so that, a potential of almost 16% reduction in total energy input use may be achieved if all farmers operated efficiently and assuming no other constraints on this adjustment. Moreover, the results revealed that sunflower production in the region showed a high sensitivity on non-renewable energy sources which may resulted in both the environmental deterioration and rapid rate of depletion of these energetic resources. Therefore, policies should emphasize on development of new technologies to substitute fossil fuels with renewable energy sources aiming efficient use of energy and lowering the environmental footprints. From this study, development of renewable energy usage technologies such as farm machinery or water pumping systems using biodiesel or solar power, applying a better machinery management technique, employing the conservation tillage methods, utilization of alternative sources of energy such as organic fertilizers may be suggested to reduce the environmental footprints

al., 2006; Omid et al., 2010). From these results it is strongly suggested that improving the usage pattern of these inputs be considered as priorities providing significant improvement in energy productivity for sunflower production in surveyed region. Improving energy use efficiency of water pumping systems, employing new irrigation systems and leveling farms properly can be suggested to prevent from electrical energy wastage by inefficient farmers. Applying a better machinery management technique, employing the conservation tillage methods and also, controlling input usage by performance monitoring can help to reduce the diesel fuel and fertilizer energy inputs and minimize their environmental impacts. Also, integrating a legume into the crop rotation, application of composts, chopped residues or other soil amendments may increases soil fertility in the medium term and so reduces the need for chemical fertilizer energy inputs.

Fig. 4. Distribution of saving energy from different sources for sunflower production.

#### **3.6 Improvement of energy indices**

The energy indices for sunflower production in target use of energy are presented in Table 8. It is evident that by optimization of energy use, both the energy ratio and energy productivity indicators can improve by 19.6%. Also, in optimum consumption of energy inputs, the net energy indicator by improvement of 5.1% would increase to 32634.4 MJ ha-1.


Table 8. Improvement of energy indices for sunflower production in Golestan, Iran.

To sum it up, oilseed sunflower is a crop with relatively high requirements on nonrenewable energy resources; its fertilizer and electrical energy requirements are high and it needs a high amount of diesel fuel consumption. The farmers mainly don't have enough knowledge on more efficient input use and there is a common belief that increased use of energy resources will increase the yield. The methodologies presented in this study demonstrate how energy use efficiency in sunflower production may improve by applying the operational management tools to assess the performance of farmers. On an average, considerable savings in energy inputs may be obtained by adopting the best practices of high-performing ones in crop production process. Adoption of more energy-efficient cultivation systems would help in energy conservation and better resource allocation.

Some strategies such as providing better extension and training programs for farmers and use of advanced technologies should be developed in order to increase the energy efficiency of agricultural crop productions in the region. The farmers should be trained with regard to the optimal use of inputs, especially, fertilizers and water for irrigation as well as employing the new production technologies. The local agricultural institutes in the region have an important role in these cases to establish the more energy efficient and environmentally healthy sunflower production systems in the region.

#### **4. Conclusion**

154 Energy Efficiency – A Bridge to Low Carbon Economy

al., 2006; Omid et al., 2010). From these results it is strongly suggested that improving the usage pattern of these inputs be considered as priorities providing significant improvement in energy productivity for sunflower production in surveyed region. Improving energy use efficiency of water pumping systems, employing new irrigation systems and leveling farms properly can be suggested to prevent from electrical energy wastage by inefficient farmers. Applying a better machinery management technique, employing the conservation tillage methods and also, controlling input usage by performance monitoring can help to reduce the diesel fuel and fertilizer energy inputs and minimize their environmental impacts. Also, integrating a legume into the crop rotation, application of composts, chopped residues or other soil amendments may increases soil fertility in the medium term and so reduces the

Fig. 4. Distribution of saving energy from different sources for sunflower production.

The energy indices for sunflower production in target use of energy are presented in Table 8. It is evident that by optimization of energy use, both the energy ratio and energy productivity indicators can improve by 19.6%. Also, in optimum consumption of energy inputs, the net energy indicator by improvement of 5.1% would increase to 32634.4 MJ ha-1.

Items Unit Quantity in optimum use Change (%) Energy ratio - 5.06 19.6 Energy productivity kg MJ-1 0.20 19.6 Specific energy MJ kg-1 4.94 -16.4 Net energy MJ ha-1 32634.4 5.1 Total input energy MJ ha-1 8028.4 -16.4 Table 8. Improvement of energy indices for sunflower production in Golestan, Iran.

need for chemical fertilizer energy inputs.

**3.6 Improvement of energy indices** 

The study describes the application of parametric and non-parametric approaches to analyze energy efficiency in agricultural production. Therefore, energy use pattern for oilseed sunflower production in Iran was investigated and the parametric method of Cob-Douglas production function and non-parametric method of data envelopment analysis were applied to analyze efficiency of farmers. These methologies helped to identify the impact of energy use from different inputs on output, measure efficiency scores of farmers, segregate efficient farmers from inefficient farmers and to find the wasteful uses of energy by inefficient farmers. The results from both the parametric and non-parametric techniques revealed that the use of machinery, human labor and seed energies had higher impacts on output. In other words, there is higher potential for increasing output by additional use of these inputs for sunflower production. On the other hand, the use of fertilizer, water for irrigation and electrical energy was inconsistent with output, indicating that use of these inputs is high, resulting in energy dissipation as well as imposing negative effects to environment and human health. The results of DEA application further indicated that there are substantial production inefficiencies for farmers; so that, a potential of almost 16% reduction in total energy input use may be achieved if all farmers operated efficiently and assuming no other constraints on this adjustment. Moreover, the results revealed that sunflower production in the region showed a high sensitivity on non-renewable energy sources which may resulted in both the environmental deterioration and rapid rate of depletion of these energetic resources. Therefore, policies should emphasize on development of new technologies to substitute fossil fuels with renewable energy sources aiming efficient use of energy and lowering the environmental footprints. From this study, development of renewable energy usage technologies such as farm machinery or water pumping systems using biodiesel or solar power, applying a better machinery management technique, employing the conservation tillage methods, utilization of alternative sources of energy such as organic fertilizers may be suggested to reduce the environmental footprints of energy inputs and to obtain sustainable food production systems.

Energy Efficiency Analysis in Agricultural Productions:

*Policy* 35:373-382

203

55:383-394.

119:135-144

35:1071-1075

*Production* 19:1464-1470

Parametric and Non-Parametric Approaches 157

Fraser, I. & Cordina, D. (1999). An application of data envelopment analysis to irrigated dairy farms in Northern Victoria, Australia. *Agricultural Systems* 59:267-282 Hamby, D.M. (1994). A review of techniques for parameter sensitivity analysis of environmental models. *Environmental Monitoring and Assessment* 32:135-154 Hatirli, S.A.; Ozkan, B. & Fert, C. (2005). An econometric analysis of energy input-output in Turkish agriculture. *Renewable and Sustainable Energy Reviews* 9:608-623 Hu, J.-L. & Kao, C.-H. (2007). Efficient energy-saving targets for APEC economies. *Energy* 

Iráizoz, B.; Rapún, M. & Zabaleta, I. (2003). Assessing the technical efficiency of horticultural

Jianbo, L. (2006). Energy balance and economic benefits of two agroforestry systems in northern and southern China. *Agriculture, Ecosystems & Environment* 116:255-262 Kallivroussis, L.; Natsis, A. & Papadakis, G. (2002). The energy balance of sunflower

Khan, S.; Khan, M.A.; Hanjra, M.A. & Mu, J. (2009). Pathways to reduce the environmental footprints of water and energy inputs in food production. *Food Policy* 34:141-149 Kitani, O. (1999). CIGR handbook of agricultural engineering, Volume 5: Energy and biomass

Kizilaslan, H. (2009). Input-output energy analysis of cherries production in Tokat Province of

Kulekci, M. (2010). Technical efficiency analysis for oilseed sunflower farms: a case study in Erzurum, Turkey. *Journal of the Science of Food and Agriculture* 90:1508-1512 Lamoureux, J., Tiersch, T.R. & Hall, S.G. (2006). Sensitivity analysis of the pond heating and temperature regulation (PHATR) model. *Aquacultural Engineering* 34:117-130 Latif, S. & Anwar, F. (2009). Effect of aqueous enzymatic processes on sunflower oil quality.

Liu, Y.; Høgh-Jensen, H.; Egelyng, H. & Langer, V. (2010). Energy efficiency of organic pear

Malana, N.M. & Malano, H.M. (2006). Benchmarking productive efficiency of selected wheat

Meul, M.; Nevens, F.; Reheul, D. & Hofman, G. (2007). Energy use efficiency of specialised

Mobtaker, H.G.; Keyhani, A.; Mohammadi, A.; Rafiee, S. & Akram, A. (2010). Sensitivity

Mohammadi, A. & Omid, M. (2010). Economical analysis and relation between energy inputs and yield of greenhouse cucumber production in Iran. *Applied Energy* 87:191-196 Mohammadi, A., Rafiee, S., Mohtasebi, S.S. & Rafiee, H. (2010). Energy inputs - yield

Moore, S.R. (2010). Energy efficiency in small-scale biointensive organic onion production in Pennsylvania, USA. *Renewable Agriculture and Food Systems* 25:181-188 Mousavi-Avval, S.H.; Rafiee, S.; Jafari, A. & Mohammadi, A. (2011a). Energy flow modeling

production in greenhouses in China. *Renewable Agriculture and Food Systems* 25:196-

areas in Pakistan and India using data envelopment analysis. *Irrigation and Drainage* 

dairy, arable and pig farms in Flanders. *Agriculture, Ecosystems & Environment*

analysis of energy inputs for barley production in Hamedan Province of Iran.

relationship and cost analysis of kiwifruit production in Iran. *Renewable Energy*

and sensitivity analysis of inputs for canola production in Iran. *Journal of Cleaner* 

production in Navarra, Spain. *Agricultural Systems* 78:387-403

engineering. ASAE Publications, St Joseph, MI

*Journal of the American Oil Chemists' Society* 86:393-400

*Agriculture, Ecosystems & Environment* 137:367-372

Turkey. *Applied Energy* 86:1354-1358

production for biodiesel in Greece. *Biosystems Engineering* 81:347-354

#### **5. References**


Adler, N.; Friedman, L. & Sinuany-Stern, Z. (2002). Review of ranking methods in the data envelopment analysis context. *European Journal of Operational Research* 140:249-265 Alam, M.S.; Alam, M.R. & Islam, K.K. (2005). Energy Flow in Agriculture: Bangladesh.

Anonymous, (2010). Annual agricultural statistics. Ministry of Jihad-e-Agriculture of Iran.

Banaeian, N.; Omid, M. & Ahmadi, H. (2010). Energy and economic analysis of greenhouse

Banker, R.; Charnes, A. & Cooper, W. (1984). Some models for estimating technical and scale inefficiencies in data envelopment analysis. Management Science 30:1078–1092 Barnes, A.P. (2006). Does multi-functionality affect technical efficiency? A non-parametric

Barr, S.I.; McCarron, D.A.; Heaney, R.P.; Dawson-Hughes, B.; Berga, S.L.; Stern, J.S. & Oparil,

Beheshti Tabar, I.; Keyhani, A. & Rafiee, S. (2010). Energy balance in Iran's agronomy (1990-

Canakci, M.; Topakci, M.; Akinci, I. & Ozmerzi, A. (2005). Energy use pattern of some field

Chalabi, Z.S. & Bailey, B.J. (1991). Sensitivity analysis of a non-steady state model of the greenhouse microclimate. *Agricultural and Forest Meteorology* 56:111-127 Chauhan, N.S.; Mohapatra, P.K.J. & Pandey, K.P. (2006). Improving energy productivity in

Coelli, T.J. (1996). A Guide to DEAP Ver. 2.1, A data envelopment analysis

Cooper, L.M.; Seiford, L.M. & Tone, K., (2006). Introduction to data envelopment analysis and

Cooper, W.; Seiford L, M. & Tone, K. (2004). Data Envelopment Analysis, A comprehensive

De, D.; Singh, R.S. & Chandra, H. (2001). Technological impact on energy consumption in rainfed soybean cultivation in Madhya Pradesh. *Applied Energy* 70:193-213 Drechsler, M. (1998). Sensitivity analysis of complex models. *Biological Conservation* 86:401-412 Erdal, G.; Esengün, K.; Erdal, H. & Gündüz, O. (2007). Energy use and economical analysis of

Esengun, K.; Erdal, G.; Gündüz, O. & Erdal, H., (2007). An economic analysis and energy use

Fadai, D. (2007). Utilization of renewable energy sources for power generation in Iran.

sugar beet production in Tokat province of Turkey. *Energy* 32:35-41

*Renewable and Sustainable Energy Reviews* 11:173-181 FAO, (2008). Food and Agricultural Organization. www.fao.org

strawberry production in Tehran province of Iran. *Energy Conversion and Management*

analysis of the Scottish dairy industry. *Journal of Environmental Management* 80:287-294

S. (2000). Effects of Increased Consumption of Fluid Milk on Energy and Nutrient Intake, Body Weight, and Cardiovascular Risk Factors in Healthy Older Adults.

crops and vegetable production: Case study for Antalya Region, Turkey. *Energy* 

paddy production through benchmarking--An application of data envelopment

(computer)program. Centre for efficiency and productivity analysis. University of

text with models, applications, references and DEA-solver software. Massachusetts,

in stake-tomato production in Tokat province of Turkey. *Renewable Energy* 32:1873-

*American Journal of Environmental Sciences* 1:213-220

*Journal of the American Dietetic Association* 100:810-817

*Conversion and Management* 46:655-666

its uses. New York: Springer

1881

USA: Kluwer Academic Publishers

2006). *Renewable and Sustainable Energy Reviews* 14:849-855

analysis. *Energy Conversion and Management* 47:1063-1085

New England <www.une.edu.au/econometrics/cepa.htm>

Cochran, W.G. (1977). Sampling techniques. 3rd Edition, John Wiley & Sons, NY, USA

**5. References** 

www.maj.ir

52:1020-1025


**Part 3** 

**Energy Efficiency in Buildings** 


**Part 3** 

**Energy Efficiency in Buildings** 

158 Energy Efficiency – A Bridge to Low Carbon Economy

Mousavi-Avval, S.H.; Rafiee, S.; Jafari, A. & Mohammadi, A. (2011b). The Functional

Mousavi-Avval, S.H.; Rafiee, S.; Jafari, A. & Mohammadi, A. (2011c). Improving energy use

Mousavi-Avval, S.H.; Rafiee, S.; Jafari, A. & Mohammadi, A. (2011d). Optimization of energy

Mousavi Avval, S.H.; Rafiee, S.; Jafari, A. & Mohammadi, A. (2011). Improving energy

Mukherjee, K. (2008). Energy use efficiency in the Indian manufacturing sector: An interstate

Nassiri, S.M. & Singh, S. (2009). Study on energy use efficiency for paddy crop using data

Omid, M.; Ghojabeige, F.; Delshad, M. & Ahmadi, H. (2010). Energy use pattern and

Pimentel, D. & Patzek, T.W. (2005). Ethanol Production Using Corn, Switchgrass, and Wood;

Rafiee, S.; Mousavi Avval, S.H. & Mohammadi, A. (2010). Modeling and sensitivity analysis of

Reig-Martínez, E. & Picazo-Tadeo, A.J. (2004). Analysing farming systems with Data Envelopment Analysis: citrus farming in Spain. *Agricultural Systems* 82:17-30 Samavatean, N.; Rafiee, S.; Mobli, H. & Mohammadi, A. (2011). An analysis of energy use and

Scheel, H. (2000). EMS: efficiency measurement system users manual, Ver. 1.3, Operations

Sherman, H.D. (1988). Service organization productivity management. The Society of

Singh, G.; Singh, S. & Singh, J. (2004). Optimization of energy inputs for wheat crop in Punjab.

Topp, C.F.E.; Stockdale, E.A.; Watson, C.A. & Rees, R.M. (2007). Estimating resource use

Turhan, S.; Ozbag, B.C. & Rehber, E. (2008). A comparison of energy use in organic and conventional tomato production. *Journal of Food Agriculture & Environment* 6:318-321 Unakitan, G.; Hurma, H. & Yilmaz, F. (2010). An analysis of energy use efficiency of canola

Zhang, X.; Huang, G.H.; Lin, Q. & Yu, H., (2009). Petroleum-contaminated groundwater

Zhou, P.; Ang, B.W. & Poh, K.L. (2008). A survey of data envelopment analysis in energy and environmental studies. *European Journal of Operational Research* 189:1-18

approach. *Journal of the Science of Food and Agriculture* 91:1885-1892

envelopment analysis (DEA) technique. *Applied Energy* 86:1320-1325

energy inputs for apple production in Iran. *Energy* 35:3301-3306

<http://www.wiso.uni-dortmund.de/lsfg/or/scheel/ems>

Management Accountants of Canada, Hamilton, Ontario

*Energy Conversion and Management* 45:453-465

*the Science of Food and Agriculture* 87:2782-2790

production in Turkey. *Energy* 35:3623-3627

*Systems with Applications* 36:5666-5672

*International Journal of Green Energy* 8:398-410

approach. *Applied Energy* 88:3765-3772

analysis. *Energy Policy* 36:662-672

Iran. *Renewable Energy* 36:1808-1813

*Energy Conversion and Management* 52:153-162

*Energy* 36:2765-2772

76

Relationship Between Energy Inputs and Yield Value of Soybean Production in Iran.

efficiency of canola production using data envelopment analysis (DEA) approach.

consumption for soybean production using Data Envelopment Analysis (DEA)

productivity of sunflower production using data envelopment analysis (DEA)

benchmarking of selected greenhouses in Iran using data envelopment analysis.

Biodiesel Production Using Soybean and Sunflower. *Natural Resources Research* 14:65-

relation between energy inputs and yield, costs and income of garlic production in

Research and Wirtschaftsinformatik, University of Dortmund. Dortmund, Germany;

efficiencies in organic agriculture: a review of budgeting approaches used. *Journal of* 

remediation systems design: A data envelopment analysis based approach. *Expert* 

**1. Introduction**

power usage of a city with 200.000 inhabitants.

used to improve model by using user knowledge.

created to monitor energy in data centers.

In the actual world frame, energy efficiency becomes a necessity. Since 1995, Kyoto protocol has highlighted that humans need to improve their energy consumption and reduce CO2 signature. Building consumption represent a third of the global energy consumption and Information Technology (IT) equipment is weighing heavily on energy expenses. For example, Google equipment consumes 0,01 percent of the world global Energy, the equivalent of the

**Energy Consumption Improvement** 

**Through a Visualization Software** 

Benoit Lange, Nancy Rodriguez and William Puech

*LIRMM Laboratory, UMR 5506 CNRS,* 

*University of Montpellier II,* 

**8**

*France* 

Solutions to reduce usage of fossil energy, called "green energies" and supported by standard industries, already exist. For example, faulty processors are removed from production and are re-factored to create solar sensors.The PUE, Power Usage Effectiveness is also a measure

As IT equipment, buildings become greener and smarter. Indeed, buildings are gap of energy and it is necessary to improve them to consume less energy. Governments establish buildings certification to help this development. For example, in France, High Quality Environmental standard has become the actual norm for new or re-factored buildings. Buildings Management Systems (BMS) are widely used to catch data from a set of building sensors. BMS are only reactive systems, there is no prediction. Therefore, it is necessary to

In this chapter, we present a project called RIDER, for Research for IT Driver EneRgy efficiencies. We are developing a green box composed by a set of IT components; able to generate efficient predictive model and building optimization. The main part of the green box is a pivot model managed by a rule engine and a standard BMS with a specific input/output protocol. Modules can extend the standard green box offering system/user

This chapter introduces the RIDER green box and details the visualization module, which is

In Section 2, we present the related work on visualization. Section 3 presents a global overview of the RIDER project. Section 4 is dedicated to the visualization method used in this project,

propose and develop smarter systems, fully focused in energy improvement.

advices, proposing improvements on data model and predicting building behavior.
