**Improving Air-Conditioners' Energy Efficiency Using Green Roof Plants**

Fulin Wang1 and Harunori Yoshida2 *1Tsinghua University, Beijing, 2Okayama University of Science, Okayama, 1China 2Japan* 

#### **1. Introduction**

18 Will-be-set-by-IN-TECH

202 Energy Efficiency – A Bridge to Low Carbon Economy

Elson, J. & Estrin, D. (2004). Sensor networks: a bridge to the physical world, pp. 3–20. European Commission (2006). Communication COM(2006) 545 final, Action Plan for Energy

European Commission (2008). Communication COM(2008) 30 final, 20 20 by 2020: Europe's

European Council (2002). Directive 2002/91/EC of the European Parliament and of the

European Union (1986). Council Resolution of 16 September 1986 concerning new

Ministerio de Medio Ambiente, y Medio Rural y Marino (2010). Inventario de Gases de Efecto

Rodríguez González, A. B., Vinagre Díaz, J. J., Caamaño, A. J. & Wilby, M. R. (2011). Towards a universal energy efficiency index for buildings, *Energy and Buildings* 43: 980–987. Uihlein, A. & Edera, P. (2010). Policy options towards an energy efficient residential building

UNEP (2007). Buildings and Climate Change: Status, Challenges and Opportunities, UNEP

United Nations (1998). Kyoto Protocol to the United Nations Framework Convention on

URL: *http://unstats.un.org/unsd/environment/air\_co2\_emissions.htm*

URL: *http://www.mma.es/secciones/calidad\_contaminacion/atmosfera/emisiones/pdf/*

of Member States, *Official Journal of the European Communities* C 241: 1–3. Gruber, T. R. (1993). A translation approach to portable ontologies, *Knowledge Acquisition*

IEA (2010a). CO2 Emissions from Fuel Combustion, 2010 edition, IEA Publications. IEA (2010b). Energy Balances of non-OECD Countries, 2010 Edition, IEA Publications. IEA (2010c). Energy Balances of OECD Countries, 2010 Edition, IEA Publications.

URL: *http://standards.ieee.org/getieee802/download/802.15.4-2003.pdf*

*Sumario\_de\_Inventario\_Nacional\_Emisiones\_GEI\_serie\_1990-2008.pdf*

Council of 16 December 2002 on the energy performance of buildings, *Official Journal*

Community energy policy objectives for 1995 and the convergence of the policies

Department of Energy (USDOE). URL: *http://www.doe2.com/*

Efficiency: Realising the Potential.

*of the European Communities* , 4 Jan. 2003, pp. 65-71.

climate change opportunity.

URL: *http://www.eia.gov/*

EIA (2011). The eia website.

5(2): 199–220.

IEEE (2006). Standard 802.15.4.

RECS (2011). The recs website.

Publications.

Invernadero de España.

URL: *http://www.eia.doe.gov/emeu/recs/*

stock in the EU-27, *Energy and Buildings* 42: 791–798.

Climate Change, United Nations Publications. United Nations Statistical Division (2009). Greenhouse Gas Emissions. Environment and energy issues are considered to be most urgent things nowadays even in future. A lot of researches have been conducted to study how to prevent global warming and reduce energy consumption. In the field of building and urban environment, green roof attracts a lot of researchers' attention because it is considered to be a good solution for improving urban thermal environment by mitigating heat island and to reduce building cooling energy consumption by reducing cooling load. Alexandria et al. (2008) analyzed how much the urban canyon temperature can be decreased due to green walls and green roofs. Takebayashi et al. (2007) compared the building surface heat transfer of green roofs with common roofs and high reflection roofs. Kumar et al. (2005) developed a mathematical model to evaluate the cooling potential and solar shading effect of green roofs. Wong et al. (2003) analyzed the thermal benefits of green roofs in tropical area. Di et al. (1999) measured an actual green wall to analyze how much cooling effect is achieved. Elena (1998) analyzed the cooling potential of green roofs. Besides studying the green roofs' benefits of heat island mitigation and thermal isolation, the cost vs. benefit is also analyzed (Clerk et al., 2008) and green roof plants selection is analyzed as well (Spala et al., 2008).

However, researchers seldom focus on how to improve air-conditioners' energy efficiency utilizing the cooling effect and solar shading of green roof plants. Therefore this chapter describes a new system combining the green roof plants with air-conditioners outdoor units for the purpose of utilizing the cooling effect and solar shading of green roof plants (Wang et al, 2008, 2009). Figure 1 shows the structure of the combination system. The outdoor units of air conditioner are set under the plants and let air flow through plants and cooled down by plants. Also the plants shade solar radiation to prevent the outdoor unit from absorbing solar energy and raising surface temperature.

#### **2. System design**

Different type of green roof plants, including tree, grass, moss, vine, etc. can be used to construct the system. Different type of plant needs different system structure.

Improving Air-Conditioners' Energy Efficiency Using Green Roof Plants 205

Grass or moss type green roof plants can be lifted up over the air-conditioner outdoor units, as shown in Figure 3. Gaps between plant blocks are needed to let cooled air flow through.

> **Outdoor Unit**

Vine type green roof plants are most suitable for combining with air-conditioner outdoor units because of the higher transpiration rate. Further, vine type plants suits for being cultivated using hydroponic technology. Figure 4 shows a typical hydroponic cultivation system, which consists of fertilizer tanks, fertilizing controller, fertilizer adjustment tank, circulation pump, and cultivation unit. Hydroponic technology can use a small area cultivation unit to grow a larger area of plants. So in the area covered by plants, only a small part needs to support the relatively heavy cultivation system, while most part only supports the plant vine, which is so light that its weight can be neglected from the viewpoint of roof supporting ability. Figure 5 shows an example of one hydroponic-cultivated tomato tree. From the figure, it can be seen that the cultivation unit only occupies a small part of the area

Compared with soil-cultivated green roof, hydroponic-cultivated green roofs are light enough to set on existing buildings, which did not consider the weight of cultivation soil during design phase so it cannot burden the weight. Further, the hydroponic-cultivated green roof plants are light enough to be lifted up over the outdoor unit of air-conditioners,

For the purpose of check the energy saving potential of the combination system, experimental devices are set up on the roof of a five-story office building in Osaka Japan.

**3. Experimental study on the cooling and shading effect of hydroponic-**

Fig. 3. Combination of air-conditioner outdoor units with hung up grass or moss.

The hung up plants can shade the solar radiation as well.

**Grass or Moss** 

**2.2 Grass and moss** 

**2.3 Vine** 

covered by the plant.

**cultivated plant** 

which makes the combination system feasible.

Fig. 1. The system combining the green roof plants with air-conditioner outdoor units.

#### **2.1 Trees**

Outdoor units of air-conditioner can be put under threes (Figure 2). Trees can shade solar radiation to prevent outdoor unit from absorbing solar radiation. Trees can cool down air as well by transpiration and the cooled air flows down and is sucked into air-conditioner outdoor units.

Fig. 2. Combination of air-conditioner outdoor units with trees.

#### **2.2 Grass and moss**

204 Energy Efficiency – A Bridge to Low Carbon Economy

**Outdoor Unit**

Fig. 1. The system combining the green roof plants with air-conditioner outdoor units.

Air-conditioner outdoor unit

Fig. 2. Combination of air-conditioner outdoor units with trees.

Outdoor units of air-conditioner can be put under threes (Figure 2). Trees can shade solar radiation to prevent outdoor unit from absorbing solar radiation. Trees can cool down air as well by transpiration and the cooled air flows down and is sucked into air-conditioner

**Shade solar radiation**

**Indoor Unit**

**2.1 Trees** 

outdoor units.

**Green roof plant**

**Cool down air by transpiration**

> **Indoor Unit**

Grass or moss type green roof plants can be lifted up over the air-conditioner outdoor units, as shown in Figure 3. Gaps between plant blocks are needed to let cooled air flow through. The hung up plants can shade the solar radiation as well.

Fig. 3. Combination of air-conditioner outdoor units with hung up grass or moss.

#### **2.3 Vine**

Vine type green roof plants are most suitable for combining with air-conditioner outdoor units because of the higher transpiration rate. Further, vine type plants suits for being cultivated using hydroponic technology. Figure 4 shows a typical hydroponic cultivation system, which consists of fertilizer tanks, fertilizing controller, fertilizer adjustment tank, circulation pump, and cultivation unit. Hydroponic technology can use a small area cultivation unit to grow a larger area of plants. So in the area covered by plants, only a small part needs to support the relatively heavy cultivation system, while most part only supports the plant vine, which is so light that its weight can be neglected from the viewpoint of roof supporting ability. Figure 5 shows an example of one hydroponic-cultivated tomato tree. From the figure, it can be seen that the cultivation unit only occupies a small part of the area covered by the plant.

Compared with soil-cultivated green roof, hydroponic-cultivated green roofs are light enough to set on existing buildings, which did not consider the weight of cultivation soil during design phase so it cannot burden the weight. Further, the hydroponic-cultivated green roof plants are light enough to be lifted up over the outdoor unit of air-conditioners, which makes the combination system feasible.

#### **3. Experimental study on the cooling and shading effect of hydroponiccultivated plant**

For the purpose of check the energy saving potential of the combination system, experimental devices are set up on the roof of a five-story office building in Osaka Japan.

Improving Air-Conditioners' Energy Efficiency Using Green Roof Plants 207

Fig. 5. A hydroponic-cultivated tomato tree (Small floor area occupied by the cultivation-

Experimental device and measurement points are shown in Figure 6. Measured items and instruments are shown in Table 1. Data are recorded with five minutes interval using a data logger. The hydroponic cultivation device is set on a stage with 1.4 meters height on the roof. The vertical walls of the stage are closed using isolation material to let air can only flow from upside and go through the plants. The upper side of the stage is covered by metal mesh, where the plants spread vines and air can flow through as well. Three ducts equipped with a fan respectively are connected to the south wall of the stage to suck air under the

Sweet potato is selected as the plant because it has high transpiration rate, grows fast, and is strong against wind. The sweet potato was planted on 22nd May 2007. One month later, it began to grow fast with a speed of 0.8 m2 of horizontal projection area per day until the end of September. Figure 7 shows the plant growing situation at the end of September. In October, its growth slowed down and withered at the beginning of November. So generally speaking, the air-cooling and solar shading effect can be utilized for two months of August and September, which are period with large cooling load. This indicates that the

For the purpose of check the cooling potential of the plants, the following experiments are

 Set the flow rate of air flowing through the plants at 2000 m3/h, 4000 m3/h, and 6000 m3/h, to check the relations between temperature decrease and air flow rate. Sprinkle water over and under the plants to check how much the air-cooling effect can be improved, focusing on cooling air temperature down as low as to wet-bulb

stage, acting as the same function of the outdoor unit fans of air-conditioner.

unit compared with the large floor area covered by the plant).

combination system is meaningful for actual application.

**3.1 Experiment setup** 

conducted.

The air temperature cooled by the green roof plants, plant transpiration rate, solar radiation shading rate, and inlet and outlet air temperatures at the outdoor unit of an air-conditioner, etc. were measured.

Fig. 4. An example hydroponic cultivation system.

Fig. 5. A hydroponic-cultivated tomato tree (Small floor area occupied by the cultivationunit compared with the large floor area covered by the plant).

#### **3.1 Experiment setup**

206 Energy Efficiency – A Bridge to Low Carbon Economy

The air temperature cooled by the green roof plants, plant transpiration rate, solar radiation shading rate, and inlet and outlet air temperatures at the outdoor unit of an air-conditioner,

Aeration

etc. were measured.

Fig. 4. An example hydroponic cultivation system.

Control panel

P

Pump

P

Concentration

sensor

Return water

Cultivation unit

Supply water

Circulating

pump

Liquid fertilizer

adjusting tank

P

Concentrated

fertilizer

feed water

Experimental device and measurement points are shown in Figure 6. Measured items and instruments are shown in Table 1. Data are recorded with five minutes interval using a data logger. The hydroponic cultivation device is set on a stage with 1.4 meters height on the roof. The vertical walls of the stage are closed using isolation material to let air can only flow from upside and go through the plants. The upper side of the stage is covered by metal mesh, where the plants spread vines and air can flow through as well. Three ducts equipped with a fan respectively are connected to the south wall of the stage to suck air under the stage, acting as the same function of the outdoor unit fans of air-conditioner.

Sweet potato is selected as the plant because it has high transpiration rate, grows fast, and is strong against wind. The sweet potato was planted on 22nd May 2007. One month later, it began to grow fast with a speed of 0.8 m2 of horizontal projection area per day until the end of September. Figure 7 shows the plant growing situation at the end of September. In October, its growth slowed down and withered at the beginning of November. So generally speaking, the air-cooling and solar shading effect can be utilized for two months of August and September, which are period with large cooling load. This indicates that the combination system is meaningful for actual application.

For the purpose of check the cooling potential of the plants, the following experiments are conducted.


Improving Air-Conditioners' Energy Efficiency Using Green Roof Plants 209

Fig. 7. Experimental system and plants growing situations at the end of September.

The experimental period is from August 15th to September 23rd. Although data are recorded 24 hours a day, the cooling affect can only be observed during the period when photosynthetic radiation is active enough to trigger plant transpiration. The observed transpiration and cooling effect are mainly in the period of 10:00 – 16:00, therefore the data

For the experimental period, the daily temperature decreases of maximum, minimum, average of 10:00 – 12:00, and average of 13:00 – 16:00 are as shown in the upper part of Figure 8. The daily average temperature decrease for clear and no-water-sprinkling days is 1.3 oC. While the average temperature decrease for rainy or water-sprinkle days is 3.0 oC, which is 2.3 times of that in clear and no-water-sprinkling day. However, the temperature

The daily sum and 10:00 – 16:00 sum of transpired water are shown in the lower part of the Figure 8. The maximum and average daily summed transpirations are 8.3 and 6.3 kg/m2 horizontal area respectively for clear and no-water-sprinkle days. For the days when water was sprinkled upon the plants, the transpiration is relatively small because leaves were wet

decrease does not differ much for the air flow rate of 2000, 4000, and 6000 m3/h.

**3.2 Experimental results** 

1) Air temperature decrease

2) Plants transpiration

during this time are used for analysis.

and the transpiration temporally stopped.


temperature. Considering it might damage the plants growth, the water sprinkling experiments are conducted twice a day on 11:00 – 12:00 and 14:00 – 15:00.

Table 1. Measured items and instruments

Fig. 6. Experimental device and measurement items.

Fig. 7. Experimental system and plants growing situations at the end of September.

#### **3.2 Experimental results**

208 Energy Efficiency – A Bridge to Low Carbon Economy

experiments are conducted twice a day on 11:00 – 12:00 and 14:00 – 15:00.

Mark Item Point Instrument WD Wind direction A vane-type WS Wind speed A 3-cup anemo meter RU Solar radiation B,D pyranometer RD Solar radiation (under green) 2,3 pyranometer PAR Photosynthetically active radiation D photon sensor NRF Net radiation (roof) C net radiometer NRG Net radiation (green) 4 net radiometer IR Infrared radiation B infrared radiometer WF Water flow F flowmeter TR Outside air temperature A thermohygrometer HR Outside air relative humidity A thermohygrometer

HU Ralative humidity (over green) 1~8 thermo recorder

TD Temperature (under green) 1~8

Infrared radiation

Table 1. Measured items and instruments

Relative humidity Solar radiation

A

A

Temperature

A

Fan

Leaf temperature

Cultivation unit

Wind direction Wind speed Temperature

> Enclosed space

Fig. 6. Experimental device and measurement items.

4

7 8

TWS1 TWS2

Solar radiation Photosynthetically active radiation

D

2 3

TWW TWE

TF

Solar radiation Solar radiation

1 5 6

TWN

B

990 3900 8100

TU Temperature (over green) 1~8

HD Ralative humidity (under green) 1~8 thermo recorder TDU Temperature (in duct) ▲ thermocouple TL Leaf temperature 1~4 thermocouple TW Wall temperature × thermocouple TF Floor temperature thermocouple TH Temperature of suction opening ○ thermocouple THO Temperature of supply opening ● thermocouple

C

流量計精度検証用回路

Net radiation

Tank Fertilizer

Duct

Temperature Relative Humidity

Temperature Relative Humidity Net radiation

1400

9300

M

散水用回路 F 流量計 F

ロガー ストレーナー 水道 バルブ メーター

水道水

Temperature of outlet air

Airconditioner

Water flow

N

Temperature of inlet air

5500

3000

N

temperature. Considering it might damage the plants growth, the water sprinkling

thermocouple thermo recorder

thermocouple thermo recorder

> The experimental period is from August 15th to September 23rd. Although data are recorded 24 hours a day, the cooling affect can only be observed during the period when photosynthetic radiation is active enough to trigger plant transpiration. The observed transpiration and cooling effect are mainly in the period of 10:00 – 16:00, therefore the data during this time are used for analysis.

#### 1) Air temperature decrease

For the experimental period, the daily temperature decreases of maximum, minimum, average of 10:00 – 12:00, and average of 13:00 – 16:00 are as shown in the upper part of Figure 8. The daily average temperature decrease for clear and no-water-sprinkling days is 1.3 oC. While the average temperature decrease for rainy or water-sprinkle days is 3.0 oC, which is 2.3 times of that in clear and no-water-sprinkling day. However, the temperature decrease does not differ much for the air flow rate of 2000, 4000, and 6000 m3/h.

#### 2) Plants transpiration

The daily sum and 10:00 – 16:00 sum of transpired water are shown in the lower part of the Figure 8. The maximum and average daily summed transpirations are 8.3 and 6.3 kg/m2 horizontal area respectively for clear and no-water-sprinkle days. For the days when water was sprinkled upon the plants, the transpiration is relatively small because leaves were wet and the transpiration temporally stopped.

Improving Air-Conditioners' Energy Efficiency Using Green Roof Plants 211

0

Hydoroponic

Fig. 9. Transpiration comparison.

Fig. 10. Solar radiation up and under plants.

plants and the solar radiation after being shaded.

Solar radiation (W/m2) .

(average)

Hydoroponic

(maximum)

Upper plants Under plants

Sedum

(water-sprinkle)

Sedum

(no-water sprinkle)

0:00 3:00 6:00 9:00 12:00 15:00 18:00 21:00

Based on experimental data, energy saving potential of the combination systems can be estimated using the measured air temperatures before and after being cooled down by the

The calculation flow for estimating the energy saving is shown in Figure 11. Firstly calculate the air cooling effect *T* caused by plants transpiration and solar shading. Then calculate air-conditioners' energy consumption using an air-conditioner energy consumption model.

**4. Estimation of energy saving effect based on experimental data** 

Lawn

Grassland

Forest

Tree

2

4

6

Transpiration rate (kg/m2.day)

8

10

The comparison of different plant transpiration rate is shown in Figure 9. The maximum daily transpiration rate (kilogram water per square meter of horizontal projection area) of hydroponoic-cultivate sweet potato is 13.8 times of sedum without water-sprinkle and 1.8 times of sedum with water-sprinkle. The maximum transpiration of hydroponoic-cultivate sweet potato is similar to a single tree, which indicates the hydroponoic-cultivate sweet potato can transpire as much water as a tree, while a tree has 20 times more leaf volume per unit horizontal projection area than sweet potato.

#### 3) Solar radiation shading

The solar radiations over and under the plants on a typical day are shown in Figure 10. Even the solar radiation is as high as 1000 w/m2, the measured solar radiation under the plants is no more than 10 W/m2. That is to say more than 99% solar radiation is shaded by the plants. Therefore hydroponic-cultivated green roof plant can shade solar radiation enough to ignore the influence of solar radiation to the outdoor unit of an air-conditioner.

Fig. 8. Experimental results of air temperature decrease (upper graph) and transpiration rate (lower graph).

Fig. 9. Transpiration comparison.

The comparison of different plant transpiration rate is shown in Figure 9. The maximum daily transpiration rate (kilogram water per square meter of horizontal projection area) of hydroponoic-cultivate sweet potato is 13.8 times of sedum without water-sprinkle and 1.8 times of sedum with water-sprinkle. The maximum transpiration of hydroponoic-cultivate sweet potato is similar to a single tree, which indicates the hydroponoic-cultivate sweet potato can transpire as much water as a tree, while a tree has 20 times more leaf volume per

The solar radiations over and under the plants on a typical day are shown in Figure 10. Even the solar radiation is as high as 1000 w/m2, the measured solar radiation under the plants is no more than 10 W/m2. That is to say more than 99% solar radiation is shaded by the plants. Therefore hydroponic-cultivated green roof plant can shade solar radiation enough to

ignore the influence of solar radiation to the outdoor unit of an air-conditioner.

2000m3/h 2000m3/h 2000m3/h 6000m3/h

2000m3/h

4000m3/h 4000m3 Various air /h

unit horizontal projection area than sweet potato.

Sprinkle water over roof planting

3) Solar radiation shading

4000m3/h 4000m3/h

volume

ave(10-12) ave(13-16)

8/15

(lower graph).

8/16

8/17

8/18

8/19

8/20

8/21

8/22

8/23

8/24

8/25

8/26

8/27

8/28

8/29

8/30

8/31

9/1

9/2

9/3

Sprinkle water under roof planting

Fig. 8. Experimental results of air temperature decrease (upper graph) and transpiration rate

9/4

9/5

9/6

9/7

9/8

9/9

9/10

9/11

9/12

9/13

Sprinkle water under roof planting

9/14

9/15

9/16

9/17

9/18

9/19

9/20

9/21

Sprinkle water over roof planting

9/22

9/23

Air temperature differences

Integrated transpitation

rate (kg/m2

)

cooled down (℃)

8/15

8/16

8/17

8/18

8/19

8/20

8/21

8/22

8/23

8/24

8/25

8/26

8/27

8/28

8/29

8/30

8/31

9/1

9/2

9/3

9/4

9/5

9/6

9/7

9/8

9/9

9/10

9/11

9/12

9/13

9/14

9/15

9/16

9/17

9/18

9/19

9/20

sum sum(10-16)

9/21

9/22

9/22

9/23

Fig. 10. Solar radiation up and under plants.

#### **4. Estimation of energy saving effect based on experimental data**

Based on experimental data, energy saving potential of the combination systems can be estimated using the measured air temperatures before and after being cooled down by the plants and the solar radiation after being shaded.

The calculation flow for estimating the energy saving is shown in Figure 11. Firstly calculate the air cooling effect *T* caused by plants transpiration and solar shading. Then calculate air-conditioners' energy consumption using an air-conditioner energy consumption model.

Improving Air-Conditioners' Energy Efficiency Using Green Roof Plants 213

energy saving is different. The daily average energy savings are shown in Figure 12. The maximum energy saving rate is 12% for the air-conditioners with high efficiency improvement ratio, and 3% for the air-conditioners whose efficiencies improve little

The summed energy savings for the experimental period are shown in Table 2. If watersprinkle is not conducted, the energy saving ratio is 4% for the air-conditioners with high efficiency improvement ratio, and 1% for the air-conditioners with low efficiency improvement ratio. If water-sprinkle is conducted for two hours a day, the energy saving ratios are 9% and 2% for the air-conditioners with high and low efficiency improvement ratio respectively. Among the total energy saving, about 10% is from solar shading and 90%

accompanying to outdoor air temperature decreasing.

is from transpiration.

0% 2% 4% 6% 8% 10% 12% 14%

Energy saving rate

8/15

sweet potato.

8/17

8/19

8/21

when outdoor air temperature decreases.

8/23

8/25

8/27

8/29

8/31

9/2

Fig. 12. Estimated daily average energy saving rate achieved by using hydroponic-cultivated

The reasons for why the energy saving ratios at the air-conditioners of the Manufacturer A are roughly 4 times less than these of the others are not further studied because the detailed information is not available about how the air-conditioner operation is tuned corresponding to the different outdoor temperature. The possible reasons might be that different manufactures use different actions or components to tune the running of airconditioner accompanying to the change of outdoor air temperature. For example if outdoor air temperature decreases, for the air-conditioner with variable speed drive compressor can decrease the compressor's rotational speed to meet the requirement of low compression rate and maintain high efficiency as well. Low compressor rotational speed directly relates to low energy consumption because the energy consumption of airconditioners is roughly linear to compressor's rotational speed. While for the airconditioner with constant speed drive compressor can only decrease compression rate through increasing refrigerant flow rate and bypassing the redundant refrigerant flow. The larger flow rate needs larger energy input, which will counteract the energy saving benefitted from the decrease of compression rate. Further the flow rate larger than rated value will cause compressor efficiency decrease. So the energy saving rate might be small

9/4

9 6/ 9/8

9/10

9/12

9/14

9/16

9/18

Manufacturer A Manufacturer B Manufacturer C Manufacturer D

9/20

9/22

Fig. 11. Calculation flow for energy consumption of air-conditioners.

#### **4.1 Air cooling effect**

The air temperature decreasing is caused by two reasons. The first is the plants transpiration. The measured temperature differences between the air over and under the plants are used for the calculation. The second is the equivalent air temperature decrease caused by solar shading. If there is no plant shading, the outdoor unit will absorb solar radiation and its surface temperature will rise. This part of heat will raise the temperature of the air sucked into the outdoor unit. The air-conditioners' energy efficiency will be decreased by the temperature increasing of sucked in air. The air temperature increase is calculated using the following equations.

$$
\Delta\theta\_s = \frac{q\_L + \alpha\_S q\_S}{m \mathcal{C}\_{Pa}} \tag{1}
$$

#### **4.2 Air-conditioner energy consumption model**

A regression model fitted using air-conditioner manufacturer's specification data is used to calculate the air-conditioner's energy consumption (Wang et al., 2005).

$$RE = \left(a\_1\theta\_{\eta^f}\,^2 + b\_1\theta\_{\eta^f} + c\_1\right)\left(a\_2\theta\_i\,^2 + b\_2\theta\_i + c\_2\right)\left(a\_3CA^2 + b\_3CA + c\_3\right) + d\tag{2}$$

$$
\theta\_{o\circ} = \theta\_o - \Delta\theta\_{\Gamma} + \Delta\theta\_s \tag{3}
$$

The energy saving is calculated for four types of typical air-conditioner made by four different manufactures, including Gas-engine Heat Pump (GHP) and Electricity-driven Heat Pump (EHP). The average nominal primary energy Coefficient of Performance (COP) is 1.4 and manufacture year is 2005. Different manufacturer's air-conditioners have different efficiency improvement ratio accompanying to outdoor temperature decreasing. So the

Green roof cooling effect *T*

Air temperature sucked into outdoor unit *Tos*

The air temperature decreasing is caused by two reasons. The first is the plants transpiration. The measured temperature differences between the air over and under the plants are used for the calculation. The second is the equivalent air temperature decrease caused by solar shading. If there is no plant shading, the outdoor unit will absorb solar radiation and its surface temperature will rise. This part of heat will raise the temperature of the air sucked into the outdoor unit. The air-conditioners' energy efficiency will be decreased by the temperature increasing of sucked in air. The air temperature increase is

Energy consumption *E*

Air-conditioner energy estimation model

*L SS <sup>s</sup> Pa q q mC* 

A regression model fitted using air-conditioner manufacturer's specification data is used to

22 2 1 1 12 2 23 3 3 ( )( )( ) *RE a b c a b c a CA b CA c d*

The energy saving is calculated for four types of typical air-conditioner made by four different manufactures, including Gas-engine Heat Pump (GHP) and Electricity-driven Heat Pump (EHP). The average nominal primary energy Coefficient of Performance (COP) is 1.4 and manufacture year is 2005. Different manufacturer's air-conditioners have different efficiency improvement ratio accompanying to outdoor temperature decreasing. So the

 

(1)

Cooling load *Rc*

*oT s* (3)

*of of i i* (2)

 

calculate the air-conditioner's energy consumption (Wang et al., 2005).

*of*

 

Fig. 11. Calculation flow for energy consumption of air-conditioners.

Outdoor Temperature *To*

**4.1 Air cooling effect** 

Indoor Temperature *Ti*

calculated using the following equations.

**4.2 Air-conditioner energy consumption model** 

energy saving is different. The daily average energy savings are shown in Figure 12. The maximum energy saving rate is 12% for the air-conditioners with high efficiency improvement ratio, and 3% for the air-conditioners whose efficiencies improve little accompanying to outdoor air temperature decreasing.

The summed energy savings for the experimental period are shown in Table 2. If watersprinkle is not conducted, the energy saving ratio is 4% for the air-conditioners with high efficiency improvement ratio, and 1% for the air-conditioners with low efficiency improvement ratio. If water-sprinkle is conducted for two hours a day, the energy saving ratios are 9% and 2% for the air-conditioners with high and low efficiency improvement ratio respectively. Among the total energy saving, about 10% is from solar shading and 90% is from transpiration.

Fig. 12. Estimated daily average energy saving rate achieved by using hydroponic-cultivated sweet potato.

The reasons for why the energy saving ratios at the air-conditioners of the Manufacturer A are roughly 4 times less than these of the others are not further studied because the detailed information is not available about how the air-conditioner operation is tuned corresponding to the different outdoor temperature. The possible reasons might be that different manufactures use different actions or components to tune the running of airconditioner accompanying to the change of outdoor air temperature. For example if outdoor air temperature decreases, for the air-conditioner with variable speed drive compressor can decrease the compressor's rotational speed to meet the requirement of low compression rate and maintain high efficiency as well. Low compressor rotational speed directly relates to low energy consumption because the energy consumption of airconditioners is roughly linear to compressor's rotational speed. While for the airconditioner with constant speed drive compressor can only decrease compression rate through increasing refrigerant flow rate and bypassing the redundant refrigerant flow. The larger flow rate needs larger energy input, which will counteract the energy saving benefitted from the decrease of compression rate. Further the flow rate larger than rated value will cause compressor efficiency decrease. So the energy saving rate might be small when outdoor air temperature decreases.

Improving Air-Conditioners' Energy Efficiency Using Green Roof Plants 215

1 *s*

*r*

( )

*s*

*<sup>g</sup>* , <sup>1</sup> *<sup>a</sup>*

*r*

max max

*T T l h*

( ) / *s s g Q f Q <sup>Q</sup> <sup>g</sup> <sup>a</sup>* 

*ol ho*

<sup>1</sup> ( ) 1(/ )*<sup>n</sup> f D D n*

The transpiration model is validated using the data measured in the former mentioned

The measured solar radiation, wind speed, ambient air temperature, plant cooled air temperature, leaf surface temperature, stomatal conductance, transpiration rate, etc. are

*TT T T f T TT TT*

*a*

*<sup>g</sup>* (5)

(9)

(8)

(10)

*ga t* 0.147 / *C ul* (6)

*Q CJ <sup>Q</sup>* (11)

max () () () *s s g g fQ fT fD* (7)

*h o o l T T*

2 1

Fig. 13. Plant transpiration mechanisms.

experiment.


Table 2. Summed annual energy saving achieved by hydroponic-cultivated sweet potato.

#### **5. Transpiration and cooling effect modeling**

Because the energy saving potential estimation results based on experimental result might depend on the local climate conditions, for the purpose of developing a universal method to estimate the energy saving potential, modeling of the transpiration and cooling effect are needed. Therefore it is necessary to develop plant transpiration model, which is used to estimate the plant transpiration rate, and boundary layer model, which is used to calculate the air temperature decreased by plant transpiration.

#### **5.1 Transpiration model**

Plants control their transpiration rate by adjusting the opening of stomata, as shown in Figure 13. So the transpiration rate can be calculated through dividing the water vapour partial pressure deficit ( ) *W W i a* by the stomatal resistance *sr* and boundary layer air resistance *ar* , as shown in Equation 4. The resistances can be calculated from their reciprocals, stomatal conductance *<sup>s</sup> g* and boundary layer air conductance *<sup>a</sup> g* , respectively (Equation 5). Boundary layer air conductance *<sup>a</sup> g* is usually used a constant value of 1.13 mol/(m2s) (Kadaira et al., 2005). Here Equation 6 (Campbell and Norman, 1998) is introduced to improve the model accuracy by considering the outdoor wind speed. Regarding the stomatal conductance *<sup>s</sup> g* Jarvis (1976) proposed a model to estimate it using the leave surface temperature *T* , photosynthetically active radiation (PAR) *Q* and water vapour saturation pressure deficit *D* (Equation 7). Based on Jarvis model, Kosugi et al. (1995) improved the models of *f Q* , *f T* , and *f D* and found that the models shown in Equation 8, 9 and 10 can suit for more plant so these models are used in this research. Further Kadaira et al. (2005) proved through experiment that in Equation 9 ambient air temperature can substitute for the leaf surface temperature *T* with acceptable accuracy. Kadaira et al. (2005) also found that the PAR can be accurately estimated by multiplying global sky radiation *J* by a correlation coefficient of *CQ* with the value of 2.1 (Equation 11). To use these models to estimate the transpiration rate, seven coefficients need to be fitted using the data of the plant that the system uses, i.e. *<sup>s</sup>*max *g* and *a* in Equation 5, *To* , *Th* , and *Tl* in Equation 6, 1 *b* and 2 *b* in Equation 10.

$$M = \frac{M\_w(\mathcal{W}\_i - \mathcal{W}\_a)}{r\_s + r\_a} \tag{4}$$

saving rate With or without green roof

Table 2. Summed annual energy saving achieved by hydroponic-cultivated sweet potato.

Because the energy saving potential estimation results based on experimental result might depend on the local climate conditions, for the purpose of developing a universal method to estimate the energy saving potential, modeling of the transpiration and cooling effect are needed. Therefore it is necessary to develop plant transpiration model, which is used to estimate the plant transpiration rate, and boundary layer model, which is used to calculate

Plants control their transpiration rate by adjusting the opening of stomata, as shown in Figure 13. So the transpiration rate can be calculated through dividing the water vapour partial pressure deficit ( ) *W W i a* by the stomatal resistance *sr* and boundary layer air resistance *ar* , as shown in Equation 4. The resistances can be calculated from their reciprocals, stomatal conductance *<sup>s</sup> g* and boundary layer air conductance *<sup>a</sup> g* , respectively (Equation 5). Boundary layer air conductance *<sup>a</sup> g* is usually used a constant value of 1.13 mol/(m2s) (Kadaira et al., 2005). Here Equation 6 (Campbell and Norman, 1998) is introduced to improve the model accuracy by considering the outdoor wind speed. Regarding the stomatal conductance *<sup>s</sup> g* Jarvis (1976) proposed a model to estimate it using the leave surface temperature *T* , photosynthetically active radiation (PAR) *Q* and water vapour saturation pressure deficit *D* (Equation 7). Based on Jarvis model, Kosugi et al. (1995) improved the models of *f Q* , *f T* , and *f D* and found that the models shown in Equation 8, 9 and 10 can suit for more plant so these models are used in this research. Further Kadaira et al. (2005) proved through experiment that in Equation 9 ambient air

can substitute for the leaf surface temperature *T* with acceptable accuracy.

Kadaira et al. (2005) also found that the PAR can be accurately estimated by multiplying global sky radiation *J* by a correlation coefficient of *CQ* with the value of 2.1 (Equation 11). To use these models to estimate the transpiration rate, seven coefficients need to be fitted using the data of the plant that the system uses, i.e. *<sup>s</sup>*max *g* and *a* in Equation 5, *To* , *Th* , and

> ( ) *wi a s a*

*MW W <sup>M</sup> r r* 

Energy consumption Energy

Energy saving rate

**5. Transpiration and cooling effect modeling** 

the air temperature decreased by plant transpiration.

Energy consumption

sum of Clear and no-water-sprinkle days sum ofwater-sprinkle days sum of rainy days

**5.1 Transpiration model** 

temperature

*Tl* in Equation 6, 1 *b* and 2 *b* in Equation 10.

With Without With Without With Without With Without 2830.5 2859.7 1.0% 2376.4 2449.9 3.0% 2673.1 2767.7 3.4% 2227.6 2317.9 3.9% 1024.9 1046.7 2.1% 905.4 970.9 6.7% 965.5 1050.3 8.1% 875.2 963.8 9.2% 1011.5 1019.6 0.8% 766.0 810.7 5.5% 857.5 917.6 6.5% 703.4 758.4 7.2%

Manufacturer A (GHP) Manufacturer B (EHP) Manufacturer C (GHP) Manufacturer D (EHP)

saving rate

Group 1 Group 2 Group 3

Energy saving rate

Energy consumption

Energy consumption Energy

(4)

Fig. 13. Plant transpiration mechanisms.

$$r\_s = \frac{1}{\mathcal{g}\_s}, \ r\_a = \frac{1}{\mathcal{g}\_a} \tag{5}$$

$$\mathbf{g}\_a = 0.147 \mathbf{C}\_t \sqrt{\mu/l} \tag{6}$$

$$\mathbf{g}\_s = \mathbf{g}\_{s\text{max}} \cdot f(\mathbf{Q}) \cdot f(T) \cdot f(D) \tag{7}$$

$$f(Q) = \frac{\mathcal{g}\_{s\max} \cdot Q}{Q + \mathcal{g}\_{s\max} / a} \tag{8}$$

$$f(T) = \left(\frac{T - T\_l}{T\_o - T\_l}\right) \left[ \left(\frac{T\_h - T}{T\_h - T\_o}\right)^{\left(\frac{T\_h - T\_o}{T\_o - T\_l}\right)} \right] \tag{9}$$

$$f(D) = \frac{1}{1 + \left(D \ne n\_1\right)^{n\_2}}\tag{10}$$

$$Q = \mathbb{C}\_Q \text{J} \tag{11}$$

The transpiration model is validated using the data measured in the former mentioned experiment.

The measured solar radiation, wind speed, ambient air temperature, plant cooled air temperature, leaf surface temperature, stomatal conductance, transpiration rate, etc. are

Improving Air-Conditioners' Energy Efficiency Using Green Roof Plants 217

When air flows along the plant leaf surface, an air boundary layer forms. If we take a micro air volume with the x-length of *x* , y-length of one meter, and z-length of boundary thickness as the analysis object, the mass conservation is shown in Equation 12 and energy conservation is shown in Equation 13. If the thermal stored in leaf is ignored, the leaf energy conservation is shown in Equation 14. If substituting Equation 14 into Equation 13, the energy conservation will be as shown in Equation 15. Because it is difficult to solve Equation 12 and 15, and what we want to know is not the temperature distribution but the average temperature, we can take the total boundary lay as the analysis object. Thus the mass conservation becomes as shown in Equation 16 and energy conservation becomes as shown in Equation 17. If substituting Equation 14 and Equation 16 into Equation 17, the energy conservation becomes as shown in Equation 18. Further, because the volume of boundary layer *Vb* is so small (it is 88.7 m3 for the green roof with a length of 10 m when calculating using boundary layer thickness Equation 20 (Kato, 1964) and boundary layer volume Equation 21) comparing to the air volume entering the boundary layer (usually several thousands m3/h) that it is safe to ignore the energy change of boundary layer air, i.e. *H* 0 . Therefore energy conservation becomes as shown in Equation 19. The physical meaning of Equation 19 is that the solar radiation energy absorbed by plant leaves becomes the energy difference between the air flowing into and flowing out of the boundary layer if

the leaves temperature change and boundary layer air energy change are ignored.

00 0 , , , [ ( ) ] *uh uh u b bx x b bx x x x b dx x <sup>w</sup>*

00 0 , , , [ ( ) ] *u h uh u b bx x b bx x x x b dx x*

 

 

> 

*h u h u ul* <sup>000</sup>

*x* 

0

*b*

*l*

 

 

 

 

 

  <sup>0</sup> () 0 *u u u ux x xx dx*

<sup>0</sup> 0 *u u ul l bl d*

*h u h u Jl t H* <sup>000</sup>

<sup>000</sup> <sup>0</sup> 0 *l bb l h u h u Jl*

<sup>0</sup> 0.380 *u x*

0.211

Further, from Equation 16, we can deduce the speed *ub* of the air flowing out of boundary layer at the end of the plant stage, as shown in Equation 22. If wind speed *u*<sup>0</sup> is small, the

*<sup>u</sup> V dx <sup>l</sup>*

*v*

1/5

1/5 0 9/5

 

> 

0 *<sup>w</sup>*

 

(12)

*h u xh q x M xr t H* (13)

*l b b bl d ql Mlr t H <sup>w</sup>* (17)

*l bb l* <sup>0</sup> (18)

(19)

(20)

(21)

*J q rM* (14)

 *h u xh J x t H* (15)

(16)

used to fit the seven coefficients mentioned above. The fitted coefficients and fitting accuracy (Root Mean Square Error, %RMSE) are shown in Table 3. Then the fitted equations are used to calculate the leaf stomatal conductance and transpiration rate. The calculate transpiration rates are compared with measured ones, as shown in Figure 14. The average and root mean square error (RMSE) are 2.9% and 18.8%, which show that the model accuracy is acceptable for simulation study the energy performance of the combination system.

Fig. 14. Comparison of calculated and measured transpiration rate.


Table 3. Coefficients fitted for calculating stomatal conductance.

#### **5.2 Boundary layer model**

After the transpiration rate is obtained, next step is to use it to calculate how much the air temperature can be decreased by the transpiration. A physical model is considered, as shown in Figure 15.

Fig. 15. Air boundary layer model.

used to fit the seven coefficients mentioned above. The fitted coefficients and fitting accuracy (Root Mean Square Error, %RMSE) are shown in Table 3. Then the fitted equations are used to calculate the leaf stomatal conductance and transpiration rate. The calculate transpiration rates are compared with measured ones, as shown in Figure 14. The average and root mean square error (RMSE) are 2.9% and 18.8%, which show that the model accuracy is acceptable for simulation study the energy performance of the combination

Transpiration R ate , S olar Radiation 20 07 08 15 ~ 2 00 70 82 1

S olar M easured

C alcu late d(W ind n ot conside re d) C alcu late d(W ind c on sidered)

07/08/15 07/08/16 07/08/17 07/08/18 07/08/19 07/08/20 07/08/21 07/08/22 <sup>0</sup>

*<sup>s</sup>*max *g a* <sup>1</sup> *b* <sup>2</sup> *b To Th Tl* %RMSE 2.464E+00 5.174E-03 4.898E-01 1.979E+00 3.748E+01 4.350E+01 2.591E+01 8.70%

After the transpiration rate is obtained, next step is to use it to calculate how much the air temperature can be decreased by the transpiration. A physical model is considered, as

Wind Speed *u*<sup>0</sup>

*ub*

*ud*

*x*

Average Speed

*x*+Δ*x* Transpiration

*ux*,*hb* Average Speed *ux*+Δ*x*,*hb*

*u*0,*h*<sup>0</sup>

Solar Radiation *J*

*ud*,*hb*

*x x*+Δ*x*

δ*+*

δ

Rate *M*

Δδ

Outdoor Unit Fan

Fig. 14. Comparison of calculated and measured transpiration rate.

Table 3. Coefficients fitted for calculating stomatal conductance.

Air Boundary Layer *<sup>u</sup>*0,*h*<sup>0</sup>

0 *l*

system.

1.5

0.5

**5.2 Boundary layer model** 

Solar Radiation *J*

*udx*,*hb*

Fig. 15. Air boundary layer model.

*x*

*ux ux*+Δ*x*,*hb* ,*hb*

Δδ

shown in Figure 15.

*z*

Transp Rate[L/(m2h)], Solar[kW/m2]

1

When air flows along the plant leaf surface, an air boundary layer forms. If we take a micro air volume with the x-length of *x* , y-length of one meter, and z-length of boundary thickness as the analysis object, the mass conservation is shown in Equation 12 and energy conservation is shown in Equation 13. If the thermal stored in leaf is ignored, the leaf energy conservation is shown in Equation 14. If substituting Equation 14 into Equation 13, the energy conservation will be as shown in Equation 15. Because it is difficult to solve Equation 12 and 15, and what we want to know is not the temperature distribution but the average temperature, we can take the total boundary lay as the analysis object. Thus the mass conservation becomes as shown in Equation 16 and energy conservation becomes as shown in Equation 17. If substituting Equation 14 and Equation 16 into Equation 17, the energy conservation becomes as shown in Equation 18. Further, because the volume of boundary layer *Vb* is so small (it is 88.7 m3 for the green roof with a length of 10 m when calculating using boundary layer thickness Equation 20 (Kato, 1964) and boundary layer volume Equation 21) comparing to the air volume entering the boundary layer (usually several thousands m3/h) that it is safe to ignore the energy change of boundary layer air, i.e. *H* 0 . Therefore energy conservation becomes as shown in Equation 19. The physical meaning of Equation 19 is that the solar radiation energy absorbed by plant leaves becomes the energy difference between the air flowing into and flowing out of the boundary layer if the leaves temperature change and boundary layer air energy change are ignored.

$$
\mu\_0 \Delta \mathcal{S} + \overline{\boldsymbol{\mu}}\_{\mathbf{x}} \boldsymbol{\delta} - \overline{\boldsymbol{\mu}}\_{\mathbf{x} + \boldsymbol{\Delta x}} (\mathcal{S} + \boldsymbol{\Delta \mathcal{S}}) - \mu\_{d\mathbf{x}} \Delta \mathbf{x} = \mathbf{0} \tag{12}
$$

$$[\rho\_0 \mu\_0 \Delta \delta \mathbf{h}\_0 + \rho\_b \overline{\mu}\_{b, \mathbf{x}} \delta \mathbf{h}\_{\mathbf{x}} - \rho\_b \overline{\mu}\_{b, \mathbf{x} + \Delta \mathbf{x}} (\delta + \Delta \delta) \mathbf{h}\_{\mathbf{x} + \Delta \mathbf{x}} - \rho\_b \mu\_{d, \mathbf{x}} \Delta \mathbf{x} \mathbf{h}\_{\mathbf{x}} - q \Delta \mathbf{x} + M \Delta \mathbf{x} r\_w] \Delta t = \Delta H \tag{13}$$

$$
\alpha \,\mathrm{J} + q - r\_w M = 0 \tag{14}
$$

$$\left[\mu\_0 \rho\_0 \Delta \delta \mathbf{h}\_0 + \rho\_b \overline{\mu}\_{b,\mathbf{x}} \delta \mathbf{h}\_{\mathbf{x}} - \rho\_b \overline{\mu}\_{b,\mathbf{x}+\mathbf{A}\mathbf{x}} (\delta + \Delta \delta) \mathbf{h}\_{\mathbf{x}+\mathbf{A}\mathbf{x}} - \rho\_b \mu\_{d,\mathbf{x}} \Delta \mathbf{x} \mathbf{h}\_{\mathbf{x}} + \alpha f \Delta \mathbf{x}\right] \Delta t = \Delta H \tag{15}$$

$$
\mu\_0 \mathcal{S}\_l - \overline{\mu}\_b \mathcal{S}\_l - \overline{\mu}\_d l = 0 \tag{16}
$$

$$\Delta \left( h\_0 \rho\_0 \mu\_0 \delta\_l - \overline{h}\_b \rho\_b \left( \overline{u}\_b \delta\_l + \overline{u}\_d l \right) - q l + M l r\_w \right) \Delta t = \Delta H \tag{17}$$

$$\left(h\_0 \rho\_0 \mu\_0 \delta\_l - \overline{h}\_b \rho\_b \mu\_0 \delta\_l + \alpha \operatorname{I}\right) \Delta t = \Delta H \tag{18}$$

$$
\lambda \, h\_0 \rho\_0 \mu\_0 \delta\_l - \overline{h}\_b \rho\_b \overline{\mu}\_0 \delta\_0 + \alpha \, l l = 0 \tag{19}
$$

$$\frac{\delta}{\nu} = 0.380 \sqrt{\left(\frac{\mu\_0 \chi}{\nu}\right)^{1/5}} \tag{20}$$

$$V\_b = \int\_0^l \delta d\mathbf{x} = 0.211 \left(\frac{u\_0}{v}\right)^{-1/5} l^{9/5} \tag{21}$$

Further, from Equation 16, we can deduce the speed *ub* of the air flowing out of boundary layer at the end of the plant stage, as shown in Equation 22. If wind speed *u*<sup>0</sup> is small, the

Improving Air-Conditioners' Energy Efficiency Using Green Roof Plants 219

Temperature D ifference20070815~ 20070821

M easured C alculated

07/08/15 07/08/16 07/08/17 07/08/18 07/08/19 07/08/20 07/08/21 07/08/22 <sup>0</sup>

The flow of predicting the energy saving of the combination system is shown in Figure 17.

Solar position Solar radiation *J* Outdoor unit data

Solar shading effect

Solar shading model

Air temperature increased by solar radiation

Air-conditioner model

*s* Cooling load simulation software

Colling load

Indoor air temperature

Combination system energy consumption *Eg*

Building data Weather data

Fig. 17. The flowchart of predicting energy consumption of the system combining the green

Energy saving *E*

roof plants with air-conditioner outdoor units.

Fig. 16. Comparison of calculated and measured air temperature difference.

The following five modules are used to calculate the energy savings.

*o*

**6. Methodology for predicting energy saving effect using simulation** 

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

**6.1 Simulation flow** 

Outdoor temperature

Transpiration cooling effect

Outdoor humidity *Ho* Solar radiation *J*

Transpiration model

Transpiration rate *M*

Boundary layer model

*b*

Conventional system energy consumption *Ec*

Air temperature

Temp Difference[oC]

*ub* might be minus. This means the air volume sucked by the air-conditioner fan is large enough to suck all the boundary layer air to the underside of the plant. When *ub* <0, the energy conservation will become as shown in Equation 23. From Equation 19 and 23, we can deduce the equation for calculating the average enthalpy of boundary layer air *<sup>b</sup> h* , as shown in Equation 24. Same as the enthalpy, the equations for calculating humidity ratio *<sup>b</sup> x* can be deduced, as shown in Equation 25. Where, the average speed of air flowing into the underside of plant is calculated using Equation 26. Further from the enthalpy definition Equation 27, we can deduce the equation for calculating air temperature, as shown in Equation 28. Thus, the air temperature changed by the plant transpiration *<sup>T</sup>* can be calculated, as shown in Equation 29.

$$
\overline{\boldsymbol{\mu}}\_b = \boldsymbol{\mu}\_0 - \overline{\boldsymbol{\mu}}\_d \frac{\boldsymbol{l}}{\boldsymbol{\delta}\_l} \tag{22}
$$

$$
\hbar \rho\_0 \rho\_0 \mu\_d l - \overline{h}\_b \rho\_b \overline{u}\_d l + \alpha \, ll = 0 \tag{23}
$$

$$\begin{aligned} \overline{u}\_b > 0 : \overline{h}\_b &= \frac{h\_0 \rho\_0 u\_0 \delta\_l + \alpha f l}{\rho\_b u\_0 \delta\_l} \\ \overline{u}\_b < 0 : \overline{h}\_b &= \frac{h\_0 \rho\_0 \overline{u}\_d + \alpha f}{\rho\_b \overline{u}\_d} \end{aligned} \tag{24}$$

$$\begin{aligned} \mu\_b > 0 : \overline{\boldsymbol{x}}\_b &= \frac{\boldsymbol{\chi}\_0 \rho\_0 \boldsymbol{u}\_0 \boldsymbol{\delta}\_l + M \boldsymbol{l}}{\rho\_b \boldsymbol{u}\_0 \boldsymbol{\delta}\_l} \\ \boldsymbol{u}\_b \prec 0 : \overline{\boldsymbol{x}}\_b &= \frac{\boldsymbol{\chi}\_0 \rho\_0 \overline{\boldsymbol{u}}\_d + M}{\rho\_b \overline{\boldsymbol{u}}\_d} \end{aligned} \tag{25}$$

$$
\overline{\mu}\_d = \frac{V}{I} \tag{26}
$$

$$\overline{h}\_b = \mathbb{C}\_{pu}\overline{\theta}\_b + \overline{\mathfrak{x}}\_b \left( \mathbb{C}\_{pw}\overline{\theta}\_b + r\_{w0} \right) \tag{27}$$

$$
\overline{\partial}\_b = \frac{\overline{h}\_b - \overline{\mathbf{x}}\_b r\_{w0}}{\mathbf{C}\_{pa} + \overline{\mathbf{x}}\_b \mathbf{C}\_{pw}} \tag{28}
$$

$$
\Delta\theta\_T = \theta\_0 - \overline{\theta}\_b \tag{29}
$$

To check the model accuracy, the measured air temperature changes are compared with the model calculated ones from August 15 to September 23. The comparison of the first one week is shown in Figure 16. The average error of the whole comparison period is 3.32% and the %RMSE is 94.02%. The model accuracy is not so high. But considering that the model is a totally physical model so it can be easily used for all situations, the model is acceptable for the further simulation study.

#### **6. Methodology for predicting energy saving effect using simulation**

#### **6.1 Simulation flow**

218 Energy Efficiency – A Bridge to Low Carbon Economy

*ub* might be minus. This means the air volume sucked by the air-conditioner fan is large enough to suck all the boundary layer air to the underside of the plant. When *ub* <0, the energy conservation will become as shown in Equation 23. From Equation 19 and 23, we can deduce the equation for calculating the average enthalpy of boundary layer air *<sup>b</sup> h* , as shown in Equation 24. Same as the enthalpy, the equations for calculating humidity ratio *<sup>b</sup> x* can be deduced, as shown in Equation 25. Where, the average speed of air flowing into the underside of plant is calculated using Equation 26. Further from the enthalpy definition Equation 27, we can deduce the equation for calculating air temperature, as shown in

*<sup>T</sup>* can be

(24)

(25)

Equation 28. Thus, the air temperature changed by the plant transpiration

0 :

*b b*

<0 :

*b b*

0 :

*b b*

*u x*

<0 :

*u x*

*b b*

*b*

*b d* 0

0 0 0 *d bbd h u l h u l Jl*

*h u Jl u h*

000

 

0 0

000

 

 

*b d*

 

*b bw*0

*pa b pw h xr C xC*

*T b* <sup>0</sup> 

To check the model accuracy, the measured air temperature changes are compared with the model calculated ones from August 15 to September 23. The comparison of the first one week is shown in Figure 16. The average error of the whole comparison period is 3.32% and the %RMSE is 94.02%. The model accuracy is not so high. But considering that the model is a totally physical model so it can be easily used for all situations, the model is acceptable for

*u*

0 0

*d V*

*h C xC r b pa b b pw b w* 

*u*

*hu J u h*

 

*uuu*

*l l*

0

 

0

*b l d*

*l*

*x u Ml*

*u xuM*

*b l d*

*u*

*b d*

*u*

*l*

 

(22)

(23)

*<sup>l</sup>* (26)

<sup>0</sup> (27)

(29)

(28)

calculated, as shown in Equation 29.

the further simulation study.

The flow of predicting the energy saving of the combination system is shown in Figure 17. The following five modules are used to calculate the energy savings.

Fig. 17. The flowchart of predicting energy consumption of the system combining the green roof plants with air-conditioner outdoor units.

Improving Air-Conditioners' Energy Efficiency Using Green Roof Plants 221

improvement ratio and 1% for air-conditioners with low efficiency improvement ratio. If water-sprinkle is conducted two hours per day, the energy savings are 9% and 2%

Further, plant transpiration model was described, which is used to calculate the transpiration rate. And boundary layer model was described as well, which is used to calculate the temperature decrease caused by the plant transpiration. The models were validated using the experimental data. The validation results show that: 1) the average and %RMSE of transpiration model are 2.9% and 18.8% respectively. The average error of boundary layer model is 3.32% and the %RMSE is 94.02%. The model accuracies are not very high, which imply that the models need further improvement. Considering that the models are physical model and it can be easily used for all situations, the models are

Finally the methodology of predicting the energy saving potential using the models is explained. Four typical climate areas in Japan are selected to simulate the cooling loads and energy saving potentials of combining green roof plant with air-conditioner outdoor units. The simulation results show that the energy saving rates are similar for the four typical climate regions, which are about 8%, 6% and 1% respectively for the air-conditioners with high, middle and low efficiency-improvement ratio accompanying to air temperature

The research described in this chapter is financially supported by Nissan Science Foundation, the Japan Society for the Promotion of Science (Project No. 17760468, 2004) and the Kansai Electric Power Co. Inc. The hydroponic cultivation system are provided and technically supported by Kyowa Corporation Ltd. Here the grateful acknowledgements are

respectively.

decreasing.

**8. Acknowledgement** 

**9. Nomenclature** 

expressed to all the supporters.

*a* : Reaction efficiency of stoma opening corresponding to light

*CA*: Cooling amount produced by an air-conditioner (kW)

*<sup>a</sup> g* : Conductance of leaf surface boundary layer, mol/m2s *<sup>s</sup>*max *g* : The maximum stomatal conductance, mol/m2s

*Cpa*: Specific heat of dry air, 1.005 kJ/(kgDAoC) *Cpw*: Specific heat of water vapor, 1.846 kJ/(kgoC) *Ct*: Outdoor wind turbulent coefficient, 1.4

*<sup>s</sup> g* : Stomatal conductance, mol/m2s *D* : Saturation pressure deficit, kPa

*h*: Air enthalpy, kJ/kgDA *J*: Global sky radiation, kW/m2 *l*: Length of green roof, m

*E*: Power consumption of air-conditioner, kW

*ai, bi, ci, d, i=1,2,3*: Coefficients fitted using manufacturer specification data

*CQ*: Correlation coefficient between PAR and sky global radiation, 2.1

acceptable for simulation study.


#### **6.2 Simulation results**

The former explained methodology is used to check the energy saving potential of the airconditioners made by six main air-conditioner manufactures. They are divided into to three groups: high, middle and low improvement of energy efficiency accompanying to the outdoor air temperature decreasing. The cooling loads of a standard office building at four typical climate areas of Sapporo, Tokyo, Osaka, and Naha in Japan are simulated and used to estimate the energy consumed by the three groups of air-conditioner at the conditions of with and without combining outdoor unit with hydroponic-cultivated sweet potato. The results are shown in Table 4. The energy saving rate is similar at four different areas. It is about 8%, 6% and 1% for the high, middle and low efficiency-improvement group respectively.


Table 4. Energy saving potentials of the combination system at four typical climate areas.

#### **7. Summary**

This chapter describes how to improve energy efficiency of air-conditioners by combining the air-conditioner outdoor units with green roof plants for the purpose of utilizing the plant transpiration cooling effect and solar shading effect.

Experimental system was set up to measure the actual plant cooling and solar shading effect. The measurement results show that: 1) the air temperature can be cooled down by the hydroponic-cultivated sweet potato by 1.3oC in average for clear day and 3oC in average when water was sprinkled; 2) more than 99% solar radiation can be shaded by the plants.

Based on the experimental results, energy saving potential of the combination system was estimated for typical air-conditioners from different manufacturers. The results show that for clear days, the energy saving ratio is about 4% for air-conditioners with high efficiency improvement ratio and 1% for air-conditioners with low efficiency improvement ratio. If water-sprinkle is conducted two hours per day, the energy savings are 9% and 2% respectively.

Further, plant transpiration model was described, which is used to calculate the transpiration rate. And boundary layer model was described as well, which is used to calculate the temperature decrease caused by the plant transpiration. The models were validated using the experimental data. The validation results show that: 1) the average and %RMSE of transpiration model are 2.9% and 18.8% respectively. The average error of boundary layer model is 3.32% and the %RMSE is 94.02%. The model accuracies are not very high, which imply that the models need further improvement. Considering that the models are physical model and it can be easily used for all situations, the models are acceptable for simulation study.

Finally the methodology of predicting the energy saving potential using the models is explained. Four typical climate areas in Japan are selected to simulate the cooling loads and energy saving potentials of combining green roof plant with air-conditioner outdoor units. The simulation results show that the energy saving rates are similar for the four typical climate regions, which are about 8%, 6% and 1% respectively for the air-conditioners with high, middle and low efficiency-improvement ratio accompanying to air temperature decreasing.

#### **8. Acknowledgement**

220 Energy Efficiency – A Bridge to Low Carbon Economy

1. The plant transpiration model, which is used to estimate the plant transpiration rate *M*.

3. Solar shading model (Equation 1), which is used to consider the solar shading effect. 4. Cooling load simulation software, which is used to simulated the cooling load of a

5. Air-conditioner model, which is used to calculate an air-conditioner's energy consumption give the outdoor air temperature, indoor air temperature, and cooling load. The regression model fitted using manufacturers' specification data is used

The former explained methodology is used to check the energy saving potential of the airconditioners made by six main air-conditioner manufactures. They are divided into to three groups: high, middle and low improvement of energy efficiency accompanying to the outdoor air temperature decreasing. The cooling loads of a standard office building at four typical climate areas of Sapporo, Tokyo, Osaka, and Naha in Japan are simulated and used to estimate the energy consumed by the three groups of air-conditioner at the conditions of with and without combining outdoor unit with hydroponic-cultivated sweet potato. The results are shown in Table 4. The energy saving rate is similar at four different areas. It is about 8%, 6% and 1% for the high, middle and low efficiency-improvement group

Table 4. Energy saving potentials of the combination system at four typical climate areas.

Use Not Use Use Not Use Use Not Use Use Not Use 32396.30 32449.91 0.2% 26876.48 28371.19 5.3% 24250.59 26250.85 7.6% 24733.34 26492.37 6.6% 6353.16 6430.23 1.2% 5396.26 5726.88 5.8% 5252.81 5690.67 7.7% 4970.74 5355.23 7.2% 10314.08 10438.78 1.2% 9595.27 10211.44 6.0% 8669.72 9465.12 8.4% 9089.36 9880.86 8.0% 12675.15 12801.64 1.0% 12050.77 12845.89 6.2% 10430.95 11445.15 8.9% 11388.72 12430.08 8.4%

Energy saving rate

Group 1 Group 2 Group 3

Energy consumption [kWh]

AE CD

Energy consumption [kWh]

Energy saving rate

Energy consumption [kWh]

Energy saving rate

This chapter describes how to improve energy efficiency of air-conditioners by combining the air-conditioner outdoor units with green roof plants for the purpose of utilizing the

Experimental system was set up to measure the actual plant cooling and solar shading effect. The measurement results show that: 1) the air temperature can be cooled down by the hydroponic-cultivated sweet potato by 1.3oC in average for clear day and 3oC in average when water was sprinkled; 2) more than 99% solar radiation can be shaded by the plants.

Based on the experimental results, energy saving potential of the combination system was estimated for typical air-conditioners from different manufacturers. The results show that for clear days, the energy saving ratio is about 4% for air-conditioners with high efficiency

plant transpiration cooling effect and solar shading effect.

Energy saving rate

Energy consumption [kWh]

*<sup>b</sup>* , which is the

2. Boundary layer model, which is used to calculate the air temperature

temperature after being decreased by plant transpiration.

given building.

(Equation 2).

**6.2 Simulation results** 

respectively.

Sapporo

Manufactuer Energy related to use or not use roof plant

> Tokyo Osaka Naha

**7. Summary** 

The research described in this chapter is financially supported by Nissan Science Foundation, the Japan Society for the Promotion of Science (Project No. 17760468, 2004) and the Kansai Electric Power Co. Inc. The hydroponic cultivation system are provided and technically supported by Kyowa Corporation Ltd. Here the grateful acknowledgements are expressed to all the supporters.

#### **9. Nomenclature**

*a* : Reaction efficiency of stoma opening corresponding to light *ai, bi, ci, d, i=1,2,3*: Coefficients fitted using manufacturer specification data *CA*: Cooling amount produced by an air-conditioner (kW) *Cpa*: Specific heat of dry air, 1.005 kJ/(kgDAoC) *Cpw*: Specific heat of water vapor, 1.846 kJ/(kgoC) *Ct*: Outdoor wind turbulent coefficient, 1.4 *CQ*: Correlation coefficient between PAR and sky global radiation, 2.1 *<sup>a</sup> g* : Conductance of leaf surface boundary layer, mol/m2s *<sup>s</sup>*max *g* : The maximum stomatal conductance, mol/m2s *<sup>s</sup> g* : Stomatal conductance, mol/m2s *D* : Saturation pressure deficit, kPa *E*: Power consumption of air-conditioner, kW *h*: Air enthalpy, kJ/kgDA *J*: Global sky radiation, kW/m2 *l*: Length of green roof, m

Improving Air-Conditioners' Energy Efficiency Using Green Roof Plants 223

Alexandria Eleftheria, Jones Phil (2008), Temperature decreases in an urban canyon due to

Clerk Corrie, Adriaens Peter, Nriantalbot Albot, Talbot F. Brian (2008), Green Roof

Campbell, G. S. & Norman, J. M. (1998) An Introduction to Environmental Biophysics

Di H. F. and Wang D. N. (1999), Cooling Effect of Ivy on a Wall, Experimental Heat

Elena Palomo Del Barrio (1998), Analysis of the green roofs cooling potential in buildings,

Jarvis, P.G. (1976) The interpretation of the variations in leaf water potential and stomatal

Kadaira, A., Yoshida, H., Ito, M. et al. (2005) A study on the climatic mitigation effect of trees

Kumar Rakesh, Kaushik S.C. (2005), Performance evaluation of green roof and shading for thermal protection of buildings Building and Environment, Vol. 40, pp. 1505–1511 Spala A., Bagiorgas H.S., Assimakopoulos M.N., Kalavrouziotis J., Matthopoulos D.,

Takebayashi Hideki, Moriyama Masakazu (2007), Surface heat budget on green roof and

Wong Nyuk Hien, Chen Yu, Ong Chui Leng, Sia Angelia (2003) Investigation of thermal

Wang Fulin, Yoshida Harunori, Masuhara Satoshi, Kitagawa Hiroaki, and Goto Kyoko

conductance found in canopies in the field. Phil.Trans.R.Soc.Lond.B, 273, pp.593-

in an area around a residential complex. Journal of Environmental Engineering,

Mihalakakou G. (2008), On the green roof system. Selection, state of the art and energy potential investigation of a system installed in an office building in Athens,

high reflection roof for mitigation of urban heat island, Building and Environment,

benefits of rooftop garden in the tropical environment, Building and Environment,

(2005), Simulation-Based Automated Commissioning Method for Air-Conditioning Systems and Its Application Case Study, Proceedings of the 9th International Building Performance Simulation Association Conference, Montreal, Canada, pp.

Environmental Science Technology, Vol. 42, pp. 2155–2161

Second edition. p101, p108, Springer, New York, US

Architecture Institute of Japan, No. 598, pp. 71-77.

Greece, Renewable Energy, Vol. 33, pp.173-177

Energy and Buildings, Vol. 27, pp.179-193

Kato, K. (1964), Heat transfer. Yokendo Co., ltd., p105.

Vol. 42, PP. 2971–2979

Vol. 38, pp.261-270

1307-1314

green walls and green roofs in diverse climates, Building and Environment, Vol. 43,

Valuation: A Probabilistic Economic Analysis of Environmental Benefits,

*l*: Length of the green roof

*L*: Long wave *S*: Solar radiation *T*: Transpiration

Superscript *t* : Time step *t* : Average

**10. References** 

pp. 480–493

610.

Transfer, Vol. 12, pp.235-245


#### Subscription


*l*: Length of the green roof *L*: Long wave *S*: Solar radiation *T*: Transpiration

Superscript

222 Energy Efficiency – A Bridge to Low Carbon Economy

: Absorption ratio to short wave radiation, leaves of sweet potato: 0.56; outdoor unit of

*of* : Final outdoor air dry bulb temperature after transpiration cooling and solar radiation

mol/m2s

*n*<sup>1</sup> : The saturation pressure deficit when stomatal conductance becomes half

*Lq* : Long wave radiation between outdoor unit and its surroundings (W)

*Wi* : Saturation partial pressure of water vapor at leaf surface, mb/mb

*<sup>s</sup>* : Short wave radiation absorption ratio of outdoor unit surface

*H* : Energy change of the air in the boundary layer, kJ

*m* : Outdoor unit fan air mass flow rate (kg/s)

*Q* : photosynthetically active radiation (PAR),

*ar* : Boundary layer resistance, m2s/mol

*Wa* : Water vapor partial pressure, mb/mb

*x* : Air humidity ratio, kg/kgDA

: Thickness of boundary layer, m

: Kinematic viscosity, m2/s

: Air dry bulb temperature, oC

*o*: Air out of the boundary layer *b*: Air in the boundary layer

*<sup>i</sup>* : Indoor air wet bulb temperature, oC

*<sup>o</sup>* : Outdoor air dry bulb temperature, oC

: Change of air dry bulb temperature, oC

*d*: Air flowing into the underside of green roof plant *g*: Use hydroponic-caltivated green roof plant

*sr* : Stomatal resistance, m2s/mol

*T* : Leaf surface temperature, oC *To* : The most proper temperature, oC *Th* : Upper applicable temperature, oC *Tl* : Lower applicable temperature, oC

*u*: Air flow speed, m/s *V*: Air volume flow rate, m3/s

air-conditioner: 0.76

: Air density, kg/m3

*t* : Time interval, s

absorption, oC

Subscription

*n*<sup>2</sup> : Curvature of the saturation pressure deficit function *q*: Sensible heat exchange between leaves and air, kW/m2

*<sup>s</sup> q* : Short wave radiation (i.e. global solar radiation) (W)

*<sup>w</sup>*<sup>0</sup> *r* : Latent heat of water evaporation at 0oC, 2501 kJ/kg

*Mw* : Molar mass of water, 0.01802 kg/mol

*M*: Transpiration rate, kg/m2s

*t* : Time step *t*

: Average

#### **10. References**


**Part 4** 

**Energy Efficiency on Supply Side** 

