Technical note

Cooling potential of ventilated PV facade and solar

air heaters combined with a desiccant

cooling machine

Li Meia,*, David Infieldb, Ursula Eickerc, Dennis Lovedaya,Volker Fuxc

aDepartment of Civil and Building Engineering, Loughborough University, Loughborough, LE11 3TU, UKbCREST, Loughborough University, Loughborough, LE11 3TU, UKcHochschule fur Technik Schellinstr 24, 70174 Stuttgart, Germany

Received 29 November 2004; accepted 12 June 2005

Available online 11 August 2005

Abstract

Thermal energy collected from a PV-solar air heating system is being used to provide cooling for

the Mataro Library, near Barcelona. The system is designed to utilise surplus heat available from the

ventilated PV facade and PV shed elements during the summer season to provide building cooling. A

desiccant cooling machine was installed on the library roof with an additional solar air collector and

connected to the existing ventilated PV facade and PV sheds. The desiccant cooling cycle is a novel

open heat driven system that can be used to condition the air supplied to the building interior. Cooling

power is supplied to the room space within the building by evaporative cooling of the fresh air supply,

and the solar heat from the PV-solar air heating system provides the necessary regeneration air

temperature for the desiccant machine. This paper describes the system and gives the main technical

details. The cooling performance of the solar powered desiccant cooling system is evaluated by the

detailed modelling of the complete cooling process. It is shown that air temperature level of the PV-

solar air heating system of 70 8C or more can be efficiently used to regenerate the sorption wheel in

the desiccant cooling machine. A solar fraction of 75% can be achieved by such an innovative system

and the average COP of the cooling machine over the summer season is approximate 0.518.

q 2005 Elsevier Ltd. All rights reserved.

Keywords: Solar energy; Ventilated PV facade/shed; Desiccant cooling; Energy balance calculation

Renewable Energy 31 (2006) 1265–1278

www.elsevier.com/locate/renene

0960-1481/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.

doi:10.1016/j.renene.2005.06.013

* Corresponding author. Tel.: C44 1509223778; fax: 44 1509610031.

E-mail address: l.mei@lboro.ac.uk (L. Mei).

L. Mei et al. / Renewable Energy 31 (2006) 1265–12781266

1. Introduction

A building integrated ventilated photovoltaic (PV) system is advantageous both from

electrical and thermal point of view. The air circulation behind the PV panel lowers the

temperature of PV module and thus improves their electrical performance. Furthermore, a

controlled air flow in the cavity between the PV panel and the surface behind leads to solar air

heating with the potential to be applied to building for winter preheating or summer cooling.

Over a number of years, the authors have investigated the thermal performance and the

application of the ventilated PV facade integrated to building in winter season [1–3].

Directly using of the warmed air from the PV facade for building heating is

environmentally sound. However, the thermal energy generated in the ventilated PV

facade is usually not used in summer when it is most abundant. It is also conceivable that

the PV facade structure might even add to the building’s cooling load. An attractive

solution to this problem is to use the heated air from the facade (which in summer is

unwanted and normally vented to the outside) to drive an adsorption cooling system for

buildings cooling in summer. The authors have recently completed an EU project which

concerned the ventilated PV-solar air heating system combining with a desiccant cooling

machine constructed and installed in the Mataro Library, near Barcelona.

The aim of this project is to design and implement an integrated solar heating and

cooling system using building integrated PV solar air heating components coupled to a

desiccant cooling system. The main scientific objective of the project is an integrated

energy analysis of such systems taking into account the thermal properties of the ventilated

PV facade and its impact on the heating and cooling loads of the building as well as the

generation of solar heated air and operation of the desiccant cooler. This has been done by

dynamic modelling and validation through an experimental evaluation of the system.

Several major innovations have been achieved in the project:

† The project proposed the first building integrated solar heating and cooling system in

Europe, using ventilated PV and solar air collectors for the heat production and

desiccant wheel technology for the solar driven cooling system

† An integrated energy analysis for the first time deliver validated results of the summer

performance of PV/solar air ventilated facades, consisting both the passive cooling

load caused the integrated components and their thermal energy production usable for

the active cooling system.

† Optimised control algorithms for solar powered desiccant systems have been

developed and tested. The control algorithm development and the integration into

standard building management system can enhance system performance and

significantly contribute to a cost reduction of such innovative cooling concepts.

2. The description of the system

The Mataro library building, near Barcelona, has a total surface area of 3200 m2,

distributed over three floor levels [4]. Both the south facade and the inclined sheds on

PV shed 1

PV shed 2

PV shed 3

PV shed 4

PV Facade

Air collector fieldmc, Tc

3000~9000m3

ms, Ts

mf, Tf

Fig. 1. Solar air system.

L. Mei et al. / Renewable Energy 31 (2006) 1265–1278 1267

the roof are equipped with ventilated PV modules (225 m2 on the south facade and a total

of 300 m2 on the roof). The original south PV facade is 6.5 m high, to which the 1.7 m high

air collectors have been added to boost the temperature levels of the ventilated PV system.

The ventilation channels of the four rows of the PV sheds on the roof (308 inclination) are

connected in parallel to the PV facade channel to supply a merged air volume to an

additional solar air collector field. This additional air collector field has the heating area of

105 m2 and is separated in three parallel strings with 35 m length and 95 mm channel

height. The further heated air from the additional air collectors is drawn to the desiccant

machine to regenerate the sorption wheel. The regeneration air can be controlled between

the maximum of 9000 m3/h and minimum of 3000 m3/h volume flow rate to keep a

necessary regeneration air temperature. The solar air system is shown in the Fig. 1

The desiccant cooling machine installed on the roof of the Mataro library is an open

heat driven cycle which comprises a sorption dehumidify wheel, a heat recovery wheel

and the humidifiers. Such system will collect outdoor air as the supply air and will first

dehumidify it with a silica-gel desiccant. By this dehumidification, the supply air warms

up and pre-cooling with the exhaust air by a heat exchanger is necessary. Thus, the

dehumidified supply air is cooled in a heat recovery wheel and further in an evaporative

humidifier. After the humidification of the supply air the maximum cooling effect is

limited by the maximum water content which is allowed to guarantee thermal comfort in

the building. The cooled supply air after humidification is passed to the indoor space to

satisfy the human environmental requirements. It is clear that the latent and sensible loads

of the room space are more efficiently handled in the desiccant cooling process than in

other cooling methods since components can be designed to handle these loads separately.

In the exhausting air stream, the silica-gel desiccant is regenerated by the solar heat air

from the PV-solar air collectors for releasing water moistures.

The desiccant wheel used in the system comprises a matrix containing parallel channels

which enables a large heat and mass transfer area to be maintained. The matrix is coated

with the silica-gel desiccants and has the 2100 mm of diameter, 360 mm of depth. The

wheel rotates slowly at the speed of 20 revolutions per hour. While, the heat recovery

wheel having the same diameter as the desiccant wheel rotates at 600 revolutions per hour

for the effective heat exchange. The volume flow rate in the supply and exhaust air streams

6 5

431

PVshed

PV facade

Air Collectors

Ambient air

Sorption wheel Recovery wheelHumidifier

2

78

Fig. 2. The simple scheme of the system.

L. Mei et al. / Renewable Energy 31 (2006) 1265–12781268

are controlled in 3000, 6000 and 12,000 m3/h for the different control strategies. The

simple scheme for this desiccant cooling system is shown in the Fig. 2.

For such a solar powered desiccant cooling system, the component models were

established and the simulation study of its performance was carried out. The thermal

building simulation tool of TRNSYS was used for simulate the desiccant cooling system

integrated with the Mataro Library building. When simulating the system operation

process including the control strategies, the energy balances for the building cooling were

yield.

3. Solar application

Since the desiccant cooling system was driven by solar air heat, the total useful heat

energy and the outlet air temperature of the ventilated PV facade and PVsheds as well as

the additionally installed solar air heater should be determined. As the air channels of the

PV facade and the PV sheds are parallel connected to supply the preheated air to

the additional air collectors, the temperature of the merged air is an important factor in the

solar cooling procedure. In general, a simplified Duffie’s solar air collector model [5] for

the both PV facade and PV sheds developed by the authors in the previously completed

project [1] was utilised to estimate the outlet air temperature of the PV facade and PV

sheds. This model is based on the standard test method for solar collectors. The

instantaneous thermal efficiencies of the PV facade and PV shed were given by:

hpf ZQu;pf

GTApf

Z 0:27K6:37Tin;pf KTa

GT

(1)

and

Tout;pf ZQu;pf

Cp _mf

CTin;pf (2)

L. Mei et al. / Renewable Energy 31 (2006) 1265–1278 1269

where Tin,pf(8C), Ta(8C) and GT (W/m2) obtained from the data logging system of the

Mataro Library are the inlet air temperature of the PV facade, the ambient air temperature

and the solar irradiance, respectively. The parameters of 0.27and 6.37 (W/Km2) were

estimated using the measured data, such as the air mass flow rate, the inlet and outlet air

temperatures of the PV channel, from Mataro Library data acquisition system during the

period 01/08/2000–06/08/2000, for 40 min periods around midday. Consequently, the

outlet air temperature of the PV facade, Tout,pf (8C)can be calculated by the inlet air

temperature, Tin,pf (8C), mass flow rate _mf (kg/s) and the total useful thermal energy Qu,

pf (W). In the Eqs. (1) and (2), the subscripts only stand for the PV facade. The same model

was used for the PV sheds and not repeated here. Then, it is clear to notice that the different

collecting areas collect the different useful thermal energy and cause the different outlet air

temperature for both the PV facade and PV sheds.

The performance characteristic of the additional solar air collectors (105 m2) was given

by the manufacture’s test data which can be expressed as:

hc

Qu;c

GTAZ 0:86K6:5

Tin;c KTa

GT

(3)

where, Tin,c (8C), the air collector inlet air temperature is the merged air temperature of the

PV facade and PV sheds. The regenerating air temperature to the desiccant cooling

machine is thus obtained by:

Tout;c ZQu;c

Cp _mc

CTin;c (4)

In order to maintain the regeneration air temperature at a significant level (above

70 8C), the ventilation air volume through the PV facade and PV sheds and the additional

solar air collectors should be controlled between 3000 and 9000 m3/h. In the simulation

model, a simple PI controller was designed for this purpose and implemented in the

simulation program. Both measured and simulated results show the total temperature

increase from ambient to the regeneration side of 35–40 K at high irradiance levels.

In the solar air system, the air volume through the PV facade and the PV sheds are

connected to the air collectors in parallel. The merged air temperature to the air collector

inlet thus approximates to the following relationship:

Tin;c Z_ms

_ms C _mf

Tout;ps C_mf

_ms C _mf

Tout;pf (5)

where, the outlet air temperatures of the PV facade and PV sheds, Tout,pf and Tout,ps are

inverse to the air mass flow rates of the PV facade and PV sheds, _mf and _ms. Thus, the solar

air collector inlet temperature, Tin;c, is not varied with the different ration of the _ms= _mf .

Furthermore, the regeneration solar energy delivered to the desiccant cooling machine is

not varied with the different ratio of the _ms= _mf .

L. Mei et al. / Renewable Energy 31 (2006) 1265–12781270

4. Desiccant cooling

The desiccant cooling machine including the sorption wheel, heat recovery wheel and

the humidifiers was modelled to simulate the summe_rtime psychrometric desiccant

processes and utilise hourly meteorological data to predict energy consumption

For the sorption wheel, the dimensionless effectiveness was used to characterise the

performance of the dehumidifier. In this study, the effectiveness of the dehumidifier was

assumed to be a constant. The humidity ratio and temperature of the air leaving sorption

wheel in the supply stream were determined by:

X2 Z X1KhdwðX1KXreÞ (6)

and

T2 Zh2K2500X2

1:004 C1:875X2

(7)

where, X (kg/kg), h (kJ/kg) and T (8C) are the humidity ratio, specific enthalpy and

temperature of air, respectively. The subscripts of 1,2 and so on in the following equations

indicate the positions shown in the Fig. 2. Xre (kg/kg) is the humidity ratio of the

regeneration air. h2 is determined by an approximate expression for the enthalpy of the

mixture air and explained in the Eq. (11).

In the heat recovery wheel, only the sensible heat exchange is in the process and the

humidity ratio is always constant. The temperature of the supply air stream leaving the

heat recovery wheel was:

T3 Z T2KhhwðT2KT6Þ (8)

For the humidifier, the enthalpy is approximately a constant in the evaporative cooling

process. The humidity ratio and temperature of the air leaving the humidifier can be

determined by:

X4 Z X3KhhdðXsKX5Þ (9)

and

T4 Zh4K2500X4

1:004 C1:875X4

(10)

where, Xs (kg/kg) is the humidity ratio at saturation state.

The modelling of condition along the entire process of desiccant cooling was carried

out for multiple psychrometric calculation. Within the calculations, the values of specific

enthalpy of the mixed humid air, h2 and h4, which are the energy content of humid air,

were calculated by an approximate expression of the sum of the enthalpy of dry air and the

enthalpy of water vapour related to the reference temperature 0 8C:

h2 Z 1:005T2 CX2ð2500 C1:875T2Þ h4 Z 1:005T4 CX4ð2500 C1:875T4Þ (11)

L. Mei et al. / Renewable Energy 31 (2006) 1265–1278 1271

The psychrometric equations from (6)–(11) were numerically solved to obtain the

steady state psychrometric values for each position in the desiccant cooling process.

In addition, the energy balance calculations based on the air temperature, humidity ratio

and specific enthalpy at the inlet of the desiccant system and the room space as well as the

solar air collector outlet were represented as:

† The cooling energy generated from the desiccant system:

Q_dsc Z _mdðh1Kh4ðWÞ (12)

† The sensible cooling load removed from the room:

Q_supply Z _mdCpðTroom KT4ÞðWÞ (13)

† The regeneration energy from solar heating system:

Q_reg Z _mcCpðTreg KT8ÞðWÞ (14)

† The auxiliary cooling energy supplied to the room:

Q_aux Z _mdCpðTroom K25:5 8CÞðWÞ (15)

† COP of the desiccant system:

COP Z COPQ_dsc

Q_reg(16)

† Solar fraction ZQ_supply

Q_load(17)

Where, _md (kg/s) is the supply air mass flow rate of the desiccant cooling machine.

5. Temperature and humidity assessment

The cooling process of the solar powered desiccant cooling system can be

illustrated by a typical steady state psychrometric process from the model simulation

results. For example, on 15th of June 2002 at 2:00pm, the warm moist air at 29.2 8C

and 0.011 kg/kg moisture content (relative humidity is about 38.9%) was vent through

the sorption wheel so that it came off at 41.3 8C and 0.0088 kg/kg moisture content

(relative humidity is about 17.7%). The supply air stream then past through the heat

recovery wheel and was sensible cooled to 24.6 8C. The cooled air then past through

an evaporative humidifier with an adiabatic efficiency of approximate 85%. In this

case, air can be supplied to the room space at 16.4 8C and 0.0108 kg/kg moisture

content (relative humidity is about 93%). On the return air side, air from the room

space at 25.5 8C and 0.0102 kg/kg moisture content (relative humidity is about 48%)

was drawn through the evaporative cooler so that it entered the thermal wheel at

approximate 18.6 8C and 0.011 kg/kg moisture content (relative humidity is about

91%). When the retune air stream past through the heat recovery wheel, it was

sensibly heated to approximate 35 8C and exhausted to the outside.

Table 1

Comparison of simulated and measured data for cooling process

Exterior

air

After

sorption

After

heat

recovery

After

humidifier

Return

air

After

return

humidifier

Air to

outside

Regener-

ation air

Simulated

8C 29.2 41.3 24.6 16.4 25.5 18.6 35 69.5

Kg/kg 0.011 0.0088 0.0068 0.0108 0.0102 0.011 0.011 –

RH 38.9% 17.7% – 93% 48% – –

Measured

8C 29.2 38.9 26.5 17.2 25.8 20.2 37.4 61.7

RH 37.6% 19.9% – 94.8% 50.5% – –

L. Mei et al. / Renewable Energy 31 (2006) 1265–12781272

On the regeneration air stream, the solar heated air at approximate 69.5 8C was vent

through the sorption wheel to regenerate the desiccants.

For validation purpose, a group of data measured from the Mataro library data

acquisition system on 5th of June 2002 at 2:00pm were recorded in the Table 1 and

compared with the simulation results given above. It can be seen that the simulated air

temperatures and humilities along the air flow stream were quite close to the measured

data. However, at some points, there were the reasonable differences between the

measured and simulated values. This perhaps related to the process modelling accuracy or

the system measuring errors. Then, in the simulation the desiccant system reached 16.4 8C

inlet air temperature, which is the design value.

In this study we also focused on the system dynamical simulation so that the solar

energy used and the cooling energy consumed for summer time can be analysed. As the

cooling machine relies purely on solar thermal heat from the ventilated PV-solar heat

system, the regeneration air temperature cannot be assumed to be constant. Thus, the

achievable solar fraction should be carefully analysed. So far a dynamical simulation

based on the component models described before and the hourly meteor weather data of

Barcelona has been carried out.

For the computer simulation, the sophisticated control strategies of the system were

simplified as the following supply/exhaust air volume control cascade:

† Supply air volume control: if the room temperature and the outside temperature are

below 24 8C and above 21 8C, the free ventilation operation will start—the supply and

exhaust air volume is 3000 m3/h. If the room temperature is above 24 8C, the desiccant

cooling starts operating and the supply/exhaust air volume is 6000 m3/h. If the room

temperature is higher than 24.5 8C, the supply and exhaust air volume will be gradually

increased to the maximum value, 12,000 m3/h.

† Auxiliary cooling supply control: if desiccant system cannot supply satisfactory

cooling energy to maintain the room temperature at 25.5 8C (or 26 8C), the auxiliary

cooling starts operation.

The supply and the exhaust air volume control cascade is illustrated in Fig. 3.

Aux. Cooler

3000m2/hr

6000m2/hr

12000m3/hr

Fig. 3. Supply and exhaust air volume control cascade.

L. Mei et al. / Renewable Energy 31 (2006) 1265–1278 1273

The complete solar powered cooling system was simulated for a volume flow on the

fresh air side of 6000–12,000 m3/h (the volume flow rate was depending on the control

strategies) and a variable volume flow of 3000–9000 m3/h on the regeneration side. The

input irradiance, temperature and humidity values are taken from the Meteonorm

database. From simulation, it can be seen that the temperature increase of PV-solar heater

can reach from the exterior to the regeneration of 35–40 K at high irradiance levels.

Approximate 70 8C of the regeneration air temperature can be obtained by a proper

controller to vary the volume flow on the regeneration air side between 3000 and

9000 m3/h. Fig. 4 shows this effect.

From Fig. 5, the room inlet air reaches 15–17 8C during cooling operation period in the

first ten days of July. On days with low irradiance levels (for example the 6th of July) the

regeneration air temperature is so low that no lower cooling effect is achieved. As ambient

air temperature levels are also rather low on that day, this should not present a problem.

During the free ventilation period (night), the room inlet air temperature is as the same as

the ambient air temperature. Depending on the control strategies, the room air temperature

can be controlled between 24 and 25.5 8C during the daytime. The maximum room

temperature in July is not higher than 25.6 8C.

Fig. 6 gives the absolute humidity of the ambient air, (supply air), room air and the air

after the sorption wheel for the first 10 days of July. For humidity simulation, there are no

control strategies implemented. So, the humidity level in the room only depends on the

ambient humidity ratio and the efficiencies of the dehumidifier and the humidifier.

However, the absolute humidity for the room inlet air is not allowed to be higher than

0

20

40

60

80

1 25 49 73 97 121 145 169 193 217 241First 10 days in July

Tem

per

atu

res

(C)

Tamb Treg

Fig. 4. Regeneration and ambient air temperatures with simulation time.

0

10

20

30

40

1 25 49 73 97 121 145 169 193 217 241First 10 days in July

Tem

per

atu

res

(C)

Tamb Troom Tsupply

Fig. 5. The room, ambient and supply air temperatures with simulation time.

0.000

0.005

0.010

0.015

0.020

1 25 49 73 97 121 145 169 193 217 241

First 10 days in July

Ab

solu

te h

um

idit

y(k

g/k

g)

Xamb Xroom Xsupply Xdhum

Fig. 6. Absolute humidity from simulation.

L. Mei et al. / Renewable Energy 31 (2006) 1265–12781274

0.01 kg/kg (RH 95%, 15 8C) when the desiccant cooling system operating. The effective

dehumidification of the sorption reaches about 0.005 kg/kg despite using the fresh air

humidifier.

6. Cooling load of inf/aud room

The solar powered desiccant cooling system in the Mataro library was designed to

supply the cooling energy to the large reading room—Inf/Aud room. The Inf/Aud room is

located in the west side and ground level of the Mataro library. The north wall, the large

part of south wall and the ground floor of the Inf/Aud room are adjoining to the soil. The

east wall and the ceiling are the interior partition. Thus, in the cooling load calculation for

the Inf/Aud room, solar radiation contributes the thermal flux only to the west wall, the

west window and the small part of the south wall.

As required in the cooling load calculation, the room dimension, the construction

material and the interior radiation of the Inf/Aud room are listed in Table 2.

Consequently, the shading influence, the adjoin soil temperature and the wind speed

were carefully considered and the Barcelona climatic data were used for the cooling load

calculation over 12 months period. The average cooling loads of the Inf/Aud room for the

August is around 9000 kWh.

Table 2

The thermal information of the Inf/Aud room

Wall description Construction Surface area U_value

(W/m2K)

West window Double glazing with 6 mm distance between two

4 mm glasses, 10% aluminium frame without any

thermal break

30.4 m!3.

2 mZ97.28 m2

4

West wall 0.25 m concrete insulation with 0.007 m steel

cover

30.4 m!1.

3 mZ39.52 m2

0.334

North wall Concrete wall completely adjoining soil (insulated) 15.5 m!4.

5 mZ69.75 m2

0.530

East wall Adiabatic two 0.02 m gypsum boards 30.4 m!4.

5 mZ136.8 m2

0.406

South wall 1 Adjoining soil, with internal common block and

face brick

15.5 m!3.

2 mZ49.6 m2

0.530

South wall 2 Not adjoining soil, with internal common block

and face brick

15.5 m!1.

3 mZ20.15 m2

0.334

Floor Concrete with insulation 30.4 m!15.

5 mZ417.2 m2

0.460

Ceiling Adiabatic 0.2 m lightweight concrete 30.4 m!15.

5 mZ417.2 m2

0.406

Room volume 30.4 m!15.

5 m!4.5 mZ2120 m3

Interior heat

radiation

40 persons, 6000 w lights and computers

L. Mei et al. / Renewable Energy 31 (2006) 1265–1278 1275

7. The energy balance

One of the important aims of the project was being capable of calculating the energy

balance of the system by the models simulation so that the valuations of system performance

and cost can be forecasted. Using the system models developed above, the energy balance has

been calculated relating to the Barcelona climatic data and the system control strategies.

For cooling purpose, the cooling energy generated by the desiccant cooling system was

calculated using the enthalpy difference between the ambient air and the room inlet air

from April to October. The total amount of 40,639 kWh cooling energy obtained refers to

the total enthalpy reduction from the ambient air to the room inlet air.

The regeneration energy for driving the desiccant cooling machine, for the period of

April to October, from ventilated PV-solar air heating system was calculated as 68,

829 kWh.

Table 3

The Cooling and regeneration power for summer season

April May June July Aug Sept Oct Total

Q_dsc (kWh) 501 2692 6562 11,831 11,395 6624 1034 40,639

Q_reg (kWh) 3356 7120 12,790 18,103 15,686 9832 1942 68,829

COP 0.149 0.378 0.513 0.654 0.726 0.674 0.533 0.518

Table 4

The supplied cooling energy to the room

April May June July Aug Sept Oct Total

Q_supply (kWh)

(at 3000 m3/h)

4285 3428 2098 1006 1214 1764 2810 16,605

Q_supply (kWh)

(at 6000 m3/h)

750 1234 2543 2592 2128 1981 753 11,981

Q_supply (kWh)

(at 6000–12,

000 m3/h)

344 1305 2026 3100 2473 1443 0 10,691

Q_supply (kWh)

(at 12,000 m3/h)

0 223 528 1739 1209 604 0 4303

Total Q_sup-

ply (kWh)

5379 6190 7195 8437 7024 5792 3563 43,580

L. Mei et al. / Renewable Energy 31 (2006) 1265–12781276

The coefficient of performance (COP), i.e. the ratio of produced cooling power to

required regeneration power, is on average 0.518 for this application. Table 3 shows the

monthly data of the regeneration energy, the desiccant cooling energy and the COP

calculated.

Using the room temperature calculated simultaneously for the aud/inf room, the cooling

energy effectively delivered to the room was obtained which depends on the enthalpy

difference between the room air and the room inlet air. The total cooling energy supplied

to the room from the desiccant cooling system is about 43,580 kWh.

Depending on the control strategies, both free ventilation and cooling ventilation were

applied to the system and three different ventilation air volumes were used. With respect to

the total cooling energy supplied to the aud/inf room, the free ventilation cooling energy

(3000 m3/h supply volume flow rate) is about 38%; the desiccant cooling energy supplied

at the volume flow rates of 6000 m3/h and between 6000 and 12,000 m3/h are 27 and 24%,

respectively; only 9% desiccant cooling energy is supplied to the room at 12,000 m3/h

volume flow rate. Table 4 shows the detailed monthly values.

With respect to the control strategies, the auxiliary cooling energy starts operating

when the room temperature is higher than 25.5 8C. Therefore, the auxiliary cooling energy

was calculated from the simulation, which can keep the room temperature at 25.5 8C

during the high cooling load periods. If considering the total amount of desiccant cooling

supply plus the auxiliary cooling energy as the total cooling demands for the aud/inf room,

the monthly solar fraction for desiccant cooling system was calculated and given

in Table 5. In average, 93% of the cooling demand for summer season can be covered by

Table 5

The solar fraction

April May June July Aug Sept Oct Total

Total Q_supply (kWh) 5379 6190 7195 8437 7024 5792 3563 43,580

Q_aux (kWh) 0 21 131 486 2099 605 0 3341

Q_supplyCQ_aux (kWh)

5379 6211 7326 8923 9123 6397 3563 46,921

Solar fraction 1.000 0.993 0.975 0.939 0.735 0.869 1.000 0.930

Table 6

Efficiency study

Efficiency Q_dsc (kWh) Q_supply (kWh) Q_aux (kWh)

Humidifier 0.8 11,831 8439 486

0.9 11,072 8673 355

0.95 10,609 8726 300

Dehumidifier 0.7 11,256 8395 533

0.8 11,831 8439 486

0.9 12,334 8481 448

Heat recovery 0.7 10,368 7979 780

0.8 11,831 8439 486

0.9 12,888 8711 278

L. Mei et al. / Renewable Energy 31 (2006) 1265–1278 1277

the solar powered cooling energy. At least, 73.5% of the cooling load can be removed by

the desiccant cooling in August.

8. Discussion

The results of the studies described above indicate that the amount of the cooling

energy supplied to the room space is sensible to the system parameters, such as the

efficiencies of the sorption wheel, heat recovery wheel and humidifier. If the efficiencies

for the three main components varied between 0.7 and 0.9 (in the above simulation, the

efficiencies are fixed as 0.8), the cooling energy supplied to the room also has G5.4%

change. This results to the changes of the auxiliary cooling supply amount and the solar

fraction value. The cooling energy generated from the desiccant system, the cooling

energy supplied to the room and the auxiliary cooling energy are given in Table 6 with the

different efficiencies for the simulation of the typical month of July.

In this study, the electrical consume of three ventilation fans was considered. The total

electric consumption of fans was calculated for the different three ventilation flow volume

rates. 13,557 kWh of electric energy could be used for summer season operation for the

desiccant cooling. This electrical consume should be added to the auxiliary cooling and the

solar fraction of the system should be analysed carefully.

It is necessary to indicate that the solar powered desiccant system can be operated in

heating mode for winter season. When operating in heating mode, the combination of solar

energy and recovered heat from the exhaust air stream appears to be capable of one-third

of the total heating energy required during the winter months. It also can be seen that most

of the auxiliary heating energy supplied to the room is during the no-solar period of the

day-evening and night.

9. Conclusion

In summary, building integrated ventilated photovoltaic facade and PV sheds can

significantly contribute to the thermal energy needs of buildings in summer season.

L. Mei et al. / Renewable Energy 31 (2006) 1265–12781278

The typical application of the thermal energy transmitted from PV facade and PV sheds is

to use it to drive a solar powered desiccant cooling system for buildings cooling. In this

paper, the establishment of the component models of a desiccant cooling system

regenerated by solar heat was described. The simulation study of the system performance

was carried out. When simulating the system operation process including the control

strategies, the energy balances for the building cooling were yield. The useful solar

thermal energy observed by ventilated PV-solar heating system and the effective cooling

energy generated from the desiccant cooling system have been calculated. With respect to

the regeneration energy from the solar air system, an average COP of 0.518 was obtained.

The supplied cooling energy from the desiccant cooling system for summer season were

estimated as 43,580 kWh which can achieve a solar fraction of 0.75 at least.

The results of the studies described in this paper demonstrate the potential of desiccant

cooling systems combined with ventilated PV facade/sheds. The results also confirmed

that the ventilated PV facade/sheds can produce the same order of magnitude of thermal

energy as the photovoltaic electric production for buildings heating and cooling.

Therefore, the concept of building integrating ventilated PV facade/sheds has the good

potential in the future.

References

[1] Infield D, Design, Study and Experimental Evaluation of an Integrated Solar Facade. E U project report

published by Europen Commision. 2000.

[2] Mei L, Infield D, Eicker U, Fux V. Thermal modelling of a building with an integrated ventilated pv facade.

Energy Building 2003;35(6):605–17.

[3] Eicker U, Fux V, Infield D, Mei L. Thermal performance of building integrated ventilated PV facade.

Proceedings of ISES Solar World Congress, Israel 1999.

[4] Lloret A, Aceves O, Sabata L, Andreu J, Merten J, Chantant M, Eicker U. Lessons learned in the electrical

system design, installation and operation of the Mataro public library. European photovoltaic solar energy

conference, Barcelona 1997 [pp. 1695–1664].

[5] Duffie JA, Beckman WA. Solar engineering of thermal processes. New York: Wiley; 1980.