Available online at www.sciencedirect.com

www.elsevier.com/locate/solener

ScienceDirect

Solar Energy 104 (2014) 82–92

A study of design options for a building integratedphotovoltaic/thermal (BIPV/T) system with glazed air collector

and multiple inlets

Tingting Yang ⇑, Andreas K. Athienitis

Dept. of Building, Civil and Environmental Engineering, Concordia University, 1455 De Maisonneuve Blvd. W., Montreal, Quebec H3G 1M8, Canada

Available online 14 March 2014

Abstract

In this paper, a prototype open loop air-based building integrated photovoltaic thermal BIPV/T system with a single inlet is studiedthrough a comprehensive series of experiments in a full scale solar simulator recently built at Concordia University. A numerical controlvolume model is developed and validated based on the results from the experiments. Improved designs of a BIPV/T system with multipleinlets and other means of heat transfer enhancement are studied through simulations. Simulation results indicate that the application oftwo inlets on a BIPV/T collector increases thermal efficiency by about 5% and increases electrical efficiency marginally. An added verticalglazed solar air collector improves the thermal efficiency by about 8%, and the improvement is more significant with wire mesh packing inthe collector by an increase of about 10%. The developed model is applied to a BIPV/T roof of an existing solar house with four sim-ulated inlets, and the thermal efficiency is improved by 7%.� 2014 Elsevier Ltd. All rights reserved.

Keywords: BIPV/T; Solar simulator; Multiple inlets; Solar air heater

1. Introduction

For a photovoltaic module, a portion of the incidentsolar energy is converted into useful electricity, while therest is either reflected or dissipated as heat. The photovol-taic module can be used as the absorber in a solar thermalcollector and such a device is also known as a photovoltaic/thermal (PV/T) collector. Photovoltaic panels can also beused as the building component provided their framingand attachment systems are modified for attachment asthe outer layer of facades or roofs. There are two waysof incorporating the PV into the building envelope–BAPV(building-added photovoltaic) and BIPV (building-inte-grated photovoltaic). In a BAPV system, the PV modulesare fixed onto the existing building envelope (such as PV

http://dx.doi.org/10.1016/j.solener.2014.01.049

0038-092X/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author.E-mail address: pvting_yang@163.com (T. Yang).

panels installed over an asphalt shingle roof). In a build-ing-integrated PV (BIPV) system, PV modules are part ofthe building envelope in a BIPV system – forming the outerlayer and also performing the function of cladding such asshedding water. In either BAPV or BIPV systems, the PVsneed to be cooled, otherwise they may overheat. Whenactive heat recovery is utilized with BIPV systems – eitherin a closed loop (like PV/T – with a liquid loop) or in anopen loop with forced air they are known as building-inte-grated photovoltaic/thermal (BIPV/T) systems. BIPV/Tsystems can be readily integrated with building envelopesand with HVAC systems (into which the recovered heator heated air can be transferred) while producing simulta-neously electricity and useful thermal energy. In additionto generating electricity and useful heat, BIPV/T systemsalso reduce the building heating loads and possibly thecooling loads compared to conventional building envelopeelements (Chow et al., 2007; Zogou and Stapountzis, 2011).

Nomenclature

cp specific heat of air, J/(kg K)dx length of a control volume, mDh hydraulic diameterF view factorG solar radiation, W/m2

h heat transfer coefficient, W/(m2 K)m mass flow rate of air, kg/sPelec electricity productionQair heat absorbed by airR thermal resistanceT temperature

Greek letters

a absorptance of PV module

e emissivityr Stefan-Boltzmann constantl viscocity

Subscripts

mix the layers of backing substrate, Tefzel, adhesiveand steel sheet in the PV module

insu insulationplate back surface of the PV moduletop heat transfer between air and the PV modulebot heat transfer between air and the insulation

T. Yang, A.K. Athienitis / Solar Energy 104 (2014) 82–92 83

The most common types of BIPV/T systems are air-based (normally open loop) or closed loop water-based.Water-based BIPV/T systems have been studied for regionssuch as Hefei (Ji et al., 2011) and Hong Kong, China(Chow et al., 2009). A dual-function BIPV/T collector (Jiet al., 2011) is sometimes used to provide passive spaceheating in winter, and water heating in warm seasonsthrough natural circulation. A variety of models have beendeveloped for BIPV/T and PV/T systems. Ji et al. (2011)developed a finite difference model considering characteris-tics of both the collector and the building. This model wasfurther validated with field experiments. Natural water cir-culation (thermosyphon) was found more efficient thanforced water circulation in terms of system thermal perfor-mance. The annual thermal and electrical efficiency reached37.5% and 9.39% respectively under the climate of HongKong. In addition, the BIPV/T system brought overall heattransmission down to 38% of that of the conventionalbuilding wall. Tripanagnostopoulos (2012) performed anextensive study on water-based PV/T systems and theirintegration with buildings.

Chen et al. (2010) designed and studied an air-basedopen-loop BIPV/T system that was thermally coupled witha ventilated concrete slab in a prefabricated, two-storeydetached low energy solar house in Quebec, Canada. Itwas found that a BIPV/T system can significantly lowerthe temperature of PV panels and showed great potentialin assisting space heating. Athienitis et al. (2011) developeda prototype BIPV/T system that was integrated withunglazed transpired collector (UTC). The value of energygenerated by the BIPV/T system was between 7% and17% higher than the UTC covering the same area, assum-ing that electricity is about four times more valuable thanheat. The concept of this prototype formed the foundationfor a full-scale demonstration project in Montreal (latitude

45N), in which crystalline silicon PV modules covered 70%of the UTC area. The BIPV/T system acts as the facade ofthe building, while also generating up to 25 kW electricityand 75 kW heat for preheating ventilation fresh air. Alsoin the cold climate of Hokkaido, Japan, Nagano et al.(2003) developed a wall-mounted BIPV/T system toreplace the flat roof incorporated system and reduce systemcost and roof leak because of snow. Their study involvedsix BIPV/T prototypes different from each other in termsof PV type (amorphous or polycrystalline silicon), PV pro-tecting material (glass or Teflon), with or without a coverglass in front of the PV. Chow et al. (2003) designed aBAPV system for a hypothetical (but typical) subtropicalhotel building in Macau, which was mounted onto thewest-facing fac�ade of the hotel. The PV modules were fixedat a gap distance of 250 mm from the building facade,allowing air movement in the gap driven by buoyancy aswell as wind-induced effects. The warm air can be collectedand pre-heat water in the hotel restaurant. ESP-r simula-tion results showed that this BAPV system could reducethe building’s cooling load compared to a BIPV systemthat had no air gap. Chow et al. (2007) studied a BIPVair system in Hong Kong, which was integrated to the win-dow of an office building. In addition to a conventionalglass assembly, a semi-transparent amorphous PV modulewas mounted as the outer layer. The bottom and top endswere left open so that the gap between the glass and PV wasnaturally ventilated. The screening effect of the PV and theair ventilating effect were shown to reduce solar heat gainof the office significantly, improve visual comfort and min-imize local thermal discomfort.

In solar thermal collecting systems, a variety of ways canbe employed to boost system performance – double chan-nels (Hegazy, 2000; Shahsavar and Ameri, 2010), fins(Kumar and Rosen, 2011; Moummi et al., 2004), slat

Fig. 1. Schematic of the two-inlet BIPV/T system connected in series withglazed air collector packed with wire mesh.

84 T. Yang, A.K. Athienitis / Solar Energy 104 (2014) 82–92

(Ammari, 2003; Ibrahim et al., 2009), packing materials(Sopian et al., 2009; Kolb et al., 1999), corrugated surfaces(Gao et al., 2007; El-Sebaii et al., 2011) and ribs(Promvonge et al., 2011). In addition to the heat transferenhancement measures mentioned above, new ways of per-formance enhancement in BIPV/T systems are emerging.Agrawal and Tiwari (2010) investigated different ways ofconnecting the air channel in a BIPV/T system and theireffects on energy and exergy production. Pantic et al.(2010) proposed three different open-loop air BIPV/T roofsystems. By comparing the system performance, they foundthat linking a short vertical solar air heater to the unglazedBIPV/T system results in a higher thermal energy produc-tion in winter. This paper aims to further investigate waysto enhance BIPV/T system performance, particularly forsloped roofs in a cold climate.

In open-loop air-cooled BIPV/T systems which ofteninvolve large-scale PV areas covering complete roof or fac�-ade surfaces, the temperature of PV arrays can rise to highvalues (exceeding 70 �C), resulting in a significant decreasein electrical efficiency and degradation of PV panels withtime. It is desirable to enhance heat removal from the PVpanels by using multiple inlets instead of a single inlet.The introduction of extra inlets breaks the exterior andinterior air boundary layers and increases the heat transfercoefficient. In winter, the air outlet temperature can be fur-ther increased by adding a vertical glazed air collector sec-tion packed with wire mesh. The vertical glazed sectiontakes advantage of the low winter sun altitude, resultingin significant air temperature increase.

This paper considers several designs as follows:

1. A prototype BIPV/T system consisting of a singleinlet studied previously outdoors by Candanedoet al. (2011) is studied through a comprehensive ser-ies of experiments in a full scale solar simulatorrecently built in Concordia University. This collec-tor simulates the BIPV/T system in the EcoTerrahouse (a low energy solar demonstration house inQuebec, Canada) (Chen et al., 2010) which has ametal roof with amorphous silicon photovoltaicpanels, with outdoor air passing through a cavityunder the metal layer, extracting heat from theBIPV/T top layer where the PV is adhered to themetal layer. The prototype is half the length ofthe EcoTerra roof (2.8 m long versus 5.6 m longfor EcoTerra) and can be tested at different anglesin the solar simulator. An explicit finite differencecontrol volume model is developed and validatedbased on the results from the solar simulator.

2. Improved designs with multiple inlets and other meansof heat transfer enhancement are studied through sim-ulations. In the design shown in Fig. 1, air is drawnfrom two inlets in the BIPV/T section connected inseries with a vertical glazed collector section packedwith wire mesh.

2. Experiments in a solar simulator

2.1. Experimental setup

The experiments with the amorphous silicon BIPV/Tsystem prototype are carried out in a recently built SolarSimulator and Environmental Chamber Laboratory(SSEC) at Concordia University in Montreal as shown inFig. 2(a). The lamp field consisting of eight special metalhalide (MHG) lamps with an artificial sky illuminates thetest area. To simulate natural sunlight, the lamps providea spectral distribution in accordance with relevant stan-dards EN 12975:2006 and ISO 9806-1:1994. The heat inputof the lamps reaches up to 27.6 kW, and the heat is takenaway by a cooling unit in the mechanical room next tothe test room. The lamps in the lamps field can be movedand continuously dimmed, allowing test areas of differentdimensions to be illuminated at various radiations levelsas uniform as possible. In the experiments conducted inthis paper, the homogeneity is controlled within ±3%.For this class B solar simulator, homogeneity of ±3%can be reached for a test surface that measures 2.0 m by2.4 m, at an irradiance level of approximately 1100 W/m2. The lamp field as well as the platform can be tilted atany angle between 0� and 90�, simulating the conditionsof different building envelop elements; Fig. 2(b) showsthe solar simulator testing the BIPV/T collector at a tiltangle of 45�. In order to remove the long-wave infraredirradiation emitted by the high-temperature lamps, an arti-ficial sky is set up between the lamps and the test area. Theartificial sky is formed by two panes of low-iron glass withanti-reflective coating, with cooled air passing in between.A pyranometer and an anemometer are mounted on anx–y scanner above the test area to measure the radiationand the wind speed, respectively.

Fig. 2. (a) BIPV/T air collector (in blue rectangle) tested horizontally in the solar simulator (in red rectangle) and air collector testing platform (in greenrectangle); (b) BIPV/T system tested at 45� slope; and (c) thermocouple positions in the BIPV/T collector (the red dots represent thermocouples attachedunder the PV cells and onto the insulation; the gray dots represent thermocouples in the BIPV/T channel). (For interpretation of the references to colour inthis figure legend, the reader is referred to the web version of this article.)

Fig. 3. Schematic of the solar simulator and the experimental setup of the BIPV/T system (the BIPV/T prototype is 2.89 m long by 0.39 m wide; the PV isattached to a metal roof layer and air is drawn under this 4 cm thick layer with a fan).

T. Yang, A.K. Athienitis / Solar Energy 104 (2014) 82–92 85

A total of forty-eight special limit T-type thermocouplesare distributed in the system. Eleven thermocouples areattached under the center of each PV cell; another elevenare installed onto the insulation surface (see Fig. 1 for gen-eral concept, but with a single inlet). Fourteen thermocou-ples measure the air temperature in the channel, andanother twelve are placed on top of the PV to measurethe temperatures just above the collector. Fig. 2(c) showsthe positions of the thermocouples. Artificial wind is blownparallel to the collector from the inlet side to the top.

A schematic of the solar simulator configuration and theexperimental setup of the BIPV/T system is presented inFig. 3.

2.2. Mathematical model (for BIPV/T with single inlet)

Flowing air in the BIPV/T system extracts heat fromboth the top and bottom surfaces of the cavity of theBIPV/T system. The total thermal energy extracted bythe air in each control volume is given by

86 T. Yang, A.K. Athienitis / Solar Energy 104 (2014) 82–92

Qair ¼ _mcp T out � T inð Þ ð1ÞThe bottom surface of the cavity receives radiative

energy from the top surface of the cavity. The radiativeheat gain by the bottom surface is discharged to air, assum-ing that heat conduction through the bottom is negligible.Thus, we have

F r T 4plate � T 4

insu

� �1

eplateþ 1

einsu� 1

¼ hbot T insu � T air

� �ð2Þ

Convective heat transfer from the top surface to air isgiven by

Qair �F r T 4

plate � T 4insu

� �1

eplateþ 1

einsu� 1

� wdx ¼ htop T plate � T air

� �� wdx

ð3ÞThe hot PV module releases heat into the ambient

through both radiative and convective heat transfer. Thecombined radiative and convective heat transfer coefficientwas decided by

hambient T pv � T wind

� �� wdx ¼ aG � wdx� P elec � Qair ð4Þ

Curve fitting was performed on the temperature read-ings of PV, air and insulation. Experimental data were fit-ted in an exponential formula:

T ðxÞ ¼ A 1� e�xB

� �þ C ð5Þ

where x denotes the distance away from the inlet.Fig. 4 is an example of the experimental measurement

and fitted curves when the incident solar radiation was1080 W/m2, the average wind speed was 1.6 m/s and theaverage air speed in the BIPV/T channel was set at0.26 m/s. Note that air temperature shows a sudden riseat the second node, which was caused by the entranceeffect.

Fig. 4. Experimental data and curve fits of the temperatures of PV,insulation and air in prototype with single inlet.

The heat transfer coefficients between air and the topand bottom surfaces of the BIPV/T channel tilted at 45�are presented in the form of local Nusselt numbers asfollows:

In the turbulent region

NutopðxÞ ¼ 8:188Re0:77Pr3:85e�x0:2

2:8Dh þ 0:061Re0:77Pr3:85 2300

< Re < 9500 ð6Þ

NubotðxÞ ¼ 4:02Re1:09Pr19:3e�x0:2

14Dh þ 0:005Re1:09Pr19:3 2300

< Re < 9500 ð7Þ

In the laminar region we have the following fittedequations

NutopðxÞ ¼ 0:6883Re0:7Pr0:8e�x0:3

6:45Dh þ 0:0124Re0:7Pr0:8 1190

< Re < 2300 ð8Þ

NubotðxÞ ¼ 50Re0:5Pr0:2e�x0:3

1:37Dh þ 0:428Re0:5Pr0:2 1190

< Re < 2300 ð9Þ

2.3. Experimental results

A comprehensive set of experiments have been con-ducted in the solar simulator. Different irradiance levelswere considered. The flow rate in the BIPV/T channel iscontrolled, with a Reynolds number ranging from 1200to 10,000, which was chosen based on previous experiencewith a BIPV/T system that the practical operating Rey-nolds number was under 10,000 (to keep friction losseslow). The artificial wind is parallel to the collector, with acontrolled speed ranging from 1.6 to 3.5 m/s. Fig. 5 com-pares the thermal efficiencies of the BIPV/T collector withdifferent wind speeds on top of the PV surface. Asexpected, with increasing wind speed, the thermal efficiencyis reduced.

Fig. 5. Thermal efficiencies of the BIPV/T system at different wind speeds,with a tilt angle of 45� and incident solar radiation of 1080 W/m2.

Fig. 6. (a) Photo of the steel bars placed underneath the PV module as structural support; and (b) dimensions of the BIPV/T channel (drawing not toscale).

T. Yang, A.K. Athienitis / Solar Energy 104 (2014) 82–92 87

In the original BIPV/T prototype, four steel bars (shownin Fig. 6(a)) were placed immediately under the amorphousPV module to support its weight. Later, these bars wereremoved from the collector. The impact of the steel supporton the thermal efficiency of the BIPV/T system is illustratedin Fig. 7. It can be seen that the steel bars act as fins, increas-ing the thermal efficiency of the BIPV/T system. The dimen-sions of the BIPV/T cavity are illustrated in Fig. 6(b).

3. Numerical modeling and verification

3.1. Mathematical model

The mathematical model for the BIPV/T system is basedon a simplified model of the amorphous siliconphotovoltaic module (other PV modules can be similarly

Fig. 7. Thermal efficiency of the BIPV/T system with and without thestructural support steel bars under the PV module, when the tilt angle is45�, the incident solar radiation is 1080 W/m2 and the average wind speedis 1.6 m/s.

modeled). As shown in Fig. 8(a), from top to bottom, thelayers of the amorphous photovoltaic module are Tefzel(encapsulation material), anti-reflective coating, silicon,backing substrate, Tefzel, adhesive and steel sheet. In a typ-ical control volume as shown in Fig. 8(b), the layers undersilicon are combined as one equivalent layer with no signif-icant thermal capacity.

The following assumptions have been applied to themodel:

� The system is in quasi steady state.� The bottom insulation and the side walls are

adiabatic.� Heat flow is assumed to be one-dimensional perpen-

dicular to the PV.

A set of energy balance equations are written for thecomponents in the BIPV/T system. For the temperaturenode on the outer surface of the PV module,

T pv � T top

RTefzel¼ hambient T top � T wind

� �ð10Þ

For the PV node in the middle of the PV module,

aG � wdx ¼ P elec þT pv � T top

RTefzel� wdxþ T pv � T plate

Rmix� wdx ð11Þ

For the temperature node on the bottom of the PVmodule,

T pv � T plate

Rmix¼ htop T plate � T air

� �þ

F r T 4plate � T 4

insu

� �1

eplateþ 1

einsu� 1

ð12Þ

For the temperature node on the insulation surface,

F r T 4plate � T 4

insu

� �1

eplateþ 1

einsu� 1

¼ hbot T insu � T air

� �ð13Þ

Fig. 8. (a) Composition of the amorphous PV module attached to steel roof layer; (b) temperature nodes of the BIPV/T system (white layer indicatesflowing air, Tplate is the temperature of the steel sheet).

88 T. Yang, A.K. Athienitis / Solar Energy 104 (2014) 82–92

For the air passing through the control volume,

_mcp T out � T inð Þ ¼ htop T plate � T air

� �� wdx

þ hbot T insu � T air

� �� wdx ð14Þ

The heat transfer process in the vertical solar air heaterpacked with wire mesh is shown in Fig. 9. Compared to theconventional solar air heater, the added wire mesh absorbsthe incident solar radiation and releases it to the passing airby convection. The following equations demonstrate thecalculation of heat transfer coefficient between air the wiremesh (Prasad et al., 2009). Similar approaches to studywire mesh packed solar air heater have been applied byMittal and Varshney (2006); Ho et al. (2013).

The porosity of the wire screen matrix is determined by

P ¼ 1� p2

nd2w

ptD1þ d2

w

p2t

� �1=2

ð15Þ

The effective heat transfer area between the wire meshand air is determined by

A ¼ 4Af Lð1� P Þdw

ð16Þ

The hydraulic radius for the packed bed duct is given by

rh ¼Pdw

4ð1� P Þ ð17Þ

Fig. 9. Cross-section view of the wire mesh packed solar air collector.

The Colburn factor is expressed by

J h ¼ 0:25631

nP

� �0:609 P t

dw

� �0:7954

Re�0:63p ð18Þ

The relation between the heat transfer coefficient andthe Stanton number is evaluated by

Stp ¼h

G0cpð19Þ

where the relative mass flow rate for a packed bed G0 isgiven by

G0 ¼_m

Af Pð20Þ

The relation between the Colburn factor and Stantonnumber is evaluated by

J h ¼ StpPr2=3 ð21ÞIn the vertical solar air heater, the convective heat trans-

fer coefficient between the plate and the air flow is decidedby Eqs. (22)–(24).

hc ¼ 1:86 � RePrDh

L

� �1=3

� lls

� �0:14

� kair

DhRe < 2300 ð22Þ

hc ¼ 0:116 � Re2=3 � 125� �

� Pr1=3 � 1þ Dh

L

� �2=3" #

lls

� �0:14

� kair

Dh2300 < Re < 6000 ð23Þ

hc ¼f8� Re� 1000ð Þ � Pr

1þ 12:7 �ffiffif8

q� Pr2=3 � 1� � 1þ Dh

L

� �2=3" #

kair

Dh6000

< Re < 106 ð24Þ

where the friction factor f is given by

f ¼ 0:790 ln Re� 1:64ð Þ�2 3000 < Re < 5� 106 ð25Þf ¼ 0:316Re�1=4 2300 < Re < 3000 ð26Þ

The electrical efficiency of PV is expressed in terms ofsolar cell temperature

gelectric ¼ 0:16 1� 0:0045 T pv;s � T ref

� � ð27Þ

where Tref is the reference temperature.

Fig. 10. Comparison between calculated and measured temperatures inthe BIPV/T system.

T. Yang, A.K. Athienitis / Solar Energy 104 (2014) 82–92 89

For laminar flow, the friction factor is calculated by

f ¼ 64

ReRe < 2300 ð28Þ

In the wire mesh packed duct, the friction factor is givenby

fp ¼ 3:57221

nP

� �1:0431 pt

dw

� �1:1507

Re�0:43p ð29Þ

The relation between pressure drop and friction factor is

Dp1 ¼ f � LDh� qV 2

2ð30Þ

For packed duct, the hydraulic radius rh is used as thehydraulic diameter, and velocity is defined as u = G0/q.

The entrance loss at the air inlet is given by

Dp2 ¼ 0:5 � qV 2

2ð31Þ

Total pressure drop is calculated by

DP ¼ Dp1 þ Dp2 ð32ÞThus, fan power can be calculated by

P p ¼_mDPqf

ð33Þ

Fig. 11. Comparison of the temperatures of the 1-inlet and 2-inlet BIPV/Tsystems.

3.2. Verification of the model

The indoor solar simulator facility enables accurate andrepeatable test conditions, allowing for the prototype beingtested under a stable environment close to room tempera-ture (at lower or higher temperatures the environmentalchamber needs to be used with a mobile solar simulator).A cooling unit maintains the ambient temperature accord-ing to the lab thermostat setting; the sunlight-simulatinglamps provide radiation close to the solar spectrum at astable; the fan creates different wind speeds parallel to thePV surface in the same direction as the flow in the cavity.At the same time, there are some special requirements.The intensity and location of each lamp (8 in total) needto be adjusted in order that the irradiance on the test sur-face is in an acceptable uniformity range. The uniformityachieved was 3%. The lab temperature was set around20 �C.

Numerical simulations were performed using the modeldescribed above and are compared against experimentallymeasured data from the simulator, as shown in Fig. 10.The operating conditions are: 1080 W/m2 of solar radia-tion, 1.6 m/s of average wind speed and 1.5 m/s of air speedin the BIPV/T cavity.

4. Results

A two-inlet BIPV/T system (without the vertical glazedcollector) was studied based on the validated model. The

inlet air flow of the second section is a mixture of the outletair of the first section and the ambient air. The local Nus-selt number for the second section is calculated assumingthat the boundary layer restarts at the second inlet.

By using the 2-inlet BIPV/T system, the thermal effi-ciency is increased by 5%. Although the electrical efficiencyincrease is marginal, it can be seen in Fig. 11 that the peakPV temperature in the 2-inlet system is lower than that inthe 1-inlet system, which means that PV degradation withhigh temperature is reduced in the 2-inlet system and byextension in a multi-inlet system. The addition of a verticalglazed solar air heater will increase the thermal output ofthe whole system, especially in low latitude areas wherethe winter sun is low.

The effect of using a glazed solar air heater is presentedin Fig. 12, which shows that using wire mesh in the glazedsolar air heater will increase thermal efficiency by 2%.However, the fan power consumption in the wire mesh

Fig. 12. (a) Thermal and electrical production in a one-inlet BIPV/T system with the glazed solar air collector with or without wire mesh; and (b) thermaland electrical production in a two-inlet BIPV/T system with the glazed solar air collector with or without wire mesh.

90 T. Yang, A.K. Athienitis / Solar Energy 104 (2014) 82–92

packed system increases noticeably when the air flow rate isincreased, resulting in a decrease in electrical efficiency.

The results also show that the vertical air collectorimproves the thermal efficiency compared to a PV coveredBIPV/T system – by about 8% with a smooth solar air col-lector and by about 10% with a wire mesh packed solar aircollector.

The thermal performance of the solar air collector canbe expressed in a Hottel-Whillier-Bliss form (Duffie andWilliam, 2006). By treating the BIPV/T system as a solarcollector, its heat removal factor can be expressed as

F R ¼Qair

Ac aG� P elec � UL T i � T að Þ½ � ð34Þ

Table 1Heat removal factors of the solar air heater with/without wire mesh.

Flow velocity (m/s) Without mesh (%) With mesh (%)

0.85 54.8 80.91.7 73.7 82.82.55 78.6 84.13.4 81.1 84.9

Fig. 13. Schematics of the original EcoTerra BIPV/T system with one air inleactual length here; the new simulated inlets are at equal distances).

Some typical heat removal factors are summarized inTable 1.

As a preliminary study, the concept of using multipleinlets is examined for the BIPV/T system installed on theEcoTerra house (Chen et al., 2010). Chen et al. (2010)reported that the typical efficiency of this BIPV/T systemwas about 20%, and the temperature of the PV panelspeaked at 65 �C with an outdoor temperature of about20 �C and low wind speed on a warm sunny day. The ori-ginal system measures 5.8 m and takes in air by one inlet(Fig. 13). The new design divides the whole length evenlyinto four parts (Fig. 13), assuming that flow rate througheach inlet equals. The temperature distributions of thePV and air flow are presented in Fig. 14. Depending oneach section, the local heat transfer coefficient betweenPV and the channel air varies approximately between 5and 50 W/(m2 K). The mixed heat transfer coefficient ofPV with wind (convection) and with surroundings isapproximately 20 W/(m2 K). It was found that the thermalefficiency is 27.1% for the four-inlet BIPV/T system.Experimental work is underway with a crystalline siliconbased BIPV/T system.

t (left) and the new BIPV/T system with four inlets (the roof length is put

Fig. 14. Temperature distributions of PV and air in the cavity along theflow direction.

T. Yang, A.K. Athienitis / Solar Energy 104 (2014) 82–92 91

5. Conclusion

A prototype BIPV/T system consisting of a single inletwas experimentally studied in a full scale solar simulator.The influence of wind speed on the thermal efficiency ofthe BIPV/T system was studied. As expected, with higherwind speed, the thermal efficiency of the BIPV/T systemis reduced significantly. Addition of structural support steelbars in the BIPV/T collector channel was shown to increasethe thermal efficiency of the BIPV/T system significantly byacting as fins to dissipate heat into the flowing air.

A control volume model for the BIPV/T system wasdeveloped and validated with experimental results. Thismodel was first used to study the performance of a BIPV/T system with two inlets. The two-inlet design increasesthermal efficiency by up to about 5% and increases electricalefficiency marginally. It was found out that by using twoinlets, the peak temperature of the PV module is reducedby about 1.5 �C for the small prototype length, and this willreduce the degradation of the PV module. For actual roofswhere the length is 5–6 m, the reduction of maximum PVtemperature is expected to be at least 5–10 �C dependingon flow rate and wind conditions.

A vertical glazed solar air collector was simulated con-nected to the end of the BIPV/T system, for the purposeof increasing thermal efficiency in winter when the solaraltitude is low. By adding wire mesh in this section, thermalperformance is further increased significantly, particularlyfor heating applications in the winter when the solaraltitude is lower, and the additional heat is needed. Theaddition of the solar air heater increases the system thermalefficiency by about 8% with a smooth air channel, and by10% with a wire mesh packed air cavity.

Acknowledgement

This project is supported by the NSERC Smart Net-Zero Energy Buildings Strategic Research Network.

References

Agrawal, Bassant, Tiwari, G.N., 2010. Optimizing the energy and exergyof building integrated photovoltaic thermal (BIPVT) systems undercold climatic conditions. Appl. Energy 87, 417–426.

Ammari, H.D., 2003. A mathematical model of thermal performance of asolar air heater with slats. Renewable Energy 28, 1597–1615.

Athienitis, Andreas K., Bambara, James, O’Neill, Brendan, Faille,Jonathan, 2011. A prototype photovoltaic/thermal system integratedwith transpired collector. Sol. Energy 85, 139–153.

Candanedo, Luis M., Athienitis, Andreas, Park, Kwang-Wook, 2011.Convective heat transfer coefficients in a building-integrated photo-voltaic/thermal system. J. Sol. Energy Eng. ASME 133.

Chen, Yuxiang, Athienitis, A.K., Galal, Khaled, 2010. Modeling, designand thermal performance of a BIPV/T system thermally coupled witha ventilated concrete slab in a low energy solar house: Part 1, BIPV/Tsystem and house energy concept. Sol. Energy 84, 1892–1907.

Chow, T.T., Hand, J.W., Strachan, P.A., 2003. Building-integratedphotovoltaic and thermal applications in a subtropical hotel building.Appl. Therm. Eng. 23, 2035–2049.

Chow, T.T., He, W., Ji, J., 2007. An experimental study of facade-integrated photovoltaic/water-heating system. Appl. Therm. Eng. 27,37–45.

Chow, T.T., Fong, K.F., He, W., Lin, Z., Chan, A.L.S., 2007. Perfor-mance evaluation of a PV ventilated window applying to officebuilding of Hong Kong. Energy Build. 39, 643–650.

Chow, T.T., Chan, A.L.S., Fong, K.F., Lin, Z., He, W., Ji, J., 2009.Annual performance of building-integrated photovoltaic/water-heat-ing system for warm climate application. Appl. Energy 86, 689–696.

Duffie, John A., William, A., 2006. Beckman. Solar Engineering ofThermal Processes, third ed. John Wiley and Sons.

El-Sebaii, A.A., Aboul-Enein, S., Ramadan, M.R.I., Shalaby, S.M.,Moharram, B.M., 2011. Investigation of thermal performance ofdouble pass flat and v-corrugated plate solar air heaters. Energy, 1076–1086.

Gao, Wenfeng, Lin, Wenxian, Liu, Tao, Xia, Chaofeng, 2007. Analyticaland experimental studies on the thermal performances of cross-corrugated and flat-plate solar air heaters. Appl. Energy 84, 425–441.

Hegazy, Adel A., 2000. Comparative study of the performances of fourphotovoltaic/thermal solar air collectors. Energy Convers. Manage.41, 861–881.

Ho, C.D., Lin, C.S., Chuang, Y.C., Chao, C.C., 2013. Performanceimprovement of wire mesh packed double-pass solar air heaters withexternal recycle. Renewable Energy 57, 479–489.

Ibrahim, Adnan, Jin, Goh Li, et al., 2009. Hybrid photovoltaic thermal(PV/T) air and water based solar collectors suitable for buildingintegrated applications. Am. J. Environ. Sci. 5, 618–624.

Ji, Jie, Luo, Chenglong, Chow, Tin-Tai, Sun, Wei, He, Wei, 2011. Thermalcharacteristics of a building-integrated dual-function solar collector inwater heating mode with natural circulation. Energy 36, 566–574.

Kolb, A., Winter, E.R.F., Viskanta, R., 1999. Experimental studies on asolar air collector with metal matrix absorber. Sol. Energy 65, 91–98.

Kumar, Rakesh, Rosen, Marc A., 2011. Performance evaluation of adouble pass PV/T solar air heater with and without fins. Appl. Therm.Eng. 31, 1402–1410.

Mittal, M.K., Varshney, L., 2006. Optimal thermohydraulicperformance of a wire mesh packed solar air heater. Sol. Energy 80,1112–1120.

Moummi, N., Youcef-Ali, S., Moummi, A., Desmons, J.Y., 2004. Energyanalysis of a solar air collector with rows of fins. Renewable Energy 29,2053–2064.

Nagano, K., Mochida, T., Shimakura, K., Murashita, K., 2003. Devel-opment of thermal-photovoltaic hybrid exterior wallboards incorpo-rating PV cells in and their winter performances. Sol. Energy Mater.Sol. Cells 77, 265–282.

Pantic, S., Candanedo, L., Athienitis, A.K., 2010. Modeling of energyperformance of a house with three configurations of building-integrated photovoltaic/thermal systems. Energy Build. 42, 1779–1789.

92 T. Yang, A.K. Athienitis / Solar Energy 104 (2014) 82–92

Prasad, B.S., Saini, S.J., Singh, M.K., 2009. Investigation of heat transferand friciton characteristics of packed bed solar air heater using wiremesh as packing material. Sol. Energy 83, 773–783.

Promvonge, P., Khanoknaiyakarn, C., Kwankaomeng, S., Thianpong, C.,2011. Thermal behaviour in solar air heater channel fitted withcombined rib and delta-winglet. Int. Commun. Heat Mass Transfer 38,749–756.

Shahsavar, A., Ameri, M., 2010. Experimental investigation and modelingof a direct-coupled PV/T air collector. Sol. Energy 84, 1938–1958.

Sopian, K., Alghoul, M.A., Alfegi, Ebrahim M., 2009. Evaluation ofthermal efficiency of double-pass solar collector with porous-nonpo-rous media. Renewable Energy 34, 640–645.

Tripanagnostopoulos, Y., 2012. Photovoltaic/thermal solar collectors. In:Sayigh, A. (Ed.), . In: Comprehensive Renewable Energy, 3. Elsevier,Oxford, pp. 255–300.

Zogou, Olympia, Stapountzis, Herricos, 2011. Energy analysis of animproved concept of integrated PV panels in an office building incentral Greece. Appl. Energy 88, 853–866.