Printer Friendly

Determination of solar heat gain coefficients for semitransparent photovoltaic windows: an experimental study.


Semitransparent photovoltaic (STPV) windows have great potential for integration in fenestration systems, adding the option of solar electricity production while still fulfilling daylighting needs. In commercial and high-rise residential buildings where trends in architecture already include large glazed facades and lighting loads constitute a significant portion of the overall energy consumption, the integration of this technology is promising.

A typical STPV glass consists of a photovoltaic (PV) layer laminated between a transparent frontsheet and backsheet (Figure 1). Depending on the PV technology used, the PV layer may be located between the encapsulation resin or monolithically deposited on the transparent conductive oxide frontsheet or backsheet. Various PV technologies are used for STPV glass applications, with the most common being opaque crystalline silicon (Si) PV cells arranged in a way that allows light to pass through the resulting space between the cells, and amorphous-microcrystalline silicon (aSi/[micro]cSi) "see-through" thin PV films. Other emerging technologies include fully transparent polymer-based PV (Li et al. 2012) and perovskites (Snaith 2013).

STPV glass can be integrated as the outer layer of insulated glazing units (referred to as STPV window herein). STPV windows can then be installed in new or retrofitted commercial and high-rise residential building facades and skylights. Their use has the potential to reduce building energy consumption through solar electricity generation and reduce solar gains by partial shading (Miyazaki et al. 2005; Ng et al. 2013) while still allowing daylight transmission and partial or full views to the outdoors (Markvart et al. 2012; Vartiainen 2001).

A deeper understanding of STPV technologies will allow the PV and window industry to provide the necessary materials and designs for high-performance windows. When solar radiation strikes a window surface, it is partly reflected ([rho]), partly transmitted ([tau]) and partly absorbed ([alpha]):

[rho]([lambda], [theta]) + [tau]([lambda], [theta]) + [alpha]([lambda], [theta]) = 1 (1)

where [lambda] is the wavelength and 0 is the solar angle of incidence. In the case of a STPV window, a fraction of the energy absorbed is transformed to solar electricity while the rest is transformed to thermal energy:

[alpha]([lambda], [theta]) = [N.sub.out] [alpha]([lambda], [theta]) + [] [alpha]([lambda], [theta]) + [[eta].sub.el] ([lambda], [theta]) (2)

where [N.sub.out] is the fraction of absorbed solar energy reemitted outwards, [] is the fraction reemitted inwards, and [[eta].sub.el] is the fraction of incident solar radiation that is absorbed and converted to solar electricity, also known as electrical conversion efficiency.

Commercially available STPV windows can convert between 5% and 20% of the incident solar radiation into electricity, while the portion of solar energy that is converted into heat (roughly 30% to 70%, depending on the optical and thermal properties and PV technology used) contributes to the increase in temperature of the PV cells. Electrical efficiency is dependent on the optical, thermal, and electrical characteristics of the STPV window, as well as the climatic conditions, air mass, solar spectrum, and solar angle of incidence (Duffie and Beckman 2006). STPV window efficiency generally decreases as the operating temperature increases, almost in a linear fashion of up to -0.55%/[degrees]C (0.306%/[degrees]F), depending on the STPV technology used (Athienitis and O'Brien 2015). Operating temperatures exceeding 75[degrees]C (167[degrees]F) have significant influence on the durability of the PV cells or films and window components such as spacers, sealants, and framing. The optical and thermal properties of the STPV window as well as the presence of low-emissivity (low-e) coatings or suspended films have a direct impact on operating temperatures, window electrical performance, and durability (Chow et al. 2010; Gaillard et al. 2014; Park et al. 2010).

Of particular importance is the solar heat gain coefficient (SHGC), which is the fraction of the solar radiation entering the space through the window assembly, consisting of the solar transmittance and the inward-flowing fraction of absorbed solar energy (ASHRAE 2013):

SHGC([lambda], [theta]) = [tau]([lambda], [theta]) + [] [alpha]([lambda], [theta]) (3)

The SHGC depends on the optical and thermal properties of the window assembly, climatic conditions, air mass, solar spectrum, and solar angle of incidence (Klems 2000; Kuhn et al. 2001). In a STPV window, the SHGC is also influenced by the electrical conversion efficiency of the STPV glass; the higher the efficiency, the lower the SHGC--more absorbed solar energy is transformed into electricity rather than into heat. Thus, a systematic study is required to experimentally determine and predict such interactions. It is essential to measure key properties (namely SHGC and U-factor) of advanced fenestration systems as they have a direct impact on building energy performance and occupancy comfort.

Considering the current advancements in the window industry, such as electrochromic windows, STPV windows, and windows incorporating angular selective coatings (Fernandes et al. 2015; Jelle et al. 2012), solar calorimetric standards should be updated to address current challenges and provide guidelines on how to test such technologies. The objective of this paper is to provide input on the development of an experimental procedure suited to determine the SHGC of STPV windows. Another important consideration in the performance of such windows is determining the temperature distribution so as to find ways to avoid excessively high temperatures for prolonged periods. Few studies have addressed the need for new testing standards for the determination of the SHGC, U-factor, and electrical performance of STPV systems (and building integrated photovoltaic technologies in general).

Fully controllable conditions, repeatability, and the ability to reach and maintain steady-state conditions are necessary to determine the SHGC and U-factor of fenestration systems. While the use of outdoor solar calorimeters provides the ability to test window technologies under realistic and dynamic climatic conditions, repeatability has proven to be challenging. Variation of the outdoor temperature, wind speed and direction, sky conditions, air mass, solar spectrum, and solar angle of incidence (Marinoski et al. 2012; Pereira and Sharples 1991) result in transient or quasi-steady test conditions. Olivieri et al. (2014a, 2014b) proposed a methodology for the optical electrical and thermal characterization of STPV windows using an outdoor calorimeter. In order to reduce measurement uncertainties due to dynamic conditions, a comparative analysis with a reference specimen of known properties was performed, tested side-by-side with the STPV specimen.

Solar tracking calorimeters have been used to achieve quasi-steady conditions by maintaining normal incidence. It has been shown that the tilt angle of the window might introduce uncertainties on the surface heat transfer coefficients and possibly distort the measurements (Harrison and Collins 1999; Tseng and Goswami 2001). The view factor of the window to ground also varies as the tilt angle of the calorimeter changes, impacting the radiative heat exchange between the window and skydome. Bloem et al. (2012) developed an outdoor calorimeter to determine the electrical and thermal output of STPV thermal windows. Free-rack mounted and rear-insulated reference modules were used for a comparative analysis.

On the other hand, there are efforts to develop an international standard (ISO/DIS 19467) for the determination of SHGCs of conventional and advanced fenestration systems such as STPV windows using a solar simulator (ISO 2015). Indoor calorimetry using a solar simulator provides the necessary control and test repeatability under steady-state or dynamic conditions. However, solar spectrum mismatch should be taken into account, while high collimation and uniformity are necessary for the test and characterization of STPV windows. Chen et al. (2012) introduced an indoor calorimetric facility for STPV window testing and determination of the SHGC and electrical performance. The facility used a continuous single-lamp solar simulator. In order to achieve high uniformity and light collimation under variable angle of incidence, the lamp was located 10 m (33 ft) away from the specimen, and a correction factor was applied to accommodate for the spectral mismatch.


The characterization and performance tests of the STPV windows were performed at the Solar Simulator and Environmental Chamber (SSEC) laboratory at Concordia University, Montreal, Quebec, Canada. The experimental study presented herein allows the determination of solar, thermal, and electrical performance of STPV windows. The experimental setup is summarized as follows.

* Solar Simulator: The Concordia University indoor solar simulator is a continuous lamp field that consists of eight metal halide (MHG) lamps emulating the sunlight and a test bench where the solar calorimeter is attached (Figure 2). The solar simulator is located in a space where the room temperature ([T.sub.amb]) is regulated at 21[degrees]C [+ or -] 1[degrees]C (69.8[degrees]F [+ or -] 1.8[degrees]F) and relative humidity of 30% [+ or -] 5%. It can be positioned from 0[degrees] (horizontal position to emulate a flat roof) to 90[degrees] (vertical position to emulate a vertical facade) and virtually to any tilt angle in between (e.g., at 30[degrees] to 45[degrees] to emulate the slope of a roof). For this study, all tests were performed in a vertical position and at normal incidence angle.

* Spectrum: The spectral quality of the lamps fulfills the specifications of ISO 9806 (ISO 2013), with approximately 80% of the emitted radiation lying in the range in which the incidence angle modifier varies by no more than 2%. An ultraviolet visible near-infrared (UV/Vis/ NIR) spectroradiometer with a spectral range of 200 to 2500 nm was used to measure the solar simulator spectrum. Figure 3 shows the normalized solar simulator spectrum in comparison to the air mass 1.5 reference spectrum (NFRC 2014a). A spectral mismatch correction factor is applied to accommodate for the difference between the solar simulator spectrum and the AM 1.5 reference spectrum (Chen et al. 2012; Harrison and Wonderen 1994).

* Irradiance Uniformity: The solar simulator irradiance intensity (S) can vary from 500 to 1200 W/[m.sup.2] with a uniformity of up to 97% (depending on the dimensions of the window) and a temporal stability of [+ or -] 1% during the testing period. A calibrated pyranometer (temperature-compensated detector) with a cosine response and a spectral range of 285 to 2800 nm is used to scan the window area on a scanning grid of a maximum spacing of 0.15 m (0.5 ft) (NFRC 2010). The spatial mean is deduced by a simple average. In addition, a calibrated mono-Si reference solar cell is also used to measure the irradiance intensity available to the poly-Si- based PV glass for electricity conversion (Dunn et al. 2012).

* Artificial Sky: An artificial sky apparatus, located in front of the lamps, maintains a surface temperature ([]) of 13[degrees]C [+ or -] 2[degrees]C (55.4[degrees]F [+ or -] 3.6[degrees]F). Its primary function is to remove the infrared radiation generated by the lamps while minimizing the influence of thermal irradiance from the adjacent surfaces to the window.

* Wind Effect: A linear, variable-speed fan is used to reproduce the wind-induced convection for still air to 14 m/s (45.9 ft/s). The fan blows ambient air parallel to the surface of the window. The wind speed is measured with a calibrated one-directional anemometer with a measurement range of up to 15 m/s (49.2 ft/s) and accuracy of [+ or -] 0.2 m/s (0.7 ft/s), on a scanning grid of a maximum spacing of 0.15 m (0.5 ft).

In addition, the convective heat transfer coefficient is measured directly with a hot-plate apparatus. The hot plate is heated with an integrated electric heater. The power output ([P.sub.pla]) of the heater is controlled with a proportional-integral-derivative (PID) controller to maintain constant plate surface temperature ([T.sub.pla]) under a given wind speed. The exterior convective film coefficient ([h.sub.o,meas]) is then calculated as follows:

[h.sub.o,meas] = [P.sub.pla]/([T.sub.pla] - [T.sub.amb]) x [A.sub.pla]

where [A.sub.pla] is the surface area of the plate. The exterior convective film coefficient is measured using a scanning grid of a spacing of 0.3 m (1 ft). Electroplated copper is chosen due to its smooth surface (similar to glass) and low emissivity. The emissivity of the plate ([[epsilon].sub.pla]) was measured at 0.030 [+ or -] 0.01 using an emissivity meter with a spectral range of 5 to 40 pm, resulting in an estimated radiative heat transfer coefficient of 0.020 W/([m.sup.2] x K) [3.52 x [10.sup.-3] Btu/(h x [ft.sup.2] x [degrees]F)]. Knowing the exterior convective heat transfer coefficient under the testing conditions allows for the correction of the measured thermal conductance of the STPV window to a value under a standard exterior convective film coefficient (NFRC 2014b) as follows:

[] = 1/1/[U.sub.meas] - 1/[h.sub.o,meas] + 1/[h.sub.o,st] (5)

where [] and [U.sub.meas] are the thermal conductances of the STPV window under standard conditions (e.g., NFRC 2014b) and test conditions, respectively, and [h.sub.o,st] and [h.sub.o,meas] are the exterior convective film coefficients under standard conditions and test conditions, respectively.

Solar Calorimeter

An indoor calibrated solar calorimeter apparatus is used to mount and test the STPV windows. The calorimeter was developed, characterized, and calibrated based on NFRC 201 (NFRC 2010). The dimensions of the calorimeter are 2.2 m long x 1.2 m wide x 0.2 m thick (283 x 63 x 8 in.), excluding the mask wall. The front (mask wall) surface of the calorimeter, where the test specimen is mounted, has dimensions of 2.6 x 1.6 x 0.06 m (102 x 63 x 2 in.), with a solar reflectance of 78%. It protrudes from the perimeter of the solar calorimeter in order to shade the rear surfaces (guard box) of the calorimeter and minimize the effects of direct solar irradiation.

The mask wall and guard box are insulated and sealed to minimize thermal losses to the ambient environment ([less than or equal to] 0.561 W/ [[m.sup.2] x K] [[less than or equal to] 0.01 Btu/[h x [ft.sup.2] x [degrees]F]]).

An absorber plate housed within the calorimeter is connected to a water-based closed loop. The closed loop is connected to a water-to-water heat exchanger capable of extracting heat in order to maintain the average absorber plate temperature ([T.sub.abs]) at desired levels. The calorimeter is attached at the test bench and placed under the solar simulator. Measurements are conducted on a tilt angle of 90[degrees] (vertical) when the apparatus reaches steady-state conditions (temperatures and emulated solar radiation levels are kept constant throughout the test period). The time constant of the calorimeter was experimentally determined at 18 min, based on NFRC 201 (NFRC 2010).

The Harrison and Dubrous method is used to determine the SHGC of the window tested (Harrison and Dubrous 1992; Harrison and Wonderen 1994). An energy balance performed on the STPV window (Figure 2c) shows that the net energy rate through the window into the calorimeter ([]) can be calculated as the sum of the solar heat gains resulting from exposure to the solar radiation ([Q.sub.sol]) and the net heat flow ([Q.sub.cond]) due to temperature gradient across the window ([DELTA][T.sub.STPV]).

[] = [Q.sub.sol] - [Q.sub.cond] = (SHGC x S x [A.sub.STPV]) - (U x [DELTA][T.sub.SPTV] * [A.sub.STPV]) (6)

where S is the solar irradiance incident on the window, [A.sub.STPV] is the surface area of the STPV window, and SHGC and U are the solar heat gain coefficient and the thermal conductance of the STPV window, respectively.

Treating the solar calorimeter apparatus as a solar thermal collector, the thermal performance factor ([[eta]]) of the window is then defined as:

[[eta]] = []/[A.sub.SPTV] x S (7)

The SHGC of the STPV window can be determined by substituting Equation 6 into Equation 7:

[[eta]] = -U x [DELTA][T.sub.STPV]/S + SHGC (8)

Through linear regression of the performance data, the SHGC is determined as the intercept of the regression line with the axis of ordinates. It should be noted that the U-factor is sensitive to window properties and assembly as well as indoor and outdoor conditions (e.g., wind speed, turbidity and direction as well as indoor and outdoor temperatures). While current solar calorimetric standards opt to derive the U-factor of the window tested using validated simulation tools ([]) in order to determine the SHGC value from Equation 8, the use of the Harrison and Dubrous method allows the measurement of the actual U-factor under the test conditions ([U.sub.meas]) thus reducing uncertainties when determining the SHGC value. When compared to the standard single-point measurement, the proposed method requires multiple point measurements (under various irradiance and temperature conditions), resulting in an increased overall test period.

The net energy flow through the STPV window into the enclosure is experimentally measured based on an energy balance on the calorimeter enclosure. The net energy flow through the window into the calorimeter is the sum of the energy rate extracted by the absorber plate ([Q.sub.abs]), the heat conducted through the mask ([Q.sub.mask]) and guard box ([Q.sub.guard]), and the heat lost due to air leakage ([Q.sub.lkg]):

[] = [Q.sub.abs] + [Q.sub.mask] + [Q.sub.guard] + [Q.sub.lkg] (9)

where [Q.sub.abs] is the product of the mass flow rate ([??]), the specific heat of the circulating water ([C.sub.P]), and the temperature rise between the inlet and outlet of the absorber plate (ATabs).

[Q.sub.abs] = [??] x [C.sub.P] x [DELTA][T.sub.abs] (10)

All front and back surface temperatures (namely those of the guard box, the mask, the absorber plate, and the STPV window tested) as well as air temperatures (ambient and inside the calorimeter) are measured using T-type thermocouples. In addition, during the assembly of each STPV window, a T-type thermocouple was installed in direct contact with the back surface of a PV cell (within the encapsulation resin). The inlet and outlet temperature of the absorber plate are measured using 1/10 DIN resistance temperature detectors, while the water flow is measured with an electromagnetic flow sensor with an accuracy of [+ or -] 0.5% of the measured value. Finally, an uncertainty analysis is contacted after each test (Harrison and Dubrous 1992).

Electronic Load and Current-Voltage Curve. The STPV windows or arrays of windows installed on a building facade are connected to a micro-inverter or central inverter, respectively. The inverter uses maximum power point tracking (MPPT) to extract maximum power from the STPV system and feed into the building or the grid (Deutsche Gesellschaft fur Sonnenenergie 2008). It is critical that during the testing period (prior to and during steady-state conditions) the STPV window tested is connected to an electronic load (or a resistor load) that functions as a current sink performing at the maximum power point ([]) to emulate realistic operation conditions. If there is no load (open circuit), the STPV glass will perform as heat-absorbing glass; the solar radiation absorbed by the PV cells or film is converted to heat only, increasing the STPV glass temperature. As the temperature increases, the SHGC rises due to increase of the inward-flowing fraction of absorbed solar energy (see Equation 3). Consequently, testing under an open circuit will produce a higher SHGC value than that observed under operating conditions with an applied load (Chen et al. 2012). An electronic load operating always under MPPT is connected to the window. A current-voltage curve is also taken to characterize the electrical performance of each specimen under test conditions.

Infrared (IR) Thermography. A thermal imaging camera is used to capture the temperature profile of the STPV window under steady-state conditions. The camera is calibrated based on the surface emissivity of the STPV window while the temperature profile is verified using five surface point temperature measurements conducted with T-type thermocouples. The IR thermography allows the detailed study of the STPV window temperature profile as well as the identification of any faults (e.g., air leakage, thermal bridging, and defective PV cells) through hot-spot detection.

Characterization of the STPV Glass Prototypes and Windows

Four STPV glass prototypes were assembled (Figure 2a). The prototypes use poly-Si PV cells arranged in such a way as to allow light to pass through the resulting space between the opaque cells. Each STPV glass assembly (Figure 1) consists of the following (outer to inner layer): (i) 3.2 mm (126 mil) tempered, antireflective-coated, white glass; (ii) ethylene-vinyl acetate (EVA) encapsulant layer; (iii) poly-Si spaced PV cells layer; (iv) EVA encapsulant layer; and (iv) polyvinyl fluoride (PVF) transparent backsheet. All prototypes are frameless, having dimensions of 1948 x 976 mm (76.69 x 38.43 in.). Various packing factors are used (packing factor f is the fraction of the glass area occupied by PV cells), resulting in various optical and electrical properties for each STPV prototype glass.

Solar-Optical Characterization. The multilayer optical properties of each STPV glass are measured based on NFRC 300 (NFRC 2014a) using a UV/Vis/NIR spectrophotometer with a spectral range of 200 to 2500 nm and equipped with an integrating sphere. Each STPV glass can be spatially separated into two parts: an opaque "PV cells" part and a "transparent encapsulant" part (Fung and Yang 2008; Zondag et al. 2002). Then, the spatially averaged "effective" optical properties of the STPV glass are determined:

[[tau].sub.STPV] = f x [[tau].sub.cell] + (1 - f) x [[tau].sub.enc] (11)

[[rho].sub.STPV] = f x [[rho].sub.cell] + (1 - f) x [[rho].sub.enc] (12)

where [[tau].sub.STPV] and [[rho].sub.STPV] are the total effective transmittance and reflectance (front or back) of the STPV glass, respectively; f is the packing factor of the STPV glass; [[tau].sub.cell] and [[rho].sub.cell] are the transmittance and reflectance of the "PV cell" part, respectively; and [[tau].sub.enc] and [[rho].sub.enc] are the transmittance and reflectance of the "encapsulant" part, respectively.

While it was found that Equations 11 and 12 can sufficiently express the spatially averaged "effective" optical properties of the STPV glass using opaque PV cells, they are not suitable for STPV glass that integrates transparent PV thin film, translucent glass, or translucent encapsulant. Instead, the ASHWAT method is recommended to calculate the effective solar-optical properties of the STPV glass (Wright et al. 2009).

The solar-optical properties are imported into Optics 6 (LBNL 2015) to calculate the effective spectral transmittance of each STPV glass prototype (Figure 4) summarized in Table 1.

Each STPV prototype glass is then integrated as the outer glass layer of a double-glazed STPV window. Each STPV window (Figure 1) consists of the following (outer to inner layer): (i) 6 mm (236 mil) STPV glass, (ii) 25.4 mm (1 in.) sealed air cavity, and (iii) 5.64 mm (222 mil) commercial clear glass with low-e coating ([[epsilon].sub.3] = 0.157). The thickness of the air cavity on the windows was not selected for optimal thermal performance but rather to accommodate the junction box located on the rear side of each STPV glass. The optical file of each STPV glass is imported into WINDOW 7.1 (LBNL 2014) to calculate the total optical properties of the STPV windows (Table 2).

Electrical Characterization. Table 3 summarizes the electrical properties measured under standard testing conditions of AM 1.5 global irradiance (ASTM 2012), 1000 W/[m.sup.2], and 25[degrees]C (77[degrees]F) PV cell temperature. These data are provided for reference purposes. However, STPV windows are unlikely to perform under these conditions. Thus, a current-voltage curve is measured under the various test conditions to study the electrical performance of STPV windows and ensure operation at maximum power point. Figure 5 illustrates the current-voltage curves for the four STPV windows. As the solar radiation incident on the STPV window increases, the short-circuit current and thus power output of the window increase almost in a linear fashion. At the same time, the PV temperature increases due to the increase on irradiance levels. In return, the diffusion current on the cells increases, leading to a reduction of the charges at the edges of the cells and reduction of the open-circuit voltage (Duffie and Beckman 2006).

Measurement of the SHGCs of STPV Windows

The SHGC value (and STPV window efficiency, profile temperatures, and U-factor) varies depending on the test conditions under which the STPV window is tested. The test conditions (Table 4) used at the SSEC laboratory are different than those in NFRC 200 (NFRC 2014c), resulting in different SHGC values. A simulation-based correction method was proposed to obtain SHGC values under a standard AM 1.5 spectrum (Chen et al. 2012; Van Wonderen 1996). Besides spectrum mismatch, the method can be extended to include measurement result correction due to variation of test conditions (namely outdoor/indoor temperatures, outdoor/indoor convective heat transfer coefficients, and irradiance intensities), thus producing SHGC values under NFRC 200 standard conditions and spectrum. An additional challenge is that existing simulation tools for the determination of the SHGCs and U-factors of window systems do not have the ability to simulate the solar electricity generation of STPV windows (Mitchell et al. 2013). Hence, the following method is proposed to measure the SHGC values of STPV windows (operating at maximum power point) under SSEC conditions and calculate the SHGC values under NFRC 200 standard conditions:

* Step 1: The solar-optical properties of the STPV window layers are measured based on NFRC 300 and imported into WINDOW 7.1 as presented previously. The software simulates the SHGC value of the STPV window assembly under NFRC 200 standard conditions ([SHGC.sub.oc,sim]).

* Step 2: The SHGC value of a STPV window under open circuit ([SHGC.sub.oc,meas]) is experimentally determined using the Harrison and Dubrous method.

* Step 3: The correction factor (c) from SSEC test conditions to NFRC 200 standard conditions is calculated as follows:

c = [SHGC.sub.oc_sim]/[SHGC.sub.co_meas] (13)

* Step 4: The SHGC value of a STPV window operating at the maximum power point ([SHGC.sub.mp_meas]) is experimentally determined using the Harrison and Dubrous method.

* Step 5: The SHGC value of a STPV window operating under maximum power point under NFRC 200 standard conditions ([SHGC.sub.mp_standard]) is then calculated as follows:

[SHGC.sub.mp_standard] = c x [SHGC.sub.oc_meas] (14)

The four STPV windows were tested and characterized following the aforementioned methodology. The temperature difference between the inlet and outlet of the absorber plate was kept at 2[degrees]C [+ or -] 0.5[degrees]C (3.6[degrees]F [+ or -] 0.9[degrees]F) to minimize the temperature differential on the STPV window, as it might affect the PV electricity generation (De Vries 1998; Tina et al. 2010). For all cases, the difference between simulated and measured SHGC values are within measurement uncertainty estimates (Table 5). Applying a correction factor, the SHGC values can then be calculated under NFRC 200 conditions for the various STPV windows operating under maximum power point conditions (Table 6).

The correction factor is strongly influenced by the spectral mismatch between the test spectrum and the standard AM 1.5 (ASTM 2012) due to the spectral response of the PV cells (Chen et al. 2012). As the area covered by PV cells increases, so does their spectral effect on the transmittance and solar heat gains (Gueymard and DuPont 2009). As a result, the correction factor is also increased. Finally, when the solar electricity generation of the STPV window is taken into account by operating the STPV window at the maximum power point rather than assuming open-circuit conditions, the SHGC is reduced between 2% (for a STPV window with visible transmittance of 40%, STPV_WIN40%) and 23% (for a STPV window with visible transmittance of 6%, STPV_WIN6%) (Figure 6). The reduction is expected to be even higher for a STPV window with no low-e coating due to significant increase of the inward-flowing absorbed solar energy.


The thermal performance of STPV windows is an area that needs urgent attention because it exerts a significant influence on the durability of the PV cells or films and other window components. Operating temperatures exceeding 75[degrees]C (167[degrees]F) have significant influence on the durability of the PV cells or films and other window components and need to be predicted either through testing or validated simulation tools. For this reason, the temperature profile of the STPV windows was measured under various irradiance intensities and wind conditions. The inner air and surface temperatures of the calorimeter were maintained at 21[degrees]C [+ or -] 1[degrees]C (69.8[degrees]F [+ or -] 1.8[degrees]F) to emulate indoor building conditions. An electronic load at maximum power point was always applied.

PV cell operating temperatures of 46.2[degrees]C to 55.3[degrees]C (115.2[degrees]F to 131.5[degrees]F) were observed for STPV_WIN40% and STPV_WIN6%, respectively, under 1000 W/[m.sup.2] and exterior convective film coefficient of 20 W/([m.sup.2] x K) (3.522 Btu/ [h x [ft.sup.2] x [degrees]F]). The average STPV window temperature is strongly affected by the solar absorbance of the STPV glass (outer layer). As the absorbance of the STPV glass increases (resulting in an increase of the electrical efficiency and reduction of the transmittance), only a small fraction (around 20% of the absorbed solar energy) is transformed to electricity, while the rest (around 80% of the absorbed solar energy) contributes to the increase of the PV cell operating temperature.

In addition, a temperature differential between the "PV cell" part and the "encapsulant" part of up to 13[degrees]C (23.4[degrees]F) was measured. This high differential temperature was observed on STPV_WIN40% and caused by the increased spacing between the opaque PV cells (a low packing factor of f = 0.46). This differential temperature is mainly driven by the variation of the solar absorbance between the "PV cell" part ([[alpha].sub.cell] = 0.991) and the "encapsulant" part ([[alpha].sub.enc] = 0.165).

However, as the spacing between the PV cells is reduced, the temperature differential is also reduced to less than 0.5[degrees]C (0.9[degrees]F) (for STPV_WIN6% the packing factor is f = 0.92). Such differential temperatures are specific to STPV windows using opaque PV cells (due to variation of optical properties between the "PV cells" part and the "encapsulant" part) and it has not been observed on STPV windows using thin film PV technologies with uniform optical properties throughout the window surface (Yoon et al. 2011).


An experimental study on the determination of solar heat gain coefficients (SHGCs) for semitransparent photovoltaic (STPV) windows was presented. An indoor solar simulator and solar calorimeter facility was used to test and characterize four STPV prototype windows. Currently, there is no commercially available simulation tool or standard test procedure able to estimate the SHGCs of STPV windows when operating at maximum power point conditions. A common practice in the building and window industry is to test or simulate STPV windows under open circuit, a condition that is witnessed only under fault operation of such a system. This study shows that when the solar electricity generation of the STPV window is taken into account by operating the STPV windows at the maximum power point rather than assuming open-circuit conditions, the SHGC is reduced between 2% (for a STPV window with visible transmittance of 40%, STPV_WIN40%) and 23% (for a STPV window with visible transmittance of 6%, STPV_WIN6%). The need to update the existing standards to provide guidelines on how to test and certify STPV window technologies is apparent. It was found that the electricity generation from the STPV windows can result in up to 23% reduction of SHGC in comparison to a heat-absorbing (e.g., tinted or fritted glass) window with the same optical and thermal properties.

During preliminary building design, simulating STPV glass as heat-absorbing glass will impact the final building energy performance and occupants' comfort, as it might result in significant overestimation of the solar heat gains and oversizing of the building cooling system.

In addition, STPV cell operating temperatures of up to 55.3[degrees]C (131.5[degrees]F) were observed with a temperature differential between the "PV cell" part and the "encapsulant" part of up to 13[degrees]C (23.4[degrees]F). This demonstrates the need to predict the operating temperatures of STPV windows through testing or simulations. The three-dimensional heat transfer phenomena that take place on a STPV window should be further studied and understood because they impact the durability of the PV cells and window components such as spacers, sealants, and framing. The development of experimentally validated heat transfer and electrical performance models for STPV windows is also required in order to provide the right tools to the building and window industry to predict the windows' overall performance and their impact on building energy and occupancy comfort. Such models can then be extended to windows integrating different PV technologies such as "see-though" thin films.


This work was funded by the Photovoltaic Innovation Network--a strategic Natural Sciences and Engineering Research Council of Canada (NSERC) research network. Additional support through an ASHRAE grant-in-aid is also acknowledged. Finally, the authors would like to thank Canadian Solar Inc. for the assembly of the STPV glass prototypes and Unicel Architectural Inc. for the assembly of the double-glazed STPV windows.

A                 = surface area, [m.sup.2] ([ft.sup.2])
c                 = adjusting factor
[C.sub.P]         = specific heat, J/(kg x K) (Btu/[lb x [degrees]F])
e                 = emissivity
f                 = packing factor
[h.sub.o]         = exterior convective film coefficient, W/([m.sup.2]
                    x K) (Btu/[h x [ft.sup.2] x [degrees]F])
[]        = closed-circuit current, A
[??]              = mass flow, kg/s (lb/s)
N                 = fraction of absorbed solar energy
P                 = power output, W
Q                 = energy flow, J (Btu)
S                 = irradiance intensity, W/[m.sup.2]
SHGC              = solar heat gain coefficient
T                 = temperature, K
U                 = thermal conductance, W/([m.sup.2] x K) (Btu/[h x
                    [ft.sup.2] x [degrees]F])
[V.sub.oc]        = open circuit voltage, V
[alpha]           = absorbance
[DELTA]T          = temperature difference, K
[eta]             = electrical efficiency; thermal performance factor
[theta]           = solar angle of incidence, [degrees]
[lambda]          = wavelength, nm
[[mu].sub.P,mp]   = temperature coefficient for maximum power point,
[rho]             = reflectance
[tau]             = transmittance


abs               = absorber plate
amb               = ambient
cond              = energy flow through the window (conductive)
el                = electrical
enc               = encapsulant part
guard             = energy flow through the guard box
in                = inward-flowing
lkg               = energy flow due to exfiltration
mask              = energy flow through the calorimeter mask
meas              = experimental measurements
mp                = at maximum power point
net               = net energy flow
oc                = open circuit
out               = outward-flowing
pla               = plate with electric heater
PV                = PV cells part
sim               = simulation
sky               = artificial sky
solar             = energy flow through the window (radiative)
st                = standard
STPV              = semitransparent photovoltaic
th                = thermal


ASHRAE. 2013. ASHRAE Handbook--Fundamentals. Atlanta: ASHRAE.

ASTM. 2012. ASTM G173-03(2012), Standard tables for reference solar spectral irradiances: Direct normal and hemispherical on 37[degrees] tilted surface. West Conshohocken, PA: ASTM International.

Athienitis, A.K., and W. O'Brien. (Eds.). 2015. Modelling, design, and optimization of net-zero energy buildings. NY: Wiley & Sons.

Bloem, J.J., C. Lodi, J. Cipriano, and D. Chemisana. 2012. An outdoor test reference environment for double skin applications of building integrated photovoltaic systems. Energy Build. 50:63-73.

Chen, F., S.K. Wittkopf, P. Khai Ng, and H. Du. 2012. Solar heat gain coefficient measurement of semi-transparent photovoltaic modules with indoor calorimetric hot box and solar simulator. Energy Build. 53:74-84.

Chow, T.T., C. Li, and Z. Lin. 2010. Innovative solar windows for cooling-demand climate. Sol. Energy Mater. Sol. Cells 94:212-20.

De Vries, D.W. 1998. Design of a photovoltaic/thermal combi-panel. Doctoral thesis, Eindhoven University, Netherlands.

Deutsche Gesellschaft fur Sonnenenergie. 2008. Planning and installing photovoltaic systems: A guide for installers, architects and engineers. London: Earthscan.

Duffie, J.A., and W.A. Beckman. 2006. Solar engineering of thermal processes, 3rd ed. Hoboken, NJ: Wiley & Sons.

Dunn, L., M. Gostein, and K. Emery. 2012. Comparison of pyranometers vs. PV reference cells for evaluation of PV array performance. Conference Record of the IEEE Photovoltaic Specialists Conference, pp. 2899-904.

Fernandes, L.L., E.S. Lee, A. McNeil, J.C. Jonsson. T. Nouidui, X. Pang, and S. Hoffmann. 2015. Angular selective window systems: Assessment of technical potential for energy savings. Energy Build. 90:188-206.

Fung, T.Y.Y., and H. Yang. 2008. Study on thermal performance of semi-transparent building-integrated photovoltaic glazings. Energy Build. 40:341-50.

Gaillard, L., C. Menezo, S. Giroux, H. Pabiou, and R. LeBerre. 2014. Experimental study of thermal response of PV modules integrated into naturally-ventilated double skin facades. Energy Procedia 48:1254-61.

Gueymard, C.A., and W.C. DuPont. 2009. Spectral effects on the transmittance, solar heat gain, and performance rating of glazing systems. Sol. Energy 83:940-53.

Harrison, S.J., and M.R. Collins. 1999. Queen's University solar calorimeter. Design, calibration and operating procedure. Presented at the Solar Energy Society of Canada Conference, Edmonton, AB.

Harrison, S.J., and F.M. Dubrous. 1992. Uncertainties in the evaluation of window SHGC and U-values measured using an indoor solar simulator facility. ASHRAE Trans. 98(1):638-45.

Harrison, S.J., and S.J. Van Wonderen. 1994. Determining window solar heat gain coefficient. ASHRAE Journal 36(8):26-32.

ISO. 2013. ISO 9806:2013, Solar energy-Solar thermal collectors--Test methods. Geneva: International Organization for Standardization.

ISO. 2015. ISO/DIS 19467, Thermal performance of windows and doors--Determination of solar heat gain coefficient using solar simulator. Geneva: International Organization for Standardization. iso/home/store/catalogue_tc/catalogue_detail.htm?csnumber=64989 (accessed 12.20.15).

Jelle, B.P.B.P., A. Hynd, A. Gustavsen, D. Arasteh, H. Goudey, and R. Hart. 2012. Fenestration of today and tomorrow: A state-of-the-art review and future research opportunities. Sol. Energy Mater. Sol. Cells 96:1-28.

Klems, J.H. 2000. Solar heat gain through fenestration systems containing shading: Procedures for estimating performace from minimal data. Berkeley, CA: Lawrence Berkeley National Laboratory.

Kuhn, T.E., C. Buhler, and W.J. Platzer. 2001. Evaluation of overheating protection with sun-shading systems. Sol. Energy 69:59-74.

LBNL. 2014. WINDOW 7.1. Berkeley, CA: Lawrence Berkeley National Laboratory. software/window/7/ (accessed 6.10.14).

LBNL. 2015. Optics 6. Berkeley, CA: Lawrence Berkeley National Laboratory. optics/optics.html (accessed 6.10.14).

Li, G., R. Zhu, and Y. Yang. 2012. Polymer solar cells. Nat. Photonics 6:153-61.

Marinoski, D.L., S. Guths, and R. Lamberts. 2012. Development of a calorimeter for determination of the solar factor of architectural glass and fenestrations. Build. Environ. 47:232-42.

Markvart, J., A. Iversen, A. Logadottir, and K. Johnsen. 2012. Indoor light and visual comfort with solar cells in glass facades. Copenhagen: Danish Building Research Institute, Aalborg University.

Mitchell, R., C. Kohler, D. Curcija, L. Zhu, S. Vidanovic, S. Czarnecki, and D. Arasteh. 2013. THERM 6.3/WINDOW 6.3 NFRC Simulation Manual. Berkeley, CA: Lawrence Berkeley National Laboratory.

Miyazaki, T., A. Akisawa, and T. KashiwagiT. 2005. Energy savings of office buildings by the use of semi-transparent solar cells for windows. Renew. Energy 30:281-304.

NFRC. 2010. NFRC 201, Procedure for interim standard test method for measuring the solar heat gain coefficient of fenestration systems using calorimetry hot box methods. Greenbelt, MD: National Fenestration Rating Council.

NFRC. 2014a. NFRC 300, Test method for determining the solar optical properties of glazing materials and systems. Greenbelt, MD: National Fenestration Rating Council.

NFRC. 2014b. NFRC 102, Procedure for measuring the steady-state thermal transmittance of fenestration systems. Greenbelt, MD: National Fenestration Rating Council.

NFRC. 2014c. NFRC 200, Procedure for determining fenestration product solar heat gain coefficient and visible transmittance at normal incidence. Greenbelt, MD: National Fenestration Rating Council.

Ng, P.K., N. Mithraratne, and H.W. Kua. 2013. Energy analysis of semi-transparent BIPV in Singapore buildings. Energy Build. 66:274-81.

Olivieri, L., E. Caamano-Martin, F. Olivieri, and J. Neila. 2014a. Integral energy performance characterization of semi-transparent photovoltaic elements for building integration under real operation conditions. Energy Build. 68:280-91.

Olivieri, L., E. Caamano-Martin, F.J. Moralejo-Vazquez, N. Martin-Chivelet, F. Olivieri, and F.J. Neila-Gonzalez. 2014b. Energy saving potential of semi-transparent photovoltaic elements for building integration. Energy 76:572-83.

Park, K.E., G.H. Kang, H.I. Kim, G.J. Yu, and J.T. Kim. 2010. Analysis of thermal and electrical performance of semi-transparent photovoltaic (PV) module. Energy 35:2681-87.

Pereira, F.O.R., and S. Sharples. 1991. The development of a device for measuring solar heat gain and shading coefficients of windows in scale models. Energy Build. 17:271-81.

Snaith, H.J. 2013. Perovskites?: The emergence of a new era for low cost, high efficiency solar cells. J. Phys. Chem. Lett 4:3623* 30.

Tina, G., F. Cosentino, and G. Notton. 2010. Effect of thermal gradient of electrical efficiency of hybrid PV/T. Presented at the 25th European Photovoltaic Solar Energy Conference and Exhibition/5th World Conference on Photovoltaic Energy Conversion, Valencia, Spain.

Tseng, C.C., and D.Y. Goswami. 2001. Effect of tilt angle and temperature difference on solar heat gain coefficient measurement of fenestration system. ASHRAE Trans. 107(1).

Van Wonderen, S.J. 1996. Solar heat gain performance evaluation of commercial solar-control glazings and shading devices. Master's thesis. Queen's University, Kingston, ON.

Vartiainen, E. 2001. Electricity benefits of day lighting and photovoltaics for various solar facade layouts in office buildings. Energy Build. 33:113-20.

Wright, J.L., M.R. Collins, N.A. Kotey, and C.S. Barnaby. 2009. Improving cooling load calculations for fenestration with shading devices. ASHRAE RP-1311 report. Atlanta: ASHRAE.

Yoon, S., J.H. Song, and J.S. Lee. 2011. Practical application of building integrated photovoltaic (BIPV) system using transparent amorphous silicon thin-film PV module. Sol. Energy 85:723-33.

Zondag, H.A., D.W. De Vries, W.G.J. Van Helden, R.J.C. Van Zolingen, and A.A. Van Steenhoven. 2002. The thermal and electrical yield of a PV-thermal collector. Sol. Energy 72:113-28.

Konstantinos Kapsis, PhD

Student Member ASHRAE

Andreas Athienitis, PhD, PEng


Stephen Harrison, PhD, PEng


Konstantinos Kapsis is a researcher and Andreas Athienitis is a professor in the Department of Building, Civil and Environmental Engineering, Concordia University, Montreal, QC. Stephen Harrison is a professor in the Department of Mechanical and Materials Engineering, Queen's University, Kingston, ON.

Caption: Figure 1 Schematic of a double-glazed STPV window integrating crystalline Si STPV glass as the outer layer.

Caption: Figure 2 (a) Indoor calibrated solar calorimeter and the four STPV glass prototypes, (b) the Concordia University solar simulator laboratory, and (c) schematic of the experimental setup.

Caption: Figure 3 Comparison of normalized solar simulator spectrum (including the artificial sky) and AM 1.5 reference spectrum.

Caption: Figure 4 Effective spectral transmittance of each STPV glass prototype.

Caption: Figure 5 Current-voltage curves for the four STPV windows under indoor/outdoor temperatures of 21[degrees]C [+ or -] 1[degrees]C (69.8[degrees]F [+ or -] 1.8[degrees]F), exterior convective film coefficient of 20 W/([m.sup.2] x K) (3.522 Btu/[h x [ft.sup.2] x [degrees]F]), and various irradiance intensities (from 838 to 1031 W/[m.sup.2]).

Caption: Figure 6 SHGC values for the four STPV windows for maximum power point and open-circuit operation under NFRC 200 (NFRC 2014c) environmental conditions.
Table 1. Optical and Thermal Properties of the Four STPV Glass

                     Optical Properties

Name                           Solar
of Glass
           [[tau].sub.front]   [[rho]].sub.front]

STPV7%           0.066               0.092
STPV21%          0.195               0.088
STPV34%          0.324               0.084
STPV48%          0.453               0.080

                    Optical Properties

Name       Visible
of Glass
           [[tau].sub.front]   [[rho]].sub.front]

STPV7%           0.070               0.059
STPV21%          0.206               0.062
STPV34%          0.342               0.064
STPV48%          0.479               0.066

                 Thermal Properties

Name                                           U-factor,
of Glass   [e.sub.front]   [e.sub.back]    W/([m.sup.2] x K)
                                          (Btu/[h x [ft.sup.2]
                                             x [degrees]F])

STPV7%         0.920          0.950              6.111
STPV21%                                         (1.046)

Notes: [tau] = transmittance, [rho] = reflectance, e = emissivity.
The STPV glass prototypes are named based on their (front) visible

Table 2. Optical and Thermal Properties of the Four Corresponding
STPV Windows

                         Optical Properties

Name of                         Solar
              [[tau].sub.front]   [[rho].sub.front]

STPV_WIN6%          0.046               0.198
STPV_WIN17%         0.135               0.181
STPV_WIN29%         0.223               0.168
STPV_WIN40%         0.312               0.158

                         Optical Properties

Name of                        Visible
              [[tau].sub.front]   [[rho].sub.front]

STPV_WIN6%          0.058               0.205
STPV_WIN17%         0.172               0.188
STPV_WIN29%         0.285               0.176
STPV_WIN40%         0.398               0.167

                   Thermal Properties             U-factor,
Name of       [e.sub.front]   [e.sub.back]   (Btu/[h x [ft.sup.2]
Window                                          x [degrees]F])

STPV_WIN6%        0.920          0.840
STPV_WIN17%                                         2.011
STPV_WIN29%                                        (0.354)

              STPV Glass
               Used as
Name of       the Outer
Window          Layer

STPV_WIN6%      STPV7%
STPV_WIN17%    STPV21%
STPV_WIN29%    STPV34%
STPV_WIN40%    STPV48%

Notes: [tau] = transmittance, [rho] = reflectance, e = emissivity.
The STPV windows are named based on their (front) visible

Table 3. Electrical Properties of the Four STPV Windows under
Standard Testing Conditions

Name of       Number of      Cell        Electrical
Window        PV Cells    Technology   Efficiency, hmp

STPV_WIN6%       72        Poly-Si          0.15
STPV_WIN17%      60                         0.13
STPV_WIN29%      48                         0.10
STPV_WIN40%      36                         0.07

Name of          Nominal       Open-Circuit    Short-Circuit
Window        Maximum Power,     Voltage,        Current,
              [], W    [V.sub.oc], V   [], A

STPV_WIN6%        294.10           45.2            8.56
STPV_WIN17%       240.40           37.61           8.52
STPV_WIN29%       187.90           29.98           8.57
STPV_WIN40%       133.30           22.28           8.48

Name of           Temperature
Window           Coefficient,
              Pp,mp, %/[degrees]C
STPV_WIN17%          -0.43
STPV_WIN29%        (-0.774)

Note: The STPV windows are named based on their (front) visible

Table 4. Environmental Conditions for the Determination of SHGC
Values through Simulation and Experiment

Parameter                   NFRC 200 Standard Conditions
                            (Simulation Using WINDOW7.1)

Irradiance intensity,                    783
Solar spectrum                         AM 1.5
Interior air temperature,             24 (75.2)
  [degrees]C ([degrees]F)
Exterior air temperature,             32 (89.6)
  [degrees]C ([degrees]F)
Wind speed, m/s (ft/s)               5.5 (18.05)

Parameter                        SSEC Conditions
                            (Initial Test Conditions)

Irradiance intensity,                  838
Solar spectrum                   SSEC lampfield
Interior air temperature,          21 (69.80)
  [degrees]C ([degrees]F)
Exterior air temperature,          21 (69.80)
  [degrees]C ([degrees]F)
Wind speed, m/s (ft/s)              5 (16.40)

Table 5. SHGC Values of the Four STPV Windows under Open Circuit,
Simulated under NFRC 200 Standard Conditions ([SHGC.sub.oc_sim]),
and Measured under SSEC Conditions ([SHGC.sub.oc_meas])

Name of       [SHGC.sub.oc_sim]    [SHGC.sub.oc_meas]
Window         (Open Circuit)        (Open Circuit)
                  Simulated             Measured

STPV_WIN6%          0.145         0.125 [+ or -] 0.022
STPV_WIN17%         0.242         0.225 [+ or -] 0.023
STPV_WIN29%         0.339         0.331 [+ or -] 0.028
STPV_WIN40%         0.436         0.438 [+ or -] 0.031

Name of           Difference        Correction
Window        [SHGC.sub.oc_sim] -    Factor,
              [SHGC.sub.oc_meas]        c

STPV_WIN6%           0.020            1.160
STPV_WIN17%          0.017            1.076
STPV_WIN29%          0.008            1.024
STPV_WIN40%         -0.002            0.995

Table 6. SHGC Values of the Four STPV Windows under Maximum Power
Point Measured under SSEC Conditions ([SHGC.sub.mp_meas]) and
Calculated under NFRC 200 Standard Conditions ([SHGC.sub.mp_sim])

Name of        [SHGC.sub.mp_meas]     [SHGC.sub.mp_sim]
Window           (Under Maximum         (Under Maximum
                  Power Point)           Power Point)
                    Measured              Calculated

STPV_WIN6%    0.096 [+ or -] 0.017   0.111 [+ or -] 0.018
STPV_WIN17%   0.203 [+ or -] 0.022   0.218 [+ or -] 0.022
STPV_WIN29%   0.314 [+ or -] 0.027   0.322 [+ or -] 0.027
STPV_WIN40%   0.428 [+ or -] 0.030   0.426 [+ or -] 0.030
COPYRIGHT 2017 American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2017 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Kapsis, Konstantinos; Athienitis, Andreas; Harrison, Stephen
Publication:ASHRAE Transactions
Date:Jan 1, 2017
Previous Article:Determination of heat transfer coefficient between heated floor and space using the principles of ansi/ashrae standard 138 test chamber.
Next Article:Development of a heat transfer model for ceiling radiant cooling panels through a combined experimental and simulation study.

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters