The active thermal absorbers for water carried heat serves three purposes; delivering heat for domestic hot water and possibly also for space heating, reducing the heat load in the interior during summer and cooling the photovoltaic cells in order to increase the electrical efficiency. The full scale window prototype was used in indoor and outdoor measurements for determination of the incidence angle dependency and
the U value of the thermal collectors. Based on these results the annual energy yield has been derived.

To estimate a U value for the prototype of the solar wall operating as a solar collector, measurements of the heat loss from the collector have been performed in a dark surrounding at different temperatures of the inlet water. The values of U0 and Ui were estimated to 4.0 W/m2K and 0.046 W/m2K2 and the resulting collector U value as a function of AT is shown in figure 7.

At AT = 30 K, the U value is 5.4 W/m2K per glazed wall area. During the measurements, the prototype was surrounded on both sides with the same temperature. In a building, however, the back side of the window will usually be surrounded with air of room temperature, which will suppress the heat losses. Approximately 10 % of the total heat losses, are estimated as border losses.

The indoor measurements have been performed by using a large solar simulator providing nearly parallel light and adjustable for solar altitude angles (Hakansson H. (2003 a and b). As described in Hakansson (2003 c) and Gajbert et al. (2004), the simulator provides relatively good parallel quality of light, though this has been achieved somewhat at the expense of the light distribution over the area. As different solar altitudes are simulated, there is a tendency of varying irradiation over the test area as the uneven light pattern moves. To continuously measure the variation of the total irradiation on the test area, a number of parallel connected photodiodes were evenly spread out over the front glass of the prototype, giving a current corresponding to the received total irradiation (Gajbert et al, 2004).

By indoor measurements of the thermal efficiency, the incidence angle dependence was derived. The prototype of the solar wall was placed perpendicular to the solar simulator and the simulator was raised in order to simulate every tenth degree of solar altitude angle, i. e. the incidence angle on the glazing projected in the transversal plane, 0T. The optical efficiency, i. e. the zero-loss efficiency, of the prototype was calculated for different solar altitude angles, giving the incidence angle dependence for transversal angles, qT (QT ). The result was normalized by an outdoor measurement where the absolute value of the optical efficiency, at 30° incidence angle was registered.

The measured zero-loss efficiency has been divided by f(Qi), the transmission of the glazing, as described in equation [3], resulting in a graph angular dependence of the reflector only, R(0r). These functions, R(0r) and f(Q), shown in figure 6, were used in MINSUN to simulate the annual thermal energy yield. The measured values of optical efficiency at high incidence angles are less reliable, due to the non-uniform light distribution on the small projected area. Therefore, for angles higher than 50°, the same theoretical values as were used for the electricity calculations, shown in figure 6, have been used. The difference between the theoretical graph of optical efficiency, calculated by ray tracing, and the calculated values is due to the reflections on the absorber, the collector efficiency, the multiple reflexes on the reflector at higher incidence angles, and on the unevenness of the reflector. Since the photovoltaic cell situated in front of the thermal absorber has an efficiency of 15%, only 85% of the in MINSUN simulated irradiation falling onto the absorber is taken into count for thermal energy yield.

The result of the simulations shows that the annual thermal energy provided by the Solar Window is 103 kWh/m2 glazed surface, calculated at an operating temperature of 50°C. For operating temperatures of 40°C and 25°C the yield would be 155 and 250 kWh/a, m2 respectively. However, as previously discussed, the real heat losses would probably be lower, because of the higher indoor temperature, thus implying higher yields.