*4.1. Validation of Energy Performance*

To validate the selection of WFG, the daily absorbed energy per unit of area was calculated using Equation (29). Figure 11 shows the performance of two WFG panels in winter in three orientations. As expected, the eastern and western panels' energy absorption was not relevant. Most of the time there were heat losses due to the little radiation and the low outdoor temperature. A different performance was shown in the southern panels, where the daily absorbed energy was 21.3 kWh in 7.8 m2, so the ratio of energy per area was 2.73 kWh/m2.

**Figure 11.** Energy absorption on eastern, southern, and western WFG facades. Sample winter day 8 January 2020.

Figure 12 shows the energy performance on a sample summer day (14 July 2020). The total absorbed energy was 30 KWh in two southern WFGs, 18.6 KWh in two western WFGs, and 15.9 KWh in two eastern WFGs. The energy-to-area ratio was 3.85 kWh/m<sup>2</sup> in the south, 2.38 kWh/m2 in the west, and 2.04 kWh/m<sup>2</sup> in the east.

**Figure 12.** Energy absorption on eastern, southern, and western WFG facades. Sample summer day 14 July 2020.

The goal of rejecting energy in the east and west was met. Despite the high solar radiation, the water heated up by 2 ◦C, and most of the infrared energy was rejected. On the south facade, the energy absorption was similar in winter and summer, and the water heated up around 3 ◦C. The reliability of the simulation tool was tested by developing real prototypes. Figure 13 illustrates the comparison between the results of the real data and the simulation data. The daily energy absorption on southern facades in winter and summer were taken from Figures 11 and 12 and compared with the simulated results from Figure 6.

**Figure 13.** Accumulated energy absorption on southern WFG facades. Sample days 14 July 2020 and 8 January 2020. Comparison between simulated and real data.

The Mean Error (*ME*), shown in Equation (24), is the difference between the measured value and simulation results. The total number of measurements was *n* = 2872.

$$ME = \frac{1}{n} \sum\_{i=1}^{n} |E\_{Si} - E\_{Ri}|\_{\prime} \tag{30}$$

where *ESi* is the simulated energy absorption, and *ERi* is the measured energy absorption. By computing *ME* on 14 July 2020 the value was 0.78. When it came to the energy absorption on 8 January 2020, the *ME* was 0.67. The reason for this might be the uncertainties about the inlet and indoor temperatures. Although the accumulated energy values were quite similar in the simulation and the real conditions, the intermediate values differed at some times of the day. The simulation tool could not work with variable inlet and indoor temperatures, which is the main goal for further research.
