In a new report, IEA PVPS evaluated electricity output and temperature for six BIPV installations across different climates and configurations
They found that existing simulation tools showed larger deviations for short-term power predictions
IEA PVPS says the study's findings on BIPV operating temperatures contributed to updates in IEC TS 63126:2025
Existing solar PV simulation tools and temperature models struggle to accurately predict the real-world performance of building-integrated photovoltaics (BIPV), according to a new IEA PVPS report.
The findings point to the need for simulation tools that better account for BIPV-specific conditions.
For the study, IEA PVPS researchers modelled and measured data from six BIPV installations covering different climates and configurations including roof-, façade- and window-integrated systems.
The report examines six BIPV installations: Varennes Library in Canada, DTU in Denmark, Fraunhofer ISE in Germany, CIEMAT in Spain, Asan Youth Library in South Korea, and BEST Lab in China. Each installation had an on-site or nearby weather station recording meteorological data at intervals of 1 to 5 minutes. The monitoring periods ranged from six months to two years.
The researchers then compared simulations from the System Advisor Model (SAM) with measured DC and AC power and energy data across the six installations. They also tested how accurately commonly used PV models could predict module temperatures.
According to the report, the findings show that existing PV modelling tools can provide reasonably accurate estimates of long-term electricity generation from BIPV systems, but predicting module temperatures and short-term performance remains challenging.
For instance, the stimulated DC generation for the Fraunhofer ISE system was 145.7 kWh/ m², slightly higher than 143.0 kWh/m² measured. The simulated AC generation was 132.9 kWh/m², slightly lower than 134.7 kWh/m² measured.
At the CIEMAT installation in Spain, the model estimated DC generation of 207.6 kWh/m², compared with 192.1 kWh/m² measured. However, simulated AC generation of 184.8 kWh/m² was almost identical to the measured 185.3 kWh/m².
These examples, says IEA PVPS, shows that the electricity estimates were generally close to actual generation, particularly at the AC level.
Nevertheless, when it comes to predicting short-term performance, the analysts found a considerable difference between simulated and measured power when it was assessed at individual time intervals rather than over a longer period.
At South Korea's Asan Youth Library, for instance, the difference between simulated and measured power output reached 26.4% for DC power and 32.4% for AC power, indicating limitations in predicting the system's short-term performance.
According to their assessment of the case studies, the researchers also found that the normal operating cell temperature (NOCT)-based model struggled to account for BIPV-specific conditions, such as limited ventilation, façade integration and non-standard mounting configurations. This contributed to inaccurate temperature and DC power estimates in some installations.
In certain cases, these errors were partly offset by differences in simulated and actual inverter conversion losses, resulting in closer AC energy estimates. However, they stress that this does not indicate a consistent bias in the inverter model.
To assess whether King’s model could address this challenge, the study evaluated its performance using different empirical coefficients to predict module back-surface temperatures. In the high-temperature assessment, only the insulated-back BIPV system at DTU exceeded the 70°C threshold in IEC TS 63126:2020.
The report findings showed that carefully selected generic coefficients performed as well as or better than system-specific ones. Analysts believe this suggests that extensive site-specific calibration may not be necessary for most BIPV applications.
The study recommended distinguishing between different BIPV configurations. IEA PVPS says these changes were incorporated into the updated IEC TS 63126:2025 standard.
“The findings highlight the need for improved representation of BIPV-specific boundary conditions in PV simulation tools and support ongoing efforts toward more differentiated modelling and standardisation approaches for BIPV systems,” concludes the report.
The IEA PVPS study titled Modelling and Simulation of Installed BIPV Systems—Analysis of Electricity Yield and Module Temperature 2026 is available for free download on its website.
