The use of solar radiation data models is widespread in energy system analysis, however a gap exists when assessing their impact in modelling large-scale solar thermal systems integrated in district heating (DH) systems. Therefore, this study presents an analysis of how using satellite-based radiation data models (SARAH), reanalysis models (CFSR, ERA and MERRA2) and other data models (Danish Reference Year) affect the modelling of these systems. Taking three DH plants in Denmark as study cases, the measured radiation between 2016 and 2019 are utilized. Using energyPRO-based mathematical models of the systems, heat outputs are calculated and compared with measured data. Moreover, the yearly DH plant operational cost is calculated to observe the economic impact of using inaccurate models. It is found that heat production assessments based on the SARAH model show a better agreement with measured data than the reanalysis-based ERA5, MERRA2 and CFSR models. The empirically-based DRY shows low errors when observing its yearly values but has a higher inaccuracy on the hourly level, providing inaccurate operation profiles of the plant. Additionally, the satellite-based solar data model SARAH is further analyzed to identify patterns of its inaccuracy. After comparing it with 18 locations in Denmark using month-hourly profiles, no error trend can be identified, supporting the robustness of the model.
