This thesis presents a data driven evaluation of photovoltaic/thermal systems in Sweden focusing on how regional and seasonal variations in solar irradiance, both under current and future climate scenarios could affect their technical and economical viability. Using machine learning models trained on daily meteorological data from 2005 to 2023, solar irradiance was forecasted across Sweden’s four electricity bidding zones (SE1–SE4) under Representative Concentration Pathways (RCP) 4.5 and 8.5. The predictions were used to simulate energy output and evaluate the financial performance of PV/T systems via discounted cash flow analysis and payback period estimation.

Results revealed a regional and seasonal disparities in solar potential, with southern zones showing slightly increased irradiance and shorter payback periods, while northern zones such as Luleå display minimal change or even declines under future scenarios. Seasonal shifts in irradiance were identified, especially reductions in winter and increases in summer. 

By combining predictive modeling with economic analysis this research provides insights into the long term feasibility of PV/T systems in Sweden and supports informed decision making in the context of an ever changing climate.