Analysis of Thermal Dynamics of Floating Photovoltaic Systems

Abstract

Floating photovoltaic systems (FPV) gain attention as a renewable energy generation source, especially in regions of limited land availability. As the technology emerges, accurate thermal modelling remains a key challenge due to complex coupling between irradiance and wind flow given varying floating structures. Since the spatial variability of wind dynamics across FPV systems remains poorly understood, this limits the accurate temperature and thus performance modelling. This study presents a scalable methodology that integrates computational fluid dynamics (CFD) derived wind decay functions into a dynamic heat transfer model to capture regional variations in module cooling. The approach is applied to a large FPV installation in the Netherlands, divided into multiple zones. The model incorporates convective, conductive, and radiative heat exchanges to calculate front, cell, and back temperatures. Validation against measured module temperatures over multiple seasons shows strong agreement, with normalised root mean square error (NRMSE) values between 6.5% to 12.5% under varying ambient conditions across daily and seasonal timescales. When compared with steady-state models such as Faiman and PVsyst, the proposed model demonstrates improved accuracy and adaptability to transient meteorological changes.