A Multi‐Domain System‐Level Framework for Floating Photovoltaic Systems With Comparative Analysis of Thermal Configurations

A multi‐domain modeling framework for the performance simulation of floating photovoltaic (FPV) systems is presented. As an initial stage of this framework, the present study implements and couples only the thermal and electrical subsystems, evaluating four interchangeable configurations of varying fidelity; the implementation of the mechanical and hydrodynamic subsystems is deferred to future work. The coupled models were validated against meteorological and module temperature data from the Open Energy Data Initiative Orlando, Florida FPV dataset. Results demonstrate that increasing structural detail in the thermal model provides negligible improvement in temperature prediction. All configurations yielded closely grouped error metrics, with annual average root‐mean‐squared error across the four configurations ranging from 3.432 to 3.611 °C. The FDM–EF configuration was found to be the most accurate, albeit marginally. A persistent negative bias observed during low‐irradiance and nighttime periods is plausibly linked to the omission of the pontoon's thermal interaction. Empirical corrections and Monte Carlo view factor integration were tested to address this gap, and results suggest that explicit modeling of back‐of‐module thermal interactions is necessary to achieve consistently accurate model performance. Furthermore, computational cost was found to depend strongly on solver tolerance settings, varying by roughly 30‐fold (5.38–148 s) across the tested range.