A simplified multi-physics approach for bifacial photovoltaic modules: Theory and validation of peculiar module layout

This work aims to develop and integrate three sub-models into a simplified multi-physics tool for simulating bifacial PV (bPV) devices. While similar tools exist, they often rely on complex modeling. In contrast, this study investigates a simpler approach that achieves comparable accuracy. The proposed models are also experimentally validated under a specific case study: a Vertical Bifacial PV (VBPV) installation. This setup is relatively novel and provides valuable insights into the feasibility of VBPV systems for agricultural and space-constrained applications, highlighting the strong dependence between environmental conditions and PV module performance. For the optical model, a 2D View Factor method is implemented, demonstrating high sensitivity to the module's surroundings. Results show that this simplified approach can achieve errors below 5%. The electrical modeling is the core of this study. Two parameter estimation methods are applied: a traditional experimental data-fitting approach and a data-driven stochastic method based on Particle Filtering. The latter is an innovative technique for this type of estimation. Three different electrical models for bPV are numerically solved and compared, showing good accuracy, with errors below 4%. Notably, a newly proposed circuit model outperforms the other two. A simplified 0-D lumped thermal model is developed and validated to complete the multi-physics framework, showing deviations of up to 7% in temperature estimation. The integration of the best-performing electrical model with the optical and thermal sub-models results in a comprehensive tool capable of estimating power and energy with errors of 5% and 2%, respectively. These findings demonstrate that a simplified approach could support the estimation of PV performance based on field measurements and weather data for VBPV installations.