Electro-thermal energy yield simulations for bifacial all-perovskite tandem modules

The energy yield of bifacial all-perovskite tandem modules is quantified for a selected outdoor testing location using electro-thermal simulations on cell and module levels combined with a model providing spectral irradiance according to geographical location and time. Cell level modelling is performed using an opto-electronic drift-diffusion-Poisson simulation on the full tandem device under consideration of direct and diffuse front illumination according to the sun position and atmosphere model, and diffuse rear illumination according to the ground albedo. On module level we combine a two-dimensional finite-element large area simulation of the electrodes in an optimized layout for a 100 cm2 monolithically interconnected module with the one-dimensional active area coupling law obtained from the simulated cell level characteristics. Thereby, thermal effects are considered regarding heat generation by light absorption, charge carrier transport and recombination, as well as heat transport through the stack and dissipation at the surface. To reduce the computational cost, operation conditions are binned with respect to limiting photocurrent, ambient temperature, and wind speed. After verification of the validity of the binning approach by comparison with full time-data modelling, it is used to compute the annual energy yield on cell level as a function of ground albedo and top cell band gap, confirming both, a large gain from bifaciality already at moderate albedo and the benefit from lower band gap top cells for high rear irradiation level due to large ground reflectivity. Finally, thermal effects and configuration-dependent cell-to-module losses are quantified via the evaluation of annual energy yield with the full electro-thermal module simulation, using the binned cell-level characteristics as well as the measured ambient temperature and wind speed as input. The results imply that an accurate assessment of upscaling losses is more critical for a reliable quantification of energy yield than consideration of the full temperature dependence.