Dynamic Agrivoltaic Power Optimization Driven by Energy–Crop Synergies: A Mechanistic Hourly Simulator Model

Most agrivoltaic (APV) models assemble stand-alone tools ex post and optimize panel orientation for energy alone. It remains an open question whether the energy resource of an APV plot (its generation, storage, and shiftable demand) can instead be optimized at the resolution of the individual module while still guaranteeing that crop needs are met. This paper addresses that question with a mechanistic, hourly, spatially resolved framework in which the panel-orientation optimizer is embedded inside the hourly loop, ahead of the radiation sub-model, so every candidate orientation is evaluated through the full geometric chain it would produce rather than from cached values. Four operating modes (pure energy, energy under a crop-yield floor, pure yield, and yield under an energy floor) make the energy–yield trade-off an explicit decision variable that can be posed against a regulatory yield threshold. Over a full growing season (1 May–31 October 2021) for a Citrus limon orchard in Valencia, Spain (39.47391° N, 0.37966° W), the modes span 45.0–85.3 MWh of generation and 74.8–100.5% relative crop yield, with Land Equivalent Ratios of 1.37–1.48 and LCOE of 0.097–0.183 EUR/kWh.