Data-driven Analysis of Crop Yield in a Controlled Agrivoltaic Installation

Agrivoltaics (APV) offer a promising solution to mitigate land-use competition between photovoltaic energy generation and agriculture, although the design requires a careful balancing of energy yield, crop productivity, regulatory constraints and economic profitability. This study presents a policy-aware, data-driven multi-objective optimization framework for APV system design, integrating experimentally validated crop yield responses, solar radiation and shading models with photovoltaic performance evaluation. A real APV micro-plant equipped with agronomic sensors was used to experimentally derive crop yield response curves, enabling the validation and specialization of relationships previously obtained from meta-analytical studies. These functions were embedded into an integrated model of APV to finally perform a multi-objective optimization simultaneously maximizing agricultural yield and electrical energy production under Italian regulatory constraints. The framework was applied to a 1 ha case study at the latitude of Naples (Italy), with 40% PV coverage and 2.5 m inter-row spacing. Panel tilt and installation height were considered as decision variables. The optimal configuration, identified using the utopia-point criterion, features a 50° tilt angle, achieving a 12.27% increase in total agricultural yield and producing 1009.6 MWh/y of electricity. An economic assessment results in a simple payback period of 11.8 years, reduced to 7.08 years when public incentives are considered.