https://www.mdu.se/

Agrivoltaic (APV) systems, where photovoltaic (PV) modules are co-located with crops, present a promising solution to land-use competition between food and energy conversion. Their successful deployment, however, requires a nuanced understanding of the trade-offs between electricity generation, crop yield, water use, and regulatory constraints. In this dissertation, a comprehensive modelling framework was developed to evaluate the technical, environmental, and economic performance of APV systems across diverse European contexts. The work builds on advances in irradiance transposition modelling, view-factor analysis, and crop yield estimation, which together enable more accurate characterisation of light distribution and photosynthetically active radiation (PAR) under different system designs. A state-of-the-art review of APV modelling approaches further situates these contributions within the wider field and identifies critical gaps in integrated assessment and validation. On this foundation, scenario-based optimisation was applied to explore system performance across varying climates, row pitches, heights, and orientations, under constraints reflecting national APV regulations. 

A novel combination of LER-based (land equivalent ratio, comparing the combined energy and crop productivity to their separate cultivation) metrics and economic indicators was used to characterise outcomes, with results evaluated over multiple weather years. The findings show that design modifications in geometric layout, particularly row pitch and orientation, can significantly shift the balance between LER_crop and LER_PV, affecting both productivity and profitability. Results further reveal that regulatory thresholds strongly influence feasibility: while strict national frameworks rendered most configurations unviable, relaxed rules enabled designs achieving LER values above 1.3 and net present values (NPVs) exceeding €2 million per hectare. If crop models were to incorporate adaptive mechanisms, even stricter regulatory scenarios could become feasible, with LER values above 1.6 in favourable contexts. Overall, this dissertation contributes transparent and generalisable modelling tools that link detailed irradiance and crop representation with system-level optimisation. The findings underscore the importance of integrating economic, agronomic, and policy perspectives, while pointing to future work on crop model realism, spectral and microclimate effects, and expanded crop databases to further enhance the robustness of APV deployment strategies.