Optimal packing and planning for large-scale distributed rooftop photovoltaic systems under complex shading effects and rooftop availabilities

Abstract

Rooftop photovoltaics (PVs) are considered a promising solution to alleviating current cities’ escalating energy usage and carbon emissions. In high-density cities, complex shading effects and rooftop availabilities (caused by diversified rooftop obstacles and irregular rooftop outlines) jointly make planning of large-scale distributed rooftop PV systems critically challenging. This study proposed an optimal packing and planning method for large-scale distributed rooftop PV systems under complex shading and rooftop availabilities, tackling the challenges by decoupling optimal packing and planning into two-step optimization. Utilizing horizontal-level genetic algorithm, the method first optimized PV-panels packing on irregular-shaped rooftops to maximize area utilization. Second, adopting sequential integer linear programming, the method optimized the planning of individual rooftop packing levels to minimize levelized cost of electricity (LCOE). Based on a 139-rooftop Hong Kong case study, the method was verified against 1 billion Monte-Carlo solutions, which reduced LCOE by 48.0% at most and achieved the lowest LCOE of 0.365 HKD/kWh. Further analysis showed that the proposed method outperformed a rule-based planning method because of its better utilization of high solar-energy-intensity areas, reducing the LCOE by 15.4%. In practice, the method can be used to facilitate deployment of large-scale distributed rooftop PV, enhancing overall system cost-effectiveness and city decarbonization.