A holistic framework to optimize embedding PV systems into building façades

In addressing fossil fuel supply concerns and their environmental impacts, the building sector, as a major energy consumer, offers an opportunity for renewable energy integration. Among renewable energy sources, solar energy through photovoltaic (PV) panels on building façades stands out as a notable option, though fully realizing their potential remains a challenge. This study introduces a framework for the automated design of PV panels integrated into the façades of existing buildings, enabling thorough assessment based on energy efficiency, economic feasibility, and environmental impact. The process involves capturing the geometry of building envelopes, a deep learning model to identify façade surfaces for PV installation, and simulations to model PV generation and energy demand. An evolutionary multi-objective optimization algorithm is then employed to determine the PV system design parameters. The results of applying the framework to two university buildings in Alberta, Canada, are presented. For these cases, when equal weights are given to economic, environmental, and energy efficiency objectives, the optimal PV layout can achieve electricity self-sufficiency of 5.16 % and 6.78 %, with greenhouse gas emission rates of 18.26 and 15.69 g CO2-eq./kWh, respectively. The analysis illustrates that adjusting objective priorities yields different optimized solutions to balance competing factors. For example, prioritizing self-sufficiency increases the number of panels while focusing on financial return results in fewer panels and shorter payback periods. Although the financial feasibility of PV systems in Alberta's energy market is currently constrained by low electricity prices, the analysis highlights opportunities for improvement through government incentives and potential electricity price increases.