Optimal deployment of distributed rooftop photovoltaic systems and batteries for achieving net-zero energy of electric bus transportation in high-density cities

• Published in: Applied energy Vol. 319; p. 119274
• Main Authors: Ren, Haoshan, Ma, Zhenjun, Ming Lun Fong, Alan, Sun, Yongjun
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.08.2022
• Subjects: Distributed rooftop PV; Electric bus; High-density city; Net-zero energy; Optimal design; Electric bus; High-density city; Distributed rooftop PV; Net-zero energy; Optimal design
• ISSN: 0306-2619; 1872-9118
• DOI: 10.1016/j.apenergy.2022.119274

[Display omitted] •Develop an optimal deployment strategy of PV and batteries to power electric buses.•Propose a GA-ILP approach for high-dimensional and nonlinear optimization problems.•Adopt 3D-GIS and DL integrated approach to accurately characterize solar potential.•Verify and analyze strategy performance by comparing to vast alternative designs. Using rooftop solar photovoltaics (PV) and batteries together to power electric buses is considered a novel and feasible approach to reducing carbon emissions and tackling street-level air pollution in high-density cities like Hong Kong. However, associated optimal deployment is highly challenging due to the involved high-dimensional system planning, non-linear component sizing, and complex solar potential accurate characterization. To address this challenge, this study proposed a strategy to minimize the payback period of the deployed rooftop PV and batteries for achieving net-zero energy of electric bus transportation by taking associated grid impacts into consideration. In this strategy, a genetic algorithm-integer linear programming (GA-ILP) approach was developed to solve the high-dimensional rooftop PV planning problem and the non-linear PV and battery sizing problem. Our 3D-GIS (geographic information system) and DL (deep learning) integrated approach was used to attain accurate spatiotemporal rooftop solar power generation profiles. To verify the proposed strategy, a case study covering 5 bus terminuses with 28 bus routes and a total of 1,224 building rooftops surrounding the terminuses was conducted in a real region of Hong Kong. The effectiveness of the developed strategy was verified by comparing the optimized deployment with 150 million Monte-Carlo-generated solutions. The optimized deployment achieved the shortest payback period of 3.98 years by effectively tackling design issues such as battery oversizing, PV misallocation, and battery misallocation. To meet the increasing requirement on peak export power reduction (representing the grid impact constraint in the study), the proposed strategy can find a cost-effective balance between shifting PVs from higher-density solar potential terminuses to lower ones and increasing battery capacity. The proposed strategy can be adopted in practice to facilitate the development of cost-effective and grid-friendly community-solar-powered electric bus networks in high-density cities.