A Support Fraction-Based PV-BESS Sizing Design for EV Charging Stations Using Hybrid GA-LP

Abstract

With the rapid growth of electrified transportation, the design of charging infrastructure and station-level energy management has become increasingly important for meeting growing power and energy demands efficiently and cost-effectively. To address this challenge, this study presents an optimal sizing framework for photovoltaic (PV) and battery energy storage system (BESS) integrated EV charging stations, using an actual battery electric bus (BEB) charging station as the case study. This work formulates the load support fraction as a planning parameter, where different load support fractions (10 to 100)% are evaluated using an annualized-cost-based NPV metric, defined as the present value of annualized net savings to quantify the economic benefits and achieve optimal PV-BESS sizing design that is most profitable over the lifetime, considering seasonal variability. A hybrid bi-level optimization approach is proposed, where the outer Genetic Algorithm (GA) searches for the best PV-BESS size combinations and the inner Linear Programming (LP) model achieves optimal hourly dispatch for each GA candidate, enabling effective energy management. The case study results from a real-world battery electric bus (BEB) charging station operated by Utah Transit Authority (UTA) in Ogden, UT, USA, demonstrate that a 40% load support fraction is optimal and robust to seasonal variations, providing the best balance between the capital costs and long-term savings, and yielding 21.2% lower annual cost compared to a charging station design without PV-BESS and 42.5% higher NPV compared to a fully PV-BESS powered design.