An Integrated Optimization Model for Large-Scale EV Fleet Deployment: Balancing Emissions Reduction and Operational Costs

Abstract

This thesis presents an integrated optimization framework designed for the large-scale deployment of electric vehicles (EVs) within commercial fleets, specifically focusing on balancing emissions reduction and operational cost efficiencies. Utilizing Verizon’s extensive fleet of over 10,000 light-duty vehicles across 1,000 sites as a case study, the research addresses the challenges and complexities in effective site selections for such a large and dispersed fleet. The research involved developing and testing several optimization models under varying scenarios, including scenarios prioritizing maximum operational savings, maximum emissions reduction, and a hybrid model employing an internal cost of carbon (ICC) to balance both operational and environmental objectives. The model essentially develops a ranking system for sites – suggesting which sites to electrify in which year and order, with how many EV conversions (from existing ICE vehicles) at each site. The results highlight the importance of tailoring EV deployment strategies to site-specific conditions, such as unique vehicle usage patterns, grid emissions profiles, regional operational costs, and available incentives. Particularly, smaller sites were found to offer greater relative benefits in terms of both cost savings and emissions reductions per unit of capital invested due to their high average mileage, making them strategic priorities for early electrification. Operational feasibility was also thoroughly examined, recommending practical constraints such as limiting the number of sites electrified annually to ensure project manageability and effectiveness. Sensitivity analyses addressed critical uncertainties such as battery degradation over the vehicle lifespan and the impact of extreme weather on EV performance. These analyses underscore the necessity of conservative battery range buffers ("safe ranges"). Robust load management strategies can be deployed to significantly reduce demand charges and optimize charging schedules based on time-of-use rates where available. Recommendations from the study advocate for implementing a hybrid optimization strategy incorporating an ICC based on corporate goals, continuous adaptive management informed by ongoing data collection, and strategic infrastructure investments to future-proof EV deployments. Policy alignment is also critical to enhance economic viability via incentives and ensure regulatory compliance. Finally, the thesis proposes future research directions, including investigation of advanced load management and integration with renewable energy sources, exploring bi-directional charging to add revenue streams, incorporating marginal operating emissions rate (MOER) data to further reduce grid emissions and exploring the resilience of EV fleets to power outages. These initiatives aim to further enhance strategic decision-making and ensure the long-term sustainability and efficiency of fleet electrification programs.