Integrated Planning and Management of Electric Vehicle Charging Infrastructure across Residential, Urban, and Freight Systems
Advisor: Professor Eleftheria Kontou
Abstract
Sustainable transportation systems are central to reducing fossil fuel dependence, improving urban air quality, and equitably improve access to opportunities. As electric vehicle (EV) adoption accelerates, planning and management of charging infrastructure have become increasingly complex. This dissertation develops integrated frameworks for siting, scheduling, and managing charging infrastructure across multiple spatial scales and application domains: multi-unit dwellings, urban curbside environments, and interstate freight corridors. Each case highlights distinct electrification challenges while offering transferable methodological and policy insights.
The first study introduces the concept of community charging hubs for residents of multi-unit dwellings (MUDs), who often lack access to home charging and face higher recharging costs and lower flexibility than single-family homeowners. To address this gap, a job-shop scheduling model is formulated and applied to minimize total system waiting time and assess hub performance under varying charger types and scales. A techno-economic assessment in the U.S. cities of Chicago, IL, New York City, NY, and Los Angeles, CA, quantifies trade-offs between charging waiting times and the levelized cost of charging, highlighting significantly higher costs but much faster service of direct current fast charging compared to Level-2 alternatives. By linking scheduling efficiency with cost outcomes, this study provides one of the first systematic evaluations of how charging hub design influences both user experience at the system level and economic feasibility in shared residential contexts.
The second study develops a rolling-horizon binary integer program for shared parking and charging management in mixed land-use MUDs, where private parking spaces in apartment complexes are offered for daytime public access. By integrating allocation rules with a demand model calibrated to EVWATTS data, the framework captures both supply-side constraints and user charging preferences. Simulation in a Chicago, IL neighborhood shows that dynamic pricing improves charging infrastructure utilization and distributes requests more evenly, while outperforming unmanaged first-come-first-served approaches in both revenue and service quality. This study makes a novel contribution by jointly modeling parking and charging in
residential settings, two domains often studied separately, demonstrating the potential of integrated parking and charging management to enhance system efficiency and financial viability.
The third study focuses on the urban context of San Francisco, CA, where curbside charging has emerged as a promising solution to limited off-street parking and charging access. A geospatial multi-criteria decision-making framework evaluates potential sites under economic, fairness, and environmental criteria, comparing four weighting scenarios (entropy, equal, high demand, high need). Results show clear trade-offs between maximizing equity and maximizing benefits in high-traffic areas. A benefit evaluation system and neighborhood clustering analysis further reveal context-specific needs and provide guidance for decision-makers on tailoring curbside charging deployment to local conditions. This study is among the first to explicitly develop a curbside charger siting framework, filling a critical research gap and addressing the unique spatial, regulatory, and social complexities of urban curbside environments.
The fourth study addresses long-haul electric freight needs, where the absence of corridor charging infrastructure remains one of the largest barriers to electrification. A network flow mixed-integer program identifies optimal corridor charging sites under a public agency's objectives of maximizing either electrified miles or electrified OD-pair coverage. An agent-based simulation translates siting plans into dynamic charging load profiles using empirical truck departure distributions and physics-based energy estimates. Results show that maximizing electrified mileage siting concentrates infrastructure on high-volume corridors, creating larger peaks in power demand, while OD-coverage strategies distribute stations more widely and reduce localized grid burdens at the cost of lower aggregate mileage. Sensitivity and robustness analyses further highlight critical sites consistently selected across near-optimal facility location scenarios. This work develops a new optimization and agent-based simulation framework for a conterminous U.S. freight charging location and energy demand estimation.
The research proposes new methods for EV charging infrastructure planning by integrating optimization, simulation, and multi-criteria evaluation across residential, urban, and freight systems, generating new knowledge relevant to transport electrification across urban and rural (i.e., freight corridors) environments. Cross-cutting insights emphasize the tension between access, charging infrastructure use, and equity; the need to couple geospatial location models with temporal operations; the critical role of pricing and ownership structures of charging hubs; and the importance of planning electric freight corridors across spatial and temporal scales. The findings provide actionable guidance for cities, utilities, transport agencies, and private charger network providers to deploy EV charging systems that are technically feasible, economically sustainable, inclusive, and aligned with tackling future energy and transport decarbonization challenges.