Comparing costs and climate impacts of various electric vehicle charging systems across the United States

The United States is pursuing large-scale public charging infrastructure to enable widespread EV adoption needed for climate targets. As EVs expand across all vehicle classes, minimizing charging dwell time is increasingly important for uninterrupted transportation. Beyond Level 2 and early DC fast charging (DCFC) deployments, medium- and heavy-duty vehicle (MDV/HDV) electrification intensifies requirements for high-power, short-dwell solutions, including 350 kW DCFC, battery swapping (BSS), and dynamic wireless power transfer (DWPT). Each technology poses distinct trade-offs in grid impact, capital intensity, standardization, social acceptance, and operational performance. A critical gap persists in understanding how the choice of charging system, location, vehicle category, and scenario jointly shape economic outcomes (TCO) and environmental outcomes (GHG intensity). This study addresses that gap by comparing nationwide TCO and cradle-to-grave GHG intensity for EVs using DCFC, BSS, and DWPT across four vehicle categories from 2031–2050 under multiple adoption and cost scenarios.

Prior work has evaluated EV charging costs, rate design, and the GHG intensity of electricity use, as well as operational challenges for DCFC such as demand charges and peak loads. Studies on BSS highlight rapid refueling potential, grid-friendly load management via off-peak charging, and deployment experience in China, while underscoring the need for battery standardization and social challenges around battery ownership and sharing. DWPT research examines in-motion charging enabling smaller batteries and eliminating charging stops, but notes high capital costs, construction disruption, limited deployment history, and lack of standards. Although performance characteristics of DCFC, BSS, and DWPT are documented, a comprehensive comparison of their system-wide economic and environmental implications across vehicle categories and U.S. geographies has remained limited, motivating this integrated analysis.

An integrated techno-economic analysis (TEA) and attributional life cycle assessment (LCA) compares widescale, independent deployments of DCFC, BSS, and DWPT across the contiguous U.S. for four vehicle categories: passenger cars, light-duty trucks (LDTs), medium-duty vehicles (MDVs), and heavy-duty vehicles (HDVs). Infrastructure is assumed built in 2030 with a 20-year operating period (2031–2050). Functional unit: one vehicle-kilometer traveled (VKT). Scenarios: optimistic, baseline, conservative for EV adoption, capital costs, electricity prices; and for ICEV/HEV fuel prices. Public charging usage modeling: For DCFC and BSS, cars/LDTs are assumed to be battery EVs using public infrastructure about 5% of energy (based on ~6% observed public charging for LDVs). For MDVs/HDVs, public charging usage is simulated by operating ranges and battery sizes with overnight depot charging; battery modules sized (90 kWh increments) to enable at most one public charge per day with state-of-charge managed between ~20–80%. VKT-weighted public energy shares are ~0.6% (MDVs) and ~14% (HDVs). For DWPT, electrified roadway maintains state-of-charge in-route for cars, LDTs, MDVs, HDVs, and buses; required electrified portion of each roadway segment is computed from vehicle efficiencies, pad power (50 kW), number of vehicle pads by class, charging efficiency (85%), speed limits, and pad failure rates (design includes redundancy). Time-of-day (ToD) resolution: Charging demand profiles are derived from trip arrival/departure distributions. Cars/LDTs: 2017 National Household Travel Survey trip datasets. MDVs/HDVs: NREL Fleet DNA (class 3–7 and class 8) weighted by operating ranges (Table 1). DCFC charging aligned with dwell periods; BSS and DWPT aligned with in-route periods. Hourly profiles inform rate selection and emissions by hour. Infrastructure deployment and allocation: Annual roadway energy demand (2031–2050) is estimated by combining FAF4 traffic (interpolated/extrapolated to 2050) and FHWA VKT shares to disaggregate by vehicle class, with EV adoption and 85% charging efficiency. Potential sites (≈122k: gas stations, surface parking lots, existing DCFC sites) are screened by proximity to sub-200 kV transmission (≤6 km), with ≤1.5 km exempt from power limits; others capped at 2.5 MW. DCFC stations use 150 or 350 kW chargers with spatial limits (up to 32 dispensers) and utilization thresholds; BSS stations include small (cars/LDTs) and large (MDVs/HDVs) swap lines with 3-min swaps and minimum daily swaps to avoid queuing. A gravity model allocates roadway charging demand to the nearest up to 30 sites proportional to site capacity and inverse square distance, iterated to honor site capacity and minimum utilization thresholds. DWPT is deployed along major roads to match in-route energy use; if one lane saturates (>50% of traffic), a second lane per direction is electrified. TEA and charging cost: For each site/segment, a discounted cash flow rate of return model determines levelized charging cost assuming 5% IRR, 50% debt at 6% interest over 10 years, federal (21%) and state corporate income taxes, sales tax, and MACRS depreciation over a 21-year cash flow (1-year build + 20-year operation). Capital costs reflect procurement and installation by technology: DCFC charger hardware and site installation (with staged procurement in 2030 and 2040); BSS building, automated storage/retrieval, cabinets, chargers (7.7 kW small; 50 kW large), and batteries added as needed; DWPT civil (rural/urban) and electronics costs per lane-km, with inverter replacements as O&M. Electricity tariffs from the US Utility Rate Database for the largest utility in each state are applied with ToD charging profiles, including demand charges, energy rates, and fixed charges (BSS/DCFC sites; DWPT per 16 lane-km segment). BSS charging is scheduled to minimize electricity cost while meeting pre-swap readiness constraints. Future electricity prices (2031–2050) use EIA AEO 2023 projections; ICE/HEV fuel prices use state-level projections. TCO: Ten-year TCO per VKT includes charging/fueling cost, depreciation, maintenance, insurance, license/registration. Vehicle purchase prices for EVs are EV glider plus battery (price per kWh by scenario) with full-size or reduced batteries (DWPT case sized for ~56 km range, ≤80% DoD). Depreciation schedules vary by class; HDV EVs may incur battery replacement within 10 years due to high VKT. Aggregate change to on-road transportation cost from ICEV to EV adoption is computed both as percentage and absolute USD, weighted by VKT and adoption. LCA: Cradle-to-grave GHG intensity (g CO2e/VKT) is computed for EVs (charging emissions, embodied infrastructure, embodied vehicle) and for ICEVs/HEVs (vehicle embodied plus feedstock/fuel and operation). Charging emissions use hourly consumption mixes for 134 Cambium 2022 zones (optimistic: 100% decarb by 2035 mid-case; baseline: mid-case; conservative: high renewables cost) tracking interconnection imports/exports and storage charging. Infrastructure inventories draw on Ecoinvent 3.8/openLCA for DCFC pedestals/cabinets and construction, BSS equipment (chargers, cabinets, AS/RS, building), and DWPT electronics and concrete-based pavement; vehicle embodied emissions from GREET 2022 using representative 2025 vehicles and lithium-ion batteries (manufactured in China). Emissions are allocated per kWh dispensed for charging/infrastructure and per VKT for vehicles.

- TCO impacts are highly location- and scenario-dependent due to local fuel and electricity prices and traffic volumes. From 2031–2050, compared to ICEVs, U.S. on-road transportation costs change by approximately -22% to +11% across scenarios. - DWPT shows the widest range in TCO change at county level (-31% to +429%) driven by utilization-sensitive capital recovery; DCFC charging cost is driven by local demand charges; BSS shows limited spatial price variability but is sensitive to capital cost and adoption. - By vehicle class, EV TCO can be lower than ICEV/HEV for cars and LDTs; MDV EVs via DCFC/BSS are cost-advantaged only under high fuel price scenarios, while DWPT-EVs for MDVs achieve lower costs across fuel price scenarios. For HDVs, DWPT-EVs (with reduced battery size) are the only EVs showing a cost advantage over ICEV/HEV due to reduced depreciation/maintenance versus full-size battery EVs. - National GHG outcomes: EVs have lower GHG intensity than ICEVs and HEVs across all scenarios and vehicle categories. Overall GHG emissions change from approximately -53% to -19% versus ICEV baselines across scenarios (2031–2050). - County-level GHG outcomes: For DWPT, low-utilization deployments can increase county emissions by up to 167%, though with limited national impact due to low VKT; with high utilization and clean grids, county-level reductions reach up to 71%. For DCFC/BSS, percent GHG change is primarily driven by local electricity mix. - GHG intensity breakdown: DCFC-EV infrastructure emissions are minimal; BSS-EV and DWPT-EV infrastructure emissions are substantial due to battery inventory (BSS) and concrete (DWPT). Reduced DWPT vehicle battery size lowers vehicle embodied emissions but does not fully offset DWPT infrastructure emissions relative to DCFC-EVs. - Grid and supply implications: Average U.S. electricity generation needs increase by ~16–38% versus 2022. Managed charging (BSS/DWPT) can reduce grid capacity upgrade needs relative to DCFC. Battery production for full-size EVs totals ~13–31 TWh (2031–2050); DWPT’s smaller batteries could reduce this by ~79%. - Capital investment (2030 builds): DCFC ≈ 23–52 billion USD; BSS ≈ 41–115 billion USD; DWPT ≈ 134 billion–1.7 trillion USD, reflecting DWPT’s scale and civil works. DWPT supplies a larger share of EV charging: cars 35–61%, LDTs 32–61%, MDVs 29–56%, HDVs 67–83%; DCFC/BSS supply about 5% (cars/LDTs), 0.6% (MDVs), and 14% (HDVs), implying substantial reliance on home/workplace/depot charging for non-DWPT scenarios.

The comparative TEA-LCA demonstrates that EV economic and environmental benefits are highly sensitive to charging system choice, infrastructure utilization, and local electricity mixes. The results quantify that EV adoption can lower national transportation GHG emissions across all scenarios while revealing spatial heterogeneity: clean-grid regions and high-utilization corridors deliver the largest benefits. For costs, high traffic volumes enhance utilization and reduce per-kWh capital recovery for DWPT and BSS, improving competitiveness, while DCFC economics are dominated by local demand charges. By class, DWPT’s smaller batteries enable HDV cost advantages not realized by DCFC/BSS EVs within 10 years due to battery replacement costs. System-level implications include increased electricity generation (16–38% above 2022) and potential distribution upgrades, with managed charging via BSS and DWPT mitigating peak impacts relative to DCFC. Battery manufacturing constraints are alleviated by DWPT’s reduced battery size, but at the expense of higher infrastructure capital and embodied emissions. Deployment feasibility varies: DCFC is mature and widely deployed for LDVs but is scarce for MDVs/HDVs; BSS is technologically mature with large-scale deployments in China but faces U.S. standardization and social acceptance barriers; DWPT is nascent with limited pilots and high capital intensity, suggesting careful, corridor-focused rollouts aligned with road refurbishments and budget constraints. Policymakers should tailor deployment by geography and vehicle class, integrating TCO, GHG intensity, and utilization to prioritize infrastructure that maximizes benefits and equity.

This work provides a comprehensive, scenario-based comparison of DCFC, BSS, and DWPT across U.S. geographies and vehicle classes, linking infrastructure deployment, charging costs, TCO, and cradle-to-grave GHG intensity from 2031–2050. Key contributions include quantifying the location- and utilization-dependence of costs and emissions; identifying DWPT’s potential for HDV cost advantages and battery manufacturing relief; and highlighting the centrality of grid decarbonization and managed charging. Future research should: (1) validate DWPT performance and costs via large-scale pilots and standardization; (2) refine behavioral/adoption models, public versus private charging splits, and dynamic pricing impacts; (3) co-optimize mixed-system deployments (DCFC, BSS, DWPT) by corridor and fleet type; (4) assess equity, financing mechanisms, and rate design to reduce demand charge impacts; and (5) incorporate evolving battery lifetimes, chemistries, and recycling pathways to reduce embodied emissions.

- The TCO includes only public charging costs from DCFC, BSS, and DWPT, excluding home, workplace, and depot charging costs; real-world TCO would reflect a mixed charging portfolio. - Infrastructure systems (DCFC, BSS, DWPT) are modeled independently for comparability; combined or hybrid deployments may yield different outcomes and synergies. - Traffic data (FAF4) exclude local roads; stationary charging VKT was scaled to FHWA totals, introducing uncertainty in spatial allocation. - DWPT deployment and performance rely on limited real-world data; costs, reliability, and consumer acceptance are uncertain, and lack of standards may affect readiness and interoperability. - Electricity price and grid mix projections (Cambium, AEO) are scenario-dependent; marginal versus average emissions and future grid policies (e.g., carbon capture allocations) are simplified or out of scope. - Maintenance of roads (DWPT) is assumed out of scope (tax-funded) and may understate lifecycle costs compared to DCFC/BSS site O&M inclusions. - Assumptions on battery lifetimes, replacement needs (notably for HDVs), and manufacturing locations/chemistries affect embodied emissions and TCO; social/standardization barriers for BSS are treated qualitatively. - Site selection uses proximity to transmission and utilization thresholds; distribution system constraints and substation siting are simplified.