Article pubs.acs.org/est
Air Quality Impacts of Electrifying Vehicles and Equipment Across the United States Uarporn Nopmongcol,*,† John Grant,† Eladio Knipping,‡ Mark Alexander,*,‡ Rob Schurhoff,§ David Young,‡ Jaegun Jung,† Tejas Shah,† and Greg Yarwood† †
Ramboll Environ, 773 San Marin Drive, Suite 2115, Novato, California 94998, United States Electric Power Research Institute, 3420 Hillview Avenue, Palo Alto, California 94304, United States § Modelytic, 4120 Dublin Boulevard, Suite 110, Dublin, California 94568, United States ‡
S Supporting Information *
ABSTRACT: U.S.-wide air quality impacts of electrifying vehicles and off-road equipment are estimated for 2030 using 3-D photochemical air quality model and detailed emissions inventories. Electrification reduces tailpipe emissions and emissions from petroleum refining, transport, and storage, but increases electricity demand. The Electrification Case assumes approximately 17% of light duty and 8% of heavy duty vehicle miles traveled and from 17% to 79% of various off-road equipment types considered good candidates for electrification is powered by electricity. The Electrification Case raises electricity demand by 5% over the 2030 Base Case but nitrogen oxide (NOx) emissions decrease by 209 thousand tons (3%) overall. Emissions of other criteria pollutants also decrease. Air quality benefits of electrification are modest, mostly less than 1 ppb for ozone and 0.5 μg m−3 for fine particulate matter (PM2.5), but widespread. The largest reductions for ozone and PM occur in urban areas due to lower mobile source emissions. Electrifying off-road equipment yields more benefits than electrifying onroad vehicles. Reduced crude oil imports and associated marine vessel emissions cause additional benefits in port cities. Changes in other gas and PM emissions, as well as impacts on acid and nutrient deposition, are discussed.
■
methodologies and assumptions, the findings consistently show that PEVs have the potential to impact ozone. A regional-scale study for the Pennsylvania, New Jersey, and Maryland region found mixed impacts of PEVs, reporting decreases of 2−6 ppb in peak 8 h ozone in urban areas and increases of up to 8 ppb in localized areas.16 A recent nationalscale study that used a life cycle inventory (accounting for upstream fuel-sector emissions) and assumed PEVs constitute 10% of 2020 vehicle miles traveled (VMT) found that ozone and PM impacts depend on how electricity for PEVs (the marginal consumption) is generated.17 A previous study by the Electric Power Research Institute (EPRI)18 assumed plug-in hybrid electric vehicles (PHEVs) constitute 40% of on-road vehicles by 2030 with marginal electricity demand supplied by existing coal-fired units. Reduced upstream emissions from the hydrocarbon fuel sector were accounted for. PHEVs were found to reduce human exposure to ozone and PM in most regions. Our study extends the previous EPRI analysis with an updated and more complete
INTRODUCTION Sales of electric vehicles and equipment are rising and it is important to understand what air quality benefits will accrue from greater electrification in coming years. Electrifying onroad and off-road mobile sources reduces their direct emissions to the atmosphere as well as associated fuel refining and distribution (upstream) emissions. Greater electricity demand for vehicle charging will tend to increase emissions except where power generation emissions are capped by regulation or incremental load is preferentially supplied by nonemitting generation. Previous studies examined emission changes for nitrogen oxides (NOx) and other criteria pollutants due to vehicle electrification in the U.S.1−10 Only a few evaluated impacts on ozone and secondary particulate matter (PM) that form in the atmosphere from emissions of several precursors. To our knowledge, none have analyzed the air quality effects of electrifying off-road equipment. The air quality impacts of electric transportation should be periodically reassessed because of changing emission characteristics of vehicles and the power sector due to regulatory and nonregulatory influences, such as Corporate Average Fuel Economy (CAFE) standards11,12 and changing natural gas prices. Previous modeling evaluated ozone impacts of using plug-in electric vehicles (PEVs).13−15 While employing different © XXXX American Chemical Society
Received: October 13, 2016 Revised: January 7, 2017 Accepted: February 10, 2017
A
DOI: 10.1021/acs.est.6b04868 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
ing Information (SI) Table S1) for the continental U.S. except in California where 2030 VMT estimates were available from EMFAC2011. All gasoline was assumed to contain 10% ethanol (E10). Both scenarios considered recent regulations omitted from the MOVES and EMFAC2011 model versions including (1) LEV-III standards within California and Tier 3 standards outside of California, (2) Light-duty greenhouse gas (GHG) rulemaking for 2017+ model year vehicles, and (3) Heavy-duty GHG rulemaking beginning with 2014 models. Base Case vehicle emissions are shown in SI Table S4. For the Electrification Case, electric vehicles were allowed for MOVES vehicle types with daily travel distances of less than 200 miles per day except for combination unit short-haul trucks. Vehicles included in the electrification analysis were passenger cars, passenger trucks, light commercial trucks, motorcycles, refuse trucks, single unit short-haul trucks, school buses, and transit buses; all other vehicle types were assumed to be entirely conventionally fueled. We used a vehicle fleet turnover model with MOVES vehicle age distribution assumptions to estimate PEV market penetration in the vehicle fleet in future years. The fleet turnover model included several different vehicle powertrain types including conventional vehicles (CV), HEV (hybrid electric vehicles), PHEV, and BEV (battery electric vehicles). The PEVs were further binned by all-electric range (AER), which is the distance that a single vehicle can travel on electricity after a full recharge. For PHEVs, alternative fuels such as biofuels, natural gas or hydrogen were not explicitly considered. A high-electrification case from the NAS Transitions to Alternative Fuels report20 was used to estimate electric vehicle market share. The PEV sales distribution (SI Figure S1) and the PEV market share create the new vehicle market share forecast by PEV type (SI Figure S2). Light duty PEV energy economy assumptions (in watthours per mile) were based upon forecasts in AEO2013 by light-duty vehicle types through 2025; energy economy improvements of 0.5% per year were assumed beyond 2025. For heavy duty PEVs, we developed energy economy estimates by multiplying the HEV fuel economy of each vehicle category by an appropriate energy efficiency ratio. The Electrification Case for on-road vehicle emissions accounts both for increasing electric vehicle market share over time for applicable vehicle types and decreasing conventional vehicle emission rates over time as standards become more stringent due to rulemakings such as Tier 3 and LEV-III. BEVs were assumed to have no exhaust or evaporative emissions. PHEVs were assumed to have no running exhaust emissions in electric mode, start emissions reduced by 80% relative to conventional vehicles, and evaporative emissions similar to conventional vehicles. Regenerative braking was assumed to reduce brake wear emissions for PEVs by 25% relative to conventional vehicles. The resulting vehicle emissions from the 2030 Electrification Case are provided in SI Table S6. Off-Road Sector Emissions. The off-road sector consists of mobile and portable equipment not certified for on-road use. This broadly includes commercial marine vessels, locomotives, aircraft, and the mobile source engines utilized in sectors such as rail, marine, airport, agricultural, port, construction, commercial, industrial, and recreational. The off-road sector 2030 Base Case emissions were developed from multiple data sources. Emissions from off-road equipment were based on EPA’s NONROAD model24 and ARB’s OFFROAD model.25 Emissions from cargo handling equipment at ports and rail
assessment of electrification for the U.S. in 2030. Fully electric vehicles and PHEVs are included along with electrification of suitable off-road equipment (e.g., forklifts, lawn, and garden equipment). Detailed emissions modeling is performed for electric generating units (EGUs), on-road vehicles, off-road equipment and upstream petroleum fuel sectors to calculate the magnitudes of emissions changes, including changes in spatial and temporal patterns of EGU emissions. We study 2030 to allow time for electrification to develop. Fine-scale 3-D photochemical air quality model simulations for a full calendar year are performed to evaluate impacts on ozone, PM, and acid and nutrient deposition. The U.S. Environmental Protection Agency (EPA) regulates ozone and PM under the Clean Air Act and sets National Ambient Air Quality Standards (NAAQS) to protect human health and welfare (40 CFR part 50). EPA is currently reviewing the secondary NAAQS for NOx and SOx in order to address potential adverse effects from acid and nutrient deposition on aquatic and terrestrial ecosystems (https://www.epa.gov/naaqs/nitrogen-dioxideno2-and-sulfur-dioxide-so2-secondary-air-quality-standards).
■
MATERIALS AND METHODS The 2030 Base Case emissions inventory uses assumptions consistent with EPA and California Air Resources Board (ARB) rulemakings related to the regulation of criteria pollutant and greenhouse gas emissions from on-road vehicles, off-road equipment, and EGUs at the time of the study. For other source sectors, the 2030 Base Case emissions were derived from the EPA’s 2007/2020 air quality modeling for the Regulatory Impact Assessment of the 2012 Final National Ambient Air Quality Standards (NAAQS) for PM2.5 (PM NAAQS).19 Recent national control programs in the PM NAAQS inventory are (1) the Final Mercury and Air Toxics (MATS) rule announced on December 21, 2011, (2) the Final Cross-State Air Pollution Rule (CSAPR) issued on July 6, 2011, and (3) NOx and Sulfur oxide (SOx) controls for Emissions Control Areas (ECAs) off North American coasts and International Marine Organization (IMO) emission standards for Class 3 commercial marine vessels (C3 marine). The inventory also includes biofuel production associated emissions based on the Energy Independence and Security Act of 2007 (EISA). The Electrification Case assumes approximately 17% of light duty and 8% of heavy duty fleet VMT in 2030 is from vehicles driving in all electric mode based on National Academy of Sciences (NAS) Transitions to Alternative Fuels report20 and up to 80% of the sales of various off-road equipment types are electric based on electrification feasibility by equipment type. Electrification case electric vehicle and electric equipment percentages are uniform across the U.S. Emission estimates are made for ozone and PM precursors including NOx, volatile organic compounds (VOC), carbon monoxide (CO), SOx, ammonia (NH3), and primary PM. Base and Electrification Case inventory methods are discussed below for key sectors. Transportation Sector Emissions. The 2030 Base Case assumed minimal electrification of the vehicle fleet with no discernible effect on vehicle emissions. Vehicle emission factors were based on the EPA MOVES model (version 2010a run with database version movesdb20100830)21 for the continental U.S. except for California where we used the ARB EMissions FACtor model (EMFAC2011)22 results. 2030 county-level VMT estimates were forecast from the 2008 National Emission Inventory (version 1)23 to 2030 based on the 2011 Annual Energy Outlook (AEO) estimates of activity growth (SupportB
DOI: 10.1021/acs.est.6b04868 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology yards relied on container and lift activities26,27 and NONROAD model emission rates. 2008 National Emission Inventory (version 1)22 aircraft emissions were forecast based on 2008 and 2030 landing-takeoff (LTO) data28 to estimate 2030 aircraft emissions. EPA’s Regulatory Impact Analysis29 provided emissions for locomotive and harbor craft and EPA’s PM NAAQS modeling platform provided ocean-going vessel emissions. Off-road equipment electrification was modeled by assuming that technology developments will lower costs, yield longer run time between recharges, and increase battery energy density which will enhance electrification for the off-road sector. Our feasibility analysis of electrification potential determined that the following equipment types could have significant electricity powered market share in the future: lawn and garden equipment (chain saws, chippers/shredders, commercial turf equipment, leaf blowers, push lawn mowers, riding lawn mowers, trimmers/edgers, snow blowers, and specialty vehicle carts), recreational equipment (golf carts, all-terrain vehicles (ATVs), motorcycles) and industrial/commercial equipment (agricultural pumps, aircraft auxiliary power units (APUs), airport ground support equipment, dredging vessels, forklifts, intermodal equipment (port cranes, yard trucks, side/top picks), shoreside power for ocean-going vessels, sweepers/ scrubbers, switching locomotives, transportation refrigeration units). The list does not include equipment types that account for a small percentage of off-road equipment emissions in the U.S. (e.g., underground mining equipment), equipment with high torque requirements (e.g., excavators), and rail. Electrified off-road equipment electricity consumption was calculated based on the same models and methodologies used to estimate 2030 Base Case emissions assuming that electrified equipment would consume the same amount of energy as the fossil-fueled equipment being replaced. Through 2020, leadacid and lower energy density lithium-ion batteries are the assumed electricity storage mediums. Electric equipment market share increases from 2020 to 2030 result from the availability of progressively better batteries with higher energy density. The market share of new electric equipment was incorporated into a fleet turnover model based on NONROAD useful life and scrappage assumptions to estimate annual fleetwide electric equipment estimates. The EPA NONROAD model estimates emissions from fossil-fueled equipment based on pre-2000 equipment population estimates;30 it does not include electric equipment. 2010+ model year electric equipment are assumed to be consistent with technology not accounted for in NONROAD model equipment population forecasts and that electric equipment power output is the same as fossil-fueled equipment being replaced. Emission reductions from electric equipment are estimated to be the same across all pollutants and ranged from 17% to 67% (SI Table S9). Yard hostlers, aircraft APUs, and marine vessel shoreside power provided the highest emission reductions from 65 to 67%. Most lawn and garden equipment types had emission reductions greater than 50%. Power Sector Emissions. EGU emissions were estimated with a bottom-up-approach using EPRI’s U.S. Regional Economy, Greenhouse Gas, and Energy Model (USREGEN).31 The US-REGEN model is an optimization tool combining a detailed power dispatch and capacity expansion model of the electric sector with a U.S. economic model covering transactions among suppliers and consumers and forecasted economic growth. The lower 48 states are
aggregated into 15 subregions (SI Figure S7). For this study, US-REGEN was calibrated to the AEO 2011 reference case, with the exception of gas prices, which were calibrated to the AEO 2013 reference case. The electric model simulates key generating technologies including coal, natural gas, wind, solar, nuclear, and biomass, the costs and performance of which are informed by EPRI technical reports. US-REGEN incorporates many existing and pending regulations such as state Renewable Portfolio Standards (as of 2012), new emissions limits for SOx and NOx under the Hazardous Air Pollutants Maximum Achievable Control Technology (HAPs MACT) regulations, and New Source Performance Standards (NSPS) for new fossil fueled generation that prevents construction of new coal units without carbon capture and storage technology, as well as the MATS and CSAPR regulations referenced above. Existing coal units may be endogenously retrofitted in response to environmental policy in US-REGEN. Emissions from retrofitted units are assumed to be at or under 0.10 lb/MMBtu for NOx and 0.15 lb/MMBtu for SO2. New coal units (limited by NSPS to those currently in operation) meet or exceed the same emission thresholds. Base case demand growth starts at 4200 TWh per year in 2010 and increases to 4850 TWh per year by 2030, consistent with the AEO 2011 reference case. Natural gas and environmental retrofitted coal units represent about 21% and 30% of the generation mix, respectively. The Electrification Case required higher demand than the Base Case by 5.0% in 2030. The additional load is met mostly by new natural gas generation along with wind and solar generation in the West and Midwest and new nuclear generation in the South. The national capacity mix modeled for the two scenarios is shown in SI Figure S11 and S12. To model hourly emissions, USREGEN was run in an 8760 h “dispatch-only” model for the year 2030, using the capacity mix chosen by the full dynamic model. This provided an hourly generation profile, from which emissions were calculated by region and technology. Hydrocarbon Fuel Sector: Crude-Oil Shipping, Refining and Gasoline Distribution Emissions. Emissions for the hydrocarbon fuel sector were obtained from the PM NAAQS inventory for the year 2020. The 2030 Base Case includes emission reductions resulting from planned vehicle fuel efficiency improvements. The Electrification case further incorporates emission reductions due to lower fuel consumption. Refinery and crude oil shipment emissions were reduced in proportion to crude oil consumption reductions. Emissions downstream of refineries and refueling were decreased proportionate to gasoline shipment reductions. We assumed that tanker traffic accounts for 17% of ocean going vessel emissions.32 Other Sectors Not Affected by Electrification. Other anthropogenic sources for which emissions are not affected by electrification were taken from the PM NAAQS 2020 inventory. Biogenic emissions were estimated using the Model of Emissions of Gases and Aerosols from Nature (MEGAN version 2.10).33,34 Emissions from wildfires and prescribed burns were an average of recent years. These emissions were held constant between the 2030 Base Case and Electrification Case. Air Quality Modeling. The Comprehensive Air Quality Model with Extensions (CAMx), version 6.0,35 was used for air quality modeling with a 12 km grid covering the entire lower 48 states for calendar year 2007. The vertical domain has 14 layers, C
DOI: 10.1021/acs.est.6b04868 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology Table 1. Annual 2030 Base Case U.S. Emissions and Percentage Emission Changes for Electrification Case base case emissions (thousand short tons/year) area
off-road
C3-marine
NOx CO VOC SO2 NH3 PM2.5 PM10
1354 4672 6107 328 3964 1549 6677
1613 15 958 1706 22 6 103 115
147 21 8 6 0 3 3
area
off-road
C3-marine
L on-road
NOx CO VOC SO2 NH3 PM2.5 PM10
0.0 0.0 −1.6 0.0 0.0 0.0 0.0
−5.3 −22.6 −17.8 −4.3 −7.2 −15.6 −15.1
−21.9 −21.8 −14.2 −23.5 0.0 −23.8 −23.5
−7.8 −9.0 −5.3 −16.4 −15.2 −8.6 −7.4
area
off-road
C3-marine
LD on-road
HD on-road
EGU
non-EGU
0.0 0.0 21.8 0.0 0.0 0.0 0.0
41.1 70.9 68.0 8.3 3.3 68.4 55.2
15.4 0.1 0.3 13.0 0.0 2.6 2.1
19.0 24.6 7.0 8.0 88.5 18.7 32.8
22.7 4.4 1.2 2.2 8.5 2.8 3.7
−1.5 −0.1 0.0 −1.2 −2.3 −0.8 −0.7
3.3 0.1 1.9 69.8 2.0 8.2 7.0
NOx CO VOC SO2 NH3 PM2.5 PM10
LD on-road
HD on-road
510 13 999 588 6 74 51 139 electrification
731 2101 83 4 12 23 49 case emissions
HD on-road
−6.5 −10.6 −6.2 −6.5 −8.9 −2.9 −2.4 contribution (%) to total
EGU
non-EGU
800 1812 213 2152 19 996 823 939 11 63 104 358 132 522 change (%) EGU 0.4 2.0 0.7 0.02 2.6 0.2 0.2 changes
total anthropogenic
natural
total
6968 39 115 9507 2127 4131 2191 7637
98 19 277 42 874 121 262 1379 1628
7765 58 392 52 382 2247 4393 3570 9264
non-EGU
total anthropogenic
natural
total
−0.4 −0.2 −0.8 −0.9 −0.4 −0.5 −0.4
−3.0 −13.0 −4.7 −0.5 −0.3 −1.1 −0.4
0.0 0.0 0.0 0.0 0.0 0.0 0.0
−2.7 −8.7 −0.9 −0.5 −0.3 −0.7 −0.3
total anthropogenic
natural
total
100 100 100 100 100 100 100
0.0 0.0 0.0 0.0 0.0 0.0 0.0
100 100 100 100 100 100 100
Figure 1. Annual 4th highest 8 h-ozone (ppb) and 8th highest 24 h-PM2.5 (μg m−3) for 2030 Base Case (a and b) and difference between Electrification Case and Base Case (c and d).
D
DOI: 10.1021/acs.est.6b04868 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
Figure 2. Summer average daily maximum 8 h ozone concentrations for (a) 2030 base case and (b) difference between Electrification Case and Base Case. Source contributions to the difference (b) are shown for (c) natural, (d) light-duty on-road, (e) heavy-duty on-road, (f) off-road, (g) EGU, (h) other U.S., and (i) non-U.S. sources.
of the reduction attributable to refinery sources in the hydrocarbon fuel sector. PM10 emission reductions of 0.4% are attributable mainly to the off-road (55%) and on-road (36%) sectors. Air quality model results are shown in Figure 1 for high ozone and PM conditions, that is, for the fourth highest 8 h ozone (4HMDA8 ozone) and 98th percentile 24 h PM2.5 (equivalent to eighth highest 24 h (8H24A) PM2.5). The 2030 Base Case shows values exceeding the current 8 h ozone standard (0.075 ppm) in several urban areas including Los Angeles, Houston, St. Louis, Baton Rouge, and Atlanta (we use the ozone standard as a reference level without suggesting whether or not cities will exceed this level in 2030). Ozone benefits related to electrification of mobile sources occur across the U.S. (Figure 1, top right) but are mostly less than 1 ppb. Larger ozone reductions are seen in urban areas, up to 4−5 ppb in Los Angeles. Decreases in off-road sector emissions drive most of these ozone reductions (Figure 2). Although the offroad equipment is predominantly concentrated in urban areas, the benefit from electrification is spread out across the U.S. C3 marine emissions reductions also boost ozone benefits near Los Angeles and the Gulf Coast. The C3 marine sector has the highest percentage reductions of NOx, SOx, and PM emissions (22−24%). Ozone increases (less than 1 ppb) occur in a few grid cells in rural areas where the Base Case H4MDA8 ozone is lower than 60 ppb. The Base Case 8H24A PM2.5 shows highest values in the west, especially in Idaho and Montana, due to wildfires which are held constant between the two scenarios. PM 2.5 concentrations are more uniform (generally below 20 μg m−3) in the eastern U.S. and are mostly comprised of sulfate and nitrate with additional organic carbon in the south. The
which span the troposphere and lower stratosphere to a pressure altitude of 100 mb. Model inputs for meteorology and initial and boundary conditions are from the EPA’s 2007 PM NAAQS modeling database.36 The 2007 emissions are from the PM NAAQS inventory except for on-road vehicle and biogenic emissions developed for this study. Model performance was evaluated for 2007 and is summarized here. The 2007 base case demonstrated acceptable ozone performance achieving the EPA’s ozone performance goals for normalized bias (