Can the grid handle the coming electric vehicle load?
Dean Murphy, Marc Chupka, Onur Aydin and Judy Chang are economists with The Brattle Group.
As electric vehicles become commonplace, will the grid be able to handle the extra load? Too many cars plugging in at once might cause disruptions and necessitate costly infrastructure upgrades. Handling the vehicle load in a smart way, however, will ensure a smooth transition to the plug-in future.
Plug-in electric vehicles (PEVs)—powered in part or fully by the electric grid—might offer an opportunity to shift much of the transport sector’s energy demands away from petroleum, reducing dependence on imported oil and improving the environment. If the electric grid can be decarbonized (still an open question), this shift will help to reduce overall carbon dioxide (CO2) emissions as well. The potential has drawn much attention from a variety of quarters—including policymakers concerned about climate change and energy security, and vehicle manufacturers wanting to capitalize on a new market. Power system planners concerned about the possible effect of PEVs on the electric system also have begun exploring the issues.
Transportation as a whole uses an immense amount of energy—about two-thirds as much primary energy as consumed by the entire U.S. electric sector—and emits about a third of the nation’s CO2. If a large share of the transport sector were to convert to electricity, it could become a huge new electric load that raises many questions, concerns and maybe opportunities for the power sector. If PEVs were to charge up during off-peak times so that they help fill the valleys in the daily load shape and improve the utilization of the existing grid infrastructure, they could be a valuable complement to the power grid. However, if PEVs are adopted in large numbers but do not coordinate with the grid (e.g., charging at times that add to peak demand or create rapid load ramps)—they could be disruptive and might necessitate significant additional infrastructure and investment.
As PEVs penetrate the transportation industry, they present a range of possible impacts on the U.S. power system.1 Because the relationships among PEV penetration, charging behavior, electric infrastructure requirements and environmental outcomes might not always be aligned, PEVs pose some interesting tradeoffs that policymakers and the power industry are just beginning to grapple with.
PEV Market Outlook
Although electric vehicles were developed in the early days of the automobile, the advantages of the internal combustion engine and petroleum fuels in terms of cost, performance and quick refueling relegated electric vehicles to the sidelines for the 20th century. The commercial introduction of hybrid electric vehicles (HEVs) about a decade ago offered a new way to capture some of the advantages of electric vehicles. HEVs don’t use electricity as a primary fuel; all their power originates with their internal combustion engine, with the battery and electric motor simply storing surplus energy and recovering it later. But they can improve fuel efficiency considerably, and their sales have grown rapidly, making up about 3 percent of U.S. light-duty vehicle sales in 2009, and more in some regions.2 Most notably, HEVs have opened the door for the plug-in hybrid electric vehicle (PHEV)—essentially an HEV with a much larger battery that can be charged with grid electricity for part of the vehicle’s primary energy needs. PHEVs might in turn re-open the potential for pure battery electric vehicles (BEVs), which forego the internal combustion engine altogether and run solely on grid electricity. For the foreseeable future, most PEVs likely will be PHEVs using a mix of electricity and gasoline.
Although the mass-market PEVs aren’t available just yet, many major auto manufacturers are planning to launch them in the next year or two. Chevrolet is introducing the Volt, Toyota will offer a plug-in version of the Prius hybrid, Nissan is launching its all-electric Leaf, and Ford is planning an electric version of the Focus. As a result, the potential for powering vehicles with grid electricity already has captured the attention of the power industry. For example, the ISO/RTO Council, a consortium of 10 North American independent system operators and regional transmission organizations, recently released an assessment of PEVs and how they might integrate with electric systems.3 Several regional collaborations of electric utilities and other entities have been formed to promote the development of electric transportation infrastructure, including the Regional Electric Vehicle Initiative in New England and the Regional Plug-In Electric Vehicle Planning process in southern California.4
Despite these developments, any forecast of PEV market penetration is highly speculative at the moment. Expectations for how quickly PEVs will penetrate the market depend on their economics compared to gasoline vehicles (including HEVs), their performance, safety, customer acceptance, and public concern about environmental impacts. Depending on assumptions about how these factors will evolve, market-share projections for PEVs range widely.
PEVs currently face a significant economic hurdle due primarily to the high cost of batteries, even accounting for recent and forecast improvements in battery cost and performance. Although electricity is much less expensive than gasoline per mile travelled, the operating cost savings are modest relative to the incremental initial purchase cost of a PEV. Fueling a 35-mpg gasoline vehicle—which would match the 2016 CAFE standard—with $3/gallon gasoline costs about 8.6¢ per mile, or $1,285 per year. In comparison, a PEV that uses 250 watt-hours per mile at 12¢/kWh, assuming it has sufficient electric range to run entirely on electricity, has fuel costs of about 3¢ per mile, or $450 per year. While this is an operating cost savings of 65 percent, or more than $800 per year, the incremental purchase cost of the PEV likely will be many thousands of dollars. Many PEVs will be plug-in hybrids that still rely on gasoline for part of their energy needs; while their smaller battery reduces the purchase cost premium, it also provides a shorter electric range and correspondingly smaller fuel savings.
The PEV’s fuel savings are smaller when compared to higher-mileage non-plug-in vehicles such as HEVs. Until battery costs decline considerably, or gasoline prices rise dramatically, it’s unlikely that PEVs will offer significant cost advantages over non-electric vehicles, so they probably will be attractive to only a limited segment of drivers. These may be the same drivers that currently favor more fuel-efficient vehicles (including HEVs) and whose driving patterns reflect a high proportion of urban travel. While this factor might complicate the estimation of gasoline savings from PEV penetration (i.e., because PEVs might tend to replace fuel-efficient conventional vehicles), it does suggest a distinct geographic pattern of initial PEV concentration—largely in urban areas on the East and West coasts. However, PEVs might be embraced for reasons beyond a strict cost advantage, and supportive public policies coupled with innovative designs and effective marketing could help to broaden their appeal.
DOE’s Energy Information Administration (EIA) has estimated that the annual sales of PEVs will grow to almost 140,000 vehicles by 2015, and 400,000 vehicles by 2030, supported by tax credits enacted in 2008—currently $2,500 per vehicle, plus $417 per kWh of battery capacity in excess of 5 kWh. An MIT report assumes plug-in hybrids will account for 2-3 percent of new vehicle sales by 2020, and 10 percent by 2030.5 A much more optimistic scenario developed by the Electric Power Research Institute (EPRI) assumes 35 percent for 2020, and 50 percent for 2030.6 Of course, for a new technology with adoption rates ramping up, its share of the overall vehicle fleet will be considerably lower than its share of new vehicle sales, because the fleet is only gradually replaced by new vehicles.
The National Research Council released a study that projected a “maximum practical” overall fleet penetration of PHEVs of about 13 percent by 2030, with a “more probable” penetration of less than 5 percent by that time.7 The ISO/RTO Council’s March 2010 report estimates 1 million PEVs will be sold within five to 10 years, also concluding that early adoption would cluster in urban centers. The faster trajectory leads to a total of about 2.5 million vehicles by 2020—about 1 percent of the light duty vehicle fleet, though with a fair amount of geographic concentration that could mean a larger impact in some areas.
Overall, the market penetration of PEVs likely will be limited to a few percent of new vehicle sales over the next decade, and a smaller share of the vehicle fleet. If some of the barriers to PEV adoption can be overcome quickly, and consumer acceptance is high, it appears possible, albeit very optimistic, that PEVs might achieve as much as a 20-percent share of new vehicle sales and 5 percent of the vehicle fleet in some urban regions by 2020, potentially growing more quickly thereafter. But there are potential grid implications of this optimistic level of PEV penetration over the next decade.
Impact on the Grid
A study by Pacific Northwest National Laboratory showed that up to 84 percent of U.S. cars, pickup trucks, and SUVs theoretically could be converted to plug-in hybrids without requiring additional electric infrastructure (i.e., by charging vehicles only at off-peak times to utilize electric generating capacity that’s otherwise idle).8 However, it’s unlikely that all PEVs would charge exclusively during off-peak periods. In fact, their demand on the regional power infrastructure depends greatly on when, and how quickly, drivers charge their vehicles. A 2008 Oak Ridge National Laboratory study estimated that the increase in energy demand would be about 1 to 2 percent in 2020, and about 2 to 5 percent in 2030, based on a very aggressive assumed fleet penetration of 10 percent by 2020, and 25 percent by 2030.9 That report also found that faster charging, if concentrated in the evening hours, could increase peak electricity demand substantially. In some extreme scenarios examined, where all vehicles utilize rapid charging coincident with system peak, peak demand could increase as much as 10 percent in 2020, and more than 25 percent in 2030. However, this study also found that if PEVs are charged at times less coincident with the existing system peak (i.e., charged later or more slowly, or not all at the same time), they would have a much more modest effect on peak load, and potentially no effect at all if charging occurs entirely off-peak.
Although it’s difficult to predict charging patterns confidently in the absence of actual customer experience with PEVs, natural diversity in drivers’ schedules and habits might make either of these extremes—all on-peak charging or all off-peak charging—unlikely.
Access to charging spots is an important factor shaping drivers’ charging patterns. Most PEVs likely will be charged at home in evening and night-time hours. The availability of public charging spots (e.g., workplaces, public garages, street charging) will increase the diversity of charge times and might increase the total amount of electricity used; charging twice daily means using more electricity and less gasoline in a PHEV, though it also will encourage more charging during high-load daytime hours.
Take for example several hypothetical charging profiles that illustrate how the incremental PEV load might be distributed across the hours of the day, with the area under each summing to 100 percent over the day (see Figure 1). The “evening concentrated” profile assumes that all drivers begin charging their batteries more or less simultaneously at 5 to 6 p.m. using rapid chargers that give a full charge in two hours. “Evening diversified” assumes that some drivers begin charging at 5 to 6 p.m. and some start several hours later, using slower chargers that spread the load across eight hours. In “increased work access,” half of the charging starts during morning at 8 to 9 a.m., and the other half starts at 5 to 6 p.m. In “off-peak,” charging occurs at night, starting from 10 to 11 p.m. and continuing through the early morning. These are purely hypothetical demonstrations of the potential incremental load shape; PEVs’ actual charging profiles will be driven by drivers’ information, habits, convenience and electric pricing schemes.
With substantial PEV penetration, the evening-concentrated profile might create concern for an electric system, since it concentrates a large new energy demand at or near system peak. Increased daytime charging access (e.g., at work) might make some PEV load coincide more directly with existing system peaks, though it also might help to diversify the charging load across time. Further, it might create additional stresses on urban distribution systems by concentrating additional loads in areas and times that already have high loads. Night-time charging seems to be the best complement to current system conditions, though if charging loads were concentrated sufficiently, they could conceivably create a secondary daily peak.
A tangible example considers the potential impact of these different charging patterns on simulated New England electricity demand in 2020. The PEV demand is overlaid on both summer and winter peak day load shapes (assuming 5-percent New England fleet penetration by PEVs that get half their energy from electricity, on average) (see Figure 2). If charging is heavily concentrated at or near system peak on a system like New England’s (e.g., the evening-concentrated charging pattern), even relatively modest PEV penetration might increase system capacity needs by several percent. Other charging patterns that involve charging later or spreading it over longer periods—some of which may occur naturally—greatly reduce or eliminate the impact on system capacity needs. The increased-work-access charging pattern adds slightly to capacity needs because it adds load at system peak times, though the increment is small because the load is distributed across many hours. The other patterns have no effect on peak load at all.
The seasonality of the daily load pattern also can be important. In New England, system load usually peaks in summer in mid-afternoon hours. This is several hours before much of the likely PEV load, if drivers charge at home after work. But the winter peak occurs later in the evening, more coincident with the likely PEV load. Since New England is a summer-peaking region overall, the coincidence of PEV loads with winter peak might not be a major concern. But on a winter-peaking system, an unmanaged PEV charging profile could increase overall system capacity requirements even at low PEV penetration, giving an incentive to more aggressively manage PEV loads.
To encourage charging during off-peak hours likely would involve technological and pricing solutions. Technological solutions include smart-charging systems that can consider electrical system preferences while meeting drivers’ needs. Pricing solutions, such as dynamic pricing of electricity, could enhance consumers’ incentives for off-peak charging by offering different prices for peak and off-peak electricity. Both approaches should help to steer drivers toward better use of the electric system, though it isn’t clear how PEV owners would respond. Depending on how prices are structured, the potential savings associated with dynamic pricing might be relatively small and could be ignored, and technological solutions might be bypassed if they don’t meet drivers’ needs.
The potential to integrate PEVs into the future smart grid so that they can respond dynamically to help balance the grid’s needs, becoming a controllable resource as well as a new customer and perhaps even compensating for the variability of renewable generation sources, is a particularly attractive long-term prospect. In fact, a recent article on variable renewable energy integration states: “Plug-in electric vehicles promise to increase minimum loads at night—making use of surplus wind-energy generation—and to offer fast and accurate response to high variability in wind net load, as needed by the system operator.”10 However, most analysts regard widespread adoption of such vehicle-to-grid (V2G) technology as many years more distant, recognizing the technological and regulatory barriers that must be overcome before it can be deployed widely.
Although the impact of likely PEV penetration rates on overall system demand is modest, if PEV adoption is highly localized (e.g., in affluent neighborhoods), the additional charging load potentially could add stresses on the local distribution system. The expected modest PEV numbers, coupled with the fact that easily observable PEV sales give advance warning of aggregate fleet penetration, might help utilities to foresee and deal with these issues. However, this is an area that would benefit from additional research and real-world experience, and active utility intervention ultimately might be needed to prevent or address localized distribution issues.
From an environmental perspective, switching from gasoline to electricity as a transport fuel means trading the emissions of a gasoline-powered internal combustion engine for those of the marginal electric generator at the time the PEV is charged.11 In addition to moving those emissions geographically from the tailpipe of the vehicle to the generator’s smokestack, the quantities emitted will differ. The CO2 emitted from a gasoline vehicle is similar to that released by charging a PEV with power from a coal-fired plant. Coal has about the highest CO2 emissions among common generation types, so in many regions and circumstances (i.e., when some other resource that emits less CO2, such as a gas plant, is on the margin), a PEV will emit less CO2 than a gasoline vehicle. This generally is the case in New England, where gas-fired units are on the margin in most hours, emitting CO2 at about half the rate of coal-fired power.
The annual CO2 emissions of a conventional gasoline vehicle (at 25 to 45 mpg) can be compared with those of a PEV under several alternative charging circumstances (see Figure 3). For a more direct comparison, the PEV is assumed to be all-electric rather than a plug-in hybrid. If the PEV is charged when a coal plant is on the margin, its CO2 emissions will be similar to those of a gasoline vehicle. Note that the range of PEV emissions when charging with coal corresponds to a range of coal plant efficiency—a less efficient coal plant emits more CO2. If a gas plant is on the margin, the PEV emits much less CO2. On most systems, the marginal emissions rate differs over the hours of the day, so the PEV’s CO2 emissions might depend on the particular charging pattern, and would be an average of the marginal emission rates during the charging period. The right-most bar on Figure 3 illustrates the CO2 emissions of a PEV charged on the simulated 2020 New England power system; the range reflects the emissions of different charging profiles, accounting for the marginal unit’s emissions in each charging hour. Because the New England power system has gas on the margin in the large majority of hours, a New England PEV would emit significantly less CO2 than a gasoline vehicle, and its emissions don’t depend greatly on the charging profile.
This might not be true in all regions, particularly those dominated by coal. In some regions such as the Midwest, where coal-fired units are often marginal in off-peak hours, off-peak charging may forego much of the potential CO2 reductions. Even in such regions, however, to the extent the power grid is decarbonized over time through the addition of less carbon-intensive resources (e.g., gas, renewables, hydro, nuclear) and reduced reliance on coal, a PEV’s emissions can fall over time.
The effect of PEVs on the emissions of other pollutants, such as nitrogen oxides (NOx) and sulfur dioxide (SO2), depends heavily on the emissions controls in place—the vehicle’s own emission controls for a gasoline vehicle, and the emission controls on the marginal generator for the PEV. As with CO2, the power system’s marginal emission rates of other pollutants can vary considerably with time—and certainly from one electric system to another—depending on which unit is on the margin. In the New England analysis, NOx emissions of a PEV generally were similar or slightly lower than those of a gasoline vehicle. SO2 emissions were a bit higher, because gasoline contains little sulfur and some generator types, particularly coal and oil-fired units, do emit SO2 and are sometimes on the margin. More generally, electric generating units might be farther from the urban areas that are most sensitive to air quality, both geographically and temporally, compared to the internal combustion engines which emissions they would displace, though again this may depend on the particular circumstances.
Time to Plan
PEV penetration of the vehicle fleet is likely to be relatively limited for the next decade and probably beyond, in part because the high initial cost of batteries likely will continue to outweigh the potential fuel-cost savings. Due to the natural lag in replacing the existing fleet, fleet penetration by PEVs will be gradual even if they achieve a relatively high share of new vehicle sales. Further, if PEV sales are high, this will be observable before PEVs take over a large share of the fleet, giving the industry some opportunity to plan accordingly.
The energy demand that PEVs might place on the power system isn’t particularly great even at higher fleet penetration levels, and will be quite modest at more likely penetration levels. The potential effect of PEVs on electric peak loads might be another matter. Even at plausible penetration levels, PEVs could affect system peak loads if many of them are charged quickly at or near peak load times, though natural diversity in users’ schedules and habits might mitigate some of this concern. This might make charging patterns and the timing of PEV load a more important factor for the power industry than overall PEV penetration. Beyond trying to understand PEV market penetration, which of course is important, we need further study to understand users’ charging behavior. It will be particularly important to understand what natural charging behavior would be, how it will respond to the electric industry’s attempts to influence it, and how this might interact with the design of vehicles and chargers. The effect in any particular region may depend importantly on the details of system load shape and seasonality, and in some cases might warrant efforts to understand and perhaps influence charging behavior even at fairly low penetration levels. Again, because of the lag between sales share and fleet penetration levels, power system planners might have an opportunity to learn from early experience with PEVs’ charging patterns in actual use, prior to their becoming a major source of load.
1. This article draws on research performed in the context of a state-level integrated resource plan (Integrated Resource Plan for Connecticut, The Brattle Group, The Connecticut Light and Power Co., and The United Illuminating Co., Jan. 1, 2010). While that research focused on planning for Connecticut and the New England electric system, the observations have similar relevance for other regions.
2. U.S. DOE, Alternative Fuels and Advanced Vehicles Data Center, at: www.afdc.energy.gov.
3. Assessment of Plug-in Electric Vehicle Integration with ISO/RTO Systems, ISO/RTO Council and KEMA Inc., March 2010.
5. Heywood, J. et al., On the Road in 2035: Reducing Transportation’s Petroleum Consumption and GHG Emissions. MIT Laboratory for Energy and the Environment, Report No. LFEE 2008-05 RP, 2008.
6. Duvall, M., E. Knipping, Environmental Assessment of Plug-In Hybrid Electric Vehicles. Volume 1: Nationwide Greenhouse Gas Emissions, Report No. 1015325, Electric Power Research Institute, 2007.
7. National Research Council Committee on Assessment of Resource Needs for Fuel Cell and Hydrogen Technologies, Transition to Alternative Transportation Technologies—Plug-in Hybrid Electric Vehicles, National Academy of Sciences, 2009.
8. Kintner-Meyer et al.,Impacts Assessment of Plug-in Hybrid Vehicles on Electric Utilities and Regional U.S. Power Grids, Part 1: Technical Analysis, Pacific Northwest National Laboratory, 2007.
9. Hadlew, S.W. and A. Tsvetkova, Potential Impacts of Plug-in Hybrid Electric Vehicles on Regional Power Generation, ORNL/TM-2007/150, Oak Ridge National Laboratory, 2008.
10. Milligan, Porter, DeMeo, Denholm, Holttinen, Kirby, Miller, Mills, O’Malley, Schuerger and Soder, “Wind Power Myth Debunked,” IEEE Power and Energy Magazine, November/December 2009.
11. A PEV’s electric emissions usually should be characterized in terms of the electric generator that is on the margin (i.e., the unit that will increase or decrease its output in response to a change in load) when the PEV is charging, rather than the system average emissions.