Interview with Mike Jacobs and Peter O’Connor
Mike Jacobs is the lead on Electricity Markets and Regulatory efforts in the Climate and Energy Program at the Union of Concerned Scientists (UCS). Prior to coming to UCS, Mr. Jacobs worked as the markets and policy director at a number of renewable energy and energy storage companies, and the American Wind Energy Association (AWEA). In these positions, he developed strategies for wind integration using battery storage, and new and existing transmission. While with AWEA, he led settlement efforts at the Federal Energy Regulatory Commission to streamline generator interconnection rules. He has served on the boards of Vineyard Power Co-cop, Solar Grid Storage, Wind on the Wires, the Wind Coalition, Interwest Energy Alliance and the Northern Maine Independent System Administrator.
Peter A. O’Connor holds a Ph.D. from Boston University’s Department of Earth and Environment. The topic of his dissertation was energy transitions in societies. He has published several articles on energy topics in peer-reviewed journals. Prior to enrolling at BU, he worked for the Global Environment & Technology Foundation (GETF) in Arlington, VA, where among other projects he conducted analysis of advanced vehicle technologies for the State of California and researched the business case for solar photovoltaics for the Solar Electric Power Association (SEPA). He has also conducted research on energy topics for the World Bank, the National Association of Regulatory Utility Commissioners, and numerous other clients.
In this interview, Jacobs and O’Connor answer questions regarding their IEEE Smart Grid webinar. To view this webinar on-demand, click here.
QUESTION: Where can we get more information on the DER maps in California?
The three largest investor-owned utilities provide DER maps.
- Pacific Gas & Electric: http://www.pge.com/en/b2b/energysupply/wholesaleelectricsuppliersolicitation/PVRFO/pvmap/index.page (registration required)
- Southern California Edison: http://on.sce.com/derim (forwards to ArcGIS site)
- San Diego Gas & Electric: http://www.sdge.com/generation-interconnections/interconnection-information-and-map (registration required)
How can EVs help reduce load during the afternoon or early evening peak periods?
EVs by themselves do not reduce late afternoon peak periods, and could increase them if “dumb charging” is used. However, there are two ways in which EVs can be used to minimize impacts or even provide a net reduction in late afternoon peaks.
In the first case, with “smart charging,” we generally avoid charging the EV during that peak period, although the consumer can override that limitation if required. The EV might instead charge late at night or in the morning hours. EVs already have enough “smart” capability to schedule charging. In order to encourage the use of that capability, energy providers may employ time-of-use rates to incentivize off-peak charging.
In the second case, with “vehicle-to-grid,” the EV charges on inexpensive power during the day, preferably clean power. Here, the surplus solar that is causing the “duck curve” and negative wholesale prices can be turned to our benefit, especially with workplace charging. The EV then returns a portion of this power to the grid during the peak period, retaining enough charge to meet the user’s stated needs. This power could be used back at home during the evening peak, with the vehicle replenishing its charge overnight. In this case, the price differential would have to be enough to justify 1) the energy losses from charging and 2) the added wear on the battery from additional charging cycles.
Is V2G a premature idea because its practice will accelerate the increase in the rate of the number of charge cycles?
V2G does cause additional cycles and thereby increases wear on the battery. Research continues on how different types of cycles impact battery life. For example, Scott Peterson’s Carnegie Mellon dissertation finds that galvanostatic discharge for V2G may be less damaging than actual driving behavior. Still, there is wear on the battery, which carries an economic cost, so it is important to identify use cases where the value gained by V2G exceeds the cost.
Currently, V2G provides relatively high-value regulation services in the PJM and California ISO regional markets. Another high-value application is backup power, which is normally called “V2H” or “V2B” for “vehicle-to-home” or “vehicle-to-building.” Nissan has a “LEAF-to-Home” option in Japan. Typically a few essential loads (such as refrigerators, lights, and security systems) would be on a circuit that an EV could supply, while larger loads such as central AC or electric stoves would be beyond the system’s capability.
As the cost of batteries comes down, the cost of V2G comes down, and so the concept will be economically viable in a broader range of scenarios.
What is the business case with V2G and how do you compensate the user for battery degradation?
As noted, regulation and backup power are two of the business cases in certain markets. As batteries become less expensive and/or more durable, more applications will present a viable business case.
Compensation for battery degradation means ensuring that the vehicle only bids in to provide services at a level outweighing the cost. It would be important for a V2G provider to disclose to EV owners the best information on costs. Currently, V2G in the regulation market is done by entities that own the fleets of participating vehicles. If a third party were to use others’ vehicles for V2G, disclosure and owner choice would be of paramount importance.
Suppose a V2G service provider aggregates a number of EVs into a resource large enough to bid into an energy-services market. It discloses to owners that cycling power will shorten the battery’s life, provides its calculations, and gives participating EV owners a default value. It uses this value to establish its bid price for regulation, demand response, or other services. In this case, suppose that a battery is estimated to last 1000 cycles, and cost $200/kWh to replace (considering the “battery second life” resale value as discussed in the next questions). The V2G service provider sets the default at $0.20/kWh, where participating owners can change that if they wish. A lower value means the EV will have a lower bid price and will participate more often in the markets. It will experience more cycling, but will earn more revenue. A higher value chosen by a conservative EV owner means their vehicle will not be called upon as often to provide services.
Longer cycle life or lower replacement cost would allow V2G in more situations.
Would dedicated large-scale grid batteries be a better solution for batteries?
All storage systems have some round-trip energy losses. That’s just the Second Law of Thermodynamics. The real question is the economics.
An EV battery used for V2G has several disadvantages compared to a dedicated large-scale battery. It is not optimized for providing grid services, must reserve enough power for its primary use (transportation), is subject to road vibration, and connects to the grid at various places using various chargers. On the other hand, it is has one major advantage: it is largely paid for by its primary use. Only the communication/control infrastructure and the incremental cost of the wear on the battery must be offset by the value of services provided, not the entire capital cost. Does this win out over stationary batteries? It is unknown, and likely to vary by application. However, besides V2G, two other configurations are important to consider.
First, “battery second life” uses batteries that have been removed from EVs due to reaching the end of their useful vehicle life. These batteries may have 70-80% of their maximum charge capacity remaining, but in a 200-mile EV that may translate to a loss of 60 miles of range. These batteries will have a relatively low cost. BMW employed a stack of used EV batteries in its ChargeForward pilot.
Second, “smart charging,” as opposed to V2G, is using a flexible load as “virtual generation.” This does not adversely impact battery life, and so it avoids that marginal cost. Imagine a future with one million EVs in California using workplace charging at a rate of 3 kW, on level 2 chargers that can vary between 0 and 6 kW. These systems could effectively provide 3 GW of swing capacity in either direction. Electric water heaters, air conditioners, and other flexible loads can also contribute. Flexible loads will compete with dedicated batteries as a means of integrating renewables. This is good news for renewables, as it will reduce their integration costs. This is not such good news for dedicated storage, as it will limit the revenues that can be earned for this service.
How can the utility planner ensure that there will be enough EVs connected to the grid at any given point in time?
With low numbers of EVs trying to provide grid services through smart charging, a resource aggregator trying to provide grid services to an ISO or a utility has to be conservative or have a backup plan. As noted, the PG&E/BMW pilot used a stack of “second life” batteries to provide power in the event that not enough vehicles were charging for the required demand reduction. In other cases, the resource simply would not bid in if not enough vehicles were participating. A pilot project using electric truck charging for regulation services in ERCOT required 10 out of its 11 trucks to be plugged in and charging in order to meet the minimum size threshold for market participation.
These issues become less significant with more EVs on the grid.
With 20 EVs, and a 50% chance of any EV being plugged in and charging at 5 kW, you may think that on average you have 50 kW of load (20 * 50% * 5 kW). However, with such a small fleet, your standard error is quite high; there is a 25% chance of having 40% or fewer of your vehicles charging. If you have 100 vehicles, there is only a 3% chance of having 40% of fewer of your vehicles charging. Larger numbers of vehicles allow much greater use of predictive analytics.
For impacts on the system load, using averages from large numbers of EVs may be sufficient. For local impacts, there may still be a need for knowing specifically which EVs are charging at what power level.
How would using local charging at off-peak hours into local storage impact the load curves?
Local storage can be used to smooth out load curves, charging to absorb surpluses (such as from home solar generation that could otherwise cause backflow of power) or discharging to shave peaks. Stationary storage can be coupled with an EV charger to reduce the strain on the distribution system, or reduce demand charges if these are present. This configuration is used by GreenCharge and ChargePoint, among others. Local distributed storage is well-suited to dealing with local network peaks rather than grid-wide system peaks.
The impact of stationary storage on load curves depends heavily on the charging algorithms used, the tariff structure, ancillary markets, and other opportunities for providing value. While storage smooth out load curves, it is not guaranteed to do so.
How much is the cost per kWh difference between peak and off-peak hours?
The cost difference varies considerably by service territory, but the difference is often a factor of two or more. Below are a few examples. We are interested to see how these will change in response to the mid-day supplies of solar and the “duck curve.”
- PG&E has several time-of-use tariffs. One plan has low-usage (Tier 1) residential rates for May-October at 15¢/kWh off-peak and 34.2¢/kWh on-peak (1 pm – 7 pm weekdays). The peak period will shift to 4 pm – 9 pm by 2022.
- SCE also has several time-of-use tariffs. One plan ranges from 12¢/kWh for “Super Off-Peak” (10 pm – 8 am) to 48¢/kWh for peak periods (2 pm – 8 pm weekdays).
- Duke Energy has a plan in North Carolina that varies from 6.5¢kWh for off-peak to 23.3¢/kWh for on-peak (1 pm – 6 pm weekdays) in the summer.
Wholesale prices can show even greater variation, sometimes going negative and sometimes spiking extremely high.
Are there any studies that have provided the value of ancillary services?
There are extensive markets for these services in different ISO territories. If the value for a particular service is high, companies will compete to provide that service at a lower cost. Storage and flexible loads are both seen as being able to provide regulation services in the lucrative PJM market, but if all of the planned projects get built this market will be saturated. Still, the demand for certain services may increase with additional variable generation such as wind and solar.
E3’s report for the California Electric Transportation Coalition identified the value of several grid benefits of EVs, including grid integration of variable renewables. Willett Kempton of the University of Delaware has done extensive work in this area. The National Renewable Energy Laboratory has looked at the value of EVs to the grid. The Rocky Mountain Institute has a series of papers examining the value of ancillary services provided by flexible loads or battery storage. Many other examples exist. It is important to keep in mind that the value of ancillary services changes as new market entrants compete to provide those services at lower cost.