Utility-Scale Energy Storage: When Free Isn’t Cheap Enough
image credit: Abbot, COVID Lockdown Graphics Dept. -- Organic Energy Storage
- Jul 10, 2020 8:30 pm GMTJul 10, 2020 8:45 pm GMT
- 7461 views
This item is part of the Special Issue - 2020-07 - Energy Storage, click here for more
Utility-Scale Energy Storage: When Free Isn’t Cheap Enough
Charles Botsford, P.E., Monrovia, California
Why Do We Need Energy Storage?
Skeptics of renewable energy quote the old adage “you don’t make power when the sun doesn’t shine or the wind doesn’t blow.” The skeptics label renewables as intermittent and non-dispatchable. All of this is true for solar and wind generation. Renewable energy sources will require a measure of energy storage to overcome these arguments and increase their value to the grid—at some point in the far future to achieve high renewables penetration. The trick is to do this so that the economics don’t suffer.
The grid requires the precise amount of energy to be generated as is being consumed, which makes generation and consumption a constant balancing act. Traditional power plants provide the generation and the capability to ramp up and down to meet demand. In a future without these baseload power plants, energy storage will be necessary to balance generation and consumption.
source: California Independent System Operator, w/labels from M. Zerega
Figure 1. 24-Hour Grid Energy Supply and Consumption
How much energy storage, exactly, do we need to have 100% renewable energy generation at some point in the future? Lots…and lots and lots. Or, again, that’s what skeptics of renewable energy would say. It’s more complicated than that and primarily has to do with capacity factor…and geography and wide grid balancing and coincident load management and many other things. For solar, the capacity factor is low (15-30%), while for land-based wind it’s medium (30-45%), and for off-shore wind it’s relatively high (40-55%). For example, the Hywind floating offshore windfarm off the coast of Scotland has reported a 12-month rolling average of 55.3% [1, 2].
For perspective, baseload generation in the form of the US nuclear fleet is on the order of 93% capacity factor. The other primary form of baseload generation in the US is fossil fuel with natural gas at about 60% (and rising over time) and coal at 48% (and dropping over time) . The low fossil fuel capacity factors, especially for coal is market-based. A large portion of the US coal fleet lies idle because of high operating costs.
Some argue that the amount of energy storage required to enable a renewable source to provide value to the grid is inversely proportional to its capacity factor. This guidance would qualitatively say that energy storage is important to solar, while not as important to floating offshore wind—again, at some point in the far future where energy storage is needed for high renewables grid penetration.
Services that Energy Storage Systems Deliver
Energy storage systems deliver services to the grid and off-grid. The services listed below depend on the market, whether for renewables integration, microgrids or for standalone systems.
- Peak shifting
- Peak shaving
- Supply capacity and firming
- Frequency regulation
- Spinning, non-spinning, and supplemental reserves
- Reactive supply and voltage control
- Transmission upgrade deferral
- Congestion relief
- Distribution upgrade deferral
- Power quality
- Power reliability
- Demand charge management
- Demand response
- Time-of-use management
Electricity markets value these services to a varying degree. For example, frequency regulation is highly valued in many markets because when grid frequency goes out of range, electrical equipment can be damaged. Battery energy storage has the fast response time necessary to provide this service [4-9].
Frequency regulation is an example of a service that provides response on the order of seconds to keep the grid at 60 Hz +/- 0.05 constantly. Traditional power generation handles this via “regulating reserves”, which provide fast up and down balancing services. Without traditional power generation, i.e., 100% renewables, some would speculate, the grid needs energy storage to provide this service. Below is a graph of frequency regulation using unidirectional electric vehicle supply equipment (EVSE) to control the EV on-board charger autonomously by monitoring the grid. This is sometimes called grid-to-vehicle (G2V or V1G) managed charging.
source: Brooks, EVS27 paper .
Figure 2. Typical grid frequency variation in the US western interconnect over a 30-min period.
Other grid services rely on sub-hourly energy markets, which operate on time steps of five minutes . Likewise, the grid requires myriad services as illustrated in the above list, not just the ride-through needed for a one-hour power outage, a four-hour emergency, or a three-week disaster such as a hurricane.
source: Abbot, COVID Lockdown Graphics Dept.
Figure 3. Southern California Windfarm with Energy Storage
Stationary Utility-Scale Battery-Based System Economics
Batteries, in particular lithium-ion batteries, are reported to be the most cost-effective stationary utility-scale energy storage systems—for storage durations up to four hours. For longer duration storage requirements, compressed air energy systems and pumped hydro systems are reported to be cost-competitive . This is primarily due to the “balance of plant” or “balance of system”, which includes inverters, converters, cabling, housing, contactors, controllers, transformers, and other components. For shorter durations, the Power Cost (balance of plant) portion of the energy system cost, measured in dollars per kilowatt ($/kW) is quite high compared to the Energy Cost (battery, dam, chamber) portion, measured in dollars per kilowatt-hour ($/kWh) [13-16].
The U.S. National Renewable Energy Laboratory (NREL) uses a simple equation to compare system costs :
Total Cost ($/kWh) = Energy Cost ($/kWh) + Power Cost ($/kW) / Duration (hr)
Many years ago, the cost of battery energy storage was $2,000/kWh. This cost was almost entirely due to the battery, not the balance-of-plant. Then the cost decreased to $1,000/kWh about a decade ago. The most recent system cost estimates are $200-400/kWh. Future systems cost projections vary widely, but some studies show a future cost on the order of $100/kWh in 2030.
However, as the capital costs decrease, other costs remain, the majority of which are not technology-based, and have little room for cost reduction:
- Engineering, procurement & construction (developer cost)
- Installation labor
- Operation and maintenance (O&M)
- Battery degradation, which depends on kWh throughput
What Does the Cost of Stationary Energy Storage Need to Be?
The $200-400/kWh is the rough cost range of current utility-scale stationary battery energy storage systems for a four-hour duration system. Systems generally cost more for shorter durations because the balance-of-plant costs dominate. The duration factor also gets to the heart of the question of what the cost needs to be for the different services and the markets that energy storage systems serve.
What would the market costs need to be to provide firming and other services for a 100% renewables grid? Again the estimates vary widely—from $5/kWh to $60/kWh to $150/kWh. The studies are numerous as are the assumptions [18-22]. Let’s say the answer is a market cost of $60/kWh. If the 2030 energy storage system cost projections are $100/kWh, then the cost doesn’t meet the market price needs.
Would free battery systems be a better economic fit?
Why is free important?
Free battery systems, in the form of electric vehicles (EVs), are currently available for use by the grid. This style of energy storage is commonly known as managed charging, or smart charging and involves charging and discharging EV batteries when the grid requires their services [23-26]. The systems are free in the sense that the capital cost is paid for by the EV owner (e.g., the EV driver, fleet owner, car share company, etc.) So are the engineering costs, installation costs, O&M, financing, and taxes. The services themselves aren’t free, but the same can be said of stationary utility-scale energy storage.
EV Battery-Based System Economics
The EV owner pays for the vehicle, which comes with the storage, and in the future, the bi-directional inverter. The EV owner will be incentivized to participate in the market via payment for services, but many will no doubt opt out because: (1) they don’t want to be bothered, (2) fear battery degradation, and (3) fear outside control of their vehicle being charged/discharged. Advantages of EV energy storage over stationary utility-scale battery energy storage are:
- Duration – In a future fleet of 20 million EVs, the duration is virtually unlimited
- Degradation – No incremental cost for battery degradation (used EV batteries become retired)
- Operation relative to markets – highly flexible, improved grid resiliency and reliability
- Other Factors and Costs – No O&M, no financing, no taxes
How to Put Power onto the Grid – AC and DC
For distributed energy resources (DERs), putting power onto the grid requires compliance with the standard International Electrical and Electronics Engineers (IEEE) 1547. The most recent version is IEEE 1547-2018. To certify that the DER complies with IEEE 1547-2018 it must be tested according to the procedures laid out in IEEE 1547.1-2020.
Why is this important? The process is well established for stationary DERs, including big utility scale batteries and even EVs that use stationary bi-directional direct current (DC). This is called vehicle-to-grid (V2G). However, EVs in the U.S. that wish to use their on-board charger to provide alternating current (AC) V2G services must comply with Society of Automotive Engineers (SAE) J3072, which references IEEE 1547-2018 and 1547.1-2020. J3072 is still in development.
Do We Need to Put Power onto the Grid?
What about exporting power from the EV to the grid? In bulk, say with millions of EVs, managed unidirectional charging, sometimes called smart charging (also known as G2V and V1G), just uses the energy storage of an EV as a reservoir for the grid. Services include up and down frequency regulation, demand response, peak shaving, peak shifting, renewables firming, and many others. All EVs in the North American market are capable of unidirectional managed charging. However, very few EVs are currently capable of V2G services because they aren’t equipped with a bi-directional on-board charger, and SAE J3072 isn’t yet in place. Almost all grid services done by V2G can be done by unidirectional managed charging, and it avoids the V2G issues of:
- roundtrip efficiency loss,
- grid interconnect protocols,
- potential EV pack degradation (not something EV drivers want), and
- complex trading schemes for revenue capture.
However, that’s for bulk services. For services closer to home where one EV sits in a carport, or at a business where a hundred EVs sit in the parking lot, V2G can be very useful [27, 28]. V2G services can include vehicle-to-home (V2H) and vehicle-to-business (V2B), which aren’t the traditional power export to the grid of V2G, but are valuable in the case of emergency power requirements including:
- public safety power shutoffs or PSPS, which is an issue in California when a utility needs to curtail power for wildfire reasons,
- earthquakes, again a major West Coast issue,
- hurricanes, Gulf and East Coasts, and
- polar vortices and bomb cyclones, definitely not California.
In the early days of EVs, when a pack might only be 24 kWh, the available storage wasn’t nearly as bountiful as current EVs with 60 or even a 100 kWh pack. A house on minimal average power draw of 3 kW during an emergency, could be served for more than a day by one EV with an 80 kWh pack.
The EV, whether for home or business non-export power, would typically need to use a power converter to supply its energy. However, the same is true for stationary energy storage, such as residential battery systems, which have only 10-15 kWh of storage.
One of the first vehicle-to-grid demonstration programs was conducted by AC Propulsion in the 2001 timeframe. The program was sponsored by California Air Resources board . More recently, the California Vehicle Grid Integration (VGI) Working Group submitted their final report to the California Public Utility Commission, which details methods, priorities, and policy recommendations for making use of unidirectional and V2G charging .
How Much Energy Storage Do We Need?
The amount of energy storage the U.S. needs is a decade-dependent question—i.e., 2020, 2030, 2040, 2050. Currently, in 2020, the U.S. doesn’t need much energy storage, relatively, because the amount of renewables penetration onto the grid is low in most markets. Standalone energy storage systems aren’t typically competitive except for frequency regulation, which is a small overall market (e.g., PJM), and demand charge management, which is a highly niche market (e.g., Southern California). As the percentage of renewables penetration increases, and fossil fuel power plants retire (i.e., coal, gas), the amount of required energy storage will increase.
2030 Power Projections for the U.S.
According to the Energy Information Administration (EIA) , by 2030 the U.S. will produce approximately 1,100 GWh of renewable electricity, with an approximate equal split between hydro, solar, and wind. EIA  estimates the current (2020) fleet capacity factors for solar at ~25%, wind at ~35%, and hydro at ~40%, or a bulk renewable 2020 capacity factor of 33%. This translates to about 380 GW of renewables by 2030. Another data point from Global Data  projects cumulative U.S. renewables at 443 GW.
If a four-hour duration were assumed, then the energy storage requirement would be ~1,600 GWh. This is a conservatively high estimate because the fleet capacity factors for solar and wind will be higher by 2030. Some would argue that a four-hour duration is too short. However, this does not consider market drivers such as the energy imbalance market (EIM), which is a cooperative effort between U.S. Balancing Authorities (BAs). The California Independent System Operator (CAISO) is the largest BA and created the EIM to minimize curtailment of renewables, which can greatly enable the penetration of renewables [34, 35]. A more likely energy storage requirement would be on the order of 1,000 GWh (1 TWh) or less. A study by Cal Berkeley  estimates 150 GW of four-hour battery energy storage, or 600 GWh, would be sufficient to meet grid demands with 90% renewables (Clean Grid) by 2035.
Stationary Utility-Scale Battery Energy Storage Cumulative Cost
At $60/kWh, the cumulative capital investment for 600 GWh would be on the order of $36B.
The above does not other costs, which include installation, O&M, taxes, battery degradation, developer costs, etc.
By 2030, the population range of EVs in the U.S. is projected to be in the range of 15-25M, depending on the study . For an average usable pack capacity of 50 kWh, this translates to 750-1,250 GWh of EV battery storage. This likely wouldn’t be enough to completely offset the energy storage requirement, considering EV driver opt-out and other factors. However, light duty pack capacity will likely be higher, probably at least 100kWh, and trucks, buses, and other medium- and heavy-duty vehicles will add major storage capacity.
Many studies predict that managed EV charging by 2030 will significantly offset potential renewables curtailment, while deferring grid system infrastructure expansion .
EV Battery Energy Storage Cumulative Cost
At $0/kWh, the cumulative capital investment would be on the order of $0 B.
Energy Storage Market Projections – Why EV Energy Storage Will Dominate
Whatever shortfalls EV energy storage might encounter, could be made up via second use of EV packs, also called EV pack retirement. This is the subject of many studies and solicitations . While a used pack may not have sufficient capacity for an EV, typically defined as 70 percent degradation, it still has many years of life for bulk energy storage. Many of the stationary utility-scale battery energy storage costs, such as balance-of-plant would still pertain, however, the battery cost portion would be minimal, possibly even negative [40, 41].
This partially addresses the issue of what to do with retired EV packs. Recycling and disposal has been the subject of many studies.
Will the U.S. and the rest of the world continue to install stationary utility-scale battery energy storage to support renewables? Yes, for as long as economic reward continues for investment. This trend will prove the viability of solar and wind projects at ever-increasing scale, and demonstrate to skeptics that allowing the retirement of natural gas and coal power plants won’t leave people in the dark, and in fact, be great for the environment.
Energy storage is needed for timeframes on the order of:
- Seconds for frequency regulation
- Minutes for sub-hourly energy markets
- Hours for peak shaving, peak shifting, and other services, and
- Days/weeks for disasters
Can stationary utility-scale battery energy storage meet the economic requirements to do this in 2030, 2040, or 2050? As EVs come of age, a low cost, robust energy storage alternative will emerge to make this question moot. Will this be an easy road? No. Besides the momentum and economic interests that favor stationary storage, primarily due to rate-based utility economics, a lot of work policy and regulatory work with public utility commissions remains to allow for the full potential of EVs to deliver their services to the grid.
The economics of renewables, and the grid value of renewables will only improve when the mass of low/zero cost energy storage from EVs hit the market.
1. Andrew ZP Smith, ORCID: 0000-0003-3289-2237; UK offshore wind capacity factors. Retrieved from https://energynumbers.info/uk-offshore-wind-capacity-factors, accessed June 18, 2020.
2. Energy Information Administration, U.S. Renewable Energy Consumption Surpasses Coal for the First Time in over 130 Years, www.eia.gov, May 28, 2020.
3. Energy Information Administration, Electric Power Monthly, Table 6.07.A. Capacity Factors for Utility Scale Generators Primarily Using Fossil Fuels https://www.eia.gov/electricity/monthly/epm_table_grapher.php?t=epmt_6_07_a
4. Finkelstein, J, S. Kane, and M Rogers, McKinsey & Company, How Residential Energy Storage Could Help Support The Power Grid, March 2019.
5. Cooper, H., S. Brauer, and D. Kalow, Battery Energy Storage in Stationary Applications, Chemical Engineering Progress, May 2020.
6. Balaraman, K., California CCAs Solicit Info on Long Duration Storage, with Possible Procurement Launch this Summer, Utility Dive, June 9, 2020.
7. IRENA, Electricity Storage and Renewables: Costs and Markets to 2030, International Renewable Energy Agency, Abu Dhabi, 2017.
8. Energy Information Administration, Large Battery Systems are Often Paired with Renewable Power Plants, www.eia.gov, May 18, 2020.
9. Energy Information Administration, Most of Hawaii’s Electric Battery Systems are Paired with Wind or Solar Power Plants, www.eia.gov, March 20, 2020.
10. Brooks, A., Vehicle Charging as a Source of Grid Frequency Regulation. EVS27, Barcelona, Spain. November 2013.
11. US Department of Energy, https://www.nrel.gov/docs/fy11osti/50169.pdf, DOE/GO-102011-3207, May, 2011.
12. Mongrid, K., etal, Energy Storage Technology and Cost Characterization, Report, HydroWire, US Department of Energy, Pacific Northwest National Laboratory, PNNL-28866, July 2019.
13. Caiazza, R., NREL Energy Storage System Cost Benchmark, Wattsupwiththat.com, July 16, 2019.
14. Weaver, J., Utility Scale Solar Power Plus Lithium Ion Storage Cost Breakdown, PV-Magazine-USA, January 2, 2019.
15. Lazard, Lazard’s Levelized Cost of Storage Analysis—Version 4.0, November 2018
16. Cole, Wesley, and A. Will Frazier. 2019. Cost Projections for Utility-Scale Battery Storage. Golden, CO: National Renewable Energy Laboratory. NREL/TP-6A20-73222. https://www.nrel.gov/docs/fy19osti/73222.pdf.
17. Fu, Ran, Timothy Remo, and Robert Margolis. 2018. 2018 U.S. Utility-Scale Photovoltaics-Plus-Energy Storage System Costs Benchmark. Golden, CO: National Renewable Energy Laboratory. NREL/TP-6A20-71714. https://www.nrel.gov/docs/fy19osti/71714.pdf.
18. Patel, P., How Inexpensive Must Energy Storage Be for Utilities to Switch to 100 Percent Renewables, IEEE Spectrum, September 16, 2019.
19. Roberts, D., Getting to 100% Renewables Requires Cheap Energy Storage. But How Cheap?, Vox.com, September 20, 2019.
20. Colthorpe, A., BloombergNEF: “Already Cheaper to Install New-Build Battery Storage Than Peaking Plants”, Energy Storage News, April 30, 2020.
21. Colthorpe, A., Behind the Numbers: The Rapidly Falling LCOE of Energy Storage, Energy Storage News, May 6, 2020.
22. Martin, J., Solar Hits New Milestones as Renewables Become World’s Cheapest Choice, PV-Tech, April 29, 2020.
23. Shahan, Z., This Stunning Chart Shows Why Electric Vehicles Win, CleanTechnica, June 10, 2020.
24. IRENA, Innovation Outlook: Smart Charging for Electric Vehicles, International Renewable Energy Agency, 2019.
25. Dayem, K., C. Mercier, and P. May-Ostendorp, Electric Vehicle Charging Control Strategies. Xergy Consulting sponsored by National Rural Electric Cooperative Association (NRECA), January 2019.
26. Enel X, Enel X_Smart Charge Hawaii Residential Case Study, 2020.
27. Chhaya, S., etal, “Distribution System Constrained Vehicle-to-Grid Services for Improved Grid Stability and Reliability”, Final Project Report, California Energy Commission, Publication Number: CEC-500-2019-027, March 2019.
28. Szinai, Julia, Colin J.R. Sheppard, Nikit Abhyankar, Anand Gopal. Reduced grid operating costs and renewable energy curtailment with electric vehicle charge management, Energy Policy (2019) https://doi.org/10.1016/j.enpol.2019.111051, Nov. 2019
29. Brooks, A., Vehicle-to-Grid Demonstration Project: Grid Regulation Ancillary Service with a Battery Electric Vehicle, California Air Resources Board Contract Number 01-313, December 2002.
30. Vehicle-Grid Integration Working Group, Final Report to the California Joint Agencies, CPUC DRIVE OIR Rulemaking R.18-12-006, June 30, 2020.
31. Energy Information Administration, Annual Energy Outlook 2019 with projections to 2050, www.eia.gov/aeo, January 2019.
32. Energy Information Administration, Electric Power Monthly, Table 6.07.B. Capacity Factors for Utility Scale Generators Primarily Using Non-Fossil Fuels, https://www.eia.gov/electricity/monthly/epm_table_grapher.php?t=epmt_6_07_b
33. Global Data, US Power Market Outlook to 2030, Update 2019 – Market Trends, Regulations and Competitive Landscape, July 2019.
34. Larson, A., How Does the Western Energy Imbalance Market Work?, Power, www.powermag.com, October 2018.
35. California Independent System Operator, Western Imbalance Benefits Report, July 31, 2017.
36. University of California Berkeley, Goldman School of Public Policy, 2035, The Report, Plummeting Solar, Wind, and Battery Costs Can Accelerate Our Clean Electricity Future, June 2020.
37. IRENA, Electric Vehicles: Technology Brief, International Renewable Energy Agency, 2017.
38. Sheppard, C, J. Szinai, N. Abhyankar, A. Gopal, Grid Impacts of Electric Vehicles and Managed Charging in California, Energy Analysis and Environmental Impacts Division, Lawrence Berkeley National Laboratory, November 2019.
39. California Energy Commission, Validating Capability of Second-life Batteries to Cost-Effectively Integrate Solar Power for Small/Medium-sized Commercial Building Applications, GFO-19-310 Solicitation, http://www.energy.ca.gov/contracts/index.html, February 2020.
40. Engel, H., P. Hertzke, and G. Siccardo, McKinsey & Company, Second-Life EV Batteries: The Newest Value Pool in Energy Storage, April 2019.
41. Anderson, M., Used EV Batteries Could Power Tomorrow’s Solar Farms, IEEE Spectrum, June 10, 2020.
Charles Botsford, PE is a professional chemical engineer in the State of California with 30 years’ experience in engineering process design, distributed generation, EV charging infrastructure, and environmental management. He has participated in California’s Vehicle Grid Integration (VGI) Working Group and participates in the Society of Automotive Engineers (SAE) J3072 AC Vehicle-to-Grid standards committee. Mr. Botsford holds a bachelor’s degree in chemical engineering from the University of New Mexico, and a master’s degree in chemical engineering from the University of Arizona.