Opportunities and Challenges for Portland General Electric’s Salem Smart Power Project (SSPP):
Energy storage is rapidly evolving and offers much promise to improve power system operations, but it also faces significant economic and regulatory challenges. The recent and rapid growth of wind and solar generation is one important driver of interest in energy storage because storage could help to mitigate the impact to the system of integrating intermittent generation resources.
But energy storage has many other potential benefits. A recent study by the staff at the California Public Utilities Commission (CPUC) identified 20 distinct services that might be provided by energy storage. However, energy storage—except for pumped hydroelectric storage, which would face environmental and siting concerns—remains a questionable economic option. Portland General Electric Co. (PGE) is among utilities that have been evaluating battery storage, while other companies have experimented with flywheel, compressed air underground and other forms of storage.
This paper analyzes potential benefits associated with PGE’s new 5-MW battery storage facility (the Salem Smart Power Project or SSPP). The CPUC staff study grouped the 20 different uses for energy storage into four scenarios: renewable support/dispatchability; distributed storage; demand-side management; and ancillary services. In this paper, we use these four scenarios as a framework to analyze potential benefits from the SSPP.
The paper begins with a Technical Analysis—a discussion of how the capabilities of the SSPP compared to the 20 services identified by the CPUC study. In the next section—Business Analysis—the paper attempts to place a dollar value upon the services the SSPP provides. Where estimating the dollar value of a particular use was not practicable, methodologies for quantifying the economic benefit of the use are recommended. Finally, the Regulatory and Policy Analysis portion of the paper discusses federal and regional efforts underway to facilitate development of energy storage, and how those efforts might affect the SSPP.
Salem Smart Power Project (SSPP)
On May 31, 2013, PGE opened its $22 Million Salem Smart Power Project, a 5 MW/1.25 MWh lithium-ion, grid-connected battery storage facility. PGE partnered with Eaton and EnerDel, Inc., and received matching funds for the project from the US Department of Energy (DOE) and Bonneville Power.
One function of the SSPP is to provide very high reliability to customers located on the Rural Feeder distribution circuit at PGE’s Oxford substation in South Salem, Oregon. Supporting this high reliability zone (HRZ) is a primary function for the SSPP. However the facility has the energy storage capabilities necessary to provide additional benefits. Of the 1.25 MWh of energy storage available for regular cycling, PGE has reserved 500 kWh for support to the HRZ; 750 kWh remains un-reserved and available to provide other services. Some services provided by batteries require, in addition to storage capacity, advanced controls such as “automated generation control”, or “AGC”. The SSPP was built with this in mind, and has state of the art AGC capability necessary for providing spinning reserve and other ancillary services. The capabilities of the facility not dedicated to supporting the HRZ remain available for research into how batteries can support operations and enhance existing service in a cost-effective manner.
Battery Use Scenarios
Previous studies on uses of battery storage identified between 17 and 21 separate uses, and grouped those applications into between four and eight use cases. While no framework is more “correct” than another, this paper adopts the framework used in the CPUC Energy Storage Proceeding (CPUC Rulemaking 10-12-007), which identified 20 separate applications of energy storage, or “end uses”:
- Ancillary Services: frequency regulation
- Ancillary Services: spinning and non-spinning reserve
- Ancillary Services: ramping
- Black start
- Real-time energy balancing
- Energy price arbitrage
- Resource adequacy
- Intermittent resource integration (ramp/voltage support)
- Intermittent resource integration (time shift, voltage sag, rapid demand support)
- Supply firming
- Peak shaving
- Transmission peak capacity support
- Transmission operation
- Transmission congestion relief
- Distribution peak capacity support (upgrade deferral)
- Distribution operation (voltage/VAR support)
- Outage mitigation: microgrid
- Time-of-use energy cost management
- Power quality
- Back-up power
No single battery can simultaneously provide all of the services above. Some of the uses are mutually exclusive while others are closely related and likely to be compatible. The CPUC grouped the 20 uses, above, into four distinct use scenarios:
1. Renewables Support/Dispatchability (uses 8, 9, and 10). Storage is used primarily to support renewable generation. Storage is used to improve dispatchability and value of renewable generator output by smoothing, firming, and time-shifting output, principally of wind and solar, while avoiding system level integration costs.
2. Distributed Storage (uses 6, 7, 11, 15, 16, and 17). Storage is used primarily to support grid operations at the distribution level (e.g. voltage/VAR; peak shaving; load shaping; etc.) Storage may also be used to arbitrage—e.g. purchase energy during low-cost periods and sell during high cost periods.
3. Demand-side Management (uses 17,18, 19, and 20). Storage is used behind the meter of the customer to shift load. Additional services (e.g. reliability, may also be provided). As noted below, the SSPP is not configured for this use.
4. Ancillary Services (uses 1,2,3, and 5). Storage is used primarily to provide generator-like services at the transmission level for ancillary service markets.
SSPP Technical Suitability for Each Battery Use Scenario
1. Renewables Support/Dispatchability. The SSPP is potentially well suited for integrating intermittent renewable resources into PGE’s system. Such integration can be broken down into four discrete categories: (1) Day-Ahead uncertainty; (2) Hour-Ahead uncertainty; (3) Load and Load-Following for Wind; and (4) Regulation. The SSPP can provide all four functions (although the total need for such resources on PGE’s system is much greater than the capacity of the SSPP).
2. Distributed Storage. The SSPP is capable of providing all of the identified distributed storage use scenario benefits. However, the SSPP was not sited for the purpose of deferring distribution upgrades, alleviating transmission constraints, or providing VAR/voltage support; a storage facility sited in a constrained area would be better suited to provide those benefits. This paper focuses on the capability of the SSPP to reduce the total cost to meet load through energy price arbitrage.
3. Demand-side Management. The SSPP is not configured for the demand-side management (DSM) use because it is not behind a retail revenue meter. This paper does not evaluate DSM applications for storage.
4. Ancillary Services. The SSPP is capable of providing ancillary services such as spinning reserve and regulation and frequency response. The size of the SSPP might prevent PGE’s use of it for sale of some or all types of ancillary services because those services are typically bought and sold in 1-MW increments.
All cost analysis in this section are based on the best data available to the authors. The resulting estimates may serve as guides for selecting more detailed analyses.
1. Renewables Support/Dispatchability. A Sandia Labs study estimates the value of using battery storage to integrate wind at between $500 and $1,000 per kW battery capacity over a ten-year period. This suggests that the SSPP could be worth up to $5 million, if used to integrate wind. However an accurate estimate of the value of the SSPP must take into account the specific characteristics of all intermittent and non-intermittent resources on PGE’s system. PGE recently developed a methodology and models necessary to estimate the cost to integrate intermittent wind resources into its system and published results from its studies in its PGE Wind Integration Study Phase II (Phase II Wind Study). PGE’s methodology uses optimization models that dispatch PGE’s available resources (including market purchase opportunities) to serve load at the lowest cost. Adding the SSPP to PGE’s stable of resources and rerunning the Phase II Wind Study optimization models is one way PGE could calculate the savings it could expect to realize by using the SSPP to integrate intermittent resources. Based on the Sandia Report and PGE’s Phase II Wind Study, the benefits are believed to be substantial; however it is possible that the SSPP will be too small to significantly impact system operations (and resulting costs). Modeling different sizes of battery resource might suggest an optimally effective level of battery storage for renewables integration.
2. Distributed Storage. Because the SSPP is sited in a robust location on PGE’s system, benefits from deferring distribution upgrades, alleviating transmission constraints, or providing VAR/voltage support are not substantial, although such benefits could be substantial at less robust locations on the PGE system.
Perfect energy arbitrage (e.g. using the SSPP to always buy at the lowest cost hour and always sell at the highest cost hour) would yield less than $10,000 per year, based on an assumed 300 annual cycles of 1.25 MWh, and a daily price difference of $25.23/MWh—the average difference at the California-Oregon Border for each day since 2007. In short, routine daily arbitrage is not a cost-effective use of the SSPP. Using the SSPP to better optimize unit commitment—and thereby reduce unit start-up costs—may yield significantly higher value. Due to the situation-dependent nature of dispatch decisions, and due to much of the data being proprietary, the authors were unable to estimate economic benefits of this alternative strategy.
3. Demand-side Management. Because the SSPP is not situated to provide demand-side management (as defined in the CPUC framework), this paper does not attempt to quantify the value of the service.
4. Ancillary Services. Spinning Reserve service has the potential to be a cost-effective use of the SSPP. If the SSPP were allowed to use up to 833 kWh of the SSPP battery’s 1.25 MWh total, the SSPP could provide 5 MW of Spinning Reserve. PGE’s OATT specifies a price for Spinning Reserve of “up to” $6.695/kW-month. At that price the SSPP can provide annual revenue of $401,700. Using the SSPP to provide Spinning Reserve might have the added benefit of substantially prolonging its life compared to other uses requiring frequent deep charges and discharges.
Eventually, PGE might be able to sell Spinning Reserve from the SSPP at market prices. Pending FERC initiatives, discussed below, will likely create easier access to third-party markets for these ancillary services, facilitating PGE’s ability to market them to other utilities in the region. A recently released study by EPRI concludes that battery storage providing primarily ancillary services could be cost-effective under CAISO market conditions within the next seven years. Under most models in the EPRI study, regulation and frequency response service proved to be the largest source of benefit.
Regulatory and Policy Analysis
Battery storage is well suited to provide ancillary services, but the Northwest currently lacks an organized market for them. FERC has recently recognized that barriers may prevent formation of markets for ancillary services outside organized markets (RTOs or ISOs). FERC has proposed new rules that would lower the barriers with the purpose of creating wider markets for ancillary services while maintaining price protections. As of the date of this paper, the rules are still pending. If implemented, the rule would pave the way for third-party battery storage operators to provide ancillary services for transmission providers providing OATT service. FERC is also proposing modifications that would recognize the greater value of faster-responding regulation resources (such as battery storage). The result could be a reduced requirement to retain regulation and frequency response service, which is now 1.3% of network load, if that service is provided by battery storage.
In addition, proposed federal legislation would create a 20-30% investment tax credit for battery storage facilities. Whether or not that is enacted, the Internal Revenue Service recently concluded that the 30% solar investment tax credit (effective through 2016) applies to battery storage if the batteries are charged solely with energy from a solar project. Last, the likely expansion of CAISO’s energy imbalance market may be a first-step towards a WECC-wide market for energy imbalance. If such a market included PGE it could be another opportunity to market SSPP’s services. In sum, several current federal initiatives have the potential to open the market for, and increase the value proposition of, battery storage in the Northwest.
The first use scenario—renewables support/dispatchability—may be implemented and deployed without any further regulatory changes. The value of these services, measured as the reduced cost to serve load, are not known, but may be estimated PGE’s existing Wind Integration Study, Phase II, optimization models, and optimization models used in its planned Solar Integration Study.
The second use scenario—distributed storage—uses the SSPP to provide local grid support, including peak shaving. PGE did not identify any present benefits associated with deferring distribution upgrades, alleviating transmission constraints, or providing VAR/voltage support, although such benefits may be substantial in other locations. Energy arbitrage based on historic hourly prices at California-Oregon Border is not a cost-effective use of the SSPP. Using arbitrage only during super peak periods may provide much greater value per MWh of energy; however, such opportunities are sporadic and somewhat unpredictable.
The third use scenario—demand side management—is not addressed by this paper because the SSPP is not behind a retail meter. The DSM use scenario groups together services that benefit the end-use customer, either by lowering rates, improving the level of service, or both. A major driver for this use scenario in California is super peak rates, which may be as great as six times the off-peak rate. Such a price spread can make battery storage a cost-effective means of reducing rates by shifting time of use. If similar rate differentials take effect in PGE territory, the DSM use scenario may become attractive to some users. Such rates might present an opportunity for PGE to install and own batteries at the end-user site and charge a monthly fee in exchange for load shaping and other benefits the batteries can provide (e.g. reliability, quality of service).
The fourth use scenario—ancillary services—has the potential to be cost-effective for battery storage. Estimates of the value using the SSPP for spinning reserve run as high as $401,700 per year. Another type of ancillary service: regulation and frequency response, appears likely to become an attractive application for battery storage in the future. Additional opportunities for marketing ancillary services are expected to arise as federal initiatives, summarized above, become reality. Such efforts are expected to lead to more liquid markets for ancillary services which generators (and energy storage facilities) can bid into. In the event FERC finalizes its proposed rule to require transmission providers to differentiate between regulation and frequency response services provided by faster-responding generation, such as battery storage, the SSPP might find its highest value when providing such services. Table 1 summarizes the SSPP’s suitability for various uses .
 Faculty Team Lead.
 Energy Storage Framework Staff Proposal, CPUC Energy Storage Proceeding R.10-12-007, p. 13 (April 3, 2012).
 The SSPP is part of the Pacific Northwest Smart Grid Demonstration Project. Portland State University, Pacific Northwest National Labs, the University of Colorado Research and potentially others will conduct research at the facility. http://www.businesswire.com/news/home/20130531005504/en/Portland-General-Electric-Opens-Salem-Smart-Power
 Energy Storage for the Electricity Grid: Benefits and Market Potential Assessment Guide–A Study for the DOE Energy Storage Systems Program, Sandia National Laboratories, p. xix, February 2010 (SAND2010-0815) (“Sandia Report”).
 PGE filed its Phase II Wind Study with the Oregon Public Utilities Commission as part of its 2011 update to its 2009 Integrated Resource Plan. The full report can be viewed at: http://www.uwig.org/PGE_Study/PGE_Phase%202_Wind_Integration_Report_9-30-11.pdf
 Cost-Effectiveness of Energy Storage in California, Application of the EPRI Energy Storage Valuation Tool to Inform the California Public Utility Commission Proceeding R. 10-12-007, Technical Update (June 2013) (“EPRI Study”).
 Third-Party Provision of Ancillary Services; Accounting and Financial Reporting for New Electric Storage Technologies, 139 FERC ¶ 61,245 (2012).
 S.1030 (20%); H.R. 1465 (30%).
 IRS PLR-121432-12
 CAISO Filing of ISO Rate Schedule No. 73, Docket No. ER13-1372-000 (April 30, 2013).