Analysis of Automatic Control Data for Storage Operations
Bulk energy storage devices and fleets of electric vehicles are being proposed as a resource for regulation services. To use energy storage for regulation services, energy storage is expected to operate seamlessly, in concert with existing regulation resources and existing regulation operation processes. This project identified, collected, and initially analyzed operating data from the California Independent System Operator (ISO) to address the use of energy storage for regulation services.
While a proposed regulation correction algorithm is applied to the collected operating data, in no way is the algorithm a certified algorithm for actual use in a production system. The collected operating data covers a full week (June 6 through June 12, 2010). The depth and breadth of the identified and collected operating data was selected to facilitate current and potential future work.
This analysis is important in leading towards a future where battery energy storage is utilized for regulation services. The documented regulation correction algorithm is leading the way for this important application. Additional analyses, including valuation and portfolio analyses, are suggested. The valuation analysis is a data mining exercise to identify specific system conditions that are correlated with “stressful” operating conditions.
Key takeaways from the report are:
This report shows how energy storage could sustainably provide Regulation service by leveraging real-time energy dispatch, and it is based on actual 4-second EMS data from the CAISO.
This study found that the provided regulation signal contains a significant negative energy bias of approximately -15 GWh over 7 days. Thus, for an energy storage system to provide all regulation services needed for this balancing area, the system would require approximately 30 GWh of storage capacity. Conversely, the corrected regulation signal has both more symmetric energy content and a significantly lower energy capacity requirement (approximately 300 MWh), potentially a 100x reduction in the energy storage requirement.
Additional results include variability, revenue, and cost analysis. The variability analysis shows how fast the different resources move in terms of several durations (from 8-second duration to 4-hour duration). The revenue and cost analysis compares regulation correction results in terms of real-time energy and regulation energy cost.
The framework for variability analysis, by technology type, could be useful for estimating the sources of imbalance (variability) that lead to the need for regulation service and other proposed Load Following services. Data for expected and actual production and load are analyzed. The results show that there are significant differences in expected and actual energy from Interchanges and wind resources not scheduled by the CAISO (non-PIRP).
For more information:
Electric Power Research Institute (EPRI) Address: 3420 Hillview Avenue Palo Alto, California United States, 94304 www.epri.com