The traditional way of modeling battery storage as pumped hydro in an integrated resource plan is outdated. Current and future energy storage technologies provide multiple benefits that can be stacked on top of each other. IRP models must analyze these multiple benefits since capacity credit values decrease as renewable penetration increases. Because of these multiple benefits, hourly models in IRPs must be replaced by sub-hourly models. Resource planners cannot afford to look at future capacity needs in hourly increments. Neither can these resource planners model IRPs in silos relative to transmission and distribution planning. It is time for IRP modelers to innovate and update their modeling.
Join us on Feb. 6, 2023 in San Diego for the inaugural GridTECH Connect Forum, the only event that brings together utility leaders and distributed energy resource developers to overcome the challenges associated with interconnection. Learn more about the new, regional event from Clarion Energy today.
Join us for discussions on virtual power plants, fleet electrification, interconnection, demand response, and more, with a focus on the California market. Register today!
Recent IRPs include a minimal amount of energy storage
It is well known that states are in the driver’s seat of the resource adequacy vehicle and each state has its Integrated Resource Planning process. After the passage of the Inflation Reduction Act (IRA), which includes tax incentives for stand-alone storage, the modeling of storage in IRP software models must be improved to accurately model storage as a future capacity resource.
Recently Detroit Edison (DTE), Xcel Energy, Minnesota power, and several Midwest utilities have filed IRPs that included energy storage, but that is the tip of the iceberg. Only DTE’s was filed after the IRA passage among those investor-owned utilities, which included 1,800 MW of battery storage (760 MW in 2023-2032 and the remaining in 2032-2042).
IRPs reflect the dynamic nature of capacity credit values
Energy storage is now a viable alternative to meet capacity needs since it has proven its worth in states like California and Texas. And the generator interconnection queues at grid operators reflect this storage trend. The California ISO has more than 100,000 MW, PJM has 77,000 MW, and ERCOT has nearly 63,000 MW of energy storage waiting to be studied. Most models predict a gradual drop in capacity credit (under an average Effective Load Carrying Capability but a sudden drop under a marginal ELCC calculation) as additional renewable energy is interconnected.
Due to decreasing capacity credit for wind, solar, and storage, the IRP models are incorporating a dynamic value for capacity credit instead of a static value. This change to dynamic values should help further the prospects of hybrid resources such as storage plus solar because storage can maintain the capacity credit for solar if interconnected at the same location.
Modeling energy storage is not the same as modeling pumped hydro storage
Capacity credit calculation for storage is one variable in the IRP models. But an important one because the IRP model should see a need for additional storage or less depending on the capacity contribution of storage alternatives. Under a simple historical IRP scenario where no storage or renewable energy was offered as an alternative to meeting capacity needs, the model would most likely choose coal or a natural gas unit because those were the only viable alternatives in the next 2-10 years. Based on how they were modeled, demand response programs are selected in the less than the 2-year mark.
Before shifting from a simple IRP scenario where there were no renewables, it is worthwhile to note that pumped hydro storage was modeled in the IRP models because some states (e.g., Michigan) have existing hydro systems. The 12-hour charging and discharging cycles of pumped storage were modeled as an energy-limited profile. Some models still treat battery energy storage like pumped hydro, which is incorrect because, unlike pumped storage, batteries are much more flexible and can provide frequency response and other ancillary services in seconds and minutes instead of hours.
That’s the reason IRP models must include sub-hourly modeling, not hourly, because of the feedback loop from the system needs to the available technologies to address those needs and back to the system. This iterative loop has become more complex due to the interconnection of renewables on the grid; hence, the IRP models cannot afford to be coarse with their hourly modeling. They must include sub-hourly modeling due to the multiple benefits of storage technologies in 5-minute increments.
Additionally, storage is not for capacity needs alone in states within organized markets. With its energy management system, storage can monitor its State Of the Charge (SOC) and participate in the energy market. In most markets, participating in the energy market is a requirement if the resource qualifies as a capacity resource. The “must offer” requirement in energy markets is another reason the IRP models must model storage correctly because the interaction of the energy market and capacity market was an afterthought in traditional IRP models. The energy market price forecast for the next 10 years is an input into the IRP process, but no modeling path was identified for a resource that could meet future capacity needs and bid into the energy market simultaneously.
Subscribe today to the all-new Factor This! podcast from Renewable Energy World. This podcast is designed specifically for the solar industry and is available wherever you get your podcasts.
Listen to the latest episode on women in clean energy featuring senior female executives from CleanCapital, BQ Energy, and Renewable IPP.
Storage charging is like the interconnection support assumption
Storage skeptics are more concerned about the “how” of charging energy and the “when” of charging schedules since energy storage depends on another resource during charging. In that modeling sense, an energy storage capacity resource is similar to a transmission line modeled in IRP models for interconnection support. If a neighboring entity has a contractual agreement to deliver x MW of capacity via a transmission line, contractual capacity and the interconnection constraint were modeled in IRP models. Similarly, to address the doubts about the how and when of energy storage charging behavior – the IRP models must account for those contracts similar to the interconnection support.
It is time to put “integrated” back into the Integrated Resource Plan
A traditional IRP brings supply-side and demand-side resources together to meet the forecasted capacity needs for the next 10-15 years. This traditional framework was an “integrated” resource plan because it integrated both short lead time demand response and energy efficiency programs and long lead time coal and natural gas. But with renewable integration and energy storage, IRP’s “integrated” nature must also include storage benefits in transmission and distribution plans.
Distribution-connected storage can provide capacity and hence needs to be modeled in the IRP model. Storage can provide multiple distribution system benefits such as power quality, voltage support, and time of use rate management. Hence, all of its capabilities must be integrated into the IRP models.
Conclusion
Energy storage modeling in the integrated resource plans must be improved because storage has proven itself. Modeling battery storage as pumped hydro storage will no longer work due to the multiple benefits that batteries provide that can be stacked on top of each other. Future long-duration storage technologies must be modeled as capacity alternatives in addition to 4-hour duration batteries. Finally, with its tax incentives for stand-alone storage, the Inflation Reduction Act forces the IRP models to update energy storage modeling.
