When the PJM Interconnection* proposed that energy storage needed 10 hours duration to qualify for capacity payments, that seemed to make no sense.
No other grid operator had proposed such a requirement—essentially disqualifying battery storage—in its plan for complying with an order from the Federal Energy Regulatory Commission to allow energy storage to participate in all markets where it is technically capable of doing so (including capacity markets).
Now we have an analysis that debunks PJM’s proposal.
Storage with 4-hour duration can provide up to 4,000 MW of capacity of “equivalent reliability value” to that supplied by conventional power plants in PJM, according to a study by Astrapé Consulting.
Astrapé used the SERVM model for its analysis, a model it has used for similar evaluations it has performed for other regional grid systems such as ERCOT, MISO, and SPP.
Astrapé’s report concluded:
A 4-hour duration requirement would correctly represent the capacity value of storage under current market conditions and would remain accurate until the amount of installed storage in PJM increases by two orders of magnitude.”
The “two orders of magnitude” refers to the 40 MW of 4-hour non-hydro storage that Astrapé determined is currently operating in the PJM region, compared to the 4,000 MW potential.
The capacity payments at issue are intended to ensure that adequate reserve capacity is available to meet demand at occasional times of extreme demand, such as extremely hot or cold days.
But when storage is excluded from the capacity market, that causes the market-clearing price for capacity to be higher, increasing customers’ electric bills. Meanwhile, lower compensation for storage can limit otherwise cost-effective deployments of solar and wind power, which pair well with storage.
In its analysis, Astrapé followed an elegant approach illustrated in Figure 1. In step 1 as shown, Astrapé modeled the addition of conventional capacity to reduce the “loss of load expectation” (LOLE) to 0.1, a “generally accepted reliability criterion” that “represents a single day of firm load shed in a 10-year period.” In step 2, Astrapé modeled the addition of energy storage, reducing LOLE below 0.1. In step 3, Astrapé modeled the removal of conventional capacity until LOLE was again 0.1. This yielded a ratio of storage capacity added to conventional capacity removed.
Recent Comments