PJM stakeholders last week unanimously endorsed the 2021 reserve requirement study (RRS) but requested the RTO conduct further analysis on the impacts of extreme weather conditions.

Patricio Rocha Garrido, of PJM’s resource adequacy department, presented the results of the RTO’s load model selection process at the Planning Committee’s meeting Aug. 10. The RTO analyzed 120 load model candidates for the 2025/26 delivery year RRS, based on the 2021 PJM Load Forecast Report released in January. (See “Load Model Selection,” PJM PC/TEAC Briefs: July 13, 2021.)

PJM recommended using a 13-year load model utilizing data from 2001 to 2013, moving the time frame back one year from that in the RSS approved for 2020.

Rocha Garrido said the load model candidates were compared to PJM’s “coincident peak 1” (CP1) distribution analysis, which represents the highest load expected for the forecast year by using two separate approaches. The model selected in 2020, which used data from 2002 to 2014, was not one of the top candidates this year because of a new CP1 distribution analysis, he said.

The 2021 curve had higher loads and was a “little bit more conservative” compared to the 2020 RRS, Rocha Garrido said. PJM used a 10-year load model (2003-2012) for several years in a row before switching to a 13-year model in 2020.

The load model selection must be conducted because the coincident peak distributions from the PJM load forecast cannot be used directly in the PRISM modeling software, which is the RTO’s primary modeling tool used for conducting resource adequacy studies.

PJM also recommended switching the peak week for the MISO, NYISO, Tennessee Valley Authority and VACAR regions, known collectively as the “world” in the analysis, to a different week in July so that it doesn’t coincide with its own peak. Rocha Garrido said PJM and the world have peaked on the same day nine times, the last coming on July 19, 2019.

“In PRISM, if you have the two regions peaking on the same week, that’s equivalent to peaking on the same day,” Rocha Garrido said.

Stakeholder Questions

Paul Sotkiewicz of E-Cubed Policy Associates asked if it would be possible for PJM to conduct similar analysis in the Multi-Area Reliability Simulation (MARS) program, which is able to be “more granular” in data than PRISM.

Rocha Garrido said PJM conducted an overall analysis several years ago of capacity benefit results using MARS, simulating several historical years. He said the resulting analysis in MARS showed to be less valuable in formulating the load model candidates, so PJM decided to continue using PRISM.

Sotkiewicz said that in light of the U.N.’s recent Intergovernmental Panel on Climate Change report that stressed that global climate change’s effects are already happening, PJM should consider re-examining its modeling methods. (See Too Late to Stop Climate Change, UN Report Says.)

He pointed to recent events in California, where the state was depending on the ability to import power from its neighbors to cover shortages, but the exporting states were having their own energy issues with high peaks.

A similar scenario could happen in the future on PJM’s borders, Sotkiewicz said, including sharing events with NYISO and ISO-NE to the north, MISO to the west and Southern states. He said he didn’t want to cause panic among stakeholders or suggest PJM’s proposed RSS was incorrect, but he asked if there was enough diversity in the modeling to account for regions all peaking at the same time.

“Maybe this is an opportunity for us to think very seriously about this before it becomes a major problem like it is out west,” Sotkiewicz said.

Rocha Garrido said he agreed with Sotkiewicz’s concerns and that PJM has looked at potential solutions. He said one of the issues is that PRISM doesn’t have the capability to model each of PJM’s neighbors separately and data must be inserted as a “monolithic” region that doesn’t capture the diversity of the regions.

Calpine’s David “Scarp” Scarpignato said he has consistently raised the peaking issue and would like to see the RTO and stakeholders come up with a solution.

In 2018, stakeholders hotly debated PJM’s proposed revisions to adjust the methodology for developing the capacity model for winter peak weeks, expressing concern about how the modifications might affect capacity procurement. (See “Adequacy Analysis Approved Despite Concerns,” PJM MRC/MC Briefs: June 21, 2018.)

Scarp also said having the historical peak load coinciding with the world nine times in the last 22 years is “pretty significant” and an issue that shouldn’t be ignored. He suggested PJM come up with a way of capturing the idea without “making it binary.”

“We can’t simply say that our modeling will assume that PJM and the world don’t peak in the same week, because that doesn’t make sense,” Scarp said. “We know from empirical data that it does peak in the same weeks.”

PJM officials agreed to continue discussions at the Resource Adequacy Analysis Subcommittee.