DETROIT — MISO’s Independent Market Monitor said the recently uncovered, eight-year-old repeat error in the RTO’s capacity market that caused a $280 million impact in this year’s auction alone is unfortunate but insisted the resulting prices were efficient.

Monitor David Patton said he thought the MISO tariff’s requirement that loss-of-load expectation (LOLE) only be contemplated during daily peak hours was outdated in the first place. He said renewable resources have shifted loss-of-load risk to MISO’s non-peak hours.

MISO discovered in summer that an unnamed vendor since 2017 has miscalculated the RTO’s LOLE using an “all-hours” methodology, rather than the tariff-defined “daily peak hour” methodology, leading this year’s auction to clear more capacity than intended. As currently defined, a day with a loss-of-load event is counted in MISO’s LOLE calculations only if the event happens during the hour with daily peak load. The coding error caused a $280 million impact on market participants in this year’s auction, with some owing more money and some getting refunds. (See MISO Discloses $280M Error, Over-procurement in 2025/26 Capacity Auction.)

Patton said that despite the mistake, MISO’s clearing prices denoted the true reliability value of capacity resources in the footprint.

“The prices are actually right from a reliability standard; they represent a true one-day-in-10 standard,” Patton told the Markets Committee of the Board of Directors, meeting during MISO Board Week on Sept. 16. “Unfortunately, the tariff is actually flawed.”

MISO entered summer with a $666.50/MW-day capacity price across all zones. (See MISO Summer Capacity Prices Shoot to $666.50 in 2025/26 Auction.) The RTO experienced average real-time prices of $48.55/MWh over the summer, a 56% increase over summer 2024. The Monitor said energy prices rose largely from a 49% increase in gas prices and a 2% increase in load.

Patton said having an LOLE limited to peak hours “made sense six to seven years ago” when MISO had fewer intermittent resources and risk hours. He said the RTO’s performance since then clearly shows that emergencies now crop up outside of the peak hour.

If MISO had set reserve margin targets and procured capacity according to a “daily peak hour” methodology, it would have only achieved a less than one-day-in-five-years loss-of-load standard, under half of the target, Patton said.

“I don’t think it’s the right answer, and MISO doesn’t think it’s the right answer either, as they have already filed to fix this,” Patton said of the existing tariff language.

MISO said it plans to adopt an all-hours calculation in its LOLE because of its more volatile risk profile and emergency conditions popping up at non-peak times. However, the RTO did not mean to impose the switch beginning in the 2018/19 planning year.

Patton said that he was encouraged to see that MISO already filed to “fix the LOLE definition in the tariff with little opposition from participants.”

Senior Vice President of Markets Todd Ramey agreed with the Monitor that the mistake resulted in a “more accurate representation” of day-to-day risk in MISO, though it “slightly overstated” risk according to tariff definitions. He explained to the board that the error affected a parameter in MISO’s LOLE calculation, which “had an effect of being at odds” with the tariff-defined LOLE calculation.

Patton said that while the resettlements may be legally required, they “undermine the integrity of the competitive markets.” He said resettlements will be “inconsistent with the information posted prior to the auction,” which market participants used to make decisions regarding supply contracts and resource retirements. “From a market standpoint, this is really unfortunate,” he said. He emphasized that it is critical that market participants can rely on the data MISO posts ahead of the auction.

Considering the tariff requirement that the RTO limit corrections on long-term errors to the past year, Ramey said MISO determined that the most “appropriate adjustment” was to resettle market participants’ positions at lower estimated capacity prices in the 2025/26 auction.

MISO has said it will not rerun or completely resettle the 2025/26 auction. It has called the process “settlement adjustments.”

Ramey said that because the auction clearing prices were the highest in summer ($666.50/MW-day), the “bulk” of financial impacts involve the summer. He said MISO would issue three separate settlement batches for the summer.

The RTO has held one-on-one meetings with affected market participants, he said.

Patton said MISO should consider tariff changes that would allow it to “avoid retroactively resettling markets in the future” when errors occur. He said he would be in favor of doing “the least destructive thing to the market.”

“I think MISO is in an impossible position, balancing its legal obligations under the tariff with the market concerns,” Patton said in summarizing the situation.

Director Nancy Lange asked about stakeholders’ reactions, as the mistake resulted in some “winners and losers” among market participants. “Do you feel like there’s grace and understanding, or some consternation?” she asked.

“No one is happy in a circumstance like this,” Ramey said. “At the end of the day, it’s an unfortunate situation we’re working through.”

Ramey said MISO had to strike a balance between mitigating the impacts of the mistake and protecting the integrity of its markets. He said the saving grace is that market participants self-supplied about 90% of their capacity needs and weren’t affected by the prices in the voluntary capacity auction. However, he said, a few market participants relied heavily on the auction for capacity.

Ramey said MISO aims to cut down on overlooked mistakes going forward by initiating reapproval of authorizations for software and “changing the approach” to testing software.

Director Robert Lurie asked for a follow-up report on MISO’s efforts to strengthen software validation.

Public Utility Commission of Texas economist Werner Roth, who is also the chair of MISO’s Resource Adequacy Subcommittee, said the loss-of-load model “exists in a black box.” He said little is known about the important calculations that planners in MISO count on to make resource decisions.

“We need more data transparency,” Roth said. “Confidence in the LOLE model results are critical, and [MISO] could benefit from additional eyes.”