By Amanda Durish Cook
MISO said last week that it is leaning against the Independent Market Monitor’s proposal to restrict the ability of offline resources to set prices based on the results of a simulation study.
Congcong Wang, market design engineer, told the Market Subcommittee on Aug. 2 that MISO “continues to recognize the value of offline pricing” and is developing alternative solutions to the Monitor’s recommendation in the second phase of the extended LMP rollout.
Using simulations, MISO found that the Monitor’s proposed expansion of price setting doesn’t result in the most efficient prices, Wang said. “It does not mean the recommendation isn’t a good one; it just means that our current software … may not maximize price efficiency,” she added.
In the State of the Market report, the Monitor said offline resources should only set prices when they are economic and can be started quickly to address a shortage.
Monitor David Patton’s ELMP recommendation was two-pronged: He also advised expanding the share of online peaking resources eligible to set prices to include those with start times of one hour or less and minimum run times of two hours or less, regardless of whether they are scheduled in the day-ahead market. (See Monitor’s State of the Market Report Seeks Changes to MISO ELMP.)
Wang said MISO ran four days of simulations: Jan. 18, 2016, with no fast-start resource participation; Jan. 4, 2016, with low participation of fast-start resources; July 17, 2015, with high participation of fast-start resources; and July 12, 2015, with scarcity conditions with offline resource participation and heavy online participation.
MISO found the Monitor’s recommended price-setting expansion resulted in price increases from $1.52/MWh to $9.42/MWh. Expanding ELMP price-setting to units with 30-minute start times resulted in price increases of $0.34/MWh to $3.50/MWh. The Monitor’s recommendation causes price divergence between day-ahead and real-time prices in as much as 85% of intervals, but the 30-minute unit expansion doesn’t affect price convergence, the RTO said.
The recommendation results in online fast-start participation in more than 99% of intervals and the amount of pricing intervals impacted by ELMP rose from 0-7% to 35-74%, according to the RTO.
Patton responded that two of the test days MISO used were already under-scheduled by as much as 6 GW. “The convergence was naturally bad to begin with,” he said.
“I think it’s important to note that this high, it’s true that ELMP will affect more intervals, but many of these intervals are moving by a few cents,” Patton said. “To me, these results suggest that the expansion is necessary.”
Patton said he discovered that offline units setting prices were actually used only 8% of the time, and a diesel unit in Michigan was allowed to set prices 50 to70 times during the period he studied for the State of the Market report without ever being started.
“I’m not sure offline pricing has a strong benefit to begin with,” said Patton, who argued to FERC in the creation of Order 825 that offline pricing can “artificially lower energy prices and obscure shortages.”
Wang pointed out that in Order 825, the commission noted that offline pricing can result in efficient prices.
She said one of the alternatives to dropping offline price setting in ELMP is shortening cost amortization intervals. MISO currently amortizes the commitment costs of offline fast start resources over four real-time intervals, or 20 minutes.
The RTO is planning a September workshop and would come back with more results at the October MSC meeting. Until then, Wang said MISO will continue to run simulations and investigate the impacts of offline price setting. MISO wants to test the second phase of ELMP in the second quarter of next year.
