New analysis from WECC suggests that Westerners should take cold comfort from the fact that grid operators were able to avert blackouts during a September heat wave that toppled records for temperatures and electricity demand.
The analysis shows that, while the region’s grid operators have significantly improved their ability to respond to extreme weather events since an August 2020 heat wave prompted California’s first rolling blackouts in two decades, other factors outside the control of operators played a key role in avoiding a repeat of the 2020 outcome.
“Things were good, but they weren’t perfect,” Tim Reynolds, WECC manager of event analysis and situational analysis, said Wednesday in presenting the findings to the regional entity’s Board of Directors.
This year’s heat wave materialized as a heat dome on Aug. 31 and lasted until Sept. 10, bringing record highs to cities throughout Northern California, such as Sacramento (116 F), Santa Rosa (115 F) and Calistoga (118 F), while temperatures to the south exceeded norms.
Over the course of the nearly two-week event, CAISO experienced persistently high demand, hitting an all-time record peak load of 52,016 MW on Sept. 6, which nudged past the previous high and far surpassed the peak of about 46,000 MW that occurred during the August 2020 heat wave.
The ISO’s own analysis, released last month, indicated that electricity imports, conservation measures and improved coordination with utilities and government agencies helped prevent blackouts this summer despite the higher demand than two years earlier. CAISO also pointed to the benefits of increased coordination with neighboring balancing areas, including through expanded membership of the ISO-run Western Energy Imbalance Market, as well as the addition of 3,500 MW of battery storage resources within its territory. (See CAISO Reports on Summer Heat Wave Performance.)
Learning Process
WECC’s examination took a wider view of conditions across the Western Interconnection, which on Sept. 6 also posted a record peak of 167,530 MW, shattering the previous high of 162,017 MW set during the 2020 heat wave.
But as the CAISO peak load figure for Sept. 6 suggests, California appeared to account for all of that increase. And that points to a key difference between the two heat waves: This year, the most extreme heat was concentrated in California, while in 2020 wide swathes of the Northwest and inland Southwest were simultaneously subject to extremes.
“So this lets us know the demand wasn’t as much as it was back in 2020 in those [Northwest and Southwest] areas, and at the same time, there are more resources that could be available,” Reynolds said.
Another key difference, according to Reynolds: This year’s heat wave saw less transmission congestion than in 2020, when planned outages limited transfers between the Pacific Northwest and California.
“Energy transfers were able to happen a lot better than … back in 2020, so that was not an issue this go-round,” Reynolds said.
And while some wildfires were burning in the West during this year’s heat wave, none of them affected systemwide reliability. The biggest impact was seen at the start of the heat wave on Aug. 31, when fires forced outages for nine transmission lines and 1,103 MW of generation throughout the interconnection. Those resources were all restored within days, before the worst of the heat.
Reynolds said Level 3 energy emergency alerts (EEA 3) were issued seven times during the September heat wave, four of which were in the same — unnamed — balancing authority area. During an EEA 3, BAs “arm” themselves to begin shedding load. But no load was shed this time around, something Reynolds partly attributed to operational improvements that the BAs adopted based on best practices developed by WECC and the region’s reliability coordinators after the 2020 blackouts. He said WECC’s analysis of the 2020 heat wave found that BAs and RCs at the time lacked clarity on how to respond to emergencies.
“We actually sat down and had several meetings to go over what were some of the best [and] common practices,” Reynolds said. “It was great to see because some of the RCs had their trainers there, and they were kind of asking each other, ‘How do you train for an EEA?’ And they’re sharing ideas and everything else, so it was a great collaboration that was going on between WECC staff and the RCs, and we collated all that information to be able to make a best practice document.”
Reynolds said the process helped inform more BAs that, during an EEA 3, they can count armed load-shedding schemes as contingency reserves, freeing them to use spinning reserves to serve real-time load.
“What was nice was [in] this go-round … we saw more balancing authorities actually doing that, once they hit that EEA 3 level,” said.
Forecasting Flaws
WECC identified continued flaws in day-ahead load forecasting during the September heat wave, a carryover from 2020, with actual peaks outpacing forecasts during both events. On the day the interconnection registered its new record peak, the actual peak exceeded the day-ahead forecast by 4%, an even wider margin than the 2 to 3% errors seen in 2020.
“One thing we’re noticing a little bit of … with the EEAs is there’s not a lot of guidance or best practices out there for the forecasting, so there’s definitely potentially some areas for improvement and sharing those forecasting best practices for the day-ahead — but also for the annual forecasts,” Reynolds said.
Wind forecasts were similarly subject to errors during the heat wave, a phenomenon WECC also identified from its 2020 analysis.
“During the times of the peak and the most intense part of the heat waves, we noticed wind generation [would] go below forecast,” he said, adding that wind output didn’t necessarily come up short of forecast during the entire heat wave.
“We are definitely recommending more analysis to kind of look into this even more,” he said.
On a positive note, battery storage was a big contributor to the grid during the heat wave, in some intervals actually outproducing the 2,200-MW nameplate capacity of the Diablo Canyon nuclear plant. About 95% of that output was from battery resources located in California, WECC determined.
Nothing to Celebrate
WECC board members were impressed with the findings. They were less pleased by their implications.
“I hope people see this as, you know, we were pretty lucky. I mean, the weather could have changed significantly, and from my point of view, we could have been right back where you started from in 2020,” Director Gary Leidich said.
Leidich encouraged WECC to publish the findings in a report that is as “neutral as possible” but makes clear that “this is not an event which we should celebrate — nor is it one that’s a disaster.”
“We need to keep pushing on those improvements to be able to, frankly, fight to keep the lights on,” he said. “I just want to see there’s a balanced perspective here, because I sense of some of the media that I read along the way [said] that people were celebrating this as some sort of a success, and I don’t think we should view it necessarily as that.”
WECC CEO Melanie Frye called Leidich’s comments “spot on.” She pointed out that Reynolds’ presentation didn’t include the fact that California at one point avoided blackouts because Gov. Gavin Newsom issued a call for emergency demand response that quickly reduced load by nearly 2,400 MW.
“And that demand response is a great tool, but that’s not the way we want to deploy that as a resource,” Frye said. “So while I think there’s a lot to be learned, and there is some recognition of … all the work that’s been done to improve over 2020, we’re not done, and we can’t just sit back and say, ‘Oh, we got this figured out.’”
“We didn’t have any major lines down, and we didn’t have any major power plants down, yet we were dangerously close to the edge,” Director Jim Avery said. “I think that’s important to highlight.”
