By Robert Mullin

California must find new approaches to long-term forecasting and collaboration to keep pace with the accelerating effects of climate change on the state’s energy system.

That was the key takeaway from a California Energy Commission workshop Thursday focusing on developing strategies for climate adaptation in the state’s energy sector.

For California, adaptation is currently focused on the threat of wildfires and the role power lines can play in igniting them. The fire season is becoming longer in duration, increasingly destructive to natural and built environment, and more disruptive — and deadly — for the state’s inhabitants. (See California Regulators OK Utility Wildfire Plans.)

“The motivation behind this whole effort is really the stuff that we’ve seen in the news,” said David Saah, managing principal and co-founder of Spatial Informatics Group (SIG). “We’ve seen a bunch of extreme wildfire events that impact the grid, and as it impacts the grid, it impacts all of us in terms of costs, safety [and] reliability.”

SIG describes itself as an “environmental think tank” that combines spatial analytics with ecological, economic and social sciences to gauge the impact of policy decisions on ecosystems. The group is working with the state’s Cal-Adapt team to “deliver updated wildfire models for improved electric utility grid resiliency and safety” and support California’s next Climate Change Assessment.

Saah, an associate professor at the University of San Francisco and director of its geospatial analysis lab, explained that while much of the science behind wildfires is well understood, there are still a lot of “known unknowns,” including how to fit California’s recent large outbreaks of tree mortality into existing wildfire models to understand how “large, dead trees” affect wildfire behavior.

“We also know that our existing fire weather forecasts underestimate really severe or extreme wildfire events,” he said. “Part of that is due to the scaling; part of that is due to the technology; part of that is due to the way we have our measurements built. We know we need to deal with that.”

Saah said that current wildfire models do not forecast “a long-term trajectory of where we’re going” and therefore fail to provide investor-owned utilities a roadmap that can inform their long-term planning.

“And all this is really needed by not only the IOUs to be able to predict these overall impacts to the way they operate their systems, but it’s also needed by the taxpayer, the resident, the environment that we all have here in California,” he said.

But Saah said development of new models is not enough: Industry stakeholders must incorporate them into scenario planning.

“Our state is changing. We have this whole wildland-urban interface that we need to think of, and that interface is changing, and where it’s locating, it’s [also] growing. And the way fire behavior moves through those communities — again, it’s one of these places that we need to do better in.”

To address that shortcoming and others, SIG is developing a three-pronged approach to wildfire planning.

The first part seeks to improve the situational awareness of extreme fire weather and tree mortality through “optimal configuration” of weather stations, examination of past extreme events, and analysis and mapping of tree mortality. The second part incorporates new scientific findings into near-term forecasts and long-term projections, while the third would create models that provide IOUs and other stakeholders with “actionable information” applicable to the time scales contained in those forecasts and projections.

Once those models are developed, Saah said their “source code” should be opened to the industry and wider public for critical examination.

“The more critics that we can get hammering away at it, the more learning we can actually get,” he said.

Shifting Paradigms

California regulators have lauded San Diego Gas & Electric as a model for how the state’s utilities can prevent wildfires in their service areas. (See Calif. Regulators to Scrutinize De-energization.) The utility credits its extensive weather monitoring system for the fact that its service territory hasn’t experienced a major fire since 2007.

Brian D’Agostino, SDG&E’s director of fire science and climate adaptation, said the utility isn’t resting on its laurels. The utility is instead effectively rebuilding what was once the world’s largest utility weather network. (The state’s larger IOUs are now poised to surpass SDG&E’s network as part of their wildfire plans.)

SDG&E plans to expand its network from 177 weather stations to 225 by the end of next year, with a focus on new installations along the wildland-urban interface that can provide data every 10 seconds to support emergency operations. The new stations will be positioned to perform a new function: minimizing the customer impact of power safety power shutoffs (PSPS) undertaken during periods of high fire danger.

“It’s not just where we find the windiest areas or where this weather information will best improve our fire models, but a big part of it is we have to work with the electric engineers on the system for PSPS events,” D’Agostino said.

SDG&E has also synchronized its fire behavior models with census and building data to identify the highest-risk areas with respect to population density.

D’Agostino also pointed out that SDG&E is also incorporating its database of 455,000 trees into its fire behavior modeling systems in order to identify every tree that has the potential to hit a power line. The utility is also simulating more than 10 million fires every day to determine the risks to its entire system.

“There is a lot of room for improvement, as we’ve heard [from Saah], so we’re looking closely to continue to collaborate with the ongoing statewide projects,” D’Agostino said, expressing excitement at the “open source” nature of the effort.

Speaking during a Q&A session at the end of the workshop, CEC Commissioner Andrew McAllister noted his agency must perform 10-year forecasts to help guide development of the state’s energy system. Pointing out that the CEC increasingly relies on scenario modeling as the effects of climate change “happen more quickly than anticipated,” he asked D’Agostino how SDG&E is considering higher-than-expected temperature increases as it maps out its own long-term transmission and distribution investments.

D’Agostino said he couldn’t directly speak to the utility’s funding priorities, but that as the head of meteorology, he could point to what his department is doing differently, including adopting an approach of focusing on only the most recent years’ weather data — rather than a long historical time horizon — to predict future temperatures and weather patterns.

Another change had to do with load forecasting. D’Agostino explained that SDG&E’s peak loads have historically occurred during periods when the hot, dry Santa Ana winds blow off the desert to the east of Southern California’s population centers. But a new pattern has emerged over the last 10 years in which hot, humid air masses coming from the south are accompanied by unusually warm water currents.

“Last year, we didn’t set a new [peak] load, but our water temperature off San Diego was supposed to be about 68, 69 degrees, and it was close to 80 for almost three weeks in a row, which kept our nighttime temperatures [from] even coming down to what our normal daytime high was,” D’Agostino said. “And that went on for weeks last summer and caused a lot of challenges in operating the electric system. So, we’re seeing a new type of load.”

Reiterating the point about the speed at which climate change is occurring, CEC Vice Chair Janea Scott asked, “What kinds of things do we need to do in this space to make sure that we’re doing our best to keep up or even get out ahead of things?”

“We’re entering into this no-analog scenario,” Saah responded. “We have no idea how this thing’s going to work. If you look at the way our scientific infrastructure’s been built for a long time, it’s been built around competitive science. I think that era’s over. I think we really need to get into collaborative science. And the place where we learn from each other as quickly as we can, we [will] change things as quickly as we can.”

D’Agostino said he seconded that view: “Our ability to work with each other at this point is really going to help us move faster.”