By Joan Koka
Wildfires, alongside other extreme weather events, remain one of the biggest threats to America’s energy system. Flames cannot only damage physical infrastructure and disrupt service for millions, but also threaten public safety.
In regions like California, the threat is particularly high. In 2020 alone, north and central regions of the state experienced record-breaking wind speeds and saw 4 million acres of land consumed by wildfires.
If utilities in the region could project wildfire risks in the future, it could help them find ways to pre-emptively strengthen or protect their infrastructure and restore service faster when crises inevitably occur.
The U.S. Department of Energy’s (DOE) Argonne National Laboratory has the expertise and computing resources needed to help utilities come up with climate projections for conditions that cause extreme weather events. That’s why Pacific Gas and Electric (PG&E) chose to partner with the laboratory to build tools for assessing wildfire risks.
PG&E has sponsored research at Argonne to create climate impact datasets that will enable the company to analyze wildfire risk through 2050. Data generated will guide the company as it looks for ways to further protect the safety of communities, maintain service and improve resiliency across its 106,681 miles of distribution lines, one third of which lie in High Fire-Threat Districts.
“Argonne’s research will really advance our knowledge of how climate change might affect our assets and operations in the future. It will help us determine where our most vulnerable electric assets are so we can continue to prepare for climate change by strengthening our infrastructure,” said Debbie Powell, interim head of Electric Operations at PG&E.
The advantages of working with a national laboratory
While numerous conditions like wind speed, relative humidity and temperature impact wildfire risk, utilities on their own can only analyze these conditions over short time periods, such as days or a couple of weeks. To fill in the gaps, Argonne researchers have developed climate model projections on how conditions associated with drought or wildfires will evolve over longer periods of time.
A unique feature of their model simulations is that that they deliver data at a much higher resolution than is typically seen among the climate models available over the same domain (North America). Such models typically evaluate climate change over grid points that are 50-by-50 kilometers in size. By comparison, Argonne’s simulations cover much of North America at a resolution of 12-by-12 kilometers.
“Twelve kilometers by 12 kilometers might still seem like a large area, but considering that our data covers almost all of North America, the resolution is still among the highest when compared to other comparable datasets,” said Argonne Atmospheric Scientist Jiali Wang. “Data from this 12-kilometer model allows PG&E’s planners and weather modelers to better understand future climate conditions at the regional and local scales where decisions must be made.”
Researchers were able to perform simulations at a much higher resolution by harnessing the supercomputing resources at the Argonne Computer Leadership Facility (ALCF), a DOE Office of Science User Facility. ALCF resources are orders of magnitude more powerful than the systems commonly used for open science. With them, researchers were able to perform simulations and measure the range of uncertainty across different scenarios and time periods in the future, and across input data from varying global climate models.
How the climate model is being used
With their climate models, Argonne researchers are delivering to PG&E detailed data on conditions that are indicative of a fire risk. This includes indicators like low humidity and high wind speeds associated with Diablo winds, which are yearly strong, dry winds that cover the San Francisco Bay Area.
“Understanding the shift in the geographical location of these winds over time, whether it’s east towards the coast or away from the coast, is one of the things we’re interested in because this information can help utilities determine where they should have more resources ready to fight fires in the future,” said Rao Kotamarthi, Senior Atmospheric Scientist and head of the Department of Atmospheric Science and Climate Research at Argonne.
PG&E meteorologists and operational experts will pair this type of data with existing science and models and use it to plan for and adapt to climate changes found likely to affect the utility’s grid now and in the future. The company may also use this information to further refine existing strategies to harden its systems.
Other applications for regional climate modeling
On a broader level, Argonne’s approach to regional climate modeling can support utilities as they decide on long-term capital investments, such as burying underground cables or installing stronger utility poles and covered power lines.
“It’s not that our simulations will tell utilities where to bury cables or install poles, but they can give decision makers some other factors to consider when they’re making decisions about these kinds of long-term investments,” Kotamarthi said.
Such data can also confer useful information about electricity generation and transmission given that environmental conditions like high temperatures impact service delivery and can signal changes in demand.
“We’re uniquely positioned to deliver projections about future climate variables that aren’t available anywhere else,” said Thomas Wall, program lead for Engineering and Applied Resilience at Argonne. “In addition to reliability and resilience, this type of information can help to ensure greater public safety in the future.”
Joan Koka is a science writer at Argonne.