Helping power utilities and others better plan for the future

Power lines can sag if temperatures rise, for example, becoming weaker, less efficient and prone to failing. Changes in air temperature and humidity can also affect power plant cooling. “These things have a real impact on the operations of a plant,” said Thomas Wall, an infrastructure and preparedness analyst in Argonne’s Risk and Infrastructure Science Center and one of the co-leads of the new initiative. “Maybe in the future I can’t generate as much power at that location because I can’t cool it as effectively.” 

After an unusually intense rain flooded one of its substations a year or so ago, a utility in the northeast approached Wall and his colleagues. The substation provides power to one of the largest employers in the state; the utility wanted to know if it would be increasingly likely to flood in the future — and if so, how badly, and what could be done to mitigate it.

Wall’s team did an analysis that took into account factors such as sea level rise and storm surge from a hurricane. They looked at the flood vulnerability of six nearby substations and used a model of the electric transmission grid to see what would happen if those six substations failed simultaneously. Would there be cascading failures across the grid that would cause a much larger problem in the region?

“The good news was the answer was no, there wouldn’t,” Wall said. But there would be local blackouts, and the substation that motivated the analysis would indeed be increasingly prone to flooding. The team recommended that the utility build a new substation on higher ground and steered them to locations in the area that would be less vulnerable.

Climate modeling has only lately developed to the point where these kinds of studies are realistic. The models dice the world into a grid of cells and calculate the state of every cell repeatedly through time. With supercomputers, researchers have the ability to model smaller cells and shorter time steps, resulting in finer resolution. And the finer the resolution, the more specific the model can be about what will happen at a given location.

At the Argonne Leadership Computing Facility, a DOE Office of Science User Facility, scientists can now simulate the climate at a resolution of a few kilometers. “It’s a huge improvement,” said Yan Feng, an Argonne climate scientist and the initiative’s other co-lead. “So it becomes easier for the climate model output to be used for decision-making.”

Feng and Argonne colleague John Krummel recently helped researchers from Nevada’s Desert Research Institute develop a high-resolution fire hazard map. The project, funded by the California Public Utilities Commission, will help utilities manage and site overhead utility infrastructure.

“They came to us because they needed to model the regional climate — winds, relative humidity and temperature — as a function of time for 10 years at a resolution of two square kilometers,” she said. “It’s not affordable for them to do that on their own computers.” The climate simulations used approximately 7 million core-hours on the Argonne Leadership Computing Facility’s Mira supercomputer.

Both Wall’s flood assessment project and the fire map effort would have been difficult to put together elsewhere, Wall said.

“There are lots of places that can do climate modeling or infrastructure modeling; there are places that have decision science and risk analysis capabilities. But what makes this unique is that we have all of these in one place and we’re all talking with one another,” he said.

John Harvey agrees: “The great thing about Argonne is we’re a multi-disciplinary lab,” he said. “Working together, we can provide a lot more value to utilities and become a wonderful resource for them. They really can’t get this information in one place anywhere else.”