Hit by three major storms in two years, a U.S. utility company was seeking a scientific model to better predict when and how to deploy its resources. 

The company was searching for something beyond weather forecasts and current disaster management platforms to improve storm-related electrical service for its 3 million customers. The utility turned to the scientific experts at Idaho National Laboratory for answers. Idaho National Laboratory, located in Idaho Falls, is a U.S. Department of Energy national laboratory equipped with an exceptional array of scientific expertise and equipment.

INL researchers Ollie Gagnon and Mary Klett set out to find a way to integrate weather data, electric infrastructure assets and the utility’s response resources to effectively predict damages, determine resource deployment needs and estimate restoration timelines. The result is Storm-DEPART (Damage Estimate Prediction and Recovery Tool). This software was introduced for the 2022 hurricane season, and the first Storm-DEPART commercial license contract was issued in 2024. 

IMPROVING EFFICIENCY 

In the event of an approaching tropical storm, this new solution assesses potential damage to power generation capacity, transmission grids, distribution networks and communications assets from hurricane force winds. Additionally, the software’s ice damage prediction functionality assesses potential damage from a winter ice event.

Storm-DEPART allows utility companies to make data-driven prioritization decisions on resource allocation before a major weather incident occurs. These resources include how many external mutual aid resources to request.

“If you have the right people in the right place at the right time, you’re able to restore power faster, which then helps get all the rest of the infrastructure up,” Gagnon says. Resource calculations determine how many crew members are needed, for example, to restore a wood pole in a heavily vegetated area versus a marshy area versus a concrete base. 

“If your calculation is off on that, you get the wrong crews in the wrong place, or you may not have enough. You could also send too many, and it’s a lot of extra money that’s being spent on the response and recovery,” Gagnon says.

Storm-DEPART also helps utility companies prioritize the restoration process based on predicted damage to critical infrastructure. “If you get a big impact in the area of generation, you may want to get that restored first because then you can send power to transmission and distribution systems,” he says.

FILLING THE GAPS

To better prepare for storm events, utility companies can identify ways to be more resilient to disruptions by assessing resilience gaps through modeling.

“Think about how you can model potential storms to determine where you may want to make resilience investments,” Gagnon says. Utilities can shorten restoration time because they’re building in resiliency by investing in the infrastructure to withstand different types of storms or stronger storms. 

Utility customers depend on electrical service for just about everything, and being without power disrupts everyday living. The Storm-DEPART data helps utilities determine when power will be restored. 

“Like in Florida, we get hit by a storm, and they’ll say, ‘Yeah, we don’t yet have an estimate on when power will be restored. You know, it could be a week.’ As a resident or business dependent on power, it’s hard to operate like that,” Gagnon says. “You want to know if it’s going to be two weeks, a week, or 24 hours.”

UPGRADES

While Storm-DEPART initially applied just to hurricanes, it now covers ice events, also. 

“It’s the same type of service provider with unique configuration data, just a different weather event,” Gagnon says. An ice event requires a different storm response than a tropical storm and potentially a different crew composition.

“Different regions have unique service requirements that can be captured to increase response flexibility and adaptiveness,” Gagnon says.

To get started with Storm-DEPART, a utility’s storm response team works with the software provider to enter configuration data, including a catalog of the utility’s infrastructure. The initial configuration takes about three months of weekly meetings and data sharing. After the initial configuration, the data is updated regularly. The next version of Storm-DEPART will have a web-based application, so storm response teams can enter their data themselves. Because the software is so new, Storm-DEPART has been used for ice events, but is yet to be tested by a hurricane.

“We haven’t had an actual storm, but we looked at legacy storms in terms of running the model and then giving them what we thought the resource needs would be. They said, ‘This is what we actually deployed,’ and it was very accurate,” Gagnon says.