In early July 2024, Hurricane Beryl made landfall in Texas, causing widespread damage to the power infrastructure serving the greater Houston area. High winds, heavy rainfall, and flooding brought down trees, poles, and lines, leaving over two million CenterPoint Energy customers without power.
In the aftermath of the event, CenterPoint ramped up its efforts to improve the resiliency of its infrastructure, investing heavily in hardening the grid and figuring out solutions to minimize storm-related outages.
Technosylva, a veteran of the wildfire tech space, had already been working with CenterPoint on wildfire mitigation and prevention. Indran Ratnathicam, Technosylva’s chief growth officer, told Latitude Media that that familiarity prompted CenterPoint to turn to them for help. “They asked us, ‘Hey, can you apply the same sort of technology to help us reduce our risk on the outage and [storm] resilience side?” he recounted.
It was great timing for Technosylva, which had just acquired KatRisk, a platform developing advanced flood mapping and risk technology for the insurance industry. So the company “diligently” set to work. Last month, it presented its progress at CenterPoint’s first Resiliency Technology Summit, which focused on tools to improve hurricane preparedness and response, alongside companies such as Neara, Pano AI, and Urbint.
But the company isn’t alone in recognizing the growing threat of the elements to energy infrastructure. Technosylva’s expansion from fires to floods is part of a trend of both tech companies and utilities approaching weather risk in a more inclusive way, according to Julia Hamm, a partner at The Ad Hoc Group, where she advises startups working with utilities, among other things.
“Utilities are thinking about weather-related risks more holistically in terms of not just looking at a single risk, but looking across the suite of extreme heat, extreme cold, floods, wind, wildfire,” Hamm said. “And on the technology side, many companies that started narrowly focused on a single risk category are now beginning to expand across multiple risk categories, because that is how utilities think about things… and having technology partners that can address multiple risks as opposed to a single risk is much more efficient.”
From fires to floods
Technosylva’s work in the wildfire mitigation and prevention space is based on AI-powered models, which process multiple data streams, including extreme weather conditions, vegetation moisture levels, wind patterns, and topographical factors. The company builds its models by running historical weather re-analysis of any given territory up to 30 years in the past, to obtain all the weather data for when extreme events occurred in the past. Technosylva runs roughly half a billion simulations per day, which are used to try to predict whether a fire could start in a utility’s service territory in the next five days, and determine its potential impact on the utility’s assets.
Jason Zumstein, who leads Technosylva’s extreme weather products, explained that the process is all about understanding the risk of something like a wildfire at the asset level. “We’re able to associate all that risk back down to the asset level for the utilities, so that they can pinpoint where they need to mitigate against risk, or take action in an operational setting,” he said.
With the acquisition of KatRisk, Technosylva was then able to incorporate flood data — things such as advanced flood maps, and data about moisture and the water absorption rate of certain terrains — into its models. Now, the company’s analysis is extended to risk related to hurricanes, tropical storms, and floods as well.
Predicting the damage of both fires and extreme weather-related floods is about “understanding the extremes” of the weather, according to Ratnathicam. And strangely, the more extreme events can be more manageable for utilities because they’re also the more predictable.
Hurricanes, cyclones, and tropical storms form days or weeks before they land, allowing utilities to prepare. But merely severe weather — thunderstorms and tornadoes, for instance — can actually do far more damage because of how randomly it can behave. While forecasters can anticipate the conditions for a thunderstorm, there’s no way of knowing where exactly that thunderstorm is going to form, and how much rain it’s going to release.
The devastating floods in Kerr County, Texas, that killed more than 130 people earlier this month, for instance, were caused by thunderstorms with low predictability.
In those cases, according to Zumstein, the best thing a utility can do is do mitigation work using 10-meter high-resolution flood prediction maps, like ones Technosylva developed “using 5 million CPU hours on the Titan supercomputer in Oak Ridge labs in Tennessee.”
“You can use these maps to plan ahead and know where those flood areas potentially are,” he said. “All those flood areas are pretty evident in the maps,” he said. “And then, of course, when severe conditions exist, vigilance is most critical.”
Other measures
Incorporating tools like Technosylva’s, which help utilities adopt forward-looking prevention strategies and organize future grid investments and maintenance, is just one of the actions utilities are taking to protect themselves from flood risk.
Another fundamental step, informed by those predictive tools, is grid hardening. That includes elevating critical equipment like control houses to higher ground, pillars, or platforms; installing flood walls; and investing in submersible equipment that can sustain damage in case of flooding. CenterPoint’s resiliency initiative for the Houston area, for example, includes putting up storm-resilient poles designed to withstand extreme winds, elevating substations, and undergrounding power lines.
Other measures, according to Hamm, include using advanced metering infrastructure and sensors “to collect real-time data on things like rainfall, river levels, grid conditions,” in order to detect emerging flood risk. She recommended implementing smart-grid solutions so that the grid can “autonomously respond to changing conditions” and reroute power if one part of the system goes down.
There’s also a growing trend of utilities hiring staff meteorologists and deploying their own weather stations, Hamm added, saying that the trend was driven by an increase in wildfire risk, but it’s beneficial to the whole weather spectrum.
“That’s very important from data perspective, because the weather data that’s typically available from the National Weather Service tends to [cover] a much broader footprint, whereas weather impacts can be very location specific,” she said, “and so utilities are now beginning to deploy weather stations in high risk areas, so that they have their own very localized data.”
Utilities have been very much aware of flood risk for years, and many of these measures, especially when it comes to grid hardening, are not necessarily novel. But while in the past, power companies have tended to implement resiliency plans after an extreme weather event — case in point: CenterPoint’s initiatives after Hurricane Beryl — that is beginning to change.
“We’re beginning to get to a point where utilities regulators and other stakeholders are recognizing they need to think more proactively about investments in the grid to prevent the most disastrous situations from happening in the first place,” Hamm said.
It’s a lesson learned through many storms, according to Ratnathicam.
“If you ask any of our utility customers, ‘What would you say to the other utilities out there?’ they would say, ‘Don’t wait,” he said. “Don’t wait for an event to take you out before you start putting together a program.”
