🔒Eversource facility aims to predict storm severity, CT threats

1 of 3

The Eversource Energy Center at UConn — a weather, climate and energy research hub — has already helped electric utilities better prepare for storms and potential power outages since it was established in the wake of Tropical Storm Irene in Aug. 2011 and the Halloween weekend storm two months later.

Those storms left hundreds of thousands of customers without electricity, some for more than a week, caused millions of dollars in economic losses and proved the need for better science to help predict storm severity, potential power disruption and assist emergency preparedness. Hurricane Sandy hit in Oct. 2012, doing significant damage as well.

After those storms, “we realized there was a need, to be able to scientifically, with precision and accuracy, predict damage that would occur on our electric grid and be able to plan for that and … ultimately build our system in a different way to help prevent that,” said Ken Bowes, Eversource Energy's vice president of transmission performance.

Eversource provided $9 million over five years to fund research and development at the center — seed money that's already attracting additional investment.

A prototype storm outage modeling program developed by UConn after the 2011 storms helped utilities better prepare for Sandy, said professor Emmanouil Anagnostou, the Eversource Energy endowed chair in environmental engineering and director of the Eversource Energy Center.

“That really motivated [state regulators], to say, yes, better information, better science or engineering technology — it's important for this industrial sector,” Anagnostou said.

That led to the decision to invest in research and form the center, which started in 2015 and is temporarily housed in the engineering school and will move into UConn's new Innovation Partnership Building later this year when it's completed. The center will occupy about 3,000 square feet of the 113,000-square-foot building that is the first in UConn's new Technology Park on the campus' north end.

The outage prediction model, the first significant program to come out of the center, combines sophisticated weather forecasting developed by UConn with historical storm outage data to create a predictive analysis program that helps forecast storm severity and outages, Anagnostou said. The impact model takes into account wind duration, speed and gusts, plus precipitation, snow and flooding.

The model required analyzing megabytes and terabytes of data related to the ground, power infrastructure and surrounding trees in a way that would most accurately predict outage locations, which is key to preparation, Anagnostou said.

“We can really have an impact on pre-staging line crews and tree crews and other resources we need … if we have several days of advance warning,” Bowes said.

The model, for example, was used to help prepare for the impact of the Nor'easter that hit the state last month with heavy snow, getting crews in place ahead of time to prepare for outages.

The predictive model has already attracted the attention of Schneider Electric, a global specialist in energy management, which signed an agreement to advance the technology by combining UConn's analytics model with Schneider's WeatherSentry Online platform to help utilities better understand storm impacts on their systems and better prepare. Commercial applications that emerge will create licensing royalties for UConn, Bowes said.

The model is one of about a dozen research projects at the center either already or soon to be examining areas that include tree endurance to wind, snow and ice, tree pruning/thinning management, flood vulnerability for electric substations that incorporates historical data and climate change analysis, electric grid reinforcements and resiliency, geomagnetic disturbances, integration of renewable power generation, and cyber and physical security.

Studies go beyond the engineering school to include the business school, for example, to run economic models and forecasts for utility resilience strategies to measure dollars spent on resilience versus return to the utility and state, Anagnostou said.

Tree and forest management is a key early area of study and includes trees equipped with instruments to better understand impacts from wind, ice and snow loads.

Electric grid resiliency is another key research area for the center, Bowes said. It goes beyond how utilities design, build and operate their systems to better withstand extreme weather, but to man-caused events as well, both cyber and physical, he said.

One area of concern, for example, is the threat of drone attacks on substations and how to prevent them, Anagnostou said.

Anagnostou said detailed flood analysis for the power infrastructure is under way and, in addition to outage predictions, the center is working to predict in real time the flood threat to power infrastructure.

Another project is looking at microgrids and how they enhance time to power restoration. A microgrid is a collection of centrally located businesses and facilities powered by a local source of generation, like a fuel cell or combined-heat-and-power unit, which can operate connected to the grid and in “island mode” during a power outage. The state turned to microgrids after the 2011 storms and awarded $18 million in grants for nine microgrid projects in 2013.

The center is having the intended effect of building partnerships to expand research and develop solutions, including with matching funds, Bowes said. Several utilities are participating and another university, expected to be named this month, will provide additional research to the center.