A Partnership of UConn and Eversource

Eversource Energy Center

 

eversource-txt-logo

  • Homepage
  • >
  • Outage Prediction Modeling and Emergency Response

Outage Prediction Modeling and Emergency Response

Highlights

  • Continue improving the UConn Outage Prediction Model for all weather caused outages (rain/wind, blizzards, ice events, hurricanes, thunderstorms, etc.)
  • Develop a procedure to incorporate weather forecasting uncertainty in the UConn outage prediction system.
  • Develop integrative tools to support optimal restoration strategies based on outage forecasts and optimal restoration strategies.
  • Support emergency managers and grid operators training based on simulated catastrophic weather-outage scenarios.

Background

Severe weather is the major cause of damages to electric distribution networks and resultant power outages in the United States. Outage prediction models, based on weather forecasts and other information such as geographical data and attributes of the electrical system, are used to predict a storm’s impact many days in advance. To use these outage predictions effectively in decision making, models must exhibit acceptable accuracy in the spatial distribution of estimated outages for all storm types. The Outage Prediction Model (OPM) developed at UConn for the New England area, currently uses more than 200 simulations of different weather events (thunderstorms, nor’easters, snow/ice and rain/wind events and hurricanes), which combined with detailed geographical and electrical system attributes and associated outage reports has demonstrated a good performance across the various storm types. In this project we aim at capitalizing on successes and unique aspects of our previous research to help utility emergency operation and planning with: (i) predicting the likely storm impact (total number of expected outages), (ii) show where damage/outages are likely to occur according to weather, infrastructure and vegetation patterns, (iii) enhance awareness of storm severity and restoration practices. Additional application include providing information about resilience, such as quantifying the value of tree trimming or other power grid hardening activity.

Needs

  • Continue enhancing UConn OPM capabilities to forecast distribution network outages from winter storms, extreme weather events and thunderstorms.
  • Pilot novel procedures to incorporate weather forecasting uncertainty in the outage prediction system.
  • Demonstrate modularity and transferability of the outage forecasting system across the different Eversource service territories.
  • Develop restoration support tools based on weather and outage forecasts and optimal restoration response strategies.
  • Develop tools for emergency managers and grid operators training and facilitate drill exercises of catastrophic weather-outage scenarios across New England.
  • Provide resiliency insights, such as quantifying the value of vegetation management and other network hardening investments in terms of storm outage reductions.
  • Provide projections of outages and outage durations under climate change projections (e.g. future storms, snow/ice events, hurricanes, sea level rise)
  • Provide operational support of Eversource incidence control managers through our on-line outage forecasts communication system.

Expected Deliverables

This work will consolidate current research that has developed methods for combining historical weather forecasts with utility-owned, or freely-available, data layers (e.g. geography, system hardening activities, electric grid attributes and outage reports) to advance the state of the art in outage modeling that provides critical information for storm recovery.

Deliverables from the new integrated project will include:

  • Enhanced OPM for winter (snow and ice) storms, severe weather events (including hurricanes) and thunderstorms.
  • Probabilistic OPM predictions by incorporating weather forecasting uncertainty.
  • A system to transfer OPM calibration from the data rich CT to less data rich MA and NH Eversource service territories.
  • A restoration optimization model that predicts restoration plans and restoration times.
  • A training restoration tool for emergency managers.
  • What-if scenarios quantifying impact of vegetation management on outage reduction.
  • A new operational outage forecast communication system that incorporates the probabilistic OPM output.
  • Training utility engineers on the use of these data.
  • Operational production of OPM and dissemination to Eversource managers.

Points of Contact

Eversource Energy Center | Innovation Partnership Building: 159 Discovery Drive, Unit 5276, Storrs, CT 06269-5276 | E-Mail: eversourceenergycenter@uconn.edu