Utility improves storm outage prediction accuracy by 20%

Data science case study
October 14, 2024

Utility improves storm outage prediction accuracy by 20%

Key takeaways

  • A Northeastern utility wasn’t prepared for severe storms due to limited forecast data
  • E Source implemented its purpose-built solution that combines practical experience with AI-powered insights
  • The utility improved outage prediction accuracy by 20% and reduced storm response costs by 25%

The challenge

A Northeastern utility’s storm preparation plans were affected by inaccurate outage prediction and reactive planning because of limited forecast data that didn’t account for utility infrastructure or vegetation conditions. Unreliable response plans and delayed mutual assistance requests meant the utility risked financial penalties, wasted resources, a longer estimated time of restoration, and customer complaints

Improve outage prediction with Storm Insight

Contact our team to learn more about our expertise and how we can help.

The solution

The utility chose E Source’s GridInform Storm Insight, a solution designed specifically to help utilities enhance their storm preparedness through AI-powered predictive analytics.

E Source reviewed and cleaned the utility’s existing data, including infrastructure data, before blending it with a comprehensive range of third-party variables to generate accurate and dynamic predictive insights.

The results

GridInform Storm Insight improved the utility’s outage prediction accuracy by 20% three days before a storm.

Additionally, it reduced storm response costs by 25% and helped the utility exceed its goal of restoring power to 95% of customers within 24 hours of a storm.

The utility has now implemented GridInform Storm Insight across its other operating companies to enhance storm readiness and operations.

Effectively manage storm-related outages with Storm Insight


© 2008 - 2025 E Source Companies LLC. All rights reserved.
Distribution outside subscribing organizations limited by license.

Source URL:https://www.esource.com/case-study/701240a96x/utility-improves-storm-outage-prediction-accuracy-20