Case Study: San Diego Gas & Electric predicts utility asset failures with PA Consulting's iPredict™

Predicting the failure of utility assets

San Diego Gas & Electric (SDG&E) faced a major reliability challenge with underground utility assets, especially T-splices, which are difficult to locate, access, and repair when failures occur. To address this, SDG&E partnered with PA Consulting and machine learning specialist Toumetis to co-develop iPredict™, a predictive maintenance solution for electrical distribution assets.

PA Consulting helped build and trial iPredict™, which analyzes high-frequency electrical data to detect precursor failure signatures and predict impending asset failures before outages happen. The solution is designed to help SDG&E proactively schedule repairs, improve crew safety, reduce operating and maintenance costs, and minimize customer outages, while proving to be a new tool for improving electric grid reliability.

San Diego Gas & Electric

Tom Bialek

Chief Engineer