Distribution Transformer Asset Health Prediction and Lifetime Extension via Integrated Real-Time Sensing and Advanced Analytics

Electric power utilities are facing increasing challenges to maintain reliability within their territories due to increased demand of electrical power combined with increased penetration of renewables, electric vehicles, and other distributed generation resources at the distribution level. At the same time, supply of distribution transformers is increasingly constrained, causing utilities to seek solutions for extending the lifetime of assets and to identify more intelligent means of real-time monitoring and condition-based maintenance schemes. Smart metering and predictive data efforts are underway to support integrating renewables, peak management, and customer operations, but existing solutions stop short of the necessary cost and performance capabilities to detect and predict equipment failures which are increasingly threatening reliability and resiliency on the electric power grid. The project will target to address the gap in real-time sensing combined with analytics to increase the reliability and enable condition-based maintenance of distribution transformers. Specifically, Sensible Photonics has licensed technologies for real-time distribution transformer sensing from University of Pittsburgh and the National Energy Technology Laboratory, developed over more than a decade through support of the DOE Grid Modernization Initiative, Office of Electricity, ARPA-E, and the Solar Energy Technology Office with an R&D100 award received in 2023. The core low-cost fiber optic sensor technology was initially developed for distribution transformer sensing applications through thermal, chemical, and acoustic monitoring functionality. In the current program, we will pursue maturation, feasibility demonstrations, and ultimately commercialization of fiber-optic based distribution transformer sensing in collaboration with Eaton and will develop new analytics capabilities to leverage real-time sensor data at the asset level for (1) incipient failure alerts, (2) lifetime predictions, and (3) overall health assessment indicators of distribution transformers. Specifically, we will design, fabricate, and test field deployable sensor prototypes in the laboratory in collaboration with University of Pittsburgh and we will integrate them into commercial distribution transformers provided by Eaton. During Phase I, testing will occur in medium voltage laboratory facilities at the Energy Innovation Center, co-located with Sensible Photonics headquarters and University of Pittsburgh GRID Institute. Phase II will further mature the sensing technology platform, develop, and demonstrate grid-edge analytics for distribution transformers, and successfully validate an intelligent transformer within the Duquesne Light Company service territory.