An Electrical Infrastructure Management Model for Long-Term Maintenance and Hazard Mitigation

Abstract

Electrical lines are critical infrastructure components. Line failures can cause catastrophic wildfires resulting in immeasurable damage to life and property. This paper discusses the implementation of an electrical infrastructure management model that tracks component degradation over time and incorporates inspection and monitoring data to update the model; the model then identifies components with the highest risk of failure to direct maintenance operations and estimate long-term line performance. The capacity of the structures is represented through fragility curves based on first-principles modeling and industry standards. Threat models, such as metal corrosion and wood decay, estimate how the fragility curve will evolve with time, based on environmental data and the age of the structure. Data-driven approaches are used to incorporate field data and historical performance into the model. Bayesian updating is used to adjust the fragility curves based on historical wind and outage data. Similarly, observed deterioration from field condition inspections are applied to further modify the model. Using a site-specific hazard model, an annualized probability of failure is estimated for each structure on an electrical line. First, an overview of the framework is provided, with a specific focus on wind hazards. To illustrate the application of the model and illustrate the challenges encountered in implementation, an example case study is discussed, exploring how the predicted probabilities of failure can be impacted by the choice of wind hazard model. This choice can result in geographic bias towards both under- and over-predicting probability of failure, and therefore can result in misdirection of maintenance resources. Future planned improvements are discussed.