[UC57] Reliability Metrics Trending and Benchmarking

The Utility Problem

Electric utilities are required to measure and report distribution reliability performance to regulators using standardized indices — most commonly SAIFI (System Average Interruption Frequency Index), SAIDI (System Average Interruption Duration Index), CAIDI (Customer Average Interruption Duration Index), CEMI (Customers Experiencing Multiple Interruptions), and CELID (Customers Experiencing Long Interruption Durations). These indices are not simply outage counts or event logs — they are normalized, customer-weighted metrics computed from outage data, designed to represent the reliability experience of the average customer on the system.

While most utilities calculate these indices at the system level for regulatory reporting, the indices are far more valuable when computed at the feeder (circuit) level. A system-level SAIFI that appears acceptable may mask one or two chronically underperforming feeders that are responsible for the majority of customer interruptions. Without feeder-level disaggregation, operations and planning teams have no systematic visibility into which circuits are driving reliability degradation, making it difficult to prioritize capital investment, maintenance programs, or targeted hardening projects.

Compounding this, reliability indices are rarely tracked as time series. Most utilities produce annual regulatory submissions but do not maintain multi-year trending of feeder-level indices, making it difficult to determine whether a particular circuit is improving, deteriorating, or seasonal in its reliability behavior. Benchmarking against regulatory targets or peer feeders is similarly ad hoc. The result is that reliability management remains largely reactive — driven by individual outage events rather than by systematic, data-driven identification of the circuits that most need intervention.

Awesense’s use cases are constantly evolving. Request free access to our Sandbox trial and GitHub repository to see the full use case descriptions and implementations, and to learn how you can create new use cases yourself.