Distribution System Reliability: Estimation Using Sampling and Statistical Analysis

Southern California Edison asked Analysis Group to help develop a more accurate method for measuring the commonly used reliability metrics System Average Interruption Duration Index (SAIDI), System Average Interruption Frequency Index (SAIFI), and Momentary Average Interruption Frequency Index (MAIFI). Developing these metrics is not always a straightforward task. Legacy methods are frequently based on approximations, but some regulators have requested that more robust methods be used. Until the real-time systems can be adopted, more accurate interim methods to measure and report SAIDI and SAIFI are required. 

The Analysis Group team, working with affiliate Dennis Aigner, used both weighted average and regression-based approaches based on stratified random sampling to develop more accurate estimates of SAIDI, SAIFI, and MAIFI. In the first part of the two-phase project, Analysis Group used stratified sampling and a weighted average estimate technique to calculate reliability metrics for the most recent historical year. In the second phase, we utilized a sophisticated regression-based approach both to calculate metrics for the most recent historical year and to forecast metrics for the current year. In addition, we estimated back-cast historical metrics for several years for use in comparison.

Southern California Edison has used our analysis and report in ratemaking filings at the California PUC and is currently using the analysis for internal planning purposes.