A reliability improvement roadmap based on a predictive model and extrapolation technique

This paper explains the development of a ten-year reliability improvement roadmap for a major distribution utility of the USA. First, a benchmark approach based on a survey of the reliability indices of 21 utilities of the USA and Canada was used to set the roadmap targets. Moreover, a historical outage analysis was performed to identify the main outage causes and potential reliability improvement options. Then, a detailed predictive reliability model was used to assess the cost-effectiveness of a broad set of reliability improvement projects for a pilot study area. Finally, the results of the study area were extrapolated to the utility distribution system by using a novel technique. Here, in order to consider the differences between the study area and the utility distribution system (representativeness error), the main characteristics of each feeder (length, number of customers per circuit mile, percentage of overhead and underground exposure, voltage level, etc) were taken into account. The reliability roadmap results for the utility system are presented and discussed.