A Scheme for Electricity Theft Detection Based on Shewhart Chart of Variance

This paper proposes a methodology for electricity theft detection based on the Shewhart s chart, which monitors the variability of the consumption data. The problem of theft detection has been split into different levels of power theft, as large power thefts may be of greater importance to utilities since they represent a higher amount of revenue lost. The proposed method consists of two stages: i) in the first stage, the number of days a particular customer is involved in fraudulent activities is detected, and ii) in the second stage, using the information of the first stage, a ranked list of the dishonest customers in a community is prepared. This methodology only uses consumption data available from smart meters installed at consumer’s homes. It does not require the installation of a central observer meter or a master meter for the process of theft detection. This method has been tested on a real-world dataset containing consumption data of residential customers. The results show that the method has good performance in terms of various metrics, especially for large power thefts.