Detecting and locating non-technical losses in modern distribution networks

Abstract

The recent addition of information and communication technologies in electric power distribution systems has introduced a new class of electricity theft or nontechnical loss. Energy consumption data can be hidden and altered through cyber-attacks that are characterized by the unauthorized access to the application database and digital tampering of smart meters. The development of cost-efficient algorithms to address these types of nontechnical losses also targets the reduction of commercial losses because the full protection of an information system is very expensive. Thus, this paper proposes a strategy to detect nontechnical losses using a multivariate control chart that establishes a reliable region for monitoring the measured variance. After the detection of nontechnical losses, a pathfinding procedure based on the A-Star algorithm is able to locate the consumption point with the non-technical loss. Moreover, a geographical information system application displays the consumption point that is the target of the cyber-attack. The numerical results demonstrate the selectivity and efficiency of the proposed methodology applied for monitoring a real distribution network.