ELECTRICITY THEFT DETECTION AND LOCALISATION IN UNKNOWN RADIAL LOW VOLTAGE NETWORK

Abstract

The distribution of electricity involves both technical and non-technical losses. One major cause of non-technical loss is the illegal abstraction of electricity which is also known as ‘Electricity Theft’. The illegal usage of electricity has many associated problems, both for utilities and consumers of electricity, implying that there is a pressing need for theft detection and localisation. Traditional methods of identifying illegal electricity consumers are time consuming and ineffective as measurements have to be performed at a large number of suspected locations. Smart metering in future electricity networks will lead to a more efficient automated system for the detection and localisation of electricity theft. This will enable immediate action to be taken by distribution network operators against the offenders and will help to improve the quality, reliability and security of electricity supply systems. The aim of this study was to analyse the performance of an electricity theft detection and localisation technique in an unknown grid. The method assumed the availability of measured voltages, currents, and powers from installed smart meters. The detection step was a power comparison process and the localisation step was a voltage comparison process. The investigation involved analysis in the presence of single and simultaneous multiple thefts. To better represent future networks with increasing penetration of renewable energy generators, distributed generation was added to the system and the capability of the detection and localisation technique was further explored. All the simulations were performed in Matlab/Simulink. It was found that the method performed satisfactorily, with a minimum stolen power of 450 W successfully detected and localised.