INTERPRETER -- Tool for non-technical losses detection

TL;DR

This paper introduces a tool within the INTERPRETER project that detects non-technical losses by analyzing smart meter data and grid models, focusing on voltage deviations to identify suspicious areas for inspections.

Contribution

The paper presents a novel hybrid method combining feature detection from smart meters and grid model analysis, with a focus on voltage deviation analysis for non-technical loss detection.

Findings

Voltage deviations in daily minimum voltage are most indicative of losses.

A heatmap visualization aids quick assessment of deviations across meters.

Six locations identified for field inspections based on analysis.

Abstract

This article presents a tool for the detection of non-technical losses, which is being developed within the European INTERPRETER project. The tool employs a hybrid method based on feature detection from smart meter data and grid model analysis. This paper focuses on the grid model analysis, where voltage deviations between the grid model (digital twin) and real-world measurements at a low-voltage pilot site have been evaluated. Energy measurements from smart meters represent hourly mean power, while voltage measurements are instantaneous with uneven time intervals. Thus, measurements are not synchronous, which poses a major challenge for grid analysis. The proposed method focuses on daily mean, minimum, and maximum voltage and results show that deviations in daily minimum voltage are the most useful ones. A heatmap is developed, which helps the DSO expert to have a quick overview of all deviations of all meters in a certain time interval (1-day time step). A total of 6 locations have been identified where field inspections will be done.