DIET - A Different Approach to Smart Meter Data Insight against Energy Theft
Identifying potential energy theft and faulty equipment by examining changes in data through time
Smart meters generate alert and error messages and send warnings about the state of the meters alongside their energy consumption readings. The daily volume of smart meter messages can be as high as that of the consumption data, but these messages are considerably more difficult to interpret, and complicated by the fact that there is no international standard on error messaging.
Drawing correlations between meter system messages and meter consumption readings to detect metering faults or tampering is a complex task and there is no effective method to do so in the industry to-date.
The objective of this project is to develop an approach using meter system messages alongside meter consumption data to identifying potential energy theft and faulty equipment by examining changes in data through time. The scope is to analyse data collected from a pool of SME electricity meters with a view to developing a reusable method for the domestic market operation. British Gas, along with partners EDMI (meter supplier), and G4S (meter operator) are working in collaboration with the e-Research Centre, who have competencies and significant engagements in energy and data analytic research.
The DIET project supports the development of new services to investigate coordinated analysis streams of consumption and logging data produced by smart meters.
These streams of meter status and error messages that are not normally retrieved give details on the physical environment of the meters and in conjunction with consumption data will enable creation of richer and more in-depth knowledge of system behaviour. The primary aim of these analytics are to discover and then recognise signatures for two different classes of events, possible meter failure scenarios and warnings on the occurrence of patterns indicative of meter attack/tamper.
British Gas, G4S, EDMI Smart Meters