Lucy Electric combines AI and digital twin tech for fault detection

Lucy Electric combines AI and digital twin tech for fault detection
Image: Lucy Electric

Advanced grid monitoring capabilities and AI are combined in a new solution for fault detection and asset management for underground cables by Lucy Electric.

The solution, named Synaps (Synchronous Analysis and Protection System), is for fault detection, classification and location using AI and machine learning to reduce faults on the LV network.

It works by using sensors at a substation and feeder level to detect anomalies in grid performance and via a ‘digital twin’ of the network pinpoints the probable location of intermittent faults.

With the solution, which was developed with UK government network innovation support in partnership with Scottish & Southern Electricity Networks and UK Power Networks, the time and cost of fault detection could be cut by up to two-thirds.

Have you read?
US lab testing new model for investigating grid faults
Energy Transitions Podcast: Improving power system efficiency – the science behind the energy transition

Early results from the solution show it can identify the probability of future failure by >95% and location accuracy >3 m, including spurs.

Paul Beck, Gridkey and Innovation Director at Lucy Electric, comments that detecting and repairing underground intermittent faults is often complex and costly.

“This exciting technology, when coupled with our existing GridKey monitoring system, allows improved fault management specifically the ability to carry out preventative maintenance before larger faults occur, delivering significant savings and reducing customer outages.”

Stewart Reid, Head of Future Networks at Scottish and Southern Electricity Networks, where Synaps is monitoring 16 circuits, says the solution is “already improving our service and benefitting our customers through a more reliable network at lower cost.”

Chino Atako, Senior Asset Engineer at UK Power Networks, which has seven circuits being monitored, says that learnings for so far indicate its potential to deliver significant ‘customer interruption’ and ‘customer minutes lost’ benefits.

The AI solution uses the ‘digital twin’ to simulate faults in millions of scenarios, with machine learning technology then comparing this to the measured network data captured when there is a current transient.

This is considered a highly effective way of locating cable faults as it works not just with large transients where there is an immediate chance of the fuse operating, but also with very small transients typical of the start of a cable fault.

Synaps is now being provided by Lucy Electric under the product name COPPsystem (Cable outage prediction and prevention).

Meanwhile trials are continuing to make it progressively more accurate and to get more experience of different cable types and network architectures.

In the future the plan is to extend its capability to the MV and HV networks, in particular offshore applications, where submarine cable faults account for up to 80% of insurance claims by operators.