AI project set to shorten storm power cuts

5th June 2020

Alex Lynn

0 0

An artificial intelligence system designed to predict the impact of a storm on the electricity network in order to shorten storm power cuts is to be trialled across the East and South East starting this winter. UK Power Networks’ ‘Storm Resilience’ project will use a machine learning algorithm to help control room staff decide where to send engineers, so they are ready to respond as storms hit.

A separate part of the project will trial lightning tracking software that could help restore power supplies caused by lightning strikes up to 90% faster.

Lightning strikes hit the ground about 300,000 times per year in the UK. The electricity network is built to be resilient and if a bolt of lightning strikes an overhead power line, or very nearby, the system ‘trips out’ in the same way as fuse boxes operate in most homes. Protection devices usually prevent the line being damaged so it can be re-energised to restore power supplies.

If successful, the new system will mean when a lightning strike trips a circuit, the control system quickly knows the grid coordinates of the location and can then automatically restore customers’ supplies in under three minutes – much faster than the current time of up to 30 minutes.

The team will use sophisticated machine learning techniques, combined with historical power cut data, to forecast the impact of gale force winds and storms on overhead power lines. This will help UK Power Networks to prepare for such weather, including moving engineers into the optimum places so they are poised to restore power supplies faster.

Ian Cameron, Head of Customer Service and Innovation at UK Power Networks, said: “This project is taking how we handle storms to a new level, by combining data science with improved customer service. Technology really is supercharging our capabilities right now, and we’re excited to see how it can help us be more efficient and, ultimately, keep the lights on even if the weather is fighting against us.”

The £660,000 trial will run until December 2021 with project partners’ global technology firm General Electric and DTN (formerly MeteoGroup).

Michael Eilts, DTN Senior Vice President-Weather, added: “We are excited to partner with UK Power Networks on this innovative and dynamic use of weather data. At DTN, we are committed to empowering our customers with actionable insights that help them manage their operation’s weather risks. UK Power Networks application of storm impact analytics is showcasing the future of utilities to mitigate impact and optimise restoration and recovery efforts.”