Scottish and Southern Energy Networks (SSEN) owns and operates electricity networks across central southern England and northern Scotland. These supply over 3.7 million homes and businesses and comprise 106,000 substations and 130,000km of overhead lines and underground cables. It is committed to giving customers the best level of service 24/7, 365 days a year.
This commitment is hampered during severe weather. SSEN has therefore prioritised innovation to modelweather impact and response plans. In so doing, it has developed statistical models that predict network faults up to 48 hours in advance using weather forecasts.
However, deploying these for operational use is challenging. They must access live weather data and deliver regular updates that users can easily understand and act upon. The risks and costs associated with such deployment can hinder innovation from becoming business-as-usual.
Bellrock Technology has helped SSEN deploy a number of weather-based fault prediction models to support planning ahead of severe weather conditions. Working closely with network managers and innovators, the models were implemented as live applications that predict faults across 13 operational regions in England and Scotland.
Users can configure the applications to receive notifications for specific locations or wider geographical regions. They are used across the company in advance of storms to prioritise the dispatch of emergency crews to locations at risk.
Lumen® notifies us when problems are expected. This helps us decide where resources should be deployed and minimise the impact to our customers.
Jim Barber, SSEPD Network Operations Manager
The fault predictors are built on Lumen®, our smart delivery platform that makes it easy to prototype, build and deploy analytics applications. The platform reduces the risk and cost of innovation by quickly delivering applications across all users’ devices and uniquely selforganises to link them with live data without manual IT integration.
Lumen® has allowed SSEN to rapidly convert complex models, used only by expert users, to live fault predictors for use by all storm managers. It saves significant time in manual analysis, which puts a greater focus on restoring power to customers as quickly as possible.