Load Forecasting on the Electricity Network

Abstract

Electricity used to be simple (well more simple than now). You made the electricity in big (often fossil fuel powered) power plants and then you transported it to where it was needed. With the move towards net zero, an increasing use of electricity rather than gas and an increasing use of smaller scale electricity generation that often depends on external factors such as wind, things are getting more complicated.

While these factors play a vital part in combating climate change, they also risk increasing prices and threatening electricity supply if we get it wrong. By better understanding electrical load on a system, we can make better decisions, increasing security and decreasing cost. This project aims specifically to forecast load at electrical substations to support the procurement of flexibility services - offerings that can move load from peak times to off peak times, supporting a more load-balanced network.

About the speaker

Dr Katrina Soderquest

Having started life in genetics and immunology, Katrina transitioned to data science, via bioinformatics, several years ago. After jobs at Tesco and Hotels.com, Katrina is now a Lead Data Scientist in the Energy and Transition Team at Faculty where she has predominantly worked with network operators (for both gas and electricity) to examine how we use data science and machine learning techniques to support the move towards net zero and the complications that brings.