Natural Power has contributed to research that has resulted in a short-term (less than four hours) power forecasting system that demonstrates up to 20% improved forecast accuracy.
The ability to accurately forecast minutes and hours ahead is essential for the continued growth of weather-dependent renewable energy generation, like wind and solar.
Rosemary Tawn completed a PhD, the result of a collaboration between the University of Strathclyde, Scotland’s Innovation Centre for data and artificial intelligence, The Data Lab, as well as Natural Power.
Iain Dinwoodie, Head of Advanced Performance Engineering at Natural Power, said: “This project really highlights the benefits of academic collaboration with industry, bringing together brilliant minds and unrivalled data.
“By applying a unique combination of cutting-edge data science with domain knowledge from statistical learning, meteorology and operational experience, we have produced a world-leading forecasting system.”
Natural Power is the biggest independent provider of control point services to generators in the UK and its ability to forecast accurately across its asset management contracts has “very significant commercial impacts” on both the individual generators and the UK grid as a whole.
The project has provided Natural Power with a robust methodology for using statistical, data driven, short-term forecasts, in conjunction with numerical weather prediction-based estimates, to create an optimum forecast for short and long-term power requirements.
As a result, the system has overcome the real-world challenge of missing data which has been a barrier to using statistical data in wind power applications until this time.
Tawn said: “Collaborating with Natural Power throughout my PhD has focused my research on current problems that are most relevant to industry and has allowed my work to have immediate impact in the real world.
“I am looking forward to continuing my career in the energy sector in Scotland and using data science to progress the energy transition.”
The new power forecasting system has been implemented with immediate effect.