A newly deployed AI-powered solar forecasting tool is expected to save the UK grid millions of pounds in balancing costs by providing more accurate predictions of PV generation.
Quartz Solar from Open Climate Fix is now being used in the National Energy System Operator (NESO) control room to forecast PV generation and inform decisions on how much reserve capacity should be purchased to ensure balanced supply and demand.
Since adoption, Quartz Solar has halved forecasting errors, according to Open Climate Fix, equivalent to saving at least GBP 30 million in imbalance cost per year, with potential to increase to GBP 150 million ($197.0 million) per year in 2035 if government solar capacity targets are met.
Developed in collaboration with NESO’s energy forecasting team, Quartz Solar combines machine learning, live satellite imagery and weather data to generate forecasts that are refined on a minute-by-minute basis. Following extensive testing, the AI-powered tool is now fully integrated into NESO’s operations.
The forecasting tool works by using machine learning techniques in conjunction with satellite imagery, weather data and historical PV data to predict solar generation within the same day, and up to 36 hours in advance.
A spokesperson from Open Climate Fix told pv magazine the non-profit company uses satellite data in 12 different spectral channels – meaning different wavelengths – which also provides a live view on clouds. This data includes infrared, visible terrain and water vapor.
Weather data is sourced from ECMWF and the United Kingdom’s Met Office, and includes irradiance, temperature, wind speed and direction at multiple discrete heights, air pressure, snow depth and cloud cover at multiple heights, and visibility.
Open Climate Fix uses PV Live as its main data source for PV generation. A spokesperson said the generation data is modeled on more than 1,000 sites and includes transmission and distribution data. “We believe this is the most accurate source of data the UK currently has for PV generation data,” they said.
Quartz Solar provides NESO with a national forecast as well as a regional breakdown based on UK grid supply points, comprising approximately 300 sites.
There are three key reasons Quartz Solar has improved upon the previous forecasting methods used by NESO, according to Open Climate Fix: speed of updates, the variety of data sources and a model that produces probabilistic output.
Traditional weather forecasting methods used by NESO are relatively slow, updating roughly every six hours. Open Solar’s machine-learning powered model updates at 15-minute intervals, which gives NESO’s control room a clearer picture of shifting patterns within the forecast.
Feeding the machine-learning model with a wide variety of data sources to train its output has resulted in a higher level of accuracy, according to the spokesperson, and more solar deployment should lead to even more accurate forecasting in the future. “The more data sources, the higher the accuracy. As the UK solar capacity increases, the data sources will increase, helping further improve accuracy over time.”
The probabilistic output produced by Open Climate Fix’s model PVNet provides a “most likely” scenario as well as 90th and 10th percentile predictions, providing NESO control room engineers with a broad view of possibilities to inform decision-making.
There are potential applications for Open Climate Fix's technology outside of the NESO control room, and a spokesperson confirmed that, while the organization is committed to its non-profit mission, it is exploring commercialization opportunities to fund further research. Open Climate Fix is also looking to enter other markets in the future, such as the Netherlands.
Founded in 2019 by Jack Kelly, formerly of Google DeepMind, and Dan Travers, formerly of Origin Energy and SunGard, Open Climate Fix is a non-profit organization engaged in energy system-related research. The UK-based organization works on a range of AI-powered energy system projects and its code is fully open source.
This content is protected by copyright and may not be reused. If you want to cooperate with us and would like to reuse some of our content, please contact: editors@pv-magazine.com.
By submitting this form you agree to pv magazine using your data for the purposes of publishing your comment.
Your personal data will only be disclosed or otherwise transmitted to third parties for the purposes of spam filtering or if this is necessary for technical maintenance of the website. Any other transfer to third parties will not take place unless this is justified on the basis of applicable data protection regulations or if pv magazine is legally obliged to do so.
You may revoke this consent at any time with effect for the future, in which case your personal data will be deleted immediately. Otherwise, your data will be deleted if pv magazine has processed your request or the purpose of data storage is fulfilled.
Further information on data privacy can be found in our Data Protection Policy.