Projects
Projects: Projects for Investigator |
||
| Reference Number | NIA_UKPN0104 | |
| Title | AI for Visibility and Forecasting of Renewable Generation | |
| Status | Started | |
| Energy Categories | Renewable Energy Sources 60%; Other Cross-Cutting Technologies or Research 10%; Other Power and Storage Technologies(Electricity transmission and distribution) 30%; |
|
| Research Types | Applied Research and Development 100% | |
| Science and Technology Fields | PHYSICAL SCIENCES AND MATHEMATICS (Computer Science and Informatics) 50%; ENGINEERING AND TECHNOLOGY (Electrical and Electronic Engineering) 50%; |
|
| UKERC Cross Cutting Characterisation | Not Cross-cutting 100% | |
| Principal Investigator |
Project Contact UK Power Networks |
|
| Award Type | Network Innovation Allowance | |
| Funding Source | Ofgem | |
| Start Date | 01 October 2024 | |
| End Date | 28 February 2026 | |
| Duration | ENA months | |
| Total Grant Value | £455,000 | |
| Industrial Sectors | Power | |
| Region | London | |
| Programme | Network Innovation Allowance | |
| Investigators | Principal Investigator | Project Contact , UK Power Networks |
| Other Investigator | Project Contact , Eastern Power Networks plc Project Contact , UKPN London Power Networks plc |
|
| Industrial Collaborator | Project Contact , UK Power Networks |
|
| Web Site | https://smarter.energynetworks.org/projects/NIA_UKPN0104 |
|
| Objectives | Introduction to Open Climate FixThis project will develop highly accurate forecasts for the renewable assets in the UK Power Networks licence areas and improving the estimate of behind the meter capacity on the UK Power Networks network.MethodThe project aims to improve our metered and behind the meter solar and wind generation forecasts to reduce our use of flexibility and associated costs and better inform our network investment. The proposed method for this project is technical research using UK Power Networks data and open data in machine learning algorithms to implement a historical and live forecast service usable by UK Power Networks.1. Metered generation: OCF will train AI algorithms on historical data and run a real-time forecast for the metered solar and wind generation on the UK Power Networks network. The AI algorithms consume data from weather forecasts, geostationary weather satellite imagery (rare in forecasting algorithms) and real-time weather-based generation readings to generate expected and probabilistic generation figures for the network. This builds on the experience OCF has in forecasting the national solar outturn, which will be extended to the solar and wind sites in UK Power Networks region. The forecasts will be evaluated against UK Power Networks existing forecasts of renewable generation to assess the improvement. The user focus will be on the forecasts being available in real-time to enable improved decision-making to reduce curtailment and help with real time network control.2. Behind the meter/unmonitored generation: OCF will estimate the half-hourly behind the meter solar and wind generation on the UK Power Networks network. This innovation will compare metered data at substation level with the theoretical wind and solar generation predicted by a machine learning algorithm using historical weather observations. Across a long history, this should enable the estimation of the renewable capacity in each region. Disaggregation techniques have been used by OCF in domestic settings and their effectiveness will be tested at the network level. The estimates of generation will be undertaken across our entire network.Through work across OCF and UK Power Networks Forecasting team, the project will provide a more nuanced and reflective estimate of the business value of improved forecasts to UK Power Networks. This project scope aims to improve forecasts for all solar and wind generation within UK Power Networks network, including both metered and behind the meter. Employing technical research and machine learning algorithms, OCF will analyse data from UK Power Networks and open sources.OCF will train AI algorithms on historical data to deliver real-time forecasts, incorporating satellite imagery for heightened accuracy. Evaluations will compare AI-generated predictions with existing UK Power Networks forecasts, enhancing decision-making in the control room.A key innovation involves estimating behind the meter solar and wind generation capacity, utilising machine learning algorithms and comparing metered data at the substation level. The DSO will be able to reduce spend on procuring flexibility services through the use of more accurate generation profiles.The forecast model will be made available to UK Power Networks via an API for use in the DSOs operations. The objective of the project is for the DSO to have improved forecasts across intra-day and day ahead horizons compared to the current model in use, to improveday ahead flexibility decisions and real time network control. This will be delivered through: Improved accuracy of generation forecasts for metered generation Improved understanding of capacity of unmonitored/behind the meter generation | |
| Abstract | AI for Visibility and Forecasting of Renewable Generation aims to improve metered and behind the meter solar and wind generation forecasts to procure flexibility and reduce curtailment more efficiently, and better inform network investment. This will consist of the development of a machine learning algorithm that takes timeseries data from commercial customers and satellite imagery, weather forecasts and other open data to provide a live forecast service. This forecast will improve on the spatial and temporal granularity of existing forecasting, supporting the more efficient procurement of flexibility services and curtailment of generation, ultimately leading to financial savings for the DSO and customers, as well as CO2 savings. | |
| Data | No related datasets |
|
| Projects | No related projects |
|
| Publications | No related publications |
|
| Added to Database | 09/04/25 | |
