World’s first carbon intensity forecast tool launches


National Grid, in partnership with Environmental Defense Fund Europe, University of Oxford Department of Computer Science and WWF, have developed the world’s first Carbon Intensity forecast with a regional breakdown.

The Carbon Intensity API uses state-of-the-art machine Llarning and sophisticated power system modelling to forecast the carbon intensity and generation mix 96+ hours ahead for each region in Great Britain.

Our OpenAPI allows consumers and smart devices to schedule and minimise CO2 emissions at a local level.

About the Carbon Intensity API

National Grid’s Carbon Intensity API provides programmatic and timely access to both forecast and estimated carbon intensity data.

The Carbon Intensity forecast includes CO2 emissions related to electricity generation only. The includes emissions from all large metered power stations, interconnector imports, transmission and distribution losses, and accounts for national electricity demand, embedded wind and solar generation.

The goal of this API service is to allow developers to produce applications that will enable consumers and/or smart devices to optimise their behaviour to minimise CO2 emissions.
Your browser does not support HTML5 video.

The carbon intensity of electricity is a measure of how much CO2 emissions are produced per kilowatt hour of electricity consumed.

Carbon intensity varies by hour, day, and season due to changes in electricity demand, low carbon generation (wind, solar, hydro, nuclear, biomass) and conventional generation.

National Grid forecasts the carbon intensity and generation mix of electricity consumed across 14 geographical regions in Great Britain. The spatial and temporal characteristics of carbon intensity can be observed in the map below.

The boundaries are defined according to Distribution Network Operator (DNO) boundaries.

The demand and generation by fuel type (gas, coal, wind, nuclear, solar etc.) for each region is forecast several days ahead at 30-min temporal resolution using an ensemble of state-of-the-art supervised Machine Learning (ML) regression models.

An advanced model ensembling technique is used to combine the ML models to generate a new optimised meta-model. The forecasts are updated every 30 mins using a nowcasting technique to adjust the forecasts a short period ahead.