The main aim of the project is to develop new methods for forecasting the daily gas demand from the National Transmission System. These would be used to improve the demand forecasts that are provided to the market and used in operational decision-making.
The national gas demand forecast includes the demand from distribution networks, power stations, industrial loads, storage sites and interconnectors. Current systems calculate demand forecasts using a “bottom-up” approach, by combining forecasts for individual elements, using regression techniques. The model is complicated and reliability of results can be variable. An extra forecast is done a day ahead using a spreadsheet model; largely based on a “top-down” forecast of the total demand, also using regression, which gives a more accurate prediction. These are used as the basis for the forecasts published and used by the Gas National Control Room.
The project will involve exploring the use of new local predictor techniques for this activity, rather than current regression-based models. Improved demand forecasts will be of benefit to gas shippers in balancing their supply and demand portfolio, enhancing the information National Grid provides to its customers. They will also support improved operational planning, and more efficient balancing actions, and reduced costs met by the industry.
Benefits
Project success would be the delivery of a new method to improve the demand forecasts that are provided to the market and to be used in operational decision-making.
