Historically, the short-term operational planning process has used forecasting tools developed using algorithms and data appropriate for the provision of sufficiently accurate end-of-day supply and demand volumes over the week-ahead time horizon, the results of which have aided in the development of effective operational strategies for the GNCC as well as market facilitation through external publication.

Whilst these tools were fit for purpose in more ‘traditional’ gas years, the gas market has become increasingly volatile in recent years. Drivers of this change include, but are not limited to

• decreasing UK continental shelf supplies

• increasing price sensitive European and global imports (e.g. interconnectors and liquefied natural gas)

• increasing price sensitive demands (e.g. interconnectors and storage)

• increasing and more volatile power station demand in response to increased wind generation

• decreasing reliance on low pressure gas storage by distribution networks

• increasing system access requirements

all leading to increasing day-to-day and within-day supply and demand volatility. As a result, current forecast accuracy has become limited by an inability to model the gas market in sufficient detail and there is risk that, without research and development into enhanced forecast methods, GSO’s ability to sufficiently plan and prepare the NTS will be compromised, leading to an increase in consumer costs through inefficient asset operation and increased constraint management and trading costs.

To mitigate this risk, this project aims to gather detailed requirements for and subsequently research, develop and deliver a new prototype solution capable of providing sufficiently accurate combined supply and demand forecasts over the week-ahead time horizon. In particular, to address those challenging market behaviours detailed above, the project aims to include within the tool

• an improved treatment of price sensitive supplies and demands

• enhanced modeling of gas and electricity market interactions

• forecast confidence levels to better quantify the level of uncertainty

• forecast within-day flow rates in addition to end-of-day volumes

• the ability to be easily or automatically updated such that it remains fit for purpose

all of which are new and as yet unproven capabilities for GSO’s short-term planning process.

Benefits

The delivery of an offline, prototype, supply and demand forecasting tool with the functionality described in the Scope (above), tested and assessed against the current range of operational forecasting tools in terms of accuracy (including against the demand forecasting incentive already in place), ability to recreate market behaviours and functionality (including compatibility and ease of use with existing downstream processes/tools).