Gas Demand Forecasting

Abstract

The paper describes the various kindsof forecasting of gas demand which are carried out within the British Gas Corporation (BGC). Forecasting is considered according to the period ahead to which it refers, and separate sections consider: within day and day-to-day; between one month and one year ahead; between one and five years ahead; beyond five years.The paper is descriptive rather than normative and it does not claim to cover every application exhaustively. The major applications are day-to-day forecasting of total gas send-out, for control purposes, and annual and peak day demand forecasting for the period one to five years ahead for planning purposes.The relationship between demand and temperature is relatively stable over periods of a few months and it has been found that Box-Jenkins models give good day-to-day forecasts. The form of these models is described and the problems of implementing them in practice are discussed.In the case of annual demand forecasting each Region makes forecasts for its own geographical area for each market sector within a framework of economic and marketing assumptions set by Headquarters. For each annual demand forecast a demand/temperature relationship is forecast which is consistent with it. The paper describes this relationship and the way it is used to forecast the peak demand corresponding to specified severe winter conditions.

Probabilistic anomaly detection in natural gas time series data

“…Regression analysis is a statistical method that is used widely for electricity and natural gas demand forecasting (Aras and Aras, 2004, Hong, 2014, Hyndman and Fan, 2010, Lyness, 1984and Nedellec et al, 2014. It has also been used in combination with a penalty function for outlier detection (Zou, Tseng, & Wang, 2014).…”

Time series analysis applied to construct US natural gas price functions for groups of states

“…With over 80 interstate, long-distance pipelines in the US (Doane and Spulber, 1994), each one providing services to regions with different specifics, it is only natural to assume that such diverse locations have different econometric data. When doing cross-regions studies of various aspects of the supply chain, such as the forecasting of demand (Gutierrez et al, 2005;Lyness, 1984), or the balancing of the pipelines after imbalances have been created by the natural gas shippers (Dempe et al, 2005;Kalashnikov and Ríos-Mercado, 2006;Keyaerts et al, 2009), or the dynamics of interstateintrastate systems (Huntington, 1978), one has to take into account the existence of different markets.…”

Identifying the Effect of Public Holidays on Daily Demand for Gas

“…Brabec et al (2009) and Gascón and Sánchez-Úbeda (2018)) and fixed effects to represent the day of the week. These are needed because gas demand follows a weekly cycle, with clear differences between weekdays and weekends (Lyness, 1984). Similarly, because demand is affected markedly by public holidays, models usually include a holiday factor (Brabec et al, 2015) or treat public holidays like days of the weekend (e.g.…”