Open AccessEkonometryczny model krótkoterminowego prognozowania zużycia gazu
Author(s) -
Tadeusz Kwilosz,
Instytut Nafty i Gazu – Państwowy Instytut Badawczy,
Bogdan Filar
Publication year - 2021
Publication title -
nafta gaz
Language(s) - English
Resource type - Journals
ISSN - 0867-8871
DOI - 10.18668/ng.2021.07.04
Subject(s) - term (time) , downtime , consumption (sociology) , econometrics , order (exchange) , econometric model , operations research , demand forecasting , computer science , economics , mathematics , social science , physics , finance , quantum mechanics , sociology , operating system
In order to develop a mathematical model of short-term gas demand, it is necessary to analyze the latest mathematical forecasting methods in order to select and adapt the right one (meeting the condition of efficiency and effectiveness). It is necessary to recognize and analyze factors (mainly environmental) affecting the result of short-term forecasts and sources of data that can be used. The result of the work is a numerical model of short-term gas demand for a selected territorial unit of the country. The developed model was calibrated and tested on historical data describing environmental conditions and real gas consumption. A heterogeneous linear econometric model was designed and calibrated on the basis of a selected set of attributes (explanatory variables). The estimated parameters of the model were statistically verified. It is worth noting that in the short term of the forecast (7 days) there are no significant changes in the gas market environment (launching new investments, connecting new users to the system, or changes in demand resulting from changing macroeconomic conditions). Other technical factors, such as production line failures at customers or industrial downtime, are difficult to predict, or knowledge about their occurrence is rarely available. For this reason, the only factors that may have an impact on changes in gas demand in the short term are weather factors, which were selected as explanatory variables for the developed model. Historical weather data was retrieved from the OpenWeatherMapHistoryBulk web service. Daily values of gas consumption for one of the voivodships of southern Poland were used as the response variable. The data was downloaded from the information exchange system of the transmission pipeline operator. The data covers a three-year period, as only such data has been made public. The explanatory variables include the daily values of weather data such as: average temperature, chilled temperature, minimum temperature, maximum temperature, atmospheric pressure, relative humidity, wind speed and wind direction.