search for




 

Forecasting analysis on daily gas supply for power generation by SARIMA-GARCH models
Journal of the Korean Data & Information Science Society 2025;36:41-57
Published online January 31, 2025;  https://doi.org/10.7465/jkdi.2024.36.1.41
© 2025 Korean Data and Information Science Society.

Siyeong Kim1 · Eunju Hwang2

12Department of Applied Statistics, Gachon University
Correspondence to: This work was supported by National Research Foundation of Korea (NRF-2023R1A2C1005395).
1 Undergraduate student, Department of Applied Statistics, Gachon University, Gyeonggi-do 13120, Korea.
2 Corresponding author: Professor, Department of Applied Statistics, Gachon University, Gyeonggi-do 13120, Korea. E-mail: ehwang@gachon.ac.kr
Received September 20, 2024; Revised November 6, 2024; Accepted November 6, 2024.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract
Accurate supply forecasting of natural gas is essential to respond to the world’s increasing energy demand and renewable energy volatility. In particular, the burden on the public is increasing due to failure to predict demand in recent years. Against this background, this study uses a time series model to more accurately predict daily gas supply for power generation. Initial analysis is performed by fitting the SARIMA model to the data. As comparing AIC, BIC, and Box-Jenkins methodologies, models with various orders are constructed. Through the diagnosis on the residuals of the SARIMA models, the existence of the conditional heteroscedasticity is revealed, form which, the SARIMA-GARCH model is proposed by fitting the residuals to the GARCH model. This model strengthens the insufficient explanatory power of the SARIMA model, whose residuals do not follow a normal distribution, and improves the accuracy of prediction. Prediction performance is compared and analyzed to derive the optimal model, which has been turned out to be the SARIMA(1,0,2)(1,1,1)7-GARCH(2,1) model. It is expected that the time series model and analysis results of this study contribute to the energy resource field by providing accurate gas supply.
Keywords : Forecasting, GARCH, natural gas for power generation, SARIMA