Daily basis mid-term demand forecast of city natural gas using univariate statistical techniques

Yıl 2020, Cilt: 35 Sayı: 2, 725 - 742, 25.12.2019

Mustafa Akpinar , Nejat Yumuşak

https://doi.org/10.17341/gazimmfd.494094

Cited By: 3

Öz

Natural gas is the most basic energy source used today in energy production, heating and cooking. With its widespread network, houses, industrial enterprises and power plants can access this energy at any time. The natural gas used in Turkey is an imported energy source and its agreements is provided by long-term contracts. Long-term contracts are submitted to the domestic market by decision-makers. In this process, natural gas supply companies and wholesale companies, provide gas supplies to city distribution companies and industrial establishments with annual contracts (mid-term). City distribution or other companies are required to report monthly and year based daily consumption demand forecasts in these contracts. This paper studies forecasting of daily and monthly demand for mid-term natural gas as contract estimations using time series decomposition, Holt-Winters and ARIMA / SARIMA models, which are statistical methods, include univariate seasonality. In the study, 365-day forecast is performed on a daily basis and 12-month forecast is performed on a monthly basis at once. As a result of daily natural gas estimation, the lowest error is realized by ARIMA(0,0,1)1(0,1,1)³⁶⁵ model with 23.68% MAPE in the year ahead prediction. In the monthly conversion, the lowest estimation model is realized by ARIMA(1,0,1)1(1,1,1)³⁶⁵ model with 11.84% MAPE. The results show that seasonal ARIMA models are the most suitable among the univariate techniques. The fact that many predictions can be made at a time and the results are acceptable allow these techniques to be used by decision-makers.

Anahtar Kelimeler

Demand forecasting, natural gas, time series decomposition, Holt-Winters model, ARIMA/SARIMA models

Günlük temelli orta vadeli şehir doğal gaz talebinin tek değişkenli istatistik teknikleri ile tahmini

Öz

Doğal gaz günümüzde enerji üretimi, ısınma ve pişirmede kullanılan en temel enerji kaynağıdır. Yaygın ağ yapısı ile birlikte evler, sanayi kuruluşları, santraller istedikleri anlarda bu enerjiye erişebilmektedir. Türkiye’de doğal gaz ithal bir enerji kaynağıdır ve uzun dönemli sözleşmeler ile anlaşmalar sağlanmaktadır. Uzun dönemli sözleşmeler karar vericiler tarafından yurtiçine arz edilir. Bu arz sürecinde doğal gaz tedarik şirketleri ve toptan satış şirketleri şehir dağıtım şirketleri ve sanayi kuruluşlarına yıllık sözleşmeler ile gaz arzı sağlarlar. Şirketler ve şehir dağıtım şirketleri bu sözleşmelerde aylık, yıl içinde de günlük tüketim talep tahminlerini bildirmekle yükümlüdür. Bu çalışma günlük ve aylık temelde orta vadeli doğal gaz talep tahminini tek değişkenli mevsimsellik içeren istatistiki yöntemler olan zaman serileri ayrıştırılması, Holt-Winters ve ARIMA/SARIMA modelleri ile gerçekleştirmiştir. Yapılan çalışmada günlük temelde 365 günlük, aylık temelde de 12 aylık tahmin bir anda gerçekleştirilmiştir. Doğal gaz tahmini sonucu günlük temelde en düşük hata yıl öncesi tahminde ARIMA(0,0,1)1(0,1,1)³⁶⁵ modeli ile 23,68%  MAPE ile gerçekleşmiştir. Aylık dönüşümde ise en düşük tahmin modeli ARIMA(1,0,1)1(1,1,1)³⁶⁵ modeli ile 11,84% MAPE ile gerçekleşmiştir. Bu sonuçlar mevsimsel ARIMA modellerinin tek değişkenli teknikler arasında en uygun olduğunu göstermiştir. Bir anda çok sayıda tahmin yapılabilmesine imkan tanıması ve sonuçlarının kabul edilebilir olması bu tekniklerin karar vericiler tarafından kullanılabilmesine olanak tanımaktadır.

Anahtar Kelimeler

Talep tahmini, doğal gaz, zaman serilerinin ayrıştırılması, Holt-Winters modeli, ARIMA/SARIMA modelleri

Kaynakça

• X. Wang, D. Luo, J. Liu, W. Wang, and G. Jie, “Prediction of Natural Gas Consumption in Different Regions of China Using a Hybrid MVO-NNGBM Model,” Math. Probl. Eng., vol. 2017, pp. 1–10, 2017.
• M. Akpinar, M. Adak, and N. Yumusak, “Day-Ahead Natural Gas Demand Forecasting Using Optimized ABC-Based Neural Network with Sliding Window Technique: The Case Study of Regional Basis in Turkey,” Energies, vol. 10, no. 6, p. 781, Jun. 2017.
• F. Scarpa and V. Bianco, “Assessing the Quality of Natural Gas Consumption Forecasting: An Application to the Italian Residential Sector,” Energies, vol. 10, no. 11, p. 1879, Nov. 2017.
• O. Laib, M. T. Khadir, and L. Mihaylova, “A Gaussian Process Regression for Natural Gas Consumption Prediction Based on Time Series Data,” in 2018 21st International Conference on Information Fusion (FUSION), 2018, pp. 55–61.
• N. Wei et al., “Short-Term Forecasting of Natural Gas Consumption Using Factor Selection Algorithm and Optimized Support Vector Regression,” J. Energy Resour. Technol., vol. 141, no. 3, p. 032701, Oct. 2018.
• I. Naim and T. Mahara, “Comparative Analysis of Univariate Forecasting Techniques for Industrial Natural Gas Consumption,” Int. J. Image, Graph. Signal Process., vol. 10, no. 5, pp. 33–44, May 2018.
• M. Akpinar and N. Yumusak, “Naïve forecasting of household natural gas consumption with sliding window approach,” Turkish J. Electr. Eng. Comput. Sci., vol. 25, no. 1, pp. 30–45, 2017.
• H. A. Es, F. Y. Kalender, and C. Hamzaçebi, “Forecasting the Net Energy Demand of Turkey by Artificial Neural Networks,” J. Fac. Eng. Archit. Gazi Univ., vol. 29, no. 3, Sep. 2014.
• F. Jiang, X. Yang, and S. Li, “Comparison of Forecasting India’s Energy Demand Using an MGM, ARIMA Model, MGM-ARIMA Model, and BP Neural Network Model,” Sustainability, vol. 10, no. 7, p. 2225, Jun. 2018.
• M. Fırat, M. A. Yurdusev, and M. Mermer, “Monthly Water Demand Forecasting by Adaptive Neuro-Fuzzy Inference System Approach,” J. Fac. Eng. Archit. Gazi Univ., vol. 23, no. 2, pp. 449–457, 2008.C. Hamzaçebi and F. Kutay, “Electric Consumption Forecasting of Turkey Using Artificial Neural Networks Up to Year 2010,” J. Fac. Eng. Archit. Gazi Univ., vol. 19, no. 3, pp. 227–233, 2004.
• S. Rehman, Y. Cai, R. Fazal, G. Das Walasai, and N. Mirjat, “An Integrated Modeling Approach for Forecasting Long-Term Energy Demand in Pakistan,” Energies, vol. 10, no. 11, p. 1868, Nov. 2017.
• M. Bulut and B. Başoğlu, “Development of a Hybrid System Based on Neural Networks and Expert Systems for Short-Term Electricity Demand Forecasting,” J. Fac. Eng. Archit. Gazi Univ., vol. 32, no. 2, Jun. 2017.D. Šebalj, J. Mesarić, and D. Dujak, “Predicting Natural Gas Consumption – A Literature Review,” in Central European Conference on Information and Intelligent Systems 2017, 2017, pp. 293–300.
• V. Prema and K. U. Rao, “Time series decomposition model for accurate wind speed forecast,” Renewables Wind. Water, Sol., vol. 2, no. 1, p. 18, Dec. 2015.
• R. A. Yaffee and M. McGee, Introduction to Time Series Analysis and Forecasting: With Applications of SAS and SPSS, 1st ed. Orlando, FL: Academic Press, 2000.
• Stephen A. DeLurgio, Forecasting principles and applications. Boston: Irwin McGraw-Hill, 1998.
• M. Akpinar and N. Yumusak, “Year Ahead Demand Forecast of City Natural Gas Using Seasonal Time Series Methods,” Energies, vol. 9, no. 9, p. 727, Sep. 2016.
• S. Makridakis, S. C. Wheelwright, and H. R. J., Forecasting : methods and applications. New York: John Wiley, 2008.
• S. Delurgio and C. Bhame, Forecasting Systems for Operations Management. New York: Irwin Professional Pub, 1991.
• H. Ma and Y. Wu, “Grey predictive on natural gas consumption and production in China,” in Proceedings of the 2009 2nd Pacific-Asia Conference on Web Mining and Web-Based Application, WMWA 2009, 2009, pp. 91–94.
• P. R. Winters, “Forecasting Sales by Exponentially Weighted Moving Averages,” Manage. Sci., vol. 6, no. 3, pp. 324–342, Apr. 1960.
• G. E. P. Box, G. M. Jenkins, and G. C. Reinsel, Time Series Analysis, 5th ed. Hoboken, NJ: John Wiley & Sons, Inc., 2008.
• M. Valipour, “Long-term runoff study using SARIMA and ARIMA models in the United States,” Meteorol. Appl., vol. 22, no. 3, pp. 592–598, Jul. 2015.
• R. Burtiev, F. Greenwell, and V. Kolivenko, “Time Series Analysis of Wind Speed and Temperature in Tiraspol, Moldova,” vol. 12, no. 1, pp. 23–33, 2013.
• S. I. Inc., SAS/ETS® 13.2 User’s Guide. North Calorina: SAS Institute Inc., 2014.
• P. S. P. Cowpertwait and A. V. Metcalfe, Introductory Time Series with R. New York: Springer, 2009.
• J. D. Hamilton, “Time Series Analysis,” Prentice Hall New Jersey 1994, vol. SFB 373, no. Chapter 5, pp. 837–900, 1994.
• W. W. S. Wei, Time Series Analysis: Univariate And Multivariate Methods, 2nd Editio. Boston: Pearson Addison Wesley, 2006.
• G. Wypych, “Solvent Use in Various Industries: Asphalt Compounding,” in Handbook of Solvents: Second Edition, vol. 2, Elsevier Inc., 2014, pp. 13–14.
• D. A. Dickey and W. A. Fuller, “Distribution of the Estimators for Autoregressive Time Series With a Unit Root,” J. Am. Stat. Assoc., vol. 74, no. 366, p. 427, Jun. 1979.
• E. Said and D. A. Dickey, “Testing for unit roots in autoregressive-moving average models of unknown order,” Biometrika, vol. 71, no. 3, pp. 599–607, 1984.
• P. C. B. Phillips and P. Perron, “Testing for a unit root in time series regression,” Biometrika, vol. 75, no. 2, pp. 335–346, 1988.
• L. Gelažanskas and K. Gamage, “Forecasting Hot Water Consumption in Residential Houses,” Energies, vol. 8, no. 11, pp. 12702–12717, Nov. 2015.
• Q. Wu and C. Peng, “Wind Power Generation Forecasting Using Least Squares Support Vector Machine Combined with Ensemble Empirical Mode Decomposition, Principal Component Analysis and a Bat Algorithm,” Energies, vol. 9, no. 4, p. 261, Apr. 2016.