SavedAdd your journal

Journals / Gazi University Journal of Science Part A: Engineering and Innovation / 2026 / Volume: 12 Issue: 2

Forecasting Natural Gas Consumption by User Type Using Machine Learning: A Comparative Study

Süleyman Tekin(IZMIR BAKIRCAY UNIVERSITY)0009-0009-4806-0671
Serhat Peker(IZMIR UNIVERSITY)0000-0002-6876-3982
Nurhan Dudaklı(IZMIR BAKIRCAY UNIVERSITY)0000-0002-5593-5335
Pages
—
Publication Date
2025-06-30
🏛️Bibliographic metadata and full text access for this article are provided via DergiPark.
DergiPark Record ↗
✨

AI Summary & Key Takeaways

AI Insight
⚡ TL;DR(Too Long; Didn't Read)

Results indicate that LightGBM and Random Forest provided the most accurate forecasts overall, with LightGBM performing best in the residential group and Random Forest slightly outperforming others in the commercial group.

📌 3 Key Takeaways

🎯
Objective & Hypothesis:This study aims to develop user-type-specific machine learning models to forecast natural gas consumption for residential and commercial user groups in İzmir, Turkey.
📊
Key Findings & Data:Multiple Linear Regression, Random Forest, LightGBM, and XGBoost algorithms were implemented, and model performance was enhanced through hyperparameter optimization.
💡
Conclusion & Impact:Results indicate that LightGBM and Random Forest provided the most accurate forecasts overall, with LightGBM performing best in the residential group and Random Forest slightly outperforming others in the commercial group.
🔬Design:Modeling & Simulation
📍Scope:Comprehensive Sample
⚙️Variables:Natural Gas Consumption · Energy Forecasting · Machine Learning

Abstract

This study aims to develop user-type-specific machine learning models to forecast natural gas consumption for residential and commercial user groups in İzmir, Turkey. Multiple Linear Regression, Random Forest, LightGBM, and XGBoost algorithms were implemented, and model performance was enhanced through hyperparameter optimization. The models were evaluated using MAE and RMSE metrics. Results indicate that LightGBM and Random Forest provided the most accurate forecasts overall, with LightGBM performing best in the residential group and Random Forest slightly outperforming others in the commercial group. In contrast, MLR underperformed due to the non-linear nature of the data. Residential consumption patterns were found to be more predictable, leading to lower error rates, whereas the commercial group exhibited higher variability and forecast challenges. The study highlights the importance of distinguishing user types and employing well-tuned machine learning algorithms for improved energy demand forecasting.

🔖

Cite This Article

Tekin, S., Peker, S., Dudaklı, N. (2025). Forecasting Natural Gas Consumption by User Type Using Machine Learning: A Comparative Study. Gazi University Journal of Science Part A: Engineering and Innovation, Volume: 12 Issue: 2. https://doi.org/10.54287/gujsa.1689278

📄 Full Text PDF Viewer

📄

PDF görüntüleyici hazırlanıyor...

References (43)

  1. Akpınar, H., & Yumuşak, S. (2020). Sakarya ilinde doğal gaz talep tahmini: SARIMA modeli uygulaması. Enerji ve Ekonomi Araştırmaları Dergisi, 5(1), 23–35.
  2. Ay, M. (2018). Yozgat ilinde doğal gaz tüketiminde eğilim analizi: ANN uygulaması. İstatistik ve Uygulamalı Bilimler Dergisi, 10(1), 45–56.
  3. Ayaz, H., & Kılıç, A. (2018). Danimarka’da doğal gaz tüketimi tahmini: Yapay sinir ağları ile bir uygulama. Uluslararası Enerji Ekonomisi ve Politika Dergisi, 8(4), 100–108.
  4. Aydemir, S., & Kılıç, M. (2020). Türkiye'de doğal gaz tüketim tahmini için makine öğrenimi yöntemlerinin uygulanması. Enerji ve Çevre Dergisi, 29(3), 150–165.
  5. Aydın, R., Yüksel, S., Silahtaroğlu, G., & Dinçer, H. (2022). Doğal gaz tüketiminin MARS algoritmasıyla tahmini ve yenilenebilir enerji ile ilişkisi. Yapay Zeka ve Enerji Sistemleri Dergisi, 8(2), 111–125.
  6. Balıkçı, A., Şahin, E., & Karaca, T. (2024). İstanbul’da şehir giriş istasyonları için doğal gaz tüketimi tahmini: ANN tabanlı bir model. Enerji Sistemleri Mühendisliği Dergisi, 12(1), 88–104.
  7. Beyca, A., Yılmaz, B., & Yıldız, S. (2019). Doğal gaz tüketim tahmini için SVR, MLR ve ANN karşılaştırması. Mühendislik Bilimleri Dergisi, 25(2), 157–168.
  8. Breiman, L. (2001). Random forests. Machine Learning, 45(1), 5–32. https://doi.org/10.1023/A:1010933404324
  9. Brown, T., & Zhang, Y. (2019). The Role of Natural Gas in Emerging Energy Markets. International Energy Journal, 17(2), 56–70.
  10. Çetin, T., Turan, M. E., & Şenol, M. E. (2025). Effects of Decision Variables Selection on Sewer Optimization Problem. Applied Sciences, 15(9), 4836. https://doi.org/10.3390/app15094836
  11. Chen, T., & Guestrin, C. (2016). XGBoost: A scalable tree boosting system. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 785–794.
  12. Chen, W., Lin, Y., & Liu, W. (2018). A sectoral comparison of natural gas demand prediction. Energy Economics, 71, 365–378.
  13. Dalkılıç, M., & Topal, M. (2014). Elazığ ilinde doğal gaz talebinin ARIMA modeli ile tahmini. Enerji ve Çevre Dergisi, 6(3), 55–63.
  14. Erdogdu, E. (2010). Natural gas demand in Turkey. Applied Energy, 87(1), 211–219.
  15. Eren, T., & Kaçtıoğlu, S. (2017). Türkiye’deki doğal gaz tüketimi ve gri tahmin metoduyla tahmin edilmesi. İstanbul Ticaret Üniversitesi Fen Bilimleri Dergisi, 16(31), 23-41.
  16. Es, F. (2021). SARIMA ve gri model karşılaştırması: Türkiye doğal gaz talep tahmini. Veri Bilimi ve Enerji Analizi Dergisi, 3(2), 90–102.
  17. Evcim, M. (2019). Derin öğrenme ile kısa vadeli enerji tüketimi tahmini. Elektrik Elektronik ve Bilgisayar Bilimleri Dergisi, 11(1), 37–47.
  18. Garcia, L., Lee, M., & Smith, J. (2020). Hyperparameter tuning in energy forecasting models: Empirical evidence from LightGBM and Random Forest. Energy Informatics, 3(1), 1–15.
  19. Harold, J., Lyne, M., & Lyons, S. (2015). Microeconomic determinants of residential gas demand in Ireland. Energy Policy, 85, 260–269.
  20. Ji, Q., Zhou, D., & Geng, J. (2018). Natural gas consumption and economic growth: Empirical evidence from China. Energy Policy, 122, 17–28.
  21. Johnson, M., Lee, S., & Kim, H. (2021). Sustainable Energy Planning and Forecasting. Journal of Energy Systems, 14(3), 234–245.
  22. Jones, R. (2019). Energy Demand Forecasting for Smart Cities. International Journal of Energy Policy Studies, 12(2), 133–149.
  23. Joshi, M. (2021). Panel regression analysis of natural gas demand in the United States. Energy Reports, 7, 889–897.
  24. Ke, G., Meng, Q., Finley, T., Wang, T., Chen, W., Ma, W., Ye, Q., & Liu, T. Y. (2017). LightGBM: A highly efficient gradient boosting decision tree. Advances in Neural Information Processing Systems, 30, 3146–3154.
  25. Khan, M. A. (2015). Demand forecasting for natural gas in Pakistan. Pakistan Journal of Energy and Environment, 4(1), 33–44.
  26. Kılınç, N. Ş., & Çoban, O. (2017). Sanayi sektöründe enerji talep esnekliklerinin tahmini: OECD ülkeleri örneği, Mehmet Akif Ersoy Üniversitesi Sosyal Bilimler Enstitüsü Dergisi, 9(20), 479-496. https://doi.org/10.20875/makusobed.342575
  27. Kutner, M. H., Nachtsheim, C. J., & Neter, J. (1998). Applied regression analysis (4th ed.). Wiley.
  28. Oruç, K., & Çelik, Ş. (2017). Isparta ili için doğal gaz talep tahmini. Süleyman Demirel Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi, 22(1), 31-42.
  29. Ozdemir, G., Aydemir, E., Olgun, M. O., & Mulbay, Z. (2016). Forecasting of Turkey natural gas demand using a hybrid algorithm. Energy Sources, Part B: Economics, Planning, and Policy, 11(4), 295-302.
  30. Palinski, A. (2019). Artificial neural networks vs. classical regression in energy prediction. Energy Management Review, 11(3), 78–91.
  31. Qiao, X., Jin, L., & Chen, M. (2020). Whale optimization algorithm-based Volterra filter for gas demand prediction. Expert Systems with Applications, 150, 113253.
  32. Qin, X., Zhang, X., & Yu, B. (2023). A federated contrastive learning approach for city-level gas consumption forecasting. Applied Energy, 321, 119312.
  33. Şengün, G. (2012). Doğal Gaz Talep Tahmini Bayburt İli Üzerine Bir Uygulama, Yayınlanmamış Yüksek Lisans Tezi, A.Ü. Sosyal Bilimler Enstitüsü, Erzurum.
  34. Shaikh, F., & Ji, Q. (2016). Forecasting China's natural gas consumption by 2035: Logistic model approach. Resources Policy, 49, 292–300.
  35. Smith, A., Brown, M., & Davis, R. (2020). Strategic energy management for urban resilience. Energy Research and Social Science, 68, 101573.
  36. Su, Y., Li, K., & Zhao, M. (2019). Hybrid short-term gas load forecasting using Bi-LSTM and GA-based wavelet transform. Energy Procedia, 158, 4517–4522.
  37. Thornton, H., Martin, C., & Evans, R. (2016). Modelling temperature effects on residential gas demand in the UK. Journal of Environmental Economics and Policy, 5(1), 48–66.
  38. Tuna, Ç. (2019). Erzurum ilinde doğal gaz talep tahmini: SARIMA modeli ile analiz. MSc Thesis, Atatürk University, Erzurum.
  39. Wadud, Z., Noman, A. M., & Ahmed, R. (2011). Forecasting gas demand in Bangladesh: An econometric approach. Energy, 36(10), 6120–6126.
  40. Wang, J., Chen, B., & Xu, T. (2021). Deep learning applications in short-term load forecasting. Energy AI, 2, 100036.
  41. Yıldız, C. (2015). Kayseri ilinde doğal gaz tüketimi tahmini: ANN, gri tahmin ve Box-Jenkins karşılaştırması. Veri Analizi ve Uygulama Araştırmaları Dergisi, 7(2), 66–78.
  42. Yu, Y., Yang, S., & Zhang, H. (2014). Regional variation in natural gas demand: Panel data analysis of Chinese cities. Resources and Energy Economics, 36(2), 291–310.
  43. Zha, D., Li, J., & Tang, Y. (2022). CNN-LSTM based hybrid model for natural gas production prediction. Applied Energy, 307, 118251.