DATA-DRIVEN FORECASTING OF STEEL INDUSTRY ENERGY CONSUMPTION
DOI:
https://doi.org/10.69980/enxr1t24Keywords:
steel industry, energy forecasting, machine learning, Random Forest, industrial energy managementAbstract
The steel industry is energy-intensive and requires accurate short-term electricity forecasting for efficient production scheduling, managing peak demand, and energy management. Data-driven modelling has a lot of potential for increasing the accuracy of forecasts at the operational level, especially with high-frequency data. This study developed and evaluated an interpretable forecasting framework for short-term steel industry energy consumption using temporal and electrical operating variables. A dataset containing 35,040 observations recorded at 15-minute intervals over one year was analysed. Following data-quality assessment, leakage-prone variables were excluded, and lagged consumption, rolling statistics, electrical measurements, and cyclical temporal features were generated. Persistence, Seasonal Naïve, Linear Regression, Random Forest, and XGBoost models were evaluated using chronological training, validation, and independent test partitions. Forecasting performance was assessed using MAE, RMSE, MAPE, sMAPE, and R2, while permutation importance was used to interpret predictor contributions. Random Forest achieved the best overall performance with an RMSE of 8.4778 kWh and an R2of 0.9265, whereas XGBoost produced the lowest MAE of 4.5443 kWh. Both ensemble models outperformed Linear Regression and baseline approaches. Lagged energy consumption, lagged reactive power, and one-hour rolling statistics were identified as the most influential predictors. The proposed framework offers accurate and interpretable short-term energy forecasts and proves to be of practical use for industrial load scheduling, production planning and energy-efficiency management purposes.
References
1.Backman, J., Kyllönen, V., & Helaakoski, H. (2019). Methods and tools of improving steel manufacturing processes: Current state and future methods. IFAC-PapersOnLine, 52(13), 1174- 1179.
2.Batra, L., Harris, D., Katsigiannakis, G., Mackovyak, J., Parmar, H., & Scheller, M. (2025). Rising current: America’s growing electricity demand. ICF Report at https://www. ourenergypolicy. org/resources/rising-current-americas-growing-electricitydemand.
3.Bermeo-Ayerbe, M. A., Ocampo-Martinez, C., & Diaz-Rozo, J. (2022). Data-driven energy prediction modeling for both energy efficiency and maintenance in smart manufacturing systems. Energy, 238, 121691.
4.Branca, T. A., Fornai, B., Colla, V., Murri, M. M., Streppa, E., & Schröder, A. J. (2020). The challenge of digitalization in the steel sector. Metals, 10(2), 288.
5.Chen, X., Shuai, C., Wu, Y., & Zhang, Y. (2021). Understanding the sustainable consumption of energy resources in global industrial sector: Evidences from 114 countries. Environmental Impact Assessment Review, 90, 106609.
6.Gu, Y., Liu, W., Wang, B., Tian, B., Yang, X., & Pan, C. (2023). Analysis and prediction of energy, environmental and economic potentials in the iron and steel industry of China. Processes, 11(12), 3258.
7.He, K., & Wang, L. (2017). A review of energy use and energy-efficient technologies for the iron and steel industry. Renewable and Sustainable Energy Reviews, 70, 1022-1039.
8.Holappa, L. (2020). A general vision for reduction of energy consumption and CO2 emissions from the steel industry. Metals, 10(9), 1117.
9.Holmberg, K., & Erdemir, A. (2017). Influence of tribology on global energy consumption, costs and emissions. Friction, 5(3), 263-284.
10.Kapp, S., Choi, J. K., & Hong, T. (2023). Predicting industrial building energy consumption with statistical and machine-learning models informed by physical system parameters. Renewable and Sustainable Energy Reviews, 172, 113045.
11.Khalid, R., Ahmed, N., & Anglani, N. (2024, October). Predictive model for high energy consumption processes in the steel manufacturing industry using machine learning tools: industry 4.0 pathway. In 2024 IEEE International Conference on Artificial Intelligence & Green Energy (ICAIGE) (pp. 1-6). IEEE.
12.Khan, Z. A., Ullah, A., Haq, I. U., Hamdy, M., Mauro, G. M., Muhammad, K., ... & Baik, S. W. (2022). Efficient short-term electricity load forecasting for effective energy management. Sustainable Energy Technologies and Assessments, 53, 102337.
13.Maaouane, M., Zouggar, S., Krajačić, G., & Zahboune, H. (2021). Modelling industry energy demand using multiple linear regression analysis based on consumed quantity of goods. Energy, 225, 120270.
14.Mubarak, H., Sanjari, M. J., Stegen, S., & Abdellatif, A. (2023). Improved active and reactive energy forecasting using a stacking ensemble approach: steel industry case study. Energies, 16(21), 7252.
15.Sathishkumar, V. E., Lee, M. B., Lim, J. H., Shin, C. S., Park, C. W., & Cho, Y. Y. (2019). Hourly steel industry energy consumption prediction using machine learning algorithms.
In Proceedings of the Korea Information Processing Society Conference (pp. 585-588).
16.Sathishkumar, V. E., Lee, M., Lim, J., Kim, Y., Shin, C., Park, J., & Cho, Y. (2020). An energy consumption prediction model for smart factory using data mining algorithms. KIPS Transactions on Software and Data Engineering, 9(5), 153-160.
17.Sathishkumar, V. E., Shin, C., & Cho, Y. (2021). Steel Industry Energy Consumption [Data set]. UCI Machine Learning Repository. https://doi.org/10.24432/C52G8C
18.Shah, I., Iftikhar, H., Ali, S., & Wang, D. (2019). Short-term electricity demand forecasting using components estimation technique. Energies, 12(13), 2532.
19.Ugli, A. T. A., Pastorelli, M., Shuxratovich, K. K., Makhmudovich, M. K., & Qizi, K. Y. U. (2025, October). Rising Of Global Electricity Demand: The Importance Of Industry And Renewable Energy Integration. In 2025 14th International Conference on Renewable Energy Research and Applications (ICRERA) (pp. 1052-1059). IEEE.
20.Xu, Z. J., Zheng, Z., & Gao, X. Q. (2021). Operation optimization of the steel manufacturing process: A brief review. International Journal of Minerals, Metallurgy and Materials, 28(8), 1274-1287.
21.Yu, W., Patros, P., Young, B., Klinac, E., & Walmsley, T. G. (2022). Energy digital twin technology for industrial energy management: Classification, challenges and future. Renewable and Sustainable Energy Reviews, 161, 112407.
22.Zhang, X., Yang, L., & Ma, D. (2025). The importance of energy consumption and the need for efficiency in the steel industry using machine learning. Journal of Electrical Engineering & Tech
23.Zhao, J., Wang, W., & Sheng, C. (2018). Data-driven prediction for industrial processes and their applications. Springer International Publishing:
