Journal of Smart Energy Systems and Sustainable Technologies

Articles

Vol. 2 (2026)

Machine Learning–Driven Global Forecasting of Electricity Generation and Consumption (1985–2035)

DOI:
https://doi.org/10.31875/2978-6436.2026.02.02
Submitted
July 2, 2026
Published
2025-07-02

Abstract

This paper examines long-term changes in global electricity generation and consumption from 1950 to 2035 using historical data from the Our World in Data (OWID) energy database "1950-2025". After addressing data inconsistencies through smoothing and gap-filling, we apply an ARIMA(1,1,1) model selected via Akaike Information Criterion minimization. While this univariate framework is deliberately simple, it provides a transparent, reproducible baseline for trend extrapolation. Forecasts from 2025 to 2035 indicate that global generation could reach approximately 35,750 TWh by 2035, with consumption slightly higher at about 36,400 TWh, implying a sustained deficit of roughly 650–700 TWh under a business-as-usual scenario. The 95% confidence intervals widen from ±2–3% in 2025 to ±7% by 2035.

This study does not claim algorithmic innovation. Its contribution lies in (i) applying a rigorously validated ARIMA framework to a uniquely long (75+ years) global electricity dataset; (ii) providing exhaustive model selection and robustness checks; and (iii) establishing a transparent baseline for benchmarking future hybrid extensions. The full code and data are provided for reproducibility.

Although this method is not a structural energy balance model, it provides a clear baseline based on historical trends, useful as a reference point for comparing policy options or more detailed modeling approaches.

References

  1. Wong, J. and J.R. Saravia, Distributed generation and alternative energy. Strategic Planning for Energy and the Environment, 2012. 32(2): p. 5. https://doi.org/10.1080/10485236.2012.10554229
  2. Sudarsanam, P., Y. Yamauchi, and P. Bharali, Heterogeneous Nanocatalysis for Energy and Environmental Sustainability, Volume 1: Energy Applications. 2022: John Wiley & Sons. https://doi.org/10.1002/9781119772057
  3. El-Hawary, M.E., Advances in Electric Power and Energy Systems: Load and Price Forecasting. 2017. https://doi.org/10.1002/9781119260295
  4. Yusoff, N.I., A.A.M. Zin, and A.B. Khairuddin. Congestion management in power system: A review. in 2017 3rd international conference on power generation systems and renewable energy technologies (PGSRET). 2017. IEEE. https://doi.org/10.1109/PGSRET.2017.8251795
  5. Misiurek, K., T. Olkuski, and J. Zyśk, Review of methods and models for forecasting electricity consumption. Energies, 2025. 18(15): p. 4032. https://doi.org/10.3390/en18154032
  6. Nti, I.K., et al., Electricity load forecasting: a systematic review. Journal of Electrical Systems and Information Technology, 2020. 7(1): p. 13. https://doi.org/10.1186/s43067-020-00021-8
  7. Box, G., Box and Jenkins: time series analysis, forecasting and control, in A Very British Affair: Six Britons and the Development of Time Series Analysis During the 20th Century. 2013, Springer. p. 161-215. https://doi.org/10.1057/9781137291264_6
  8. Hyndman, R.J. and G. Athanasopoulos, Forecasting: principles and practice. 2018: OTexts.
  9. Ouyang, Y., Self-organising mixture neural networks for enhanced modelling and forecasting of Fx time series. 2016: The University of Manchester (United Kingdom).
  10. Mughees, N., et al., Bi-LSTM-Based Deep Stacked Sequence-to-Sequence Autoencoder for Forecasting Solar Irradiation and Wind Speed. Computers, Materials & Continua, 2023. 75(3). https://doi.org/10.32604/cmc.2023.038564
  11. Antoniadis, A., et al., Statistical Learning Tools for Electricity Load Forecasting. 2024: Springer. https://doi.org/10.1007/978-3-031-60339-6
  12. Dehghanian, S.A., et al., Deep learning for solar power forecasting: A robust stacked LSTM algorithm for operational applications. Edelweiss Applied Science and Technology, 2025. 9(10): p. 1591-1600. https://doi.org/10.55214/2576-8484.v9i10.10722
  13. Jüngst, N., et al., Applications in Energy and Combustion Science. 2023.
  14. Kokoni, S., Estimation and projection of the demand for heating systems in the residential sector in England and Wales: economic theory, energy modelling and policy implications for heat pumps. 2020, University of Surrey.
  15. Forbes, K.F. and E.M. Zampelli, Accuracy of wind energy forecasts in Great Britain and prospects for improvement. Utilities Policy, 2020. 67: p. 101111. https://doi.org/10.1016/j.jup.2020.101111
  16. Roberts, E.M., Reducing Energy Consumption in Everyday Life: A study of landscapes of energy consumption in rural households and communities in North Wales. 2016, Cardiff University.
  17. Babrowski, S., P. Jochem, and W. Fichtner, Electricity storage systems in the future German energy sector: An optimization of the German electricity generation system until 2040 considering grid restrictions. Computers & Operations Research, 2016. 66: p. 228-240. https://doi.org/10.1016/j.cor.2015.01.014
  18. Vale, R., The Hockerton housing project, England, in Living within a fair share ecological footprint. 2013, Routledge. p. 262-274. https://doi.org/10.4324/9780203126448
  19. Eidiani, M. and M. Kargar. Frequency and voltage stability of the microgrid with the penetration of renewable sources. in 2022 9th Iranian Conference on Renewable Energy & Distributed Generation (ICREDG). 2022. IEEE. https://doi.org/10.1109/ICREDG54199.2022.9804542
  20. Klyuev, R.V., et al., Methods of forecasting electric energy consumption: A literature review. Energies, 2022. 15(23): p. 8919. https://doi.org/10.3390/en15238919
  21. Krstev, S., J. Forcan, and D. Krneta, An overview of forecasting methods for monthly electricity consumption. Tehnički vjesnik, 2023. 30(3): p. 993-1001. https://doi.org/10.17559/TV-20220430111309
  22. Heidary, S., et al., Carbon-aware machine learning for energy-efficient quantum data centers: Joint optimization of workload scheduling and cooling. The Global Environmental Engineers, 2025. 12: p. 61-81. https://doi.org/10.15377/2410-3624.2025.12.5
  23. Marisetty, N., Forecasting Selected International Stock Indices Returns by Using ARIMA Model. Available at SSRN 4898668, 2024. https://doi.org/10.2139/ssrn.4898668
  24. Heidary, S., R. Zahedi, and A. Ahmadi. AI-driven optimization of energy efficiency in HVAC systems through waste heat recovery and thermal energy storage. in 2025 6th International Conference on Optimizing Electrical Energy Consumption (OEEC). 2025. IEEE. https://doi.org/10.1109/OEEC66525.2025.11100185
  25. Elgqvist, E.M., Energy Storage Versus Back-up Generation: Energy Storage Overview. 2018, National Renewable Energy Lab.(NREL), Golden, CO (United States).
  26. Carmichael, R., R. Gross, and A. Rhodes, Unlocking the potential of residential electricity consumer engagement with Demand Response. Energy Futures Lab (Imperial College London), Briefing Paper, 2018. 120.
  27. Van der Meer, D.W., et al., Probabilistic forecasting of electricity consumption, photovoltaic power generation and net demand of an individual building using Gaussian Processes. Applied energy, 2018. 213: p. 195-207. https://doi.org/10.1016/j.apenergy.2017.12.104
  28. Amber, K., et al., Intelligent techniques for forecasting electricity consumption of buildings. Energy, 2018. 157: p. 886-893. https://doi.org/10.1016/j.energy.2018.05.155
  29. Bâra, A. and S.V. Oprea, Electricity Consumption and Generation Forecasting. Advanced Applications for Artificial Neural Networks, 2018: p. 119. https://doi.org/10.5772/intechopen.71239
  30. McGhee, R. and K. Svehla, Opportunity mapping for urban scale renewable energy generation. Renewable Energy, 2020. 162: p. 779-787. https://doi.org/10.1016/j.renene.2020.08.060
  31. Ahmadi, L., et al., Energy, exergy, exergoeconomic and exergoenvironment (4E) analysis method in energy systems engineering: advanced bibliometric mapping. Journal of Thermal Analysis and Calorimetry, 2025. 150(20): p. 15903-15918. https://doi.org/10.1007/s10973-025-14877-3
  32. Thota, V.R., Comparative Study of Time Series Forecasting for Trucking Shipments: Evaluating Arima, LSTM, and Hybrid Arima-LSTM Models. 2025, State University of New York at Binghamton.
  33. Chen, G., et al., Leveraging graph convolutional-LSTM for energy-efficient caching in blockchain-based green IoT. IEEE Transactions on Green Communications and Networking, 2021. 5(3): p. 1154-1164. https://doi.org/10.1109/TGCN.2021.3069395
  34. Riswandi, B.A., et al., AI Applications for Clean Energy and Sustainability. 2024: IGI Global. https://doi.org/10.4018/979-8-3693-6567-0
  35. Babaei, A.E., 2025 12th Iranian Conference on Renewable Energies and Distributed Generation (ICREDG), 26 February 2025, Qom University of Technology (QUT), Qom, Iran.
  36. Hiralal, P., et al., Nanomaterial-enhanced all-solid flexible zinc− carbon batteries. ACS nano, 2010. 4(5): p. 2730-2734. https://doi.org/10.1021/nn901391q
  37. Duxbury, D.F., The photochemistry and photophysics of triphenylmethane dyes in solid and liquid media. Chemical Reviews, 1993. 93(1): p. 381-433. https://doi.org/10.1021/cr00017a018