Performance Analysis of a Grid-Connected Solar PV System for Industrial Loads Using AI-Assisted Techniques

Bibi Ayesha; Bilal Ur Rehman; Muhammad Amir; Kifayat Ullah; Waqar Ishaq

Authors

Bibi Ayesha Department of Electrical Engineering, University of Engineering and Technology, Peshawar, Pakistan
Bilal Ur Rehman Department of Electrical Engineering, University of Engineering and Technology, Peshawar, Pakistan
Muhammad Amir Department of Electrical Engineering, University of Engineering and Technology, Peshawar, Pakistan
Kifayat Ullah Department of Electrical Engineering, University of Engineering and Technology, Peshawar, Pakistan
Waqar Ishaq Department of Telecommunication, Hazara University, Mansehra, Pakistan

Keywords:

Machine Learning, Grid-Connected Solar Photovoltaic (PV) System, Industrial Energy Demand Forecasting, Carbon Emission Reduction, Pakistan Energy Sector

Abstract

The industrial sector in Pakistan faces increasing challenges due to rising electricity costs, grid instability, and growing environmental concerns. Although grid-connected solar photovoltaic (PV) systems offer a promising solution, conventional performance assessment approaches often fail to capture the dynamic behavior of industrial loads and solar energy generation. This study proposes an Artificial Intelligence (AI)-assisted framework for evaluating the technical, economic, and environmental performance of a grid-connected PV system supplying industrial loads. Historical meteorological data from the Pakistan Meteorological Department (PMD) and industrial electricity demand data from the Lahore Electric Supply Company (LESCO) for 2023 were utilized. Four machine-learning models, namely Artificial Neural Network (ANN), Support Vector Regression (SVR), Random Forest Regression (RFR), and Long Short-Term Memory (LSTM), were developed and compared for load and solar power forecasting. Among the evaluated models, LSTM achieved the highest forecasting accuracy, with Root Mean Square Error (RMSE) values of 32.5 kW for load forecasting and 18.5 kW for solar forecasting, representing approximately 48% improvement over the baseline Linear Regression model. Based on LSTM forecasts, a 500 kWp grid-connected PV system generated 825,000 kWh annually, achieved 78% self-consumption, and reduced grid electricity imports by 643,500 kWh/year. Economic analysis revealed annual net savings of Rs 20.80 million, a payback period of 3.37 years, a Net Present Value (NPV) of Rs 112 million, and an Internal Rate of Return (IRR) exceeding 28%. Furthermore, the system reduced carbon emissions by approximately 412.5 tonnes CO₂ annually. (or 412.5 metric tons of CO₂ annually. These findings demonstrate the effectiveness of AI-assisted forecasting for reliable industrial PV-system evaluation and investment planning.

References

A. E. Cruceru et al., “Exploring the Dimensions of Energy Security in Relation to the National Power Grid,” Curr. Approaches Eng. Res. Technol. Vol. 10, pp. 54–69, Dec. 2024, doi: 10.9734/BPI/CAERT/V10/3282.

“Electricity Mid-Year Update - July 2024 – Analysis - IEA.” Accessed: Jun. 26, 2026. [Online]. Available: https://www.iea.org/reports/electricity-mid-year-update-july-2024

H. Ritchie, “Sector by sector: where do global greenhouse gas emissions come from?,” Our World Data, Sep. 2020, Accessed: Jun. 26, 2026. [Online]. Available: https://ourworldindata.org/ghg-emissions-by-sector

“Explore the 2024 State of the Industry reports”, [Online]. Available: https://gfi-apac.org/state-of-the-industry-2024/

“Renewable Power Generation Costs in 2023”, [Online]. Available: https://www.irena.org/-/media/Files/IRENA/Agency/Publication/2024/Sep/IRENA_Renewable_power_generation_costs_in_2023.pdf

“Govt notifies cuts in industrial electricity tariffs by up to Rs4.58/unit, providing relief to businesses - Profit by Pakistan Today.” Accessed: Jun. 26, 2026. [Online]. Available: https://profit.pakistantoday.com.pk/2026/02/17/govt-notifies-cuts-in-industrial-electricity-tariffs-by-up-to-rs4-58-unit-providing-relief-to-businesses

Sameer Al-Dahidi, Manoharan Madhiarasan, “Forecasting Solar Photovoltaic Power Production: A Comprehensive Review and Innovative Data-Driven Modeling Framework,” Energies, vol. 17, no. 16, p. 4145, 2024, doi: https://doi.org/10.3390/en17164145.

A. Aronescu, J. Appelbaum, “Design optimization of photovoltaic solar fields-insight and methodology,” Renew. Sustain. Energy Rev., vol. 76, pp. 882–893, 2017, doi: https://doi.org/10.1016/j.rser.2017.03.079.

Biswajit Biswal, Subhasish Deb, “Review on smart grid load forecasting for smart energy management using machine learning and deep learning techniques,” Energy Reports, vol. 12, pp. 3654–3670, 2024, doi: https://doi.org/10.1016/j.egyr.2024.09.056.

S. Hochreiter and J. Schmidhuber, “Long Short-Term Memory,” Neural Comput., vol. 9, no. 8, pp. 1735–1780, Nov. 1997, doi: 10.1162/NECO.1997.9.8.1735.

“The Python Language Reference — Python 3.14.6 documentation.” Accessed: Jun. 26, 2026. [Online]. Available: https://docs.python.org/3/reference/index.html

“The Paris Agreement | UNFCCC.” Accessed: Jun. 26, 2026. [Online]. Available: https://unfccc.int/process-and-meetings/the-paris-agreement

“(PDF) Techno-economic-Analysis of Net-metered solar Photovoltaic Systems for Residential Households in Pakistan.” Accessed: Jun. 26, 2026. [Online]. Available: https://www.researchgate.net/publication/351625235_Techno-economic-Analysis_of_Net-metered_solar_Photovoltaic_Systems_for_Residential_Households_in_Pakistan

Osama Ayadi, Rami Al-Assad, “Techno-economic assessment of a grid connected photovoltaic system for the University of Jordan,” Sustain. Cities Soc., vol. 39, pp. 93–98, 2018, doi: https://doi.org/10.1016/j.scs.2018.02.011.

M. U. Nazir and J. Li, “Forecasting of electricity consumption in Pakistan based on integrating machine learning algorithms and Monte Carlo simulation,” Electr. Eng., vol. 107, no. 6, pp. 7539–7559, Jun. 2025, doi: 10.1007/S00202-024-02923-6.

M. Q. Raza and A. Khosravi, “A review on artificial intelligence based load demand forecasting techniques for smart grid and buildings,” Renew. Sustain. Energy Rev., vol. 50, pp. 1352–1372, Jun. 2015, doi: 10.1016/J.RSER.2015.04.065.

Ezekiel Ezekiel Smart, Lois Oyindamola Olanrewaju, “Artificial Intelligence (AI) in renewable energy forecasting and optimization,” World J. Adv. Eng. Technol. Sci., vol. 15, no. 2, pp. 1100–1112, 2025, doi: 10.30574/wjaets.2025.15.2.0300.

Natei Ermias Benti, Mesfin Diro Chaka, “Forecasting Renewable Energy Generation with Machine Learning and Deep Learning: Current Advances and Future Prospects,” Sustainability, vol. 15, no. 9, p. 7087, 2023, doi: https://doi.org/10.3390/su15097087.

Pratima Kumari, Durga Toshniwal, “Deep learning models for solar irradiance forecasting: A comprehensive review,” J. Clean. Prod., vol. 318, 2021, [Online]. Available: https://www.sciencedirect.com/science/article/pii/S0959652621027736

Holger R. Maier, Stefano Galelli, “Exploding the myths: An introduction to artificial neural networks for prediction and forecasting,” Environ. Model. Softw., vol. 167, p. 105776, 2023, doi: https://doi.org/10.1016/j.envsoft.2023.105776.

V. N. Vapnik, “The Nature of Statistical Learning Theory,” Nat. Stat. Learn. Theory, 2000, doi: 10.1007/978-1-4757-3264-1.

M. Zulfiqar, Kelum A.A. Gamage, “Hyperparameter Optimization of Bayesian Neural Network Using Bayesian Optimization and Intelligent Feature Engineering for Load Forecasting,” Sensors, vol. 22, no. 12, p. 4446, 2022, doi: https://doi.org/10.3390/s22124446.

H. A. Manzoor, M. Ellahi, S. U. Rehman, M. Z. Babar, W. Arif, and A. Saeed, “Forecasting of Domestic and Commercial Power Demands Using Machine Learning for the LESCO Region in Pakistan,” CIGRE Sci. Eng., no. 38, pp. 1–26, 2025, Accessed: Jun. 26, 2026. [Online]. Available: https://researchportal.hw.ac.uk/en/

publications/forecasting-of-domestic-and-commercial-power-demands-using-machin/

V. Gayathry, Deepa Kaliyaperumal & Surender Reddy Salkuti, “Seasonal solar irradiance forecasting using artificial intelligence techniques with uncertainty analysis,” Sci. Rep., 2024, [Online]. Available: https://www.nature.com/articles/s41598-024-68531-3

“Support Vector Regression - an overview | ScienceDirect Topics.” Accessed: Jun. 26, 2026. [Online]. Available: https://www.sciencedirect.com/topics/computer-science/support-vector-regression

Leo Breiman, “Random Forests,” Mach. Learn., vol. 45, 2001.

Sheeza Mariam Nawaz, Ubaid Ahmed, “Medium-term load forecasting with Power Market Survey: GEPCO case study,” Acad. Green Energy, vol. 1, no. 1, 2024, doi: 10.20935/AcadEnergy6257.

Pablo Reina-Jiménez, María Martínez-Ballesteros, “Tree-based vs. deep learning in time series: A prescriptive framework for performance-driven model selection,” Appl. Soft Comput., vol. 195, p. 114983, 2026, doi: https://doi.org/10.1016/j.asoc.2026.114983.

J. S. Sepp Hochreiter, “Long Short-Term Memory,” Neural Comput., vol. 9, no. 8, pp. 1735–1780, 1997, doi: 10.1162/neco.1997.9.8.1735.

Fouzi Harrou, Farid Kadri, “Forecasting of Photovoltaic Solar Power Production Using LSTM Approach,” Adv. Stat. Model. Forecast. Fault Detect. Renew. Energy Syst., 2020, [Online]. Available: https://www.researchgate.net/publication/340265680_Forecasting_of_Photovoltaic

_Solar_Power_Production_Using_LSTM_Approach

S. Siami-Namini, N. Tavakoli, and A. Siami Namin, “A Comparison of ARIMA and LSTM in Forecasting Time Series,” Proc. - 17th IEEE Int. Conf. Mach. Learn. Appl. ICMLA 2018, pp. 1394–1401, Jul. 2018, doi: 10.1109/ICMLA.2018.00227.

Sheeraz Iqbal, Md Shafiullah, “Forecasting large-scale solar power plant energy production based on Monte Carlo simulations and long-short-term memory,” Results Eng., vol. 27, p. 106269, 2025, doi: https://doi.org/10.1016/j.rineng.2025.106269.

Faisal Ahmed, Asif Uzzaman, “AI-Driven Microgrid Solutions for Enhancing Energy Access and Reliability in Rural and Remote Areas: A Comprehensive Review,” Control Syst. Optim. Lett., vol. 3, no. 1, pp. 110–116, 2025, doi: 10.59247/csol.v3i1.183.

Moez Krichen, Alaeddine Mihoub, “Long Short-Term Memory Networks: A Comprehensive Survey,” AI, vol. 6, no. 9, 2025, doi: https://doi.org/10.3390/ai6090215.

Hai Xiang Zhao, Frédéric Magoulès, “A review on the prediction of building energy consumption,” Renew. Sustain. Energy Rev., vol. 16, no. 6, pp. 3586–3592, 2012, doi: 10.1016/j.rser.2012.02.049.

M. Irfan, Z.-Y. Zhao, M. Ahmad, and M. C. Mukeshimana, “Solar Energy Development in Pakistan: Barriers and Policy Recommendations,” Sustain. 2019, Vol. 11, Page 1206, vol. 11, no. 4, p. 1206, Feb. 2019, doi: 10.3390/SU11041206.

Charles R. Harris, K. Jarrod Millman, Stéfan J. van der Walt, “Array programming with NumPy,” Nature, vol. 585, pp. 357–362, 2020, [Online]. Available: https://www.nature.com/articles/s41586-020-2649-2

“Data Structures for Statistical Computing in Python - SciPy Proceedings.” Accessed: Jun. 26, 2026. [Online]. Available: https://proceedings.scipy.org/articles/Majora-92bf1922-00a

PedregosaFabian et al., “Scikit-learn: Machine Learning in Python,” J. Mach. Learn. Res., Nov. 2011, doi: 10.5555/1953048.2078195.

Martín Abadi, Paul Barham, Jianmin Chen, “TensorFlow: A system for large-scale machine learning,” arXiv:1605.08695, 2016, [Online]. Available: https://arxiv.org/abs/1605.08695

“Keras: Deep Learning for humans.” Accessed: Nov. 01, 2025. [Online]. Available: https://keras.io/

J. D. Hunter, “Matplotlib: A 2D graphics environment,” Comput. Sci. Eng., vol. 9, no. 3, pp. 90–95, 2007, doi: 10.1109/MCSE.2007.55.

Waskom, Michael, “seaborn: statistical data visualization,” J. Open Source Softw., vol. 6, 2021, [Online]. Available: https://seaborn.pydata.org/

L. Breiman, “Random Forests,” Mach. Learn. 2001 451, vol. 45, no. 1, pp. 5–32, Oct. 2001, doi: 10.1023/A:1010933404324.