A Machine Learning Models for Predicting Power Output in Gas Turbines: A Case Study from Perdawood Power Station, Erbil, Iraqi Kurdistan

Asaad Saber Hamad Ameen; Ibrahim Ismail Hamarash

doi:10.21271/ZJPAS.37.5.9

Authors

Asaad Saber Hamad Ameen Department of Electrical Engineering, College of Engineering, Salahaddin University-Erbil, Erbil ,Kurdistan Region, Iraq
Ibrahim Ismail Hamarash Department of Electrical Engineering, College of Engineering, Salahaddin University-Erbil, Erbil ,Kurdistan Region, Iraq

DOI:

https://doi.org/10.21271/ZJPAS.37.5.9

Keywords:

Gas Turbine (GT), Machine Learning (ML), Support Vector Regression (SVR), Power Generation Forecasting, Dataset.

Abstract

Gas Turbines (GT) are essential generation units in both centralized and distributed power systems. In many regions worldwide, GTs are the primary source of electrical energy due to their affordability, efficiency, and flexibility. GT can quickly reach peak output, making them ideal as auxiliary units, additionally, they are often integrated with renewable energy sources to improve system adequacy and reliability. Accurate forecasting of GT power output is crucial for achieving efficient management and smart operation of these units.

This study presents the first application of machine learning techniques for electrical power output prediction from a combined cycle gas power plant in the Iraqi Kurdistan region. A novel dataset comprising 7,893 data points collected during base load operations over four years, from 2019 to 2023 at the Perdawood Electrical Power Station in Erbil, Kurdistan Region of Iraq. The dataset includes eight features - Ambient Temperature (AT), Ambient barometric Pressure (AP), Specific Humidity (CH), grid frequency (DF), Gas Fuel Flow (GP), Compressor Pressure Discharge (CPD), Compressor Pressure Ratio (CPR), and Exhaust Temperature (TTXM), and the Electrical Power Output (EP) as the target variable.

The dataset has been preprocessed, and the Support Vector Regression (SVR) machine learning algorithm is applied to models and trains the dataset. The model's accuracy is carefully evaluated using the effective metrics such as Mean Squared Error (MSE), R-squared (R²), Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE). The results demonstrate the effectiveness of the Support Vector Regression (SVR) algorithm for forecasting the power output of gas turbines, that authenticates the practicability and value of applying machine learning to real-world GT operation in the Kurdistan region. This work provides a strong foundation for introducing a new strategy of smart energy management in the power sector in the region.

References

Abad, L. A., Sarabia, S. M., Yuzon, J. M. & Pacis, M. C. A Short-Term Load Forecasting Algorithm Using Support Vector Regression & Artificial Neural Network Method (SVR-ANN). 2020 11th IEEE Control and System Graduate Research Colloquium (ICSGRC), 8-8 Aug. 2020 2020. 138-143.

Alketbi, S., Nassif, A. B., Eddin, M. A., Shahin, I. & Elnagar, A. Predicting the power of a combined cycle power plant using machine learning methods. 2020 International conference on communications, computing, cybersecurity, and informatics (CCCI), 2020. IEEE, 1-5.

Boyce, M. P. 2011. Gas turbine engineering handbook, Elsevier.

Chen, G., Hu, Q., Wang, J., Wang, X. & Zhu, Y. 2023. Machine-Learning-Based Electric Power Forecasting. 15, 11299.

Chiman Haydar, S. & Abbas, M. A. 2022. Comprehensive Study for Breast Cancer Using Deep Learning and Traditional Machine Learning. Zanco Journal of Pure and Applied Sciences, 34, 22-36.

Das, U. K., Tey, K. S., Seyedmahmoudian, M., Idna Idris, M. Y., Mekhilef, S., Horan, B. & Stojcevski, A. 2017. SVR-Based Model to Forecast PV Power Generation under Different Weather Conditions. 10, 876.

Farooqi, B. A., Kazmi, A. A. & Janjua, A. K. Machine learning in power markets. 2019 2nd International Conference on Computing, Mathematics and Engineering Technologies (iCoMET), 2019. IEEE, 1-6.

González-Briones, A., Hernández, G., Corchado, J. M., Omatu, S. & Mohamad, M. S. Machine Learning Models for Electricity Consumption Forecasting: A Review. 2019 2nd International Conference on Computer Applications & Information Security (ICCAIS), 1-3 May 2019 2019. 1-6.

Haider Abdula, H. 2023. Online Student Feedback System Using Machine Learning. Zanco Journal of Pure and Applied Sciences, 35, 78-85.

Hasan, M. S., Tarequzzaman, M., Moznuzzaman, M. & Ahad Juel, M. A. 2025. Prediction of energy consumption in four sectors using support vector regression optimized with genetic algorithm. Heliyon, 11.

Hong, W.-C. 2009. Hybrid evolutionary algorithms in a SVR-based electric load forecasting model. International Journal of Electrical Power & Energy Systems, 31, 409-417.

Hong, W.-C. J. E. 2011. Electric load forecasting by seasonal recurrent SVR (support vector regression) with chaotic artificial bee colony algorithm. 36, 5568-5578.

Ibrahim, T. K. & Rahman, M. J. I. J. O. E. E. 2012. Effect of compression ratio on performance of combined cycle gas turbine. 2, 9-14.

Kabengele, K. T., Olayode, I. O. & Tartibu, L. K. J. a. S. 2023. Analysis of the Performance of a Hybrid Thermal Power Plant Using Adaptive Neuro-Fuzzy Inference System (ANFIS)-Based Approaches. 13, 11874.

Kaledio, P. J. R. & Sustainability. 2024. Machine Learning Applications in Electric Power Systems: Enhancing Efficiency, Reliability, and Sustainability.

L Blinov, V. & Deryabin, G. Estimation of gas turbine technical condition using machine learning methods. IOP Conference Series: Earth and Environmental Science, 2022. IOP Publishing, 012170.

Liu, Z. & Karimi, I. a. J. E. 2020. Gas turbine performance prediction via machine learning. 192, 116627.

Lobo, J. L., Ballesteros, I., Oregi, I. & Del Ser, J. J. a. P. A. 2019. Real-time Electrical Power Prediction in a Combined Cycle Power Plant.

Mardin Abdullah, A., Ghassan Akram, Q. & Abbas Mohamad, A. 2024. Ocular Disease Classification Using Different Kinds of Machine Learning Algorithms. Zanco Journal of Pure and Applied Sciences, 36, 25-34.

Mohapatra, A. & Sanjay, P. 2014. Comparative Analysis of Inlet Air Cooling Techniques Integrated To Cooled Gas Turbine Plant. Journal of the Energy Institute, 88.

Pachauri, N. 2024. An emission predictive system for CO and NOx from gas turbine based on ensemble machine learning approach. Fuel, 366, 131421.

Qi, Z., Feng, Y., Wang, S. & Li, C. 2025. Enhancing hydropower generation Predictions: A comprehensive study of XGBoost and Support Vector Regression models with advanced optimization techniques. Ain Shams Engineering Journal, 16, 103206.

Rawal, K. & Ahmad, A. A Comparative Analysis of Supervised Machine Learning Algorithms for Electricity Demand Forecasting. 2022 Second International Conference on Power, Control and Computing Technologies (ICPC2T), 1-3 March 2022 2022. 1-6.

Santarisi, N. S. & Faouri, S. S. J. E.-E. J. O. E. T. 2021. Prediction of combined cycle power plant electrical output power using machine learning regression algorithms. 6, 114.

Shaik, N. B., Jongkittinarukorn, K. & Bingi, K. 2024. XGBoost based enhanced predictive model for handling missing input parameters: A case study on gas turbine. Case Studies in Chemical and Environmental Engineering, 10, 100775.

Sharifzadeh, M., Sikinioti-Lock, A., Shah, N. J. R. & Reviews, S. E. 2019. Machine-learning methods for integrated renewable power generation: A comparative study of artificial neural networks, support vector regression, and Gaussian Process Regression. 108, 513-538.

Solyali, D. 2020. A Comparative Analysis of Machine Learning Approaches for Short-/Long-Term Electricity Load Forecasting in Cyprus. Sustainability [Online], 12.

Wankhede, A. & Ghate, V. Prediction of power of a combine gas and steam turbine using artificial neural network. 2018 3rd IEEE International Conference on Recent Trends in Electronics, Information & Communication Technology (RTEICT), 2018. IEEE, 1103-1108.

Yi, Q., Xiong, H. & Wang, D. J. E. 2023. Predicting Power Generation from a Combined Cycle Power Plant Using Transformer Encoders with DNN. 12, 2431.