PREDICTING THE FEASIBILITY OF SOLAR PARABOLIC TROUGH SYSTEMS IN SUB-SAHARAN AFRICA USING MACHINE LEARNING ALGORITHMS

Öz

The continuous increase in global energy demand has intensified reliance on fossil fuels. This dependency leads to significant challenges related to economics, environmental degradation, and energy security. As a result, the shift toward renewable energy sources has become increasingly important. Among these, solar energy stands out as one of the most abundant and sustainable resources. Concentrated Solar Power (CSP) systems offer high thermal efficiency through solar radiation concentration and the capability to store thermal energy. This study presents a thermodynamic performance and feasibility analysis of a conceptual 100 MW parabolic trough collector (PTC) CSP plant for three Sub-Saharan African countries: Nigeria, Ethiopia, and Kenya. These countries were selected based on their high solar potential and diverse climatic conditions. Using local meteorological data, including direct normal irradiance (DNI), ambient temperature, and wind speed, the required number of collectors to generate 100 MW was calculated. The results show that Ethiopia requires 1606 collectors, Kenya 1669, and Nigeria 2430. Ethiopia demonstrates the best performance due to its high elevation and reduced thermal losses. In addition to thermodynamic analysis, the study incorporates machine learning-based regression models using experimental data. The collector outlet temperature was predicted using four environmental variables. Six regression algorithms were tested. The K-Nearest Neighbors (KNN) model provided the highest accuracy with an R² value of 0.8896, a mean absolute error (MAE) of 1.52, and a root mean square error (RMSE) of 2.61. The Decision Tree model followed with an R² of 0.8632 and the lowest MAE of 1.40. Other models such as Ridge, Lasso, ElasticNet, and Multiple Linear Regression (MLR) showed consistent but slightly lower performance, with R² values around 0.83. The results confirm that PTC systems are technically feasible for Sub-Saharan African conditions. The use of machine learning algorithms has also proven effective in predicting thermal performance. These findings suggest that CSP investments, supported by appropriate policies and regional cooperation, have strong potential to contribute to sustainable electricity generation and energy security in the region.

Anahtar Kelimeler

• Renewable Energy
• Parabolic Trough Collector
• Sub-Saharan Africa
• Machine Learning.

GÜNEŞ ENERJİLİ PARABOLİK OLUK SİSTEMLERİNİN SAHRA ALTI AFRİKA’DAKİ UYGULANABİLİRLİĞİNİN MAKİNE ÖĞRENMESİ ALGORİTMALARI İLE TAHMİNLENMESİ

Öz

Küresel enerji talebindeki sürekli artış, fosil yakıtlara bağımlılığı artırmakta ve bu durum ekonomik, çevresel ve arz güvenliği sorunlarını beraberinde getirmektedir. Bu bağlamda, güneş enerjisi gibi yenilenebilir kaynaklara yönelim giderek önem kazanmaktadır. Özellikle yoğunlaştırıcı güneş enerjisi (CSP) sistemleri, güneş ışığını odaklayarak yüksek sıcaklıklara ulaşabilmeleri ve ısıl enerjiyi depolayabilmeleri sayesinde öne çıkmaktadır. Bu çalışma, Sahra Altı Afrika'daki üç ülke (Nijerya, Etiyopya, Kenya) için 100 MW kapasiteli kavramsal bir parabolik oluk kolektör (PTC) santralinin termodinamik performansını ve uygulanabilirliğini analiz etmektedir. Direkt normal ışınım, ortam sıcaklığı ve rüzgâr hızı gibi yerel meteorolojik verilerle yapılan hesaplamalarda, 100 MW güç üretimi için gerekli kolektör sayısı Etiyopya’da 1606, Kenya’da 1669, Nijerya’da ise 2430 olarak bulunmuştur. Bu sonuçlar, yüksek rakım ve düşük ısı kaybı sayesinde Etiyopya’nın en verimli ülke olduğunu göstermektedir. Ayrıca, çalışmaya deneysel verilere dayalı makine öğrenmesi regresyon modelleri entegre edilmiştir. Kolektör çıkış sıcaklığı, dört çevresel değişken kullanılarak tahmin edilmiş ve 6 farklı regresyon modeli test edilmiştir. K-En Yakın Komşu (KNN) algoritması R² = 0.8896, MAE = 1.52 ve RMSE = 2.61 ile en yüksek doğruluk değerlerine ulaşırken, Decision Tree algoritması R² = 0.8632, MAE = 1.40 ile düşük hata oranları sunmuştur. Ridge, Lasso, ElasticNet ve MLR gibi doğrusal modeller ise R² ≈ 0.83 seviyelerinde kalmıştır. Elde edilen sonuçlar, PTC sistemlerinin Sahra Altı Afrika’da teknik olarak uygulanabilir olduğunu ve makine öğrenmesi algoritmalarının ısıl performans tahminlerinde başarılı bir şekilde kullanılabileceğini göstermektedir. Bu bağlamda, CSP yatırımlarının bölgesel kalkınma ve sürdürülebilir enerji arzı açısından önemli fırsatlar sunduğu ortaya konmuştur.

Anahtar Kelimeler

• Yenilenebilir Enerji
• Parabolik Oluk Kolektör
• Sahra Altı Afrika
• Makine Öğrenmesi.

Kaynakça

1. 1. Ambrose, J., “Fossil fuel use reaches global record despite clean energy growth”, https://www.theguardian.com/environment/article/2024/jun/20/fossil-fuel-use-reaches-global-record-despite-clean-energy-growth , July 25, 2025.
2. 2. Tesfaye, M., “The hub goes to the sun insights from to our visit to world’s largest CSP project”, https://energyforgrowth.org/article/the-hub-goes-to-the-sun/ , April 2, 2025.
3. 3. Rapid Transition Alliance, “Doing development differently: How Kenya is rapidly emerging as Africa’s renewable energy superpower”, https://rapidtransition.org/stories/doing-development-differently-how-kenya-is-rapidly-emerging-as-africas-renewable-energy-superpower/ , August 01, 2025.
4. 4. Ogunmodimu, O.O., “CSP technology and its potential contribution to electricity supply in northern Nigeria”, International Journal of Renewable Energy Research, Vol. 3, Issue 3, Pages 529-537, 2013. 5. Aly, A., Bernardos, A., Fernandez-Peruchena, C.M., Jensen, S.S., Pedersen, A.B., “Is concentrated solar power (CSP) a feasible option for sub-Saharan Africa?: Investigating the techno-economic feasibility of CSP in Tanzania”, Renewable Energy, Vol. 135, Pages 1224-1240, 2019.
5. 6. National Renewable Energy Laboratory, “Solar energy basics”, https://www.nrel.gov/research/re-solar , June 21, 2025.
6. 7. Okeke, C.J., Egberibine, P.K., Edet, J.U., Wilson, J., Blanchard, R.E., “Comparative assessment of concentrated solar power and photovoltaic for power generation and green hydrogen potential in West Africa: A case study on Nigeria”, Renewable and Sustainable Energy Reviews, Vol. 215, Pages 1-30, 2025.
7. 8. International Energy Agency, “Africa Energy Outlook 2022”, https://www.iea.org/reports/africa-energy-outlook-2022 , May 7, 2025.
8. 9. South African Department of Energy, “The South African renewable energy IPP procurement programme”, https://www.ipp-projects.co.za/Publi cations/, June 11, 2025.

9. 10. Akello, O.O.P., Saoke, O.C., Kamau, N.J., Ndeda, H.O.J., “Modeling and performance analysis of solar parabolic trough collectors for hybrid process heat application in Kenya’s tea industry using system advisor model”, Sustainable Energy Research, Vol. 10, Issue 7, Pages 1-15, 2023.
10. 11. World Bank Group, & ESMAP, “Global solar atlas: country DNI reports (Nigeria, Ethiopia, Kenya)”, https://globalsolaratlas.info/download/cou ntry, April 28, 2025.
11. 12. Gonzalez Gonzalez, A., Alvarez Cabal, J.V., Vigil Berrocal, M.A., Peón Menéndez, R., Riesgo Fernández, A., “Simulation of a CSP solar steam generator, using machine learning”, Energies, Vol. 14, Issue 12, 3613, 2021.
12. 13. Asif Iqbal Turja, Hasan, M.M., Ehsan, M.M., Khan, Y., “Multi-objective performance optimization & thermodynamic analysis of solar powered supercritical CO₂ power cycles using machine learning methods & genetic algorithm”, Energy and AI, Vol. 15, 100327, 2024.
13. 14. Peer, M.S., Melesse, T.Y., Orrù, P.F., Braggio, M., Petrollese, M., “Next-generation CSP: The synergy of nanofluids and Industry 4.0 for sustainable solar energy management”, Energies, Vol. 18, Issue 8, 2083, 2025.
14. 15. SolarMillennium, “The parabolic trough power plants andasol 1-3. the largest solar power plant in the World-Technology Premiere in Europe”, http://large.stanford.edu/publications/power/references/docs/Andasol1-3engl.pdf, April 12, 2025.
15. 16. Eylence, M., Aksoy, B., Özsoy, K., “Makine öğrenmesi ile 3 boyutlu yazıcı plastik filamentlerinin ergime noktası ve esneklik özelliklerine dayalı çekme dayanımının tahmini”, Yekarum, Cilt 9, Sayı 2, Sayfa 91-107, 2024.
16. 17. Kaya, Z., Aksoy, B., Özsoy, K., “Eklemeli imalat yöntemiyle üretilen altı eksenli robot kol ile görüntü işleme ve yapay zekâ tabanlı ürünlerin tasniflemesi”, Journal of Materials and Mechatronics: A, Cilt 4, Sayı 1, Sayfa 193-210, 2023.
17. 18. Altman, N. S., “An Introduction to Kernel and Nearest-Neighbor Nonparametric Regression”, The American Statistician, Vol. 46, Issue 3, Pages 175-185, 1992.
18. 19. Hoerl, A. E. and Kennard, R. W., “Ridge regression: Biased estimation for nonorthogonal problems”, Technometrics, Vol. 12, Issue 1, Pages 55-67, 1970.
19. 20. Tibshirani, R., “Regression shrinkage and selection via the lasso”, Journal of the Royal Statistical Society: Series B (Methodological), Vol. 58, Issue 1, Pages 267-288, 1996.
20. 21. Zou, H. and Hastie, T., “Regularization and variable selection via the elastic net”, Journal of the Royal Statistical Society: Series B (Statistical Methodology), Vol. 67, Issue 2, Pages 301-320, 2005.
21. 22. Nayak, J., Vakula, K., Dinesh, P., Naik, B. and Pelusi, D., “Intelligent food processing: Journey from artificial neural network to deep learning”, Computer Science Review, Vol. 38, Pages 1-28, 2020.
22. 23. Khan, M. I. H., Sablani, S. S., Nayak, R. and Gu, Y., “Machine learning‐based modeling in food processing applications: State of the art”, Comprehensive Reviews in Food Science and Food Safety, Vol. 21, Issue 2, Pages 1409-1438, 2022.
23. 24. Chicco, D., Warrens, M. J. and Jurman, G., “The coefficient of determination R-squared is more informative than SMAPE, MAE, MAPE, MSE and RMSE in regression analysis evaluation”, PeerJ Computer Science, Vol. 7, Pages 1-24, 2021.
24. 25. John, A., and Oyekale, J. “Techno-economic feasibility analysis of a large-scale parabolic trough thermal power plant in Effurun-Warri, Nigeria”, International Journal of Energy and Smart Grid, Vol. 8, Issue 1, Pages 1-11, 2023.
25. 26. T. E. Boukelia, M. S. Mecibah and I. E. Meliche, "Design, optimization and feasibility study of parabolic trough solar thermal power plant with integrated thermal energy storage and backup system for Algerian conditions", 3rd International Symposium on Environmental Friendly Energies and Applications (EFEA), Pages 1-5, Paris, 2014.
26. 27. Aguilar‑Jiménez, J. A., Velázquez, N., Acuña, A., Cota, R., González, E., González, L., López, R., and Islas, S., “Techno‑economic analysis of a hybrid PV‑CSP system with thermal energy storage applied to isolated microgrids”, Solar Energy, Vol. 174, Pages 55-65, 2018.
27. 28. Ding, Y., Vijay, A., Neal, D., & McCulloch, M. D.,”Predictive control of rural microgrids with temperature-dependent battery degradation cost.”, 2020 IEEE PES Innovative Smart Grid Technologies Europe (ISGT-Europe), Pages 509-513, Netherlands, 2020.
28. 29. Yahya, A., Yessef, M., Bennouna, E., Lagrioui, A., and Boutammachte, N., “Parabolic Trough Plant Performance in Morocco for Techno-economic Evaluation: A Case Study of Noor 1”, 2024 16th International Conference on Electronics, Computers and Artificial Intelligence (ECAI), Pages 1-4, Romania, 2024.
29. 30. Suwa, T., “Transient heat transfer performance prediction using a machine learning approach for sensible heat storage in parabolic trough solar thermal power generation cycles”, Journal of Energy Storage, Vol. 56, Pages 1-11, 2022.
30. 31. Kottala, R., Balasubramanian, K. R., Jinshah, B. S., Divakar, S., and Kumar, R., “Experimental investigation and machine learning modelling of phase change material-based receiver tube for natural circulated solar parabolic trough system under various weather conditions”, J Therm Anal Calorim, Vol. 148, Pages 7101-7124, 2023.
31. 32. Alhamayani, A., “Performance analysis and machine learning algorithms of parabolic trough solar collectors using Al2O3-MWCNT as a hybrid nanofluid”, Case Studies in Thermal Engineering, Vol. 57, Pages 1-16, 2024.