- All Journals
- Al-Qadisiyah University
- Basrah University
- Kirkuk University
- Mosul University
- The Iraqi Borad for Medical Specialization
- University of Anbar
- University of Technology
- Al Mustafa University
- Al-Zahraa University for Women
- Alnoor University
- Kunooz University
- Misan University
- Ninevah University
- Tikrit University
- University of Al-Hamdaniya
- University of Kerbala
Performance Prediction of Reverse Osmosis in Large Basra Water Desalination Project Using Artificial Neural Networks | ||
| Al-Qadisiyah Journal for Engineering Sciences | ||
| Articles in Press, Accepted Manuscript, Available Online from 14 September 2025 PDF (7.13 M) | ||
| Document Type: Research Paper | ||
| DOI: 10.30772/qjes.2025.163627.1674 | ||
| Authors | ||
| Zaman Imad Nazzal; Ala'a Abdulrazaq Jassim* | ||
| Department of Chemical Engineering, College of Engineering, University of Basra, Al-Basrah, Iraq | ||
| Abstract | ||
| The study evaluates the Large Basrah Water Project's (LBWP) operational performance from Feb 1, 2023, to Dec 31, 2024, using an artificial neural network to predict reverse osmosis processes, analyzing factors influencing permeability and concentration polarization. This trains and tests the artificial neural network (ANN) model using a dataset comprising 700 items and divided it into three groups: 80% for training, 10% for validation, and 10% for testing. The developed neural network model successfully predicts the output variables QP and CP based on the six input variables Feed Pressure, Temperature, QF, CF, Turb, and PH. Using Bayesian regularization backpropagation, the model demonstrated excellent predictive performance for QP, with high correlation (R=0.98268) and low error metrics (RMSE =27.5389). While the prediction for CP was slightly less accurate (R=0.95464 and RMSE=6.9029), the overall model performance remains robust and reliable. This approach provides a valuable predictive tool for understanding and optimizing the underlying system behavior based on the selected input parameters. Furthermore, the ANN model indicates that the related weights for temperature, pressure, feed water flow rates, feed water salinity, turbidity, and pH are 17%, 2.94%,42.94%, 28.23%, 6.72% and 2.17% respectively. These results imply that using the training datasets, the model fairly forecasts the concentration and flow of permeate. | ||
| Keywords | ||
| Reverse osmosis; Artificial neural networks; Back Propagation; Basrawater; Correlation coefficient | ||
|
Statistics Article View: 210 PDF Download: 137 |
||