Modeling water flux in forward osmosis and pressure retarded osmosis accounting for reverse solute flux

Accurate water flux prediction is essential for optimizing forward osmosis (FO) and pressure retarded osmosis (PRO) membrane processes. However, many models neglect the role of salt flux for simplicity, leading to uncertainties regarding the suitability for localized predictions within the membrane...

Full description

Bibliographic Details
Main Authors: Tan, Jia Xin, Abdul Latif, Ahmad, Li, Mengheng, Lau, Woei Jye, Mohd Hasbi, Ab Rahim, Fimbres Weihs, G. A., Liang, Y. Y.
Format: Article
Language:English
Published: Elsevier Ltd 2025
Subjects:
Online Access:http://umpir.ump.edu.my/id/eprint/44012/
http://umpir.ump.edu.my/id/eprint/44012/1/Modeling%20water%20flux%20in%20forward%20osmosis%20and%20pressure%20retarded%20osmosis%20accounting.pdf
Description
Summary:Accurate water flux prediction is essential for optimizing forward osmosis (FO) and pressure retarded osmosis (PRO) membrane processes. However, many models neglect the role of salt flux for simplicity, leading to uncertainties regarding the suitability for localized predictions within the membrane systems. This study properly accounts the solute permeability (B) to achieve more accurate water flux predictions in osmotically driven membrane processes. The validity of models with and without B is evaluated against experimental data by using computational fluid dynamics to implement them within a membrane channel, predicting both average and local water flux under various operating and membrane conditions. The applicability of the models is further examined at the longer membrane channel. The results show that both models accurately predict water flux in FO systems in terms of local and area-averaged flux. However, in PRO systems, significant differences are observed between the models, with differences reaching up to 116 % in empty channels and 85 % in spacer-filled channels. We also found that both models accurately predict concentration polarization with discrepancies under 1 % but the model without B underestimates concentrative internal concentration polarization in PRO up to 5 %. Even so, errors for both membrane processes stay below 10 %, proving the models are generally reliable for evaluating transmembrane osmotic pressure in FO and PRO. This highlights the necessity of including B in the model for accurate predictions, especially for large-scale membrane separation system. A simplified linear model which includes the B is proposed for FO and PRO system to predict water flux.