Concentration polarization model of spiral-wound membrane modules with application to batch-mode RO desalination of brackish water

Article history: Received 8 July 2014 Received in revised form 23 December 2014 Accepted 24 December 2014 Available online 22 January 2015

For inland applications, batch-RO desalination has shown advantages in BW desalination over conventional continuous BWRO because of the high recovery ratio achievable without high energy input. Moreover, batch-RO allows greater control over CP through adjustment of the re-circulation flow rate. Thorough examinations of CP in both batch and continuous mode RO processes were carried out. It was demonstrated that the film/KS model gave an adequate description. To quantitatively assess the impact of flow rate on the CP, we derived a Sherwood correlation as a means of predicting the degree of CP and the permeate flux. The herein proposed Sherwood correlation (Eqs. (18), (19) and (20)) was fully validated by the experimental observations, indicating possible applications to SWM ROs with different geometrical properties (e.g., different feed channel heights and lengths). It was found that the newly established Sherwood correlation for batch-RO gave accurate CPF within less than 10% deviation; for continuous RO, the correlation predicted CP more accurately than other popular correlations. The optimum re-circulation rate for DesaLink operation was investigated based on the specific mass transport characterization for batch-RO operation. It may be argued that the optimum feed flow rate determined here may lead to a lower performance for other batch-RO at different scales and using different components, thus limiting its wider applicability. It was, nevertheless, satisfactory in the scope of this work. Furthermore, with the established Sherwood correlation, we were able to evaluate the CPFs under the different feed flow rates and to quantify the total volume and concentration of the permeate. Acknowledgments SECpump 1.5 SEC (kWh/m ) CPF · SECideal, decreases with the feed flow rate u exponentially. The second term, SECpump, increases with u linearly. Therefore, there exists an optimum point at which the minimum SEC is achieved. Fig. 17 thus illustrates the changes of the CPF · SECideal, SECpump and their sum respectively, and a value of 2 l min− 1 for feed flow rate can be identified graphically to yield the optimum SEC. When the optimal re-circulation flow (i.e., 2 l min−1) is applied, an average value of 1.25 was obtained for CPF; this value suggests 25% energy loss due to CP. The fact that there exists an optimal re-circulation flow for our batch-RO system means that the lowest SEC is achieved when the ratio between re-circulation flow and produced permeate flow is 3. A larger re-circulation flow would result in energy inefficiency due to increased energy consumption by the re-circulation pump, while a smaller flow would lead to an increased CP layer which also decreases the overall system efficiency. Likewise, this optimal ratio between re-circulation flow and permeate flow can be applied in operation for similar systems with different sizes.