Desalination 352 (2014) Contents lists available at ScienceDirect. Desalination. journal homepage:

Transcription

1 Desalination 35 (1) Contents lists available at ScienceDirect Desalination journal homepage: A novel analysis of reverse draw and feed solute fluxes in forward osmosis membrane process Bongchul Kim, Sangyoup Lee, Seungkwan Hong School of Civil, Environmental & Architectural Engineering, Korea University, 1-5 Ga, Anam-Dong, Seongbuk-Gu, Seoul , Republic of Korea HIGHLIGHTS GRAPHICAL ABSTRACT A novel method to determine draw and feed solute fluxes in FO was developed. Draw and feed fluxes were not well predicted with the existing solute permeability. Apparent draw (B d ) and feed (B f ) solute permeability were proposed. B d and B f were applied to analyze draw solute loss and permeate water quality. article info abstract Article history: Received 5 April 1 Received in revised form 11 August 1 Accepted 1 August 1 Available online 7 September 1 Keywords: Forward osmosis (FO) membrane Reverse draw solute flux Feed solute flux Solute permeability coefficient (B) Internal concentration polarization (ICP) A novel method to determine reverse draw and forward feed solute fluxes in forward osmosis (FO) membrane was developed to analyze FO performance more accurately. Specifically, apparent draw solute permeability (B d )andfeed solute permeability (B f ) were proposed, instead of relying on single solute permeability (B). Our results clearly demonstrated that both draw and feed fluxes were not well predicted with the solute permeability (B) measured by RO mode experiment, typically employed in FO membrane characterization. In this study, the draw and feed solute permeabilities were evaluated independently by the experimental protocols which simulated actual FO operation more closely. Much better agreement between experimental observations and theoretical predictions was obtained when both B d and B f were applied for the analysis of draw and feed solute fluxes, respectively. Thus, the utilization of apparent draw and feed solute permeabilities provides more precise assessment of draw solute loss and permeate water quality, which are very important for FO membrane process design and operation. 1 Elsevier B.V. All rights reserved. 1. Introduction Corresponding author. Tel.: ; fax: address: skhong1@korea.ac.kr (S. Hong). Membrane processes represent one of the most feasible options for water shortage alleviation and water supply augmentation [1].Such examples include microfiltration (MF) and ultrafiltration (UF) for membrane bioreactors in wastewater treatment [] and for pre-treatments in seawater desalination [3] as well as nanofiltration (NF) and reverse osmosis (RO) in brackish water [] and seawater desalination [5 7]. The forward osmosis (FO) process has also been attracting great attention for its potential applications in seawater desalination [,9], wastewater reclamation [1,11], and industrial wastewater treatment such / 1 Elsevier B.V. All rights reserved.

2 B. Kim et al. / Desalination 35 (1) as shale gas produced wastewater [1 1]. Particularly, in terms of membrane fouling/cleaning, FO process is assumed to be more preferable to RO process [9,1,15 1]. When describing the performance of FO membranes, three parameters are commonly employed: the pure water permeability coefficient (A) and solute permeability coefficient (B), which describe mass transport across the membrane active layer, and the structural parameter (S) which governs the transport phenomena across the membrane support layer. These parameters are typically used to describe the permeate water and solute fluxes of FO process, and thus represent the standard criteria for determining FO membrane characteristics and comparing their performance. The existing approaches for measuring FO performance parameters involve at least two separate independent experiments. Firstly, the parameters related to the active layer of the FO membrane (i.e., A and B) are measured by applying hydraulic pressure in RO mode experiments. Subsequently, the support layer structural parameter (S) is determined by a FO mode experiment which uses osmotic pressure as the driving force [,1,19]. In addition, the measurements made in pressure retarded osmosis (PRO) mode experiments, with the draw solution facing the active side of the membrane, may be used to complement or substitute those made in FO mode experiments [,1]. As presented in a recentstudy[], by changing the concentration of the draw solution in each FO experiment set, water flux and reverse solute flux are determined, and then membrane performance parameters are estimated through non-linear regression, where A, B, and S are treated as adjustable parameters to fit the FO transport equations to the experimental data. All of the aforementioned conventional approaches to determine the performance parameters of FO membranes consider only one way transport behavior. Whereas RO mode experiments, for example, evaluate only feed solute flux, PRO and FO mode experiments evaluate only reverse draw solute flux. FO processes, however, intrinsically have simultaneous two-way transport behaviors across the active and support layers of FO membranes. Therefore, to delineate the solute transport behaviors of FO process more accurately, it may be essential to assess both forward feed and reverse draw solute fluxes separately. The objective of this study is to demonstrate new experimental ways of determining membrane performance parameters and apply them for a better understanding of the FO process in terms of permeate water quality and draw solute loss. For the first time in the literature, the solute permeability coefficient (B), typically measured by single RO and/or FO mode experiments, was separated into apparent draw (B d ) and feed (B f ) solute permeabilities. These parameters were then employed to describe draw and feed solute transport in FO membrane process. The method and analysis presented in this study are expected to provide a useful tool to predict permeate water quality and draw solute loss in various FO applications. membrane for co-current flows of the feed and draw solutions. More specifically, the crossflow FO cell was custom-built with a channel 77 mm long, mm wide and 3 mm deep, creating an effective membrane area of. cm. No feed spacers were used in the feed or draw solution channels of the FO cell. Variable speed gear pumps (Micropump, Vancouver, WA) were used to deliver the liquids in a closed loop. The cross-flows of feed and draw solutions were fixed at 1. cm/s (Reynolds number of this system is 111). We assumed a constant diffusion coefficient of NaCl solution for the range of concentrations of the solution considered in this work, namely M. Over this range, diffusion coefficients of NaCl solutions varied by less than 3% in a previous study []. A constant temperature water bath was used to maintain both feed and draw solution temperatures at 5 ± 1. C. After the initial flux was stabilized, which took about 3 min, the weight of draw solution tank was measured over time to determine the permeate water flux by computer and digital scale (CAS, Korea). The conductivity of feed and draw sides was measured with a calibrated conductivity meter (model 3, YSI Incorporated, Yellow Springs, OH) to determine solute flux..3. Water flux measurements in FO membrane process Typical FO experiment was performed to measure water flux through the FO membrane. In this mode (i.e., active layer facing feed solution, AL-FS mode), water flowed from the active layer to the support layer, while draw solutes diffused from the support layer to the active layer. Concentrated stock solution (3 M NaCl or dextrose) was added to the draw side to establish the desired osmotic driving force, and the resulting permeate water flux was measured. Osmotic pressure was calculated using software from OLI (Morris Plains, NJ)... Determination of FO membrane performance parameters..1. Solute permeability measured by RO system The feed solute permeability of the FO membrane was determined using a laboratory scale cross-flow RO test unit. A diagram depicting this method with a FO membrane is given in Fig. 1(a). Initially, the membrane was equilibrated with DI water at the applied hydraulic pressure (ΔP) of 17.5 bar until the permeate flux reached a steady value. After equilibration, NaCl rejection (R) was determined at the same applied pressure. Using 5 mm NaCl feed solution, the rejection was determined from the difference in bulk feed (C f ) and permeate (C p ) solute concentrations, using the equation, R = 1 C p /C f.the. Materials and methods.1. FO membrane The cartridge type FO membrane used in this study was provided by Hydration Technologies Innovations (Albany, OR) and contained a woven polyester mesh embedded in thin film of cellulose acetate. A detailed description regarding structure and properties of this membrane is available in a previous study [9]... RO and FO systems Typical bench-scale crossflow RO and FO systems were used in this study, and their schematic diagrams can be found elsewhere [ 1]. The membrane cells in both systems had the same geometry, except that the FO cell had two symmetric channels on both sides of the Fig. 1. Determination of solute permeability (B) by (a) RO mode experiment and (b) FO mode experiment. The direction of water and solute fluxes is schematically illustrated with driving force of each system. Experimental procedures are described in Sections..1 and...

3 13 B. Kim et al. / Desalination 35 (1) Draw solute flux (gmh) Permeate water flux (LMH) 5 1 Permeate water flux Draw solute flux 1 Reverse flux selectivity Draw solution concentration (M) Fig.. Variation of permeate water flux (J w ), reverse draw solute flux (J s ), and reverse flux selectivity (J w /J s ) with increasing draw solute concentration (.1 to 1.5 M NaCl) in the FO mode experiments. The cross-flow rate and temperature were kept at 1. cm/s and 5 C for both feed and draw solutions. The mass transfer coefficient of NaCl draw solutes was m/s. solute permeability parameter (B) was determined using Eq. (1) [,9], 1 R B ¼ J w exp J w : ð1þ R k In this equation, J w and k correspond to the permeate flux and mass transfer coefficient of NaCl in the feed solution, respectively.... Solute permeability measured by FO system The FO experiment started with DI water of the same temperature on both sides of the membrane. Then, the NaCl stock solution was added to the draw solution to reach the desired draw solution concentration, specifically in the range of M. When steady state was established, permeate water flux (J w ) was determined by weighing draw solution. The NaCl concentration in the feed was also measured at every 3 min interval, and the reverse solute flux (J s ) was calculated from the change of NaCl concentration in the feed. A diagram depicting this method with a FO membrane is given in Fig. 1(b)...3. Measurement of S parameter The structural parameter of the membrane support layer, S, determines the extent of internal concentration polarization in FO membrane, and it is defined as the product of the thickness (t) and tortuosity (τ), divided by the porosity (ε), of the porous support layer (i.e., S = tτ / ε). Typical experiments employing FO membrane apparatus were conducted to calculate S parameter, following the protocol S parameter (µm) RO FO Operating mode determing solute permeability Reverse flux selectivity (L/g) described in earlier studies [19,]. Specifically, employing M NaCl draw solution and DI water feed solution, the water flux was measured in FO mode. The membrane support structural parameter was then determined using Eq. () C d;b exp J ws C D f;b exp J 9 w >< k >= J s ¼ B 1 þ B exp J w exp J ðþ >: ws >; J w k D where D is the diffusivity of the draw solute, C d,b is the bulk concentration of the draw solution, and C f,b is the bulk concentration of the feed solution (zero for deionized water feed). As described in the previous sections, the solute permeability (B) was determined from RO and/or FO mode experiments, and then used for the determination of S parameter. 3. Results and discussions 3.1. Prediction of solute transport behavior in FO membrane process Solute transport from draw to feed solution Permeate water and solute flux behaviors obtained from typical FO experiment are shown in Fig.. In this AL-FS mode operation, DI water was used as the feed solution and NaCl was added as draw Theoretical draw solute flux (gmh) Theoretical draw solute flux (gmh) 1 (a) 1 1 (b) Experimenatal draw solute flux (gmh) 1 Experimenatal draw solute flux (gmh) Fig. 3. Comparison of S parameter variations estimated by RO and FO mode experiments. The S parameter was calculated based on Eq. () by employing solute permeability (B) determined from both RO and FO mode experiments. The average B coefficients obtained by RO and FO systems were.9 ±.1 gmh and.35 ±. gmh, respectively. Fig.. Comparison of experimental draw solute fluxes (Fig. ) with theoretical draw solute fluxes calculated using each B coefficient obtained by (a) RO system and (b) FO system. The dashed line (slope = 1) represents perfect agreement between theoretical prediction and experimental observation.

4 B. Kim et al. / Desalination 35 (1) (a) Feed (NaCl) (Increased) Water Feed Solute (NaCl) Draw (Dextrose) (Fixed) Draw Solute (Dextrose) Permeate water flux (LMH) 1 Permeate water flux Feed solute flux NaCl feed solution concentration (M) 1 Feed solute flux (gmh) (b) Feed (NaCl) (Fixed) Water Feed Solute (NaCl) Draw (Dextrose) (Increased) Draw Solute (Dextrose) Permeate water flux (LMH) 1 Permeate water flux Feed solute flux Dextrose draw solution concentration (M) 1 Feed solute flux (gmh) Fig. 5. Behavior of permeate water and feed solute fluxes obtained during FO system operations with various draw and feed solutions at 5 C: (a) feed solute (NaCl) concentration was increased at fixed draw solute concentration and (b) draw solute (dextrose) concentration was increased at fixed feed solute concentration. solutes. When NaCl concentration in the draw solution increased, stronger driving forces for both water permeation and reverse solute diffusion caused permeate water and reverse draw solute fluxes to increase. These results are in good accordance with previous studies [,1]. The experimental observation in Fig. also showed that J w /J s remained approximately constant throughout the concentration of draw solution tested, as expected from the theory concerning reverse flux selectivity. It should be noted that the ratio of water and reverse solute fluxes only depends on intrinsic membrane active layer characteristics, water permeability (A) and solute permeability (B) [1]. In addition to A and B parameters, the S parameter of support layer was also evaluated by utilizing the experimental results obtained in Fig.. Specifically, the structural parameter was estimated based on J w J d, s J f, s C d, b Theoretical feed solute flux (gmh) Experimenatal feed solute flux (gmh) Fig.. Comparison of experimental feed solute fluxes (Fig. 5) with theoretical feed solute fluxes calculated using B coefficient obtained by FO system. The dashed line (slope = 1) represents perfect agreement between theoretical prediction and experimental observation. C f, b C f, m z δ t a C d, m Fig. 7. A conceptual illustration of the draw and feed solute concentration profile for a FO membrane, incorporating both internal and external concentration polarizations. C d,b is the bulk draw concentration. C d,m is the draw concentration on the active layer at the support side. C f,b is the bulk feed concentration. C f,m is the feed concentration on the active layer at the feed side. t s x

5 13 B. Kim et al. / Desalination 35 (1) Eq. () by employing solute permeability (B) determined from both RO and FO mode experiments, respectively. Despite S parameters being intrinsic FO membrane characteristics, they were varied significantly with draw solute concentrations depending on the solute permeability used. The results are summarized in Fig. 3. As shown, the S values calculated using B determined from RO system (51 ± μm) exhibited larger deviation than those calculated using B determined from FO system (39 ± 7.9 μm). This analysis clearly demonstrated the importance of solute permeability for the characterization of FO membrane. The reverse draw solute fluxes can be predicted theoretically based on FO membrane characteristics (i.e., B and S). Using B values obtained from each RO and FO mode experiment, theoretical draw solute fluxes were determined and compared with those observed in FO operation. As presented in Fig. (a), a poor correlation between theoretical and experimental draw solute fluxes was obtained for the case of using B coefficient determined from RO system (R =.75). Particularly, the disagreement became larger when higher draw solute fluxes were compared. However, theoretical prediction based on the B value from FO system (Fig. (b)) showed good agreement with experimental draw solute flux (R =.99). On the basis of these results, it may be speculated that FO membrane compaction due to high pressure applied in the RO system may lead to larger variation in S parameters and poor prediction of draw solute flux. Thus, it was suggested that draw solute permeability should be determined from FO mode experiment simulating actual draw solute transport Solute transport from feed to draw solution In order to investigate feed solute transport in FO membrane, two different series of FO experiments were performed and their results are presented in Fig. 5. In these experiments, NaCl and dextrose solutions were used as feed and draw solutions, respectively. Under this condition, water and NaCl solutes move across the FO membrane in the same direction, from feed to draw side. The selection of dextrose for draw solutes was made to prevent draw solutes from interfering and/or hindering feed solute transport, since dextrose molecules hardly diffuse from the support to the active layer due to their large hydrated radius [,9]. As shown in Fig. 5(a), NaCl concentration in the feed water was first varied from.1 to.3 M, while dextrose concentration in the draw solution was fixed at 1.5 M, to study the effect of feed solution concentration on the solute flux. In Fig. 5(b), on the other hand, feed solution was fixed at.1 M of NaCl and the dextrose draw concentration was increased from.5 to 1.5 M to investigate the effect of draw concentration on feed solute flux. When NaCl feed concentration increased as shown in Fig. 5(a), feed solute transport was greatly enhanced, but much less water permeated through FO membrane due to decreased osmotic pressure gradient. It was also interesting to notice that much higher NaCl solute transport from feed to draw was observed, compared to NaCl reverse solute transport from draw to feed as shown in Fig., under similar NaCl concentration gradients across the FO membrane (e.g.,.1,. and.3 M). This can be easily explained by the greater influence of internal concentration polarization (ICP) on draw solute flux. Contrary to draw solute flux, feed solute flux was affected only by external concentration polarization (ECP), resulting in higher NaCl transport from feed to draw side. Lastly, when dextrose concentration was increased in the draw solution (Fig. 5(b)), more water permeates through the FO membrane as expected from increased osmotic pressure. The transport of NaCl feed solutes, however, only increased slightly due to ECP since there was no change in NaCl concentration gradient. It was also confirmed that dextrose solute did not interfere with NaCl feed transport behavior. Fig.. Experimental protocols for measuring apparent reverse draw and feed solute permeabilities by FO mode experiments. The detailed experimental procedures are described in Section 3...

6 B. Kim et al. / Desalination 35 (1) For further verifying our observation and explanation above, we calculated theoretical feed solute fluxes using membrane performance parameters (B and S) measured by FO system. The results are illustrated in Fig., and showed that calculated feed solute fluxes were lower than experimental data in Fig. 5, resulting in poor correlation (R =.). This observation was attributed to the fact that forward feed solute transport from feed to draw side was higher than reverse draw solute transport, indicating that solute permeability measured from draw solute flux experiment might not be appropriate for the prediction of feed solute flux. All of the analyses performed in this study demonstrated the necessity of more complex approaches which could account for both apparent draw and feed solute fluxes in FO membrane process. 3.. Determination of apparent draw and feed solute permeabilities The concept of apparent reverse draw solute permeability (B d )and forward feed solute permeability (B f ) was developed for more accurate prediction of permeate water quality and draw solute loss in the FO membrane process. In the following subsections, theoretical foundation behind this concept is first illustrated, and then experimental protocols to determine draw and feed solute permeabilities are described in detail Theoretical foundation for draw and feed solute permeabilities In the FO process, concentration polarization (CP) occurs on both feed and draw sides of the membrane. Specifically, feed solutes are concentrated on the active layer of FO membrane, while draw solutes are diluted inside the porous support layer [19 1]. A schematic illustration of concentrative external concentration polarization (ECP) and dilutive internal concentration polarization (ICP) is provided in Fig. 7. Incorporating these CP phenomena, the solute flux across the FO membrane is described as Eq. () (refer to Section..3). In this equation, the terms exp(j w /k) and exp( J w S/D) account for concentrative ECP and dilutive ICP, respectively. However, Eq. () often failed to accurately predict both draw and feed solute transport behaviors depending on the experimental method of measuring B coefficient, as discussed in the previous sections. Assuming the draw solute is not present in the feed solution, the reverse draw solute flux (J d,s )acrosstheactivelayerissimplyexpressedby J d;s ¼ B d C d;m : The solute flux across the porous support, on the other hand, is the sum of the diffusive component driven by the solute concentration gradient, and the convective component arising from the permeation of water through the membrane; J d;s ¼ D s dcðxþ dx J wcx ð Þ where D s is the effective diffusion coefficient of draw solutes in the porous support, and can be related to the bulk diffusion coefficient (D), by accounting for the porosity and tortuosity of the support layer (i.e., D s =Dε / τ). At steady-state, the solute fluxes across the active and support layers are equal; D s dcðxþ dx J wcx ð Þ ¼ B d C d;m : Integrating this equation across the support layer thickness, the following equation is derived with proper boundary conditions; C d;m ¼ C d;b exp J ws B d C D J d;m w 1 exp J ws D ð3þ ðþ ð5þ : ðþ The solute concentration at the active-support interface (C d,m ) is, however, not experimentally accessible. To circumvent this, C d,m was eliminated through rearrangement of Eq. () with Eq. (3), andthedraw solute flux was finally derived as Eq. (7); 9 J C d;b exp w S >< D >= J d;s ¼ B d 1 þ B d 1 exp J : ð7þ >: ws >; D J w Similarly, the feed solute flux (J f,s ) can be also derived considering concentrative ECP and assuming the feed solute is not present in the draw solution. The resulting equation is as follows; 9 J C f;b exp w >< k >= J f;s ¼ B f 1 þ B f 1 exp J : ðþ >: w >; J w k On the basis of these theoretical approaches, the solute permeability (B) was separately evaluated to be apparent draw (B d ) and feed (B f ) permeabilities, and their applications were verified and discussed in the later sections Experimental protocol for measuring apparent draw and feed solute permeabilities Based on all of the experimental and theoretical results discussed in the study, new experimental method was proposed to determine apparent draw solute permeability as well as apparent feed solute Draw solute flux (gmh) Feed solute flux (gmh) 1 Experiemantal data Draw solute flux by B f Draw solute flux by B d (a) NaCl draw solution concentration 1 Experiemantal data feed solute flux by B f feed solute flux by B d Dextrose draw solution concentration (b) Fig. 9. Verification of apparent draw solute permeability (B d ) and feed solute permeability (B f ) with experimental observations in FO system: (a) draw solute flux and (b) feed solute flux as a function of the osmotic pressure difference between draw and feed solutions. Predicted solute fluxes were obtained using the methods described in Section 3.: Eq. (7) for draw solute flux and Eq. () for feed solute flux.

7 13 B. Kim et al. / Desalination 35 (1) permeability. Their experimental protocols are schematically summarized in Fig.. For the assessment of apparent draw solute permeability, FO mode experiment was performed systematically. Specifically, the draw solution concentration was increased incrementally from.1 to. M NaCl by adding appropriate volumes of the NaCl stock solution (3 M). After steady state was attained with each draw concentration, the water flux was determined from the rate of change in the weight of draw solution, and the solute flux was evaluated from measuring NaCl concentration in the feed. Based on J w and J s determined from this FO mode experiment, the apparent draw solute permeability (B d )wasfinally determined using Eq. () described in Section... For the determination of apparent feed water permeability, similar FO mode experiment was conducted, but dextrose was used as draw solution instead of NaCl. Specifically, 1. and 1.5 M draw concentrations were made using 3 M of dextrose stock solution. The feed solution concentration was fixed with.1 M NaCl. NaCl rejection was evaluated at each concentration of dextrose draw solution. In order to accurately determine NaCl concentration in the draw solution, NaCl concentration in 1. and 1.5 M dextrose solutions was measured by a conductivity meter, and NaCl rejection was calculated using a pre-determined calibration curve. Lastly, the apparent feed solute permeability (B f )wasdeterminedusingeq.(1) presented in Section Application of apparent draw and feed solute permeabilities In order to verify the applicability of apparent draw and feed solute permeabilities, draw and feed solute fluxes measured experimentally using NaCl and dextrose as draw solutions were compared with theoretical predictions calculated using B d and B f, and their results are presented as a function of draw solute concentration in Fig. 9. As expected, reverse draw solute flux increased with increasing NaCl draw concentration by enhanced NaCl diffusion through the FO membrane, while feed solute flux was almost constant regardless of dextrose draw solution concentration. Fig. 9(a) revealed that the draw solute flux was predicted much better when B d was adopted, implying that B d rather than B f reflected more precisely the influence of ICP on reverse solute flux. The feed solute flux, on the other hand, showed better agreement when B f was used in the prediction as presented in Fig. 9(b). Therefore, it can be concluded that the analysis of draw and feed fluxes should be implemented using separate apparent solute permeabilities (i.e., B d and B f ), respectively. Lastly, the concept of two split B parameters is recapitulated in Fig. 1. The deterioration of permeate water quality by feed solute transport may be underestimated with the solute permeability measured in the RO experiment. The draw solute loss in the FO process, on the other hand, may be overestimated when the solute permeability (B), typically measured by RO mode experiment, is employed. The draw solutes are transported much less through FO membrane due to ICP, and thus the draw solute permeability (B d ) measured in the FO mode experiment needs to be used. This can be explained possibly by FO membrane compaction during RO mode experiment which may result in tighter active layer structure. Thus, it is suggested that apparent draw and feed solute permeabilities should be determined by the experiments simulating real operating conditions including not only operating modes (i.e., FO/RO operations) but also feed and draw solute types, in order to more precisely predict permeate water quality and draw solute loss in the FO membrane process. This study has provided such an example.. Conclusions The feasibility of FO membrane process in the water treatment applications depends on not only water productivity but also permeate Water flux Water flux Feed solute flux Apparent B f > B Draw solute flux Apparent B d < B Prediction of permeate water quality Prediction of draw solute loss Fig. 1. A conceptual illustration of apparent draw solute permeability (B d ) and feed solute permeability (B f ) compared with the solute permeability (B) obtained by typical RO mode experiment. The average values of B f,b,andb d were.37 gmh,.9 gmh and.35 gmh, respectively. Experimental procedure measuring B was described in Section..1. Experimental procedures determining B f and B d were described in Section 3.. and schematically summarized in Fig..

8 B. Kim et al. / Desalination 35 (1) water quality and draw solute loss. Thus it is important to elucidate the mechanisms of draw and feed solute transport in the FO process. In this study, by varying the types and concentrations of draw and feed solutions as well as the operating mode of filtration, FO membrane performance parameters, primarily solute permeability, were systematically investigated for the analysis of draw and feed solute fluxes. This study demonstrated differences in solute permeability depending on the method used to determine draw and feed solute fluxes. Especially, reverse draw solute flux was much less than the estimations made based on the solute permeability measured by RO mode experiment, typically practiced in the current literature. Furthermore, feed solute flux was not well predicted even with the solute permeability determined from draw solute flux measurements in the FO mode. These experimental observations suggested that draw and feed solute fluxes should be modeled by employing solute permeability determined from the experiments simulating actual FO operation. Accordingly, apparent draw permeability (B d ) and feed solute permeability (B f ) were determined based on the newly developed experimental protocol in this study. Experimental draw and feed fluxes were predicted well when both B d and B f were employed, respectively. This result was attributed to the fact that all of the factors affecting the transport of draw and feed solutes such as ICP and ECP were reflected in the determination of apparent draw and feed solute permeabilities. Findings from this study are expected to help more precisely evaluate permeate water quality and draw solute loss in the real-scale FO applications. Acknowledgments This research was supported by a grant from the Fundamental R&D Program for Technology of World Premier Materials funded by the Ministry of Knowledge Economy, South Korea (137715). References [1] M. Elimelech, W.A. Phillip, The future of seawater desalination: energy, technology, and the environment, Science 333 (11) [] G. Di Bella, D. Di Trapani, M. Torregrossa, G. Viviani, Performance of a MBR pilot plant treating high strength wastewater subject to salinity increase: analysis of biomass activity and fouling behaviour, Bioresour. Technol. 17 (13) 1 1. [3] A.R. Guastalli, F.X. Simon, Y. Penru, A. de Kerchove, J. Llorens, S. Baig, Comparison of DMF and UF pre-treatments for particulate material and dissolved organic matter removal in SWRO desalination, Desalination 3 (13) [] S. Hong, M. Elimelech, Chemical and physical aspects of natural organic matter (NOM) fouling of nanofiltration membranes, J. Membr. Sci. 13 (1997) [5] B. Peñate, L. García-Rodríguez, Current trends and future prospects in the design of seawater reverse osmosis desalination technology, Desalination (1) 1. [] C. Park, Y.H. Lee, S. Lee, S. Hong, Effect of cake layer structure on colloidal fouling in reverse osmosis membranes, Desalination () [7] L. Zhao, W.S.W. Ho, Novel reverse osmosis membranes incorporated with a hydrophilic additive for seawater desalination, J. Membr. Sci. 55 (1) 5. [] C. Kim, S. Lee, H.K. Shon, M. Elimelech, S. Hong, Boron transport in forward osmosis: measurements, mechanisms, and comparison with reverse osmosis, J. Membr. Sci. 19 (1). [9] C. Boo, S. Lee, M. Elimelech, Z. Meng, S. Hong, Colloidal fouling in forward osmosis: role of reverse salt diffusion, J. Membr. Sci (1) 77. [1] C. Boo, M. Elimelech, S. Hong, Fouling control in a forward osmosis process integrating seawater desalination and wastewater reclamation, J. Membr. Sci. (13) [11] L. Chen, Y. Gu, C. Cao, J. Zhang, J.-W. Ng, C. Tang, Performance of a submerged anaerobic membrane bioreactor with forward osmosis membrane for low-strength wastewater treatment, Water Res. 5 (1) [1] K.L. Hickenbottom, N.T. Hancock, N.R. Hutchings, E.W. Appleton, E.G. Beaudry, P. Xu, T.Y. Cath, Forward osmosis treatment of drilling mud and fracturing wastewater from oil and gas operations, Desalination 31 (13). [13] R.L. McGinnis, N.T. Hancock, M.S. Nowosielski-Slepowron, G.D. McGurgan, Pilot demonstration of the NH 3 /CO forward osmosis desalination process on high salinity brines, Desalination 31 (13) 7 7. [1] B.D. Coday, P. Xu, E.G. Beaudry, J. Herron, K. Lampi, N.T. Hancock, T.Y. Cath, The sweet spot of forward osmosis: treatment of produced water, drilling wastewater, and other complex and difficult liquid streams, Desalination 333 (1) [15] Y. Kim, M. Elimelech, H.K. Shon, S. Hong, Combined organic and colloidal fouling in forward osmosis: fouling reversibility and the role of applied pressure, J. Membr. Sci. (1) 1. [1] S. Lee, C. Boo, M. Elimelech, S. Hong, Comparison of fouling behavior in forward osmosis (FO) and reverse osmosis (RO), J. Membr. Sci. 35 (1) [17] S. Phuntsho, H.K. Shon, S. Hong, S. Lee, S. Vigneswaran, A novel low energy fertilizer driven forward osmosis desalination for direct fertigation: evaluating the performance of fertilizer draw solutions, J. Membr. Sci. 375 (11) [1] D.L. Shaffer, N.Y. Yip, J. Gilron, M. Elimelech, Seawater desalination for agriculture by integrated forward and reverse osmosis: improved product water quality for potentially less energy, J. Membr. Sci (1) 1. [19] T.Y. Cath, M. Elimelech, J.R. McCutcheon, R.L. McGinnis, A. Achilli, D. Anastasio, A.R. Brady, A.E. Childress, I.V. Farr, N.T. Hancock, J. Lampi, L.D. Nghiem, M. Xie, N.Y. Yip, Standard methodology for evaluating membrane performance in osmotically driven membrane processes, Desalination 31 (13) [] J.R. McCutcheon, M. Elimelech, Influence of concentrative and dilutive internal concentration polarization on flux behavior in forward osmosis, J. Membr. Sci. () [1] W.A. Phillip, J.S. Yong, M. Elimelech, Reverse draw solute permeation in forward osmosis: modeling and experiments, Environ. Sci. Technol. (1) [] A. Tiraferri, N.Y. Yip, A.P. Straub, S. Romero-Vargas Castrillon, M. Elimelech, A method for the simultaneous determination of transport and structural parameters of forward osmosis membranes, J. Membr. Sci. (13)