- Email: [email protected]
Toward a combined system of forward osmosis and reverse osmosis for seawater desalination
Desalination 247 (2009) 239–246
Toward a combined system of forward osmosis and reverse osmosis for seawater desalination Yong-Jun Choia,d, June-Seok Choia, Hyun-Je Oha, Sangho Leea,* Dae Ryook Yangb, Joon Ha Kimc a
Korea Institute of Construction Technology, Gyeonggi-Do, 411-712, Korea Tel. +82-31-910-0320; Fax: +82-31-910-0291; email: [email protected] b Department of Chemical & Biological Engineering, Korea University, Seoul, 136-701, Korea c Department of Environmental Science and Engineering, Gwangju Institute of Science and Technology (GIST), Gwangju, 500-712, Korea d University of Science and Technology, 113 Gwahangno, Yuseong-gu, Daejeon, Korea Received 13 November 2008; revised 17 December 2008; accepted 24 December 2008
Abstract Forward osmosis (FO) is an osmotic process that uses a semi-permeable membrane to effect separation of water from dissolved solutes by an osmotic pressure gradient. Unlike reverse osmosis (RO), FO does not require high pressure for separation, allowing low energy consumption to produce water. However, the internal concentration polarization in FO is an important factor affecting the performance of FO processes. This paper was intended to investigate the characteristics of FO and RO processes. A simple film theory model was applied to consider concentration polarization in FO and RO processes. This model allows the estimation of internal and external concentration polarization effects in FO process. A laboratory-scale FO device was used to find the model parameters for further calculations. The calculated flux was compared with experimental flux under a variety of operating conditions. It was found that the combination of FO and RO may result in a higher flux than FO-only process under some operating conditions. Further research will be required to investigate the effect of membrane materials on energy efficiency of FO and RO hybrid system. Keywords: Desalination; Forward osmosis; Reverse osmosis; Concentration polarization; Combined system
*Corresponding author. Presented at the 2nd joint workshop between the Center for Seawater Desalination Plant and the European Desalination Society, Gwangju Institute of Science and Technology, Korea, October 8–9, 2008. 0011-9164/09/$– See front matter © 2009 Published by Elsevier B.V. doi:10.1016/j.desal.2008.12.028
240
Y.-J. Choi et al. / Desalination 247 (2009) 239–246
1. Introduction Most membrane processes, such as reverse osmosis (RO), are pressure-driven systems where permeate water flux and recovery are controlled by the hydraulic pressure applied to the feed water. Unlike these pressure-driven systems, forward osmosis (FO) processes operate on the principle of osmotic transport of water across a semi-permeable membrane from a dilute feed solution into a concentrated draw solution [1,2]. Since FO does not require high pressure for separation, it has potential to allow lower energy consumption to produce water than RO systems. Therefore, FO has recently drawn attention as a novel method for wastewater treatment, food processing, and sea water and brackish water desalination [3,4]. However, a major limiting factor of FO system performance is a permeate flux decline due to concentration polarization [5]. Two types of concentration polarization in FO on both sides of the membrane play a prominent role in reducing the effective transmembrane osmotic pressure across asymmetric membranes: (1) The external concentration polarization occurs on the side of active layer and (2) the internal concentration polarization occurs on the side of porous support layer. The external concentration polarization may be reduced if turbulence is induced near the membrane surface, facilitating the diffusion of the concentrated solute back into the bulk solution. However, the internal concentration polarization cannot be mitigated by increased shear stress or turbulence because of the stagnant environment inside the porous support layer [6]. The internal concentration polarization is especially dominant in typical thin film composite membranes. These membranes comprise a polymer porous support layer cast upon a thick fabric backing layer, which provides
mechanical strength. Recent development of new FO membrane allows reducing the internal concentration polarization by making the support layer thinner with embedded mesh structures [1,7]. The reduced internal concentration polarization in the new FO membrane results in a greater utilization of the osmotic driving force and a higher water flux. Unfortunately, this type of FO membrane is not common and expensive compared to typical RO membranes. In this study, we focused on the combined use of FO and RO processes in seawater or brackish water desalination. A simple film theory model was applied to consider the effect of internal and external concentration polarizations on FO and RO processes. Preliminary experiments were carried out using asymmetric RO membrane in FO system. The final goal of this research is to develop an optimum hybrid system of FO and RO for Seawater Desalination. 2. Theoretical analysis We have applied the solution-diffusion model modified with the film theory model to analyze the performance of FO and RO systems. For an RO system, in the absence of salt passage, the generalized flux equation is: ⎛ ⎛ J ⎞⎞ Jw = Lv ⎜ ΔP − π F ,b exp ⎜ w ⎟ ⎟ ⎝ kF ⎠ ⎠ ⎝
(1)
where Jw is the permeate flux, Lv is the water transport parameter, ΔP is the transmembrane pressure, πF,b is the osmotic pressure on the feed side, and kF is the mass transfer coefficient for external concentration polarization. The standard flux Equation for FO is given as [8]: ⎛ ⎛ J ⎞⎞ ⎛ J ⎞ Jw = Lv ⎜ π D ,b exp ⎜ − w ⎟ − π F ,b exp ⎜ w ⎟ ⎟ ⎝ kF ⎠ ⎠ ⎝ kD ⎠ ⎝
(2)
Y.-J. Choi et al. / Desalination 247 (2009) 239–246
where πD,b is the osmotic pressure on the draw solution side and kD is the mass transfer coefficient for internal concentration polarization. Based on the mass transfer correlations, kF and kD are given as [9]: kF = 1.85
D ( Re Sc)0.33 d h L0.33 0.67
(3)
for plate-and-frame module under laminar flow kD =
Dε τl
(4)
where D is the diffusion coefficient, dh is the hydraulic diameter, L is the channel length, Re is the Reynolds number, Sc is the Schmidt number, ε is the porosity of support layer, l is the thickness of support layer, and τ is the tortuosity of support layer. Finally, the flux of combined system using FO and RO is: ⎛ ⎛ J ⎞ Jw = Lv ⎜ ΔP + π D ,b exp ⎜ − w ⎟ ⎝ kD ⎠ ⎝
(5)
⎛ J ⎞⎞ − π F ,b exp ⎜ w ⎟ ⎟ ⎝ k ⎠⎠ F
3. Experiments The test system shown in Figure 1(a) was used to measure characteristics of FO systems. A plate-and-frame membrane module, which was especially designed to have channels on both sides of the membrane, was used for FO tests. As shown in Figure 1(b), the draw solution is flowing on the permeate side and the feed solution on the feed (active layer) side. Co-current flow is used to reduce strain on the suspended membrane. The channel has dimensions of 8 cm length, 2.5 cm width, and 0.3 cm height, providing an effective membrane area
241
of 20.0 cm2. Mesh spacers are inserted within both channels to promote turbulence and mass transport. An electronic balance connected to a personal computer was used to measure the water flux, which was determined by measuring the weight change of the draw solution over a selected time period at the initial stage of the process. The system can be operated as RO operation mode as well as FO operating mode by adjusting the back pressure valve. In addition to typical FO tests, experiments were carried out under a moderate feed pressure (2.2 bar) to increase permeate flux by combining FO with RO. A commercially available membrane (SW30, Filmtec, USA) was used for both FO and RO tests. The volumes of feed tank and draw solution tank were 10 L. The draw solution is flowing on the permeate side and the feed solution on the feed (active layer) side. NaCl solutions with different concentrations were used for feed and draw solutions. Two variable speed gear pumps were used to circulate feed and draw solutions. The flow rates of feed and draw solutions were monitored using the flow meters in the front panel. The temperature of draw solution was maintained to be constant by a temperature control unit. An electronic balance connected to a personal computer measures the mass of water permeating into the draw solution, from which permeate water flux is calculated. The osmotic pressure of NaCl solution was calculated using the empirical relation in Figure 2, which is based on the calculation using the OLI software (OLI systems, Inc., Morris Plains, NJ). 4. Results and discussion 4.1. Water flux in FO mode To begin, the permeate water flux was evaluated for the RO membrane using the DI
Y.-J. Choi et al. / Desalination 247 (2009) 239–246
242 (a)
(b)
Pressure gauge Flow meter
Pressure gauge Flow meter
Membrane Module
Draw tank
Feed tank
Heat exchanger
Electronic balance
Pump
Pump
Fig. 1. Schematic diagram and photograph of experimental setup for FO (a) Overview of the system (b) Schematic diagram.
water as feed water and NaCl solution as draw solution. The feed solution was directed to the active layer of the membrane and the draw solution was directed against the support layer. The operation was done in crossflow mode. Figure 3 shows the permeate flux as
a function of draw solution concentrations. The permeate flux increases with increasing draw solution concentration because of an increased driving force (osmotic pressure difference between feed and draw solution sides). The permeate flux ranges from 2 to
243
Y.-J. Choi et al. / Desalination 247 (2009) 239–246 3.6
Experimental data (L/m2-hr)
Osmotic pressure (bar)
120 100 80 60 40 20
3.4 3.2 3.0 2.8 2.6 2.4 2.2 2.0
0 0.0
0.5
1.0
1.5
2.0
2.0
2.5
2.2
2.4
2.6
2.8
3.0
3.2
3.4
3.6
2
NaCl concentration (M)
Model fit (L/m -hr)
Fig. 2. Correlation between NaCl concentration and osmotic pressure. (a) 4
Fig. 4. Comparison of model fit with experimental data (in Figure 3). (a) 4 Draw (2 M) - Feed (0.5 M) - 0.1 m/s Draw (2 M) - Feed (0.5 M) - 0.3 m/s 3 2
Flux (L/m -hr)
Flux (L/m 2-hr)
3
2
Draw (1 M) - Feed (0 M) Draw (2 M) - Feed (0 M) Draw (3 M) - Feed (0 M) Draw (4 M) - Feed (0 M)
1
1
0 0
5
10
15
20
2
0
25
0
20
40
Time (hour)
60
80
100
120
Time (min)
(b) 4
(b) 4 Draw (3 M) - Feed (0.5 M) - 0.1 m/s Draw (3 M) - Feed (0.5 M) - 0.2 m/s Draw (3 M) - Feed (0.5 M) - 0.3 m/s
Flux (L/m -hr)
3
2
Flux (L/m -hr)
3
2
2
1
Draw (1 M) - Feed (0 M) Draw (2 M) - Feed (0 M) Draw (3 M) - Feed (0 M)
2
1
0 0
5
10
15
20
25
Time (hour)
Fig. 3. Permeate flux driven by osmotic pressure difference in FO operation (Feed solution: DI water). (a) Crossflow velocity = 0.2 m/s (b) Crossflow velocity = 0.3 m/s.
0 0
20
40
60
80
100
120
Time (min)
Fig. 5. Effect of crossflow velocity on permeate flux in FO operation (Feed solution: 0.5 M). (a) Draw solution: 2 M (b) Draw solution: 3 M.
244
Y.-J. Choi et al. / Desalination 247 (2009) 239–246
3 L/m2-hr, which is similar to the reported values for FO tests using asymmetric RO membranes [10]. Using the experimental results in Figure 3 and Eq. (2), the mass transfer coefficient for internal concentration polarization, kD, is estimated to be 2 × 107 m/sec. The typical kD value for FO membrane is reported as 5 × 107 m/sec [8], which is 20 times larger than the kD value in this study. This is because
of thicker support layer in the RO membrane than that in FO membrane. Figure 4 compares the experimental data and model fit using the calculated kD. The model matches the experimental data well. 4.2. FO operation of salt water The FO system was operated using 0.5 M NaCl feed solution, which is representative of
(a) 4 Draw (1 M) - Feed (0.5 M) - 2.2 bar Draw (2 M) - Feed (0.5 M) - 2.3 bar Draw (3 M) - Feed (0.5 M) - 2.4 bar
2
Flux (L/m -hr)
3
2
1
0 0
20
40
60
80
100
120
Time (min)
(b)
3M-0.18m/s-2.4bar 3M-0.3 m/s 3M-0.1 m/s 2M-0.18m/s-2.3bar 2M-0.3 m/s 2M-0.1 m/s
0.0
0.5
1.0
1.5
2.0
2.5
2
Permeate flux (L/m -hr)
Fig. 6. Effect of applied pressure on permeate flux in combined FO+RO operation (Feed solution: 0.5 M). (a) Permeate flux at different draw solution concentrations (b) Comparison of permeate flux in FO with that in FO+RO.
245
Y.-J. Choi et al. / Desalination 247 (2009) 239–246
4.3. Combination of FO and RO Since the permeate flux in FO system using RO membrane is relatively low, an external pressure was applied to increase the flux through the membrane. This was done by increasing the feed-side pressure using the back-pressure value. The maximum pressure applied during this test was 2.4 bar. Figure 6(a) shows the permeate flux in FO+RO mode for different draw solution concentrations. Compared with the results in Figure 5, the permeate flux is much higher in FO+RO mode. To further investigate the effect of applied pressure on permeate flux, the average permeate fluxes in FO and FO+RO modes were compared in Figure 6(b). At 2 M of draw solution, the flux improvement by applying 2.2 bar of external pressure was substantial. The effect of applied pressure is less important at 3 M of draw solution. This suggests that the combination of FO with RO may provide a better performance than FO-only or RO-only systems under a certain conditions. Further work should be done to quantify the combined effect of FO and RO. Figure 7 shows the model predictions for the experimental data in Figs. 5 and 6. kD, is assumed to be constant at 2 × 107 m/sec. It is likely that the model fails to match the experimental data under some conditions. The deviations of model
3.0
Experimental data (L/m2-hr)
seawater. Figure 5 shows the permeate flux as a function of time under various draw solution concentrations and crossflow velocities. Compared with Figure 3, the permeate flux is significantly reduced due to a decrease in osmotic pressure difference. The permeate flux can be increased by an increase in draw solution concentration or crossflow velocity. Nevertheless, the permeate flux was less than 1.5 L/m2-hr even at high draw solution concentration (3 M) and high crossflow velocity (0.3 m/sec).
2.5 2.0 1.5 1.0 0.5 0.0 0.0
0.5
1.0
1.5
2.0
2.5
3.0
Model fit (L/m2-hr)
Fig. 7. Comparison of model fit with experimental data in Figs. 5 and 6.
prediction from experimental data are relatively large at low crossflow velocity. 5. Conclusions In this work, the characteristics of FO and RO processes were investigated using a simple film theory model and laboratory scale tests. The following conclusions can be drawn from this work: (1) The simple film theory mode can match the FO experimental data well. The permeate flux of RO membrane in FO mode was less than 3 L/m2-hr under 3 M of draw solutions due to internal concentration polarization. (2) As the salt concentration of feed water increases, the permeate flux decreases even at high draw solution concentrations. Increasing crossflow velocity may result in a higher flux by decreasing concentration polarization. (3) Combination of FO and RO allows an increase in permeate flux under some operating conditions. Further work will be required to optimize the design and operation conditions of FO+RO systems.
246
Y.-J. Choi et al. / Desalination 247 (2009) 239–246
Acknowledgements This research was supported by a grant (code# C106A152000106A085700220) from Plant Technology Advancement Program funded by Ministry of Construction & Transportation of Korean government and a research grant (no. 2008-0206-49-1) from Korea Institute of Construction Technology.
[5]
[6]
[7]
References [1] T.Y. Cath, A.E. Childress and M. Elimelech, Forward osmosis: Principles, applications, and recent developments, J. Membr. Sci., 281 (2006) 70–87. [2] R.W. Holloway, T.Y. Cath, K.E. Dennett and A.E. Childress. Forward osmosis for concentration of anaerobic digester centrate. in Proceedings of the AWWA Membrane Technology Conference and Exposition, (2005) Phoenix, Arizona, USA. [3] E. board, Yale constructs forward osmosis desalination pilot plant, Membrane Technology, January (2007) 7–8. [4] R.L. McGinnis and M. Elimelech, Energy requirements of ammonia–carbon dioxide forward
[8]
[9]
[10]
osmosis desalination, Desalination, 207 (2007) 370–382. J.L. McCutcheon, R.L. McGinnis and M. Elimelech, The ammonia-carbon dioxide forward osmosis desalination process, Water Conditioning and Purification, October (2006) 1–4. J.R. McCutcheona and M. Elimelech, Influence of membrane support layer hydrophobicity on water flux in osmotically driven membrane processes, J. Membr.Sci., 318 (2008) 458-466. K.Y. Wang, T.-S. Chung and J.-J. Qin, Polybenzimidazole (PBI) nanofiltration hollow fiber membranes applied in forward osmosis process, J. Membr. Sci., 300 (2007) 6–12. J.R. McCutcheon and M. Elimelech, Modeling water flux in forward osmosis: Implications for improved membrane design, AIChE Journal, 53 (2007) 1736–1744. J.R. McCutcheon, R.L. McGinnis and M. Elimelech, Desalination by ammonia-carbon dioxide forward osmosis: Influence of draw and feed solution concentrations on process performance, J. Membr. Sci., 278 (2006) 114–123. J.R. McCutcheona, R.L. McGinnis and M. Elimelech, A novel ammonia-carbon dioxide forward (direct) osmosis desalination process, Desalination, 174 (2005) 1–11.
Toward a combined system of forward osmosis and reverse osmosis for seawater desalination Yong-Jun Choia,d, June-Seok Choia, Hyun-Je Oha, Sangho Leea,* Dae Ryook Yangb, Joon Ha Kimc a
Korea Institute of Construction Technology, Gyeonggi-Do, 411-712, Korea Tel. +82-31-910-0320; Fax: +82-31-910-0291; email: [email protected] b Department of Chemical & Biological Engineering, Korea University, Seoul, 136-701, Korea c Department of Environmental Science and Engineering, Gwangju Institute of Science and Technology (GIST), Gwangju, 500-712, Korea d University of Science and Technology, 113 Gwahangno, Yuseong-gu, Daejeon, Korea Received 13 November 2008; revised 17 December 2008; accepted 24 December 2008
Abstract Forward osmosis (FO) is an osmotic process that uses a semi-permeable membrane to effect separation of water from dissolved solutes by an osmotic pressure gradient. Unlike reverse osmosis (RO), FO does not require high pressure for separation, allowing low energy consumption to produce water. However, the internal concentration polarization in FO is an important factor affecting the performance of FO processes. This paper was intended to investigate the characteristics of FO and RO processes. A simple film theory model was applied to consider concentration polarization in FO and RO processes. This model allows the estimation of internal and external concentration polarization effects in FO process. A laboratory-scale FO device was used to find the model parameters for further calculations. The calculated flux was compared with experimental flux under a variety of operating conditions. It was found that the combination of FO and RO may result in a higher flux than FO-only process under some operating conditions. Further research will be required to investigate the effect of membrane materials on energy efficiency of FO and RO hybrid system. Keywords: Desalination; Forward osmosis; Reverse osmosis; Concentration polarization; Combined system
*Corresponding author. Presented at the 2nd joint workshop between the Center for Seawater Desalination Plant and the European Desalination Society, Gwangju Institute of Science and Technology, Korea, October 8–9, 2008. 0011-9164/09/$– See front matter © 2009 Published by Elsevier B.V. doi:10.1016/j.desal.2008.12.028
240
Y.-J. Choi et al. / Desalination 247 (2009) 239–246
1. Introduction Most membrane processes, such as reverse osmosis (RO), are pressure-driven systems where permeate water flux and recovery are controlled by the hydraulic pressure applied to the feed water. Unlike these pressure-driven systems, forward osmosis (FO) processes operate on the principle of osmotic transport of water across a semi-permeable membrane from a dilute feed solution into a concentrated draw solution [1,2]. Since FO does not require high pressure for separation, it has potential to allow lower energy consumption to produce water than RO systems. Therefore, FO has recently drawn attention as a novel method for wastewater treatment, food processing, and sea water and brackish water desalination [3,4]. However, a major limiting factor of FO system performance is a permeate flux decline due to concentration polarization [5]. Two types of concentration polarization in FO on both sides of the membrane play a prominent role in reducing the effective transmembrane osmotic pressure across asymmetric membranes: (1) The external concentration polarization occurs on the side of active layer and (2) the internal concentration polarization occurs on the side of porous support layer. The external concentration polarization may be reduced if turbulence is induced near the membrane surface, facilitating the diffusion of the concentrated solute back into the bulk solution. However, the internal concentration polarization cannot be mitigated by increased shear stress or turbulence because of the stagnant environment inside the porous support layer [6]. The internal concentration polarization is especially dominant in typical thin film composite membranes. These membranes comprise a polymer porous support layer cast upon a thick fabric backing layer, which provides
mechanical strength. Recent development of new FO membrane allows reducing the internal concentration polarization by making the support layer thinner with embedded mesh structures [1,7]. The reduced internal concentration polarization in the new FO membrane results in a greater utilization of the osmotic driving force and a higher water flux. Unfortunately, this type of FO membrane is not common and expensive compared to typical RO membranes. In this study, we focused on the combined use of FO and RO processes in seawater or brackish water desalination. A simple film theory model was applied to consider the effect of internal and external concentration polarizations on FO and RO processes. Preliminary experiments were carried out using asymmetric RO membrane in FO system. The final goal of this research is to develop an optimum hybrid system of FO and RO for Seawater Desalination. 2. Theoretical analysis We have applied the solution-diffusion model modified with the film theory model to analyze the performance of FO and RO systems. For an RO system, in the absence of salt passage, the generalized flux equation is: ⎛ ⎛ J ⎞⎞ Jw = Lv ⎜ ΔP − π F ,b exp ⎜ w ⎟ ⎟ ⎝ kF ⎠ ⎠ ⎝
(1)
where Jw is the permeate flux, Lv is the water transport parameter, ΔP is the transmembrane pressure, πF,b is the osmotic pressure on the feed side, and kF is the mass transfer coefficient for external concentration polarization. The standard flux Equation for FO is given as [8]: ⎛ ⎛ J ⎞⎞ ⎛ J ⎞ Jw = Lv ⎜ π D ,b exp ⎜ − w ⎟ − π F ,b exp ⎜ w ⎟ ⎟ ⎝ kF ⎠ ⎠ ⎝ kD ⎠ ⎝
(2)
Y.-J. Choi et al. / Desalination 247 (2009) 239–246
where πD,b is the osmotic pressure on the draw solution side and kD is the mass transfer coefficient for internal concentration polarization. Based on the mass transfer correlations, kF and kD are given as [9]: kF = 1.85
D ( Re Sc)0.33 d h L0.33 0.67
(3)
for plate-and-frame module under laminar flow kD =
Dε τl
(4)
where D is the diffusion coefficient, dh is the hydraulic diameter, L is the channel length, Re is the Reynolds number, Sc is the Schmidt number, ε is the porosity of support layer, l is the thickness of support layer, and τ is the tortuosity of support layer. Finally, the flux of combined system using FO and RO is: ⎛ ⎛ J ⎞ Jw = Lv ⎜ ΔP + π D ,b exp ⎜ − w ⎟ ⎝ kD ⎠ ⎝
(5)
⎛ J ⎞⎞ − π F ,b exp ⎜ w ⎟ ⎟ ⎝ k ⎠⎠ F
3. Experiments The test system shown in Figure 1(a) was used to measure characteristics of FO systems. A plate-and-frame membrane module, which was especially designed to have channels on both sides of the membrane, was used for FO tests. As shown in Figure 1(b), the draw solution is flowing on the permeate side and the feed solution on the feed (active layer) side. Co-current flow is used to reduce strain on the suspended membrane. The channel has dimensions of 8 cm length, 2.5 cm width, and 0.3 cm height, providing an effective membrane area
241
of 20.0 cm2. Mesh spacers are inserted within both channels to promote turbulence and mass transport. An electronic balance connected to a personal computer was used to measure the water flux, which was determined by measuring the weight change of the draw solution over a selected time period at the initial stage of the process. The system can be operated as RO operation mode as well as FO operating mode by adjusting the back pressure valve. In addition to typical FO tests, experiments were carried out under a moderate feed pressure (2.2 bar) to increase permeate flux by combining FO with RO. A commercially available membrane (SW30, Filmtec, USA) was used for both FO and RO tests. The volumes of feed tank and draw solution tank were 10 L. The draw solution is flowing on the permeate side and the feed solution on the feed (active layer) side. NaCl solutions with different concentrations were used for feed and draw solutions. Two variable speed gear pumps were used to circulate feed and draw solutions. The flow rates of feed and draw solutions were monitored using the flow meters in the front panel. The temperature of draw solution was maintained to be constant by a temperature control unit. An electronic balance connected to a personal computer measures the mass of water permeating into the draw solution, from which permeate water flux is calculated. The osmotic pressure of NaCl solution was calculated using the empirical relation in Figure 2, which is based on the calculation using the OLI software (OLI systems, Inc., Morris Plains, NJ). 4. Results and discussion 4.1. Water flux in FO mode To begin, the permeate water flux was evaluated for the RO membrane using the DI
Y.-J. Choi et al. / Desalination 247 (2009) 239–246
242 (a)
(b)
Pressure gauge Flow meter
Pressure gauge Flow meter
Membrane Module
Draw tank
Feed tank
Heat exchanger
Electronic balance
Pump
Pump
Fig. 1. Schematic diagram and photograph of experimental setup for FO (a) Overview of the system (b) Schematic diagram.
water as feed water and NaCl solution as draw solution. The feed solution was directed to the active layer of the membrane and the draw solution was directed against the support layer. The operation was done in crossflow mode. Figure 3 shows the permeate flux as
a function of draw solution concentrations. The permeate flux increases with increasing draw solution concentration because of an increased driving force (osmotic pressure difference between feed and draw solution sides). The permeate flux ranges from 2 to
243
Y.-J. Choi et al. / Desalination 247 (2009) 239–246 3.6
Experimental data (L/m2-hr)
Osmotic pressure (bar)
120 100 80 60 40 20
3.4 3.2 3.0 2.8 2.6 2.4 2.2 2.0
0 0.0
0.5
1.0
1.5
2.0
2.0
2.5
2.2
2.4
2.6
2.8
3.0
3.2
3.4
3.6
2
NaCl concentration (M)
Model fit (L/m -hr)
Fig. 2. Correlation between NaCl concentration and osmotic pressure. (a) 4
Fig. 4. Comparison of model fit with experimental data (in Figure 3). (a) 4 Draw (2 M) - Feed (0.5 M) - 0.1 m/s Draw (2 M) - Feed (0.5 M) - 0.3 m/s 3 2
Flux (L/m -hr)
Flux (L/m 2-hr)
3
2
Draw (1 M) - Feed (0 M) Draw (2 M) - Feed (0 M) Draw (3 M) - Feed (0 M) Draw (4 M) - Feed (0 M)
1
1
0 0
5
10
15
20
2
0
25
0
20
40
Time (hour)
60
80
100
120
Time (min)
(b) 4
(b) 4 Draw (3 M) - Feed (0.5 M) - 0.1 m/s Draw (3 M) - Feed (0.5 M) - 0.2 m/s Draw (3 M) - Feed (0.5 M) - 0.3 m/s
Flux (L/m -hr)
3
2
Flux (L/m -hr)
3
2
2
1
Draw (1 M) - Feed (0 M) Draw (2 M) - Feed (0 M) Draw (3 M) - Feed (0 M)
2
1
0 0
5
10
15
20
25
Time (hour)
Fig. 3. Permeate flux driven by osmotic pressure difference in FO operation (Feed solution: DI water). (a) Crossflow velocity = 0.2 m/s (b) Crossflow velocity = 0.3 m/s.
0 0
20
40
60
80
100
120
Time (min)
Fig. 5. Effect of crossflow velocity on permeate flux in FO operation (Feed solution: 0.5 M). (a) Draw solution: 2 M (b) Draw solution: 3 M.
244
Y.-J. Choi et al. / Desalination 247 (2009) 239–246
3 L/m2-hr, which is similar to the reported values for FO tests using asymmetric RO membranes [10]. Using the experimental results in Figure 3 and Eq. (2), the mass transfer coefficient for internal concentration polarization, kD, is estimated to be 2 × 107 m/sec. The typical kD value for FO membrane is reported as 5 × 107 m/sec [8], which is 20 times larger than the kD value in this study. This is because
of thicker support layer in the RO membrane than that in FO membrane. Figure 4 compares the experimental data and model fit using the calculated kD. The model matches the experimental data well. 4.2. FO operation of salt water The FO system was operated using 0.5 M NaCl feed solution, which is representative of
(a) 4 Draw (1 M) - Feed (0.5 M) - 2.2 bar Draw (2 M) - Feed (0.5 M) - 2.3 bar Draw (3 M) - Feed (0.5 M) - 2.4 bar
2
Flux (L/m -hr)
3
2
1
0 0
20
40
60
80
100
120
Time (min)
(b)
3M-0.18m/s-2.4bar 3M-0.3 m/s 3M-0.1 m/s 2M-0.18m/s-2.3bar 2M-0.3 m/s 2M-0.1 m/s
0.0
0.5
1.0
1.5
2.0
2.5
2
Permeate flux (L/m -hr)
Fig. 6. Effect of applied pressure on permeate flux in combined FO+RO operation (Feed solution: 0.5 M). (a) Permeate flux at different draw solution concentrations (b) Comparison of permeate flux in FO with that in FO+RO.
245
Y.-J. Choi et al. / Desalination 247 (2009) 239–246
4.3. Combination of FO and RO Since the permeate flux in FO system using RO membrane is relatively low, an external pressure was applied to increase the flux through the membrane. This was done by increasing the feed-side pressure using the back-pressure value. The maximum pressure applied during this test was 2.4 bar. Figure 6(a) shows the permeate flux in FO+RO mode for different draw solution concentrations. Compared with the results in Figure 5, the permeate flux is much higher in FO+RO mode. To further investigate the effect of applied pressure on permeate flux, the average permeate fluxes in FO and FO+RO modes were compared in Figure 6(b). At 2 M of draw solution, the flux improvement by applying 2.2 bar of external pressure was substantial. The effect of applied pressure is less important at 3 M of draw solution. This suggests that the combination of FO with RO may provide a better performance than FO-only or RO-only systems under a certain conditions. Further work should be done to quantify the combined effect of FO and RO. Figure 7 shows the model predictions for the experimental data in Figs. 5 and 6. kD, is assumed to be constant at 2 × 107 m/sec. It is likely that the model fails to match the experimental data under some conditions. The deviations of model
3.0
Experimental data (L/m2-hr)
seawater. Figure 5 shows the permeate flux as a function of time under various draw solution concentrations and crossflow velocities. Compared with Figure 3, the permeate flux is significantly reduced due to a decrease in osmotic pressure difference. The permeate flux can be increased by an increase in draw solution concentration or crossflow velocity. Nevertheless, the permeate flux was less than 1.5 L/m2-hr even at high draw solution concentration (3 M) and high crossflow velocity (0.3 m/sec).
2.5 2.0 1.5 1.0 0.5 0.0 0.0
0.5
1.0
1.5
2.0
2.5
3.0
Model fit (L/m2-hr)
Fig. 7. Comparison of model fit with experimental data in Figs. 5 and 6.
prediction from experimental data are relatively large at low crossflow velocity. 5. Conclusions In this work, the characteristics of FO and RO processes were investigated using a simple film theory model and laboratory scale tests. The following conclusions can be drawn from this work: (1) The simple film theory mode can match the FO experimental data well. The permeate flux of RO membrane in FO mode was less than 3 L/m2-hr under 3 M of draw solutions due to internal concentration polarization. (2) As the salt concentration of feed water increases, the permeate flux decreases even at high draw solution concentrations. Increasing crossflow velocity may result in a higher flux by decreasing concentration polarization. (3) Combination of FO and RO allows an increase in permeate flux under some operating conditions. Further work will be required to optimize the design and operation conditions of FO+RO systems.
246
Y.-J. Choi et al. / Desalination 247 (2009) 239–246
Acknowledgements This research was supported by a grant (code# C106A152000106A085700220) from Plant Technology Advancement Program funded by Ministry of Construction & Transportation of Korean government and a research grant (no. 2008-0206-49-1) from Korea Institute of Construction Technology.
[5]
[6]
[7]
References [1] T.Y. Cath, A.E. Childress and M. Elimelech, Forward osmosis: Principles, applications, and recent developments, J. Membr. Sci., 281 (2006) 70–87. [2] R.W. Holloway, T.Y. Cath, K.E. Dennett and A.E. Childress. Forward osmosis for concentration of anaerobic digester centrate. in Proceedings of the AWWA Membrane Technology Conference and Exposition, (2005) Phoenix, Arizona, USA. [3] E. board, Yale constructs forward osmosis desalination pilot plant, Membrane Technology, January (2007) 7–8. [4] R.L. McGinnis and M. Elimelech, Energy requirements of ammonia–carbon dioxide forward
[8]
[9]
[10]
osmosis desalination, Desalination, 207 (2007) 370–382. J.L. McCutcheon, R.L. McGinnis and M. Elimelech, The ammonia-carbon dioxide forward osmosis desalination process, Water Conditioning and Purification, October (2006) 1–4. J.R. McCutcheona and M. Elimelech, Influence of membrane support layer hydrophobicity on water flux in osmotically driven membrane processes, J. Membr.Sci., 318 (2008) 458-466. K.Y. Wang, T.-S. Chung and J.-J. Qin, Polybenzimidazole (PBI) nanofiltration hollow fiber membranes applied in forward osmosis process, J. Membr. Sci., 300 (2007) 6–12. J.R. McCutcheon and M. Elimelech, Modeling water flux in forward osmosis: Implications for improved membrane design, AIChE Journal, 53 (2007) 1736–1744. J.R. McCutcheon, R.L. McGinnis and M. Elimelech, Desalination by ammonia-carbon dioxide forward osmosis: Influence of draw and feed solution concentrations on process performance, J. Membr. Sci., 278 (2006) 114–123. J.R. McCutcheona, R.L. McGinnis and M. Elimelech, A novel ammonia-carbon dioxide forward (direct) osmosis desalination process, Desalination, 174 (2005) 1–11.