The most commonly explored salinity-gradient scheme in the pressure retarded osmosis (PRO) process is based on the pairing of seawater and river water as a salinity gradient resource. However, due to the lack of a high-performance PRO membrane, a sufficient osmotic pressure differential across the membrane active layer to produce high power density cannot be attained with this seawater–river water pair. While high-performance PRO membranes have to be developed, there are other approaches to increase power density that combine several salinity gradient resources more efficiently. This study analyzes scenarios of osmotic power generation by PRO based on a variety of salinity gradient resources. Brine from RO (or future FO) desalination plants and municipal wastewater effluent (or brackish water) may be used as a high-salinity draw solution and low-salinity feed solution, respectively. The use of high salinity brines (as draw solution) from future FO desalination plants paired with seawater (as feed solution) may be an especially viable approach for a hybrid process of FO desalination and PRO power generation. In this approach, seawater is the only input resource, and there is no need to have separate intake and pre-treatment for the feed and draw solutions as in the conventional seawater–river water PRO scheme. From the analysis of these scenarios, we performed experiments with seawater (0.5 M NaCl) as feed solution to investigate the feasibility of the PRO process with the various proposed schemes. For the 2 M draw/0.5 M feed solution scheme at a solution temperature of 30 °C, a water flux of 13.9 L m−2 h−1 and corresponding projected power density of 4.7 W/m2 were obtained at a hydraulic pressure difference of 12.5 bar using a commercial cellulose triacetate FO membrane. This study demonstrates that osmotic power generation by pressure retarded osmosis using seawater as a feed solution is potentially viable through the introduction of a hybrid process of FO desalination and PRO osmotic power generation.
► Innovative salinity-gradient schemes are analyzed for power generation by PRO. ► Feasibility of using seawater as feed solution for PRO is examined. ► High power densities are feasible by pairing seawater feed with desalination plant brines. ► Internal concentration polarization is a major hindrance for high power densities with seawater feed.
Recently, osmotic power has attracted a lot of attention as a potential renewable energy source [1], [2], [3], [4], [5] and [6]. Osmotic power is analogous to hydropower in that both use hydroturbines to generate electricity. But there are two major differences in terms of the water resource type and energy conversion means [7] and [8] each system uses. A hydroelectric power plant exploits river water and a dam, while the energy produced from mixing river water and seawater is harnessed in an osmotic power plant, using semipermeable membranes.
Pressure retarded osmosis (PRO) is a membrane-based process that generates power from salinity gradients [1], [2], [9], [10], [11] and [12]. PRO has the potential to sustainably produce electric power because seawater as a salinity source is vast. Unlike the forward osmosis (FO) process, PRO requires a back pressure on the high-salinity draw solution side that retards the permeate water flow and generates power by depressurizing the solution through a turbine. This water permeation against the applied hydraulic pressure depends on the osmotic pressure difference, which thereby converts the chemical potential energy into mechanical work. If a low-salinity source like river water is available, seawater can serve as the high salinity source, and energy can be extracted by the controlled mixing of the two streams through a semi-permeable membrane.
The osmotic pressure difference between two streams is an important factor in PRO to induce water permeation under the applied hydraulic pressure. Permeate water flux increases at higher osmotic pressure differences and the hydraulic pressure applied to the draw solution side can also be increased to obtain higher power density. In osmotic power generation by the PRO process, there is an optimal hydraulic pressure for a given osmotic pressure difference to obtain peak power density.
The majority of recent PRO studies have focused on the seawater/river water resource scheme [2], [4], [12] and [13]. However, the water flux and resulting power density based on this scheme is not high enough with current commercial membranes because of their relatively low water permeability and the suboptimal structure of their support layer that substantially reduce the available osmotic pressure difference [3], [13], [14] and [15]. Because brines from seawater desalination plants (∼70,000 ppm) paired with river water produce higher osmotic pressure differentials, this combination can be seen as a potential resource scheme for PRO [3], [5], [6], [8] and [16]. However, it is difficult to find places where this scheme can be implemented because desalination plants that produce high salinity brines are often located in arid or semi-arid regions where rivers with sufficient fresh water flow are scarce.
In this study, we analyze possible scenarios for PRO osmotic power generation based on a variety of salinity gradient resources. We also introduce a hybrid process of FO desalination and PRO power generation, where brine from an FO desalination plant is used as a high-salinity draw solution and seawater as feed solution to generate the salinity gradient for PRO. The potential of such a PRO osmotic power generation scheme is investigated via laboratory-scale PRO performance experiments using a commercial FO membrane.
The PRO process for osmotic power generation requires the availability of two streams with different salinities. Fig. 1 depicts possible scenarios of PRO osmotic power generation aimed at increasing power density that combine several salinity gradient resources more efficiently. Presently, the most commonly explored salinity-gradient scheme is the pairing of seawater with river water, which can be implemented at river mouths in coastal regions [2], [4], [12] and [13]. Fig. 1a shows low-salinity feed solution candidates that could be paired with a seawater draw solution. In addition to river water (100–600 mg/L, [5] and [8]), municipal wastewater effluent and brackish water could be selected as counterparts for seawater. The total dissolved solids (TDS) concentration of municipal wastewater effluent is typically in the range of 500–3000 mg/L [5] and [17]. If pretreatment can successfully prevent membrane fouling, municipal wastewater effluent could be an alternative to river water. Brackish water, which typically has a TDS between 1000 and 5000 mg/L [18], [19], [20] and [21], could also be considered as a low salinity source. Brackish groundwater is typically of high quality [18] and [19] and thereby would require minimal pretreatment if used for PRO power generation. However, the available amount of brackish water will likely be the limiting factor for use with seawater in PRO [22].
Fig. 1. Possible scenarios of osmotic power generation by PRO based on a variety of salinity gradient resources. (a) Seawater as draw solution and river water, municipal wastewater effluent, or brackish water as feed solution. (b) Brine from RO desalination plant as draw solution and municipal wastewater effluent or brackish water as feed solution. (c) Brine from FO desalination plant as draw solution and municipal wastewater effluent or brackish water as feed solution and (d) Seawater as a feed solution to a desalination plant, with the generated brine being used as a draw solution; in this case, there is no need to perform separate intake and pre-treatment processes for the feed and draw solutions.
Locating PRO plants in estuaries has several challenges, including potential ecological impacts, the need for intake structures and pre-treatment facilities to prevent membrane fouling, and insufficient salinity gradient for high power density. As with seawater desalination, where the reliability of the process is dependent on the intake [18], [19] and [22], a site-specific osmotic power plant would be highly dependent on the intake of two water resources (seawater and river water). Pretreatment of feed and draw streams would likely be similar to that required in conventional membrane plants to prevent fouling of the membranes [18], [19] and [20]. The osmotic pressure difference is about 25 bar and the concentration of seawater near the river mouth would likely be lower because of dilution of seawater at the estuary.
Because of the limitations of the seawater/river water salinity gradient scheme discussed above, it is necessary to explore other salinity-gradient resources. For example, concentrated brines from seawater RO desalination plants (70,000 ppm for 50% recovery) can be highly valuable resources for PRO power generation [3], [5], [6], [8], [16], [23], [24] and [25]. Fig. 1b shows low-salinity feed solution candidates—municipal wastewater effluent and brackish water—that can be paired with RO brine. Here, given that abundant fresh water sources, like river water, are often not available where desalination plants are located, PRO schemes that employ brines from desalination plants and require river water are not practical and hence excluded.
Recently, new desalination processes that can achieve high water recovery, like FO [26] and [27], membrane distillation (MD) [28], and RO/MD hybrid [29] and [30] have been proposed. The concentration of brine generated from these processes that are currently under development can be as high as 140,000 ppm (i.e., for 75% recovery). Instead of RO brines (Fig. 1b), these highly concentrated brines could be used as a draw solution for PRO as shown in Fig. 1c to produce higher power densities.
Instead of the low-salinity feed solutions mentioned above, seawater could be used as the low-salinity stream resource and paired against an RO or FO brine as the high-salinity resource. In such a case, there is no need to have separate intake and pretreatment for the feed and draw solutions as shown in Fig. 1d. The osmotic pressure difference between seawater and a brine from an RO desalination plant is high enough, but internal concentration polarization (ICP) would substantially reduce the osmotic pressure difference because the seawater feed would face the porous support layer [31]. This high concentration brine/seawater resource scheme may be an attractive approach if the ICP effect could be controlled through the development of high-performance PRO membranes.
Fig. 2 shows specifically a hybrid process of FO desalination and PRO power generation [32]. The PRO system on the left represents the osmotic power generation process using seawater as feed solution and concentrated brine from an FO desalination plant as draw solution. A turbine, high-pressure pump, and energy recovery device (ERD) are required for PRO power generation. The FO system on the right represents the osmotically driven desalination process. Two main configurations are possible with different input streams. Fig. 2a presents a hybrid configuration in which pretreated seawater is used as the feed solution for both the FO desalination and PRO power generation systems. In the second hybrid configuration (Fig. 2b), pretreated seawater is used as the feed solution for PRO, the brine from PRO is used as the feed solution for FO, and the brine from FO is used as the draw solution for PRO.
Fig. 2. Hybrid process of FO desalination and PRO power generation. The PRO part is the osmotic power generation process using seawater as feed solution and concentrated brine from an FO desalination plant as draw solution. Two main configurations are possible with different input streams. (a) Pretreated seawater is used as the feed solution for both FO desalination and PRO power generation and (b) Pretreated seawater is used as the feed solution for PRO, the brine from PRO is used as the feed solution for FO, and the brine from FO is used as the draw solution for PRO.
The membrane employed in our PRO experiments was a flat-sheet, cellulose-based FO membrane obtained from Hydration Technology Innovations (HTI), because there are no commercial membranes designed specifically for PRO. The FO membrane is reinforced by an embedded polyester mesh and is relatively thin (93 μm). The flat-sheet FO membrane coupon was loaded into a PRO crossflow test cell having two flow channels with the active layer facing the high salinity draw solution. The effective membrane surface area was 20.02 cm2.
A feed channel spacer is required in PRO experiments to maintain the channel geometry when hydraulic pressure is applied to the draw solution side. Without the feed channel spacer, the membrane would deform under the high hydraulic draw pressure and block the channel. Accordingly, we used commercial mesh-type spacers on both the feed and draw channels as described in our earlier publication [6]. The spacers are composed of two levels of polyethylene filaments forming a diamond-type spacer geometry.
A schematic diagram of our bench-scale experimental PRO unit is presented in Fig. 3. The crossflow test cell had two inlets and two outlets for the feed and draw solution streams. The feed and draw solution channels were 26 mm wide, 3 mm deep, and 77 mm long. A variable speed gear pump (Cole-Parmer, Vernon Hills, IL) and a high-pressure pump (Hydra-cell pump, Wanner Engineering, Inc., Minneapolis, MN) were used to circulate the feed and draw solutions in a closed loop, respectively. A bypass valve (Swagelok, Solon, OH) connected to the high pressure pump and a backpressure valve (Swagelok, Solon, OH) installed at the outlet on the draw side were manipulated to control the flow rate and inlet draw pressure. A water bath (Neslab, Newington, NH) was used for temperature control of both the feed and draw solutions. The weight of the feed solution was measured by a digital balance to obtain water permeate flux.
Fig. 3. Schematic diagram of the bench-scale pressure-retarded osmosis (PRO) experimental unit. To determine the water flux in PRO, the weight change of the feed solution was measured using a balance connected to a computer. The draw flow rate and inlet pressure were adjusted by using a bypass valve connected to the high pressure (HP) pump and a back-pressure valve on the draw outlet.
For PRO experiments, 1, 1.5, and 2 M NaCl solutions were used as high salinity draw solutions and a 0.5 M NaCl solution was used as a low salinity feed solution. The flow rates and temperatures of the two solutions were maintained at 0.5 L/min and 20 °C, respectively. The applied hydraulic pressure differences (ΔP) were 0.48, 2.88, 6.01, 9.25, and 12.61 bar. The weight of the feed solution was recorded in a data-logging program every 0.5 min and the average water flux was calculated over 120 data points (i.e., 60 min). Osmotic pressures of the NaCl solution were calculated using a commercial software program (Stream Analyzer, OLI Systems, Inc., Morris Plains, NJ).
According to our recent study [6], as the hydraulic pressure difference is increased in a PRO experiment, the membrane surface is compressed against the feed channel spacer mesh, resulting in membrane deformation and affecting membrane performance. Hence, to accurately predict PRO performance (water flux and corresponding power density), we used the water and salt permeability coefficients (A and B, respectively) and salt rejection (R) of the membrane obtained from a spacer-filled PRO test cell, as opposed to an RO test cell as is commonly used [3], [6], [33] and [34]. Specifically, we used A, B, and R of 1.23 L m−2 h−1 bar−1, 2.62 L m−2 h−1, and 76%, respectively, which account for membrane deformation, as reported in our previous study [6]. We note that the membrane A, B, and R, excluding membrane deformation, were reported to be 0.36 L m−2 h−1 bar−1, 0.32 L m−2 h−1, and 94%, respectively [34]. As for the mass transfer coefficient (k), we used a value of 8.62×10−5 m/s as determined from our study discussed above [6].
To determine the membrane structural parameter (S), we first determine the solute resistance to diffusion within the membrane support layer (K) and use S=KD, where D is the salt diffusion coefficient. We applied the parameter values of A, B, and k indicated above and the water flux values (JW) obtained from PRO experiments at ΔP=0.48 bar to the PRO water flux equation and solved numerically for K[4]:
For osmotic power generation, several types of salinity gradient resources may exist (Fig. 1). Based on the salinity of draw and feed solutions, possible salinity gradient resources are categorized in Table 1. Seawater, brine from seawater RO desalination at 50% recovery, and brine from seawater FO (or MD) desalination at 75% recovery are investigated in this study as high-salinity draw solutions and are simulated by concentrated solutions of 0.5, 1, and 2 M NaCl, respectively. River water, municipal wastewater effluent, brackish water, and seawater are considered as low-salinity feed solutions (simulated as 0.01, 0.05, 0.08, and 0.5 M NaCl solutions, respectively), to generate salinity gradients in combination with the above high salinity draw solutions.
Table 1. Possible resource schemes for PRO power generation based on the various salinity gradient resources described in Fig. 1. The various draw and feed streams were simulated as NaCl solutions and their corresponding osmotic pressures were determined from the OLI Stream Analyzer software (Morris Plains, NJ). The modeled results for water flux and projected power density for these resource schemes are shown in Fig. 4.
| Draw | Feed | |||
|---|---|---|---|---|
| River water (0.01 M) | Municipal wastewater (0.05 M) | Brackish water (0.08 M) | Seawater (0.5 M) | |
| Seawater (0.5 M) | Δπ=22.28 bar Fig. 4(a) | Δπ=20.45 bar Fig. 4(a) | Δπ=19.10 bar Fig. 4(a) | |
| Brine (from RO) (1 M) | Δπ=44.45 bar Fig. 4(b) | Δπ=43.09 bar Fig. 4(b) | Δπ=24.00 bar Fig. 4(b) | |
| Brine (from FO) (2 M) | Δπ=98.14 bar Fig. 4(c) | Δπ=96.79 bar Fig. 4(c) | Δπ=77.69 bar Fig. 4(c) | |
The PRO water flux model (Eq. (1)) was solved numerically to determine the theoretical water flux (JW) and the corresponding power densities (W=JWΔP) were calculated over a range of hydraulic pressure differences (ΔP). Fig. 4 presents the model results based on the parameter values of A, B, k, and S determined in the previous section for a spacer-filled channel loaded with the commercial FO membrane. Ideally, the power density reaches a maximum when the hydraulic pressure difference (ΔP) is half of the osmotic pressure difference, and the water flux vanishes when the hydraulic pressure difference is close to the osmotic pressure difference [3] and [35]. However, maximum power density in Fig. 4 occurs at ΔP<Δπ/2 and the theoretical flux reversal pressure was significantly lower than the ideal flux reversal pressure (ΔP=Δπ). This observation is attributed to the detrimental effects of ECP, ICP, and reverse salt diffusion (RSD) [13], [16] and [33]. Sensitivity analysis (data not shown) revealed that ICP followed by reverse salt diffusion were the major contributors for the deviation between the model results and the ideal case based on bulk osmotic pressure difference.
Fig. 4. Modeled water flux (JW) and respective projected power density (W) as a function of applied hydraulic pressure difference (ΔP) for the possible salinity gradient resources of osmotic power generation described in Fig. 1 and Table 1. River water, municipal wastewater effluent, brackish water, seawater, brine from a RO desalination plant, and brine from a FO desalination plant were simulated as 0.01, 0.05, 0.08, 0.5, 1, and 2 M NaCl solutions, respectively. Results are shown for three draw solutions: (a) seawater (0.5 M NaCl solution), (b) brine from a RO desalination plant (1 M NaCl solution), and (c) brine from a FO desalination plant (2 M NaCl solution). The bulk osmotic pressure differences (Δπ), calculated from the OLI Stream Analyzer software, are indicated by open circles on the horizontal axes of each graph.
For the 0.5 M NaCl draw solution (representative of seawater), the maximum power densities for each feed solution (0.01, 0.05, and 0.08 M NaCl solution) were calculated to be 1.51, 1.11, and 0.89 W/m2, respectively (Fig. 4a). These power densities are significantly lower than what is generally considered as an economically viable power density (∼5 W/m2) for PRO [8] and [12]. However, when the concentration of draw solution was increased up to 1 or 2 M NaCl (Fig. 4b and c), the power density significantly increased. As higher concentration draw solutions are used, the permeate flow rate increases as well as the hydraulic pressure difference necessary to generate maximum power density [36].
Interestingly, even though the osmotic pressure difference between the 1 M draw and 0.5 M feed solutions (Δπ=24.0 bar, Fig. 4b) was slightly larger than that between the 0.5 M draw and 0.01 M feed solutions (Δπ=22.3 bar, Fig. 4a), the projected power density obtained with the 1 M draw and 0.5 M feed solutions (1.03 W/m2) was much lower than that calculated for the 0.5 M draw and 0.01 M feed solutions (1.51 W/m2). This observation is attributed to the severe ICP that occurs within the membrane porous support layer when high salinity feed solutions are used [15], [31] and [37]. Accordingly, higher salinity draw solutions should be exploited to obtain the target power density when seawater feed solution is considered. When a 2 M NaCl draw solution was used, the projected power density increased to 6.76 W/m2 at a hydraulic pressure difference of 32.5 bar, in spite of the use of 0.5 M NaCl feed solution. If municipal wastewater effluent or brackish water (simulated as 0.05 and 0.08 M NaCl, respectively), instead of seawater (0.5 M NaCl), are paired with RO brine or FO brine (1 and 2 M NaCl, respectively), higher power density will be obtained as shown in Fig. 4b and c.
Fig. 5 shows a contour plot of projected power density as a function of salinity difference between the feed and draw solutions. The concentration of the high-salinity draw solution ranges from 1 to 2 M NaCl, while that of the low-salinity feed solution varies from 0 to 0.5 M NaCl. The right lower corner of the graph indicates the highest salinity gradient area and the left higher corner of the graph indicates the lowest salinity gradient area. At a hydraulic pressure difference of 10 bar (Fig. 5a), the contour line of projected power density increased from 0.84 to 8.6 W/m2. In this case, the maximum power density cannot be obtained over the entire range of salinity gradient because the optimal hydraulic pressure difference for peak power density varies depending on the salinity gradient. Only the left top point indicating 1 M draw and 0.5 M feed represents the peak power density at ΔP=10 bar. On the other hand, at a hydraulic pressure difference of 32.5 bar (Fig. 5b), the contour line of the projected power density rises from 0 to 19.2 W/m2. The peak power density at 32.5 bar exists at the right top point (2 M draw and 0.5 M feed). The red triangle in Fig. 5b indicates an unrealistic and meaningless region because the calculated power density is based on negative water fluxes as the hydraulic pressure differences exceed the flux reversal pressure (i.e., ΔP≥Δπ).
Fig. 5. Projected power density (W) as a function of draw and feed solution concentrations at a hydraulic pressure difference (ΔP) of (a) 10 bar and (b) 32.5 bar. The red triangle in (b) indicates the calculated results based on the water fluxes at hydraulic pressure differences above the flux reversal point (ΔP≥Δπ). The open red circle on each graph represents the peak power density of the salinity gradient. Each hydraulic pressure difference is the optimal pressure difference of the salinity gradient. The black line on the scale bar in (b) represents a power density of 0 W/m2. The osmotic pressures (π) of the draw and feed solutions were determined by OLI Stream Analyzer software. (For interpretation of references to color in this figure, the reader is referred to the web version of this article.)
To investigate the feasibility of PRO power generation by seawater feed solution with several possible resources, we used a 0.5 M NaCl solution as the feed solution and concentrated solutions of 1, 1.5, and 2 M NaCl as draw solutions. Experimentally measured water flux (JW) data obtained in the PRO test cell and the corresponding projected power density (W) based on water flux as a function of the hydraulic pressure difference for the three draw solution concentrations are presented in Fig. 6.
Fig. 6. Modeled and experimental water flux (JW) and the respective calculated power density (W) as a function of applied hydraulic pressure difference (ΔP). (a) Modeling results for the indicated draw solutions. (b–d) Experimental and modeling results for 1, 1.5, and 2 M NaCl draw solutions. The osmotic pressures (π) of the 1, 1.5, and 2 M NaCl draw solutions were 46.75, 72.72, and 100.44 bar, respectively, as determined by OLI Stream Analyzer software. Also, the osmotic pressure (π) of the 0.5 M NaCl feed solution was 22.75 bar. For each graph, symbols (open squares and circles) represent experimental water fluxes and the corresponding calculated power densities, respectively, and lines represent model results. All experiments were performed with solutions at a fixed temperature of 20 °C.
Fig. 6a compares modeling results for the indicated draw solutions (1, 1.5, and 2 M NaCl) paired with a 0.5 M NaCl feed solution. The water flux decreases linearly and power density shows a quadratic function curve with a peak point as a function of the hydraulic pressure difference [3] and [35]. The power density depends on both the water flux and the hydraulic pressure difference between both sides of the membrane (W=JWΔP). The theoretical model results, however, were significantly lower than the ideal case based on the bulk osmotic pressure difference. As discussed earlier, this deviation is mostly attributed to the performance limiting effects of ICP and reverse salt diffusion.
The experimental water flux and corresponding projected power densities (symbols) are compared with model predictions (lines) in Figs. 6b through 6d. The theoretical maximum power densities for each draw solution concentration (1, 1.5, and 2 M) are 1.03, 3.17, and 6.76 W/m2, respectively. For the osmotic pressure difference between the 1 M draw and 0.5 M feed solution (Δπ=24 bar, Fig. 6b), we observed a peak power density of 0.73 W/m2 at a hydraulic pressure difference of 9.30 bar. However, for the higher salinity gradient conditions using 1.5 and 2 M NaCl draw solutions, we were able to confirm only the increasing trend of the projected power density, because PRO experiments were not possible in the PRO test cell loaded with the commercial FO membrane coupon at pressures greater than 12.6 bar due to the PRO test cell's structural limitations. At a hydraulic pressure difference of 12.6 bar, water fluxes of 5.91 and 9.23 L m−2 h−1 (1.64×10−6 and 2.56×10−6 m/s) and corresponding projected power densities (W) of 2.07 and 3.22 W/m2 were obtained for the 1.5 and 2 M NaCl draw solutions, respectively.
The model results, obtained with the parameters derived from a spacer-filled channel, closely matched the experimental data at the lowest hydraulic pressure difference used (ΔP=0.47 bar). However, as the hydraulic pressure difference increased, the deviation between the theoretical and experimental values also increased. The theoretical results accounted for the effect of membrane deformation induced by the feed channel spacer because they are calculated on the basis of the parameter values (A, B, k, and S) determined in a spacer-filled channel [6]. However, the predictions do not account for the spacer “shadow effect”, that is, the blocking of available membrane area for water permeation by the spacer strands under high hydraulic pressure [6]. Accordingly, we attribute the lower-than-predicted water fluxes to the “shadow effect.”
In pressure-driven membrane processes, temperature control can be considered as a way to enhance water flux. Similarly, in osmotically driven membrane processes like FO and PRO, increasing temperature is an important factor for water flux enhancement due to an increase of the water permeability coefficient [21] and [38]. Increasing temperature results in a concomitant increase in the salt permeability coefficient [16] and [38], which can be detrimental for RO and FO applications. However, for PRO applications, there is no need for a low salt permeability coefficient as is required in FO and RO [13], thereby adding another degree of freedom for PRO optimization.
To investigate the effect of solution temperature on the water flux and power density in PRO, we performed PRO experiments at two different temperatures as presented in Fig. 7. Increasing the solution temperature had an obvious effect on the water flux across the membrane. When the temperature increased from 20 °C to 30 °C, the water flux and the corresponding projected power density increased significantly. The obtained water flux increased from 9.23 to 13.89 L m−2 h−1 and the corresponding projected power density increased from 3.22 to 4.72 W/m2 for the 2 M draw and 0.5 M feed solutions (Fig. 7c).
Fig. 7. Effect of solution temperature on measured water flux (JW) in PRO and the corresponding calculated power density (W). (a) 1 M NaCl draw–0.5 M NaCl feed solution, (b) 1.5 M NaCl draw–0.5 M NaCl feed solution and (c) 2 M NaCl draw–0.5 M NaCl feed solution. All experiments were performed with solutions at two temperatures (20 and 30 °C). The results were compared at a hydraulic pressure difference of 12.5 bar.
The temperature of seawater normally ranges from 12 to 35 °C [19]. Because desalination plants are often located in hot regions, the seawater temperature in such areas is over 25 °C. We also note that the temperature of brines discharged from desalination plants is generally higher than that of raw seawater [39]. Also, if evaporation ponds are used for the discharged brine, the brine temperature will increase with increasing salinity [40] and [41]. Lastly, if heat supply by renewable solar energy and waste heat from a power plant is available, the enhancement of water flux in the PRO process may be feasible.
To date, seawater has not been considered as a feed solution for the PRO process because of severe ICP effects in current generation membranes. However, once high-performance PRO membranes are developed, high concentration brines paired with seawater can be an attractive source for salinity gradient. Future PRO membranes should not only have porous support layers with low ICP effects, but they should also withstand the high hydraulic pressures that are necessary to produce higher power density at high salinity gradients. There is also a critical need to optimize feed channel spacers to minimize membrane deformation and the “shadow effect” in order to achieve high power densities. We therefore conclude that both membrane and module designs remain a major hurdle for PRO osmotic power generation. Future PRO installations must also address pre-treatment of the low and high concentration streams as well as membrane fouling both at the membrane active layer and within the membrane support layer.
Dr. Yu Chang Kim was supported by the Postdoctoral Program of the Korea Institute of Machinery and Materials (KIMM). Professor Elimelech acknowledges the support of the World Class University (WCU) Program (Case III) through the National Research Foundation of Korea and the Ministry of Education, Science and Technology (R33-10046).
Membrane-based processes for sustainable power generation using water
Nature, 488 (2012), pp. 313–319
The potential for power production from salinity gradients by pressure retarded osmosis
J. Membr. Sci., 335 (2009), pp. 103–110
Power generation with pressure retarded osmosis: An experimental and theoretical investigation
J. Membr. Sci., 343 (2009), pp. 42–52
Thin-film composite pressure retarded osmosis membranes for sustainable power generation from salinity gradients
Environ. Sci. Technol., 45 (2011), pp. 4360–4369
Thin-film composite hollow fiber membranes for pressure retarded osmosis (PRO) process with high power density
J. Membr. Sci., 389 (2012), pp. 25–33
Adverse impact of feed channel spacers on the performance of pressure retarded osmosis
Environ. Sci. Technol., 46 (2012), pp. 4673–4681
Pressure retarded osmosis: From the vision of Sidney Loeb to the first experimental installation–review
Desalination, 261 (2010), pp. 205–211
Membrane-based production of salinity-gradient power
Energy Environ. Sci., 4 (2011), pp. 4423–4434
Energy production at the Dead sea by pressure-retarded osmosis: challenge or chimera
Desalination, 120 (1998), pp. 247–262
One hundred and thirty benign and renewable megawatts from Great Salt Lake? The possibilities of hydroelectric power by pressure-retarded osmosis
Desalination, 141 (2001), pp. 85–91
Large-scale power production by pressure-retarded osmosis, using river water and sea water passing through spiral modules
Desalination, 143 (2002), pp. 115–122
Osmotic power—power production based on the osmotic pressure difference between waters with varying salt gradients
Desalination, 220 (2008), pp. 476–482
Performance limiting effects in power generation from salinity gradients by pressure retarded osmosis
Environ. Sci. Technol., 45 (2011), pp. 10273–10282
Effect of porous support fabric on osmosis through a Loeb-Sourirajan type asymmetric membrane
J. Membr. Sci., 129 (1997), pp. 243–249
Internal concentration polarization in forward osmosis: role of membrane orientation
Desalination, 197 (2006), pp. 1–8
Osmotic power production from salinity gradient resource by pressure retarded osmosis: effects of operating conditions and reverse solute diffusion
J. Membr. Sci., 401–402 (2012), pp. 262–273
Indirect desalination of Red Sea water with forward osmosis and low pressure reverse osmosis for water reuse
Desalination, 280 (2011), pp. 160–166
State-of-the-art of reverse osmosis desalination
Desalination, 216 (2007), pp. 1–76
Reverse osmosis desalination: water sources, technology, and today's challenges
Water Research, 43 (2009), pp. 2317–2348
Feed water pretreatment in RO systems: unit processes in the Middle East
Desalination, 150 (2002), pp. 235–245
Effects of working temperature on separation performance, membrane scaling and cleaning in forward osmosis desalination
Desalination, 278 (2011), pp. 157–164
Seawater intakes for desalination plants
Desalination, 156 (2003), pp. 249–256
Production of energy from concentrated brines by pressure-retarded osmosis. I. Preliminary technical and economic correlations
J. Membr. Sci., 1 (1976), pp. 49–63
Production of energy from concentrated brines by pressure-retarded osmosis. II. Experimental results and projected energy costs
J. Membr. Sci., 1 (1976), pp. 249–269
Performance of permasep B-9 and B-10 membranes in various osmotic regions and at high osmotic pressures
J. Membr. Sci., 4 (1979), pp. 335–349
A novel ammonia-carbon dioxide forward (direct) osmosis desalination process
Desalination, 174 (2005), pp. 1–11
Desalination by ammonia–carbon dioxide forward osmosis: ınfluence of draw and feed solution concentrations on process performance
J. Membr. Sci., 278 (2006), pp. 114–123
Desalination using membrane distillation: experimental studies on full scale spiral wound modules
J. Membr. Sci., 375 (2011), pp. 104–112
Osmotically and thermally driven membrane processes for enhancement of water recovery in desalination processes
Desalination Water Treat., 15 (2010), pp. 279–286
High recovery of concentrated RO brines using forward osmosis and membrane distillation
J. Membr. Sci., 331 (2009), pp. 31–39
Internal polarization in the porous substructure of a semipermeable membrane under pressure-retarded osmosis
J. Membr. Sci., 4 (1978), pp. 261–265
Reverse draw solute permeation in forward osmosis: modeling and experiments
Environ. Sci. Technol., 44 (2010), pp. 5170–5176
High performance thin-film composite forward osmosis membrane
Environ. Sci. Technol., 44 (2010), pp. 3812–3818
Membranes for power generation by pressure-retarded osmosis
J. Membr. Sci., 8 (1981), pp. 141–171
A novel ammonia–carbon dioxide osmotic heat engine for power generation
J. Membr. Sci., 305 (2007), pp. 13–19
Influence of concentrative and dilutive internal concentration polarization on flux behavior in forward osmosis
J. Membr. Sci., 284 (2006), pp. 237–247
Impacts of operating conditions and solution chemistry on osmotic membrane structure and performance
Desalination, 287 (2012), pp. 340–349
Salt production by the evaporation of SWRO brine in Eliat: a success story
Desalination, 205 (2007), pp. 374–379
Membrane concentrate management options: a comprehensive critical review
Can. J. Civ. Eng., 36 (2009), pp. 1107–1119
Copyright © 2012 Elsevier B.V. Published by Elsevier B.V. All rights reserved.
