Abstract

Seawater reverse osmosis (SWRO) desalination technology accounted for 78% of annual new contracted desalination capacity from 2006 to 2012, due of its lower electrical power consumption (3–5 kWh/m3) versus thermal desalination technologies (up to 18 kWh/m3), such as multistage flash and multiple effects. However, the existing SWRO desalination process still needs further improvement to lower the energy consumption. Recently, a novel hybrid SWRO desalination system using pressure-retarded osmosis (PRO) technology has been studied, which can recover a large amount of osmotic power from the concentrated brine. In this study, GS Engineering & Construction Corp. (GS E&C) developed an advanced SWRO-PRO process to economically couple this PRO technology with a conventional SWRO desalination process. To investigate the SWRO-PRO process, pilot and demonstration plants, of 20 m3/d and 240 m3/d PRO treatment capacity, were constructed and thin-film composite spiral-wound PRO membrane modules (8 inches) were assessed. The operating parameters of the pilot plants, such as pressure, temperature, and flow rates of the draw solution and the feed solution, were found to be important factors determining the plant energy consumption and operating efficiency. An economic analysis of a large-scale SWRO-PRO hybrid desalination is also described.

INTRODUCTION

Global water scarcity problems are becoming more severe because of global warming, environmental contamination, industrialization, and population growth. It is anticipated that 1.8 billion people will suffer from absolute water scarcity, and two-thirds of the global population will live under water-stressed conditions by 2025 (Few Resources, www.fewresource.org). Securing available water resources, especially in North Africa, the Middle East, India, Southeast Asia, and North China, will be a major issue by 2020 and there is a high chance of regional conflicts caused by water scarcity issues. Thus, the desalination market will grow from an estimated 9,882 million USD in 2013 to 15,274 million USD by 2018 with compound annual growth rate of 9.1% (Global Trends and Forecasts 2014; Lu et al. 2015).

Compared with thermal desalination technologies, such as multistage flash and multiple effects, membrane-based (i.e. reverse osmosis (RO)) desalination is more energy efficient because of the development of high-performance RO membranes and modules, and energy recovery devices (ERDs). However, RO technology still needs to be improved further to lower its still high energy consumption. Recently, a new-generation desalination technology using pressure-retarded osmosis (PRO) has been investigated in an attempt to minimize the energy required for desalination (Prante et al. 2014; Wan & Chung 2015; Chung et al. 2016).

PRO is a promising osmotic energy recovery technology to extract osmotic energy (or pressure) by transporting water through a semi-permeable membrane, from a low-salinity feed solution (FS; e.g. wastewater effluent) to a high-salinity draw solution (DS; e.g., brine) against an applied hydraulic pressure. Many PRO studies have focused on the fundamental mechanisms of the salt and water transfer kinetics in laboratory-scale systems, with the goal of clarifying the properties of the PRO membrane and the power density (She et al. 2012; Helfer et al. 2014). Although these laboratory-scale studies are necessary to estimate the relative performance of various membranes, and to clarify the energy recovery mechanism in PRO systems, few pilot-scale studies with a commercial-size PRO membrane module have been conducted to investigate the key design and operational parameters of a PRO system.

In this study, large-scale seawater reverse osmosis (SWRO)-PRO pilot and demonstration plants (20 m3/d and 240 m3/d of PRO treatment capacity) were constructed and operated with thin-film composite (TFC) spiral-wound membrane modules (8 inches). These pilot studies were conducted under various operating conditions. The effects of hydraulic pressure, temperature, and various versions of a PRO membrane module on PRO system performance were investigated. Moreover, an economic analysis of a large-scale SWRO-PRO hybrid desalination plant (100,000 m3/d) was conducted to elucidate the effects of important plant design parameters on plant Capex and Opex.

MATERIAL AND METHODS

Pilot plant description and experimental conditions

As shown in Figure 1, the pilot plant has 20 m3/d of PRO treatment capacity and it consists essentially of SWRO and PRO systems, and a control office. The PRO system includes the feed pumps and tanks for DS and FS, four 8-inch PRO membrane modules, and a Pelton turbine to generate electricity from the PRO system. The DS and FS were injected using a high- (plunger type) and low-pressure (horizontal centrifugal type) pump, respectively. The hydraulic pressure and flow rate of the DS and FS were controlled with flow control valves. The pressure, flow rate, concentration, conductivity, and temperature were monitored using the human-machine interface (HMI) of the SWRO-PRO system. To ensure accurate conversion of conductivity into concentration, the conductivity meters were calibrated before the experiment. Data were collected every 10 s.

Figure 1

Pilot plant with 20 m3/d of pressure-retarded osmosis (PRO) treatment capacity: (a) pilot plant composition diagram, (b) an inside view of the pilot plant, and (c) pilot plant schematic flow diagram.

Figure 1

Pilot plant with 20 m3/d of pressure-retarded osmosis (PRO) treatment capacity: (a) pilot plant composition diagram, (b) an inside view of the pilot plant, and (c) pilot plant schematic flow diagram.

The PRO membrane modules were manufactured by Toray Chemical Korea (Seoul, Korea). The 8-inch modules are 40 inches in length and 18 m2 in total membrane area. The physical characteristics of the membrane were presented previously (Jeon et al. 2015). In the PRO experiments, the DS and FS were injected into the module in a co-current direction. Unless specified otherwise, all reagents and chemicals were of analytical grade. Certified ACS-grade NaCl (Fisher Scientific, Waltham, MA, USA) was used to prepare the SWRO FS at 35,000 mg/L. The SWRO brine concentration with 70,000 mg/L was used as a high-salinity DS and a SWRO permeate was used as a low-salinity FS. The flow rates of the DS and FS were 10 L/min each. The applied hydraulic pressure differences ranged from 5 to 35 bar.

Mathematical estimation of PRO system performance

The PRO performance was evaluated based on the following Equations (1) to (7). The water flux, across a PRO membrane is defined as:  
formula
(1)
where A is the water permeability coefficient of the membrane, is the transmembrane osmotic pressure difference, and is the transmembrane hydraulic pressure difference. The values of the characteristic parameters A and B of PRO membrane modules were obtained from a previous study (Jeon et al. 2015) or were provided by Toray Chemical Korea (Table 1).
Table 1

PRO membrane properties and structure parameters

Model NameCSM-PRO-1CSM-PRO-2CSM-PRO-3
Type TFC TFC TFC 
Thickness (μm) 165 145 125 
A (L/m2/hr/bar) 1.8 1.55 2.85 
B (L/m2/hr) 0.714 0.187 0.466 
S (mm) 2.53 2.098 0.926 
Model NameCSM-PRO-1CSM-PRO-2CSM-PRO-3
Type TFC TFC TFC 
Thickness (μm) 165 145 125 
A (L/m2/hr/bar) 1.8 1.55 2.85 
B (L/m2/hr) 0.714 0.187 0.466 
S (mm) 2.53 2.098 0.926 
The salt permeability coefficient (B) was determined from Equation (2) (Lee et al. 1981):  
formula
(2)
where R is the salt rejection of the module, is the water flux, and k is the mass transfer coefficient.  
formula
(3)
where D is the diffusion coefficient of NaCl, which is presented in a previous study (Chung et al. 2016), is the hydraulic diameter, and Sh is the Sherwood number (Lee et al. 2015):  
formula
(4)
The solute resistivity to salt transport in a porous substrate , which is a function of the structural parameters (S) and the salt diffusion coefficient (D), , was calculated using Equation (5) (Yip & Elimelech 2011) and the water flux:  
formula
(5)
Water flow across a PRO membrane is induced by the transmembrane osmotic pressure difference . The water flux decreases as the pressure difference increases (Gray et al. 2006; McCutcheon & Elimelech 2006) because the pressure on the PRO membrane is exerted in the direction opposite to the water flow. The power density is defined as below:  
formula
(6)
The water recovery rate, was calculated using:  
formula
(7)
where is the initial volume of the FS and is its change in volume.

RESULTS AND DISCUSSION

Effect of hydraulic pressure on PRO system performance

It is known that the hydraulic pressure of DS is one of the most important operating factors determining PRO system performance, such as power density, recovery of the FS, and ultimately the extractable energy. In this study, the flow rate and ratio of the DS and FS were 10 LPM and 1:1, respectively. As shown in Figure 2, as the hydraulic pressure increased from 5 bar, the power density (=JW × ΔP) of the PRO membrane module increased up to 3.4 W/m2 at 24 bar and then decreased to 2.9 W/m2 at 34 bar. The optimal pressure generating the maximum power density of the PRO membrane module was lower than the theoretical optimal pressure (ΔP = Δπ/2) (Jeon et al. 2015) because it may be induced by the reduction of the osmotic pressure through the membrane module, resulting from the changes in the DS and FS concentrations. Unlike the power density results, the recovery rate of the FS decreased continuously with the hydraulic pressure increase due to the reduction in the driving force of FS permeation (i.e. the hydraulic and osmotic pressure differences) across the membrane. In terms of system performance, extractable energy is the more important parameter, determined by the FS recovery and hydraulic pressure of the PRO system. It was found that the relative extractable energy (= hydraulic pressure × recovery) was highest at 20 bar with the CSM-PRO-2 module (the second version of the PRO membrane module).

Figure 2

Effects of hydraulic pressure of PRO DS on power density and recovery (with the second version 8-inch PRO membrane module, CSM-PRO-2).

Figure 2

Effects of hydraulic pressure of PRO DS on power density and recovery (with the second version 8-inch PRO membrane module, CSM-PRO-2).

Effect of the PRO membrane and module on PRO system performance

To enhance the economic feasibility of the SWRO-PRO hybrid process, it is essential to have a high-performance PRO membrane and module. Three versions of PRO membrane modules (CSM-PRO-1, CSM-PRO-2, CSM-PRO-3) have been developed by Toray Chemical Korea with different types and structural parameters of the PRO membrane support layer (Jeon et al. 2015). The PRO membrane properties and structure parameters are shown in Table 1. The manufactured 8-inch PRO membrane modules were investigated in the pilot plant. The performance of the PRO membrane modules were compared at the DS and FS rate of 10 LPM (liter per minute) under the applied hydraulic pressure of 20 bar. The DS was RO brine (70,000 ppm) produced from artificial seawater (35,000 ppm) and the FS was RO permeate (500 ppm). Figure 3 shows the results on the flux and recovery for the three kinds of the PRO membrane modules. The first version of the PRO membrane module showed 2.3 LMH of permeate flux and 9.2% FS recovery. These performance values were improved greatly with the second and third versions: 300% and 574% in flux, and 224% and 440% in recovery, respectively. The single-element module of the latest version, CMS-PRO-3, could achieve 13.2 LMH flux and 40.2% recovery. It is expected that the membrane module performance will be enhanced further by improvements in the membrane support layer and module design, and by enlargement of the membrane surface area.

Figure 3

Flux and recovery of PRO modules (CSM-PRO-1, CSM-PRO-2, and CSM-PRO-3).

Figure 3

Flux and recovery of PRO modules (CSM-PRO-1, CSM-PRO-2, and CSM-PRO-3).

Simulation of temperature effect on PRO performance

To simulate the power density of the PRO membrane module affected by temperature, the water permeability (A), salt permeability (B), and membrane structure parameter (S) were used (She et al. 2012). The values of the characteristic factors, A and B, were obtained from previous studies using Equations (1) and (2) (Jeon et al. 2015). The relationship between the water permeability (A) and temperature is (Kim & Park 2011; Lee et al. 2015):  
formula
(8)
The relationship between the salt permeability (B) and temperature is:  
formula
(9)
Because both water permeability (A) and salt permeability (B) are dependent on temperature, the power density of the PRO modules is also temperature-dependent (Anastasio et al. 2015; Chung et al. 2016). Figure 4 presents the simulated power density of the three versions of the PRO module at different temperatures. Clearly, PRO performance improved significantly as the temperature increased from 5 to 35 °C, similar to previous studies (Zhao & Zou 2011; Anastasio et al. 2015). At higher temperatures, the diffusivity of the sodium chloride is increased from 0.80 to 1.68 × 10−9 m2/s, which corresponds to enhance the mass transfer coefficient (Chung et al. 2016). Moreover, the later versions of the PRO membrane module also showed better performance, in the order CSM-PRO-3 > CSM-PRO-2 > CSM-PRM-1, which resulted from the change on the PRO membrane structure parameter, S. According to these results, the maximum simulated power density achievable at 35 °C was 13 W/m2 with the CSM-PRO-3.
Figure 4

Simulated power density of PRO modules (CSM-PRO-1, CSM-PRO-2, and CSM-PRO-3).

Figure 4

Simulated power density of PRO modules (CSM-PRO-1, CSM-PRO-2, and CSM-PRO-3).

Economic analysis of a SWRO-PRO hybrid system

There are two processes available to recover the osmotic energy in a SWRO-PRO hybrid desalination system: use of (1) a hydraulic turbine and (2) an ERD. Because of the lower mechanical performance efficiency of a hydraulic turbine (80–85%) versus a pressure exchange-type ERD (93–98%), the SWRO-PRO hybrid desalination process with two ERDs, shown in Figure 5(a), was more efficient. ERD (e.g. a rotary pressure exchanger with a cylindrical rotor with longitudinal ducts parallel to its rotational axis) is used to pressurize the low-pressure seawater from high-pressure brine. This process was systemized in the demonstration plant (240 m3/d, PRO treatment capacity) in Busan, Korea, and it has been operated since November 2015.

Figure 5

(a) A process diagram for the seawater reverse osmosis (SWRO)-PRO hybrid desalination process and (b) an outside view of the demonstration plant in Busan, Korea.

Figure 5

(a) A process diagram for the seawater reverse osmosis (SWRO)-PRO hybrid desalination process and (b) an outside view of the demonstration plant in Busan, Korea.

For the economic analysis of the SWRO-PRO hybrid desalination process, the operating results of the pilot and demonstration plants were used with several assumptions. The hybrid plant was assumed to have 100,000 m3/d of production capacity and the unit costs of the PRO membrane module and pressure vessel were the same as those of an RO system. It was assumed that 20% of the PRO membrane modules will be replaced each year. The total construction cost of the SWRO desalination plant (100,000 m3/d) was 110 million USD, and the breakdown cost of unit systems is given in Table 2. The plant energy consumption rate was set at 3.5 kWh/m3 and the electricity cost is 50% of the total plant operating cost (Opex).

Table 2

Construction cost (Capex) of a SWRO desalination plant with 100,000 m3/d of production capacity

Capex (SWRO desalination plant)
ItemsProportion (%)Amount (Million USD)
Engineering 5.5 
Intake/outfall 8.8 
Pre-treatment 20 22 
RO system 22 24.2 
Post-treatment 2.2 
Electric 9.9 
Others 6.6 
Construction 23 25.3 
Commissioning 5.5 
Total 100 110 
Capex (SWRO desalination plant)
ItemsProportion (%)Amount (Million USD)
Engineering 5.5 
Intake/outfall 8.8 
Pre-treatment 20 22 
RO system 22 24.2 
Post-treatment 2.2 
Electric 9.9 
Others 6.6 
Construction 23 25.3 
Commissioning 5.5 
Total 100 110 

As shown in Figure 6, the plant construction cost with the first-version PRO membrane module was 194 million USD and this represented a 76.5% increase in Capex versus an SWRO desalination plant. This cost increment was reduced with the second- and third-version PRO membrane modules, to 47.5% and 33.0%, respectively. Other important factors in determining plant Capex and Opex are the flow rates of the DS and FS into the PRO pressure vessel. With a fixed recovery in the PRO system, a higher flow rate can reduce the number of PRO membrane modules and pressure vessels required, which can ultimately reduce the plant Capex and Opex. Accordingly, the break-even point (BEP), a balanced point making neither a profit nor a loss, can be reduced (Figure 7). Also, the higher energy saving rate of the PRO system can also result in a lower BEP. With a 25% energy saving rate and 20 LPM for the DS and FS flow rates, the BEP can decrease to 7.3 years. Additionally, because desalination is a highly energy-dependent process, the Opex is affected greatly by energy costs (power tariffs). The results with power tariff variation from 0.05 to 0.30 USD/kWh are presented in Figure 8. It is found that the 87.7-year BEP at 0.05 USD/kWh could be reduced markedly, to 3.7 years, at 0.30 USD/kWh.

Figure 6

Estimated total construction cost (Capex) with three versions of the PRO membrane module.

Figure 6

Estimated total construction cost (Capex) with three versions of the PRO membrane module.

Figure 7

Estimated total construction cost (Capex) and break-even point (BEP) with three versions of the PRO membrane module and energy saving percentages.

Figure 7

Estimated total construction cost (Capex) and break-even point (BEP) with three versions of the PRO membrane module and energy saving percentages.

Figure 8

Estimated BEP with various power tariffs (25% energy saving rate and 15 LPM draw and FS flow rates).

Figure 8

Estimated BEP with various power tariffs (25% energy saving rate and 15 LPM draw and FS flow rates).

CONCLUSIONS

Although a SWRO-PRO hybrid desalination process has been a ‘hot’ research topic in recent years because of its potential for renewable energy generation from RO brine, a desalination by-product, most of the experimental and analytical work has been conducted at the laboratory scale. However, in this study, a pilot plant with commercial-scale PRO membrane modules (8 inches) was investigated and the results showed the economic potential of a SWRO-PRO hybrid system, which can be further enhanced with higher-performance PRO membrane modules, operating temperatures, treatment capacity, and energy costs. In future studies, the PRO membrane module will be further improved for higher treatment capacities. Moreover, the effects of various types of feed water solutions will be investigated. Also, an economic sensitivity analysis needs to be conducted with diverse PRO pre-treatment options, the benefits of brine dilution, carbon taxes, and environmental impact minimization.

ACKNOWLEDGEMENTS

This research was supported by a grant (code 16IFIP-B065893-04) from Industrial Facilities and Infrastructure Research Program funded by the Ministry of Land, Infrastructure and Transport of the Korean government.

REFERENCES

REFERENCES
Anastasio
,
D. D.
,
Arena
,
J. T.
,
Cole
,
E. A.
&
McCutcheon
,
J. R.
2015
Impact of temperature on power density in closed-loop pressure retarded osmosis for grid storage
.
Journal of Membrane Science
479
,
240
245
.
Chung
,
K.
,
Yeo
,
I. H.
,
Lee
,
W.
,
Oh
,
Y. K.
,
Kim
,
D.
&
Park
,
Y. G.
2016
Investigation into design parameters in seawater reverse osmosis (SWRO) and pressure retarded osmosis (PRO) hybrid desalination process: a semi-pilot scale study
.
Desalination Water and Technology
57
(
51
),
24636
24644
.
Global Trends and Forecasts
2014
Water desalination equipment market by technology (RO, MSF, MED, others), by application (municipal, industrial, others), by products (membrane systems, pumps, evaporators, others) and by geography (Europe, the U.S., Asia-Pacific, the Middle East, and the Rest of the World), Seattle, WA, USA
.
Gray
,
G. T.
,
McCutcheon
,
J. R.
&
Elimelech
,
M.
2006
Internal concentration polarization in forward osmosis: role of membrane orientation
.
Desalination
197
,
1
8
.
Helfer
,
F.
,
Lemckert
,
C.
&
Anissimov
,
Y. G.
2014
Osmotic power with pressure retarded osmosis: theory, performance and trends – A review
.
Journal of Membrane Science
453
,
337
358
.
Jeon
,
E. J.
,
Sim
,
J. H.
&
Lee
,
J. H.
2015
Development of thin-film composite PRO membranes with high power density
.
Desalination Water and Technology
57
,
10093
10100
.
Kim
,
Y. C.
&
Park
,
S. J.
2011
Experimental study of a 4040 spiral-wound FO membrane module
.
Environmental Science and Technology
45
,
7737
7745
.
Lee
,
K. L.
,
Baker
,
R. W.
&
Lonsdale
,
H. K.
1981
Membranes for power generation by pressure-retarded osmosis
.
Journal of Membrane Science
8
,
141
171
.
Lee
,
S.
,
Kim
,
Y. C.
,
Park
,
S. J.
,
Lee
,
S. K.
&
Choi
,
H. C.
2015
Experiment and modelling for performance of a spiral-wound pressure-retarded osmosis membrane module
.
Desalination Water and Technology
57
,
10101
10110
.
Lu
,
X.
,
Chavez
,
L. H. A.
,
Castrillon
,
S. R. V.
,
Ma
,
J.
&
Elimelech
,
M.
2015
Influence of active layer and support layer surface structures on organic fouling propensity of thin film composite forward osmosis membrane
.
Environmental Science and Technology
49
,
1436
1444
.
Prante
,
J. L.
,
Ruskowitz
,
J. A.
,
Childress
,
A. E.
&
Achilli
,
A.
2014
RO-PRO desalination: an integrated low-energy approach to seawater desalination
.
Applied Energy
120
,
104
114
.