DOI 10.1515/reveh-2014-0019 Rev Environ Health 2014; 29(1-2): 67–70
Vasiliki Karanikola*, Andrea F. Corral, Patrick Mette, Hua Jiang, Robert G. Arnoldand and Wendell P. Ela
Solar membrane distillation: desalination for the Navajo Nation Abstract: Provision of clean water is among the most serious, long-term challenges in the world. To an ever increasing degree, sustainable water supply depends on the utilization of water of impaired initial quality. This is particularly true in developing nations and in water-stressed areas such as the American Southwest. One clear example is the Navajo Nation. The reservation covers 27,000 square miles, mainly in northeastern Arizona. Low population density coupled with water scarcity and impairment makes provision of clean water particularly challenging. The Navajos rely primarily on ground water, which is often present in deep aquifers or of brackish quality. Commonly, reverse osmosis (RO) is chosen to desalinate brackish ground water, since RO costs are competitive with those of thermal desalination, even for seawater applications. However, both conventional thermal distillation and RO are energy intensive, complex processes that discourage decentralized or rural implementation. In addition, both technologies demand technical experience for operation and maintenance, and are susceptible to scaling and fouling unless extensive feed pretreatment is employed. Membrane distillation (MD), driven by vapor pressure gradients, can potentially overcome many of these drawbacks. MD can operate using low-grade, sub-boiling sources of heat and does not require extensive operational experience. This presentation discusses a project on the Navajo Nation, Arizona (Native American tribal lands) that is designed to investigate and deploy an autonomous (off-grid) system to pump and treat brackish groundwater using solar energy. Βench-scale, hollow fiber MD experiment results showed permeate water fluxes from 21 L/m2·d can be achieved with transmembrane temperature differences between 40 and 80˚C. Tests run with various feed salt concentrations indicate that the permeate flux decreases only about 25% as the concentration increases from 0 to 14% (w/w), which is four times seawater salt concentration. The quality of the permeate water remains constant at about 1 mg/L regardless of the changes in the influent salt concentration. A nine-month MD field trial, using hollow fiber membranes and completely offthe-shelf components demonstrated that a scaled-up solar-driven MD system was practical and economically
viable. Based on these results, a pilot scale unit will be constructed and deployed on the tribal lands. Keywords: desalination; membrane distillation; micro porous hydrophobic membrane. *Corresponding author: Vasiliki Karanikola, Department of Chemical and Environmental Engineering, The University of Arizona, 1133 E. James E. Rogers Way, Harshbarger 108, Tucson, AZ 85721, USA, Phone: +858 429-8112, E-mail: [email protected] Andrea F. Corral, Patrick Mette, Hua Jiang, Robert G. Arnoldand and Wendell P. Ela: Department of Chemical and Environmental Engineering, The University of Arizona, 1133 E. James E. Rogers Way, Harshbarger 108, Tucson, AZ 85721, USA
Membrane distillation (MD) utilizes a microporous hydrophobic membrane that is in contact with aqueous solutions at different temperatures and/or compositions (1). The hydrophobic nature of the membrane prevents the passage of liquid water through the pores while allowing the passage of water vapor. The temperature difference produces a vapor pressure gradient which causes water vapor to pass through the membrane and then condenses on a colder surface on the membranes permeate side (Figure 1). The result is a distillate of very high purity (2). The first MD was patented in 1963. At that time, interest in this technology was low due to the growing popularity of reverse osmosis (3). MD interest grew in the early 1980s with the development of improved membranes. The fact that alternative sources of energy, such as solar energy, could drive the process increased the attractiveness of MD. This interest motivated development of different MD configurations that are described later. From the late 1990s to now, interest has steadily increased, especially in MD powered by renewable energies. The MD advantage relies on the fact that a lower temperature and hydrostatic pressure are needed for operation. The membrane acts as a barrier to hold the liquid/ vapor interfaces at the pore entrance. Pore sizes of the membranes used in MD lies between 10 nm and 1 μm (4). To avoid pore wetting, the membrane material must
Brought to you by | New York University Bobst Library Technical Services Authenticated Download Date | 6/4/15 5:05 PM
68 Karanikola et al.: Solar membrane distillation: desalination for the Navajo Nation
Vapor flux
Tf
3. Sweep gas membrane distillation: An inert gas (e.g., air) sweeps the permeate side of the membrane carrying the vapor molecules, and condensation occurs outside the membrane module. 4. Vacuum membrane distillation: The total applied pressure under vacuum on thepermeate side is lower than the saturation pressure of volatile molecules to be separated from the feed solution. As in sweep gas membrane distillation, condensation occurs outside of the membrane module.
Convection
Heat flux Tfm Pf
Tpm
Convection
Tp
Diffusion
Pp
Figure 1 Membrane distillation process (1).
Among the advantages of MD as a separation technology, we can mention that: i. the operating temperature of the MD process is in the range of 60–80°C, which is a temperature level at which thermal solar collectors perform well, ii. it is not necessary to chemically pretreat the feed water, iii. intermittent operation of the module is possible contrary to RO, and iv. system efficiency and high product water quality are virtually independent of the salinity of the feed water (4). However, MD has not yet been widely implemented in the water treatment industry. The main obstacles for the commercial implementation of MD are: i. the relative low permeate flux compared to other separation techniques like RO, ii. permeate flux decay due to temperature and concentration polarization effects, iii. membrane and module design for MD, and iv. high thermal energy consumption: uncertain energy and economic costs for each MD configuration and application (3).
be hydrophobic with high water contact angle and small maximum pore size. Polypropylene, polyethylene, polytetrafluoroethylene, and polyvinylidene fluoride are examples of membrane materials that meet these requirements (1). Because MD membranes are relatively loose and chemically inert (for instance with respect to free chlorine and low pH), their economics and robustness are favorable compared to the reverse osmosis (RO) and nanofiltration membrane alternatives. The MD driving force is the transmembrane vapor pressure difference, which can be established in any of the following reactor configurations (Figure 2) (3). 1. Direct contact membrane distillation: an aqueous solution colder than the feed solution is in direct contact with the permeate side of the membrane. A vapor pressure difference is induced by the transmembrane temperature difference. As a result, water evaporates at the hot liquid/vapor interface, crosses the membrane in vapor phase and condenses at the cold liquid/vapor interface. 2. Air gap membrane distillation: A stagnant air gap is interposed between the membrane and the condensation surface. In this case, the evaporated fluid or solute crosses both the membrane pores and the air gap to finally condense on a cold surface inside the membrane module.
Feed out
Permeate in
Coolant out
Feed out
Air gap
Membrane Membrane
Research objective The objective of this research was to develop a MD configuration for desalination within the constraints that it be ready for immediate field-scale deployment (robust, generic, commercially available components); operate autonomously off the energy grid; and be broadly scalable
Feed in
Sweep gas out Condenser
Membrane
Condensing plate Feed in
Feed in
Permeate out
DCMD
Coolant in
Product
AGMD
Feed in
Condenser
Membrane
Permeate
Feed outt
Sweep gas in
SGMD
Vacuum
Permeate
Feed out
VMD
Figure 2 Different types of MD configurations (3).
Brought to you by | New York University Bobst Library Technical Services Authenticated Download Date | 6/4/15 5:05 PM
Karanikola et al.: Solar membrane distillation: desalination for the Navajo Nation 69
from 50 to 5000 gallon per day (gpd) product and from 50% to 98+% recovery. With this in mind, a bench-scale hybrid, vacuum-driven sweeping gas, MD configuration was constructed using only off-the-shelf components. The challenge was to develop an appropriate system, determine its capabilities and constraints, and transition the setup from benchscale to field pilotscale. Such a system has the advantage that can be deployed in any remote location that faces water scarcity and where the only water available would be water of impaired quality. More specifically this first field scale is going to be deployed in the Navajo Nation. The community has a population density of approximately 4 people/km2 and the only source of water in the area is brackish groundwater. The community currently hauls water from the nearby towns with the closest one at approximately 70 km. The solution proposed will reduce the cost and ease the access of water for the community. The bench-scale vacuum-driven sweeping gas MD experimental setup is schematically shown in Figure 3. It consists of three fluid circulation loops: a brine loop, air loop, and a condenser cooling water loop. In the brine loop, the salty water is heated, circulated past the hydrophobic porous membrane, and returned to the heating tank. In the air loop, air is sucked by a compressor creating a vacuum and acting as a sweep gas to carry the vapor to the condenser. Finally, in the condenser cooling water loop, water flows in counter current flow with the vapor and condenses it as purified water. Due to the vapor pressure difference created on both sides of the membrane between the heated brine and the air, the feed water evaporates through the membrane pores and the air loop sweeps the vapor from the reactor in the air
1
stream. As the air is cooled down to near-ambient temperature, the water vapor moves to the condenser loop where it condenses and is finally collected as pure product. The MD performance was determined as a function of the feed water temperature and salinity. Bench-scale, hollow fiber MD experiments show that permeate water fluxes increased 5 times with small changes in the transmembrane temperature between 40% and 70°C. Tests run with various feed salt concentrations show that the permeate flux decreases about 25% as the concentration increases from 0% to 14% (w/w), which is four times seawater salt concentration. The quality of the permeate water remains constant at about 1 mg/L regardless of the changes in the influent salt concentration. Some representative results can be shown in Figure 4. In addition, similar research has been conducted in Australia and more specifically at the National Centre of Excellency in Desalination Australia, Perth. The objective of the project is to provide clean water to a community of 150 people located 800 km east of Kalgoorlie, called Tjuntjuntjarra. The community is located in a very remote location, with limited access to potable water. The only water that can be used is hypersaline groundwater with Total Dissolved Solids (TDS) of 47,000 mg/L. The system to be deployed is Solar Membrane Distillation (SMD); however, it is operated under a different configuration, vacuum membrane distillation. The system is fully built by MemSYS and the water is heated up by solar PV panels provided by CoGenra (Mountain View, CA, USA). A system schematic can be seen below in Figure 5 that was obtained from the MemSYS (Thermal Separation company based in Singapore and Germany) operational manual for the Vacuum Membrane Distillation (V-MEMD) system. This similar project is an indication of the need for alternative ways of water purification in arid and semiarid locations.
Cooling water
2 4
3
Cold reservoir Vapor
Brine/hot reservoir
5
Compressor
Figure 3 Bench-scale vacuum-driven sweeping gas MD experimental setup.
Permeate molar flowrate, mol/min-m2
Membrane module
Hot brine
Air in 1.0 0.8 0.6 0.4 Brine from 0% to 14% 0.2 0
T from 40˚C to 70˚C 0
0.1
0.2 Vapor pressure, atm
0.3
0.4
Figure 4 Vapor pressure vs. water production for varying temperatures and salt concentrations. The solid line denotes predicted data.
Brought to you by | New York University Bobst Library Technical Services Authenticated Download Date | 6/4/15 5:05 PM
70 Karanikola et al.: Solar membrane distillation: desalination for the Navajo Nation
Memsys V-MEMD unit Vacuum system Droplet catcher
Memsys-module
Heating loop
Steam raiser
Stage 1
Stage 2
...
Stage N Condenser
Cooling loop
Recirculation
Feed
Brine
Distillate
Figure 5 Multistage V-MEMD MemSYS system setup.
Currently, all SMD systems have only been studied in pilot scale studies and positive results (maximum purified water production with less cost compared to previous purification methods) would enhance the possibility for duplication and adaptation of the system in similar
remote locations and communities in need of improvement of quality of life. Received January 16, 2014; accepted January 17, 2014; previously published online February 19, 2014
References 1. Curcio E, Drioli E. Membrane distillation and related operations – A review. Sep Purif Rev 2005;34:35–86. 2. Hogan P, Sudjito A, Fane AG, Momson GL. Desalination by solar heated membrane distillation. Desalination 1991;81: 81–90.
3. El-Bourawi, Ding Z, Maa R. A framework for better understanding membrane distillation separation process. J Membrane Sci 2006;285:4–29. 4. Koschikowski J, Wieghaus M, Rommel M. Solar thermaldriven desalination plants based on membrane distillation. Desalination 2003;156:295–304.
Brought to you by | New York University Bobst Library Technical Services Authenticated Download Date | 6/4/15 5:05 PM
Vasiliki Karanikola*, Andrea F. Corral, Patrick Mette, Hua Jiang, Robert G. Arnoldand and Wendell P. Ela
Solar membrane distillation: desalination for the Navajo Nation Abstract: Provision of clean water is among the most serious, long-term challenges in the world. To an ever increasing degree, sustainable water supply depends on the utilization of water of impaired initial quality. This is particularly true in developing nations and in water-stressed areas such as the American Southwest. One clear example is the Navajo Nation. The reservation covers 27,000 square miles, mainly in northeastern Arizona. Low population density coupled with water scarcity and impairment makes provision of clean water particularly challenging. The Navajos rely primarily on ground water, which is often present in deep aquifers or of brackish quality. Commonly, reverse osmosis (RO) is chosen to desalinate brackish ground water, since RO costs are competitive with those of thermal desalination, even for seawater applications. However, both conventional thermal distillation and RO are energy intensive, complex processes that discourage decentralized or rural implementation. In addition, both technologies demand technical experience for operation and maintenance, and are susceptible to scaling and fouling unless extensive feed pretreatment is employed. Membrane distillation (MD), driven by vapor pressure gradients, can potentially overcome many of these drawbacks. MD can operate using low-grade, sub-boiling sources of heat and does not require extensive operational experience. This presentation discusses a project on the Navajo Nation, Arizona (Native American tribal lands) that is designed to investigate and deploy an autonomous (off-grid) system to pump and treat brackish groundwater using solar energy. Βench-scale, hollow fiber MD experiment results showed permeate water fluxes from 21 L/m2·d can be achieved with transmembrane temperature differences between 40 and 80˚C. Tests run with various feed salt concentrations indicate that the permeate flux decreases only about 25% as the concentration increases from 0 to 14% (w/w), which is four times seawater salt concentration. The quality of the permeate water remains constant at about 1 mg/L regardless of the changes in the influent salt concentration. A nine-month MD field trial, using hollow fiber membranes and completely offthe-shelf components demonstrated that a scaled-up solar-driven MD system was practical and economically
viable. Based on these results, a pilot scale unit will be constructed and deployed on the tribal lands. Keywords: desalination; membrane distillation; micro porous hydrophobic membrane. *Corresponding author: Vasiliki Karanikola, Department of Chemical and Environmental Engineering, The University of Arizona, 1133 E. James E. Rogers Way, Harshbarger 108, Tucson, AZ 85721, USA, Phone: +858 429-8112, E-mail: [email protected] Andrea F. Corral, Patrick Mette, Hua Jiang, Robert G. Arnoldand and Wendell P. Ela: Department of Chemical and Environmental Engineering, The University of Arizona, 1133 E. James E. Rogers Way, Harshbarger 108, Tucson, AZ 85721, USA
Membrane distillation (MD) utilizes a microporous hydrophobic membrane that is in contact with aqueous solutions at different temperatures and/or compositions (1). The hydrophobic nature of the membrane prevents the passage of liquid water through the pores while allowing the passage of water vapor. The temperature difference produces a vapor pressure gradient which causes water vapor to pass through the membrane and then condenses on a colder surface on the membranes permeate side (Figure 1). The result is a distillate of very high purity (2). The first MD was patented in 1963. At that time, interest in this technology was low due to the growing popularity of reverse osmosis (3). MD interest grew in the early 1980s with the development of improved membranes. The fact that alternative sources of energy, such as solar energy, could drive the process increased the attractiveness of MD. This interest motivated development of different MD configurations that are described later. From the late 1990s to now, interest has steadily increased, especially in MD powered by renewable energies. The MD advantage relies on the fact that a lower temperature and hydrostatic pressure are needed for operation. The membrane acts as a barrier to hold the liquid/ vapor interfaces at the pore entrance. Pore sizes of the membranes used in MD lies between 10 nm and 1 μm (4). To avoid pore wetting, the membrane material must
Brought to you by | New York University Bobst Library Technical Services Authenticated Download Date | 6/4/15 5:05 PM
68 Karanikola et al.: Solar membrane distillation: desalination for the Navajo Nation
Vapor flux
Tf
3. Sweep gas membrane distillation: An inert gas (e.g., air) sweeps the permeate side of the membrane carrying the vapor molecules, and condensation occurs outside the membrane module. 4. Vacuum membrane distillation: The total applied pressure under vacuum on thepermeate side is lower than the saturation pressure of volatile molecules to be separated from the feed solution. As in sweep gas membrane distillation, condensation occurs outside of the membrane module.
Convection
Heat flux Tfm Pf
Tpm
Convection
Tp
Diffusion
Pp
Figure 1 Membrane distillation process (1).
Among the advantages of MD as a separation technology, we can mention that: i. the operating temperature of the MD process is in the range of 60–80°C, which is a temperature level at which thermal solar collectors perform well, ii. it is not necessary to chemically pretreat the feed water, iii. intermittent operation of the module is possible contrary to RO, and iv. system efficiency and high product water quality are virtually independent of the salinity of the feed water (4). However, MD has not yet been widely implemented in the water treatment industry. The main obstacles for the commercial implementation of MD are: i. the relative low permeate flux compared to other separation techniques like RO, ii. permeate flux decay due to temperature and concentration polarization effects, iii. membrane and module design for MD, and iv. high thermal energy consumption: uncertain energy and economic costs for each MD configuration and application (3).
be hydrophobic with high water contact angle and small maximum pore size. Polypropylene, polyethylene, polytetrafluoroethylene, and polyvinylidene fluoride are examples of membrane materials that meet these requirements (1). Because MD membranes are relatively loose and chemically inert (for instance with respect to free chlorine and low pH), their economics and robustness are favorable compared to the reverse osmosis (RO) and nanofiltration membrane alternatives. The MD driving force is the transmembrane vapor pressure difference, which can be established in any of the following reactor configurations (Figure 2) (3). 1. Direct contact membrane distillation: an aqueous solution colder than the feed solution is in direct contact with the permeate side of the membrane. A vapor pressure difference is induced by the transmembrane temperature difference. As a result, water evaporates at the hot liquid/vapor interface, crosses the membrane in vapor phase and condenses at the cold liquid/vapor interface. 2. Air gap membrane distillation: A stagnant air gap is interposed between the membrane and the condensation surface. In this case, the evaporated fluid or solute crosses both the membrane pores and the air gap to finally condense on a cold surface inside the membrane module.
Feed out
Permeate in
Coolant out
Feed out
Air gap
Membrane Membrane
Research objective The objective of this research was to develop a MD configuration for desalination within the constraints that it be ready for immediate field-scale deployment (robust, generic, commercially available components); operate autonomously off the energy grid; and be broadly scalable
Feed in
Sweep gas out Condenser
Membrane
Condensing plate Feed in
Feed in
Permeate out
DCMD
Coolant in
Product
AGMD
Feed in
Condenser
Membrane
Permeate
Feed outt
Sweep gas in
SGMD
Vacuum
Permeate
Feed out
VMD
Figure 2 Different types of MD configurations (3).
Brought to you by | New York University Bobst Library Technical Services Authenticated Download Date | 6/4/15 5:05 PM
Karanikola et al.: Solar membrane distillation: desalination for the Navajo Nation 69
from 50 to 5000 gallon per day (gpd) product and from 50% to 98+% recovery. With this in mind, a bench-scale hybrid, vacuum-driven sweeping gas, MD configuration was constructed using only off-the-shelf components. The challenge was to develop an appropriate system, determine its capabilities and constraints, and transition the setup from benchscale to field pilotscale. Such a system has the advantage that can be deployed in any remote location that faces water scarcity and where the only water available would be water of impaired quality. More specifically this first field scale is going to be deployed in the Navajo Nation. The community has a population density of approximately 4 people/km2 and the only source of water in the area is brackish groundwater. The community currently hauls water from the nearby towns with the closest one at approximately 70 km. The solution proposed will reduce the cost and ease the access of water for the community. The bench-scale vacuum-driven sweeping gas MD experimental setup is schematically shown in Figure 3. It consists of three fluid circulation loops: a brine loop, air loop, and a condenser cooling water loop. In the brine loop, the salty water is heated, circulated past the hydrophobic porous membrane, and returned to the heating tank. In the air loop, air is sucked by a compressor creating a vacuum and acting as a sweep gas to carry the vapor to the condenser. Finally, in the condenser cooling water loop, water flows in counter current flow with the vapor and condenses it as purified water. Due to the vapor pressure difference created on both sides of the membrane between the heated brine and the air, the feed water evaporates through the membrane pores and the air loop sweeps the vapor from the reactor in the air
1
stream. As the air is cooled down to near-ambient temperature, the water vapor moves to the condenser loop where it condenses and is finally collected as pure product. The MD performance was determined as a function of the feed water temperature and salinity. Bench-scale, hollow fiber MD experiments show that permeate water fluxes increased 5 times with small changes in the transmembrane temperature between 40% and 70°C. Tests run with various feed salt concentrations show that the permeate flux decreases about 25% as the concentration increases from 0% to 14% (w/w), which is four times seawater salt concentration. The quality of the permeate water remains constant at about 1 mg/L regardless of the changes in the influent salt concentration. Some representative results can be shown in Figure 4. In addition, similar research has been conducted in Australia and more specifically at the National Centre of Excellency in Desalination Australia, Perth. The objective of the project is to provide clean water to a community of 150 people located 800 km east of Kalgoorlie, called Tjuntjuntjarra. The community is located in a very remote location, with limited access to potable water. The only water that can be used is hypersaline groundwater with Total Dissolved Solids (TDS) of 47,000 mg/L. The system to be deployed is Solar Membrane Distillation (SMD); however, it is operated under a different configuration, vacuum membrane distillation. The system is fully built by MemSYS and the water is heated up by solar PV panels provided by CoGenra (Mountain View, CA, USA). A system schematic can be seen below in Figure 5 that was obtained from the MemSYS (Thermal Separation company based in Singapore and Germany) operational manual for the Vacuum Membrane Distillation (V-MEMD) system. This similar project is an indication of the need for alternative ways of water purification in arid and semiarid locations.
Cooling water
2 4
3
Cold reservoir Vapor
Brine/hot reservoir
5
Compressor
Figure 3 Bench-scale vacuum-driven sweeping gas MD experimental setup.
Permeate molar flowrate, mol/min-m2
Membrane module
Hot brine
Air in 1.0 0.8 0.6 0.4 Brine from 0% to 14% 0.2 0
T from 40˚C to 70˚C 0
0.1
0.2 Vapor pressure, atm
0.3
0.4
Figure 4 Vapor pressure vs. water production for varying temperatures and salt concentrations. The solid line denotes predicted data.
Brought to you by | New York University Bobst Library Technical Services Authenticated Download Date | 6/4/15 5:05 PM
70 Karanikola et al.: Solar membrane distillation: desalination for the Navajo Nation
Memsys V-MEMD unit Vacuum system Droplet catcher
Memsys-module
Heating loop
Steam raiser
Stage 1
Stage 2
...
Stage N Condenser
Cooling loop
Recirculation
Feed
Brine
Distillate
Figure 5 Multistage V-MEMD MemSYS system setup.
Currently, all SMD systems have only been studied in pilot scale studies and positive results (maximum purified water production with less cost compared to previous purification methods) would enhance the possibility for duplication and adaptation of the system in similar
remote locations and communities in need of improvement of quality of life. Received January 16, 2014; accepted January 17, 2014; previously published online February 19, 2014
References 1. Curcio E, Drioli E. Membrane distillation and related operations – A review. Sep Purif Rev 2005;34:35–86. 2. Hogan P, Sudjito A, Fane AG, Momson GL. Desalination by solar heated membrane distillation. Desalination 1991;81: 81–90.
3. El-Bourawi, Ding Z, Maa R. A framework for better understanding membrane distillation separation process. J Membrane Sci 2006;285:4–29. 4. Koschikowski J, Wieghaus M, Rommel M. Solar thermaldriven desalination plants based on membrane distillation. Desalination 2003;156:295–304.
Brought to you by | New York University Bobst Library Technical Services Authenticated Download Date | 6/4/15 5:05 PM
