Innovative Concentrate Treatment And Recovery Technologies
Innovative Concentrate Treatment And Recovery Technologies Ufuk G. Erdal, Ph.D., PE CH2M HILL [email protected] Raed A. Bkayrat, Ph.D., PE KAUST [email protected]
Reverse osmosis (RO) and electrodialysis reversal (EDR) are widely used technologies for TDS removal from brackish waters and more recently from wastewater in water reuse projects. While both processes are highly effective for removing TDS from water, they generate a concentrate stream which requires proper disposal. Disposal of the concentrate streams is economically and environmentally a big challenge. The challenge gets even bigger in inland areas where surface discharge is not possible. Feasibility of implementing ultimate disposal approaches (i.e., surface discharge, deep well injection, solar evaporation coupled with landfill, etc.) is site/location specific. Often, a concentrate volume reduction technology is used prior to ultimate disposal to reduce project costs. Numerous research studies have been conducted to develop low carbon foot-print technologies and technologies that increase recovery, use renewable energy and low grade waste heat to desalinate wastewater. The majority of these technologies have been developed for brackish water or seawater desalination; the same technologies can also be used for concentrate volume reduction technology. This article highlights the innovative/developmental technologies, presents state of development, and what improvements need to be done for commercial success. The innovative/developmental technologies reviewed in this article included: • • • • •
•
Adsorption/desorption desalination Capacitive Deionization Dewvaporation Forward osmosis High Recovery RO Processes o ARROW o OPUS o SPARRO o Zero Discharge Desalination Membrane Distillation
Review of the Technologies Silica Based Adsorption/Desorption Desalination Technology Silica gel based adsorption technology was developed in National University of Singapore (NUS) to reduce scaling and corrosion issues associated with thermal processes and to lower the carbon footprint of desalination using solar energy or waste heat, and to capture the generated cool air for district cooling. The technology was originally developed for seawater desalination but it can also be applied for brackish water desalination. Currently, NUS and King Abdullah University of Science and Technology (KAUST) are collaborating to optimize system performance and evaluate fouling/scaling behavior of the technology in a demonstration unit (35 kW) treating seawater. A picture of the demo unit at KAUST is presented in Figure1.
Figure 1. Picture of a Demo Unit at KAUST (Adapted from CH2MHILL, 2011) The silica gel adsorption desalination and cooling unit comprises a host of stationary units, namely an evaporator, condenser and adsorber/desorber beds. After de-aeration, saline or brackish water is fed intermittently into the evaporator where desalting is achieved at low system pressure (1–5 kPa) and temperatures (5-10oC) which mitigates the problems associated with scaling and corrosion. During adsorption processes, water vapor is adsorbed by the silica gel due to its high affinity to water at low temperatures. Concomitantly, a heat source such as hot water (typically from 50 to 80oC) is supplied to the desorption beds, containing the saturated silica gel from the previous cycle, to expel the water vapor from the adsorbent. The desorbed vapor condenses on the cooler surfaces of condenser, cooled by recirculating coolant from a seawater cooling tower. Highly purified potable water is produced at low temperatures (50-80oC), which enable the technology to have a cooling application. The heat to drive desorption can be obtained from waste heat as well as renewable sources including solar and geothermal. Potential Advantages over Established Technologies: • Very low specific energy cost. (It is reported to be 1.38 kWh/m3 which is very close to the lowest theoretical amount (1.0 kWh/m3) achieved by any desalination system (Thu et.al. 2010) • Higher water recoveries than RO • Capability to use solar energy results in great flexibility to locate the AD facilities • Does not require pretreatment and post treatment chemicals • Low temperature operation minimizes scaling and corrosion problems Potential Disadvantages Compared to Established Technologies: • •
Requires large space (Use of tall silos can reduce the foot-print but increases the capital cost of the facility) Availability of low temperature waste heat sources is important but finding these sources may be a challenge for siting large-scale AD facilities
Currently, Advon Singapore Pte Ltd. is providing demo units. A four bed small scale demonstration unit has been successfully operated in Singapore for more than 6 years (Wang and Ng. 2005). The electricity consumption and low O&M requirement are the two major attractive features of this technology. In addition, this technology can be retrofitted into an existing MED plant to increase
efficiency of the system as reflected in gain-output-ratio (GOR) value. This technology is potentially best fit to locations where there is need for desalination, demand for cooling and waste heat sources are available. Technology has already been commercialized and large demonstration projects are underway.
Capacitive Deionization (CD) The original CD process was developed and patented at Lawrence Livermore National Laboratory in the late 1980s. CD is a low-pressure, non-membrane desalination technology that uses electrostatic forces to remove dissolved ions from solution. An aqueous solution of soluble salts is passed through pairs of electrodes held at a potential difference of 1.2 volts (V). The electrodes consist of porous carbon aerogel, with a specific surface area of 400 to 1,100 m2/g and with very low electrical resistivity (less than 40 kilohm-meters). Ions are adsorbed to the electrode of opposing charge in a semi-batch process. Eventually, the electrodes become saturated with ions and must be regenerated. To regenerate CD, the applied potential is removed, and the ions attached to the electrodes are released and flushed from the system. Flushing of cells using a small quantity of product water generates a concentrate stream, as shown in Figure 2. Unlike ion exchange processes, no additional chemicals are required for regeneration of the electrosorbent in this system.
Figure 2. Schematic Illustration of CD Operation (Top) and Regeneration (Bottom) (Adapted from CH2MHILL, 2011) Carbon aerogel is an ideal electrode material because of its high electrical conductivity, high specific surface area, and controllable pore size distribution. However, despite recent advancements, carbon aerogel electrodes are still expensive and their ion storage capacity is relatively low. Original designs of CD systems were limited to the treatment of relatively low ionic strength solutions (TDS< 3,000 mg/L). The reason for their limited application has been identified as the high pore volume to surface area characteristic of the carbon electrode material (Seed et al., 2006). The technology has been investigated in several academic and research institutions. In the USA, Colorado School of Mines researchers concluded that this technology could yield results comparable to those of conventional RO for desalination of streams with TDS of less than 3,000 mg/L (Kristen, 2006). This is mainly due to the high cost of CD modules with increased feed water TDS concentrations.
Capacitive Deionization Technology Systems, Inc. (CDT Inc.) is the worldwide licensee for the patented CD technology for water purification and desalination applications. Independently, ENPAR Technologies Inc. of Canada has developed its DesEL System, which uses principles of CD to remove TDS. Potential Advantages over Established Technologies: • •
• • •
CD requires less energy than EDR, RO, and mechanical/thermal processes Membrane technologies require more advanced operation and construction considerations for high-pressure pumps and clean-in-place (CIP) systems. The CD operates at ambient conditions, and there are no requirements for high-pressure pumps and CIP systems CD uses electrostatic regeneration and requires minimal or no chemicals for electrode fouling and scaling controls Less scaling propensity compared to RO Silica does not limit the recovery compared to RO
Potential Disadvantages Compared to Established Technologies: •
•
• •
•
CD is still under development. Knowledge about treatment efficiencies of larger-scale installations, economics, and short- and long-term fouling/scaling issues of CD systems has not been established. Lower TDS removal than RO and mechanical/thermal processes. The process cannot remove uncharged molecules (such as boron, silica, and non-polar organic compounds. Solids and pathogens). CD recovers lower amounts of water than conventional membrane processes Adsorption of total organic carbon (TOC) to the aerogel material during regeneration when the cell is uncharged could result in electrode fouling if organic matter clogs the pores of carbon aerogel material Lengthy down-time period during cleaning of electrodes
Currently, there is no known full-scale CD facility. In a bench-scale study, Seed et al. (2006) showed that CD can achieve 80 percent TDS removal with high water recoveries (up to 95 percent). However, the feed water TDS was only 460 mg/L during bench-scale testing. Tao et al. (2009) used a pilot-scale CD unit to treat and recover an RO concentrate stream generated from the Kranji NeWater Reuse RO facility in Singapore. Biological activated carbon pretreatment followed by CD resulted in 85 percent TDS removal efficiency at approximately 85 percent recovery. Tao et al. (2009) reported a cell energy consumption of 0.7 KWh/m3 based on the pilot CD testing results. Similar to the previous study, the feed TDS concentration of CD was relatively low (1,250 mg/L). This covers only lower range of TDS found in brackish water across USA. A comprehensive evaluation performed by the Colorado School of Mines (USBR, 2009) concluded that the energy consumption of CD is similar to that of RO (4 kWh/1,000 gallons, or 1.1 kWh/m3) for low flow systems (0.7 mL/min). The efficiency and recovery of the process need to be improved before CD becomes economically feasible to treat relatively high TDS water (>5,000 mg/L). CD could probably be implemented within the next 5 to 10 years, if the following improvements are made (USBR, 2009): • •
Higher capacitance and lower cost of electrode materials Recovery of residual electricity
• • •
Shorter regeneration time to reduce process down-time Reduced carryover volume after regeneration Cost-effective and low carbon footprint pretreatment technologies
Dewvaporation Dewvaporation uses a humidification-dehumidification cycle to produce distilled product water. Feed water is evaporated by heated air, and freshwater is condensed on the opposite side of a heat transfer wall. Each Dewvaporation tower contains a heat transfer wall made of plastic. The wall divides the module into two compartments, one for evaporation and one for dew formation. The energy needed for evaporation is partially supplied by the recovered energy released during condensation. Heat sources can be combustible fuel, solar, or low-grade heat from various sources. Using waste heat or low-grade heat can reduce O&M costs significantly, thereby making it a very attractive desalination technology. The tower unit is built of thin plastic films to avoid corrosion and to minimize equipment costs. Tower construction is relatively inexpensive because the towers operate at atmospheric pressure. The Dewvaporation concept was developed at Arizona State University, USA, in conjunction L'Eau LLC, the company that owns the patent rights to the process. The process has been marketed under license by Altela Inc. since mid-2006. Altela, Inc. has designed, manufactured, and tested several AltelaRainTM prototype systems based on the Dewvaporation process. A schematic of the AltelaRainTM process is shown in Figure 3.
Figure 2. AltelaRainTM Dewvaporation Process Schematic
Potential Advantages over Established Technologies: • Can treat very high TDS-containing streams (up to 60,000 mg/L) • Use of solar or waste heat significantly reduces operating cost • Operates in atmospheric pressure, so there are no requirements for high-pressure pumps and expensive sturdy towers • Relatively high recovery (such as 90 percent, based on pilot tests conducted in New Mexico, USA) compared to NF/RO and EDR for treating high-TDS streams • Less complex than mechanical/thermal evaporative processes • Water quality has little impact on process performance • Plastic heat transfer walls reduce capital cost and eliminate corrosion concerns • Less fouling/scaling propensity compared to NF/RO • Silica does not limit the recovery compared to NF/RO • No chemical usage for pretreatment of feed water Potential Disadvantages Compared to Established Technologies: • Dewvaporation is still under development. Knowledge about the following has not been established: treatment efficiencies of larger-scale installations, economics, and short- and long-term performance of the system. • Like other evaporative processes, the energy consumption of the dewvaporation system is high • As in most desalination systems, post-treatment is required for stabilization and mineralization of the water A 5,000-gpd Dewvaporation pilot plant was operated at the 23rd Avenue Wastewater Treatment Plant (WWTP) in Phoenix, Arizona. The pilot plant feed was concentrate from a Tactical Water Purification System (TWPS) RO unit with ultrafiltration pretreatment. A 2,000-mg/L TDS wastewater RO concentrate stream was treated by the pilot plant to more than 45,000-mg/L TDS brine and 10 mg/L TDS distillate, yielding a recovery of up to 95 percent and salt rejection of more than 99 percent. Thermal multiple effects varied from 2.0 to 3.5, which was less than the 5.0 effects demonstrated prior to transport to the WWTP site. Three full-scale AltelaRain ARS-4000 systems were operated at natural gas wells in the San Juan basin near Farmington, New Mexico, USA (Colorado School of Mines, 2009). The ARS-4000 system processed approximately 4,000 gpd of produced water. The AltelaRainTM System produced distilled water with TDS of approximately 100 mg/L while processing a waste stream containing approximately 42,000 mg/L TDS. One unit reduced effluent disposal volumes by as much as 90 percent (Colorado School of Mines, 2009). This technology has shown to be effective for producing high-quality water from high-TDS streams; it can be used for treating and recovering RO/EDR concentrate/reject streams. Commercial units are already available and can be used in small applications. For large-scale applications, custom design is essential to reduce the capital investment. Project locations where waste heat or low-grade heat is available (such as refineries) are favorable locations for implementation of this technology.
Forward Osmosis (FO) A 2010 desalination market survey completed by Global Water Intelligence (GWI) notes that “three areas of new technology are likely to be the focus of commercialization over the next four years” forward osmosis was one of them. FO is an innovative membrane-based technology that has the potential to reduce the costs and environmental impacts of desalination. This is an osmotic process that uses a semi-permeable membrane to separate salts from water. FO uses an osmotic pressure gradient
(∆п) instead of hydraulic pressure (∆P), which is used in RO, to create the driving force for water transport through the membrane. No energy is needed to drive the water flux of an FO process, as the water flux is the natural tendency of the system. The concentrated solution, or draw solution, is the source of the driving force in the FO process. A selective permeable membrane allows passage of water, but rejects solute molecules and ions. Osmotic driving forces in FO can be significantly greater than hydraulic driving forces in RO. This results in the potential for higher water flux rates and recoveries. The selection of an appropriate draw solution is the key to FO performance. Example draw solutions include magnesium chloride, calcium chloride, sodium chloride, potassium chloride, ammonium carbonate, and sucrose. A simplified process schematic of an FO process using ammonium carbonate as a draw solution is presented in Figure 3.
Figure 3. Process Schematic of FO (Adapted from McCutcheon et al., 2007) Potential Advantages over Established Technologies: • • •
Operates around 1 atm, which results in much lower energy consumption compared to conventional membrane and mechanical/thermal evaporative desalination technologies Membrane compaction is not typically an issue Less fouling propensity compared to RO
Potential Disadvantages Compared to Established Technologies: •
•
•
Still under development. Knowledge regarding treatment efficiencies of larger-scale installations, economics, and short- and long-term performance and fouling/scaling has not been established Requires special membranes. Existing commercially available RO membranes are not suitable for FO because such membranes have a relatively low product water flux, which can be attributed to severe internal concentration polarization in the porous support and fabric layers of RO membranes Use of ammonium carbonate as draw solution may provide desired osmotic pressure. However, ammonia diffuses back to the permeate stream and it should be removed using a cost effective technology (such as waste heat to strip ammonia)
Various approaches for FO-based desalination exist. In the USA, a pilot-scale FO unit was built and has been operated at the Yale University laboratory since 2005. The Yale pilot study utilizes an ammonium carbonate solution as the draw solution. To recover freshwater, the diluted ammonium carbonate solution is heated to approximately 55oC, where ammonium carbonate undergoes thermal decomposition. The Yale research is being commercialized by Oasys Water Inc. Modern Water PLC has constructed a 25,000 gpd pilot facility in Gibraltar for sea water as a desalination demonstration (Water Desalination Report, 2008). Additional facilities located in Oman are being developed by Modern Water. Several researchers have investigated the use of FO as a potential energy source using pressure restrained osmosis (PRO). Loeb (1998) first demonstrated the feasibility of energy production by PRO. In 2009, Statkraft, the national energy company of Norway, commissioned a demonstration PRO power plant in Tofte, Norway. With 2-kW capacity, the facility is a proof of concept test bed. More recently, membrane manufacturer HTI Water, of Arizona (USA), launched large-scale commercialization of FO as a low-energy means of treating waste flows such as hydraulic fracturing fluids. Modern Water PLC has already established three installations using “manipulated osmosis” as a desalination process. The following advancements are needed for consideration of this technology in full-scale applications. • • • •
Identifying effective and economical draw solutions and technologies/approaches that remove draw solutions economically (such as using waste/low-grade heat). Developing new and additional membrane sources. Currently, a limited number of commercially available membranes are on the market using cellulose triacetate. Addressing mass transfer limitations resulting from concentration polarization within the membrane support layer. Developing new modules suitable for full-scale implementation. To date, most applications have used flat-sheet, plate and frame elements.
High Recovery RO Processes The major issue for operating an RO process at higher recoveries is the precipitation of sparingly soluble inorganic salts, most notably barium sulfate (BaSO4), calcium carbonate (CaCO3), calcium phosphate (Ca3(PO4)2), calcium sulfate (CaSO4) and silica. Inorganic salt precipitation can be controlled at lower recoveries by using an appropriate antiscalant and by acidifying feed water pH (effective for CaCO3 and (Ca3(PO4)2 control). At higher recoveries (greater than 85 percent), the concentration of sparingly soluble salts can exceed the effective range of antiscalants, and pH control does not prevent precipitation of some problematic minerals such as silica, BaSO4 and CaSO4 (that may not be as effectively removed by chemical cleaning). To improve RO recovery, one or more components of these scale-forming salts need to be lowered. High-recovery RO processes were developed to alter water chemistry prior to RO to allow higher water recoveries. These processes are applicable for treating RO/EDR concentrate streams with very high recoveries, in some cases approaching zero liquid discharge (ZLD). High-recovery RO/concentrate recovery processes include Advanced Reject Recovery of Water, OPUSTM, SPARRO, and ZDDTM.
a. Advanced Reject Recovery of Water (ARROWTM) ARROWTM is a high-recovery, advanced membrane system that couples a softening process with RO to increase water recovery. This is a proprietary technology marketed by Advanced Water Solutions and O’Brien & Gere. ARROWTM has a number of configurations that can be adjusted depending on the flow rate, hardness, concentration of silica relative to other hardness precursors, and TDS concentration. The ARROW process is illustrated in Figure 4 and includes the following steps: Pretreatment - Pretreatment typically includes acid and antiscalant addition
First-Stage RO - ARROWTM produces a permeate stream of 60 to 75 percent of the flow, while 40 to 25 percent of the stream is RO concentrate. Second-Stage RO - Concentrate from the first-stage RO is treated and combined with an appropriate flow of recycled stream from the second-stage RO concentrate. Softening of RO Concentrate Stream from Second-Stage RO - ARROW uses either chemical precipitation to reduce calcium, magnesium, and silica hardness or ion exchange (IX) softening containing strongly acidic cation exchange resins, if silica hardness is not a concern. Chemical precipitation uses caustic soda or soda ash depending upon the ratio of alkalinity to calcium hardness. Recovery - A small amount of flow from second–stage RO concentrate and a small reject stream from the bottom of the clarifier or from the IX system is sent to a solar evaporator or thermal crystallizer. The combined volume of the two reject streams is less than 5 percent, giving an overall process recovery of greater than 95 percent.
Figure 4. ARROWTM Process Schematic Potential Advantages over Established Technologies:
• • • •
High-quality product water compared to CD and EDR Applicable for treating high silica content streams Pretreatment reduces fouling and scaling precursors, resulting in much higher RO system recovery rates (e.g., 90 percent) compared to EDR, CD, and conventional membrane-based systems for treating high-TDS water. Pretreatment also reduces feed pressure and energy requirements. Compact skid-mounted system, which reduces equipment delivery and installation time
Potential Disadvantages Compared to Established Technologies: •
Process is still under development; no full-scale applications exist in municipal water or wastewater treatment
• • • • •
High cost of chemicals used for pretreatment and softening of water Requires skilled operation Available from only a single supplier and so does not allow for competitive bidding Pilot testing may be required to determine key design criteria Sludge from softening might require separate disposal, creating additional challenge and expense
No full-scale operation. There is one demonstration project in the industrial water treatment field. This project is a 33-gallon-per-minute (gpm) unit in New Jersey, USA. Knowledge about treatment efficiencies, economics (including sludge disposal and concentrate disposal costs), and long-term system performance is not established. Existing skid-mounted units, which are easy to implement, are applicable only to very small systems (up to 0.25 mgd) Potential applications are currently limited to small industrial treatment and small brine/concentrate recovery applications. Process performance and robustness under varying feed water quality conditions should be proven, and implementation costs should be comparable to those of other high recovery technologies for this technology to be considered for wider size and purpose applications. b. OPUSTM N.A. Water Systems, Veolia Water Solutions and Technology, France, designed the OPUS™ system. OPUS™ is a proprietary optimized pretreatment which incorporates unique separation processes for desalination of water with high concentrations of sparingly soluble solutes (e.g., silica, CaSO4, and Mg(OH)2), organics, and boron. The system is able to achieve high recovery with high purity product water through the use of extensive pretreatment processes prior to water being processed through IX and RO subsystems (RPSEA, 2009). A process schematic is shown in Figure 5.
Figure 5. OPUSTM Process Schematic (Adapted from CH2M HILL, 2011)
The first step of the process includes acidification and degasification of the raw feed water. This is followed by a conventional coagulation, flocculation, and high-rate plate settler sedimentation process, which is termed Multiflo™. After this step, the flow stream should be devoid of nearly all highmolecular-weight organic molecules and oxidized metals (particularly iron and manganese). Additionally, colloidal silica is partially removed by co-precipitation. Decant from the sedimentation basin is then filtered by a packed-bed media filtration column, which removes any microflocs and most
suspended solids that pass through the plate settlers. The media filter may also achieve additional removal of low to medium molecular weight hydrophobic organic molecules, including oil and grease (Colorado School of Mines, 2009). Filtrate from the media filter is then processed through a mixed, packed-bed IX column for further water softening. A cartridge filter is then used to remove any IX resin or remaining particulate material prior to RO. The water is then pressurized and treated by brackish water RO (BWRO) membranes at an elevated pH. Operating the RO elements under these conditions reduces the fouling propensity of silica and increases the rejection of both silica and boron. Potential Advantages over Established Technologies: • Could produce very high quality water (product literature reports greater than 99 percent rejection of TDS and most multivalent solutes and achievement of additional silica and boron removal with high pH operation) • Pretreatment reduces fouling and scaling precursors, resulting in much higher RO system recovery rates (e.g., 90 percent) compared to EDR, CD, and conventional membrane-based systems for treating high-TDS water. Pretreatment also reduces feed pressure and energy requirements. Potential Disadvantages Compared to Established Technologies: • OPUSTM is still under development. Knowledge about the following has not been established: treatment efficiencies of larger-scale installations, economics, and short- and long-term performance of the system. • Despite the high water recoveries, OPUSTM generates multiple waste streams and a concentrate stream that needs to be disposed off or treated; sludge from the sedimentation basin requires dewatering and landfill application. • Larger footprint than conventional RO or IX systems due to inclusion of pretreatment, chemical feed and storage, and dewatering facilities (if included). • Complex operation. Requires skilled labor and sophisticated process automation. • Available from only a single supplier and so does not allow for competitive bidding • Multiple chemicals to handle, including acids, bases, hydrolyzing metal coagulants, and polymerbased coagulants. • Available from only a single supplier and so does not allow for competitive bidding Currently, there are no full-scale applications. The OPUS™ was field tested at a steam-enhanced oil production field in San Ardo, California, USA. Field trials have demonstrated that this system can treat feed water with TDS levels up to 10,000 mg/L. Veolia Water Solutions Technologies claims that it could treat streams with TDS levels up to 30,000 mg/L. The treatment process permeate quality is dependent on feed water salinity and operating conditions. However, product literature reports greater than 99 percent rejection of TDS and most multivalent solutes (Colorado School of Mines, 2009). The technology is still under development by Veolia. The technology has not been pilot-tested extensively and requires third-party independent evaluation to establish/prove the following: • Life cycle costs of the treatment • Robustness of the process under varying feed water and operating conditions • Design and operating parameters to optimize process performance and to lower chemical usage
c. Slurry Precipitation and Reverse Osmosis (SPARRO)
SPARRO involves circulating a slurry of seed crystals within the RO system, which serve as preferential growth sites for calcium sulfate and other calcium salts and silicates. The scaling compounds are precipitated on the seed crystals instead of on the membrane. This process is confined to the use of tubular membranes due to the need to circulate the slurry within the membranes without plugging them as presented in Figure 6. The water to be desalted is mixed with a stream of recycled concentrate containing the seed crystals and fed to the RO process. The concentrate with seed crystals is processed in a cyclone separator to separate the crystals, and the desired seed concentration is maintained in a reactor tank by controlling the rate of wasting the upflow and/or underflow streams from the separator. The combined recovery of the process is estimated to be greater than 90 percent (CH2M HILL, 2009).
Figure 6. Illustration of Tubular Membranes Used in SPARRO (Adapted from CH2M HILL, 2009) Potential Advantages over Established Technologies: • Low energy requirement compared to RO and thermal/evaporative desalination processes • Higher recoveries compared to EDR and RO, especially treating high scaling potential waters (industrial, RO concentrate, etc.) Potential Disadvantages Compared to Established Technologies: • Still under development. Knowledge about the following has not been established: treatment efficiencies of larger-scale installations, economics, and short- and long-term performance of the system. • Larger footprint; requires large reaction tanks and large area for membrane systems due to lower specific surface area of membranes • Tubular membranes have lower TDS rejections (TOC and TN rejections may also be lower than with traditional RO membranes) • Additional chemicals to handle • Relatively complex operation
Currently there are no full-scale applications. This technology has been tested at a pilot scale on several applications, including treating mining wastewater in South Africa, and treating primary and secondary RO concentrate streams from the Eastern Municipal Water District (EMWD), California, USA, zero liquid discharge (ZLD) pilot project (USBR, 2008). The combined recovery of the process was greater than 90 percent according to the study performed in South Africa (CH2M HILL, 2009). The results from the EMWD study indicate 60 to 70 percent recovery. Although this recovery seems moderate, it is in fact remarkably high considering the water quality matrix which prohibits use of RO/EDR without extensive pretreatment (softening and IX). Knowledge about treatment efficiencies, economics, and long-term performance of the system has not been established. Process performance and robustness under varying feed water quality conditions should be proven and cost should be comparable to those of other high recovery technologies before this technology could be considered in mid- to large-scale applications.
d. Zero Discharge Desalination (ZDD™) Zero Discharge Desalination (ZDD™), marketed by Veolia Water Solutions and Technology, France, was developed to reduce concentrate volumes generated from the EDR/RO facilities. In ZDDTM, the concentrate stream from a conventional RO system is fed to an electrodialysis metathesis (EDM) stack consisting of ion exchange membranes and thin solution compartments. A direct electric potential is applied to the ends of the stack, resulting in a direct current that is carried by ions migrating through the membranes and solution compartments. The DC potential pushes ions through membranes from a lower concentration to a more concentrated solution. Water flows tangentially to the membrane, while the flow of ions is perpendicular to the membrane (Biagini et al., 2010). An electrodialysis stack consists of anion membranes containing many fixed positive charges, usually quaternary amines, that are loosely associated with mobile negatively charged ions that permeate the membrane, and cation membranes with fixed negative charges, usually sulfonic acids, that allow positively charged ions to permeate. In the ZDD technology, the calcium salts that cause scaling in a high recovery RO process are removed in an EDM process that is a variant of ordinary EDR. The concentrate recovery mode of operation assumes that the RO concentrate from an existing RO system would be treated by the EDM process to produce highly concentrated salt streams while reducing the salt concentrations and the scaling tendency of this RO concentrate. If necessary, a silica removal system can be incorporated as a slip stream treatment, as presented in the process flow schematic shown in Figure 7. From an energy perspective, it is not practical to reduce the salt concentration to potable levels by the EDM process, so RO would be used to further reduce the TDS of the EDM treated water. Potential Advantages over Established Technologies: • Potentially very high recovery (97 percent recovery, including dewatering) was reported in Brackish Groundwater National Desalination Research Facility in Alamogordo, New Mexico (USA) • Less waste to handle compared to IX-RO or lime softening-RO-based processes • Requires less space than lime softening RO Potential Disadvantages Compared to Established Technologies: • ZDDTM is still under development. Knowledge about the following has not been established: treatment efficiencies of larger-scale installations, economics, and short- and long-term performance of the system.
•
Crystallization of sodium bicarbonate in a sodium chloride solution limits the possibility of increasing the water recovery in waters that contain substantial quantities of chloride and alkalinity
Figure 7. ZDD Process Schematic (Adapted from CH2M HILL, 2011) Currently there are no full-scale applications. A large-scale pilot test was conducted at the Brackish Groundwater National Desalination Research Facility in Alamogordo, New Mexico, USA. The pilot testing results indicated that a recovery of 94 percent was achievable without dewatering (97 percent with dewatering). The cost projections performed by Veolia Water Systems indicated much lower capital and O&M costs compared to the lime softening-RO systems. The technology is still under development by Veolia. The technology has not been pilot-tested extensively and requires third-party independent evaluation to establish/prove the following: • • •
Life cycle costs of the treatment including solids disposal Robustness of the process under varying feed water and operating conditions Design and operating parameters to optimize process performance
Project costs should be comparable to those of other high recovery technologies before this technology could be considered for mid- to large-scale applications.
Membrane Distillation (MD) Membrane distillation (MD) is another technology that GWI identified combines membrane technology and evaporation processing in a single unit. The appeal of MD is that it functions at atmospheric pressure and requires only relatively low feed temperatures of 70ºto 90ºC. MD transports water vapor through the pores of hydrophobic membranes using the temperature difference across the membrane. The membrane allows water vapor to penetrate the hydrophobic surface while repelling the liquid. The clean vapor is carried away from the membrane and condensed as pure water, either within the membrane package or in a separate condenser system.
MD differs from other membrane technologies in that the driving force that pushes the water through the membrane is not feed pressure but temperature. In MD units, vapor production is enhanced by heating the feed water, which increases the vapor pressure and penetration rate. MD requires the same amount of energy input to heat and condense vapor as traditional evaporation; however, it does not require boiling water and is operated at ambient pressure. The energy requirement for MD is lower than those of conventional evaporation systems. MD is most efficient on low-grade or waste heat, such as industrial heat streams or even solar energy (Huehmer and Wang, 2009). Also, the efficiency of the unit can be improved with heat recovery. MD membranes must be microporous (pore diameters of 0.05 to 0.2 μm) and non-wettable by the feed. A variety of arrangements and configurations can be used to induce the vapor through the membrane and to condense penetrant gas; however, the feed water must always be in direct contact with the membrane. Condensation is typically achieved via two major process configurations (Salamero, 2004):
•
Direct-Contact Membrane Distillation: The cool condensing solution directly contacts the membrane and flows countercurrent to the raw water. This is the simplest configuration and is best suited for applications such as desalination and concentration of aqueous solutions (for example, juice concentrates).
•
Air-Gap Membrane Distillation: An air gap is followed by a cool surface. The use of an air gap configuration allows larger temperature differences to be applied across the membrane, which can compensate in part for the greater transfer resistances. The air gap configuration is the most common and can be used for any application, including desalination. A schematic illustration of an air gap MD is shown in Figure 8.
Figure 8. Schematic of Air Gap MD (CH2MHILL, 2011) The thermal efficiency of MD declines with increasing salinity (TDS levels), because highly saline water requires a greater temperature drop across the air gap, leading to greater losses of heat conduction through the air gap. Similarly, as salinity is increased, lower fluxes can be achieved due to reduced heat transfer with highly saline water. The thermal efficiency and operating flux are estimated as a function of water salinity (Scott et al., 2007). Memstill, a patented MD technology of TNO Environment Energy and Process Innovation, Netherlands, is an innovative concept that combines high transport of water vapor and high transfer of evaporation heat into one membrane module. Because a Memstill module was designed to house a continuum of evaporation stages in an almost ideal countercurrent flow
process, a very high recovery of evaporation heat is possible. The process was intended to decrease desalination costs to well below $0.65/m³, using low-grade waste steam or heat as the driving force (Huehmer and Wang, 2009). Potential Advantages over Established Technologies: • • • • • •
Ability to utilize low-grade heat (solar collector, waste heat, etc.) Less organic fouling propensity compared to RO Less pretreatment compared to conventional membrane desalination processes The only membrane process that can maintain process performance (such as water flux and solute rejection) almost independently of feed solution TDS concentration MD membranes are more chemically inert and resistant to oxidation than traditional RO and NF membranes, which allows for more efficient, chemically aggressive cleaning Produces higher-quality water than NF/RO, EDR and CD.
Potential Disadvantages Compared to Established Technologies: •
•
•
Still under development. Knowledge about the following has not been established: treatment efficiencies of larger-scale installations, economics, short- and long-term performance, and fouling/scaling of MD Requires special hydrophobic membranes. Existing commercially available RO membranes are not suitable for MD because such membranes have a relatively low product water flux, which can be attributed to severe internal concentration polarization in the porous support and fabric layers of RO membranes Contamination of distillate occurs when the membrane fouls and wets the membrane pores
A pilot test of MD using RO concentrate generated from a groundwater desalination facility operated by Eastern Municipal Water District (EMWD), California, USA was performed in 2009 (USBR, 2008). The pilot test showed that the operating flux was between 1.2 and 2.4 gfd at feed and permeate temperatures of 40 and 20 degrees Celsius (°C), respectively. The water recoveries were marginal (i.e. 60%) during pilot testing. However, if the higher temperature differences (∆T) were maintained between feed and permeate side, the recovery could be higher. The pilot MD exhibited excellent salt rejections (that is, 99 percent or greater) during testing. No known full-scale application of MD. A 2010 desalination market survey completed by Global Water Intelligence (GWI) also identified that MD is likely to be the focus of commercialization over the next four years.” MD was one of them. The market interest captured by the recent survey is a critical support for commercial success. The availability of low-grade or waste heat source or renewable energy is another important factor that favors implementation of MD in desalination projects. The following advances are required for commercial success: • •
Development of micro-porous membranes that have the desired porosity, hydrophobicity, low thermal conductivity, and low potential for fouling. Development of membrane modules to reduce footprint.
As with many membrane technologies, MD systems are highly modular and can be applicable at small to large-scale facilities. Project locations where waste heat or low-grade heat is available (such as refineries) are favorable locations for implementation of this technology.
References 1.
Beckman J.R, Dewvaporation Desalination 5,000-Gallon-Per-Day Pilot Plant. 2008. DWPR Report No. 120. Denver, Colorado.: U.S. Bureau of Reclamation
2.
Biagini, B., B. Mack and T.A. Davis. 2010. Zero Discharge Desalination (ZDD) Technology – High Recovery Solution for Inland Desalination to Significantly Reduce Brine Disposal. Proceedings of AMTA Annual Conference and Exposition. July 2010.
3.
Cath T.Y., J.E. Drewes and C. Lundin. 2009. A Novel Hybrid Forward Osmosis – Reverse Osmosis Process for Water Purification and Reuse, using Impaired and Saline Water. Proceedings of American Water Works Association Membrane Technology Conference. April 2009.
4.
CH2M HILL, 2009. Brine-Concentrate Treatment and Disposal Options Report. Prepared for United States Department of the Interior Bureau of Reclamation. October 2009.
5.
Colorado School of Mines. 2009. An Integrated Framework for Treatment and Management of Produced Water. Technical Assessment of Produced Water Treatment Technologies. RPSEA Project 07122-12. November 2009.
6.
Christen, K, 2006. Desalination Technology Could Clean Up Wastewater from Coal-bed Methane Production. Environmental Science & Technology, January 11 2006.
7.
Fan, Y., Hu, H., Liu, H. 2007. Enhanced Coulombic Efficiency and Power Density of Air-cathode
8.
Gutierrez G., A. Lobo, D. Allende, A. Cambiella, C. Pazos, J. Coca, and J.M. Benito (2008). Influence of Coagulant Salt Addition on the Treatment of Oil-in-Water Emulsions by Centrifugation, Ultrafiltration, and Vacuum Evaporation. Separation Science and Technology 43. 1884 – 1895.
9.
Hancock N.T. and T.Y. Cath, 2009. Novel Performance Modeling of Forward Osmosis – Reverse Osmosis Integrated Systems. Proceedings of American Water Works Association Membrane Technology Conference. April 2009.
10. Huehmer R. and F. Wang, 2009. Energy in Desalination: Comparison of Energy Requirements for Developing Desalination Techniques. Proceedings of American Water Works Association Membrane Technology Conference and Exposition. April 2009. 11. Kristen, K. 2007. Environmental Costs of Desalination. Environ. Sci. Technol. 41(16), 5576-5579. 12. Lattemann, S., Kennedy, M., Schippers, J. and Amy, G. 2010. Chapter 2 Global Desalination Situation, Sustainability Science and Engineering. 2, 7-39. 13. Loeb S. (1998). Energy Production at the Dead Sea by Pressure-Retarded Osmosis: Challenge or Chimera? Desalination 120: 247–262. 14. McCutcheon J.R., R.L. McGinnis, and M. Elimelech. 2005. A Novel Ammonia-Carbon Dioxide Forward (direct) Osmosis Desalination Process, Desalination 174, 1-11. 15. Salamero F. D. 2004. Modelling of Membrane Distillation Processes. Computer Aided Process Engineering Center, Department of Chemical Engineering, Technical University of Denmark, Denmark. 16. Scott, L., M Sivakumar, and H. Dharmappa. 2007. Optimising Membrane Distillation Using Hollow Fibres. Sustainable Earth Research Centre, Environmental Engineering University of Wollongong, Australia
17. Seed, L.P., Daren D. Yetman, Yuri Pargaru and Gene S. Shelp. 2006. The DESEL System- Capacitive Deionization for the Removal of Ions from Water. Proceedings of 79th Annual Water Environment Federation Technical Exposition and Conference. October 2006. 18. Tao G., B. Viswanath, K. Kekre, L.Y.Lee, H.Y. Ng, W.C.L. Lay and H. Seah. 2009. RO Brine Treatment by a Capacitive Deionization Based Process to Increase Water Recovery. Proceedings of 13th Annual Water Reuse and Desalination Research Conference. May 2009. 19. Thu K., A. Chakraborty, B.B. Saha, W. G.Chun, K.C. Ng. 2010. Life-cycle Cost Analysis of Adsorption Cycles for Desalination. Desalination and Water Treatment. Vol. 20, Issue 3. August 2010. 20. U.S. Bureau of Reclamation (USBR). 2008. Evaluation and Selection of Available Processes for a Zero Liquid Discharge System for the Perris, California Ground Water Basin. Desalination and Water Purification Research and Development Program Report No. 149. April 2008. 21. U.S. Bureau of Reclamation (USBR). 2009. Multibeneficial Use of Produced Water Through High-Pressure Membrane Treatment and Capacitive Deionization Technology. Desalination and Water Purification Research and Development Program Report No. 133. September 2009. 22. Wang X. and K.C. Ng. 2005. Experimental Investigation of an Adsorption Desalination Plant Using Low-temperature Waste Heat. Applied Thermal Engineering. 25, 2780–2789. 23. Water Desalination Report. 2008. Volume 44, Number 10. April 2008. 24. Water Desalination Report. 2009. Volume 45, Number 27. July 2009. 25. Xu P., J.E. Drewes, D. Heil, and G. Wang (2008 a) Treatment of Brackish Produced Water Using Carbon Aerogelbased Capacitive Deionization Technology, Water Research, 40, 2605-2617. 26. Xu P. and J.E. Drewes (2008 b), Sustainability of Implementing Desalination Technology. Proceedings of Water Reuse and Desalination Research Conference. May 2008.
Reverse osmosis (RO) and electrodialysis reversal (EDR) are widely used technologies for TDS removal from brackish waters and more recently from wastewater in water reuse projects. While both processes are highly effective for removing TDS from water, they generate a concentrate stream which requires proper disposal. Disposal of the concentrate streams is economically and environmentally a big challenge. The challenge gets even bigger in inland areas where surface discharge is not possible. Feasibility of implementing ultimate disposal approaches (i.e., surface discharge, deep well injection, solar evaporation coupled with landfill, etc.) is site/location specific. Often, a concentrate volume reduction technology is used prior to ultimate disposal to reduce project costs. Numerous research studies have been conducted to develop low carbon foot-print technologies and technologies that increase recovery, use renewable energy and low grade waste heat to desalinate wastewater. The majority of these technologies have been developed for brackish water or seawater desalination; the same technologies can also be used for concentrate volume reduction technology. This article highlights the innovative/developmental technologies, presents state of development, and what improvements need to be done for commercial success. The innovative/developmental technologies reviewed in this article included: • • • • •
•
Adsorption/desorption desalination Capacitive Deionization Dewvaporation Forward osmosis High Recovery RO Processes o ARROW o OPUS o SPARRO o Zero Discharge Desalination Membrane Distillation
Review of the Technologies Silica Based Adsorption/Desorption Desalination Technology Silica gel based adsorption technology was developed in National University of Singapore (NUS) to reduce scaling and corrosion issues associated with thermal processes and to lower the carbon footprint of desalination using solar energy or waste heat, and to capture the generated cool air for district cooling. The technology was originally developed for seawater desalination but it can also be applied for brackish water desalination. Currently, NUS and King Abdullah University of Science and Technology (KAUST) are collaborating to optimize system performance and evaluate fouling/scaling behavior of the technology in a demonstration unit (35 kW) treating seawater. A picture of the demo unit at KAUST is presented in Figure1.
Figure 1. Picture of a Demo Unit at KAUST (Adapted from CH2MHILL, 2011) The silica gel adsorption desalination and cooling unit comprises a host of stationary units, namely an evaporator, condenser and adsorber/desorber beds. After de-aeration, saline or brackish water is fed intermittently into the evaporator where desalting is achieved at low system pressure (1–5 kPa) and temperatures (5-10oC) which mitigates the problems associated with scaling and corrosion. During adsorption processes, water vapor is adsorbed by the silica gel due to its high affinity to water at low temperatures. Concomitantly, a heat source such as hot water (typically from 50 to 80oC) is supplied to the desorption beds, containing the saturated silica gel from the previous cycle, to expel the water vapor from the adsorbent. The desorbed vapor condenses on the cooler surfaces of condenser, cooled by recirculating coolant from a seawater cooling tower. Highly purified potable water is produced at low temperatures (50-80oC), which enable the technology to have a cooling application. The heat to drive desorption can be obtained from waste heat as well as renewable sources including solar and geothermal. Potential Advantages over Established Technologies: • Very low specific energy cost. (It is reported to be 1.38 kWh/m3 which is very close to the lowest theoretical amount (1.0 kWh/m3) achieved by any desalination system (Thu et.al. 2010) • Higher water recoveries than RO • Capability to use solar energy results in great flexibility to locate the AD facilities • Does not require pretreatment and post treatment chemicals • Low temperature operation minimizes scaling and corrosion problems Potential Disadvantages Compared to Established Technologies: • •
Requires large space (Use of tall silos can reduce the foot-print but increases the capital cost of the facility) Availability of low temperature waste heat sources is important but finding these sources may be a challenge for siting large-scale AD facilities
Currently, Advon Singapore Pte Ltd. is providing demo units. A four bed small scale demonstration unit has been successfully operated in Singapore for more than 6 years (Wang and Ng. 2005). The electricity consumption and low O&M requirement are the two major attractive features of this technology. In addition, this technology can be retrofitted into an existing MED plant to increase
efficiency of the system as reflected in gain-output-ratio (GOR) value. This technology is potentially best fit to locations where there is need for desalination, demand for cooling and waste heat sources are available. Technology has already been commercialized and large demonstration projects are underway.
Capacitive Deionization (CD) The original CD process was developed and patented at Lawrence Livermore National Laboratory in the late 1980s. CD is a low-pressure, non-membrane desalination technology that uses electrostatic forces to remove dissolved ions from solution. An aqueous solution of soluble salts is passed through pairs of electrodes held at a potential difference of 1.2 volts (V). The electrodes consist of porous carbon aerogel, with a specific surface area of 400 to 1,100 m2/g and with very low electrical resistivity (less than 40 kilohm-meters). Ions are adsorbed to the electrode of opposing charge in a semi-batch process. Eventually, the electrodes become saturated with ions and must be regenerated. To regenerate CD, the applied potential is removed, and the ions attached to the electrodes are released and flushed from the system. Flushing of cells using a small quantity of product water generates a concentrate stream, as shown in Figure 2. Unlike ion exchange processes, no additional chemicals are required for regeneration of the electrosorbent in this system.
Figure 2. Schematic Illustration of CD Operation (Top) and Regeneration (Bottom) (Adapted from CH2MHILL, 2011) Carbon aerogel is an ideal electrode material because of its high electrical conductivity, high specific surface area, and controllable pore size distribution. However, despite recent advancements, carbon aerogel electrodes are still expensive and their ion storage capacity is relatively low. Original designs of CD systems were limited to the treatment of relatively low ionic strength solutions (TDS< 3,000 mg/L). The reason for their limited application has been identified as the high pore volume to surface area characteristic of the carbon electrode material (Seed et al., 2006). The technology has been investigated in several academic and research institutions. In the USA, Colorado School of Mines researchers concluded that this technology could yield results comparable to those of conventional RO for desalination of streams with TDS of less than 3,000 mg/L (Kristen, 2006). This is mainly due to the high cost of CD modules with increased feed water TDS concentrations.
Capacitive Deionization Technology Systems, Inc. (CDT Inc.) is the worldwide licensee for the patented CD technology for water purification and desalination applications. Independently, ENPAR Technologies Inc. of Canada has developed its DesEL System, which uses principles of CD to remove TDS. Potential Advantages over Established Technologies: • •
• • •
CD requires less energy than EDR, RO, and mechanical/thermal processes Membrane technologies require more advanced operation and construction considerations for high-pressure pumps and clean-in-place (CIP) systems. The CD operates at ambient conditions, and there are no requirements for high-pressure pumps and CIP systems CD uses electrostatic regeneration and requires minimal or no chemicals for electrode fouling and scaling controls Less scaling propensity compared to RO Silica does not limit the recovery compared to RO
Potential Disadvantages Compared to Established Technologies: •
•
• •
•
CD is still under development. Knowledge about treatment efficiencies of larger-scale installations, economics, and short- and long-term fouling/scaling issues of CD systems has not been established. Lower TDS removal than RO and mechanical/thermal processes. The process cannot remove uncharged molecules (such as boron, silica, and non-polar organic compounds. Solids and pathogens). CD recovers lower amounts of water than conventional membrane processes Adsorption of total organic carbon (TOC) to the aerogel material during regeneration when the cell is uncharged could result in electrode fouling if organic matter clogs the pores of carbon aerogel material Lengthy down-time period during cleaning of electrodes
Currently, there is no known full-scale CD facility. In a bench-scale study, Seed et al. (2006) showed that CD can achieve 80 percent TDS removal with high water recoveries (up to 95 percent). However, the feed water TDS was only 460 mg/L during bench-scale testing. Tao et al. (2009) used a pilot-scale CD unit to treat and recover an RO concentrate stream generated from the Kranji NeWater Reuse RO facility in Singapore. Biological activated carbon pretreatment followed by CD resulted in 85 percent TDS removal efficiency at approximately 85 percent recovery. Tao et al. (2009) reported a cell energy consumption of 0.7 KWh/m3 based on the pilot CD testing results. Similar to the previous study, the feed TDS concentration of CD was relatively low (1,250 mg/L). This covers only lower range of TDS found in brackish water across USA. A comprehensive evaluation performed by the Colorado School of Mines (USBR, 2009) concluded that the energy consumption of CD is similar to that of RO (4 kWh/1,000 gallons, or 1.1 kWh/m3) for low flow systems (0.7 mL/min). The efficiency and recovery of the process need to be improved before CD becomes economically feasible to treat relatively high TDS water (>5,000 mg/L). CD could probably be implemented within the next 5 to 10 years, if the following improvements are made (USBR, 2009): • •
Higher capacitance and lower cost of electrode materials Recovery of residual electricity
• • •
Shorter regeneration time to reduce process down-time Reduced carryover volume after regeneration Cost-effective and low carbon footprint pretreatment technologies
Dewvaporation Dewvaporation uses a humidification-dehumidification cycle to produce distilled product water. Feed water is evaporated by heated air, and freshwater is condensed on the opposite side of a heat transfer wall. Each Dewvaporation tower contains a heat transfer wall made of plastic. The wall divides the module into two compartments, one for evaporation and one for dew formation. The energy needed for evaporation is partially supplied by the recovered energy released during condensation. Heat sources can be combustible fuel, solar, or low-grade heat from various sources. Using waste heat or low-grade heat can reduce O&M costs significantly, thereby making it a very attractive desalination technology. The tower unit is built of thin plastic films to avoid corrosion and to minimize equipment costs. Tower construction is relatively inexpensive because the towers operate at atmospheric pressure. The Dewvaporation concept was developed at Arizona State University, USA, in conjunction L'Eau LLC, the company that owns the patent rights to the process. The process has been marketed under license by Altela Inc. since mid-2006. Altela, Inc. has designed, manufactured, and tested several AltelaRainTM prototype systems based on the Dewvaporation process. A schematic of the AltelaRainTM process is shown in Figure 3.
Figure 2. AltelaRainTM Dewvaporation Process Schematic
Potential Advantages over Established Technologies: • Can treat very high TDS-containing streams (up to 60,000 mg/L) • Use of solar or waste heat significantly reduces operating cost • Operates in atmospheric pressure, so there are no requirements for high-pressure pumps and expensive sturdy towers • Relatively high recovery (such as 90 percent, based on pilot tests conducted in New Mexico, USA) compared to NF/RO and EDR for treating high-TDS streams • Less complex than mechanical/thermal evaporative processes • Water quality has little impact on process performance • Plastic heat transfer walls reduce capital cost and eliminate corrosion concerns • Less fouling/scaling propensity compared to NF/RO • Silica does not limit the recovery compared to NF/RO • No chemical usage for pretreatment of feed water Potential Disadvantages Compared to Established Technologies: • Dewvaporation is still under development. Knowledge about the following has not been established: treatment efficiencies of larger-scale installations, economics, and short- and long-term performance of the system. • Like other evaporative processes, the energy consumption of the dewvaporation system is high • As in most desalination systems, post-treatment is required for stabilization and mineralization of the water A 5,000-gpd Dewvaporation pilot plant was operated at the 23rd Avenue Wastewater Treatment Plant (WWTP) in Phoenix, Arizona. The pilot plant feed was concentrate from a Tactical Water Purification System (TWPS) RO unit with ultrafiltration pretreatment. A 2,000-mg/L TDS wastewater RO concentrate stream was treated by the pilot plant to more than 45,000-mg/L TDS brine and 10 mg/L TDS distillate, yielding a recovery of up to 95 percent and salt rejection of more than 99 percent. Thermal multiple effects varied from 2.0 to 3.5, which was less than the 5.0 effects demonstrated prior to transport to the WWTP site. Three full-scale AltelaRain ARS-4000 systems were operated at natural gas wells in the San Juan basin near Farmington, New Mexico, USA (Colorado School of Mines, 2009). The ARS-4000 system processed approximately 4,000 gpd of produced water. The AltelaRainTM System produced distilled water with TDS of approximately 100 mg/L while processing a waste stream containing approximately 42,000 mg/L TDS. One unit reduced effluent disposal volumes by as much as 90 percent (Colorado School of Mines, 2009). This technology has shown to be effective for producing high-quality water from high-TDS streams; it can be used for treating and recovering RO/EDR concentrate/reject streams. Commercial units are already available and can be used in small applications. For large-scale applications, custom design is essential to reduce the capital investment. Project locations where waste heat or low-grade heat is available (such as refineries) are favorable locations for implementation of this technology.
Forward Osmosis (FO) A 2010 desalination market survey completed by Global Water Intelligence (GWI) notes that “three areas of new technology are likely to be the focus of commercialization over the next four years” forward osmosis was one of them. FO is an innovative membrane-based technology that has the potential to reduce the costs and environmental impacts of desalination. This is an osmotic process that uses a semi-permeable membrane to separate salts from water. FO uses an osmotic pressure gradient
(∆п) instead of hydraulic pressure (∆P), which is used in RO, to create the driving force for water transport through the membrane. No energy is needed to drive the water flux of an FO process, as the water flux is the natural tendency of the system. The concentrated solution, or draw solution, is the source of the driving force in the FO process. A selective permeable membrane allows passage of water, but rejects solute molecules and ions. Osmotic driving forces in FO can be significantly greater than hydraulic driving forces in RO. This results in the potential for higher water flux rates and recoveries. The selection of an appropriate draw solution is the key to FO performance. Example draw solutions include magnesium chloride, calcium chloride, sodium chloride, potassium chloride, ammonium carbonate, and sucrose. A simplified process schematic of an FO process using ammonium carbonate as a draw solution is presented in Figure 3.
Figure 3. Process Schematic of FO (Adapted from McCutcheon et al., 2007) Potential Advantages over Established Technologies: • • •
Operates around 1 atm, which results in much lower energy consumption compared to conventional membrane and mechanical/thermal evaporative desalination technologies Membrane compaction is not typically an issue Less fouling propensity compared to RO
Potential Disadvantages Compared to Established Technologies: •
•
•
Still under development. Knowledge regarding treatment efficiencies of larger-scale installations, economics, and short- and long-term performance and fouling/scaling has not been established Requires special membranes. Existing commercially available RO membranes are not suitable for FO because such membranes have a relatively low product water flux, which can be attributed to severe internal concentration polarization in the porous support and fabric layers of RO membranes Use of ammonium carbonate as draw solution may provide desired osmotic pressure. However, ammonia diffuses back to the permeate stream and it should be removed using a cost effective technology (such as waste heat to strip ammonia)
Various approaches for FO-based desalination exist. In the USA, a pilot-scale FO unit was built and has been operated at the Yale University laboratory since 2005. The Yale pilot study utilizes an ammonium carbonate solution as the draw solution. To recover freshwater, the diluted ammonium carbonate solution is heated to approximately 55oC, where ammonium carbonate undergoes thermal decomposition. The Yale research is being commercialized by Oasys Water Inc. Modern Water PLC has constructed a 25,000 gpd pilot facility in Gibraltar for sea water as a desalination demonstration (Water Desalination Report, 2008). Additional facilities located in Oman are being developed by Modern Water. Several researchers have investigated the use of FO as a potential energy source using pressure restrained osmosis (PRO). Loeb (1998) first demonstrated the feasibility of energy production by PRO. In 2009, Statkraft, the national energy company of Norway, commissioned a demonstration PRO power plant in Tofte, Norway. With 2-kW capacity, the facility is a proof of concept test bed. More recently, membrane manufacturer HTI Water, of Arizona (USA), launched large-scale commercialization of FO as a low-energy means of treating waste flows such as hydraulic fracturing fluids. Modern Water PLC has already established three installations using “manipulated osmosis” as a desalination process. The following advancements are needed for consideration of this technology in full-scale applications. • • • •
Identifying effective and economical draw solutions and technologies/approaches that remove draw solutions economically (such as using waste/low-grade heat). Developing new and additional membrane sources. Currently, a limited number of commercially available membranes are on the market using cellulose triacetate. Addressing mass transfer limitations resulting from concentration polarization within the membrane support layer. Developing new modules suitable for full-scale implementation. To date, most applications have used flat-sheet, plate and frame elements.
High Recovery RO Processes The major issue for operating an RO process at higher recoveries is the precipitation of sparingly soluble inorganic salts, most notably barium sulfate (BaSO4), calcium carbonate (CaCO3), calcium phosphate (Ca3(PO4)2), calcium sulfate (CaSO4) and silica. Inorganic salt precipitation can be controlled at lower recoveries by using an appropriate antiscalant and by acidifying feed water pH (effective for CaCO3 and (Ca3(PO4)2 control). At higher recoveries (greater than 85 percent), the concentration of sparingly soluble salts can exceed the effective range of antiscalants, and pH control does not prevent precipitation of some problematic minerals such as silica, BaSO4 and CaSO4 (that may not be as effectively removed by chemical cleaning). To improve RO recovery, one or more components of these scale-forming salts need to be lowered. High-recovery RO processes were developed to alter water chemistry prior to RO to allow higher water recoveries. These processes are applicable for treating RO/EDR concentrate streams with very high recoveries, in some cases approaching zero liquid discharge (ZLD). High-recovery RO/concentrate recovery processes include Advanced Reject Recovery of Water, OPUSTM, SPARRO, and ZDDTM.
a. Advanced Reject Recovery of Water (ARROWTM) ARROWTM is a high-recovery, advanced membrane system that couples a softening process with RO to increase water recovery. This is a proprietary technology marketed by Advanced Water Solutions and O’Brien & Gere. ARROWTM has a number of configurations that can be adjusted depending on the flow rate, hardness, concentration of silica relative to other hardness precursors, and TDS concentration. The ARROW process is illustrated in Figure 4 and includes the following steps: Pretreatment - Pretreatment typically includes acid and antiscalant addition
First-Stage RO - ARROWTM produces a permeate stream of 60 to 75 percent of the flow, while 40 to 25 percent of the stream is RO concentrate. Second-Stage RO - Concentrate from the first-stage RO is treated and combined with an appropriate flow of recycled stream from the second-stage RO concentrate. Softening of RO Concentrate Stream from Second-Stage RO - ARROW uses either chemical precipitation to reduce calcium, magnesium, and silica hardness or ion exchange (IX) softening containing strongly acidic cation exchange resins, if silica hardness is not a concern. Chemical precipitation uses caustic soda or soda ash depending upon the ratio of alkalinity to calcium hardness. Recovery - A small amount of flow from second–stage RO concentrate and a small reject stream from the bottom of the clarifier or from the IX system is sent to a solar evaporator or thermal crystallizer. The combined volume of the two reject streams is less than 5 percent, giving an overall process recovery of greater than 95 percent.
Figure 4. ARROWTM Process Schematic Potential Advantages over Established Technologies:
• • • •
High-quality product water compared to CD and EDR Applicable for treating high silica content streams Pretreatment reduces fouling and scaling precursors, resulting in much higher RO system recovery rates (e.g., 90 percent) compared to EDR, CD, and conventional membrane-based systems for treating high-TDS water. Pretreatment also reduces feed pressure and energy requirements. Compact skid-mounted system, which reduces equipment delivery and installation time
Potential Disadvantages Compared to Established Technologies: •
Process is still under development; no full-scale applications exist in municipal water or wastewater treatment
• • • • •
High cost of chemicals used for pretreatment and softening of water Requires skilled operation Available from only a single supplier and so does not allow for competitive bidding Pilot testing may be required to determine key design criteria Sludge from softening might require separate disposal, creating additional challenge and expense
No full-scale operation. There is one demonstration project in the industrial water treatment field. This project is a 33-gallon-per-minute (gpm) unit in New Jersey, USA. Knowledge about treatment efficiencies, economics (including sludge disposal and concentrate disposal costs), and long-term system performance is not established. Existing skid-mounted units, which are easy to implement, are applicable only to very small systems (up to 0.25 mgd) Potential applications are currently limited to small industrial treatment and small brine/concentrate recovery applications. Process performance and robustness under varying feed water quality conditions should be proven, and implementation costs should be comparable to those of other high recovery technologies for this technology to be considered for wider size and purpose applications. b. OPUSTM N.A. Water Systems, Veolia Water Solutions and Technology, France, designed the OPUS™ system. OPUS™ is a proprietary optimized pretreatment which incorporates unique separation processes for desalination of water with high concentrations of sparingly soluble solutes (e.g., silica, CaSO4, and Mg(OH)2), organics, and boron. The system is able to achieve high recovery with high purity product water through the use of extensive pretreatment processes prior to water being processed through IX and RO subsystems (RPSEA, 2009). A process schematic is shown in Figure 5.
Figure 5. OPUSTM Process Schematic (Adapted from CH2M HILL, 2011)
The first step of the process includes acidification and degasification of the raw feed water. This is followed by a conventional coagulation, flocculation, and high-rate plate settler sedimentation process, which is termed Multiflo™. After this step, the flow stream should be devoid of nearly all highmolecular-weight organic molecules and oxidized metals (particularly iron and manganese). Additionally, colloidal silica is partially removed by co-precipitation. Decant from the sedimentation basin is then filtered by a packed-bed media filtration column, which removes any microflocs and most
suspended solids that pass through the plate settlers. The media filter may also achieve additional removal of low to medium molecular weight hydrophobic organic molecules, including oil and grease (Colorado School of Mines, 2009). Filtrate from the media filter is then processed through a mixed, packed-bed IX column for further water softening. A cartridge filter is then used to remove any IX resin or remaining particulate material prior to RO. The water is then pressurized and treated by brackish water RO (BWRO) membranes at an elevated pH. Operating the RO elements under these conditions reduces the fouling propensity of silica and increases the rejection of both silica and boron. Potential Advantages over Established Technologies: • Could produce very high quality water (product literature reports greater than 99 percent rejection of TDS and most multivalent solutes and achievement of additional silica and boron removal with high pH operation) • Pretreatment reduces fouling and scaling precursors, resulting in much higher RO system recovery rates (e.g., 90 percent) compared to EDR, CD, and conventional membrane-based systems for treating high-TDS water. Pretreatment also reduces feed pressure and energy requirements. Potential Disadvantages Compared to Established Technologies: • OPUSTM is still under development. Knowledge about the following has not been established: treatment efficiencies of larger-scale installations, economics, and short- and long-term performance of the system. • Despite the high water recoveries, OPUSTM generates multiple waste streams and a concentrate stream that needs to be disposed off or treated; sludge from the sedimentation basin requires dewatering and landfill application. • Larger footprint than conventional RO or IX systems due to inclusion of pretreatment, chemical feed and storage, and dewatering facilities (if included). • Complex operation. Requires skilled labor and sophisticated process automation. • Available from only a single supplier and so does not allow for competitive bidding • Multiple chemicals to handle, including acids, bases, hydrolyzing metal coagulants, and polymerbased coagulants. • Available from only a single supplier and so does not allow for competitive bidding Currently, there are no full-scale applications. The OPUS™ was field tested at a steam-enhanced oil production field in San Ardo, California, USA. Field trials have demonstrated that this system can treat feed water with TDS levels up to 10,000 mg/L. Veolia Water Solutions Technologies claims that it could treat streams with TDS levels up to 30,000 mg/L. The treatment process permeate quality is dependent on feed water salinity and operating conditions. However, product literature reports greater than 99 percent rejection of TDS and most multivalent solutes (Colorado School of Mines, 2009). The technology is still under development by Veolia. The technology has not been pilot-tested extensively and requires third-party independent evaluation to establish/prove the following: • Life cycle costs of the treatment • Robustness of the process under varying feed water and operating conditions • Design and operating parameters to optimize process performance and to lower chemical usage
c. Slurry Precipitation and Reverse Osmosis (SPARRO)
SPARRO involves circulating a slurry of seed crystals within the RO system, which serve as preferential growth sites for calcium sulfate and other calcium salts and silicates. The scaling compounds are precipitated on the seed crystals instead of on the membrane. This process is confined to the use of tubular membranes due to the need to circulate the slurry within the membranes without plugging them as presented in Figure 6. The water to be desalted is mixed with a stream of recycled concentrate containing the seed crystals and fed to the RO process. The concentrate with seed crystals is processed in a cyclone separator to separate the crystals, and the desired seed concentration is maintained in a reactor tank by controlling the rate of wasting the upflow and/or underflow streams from the separator. The combined recovery of the process is estimated to be greater than 90 percent (CH2M HILL, 2009).
Figure 6. Illustration of Tubular Membranes Used in SPARRO (Adapted from CH2M HILL, 2009) Potential Advantages over Established Technologies: • Low energy requirement compared to RO and thermal/evaporative desalination processes • Higher recoveries compared to EDR and RO, especially treating high scaling potential waters (industrial, RO concentrate, etc.) Potential Disadvantages Compared to Established Technologies: • Still under development. Knowledge about the following has not been established: treatment efficiencies of larger-scale installations, economics, and short- and long-term performance of the system. • Larger footprint; requires large reaction tanks and large area for membrane systems due to lower specific surface area of membranes • Tubular membranes have lower TDS rejections (TOC and TN rejections may also be lower than with traditional RO membranes) • Additional chemicals to handle • Relatively complex operation
Currently there are no full-scale applications. This technology has been tested at a pilot scale on several applications, including treating mining wastewater in South Africa, and treating primary and secondary RO concentrate streams from the Eastern Municipal Water District (EMWD), California, USA, zero liquid discharge (ZLD) pilot project (USBR, 2008). The combined recovery of the process was greater than 90 percent according to the study performed in South Africa (CH2M HILL, 2009). The results from the EMWD study indicate 60 to 70 percent recovery. Although this recovery seems moderate, it is in fact remarkably high considering the water quality matrix which prohibits use of RO/EDR without extensive pretreatment (softening and IX). Knowledge about treatment efficiencies, economics, and long-term performance of the system has not been established. Process performance and robustness under varying feed water quality conditions should be proven and cost should be comparable to those of other high recovery technologies before this technology could be considered in mid- to large-scale applications.
d. Zero Discharge Desalination (ZDD™) Zero Discharge Desalination (ZDD™), marketed by Veolia Water Solutions and Technology, France, was developed to reduce concentrate volumes generated from the EDR/RO facilities. In ZDDTM, the concentrate stream from a conventional RO system is fed to an electrodialysis metathesis (EDM) stack consisting of ion exchange membranes and thin solution compartments. A direct electric potential is applied to the ends of the stack, resulting in a direct current that is carried by ions migrating through the membranes and solution compartments. The DC potential pushes ions through membranes from a lower concentration to a more concentrated solution. Water flows tangentially to the membrane, while the flow of ions is perpendicular to the membrane (Biagini et al., 2010). An electrodialysis stack consists of anion membranes containing many fixed positive charges, usually quaternary amines, that are loosely associated with mobile negatively charged ions that permeate the membrane, and cation membranes with fixed negative charges, usually sulfonic acids, that allow positively charged ions to permeate. In the ZDD technology, the calcium salts that cause scaling in a high recovery RO process are removed in an EDM process that is a variant of ordinary EDR. The concentrate recovery mode of operation assumes that the RO concentrate from an existing RO system would be treated by the EDM process to produce highly concentrated salt streams while reducing the salt concentrations and the scaling tendency of this RO concentrate. If necessary, a silica removal system can be incorporated as a slip stream treatment, as presented in the process flow schematic shown in Figure 7. From an energy perspective, it is not practical to reduce the salt concentration to potable levels by the EDM process, so RO would be used to further reduce the TDS of the EDM treated water. Potential Advantages over Established Technologies: • Potentially very high recovery (97 percent recovery, including dewatering) was reported in Brackish Groundwater National Desalination Research Facility in Alamogordo, New Mexico (USA) • Less waste to handle compared to IX-RO or lime softening-RO-based processes • Requires less space than lime softening RO Potential Disadvantages Compared to Established Technologies: • ZDDTM is still under development. Knowledge about the following has not been established: treatment efficiencies of larger-scale installations, economics, and short- and long-term performance of the system.
•
Crystallization of sodium bicarbonate in a sodium chloride solution limits the possibility of increasing the water recovery in waters that contain substantial quantities of chloride and alkalinity
Figure 7. ZDD Process Schematic (Adapted from CH2M HILL, 2011) Currently there are no full-scale applications. A large-scale pilot test was conducted at the Brackish Groundwater National Desalination Research Facility in Alamogordo, New Mexico, USA. The pilot testing results indicated that a recovery of 94 percent was achievable without dewatering (97 percent with dewatering). The cost projections performed by Veolia Water Systems indicated much lower capital and O&M costs compared to the lime softening-RO systems. The technology is still under development by Veolia. The technology has not been pilot-tested extensively and requires third-party independent evaluation to establish/prove the following: • • •
Life cycle costs of the treatment including solids disposal Robustness of the process under varying feed water and operating conditions Design and operating parameters to optimize process performance
Project costs should be comparable to those of other high recovery technologies before this technology could be considered for mid- to large-scale applications.
Membrane Distillation (MD) Membrane distillation (MD) is another technology that GWI identified combines membrane technology and evaporation processing in a single unit. The appeal of MD is that it functions at atmospheric pressure and requires only relatively low feed temperatures of 70ºto 90ºC. MD transports water vapor through the pores of hydrophobic membranes using the temperature difference across the membrane. The membrane allows water vapor to penetrate the hydrophobic surface while repelling the liquid. The clean vapor is carried away from the membrane and condensed as pure water, either within the membrane package or in a separate condenser system.
MD differs from other membrane technologies in that the driving force that pushes the water through the membrane is not feed pressure but temperature. In MD units, vapor production is enhanced by heating the feed water, which increases the vapor pressure and penetration rate. MD requires the same amount of energy input to heat and condense vapor as traditional evaporation; however, it does not require boiling water and is operated at ambient pressure. The energy requirement for MD is lower than those of conventional evaporation systems. MD is most efficient on low-grade or waste heat, such as industrial heat streams or even solar energy (Huehmer and Wang, 2009). Also, the efficiency of the unit can be improved with heat recovery. MD membranes must be microporous (pore diameters of 0.05 to 0.2 μm) and non-wettable by the feed. A variety of arrangements and configurations can be used to induce the vapor through the membrane and to condense penetrant gas; however, the feed water must always be in direct contact with the membrane. Condensation is typically achieved via two major process configurations (Salamero, 2004):
•
Direct-Contact Membrane Distillation: The cool condensing solution directly contacts the membrane and flows countercurrent to the raw water. This is the simplest configuration and is best suited for applications such as desalination and concentration of aqueous solutions (for example, juice concentrates).
•
Air-Gap Membrane Distillation: An air gap is followed by a cool surface. The use of an air gap configuration allows larger temperature differences to be applied across the membrane, which can compensate in part for the greater transfer resistances. The air gap configuration is the most common and can be used for any application, including desalination. A schematic illustration of an air gap MD is shown in Figure 8.
Figure 8. Schematic of Air Gap MD (CH2MHILL, 2011) The thermal efficiency of MD declines with increasing salinity (TDS levels), because highly saline water requires a greater temperature drop across the air gap, leading to greater losses of heat conduction through the air gap. Similarly, as salinity is increased, lower fluxes can be achieved due to reduced heat transfer with highly saline water. The thermal efficiency and operating flux are estimated as a function of water salinity (Scott et al., 2007). Memstill, a patented MD technology of TNO Environment Energy and Process Innovation, Netherlands, is an innovative concept that combines high transport of water vapor and high transfer of evaporation heat into one membrane module. Because a Memstill module was designed to house a continuum of evaporation stages in an almost ideal countercurrent flow
process, a very high recovery of evaporation heat is possible. The process was intended to decrease desalination costs to well below $0.65/m³, using low-grade waste steam or heat as the driving force (Huehmer and Wang, 2009). Potential Advantages over Established Technologies: • • • • • •
Ability to utilize low-grade heat (solar collector, waste heat, etc.) Less organic fouling propensity compared to RO Less pretreatment compared to conventional membrane desalination processes The only membrane process that can maintain process performance (such as water flux and solute rejection) almost independently of feed solution TDS concentration MD membranes are more chemically inert and resistant to oxidation than traditional RO and NF membranes, which allows for more efficient, chemically aggressive cleaning Produces higher-quality water than NF/RO, EDR and CD.
Potential Disadvantages Compared to Established Technologies: •
•
•
Still under development. Knowledge about the following has not been established: treatment efficiencies of larger-scale installations, economics, short- and long-term performance, and fouling/scaling of MD Requires special hydrophobic membranes. Existing commercially available RO membranes are not suitable for MD because such membranes have a relatively low product water flux, which can be attributed to severe internal concentration polarization in the porous support and fabric layers of RO membranes Contamination of distillate occurs when the membrane fouls and wets the membrane pores
A pilot test of MD using RO concentrate generated from a groundwater desalination facility operated by Eastern Municipal Water District (EMWD), California, USA was performed in 2009 (USBR, 2008). The pilot test showed that the operating flux was between 1.2 and 2.4 gfd at feed and permeate temperatures of 40 and 20 degrees Celsius (°C), respectively. The water recoveries were marginal (i.e. 60%) during pilot testing. However, if the higher temperature differences (∆T) were maintained between feed and permeate side, the recovery could be higher. The pilot MD exhibited excellent salt rejections (that is, 99 percent or greater) during testing. No known full-scale application of MD. A 2010 desalination market survey completed by Global Water Intelligence (GWI) also identified that MD is likely to be the focus of commercialization over the next four years.” MD was one of them. The market interest captured by the recent survey is a critical support for commercial success. The availability of low-grade or waste heat source or renewable energy is another important factor that favors implementation of MD in desalination projects. The following advances are required for commercial success: • •
Development of micro-porous membranes that have the desired porosity, hydrophobicity, low thermal conductivity, and low potential for fouling. Development of membrane modules to reduce footprint.
As with many membrane technologies, MD systems are highly modular and can be applicable at small to large-scale facilities. Project locations where waste heat or low-grade heat is available (such as refineries) are favorable locations for implementation of this technology.
References 1.
Beckman J.R, Dewvaporation Desalination 5,000-Gallon-Per-Day Pilot Plant. 2008. DWPR Report No. 120. Denver, Colorado.: U.S. Bureau of Reclamation
2.
Biagini, B., B. Mack and T.A. Davis. 2010. Zero Discharge Desalination (ZDD) Technology – High Recovery Solution for Inland Desalination to Significantly Reduce Brine Disposal. Proceedings of AMTA Annual Conference and Exposition. July 2010.
3.
Cath T.Y., J.E. Drewes and C. Lundin. 2009. A Novel Hybrid Forward Osmosis – Reverse Osmosis Process for Water Purification and Reuse, using Impaired and Saline Water. Proceedings of American Water Works Association Membrane Technology Conference. April 2009.
4.
CH2M HILL, 2009. Brine-Concentrate Treatment and Disposal Options Report. Prepared for United States Department of the Interior Bureau of Reclamation. October 2009.
5.
Colorado School of Mines. 2009. An Integrated Framework for Treatment and Management of Produced Water. Technical Assessment of Produced Water Treatment Technologies. RPSEA Project 07122-12. November 2009.
6.
Christen, K, 2006. Desalination Technology Could Clean Up Wastewater from Coal-bed Methane Production. Environmental Science & Technology, January 11 2006.
7.
Fan, Y., Hu, H., Liu, H. 2007. Enhanced Coulombic Efficiency and Power Density of Air-cathode
8.
Gutierrez G., A. Lobo, D. Allende, A. Cambiella, C. Pazos, J. Coca, and J.M. Benito (2008). Influence of Coagulant Salt Addition on the Treatment of Oil-in-Water Emulsions by Centrifugation, Ultrafiltration, and Vacuum Evaporation. Separation Science and Technology 43. 1884 – 1895.
9.
Hancock N.T. and T.Y. Cath, 2009. Novel Performance Modeling of Forward Osmosis – Reverse Osmosis Integrated Systems. Proceedings of American Water Works Association Membrane Technology Conference. April 2009.
10. Huehmer R. and F. Wang, 2009. Energy in Desalination: Comparison of Energy Requirements for Developing Desalination Techniques. Proceedings of American Water Works Association Membrane Technology Conference and Exposition. April 2009. 11. Kristen, K. 2007. Environmental Costs of Desalination. Environ. Sci. Technol. 41(16), 5576-5579. 12. Lattemann, S., Kennedy, M., Schippers, J. and Amy, G. 2010. Chapter 2 Global Desalination Situation, Sustainability Science and Engineering. 2, 7-39. 13. Loeb S. (1998). Energy Production at the Dead Sea by Pressure-Retarded Osmosis: Challenge or Chimera? Desalination 120: 247–262. 14. McCutcheon J.R., R.L. McGinnis, and M. Elimelech. 2005. A Novel Ammonia-Carbon Dioxide Forward (direct) Osmosis Desalination Process, Desalination 174, 1-11. 15. Salamero F. D. 2004. Modelling of Membrane Distillation Processes. Computer Aided Process Engineering Center, Department of Chemical Engineering, Technical University of Denmark, Denmark. 16. Scott, L., M Sivakumar, and H. Dharmappa. 2007. Optimising Membrane Distillation Using Hollow Fibres. Sustainable Earth Research Centre, Environmental Engineering University of Wollongong, Australia
17. Seed, L.P., Daren D. Yetman, Yuri Pargaru and Gene S. Shelp. 2006. The DESEL System- Capacitive Deionization for the Removal of Ions from Water. Proceedings of 79th Annual Water Environment Federation Technical Exposition and Conference. October 2006. 18. Tao G., B. Viswanath, K. Kekre, L.Y.Lee, H.Y. Ng, W.C.L. Lay and H. Seah. 2009. RO Brine Treatment by a Capacitive Deionization Based Process to Increase Water Recovery. Proceedings of 13th Annual Water Reuse and Desalination Research Conference. May 2009. 19. Thu K., A. Chakraborty, B.B. Saha, W. G.Chun, K.C. Ng. 2010. Life-cycle Cost Analysis of Adsorption Cycles for Desalination. Desalination and Water Treatment. Vol. 20, Issue 3. August 2010. 20. U.S. Bureau of Reclamation (USBR). 2008. Evaluation and Selection of Available Processes for a Zero Liquid Discharge System for the Perris, California Ground Water Basin. Desalination and Water Purification Research and Development Program Report No. 149. April 2008. 21. U.S. Bureau of Reclamation (USBR). 2009. Multibeneficial Use of Produced Water Through High-Pressure Membrane Treatment and Capacitive Deionization Technology. Desalination and Water Purification Research and Development Program Report No. 133. September 2009. 22. Wang X. and K.C. Ng. 2005. Experimental Investigation of an Adsorption Desalination Plant Using Low-temperature Waste Heat. Applied Thermal Engineering. 25, 2780–2789. 23. Water Desalination Report. 2008. Volume 44, Number 10. April 2008. 24. Water Desalination Report. 2009. Volume 45, Number 27. July 2009. 25. Xu P., J.E. Drewes, D. Heil, and G. Wang (2008 a) Treatment of Brackish Produced Water Using Carbon Aerogelbased Capacitive Deionization Technology, Water Research, 40, 2605-2617. 26. Xu P. and J.E. Drewes (2008 b), Sustainability of Implementing Desalination Technology. Proceedings of Water Reuse and Desalination Research Conference. May 2008.
