Selective Removal Of Sodium And Chloride? Mono-Valent Selective Ion Exchange Membrane For Desalination And Reuse Enhancement Charlie (Qun) He, Carollo Engineers, Inc., CHE@Carollo.com, 4600 E Washington Street, Suite 500, Phoenix, AZ 85034 Pei Xu, Nirmala Khandan, New Mexico State University Abstract Desalination and water reuse has become a key solution to addressing water shortage and a critical component of water sustainability. Desalting technologies such as reverse osmosis (RO) are a primary method for treating brackish groundwater as they are effective in removing most contaminants. The primary shortfall of RO is management and disposal of the highly saline concentrate laden with accumulated contaminants. This brine stream represents a significant loss of water resource and is often associated with expensive concentrate treatment. Reclaimed water with high salinity and high concentrations of certain inorganic contaminants is causing reuse challenges. Conventional RO membranes cannot remove monovalent ions selectively. Some NF membranes has selectivity. These membranes are often looser and reject more larger ions (e.g., Calcium and sulfate) than smaller ions (e.g., sodium, chloride), therefore worsening the situation. This paper will provide a synopsis on technologies that are capable of selectively removing sodium and chloride ions. ED and EDR use electrical potential to separate ions from water, and have been widely used for desalination of groundwater and wastewater. One of the benefits of ED/EDR is the selectivity of removing monovalent ions (such as Na+, NH4+, Cl-, and NO3-) over multivalent ions, which can be achieved by using monovalent selective IX membranes or by lowering the electrical current. The ongoing project is funded by USBR Desalination and Water Purification Research and Development Program (DWPR). It includes bench testing and pilot testing of the newly developed monovalent selective membranes (sodium selective membrane CR671 and nitrate/chloride selective membrane AR112B by GE Water) on a 10-15 gallon per minute EDR unit at two test sites: treating tertiary reclaimed water at the Water Campus, City of Scottsdale, Arizona; and treating reverse osmosis (RO) concentrate at the Kay Bailey Hutchison Desalination Plant (KBHDP), El Paso, Texas. Preliminary results demonstrated that these membranes provide unique advantages of removing sodium, chloride and nitrate. This presentation will include results of ongoing bench testing and pilot testing in Scottsdale, which is expected to be completed by August 2015. Introduction Currently our water industry is facing unprecedented challenges in supply water. Due to continued population growth, economic development, frequently occurred droughts, the demand for freshwater increases. However, freshwater sources are very limited. That is, less than 0.5
percent of the world s entire water resources; while seawater constitutes 97 percent. The existing limited freshwater sources are further declining due to salinity buildup, contamination and overdraft. The difficulty to meet the increasing demand for new freshwater resources has motivated many municipalities and water utilities to explore the desalination of seawater, brackish waters, and reclaimed water as alternative water supplies. Desalination and water reuse has become a key solution to addressing water shortage and a critical component of water sustainability. Desalting technologies such as reverse osmosis (RO) are a primary method for treating brackish groundwater as they are effective in removing most contaminants. The primary shortfall of RO is management and disposal of the highly saline concentrate laden with accumulated contaminants. This brine stream represents a significant loss of water resource and is often associated with expensive concentrate treatment. Reclaimed water with high salinity and high concentrations of certain inorganic contaminants is causing reuse challenges. For example, in Arizona and the entire southwest region, reclaimed water with high sodium concentration or high sodium adsorption ratio (SAR) poses a critical issue for irrigation. Effluent with high Total Dissolved Solids (TDS) and chloride concentrations often fails the Whole Effluent Toxicity (WET) test and result in compliance issues for discharge. Many utilities uses or plans for RO and nanofiltration (NF) to desalinate the reclaimed water, which generates a brine that cannot be easily disposed of. A treatment process that can selectively removes sodium or chloride ions, instead of all salts, is ideal but not available. Conventional RO membranes cannot remove monovalent ions selectively. Some NF membranes has selectivity. These membranes are often looser and reject more larger ions (e.g., Calcium and sulfate) than smaller ions (e.g., sodium, chloride), therefore worsening the situation. This paper will provide a synopsis on technologies that are capable of selectively removing sodium and chloride ions. ED and EDR use electrical potential to separate ions from water, and have been widely used for desalination of groundwater and wastewater. One of the benefits of ED/EDR is the selectivity of removing monovalent ions (such as Na+, NH4+, Cl-, and NO3-) over multivalent ions, which can be achieved by using monovalent selective IX membranes or by lowering the electrical current. ED process can be controlled to partially reduce TDS concentration of the water rather than produce water with very low TDS. This can minimize the complexity of product water blending and stabilization, and allow more options for concentrate management and disposal. Compared to RO membranes, ED and EDR membranes are less stringent to pre-treatment requirements and more tolerant to waters with high silica, hardness, suspended solids, and organics (Reahl, 2006; Trussell and Williams, 2012; Turek et al., 2009). Therefore ED and EDR offer significant advantages over conventional RO and can operate at substantially lower cost and higher water recovery to treat reclaimed water, brackish groundwater and RO concentrate. Monovalent selective membrane is a leading edge technology that is available from only a few suppliers. For this project, monovalent selective membranes from two manufacturers were tested: NEOSEPTA membranes from Japan and newly developed membranes from GE. The NEOSEPTA membrane is commercially available and has been widely tested and used. GE s monovalent anion membranes were developed and tested in 1997, but the monovalent cation membranes were not fully developed and tested until recently.
Newly developed IX membranes enhance the selective removal of monovalent ions. Using these new membranes, sodium ions can be exchanged with calcium or magnesium ions between two process streams across an IX membrane. Exchanging calcium for sodium in such a process represents a potential solution to reuse water salinity and SAR problems without generating a brine. For example, a saline reclaimed water stream (~1,200 mg/l TDS, ~250 mg/l sodium, SAR > 5) is not suitable for irrigation. When treated with an EDR system using monovalent membrane, the "product" stream, (about 50-70% of the feed flow, < 900 mg/l TDS, <110 mg/l sodium) would be ideal for irrigation. Instead of creating a brine waste that must be disposed of appropriately, the concentrated stream from this process (30-50% of flow, 2,000-3,000 mg/l TDS, ~500 mg/l sodium) is still suitable for uses such as urban lakes, wetland creation, and dust control. Other application examples will also be discussed in the presentation, including using the monovalent selective membrane to treat brackish water groundwater or RO concentrate. An ongoing project is funded by USBR Desalination and Water Purification Research and Development Program (DWPR). It includes bench testing and pilot testing of the newly developed monovalent selective IX membranes (sodium selective membrane CR671 and nitrate/chloride selective membrane AR112B by GE Water) using a 10-15 gallon per minute EDR unit at two test sites: treating tertiary reclaimed water at the Water Campus, City of Scottsdale, Arizona; and treating RO concentrate at the Kay Bailey Hutchison Desalination Plant (KBHDP), El Paso, Texas. Preliminary results demonstrated that these membranes provide unique advantages of removing sodium, chloride and nitrate. The anticipated performance and engineering cost estimates for various treatment schemes utilizing this unique new membrane will be presented. How Does It Work As illustrated in Figure 1, a conventional electrodialysis (ED) stack comprises a series of alternating cation and anion permselective membranes between a cathode and anode. Cations are drawn toward the negatively charged cathode passing through negatively charged cation exchange membrane and being rejected by the positively charged anion exchange membrane. Similarly, anions are drawn toward the positively charged anode passing through the anion exchange membrane and being rejected by the cation exchange membrane.
Figure 1: Electrodialysis (ED) with Selective Ion Exchange Membranes In an electrodialysis stack, the diluate (D) feed stream, brine or concentrate (C) stream, and electrode (E) stream are allowed to flow through the appropriate cell compartments formed by the ion exchange membranes. Under the influence of an electrical potential difference, the negatively charged ions (e.g., chloride and sulfate) in the diluate stream migrate toward the positively charged anode. These ions pass through the positively charged anion exchange membrane, but are prevented from further migration toward the anode by the negatively charged cation exchange membrane and therefore stay in the C stream, which becomes concentrated with the anions. The positively charged species (e.g., sodium and calcium) in the D stream migrate toward the negatively charged cathode and pass through the negatively charged cation exchange membrane. These cations also stay in the C stream, prevented from further migration toward the cathode by the positively charged anion exchange membrane. As a result of the anion and cation migration, electric current flows between the cathode and anode. Only an equal number of anion and cation charge equivalents are transferred from the D stream into the C stream and so the charge balance is maintained in each stream. The overall result of the electrodialysis process is an ion concentration increase in the concentrate stream with a depletion of ions in the diluate solution feed stream. The E stream is the electrode stream that flows past each electrode in the stack. This stream may consist of the same composition as the feed stream (e.g., sodium chloride) or may be a separate solution containing a different species (e.g., sodium sulfate). Depending on the stack configuration, anions and cations from the electrode stream may be transported into the C stream, or anions and cations from the D stream may be transported into the E stream. In each case, this transport is necessary to carry current across the stack and maintain electrically neutral stack solutions. The monovalent valent selective membrane works similar to regular ED/EDR, with the exception that the selective membrane rejects more sodium than calcium and magnesium, more chloride and nitrate than sulfate and phosphate. Bench and pilot scale testing was conducted to investigate the selectively of the membrane and its long term performance.
Materials and Methods - Bench Scale The ED bench testing used a single-stage ED-1 electrodialysis stack from GE Water. Ten pairs of IX membranes provided a total surface area of 900 in 2 with an individual membrane area of 90 m 2. The stack is integrated in a lab-scale ED-unit that provides the flow for three compartments (feed, concentrate, and electrode rinse) and the electrical power. The feed and concentrate streams were operated in a once-through mode to simulate 60 percent recovery. The working pressure was 2-4 pounds per square inch (psi). Figure 2 provides a photo of the apparatus. Membranes utilized for the testing include GE normal grade membranes (AR204 and CR67), GE Monovalent selective membranes (AR112B and CR671), NEOSEPTA normal grade membranes (AMX and CMX-SB) and NEOSEPTA monovalent permselective membranes (ACS and CMX- S). Microfiltration treated reclaimed water collected from the City of Scottsdale Water Campus Advanced Water Treatment Plant. The facility is treating reclaimed water from the Water Campus Wastewater Treatment Plant using microfiltration and RO (85% recovery). Table 1 summarizes the reclaimed water quality.
Table 1 Reclaimed Water Quality Analytes Reclaimed water Calcium, mg/l 80.0 Magnesium, mg/l 30.0 Sodium, mg/l 235.0 Potassium, mg/l 33.0 Bicarbonate, mg/l 148.8 Sulfate, mg/l 272.0 Chloride, mg/l 311.0 Fluoride, mg/l 0.5 Nitrate, mg/l 13.3 Total PO4, mg/l 15.7 SAR 5.68 TDS, mg/l 1153.7 Conductivity, us/cm 1787.9 ph 7.12
Figure 2 Bench Scale Electrodialysis Reversal (EDR) Apparatus Pilot Testing Site and Equipment Two phases of pilot testing are included for this project. The first phase was conducted at Scottsdale Water Campus inside its Microfiltration Building. The testing started in July 2015 and completed in October 2015. The testing utilized rental EDR equipment (Aquamite IV) from GE Waters. The stack has a two stage configuration, with 50 cell pairs each stage. Figure 3 showed photos of the testing site and the stack. Figure 3 Pilot Scale Testing Site at Scottsdale Water Campus
Preliminary Testing Results A series of testing was completed using the bench scale apparatus. Figure 4 presents selected data on the sodium membrane selectivity. The data was presented in selectivity over sodium (defined as the rejection of a given cation over the rejection of sodium ion) versus percent polarization. The results showed that potassium, another monovalent ion has slightly higher rejection rates compared to sodium ion (with a selectivity around 1.2) across a wide range of percent polarization. The membrane rejected far less calcium and magnesium, especially under lower polarization. For example, at a polarization of 65%, the rejection rate of magnesium ions is only 5% of that for the sodium ion. At a polarization of 85%, the rejection of calcium ions is only 10% of that for the sodium. This data demonstrated that the monovalent membrane has potentially good selectivity. Figure 4 Selected Bench Testing Results for Selective Membrane - Cation Selectivity A series of pilot testing was completed at Scottsdale Water Campus, using both normal grade membranes and monovalent membranes for comparison. At the time of this paper, pending the water quality analysis from the lab, only testing conducted with normal grade membranes are available. But all data will be available at the time of the conference. As demonstrated in Figure 5 below, the actual salt rejection was slightly higher than the GE EDR performance model (WATSYS) projections. The rejection rates decreased with increases in percent polarization.
Figures 6 through 8 presents the actual and projected product water sodium, calcium and chloride by normal grade membranes. The results showed that WATSYS overestimated the rejection of sodium slightly, but significantly underestimated the rejection of divalent ions (i.e. Calcium). It also slightly underestimate the rejection of chloride. Based on preliminary testing results, it is expected that the monovalent membrane would reject much more sodium and chloride than calcium and sulfate. Figure 5 Selected Pilot Scale Testing Results for Normal Grade Membrane - Actual and Modeled Conductivity in Product Water
Figure 6 Selected Pilot Scale Testing Results for Normal Grade Membrane - Actual and Modeled Sodium in Product Water Figure 7 Selected Pilot Scale Testing Results for Normal Grade Membrane - Actual and Modeled Calcium in Product Water
Figure 8 Selected Pilot Scale Testing Results for Normal Grade Membrane - Actual and Modeled Chloride in Product Water Conclusion The following conclusions can be drawn from the testing: Monovalent selective membranes are more effective to remove Na+ than Ca 2+ and Mg 2+ at lower polarization and higher flow rate. EDR with GE normal grade membranes can meet the required water quality for irrigation. Long-term pilot-testing is ongoing to demonstrate the viability of monovalent selective membranes for beneficial use of reclaimed water achieving ZLD.