Electrodialysis and Electrodialysis Reversal

Home , Electrical polarity, Electrodialysis, Electrodialysis reversal

AWWA MANUAL M38 First Edition©

FOUNDED 1881 American Water Works Association

Preface, v

Acknowledgments, vii

Chapter 1 Introduction ...... 1 Basic Water Chemistry Concepts, 1 Operating Principles of ED and EDR, 3 Development of ED and EDR Systems, 5 Applications, 10

Chapter 2 Design ...... 13 Components of ED and EDR Design, 13 Staging, 20 Limiting Parameters, 22 Water Recovery, 25 Pretreatment, 26 Operating Principles for Design, 29 Posttreatment, 31 Concentrate Disposal, 32 References, 35

Chapter 3 Equipment and Installation ...... 37 Equipment Subsystems, 37 Installation, 41 Costing, 42 References, 44

Chapter 4 Operation and Maintenance ...... 45 Operation Procedures, 45 Maintenance Requirements, 47 Safety, 52

Abbreviations, 55

Additional© Sources of Information, 57

Index, 59

iii

Chapter 1 Introduction

Electrodialysis (ED) is an electrically driven membrane process used to demineralize brackish water. Brackish waters lie under approximately two thirds of the United States, and inland rivers, such as the Rio Grande and the lower reaches of the Colorado, also contain high levels of salinity. Water is classified as brackish when mineral content ranges between that of fresh drinking water and that of seawater. Brackish water contains more than 500 mg/L of total dissolved solids (TDS) and seawater more than 30,000 mg/L TDS. ED and electrodialysis reversal (EDR) reduce TDS in brackish source water by electrically removing contaminants that exceed acceptable levels for drinking and process water. An overview of membrane process applications based on the molecular weights of contaminants appears in Figure 1-1. The ED and EDR processes are competitive with reverse osmosis (RO) in treating brackish waters. Typical ED systems include chemical feed systems for antiscalant and perhaps acid addition, a cartridge filter for prefiltration, the ED unit, and equipment for aeration, disinfection, and stabilization. EDR systems can often operate without fouling and scaling chemical feed, and they can treat high-fouling sources more efficiently than RO. However, it is important to remember that the types of membranes used in ED and EDR systems do not provide a barrier to remove microorganisms as do RO, nanofiltration© (NF), ultrafiltration (UF), and microfiltration (MF) membranes. BASIC WATER CHEMISTRY CONCEPTS ______A basic understanding of salts and water is necessary to understand the design, operation, and maintenance of a water demineralization system. A review of water chemistry concepts is provided here. Ionic Solutions An ion is a charged atom, molecule, or radical, the migration of which affects the transport of electricity through an electrolyte solution. For example, common table salt is a typical ionic compound. The chemical name for this crystal is sodium chloride, and the chemical symbol is NaCl. The crystal consists of two types of

Metal Ions Aqueous Salts Viruses Humic Acids Bacteria Cysts Antimony Sodium Salts Infectious Trihalomethane Salmonella Protozoa Arsenic Sulfate Salts Hepatitis Precursors Shigella Giardia Nitrate Manganese Salts Vibrio cholerae Cryptosporidium Nitrite Aluminum Salts Cyanide etc.

Figure 1-1 Membrane processes overview

charged atoms, sodium and chloride, that are held together by electrically attractive forces. If a crystal of salt is dissolved in water, water molecules will orient themselves around the charged atoms and nullify the attractive force between them. This is known as the solution and dissociation (dissolving) of a salt in water. When this occurs, two electrically charged particles are formed, one with a positive charge (sodium, represented as Na+) and one with a negative charge (chloride, represented ©– as Cl ). The subatomic particle responsible for the electrical charge is called an electron. An electron, by convention, has an assigned charge of negative one (–1). An atom that accepts an electron during the dissociation process will have a net charge of –1. An atom that gives up an electron during the process will have a net charge of +1. These resulting charged particles are ions. The positively charged ions are called cations, and the negatively charged ions are called anions. These two types of ions are completely dissociated and mobile in water. In the same manner as salts, minerals and acids may also dissociate into ions in solution. Some common ions that may be found in natural water are shown in Table 1-1. Some of the ions listed in Table 1-1 have more than one positive or negative

Table 1-1 Common ions found in natural waters

Cations Anions Sodium (Na+) Chloride (Cl–) +2 – Calcium (Ca ) Bicarbonate (HCO3 ) +2 –2 Magnesium (Mg ) Sulfate (SO4 ) + – Potassium (K )Nitrate (NO3 )

charge associated with them (e.g., calcium has a charge of +2). In these cases, the ion has accepted or given up more than a single electron during the dissociation process. Electrical Conductivity The most important property of an ionic solution is its ability to conduct electricity. When two electrodes are connected to a direct current (DC) power supply and immersed in pure water, no electric current passes between the electrodes because no ions exist in the solution to transport the current. In an ionic solution, however, the dissociated ions transport the electric charge between the two electrodes. The ability of a solution to carry an electric charge is known as conductivity and is measured in either micromhos per centimetre (µmho/cm) or microsiemens per centimetre (µS/cm). Conductivity is affected by the concentration of ions, the ionic composition, and the temperature of the solution in the following ways: • Increasing ion concentration results in increased electric conductivity. • Smaller ions and those with more than one electric charge tend to move through the solution more quickly. • Raising the temperature increases ion mobility, resulting in an increase in conductivity. OPERATING PRINCIPLES OF ED AND EDR ______Electrodialysis is an electrochemical separation process in which ions are transferred through ion exchange membranes by means of a DC voltage. This process can be understood more clearly by referring to Figure 1-2, which shows a tank filled with an NaCl solution and electrodes (cathode and anode) placed at either end. When DC potential is applied across the electrodes, the following take place: • Cations (Na+) are attracted to the cathode, or negative electrode. • Anions (Cl–) are attracted to the anode, or positive electrode. ©• Pairs of water molecules break down (dissociate) at the cathode to produce – two hydroxyl (OH ) ions plus hydrogen gas (H2). • Pairs of water molecules dissociate at the anode to produce four hydrogen ions + – (H ), one molecule of oxygen (O2), and four electrons (e ). • Chlorine gas (Cl2) may be formed at the anode. The movement of ions in the tank can be controlled by the addition of ion exchange membranes that form watertight compartments, as shown in Figure 1-3. The two types of ion exchange membranes used in electrodialysis are • anion transfer membranes (A in Figure 1-3), which are electrically conductive membranes that are water impermeable and allow only negatively charged ions to pass through

Source: Ionics Inc. Figure 1-2 Sodium chloride solution under the influence of a DC potential

Source: Ionics Inc. Figure 1-3 Ion exchange membranes in an NaCl solution (DC circuit open)

• cation transfer membranes (C in Figure 1-3), which are electrically conductive membranes that are water impermeable and allow only positively charged ions to pass through Varieties of these basic types of membranes exist that are selective to ions that are either monovalent (having a charge magnitude of 1) or divalent (having a charge magnitude of 2). Other types can be formulated to enhance the passage rates of selected© ions. For example, membranes exist that show an affinity for nitrate passage over other anions. In Figure 1-3 there is no DC potential applied to the electrodes and no movement of ions. Figure 1-4 shows what occurs when DC potential is applied across the electrodes. The figure shows six compartments separated by ion exchange membranes. The membranes influence ion behavior as follows: 1. Compartments 1 and 6 — Compartments 1 and 6 contain metal electrodes where reduction and oxidation occur. 2. Compartment 2 — Cl– ions pass through the anion membrane (A) into compartment 3, while Na+ ions move through the cation membrane (C) into compartment 1.

Source: Ionics Inc. Figure 1-4 DC potential applied across electrodes for an NaCl solution with ion exchange membrane

3. Compartment 3 — The Na+ ions cannot move through the anion membrane and remain in compartment 3. The Cl– ions cannot pass through the cation membrane and also remain in compartment 3. 4. Compartment 4 — The Cl– ions pass through the anion membrane into compartment 5, while Na+ ions pass through the cation membrane into compartment 3. 5. Compartment 5 — The Na+ ions cannot pass through the anion membrane and remain in compartment 5. The Cl– ions cannot pass through the cation membrane and remain in compartment 5. Compartments 2 and 4 are depleted of ions, whereas compartments 3 and 5 have a higher concentration of ions. When these membranes are properly arranged, two major and separate streams are produced (demineralized and concentrated), as well as two minor streams from the electrode compartments. For water treatment, several hundred of these compartments are assembled into a membrane stack, forming the heart of an ED system. DEVELOPMENT OF ED AND EDR SYSTEMS ______ED selectively removes dissolved solids, based on their electrical charge, by transferring the brackish water ions through a semipermeable ion exchange membrane charged with an electrical potential. Figure 1-5 shows a schematic of an entire ED system. It points out that the feedwater becomes separated into the following© three types of water: (1) product water, which has an acceptably low TDS level; (2) brine, or concentrate, which is the water that receives the brackish water ions; and (3) electrode feedwater, which is the water that passes directly over the electrodes that create the electrical potential. EDR involves reversing the electrical charge to a membrane after a specific interval of time. As described later, this polarity reversal helps prevent the formation of scale on the membranes. Figure 1-6 shows a schematic of an EDR system. The setup is very similar to an ED system except for the presence of reversal valves. Demineralization of brackish water using ED was pioneered in the 1950s. ED has been used successfully over the past 40 years to treat municipal and process water supplies. ED process technology has advanced rapidly since its inception because of improved ion exchange membrane properties, better materials of

Legend: C Conductivity Controller PRV Pressure-Regulating Valve

Source: Ionics Inc. Figure 1-5 Electrodialysis© system flow diagram

Legend: C Conductivity Controller PRV Pressure-Regulating Valve

Source: Ionics Inc. Figure 1-6 Electrodialysis reversal system flow diagram

construction, advances in technology, and the evolution of polarity reversal. According to IDA Desalting Plants Inventory,* the installed worldwide capacity of ED and EDR membrane treatment plants increased from 2 mgd (7.5 ML/d) in 1955 to more than 200 mgd (750 ML/d) in 1992. Custom-designed and prepackaged ED and EDR plants provide water at predetermined TDS or salt-removal levels with high water recovery rates (i.e., with low amounts of feedwater being sent to waste). Additional production can be achieved by adding process trains or by operating the units in parallel (side by side) rather than in series (one after the other). The desalting capacity can be increased with additional stages of membranes in series. ED and EDR systems are capable of treating variable source water quality while producing a consistent finished water quality. ED and EDR plants can be designed to remove from 50 to 99 percent of source water contaminants or dissolved solids. Source water salinities of less than 100 mg/L up to 12,000 mg/L TDS can be successfully treated to produce finished water of less than 10 mg/L TDS. Batch and Continuous Electrodialysis The first type of commercial ED system was the batch system. In this type of ED system, source water is recirculated from a holding tank through the demineralizing spacers of a single membrane stack and back to the holding tank until the final purity is obtained. The production rate is dependent on the dissolved minerals concentration in the source water and on the degree of demineralization required. The concentrate stream is also recirculated to reduce wastewater volume, and continuous addition of acid is required to prevent membrane stack scaling. The second type of commercially available system was the unidirectional continuous-type ED. In this type of system, the membrane stack contains two stages in series; each stage helps demineralize the water. The demineralized stream makes a single pass through the stack and exits as product water. The concentrate stream is partially recycled to reduce wastewater volume and is injected with acid to prevent scaling. ED systems are unidirectional in the sense that cations move only toward the cathode and anions move only toward the anode. The current polarity does not reverse. (However, the direction of flow could reverse, and some commercial systems use this technique to deter the buildup of slime and foulants.) In unidirectional ED systems, scale prevention is achieved either by the use of scale inhibitors for calcium sulfate (CaSO4) control and/or acids for carbonates control, or through the use of permselective membranes. Permselective membranes can be tailored to inhibit the passage of divalent anions or cations, such as sulfates, calcium, and magnesium. Permselective© refers to the ability of an ED membrane to discriminate between different ions to allow passage or permeation through the membrane. For example, the AST-type membranes show good permeation or high transport numbers for monovalent anions, such as Cl– or NO–2, but have low transport numbers and show –2 –2 very low permeation rates for divalent or trivalent ions, such as SO4 , PO4 , or similar anions. This is achieved by specially treating the anion membrane, and the effect can be exploited to separate various ions. Existing commercial membranes are monovalent anion specific, monovalent cation specific, or hydrogen ion specific. The relative specificities vary, with the monovalent anion membrane showing the greatest

*Available from the International Desalting Association, Topsdale, Mass.

specificity, for example, the ratio of chloride to sulfate ion transport numbers. Through the use of proper staging, with monovalent and divalent permselective membranes, the development of high calcium sulfate in the concentrate side of the membranes can be forestalled and scale formation prevented. Figures 1-7 and 1-8 illustrate. Figure 1-7 illustrates how a combination of monovalent anion selective membranes in a first stage, followed by a second stage containing monovalent cation permselective membranes, can be used to concentrate solutions well past the normal calcium sulfate solubility limits. In the first stage, no sulfate passes through the membrane, and so the concentrate is rich in calcium chloride. Rather than passing this concentrate to stage 2, the stage 2 system concentrate is made up from fresh feed or another source. Here, the passage of calcium ions is retarded and the concentrate is rich in sodium sulfate. Neither stage ever exceeds the calcium sulfate solubility limits. Yet, when the two concentrate streams are combined, together they can far exceed the calcium sulfate limit. In fact, precipitation can result on mixing. Figure 1-8 is a detail of the first stage from Figure 1-7 showing the use of a standard membrane with a monovalent anion permselective membrane. The concentrate stream is very low in sulfate, about equal to or slightly greater than the feedwater, while the chloride and sodium concentrations, for example, could be many times higher than the feedwater. In other schemes, the concentrate can be made up from a separate water source that is already low in sulfate (for example, reverse osmosis permeate or ED dilute water) to increase water recovery. Colloidal particles or slimes that are slightly electronegative may accumulate on the anion membrane and cause membrane fouling. This problem is common to all classes of ED systems. These fouling agents are removed by flushing with cleaning systems. Control of scale and fouling is critical to all membrane systems — ED, EDR, RO, UF,© and others. Costs to install, operate, and maintain chemical feed systems as well

Source: Thomas D. Wolfe. Figure 1-7 Use of monovalent permselective ED membranes for high recovery (concentration of calcium sulfate in saturated waters)

------

Source: Thomas D. Wolfe. Figure 1-8 Principle of monovalent permselective electrodialysis

as chemical storage facilities can significantly add to the costs of any membrane- based system. Electrodialysis Reversal Electrode compartments in EDR perform differently from those in unidirectional ED. EDR systems, first developed in the 1960s, incorporate electrical polarity reversal to control membrane scaling and fouling. These systems are designed to produce demineralized water continuously without continuous chemical addition during normal operation. In EDR systems, the polarity of the electrodes is reversed two to four times each hour. When polarity is reversed, chemical reactions at the electrodes are reversed. At the negative electrode, reactions produce hydrogen gas and hydroxide ions. Hydroxide raises the pH of the water, causing calcium carbonate (CaCO3) precipitation. At the positive electrode, reactions produce acid, oxygen, and some chlorine. The acid tends to dissolve any calcium carbonate present to inhibit scaling. ©Valves in the electrode streams automatically switch flows in the two types of compartments. Streams that were in demineralizing compartments become concentrate streams, and concentrate streams become demineralizing streams, as shown in Figure 1-9. Because of the corrosive nature of the anode compartments, electrodes are constructed of an inert metal, usually platinum coated. The current-reversal process affects the operation of a membrane system by • detaching polarization films • breaking up freshly precipitated scale or seeds of scale before they can cause damage

• reducing slime formations on membrane surfaces • reducing problems associated with the use of chemicals • cleaning electrodes with acid automatically during anodic operation APPLICATIONS ______Both ED and EDR are electrically driven membrane processes that selectively remove soluble ionic constituents carrying electrical charges that pass through permeable ion exchange membranes. The natural electrical conductivity of water allows ED and EDR processes to be applied to a wide range of water treatment objectives. In ED and EDR systems, the membranes are impermeable only to water and to particles that have a particular characteristic (e.g., a certain charge), so these systems do not present a barrier to remove bacteria or noncharged organic contaminants. In contrast, RO, NF, UF, and MF systems filter water through membranes designed to remove contaminants in the molecular and ionic size ranges, effectively removing Giardia cysts and enteric viruses. RO can be used in combination with ED and EDR to remove these contaminants and to further concentrate the waste stream. Reduction of Total Dissolved Solids Reduction of TDS to meet drinking water standards is the most common application of ED and EDR technology. Plants treating brackish sources that contain up to 10,000 mg/L TDS can reliably and economically yield product water containing less than 500 mg/L TDS. For example, an EDR plant in Sarasota County, Fla., treats brackish well water with 2,500 mg/L TDS and yields product water of less than 350 mg/L TDS — an overall reduction in dissolved solids of 86 percent — with 85 percent recovery. The same plant can produce product water with 500 mg/L TDS from source waters containing as much as 3,600 mg/L TDS at the same rate of recovery. This flexibility is particularly important in applications for which multiple or variable source waters are used. Control of Inorganics and Ionized Contaminants ED and EDR also control specific inorganic constituents or ionized contaminants in water. Common applications include the reduction of naturally occurring levels of sodium, chloride, fluoride, or sulfate to below the US Environmental Protection Agency (USEPA) regulatory levels. ED and EDR can be used to remove or reduce some of the following common ionized constituents: • TDS ©• chromium • sodium • mercury • chloride • copper • sulfate • uranium • fluoride • nitrate and nitrite • iron

A. Before polarity reversal

B. After polarity reversal

Source: Ionics Inc. Figure 1-9 Reversed polarity in EDR systems

• selenium • hardness • barium • bicarbonate • cadmium ©• strontium Removal rates for these ionized constituents are similar to TDS removal rates. In addition to overall TDS reduction, ED and EDR systems effectively treat source water problems, such as saltwater intrusion; high nitrate–nitrite and selenium levels from agricultural contamination; high, naturally occurring fluoride levels; and heavy metals contamination.