What Is Electrodialysis and Why Use It for Salt Removal?
Electrodialysis (ED) removes salts from industrial wastewater by driving dissolved ions through selective cation- and anion-exchange membranes under an applied electric field, concentrating them into a brine stream while producing a desalted permeate. ED is most cost-effective at feed TDS of roughly 1–10 g/L, where it delivers 70–95% salt recovery at 0.5–1.5 V per membrane pair, making it a strong alternative or complement to reverse osmosis in textile, chemical, and metal-finishing applications.
Mechanistically, an ED stack alternates cation-exchange membranes (CEM), which pass only positively charged ions such as Na⁺ and Ca²⁺, with anion-exchange membranes (AEM), which pass only negatively charged ions such as Cl⁻ and SO₄²⁻. A DC field applied across the stack shuttles those ions out of one flow channel and into the next, so a desalted "diluate" stream and a concentrated "brine" stream leave the same skid simultaneously (per S3, S4). Because the driving force is voltage rather than pressure or heat, ED sits alongside reverse osmosis (pressure-driven) and multi-effect evaporation (heat-driven) as the third major route to industrial desalination.
ED is documented across textile, metal/electroplating, chemical, food, and livestock effluents (per S3), and the academic literature positions it as a sustainable desalination method with strong energy efficiency at moderate feed salinities (per S4). For an engineer weighing RO versus ED in 2026, that positioning is the deciding fact: ED does not compete head-to-head with seawater RO at 35 g/L; it competes at the 1–10 g/L middle band where RO energy and brine penalties are steepest.
How an Electrodialysis Stack Is Built
A commercial ED membrane stack is a plate-and-frame device in which the smallest repeating unit is a single cell pair: one cation-exchange membrane, one anion-exchange membrane, two flow spacers (one for diluate, one for concentrate), and the sealing frames that hold them apart. At each end of the stack sits an electrode — typically titanium with a platinum or mixed-metal-oxide coating — fed by an electrolyte rinse loop that carries the small amount of gas and reaction products generated at the electrode surface. Across one cell pair the applied DC voltage is normally 0.5–1.5 V; multiply by cell-pair count and the stack DC bus becomes a meaningful electrical design point.
Two process streams flow simultaneously through the stack. The diluate stream enters with the feed TDS and exits with the desalted product; the concentrate stream enters with low TDS (often recirculated from the brine tank) and exits as a more concentrated brine. The electrode rinse loop is a third, separate stream that never contacts the product water, which keeps the iron, chloride, and hydrogen/peroxide generated at the electrodes out of the permeate.
Industrial stacks typically contain 50–600 cell pairs. Frames are commonly molded PVC or polypropylene, spacers are polyethylene or PVC with diamond or woven flow-net profiles, and the ion-exchange membranes themselves are heterogeneous or homogeneous polymer sheets 0.3–0.6 mm thick. With proper pretreatment — turbidity below 1 NTU, free chlorine below 0.5 mg/L, and a routine CIP cycle — industrial ED membrane life is in the 3–7 year range. Stack frames and electrodes routinely last 10+ years.
2026 Process Design Parameters for Industrial ED Systems
The numeric envelope below is what you should expect a reputable ED vendor to quote in 2026. If a proposal sits well outside these bands, push back: either the application is unusual or the design is wrong.
| Parameter | Typical industrial range (2026) | Design note |
|---|---|---|
| Feed TDS | 1–10 g/L | Sweet spot for capital efficiency; below 1 g/L favors RO, above 10 g/L favors ED as a pre-concentrator to RO or evaporator |
| Operating current density | 100–500 A/m² | Set by limiting current for the feed; higher density = smaller stack area but more energy per kg salt removed |
| Per-membrane-pair voltage | 0.5–1.5 V DC | At 300 cell pairs, total stack bus ≈ 150–450 V DC |
| Salt removal per pass | 25–60% | Single-pass figure; staged or batch ED reaches 70–95% |
| Overall water recovery | 70–95% | Higher than single-pass RO at the same feed TDS |
| Specific energy consumption | 1–7 kWh/m³ treated | At 2–5 g/L feed TDS; scales roughly with log of TDS reduction |
| Feed turbidity | <1 NTU | Mandatory to protect membrane surface |
| Free chlorine in feed | <0.5 mg/L | Chlorine degrades anion-exchange membranes rapidly |
| Feed temperature | 5–40 °C | Higher temperature lowers resistance but may damage membranes above 45 °C |
| Membrane lifetime | 3–7 years | With proper pretreatment and CIP discipline |
Two engineering points that are easy to miss: first, current density is not free — running above the limiting current for a given feed causes water splitting, pH drift, and irreversible membrane fouling. Second, per-membrane-pair voltage is the metric that scales with the rectifier, so a 600-pair stack at 1.0 V/cell pair needs a 600 V DC rectifier, which moves the skid into special electrical-design territory.
Electrodialysis vs Reverse Osmosis: Head-to-Head for Salt Removal
Both ED and RO are membrane processes, but they reject salt through different mechanisms: ED uses an electric field and ion-selective membranes, RO uses hydraulic pressure and a dense salt-rejecting layer. That single difference cascades through every downstream design choice, from pretreatment to energy recovery to brine handling.
| Criterion | Electrodialysis (ED / EDR) | Reverse Osmosis (RO) |
|---|---|---|
| Driving force | DC voltage across ion-exchange membranes | Hydraulic pressure across a dense membrane |
| Best feed TDS | 1–10 g/L (industrial sweet spot) | <2 g/L brackish; up to 35 g/L with seawater RO |
| Water recovery | 70–95% | 60–85% (single pass) |
| Salt removal per pass | 25–60% (staged to 70–95%) | 95–99.5% |
| Brine volume | Smaller at moderate TDS; can be tuned | Larger; fixed by recovery and feed TDS |
| Scaling sensitivity (Ca, Si, Ba) | Lower; EDR reverses polarity to break scale | Higher; requires anti-scalant and tight recovery limits |
| Energy form | Electrical (kWh) | Electrical (kWh) plus high-pressure pump work |
| Membrane cost per m² | Higher | Lower, but replacement is more frequent on harsh feeds |
| Typical use | Salt concentration, brine recovery, ZLD pre-concentrator | High-purity reuse water, final polish |
For a 2026 plant project, the practical rule is: if feed TDS is 1–10 g/L, organics are moderate, and the goal is salt concentration or partial desalination (not deionization), ED is usually the better primary unit. If the goal is reuse-quality permeate with <500 mg/L TDS, RO wins. Many plants run a hybrid train — ED to do the bulk salt reduction at high recovery, then industrial RO polishers for ED permeate to push the final TDS down for reuse. That hybrid is also how the high-salinity wastewater hybrid process designs most plants are spec'ing in 2026 actually look like in practice.
Where Electrodialysis Wins in 2026: Application Snapshots
Textile dye-bath reuse is the clearest ED win. A hybrid UF + ED process has been demonstrated for removal of COD, salt, and color so that the permeate can be returned to the dye house — the salt fraction that fouls downstream reuse loops is removed without thermally degrading the dye chemistry (per S1). For a textile plant discharging 2,000–5,000 m³/day of dye-bath wastewater in the 3–8 g/L TDS range, ED is typically a smaller energy consumer than evaporative salt removal and avoids the color breakthrough that plagues RO.
Metal-finishing and electroplating rinsewater is the second anchor case. ED concentrates nickel, copper, and zinc ions into a brine that can be returned to the plating bath, while producing a diluate that meets sewer discharge limits for heavy metals (per S2). For mining and metals operations, ED applied to acidic drainage can split the stream into a metal-recovery concentrate and a reusable water stream — a configuration increasingly common in desalination market regional analysis 2026 for MENA and Latin American mining sites.
Chlor-alkali and chemical process brines are a third fit: ED or EDR is used to upgrade spent brine for reuse and reduce salt consumption in chlor-alkali cells. Food and dairy applications — cheese whey deashing, soy sauce desalination, sugar refining — exploit ED's low operating temperature, which protects heat-sensitive products. In every one of these, the underlying economic logic is the same: both the desalted water and the concentrated salt have value, so a process that delivers them simultaneously is more efficient than one that destroys one to make the other.
Pretreatment, Hybrids, and Where Equipment Fits in the ED Train
An ED stack is only as good as the water in front of it. The typical 2026 ED train runs: equalization → coarse screening → DAF pretreatment upstream of an ED stack or an MBR biological step ahead of electrodialysis → multi-media filter for ED feed polishing → 5 µm cartridge filter → ED stack → post-treatment pH correction. For salt-laden streams carrying FOG, fibers, or high TSS (textile, food, pulp & paper), DAF upstream protects the cartridge and ED membranes from organic blinding. For organic-rich feeds, an MBR upstream drops COD and TSS so the ED stack sees <1 NTU feed water with stable conductivity.
Chemical conditioning is handled by an automatic anti-scalant and pH dosing for ED trains skid mounted upstream of the stack. After the ED stack, the desalted permeate is often polished through RO if reuse-quality water is the end target, and the brine from ED can be further concentrated in a mechanical vapor recompression evaporator or a crystallizer for plants targeting zero liquid discharge. For sites with limited footprint, an integrated water purification package combining clarification, filtration, and chemical dosing can replace three separate skids in the upstream section.
One operating point worth flagging: EDR (electrodialysis reversal) flips the polarity of the stack every 15–60 minutes, which breaks scale off the membrane surface and dramatically extends CIP intervals on hard-water feeds. If your feed has Ca²⁺ + SO₄²⁻ or silica scaling potential, specify EDR — the incremental cost is small compared with the avoided fouling.
2026 Selection Framework: Should You Specify Electrodialysis?
Use this five-step framework in your next project meeting to decide whether ED belongs on the flow sheet.
- Characterize the feed. Measure TDS, hardness, silica, organics (COD/BOD), TSS, and temperature. If TDS is 1–10 g/L and organics are moderate, ED moves to the shortlist. Above 10 g/L, ED is still viable as a pre-concentrator feeding RO or an evaporator for ZLD, per the brackish water RO system specifications used as a reference for hybrid trains.
- Define the product target. Are you producing reuse water, meeting a sewer discharge limit, or recovering a saleable salt or metal? ED is strongest when both the desalted water and the concentrated stream are valuable. If only the water has value and the brine is a disposal cost, RO usually wins.
- Evaluate the hybrid option. For reuse-quality effluent, run UF/MBR → ED → RO, with ED handling the bulk salt load at 80%+ recovery and RO polishing to <500 mg/L. For ZLD, run ED → evaporator/crystallizer, with ED cutting the evaporator feed volume by 60–80%. The industrial water reuse and RO/ED market trends for 2026 show hybrid ED trains are now the default for new textile and metal-finishing projects in Asia and MENA.
- Score vendors. Compare membrane area per stack (m²), current density rating, whether EDR is standard or an option, materials of construction, and reference list inside your industry. Ask for the same vendor's last three startup reports and CIP history.
- Run a 90-day pilot. Most reputable ED suppliers offer containerized pilot units. Require at least 2,500 operating hours and steady-state data across at least one seasonal temperature swing before you sign a full-scale PO. Where space is constrained, an integrated packaged pretreatment skid can sit beside the pilot container and be reused in the full-scale plant.
Frequently Asked Questions
How does electrodialysis remove salt?
Electrodialysis removes salt by applying a DC voltage across a stack of alternating cation- and anion-exchange membranes. Cations such as Na⁺ migrate toward the cathode and pass only through the cation-exchange membranes; anions such as Cl⁻ migrate toward the anode and pass only through the anion-exchange membranes. The result is one set of flow channels that becomes progressively desalted (the diluate) and an alternating set that becomes progressively concentrated (the brine) (per S3, S4).
Is electrodialysis cheaper than reverse osmosis?
It depends on the feed. At 1–10 g/L feed TDS, ED typically has lower total energy and lower membrane replacement cost than RO, because the driving force is voltage rather than high-pressure pumping and the membranes tolerate scaling better — especially in EDR mode. Below 1 g/L feed or when permeate TDS must drop below 500 mg/L, RO is almost always cheaper per m³ of product water. Many 2026 plants run both, with ED handling the bulk salt reduction and RO polishing the ED permeate to reuse quality.
What TDS range is best for electrodialysis?
The industrial sweet spot for stand-alone ED is 1–10 g/L feed TDS. Below 1 g/L, the energy per kg of salt removed rises because the ionic current falls. Above 10 g/L, single-pass voltage drop across each cell pair climbs and the stack becomes uneconomical without staging; at very high TDS, ED is better deployed as a pre-concentrator feeding RO or an evaporator rather than as a stand-alone desalination step.
Can electrodialysis recover metals from wastewater?
Yes. ED and EDR are used to concentrate nickel, copper, zinc, and other metal ions from electroplating rinsewater, mining effluents, and acidic drainage, producing a brine that can be returned to the process and a diluate that meets discharge limits (per S2). For acidic metal-bearing streams, ED's ability to operate across a wide pH range — combined with selective membranes — makes it one of the few membrane processes that can both recover the metal and reuse the water in a single unit operation.
How much electricity does an electrodialysis system use?
A well-designed industrial ED system uses roughly 1–7 kWh per m³ of treated water at 2–5 g/L feed TDS, with the figure rising as the required salt removal increases. The dependence is roughly logarithmic: doubling the salt removed increases specific energy consumption by about 50–70%, not 100%, which is why ED holds up economically at high recovery. Including auxiliary loads (pumps, dosing, controls), a typical full-scale plant with ED, pretreatment, and RO polish runs 3–10 kWh/m³ total.