What Electrodialysis Does in Industrial Water Treatment
Electrodialysis (ED) is an electrically driven, membrane-based separation that moves charged species out of a feed stream under a DC potential, leaving neutral molecules behind. Unlike thermal desalination, which pays the enthalpy of vaporizing water, ED does not require a phase change. Unlike reverse osmosis, which uses hydraulic pressure to push water past a size-selective barrier, ED uses ion-exchange membranes to channel cations and anions out of the feed under electrostatic force. The result is a low-TDS dilute stream and a high-TDS concentrate stream, plus a third outgoing stream — the electrodes' own electrolyte loop — that an engineer must also account for in the mass balance. ED fits cases where RO recovery stalls, where a specific ion must be removed without taking the rest of the salt profile with it, and where the feed is brackish to mid-salinity rather than full-strength seawater.
The Core Working Principle: Electric Field, Electrodes, and Ion Migration
- Apply a DC potential across two electrodes. An anode (positive) and a cathode (negative) are placed at opposite ends of the stack, and a rectifier supplies the voltage. The electric field established between them is the driving force for everything that follows.
- Pull cations toward the cathode and anions toward the anode. Cations — Na⁺, Ca²⁺, Mg²⁺, K⁺ — migrate in the direction of the negative electrode. Anions — Cl⁻, SO₄²⁻, NO₃⁻, HCO₃⁻ — migrate toward the positive electrode. This is Coulomb's law operating at the ionic scale.
- Let neutral species pass through unaffected. Uncharged organics, dissolved silica that is not ionized, and the water molecule itself are not driven by the field. They remain in the bulk flow, which is why ED is described as selective by charge rather than by size.
- Maintain continuous flow so the streams keep refreshing. Feed enters one set of compartments while dilute and concentrate leave continuously. Continuous operation makes ED suitable for industrial throughput rather than batch processing.
Inside the Membrane Stack: How CEM and AEM Create the Dilute and Concentrate Streams
The physics above is generic — any two electrodes in a salt solution will move ions. The alternation of two membrane types inside the stack turns that physics into separation. An ED cell pair consists of one cation-exchange membrane (CEM) and one anion-exchange membrane (AEM) framing a feed compartment. The CEM carries fixed negative charges and passes only cations while rejecting anions. The AEM carries fixed positive charges and passes only anions while rejecting cations. When the stack repeats this pair — CEM, feed, AEM, feed, CEM — and a DC field is applied, every other compartment becomes ion-depleted (dilute) and the compartments in between become ion-enriched (concentrate). The alternating membrane pattern physically creates the two streams with no moving parts; an engineer reading a CEM/AEM drawing should picture a repeating salt-and-pepper pattern of cells, each becoming either a dilute or concentrate channel depending on the local field direction.
| Membrane | Fixed charge | Passes | Blocks | Faces which electrode | Resulting outlet stream |
|---|---|---|---|---|---|
| CEM (cation-exchange membrane) | Negative | Cations (Na⁺, Ca²⁺, Mg²⁺) | Anions (Cl⁻, SO₄²⁻) | Cathode side of the dilute cell | Dilute (product) |
| AEM (anion-exchange membrane) | Positive | Anions (Cl⁻, SO₄²⁻, NO₃⁻) | Cations (Na⁺, Ca²⁺) | Anode side of the dilute cell | Dilute (product) |
Because both ions must travel through their respective membrane to escape a cell, the separation is selective by charge rather than by molecular size. This distinction is why ED is positioned as a "selective ion removal" technology rather than a generic desalination one.
From Cell Pair to Industrial Stack: What Engineers Actually Specify
An industrial ED skid is the textbook cell pair multiplied many times over. Cell pairs are stacked in parallel hydraulic paths between a single pair of end electrodes; more cell pairs mean higher salt-removal capacity per pass and higher throughput at the same applied voltage. Operating parameters that should be specified at the inquiry stage include the feed TDS window, target product TDS, recovery ratio, flow per cell pair, applied current density, and pressure drop per stage. The buyer should request these from the vendor with the same diligence applied to an RO data sheet, because they govern both capex (number of cell pairs, membrane area, rectifier sizing) and opex (energy, cleaning frequency, membrane life).
Two variants extend the basic architecture. Electrodialysis reversal (EDR) periodically swaps electrode polarity, which inverts the dilute and concentrate channels. The reversal flushes scale, biofouling, and loosely attached organics off the membrane surfaces. Bipolar membrane electrodialysis (BMED) uses a three-layer membrane (cation layer / water-splitting interface / anion layer) to split a dissolved salt into its constituent acid and base — for example, recovering NaOH and HCl from a NaCl-rich sidestream rather than just concentrating it.
Electrodialysis vs Reverse Osmosis: Choosing the Right Separation
ED uses electricity to move ions, while RO uses pressure to filter water. ED separates by charge and can be tuned to target specific ions; RO separates by size and rejection cutoff and treats nearly everything above that cutoff as one stream. For high-salinity feeds such as seawater, RO is the default because ED's electrical energy demand rises sharply with TDS; for brackish feeds, selective ion targets, and brine-minimization duties, ED is often the lower-energy option because no phase change is involved and no high-pressure pump is required. Pretreatment philosophies differ: an industrial RO system is most sensitive to particulates, free chlorine, and scaling indices, while an ED stack is most sensitive to organics, hardness scaling, and microbial growth on the membrane face.
| Parameter | Electrodialysis (ED / EDR) | Reverse Osmosis (RO) |
|---|---|---|
| Driving force | DC electric field | Hydraulic pressure |
| Selectivity basis | Charge of the ion | Size / rejection cutoff |
| Best feed salinity window | Brackish to mid-salinity | Brackish through seawater |
| Phase change required | No | No |
| Primary fouling concern | Organics, scale, microbial growth | Particulates, oxidizers, scaling |
| Typical pretreatment | Filtration to prevent clogging | Multimedia + antiscalant + cartridge |
| Energy driver | Rectifier kWh per m³ of product | High-pressure pump kWh per m³ of permeate |
| Target ions | Selective — specific cations or anions | Bulk — all species above cutoff |
Where ED Fits in an Industrial Treatment Train
On a P&ID, the ED skid sits between a feed-conditioning stage and a downstream polishing or disposal stage. Feed should be screened and filtered so that suspended solids, oil, and large organics do not reach the membrane face; a multimedia or ultrafiltration pre-treatment skid is the typical guard stage in industrial plants. Downstream, the dilute stream is commonly polished: by an EDI polishing stack when ultra-pure water is the target (pharma, semiconductor, power-plant boiler feed), or by a second RO pass when the duty is high-purity reuse with tight conductivity limits. The concentrate stream is a separate engineering decision — crystallization, evaporation, or controlled discharge — and it should not be left to the ED vendor alone because concentrate volume, scaling tendency, and local disposal limits are site-specific. A well-placed ED skid reduces the load on the RO that follows it, which lowers the RO's specific energy and raises overall recovery.
Operating Advantages and Honest Limitations
The advantages include lower energy use than thermal desalination; selective removal of specific ions rather than bulk salt rejection; minimal chemical additives in normal operation; scalability from small to large industrial capacities; and long membrane service life when CIP and reversal schedules are respected. The limitations are consistent: ED is less efficient than RO for high-salinity feeds such as seawater; membranes foul from organics, scale, and microbial growth; capex for membranes and the rectifier/instrumentation envelope is significant; and pretreatment is mandatory. Translated into operator-facing actions, this means scheduled polarity reversal (for EDR units), CIP routines on a defined interval tied to pressure-drop trends, scaling-index monitoring (Langelier or Stiff-Davis) on the concentrate side, and a realistic membrane-replacement line item in the depreciation model. For broader system care beyond ED, the same disciplined approach is laid out in a membrane system O&M guide that applies the same logic to biological and UF skids.
Frequently Asked Questions
What TDS window is an electrodialysis system suited to?
ED is best matched to brackish and low-to-mid salinity feeds. ED is less efficient for seawater desalination due to the high concentration of dissolved salts, which sets the practical upper bound; the lower bound is set by the user's product-water target and by the diminishing return on electrical energy as feed TDS drops toward the product specification.
What pre-treatment does an ED system need?
Feed must be filtered before the membrane stack to prevent clogging. In industrial practice, this typically means a multimedia filter or an ultrafiltration guard stage ahead of the ED skid, plus any chemical conditioning (antiscalant, pH adjustment) dictated by the concentrate's scaling index.
What does an electrodialysis system cost to operate?
Operating cost is dominated by rectifier kWh per cubic metre of product, membrane replacement on a multi-year cycle, CIP chemicals, and concentrate disposal. Site-specific energy use and membrane life vary with feed TDS, temperature, and fouling load, so the buyer should request a vendor-validated model with those four inputs — a useful reference is the 2026 electrodialysis OPEX breakdown, but its numbers must be checked against the actual feed analysis and duty cycle.
How do I size and select an ED supplier for my site?
Size the skid from four inputs: feed TDS, target product TDS, recovery target, and feed temperature; then ask each vendor for the resulting cell-pair count, membrane area, rectifier kW, and concentrate volume. When comparing suppliers, evaluate membrane sourcing transparency (homogeneous vs heterogeneous, who manufactures them), the willingness to run a pilot on the actual site water before purchase, and the after-sales service envelope for spare membranes and CIP chemistry.
