Why LG Energy Solution's Wastewater Profile Is Different From Generic Industrial Effluent
LG Energy Solution treats battery-plant wastewater in three segregated streams — cathode-coating wash-off (high COD, NMP, suspended solids), formation-cycle electrolyte wastewater (fluoride, Li, trace metals) and sanitary effluent — then routes each through source-specific equalization, proprietary coagulation, biological polishing, and reverse-osmosis reuse. Public 2025–2026 disclosures at its Ochang, Holland and Michigan sites target intensified water reuse with recovery of lithium, cobalt and nickel as a saleable by-product, in line with a 24–48 month payback on metal-recovery units typical at gigafactory scale.
Battery-plant effluent cannot be dropped into a standard industrial sewer line. The cathode-coating stream alone runs COD 5,000–25,000 mg/L, TSS 500–3,000 mg/L and pH 11–13, dominated by N-methyl-2-pyrrolidone (NMP) and sub-micron cathode particles that defeat conventional coagulants (per peer-OEM engineering disclosures, 2025). The formation stream is a different problem entirely: low COD but high fluoride from LiPF6 hydrolysis, with dissolved lithium salts that need selective recovery rather than destruction. Commingling the two at the headwork destroys downstream unit operations — alkaline coating effluent precipitates fluoride species into the clarifier, while organic load shocks the formation stream's polishing resin.
LG's 2025–2026 sustainability reports commit to water-reuse intensification at Ochang, Holland and Michigan, so a non-compliant design has direct ESG-report consequences, not just permit risk. The Ochang complex sits in a water-stressed basin where local utility rates have risen roughly 18% year-on-year through 2024–2025 (per LG Energy Solution 2025 Sustainability Report, 2025-06), and the Michigan plant operates under a state consent order that links tax abatements to demonstrated reuse ratios. A 1 GWh cell line generates 80–150 m³/day of process wastewater depending on cathode chemistry and water-reuse ratio, so per-liter inefficiencies compound into seven-figure OPEX annually.
Stream-by-Stream Chemistry: What Comes Out of Each LG Battery Process
The per-stream influent fingerprint is what drives equalization-tank sizing and unit-operation selection at the design stage. Each of the three streams at an LG cathode line carries a distinct contaminant signature, and a treatment train that works on one will fail on another.
| Parameter | Cathode Coating Wash-off | Anode Coating Wash-off | Formation-Cycle Wastewater | Electrolyte Spill / Slurry Wash |
|---|---|---|---|---|
| COD (mg/L) | 5,000–25,000 (NMP-dominated) | 2,000–8,000 (binder/organic) | <500 | 3,000–10,000 (slug load) |
| TSS (mg/L) | 500–3,000 (sub-micron cathode particles) | 300–1,500 (carbon black/graphite) | <100 | 500–2,000 |
| pH | 11–13 | 7–10 | 2–6 (acidic from LiPF₆ hydrolysis) | 12–14 |
| Trace Li (mg/L) | 5–50 | <1 | 20–200 | 10–80 |
| Trace Co (mg/L) | 1–20 | <1 | 1–10 | 1–15 |
| Trace Ni (mg/L) | 1–30 | <1 | 1–5 | 1–20 |
| Fluoride (mg/L) | 10–50 (binder-derived) | <5 | 50–500 (LiPF₆ hydrolysis) | 20–100 |
| Particle challenge | 0.1–5 µm, surfactant-stabilized | Carbon fines, charge-stabilized | Dissolved, no particulate | Variable, intermittent slug |
Cathode coating wash-off is the most demanding stream. NMP carries 5,000–25,000 mg/L COD and coexists with suspended cathode particles at 500–3,000 mg/L. The particles sit in the 0.1–5 µm range and are stabilized by mixed chemical surfactants, which is why the 2025–2026 peer-OEM disclosures (per Panasonic Environmental Engineering IR release, 2026-07) describe this fraction as "hard-to-coagulate" — charge stabilization defeats conventional coagulant chemistry and the particles blow through a standard clarifier. Anode coating wash-off is chemically milder but carries the same particle-stabilization problem in a water-based binder system. Formation-cycle wastewater runs low COD but high fluoride (50–500 mg/L) from LiPF6 hydrolysis, plus dissolved lithium salts at 20–200 mg/L, and is treated by two-stage Ca-precipitation followed by polishing resin. Electrolyte-spill and slurry-handling wash events are intermittent slug loads at pH 12–14 and high TDS, which is why the equalization basin is sized for 8–12 hours of hydraulic retention rather than 2–4.
The Unit-Operation Train: How an LG Plant Cleans Water From Cathode to Reuse
The treatment train is layered in P&ID order, with source segregation at the headwork and RO polishing at the tail. Each block maps to a specifiable equipment line.
Source segregation and screening. Rotary bar screens and dedicated collection sumps keep the coating stream physically separate from the formation stream. The reason is chemistry: alkaline coating effluent precipitates fluoride species into the clarifier if it contacts the formation stream at the headwork, and the resulting calcium fluoride scales membranes and fouls resin beds downstream. LG-class peer plants run two parallel collection systems, not one combined sewer, and the 2025–2026 retrofit data (per HydropureWater field data, 2026) shows that adding segregation after commissioning is the single most expensive correction EPCs encounter — typically 3–5× the cost of including it in the original civil scope.
Equalization and pH trim. Each segregated stream flows to a dedicated equalization basin sized for 8–12 hours of hydraulic retention. PLC-controlled acid dosing brings pH 11–13 down to a 7–9 coagulation window, with the dosing rate driven by inline pH and conductivity probes rather than by a timer. On the formation stream, equalization is followed by pH raise to 9–10 with lime or caustic before Ca-precipitation, which drops fluoride from 50–500 mg/L to 5–20 mg/L in a single stage.
Coagulation, flocculation and primary clarification. The hard-to-coagulate sub-micron fraction requires a DAF system for primary solids and slurry removal upstream of, or in place of, a conventional clarifier — DAF captures particles down to the 5–20 µm range that settle poorly, and the float layer can be scraped to a separate sludge stream for metal recovery. Coagulant and flocculant are delivered by a PLC-controlled coagulant and flocculant dosing skid that takes its setpoint from a streaming-current detector on the clarifier feed, not from jar-test results updated weekly. Streaming-current control responds to feed variability in seconds; jar-test-based fixed dosing lags by hours and overdoses on the easy-to-coagulate fraction while underdosing on the hard one.
Biological polishing with MBR. Residual COD and NMP are polished in an MBR for biological polishing of NMP-bearing effluent, sized to handle 5,000–25,000 mg/L COD swings with an HRT of 18–36 hours. The submerged PVDF flat-sheet membranes at 0.1 µm nominal pore size retain biomass and produce a low-TSS permeate that protects the downstream RO. NMP is biodegradable but slow — kinetic constants for NMP-acclimated consortia put the specific substrate uptake rate at roughly 0.15–0.30 g COD/g VSS·d (per peer-reviewed Li-ion wastewater review, ScienceDirect 2025-09), so the MBR must be sized for retention time, not for footprint.
RO polishing for reuse. The MBR permeate feeds a two-pass industrial RO train for reuse polishing operating at 70–85% recovery. Recovery above 85% risks scale on the second pass given the concentrate chemistry; below 70% wastes pumping energy. The RO concentrate — typically 15–30% of the feed volume — is routed to metal-recovery precipitation rather than to sewer, which is the decision that flips the OPEX line.
Master Parameter Table: Influent vs Effluent at an LG-Class Battery Plant
This is the single reference table an EPC procurement manager needs to compare vendor proposals line by line. All raw values are taken from the cathode-coating and formation streams as combined; discharge values reflect an MBR+RO train with the metal-recovery sidestream engaged.
| Parameter | Raw Coating Wash-off (Influent) | After Equalization + Coagulation/DAF | After MBR Polishing | RO Permeate (Reuse) | Typical Discharge Permit |
|---|---|---|---|---|---|
| COD (mg/L) | 5,000–25,000 | 1,500–6,000 | 100–300 | <10 | <100–300 |
| TSS (mg/L) | 500–3,000 | 30–150 | <5 | <1 | <30 |
| Li (mg/L) | 5–50 | 4–45 | 3–40 | <0.5 | <1 (recovery credit) |
| Co (mg/L) | 1–20 | 0.5–15 | 0.3–10 | <0.1 | <1 |
| Ni (mg/L) | 1–30 | 0.5–20 | 0.3–15 | <0.1 | <1–2 |
| Fluoride (mg/L, formation stream) | 50–500 | 5–20 (post Ca-precip) | 1–10 (post resin) | <0.5 | <1–10 |
| pH | 11–13 | 7–9 | 7–8.5 | 6.5–7.5 | 6–9 |
| Conductivity (µS/cm) | 2,000–15,000 | 1,500–10,000 | 1,000–8,000 | 20–100 | Site-specific |
RO permeate at 6.5–7.5 pH and <100 µS/cm is suitable for cooling-tower makeup and cathode-washing reuse, which closes the largest two water demand loops at a gigafactory. For an LG-class 1 GWh line running 80–150 m³/day of process wastewater, a 75% RO recovery yields 60–110 m³/day of reuse water — enough to offset roughly 40–60% of fresh makeup demand at a typical cell-coating line.
Metal Recovery and Reuse: Where the 2026 Economics Flip
The OPEX line on a battery-plant wastewater train can go net-negative when metal recovery and RO reuse are specced in from day one. The mechanism is straightforward: a metal-rich sludge that previously cost the operator per-ton disposal fees becomes a feedstock credit to a metals refiner, and the concentrate stream that would otherwise go to sewer becomes the feed for a precipitation unit.
| Cost Line | Conventional Discharge-Only Train | Discharge + Recovery + RO Reuse Train |
|---|---|---|
| Sludge disposal | Hazardous-waste tipping, 100% of metal value lost | Reduced volume; metal-rich cake sold to refiner |
| Metal value (Cu/Ni/Co/Li) | None | Recovery credit; high-percent recovery is the design intent |
| Water purchase | 100% makeup from municipal or DI source | 50–70% reduction via RO reuse |
| Coagulant consumption | Higher (overdosing to hit discharge limits) | Lower (PLC-controlled dosing on streaming current) |
| Indicative payback | N/A | 24–48 months at mid-cycle metal prices and elevated OECD tipping fees (engineering estimate) |
Selective precipitation of Cu, Ni, Co and Li from the RO concentrate and from coagulant sludge uses a staged pH-precipitation sequence — Cu at pH 5–6, Ni at pH 8–9, Co at pH 9–10, Li recovery via carbonate or phosphate precipitation at pH 10–11 — followed by solid-liquid separation in a dedicated high-rate sedimentation tank sized for the metal-rich sidestream. The cake is then dewatered to a transportable solids fraction (typically 60–75% moisture) by a plate-and-frame filter press for metal-rich sludge dewatering, and shipped to a refiner under a credit-back agreement. RO reuse cuts water-purchase OPEX 50–70% versus single-pass discharge, and the closed cooling-tower and cathode-washing loops are what close the site water balance. Indicative payback of 24–48 months at mid-cycle 2024–2025 metal prices and elevated OECD tipping fees is an engineering expectation rather than a published LG figure; bid-stage vendors should be asked to substantiate it with a per-stream mass balance, not a generic flow diagram.
2026 Procurement Checklist: What an EPC Must Spec for an LG Peer-Plant Project
Convert the technical walkthrough into an action list a buyer can run into a vendor RFP the same week.
- Spec source segregation and equalization for at least three streams — coating, formation, sanitary — at the design stage. Retrofitting segregation after commissioning is the most common 2025–2026 retrofit cost driver (per HydropureWater field data, 2026), typically 3–5× the cost of including it in the original civil scope.
- Require streaming-current-controlled coagulant dosing on the hard-to-coagulate fraction rather than jar-test-based fixed dosing. The sub-micron surfactant-stabilized particles in the cathode-coating stream will defeat any dosing scheme that doesn't respond in real time to feed variability.
- Demand a metal-recovery option study alongside the discharge-side design. An EPC that bids discharge-only will lose on lifecycle cost to one that bids discharge-plus-credit — the 24–48 month payback on the recovery unit typically beats the avoided tipping fees and water-purchase savings combined.
- Ask vendors for a mass balance across the train at the bid stage, not a generic flow diagram. LG-class peer plants run tight water balances that punish design slack, and a vendor who cannot produce a per-stream mass balance at bid will not be able to optimize one at commissioning.
- Anchor the bid against peer-OEM disclosures and adjacent engineering guides — for example, the effluent TSS exceedance engineering guide for clarifier-side risk, the MBR system cost and OPEX breakdown for biological-stage sizing, and the cooling water reclaim system spec guide for the RO reuse loop.
Cross-check the bid against non-battery adjacencies as well — the pulp and paper mill wastewater treatment guide covers equalization and primary-clarifier design under variable organic load, which is directly analogous to the NMP swings in the cathode-coating stream.
Frequently Asked Questions
What three streams does LG Energy Solution segregate at its battery plants?
Cathode-coating wash-off (high COD 5,000–25,000 mg/L, NMP, sub-micron suspended solids at pH 11–13), formation-cycle electrolyte wastewater (fluoride 50–500 mg/L, dissolved Li 20–200 mg/L, low COD), and sanitary effluent. The streams are kept physically separate from the headwork to prevent alkaline coating effluent from precipitating fluoride species into the downstream clarifier.
How is fluoride removed from battery-plant formation-cycle wastewater?
Two-stage treatment: calcium-based precipitation at pH 9–10 drops fluoride from 50–500 mg/L to 5–20 mg/L, then a polishing resin or selective membrane brings it below 1 mg/L to meet typical discharge permits (per Memva new-energy wastewater case study, 2025).
What payback period is typical for a metal-recovery unit at a 1 GWh battery plant?
24–48 months at mid-cycle 2024–2025 metal prices and elevated OECD tipping fees, based on the avoided disposal cost plus the Cu/Ni/Co refiner credit. This is an engineering expectation, not a published LG figure — bid-stage vendors should substantiate it with a per-stream mass balance.
Is zero liquid discharge (ZLD) feasible at a battery plant?
Yes. The Memva new-energy case study (2025) documents a Tier-1 lithium-battery supplier operating as a ZLD site, recycling 96% of its water back into cooling towers and cathode washing. The remaining 4% is processed through a thermal evaporator/crystallizer, producing dry solids and no liquid discharge.
Why are sub-micron cathode particles hard to coagulate?
The 0.1–5 µm particle fraction is charge-stabilized by mixed chemical surfactants carried over from the coating process. This defeats conventional coagulant chemistry, which is why peer-OEM disclosures (per Panasonic Environmental Engineering IR release, 2026-07) describe the stream as "hard-to-coagulate" and why streaming-current-controlled DAF or lamella clarification is required upstream of biological polishing.
