What Makes Battery Electrolyte Wastewater So Hard to Treat
Electrolyte wastewater is a four-contaminant problem, not a single-stream problem. The lithium hexafluorophosphate (LiPF6) salt hydrolyses the moment a spent cell is cracked open over the equalization tank, the organic carbonates (EC, DMC, EMC) show up as a hard COD load, additives like vinylene carbonate (VC) and fluoroethylene carbonate (FEC) add a fouling layer to downstream membranes, and trace Li⁺ plus residual PF6⁻ salts round out the matrix (AOT Battery, 2024). Conductivity lands in the 15,000–25,000 µS/cm window, and the BOD5/COD ratio sits at 0.05–0.15 — well below the 0.4 threshold where conventional activated sludge is viable without a chemical pre-oxidation step (Zhongsheng field data, 2026).
The hydrolysis mechanism is what makes this wastewater unique. LiPF6 reacts with water in two steps: LiPF6 + H2O → LiF↓ + PF5↑, then PF5 + H2O → PO3F²⁻ + 2HF. Every mole of LiPF6 dumped from a faulty cell can therefore release up to 5 moles of fluoride-bearing species and drop the pH below 2, which is why F⁻ spikes to 1,500–4,000 mg/L in raw wastewater and why any concrete tank without a chemical-resistant liner (HDPE or GRP) starts to soften inside 18 months (AOT Battery, 2024; CNKI patent CN114560542A, 2024-11).
The low BOD/COD ratio is the operational tell. A 0.10 ratio means roughly 90% of the COD is non-biodegradable on the timescale of a standard aeration basin; the organic carbonates are not food for heterotrophs, they are a chemical-oxidation job. That fact dictates the entire downstream train.
2026 Discharge & Reuse Standards the System Must Hit
Two regulatory envelopes govern every tender in this space, and the engineer must pick the tighter one before specifying a single tank. China's GB 39731-2020 sets COD ≤ 500 mg/L, BOD ≤ 300 mg/L, SS ≤ 400 mg/L, F⁻ ≤ 10 mg/L, NH3-N ≤ 45 mg/L, pH 6–9 for direct discharge from electronics manufacturing (per the 2026 China GB 39731 electronics wastewater discharge limits). The EU IED 2010/75/EU BAT-AELs for waste battery treatment set TOC ≤ 60 mg/L and F⁻ ≤ 25 mg/L for discharge to sewer. Where the plant's end-product water is going back to cell rinsing, the World Bank EHS Guidelines for battery manufacturing tighten the reuse spec to conductivity ≤ 250 µS/cm — a number only two-pass RO can guarantee.
| Standard | COD / TOC | F⁻ | NH3-N | SS | Conductivity |
|---|---|---|---|---|---|
| GB 39731-2020 (China, direct discharge) | COD ≤ 500 mg/L | ≤ 10 mg/L | ≤ 45 mg/L | ≤ 400 mg/L | — |
| EU IED BAT-AEL (waste battery treatment, sewer) | TOC ≤ 60 mg/L | ≤ 25 mg/L | — | — | — |
| World Bank EHS (cell-rinse reuse) | — | — | — | — | ≤ 250 µS/cm |
Two operational rules follow. First, the F⁻ ≤ 10 mg/L number is the binding constraint on chemical precipitation — anything above that and the plant fails the discharge test. Second, the ≤ 250 µS/cm reuse target is the binding constraint on RO selection: a single-pass brackish-water unit at 75–85% recovery will not get there for feed above 12,000 µS/cm, which forces a two-pass design.
The Five-Stage Process Flow That Actually Works
The train that survives a 2026 gigafactory audit reads, in order: equalization → Fenton oxidation → defluoridation/chemical precipitation → DAF → MBR → RO → optional MVR. Each stage has a job and a datasheet, and skipping any one of them leaves a parameter the next stage cannot clean up.
Stage 1 — Equalization and solvent skimming. 8–24 hours of HRT buffers the spike when a bad batch drops a few hundred kilograms of electrolyte into the tank. A pH probe in the loop, with NaOH dosing, holds pH above 6 to keep HF from volatilising into the workshop. A top-mounted skimmer pulls the free DMC/EMC layer, which can be sold back as a solvent-grade product at 70–85% recovery (Zhongsheng field data, 2026).
Stage 2 — Fenton oxidation. The acidic Fenton intermediate is the workhorse. Dose FeSO4·7H2O at 500–1,500 mg/L and H2O2 (30%) at 1,500–4,000 mg/L, hold pH 3.0–3.5, 2–4 hours HRT (CNKI patent CN114560542A, 2024-11). This stage achieves 50–70% COD reduction and breaks fluorinated organic chains, but the effluent is acidic and iron-laden — that is the next stage's problem.
Stage 3 — Defluoridation and chemical precipitation. Raise pH to 8.5–9.5 with Ca(OH)2, dose CaCl2 to bring Ca²⁺ to 2,000–3,000 mg/L. Calcium fluoride has a Ksp of roughly 3.9 × 10⁻¹¹, which is low enough to drag F⁻ from 4,000 mg/L down to <20 mg/L in a single pass with a lamella clarifier for the neutralization step. The clarifier underflow becomes the fluoride-bearing sludge that the DAF unit for fluoride-bearing sludge then thickens ahead of dewatering.
Stage 4 — MBR polishing. A submerged PVDF flat-sheet or hollow-fiber membrane at 0.1 µm pore size, MLSS 8,000–12,000 mg/L, SRT 30–60 days. Effluent SS drops below 5 mg/L and residual COD lands in the 200–400 mg/L range — clean enough to feed RO without fouling the membranes. The submerged MBR polishing stage also acts as a biological safety net for any biodegradable COD that survived the Fenton stage.
Stage 5 — RO and optional MVR. A brackish-water RO at 75–85% recovery, with a two-pass RO for cell-rinse reuse when the conductivity target is ≤ 250 µS/cm. The 5–8% reject stream is sent to a mechanical vapour recompression (MVR) evaporator at 0.18–0.24 kWh per litre evaporated for plants chasing zero-liquid-discharge. All chemical feed points are handled by a PLC-controlled chemical dosing skid.
| Stage | Key reagent / parameter | Operating range | Design duty |
|---|---|---|---|
| 1. Equalization | NaOH (pH control) | 8–24 h HRT, pH 6–8 | Buffer & solvent skim (70–85% recovery) |
| 2. Fenton | FeSO4·7H2O / H2O2 30% | 500–1,500 / 1,500–4,000 mg/L, pH 3.0–3.5, 2–4 h | 50–70% COD reduction |
| 3. Defluoridation | Ca(OH)2 + CaCl2 | pH 8.5–9.5, Ca²⁺ 2,000–3,000 mg/L | F⁻ 4,000 → <20 mg/L |
| 4. MBR | PVDF 0.1 µm | MLSS 8,000–12,000 mg/L, SRT 30–60 d | SS < 5 mg/L, COD 200–400 mg/L |
| 5. RO (+ MVR) | Brackish membrane | 75–85% recovery, two-pass for reuse | Permeate < 250 µS/cm; MVR at 0.18–0.24 kWh/L |
Influent and Effluent Parameters Across the Train
Five parameters make or break a tender: COD, F⁻, NH3-N, SS, and conductivity. The mass-balance table below is the reference an engineer can hand to a junior to size pumps, blowers, and chemical tanks without re-running the chemistry.
| Parameter | Raw wastewater | After Fenton | After defluoridation + DAF | After MBR | After RO permeate |
|---|---|---|---|---|---|
| COD (mg/L) | 5,000–25,000 | 2,000–6,000 | 1,500–4,500 | 200–400 | < 30 |
| F⁻ (mg/L) | 1,500–4,000 | 1,200–3,500 | 10–20 | 5–15 | < 1 |
| NH3-N (mg/L) | 200–800 | 150–600 | 150–600 | < 45 | < 5 |
| SS (mg/L) | 300–1,200 | 250–900 | < 50 | < 5 | < 1 |
| Conductivity (µS/cm) | 15,000–25,000 | 14,000–22,000 | 12,000–18,000 | 11,000–16,000 | < 250 |
Two numbers decide the project's fate. F⁻ is the hardest to clear because the Fenton stage barely touches it — the mass balance shows F⁻ going from 1,500–4,000 mg/L down to only 1,200–3,500 mg/L after Fenton. The real F⁻ work happens in Stage 3, where Ca²⁺ chemistry takes it from the thousands to the tens. Conductivity is the second pivot: the train only removes about 30% of total dissolved solids before RO, which is why every plant chasing reuse-grade water must commit to a two-pass RO and cannot shortcut it with ion exchange alone (Zhongsheng field data, 2026). For plants running a related cathode line, the same Fenton-then-neutralize logic carries over to the ternary precursor coprecipitation wastewater treatment stream, where the ammonia load is the controlling parameter instead of F⁻.
Recycle vs. Discharge: 2026 CAPEX and OPEX Compared
The procurement decision comes down to one question: do you pay the wastewater hauler, or do you pay the membrane? Two reference plants of 10 m³/h, both starting from the same equalization + Fenton + DAF front end, give a defensible number for a 2026 tender.
Discharge-only path — Fenton + DAF + MBR only. CAPEX lands at $180K–$320K, OPEX at $0.42–$0.68 per m³ treated, dominated by NaOH, FeSO4, H2O2, sludge disposal, and power. Meets GB 39731-2020 direct-discharge limits without further polishing.
Recycling path — same front end, plus RO and MVR. CAPEX jumps to $600K–$1.1M, OPEX to $0.55–$0.95 per m³. The offset is freshwater savings: at 2026 industrial tariffs of $1.10–$1.60 per m³, a 10 m³/h plant drawing 250 m³/day or more recovers the RO+MVR premium in 2.5–4.0 years (Zhongsheng tender benchmark, 2026).
The hidden line item is the fluoride-bearing sludge. CaF2 sludge from chemical precipitation is classified as hazardous waste in the EU under EWC 19 02 07 and runs $180–$320 per tonne to landfill in 2026. For a 20 m³/h plant that produces 8–12 t/day of wet sludge, that single cost line can flip the payback calculation by 12–18 months and is the first sensitivity variable any credible vendor should be asked to model (per resin adsorption for residual COD polishing cost references, 2026). The same sludge dewatering question drives the choice of a filter press for fluoride sludge dewatering to drop cake moisture below 65% and cut haulage cost.
| Path | CAPEX (10 m³/h) | OPEX per m³ | Freshwater offset per m³ | Payback window |
|---|---|---|---|---|
| Discharge (Fenton + DAF + MBR) | $180K–$320K | $0.42–$0.68 | — | — |
| Recycle (+ RO + MVR) | $600K–$1.1M | $0.55–$0.95 | $1.10–$1.60 saved | 2.5–4.0 years at >250 m³/day |
How to Choose the Right Skid for Your Plant
Run these four questions in a 30-minute meeting and the vendor shortlist usually narrows to two.
1. Daily flow. Below 50 m³/day, a packaged skid with a 6 m container footprint will land faster and cheaper than any civil scope. Above 200 m³/day, the OPEX savings of a custom concrete-tank build, plus the ability to integrate a lamella clarifier for the neutralization step in-line, justify the engineering hours.
2. Local water tariff. Below $0.50/m³, the discharge path wins on simple payback. Above $0.90/m³, the recycle path is the default. In between, run the payback calculator with the fluoride-sludge disposal cost as the sensitivity variable — that single number is the swing factor in 2026.
3. End-product water use. Floor scrubbing, landscaping, or toilet flushing tolerates MBR-only effluent. Cell-rinse water requires RO permeate at < 250 µS/cm, which forces a two-pass RO and a polishing carbon stage.
4. Operator skill level. Fenton + defluoridation + DAF + MBR needs about one hour per day from a skilled operator. Adding RO + MVR needs 2–3 hours per day plus a process chemist on call; if the plant does not have that headcount, the MVR will underperform within a year.
Frequently Asked Questions
What Fenton dosage range is standard for battery electrolyte wastewater? FeSO4·7H2O at 500–1,500 mg/L paired with H2O2 (30%) at 1,500–4,000 mg/L at pH 3.0–3.5 for 2–4 hours HRT, achieving 50–70% COD reduction and breaking fluorinated organic chains before defluoridation (CNKI patent CN114560542A, 2024-11).
Why is MBR alone insufficient for battery electrolyte wastewater? MBR removes SS to < 5 mg/L and polishes biodegradable COD, but it does not touch the F⁻ load of 1,200–3,500 mg/L coming out of the upstream Fenton stage; chemical precipitation with Ca(OH)2 and CaCl2 must precede MBR to drop F⁻ below 20 mg/L.
How does defluoridation chemistry work in this train? Raising pH to 8.5–9.5 with Ca(OH)2 and dosing CaCl2 to 2,000–3,000 mg/L Ca²⁺ precipitates CaF2 (Ksp ≈ 3.9 × 10⁻¹¹), pulling F⁻ from 4,000 mg/L down to < 20 mg/L in a single stage ahead of DAF sludge thickening.
What are the 2026 economics of zero-liquid-discharge versus discharge for a 10 m³/h plant? Discharge-only CAPEX $180K–$320K with OPEX $0.42–$0.68 per m³; recycle path CAPEX $600K–$1.1M with OPEX $0.55–$0.95 per m³ offset by $1.10–$1.60 per m³ freshwater savings, paying back in 2.5–4.0 years for plants drawing > 250 m³/day.
Which 2026 discharge standard applies to electronics-industry battery wastewater in China? GB 39731-2020 sets COD ≤ 500 mg/L, BOD ≤ 300 mg/L, SS ≤ 400 mg/L, F⁻ ≤ 10 mg/L, NH3-N ≤ 45 mg/L, and pH 6–9 for direct discharge from electronics manufacturing, with the F⁻ ≤ 10 mg/L limit being the binding constraint on the design.
