Why CATL Mexico’s ZLD Capacity Is Not Yet a Verified Number
The available research does not establish a verified ZLD capacity in m³/day for CATL’s planned Mexican battery gigafactory. Capacity must be calculated from site-specific wastewater generation, RO recovery and concentrate flow—not from battery GWh alone. HydropureWater’s 2026 CATL concept indicates that MBR effluent should reach TSS below 1 mg/L and SDI below 3 before RO, with RO recovering 70–80% and sending the remaining 20–30% concentrate to ZLD. The final MVR and crystallizer duty must also cover peak batch discharges, expected recovery, uptime and the final salt-management route.
That boundary matters because the project has a global manufacturing context and a published treatment concept, but no verified Mexico-site wastewater flow or ZLD design flow. Battery cell capacity in GWh does not establish a hydraulic duty. The manufacturing boundary, coating intensity, water reuse, batch practices, recovery performance and excluded flows all affect the battery manufacturing water balance.
The starting inventory is the four-stream treatment concept in HydropureWater’s 2026 CATL article: cathode coating wash water, anode coating wash water, electrolyte salt flush and general plant effluent. These streams should not be treated as one generic industrial wastewater. The CATL battery-factory wastewater framework explains why segregation, solvent recovery and stream-specific chemistry precede the decision on discharge or ZLD.
Wastewater treatment capacity and ZLD capacity are different procurement quantities. Segregation, NMP recovery, chemical pretreatment, MBR and RO handle the upstream wastewater and produce the reusable-water and concentrate streams. ZLD receives the residual RO concentrate, then converts it into distillate, concentrated brine and, if justified, solid salt. A m³/day figure for gross wastewater therefore cannot be used directly as the evaporator capacity.
Build the CATL Mexico Water Balance from the Manufacturing Boundary
CATL must define which manufacturing operations occur inside the Mexican site before applying any wastewater factor. Electrode coating, mixing and drying, electrolyte handling, cell assembly and cleaning can have different water demands and contaminant profiles. If coating is performed by a Tier 1 supplier rather than CATL, the site boundary and incoming wastewater chemistry change; they cannot be assumed away.
The required sequence is to segregate, meter, sample and balance each stream:
- Confirm the inside-factory manufacturing boundary.
- Separate cathode, anode, electrolyte and general plant drainage.
- Identify batch dumps, startup, shutdown and abnormal discharges.
- Separate domestic sewage, stormwater and construction runoff from collectible process wastewater.
- Apply NMP recovery where cathode coating occurs on site.
- Calculate the accepted RO feed after upstream treatment and exclusions.
HydropureWater’s 2026 CATL concept reports cathode wastewater at COD 5,000–25,000 mg/L, fluoride 50–500 mg/L, suspended solids 500–3,000 mg/L and pH 2–4. These are source-specific characterization values, not assumed conditions for every Mexican stream. The concept also identifies roughly 1–5 wt% NMP in cathode coating wastewater and vacuum distillation at 50–100 mbar and 80–120 °C for recovery above 99.5%.
The equalization criterion in the same 2026 concept is 8–24 hours. CATL should apply that criterion only after obtaining hourly and batch-flow data. Coarse protection can be provided with a rotary mechanical bar screen, but screening does not establish flow, equalization volume or ZLD duty.
| Water-balance input | Evidence currently available | CATL data required before sizing |
|---|---|---|
| Cathode coating wash water | COD 5,000–25,000 mg/L; fluoride 50–500 mg/L; TSS 500–3,000 mg/L; pH 2–4 | Flow by batch, NMP concentration, production schedule and batch-dump profile |
| NMP recovery | Approximately 1–5 wt% NMP; recovery above 99.5% at 50–100 mbar and 80–120 °C | Whether coating occurs on site, solvent inventory and aqueous residual flow |
| Anode coating wash water | Stream identified; site-specific flow and chemistry not supplied | Average and peak flow, COD, TSS, binder residues and reuse potential |
| Electrolyte salt flush | Stream identified; fluoride and phosphate associated with LiPF₆ hydrolysis | Flush frequency, discharge volume, fluoride, phosphate and metals |
| General plant effluent | Includes boiler blowdown, HVAC condensate and domestic sewage | Separate flow, temperature, chemistry and permitted destination |
| Equalization | 8–24 hours stated in HydropureWater’s 2026 concept | Hourly peaks, batch dumps, usable storage volume and mixing constraints |
Calculate the ZLD Feed from RO Recovery
The ZLD feed is the RO non-permeate stream, not the gross process wastewater. The governing calculation is:
ZLD feed flow = RO feed flow × (1 − RO recovery)
RO feed flow must be built from collectible process wastewater after accepted upstream losses, stream exclusions and treatment decisions. RO permeate may be reused or discharged depending on site conditions. The remaining fraction becomes the nominal concentrate entering ZLD, subject to the actual recovery achieved under design-feed chemistry.
HydropureWater’s 2026 CATL concept specifies 70–80% RO recovery, which makes the nominal concentrate 20–30% of the RO feed. It does not provide a verified Mexican RO feed flow, so assigning an m³/day ZLD capacity would be an unsupported assumption. Bidders should guarantee recovery using the specified feed chemistry, temperature and fouling conditions, not a clean-water test.
The RO feed also needs membrane-compatible quality. The 2026 concept identifies MBR effluent below 1 mg/L TSS and SDI below 3 before RO. An integrated MBR membrane bioreactor addresses biological treatment and solids separation; an industrial RO system then determines the concentrate quantity that becomes the ZLD design feed.
| Calculation step | Equation or basis | Procurement requirement |
|---|---|---|
| Collectible wastewater | Measured process flow minus verified exclusions and upstream losses | Metered flow records and batch discharge data |
| RO feed | Collectible wastewater accepted for RO | Upstream treatment, reuse and stream-disposition decisions |
| RO recovery | 70–80% in HydropureWater’s 2026 CATL concept | Guarantee at the specified design-feed chemistry |
| Nominal RO concentrate | RO feed × (1 − recovery) | Vendor calculation with temperature, pressure, flux and fouling conditions |
| ZLD design feed | Nominal concentrate plus agreed hydraulic and operating allowances | Peak-flow method, storage basis and exclusions stated |
| RO feed quality | TSS below 1 mg/L and SDI below 3 in the 2026 concept | Integrated MBR-to-RO performance test |
Size MVR and Crystallization for Peak and Variable Loads
MVR and crystallizer capacity must be separated into nominal feed, peak feed, equalization storage and guaranteed daily throughput. Average daily flow will not capture a coating wash batch, electrolyte flush, startup discharge or chemistry excursion. The bid should therefore state exactly how peak flow enters storage, how long peak discharge can persist and whether treatment remains available during cleaning or maintenance.
HydropureWater’s 2026 field-data range for MVR is 20–35 kWh per m³ of distillate. That is an energy-screening value, not a universal consumption guarantee. Mexico-site electricity pricing, power quality, available generation, steam or hot-water integration, cooling-water conditions and condenser capacity must be checked before comparing operating costs.
The treatment sequence in the 2026 CATL concept routes RO concentrate to MVR. The evaporator produces reuse-grade distillate and a concentrated brine reported at 200,000–250,000 mg/L TDS, which feeds a forced-circulation crystallizer. The crystallizer produces a reported mixed Na₂SO₄/NaCl salt cake and a lithium-enriched mother liquor. Any lithium-recovery claim remains conditional: the 2026 academic review identifies lithium, cobalt and nickel recovery as a potential ZLD benefit, but it also warns that unmanaged concentrate can create risks greater than those of the original wastewater.
Final solids capacity must be linked to the authorized salt destination. Cake mass depends on the actual concentrate chemistry, recovery, purge and crystal composition; no supplied source establishes those Mexican inputs. Equipment must therefore be selected around scale, fouling, scaling, cleanout frequency, crystallizer availability and operating reserve rather than a headline production rate.
| ZLD element | Duty to size | Key missing or conditional input |
|---|---|---|
| MVR | Evaporation of variable RO-concentrate flow and production of reuse-grade distillate | RO feed, recovery, feed temperature, peak profile, utility conditions and fouling assumptions |
| Forced-circulation crystallizer | Concentration of MVR brine and controlled salt-cake production | Brine TDS, salt composition, purge, recovery, cooling and authorized residual route |
| Equalization | Buffering of batch discharges and chemistry variation | 8–24 hours is a concept criterion; actual usable volume requires CATL flow data |
| Solids handling | Collection, dewatering, characterization, transport and disposal | Cake moisture, hazardous-waste classification and licensed destination |
Compare Discharge, High Reuse and ZLD Decisions
CATL should compare permitted discharge, high internal reuse and ZLD as distinct treatment architectures, not as interchangeable equipment packages. Each route changes pretreatment, monitoring, energy demand, residual handling, redundancy and compliance exposure. The selected route must be decided before the RO skid is ordered because the concentrate destination affects the whole treatment train.
HydropureWater’s 2026 CATL concept reports an overall water balance of 70–90% reuse and 10–30% solid salt cake. That is a reference concept; it is not a transferable Mexican design balance. CATL must recalculate reuse against actual process demand and recalculate salt generation from the site’s water and contaminant balances.
Permitted discharge may be defensible only where the authorization, receiving-water conditions and industrial-user or pretreatment requirements cover the complete wastewater inventory. High reuse may reduce freshwater demand but requires a defined service and verified water-quality compatibility. ZLD is the conservative route where discharge salinity, residual contaminant risk or receiving-water acceptance remains uncertain.
Alfa Laval’s observation that ZLD can enable water reuse and material recovery is strategically relevant, but it does not supply CATL-specific flow, chemistry or permit data. Any options analysis should include RO-reject management before comparing capital cost. The same discipline applies to a Mexican industrial wastewater capacity assessment: site evidence must support the calculation rather than the project label.
| Decision criterion | Permitted discharge | High internal reuse | ZLD |
|---|---|---|---|
| Primary benefit | Conventional liquid disposal route | Freshwater savings and identified process reuse | Water recovery with a controlled final residual |
| Main compliance exposure | Permit, receiving authority and complete wastewater inventory | Reuse-water quality and operational control | Concentrate, energy, solids characterization and final destination |
| Key equipment | Pretreatment and discharge monitoring | MBR, RO, reuse distribution and controls | MBR, RO, MVR, crystallizer and residual handling |
| Decision condition | Written acceptance for the actual flow and chemistry | Confirmed demand and water-quality compatibility | Liquid discharge unsuitable and brine can be lawfully managed |
Requirements to Turn the Estimate into a Procurement Guarantee
A defensible guarantee requires a defined design basis, not an equipment list. CATL should request twelve months of representative or design-basis data where available, together with production schedules, batch dumps, water-reuse demand, recovery assumptions and chemical-use records. Composite samples alone will not characterize variable coating and electrolyte-flush events.
The influent dataset should cover COD, TOC, fluoride, phosphate, nickel, cobalt, manganese, lithium, suspended solids, oil and grease, conductivity, temperature and pH. HydropureWater’s 2026 CATL framework recommends sampling by production area and operating mode, including batch dumps and startup conditions. This dataset becomes the envelope attached to the MBR, RO, MVR and crystallizer guarantees.
Each bidder should disclose design flow, peak flow, storage assumptions, operating reserve, energy consumption, cleanout frequency, redundancy and performance exclusions. The acceptance protocol should include clean-water checks, individual-equipment tests, integrated run testing, representative sampling and operator training. For chemical handling, the automatic chemical dosing system should be tied to calibrated pH, flow and process chemistry; for sludge, the plate-and-frame filter press performance should be demonstrated with representative hydroxide sludge.
The residual fluoride target below 10 mg/L cited with China GB 30485-2013 and EU IED 2010/75/EU BAT-AEL in HydropureWater’s 2026 concept is source-specific context, not a Mexican permit limit. The Mexican design must use current site permits, receiving-authority requirements and representative testing. Where an authorized residual route is uncertain, ZLD selection and final salt handling remain unresolved commercial and compliance risks.
Frequently Asked Questions
What ZLD capacity is required for CATL’s new battery gigafactory in Mexico?
No verified m³/day figure can be calculated from the available research because the Mexican site wastewater flow, RO feed flow, actual recovery, batch peaks and salt-management route are missing. The calculation must proceed from segregated stream data to RO feed, then multiply RO feed by one minus the guaranteed recovery. MVR and crystallizer duties must then be checked against peak discharge, operating hours and utility conditions.
How much will CATL’s ZLD system cost?
No defensible budget can be calculated from the supplied evidence. A request for quotation should specify average and peak wastewater flow, stream chemistry, RO recovery, storage, residual route, utility conditions, required redundancy and operating reserve. The budget should also cover MBR, RO, MVR, crystallization, chemicals, membrane replacement, sludge and salt disposal, monitoring, operator labor and contingency for production-chemistry changes.
How should CATL select an MVR and crystallizer supplier?
Require the supplier to guarantee distillate production, concentrate quality and solids handling against the specified feed envelope, not clean water. The bid should state electrical consumption, scaling and fouling assumptions, cleanout frequency, availability, redundancy and exclusions. HydropureWater’s 2026 field-data range of 20–35 kWh per m³ of distillate is a screening reference, not a project-specific consumption guarantee.
What site data must be confirmed before equipment procurement?
CATL must confirm representative influent chemistry, production schedules, batch dumps, startup and shutdown flows, reuse demand, collection exclusions, RO recovery, utility availability and the authorized destination for concentrate and salt. The final capacity depends on those inputs together with recovery guarantees, peak-flow treatment, uptime, current permits and the selected disposal or recovery route. If any item remains unresolved, a fixed equipment guarantee would carry an undefined basis.
