Start with a complete wastewater stream inventory
A Mexicali data center should develop separate treatment and disposal paths for each liquid stream, because cooling-tower blowdown, glycol-bearing closed-loop waste, utility wastewater, emergency wastewater, generator waste, sanitary sewage and office wastewater have different chemical compositions, regulatory destinations and operational consequences. The first engineering deliverable is therefore a source-by-source mass balance, not a single combined wastewater skid. It must distinguish normal operation, maintenance drainage, upset conditions and emergency storage.
Cooling-tower blowdown is the principal continuous industrial stream. It contains concentrated dissolved minerals, corrosion and scale products, suspended solids, biocides, corrosion inhibitors, scale inhibitors and dispersants. A closed glycol loop does not normally produce cooling blowdown, but drains from outdoor piping, dry coolers and liquid-cooling loops require a separate disposal route. Maintenance wastewater and accidental releases should not be admitted to the cooling-blowdown system unless a documented mass-balance and treatment review proves that their chemistry can be accommodated.
Valicor’s 2026 data-center water-management material identifies cooling-tower blowdown, glycol loops, onsite pretreatment, emergency-wastewater hauling and utility-system liquid wastes as distinct services requiring compliant handling. The project engineer should request the following before preparing an equipment budget:
- Cooling-system type and topology, including whether it is open evaporative, hybrid wet/dry, air-cooled, closed glycol or a combination.
- Design cycles of concentration, average and peak blowdown flow, wet-mode duty, available storage and future expansion provisions.
- Make-up-water source and quality, complete chemical program and intended discharge or reuse destination.
- Separate generators for sanitary sewage, office wastewater, utility wastewater, emergency wastewater, generator waste and glycol-bearing drains.
| Waste stream | Expected characteristics | Initial route decision |
|---|---|---|
| Cooling-tower blowdown | Concentrated salts, scale and corrosion products, suspended solids and treatment chemicals | Discharge, reuse train or ZLD after characterization |
| Glycol-bearing closed-loop waste | Spent glycol and glycol-contaminated drainage; compatibility depends on the fluid and concentration | Segregated collection and approved treatment or disposal |
| Utility wastewater | System-specific; may include filter backwash, wash water and maintenance drainage | Characterize by source before combining |
| Emergency wastewater | Intermittent, potentially high-contaminant event flow | Impermeable storage, response plan and standby treatment |
| Generator waste | Oil-contaminated water, sludge and other generator-service residues | Segregated waste characterization and disposal route |
| Sanitary and office wastewater | Sewage and greywater with organic and solids loads | Sanitary sewer or separately approved treatment |
Define the Mexicali compliance and water-resilience basis
NOM-001-SEMARNAT-2021 is the current industrial-discharge regulation identified in the supplied research for discharges to sewer or surface water. It does not replace the project-specific agreement with the receiving sewer authority. The engineer should obtain current written limits and permit conditions from the authority responsible for the proposed discharge point, including the destination, permitted flow, analytical methods, monitoring requirements and whether reuse requires separate approval.
Written pretreatment limits are necessary because pH, residual chlorine, oil and grease, metals, suspended matter and TDS may be evaluated separately from the national standard. The 2026 Mexican industrial wastewater compliance guide identifies the need to establish both the applicable national basis and written local sewer requirements. It also identifies the current concession allocation, current environmental allocation, written pretreatment limits, published water and ESG commitments, and drought-related operating conditions as inputs to collect before engineering design.
Reuse is a water-resilience strategy as well as a treatment option. In a water-stressed setting, it can reduce freshwater withdrawal and the volume requiring discharge while aligning with the operator’s water-stewardship commitments. However, the supplied research does not establish a particular current Mexicali discharge prohibition, so the project should not assume one. Instead, it should record current source-water conditions, community scrutiny, the operator’s commitments and any drought-related operating conditions in force when the facility is commissioned.
Before selecting technology, request:
- The authority’s written discharge destination, flow limit and quality limits.
- Requirements for reuse connections, sampling, monitoring and cross-connection control.
- Rules for glycol, generator waste, hazardous waste and other segregated residues.
- Current freshwater allocation and applicable drought-related operating conditions.
Characterize cooling blowdown before sizing the treatment train
The supplied 2026 Tijuana data-center blowdown treatment guide reports a reference evaporative cooling tower at four cycles of concentration with blowdown TDS of 1,200–6,000 mg/L and suspended solids of 10–50 mg/L. These are reference values, not Mexicali site measurements, so they are appropriate for preliminary comparison but not final equipment sizing.
Evaporation removes water while leaving dissolved salts behind. Periodic blowdown then controls salt, scale and corrosion accumulation in the recirculating loop. The same 2026 guide cites a 25–30% loss of make-up water as blowdown at four cycles of concentration. That relationship makes operating optimization a first-stage water-reduction measure. It also shows why a minor error in design flow can create a disproportionate contaminant load.
Cooling topology directly controls the treatment duty. A hybrid wet/dry system has a different blowdown profile from a continuously wet tower and must be sized for peak wet-mode blowdown, not its annual average. A closed glycol loop does not generate cooling blowdown during normal operation, so including a large blowdown treatment skid for that configuration would be an error.
Obtain a representative chemical analysis covering TDS, silica, calcium, hardness, alkalinity, suspended solids, pH, residual oxidant, conductivity and temperature. Also obtain the complete cooling program: biocide, scale-inhibitor, corrosion-inhibitor and dispersant products, dosages and application frequencies. These chemicals affect discharge quality and can foul membranes or attack materials. The final treatment specification should therefore be reviewed against the actual multimedia filtration equipment selected for suspended-solids control.
| Input | Why it controls treatment | Required evidence |
|---|---|---|
| Cooling topology and wet-mode duty | Defines when and how much blowdown exists | Approved cooling design and operating modes |
| Cycles of concentration | Sets the blowdown-to-make-up water relationship | Design and operating range |
| Average and peak flow | Sets hydraulic and membrane capacity | Seasonal and upset water balance |
| TDS, hardness, silica and alkalinity | Defines osmotic load and scaling risk | Representative laboratory analysis |
| Suspended solids and residual oxidant | Controls pretreatment and oxidant demand | Representative analysis and chemical program |
| Cooling chemical inventory | Determines membrane and materials compatibility | SDS, dosage and application method |
Choose between discharge, partial reuse, high-recovery reuse and ZLD
The project should select one treatment objective before comparing equipment: compliant treatment and discharge, partial reuse, high-recovery cooling make-up reuse, or zero liquid discharge. Each option has a different balance of capital, freshwater reduction, discharge volume, concentrate liability and operating complexity.
The 2026 Tijuana guide cites 60–85% recovery for cooling-tower make-up reuse, with permeate blended back into the cooling system to reduce fresh make-up and discharge. It also cites RO removal of 95–99% of TDS, hardness and silica, with permeate at 10–50 mg/L TDS reported as suitable for direct cooling-tower make-up. An industrial reverse osmosis system for cooling-water reuse should therefore be evaluated against feed chemistry, scaling limits and the guaranteed make-up-water specification—not headline recovery alone.
IDE Tech’s 2026 cooling-blowdown article reports that conventional brackish-water RO typically plateaus at 75–80% recovery because of scaling. It also describes a supplier-specific MAXH₂O brine-desalter configuration operating around 95% recovery with approximately 1 mg/L silica in the permeate. That configuration combines conservative local RO operation, controlled salt precipitation in a fluidized-bed reactor and dynamic membrane operation; it is not a generic performance guarantee for a Mexicali installation.
The same 2026 Tijuana guide cites 95–99% overall recovery for full ZLD and an 80–90% concentrate-volume reduction for partial ZLD. Partial ZLD still leaves a residual stream requiring deeper treatment, permitted disposal, deep-well injection, haul-off or another approved route. These figures are benchmark references, not guaranteed Mexicali performance. Final recovery depends on feed-water chemistry, pretreatment, scaling controls, equipment configuration and the selected concentrate-management route.
| Treatment outcome | Reference recovery | Primary benefit | Principal unresolved input |
|---|---|---|---|
| Treatment and discharge | Not established by the supplied research | Meets the selected discharge objective | Written authority limits |
| Partial cooling make-up reuse | 60–85% in the 2026 Tijuana guide | Reduces make-up and discharge volume | Peak wet-mode flow and blend limits |
| High-recovery reuse | Conventional BWRO: 75–80%; IDE Tech brine desalter: around 95% in its published example | Further reduces freshwater demand and concentrate volume | Feed chemistry, scaling design and concentrate route |
| Full ZLD | 95–99% overall in the 2026 Tijuana guide | Minimizes liquid discharge | Residual handling, energy demand and long-term operation |
Build a pretreatment and reuse process that protects uptime
A defensible treatment train starts with segregated collection and equalization, followed by cooling chemistry control, side-stream filtration, the fine filtration specified by the membrane supplier, reverse osmosis, permeate storage and controlled make-up blending. Concentrate management completes the process. Equalization must absorb expected flow variation without blending streams whose chemistry or regulatory status has not been evaluated.
Filtration protects the downstream treatment train from suspended solids, colloids and biological growth. The 2026 Tijuana guide describes self-cleaning spiral or multimedia filters for pretreatment and identifies suspended-solids and biological-load reduction as their role. The final filter specification, including micron rating and backwash or cleaning method, should be approved by the membrane and chemical suppliers rather than copied from a generic data sheet.
Chemical injection should be coordinated with the cooling-water program. A PLC-controlled chemical dosing skid can meter coagulants, flocculants, pH adjusters and specialty chemicals, but compatibility must be checked against biocide, scale-inhibitor and corrosion-inhibitor chemistry. The ultrafiltration system for pretreatment is one possible fine-filtration stage; it is not mandatory until the membrane supplier confirms the required configuration.
High recovery requires active scale management. IDE Tech’s 2026 supplier-specific approach uses controlled salt precipitation in a fluidized-bed reactor and dynamic RO operation to manage supersaturation. Bidder submissions should clearly identify which parts are standard equipment, which are proprietary and how the system remains stable when feed chemistry changes.
The concentrate route must be selected before equipment bids are compared. Depending on chemistry and regulatory acceptance, it may require further treatment, permitted disposal, haul-off or crystallization. For mission-critical cooling, specify redundant equipment, protected bypasses, online conductivity and flow monitoring, membrane-cleaning provisions, automatic permeate diversion, leak detection and an operating response for off-specification feed or permeate.
| Process step | Primary function | Supplier information required |
|---|---|---|
| Segregation and equalization | Controls hydraulic variation and protects treatment stages | Working volume, standby capacity and upset response |
| Cooling chemistry control | Limits residual chemical loading | Compatibility review and dosing limits |
| Side-stream and fine filtration | Removes solids, colloids and biological material | Filter rating, cleaning method and differential-pressure limits |
| RO | Produces reusable permeate | Projected flow, recovery, salt passage and membrane life |
| Permeate storage and blending | Matches make-up demand and water chemistry | Storage duration, blending controls and quality instrumentation |
| Concentrate management | Defines the residual route | Composition, volume, destination and regulatory basis |
Estimate capacity and budget from verified project inputs
The treatment train should be sized from the verified site water balance, not from a generic data-center flow. The 2026 Tijuana guide cites a relationship in which a cooling tower at four cycles of concentration loses approximately 25–30% of make-up water as blowdown. That is a reference point, not a universal factor; the final calculation must account for the actual cycles, evaporation, wet-mode operation, drift, makeup and operating variability.
The same guide provides a reference case: 5 million gallons per day of make-up at four cycles produces approximately 1.25–1.5 MGD of blowdown, and a 75% reuse objective implies treating roughly 1.25 MGD before the detailed water balance is checked. This example should be scaled only after confirming that the Mexicali facility has comparable flow and operating conditions.
The cited 2026 equipment benchmarks are side-stream filtration at $50,000–$200,000, a 50,000 GPD RO skid at $250,000–$500,000, MVC at 10,000–30,000 GPD for $1–3 million, and a full data-center ZLD package at $3–8 million. These figures come from the supplied Tijuana guide and are not quotations for a Mexicali project. Currency, taxes, freight, civil works and local installation must be confirmed during procurement.
The estimate should separate equipment CAPEX from tanks and piping, chemical storage, electrical works, instrumentation, redundancy, commissioning, operator training, O&M, membrane replacement and residual handling. The cited full-ZLD OPEX is $5–15 per 1,000 gallons treated. Bidders should separately report energy, antiscalant, cleaning chemicals, filter media, membrane replacement, labour and concentrate disposal.
| Cost or capacity item | 2026 reference in the Tijuana guide | Mexicali verification required |
|---|---|---|
| Side-stream filtration | $50,000–$200,000 | Flow, solids loading, filter configuration and installation |
| RO skid | $250,000–$500,000 for 50,000 GPD | Feed chemistry, recovery, staging and redundancy |
| MVC | $1–3 million for 10,000–30,000 GPD | Concentrate flow, fouling, energy and fouling-control design |
| Full ZLD | $3–8 million | Definition of included equipment, scale and residual disposal |
| Full-ZLD OPEX | $5–15 per 1,000 gallons treated | Energy, consumables, labour and residual-management breakdown |
Issue a supplier package that reduces compliance and performance risk
A bid should be evaluated against guaranteed process outputs and documented design assumptions, not equipment names or headline recovery. Each bidder should state the design feed flow, product-water quality, overall recovery, concentrate composition, concentrate volume and final disposal route. These guarantees must identify whether they apply to average flow, peak flow, worst-case chemistry or a defined operating envelope.
Request complete membrane projections, scaling calculations, chemical-cleaning design, normalized energy and chemical consumption, expected membrane-replacement intervals and the impact of feed-temperature variation. A supplier should also provide details on automatic controls, fail-safe bypasses, permeate diversion, leak detection, remote monitoring, local service support and response times.
Startup performance testing should use site-representative feed water and define acceptance limits for flow, conductivity, silica, recovery and concentrate. The test protocol should also address membrane cleaning, chemical consumption and operation during wet-mode peak conditions.
Commercial comparison should include life-cycle cost, residuals handling, compliance support, operator training, warranties, spare-parts strategy and local service—not purchase price alone. The supplied research emphasizes that the selected treatment system must protect both uptime and compliance; neither objective is fully demonstrated by a laboratory permeate value.
The bidder package should also require each supplier to identify incompatible cooling chemicals, missing feed data and unresolved discharge or concentrate constraints before the bill of materials is frozen. The 2026 Tijuana guide specifically identifies the full cooling chemical program as an input because it drives membrane compatibility and discharge limits. For procurement of the physical treatment train, the scope should also specify the required membranes and filter elements, replacement schedule and approved spare parts.
Frequently Asked Questions
What treatment does a Mexicali data center need for cooling-tower blowdown?
Start with segregated collection, equalization and characterization, then select filtration, chemistry control and a treatment objective. If the blowdown will be discharged, the treatment train must satisfy NOM-001-SEMARNAT-2021 and the receiving authority’s written pretreatment limits. If it will be reused, the 2026 Tijuana guide cites 60–85% recovery for cooling-tower make-up reuse, but the equipment must be sized from verified peak wet-mode flow and feed chemistry.
Does the data center need reverse osmosis if it only plans to discharge the blowdown?
Not automatically. First obtain the written discharge limits, then determine whether filtration, chemistry control, pH adjustment and removal of specific constituents can meet them. If low-TDS make-up reuse is part of the project, an industrial reverse osmosis system for cooling-water reuse may be justified. The decision should compare the required discharge quality with the cost and residual burden of each train.
What is the budget for cooling-blowdown reuse and ZLD in 2026?
The supplied 2026 Tijuana guide cites $50,000–$200,000 for side-stream filtration, $250,000–$500,000 for a 50,000 GPD RO skid, $1–3 million for 10,000–30,000 GPD MVC, and $3–8 million for full data-center ZLD. Full-ZLD OPEX is cited at $5–15 per 1,000 gallons treated. These are benchmarks, not Mexicali quotations. Request a site-specific quotation stating flow, recovery, feed chemistry, concentrate route, exclusions and O&M assumptions.
How should a supplier be selected for a Mexicali data-center treatment system?
Require guaranteed feed flow, permeate quality, overall recovery and concentrate composition, supported by membrane projections and scaling calculations. Check automatic controls, fail-safe bypass, permeate diversion, leak detection, remote monitoring and local service. Then compare life-cycle cost, startup testing, operator training, warranties and residual handling—not just purchase price.
Can sanitary sewage, glycol and cooling blowdown share one treatment system?
Do not combine them by default. Sanitary sewage has an organic and solids load; glycol waste may require a separate disposal route; and cooling blowdown contains a distinct concentrate of salts and treatment chemicals. Segregate collection, characterize each stream and obtain written acceptance for any proposed combined treatment. Valicor’s 2026 data-center water-management material identifies these as distinct liquid-waste streams requiring compliant handling.
