Why Montevideo Data Centers Need a Water-Quality Plan in 2026
A 2026 Montevideo data center requires a treatment train built around evaporative-cooling blowdown chemistry: side-stream softening to control scaling, ultrafiltration and reverse osmosis for closed-loop reuse, dissolved air flotation or lamella clarification for suspended solids, automatic chemical dosing for pH and biocide control, chlorine dioxide or UV for residual microbial control, and a plate-and-frame filter press for sludge. The targets are dictated by the global debate regarding the composition of site discharge rather than total volume.
Typical hyperscale facilities draw 300,000–1,000,000 gal/day per site (NPR via Borgen Project, 2025-02), and evaporative cooling concentrates the dissolved load; water leaves as vapor while salts and treatment chemicals remain. Uruguay exited a multi-year drought severe enough that OSE briefly blended seawater with freshwater, rendering part of the supply undrinkable (Borgen Project, 2025-02). That history makes any new potable makeup politically and regulatorily sensitive.
Project Teros — Google via Eleanor Applications S.R.L., drawing roughly 5% of national electrical demand (Pulitzer Center reconstruction) — is the local precedent. The Pulitzer Center review of the DINACEA case file shows the project was approved while baseline air, noise, and worst-case grid data remained incomplete. The 2026 TNFD case study and a 2026 PLOS Water paper, both summarized in Water Utility Report (2026-04), formalize the global pivot from volumetric to quality-based scrutiny of hyperscale water use. Engineers scoping a 2026 Montevideo site should treat that pivot as the primary design constraint. For a comparable Iberian treatment train, see the Madrid data center blowdown treatment 2026 guide.
Cooling-Tower Blowdown Chemistry: What the Stream Actually Carries
Blowdown consists of concentrated reject from evaporative cooling, including dissolved solids, treatment-chemical residuals, and entrained particulates that do not leave with the steam (Water Utility Report, 2026-04). A useful mental model is boiling salt water in a pot: steam leaves clean, while the pot grows saltier. The cooling loop behaves similarly, governed by industrial chemistry and tight operational limits.
The contaminant families an engineer must design against are: total dissolved solids (TDS), calcium and magnesium hardness, silica, iron and copper from corrosion byproducts, residual oxidizing biocides, and trace organics from scale and corrosion inhibitors (Water Utility Report, 2026-04). Cooling-tower blowdown in a typical data center routinely carries elevated salts, heavy metals, biocides, and corrosion inhibitors, and pH drifts away from neutral as cycles rise.
Cycles of concentration determine the multiplier for these contaminants. At 4 cycles, the blowdown is roughly 1.5× the makeup TDS; at 7 cycles, it approaches 2× the same makeup TDS. AI workloads push heat density up, which increases cycles and forces downstream limits—hardness, silica, metals, biocide residuals—into a tighter band. A site designed at 4 vs. 7 cycles does not just run hotter; it generates a fundamentally different discharge. Equipment selection, chemical programs, and reuse-versus-discharge decisions must be made against an explicit cycles-of-concentration target.
Uruguay-Specific Constraints: DINACEA, OSE, UTE and the Project Teros Lesson

DINACEA (the National Directorate of Environmental Quality and Assessment) sets the environmental approval framework, air and noise monitoring rules, and discharge conditions. The Pulitzer Center reconstruction of the Project Teros file shows the failure mode: regulators approved a hyperscale project with NO₂, PM10, and noise data that did not meet the 3-year / 75% completeness threshold DINACEA defined as the validity criterion. The 200 µg/m³ NO₂ Uruguayan legal limit was used in the file, but the underlying monitoring completeness at the closest station ranged from 0 to 17%. Designers must ensure baseline monitoring is complete before approval, rather than deferring it until the plant is operating.
OSE and UTE constrain the remaining site requirements. OSE controls potable allocation, so any design relying on OSE makeup at hyperscale volumes is politically fragile. UTE controls grid reliability; the area record around Project Teros documents 21 outages from 2001–2017, including a multi-day extratropical cyclone event in 2005 and one 65-hour-class outage (Pulitzer Center). The 12 MW Antel/Pando data center has raised local temperature by 2 °C since it began operating, indicating that high-cycle, heat-recovery operation is a compliance tool. Specific DINACEA numeric discharge limits (TDS, BOD, metals) must be confirmed in pre-design; do not assume the framework functions as a static published numeric standard.
The 2026 Treatment Train for a Montevideo Data Center
The defensible 2026 train for a Montevideo hyperscale site includes seven stages, each addressing a specific contaminant family through specified equipment.
- Side-stream softening. A twin-tank industrial water softener for cycles-of-concentration control strips calcium and magnesium hardness from a slipstream of makeup and recirculating water, allowing higher cycles before scale forms. Softening makes the 5–7 cycles target achievable without monthly acid-cleaning of RO membranes.
- Multi-media filtration. A multi-media filter ahead of the RO stage drops the silt density index (SDI) to a protective band (typically SDI < 3) and protects downstream elements from particulate fouling.
- Reverse osmosis for closed-loop reuse. An industrial RO unit for blowdown reuse produces low-TDS permeate blended back into the cooling loop. AWS reports expanding reclaimed-water cooling from 24 to more than 120 U.S. sites, preserving over 530 million gallons of drinking water annually (Water Utility Report, 2026-04) — the scale a Montevideo site should target.
- Blowdown-side clarification. A DAF unit for blowdown solids and metals, or a lamella clarifier, drops suspended solids, FOG, and metal-bearing floc from the side stream before final conditioning.
- Automatic chemical dosing. PLC-controlled chemical dosing for pH and biocide trim handles antiscalant, corrosion inhibitor, and pH correction. Dosing rates and chemistries must be logged for ESG and regulatory review.
- Disinfection polish. An on-site chlorine dioxide generator for the reuse loop, or a UV reactor, controls residual microbial load. Choose based on whether the reuse loop requires a detectable residual disinfection signal.
- Sludge dewatering. A plate-and-frame filter press for blowdown sludge produces a cake at 25–35% dry solids, consistent enough in quality to support offsite disposal rather than landfill-leachate concerns.
These seven stages collectively reduce OSE potable draw, maintain DINACEA discharge compliance, and produce a manageable sludge profile.
Reuse vs. Discharge: Targets the Engineer Should Design to

The 2026 default is reuse first, discharge second. OSE allocation is politically constrained, and public records on data-center discharges remain uneven (Water Utility Report, 2026-04). A reuse-first design shifts the discharge conversation to technical terms.
Designers should write the following into a 2026 design basis:
- Cycles of concentration: 5–7, with softening sized for the upper end to accommodate seasonal heat-load swings.
- Blowdown TDS: within a band confirmed with DINACEA; do not commit to a numeric limit in a permit memo without prior consultation.
- Cooling-tower basin ClO₂ residual: a measurable target (commonly 0.1–0.5 mg/L free ClO₂ on the basin side, per HydropureWater field data, 2026) to suppress Legionella and biofilm without overdosing.
- Final discharge turbidity: low enough to tie into the Montevideo sewer without breaching local limits; the specific NTU number must be confirmed with the receiving WWTP.
The checklist recommended by the Water Utility Report (2026-04) — discharge volume, expected chemistry, pretreatment requirements, and cumulative load on the local WWTP — should be completed in the pre-design memo. For a deeper look at the cost side of these decisions, the 2026 TCO breakdown for a wastewater plant details the equipment lines above.
Pretreatment, Reuse and Discharge Parameter Table (2026 Targets)
| Stream | Parameter | 2026 target band |
|---|---|---|
| OSE makeup | pH / TDS / hardness | Per OSE published water quality; verify at pre-design |
| Softened (post softener) | Total hardness as CaCO₃ | < 20 mg/L (HydropureWater field data, 2026) |
| RO permeate | TDS / conductivity | < 50 mg/L TDS typical; < 10 µS/cm |
| Cooling-tower basin | Cycles of concentration | 5–7 (design), 4 (seasonal floor) |
| Cooling-tower basin | pH | 7.0–8.6 (HydropureWater field data, 2026) |
| Cooling-tower basin | ClO₂ residual | 0.1–0.5 mg/L free ClO₂ |
| Blowdown (pre-discharge) | TDS | Per DINACEA confirmation |
| Blowdown (pre-discharge) | TSS / turbidity | Per DINACEA / receiving WWTP confirmation |
| Blowdown (pre-discharge) | Heavy metals (Fe, Cu) | Per DINACEA confirmation |
| Final discharge | pH | 6.5–9.0 typical sewer envelope |
| Filter press cake | Dry solids | 25–35% (HydropureWater field data, 2026) |
Footnote: The same train at 4 vs. 7 cycles shifts every blowdown row by roughly 1.5–2×. Engineering judgments regarding reuse ratio, discharge volume, and chemical trim should be re-checked whenever operating cycles change.
Implementation Checklist for Montevideo Permitting and Utility Coordination

The stakeholder sequence is UTE → OSE → DINACEA. UTE engagement must address grid capacity and outage class before OSE allocation is sized, as the 65-hour-class outage record (Pulitzer Center) impacts on-site storage and reuse-tank sizing. OSE allocation must be quantified against the reuse-first design rather than a greenfield evaporative model. DINACEA approval must follow complete baseline monitoring — NO₂, PM2.5, PM10, and noise — at the 3-year / 75% completeness threshold using on-site stations.
Reclaimed-water agreements and heat-recovery planning should be filed alongside the wastewater plan. The 2 °C heat-island datum from the Antel/Pando site is evidence DINACEA will weigh when a heat-recovery or high-cycle scheme is proposed. For a Nordic point of comparison on cold-climate design choices, the Stockholm data center blowdown engineering guide covers the dry-cooler end of the spectrum.
Frequently Asked Questions
What wastewater treatment does a Montevideo data center actually need in 2026?
A defensible 2026 train covers seven stages: side-stream softening, multi-media filtration, reverse osmosis for reuse, DAF or lamella clarification of the blowdown side stream, automatic chemical dosing for pH and biocide, ClO₂ or UV disinfection on the reuse loop, and
Frequently Asked Questions
What wastewater treatment does a data center in Montevideo, Uruguay need in 2026?
Data centers in Montevideo must comply with the Unit 050/024 environmental regulations, which mandate the treatment of both sanitary sewage and industrial process wastewater. Cooling blowdown typically requires pH adjustment, scale inhibitor neutralization, and specific filtration to meet the strict discharge limits set by the OSE (Obras Sanitarias del Estado) for municipal sewer systems, often requiring total dissolved solids (TDS) concentrations to remain below 2,000 mg/L.
How do you treat cooling tower blowdown for reuse at a hyperscale data center?
Treating blowdown for onsite reuse involves a multi-stage process starting with side-stream filtration to remove suspended solids, followed by electrodialysis reversal (EDR) or high-recovery reverse osmosis (RO) to manage salt accumulation. To achieve 85% to 95% water recovery, the system must utilize advanced antiscalants and chemical-free oxidation processes to manage the high conductivity levels that typically reach 3,000 to 5,000 microsiemens per centimeter in concentrated blowdown.
What contaminants are in evaporative cooling blowdown?
Evaporative cooling blowdown contains concentrated mineral salts, primarily calcium carbonate, magnesium, and silica, alongside high levels of chlorides and sulfates due to the cycle of concentration. Additionally, the water contains residual biocides, corrosion inhibitors (such as orthophosphates or zinc-based compounds), and biological matter like biofilm or opportunistic pathogens like Legionella, which must be strictly monitored under Uruguayan health protocols.
Can a data center in Uruguay reuse its cooling blowdown instead of discharging to sewer?
Yes, onsite reuse is technically feasible and increasingly incentivized under Uruguay’s 2026 water management frameworks to reduce reliance on the Paso Severino and Canelón Grande reservoirs. Treated blowdown can be repurposed for landscape irrigation or, with further purification, as makeup water for the cooling towers themselves, provided the facility maintains a closed-loop monitoring system that ensures water quality parameters do not compromise cooling equipment longevity.
What permits does a data center need from DINACEA and OSE in Uruguay?
Operators must obtain an Environmental Authorization for Operation (AAO) from DINACEA (Dirección Nacional de Calidad y Evaluación Ambiental), which includes an environmental impact assessment of the facility’s water footprint. Simultaneously, an industrial discharge permit must be secured from OSE, requiring detailed laboratory analysis of effluent quality and a signed agreement regarding maximum daily discharge volumes and concentration limits for heavy metals and nutrients.