Data Center Wastewater & Cooling Blowdown Treatment in Barranquilla, Colombia (2026 Engineering Guide)

Why Barranquilla changes the water math for a data center

Generic evaporative-cooling guidance underestimates what the Colombian Caribbean coast does to the water budget. Globally, an estimated 75% to 90% of data centers primarily use water-based cooling, with approximately 57% of all facilities drawing cooling water from potable supplies, per the University of Georgia CAES Field Report (S3). That ratio was reached in temperate U.S. climates; Barranquilla layers a hotter, drier dry season, hyperscale build-out and tighter community scrutiny on top of the same baseline.

The CAES source documents the regulatory direction regulators are taking toward large data centers: scrutiny of thermal discharge, chemical additives including biocides and PFAS-bearing coolants, and groundwater drawdown (S3). In Colombia, that pressure maps onto the Ministerio de Ambiente, the Autoridad Nacional de Licencias Ambientales (ANLA), and regional bodies such as the Corporación Autónoma Regional del Atlántico (CRA). The S3 source further notes that specific facility-level water and energy data is often concealed through nondisclosure agreements, and that U.S. state legislatures such as Georgia with Senate Bill 421 (introduced January 2026) are moving to make that data public; coastal Colombian projects adjacent to residential or tourist areas should plan for comparable disclosure pressure.

The economics of blowdown recovery are the same ones Genesis Water Technologies (S2) frames for water-stressed U.S. regions: at 4 cycles of concentration (COC), approximately 25–30% of makeup water leaves the cooling loop as blowdown, so reuse capex is paid back out of saved freshwater, reduced discharge exposure and reduced permit risk. For a 2026 Barranquilla plant, the design therefore has to start with source-water characterization — Magdalena River intake, Caribbean groundwater, or municipal supply — measuring TDS, hardness, chloride, silica and biological loading, before any unit operation is selected.

The data center blowdown treatment train is not a copy of a U.S. or European reference design; it is a Caribbean-tuned envelope. Engineers who skip that envelope step end up re-engineering the train during commissioning.

For project leads comparing this scope to a coastal Peruvian alternative, the parallel Callao data center blowdown guide walks the same logic on a Pacific coast TDS profile.

What is actually in cooling tower blowdown at a coastal site

Total dissolved solids in blowdown typically range 1,200–6,000 mg/L, which is 4–8 times the concentration of the makeup water, per Genesis Water Technologies (S2). The exact position inside that band is set by cycles of concentration: a Barranquilla plant pushed to higher COC to cut freshwater demand will sit at the upper end and bring harder scaling chemistry with it. Makeup water on the Caribbean coast is itself higher in TDS and chloride than the U.S. references that the S2 source draws on, which raises the absolute floor of the band.

Scaling minerals are concentrated along with TDS. Calcium, magnesium, silica and alkalinity are the species that drive antiscalant selection and pretreatment sizing on any recovery loop, per S2. In a closed Caribbean cooling loop with seawater-influenced makeup, the chloride loading also increases corrosion rate on standard metallurgy, which feeds suspended solids through corrosion products and raises the load on downstream membranes.

Treatment chemicals accumulate in the blowdown stream. Biocides, corrosion inhibitors, scale inhibitors and dispersants all concentrate at the blowdown end of the loop (S2). Legacy systems using chromates or high-phosphate chemistries are particularly hostile to reuse or discharge (S2), and a Barranquilla site inheriting an older chemistry should plan for a phased changeover, not a membrane retrofit on top of the legacy program.

Suspended solids of 10–50 mg/L are normal even with basin filtration (S2). The load is a mix of corrosion products, biofilm fragments and airborne particulates — the last of these is non-trivial at a coastal site with consistent onshore wind and periodic Saharan dust events. Biological content — planktonic bacteria, algae and biofilm-forming organisms — must be assumed present in any open evaporative system, which dictates both the disinfection step and the membrane pretreatment choice (S2).

The table is built on S2's range data and the qualitative points S2 attaches to it; numeric values for the Barranquilla-specific rows are inputs a buyer must obtain from a site water characterization, not numbers that exist in the research.

Pretreatment: side-stream filtration and ultrafiltration

Side-stream filtration is the workhorse pretreatment. Self-cleaning spiral filtration units at 10–25 micron rating let the cooling loop run at higher cycles of concentration and they cut blowdown suspended solids to a level that downstream membranes can tolerate, per Genesis Water Technologies (S2). The S2 source positions side-stream filtration as the technology that enables everything else — heat exchanger surfaces stay cleaner, biological activity drops because biofilm attachment sites are minimized, and blowdown suspended solids fall to a level that does not over-foul the recovery train.

Implementation envelope from S2: install filtration capacity equivalent to 1–5% of total circulation flow, with capital costs in the $50,000–200,000 range for a typical data center installation based on flow rates. The exact number for a Barranquilla site is driven by circulation flow, target COC and source-water TSS, not by the envelope itself; the S2 range is the order of magnitude to plan against.

Ultrafiltration is the membrane that protects every downstream asset. Per S2, UF with 0.01–0.1 micron pore size removes suspended solids, bacteria, viruses and high-molecular-weight organics while passing dissolved salts. UF operates at 10–30 psi, reaches 90–95% recovery, and requires chemical cleaning every 1–3 months depending on feed water quality. For a Barranquilla plant, UF also acts as a barrier against the higher biological and particulate load carried in from coastal air, which is why the S2 guidance positions UF ahead of any RO train.

Side-stream filtration and UF are not competing choices; they are sequential. The spiral filter takes the bulk suspended load and keeps the cooling loop clean at high COC; the UF polishes to a membrane-grade feed for RO. A site that skips UF and feeds RO directly from a side-stream filter can expect accelerated membrane fouling and shortened membrane life, especially with the Caribbean biological load.

A hollow-fiber UF system sized to the data center blowdown envelope maps directly onto this duty as the RO pretreatment stage, including the automatic backwash and air-scour cycle that keeps the membrane operating between chemical cleans.

Reverse osmosis and nanofiltration for cooling-tower makeup reuse

Reverse osmosis is the highest-value reuse path. Per Genesis Water Technologies (S2), RO removes 95–99% of dissolved solids, hardness, silica and most treatment chemicals, and produces permeate of 10–50 mg/L TDS suitable for direct return to the cooling tower or for blending with a small makeup stream. The permeate quality number is the one that sets reuse apart from the alternatives.

The constraints are real and they are written into the S2 envelope. RO on blowdown is limited to 50–85% recovery by scaling potential, with operating pressure 150–400 psi to overcome the osmotic pressure of the concentrate. Antiscalant injection is essential — a hybrid formulation combining scale inhibition with catalytic properties is the technology S2 highlights. A 50,000 GPD RO unit treating blowdown installs in the $250,000–500,000 band and runs at $1.50–3.00 per thousand gallons including energy, chemicals, membrane replacement and maintenance (S2).

Nanofiltration is the right answer when hardness, not total TDS, is the bottleneck to higher cycles of concentration. NF operates at 75–150 psi with 70–85% recovery and passes some monovalent ions while removing hardness and sulfate (S2). NF permeate TDS is typically 30–50% of feed water concentration, which is a much weaker reuse stream than RO but a much cheaper one to operate. The decision rule S2 implies is straightforward: high-silica or high-TDS makeup water benefits from RO; calcium/magnesium-limited water may achieve objectives with NF at lower cost.

Cooling-tower makeup specification is itself a design constraint. Per Genesis Water Technologies' alternative-water guidance (S5), cooling-tower makeup is normally specified at 500–1,500 mg/L TDS with suspended solids limited to 10–25 mg/L. An RO permeate at 10–50 mg/L is therefore well below the spec and is normally blended with a small makeup stream — a fact that has direct consequences for the size of the RO unit and the volume of concentrate that downstream processes have to handle.

An industrial RO platform — recovery up to 95% with full PLC automation and minimal chemical consumption — is the unit operation that actually turns blowdown into a closed-loop cooling resource at a Barranquilla site. The permeate TDS range, the recovery range, and the operating-pressure range above are the envelope to defend in front of a project sponsor, and the S2 capital and operating cost band is the order-of-magnitude figure to plan against — site-specific pricing must be requested from the equipment vendor for a firm budget.

Zero liquid discharge and partial ZLD: when MVC and crystallization earn their capex

Mechanical vapor compression is the workhorse thermal step. Per Genesis Water Technologies (S2), MVC systems use mechanical energy to compress water vapor to provide heat for evaporation, producing high-purity distillate suitable for cooling-tower makeup. The envelope from S2: 95–98% recovery from an RO concentrate, distillate below 10 mg/L TDS, 20–30% dissolved-solids residual brine, $1–3 million capital for 10,000–30,000 GPD systems, and 15–25 kWh per 1,000 US gallons of distillate.

A full ZLD train combines RO concentration (50–75% recovery per S2), an MVC or brine concentrator on the concentrate (95% recovery of that concentrate), and a crystallizer to solid salt cake. System-level water recovery reaches 95–99%, with solid waste under 1% of original blowdown volume (S2). The S2 cost band — $3–8 million capital and $5–15 per thousand gallons operating — is the number to plan against; site-specific cost must be confirmed with vendors because feed-water characteristics and energy costs move both numbers.

Partial ZLD is the realistic answer for most 2026 Barranquilla projects. An RO unit at 50–75% recovery feeding an MVC unit at 95% recovery of that concentrate gives 85–95% system-level water recovery without the full crystallizer capex (S2). The remaining concentrated brine is small enough to manage by deep-well injection, off-site haulage to an approved facility, or periodic discharge under a special permit — though each of those routes is its own permit conversation in Colombia.

The right trigger for full ZLD at a Barranquilla site is either a regional prohibition on liquid discharge of high-TDS streams or a water-stress business case that genuinely supports the capex. A generic corporate sustainability target is not an economic justification, and the S2 cost band is high enough that the alternative cost of freshwater plus a real discharge prohibition has to clear it on its own.

Both rows in the table are anchored in the S2 envelope. Specific capex and opex for a given Barranquilla site are inputs the equipment vendor must provide against an actual feed analysis; the S2 numbers are the order of magnitude to plan against.

Cooling chemistry has to be designed with the recovery train, not after it

Legacy chromate or high-phosphate cooling chemistries are particularly hostile to blowdown recovery and discharge compliance, per Genesis Water Technologies (S2). A Barranquilla site inheriting an older chemistry should plan for a phased changeover, not a membrane retrofit on top of the legacy program. The changeover is a permit and operations conversation, not a hardware decision.

Tablet-based, low-phosphate, non-fouling cooling programs — the Genclean-S approach cited in S2 — deliver scale inhibition, corrosion protection and biocide control with chemistries chosen for compatibility with downstream membrane treatment. The S2 source is explicit that the program reduces the chemical load that the recovery train has to handle, which is itself a capex lever: a smaller antiscalant dose on the RO means longer membrane life and longer cleaning intervals.

The right operational picture is one problem, not three. Well-managed cooling chemistry raises cycles of concentration, which reduces blowdown volume, which makes the recovery train smaller and cheaper. Chemistry, hydrology and membrane design are coupled, and they need a single integrated scope. A hydropureWater automatic chemical dosing system delivers the PLC-controlled precision this needs — antiscalant matched to the RO membrane, biocide matched to the UF, and a chemistry program sized for the COC the recovery train is designed around.

Legionella and biofilm control sits in this section because biofilm fragments are part of the suspended-solids load on the recovery membranes (S2, S5) and because biofilm control is part of the public-health story for a coastal data center. A chemistry program that ignores the biofilm question produces a clean cooling loop and a fouled RO train.

2026 Colombian compliance and discharge map for a Barranquilla data center

Cooling-system blowdown is treated as industrial wastewater under Colombian environmental law. Permitting is routed through the Ministerio de Ambiente, the Autoridad Nacional de Licencias Ambientales (ANLA), and the regional autonomous corporation for Atlántico. Water-use concessions involve the Autoridad Nacional del Agua (ANA), and tariff and reuse rules sit with the Comisión de Regulación de Agua Potable y Saneamiento Básico. Specific numeric effluent limits are not part of the source research and must be confirmed with each authority at the time of application.

The CAES Field Report (S3) shows the global direction of travel: regulators are tightening scrutiny of thermal discharge, chemical additives including biocides and PFAS-bearing coolants, and groundwater drawdown. A Barranquilla project should expect questions on all three even if specific local numeric limits are still being updated. The S3 source also documents the U.S. trend toward mandatory disclosure of data center water and electricity use, and Colombian projects adjacent to residential or tourist areas on the Caribbean coast should expect comparable pressure to publish water metrics.

Discharge options to evaluate in 2026 are: (a) sewer discharge to the local wastewater utility with pretreatment, (b) surface-water discharge under a regional permit, (c) off-site haulage of concentrate, and (d) on-site recovery or ZLD. The Black & Veatch project described in S4 is a direct analog: the firm designed a 2½-mile wastewater pipeline for cooling-water blowdown from a data center to the local wastewater treatment plant, with NPDES permit modification and multi-agency coordination on the critical path. The S4 source also describes an alternatives analysis on chlorinated discharge — can the blowdown be discharged directly, can it go to the local wastewater treatment plant without additional treatment, will the company need an on-site water treatment system, can the local interceptor handle the volume — and a separate conceptual design for getting recycled water to the site. That is the same set of questions a Barranquilla project sponsor will be asked, in a different permit vocabulary.

For cross-jurisdictional orientation, the parallel industrial discharge permit guidance for Malaysia in 2026 walks a comparable permit conversation in a different regional framework.

Reuse, partial ZLD or full ZLD: a 2026 decision framework for Barranquilla

Reuse for cooling-tower makeup is the default. Per Genesis Water Technologies (S2), 60–85% recovery of blowdown with RO permeate at 10–50 mg/L TDS is the highest-value reuse path because it has the shortest payback and the smallest operating burden. It is the right answer wherever the local discharge authority will accept a reduced-volume, lower-TDS effluent. At a Barranquilla site, the practical trigger is whether the Corporación Autónoma Regional del Atlántico will accept a low-volume, low-TDS discharge to sewer or surface water under a regional permit — a question the project sponsor has to put to the authority in writing before the equipment list is locked in.

Discharge compliance — TDS reduction, metals removal, biocide neutralization to meet a local permit — is the right answer only when reuse is physically impossible or when the avoided discharge fees plus avoided freshwater are genuinely lower than the recovery capex. The S4 Black & Veatch hyperscale project models this exact trade-off and concludes that the cost of an alternatives analysis, a permit modification and pipeline design can be lower than a full on-site treatment train for a specific site. For a Barranquilla site, that conclusion transfers only after a current regional discharge-fee picture is obtained, and the source research does not provide one.

Full ZLD is justified only where the alternative cost of freshwater plus a real prohibition on liquid discharge is high enough to amortize the capital. Per S2, full ZLD costs $3–8 million with $5–15 per thousand gallons operating cost and reaches 95–99% water recovery; that is the band the alternative cost has to clear. For most 2026 Barranquilla projects, reuse with a contingency plan is the better economic answer, and ZLD enters the design only as an escalation step rather than a starting point.

A practical 2026 sequence for a Barranquilla project is: characterize source water (Magdalena, Caribbean groundwater or municipal supply) for TDS, hardness, chloride, silica and biological loading; size the cooling loop and target cycles of concentration; run a reuse design with hollow-fiber UF followed by an industrial RO; test whether a discharge permit is achievable with the residual concentrate; escalate to partial ZLD if not; escalate to full ZLD only as a last step.

The capex and opex columns are anchored in the S2 envelope. Site-specific pricing is an input the buyer must request from equipment vendors against an actual feed analysis, not a number the research provides.

Frequently Asked Questions

What cooling-tower blowdown TDS should a Barranquilla data center design for in 2026?

The defensible design envelope is 1,200–6,000 mg/L TDS at 4 cycles of concentration, with the upper end reached as cycles of concentration are pushed higher, per Genesis Water Technologies (S2). Makeup water on the Caribbean coast is itself higher in TDS and chloride than the U.S. references the S2 source draws on, so the floor of that band is higher for a Barranquilla site than for a Phoenix or Northern Virginia site. A specific number must be confirmed with a site water characterization before any unit operation is sized.

Is reverse osmosis the right core process for cooling-tower blowdown in Barranquilla?

For cooling-tower makeup reuse, yes. RO delivers 10–50 mg/L TDS permeate, removes 95–99% of dissolved solids, hardness, silica and most treatment chemicals, and runs at 50–85% recovery with 150–400 psi operating pressure, per S2. Nanofiltration is the alternative when hardness is the limiting species, with 75–150 psi operating pressure and 70–85% recovery (S2). The choice between them is set by the makeup water analysis, not by a generic preference.

When does zero liquid discharge actually pay back at a Colombian data center?

Only when the alternative cost of freshwater plus a credible prohibition on liquid discharge is high enough to support $3–8 million of capex and $5–15 per thousand gallons of operating cost, per S2. For most 2026 Barranquilla projects, reuse with a contingency plan is the better economic answer; full ZLD enters the design only as a last escalation step. The freshwater cost and the specific permit position must both be confirmed against the site, because the source research does not provide either number for Colombia.

How should cooling-water chemistry be specified alongside the recovery train?

As a single integrated scope: low-phosphate, non-fouling, biocide-compatible chemistry sized for the cycles of concentration the recovery train is designed around, with the antiscalant program matched to the RO membrane, per S2. Legacy chromate or high-phosphate programs are particularly hostile to blowdown recovery and discharge compliance (S2), and a phased changeover should be planned rather than a membrane retrofit on top of the legacy program. An automatic chemical dosing system delivers the PLC-controlled precision this scope needs.

Related Equipment

• industrial RO platform — specifications, capacity range, and technical data
• multi-media filter — specifications, capacity range, and technical data

Further Reading

• data center cooling tower blowdown recycling specs