What a Port-au-Prince Data Center Actually Discharges
A sub-10 MW facility in Port-au-Prince produces four liquid streams the design must account for, and they behave differently from a temperate-climate template. Cooling-tower blowdown is the dominant volume and represents 20–40% of intake water according to Genesis Water Technologies' 2025 overview of cooling-tower inefficiency. In a Caribbean humidity regime, that fraction sits at the upper end because free-air economizer hours drop and the tower runs in evaporative mode for more hours of the year. The blowdown is already-conditioned water that has picked up dissolved solids, residual biocides, corrosion inhibitors, and trace metals from corrosion — exactly the parameter set HOBI catalogues in its 2025 community-and-water review of data center effluents.
Closed-loop glycol-water mixes sit behind the scenes until a maintenance drain or dry-cooler leak exposes them. Both ethylene and propylene glycol create high oxygen demand in surface water if released, and ethylene carries higher toxicity (HOBI, 2025). Port-au-Prince sites that serve outdoor plant rooms, dry coolers, or a liquid-to-air heat-rejection loop must therefore plan segregated capture and licensed removal, not a route to the floor drain.
Sanitary discharge from welfare areas is conventional and should be kept on a separate drainage run so it never mixes with the process streams. On-site diesel generators add a fourth, often-forgotten stream: their jacket-water or radiator loop may need its own blowdown during extended grid-outage operation, as flagged in the CRS R49057 FAQ on data center water use (2025). Site engineers should verify the generator's secondary-loop draining points during the audit and route that water through the same treatment train rather than discharging it raw.
Caribbean Constraints That Reshape the Design
A standard "data center water reuse" template assumes a temperate climate, a reliable municipal supply, and a sewer that will accept anything below a few hundred ppm of TDS. Port-au-Prince breaks every one of those assumptions. Bluefield Research, cited in CRS R49057 (2025), estimates that public water systems supply 97% of U.S. data centers; Haitian municipal supply is less reliable, so on-site storage and reuse carry a different risk weight and a different cost-of-water curve. The Port-au-Prince sewer has limited pretreatment capacity, so discharging blowdown with elevated TDS, biocides, or trace metals can exceed typical small-plant limits — a point HOBI (2025) flags for any municipality where a "few large projects can dominate local concern."
Climate forces chemistry changes. Higher wet-bulb temperatures push cooling towers harder, and the biological and scaling risk that Genesis Water Technologies (2025) describes as growing "exponentially above 5–6 CoC without advanced treatment" arrives earlier in the operating envelope. Right-sizing matters: hyperscale RO/IX blowdown-to-makeup trains cost 3–4× more per gallon at 5 MW than at hyperscale and often sit idle when not run by a dedicated treatment operator (Genesis Water Technologies, 2025). For a sub-10 MW facility, modular physical-treatment trains are the economic answer.
Grid-outage frequency is the third Caribbean variable. When the public supply or power drops, the diesel plant carries the cooling load, and operators may shift to once-through or non-tower cooling events that bypass the normal blowdown path. The research characterises this as an "unmeasured loss" category, which means the engineer must size equalisation and bypass routing explicitly. None of the top U.S. search results address this combined Haitian set of constraints, which is why a Port-au-Prince-specific blueprint is needed.
Cooling-Tower Blowdown Characterisation and Target Parameters
The first engineering task is to fix the parameters of the stream you must treat. HOBI (2025) and Genesis Water Technologies (2025) agree that blowdown carries high TDS, suspended solids, scale inhibitors, biocides, and corrosion metals, but neither publication supplies a generic numerical envelope — those depend on makeup-water quality, cycles of concentration, and the chemistry program. The buyer must therefore request the values below from the supplier and from the receiving authority before any equipment is ordered.
| Parameter | Why it matters | Source / what to confirm |
|---|---|---|
| TDS (mg/L) | Sets the discharge ceiling; drives whether reuse is viable | Local sewer or surface-water permit; HOBI 2025 notes elevated TDS as a primary concern |
| Suspended solids (mg/L) | Defines DAF or lamella loading | Site audit, Genesis Water Technologies 2025 |
| Hardness, silica, trace metals | Drives Stage 3 precipitation and pH adjustment | Site audit, Genesis Water Technologies 2025 modular-train logic |
| Residual oxidiser / biocide | Sets UV dose and ClO₂ demand | HOBI 2025; chemistry program data |
| pH and temperature | Limits biological growth and metal solubility | Site audit; receiving authority |
| Blowdown ratio (1/(CoC−1)) | 25% at 4 CoC; 20% at 6 CoC | Genesis Water Technologies 2025 |
| WUE benchmark (context) | 0.47–0.65 Gal (1.8–2.5 L)/kWh industry range | Genesis Water Technologies 2025 |
The arithmetic of cycles is worth pinning down because it is widely misread. At 4 CoC, blowdown is 25% of makeup; at 6 CoC, it is 20% — a 5-percentage-point absolute drop, not a 50% reduction (Genesis Water Technologies, 2025). DOE guidance cited by HOBI (2025) states that raising cycles from 3 to 6 cuts makeup water by approximately 20% and blowdown by approximately 50%. The two figures are consistent once you account for the different baselines, and they tell the engineer the same thing: the second cycle of efficiency gains delivers less than the first, and the chemistry cost rises as the cycle count rises.
Haitian-specific discharge limits (TDS, temperature, residual chlorine, metals) were not found in the research. The buyer must request them in writing from the receiving sewer authority or, for direct surface-water discharge to the Bay of Port-au-Prince, from the environmental ministry. Without those limits, the polish stage cannot be sized.
Recommended Process Train for Sub-10 MW Facilities
For a Port-au-Prince sub-10 MW facility, the train that matches the waste stream and the local discharge reality is staged physical separation followed by targeted polishing, sized as a modular skid. Stage 1 is suspended-solids and FOG removal using a dissolved air flotation system or, where footprint is constrained, a lamella clarifier — the latter trades residence time for plate area and suits sites where civil works are limited. Both protect downstream media and reduce the load on the polish stages. Genesis Water Technologies (2025) recommends physical separation first because the alternative — aggressive chemistry — increases dissolved-solids loading in the blowdown and makes reuse harder, not easier.
Stage 2 is a multi-media filter sized to bring the silt density index into a range that protects any downstream polishing and that gives the UV stage adequate transmittance. Stage 3, optional but recommended where blowdown is destined for cooling-tower makeup, is chemical precipitation and pH adjustment for hardness, silica, and trace metals, fed by an automatic chemical dosing skid. Stage 4 is microbial control: a UV steriliser for low-residual, non-contact disinfection, with a chlorine dioxide generator available where a residual is required by the receiving authority. Genesis Water Technologies (2025) explicitly recommends the two-stage physical-separation-plus-targeted-treatment approach for enterprise and colocation facilities, with modular 100–300 GPM trains as the right size band.
Glycol capture runs on a parallel workstream: containment, leak detection, segregated drain routing, and licensed removal during maintenance drains, as HOBI (2025) recommends. The discharge-route decision — sewer with pretreatment versus surface-water discharge under permit — is the dominant fork and should be settled before Stage 3 chemistry is locked. Haitian surface-water permits are the default path where the municipal sewer cannot accept the volume, and the engineer should confirm acceptance with the receiving authority before specifying polish quality.
Cooling Chemistry First, Reuse Second
Reuse projects fail when the chemistry program is not simplified first. Genesis Water Technologies (2025) is direct on the sequence: aggressive chemical programs — phosphonates, dispersants, corrosion inhibitors, biocides, scale inhibitors in rotating schedules — raise the dissolved-solids load in blowdown and make downstream treatment more expensive; the same source recommends non-oxidising microbiological control such as solid-tablet programs to avoid persistent organic and heavy-metal accumulation in circulating water. Simpler chemistry is what allows a higher effective CoC and a cleaner blowdown for reuse.
This sequencing is the reason most facilities stall between Stages 2 and 3 of a water-efficiency roadmap: they have optimised the existing system and bought the modular equipment, but they have not yet replaced the legacy chemistry program (Genesis Water Technologies, 2025). For a Port-au-Prince site, the practical implication is that any capex submission to the finance committee should include the automatic chemical dosing skid plus the chemistry simplification in the same line item, or the equipment will underperform. If the chemistry change is deferred, the treatment train has to be over-sized, the blowdown reuse rate falls, and the payback slips.
Building the Business Case for a Port-au-Prince Facility
Finance committees in the Caribbean respond to the same total-cost-of-water logic used in temperate markets, but the inputs differ. The reference case in Genesis Water Technologies (2025) is a 15 MW facility spending $200,000 to recover 60% of blowdown (3 million gallons per year), which gives a 6.7-year simple payback on water alone; the same source notes that adding avoided sewer fees, discharge compliance, and risk typically compresses payback to 3–5 years. The volumetric anchor from the same reference case: a 10 MW evaporative-cooled facility at 4 CoC can produce 3.75 million gallons of recoverable blowdown per month.
U.S. context helps when communicating scale to municipal stakeholders, but it should be flagged as such. CRS R49057 (2025) cites the LBNL 2024 report showing U.S. data center direct water use reached 17 billion gallons in 2023, up from 5.6 billion in 2014. IEA, via the same CRS report, estimates that a 100 MW U.S. data center can directly consume roughly the same amount of water per day as 2,600 households. Neither figure is a Haitian number, but both help a finance committee or a port-city environmental authority visualise why a sub-10 MW Caribbean site still needs a deliberate water plan rather than an improvised one.
What the research does not supply is a Haitian-specific water tariff, sewer-acceptance fee, or avoided-discharge-cost value. Those must be requested from the local utility and the receiving sewer authority before the capex paper is finalised. The engineer should treat the 3–5 year payback range as a U.S.-tuned indication, then run the same model with the local tariff and the local avoided-discharge line to produce the figure the finance committee will see.
Implementation Checklist for a Port-au-Prince Site
Stage 1 of the work is measurement, not procurement. Install metering on makeup, blowdown, evaporation, and conductivity; calculate true consumption versus usage; and identify the unmeasured-loss gaps that Genesis Water Technologies (2025) warns typically run 15–30% above theoretical blowdown. Stage 2 is the audit: repair leaks, eliminate once-through cooling events, optimise control sequences to prevent premature blowdown triggers, and add self-cleaning filtration so the tower is not blowing down for clarity alone. The capital line only opens once Stages 1 and 2 are closed out.
Stage 3 is the modular treatment train sized to the 100–300 GPM band appropriate for sub-10 MW facilities, with a plate and frame filter press handling the sludge produced by the DAF or lamella stage, and treatment-plant spares held on site for the consumables that drive uptime in a Caribbean logistics environment. Stage 4 is the discharge-route confirmation: written acceptance criteria from the receiving authority for either sewer or surface-water discharge, locked before chemistry is finalised. Stage 5 is the drought and grid-outage operating plan, defining minimum cooling modes and discharge restrictions during a public-supply or power outage — the same "community-questions" framework HOBI (2025) recommends for new builds. Glycol capture, containment, and licensed removal sit on a separate workstream and should not be folded into the blowdown specification.
Frequently Asked Questions
What does a modular 100–300 GPM blowdown treatment system cost and what is the realistic payback for a Port-au-Prince facility?
The only published capital figure in the research is a U.S. reference case: 15 MW, $200,000 capex for 60% blowdown recovery, giving a 6.7-year simple payback on water alone and 3–5 years once avoided sewer fees, discharge compliance, and risk are included (Genesis Water Technologies, 2025). The buyer must request a Haitian water tariff, a sewer-acceptance fee, and a diesel-runtime-adjusted avoided-discharge value from the local utility before converting that range into a Port-au-Prince number.
How do we choose a supplier for a sub-10 MW Caribbean site, and what lead time should we expect?
Caribbean lead times and limited on-site operator headcount both favour modular skid-mounted systems with pre-wired controls rather than stick-built plants, and Genesis Water Technologies (2025) is explicit that the modular 100–300 GPM band suits enterprise and colocation facilities. The actionable check for a Port-au-Prince buyer is to require a packaged scope that includes the dosing skid, the DAF or lamella unit, the multi-media filter, the UV or ClO₂ stage, the filter press for sludge, and a documented spares holding — all from one integrator — and to confirm a written lead time against the Haitian import pathway before signing.
What is the compliance risk if we discharge blowdown without pretreatment, and what is the lowest-risk baseline?
Discharging blowdown to a sewer or surface water without pretreatment can breach local acceptance limits on TDS, residual chlorine, metals, or temperature, and HOBI (2025) is clear that the risk is real even where the receiving plant is small. The lowest-risk baseline for a sub-10 MW Caribbean facility is a modular DAF plus multi-media filter train ahead of UV or ClO₂ polishing, with the chemistry simplified so the blowdown already meets the receiving limit; the final compliance position must be confirmed with the local authority in writing.
What site-specific data must we obtain locally before finalising the design?
Three numbers are missing from the research and must be sourced in Haiti before specification: the local water tariff, the sewer or surface-water acceptance limits, and the grid-outage profile that drives diesel runtime and therefore the once-through or non-tower cooling events that bypass normal treatment. Without those three inputs, the cycles of concentration, the polish stage, and the equalisation volume cannot be sized with confidence.
