Why Dakar Is a Special Case for Data Center Water
Senegal's coastal aquifer serves a population of roughly 3.9 million in the Dakar metropolitan area, and the city's water-stress pledge is now embedded directly in the PAIX Dakar data center design brief. The operator has publicly committed to "saving cooling water consumption in water-stressed regions" (datacenterdynamics.com, 2026) and to 100% renewable-energy construction by 2030 (connectingafrica.com, 2026), which means water efficiency is an upfront engineering constraint, not a CSR add-on. The reference facility sits in Les Mamelles, with 900+ m² of colocation floor, 330 bays, and a design point of up to 1.2 MW of critical power, with the first phase scheduled to come online in 2026 (connectingafrica.com, 2026). Four submarine cables — ACE, MainOne, SAT-3, and SHARE — already land in Dakar, and the 2Africa cable is being added through the Onix facility, which is what makes the site a commercial West African hub but also concentrates the resilience risk on a single municipal supply (datacenterdynamics.com, 2026).
That combination — a water-stressed coastal aquifer, seasonal demand spikes during the harmattan, and SONABEL grid-water hardness typically in the 200-400 mg/L as CaCO₃ range — means discharging untreated blowdown to the municipal sewer is increasingly expensive and, in some jurisdictions, already non-compliant: total dissolved solids limits below 1,500 mg/L are appearing in industrial discharge frameworks and effectively prohibit disposal of concentrated blowdown without polishing (genesiswatertech.com, 2026). For a 1.2 MW Dakar hall, the lowest-hanging operational lever before any new source is developed is to recover cooling-tower blowdown on-site, which is also the only path that keeps the operator's water-stress pledge from becoming a permit issue later.
What Cooling-Tower Blowdown Actually Looks Like in 2026
Cooling-tower blowdown is the bleed stream drawn from the recirculating loop to keep dissolved solids, corrosion byproducts, and biological growth below their scaling and fouling thresholds. The defining operating parameter is cycles of concentration (CoC) — the ratio of dissolved solids in the circulating water to dissolved solids in the makeup. At 4 CoC, blowdown is 25-30% of makeup water; for a 10 million gallon/month facility that is 2.5-3.0 million gallons discharged each month (genesiswatertech.com, 2026). Pushing to 6-8 CoC cuts that volume in half but multiplies the scaling and membrane-fouling risk on the downstream treatment train.
Typical 2026 blowdown chemistry, drawn from genesiswatertech.com (2026) field data: TDS in the 1,200-6,000 mg/L band (4-8× makeup), suspended solids 10-50 mg/L, plus biocides, scale and corrosion inhibitors that carry through. Legacy chromate or high-phosphate cooling-water programs layer an additional disposal hazard on top of that baseline and should be retired before the recovery train is commissioned.
| Cycles of Concentration (CoC) | Expected Blowdown TDS (mg/L) | Expected Suspended Solids (mg/L) | Typical Contaminants of Concern |
|---|---|---|---|
| 3-4 | 1,200-2,500 | 10-30 | Hardness, alkalinity, free chlorine/bromine residuals |
| 5-6 | 2,500-4,000 | 20-40 | Silica, calcium phosphate, biofilm fragments |
| 7-8 | 4,000-6,000 | 30-50 | High silica, Ca/Mg scaling risk, accumulated corrosion inhibitors |
For a Dakar 1.2 MW site, the safe design assumption is 5-6 CoC as the steady-state operating point, with the membrane train sized to tolerate the upper end of that TDS band during shoulder-season operation.
The 2026 Reference Treatment Train for a Dakar Data Center
Cooling-tower makeup reuse is the default scope for any 2026 Dakar reference design; it gives 60-85% recovery of blowdown as high-quality makeup, reduces freshwater draw, and shrinks the discharge stream to a manageable concentrate (genesiswatertech.com, 2026). The train is built in four stages, each independently defensible in a spec.
Step 1 — Side-stream filtration on the recirculating loop, sized at 1-5% of total circulation flow. A self-cleaning 10-25 µm screen (typically spiral-wound or backwashable) drops suspended solids before they concentrate in the basin; installed CAPEX sits in the USD 50,000-200,000 range for a 1.2 MW site (genesiswatertech.com, 2026), and the protection this gives the downstream membrane train justifies the line item on its own. A multi-media filter as side-stream filtration is a viable alternative where the cooling-tower basin already runs clean and a deeper multimedia bed is preferred over a self-cleaning screen.
Step 2 — Ultrafiltration (UF) as RO pretreatment. PVDF hollow-fiber UF at 0.01-0.1 µm pore size removes the residual colloids, bacteria, and biofilm fragments that pass the side-stream screen; it runs at low pressure (10-30 psi), tolerates the blowdown feed without extensive pretreatment, and reaches 90-95% recovery on the UF stage itself with a periodic permeate backwash and a 1-3 month CIP cycle (genesiswatertech.com, 2026). A packaged PVDF ultrafiltration skid for blowdown pretreatment at this duty point is a standard 2026 deliverable.
Step 3 — Reverse osmosis. RO is the workhorse: 95-99% rejection of dissolved solids, hardness, and silica, with permeate at 10-50 mg/L TDS suitable for direct return to the cooling tower as blended makeup (genesiswatertech.com, 2026). Recovery is 50-85% on the RO skid itself, capped by the scaling potential of the concentrate; operating pressure lands at 150-400 psi to overcome osmotic pressure, and antiscalant injection — often a hybrid catalytic formulation — is non-optional on high-silica or high-Ca/Mg feed. An industrial RO system for blowdown recovery at 50,000 GPD is the published 2026 reference size for this class of facility.
Step 4 (optional) — Mechanical vapor compression (MVC) on the RO concentrate, or full zero liquid discharge (ZLD) with a crystallizer. MVC alone hits 85-95% overall system recovery with distillate below 10 mg/L TDS, at USD 1-3 million for 10,000-30,000 GPD of distillate capacity (genesiswatertech.com, 2026). Full ZLD combines RO, MVC, and a crystallizer to reach 95-99% overall water recovery and converts the residual brine to a manageable solid cake, at USD 3-8 million installed for a data-center-scale plant (genesiswatertech.com, 2026). ZLD is a future-proof option, not a 2026 default — but it is the right scope to keep open for any water-positive Dakar site.
| Stage | Function | Key Operating Parameter | Indicative CAPEX (USD) |
|---|---|---|---|
| Side-stream filtration | 1-5% of circulation, 10-25 µm | Continuous, self-cleaning | 50,000-200,000 |
| UF pretreatment | 0.01-0.1 µm PVDF | 10-30 psi, 90-95% recovery | 150,000-400,000 |
| RO | 95-99% rejection | 150-400 psi, 50-85% recovery, 10-50 mg/L TDS permeate | 250,000-500,000 (50,000 GPD) |
| MVC (optional) | Concentrate reduction | 15-25 kWh/1,000 gal distillate | 1,000,000-3,000,000 |
| ZLD (optional) | Full water recovery + crystallizer | 95-99% overall recovery | 3,000,000-8,000,000 |
For a 2026 Dakar project where reuse is not yet contractually required, the realistic base scope is steps 1-3, with the concentrate routed to sewer under a TDS permit. Where reuse is constrained, polishing for TDS compliance is cheaper than an MVC evaporator at this scale.
Sizing the Train to a 1.2 MW PAIX-Dakar-Class Hall
Translating the 1.2 MW / 330-bay benchmark into a flow band is an order-of-magnitude exercise, not a point estimate. A 1.2 MW critical-power facility with a PUE of 1.4-1.6 and a Dakar wet-bulb around 26-28 °C for much of the year will reject 1.5-2.0 MW of heat through the cooling loop; the cooling-tower makeup demand lands in the 30,000-80,000 L/day range depending on ambient conditions and CoC setpoint. At 5-6 CoC, blowdown runs 25-40% of makeup, which puts the design-point RO feed at roughly the 50,000 GPD (≈190 m³/day) mark — the published 2026 reference size (genesiswatertech.com, 2026).
At that size, the 50,000 GPD blowdown RO skid sits at USD 250,000-500,000 installed (genesiswatertech.com, 2026), and the recovery arithmetic is straightforward: 60-85% of the feed returns to the cooling loop as 10-50 mg/L TDS permeate, and the remaining 15-40% goes to sewer as concentrate or, in a ZLD build, forward to the evaporator. Antiscalant choice is the single biggest OPEX driver on the RO stage after energy, and that is where an automatic antiscalant and biocide dosing skid earns its place on the equipment list. Membrane replacement is managed through the RO and UF membrane replacement program sized to a 3-5 year rotation.
For a parallel engineering reference, see the Warsaw data center blowdown treatment 2026 guide and the Manila data center cooling blowdown guide — both walk through the same side-stream-plus-UF-plus-RO logic against a different municipal water profile, which is useful when justifying the train to a Senegalese reviewer who is more familiar with European reference designs. The DOE FEMP on-site wastewater reuse guidance (2026-09) is the cleanest external validation of the reuse business case for institutional owners and is worth citing in any pre-FEED memo.
Senegal Compliance, Discharge, and Reuse Pathway
Senegal's industrial discharge framework sits under the NS 05-061 series of standards administered through the Direction de l'Environnement, with the pre-FEED touchpoints being the local Préfecture, the national water utility, and the operator of the receiving sewer or receiving water body. The buyer should confirm the current limits in writing at the start of FEED rather than assume them at commissioning; the article assumes a discharge envelope of TDS, BOD, residual chlorine/biocide, and selected heavy metals consistent with the NS 05-061 framework, but the exact thresholds should be re-validated for any given site. Where the receiving body is the municipal sewer, some jurisdictions have already moved to <1,500 mg/L TDS limits on industrial discharge, which effectively prohibits disposal of unpolished blowdown (genesiswatertech.com, 2026).
Direct discharge fees in water-stressed regions already reach USD 5-15 per 1,000 gallons (genesiswatertech.com, 2026), and that is the dollar case for reuse over disposal: the avoided discharge cost alone can carry a meaningful share of the RO OPEX. On-site reuse paths in priority order are cooling-tower makeup (the highest-value use, and the one PAIX Dakar is implicitly committing to), landscape irrigation around the perimeter, and toilet flushing in the facility — each with its own regulatory nuance for non-potable reuse in a commercial building under Senegal's framework, and each needing sign-off from the Direction de l'Environnement and the facility's health-and-safety authority.
CAPEX and OPEX Bands for a Dakar 2026 Project
The CAPEX envelope for a 1.2 MW Dakar reference facility, in 2026 dollars, sits in the following bands (genesiswatertech.com, 2026): side-stream filtration USD 50,000-200,000; UF pretreatment USD 150,000-400,000; 50,000 GPD blowdown RO USD 250,000-500,000 installed; MVC USD 1-3 million; and full ZLD USD 3-8 million. OPEX is dominated by energy and membranes on the RO stage, running USD 1.50-3.00 per 1,000 gallons treated; ZLD OPEX climbs to USD 5-15 per 1,000 gallons because of the thermal energy demand (genesiswatertech.com, 2026). Offsetting that, avoided discharge fees of USD 5-15 per 1,000 gallons in water-stressed regions turn the payback math into a discharge-cost-avoidance story, not just a water-cost story.
| Line Item | CAPEX (USD) | OPEX Driver | Indicative OPEX (USD / 1,000 gal treated) |
|---|---|---|---|
| Side-stream filtration | 50,000-200,000 | Solids disposal, maintenance | < 0.50 |
| UF pretreatment | 150,000-400,000 | CIP chemicals, membrane replacement (3-5 yr) | 0.50-1.00 |
| RO (50,000 GPD) | 250,000-500,000 | Energy, antiscalant, membrane replacement | 1.50-3.00 |
| MVC (optional) | 1,000,000-3,000,000 | Thermal energy at 15-25 kWh/1,000 gal | 4.00-8.00 |
| Full ZLD (optional) | 3,000,000-8,000,000 | Energy + crystallizer chemicals + solids disposal | 5.00-15.00 |
| Avoided discharge cost (offset) | — | Permit fees, sewer discharge | 5.00-15.00 (credit) |
As a sanity check on budget allocation, water treatment typically lands at 0.5-1.5% of total facility CAPEX for a colocation build, so a USD 1-1.5M base scope (side-stream + UF + RO at 50,000 GPD) is the line item to defend against the overall USD 80-150M project envelope, with MVC or ZLD added on top only when the water-stress pledge or the permit pathway demands it.
Frequently Asked Questions
What wastewater and cooling blowdown treatment does a data center in Dakar, Senegal need?
The 2026 reference train for a Dakar 1.2 MW colocation hall is side-stream filtration (1-5% of circulation flow at 10-25 µm) followed by UF pretreatment (0.01-0.1 µm PVDF, 90-95% recovery) and reverse osmosis (95-99% rejection, 10-50 mg/L TDS permeate, 50-85% recovery). At a 50,000 GPD design point — the published 2026 reference size for a PAIX-Dakar-class hall — the RO skid lands at USD 250,000-500,000 installed (genesiswatertech.com, 2026), and the train returns 60-85% of blowdown to the cooling loop as makeup.
Is it cheaper to reuse cooling-tower blowdown or to discharge it in Senegal?
For a water-stressed coastal site like Dakar, reuse wins on OPEX even before any water-cost benefit is counted. RO OPEX runs USD 1.50-3.00 per 1,000 gallons treated, while avoided discharge fees in water-stressed regions already reach USD 5-15 per 1,000 gallons (genesiswatertech.com, 2026), and some jurisdictions have moved to <1,500 mg/L TDS discharge limits that effectively prohibit unpolished blowdown. The payback is driven by discharge-cost avoidance, not by the freshwater tariff alone.
When does a data center in Dakar need full zero liquid discharge instead of RO plus sewer discharge?
Full ZLD — RO plus MVC plus crystallizer, 95-99% overall recovery, USD 3-8 million installed, USD 5-15 per 1,000 gallons OPEX (genesiswatertech.com, 2026) — is justified only when the receiving body will not accept concentrate under any permit pathway, or when the operator has committed to a water-positive target. For a 2026 PAIX-Dakar-class hall, the defensible default is side-stream filtration + UF + RO with concentrate to sewer under a TDS permit, with the ZLD scope held as a future-proof option rather than a base build.
Related Equipment
- industrial RO system for blowdown recovery — specifications, capacity range, and technical data
