Separate the Wastewater Streams Before Choosing Equipment
A data center in Calumbo, Angola, should separate cooling-tower blowdown, sanitary/process wastewater, and construction fill-and-flush water before selecting equipment. Cooling blowdown is a concentrated, chemistry-constrained residual; sanitary/process wastewater generally needs biological treatment; construction flush water may require dedicated assessment and disposal. The three streams should not enter one treatment train without flow, chemistry and acceptance data. Calumbo-specific permits, receiving-system requirements, influent quality, temperature and reuse targets must be confirmed before equipment is sized.
Cooling-tower blowdown is the residual released after evaporation concentrates dissolved salts and minerals in the circulating water. It can also carry corrosion products, biofilm, airborne dust and treatment chemicals. Its treatment objective is usually recovery for cooling-tower make-up, with protection against cooling blowdown silica scaling.
Sanitary/process wastewater is different. Kitchen and greywater, humidification bleed-off, generator coolant and other process flows can contain surfactants, food oils, nitrogen and substantially higher BOD/COD. HydropureWater’s 2026 Luanda benchmark reports that sanitary/process wastewater carries 5–10 times more BOD/COD per cubic metre than cooling blowdown. That supports a separate biological train rather than blending it into the blowdown treatment system.
Construction fill-and-flush water must also remain separate during commissioning. E&E News reports that chemicals used to clean closed-loop piping may not be suitable for wastewater treatment; utilities require testing and may direct the water to an appropriate disposal facility instead.
Calumbo is in Viana municipality, Luanda province, but it is not interchangeable with metropolitan Luanda for permitting. The Luanda Municipal Water Authority process described in the benchmark cannot be assumed to govern a Calumbo discharge point. The Pasig effluent-treatment supplier comparison provides a useful procurement structure, but not Calumbo acceptance criteria.
| Stream | Primary contaminants or risks | Treatment objective | Initial routing decision |
|---|---|---|---|
| Cooling-tower blowdown | Dissolved salts, silica, hardness, corrosion products, suspended solids and treatment chemicals | Produce cooling make-up and control concentrate | Side-stream pretreatment followed by UF and RO or NF where reuse is justified |
| Sanitary/process wastewater | BOD/COD, surfactants, oils, nutrients and variable pH | Biological treatment and documented reuse or discharge quality | Separate biological train, such as an integrated MBR wastewater treatment system where justified |
| Construction fill-and-flush water | Cleaning chemicals, corrosion inhibitors, sediments and unidentified contaminants | Protect the sewer and receiving treatment system | Temporary collection, sampling and written acceptance or off-site disposal |
Build the Calumbo Cooling-Water Balance
The blowdown ratio provides the first quantitative check on the Calumbo design: HydropureWater’s 2026 Luanda benchmark gives the relationship as 1/(CoC − 1). At 4 cycles of concentration, blowdown is 25% of make-up; at 6 CoC, it is 20%. The move from 4 to 6 CoC therefore reduces blowdown by five percentage points, or 20% relative to the 4-CoC discharge—not 50%.
IDE Tech’s 2026 benchmark says a 100 MW facility may require up to 2 million litres of water per day for evaporative cooling. This is a scale indicator, not a Calumbo design flow. The project’s actual requirement must be calculated from its cooling configuration, heat-rejection duty, wet-bulb conditions and operating schedule. A similar method is set out in the Bulawayo data center cooling-blowdown guide, but Calumbo’s climate and system data remain necessary.
The baseline water balance should account for make-up, evaporation, blowdown, drift, humidification bleed, maintenance use, leaks and emergency dumps. HydropureWater’s 2026 Luanda benchmark states that actual blowdown can run 15–30% above theoretical because of unmeasured losses and emergency discharges. Flow meters should therefore be reconciled against evaporation estimates before equipment capacity is fixed.
Raising CoC also changes the control burden. HydropureWater’s 2026 Luanda benchmark warns that above 5–6 CoC, biological growth, scaling and microbiologically influenced corrosion rise sharply unless side-stream filtration and non-oxidant chemistry are already established. Cooling blowdown silica scaling, corrosion chemistry and biocide compatibility must therefore be evaluated together.
| Water-balance term | Measurement or design requirement | Engineering decision supported |
|---|---|---|
| Make-up water | Continuous flow measurement plus source-water analysis | Freshwater demand, pretreatment requirements and reuse value |
| Evaporation | Calculated from heat rejection and measured wet-bulb conditions | Theoretical mass balance |
| Blowdown | Continuous flow, conductivity and automatic level control | Treatment capacity and concentrate volume |
| Drift and humidification bleed | Separate metering where practicable | Determine whether bleed can join blowdown treatment |
| Leaks, maintenance use and emergency dumps | Daily operating log and temporary containment | Reconcile measured discharge with theoretical blowdown |
| Cycles of concentration | Calculated from measured cycles, not assumed as a nameplate value | Balance freshwater savings against scaling and fouling controls |
Select a Treatment Train for Each Calumbo Stream
Cooling blowdown treatment should follow the contaminant pathway: remove solids, protect the membrane, remove dissolved salts and then apply microbial control. Side-stream filtration handles suspended material; DAF or lamella clarification is appropriate where colloids or floatable solids require separation; hollow-fibre ultrafiltration pretreatment protects RO or NF; the selected membrane removes dissolved solids; chlorine dioxide or UV provides final microbial control.
HydropureWater’s 2026 Luanda benchmark gives side-stream filtration at 10–25 µm, hollow-fibre UF at 0.01–0.1 µm with 90–95% recovery, and UV polishing at 30–40 mJ/cm². These are benchmark values for preliminary concept development, not Calumbo guarantees. Feed chemistry, temperature, membrane compatibility and the required reuse quality determine the final specification.
Nanofiltration versus reverse osmosis should be decided by the limiting contaminant. The same 2026 benchmark gives NF recovery of 70–85% when hardness is limiting and RO recovery of 75–80% on silica-rich cooling blowdown. NF is the candidate where calcium and magnesium determine reuse quality; RO is required where silica or total dissolved solids must be removed more effectively. The Cairo data-center treatment guide provides a comparable membrane-selection framework, while the Adana data-center guide illustrates how reuse and discharge constraints affect process selection.
Sanitary/process wastewater requires biological treatment. For near-reuse quality and a compact installation, the 2026 Luanda benchmark describes submerged PVDF MBR at 0.1 µm and a 60% smaller footprint than conventional activated sludge. MBR should not automatically be used for cooling blowdown or construction water: it addresses organics and solids, not the principal dissolved-salt problem.
Construction fill-and-flush water belongs in temporary storage with sampling, laboratory assessment and written discharge acceptance. It should not be assumed acceptable merely because it resembles cooling water.
| Equipment | Benchmarked operating or capacity value | Most appropriate duty | Selection constraint |
|---|---|---|---|
| Side-stream filtration | 10–25 µm | Remove suspended solids before clarification or membranes | Match filtration to measured particle loading |
| DAF | 4–300 m³/h in HydropureWater’s 2026 Luanda benchmark | Remove colloids, oils and floatable solids | Capacity follows Calumbo flow and chemistry |
| UF | 2,000–40,000 L/h; 0.01–0.1 µm; 90–95% recovery in the 2026 benchmark | Protect RO or NF | Validate feed quality and temperature |
| NF | 70–85% recovery where hardness is limiting | Remove hardness and reduce TDS for make-up blending | Confirm silica concentration and scaling potential |
| RO | 75–80% recovery on silica-rich cooling blowdown in the 2026 benchmark | Remove dissolved salts when silica or TDS is limiting | Define concentrate management |
| MBR | Submerged PVDF at 0.1 µm; 60% footprint reduction versus conventional activated sludge in the 2026 benchmark | Treat sanitary/process wastewater biologically | Apply only to the organic-bearing stream |
| UV or chlorine dioxide | UV at 30–40 mJ/cm² in the 2026 benchmark | Final microbial barrier for reuse water | Match disinfection to the reuse specification and residual controls |
Decide Between Reuse, Partial Reuse, and Zero Liquid Discharge
Cooling-tower make-up reuse is the preferred first option because HydropureWater’s 2026 Luanda benchmark identifies it as the highest-value, lower-risk strategy. The engineering question is not whether ZLD is technically possible; it is whether the recovered water value, discharge restriction and concentrate-management obligation justify the additional energy and equipment.
HydropureWater’s 2026 Luanda benchmark gives a preliminary operating rule: if make-up costs below about $1/m³ and discharge is unconstrained, remain at 4–5 CoC and prioritise compliance. If make-up cost rises above about $3/m³ or discharge fees bind, evaluate 6–8 CoC with side-stream filtration and tablet-based non-oxidant chemistry. These thresholds must be replaced with Calumbo’s actual water tariff, discharge conditions and operating risk.
Conventional BWRO generally plateaus at 75–80% recovery on scale-prone cooling blowdown. IDE Tech reports an industrial desalter operating at around 95% recovery with permeate silica of about 1 mg/L. That case demonstrates a controlled salt-precipitation route; it is not evidence that every Calumbo blowdown can achieve the same result.
ZLD should remain an option rather than the default. The 2026 Luanda benchmark reports 95–99% overall recovery for BWRO plus MVC and crystallisation, with MVC energy of 15–25 kWh per 1,000 gallons of distillate. Open Engineering Inc’s 2026 life-cycle study found that its wastewater scenario had about twice the GWP of the freshwater case, with treatment energy accounting for roughly 80% of the difference. Generator-waste-heat availability and grid carbon intensity therefore belong in the comparison. The data center and semiconductor wastewater design guide provides a broader treatment-selection context.
| End-use strategy | Evidence-based decision boundary | Principal risk |
|---|---|---|
| Cooling make-up reuse | Preferred initial option under the 2026 Luanda benchmark | Cooling chemistry, scaling and microbial control |
| Partial reuse | Use validated permeate for a defined duty or blend it with another approved source | Variable feed quality may affect reuse |
| ZLD | Consider where liquid discharge is unacceptable or residual management is controlled | High energy demand and concentrate or solid disposal burden |
Plan Compliance, Sizing, and Supplier Evaluation
The procurement package should begin with the permit that applies to the Calumbo discharge point, not the Luanda permit. HydropureWater’s 2026 Luanda benchmark cites Angola Decree 7/09 screening values of COD below 250 mg/L, TSS below 60 mg/L and pH 6–9. These values are preliminary references only until the responsible authority confirms applicability and the Calumbo receiving-system requirements.
Request at least 8–12 weeks of site characterisation covering conductivity, TDS, silica, hardness, calcium, alkalinity, TSS, COD, BOD, temperature, pH, biocide products and flow variation. The sampling plan should distinguish steady-state blowdown from start-up, shutdown, maintenance and emergency discharge.
Ask each bidder to size every stream separately and disclose design flow, recovery, concentrate volume, redundancy, cleaning strategy, membrane replacement assumptions and power demand. The Angola wastewater equipment supplier selection guide provides a framework for comparing local execution and imported skid risks.
HydropureWater’s 2026 Luanda estimate places 50–500 m³/h turnkey packages at $1.2 million–$8.5 million and flags 15–25% logistics premiums on imported skids, plus seven days of on-site chemical storage. These are regional benchmark figures, not a Calumbo quotation. Power, foundations, access, import logistics, taxes and permit conditions must be priced for the actual site.
The supply contract should include a performance test for permeate quality, recovery, energy consumption, availability and concentrate disposition. Supplier selection should be based on the complete package and measured site performance—not an unsupported turnkey price.
Implement the Project in Controlled Phases
A phased implementation reduces the risk of sizing equipment around an unverified water balance. The first phase establishes facts; later phases release capital only when chemistry, permits and reuse performance support the decision.
- Months 0–3: Establish the make-up, blowdown, evaporation, conductivity, pH and TSS baseline. Reconcile measured flow with the theoretical balance and investigate discrepancies.
- Months 3–6: Repair leaks, tune controls and optimise chemistry. Test whether higher CoC can safely reduce blowdown without unacceptable scaling, biological growth or corrosion.
- Months 6–12: Deploy modular side-stream filtration and solids clarification while discharge permissions and Calumbo acceptance requirements are resolved.
- Months 12–18: Install and commission the selected UF, NF or RO, MBR and disinfection equipment. Blend validated reuse water into the cooling system under controlled operating conditions.
- Conditional ZLD stage: Add MVC and crystallisation only after power availability, generator-waste-heat integration, concentrate treatment and residual disposal have been validated.
Construction fill-and-flush water remains under a separate commissioning procedure throughout: segregate, sample, assess in an accredited laboratory, obtain written acceptance and use off-site disposal if acceptance is refused. The staged data center wastewater implementation guide provides a useful sequence for coordinating these activities.
| Phase | Decision released | Evidence required before proceeding |
|---|---|---|
| Baseline | Measured water balance | Reliable flow, conductivity, pH, TSS and operating records |
| Optimisation | Safe operating CoC | Scaling, corrosion, biological and control assessment |
| Pre-treatment | Feasibility of UF and RO or NF | Representative chemistry and confirmed permit route |
| Reuse | Validated cooling make-up | Permeate quality, recovery, energy and reliability tests |
| ZLD | Residual liquid disposal | Power, waste heat, concentrate and solids-disposal assessment |
Frequently Asked Questions
What treatment is needed for cooling-tower blowdown in a Calumbo data center?
Start with side-stream filtration, then add DAF or lamella clarification if solids or colloids require it, UF for membrane protection, RO or NF for dissolved solids, and chlorine dioxide or UV for final microbial control. The choice between NF and RO depends on whether hardness or silica is limiting. The data center cooling-blowdown treatment guide explains the process sequence, while Calumbo testing determines the final design.
Should a Calumbo data center use MBR for all wastewater?
No. MBR is appropriate for sanitary/process wastewater containing biodegradable organics and solids; it does not replace the dissolved-salt removal required for cooling blowdown. Construction fill-and-flush water should remain separately assessed because cleaning chemicals and unidentified contaminants may prevent sewer acceptance.
How much does wastewater and cooling blowdown treatment cost for a data center in Angola?
There is no defensible Calumbo price without stream flows, chemistry, reuse targets, permit conditions and site logistics. HydropureWater’s 2026 Luanda benchmark gives $1.2 million–$8.5 million for 50–500 m³/h turnkey packages, but that regional range must not be presented as a Calumbo quotation. Request an itemised bid covering equipment, membranes, civil works, power, installation, commissioning, chemicals, spares, concentrate handling and performance guarantees.
What site data is needed to select a UF, NF, or RO system?
Provide at least 8–12 weeks of representative data covering flow, conductivity, TDS, silica, hardness, calcium, alkalinity, TSS, temperature, pH and treatment-chemical history. Also obtain the intended reuse specification, cooling-system chemistry limits and membrane compatibility requirements. These inputs determine pretreatment, recovery and concentrate volumes; they cannot be replaced by a generic package capacity.
Can construction fill-and-flush water enter the municipal sewer?
Only after testing and written acceptance from the authority responsible for the Calumbo receiving system. E&E News reports that some utilities require testing and may direct construction flush water to an appropriate disposal facility because treatment chemicals may not suit the wastewater plant. Until that acceptance exists, segregate and contain the flow and arrange off-site disposal if necessary. The Bulawayo data center blowdown guide provides supporting context for segregating commissioning water.
