Why Riyadh Ambient Conditions Drive the Treatment Train
Riyadh design ambient of 52°C dry-bulb and 28–30°C wet-bulb at 5% relative humidity forces cooling-tower evaporation to dominate the water balance, which in turn forces cycles of concentration (CoC) down and blowdown up relative to a temperate site. A 2026 hyperscale hall in the Kingdom typically consumes 15–50 million liters per year for cooling (per S1, 2024–2025 EPC bidding data), which converts to a per-MW figure of roughly 150–500 m³/MW·yr at a 1.2 PUE with evaporative or adiabatic cooling. For a 100 MW IT load the worked numbers look like this: evaporation ≈ 6–8% of circulation, drift ≈ 0.001–0.002% of circulation (ASHRAE 188 / CTI ATC-140 range), and mandatory blowdown of 3–7% of circulation at 4–6 cycles of concentration. The governing identity is blowdown % = evaporation % / (CoC − 1), so if evaporation is 7% and the operator holds 5 CoC, blowdown is 7 / 4 = 1.75% of circulation. Vision 2030 hyperscale campuses and NEOM The Line / Oxagon explicitly require renewable-powered, water-efficient cooling with zero-liquid-discharge or near-ZLD tender language (S1), so the polishing train is not optional. The clarifier and DAF upstream of the cooling tower must therefore be sized for the higher solids loading that comes from running close to the scale threshold, which is where a ZSQ dissolved air flotation system and a HydropureWater multi-media filter become the two pieces of equipment that hold the rest of the train together.
| Parameter | Typical value at Riyadh ambient (52°C DB) | Source / note |
|---|---|---|
| Evaporation | 6–8% of circulation | Per S1 2024–2025 EPC data |
| Drift | 0.001–0.002% of circulation | CTI ATC-140 induced-drift range |
| Blowdown at 4–6 CoC | 3–7% of circulation | CoC = TDS(circ) / TDS(makeup) |
| LSI target on circulating water | 0 to +0.5 | HydropureWater field data, 2025 |
| Conductivity before blowdown | < 2,500 µS/cm | HydropureWater field data, 2025 |
| Annual water use, 100 MW IT load | 15,000–50,000 m³/yr | Per S1 2024–2025 EPC data |
Makeup Water Treatment: From Tap or TSE to Cooling-Tower Feed
Three realistic feed sources exist for a Riyadh site: municipal potable from the National Water Company (chloride 250–500 mg/L, TDS 600–1,200 mg/L), treated sewage effluent / TSE from STC-operated plants (phosphate 2–10 mg/L, ammonia 5–20 mg/L, variable TSS 5–60 mg/L), and desalinated RO permeate from Marafiq or SWCC (TDS < 100 mg/L, very low hardness, aggressive LSI). For any of these, the first unit operation is a HydropureWater multi-media filter with anthracite over sand over garnet, which delivers an SDI < 5 to protect downstream RO and a TSS < 2 mg/L feed to the cooling tower. If the source carries oil from a generator hall or FOG from an adjacent food-service block, the EPC should put a ZSQ dissolved air flotation system ahead of the multimedia filter, sized from the catalog range of 4–300 m³/h across 13 standard models. For RO permeate makeup, an industrial water softener in the KJ-WT 1–45 T/h class is required to keep CaCO₃ scale off the fill at 4–6 cycles in Riyadh ambient, where the saturation index climbs fast. For TSE or any source with high microbial load, an on-site ClO₂ generator sized 50 g/h to 20,000 g/h is preferred over chlorine because chlorine forms trihalomethanes rapidly at loop temperatures above 35°C, and PME discharge rules penalize halogenated organics.
Cooling-Loop Side-Stream Filtration and Microbial Control

Side-stream filtration at 5–10% of circulation flow is the EPC's main lever against suspended solids, biofilm, and scale — the three chronic problems named in S1. For a 10,000 m³/h loop, that is 500–1,000 m³/h of side-stream, which is well inside the ZSQ DAF envelope. The recommended biocontrol program is catalytic ozone ahead of the side-stream filter, which oxidizes biofilm precursors and cuts biocide demand; this train runs 10–20% higher CAPEX but 30–50% lower OPEX than chemical-only treatment (per S1, 2024–2025 EPC bidding). Where the client will not accept ozone off-gas risk — typically because of adjacent office space or a tight PME air permit — fall back to a ClO₂ generator on a PLC-controlled dosing loop with residual ORP held at 650–750 mV. Both programs share the same PLC-controlled chemical dosing skid for coagulant, scale inhibitor, and corrosion inhibitor, sized for the actual circulation flow and the LSI target. Control targets for a Riyadh site: LSI 0 to +0.5, CoC 4–6, free ClO₂ 0.1–0.5 mg/L (or ORP ~700 mV), and conductivity < 2,500 µS/cm on the blowdown line. Holding these targets is what keeps the blowdown within RO recovery limits and the sludge within filter-press capacity.
Choosing the Right Clarifier for the Side-Stream
DAF, lamella, and conventional rectangular clarifiers all show up in Riyadh tenders; the EPC's job is to pick by flow, FOG, and footprint, not by habit. A ZSQ dissolved air flotation system removes > 90% TSS at 4–300 m³/h (catalog spec) and floats oil and grease in the same pass. A lamella clarifier runs a surface loading of 20–40 m/h (catalog spec) and is the right call when the side-stream is high-flow, low-FOG, and already chemically conditioned. A conventional rectangular clarifier is the fallback only when the site cannot accept the height of a DAF or the inclined-plate inventory of a lamella. The decision rule the EPC should put in the tender response: if FOG > 20 mg/L or TSS > 150 mg/L on the side-stream, specify DAF; if flow > 200 m³/h and FOG < 20 mg/L, specify lamella; otherwise a multimedia side-stream filter is sufficient. In NEOM and other Vision 2030 tenders, containerized DAF skids are increasingly preferred because they shorten on-site installation to 2–4 weeks (per S1 2024–2025 EPC data) and they ship pre-wired for IKTVA documentation.
| Parameter | ZSQ DAF | Lamella clarifier | Conventional rectangular clarifier |
|---|---|---|---|
| Surface loading rate | 10–25 m/h (hydraulic) | 20–40 m/h (catalog) | 1–2 m/h |
| Footprint at 500 m³/h | ~25 m² (skidded) | ~15 m² | ~250 m² |
| Chemical demand | Low–moderate (polymer) | Moderate (coag + floc) | High |
| TSS removal | > 90% (catalog) | 70–85% | 50–70% |
| Startup time | 2–4 weeks (containerized) | 4–8 weeks | 8–16 weeks |
Blowdown Polishing: RO, Electrodialysis, or Evaporation to ZLD

The regulatory floor for any surface discharge in KSA is set by the Presidency of Meteorology and Environment (PME) and the National Center for Environmental Compliance (NCEC) on TDS, chloride, and temperature, with RCJY at KAEC and NEOM adding explicit ZLD or near-ZLD clauses for hyperscale tenants (per S1, 2024–2025 EPC bidding). Three polishing options are credible: an industrial RO system at 75–95% recovery (catalog: up to 95% recovery), electrodialysis reversal (EDR) for 3,000–10,000 mg/L feed TDS, and mechanical vapor recompression or thermal crystallization to true ZLD. RO is the right answer for roughly 80% of Riyadh sites because the permeate (typically < 50 mg/L TDS) can be recycled as cooling-tower makeup, and the concentrate (15,000–40,000 mg/L) is a small, manageable stream for a downstream brine evaporator. Design basis the EPC should put on the P&ID: blowdown at 2,500–3,500 mg/L TDS at 4–6 cycles, RO permeate < 50 mg/L TDS, RO brine 15,000–40,000 mg/L. For sites with hypersaline intake or high colloidal load, add a PVDF ultrafiltration system at 0.03 µm cutoff, 2,000–40,000 L/h, which the catalog spec accepts at up to 300 ppm turbidity. For very high-purity polish water, an electrodeionization system downstream of the RO can drop resistivity past 15 MΩ·cm, which is useful for the adiabatic cooler make-up line.
| Polishing option | Feed TDS range | Recovery / output | Best-fit Riyadh site |
|---|---|---|---|
| Industrial RO | 2,000–10,000 mg/L | 75–95% recovery, permeate < 50 mg/L TDS | Most sites, permeate to cooling tower |
| Electrodialysis reversal (EDR) | 3,000–10,000 mg/L | 80–90% water recovery, concentrate 30,000+ mg/L | Sites scaling back RO brine volume |
| MVR / thermal crystallization | 15,000–250,000 mg/L | > 95% water recovery, solid salt cake | NEOM / Oxagon ZLD-mandated sites |
Sludge Handling and Reject Management
The sludge line is a real tender scoring item in KSA and most competitor pages ignore it. Clarifier or DAF sludge at 1–3% dry solids (DS) should be thickened in a lamella or DAF thickener, then dewatered in a plate and frame filter press from the catalog range of 1–500 m² filtration area, manual through PLC. Typical cake dryness off a filter press on cooling-tower sludge is 25–35% DS (HydropureWater field data, 2025), which lets the cake go to lined landfill or, at NEOM, to a cement co-processing facility under NEOM's waste-to-material mandate. RO brine from the blowdown polish step is handled separately, in the evaporator or crystallizer line, not in the sludge press — the two streams should not be combined. Sludge yield, for the EPC's mass balance: roughly 0.3–0.8 kg DS per m³ of blowdown for a clarifier, and 0.1–0.3 kg DS per m³ for a DAF side-stream. A lamella thickener ahead of the press is the right call when the clarifier underflow drops below 1% DS.
CAPEX and OPEX Benchmarks for a Riyadh Hyperscale Plant

The ozone-based train runs 10–20% higher CAPEX but 30–50% lower OPEX than chemical-only treatment, with some configurations showing 15–25% lower CAPEX (per S1, 2024–2025 EPC bidding). Riyadh-specific uplifts the EPC should add: +10–15% for IKTVA local assembly, +5–10% for desert-rated enclosures (IP55, 50°C ambient rating), and +5% for HSE, SEC earthing, and Saudi Aramco-style inspection regime. The representative CAPEX envelope below is for a 100 MW hyperscale hall's full water scope — makeup, cooling-loop side-stream, blowdown polish, and sludge dewatering — expressed as an indexed band so the reader can scale to their IT load. OPEX drivers, in descending order: power for blowdown pumps and the RO high-pressure pump, chemical or ozone consumables, sludge disposal, and IKTVA documentation overhead. The numbers below are S1 2024–2025 EPC bidding data and should be escalated for 2026 tender inflation rather than frozen.
| Cost line | Chemical-only train (Riyadh, 100 MW) | Ozone + DAF train (Riyadh, 100 MW) | Notes |
|---|---|---|---|
| Makepretreatment + multimedia + softener | Index 1.0 (baseline) | Index 1.0 | Same scope both trains |
| Side-stream clarification (DAF / lamella) | Index 0.8 | Index 1.0 | DAF adds air-saturation package |
| Biocontrol (chem dosing vs ozone) | Index 1.0 | Index 1.15–1.20 CAPEX | S1 CAPEX delta, 2024–2025 |
| Blowdown RO + (optional) UF | Index 1.0 | Index 1.0 | Same for both |
| Sludge dewatering (filter press) | Index 1.0 | Index 1.0 | Same for both |
| Annual OPEX (power + chem + sludge) | Index 1.0 | Index 0.50–0.70 | S1 OPEX delta, 2024–2025 |
| IKTVA / desert / HSE uplift | +20–30% on total | +20–30% on total | HydropureWater field data, 2025 |
Vision 2030, IKTVA, and Saudi Compliance Checklist
Saudi Vision 2030 names data centers as a cloud-hub pillar, with gigawatt-scale clusters under construction in Riyadh, Jeddah, Dammam, and NEOM (per S1, 2024–2025 EPC bidding). NEOM The Line and Oxagon tender language specifically demands ZLD or near-ZLD, DAF + ozone pretreatment, and local assembly under IKTVA localization rules. The compliance items the EPC should expect to tick in any 2026 Riyadh hyperscale tender: PME / NCEC discharge permit, RCJY environmental approval (for KAEC sites), SEC / SWCC water-supply coordination, Saudi Civil Defense sign-off on chemical storage (especially ClO₂ and ozone generator rooms), and an IKTVA local-content score that is now a scored tender line. Containerized, skid-mounted, factory-tested packages — the same scope as the DAF, RO, and dosing skids referenced in this article — shorten the IKTVA documentation effort because assembly happens in-region, which is the easiest way to pick up the local-content points without re-engineering the equipment.
Frequently Asked Questions
What is the minimum treatment train for a 5 MW Riyadh edge data center?
Municipal or TSE makeup through a multi-media filter (SDI < 5) plus a water softener, a side-stream filter at 5–10% of circulation, a PLC-controlled chemical dosing skid for scale and corrosion inhibitor, and blowdown routed to a small industrial RO at 75–80% recovery with permeate recycled as cooling-tower makeup. ZLD is not typically required below 5 MW; PME discharge limits apply.
Is catalytic ozone cooling-water treatment allowed under PME rules in Saudi Arabia?
Yes. PME and NCEC regulate discharge quality, not the in-loop oxidant, and catalytic ozone decomposes to O₂ with no persistent chemical residual. The EPC must still meet the PME off-gas and ambient air rules for the ozone generator room and provide a destruct unit on the off-gas.
When is ZLD mandatory versus near-ZLD for a Riyadh data center?
NEOM The Line and Oxagon tenants are tendered to true ZLD (MVR + crystallization). RCJY at KAEC and most Vision 2030 hyperscale campuses accept near-ZLD with RO at 75–95% recovery and brine to a small evaporator. Conventional PME-jurisdiction sites outside these zones can usually discharge RO permeate to sanitary sewer under an NCEC permit.
What IKTVA minimum local-content score should an EPC target for a 2026 Riyadh hyperscale tender?
Target a Local Content Value (LCV) score in the 40–60% band for water-treatment skids by sourcing assembly, enclosure fabrication, and PLC integration in-region, and request the IKTVA scorecard from the client during the bid clarification phase because the threshold varies by client (Aramco, STC, NEOM all score differently).
How many weeks does a containerized DAF + RO skid take to commission in Riyadh?
Plan 2–4 weeks for a containerized DAF + ozone skid and 6–10 weeks for a full containerized RO + UF + dosing train, including loop turnover, SAT, and IKTVA documentation. The containerized DAF + RO scope aligns with the S1 2024–2025 EPC commissioning benchmark for compact plants.