Why Nagpur is a special case for data-centre water
Nagpur sits in Vidarbha's hot semi-arid belt, where peak summer cooling demand collides with seasonal freshwater stress, so any blowdown design has to be biased toward reuse from day one. India-wide, data centres used about 150 billion litres of water in 2024, and that figure is expected to more than double by 2030, with roughly a quarter of installed capacity in Mumbai and the balance spread across Chennai, Hyderabad, Bengaluru and emerging tier-2 hubs (Earth5r, citing CEEW and SYSTEMIQ, 2025).
CEEW also finds 57% of Indian districts at high to very high risk of extreme heat, including many data-centre hubs, so cooling load peaks in the months when surface and groundwater are scarcest. India still has no national data-centre water-efficiency standard and no mandatory sectoral water-use reporting; of the 15 states with data-centre policies, only five include explicit sustainability provisions (Earth5r, 2025). For a Nagpur facility, the design is therefore driven by MPCB discharge consent, BRSR Principle 6 disclosure, and a 2026 commercial push from hyperscale tenants who expect a water-stewardship plan as part of site selection.
The two wastewater streams every Nagpur data centre must handle
Two physically and chemically distinct streams arrive at the treatment boundary, and they should never share a single plant. Stream 1 is cooling tower blowdown (CTBD) — the periodic purge from the recirculating cooling loop needed to stop dissolved solids, corrosion byproducts and biological growth from concentrating past safe limits. At about 4 cycles of concentration, 25–30% of makeup water leaves as blowdown, and the stream typically carries 1,200–6,000 mg/L TDS (4–8× the makeup), 10–50 mg/L suspended solids, and an accumulated load of biocides, corrosion inhibitors, scale inhibitors and dispersants, with legacy chromate and high-phosphate chemistries adding reuse and discharge complications (Genesis Water Tech). Stream 2 is on-site domestic sewage from offices, cafeteria and any staff housing, which is lower in flow but high in BOD and must meet MPCB effluent conditions. CTBD needs a physico-chemical and membrane train; sewage needs biological treatment. Routing them together degrades both trains, fouls membranes, and breaches typical MPCB TDS, BOD and total-phosphorus limits.
| Parameter | Cooling tower blowdown (CTBD) | On-site staff sewage |
|---|---|---|
| Primary source | Purge from evaporative cooling loop | Offices, cafeteria, staff housing |
| Flow share at a 10 MW site | Dominant (most of liquid waste) | Minor (CPHEEO norm-based) |
| Typical TDS | 1,200–6,000 mg/L (4–8× makeup) | Low (municipal-grade) |
| Suspended solids | 10–50 mg/L, plus corrosion products and biofilm | Standard domestic range |
| Key contaminants | Biocides, corrosion/scale inhibitors, legacy chromate/phosphate, silica, hardness | BOD, COD, nutrients, pathogens |
| Treatment train | Side-stream filtration → UF → RO (± MVC) for reuse | Packaged A/O or MBR to MPCB consent |
| Consent reporting | Discharge by destination and level of treatment, ZLD status (BRSR Principle 6) | Separate domestic effluent line under MPCB consent |
Sizing blowdown for a 5–20 MW Nagpur site
The working benchmark is the IEA figure cited by CEEW: a 100 MW hyperscale facility uses roughly 0.8–2 million litres a day for on-site cooling, depending on design and climate, and the upper end is the realistic Nagpur assumption given Vidarbha summer peaks. Rack density has migrated from about 1–3 kW per rack in the early 2000s to around 8 kW today (Earth5r, 2025), which lifts cooling intensity per megawatt and pushes the per-MW water figure toward the higher end of the band in Vidarbha. For a 10 MW reference site, the sizing path is: pick a per-MW water use within the 8,000–20,000 L/day range, then apply a 20–30% blowdown fraction on makeup water (KETOS; Genesis Water Tech) to estimate blowdown volume. The table below shows the order-of-magnitude envelope; final numbers must come from the project PUE, WUE target, cycles of concentration, makeup source (Nagpur Municipal Corporation vs borewell vs recycled sewage), and local wet/dry-bulb data.
| Site size (IT load) | Cooling makeup envelope (L/day) | Blowdown at 20–30% of makeup (L/day) | Notes for Nagpur context |
|---|---|---|---|
| 5 MW edge facility | ~40,000–100,000 | ~8,000–30,000 | Lower end if free-cooling hours are high; upper end in peak summer |
| 10 MW reference site | ~80,000–200,000 | ~16,000–60,000 | Plan reuse train against upper end; pre-allocate ZLD footprint |
| 20 MW expansion | ~160,000–400,000 | ~32,000–120,000 | MPCB consent negotiation and concentrate path become critical |
Inputs the engineer must collect before sizing: contracted IT load, design PUE, target WUE (L/kWh), cooling-tower cycles of concentration, makeup source and quality, and Nagpur-specific wet-bulb and dry-bulb design conditions. Without these, any sizing is a guess.
Reuse-first treatment train for cooling tower blowdown
The 2026 default for Indian data centres is a reuse-first train that returns blowdown to the cooling-tower make-up stream rather than sending it to a sewer or a ZLD system. The sequence is engineered in four stages, with each stage removing a specific class of foulant or scale-former so the next stage can run at its design recovery.
- Side-stream filtration. Self-cleaning spiral or disc filters at 10–25 micron on 1–5% of circulation flow drop suspended solids, corrosion products and biofilm fragments before they reach membranes; capital cost $50,000–200,000 for typical data-centre flow rates (Genesis Water Tech). A multi-media filter is often paired here to take the load off downstream membranes.
- Ultrafiltration (UF). 0.01–0.1 micron UF membranes remove residual colloids, bacteria and high-molecular-weight organics at 90–95% recovery and low pressure (10–30 psi), producing an RO-quality feed (Genesis Water Tech). An ultrafiltration system for blowdown pretreatment is the standard choice at this step.
- Reverse osmosis (RO). RO delivers 95–99% rejection of dissolved solids, hardness and silica, with permeate at 10–50 mg/L TDS and system recovery of 50–85% on blowdown; a 50,000 GPD unit is installed at $250,000–500,000 with operating cost of $1.50–3.00 per 1,000 gallons (Genesis Water Tech). An industrial RO system for cooling tower make-up is sized against the upper end of the project's blowdown envelope, not the average.
- Concentrate management. Options are discharge to an MPCB-approved sewer where the consent allows, send to mechanical vapour compression (MVC) for further concentration, or haul to an authorised facility. MVC produces <10 mg/L TDS distillate at 95–98% recovery for 15–25 kWh per 1,000 US gallons (Genesis Water Tech). For sites targeting >80% overall recovery, the IDE Tech MAXH₂O design philosophy — controlled precipitation of silica, calcium carbonate and calcium sulfate in a fluidised-bed reactor plus dynamic RO cycling — avoids the multi-stage RO stacks that traditional high-recovery designs rely on (IDE Tech).
Reuse is no longer experimental: Amazon Web Services was cooling 20 data centres on purified wastewater by 2023, and Google uses reclaimed or non-potable water at over 25% of its data-centre campuses (KETOS).
| Stage | Function | Key spec | Capex / opex band |
|---|---|---|---|
| Side-stream filtration | Drop TSS, corrosion products, biofilm | 10–25 micron, 1–5% of circulation flow | $50,000–200,000 capex; low opex (Genesis Water Tech) |
| Ultrafiltration | Colloids, bacteria, organics removal | 0.01–0.1 micron, 90–95% recovery, 10–30 psi | Sized to flow; CIP every 1–3 months |
| Reverse osmosis | Dissolved solids, hardness, silica rejection | 95–99% rejection, 50–85% recovery, permeate 10–50 mg/L TDS | 50,000 GPD unit: $250,000–500,000 installed; $1.50–3.00 per 1,000 gal opex (Genesis Water Tech) |
| MVC (concentrate polishing) | Brine volume reduction, high-purity distillate | 95–98% recovery, <10 mg/L TDS distillate | 10,000–30,000 GPD unit: $1–3M capex; 15–25 kWh per 1,000 US gal (Genesis Water Tech) |
When full ZLD or partial ZLD is justified at a Nagpur site
Full ZLD pairs RO with MVC or a brine concentrator and a crystallizer, achieving 95–99% overall recovery at $3–8 million in capex and $5–15 per 1,000 gallons in opex (Genesis Water Tech). Partial ZLD uses MVC to reduce concentrate volume by 80–90%, with the residual brine sent to authorised haulers, deep-well injection, or a special discharge permit — capturing most of the reuse benefit without the full ZLD bill. The decision is policy- and consent-driven rather than vendor-driven. India's five sustainability-flagged state policies are the current ceiling: Rajasthan's 2025 policy is the most complete, requiring wastewater recycling, ZLD, rainwater harvesting and groundwater recharge (Earth5r). For Nagpur, full ZLD is generally justified only if (a) the MPCB consent prohibits discharge, (b) makeup is from a stressed borewell with no reuse alternative, or (c) the operator has signed a water-positive pledge; otherwise the reuse-first train with concentrate-to-sewer under consent is the lower-risk 2026 default.
| Path | Overall recovery | Capex band | Opex band | Typical trigger |
|---|---|---|---|---|
| Reuse (RO permeate → cooling make-up) | 50–85% on blowdown | $250,000–500,000 for a 50,000 GPD RO | $1.50–3.00 per 1,000 gal (Genesis Water Tech) | Default for 5–20 MW Nagpur sites with MPCB-approved sewer for concentrate |
| Partial ZLD (RO + MVC, brine hauled) | 85–95% system | $1–3M for MVC skid | Energy-dominated; MVC at 15–25 kWh per 1,000 US gal | Consent restricts discharge volume or TDS |
| Full ZLD (RO + MVC/brine concentrator + crystallizer) | 95–99% | $3–8M system-level | $5–15 per 1,000 gal (Genesis Water Tech) | Discharge prohibited; water-positive pledge; stressed borewell only |
On-site sewage: a separate, smaller train for staff and cafeteria
Staff sewage is sized against the CPHEEO urban norm of 135 litres per person per day, cross-checked against the Earth5r estimate that a 100 MW facility's water footprint is equivalent to the daily domestic supply of roughly 6,000–15,000 people. At 5–20 MW the sewage flow is modest, and the right plant depends on footprint and reuse intent: a packaged A/O plant of the WSZ type, buried where land is at a premium, suits campus flows in the 1–80 m³/h range, while an MBR (10–2,000 m³/day, sub-micron filtrate) is the right pick when the operator wants reuse for landscaping or toilet flushing. The packaged plant must meet MPCB effluent parameters and is reported separately from CTBD under BRSR Principle 6, which captures withdrawal, consumption, discharge by destination, level of treatment and ZLD status (Earth5r). Routing sewage into the blowdown RO train must be avoided — fouling, biocide interaction and TDS swings will destroy membrane life and breach consent. A packaged sewage treatment plant for staff flows and an MBR system for on-site sewage reuse are the two stock configurations; a ClO₂ generator for cooling-loop biocide control belongs on the CTBD side, not the sewage side.
India-specific consent, BRSR and water-stewardship triggers
There is still no national data-centre water-efficiency standard and no mandatory sectoral water-use reporting in India; disclosure is voluntary unless the operator falls inside SEBI's top-1,000 listed companies and is bound by the Business Responsibility and Sustainability Report (Earth5r, 2025). BRSR Principle 6 requires disclosure of water withdrawal by source, consumption, water intensity, discharge by destination, level of treatment, and ZLD status — so the way a facility treats cooling-tower blowdown is a direct disclosure line item, not a footnote. On the state-policy side, 15 states have data-centre policies but only five include explicit sustainability provisions, with Rajasthan's 2025 policy the current benchmark on recycling, ZLD, rainwater and recharge. Hyperscale tenants increasingly require a water-stewardship plan as a site-selection input, so designing with BRSR in mind is a commercial lever as well as a compliance one — a useful tie-back for the bid-stage narrative. Engineers should map the project against the CEEW regional heat-risk data, the local MPCB consent conditions, and the operator's BRSR scope before locking the train.
Nagpur data centre: 2026 decision matrix for blowdown and sewage
The matrix below is the single artefact a facilities engineer or EPC PM can carry into a bid review. Cost bands are drawn from Genesis Water Tech; reuse is the lowest-risk 2026 default for new Nagpur builds, with ZLD retained as a future-proofing option if MPCB tightens discharge TDS limits (some jurisdictions already cap blowdown TDS below 1,500 mg/L). For sewage, the choice is footprint-driven: WSZ where land is constrained, MBR where reuse for landscaping or toilet flushing is intended.
| Decision lever | Discharge to MPCB-approved sewer | Reuse as cooling-tower make-up (RO) | Full or partial ZLD |
|---|---|---|---|
| Best for | 5–20 MW site with consent allowing concentrate discharge; makeup from NMC or a low-stress source | 5–20 MW Nagpur site as the 2026 default; water-stress sites; hyperscale tenants expecting reuse | Consent prohibits discharge; stressed borewell only; operator has a water-positive pledge |
| Capex band (blowdown train) | Minimal — equalisation + filtration only | ~$250,000–500,000 for a 50,000 GPD RO (Genesis Water Tech); plus side-stream and UF ancillaries | Full ZLD: $3–8M; partial ZLD: $1–3M for MVC plus RO (Genesis Water Tech) |
| Opex band | Discharge fees and consent monitoring | $1.50–3.00 per 1,000 gal treated (Genesis Water Tech) | $5–15 per 1,000 gal for full ZLD; energy-dominated for partial ZLD (Genesis Water Tech) |
| MPCB / BRSR posture | Discharge by destination and level of treatment reported under BRSR Principle 6 | Reuse destination and recovery rate reported; aligns with hyperscale tenant stewardship asks | ZLD status reported; meets Rajasthan's 2025-style policy bar if the operator is state-mandated |
| Recommended default for 5–20 MW Nagpur | Only as concentrate path under consent | Yes — side-stream filtration + UF + RO, concentrate to sewer under consent | Retain as future-proofing option; full ZLD not justified on cost alone |
Frequently Asked Questions
How much will the blowdown treatment train cost for a 10 MW Nagpur data centre?
For a reuse-first train, a 50,000 GPD RO unit treating blowdown is in the $250,000–500,000 installed range, with operating cost of $1.50–3.00 per 1,000 gallons treated (Genesis Water Tech). Add side-stream filtration ($50,000–200,000), UF ancillaries, and a concentrate path, and the blowdown train alone typically lands in the low-six-figure to low-seven-figure USD band before MVC. Final pricing depends on the project blowdown volume, Nagpur ambient conditions, and consent path — request a sizing backed by your PUE, WUE target and cycles of concentration.
How do I pick the right blowdown treatment supplier for an Indian data centre?
Shortlist suppliers that can document both an India-relevant reference list and a working MVC or high-recovery RO design. Ask for hyperscale precedents analogous to the AWS (20 sites on purified wastewater by 2023) and Google (reclaimed water at 25%+ of campuses) cases cited by KETOS, and request evidence that the proposed train meets BRSR Principle 6 disclosure and MPCB consent parameters for your site. Confirm lead time on the MVC or RO skids, membrane-replacement logistics into Vidarbha, and local service coverage before signing.
Do I really need a separate sewage plant if the site already has a blowdown RO train?
Yes. Blowdown carries biocides, corrosion inhibitors, scaling ions and high TDS; staff sewage carries BOD, nutrients and pathogens. Running them through one plant fouls the RO membranes, swings the feed TDS, and breaches the typical MPCB effluent split between industrial and domestic streams. Use a packaged WSZ plant for flows where footprint is the constraint, or an MBR if you intend to reuse the effluent for landscaping or toilet flushing, and report it separately under BRSR Principle 6.
When is ZLD the right answer for a Nagpur site?
ZLD is justified only when the alternative paths are blocked: MPCB consent prohibits discharge, the only available makeup is a stressed borewell, or the operator has committed to a water-positive pledge and needs to back it with hardware. Full ZLD runs $3–8 million in capex and $5–15 per 1,000 gallons in opex (Genesis Water Tech); partial ZLD using MVC captures most of the recovery at lower cost. For most 5–20 MW Nagpur builds in 2026, a reuse-first train with concentrate-to-sewer under consent is the lower-risk default, with ZLD held as a future-proofing option.
Related Equipment
- ultrafiltration system for blowdown pretreatment — specifications, capacity range, and technical data
- industrial RO system for cooling tower make-up — specifications, capacity range, and technical data
- multi-media filter for blowdown side-stream duty — specifications, capacity range, and technical data
- MBR system for on-site sewage reuse — specifications, capacity range, and technical data
- packaged sewage treatment plant for staff flows — specifications, capacity range, and technical data
- ClO₂ generator for cooling-loop biocide control — specifications, capacity range, and technical data
