Why Lucknow's Water Source Changes the CTBD Treatment Question
Lucknow data centers are typically fed from Ganga-basin groundwater or municipal supply that runs high in calcium, magnesium, and bicarbonate hardness. For membrane sizing the engineer should default to the upper end of the 1,200–6,000 mg/L CTBD TDS range until a site-specific water analysis is in hand (S1, S2). That range is the blowdown envelope at 4–8× makeup concentration; Lucknow's Ganga-fed aquifers sit at the harder end of municipal supply, so ignoring local water analysis before locking membranes is the most common source of an undersized RO.
The regulatory driver sits in the same place. The Uttar Pradesh Pollution Control Board consent envelope and the assimilative capacity of the receiving drain or the Gomti basin now dictate whether on-site treatment is optional or mandatory, and consent conditions typically include TDS, temperature, and reuse obligations rather than just flow volume (S2). Treat the consent conditions, not the equipment catalogue, as the design driver.
The math that ties the two together is cycles of concentration. At 4 COC a data-center cooling tower loses 25–30% of its makeup water as blowdown; pushing toward 5–6 COC shrinks the blowdown volume and therefore the downstream treatment capex, but every dissolved species is now more concentrated, so the 4–8× concentration factor on makeup TDS must be carried through the design (S1, S2). Blowdown fraction drops as COC rises, but the chemistry intensity rises faster than the volume falls.
Mapping Every Wastewater Stream on a Lucknow Data Center Campus
CTBD is the largest by volume and the most chemistry-constrained stream, but it is not the only one. A 20–60 MW Lucknow campus typically has to handle six distinct lines: cooling-tower blowdown, secondary-loop bleed from any liquid-immersion or direct contact liquid cooling rack cooling, sanitary and pantry wastewater, equipment-wash water, RO reject from the freshwater treatment plant, and DG-set cooling water (S1, S2). Treating them as a single combined waste stream is the most common design error, because each line carries a different chemistry and a different reuse-or-discharge obligation.
CTBD is brackish, silica-rich, and chemically dosed with biocides, scale inhibitors, and dispersants. Secondary-loop bleed carries dielectric-fluid residuals that have to be confirmed against the OEM's fluid-handling guidance before being blended anywhere. Sanitary and equipment-wash streams are biodegradable but low-flow, and they belong on a polishing train that is not sized for the silica and hardness load of CTBD. Mixing the streams complicates membrane chemistry, breaks reuse accounting, and risks fouling an RO that was sized only for the CTBD fraction (S2).
The volume hierarchy is set by the cooling tower. For a 100 MW reference facility, daily water demand can reach 2 million liters with roughly 60% lost to evaporation and the balance cycling through blowdown, so the 20–30% blowdown share reported across the industry is the engineering baseline for sizing the CTBD line on a 20–60 MW campus (S2, S3). Sanitary, wash, and DG cooling are an order of magnitude smaller and should be sized separately, not folded into the CTBD envelope.
| Stream | Typical chemistry flag | Volume share on a 20–60 MW campus | Default train |
|---|---|---|---|
| Cooling-tower blowdown (CTBD) | Brackish, silica-rich, dosed | Largest; sized by COC at 4–8× makeup TDS | Side-stream filtration → UF → RO, optional MVC |
| Secondary-loop bleed (immersion / DCLC) | Dielectric-fluid residuals possible | Site-specific; smaller than CTBD | OEM-compatible separation, then to CTBD or dedicated line |
| Sanitary and pantry | Biodegradable, low-flow | Small; sized to headcount | MBR polishing line |
| Equipment-wash water | Suspended solids, occasional oil | Intermittent; duty-based | MBR or physicochemical + MBR |
| RO reject from freshwater plant | Lower TDS, lower silica than CTBD | Small; 15–30% of RO feed | Blend upstream of UF/RO after characterization |
| DG-set cooling water | Low-TDS, heat-load driven | Small, intermittent | Bleed-and-feed recovery where possible |
The Baseline 2026 CTBD Treatment Train: Side-Stream Filtration, UF, and RO
The standard reuse train pairs side-stream filtration with a hollow-fiber ultrafiltration system in front of an industrial RO system, with antiscalant and pH conditioning as mandatory pretreatment. A multi-media filter as a polishing stage in front of UF materially reduces membrane fouling rates on Lucknow-type silica-rich makeup, and should be priced in the EPC bill of materials rather than treated as optional.
Side-stream filtration is sized to 1–5% of total circulation flow, with self-cleaning spiral units rated to 10–25 µm that continuously remove suspended solids and biological load so the blowdown becomes membrane-friendly; capex is USD 50,000–200,000 for typical data-center installations (S1, S2). The cascade benefit is cleaner heat exchangers, lower biological load, and blowdown that is membrane-ready without excessive pretreatment loading downstream.
UF is the workhorse pretreatment. UF membranes at 0.01–0.1 µm pore size, operating at 10–30 psi, remove suspended solids, bacteria, viruses, and high-molecular-weight organics with 90–95% recovery, and they are also a defensible standalone option when only biological and particulate removal are required (S1, S2). The UF is backwashed with permeate and chemically cleaned every 1–3 months; replacement is a coordinated spares package with the RO elements rather than a per-train purchase.
RO removes 95–99% of dissolved solids, hardness, silica, and most treatment chemicals, producing 10–50 mg/L TDS permeate suitable for direct cooling-tower makeup. Recovery on blowdown is 50–85%, limited by silica, calcium carbonate, and calcium sulfate scaling as concentrate TDS rises; an industrial RO system in this service requires operating pressures of 150–400 psi to overcome osmotic pressure, with feed filtered to < 10–15 µm and an automatic antiscalant and pH dosing skid ahead of the high-pressure pump (S1, S2). For a 50,000 GPD blowdown RO train, installed cost is USD 250,000–500,000 with operating cost of USD 1.50–3.00 per 1,000 gallons treated (S1, S2).
Nanofiltration at 75–150 psi and 70–85% recovery is the cost-optimized alternative when hardness and sulfate, not total TDS, drive the discharge or reuse limit, with permeate TDS at roughly 30–50% of feed (S1, S2). High-silica Lucknow-type makeup waters still favor RO for complete silica removal, and NF is best evaluated as a first stage in a hardness-bound scenario, not as a silica-bound replacement.
| Stage | Key parameter | Operating envelope | Capex envelope (USD, research range) |
|---|---|---|---|
| Side-stream filtration | 1–5% of circulation flow, 10–25 µm | Continuous, self-cleaning | 50,000–200,000 (S1, S2) |
| Ultrafiltration | 0.01–0.1 µm pore size, 10–30 psi | 90–95% recovery, CIP every 1–3 months | Included in train envelope (S1, S2) |
| Reverse osmosis | 150–400 psi, permeate 10–50 mg/L TDS | 50–85% recovery, antiscalant + pH conditioning | 250,000–500,000 installed for 50,000 GPD (S1, S2) |
| Nanofiltration (alternative) | 75–150 psi, permeate 30–50% of feed TDS | 70–85% recovery, hardness / sulfate selective | Lower than RO at same capacity (S1, S2) |
When to Add MVC or Move Toward Partial / Full ZLD
Conventional brackish-water RO on CTBD plateaus at 75–80% recovery because silica, calcium carbonate, and calcium sulfate reach scaling thresholds beyond that point (S2). When a Lucknow site is pushing COC toward 5–6 and the consent envelope forces higher overall recovery, the next decision is whether to add thermal concentration or accept the RO ceiling.
Mechanical vapor compression (MVC) reaches 95–98% recovery on the RO concentrate and produces distillate with TDS below 10 mg/L, leaving a 20–30% dissolved-solids brine for disposal. MVC capex is USD 1–3 million for systems processing 10,000–30,000 GPD, with energy consumption of 15–25 kWh per 1,000 US gallons of distillate (S1, S2). Brine concentrators driven by waste heat from generators or chillers can lower MVC operating cost, but few data centers have enough recoverable waste heat without purpose-built heat supply, so the 15–25 kWh figure should be the planning assumption (S1, S2).
Partial ZLD without a crystallizer concentrates blowdown to reduce discharge volume by 80–90% and captures most of the recovery benefit at a fraction of full ZLD cost; the remaining concentrated brine may qualify for deep-well injection, hauling to approved disposal facilities, or periodic discharge under special permits (S1, S2). On a Lucknow campus, partial ZLD is the more defensible first move because Gomti-basin discharge caps are tightening but full prohibition is not yet typical, and a crystallizer is a capex line that is hard to justify without a discharge prohibition in the consent.
Full ZLD at 95–99% overall recovery combines RO, evaporative concentration, and a crystallizer producing solid salt cake for disposal, with the solid waste representing less than 1% of original blowdown volume (S1, S2). Capex is USD 3–8 million for data-center scale and opex is USD 5–15 per 1,000 gallons treated, economically rational only where freshwater is unavailable, discharge is fully prohibited, or the receiving WWTP cannot accept the daily blowdown volume (S1, S2). For a 20–60 MW Lucknow campus, full ZLD is a 2026 planning option but rarely the default — it is a force majeure decision, not an efficiency decision.
One chemistry-management alternative worth pricing is the IDE MAXH₂O architecture: a brine-desalter that intentionally deactivates scale inhibitors inside a fluidized-bed reactor, precipitating silica and calcium salts as dense pellets on seed material so the remaining brine is dominated by sodium chloride and can be looped to very high overall recovery, with a published case study reporting ~95% recovery with permeate silica at about 1 mg/L (S2, S3). For a silica-rich Ganga-basin makeup, that approach is a vendor option to evaluate against an MVC train, not a replacement for the upstream side-stream filtration and UF stages.
| Train configuration | Overall recovery | Discharge outcome | Capex envelope (USD, research range) |
|---|---|---|---|
| RO + MVC (partial ZLD, no crystallizer) | 85–95% | Brine reduced by 80–90%, residual managed | MVC 1–3 M for 10k–30k GPD + RO 250k–500k (S1, S2) |
| RO + MVC + crystallizer (full ZLD) | 95–99% | Salt cake <1% of blowdown volume | 3–8 M for data-center scale (S1, S2) |
| RO + brine-desalter (closed-loop chemistry) | ~95% with permeate silica ~1 mg/L | Pellet waste, no thermal brine | Pricing on vendor proposal; treat as option (S2, S3) |
Cost Envelope in INR and the Reuse-vs-Discharge Decision
Cooling-tower makeup reuse is the highest-value route, with 60–85% recovery that directly cuts freshwater intake and discharge volume; the economic driver is freshwater savings, not avoided discharge fees, and the avoided cost of municipal water plus pumping is the line item that closes the business case (S1, S2). On a Lucknow campus where groundwater pumping is metered and rising, the avoided freshwater cost is material enough to anchor a reuse-first design.
Discharge compliance is justified only where avoided fees are material. In water-stressed regions direct discharge fees exceed USD 5–15 per 1,000 gallons and some jurisdictions cap TDS below 1,500 mg/L, both conditions increasingly plausible for UPPCB consents in the Gomti basin (S1, S2). The engineer should request the actual consent conditions and the receiving WWTP's daily-volume capacity before locking the train, because on-site treatment becomes mandatory once municipal limits, water-scarcity mandates, or WWTP capacity are binding constraints (S2).
The USD ranges should be treated as planning envelopes, not turnkey quotes. The research gives side-stream filtration at USD 50,000–200,000, a 50,000 GPD blowdown RO at USD 250,000–500,000 installed, MVC at USD 1–3 million for 10,000–30,000 GPD, and full ZLD at USD 3–8 million (S1, S2). INR conversions on a like-for-like basis run roughly an order of magnitude higher at current exchange, but the EPC should request INR-denominated vendor proposals for membranes, antiscalants, and energy rather than treat the USD benchmarks as a quotation, because freight, duties, and local installation labour move the number materially (S2).
Decision rule: drive the choice from the consent envelope and the local discharge-fee structure, not from equipment preference. Confirm the receiving WWTP's daily-volume capacity and the TDS cap before locking the train, and escalate to partial ZLD only when reuse cannot absorb the brine or when the consent forces it. A useful peer-engineer check is to ask whether the avoided freshwater plus avoided discharge fee can carry the incremental capex over a 5-year payback — if the answer is no, the train is over-specified for the consent envelope.
| Path | Capex envelope (USD, research) | Capex envelope (INR, planning estimate) | Decision driver |
|---|---|---|---|
| Reuse (side-stream + UF + RO) | 300,000–700,000 (S1, S2) | Request INR vendor proposal; treat as planning | Freshwater cost reduction; ROI against pumping + municipal supply (S1, S2) |
| Discharge compliance (RO + polishing) | 250,000–500,000 + polishing (S1, S2) | Request INR vendor proposal; treat as planning | Avoided fees > USD 5–15 / 1,000 gal; TDS cap <1,500 mg/L (S1, S2) |
| Partial ZLD (RO + MVC) | MVC 1–3 M + RO (S1, S2) | Request INR vendor proposal; treat as planning | Volume reduction 80–90% without full ZLD capex (S1, S2) |
| Full ZLD (RO + MVC + crystallizer) | 3–8 M (S1, S2) | Request INR vendor proposal; treat as planning | Discharge prohibited or freshwater unavailable (S1, S2) |
Sanitary, Wash, and RO Reject: How to Polish the Non-CTBD Streams
Sanitary and equipment-wash streams route to an MBR polishing line sized to the relevant reuse or discharge standard. MBRs deliver near-reuse-quality effluent at sub-1 µm filtration in a footprint materially smaller than conventional activated sludge, with daily capacities scaled to the campus headcount and equipment-wash duty; this is the standard polishing stage before reuse or discharge for the non-CTBD lines (S2).
RO reject from the freshwater treatment plant is a lower-TDS, lower-silica concentrate than CTBD and can often be blended upstream of the UF/RO train for partial volume reduction, but it must be characterized separately because pretreatment-chemistry carryover can foul the CTBD membranes (S2). Sourcing RO and UF membrane elements as a coordinated spares package keeps replacement cycles aligned across the two trains and prevents chemistry drift between the freshwater-side and CTBD-side membranes.
DG cooling water and secondary-loop bleed are typically low-TDS and amenable to bleed-and-feed recovery, but dielectric-fluid residuals in immersion or DCLC secondary loops must be confirmed against the OEM's fluid-handling guidance before being blended with the CTBD line. A 20–60 MW campus in 2026 is increasingly likely to carry at least a partial immersion or DCLC rack load, and the fluid-handling constraint is the gating item, not the hydraulic capacity.
Frequently Asked Questions
What capex envelope should a Lucknow EPC carry for a 20–60 MW data center CTBD train in 2026?
Use the research ranges as planning envelopes, not turnkey quotes. Side-stream filtration is USD 50,000–200,000, a 50,000 GPD blowdown RO is USD 250,000–500,000 installed, MVC for 10,000–30,000 GPD is USD 1–3 million, and full ZLD is USD 3–8 million (S1, S2). The EPC should request INR-denominated vendor proposals for membranes, antiscalants, and energy, and size each stage against the upper end of the 1,200–6,000 mg/L blowdown TDS range until site-specific water analysis is in hand (S1, S2). A 20–60 MW reuse-first baseline (side-stream + UF + RO) is the lowest defensible envelope, with partial ZLD priced as an option against the consent.
How should a Lucknow developer select a CTBD treatment supplier?
Shortlist suppliers on three documented inputs: a written guarantee on permeate TDS at 10–50 mg/L for the silica-rich Ganga-basin makeup, a coordinated membrane spares package covering UF and RO elements on aligned replacement cycles, and a reference list on at least one Indian data center or comparable silica-bound cooling-water site (S1, S2). Ask for an INR-denominated proposal split by equipment, membranes, antiscalant program, and commissioning, and verify the supplier's experience with UPPCB consent documentation rather than only CPCB-level submissions. A vendor that can support both the process datasheet and the consent pack is materially lower delivery risk than one that hands over a P&ID and walks away.
When does the UPPCB consent envelope force on-site treatment at a Lucknow site?
On-site treatment becomes mandatory under three documented conditions: when effluent exceeds municipal limits for TDS or temperature, when the site sits in a water-scarce basin with reuse mandates, and when the local municipal treatment plant cannot accept the daily blowdown volume (S2). All three conditions are increasingly plausible for a Gomti-basin campus, and the consent envelope should be requested and read before the train is locked, because the consent conditions drive whether reuse, discharge compliance, or partial or full ZLD is the correct economic case.
Why is nanofiltration not the default choice for a Lucknow Ganga-basin makeup?
NF at 75–150 psi and 70–85% recovery is the cost-optimized choice when hardness and sulfate, not total TDS, drive the discharge or reuse limit, with permeate TDS at roughly 30–50% of feed (S1, S2). High-silica Lucknow-type makeup waters still favor RO for complete silica removal because NF does not reject silica to the same extent and a silica-saturated RO concentrate is the actual scaling bottleneck, not hardness. NF is best evaluated as a first stage in a hardness-bound scenario, not as a replacement for RO on a silica-bound Ganga-basin site.
