Cooling tower blowdown typically leaves the basin at 1,200–6,000 mg/L TDS at 4 cycles of concentration. Plants recover 70–95% of that stream with RO, or push to zero liquid discharge when permits block brine release. A 100 m³/h RO train often starts near $250K CapEx with OpEx of $0.80–$1.20/m³ treated. Full ZLD for the same flow commonly sits at $1.2M–$2M CapEx and $2.50–$4.00/m³ OpEx.
Why cooling tower blowdown recovery is a 2026 plant priority
Industrial cooling uses about 40% of industrial water, and blowdown is typically 25–30% of makeup at common cycles. At 4 cycles, blowdown TDS is usually 1,200–6,000 mg/L and can exceed 10,000 mg/L with poor makeup or aggressive concentration. RO recovers 70–90% below about 6,000 mg/L TDS; ZLD targets 95–99% when discharge of brine is barred.
Facility teams feel the squeeze from rising freshwater prices and tighter dissolved-solids permits. US EPA 40 CFR Part 423 remains the common US reference frame for steam-electric and related cooling discharges. Many sites now compare local sewer surcharges against regional industrial wastewater compliance strategies before they lock CapEx.
Most plants we size still run 3–6 cycles. They only push higher when silica and sulfate scaling stay manageable on exchangers and membranes. A published 10 MW data-center case in Arizona cut freshwater use by 35% after RO recovery. That project also reported about $120,000/year savings (Genesis Water Technologies, 2025).
Untreated brine drives corrosion, exchanger scale, and municipal surcharges for high-TDS discharge. Those hidden costs often exceed the visible makeup-water bill within two or three seasons. RO systems for blowdown recovery address the load when feed TDS stays in the brackish window. Corporate water-stewardship targets for 2026 make the same projects easier to justify to finance teams on multi-year budgets.
Blowdown chemistry: TDS, cycles of concentration, and foulants
Blowdown TDS at 4 cycles of concentration usually falls between 1,200 and 6,000 mg/L. It can climb past 10,000 mg/L when makeup is poor or cycles are pushed hard (Genesis Water Technologies, 2025). Chlorides often sit at 500–3,000 mg/L, and sulfates at 800–4,000 mg/L. Silica commonly ranges 50–200 mg/L, with hardness at 300–1,500 mg/L as Ca and Mg.
Cycles of concentration equal blowdown TDS divided by makeup TDS. A tower with 300 mg/L makeup TDS and 1,500 mg/L blowdown TDS runs at 5 cycles (1,500 / 300 = 5). Higher cycles save makeup water but intensify scale formers that limit membrane recovery. TSS of 20–100 mg/L is enough to foul RO unless media or membrane pretreatment is in place.
Data centers usually show low organics and high evaporative TDS. Power plants often carry elevated silica that hard-scales heat-transfer surfaces. Chemical plants may swing pH, metals, and residual organics from process leaks. Those profile differences decide whether a standard brackish RO skid is enough or whether hybrid thermal duty is required.
| Contaminant | Typical Range (mg/L) | Impact on Treatment | Industry Example |
|---|---|---|---|
| Total Dissolved Solids (TDS) | 1,200–10,000+ | Osmotic pressure for RO, scaling, discharge limits | All industrial cooling towers |
| Chlorides (Cl-) | 500–3,000 | Corrosion, limits RO recovery | Coastal power plants, chemical plants |
| Sulfates (SO42-) | 800–4,000 | Scaling (CaSO4, BaSO4), membrane fouling | Mining, power generation |
| Silica (SiO2) | 50–200 | Hard scaling on membranes and heat exchangers | Power plants, geothermal |
| Hardness (Ca2+, Mg2+) | 300–1,500 | Carbonate and sulfate scaling | General manufacturing, food & beverage |
| Biological Growth (Bacteria, Algae) | Variable (high CFU/mL) | Biofouling of membranes, Legionella risk | Data centers, HVAC |
| Suspended Solids (TSS) | 20–100 | Physical fouling, pre-treatment necessity | Steel mills, heavy industry |
Treatment options compared: RO, ZLD, ponds, and hybrids
RO is the default reuse path for blowdown below about 6,000 mg/L TDS. It delivers 70–90% recovery at roughly $0.80–$1.20/m³ OpEx (Saltworks Technologies, 2023). Brackish-water membranes suit moderate TDS. Higher-rejection elements are selected when chloride or silica force tighter permeate specs.
When permits ban liquid brine, ZLD trains target 95–99% recovery at about $2.50–$4.00/m³ OpEx. Those trains rely on thermal steps such as evaporation crystallization for high-TDS blowdown treatment or membrane distillation (Pall Corporation, 2012). Brine concentrators typically reclaim 90–95% of RO reject, and crystallizers finish dry solids.
Evaporation ponds stay low CapEx at about $50K–$200K for a 100 m³/h equivalent footprint class. They recover no water and need land plus liner compliance. Hybrid RO plus thermal ZLD often lands between $800K and $1.5M CapEx for 100 m³/h. OpEx lands near $1.50–$3.00/m³ because RO shrinks the thermal load.
That RO-first split is how most EPC packages we review control energy when feed TDS sits near 6,000–10,000 mg/L. Sending the full flow to a crystallizer wastes steam or MVR power on water that membranes could have removed. Pretreatment still matters: hardness precipitation and silica conditioning protect both membranes and heat-transfer surfaces. Skip that step and flux decline shows up within weeks, not years.
| Technology | Water Recovery (%) | Typical OpEx ($/m³) | CapEx (100 m³/h) | Pros | Cons | Use Case |
|---|---|---|---|---|---|---|
| Reverse Osmosis (RO) | 70–90 | 0.80–1.20 | $250K–$800K | High recovery, lower energy than ZLD | Sensitive to fouling, TDS limit <6,000 mg/L | Moderate TDS blowdown, water reuse |
| Zero Liquid Discharge (ZLD) | 95–99 | 2.50–4.00 | $1.2M–$2M | Eliminates liquid discharge, maximum recovery | High energy consumption, complex operation | High TDS blowdown, strict discharge limits |
| Evaporation Ponds | 0 (disposal) | 0.10–0.30 | $50K–$200K | Low CapEx, simple operation | High land use, regulatory hurdles, no recovery | Remote areas with ample land, lax regulations |
| Hybrid (RO + ZLD) | 90–98 | 1.50–3.00 | $800K–$1.5M | Reduced ZLD energy, high recovery for high TDS | Higher complexity than standalone RO | Very high TDS blowdown, cost optimization |
What limits ZLD reclaim in semiconductor plants?
Silica, calcium sulfate, and high ionic strength set the practical ceiling for semiconductor ZLD reclaim. Nameplate recovery is rarely the binding constraint on a live fab utility. Fab cooling loops and UPW reject streams concentrate silica into the 50–200 mg/L class typical of aggressive blowdown. Hard scale on membranes and exchangers then forces antiscalant, softeners, or lower recovery.
Scaling challenges grow when chloride-rich blowdown limits RO recovery. More volume then enters evaporators that already draw 15–25 kWh/m³. Most fabs we support stage RO first, then thermal concentration. Raw blowdown rarely goes straight to a crystallizer unless the flow is tiny.
For UPW-quality reuse of recovered water, polishing after RO is mandatory. See ultrapure water treatment for boiler feed reuse when the reuse target is boiler or rinse makeup rather than tower return. Tower return can accept higher residual TDS than UPW loops. Mixing those specs in one P&ID is a common design error.
Engineering specs for RO and thermal blowdown trains
Brackish RO for blowdown typically runs 15–25 LMH flux at 10–15 bar. Single-pass recovery of 75–85% holds when pretreatment keeps SDI and scale indices in range. Antiscalant dose is commonly 2–5 mg/L. Biocide residual is often 0.5–1 mg/L as ClO2 to limit biofouling on membranes.
An automatic chemical dosing system keeps those setpoints stable when blowdown chemistry swings with weather and cycles. Sites that generate oxidant on demand often add on-site chlorine dioxide generators for biocide dosing instead of relying on aged bulk stock. Thermal ZLD after RO reject operates under vacuum near 0.1–0.5 bar abs.
Thermal energy use is roughly 15–25 kWh/m³ versus 0.5–1.5 kWh/m³ for the RO stage. MF or UF ahead of RO removes TSS that would otherwise cake spacers. Chemical conditioning of hardness and silica before membranes is standard on high-cycle towers. Designers who omit redundancy on high-pressure pumps learn that lesson on the first unplanned outage.
| Parameter | Reverse Osmosis (RO) | Thermal ZLD (Brine Concentrator/Crystallizer) |
|---|---|---|
| Membrane Flux Rate (LMH) | 15–25 (BWRO) | N/A (Thermal) |
| Operating Pressure (bar) | 10–15 (BWRO) | Vacuum (0.1–0.5 bar abs) |
| Water Recovery (%) | 75–85 (single pass) | 90–99 (from RO reject) |
| Antiscalant Dosing (mg/L) | 2–5 | N/A (Pretreatment for RO) |
| Biocide Dosing (mg/L) | 0.5–1 (e.g., ClO2) | N/A (Pretreatment for RO) |
| Energy Consumption (kWh/m³) | 0.5–1.5 | 15–25 |
| Pre-treatment Requirement | MF/UF, chemical dosing | RO (for feed), chemical conditioning |
What do data center blowdown systems cost?
Data-center blowdown recovery CapEx for a 100 m³/h RO package typically falls between $250K and $800K. OpEx sits near $0.80–$1.20/m³ treated. Payback is often quoted at 3–5 years when water and sewer rates are high (Saltworks Technologies, 2023). Full ZLD for the same hydraulic class is usually $1.2M–$2M CapEx and $2.50–$4.00/m³ OpEx.
ZLD payback windows of 5–8 years are common in the same cost tables. Hybrid RO plus ZLD packages for high-TDS campuses often bid at $800K–$1.5M CapEx. Their OpEx band is about $1.50–$3.00/m³. Energy, membrane replacement, and chemical spend dominate RO OpEx.
Steam or mechanical vapor compression energy and maintenance dominate ZLD OpEx. Pond disposal can look cheap at $50K–$200K CapEx and $0.10–$0.30/m³. It buys no reuse credit and fails where land or liners are constrained. Campus owners chasing LEED or corporate water KPIs usually reject ponds for that reason.
| Technology | CapEx (100 m³/h system) | OpEx ($/m³ treated) | Typical ROI/Payback | Key Cost Drivers |
|---|---|---|---|---|
| Reverse Osmosis (RO) | $250K–$800K | $0.80–$1.20 | 3–5 years | Energy, membranes, chemicals |
| Zero Liquid Discharge (ZLD) | $1.2M–$2M | $2.50–$4.00 | 5–8 years | Energy, maintenance, labor |
| Evaporation Ponds | $50K–$200K | $0.10–$0.30 | N/A (disposal, not recovery) | Land, liner replacement, regulatory fees |
| Hybrid (RO + ZLD) | $800K–$1.5M | $1.50–$3.00 | 4–7 years | Energy, membranes, ZLD components |
How do UPW design decisions cut long-term cost?
Documented design decisions on recovery setpoints, silica limits, and polishing stages cut long-term UPW and blowdown cost. They prevent chronic membrane replacements and unplanned thermal duty. Locking a realistic RO recovery—often 75–85% single pass on blowdown—avoids chasing 90%+ recovery. That chase collapses when silica hits 150–200 mg/L.
Writing those limits into the basis of design clarifies reuse paths. It shows when recovered permeate can return to the tower. It also shows when permeate must feed a polishing train before boiler or process use. Plants that skip that documentation usually oversize crystallizers or burn antiscalant trying to force recovery the chemistry will not support.
Keep a living decision log: design TDS, silica ceiling, recovery target, and polish train. Update it after every pilot campaign. Finance teams trust CapEx requests that trace each dollar to a measured foulant, not a brochure recovery claim. That habit is cheap insurance on a multi-million-dollar utility project.
Discharge limits and compliance baselines for blowdown
Surface-water discharge planning in the United States often starts from a 500 mg/L TDS reference under EPA 40 CFR Part 423 discussions. Site NPDES permits still set the enforceable chloride and sulfate caps. EU examples such as Germany may allow about 2,000 mg/L TDS under Industrial Emissions Directive BAT framing. China GB 8978-1996 lists 2,000 mg/L TDS with industry-specific chloride and sulfate notes.
Saudi SASO 2005 cites 2,000 mg/L TDS, 500 mg/L chloride, and 1,000 mg/L sulfate for some industrial reuse cases. Original article notes flag possible 2024 tightening discussions around Chinese limits. Treat that as a watch item and verify the current local standard before design freeze. Daily non-compliance exposure cited for EPA Part 423 contexts can reach about $25,000 per violation per day (IDE Tech, 2024).
Repeat violations erase a multi-year RO payback faster than membrane replacements ever will. Build the treatment train to the permit, then add operating margin for seasonal TDS spikes. Do not design to the annual average alone. Summer cycles and poor makeup years are what trip analyzers and inspectors.
| Region/Country | TDS Limit (mg/L) | Chloride Limit (mg/L) | Sulfate Limit (mg/L) | Notes |
|---|---|---|---|---|
| US (EPA 40 CFR 423) | 500 | Varies by permit | Varies by permit | For surface water discharge; site-specific NPDES permits apply |
| EU (Germany example) | 2,000 | Varies locally | Varies locally | Under Industrial Emissions Directive (BAT) |
| China (GB 8978-1996) | 2,000 | Varies by industry | Varies by industry | Updates in 2024 may introduce stricter limits |
| Saudi Arabia (SASO 2005) | 2,000 | 500 | 1,000 | For industrial reuse applications |
How to select a blowdown recovery train
Selection starts with a full blowdown assay—TDS, chloride, sulfate, silica, pH, turbidity, COD, and hardness. A single grab TDS number is not a design basis. A representative lab set might read 4,500 mg/L TDS, 1,800 mg/L chloride, and 150 mg/L silica. Those three values alone decide whether RO, hybrid, or thermal ZLD is credible.
If biological loading is high, evaluate MBR systems for biological pretreatment of blowdown before membranes. Biofouling can erase an otherwise sound RO design within one warm season. Pair biology control with solids removal and you protect flux much longer. Most plants we commission regret cheap pretreatment more than expensive membranes.
- Characterize the water. Map scale indices and foulants at the design cycles you actually run, not the brochure maximum.
- Set recovery and reuse quality. Tower makeup may accept RO permeate; boiler or rinse reuse needs polishing after TDS removal.
- Fix CapEx, OpEx, and payback gates. Many owners target ~3-year RO payback or ~5-year ZLD payback before approving spend.
- Read the permit. If liquid discharge is banned, ZLD is not optional.
- Pilot on real blowdown. A multi-month RO pilot confirms flux, dose, and energy before full-scale purchase.
- Check utilities. Confirm power, steam, and waste-solids handling capacity for thermal duty.
- Plan residuals. Concentrator brine or crystallizer salt needs a disposal path before startup.
Decision rules we use on most bids:
- Blowdown TDS above 6,000 mg/L → plan hybrid or thermal ZLD rather than standalone RO.
- Blowdown TDS below 3,000 mg/L → RO is usually the cost-effective reuse path.
- Abundant land and weak recovery drivers → ponds may dispose brine but reclaim nothing.
- Water scarcity or zero-discharge permits → ZLD becomes the compliance path.
Keep spare membrane elements and a documented CIP recipe on site before summer peak. Most plants we commission lose more water to delayed cleaning than to steady-state reject.That operating discipline protects both CapEx recovery and permit headroom on every shift.
Who this is for and next step
This guide is for plant engineers, EPC process leads, and procurement managers sizing reuse or ZLD for industrial and data-center towers. Look elsewhere if you only need a once-through softener for potable makeup with no blowdown recovery goal. If you already have flow, TDS, silica, and permit limits, request a blowdown treatment design and budget quote with those numbers attached so sizing stays tied to your chemistry.
Frequently Asked Questions
What TDS should I expect in tower blowdown?
Expect about 1,200–6,000 mg/L TDS at 4 cycles of concentration on typical makeup. Poor makeup quality or aggressive cycles can push blowdown past 10,000 mg/L TDS. Measure chloride, sulfate, and silica with TDS before you pick RO versus thermal concentration, because those ions set recovery limits more than TDS alone (Genesis Water Technologies, 2025).
How much does a 100 m³/h blowdown treatment system cost?
CapEx for a 100 m³/h train commonly runs $250K–$800K for RO and $1.2M–$2M for ZLD, with hybrid packages often $800K–$1.5M. OpEx is about $0.80–$1.20/m³ for RO, $1.50–$3.00/m³ for hybrids, and $2.50–$4.00/m³ for full ZLD under the cost bands used in this article (Saltworks Technologies, 2023).
Which technology fits blowdown above 6,000 mg/L TDS?
Blowdown above 6,000 mg/L TDS usually needs hybrid RO plus ZLD or a full thermal ZLD train rather than standalone brackish RO. RO still helps as a volume reducer when chemistry allows a first pass, then brine concentrators and crystallizers finish recovery toward 95–99% (Pall Corporation, 2012).
Can treated blowdown become boiler feed water?
Yes, treated blowdown can become boiler feed if RO permeate is polished to boiler silica and conductivity limits—often below 1 mg/L silica for high-pressure boilers. Typical polishing uses ion exchange or electrodeionization after RO. Tower-return reuse needs less polishing than boiler makeup, so match the train to the reuse specification before you buy membranes.
What compliance risk does untreated blowdown create?
Untreated high-TDS blowdown can trigger permit exceedances for TDS, chloride, or sulfate and municipal sewer surcharges. Source material tied to EPA 40 CFR Part 423 contexts cites fine exposure on the order of $25,000 per violation per day, which can erase reuse project economics if violations repeat (IDE Tech, 2024).
