Technology comparison
Cooling tower blowdown treatment technologies, compared.
Seven ways to cut what a cooling tower sends to drain: what each removes, what it needs, what it leaves behind, how much water it returns and where it stops working. Nona makes one of them.
Last reviewed
Short answer
Start with controls and cycles of concentration, then filtration for suspended solids and softening where hardness is the limit. If dissolved salts still set the limit, nanofiltration, reverse osmosis or ICP can return water from the blowdown and leave a concentrate to manage. Thermal systems go furthest and suit sites where no liquid can leave.
New to the topic? Read the cooling tower blowdown recovery guide, or put your own numbers in the savings calculator.
Side by side
Seven blowdown treatment options compared
Figures are labelled by where they came from. Vendor-reported means a supplier's published claim, not an independent measurement. Where no sourced figure exists, the cell describes the behaviour instead. Nona wrote the ICP row about its own technology.
| Technology | What it removes | What it needs | Waste stream | Typical recovery | Where it fits | Limits |
|---|---|---|---|---|---|---|
| Raising cycles of concentration | Nothing. It keeps water in the loop longer by bleeding less, so dissolved solids rise. | A conductivity controller, a treatment program for scale and corrosion, and logs of makeup and blowdown. | Blowdown continues in a smaller, more concentrated volume. | Not a recovery process. EPA: going from three to six cycles cuts makeup 20% and blowdown 50%. | Towers running at two to four cycles, the range EPA says many systems operate in, where chemistry allows more. | Bounded by makeup and tower water chemistry. Higher cycles raise scale and corrosion risk, and discharged water must still meet allowable standards. |
| Side-stream filtration | Suspended solids, organics and silt from a portion of the circulating water. | Typically 20 to 30 psi supply pressure. Generally sized for 3 to 10% of system flow, up to 20%. | Backwash or purge water. Volume varies by filter type: centrifugal separators lose little, pressure sand filters need significant water. | Does not treat dissolved solids. DOE says it often reduces the water rejected from the system. | Unclarified makeup sources, fouling-prone systems, windblown dirt in the sump, heat exchangers that need frequent mechanical cleaning. | Works on particles, so dissolved salts that set the cycles limit remain. Sand filter backwash needs a clean water source. |
| Softening (ion exchange) | Hardness (calcium and magnesium), exchanged for sodium on a resin. Applied to makeup water. | Ion exchange resin and salt for regeneration. | Salt brine from regeneration, which some utilities ask customers to limit because it adds chloride to wastewater. | Does not return water by itself. It can allow operation at higher cycles of concentration. | Sites where hardness is the limiting factor on cycles. | Sodium replaces the hardness rather than removing salt from the water overall. Check local limits on brine and chloride before specifying. |
| Nanofiltration | A selective membrane cut, with pores typically larger than reverse osmosis. One cooling tower blowdown study measured about 50% salt rejection, against 98% for RO. | Pressure-driven membrane. Vendor-reported 75 to 150 psi; one hybrid makeup study ran at 7 bar with lime and coagulant pretreatment. | Concentrate. One published data center design sends a 10% reject stream to lime treatment and gypsum precipitation. | Vendor-reported 70 to 85%. One published data center design targets 90%, a design figure. | Selective removal, such as sulfate or hardness, where full desalting is more than the reuse needs. | Lower salt rejection can rule out reuse as makeup, as it did in that blowdown study. Silica rejection may be limited deliberately to manage scaling. |
| Reverse osmosis, with brine management | Dissolved constituents. One cooling tower blowdown study measured 98% salt rejection. | Pressure above osmotic pressure: vendor-reported 150 to 400 psi on blowdown. PNNL describes extensive pretreatment. Blowdown adds fouling from biology, treatment chemicals and airborne debris. | Concentrated brine that must be disposed of properly, for example in evaporation basins or by further concentration. | Vendor-reported 50 to 85%. One campus study recommended UF and RO at 45% over high-efficiency RO at 75%. NREL measured 53% blowdown savings on a train that includes RO. | Reuse that needs high quality water, at sites with stable feed and the pretreatment and operating skills to run membranes. | Membrane fouling, energy use and concentrate disposal, all named by PNNL. RO water can be aggressive and cause corrosion, so check the return chemistry. |
| Thermal concentration or MVC | Water is evaporated away from nearly all dissolved solids, leaving a highly concentrated brine or solids. | Mechanical vapor compression or thermal crystallizers, usually downstream of a membrane step. Heat or compressor energy replaces membrane pressure. | Very concentrated brine or solids, which avoids brine disposal to sewer in a full zero liquid discharge design. | The highest of this group. Vendor-reported 95 to 98% from concentrate streams, and 85 to 95% overall when paired with RO. | Sites with strict liquid discharge limits or no permitted route for concentrate. | Nature Reviews Clean Technology calls zero liquid discharge expensive and energy-intensive. Solids still need handling. |
| Ion concentration polarization (ICP), Nona RECOVER | Charged dissolved ions, separated by an electric field into a cleaner stream and a concentrate. | An electric field rather than high membrane pressure, plus prefiltration. | A concentrated brine stream, subject to your discharge permit. | 90% of cooling tower blowdown recovered in on-site testing. | Towers where dissolved salts set the cycles limit and water and sewer costs justify recovery. | Newer than the other options. Leaves a concentrate to manage. Not a source of high-purity water. |
Where ICP is not the best fit
We sell ICP, so here is the short list of sites where we would point you elsewhere first. For one technology head to head, see ICP versus reverse osmosis.
- You must discharge no liquid
- ICP leaves a concentrated brine. Look at thermal concentration or a crystallizer.
- You need high-purity water
- Look at reverse osmosis, which measured 98% salt rejection in one blowdown study.
- Controls or filtration have not been addressed
- Start there. Those steps are simpler and may be enough.
- Water and sewer are cheap
- Recovery of any kind may not pay back. Model your own rates first.
Blowdown treatment technologies: common questions
Direct answers for engineers choosing how to treat or reduce cooling tower blowdown.
What is the first thing to try before treating blowdown?
Check controls and cycles of concentration. EPA says many systems run at two to four cycles and that going from three to six cuts makeup by 20% and blowdown by 50%. How far you can go depends on your makeup and tower water chemistry.
What is the difference between softening and desalting blowdown?
Softening swaps hardness for sodium on a resin, so it can allow higher cycles but does not return water by itself. Nanofiltration, reverse osmosis and ICP separate blowdown into a cleaner stream to reuse and a concentrate to dispose of.
How much of its feed does reverse osmosis recover on blowdown?
It varies with chemistry and pretreatment. One vendor reports 50 to 85%, a campus study recommended a 45% design over a 75% option, and an NREL evaluation measured 53% blowdown savings on a train that includes RO.
Can any option leave no liquid discharge?
Yes. Thermal concentration and crystallization can reach zero liquid discharge. A 2025 review in Nature Reviews Clean Technology describes that route as expensive and energy-intensive, so it usually suits sites with strict discharge limits.
When is ICP not the right choice?
When you need zero liquid discharge or high-purity water, when controls and filtration have not been addressed, or when water and sewer are cheap.
References
Public sources for the table. ICP figures come from Nona's own on-site testing and the MIT sources listed. Vendor sources are marked as vendor-reported in the table.
- WaterSense at Work, Section 6.3: Cooling Towers
US EPA (2012). Cycles of concentration, blowdown and makeup.
- Best Management Practice #10: Cooling Tower Management
US DOE Federal Energy Management Program. Softening, side-stream filtration and acid treatment.
- Technical Evaluation of Side Stream Filtration for Cooling Towers
US DOE Federal Energy Management Program. Pressure, sizing, filter types and backwash.
- Emerging Technologies Review: Water Reuse Systems for Cooling Tower Applications (PNNL-34788)
Pacific Northwest National Laboratory (2023). Reverse osmosis, nanofiltration, waste handling and technology maturity.
- GPG 052: Blowdown Recovery System for Cooling Towers
US GSA and NREL (2023). Measured blowdown savings on a system that includes reverse osmosis.
- Recovery of cooling tower blowdown water through reverse osmosis
Ahmed, Jamal and Shujaatullah, Desalination and Water Treatment 189 (2020). Fouling and pretreatment review.
- Comparison of Nanofiltration and Reverse Osmosis for Cooling Tower Blowdown
Pungsang, Chulalongkorn University thesis (2016). Salt rejection on blowdown at a cogeneration plant.
- Development of combined physico-chemical processes to remove hardness and silica
Thai Environmental Engineering Journal 32(3) (2018). Nanofiltration at 7 bar on cooling makeup groundwater.
- Sulfate reduction without biology
Brown and Caldwell, Water Technology (2026). A nanofiltration design for data center blowdown.
- Cooling Tower Blowdown Recovery
Bellardine, Jasper and Freeman, International District Energy Association. Campus UF and RO design choice.
- Brine management with zero and minimal liquid discharge
Tong et al., Nature Reviews Clean Technology 1 (2025). Mechanical vapour compression and crystallizers.
- Advanced Blowdown Treatment Technologies for Data Center Water Recovery
Genesis Water Technologies (2025). Source of the vendor-reported pressure and recovery ranges.
- AB 334 bill analysis, California Assembly
How softener resins exchange calcium for sodium.
- Help in reduction of chloride discharge
Village of Union Grove, Wisconsin. Softener brine as a source of effluent chloride.
- Ion concentration polarization electrocoagulation
MIT Technology Licensing Office. Mechanism and concentrated stream.
- Return flow system for ICP desalination (US Patent 11,780,754)
Han and Yoon, Massachusetts Institute of Technology (2023).
Check the options against your own water.
Bring your water analysis, cycles of concentration, blowdown flow and discharge limits. We will tell you whether ICP fits and when another option makes more sense.