Introduction: no alarm, but less permeate
When feedwater temperature falls, plant operators often report the same thing: the high-pressure pump is running, the control panel shows no alarm, pretreatment has not changed, yet the permeate flow meter reads lower than it did in warmer months. The first reaction is usually to assume the membranes have failed.
In most cases they have not. RO membrane output is strongly temperature dependent, and the loss that comes with colder water is a physical effect that can be estimated in advance. The real risk is not the lower flow itself. It is misreading that normal temperature effect as membrane damage and then raising pressure or recovery to compensate - which can push the system toward scaling or beyond its operating limits.
This article breaks the problem into three parts: how much output you should expect to lose, how to use three readings to separate a normal temperature effect from a genuine fault, and what to change in day-to-day operation and in the design of a new plant.
1. Key takeaways
The loss is predictable. Using 25 C (degrees Celsius) as the common design baseline, membrane flux falls by roughly 2-3% for every 1 C drop at constant pressure; at 5 C, permeate output is typically about half of the 25 C value.
Normalize the data before judging. Only when normalized permeate flow - corrected to the same reference temperature, pressure, recovery and salinity - is still falling should fouling or scaling be suspected.
Do not simply add pressure. Higher operating pressure is bounded by the membrane element maximum pressure, the pressure vessel rating, and the pump and motor ratings. Raising recovery to make up the shortfall directly increases scaling risk.
Fix it at design stage. Size the plant for the coldest-month feedwater temperature rather than the annual average, and allow margin in membrane area, pump power and insulation.
2. Why cold water reduces permeate output
Membrane flux can be understood, in simplified terms, as proportional to net driving pressure and inversely proportional to the dynamic viscosity of water:
Jw ~ NDP / mu (Jw = membrane flux, NDP = net driving pressure, mu = dynamic viscosity of water)
As temperature falls, water viscosity rises significantly - at 5 C it is about 1.7 times the value at 25 C - so water molecules meet more resistance passing through the membrane. At unchanged feed pressure, more of the net driving pressure is spent overcoming that resistance and permeate flow falls. Flow resistance on the feed side also increases, so the stage pressure drop can rise slightly.
Engineers use a temperature correction factor (TCF) to translate membrane performance at different water temperatures to the 25 C baseline. The table below gives general engineering ranges, useful for checking whether the observed loss is within the expected band:
| Feedwater temperature | Temperature correction factor TCF (reference) | Relative permeate flow at constant pressure (reference) | Operating note |
|---|---|---|---|
| 25 C | 1.00 | 100% | Common design baseline |
| 20 C | 1.12-1.15 | about 87-89% | Cooling begins; the loss becomes visible |
| 15 C | 1.32-1.36 | about 74-76% | Surface water and groundwater turning cold |
| 10 C | 1.58-1.65 | about 61-63% | Check pump margin and membrane area |
| 5 C | 1.90-2.05 | about 49-53% | Low-temperature duty; consider feed heating or insulation |
Note: These are general engineering reference ranges. Results vary with membrane model, feed salinity, recovery and pressure. Use the membrane supplier's design software and the project design documents as the basis for calculation; this table is neither a product performance commitment nor a design basis.
One companion effect is easy to miss: at lower temperatures salt passage normally decreases slightly, so permeate conductivity often falls rather than rises. If conductivity climbs during a cold period, look first at membrane oxidation, O-ring or permeate tube seals, and concentrate-side leakage - not at temperature.
3. Three readings that separate normal loss from a real fault
- Normalized permeate flow: the flow corrected to reference conditions. It is the meaningful indicator of membrane condition; comparing raw flow readings from different months is not.
- Differential pressure (feed to concentrate): a sustained rise points to channel blockage, scaling or biofouling; a stable value supports the temperature explanation.
- Permeate conductivity: helps rule out seal and membrane integrity problems; sample individual pressure vessels if needed.
Together they give a practical troubleshooting checklist:
| Observation | Readings to check first | Likely causes | Suggested action |
|---|---|---|---|
| Permeate flow down, differential pressure stable, conductivity stable or slightly lower | Normalized permeate flow, feed temperature | Physical effect of colder feedwater | Compare against the temperature curve; do not add pressure blindly; check against design margin |
| Permeate flow down, stage pressure drop rising | Pressure drop trend, SDI15 | Membrane fouling, scaling, plugged cartridge filter | Review pretreatment and dosing; inspect and replace cartridge filters; assess cleaning per membrane guidelines |
| Permeate flow down, feed pressure also down | Pump discharge pressure, VFD frequency, valve position | Pump performance loss, insufficient feed supply, blocked upstream filter | Check the pump curve and the differential pressure across upstream filters |
| Permeate conductivity rising | Conductivity of each pressure vessel | Membrane oxidation, failed O-ring or permeate tube seal, concentrate leakage | Sample vessel by vessel to locate the fault; check free chlorine and reducing-agent dosing |
| Recovery too high, concentrate flow too low | Concentrate flow meter, recovery setpoint | Recovery raised to compensate for lower output | Return to design recovery; reassess scaling tendency and antiscalant dose |
| Flow fluctuating during start-up | Pressure ramp, venting condition | Air in lines at low temperature, membranes not yet stabilized | Flush at low pressure to vent air, then ramp pressure slowly to the setpoint |
| Abnormal flow after a shutdown | Shutdown log, ambient temperature | Frozen membrane elements, residual water icing, microbial growth during standby | Follow freeze protection and shutdown preservation procedures; test membrane integrity if needed |
| Coagulation or dosing less effective | SDI15, free chlorine, turbidity | Slower coagulation reactions and changed chemical solubility at low temperature | Review chemical selection and dose; allow longer mixing and reaction time |
4. Five actions for low-temperature operation
1. Keep pressure and recovery within limits. If higher operating pressure is genuinely needed to compensate, first confirm margin in the membrane element maximum pressure, pressure vessel rating, high-pressure pump head and motor power. Recovery should not be raised to make up the shortfall, since the concentrate side is already closer to sparingly soluble salt saturation in cold conditions.
2. Recalculate pretreatment for cold-water quality. Keep feed SDI15 below 5 (below 3 where practical); for polyamide composite membranes, keep free chlorine below 0.1 mg/L; replace cartridge filters by differential pressure (a common practical value is 0.05-0.1 MPa); and recalculate antiscalant dosing from a full cold-season water analysis and the actual recovery instead of carrying over warmer-month settings.
3. Clean on condition, not on the calendar. Consider chemical cleaning when normalized permeate flow has fallen by about 10-15%, stage pressure drop has risen by about 15%, or salt rejection has clearly declined. Identify the foulant type (inorganic scale, organic matter or biofilm) before choosing the cleaning chemistry and temperature, and follow the membrane supplier's guidelines.
4. Freeze protection matters more than insulation alone. Keep the equipment room or enclosure above 5 C; provide low-point drains and heat tracing on pipework; flush at low pressure during short stops, and follow the membrane supplier's preservation procedure for long shutdowns. Do not add antifreeze on your own initiative - if it is required, confirm compatibility with membrane materials and flush thoroughly before restart.
5. Record trends, not single readings. Log feed temperature, feed/concentrate/permeate pressures, permeate and concentrate flows, permeate conductivity and SDI15 at the same frequency. Low-temperature diagnosis depends on a continuous trend; one set of readings rarely proves anything.
5. Design implications for new projects
If a project is still at the selection stage, the following checks are more economical than any operational remedy later:
| Design item | What to check for low feedwater temperature |
|---|---|
| Design temperature basis | Use the coldest-month feedwater temperature, not the annual average |
| Membrane configuration | Allow spare membrane area or elements; reserve space for an extra stage if needed |
| High-pressure pump | Check head, flow and motor power against the pressure required at low temperature |
| Intake and feed | Evaluate intake warming, feed heating or heat exchange |
| Enclosure and pipework | Insulation for containerized/skid units, heat tracing, low-point drains, frost-safe slopes |
| Integration scope | House pretreatment, dosing, cleaning and controls in a heated enclosure to limit on-site cold exposure |
| Technical appendix | State the permeate output basis and acceptance conditions at low temperature to avoid disputes at handover |
For export projects, defining the low-temperature permeate output in the technical appendix up front is usually far easier than explaining a temperature correction factor after commissioning.
6. What HONCO can provide
HONCO supplies complete commercial and industrial reverse osmosis systems for overseas projects, including skid-mounted RO plants from 0.5 to 100 m3/h, EDI units, pretreatment stages (multimedia filtration, activated carbon, softening, ultrafiltration, cartridge filtration), dosing and chemical cleaning packages, containerized integrated systems, and replacement filter cartridges.
For projects in cold regions, HONCO can verify the design against the coldest-month feedwater temperature, a full feedwater analysis, target permeate output and required recovery, and provide a configuration list, recommended operating parameters and a technical appendix. Commissioning and remote operating guidance are also available. Final configuration and product-water figures are subject to the project design documents agreed by both parties.
7. Scope and limitations
- The temperature and pressure values in this article are general engineering reference ranges, not an HONCO product performance commitment and not a design basis.
- Actual results vary with membrane model, feed salinity, recovery and pressure. The membrane supplier's design software and the project design documents govern the calculation.
- Where product water is intended for drinking, compliance must be judged against the regulations applicable at the project location and actual test reports.
- Chemical cleaning and dosing should be carried out by trained personnel following the membrane supplier's guidelines. This article does not replace the equipment operating manual.
Conclusion
A seasonal drop in permeate output is not the problem. The two costly mistakes are treating a temperature effect as membrane failure and cleaning repeatedly, and pushing pressure and recovery outside their limits to recover the missing flow. Normalize the data first, then adjust operating parameters. Once the coldest-month feedwater temperature is written into the design input, a cold period becomes a predictable curve rather than an unexpected fault.
Reviewing a design for low-temperature operation?
Send us your feedwater analysis, target permeate output, required recovery and local low-season water temperature. The HONCO technical team will prepare a configuration proposal and technical appendix.