Set the reference before calculating
Cold water presents greater viscous resistance. A decrease in measured permeability can reflect cooling as well as a change in membrane condition. The IntegraTec manual uses 25 °C as its reference and expresses permeability as flux per unit area divided by transmembrane pressure. Its correction is based on water viscosity. IAPWS provides a correlation for liquid-water viscosity at 0.1 MPa.
Before comparing campaigns, document the reference and formula used for the installed module. A sheet normalised to 20 °C cannot be directly overlaid on a series normalised to 25 °C. Do not automatically transfer an RO correction to UF either: use the documentation for the relevant process and selected membrane.
The formula and the physical meaning of its direction
Define J as filtration flux in L/(m²·h), TMP in bar and Lp = J/TMP in L/(m²·h·bar). Under the cited convention, TCF(T) = µ25/µT is dimensionless, and Lp25 = Lp/TCF. Equivalent TMP at 25 °C, holding J constant, is TMP25 = TMP × TCF. Both viscosities in the ratio must use the same unit.
Below 25 °C, TCF is less than one, so the correction raises permeability to its equivalent reference value. Above 25 °C, TCF exceeds one and the correction reduces permeability. This directional check helps reveal formula errors. If another manufacturer defines the inverse factor, the operation also changes; keep the definition, units and equation together rather than copying only the name TCF.
Illustrative example: pressure alone can mislead
Our assumptions for this calculation are the same module and active area, J of 60 L/(m²·h), readings at the same filtration-cycle phase and a 25 °C reference. The rounded TCF values from the IntegraTec table are used: 0.78 at 15 °C and 1.12 at 30 °C. These are hypothetical operating data, not measurements or performance results from a Neptuno plant.
- Cold: 60/0.39/0.78 = 197.2; approximately −1.4% relative to baseline.
- Warm: 60/0.30/1.12 = 178.6; approximately −10.7% relative to baseline.
- The warm reading restores baseline TMP, but its normalised permeability remains lower.
| Illustrative condition | Actual TMP, bar | Actual Lp, L/(m²·h·bar) | TCF | Lp25, L/(m²·h·bar) |
|---|---|---|---|---|
| Baseline, 25 °C | 0.30 | 200.0 | 1.00 | 200.0 |
| Cold reading, 15 °C | 0.39 | 153.8 | 0.78 | 197.2 |
| Warm reading, 30 °C | 0.30 | 200.0 | 1.12 | 178.6 |
What to record when interpreting the trend
Compare values at the same elapsed time since backwashing and under equivalent conditions. Record temperature, flow, the area actually in service, pressures used to calculate TMP, feed quality and cleaning changes. Also state whether a reading comes from the beginning or end of a cycle. A curve combining both phases can suggest deterioration when the measurement procedure has changed.
Normalisation removes a temperature contribution under the selected model; it does not correct sensor errors, changes in active area or changes in feedwater. To assess recovery after cleaning, apply the same method before and after and compare the trend with integrity and filtrate quality. A buyer can request both measured and normalised data, along with the formula, so the assessment remains reproducible.