Choose a baseline that can be reconstructed
Keep a stable operating dataset after commissioning or after cleaning with documented recovery. Record membrane models, stage arrangement, recovery, temperature, pressures and conductivities. A baseline collected during start-up, while air remains trapped or product quality is unstable, can turn normal operation into a misleading alarm.
Distinguish comparison with the design from comparison with initial service. The former checks a specified duty under its conditions; the latter follows deterioration over time. Do not quietly reset the reference after each incident, because doing so removes the accumulated performance history.
Validate the input measurements first
The net pressure driving water through the membrane depends on hydraulic pressure, permeate backpressure and osmotic pressure difference. Correcting temperature alone leaves changes in salinity or recovery unexplained. Use the membrane supplier’s procedure or a normalisation tool appropriate for the installed arrangement.
Collect simultaneous readings and check that feed flow equals permeate plus concentrate within instrument uncertainty. A fouled conductivity probe or incorrectly configured flow meter can generate a convincing but false trend. Retain measurement units, temperature-compensation settings and calibration records alongside the calculated values.
Interpret relative changes consistently
For an arithmetic example, if baseline normalised production is 20 m³/h and the current value is 18 m³/h, the change is (18/20 − 1) × 100 = −10%. This describes a trend; it does not establish the cause or prescribe a cleaning operation.
If normalised salt passage moves from 1% to 1.2%, its relative increase is 20%, although the absolute change is only 0.2 percentage points. Confusing these expressions can delay investigation. Keep the same definition when comparing trends, and distinguish salt passage, rejection and measured conductivity.
| Indicator | Useful comparison | Frequent mistake |
|---|---|---|
| Normalised flow | Same baseline and method | Comparing measured output alone |
| Normalised salt passage | Relative change | Confusing percentages and percentage points |
| Pressure drop | Same flow and stage | Assuming every increase is scale |
Turn the trend into a focused investigation
A sustained decrease in normalised flow accompanied by greater pressure drop points towards deposits or hydraulic restrictions. Higher salt passage also calls for checks of seals, chemical damage and measurement conditions. These patterns are clues that need confirmation by stage and against the feedwater history.
Mark interventions on the trend: a chemical change, new water source, cleaning or element replacement. This makes it possible to assess which action restored performance. Apply the installed membrane manufacturer’s criteria and retain comparable before-and-after datasets whenever the operating arrangement changes.
Common questions
Does normalisation increase output?
No. It supports interpretation. It helps distinguish operating changes from performance loss so the next intervention can address a plausible cause.
Are temperature and flow readings enough?
Usually not. Pressure, salinity, recovery and permeate backpressure also matter. Required inputs depend on the calculation method and membrane arrangement.