Define the boundary being tested
Identify the modules, manifolds and valves included in the isolated test volume. A rack test assesses that assembly, while an individual module test can help locate an anomaly. Permeate turbidity provides useful indirect monitoring but is different from a direct test with documented sensitivity and an approved interpretation.
Where a microbiological barrier is claimed, test sensitivity must be related to the required removal performance. The EPA guidance distinguishes these concepts. Applying compressed air and watching a gauge does not itself establish a health-related removal credit or demonstrate a particular virus reduction.
Use the approved preparation and test conditions
The test requires correctly wetted membranes and the pressure, timing and preparation specified for the module. Surface tension, temperature and cleaning residues can affect the response. Trapped air and incomplete stabilisation of the isolated volume may also change the observed pressure trend.
Verify the compressed-air quality and instrumentation specification, isolate the train correctly and prevent overpressure throughout the sequence. A pressure used for one manufacturer is not automatically suitable for another. An automatic routine must include verified valve states and controlled depressurisation before filtration resumes.
Compare with a meaningful commissioning baseline
Retain the baseline measured on compliant modules in the assembled rack during commissioning. Record initial pressure, actual test duration, temperature and assembly volume. A decay slope needs those conditions for interpretation: racks with different trapped volumes can exhibit different pressure responses to a similar leak.
As an arithmetic example only, a fall from 1.00 to 0.97 bar over three minutes is 0.01 bar/min. Interpret it against the approved criterion and instrument uncertainty. This number is not a universal acceptance limit, and increasing a limit simply to hide a deteriorating trend does not resolve the problem.
Locate the cause before replacing modules
Following a failed test, manage the train under the agreed water-quality procedure. Check valves, fittings, seals and the sensor before assigning the fault to membrane fibres. Where authorised, testing smaller groups or individual modules can narrow the source. Document any repair, plugged fibre or isolated module.
Repeat the required test after intervention and confirm quality before returning the train to service. Keep the historical record to identify progressive deterioration or changes following cleaning and maintenance. A valid repair must preserve the agreed barrier and capacity, rather than merely clearing a controller alarm.
Common questions
Does pressure decay prove a broken fibre?
No. Diffusion, temperature, wetting and leaks in the isolated circuit can also affect the result.
Does low turbidity confirm integrity?
No. Turbidity is an indirect indicator and must be combined with the tests and sensitivity required for that installation.