Sterilising filter assembly and integrity tester
Sterile filtration · 0.22 µm filter · Sterilising-grade filter · PUPSIT
The 0.22 µm filter train that sterilises a liquid that cannot be terminally sterilised, and the integrity tests before and after use that show the filter stayed intact.
What this page does not claim
An equipment family page teaches how a kind of equipment works and how it is assured. It is not a qualification of any site’s equipment, sets no universal limits, and was checked by SPEQ against the sources it lists, not by an independent engineer.
WHAT IT DOES
When a product cannot be sterilised in its final container, it is sterilised by filtration through "a sterile sterilising grade filter (with a nominal pore size of a maximum of 0.22 µm…)" and aseptically filled into sterilised containers (Annex 1, 8.79). Pre-filters may lower the bioburden first, and a second sterilising filter close to the point of fill should be considered (8.80).
The sterilised assembly is integrity tested before use (pre-use post-sterilisation, PUPSIT) and the filter is tested non-destructively after use before it is removed from its housing (Annex 1, 8.87). The integrity test is validated and correlated with the retention shown in bacterial challenge testing.
Where it sits: Between formulation and filling: usually a bioburden-reduction filter, then the final sterilising filter as close to the filling needles as practical.
WHERE YOU FIND IT
- Solutions and liquids that cannot be terminally sterilised
- Vent filters on product vessels
- Sterile gas supplies to aseptic processes
In the reference facility: Processing & formulation suite, Aseptic filling & isolator line
Not covered on this page: Virus filtration; HVAC filters, covered on the HEPA filtration page.
MAJOR COMPONENTS
- Bioburden-reduction pre-filter. Lowers and controls bioburden before the final filter (Annex 1, 8.80).
- Final sterilising-grade filter in its housing or capsule. Produces the sterile filtrate; 0.22 µm nominal or smaller.
- Redundant filter (optional). A backup filter; only tested after use if the primary fails (Annex 1, 8.92).
- Sterile connections and transfer line to the filler. Kept to a minimum between the final filter and the fill (Annex 1, 8.82).
- Integrity tester and test connections. Allow in-place testing, preferably as a closed system, before and after filtration (Annex 1, 8.82).
- Bioburden sampling point. Immediately before the final filter, or before the first filter in a redundant set-up (Annex 1, 8.93).
Connects to: Formulation vessel; Filling line; Autoclave or SIP for the assembly; Batch record.
INSTRUMENTS AND WHAT THEY MEASURE
- Filter integrity tester. Bubble point, diffusive flow, pressure hold or water intrusion, before and after use (Annex 1, 8.87). Watch for: membrane not fully wetted, giving a false failure.
- Pressure transmitter (gauge or absolute). Pressure upstream and downstream, for the pressure difference across the filter (Annex 1, 8.86). Watch for: gauge and absolute values compared directly.
- Temperature sensor (RTD or thermocouple). Fluid temperature during filtration, where it is a validated parameter (Annex 1, 8.85). Watch for: sensor not at the location that matters (cold spot, load centre).
PARAMETERS
SETPOINT · MEASURED · LIMIT · ATTRIBUTE- Sterilising filter integrity result (limit, bar or psi (bubble point) or mL/min (diffusion, water intrusion)). Against the integrity test value established in validation for the product as wetting fluid (Annex 1, 8.85, 8.87).
- Pressure differential (limit, Pa). Across the filter, within the maximum the site validated for the process, recorded in the batch record (Annex 1, 8.85, 8.86).
- Filtration time and volume (limit, min or L). Minimum time to filter a known volume, maximum contact time and volume, all established by the site in validation (Annex 1, 8.85, 8.86).
- Pre-filtration bioburden (limit, CFU/100 mL or CFU/mL). Sampled from the bulk and before the final filter, against the site’s limit (Annex 1, 8.93).
SPEQ publishes no universal alarm or acceptance values: they depend on the process, the product and the site’s own qualification.
CONTROLS, ALARMS AND INTERLOCKS
- Procedure: PUPSIT and post-use integrity test. Before use after sterilisation and after use before removal; an alternative to PUPSIT only with a documented risk assessment and controls (Annex 1, 8.87).
- Procedure: Single-batch use. Liquid filters are discarded after one batch and not used for more than one working day unless validated (Annex 1, 8.94).
- Procedure: Batch record of filtration parameters. Time to filter a known volume and the pressure difference are recorded and deviations investigated (Annex 1, 8.86).
- Alarm: Pressure difference above limit. Warns of operation outside validated conditions.
An alarm asks for a response; an interlock stops an action; a safety function protects people and has its own risk basis.
HOW IT FAILS
- Filter damaged during sterilisation or installation. A non-integral filter used for the batch; PUPSIT exists to find it before use (Annex 1, 8.87).
- Masking of a defect by product during filtration. A post-use test passes a filter that leaked; a reason given for PUPSIT.
- Incomplete wetting at test. False failure, and pressure to retest until a pass.
- Downstream contamination during a pre-use test. The sterile side opened or exposed.
- High bioburden or long contact time. A challenge beyond what the filter was validated for.
ASSURANCE ACROSS THE LIFECYCLE
EXAMPLES, NOT A MANDATORY LIST- Selection and validation (Annex 1, 8.81–8.85). Compatibility, adsorption and extractables; Bacterial retention with the product or a justified surrogate; Integrity test value for the product as wetting fluid; Worst-case process conditions.
- Each batch. Pre-use and post-use integrity results; Filtration time, volume and pressure difference; Pre-filtration bioburden.
- Redundant and multi-filter systems. All filters of a validated multi-filter unit pass after use (Annex 1, 8.91); Redundant filter tested if the primary fails, with investigation (8.92).
- Campaigns. Validated maximum duration of use with controls and records (Annex 1, 8.95).
WHAT THIS DOES NOT PROVE
- A passing post-use integrity test does not prove the filter was intact when filtration started. Instead: The pre-use post-sterilisation test, or the documented risk assessment that replaces it (Annex 1, 8.87).
- A filter labelled 0.22 µm does not prove it retains bacteria in this product. Instead: Bacterial retention validation with the product or a justified surrogate (Annex 1, 8.84).
- An integral filter does not prove the filtrate stayed sterile to the filling needle. Instead: Sterility of the assembly downstream and the aseptic connections, shown by media fills.
SOURCES
2 CITED- EU GMP Annex 1: Manufacture of Sterile Medicinal Products — European Commission (EudraLex Volume 4), 2022 revision, in operation 25 August 2023. EU / EEA; revised jointly with the PIC/S Committee; in force. Supports: Sterile filtration and pre-filters (8.79–8.81); system design and in-place testing (8.82); validation and parameters (8.83–8.86); PUPSIT and post-use testing (8.87); vent and gas filters (8.88–8.90); multiple and redundant filters (8.91, 8.92); bioburden sampling (8.93); duration of use (8.94, 8.95).
- EU GMP Annex 15: Qualification and Validation — European Commission (EudraLex Volume 4), 2015. EU / EEA; in force. Supports: Process validation approach.
HOW THIS PAGE WAS CHECKED
Researched and checked against the sources above by SPEQ (Claude) on 2026-10-21. It has not been reviewed by an independent engineer. It teaches how this kind of equipment works and is assured; it does not say any site, asset or process is qualified, validated or compliant.
- No integrity test value is given: it comes from the filter manufacturer’s validation for the specific product.
Spotted something wrong? Tell us.