A failed leak check on an isolation damper raises an immediate question that the test report itself may not answer: does the result tell you where a seal is failing, or does it tell you whether the assembly as a whole holds its rating? Choosing the wrong method for the question at hand produces a result that looks conclusive but does not support the decision the project actually needs to make.
Leak question each test method can answer
| Decision question | Bubble test | Pressure-decay test | Decision boundary |
|---|---|---|---|
| Where is leakage occurring at an accessible seal or joint? | Locates leakage under an agreed test condition. | Evaluates the sealed assembly response rather than locating a specific leak. | Use local leak location as diagnostic evidence. |
| How does the sealed assembly perform over time? | Leak location does not quantify overall assembly tightness. | Provides assembly-level response over time when pressure, volume, stabilization, temperature, duration, and instrument resolution are defined. | Interpret the result only under the stated test conditions. |
| Can the result document acceptance? | Leak detection alone is not final closure. | Can provide quantitative evidence when the protocol defines the conditions and acceptance criteria. | State whether the result is diagnostic, quantitative, or both; exact acceptance criteria remain project-specific. |
Before specifying a test, the project team has to decide what question the result needs to answer. A damper acceptance protocol can be built to answer “where is this leaking,” “how does this sealed assembly behave over a defined period,” or both — but the same test run does not automatically answer all three. Where the open question is the location of a defect at an accessible seal or joint, a diagnostic method that responds to local leakage under an agreed test condition is the appropriate tool. Where the open question is the sealed assembly’s overall performance, a method that reports a measurable response over time, under stated pressure and environmental conditions, is what the acceptance record needs to show.
This distinction matters because a protocol that asks for one kind of evidence and receives the other leaves the acceptance decision unsupported. A located leak at a joint tells an installer where to direct corrective work, but it does not by itself quantify how the complete assembly performs once that joint is restored. Conversely, an assembly-level figure showing an acceptable response does not identify which seal, gasket, or joint contributed to that result if the figure had failed — it would only confirm that something in the assembly needs attention, not what or where. Writing the test requirement into the URS or acceptance protocol means stating, for each required test, whether the output is expected to be diagnostic, quantitative, or both, and what evidence supports a closure decision for each.
This framing also matters for coordinating test stages across the project. A factory check performed before dispatch, a pre-commissioning check at the site boundary, and a periodic check during operating life can each call for a different balance of diagnostic and quantitative evidence, depending on what is accessible at that stage and what decision the result needs to support. ISO 10648-2:1994, which classifies containment enclosures by leak tightness and identifies factory, acceptance, pre-commissioning, and periodic checks as distinct points in an enclosure’s life cycle, supports treating these as separate decision points rather than a single pass/fail event, though the standard’s enclosure-level scope does not itself set damper-specific test parameters or class limits — those remain for the project’s own acceptance criteria to define.
Bubble testing for locating seal and joint leakage
A method built around locating leakage at an accessible seal or joint works by applying a test medium and an agreed test condition, then observing a local response at the boundary being checked. Its value is diagnostic: it tells the project team where a defect is occurring, not how the assembly performs as a whole once put back into service. This makes it a natural fit for troubleshooting after an assembly-level result has already flagged a problem, or for checking an individual joint during installation or maintenance before the full assembly is challenged as a unit.
The condition that changes how useful this method is in a given project is accessibility. A seal or joint that can be reached and observed directly under the agreed test condition supports a conclusive local finding. Where a boundary has been sealed as part of final assembly, or where penetrations sit behind cladding or within ductwork that cannot be opened without disturbing the system, local leak location becomes harder to apply, and the team has to decide at what stage of assembly the check needs to happen in order to remain accessible.
Guidance written for fumigatable boundaries — such as the Swiss Expert Committee for Biosafety’s recommendations for BSL-3 laboratories — treats penetrations through a containment boundary as locations that need attention for seal chemical resistance, local leak location, final tightness testing, and retest after repair. That framing supports using a local leak-location method as one step within a broader boundary-verification sequence, rather than as a stand-alone acceptance test; it is nonbinding guidance written for a specific jurisdiction and does not set a transferable acceptance threshold or material proof for isolation dampers outside that context.
Because this method identifies rather than quantifies, a finding of “no visible leakage” at the joints checked is not equivalent to a statement that the assembly meets an overall leak-tightness class. The project record needs to state clearly what was checked, under what condition, and what the absence of a local indication does and does not establish about the assembly as a whole.
Pressure-decay testing for assembly-level leakage evidence
A method built around the sealed assembly’s response over time works differently: rather than observing a local indication at a joint, it monitors how the assembly as a whole responds to a defined test pressure across a stabilization and test duration. The output is a measured response, not a location. That makes it the method suited to answering whether the assembly meets the quantitative performance the project has specified, and it is the form of evidence most naturally referenced in an acceptance record that states a defined pass/fail criterion.
The value of this method depends entirely on how completely its governing conditions are defined. A response measured without a clearly stated test pressure, enclosure volume, stabilization period, temperature, test duration, and instrument resolution is not comparable from one test event to another, and is not defensible as acceptance evidence if challenged during a later audit or qualification review. Where the project’s qualification documentation references predefined, justified acceptance criteria — the approach EudraLex Volume 4 Annex 15 describes for qualification and validation activities generally — the assembly-level result is the figure that gets compared against that criterion, with the URS serving as the life-cycle reference point and supplemented FAT/SAT evidence documented against it; the exact acceptance figures and staging remain specific to each project’s own protocol rather than fixed by the general qualification framework.
A result from this method that fails the stated criterion tells the team that the assembly does not perform as required, but it does not by itself tell them which seal, fitting, or joint is responsible. That is the condition under which pairing this method with a diagnostic, location-based check becomes useful — the assembly-level figure establishes that corrective work is needed, and a local check identifies where to direct it.
Method controls, instrumentation, and stabilization conditions
Both methods depend on a set of conditions being defined and held consistent, and the choice of which conditions matter shifts according to which method is in use. For a pressure-decay-type result to be interpretable, the protocol needs to fix the test pressure applied, the enclosure or assembly volume against which the response is being measured, the stabilization period allowed before readings are taken, the ambient and process temperature during the test, the total test duration, and the resolution of the instrument recording pressure change. A change in any one of these — a shorter stabilization period, a different ambient temperature, a coarser instrument resolution — changes what the resulting figure means and whether it can be compared against a stated acceptance criterion or against a prior test event on the same unit.
For a location-based check, the governing control is the agreed test condition under which the medium is applied and the local response observed — what pressure or concentration is used, how the boundary is accessed, and what counts as an observable indication at that boundary. Because this method does not produce a single comparable figure the way an assembly-level response does, its controls center on ensuring the check is repeatable and that the boundary being observed is genuinely representative of the joint in question once the unit returns to its installed configuration.
Where a project specifies both forms of evidence, the protocol needs to state which conditions apply to each so that a reviewer can tell, from the record alone, which figure answers which question. Writing stabilization time or instrument resolution once, in a way that is silently assumed to apply to both test types, creates ambiguity that surfaces later when a result is challenged. The information QUALIA requests during project review — the intended test sequence, the stage at which each check occurs, and which conditions the project has already fixed — feeds directly into how an isolation damper is configured and quoted for the acceptance evidence the project actually needs.
Acceptance and retest rules after a failed result
| Stage | Evidence or action | Closure boundary |
|---|---|---|
| Diagnostic leak location | Identify leakage at an accessible seal or joint under the agreed test condition. | A located leak supports troubleshooting but does not establish final acceptance. |
| Corrective work | Record each repair location. | The repair record does not replace the applicable acceptance test. |
| Post-repair verification | Repeat the applicable acceptance test after corrective work. | Judge closure against the protocol’s predefined, project-specific acceptance criteria. |
A failed result from either method is evidence of a condition that needs correction, not a closed finding on its own. Where a local check identifies leakage at an accessible seal or joint, that finding supports directing corrective work to a specific location, but it does not establish that the assembly meets its required performance once that work is done — only a repeat of the applicable acceptance test, under the same defined conditions as the original, can establish that. Where an assembly-level result fails against a stated criterion, the figure confirms that the assembly as a whole does not meet the requirement, but identifying the cause typically calls for the kind of location-based check described earlier before corrective work can be targeted.
Recording the repair location and method is part of the project’s own documentation discipline, separate from and in addition to repeating the test. A repair log establishes what was changed, but it is the repeated test result — evaluated against the protocol’s predefined acceptance criteria — that supports a documented acceptance decision. Treating a repair as self-evidently sufficient, without a subsequent test event matching the one that originally failed, leaves the acceptance record without the comparison it needs.
This sequencing also matters for how the project schedules return-to-service and for how the damper’s calibration and certification records stay consistent over its operating life — the questions addressed when reviewing how a damper was certified initially or how it is recalibrated over time apply again each time a repair reopens the acceptance question. Where the original protocol specified an assembly-level criterion, retesting against a local check alone does not close that requirement, and where it specified a local check, an assembly-level retest does not substitute for confirming the specific joint was corrected. The protocol’s predefined, project-specific acceptance criteria remain the reference for deciding whether the corrected unit may return to service.
Frequently Asked Questions
Q: Do I need both bubble testing and pressure-decay testing?
A: Use both when the project needs leak-location evidence as well as assembly-level quantitative evidence. Bubble testing can guide troubleshooting at accessible seals or joints, while pressure-decay testing can evaluate the sealed assembly under defined conditions; the protocol should identify which result is diagnostic, quantitative, or both.
Q: What must be defined before a pressure-decay result can be interpreted?
A: Define the test pressure, sealed volume, stabilization period, temperature, test duration, instrument resolution, and project-specific acceptance criteria. Record these conditions so the result is judged within the agreed protocol and any later retest has a clear basis for comparison.
Q: Does finding a leak with a bubble test mean the damper has passed?
A: No. A located leak supports diagnosis and corrective work, but it does not establish final acceptance. Record each repair location and repeat the applicable acceptance test against the protocol’s predefined criteria.
Q: Can a factory test result replace site acceptance testing?
A: Not automatically. The qualification plan should define whether a justified factory acceptance test is sufficient for a particular requirement or must be supplemented by site acceptance testing, using the user requirements specification as the continuing reference.
