An airflow visualization study, commonly called a smoke study, is the operational evidence that the cleanroom is behaving as the design intended. The qualification documents show that the filters are in spec, the pressure cascade is correct, and the air change rate meets the design value. None of that proves that the air actually moves the way the model assumed. A smoke study is the test that closes the gap between the documented design and the physical room.
This article is a working guide to designing, executing, and documenting an airflow visualization study that survives both an internal review and a regulatory audit. It assumes the cleanroom is already running with a defined contamination control strategy, a working cleaning rotation, and a documented gowning procedure. The smoke study is the verification that those three operational programs are running inside an airflow envelope that supports them.
What the Study Proves โ and What It Does Not
A smoke study proves three things and only three things. It proves that the air moves from the clean zone to the less-clean zone, that there are no stagnant zones in the working area, and that the air sweep over the critical work area is sufficient to clear particles generated at the work surface. It does not prove that the room is operating within particle limits โ that is the job of the environmental monitoring program. It does not prove that the disinfection program is working โ that is the job of the contact plates and the residue trend. It does not prove that the operators gown correctly โ that is the job of the gowning qualification.
The mistake most teams make on the first smoke study is trying to prove too much. The report ends up mixing airflow observations, particle data, and operational observations, and none of the three is as clean as it should be. A working report is structured around the three things the study actually proves, with the operational data cross-referenced to the documents that own it. The cleaner the boundary, the easier the audit.
The Three Protocols: Static, Dynamic, and Recovery
A complete smoke study has three protocols, executed in this order. Each protocol has a defined scope, a defined pass criterion, and a defined output. Skipping a protocol โ typically the recovery protocol โ is one of the most common audit findings.
- Static (at-rest) protocol. The cleanroom is operating, the equipment is in place, but there are no operators and no active work. The smoke is released at a fixed grid of points across the room, and the airflow pattern is recorded. The output is a pass/fail against the design assumption: first air at the critical work area, no reverse flow, no stagnation. Static is the baseline protocol and the one that almost always passes if the room is qualified correctly.
- Dynamic (in-operation) protocol. The cleanroom is operating with the actual process, the actual number of operators, and the actual material flow. The smoke is released at the same grid, and the airflow is recorded with the operators moving. The output is a pass/fail against the same design assumption, but the test is harder because the operators and the process are the contamination source the design has to handle. Dynamic is the protocol that finds the design problems the static protocol missed.
- Recovery protocol. The cleanroom is challenged with a defined release of a surrogate contaminant (typically a fixed mass of particles or a fixed volume of smoke), and the time to clear the room back to the operating limit is measured. The output is a recovery time, compared to the design recovery time. Recovery is the protocol that proves the room can recover from a process upset, and it is the one most often skipped because it is operationally harder to run than the other two.
The three protocols are linked. Static without dynamic is a partial study. Dynamic without recovery is a partial study. Recovery without the first two has no baseline. A complete study runs all three, in this order, on a defined cadence โ typically annually for the dynamic and recovery protocols, and at the requalification cadence for the static protocol.
Smoke Generation: Tools, Methods, and What the Camera Sees
Smoke studies use either a fog generator (typically a glycol or glycerin-based fluid that produces a visible white aerosol) or a smoke pencil (a small handheld tube that produces a thin, controllable plume). The two tools answer different questions. The fog generator fills a zone and shows the bulk flow pattern; the smoke pencil traces a single streamline and shows whether the air at a specific point is moving in the expected direction. A working study uses both, in sequence: the fog generator to map the bulk flow, and the smoke pencil to test the specific points the bulk flow raised as questionable.
The camera is the third tool. A video record of the smoke release is the audit evidence, and the report is built from the video. The video has to be time-stamped, the location of the camera has to be documented, and the lighting has to be sufficient to see the smoke against the back wall of the room. A study that produced a written report but no video is the second most common audit finding.
The smoke is neutral. It does not represent particles, it does not represent microbes, and it does not represent process byproducts. It is a visible tracer for the air that carries all of those things. A study that confuses the smoke with the contamination is the third most common audit finding. The report has to state, in plain language, that the smoke is a tracer and that the conclusions are about airflow, not about the contamination itself.
The Pass Criteria: Six Things That Must Be True
A working smoke study has a small number of pass criteria, written down before the study starts, and applied consistently. Six criteria cover the bulk of the audit expectations. A study that does not have a written pass criterion is the fourth most common audit finding.
- First air at the critical work area. The smoke released 30 cm upstream of the critical work area (typically the open vial, the open container, or the active tool) reaches the work area before it reaches any other surface. First air is the design intent, and the study has to demonstrate it visually.
- Sweep across the work area. The smoke released at the work area moves away from the work area in a defined direction (typically toward the return air grille), without recirculation back to the work area.
- No reverse flow at the airlock doors. The smoke released on the higher-pressure side of an airlock door does not pass to the lower-pressure side when the door is closed. Reverse flow defeats the cascade and is one of the most consequential design problems a smoke study can find.
- No stagnation in the working zone. The smoke released at the grid points in the working zone continues to move for the duration of the video clip. A stagnation is defined as a cloud that remains visible at a fixed location for more than 30 seconds without dispersion.
- Operator shadow effect is bounded. The smoke released downstream of an operator shows a defined wake that does not extend to the critical work area. An operator that disrupts the airflow to the critical work area is an operational problem the study has to surface.
- Recovery time within design. The recovery time measured in the recovery protocol is within the design value, typically 15 to 20 minutes for an ISO 7 / Grade C cleanroom. A recovery time that has lengthened since the last study is a signal of a filter or air balance issue.
The Dynamic Protocol: How to Run It Honestly
The dynamic protocol is the one that finds the real problems, and the one that is hardest to run honestly. The temptation is to script the operator movements โ “stand here, move there, sit still” โ and to release the smoke at the same points the static protocol used. That approach produces a clean report that misses the operational reality. The honest dynamic protocol uses the actual process, the actual number of operators, and the actual material flow, with the smoke released at the points the static protocol raised as questionable.
A working dynamic protocol has the operators perform a representative process for the duration of the video clip โ typically 10 to 15 minutes โ with the smoke released at the critical work area, at the airlock doors, and at the points the static protocol flagged. The video records the airflow, the report identifies the observations, and the observations are either pass (consistent with the design intent) or fail (not consistent, with a defined corrective action). The corrective action is a CAPA, the CAPA is closed before the next study, and the next study has to demonstrate that the corrective action worked.
The Recovery Protocol: The One Most Often Skipped
The recovery protocol is the one most often skipped because it is operationally expensive. The protocol requires challenging the room with a defined release of a surrogate contaminant and measuring the time to return to the operating limit. The challenge itself is disruptive โ the room has to be filled with particles or smoke, and the time to clear is the time the room is not available for production. The cost of running the protocol is real, and the cost of not running it is the audit finding.
A working recovery protocol uses a defined release (typically 100x to 1000x the operating limit of an ISO 7 / Grade C room), a defined measurement (the time for the continuous particle counters to return to within the operating limit), and a defined pass criterion (the design recovery time, typically 15 to 20 minutes). The protocol is run annually, and the trend is plotted. A lengthening recovery time is a signal of an HVAC or filter issue that the static and dynamic protocols will not detect.
The Report: Six Sections, One Conclusion
The report from a complete smoke study has six sections, in this order. The structure mirrors the structure of the contamination control strategy, and the cross-reference is intentional. The auditor reads the CCS first and the smoke study second, and the two documents have to support each other.
- Protocol summary. A one-page description of the three protocols, the pass criteria, the tools used, and the date and time of the study.
- Static protocol results. The video clips, the grid of release points, and the pass/fail observation for each point. Failures are highlighted and carried forward to the corrective action section.
- Dynamic protocol results. The video clips with the operators in motion, the observations at the critical work area, the airlock doors, and the points the static protocol raised as questionable. Failures are highlighted and carried forward.
- Recovery protocol results. The release profile, the time-to-clear measurement, and the comparison to the design recovery time. A recovery time that has lengthened is highlighted.
- Corrective actions. The list of failures from the static and dynamic protocols, the CAPA opened for each, the owner, the target close date, and the status. The status is either closed (with evidence) or open (with a reason).
- Conclusion. A one-paragraph statement: either “the cleanroom is operating in conformance with the design intent and the CCS” or “the cleanroom is operating in conformance with the design intent, with the following open CAPAs that will be closed before the next study.” The conclusion is signed by QA, by manufacturing, and by engineering.
Closing: The Smoke Study Is a Verification, Not a Validation
The mental shift that makes the smoke study work is to stop treating it as a validation event and start treating it as a periodic verification. The room was validated at qualification. The smoke study is the evidence that the room is still behaving the way the validation assumed. The cost of running the study annually is small compared to the cost of a particle excursion that the static and dynamic protocols would have caught if they had been run. A working smoke study is a 30-minute video and a six-page report, run annually, with the conclusions signed and the corrective actions closed.
If you are planning a smoke study or reviewing an existing one, we can share a draft protocol template, a pass-criteria checklist, and a report outline, typically within two business days. Reach out with your cleanroom class, the date of your most recent study, and the pass/fail observations from the last report.