Cleanroom recovery and contamination control testing

A cleanroom that passes a particle count at rest has proven only one thing: on that day, with no people and no process running, the air was clean enough. Contamination control testing asks harder questions. How quickly does the room clean itself after a burst of particles? Are the filters and their seals intact? Does air move from clean to less clean areas, or does it eddy and stall around equipment?

These tests are described in ISO 14644-3, the test methods part of the ISO 14644 series, and many are referenced by EU GMP Annex 1 and Health Canada's GMP guidance for sterile products. This guide explains what each test does, how it is performed and what the results mean for an owner or operator.

Why classification alone is not enough

ISO 14644-1 classification is a snapshot of airborne particle concentration at defined locations. It does not tell you whether the airflow pattern protects the product, whether the room recovers after a disturbance, or whether a filter seal is quietly bypassing unfiltered air. For that reason, a full certification or qualification normally combines classification with a set of supporting tests. The core ones are summarized below.

Recovery testing

What recovery time means

Recovery time is the time needed for the particle concentration at a location to fall by a stated ratio after the room is deliberately contaminated. ISO 14644-3 defines a 100:1 recovery time, meaning the time for the concentration to drop to one hundredth of the initial challenge level. Where a high enough challenge cannot be generated without overloading the counter, a smaller ratio such as 10:1 can be used, provided the method is documented.

Recovery testing applies mainly to non-unidirectional (mixed airflow) rooms. In a unidirectional zone, such as a Grade A filling area, contamination is swept away almost immediately and the more meaningful test is airflow visualization.

How the test is performed

1. Confirm the room is operating at its normal air change rate and pressure, with the HVAC system stable.
2. Place a particle counter at the test location, usually away from supply diffusers and representative of the critical work area.
3. Generate a challenge aerosol, typically with a particle generator or by briefly disturbing the room, until the concentration is well above the class limit.
4. Stop the challenge and log particle counts at short, fixed intervals.
5. Record the time for the concentration to fall by the target ratio, and plot the decay curve.

Reading the result

In a well-mixed room, particle decay is roughly exponential. The theoretical time for a 100:1 reduction is about 4.6 divided by the air change rate, expressed in hours. At 20 air changes per hour that is about 14 minutes; at 40 ACH, about 7 minutes. A measured result much longer than the theoretical value points to poor mixing, short-circuiting between supply and return, or dead zones behind equipment.

EU GMP Annex 1 also uses the related idea of a clean-up period: the time to return from in-operation conditions to the at-rest limits, with a guidance value of 15 to 20 minutes. Recovery data supports the airflow and air change design discussed in air change rates and airflow in cleanrooms.

Installed filter leak testing

A HEPA filter that met its factory efficiency rating can still leak after installation. Damage during handling, a poorly seated gasket, a gap in a gel seal or a cracked frame all let unfiltered air bypass the media.

The installed leak test introduces an upstream aerosol challenge, commonly a polyalphaolefin (PAO) or similar oil aerosol, and scans the entire downstream face and perimeter with a probe. Two detection methods are common:

• Aerosol photometer: measures the downstream concentration as a percentage of upstream. A reading above the acceptance threshold, often 0.01 percent of upstream for HEPA filters, is treated as a leak.
• Discrete particle counter: counts particles in a size range downstream. It needs a much lower upstream challenge, which suits rooms where aerosol deposition is a concern.

Scanning too fast or too far from the face can miss pinhole leaks. Repairs are usually limited to a small percentage of filter face area by a sealant approved by the filter manufacturer; larger damage means replacement. More on filter types and housings is in HEPA and ULPA filter systems.

Airflow visualization

Smoke studies, or airflow pattern visualization, show how air actually moves. Operators release a visible tracer, often water-based fog or a glycol fog, at supply outlets and around critical points, then record video.

What testers look for:

• Air moving smoothly downward and away from exposed product in unidirectional zones.
• No reverse flow or eddies over open containers, fill needles or transfer points.
• Air flowing outward through doors and pass-throughs from the cleaner room to the less clean one.

Annex 1 expects airflow visualization studies for Grade A and B areas, ideally performed both at rest and under simulated operation.

Containment leak testing

Cleanrooms are usually pressurized so that air leaks out, not in. Containment leak testing checks that particles are not entering through ceiling grid joints, light fixtures, wall panel seams, conduit penetrations or door frames.

The tester raises the particle concentration on the outside of the barrier, then scans the inside surface with a particle counter probe. A significant rise in downstream counts near a joint indicates a leak path.

Pressure, velocity and supporting measurements

Recovery and visualization results only make sense when the room is operating as designed. Supporting measurements typically include:

1. Supply airflow and air change rate: measured by capture hood at each diffuser or filter, or by velocity traverse at filter faces.
2. Unidirectional velocity and uniformity: measured on a grid below filters. A typical design range for unidirectional zones is about 0.36 to 0.54 m/s (70 to 105 fpm) under EU GMP guidance, but the validated value is set by the user's requirements.
3. Pressure differentials: measured between each room and its neighbours. Common design steps are 10 to 15 Pa (0.04 to 0.06 in. w.c.) between grades, as described in cleanroom pressurization cascades.

Test frequency and states

ISO 14644-2 sets out monitoring and periodic testing intervals based on risk. Classification for cleaner classes is commonly repeated at intervals of 6 to 12 months, with supporting tests on a frequency set by the site's contamination control strategy. Tests should also be repeated after any change that could affect airflow: filter replacement, layout changes, new equipment, or modifications to the air handling unit.

Every result should state the occupancy state: as-built, at rest or operational.

Frequently asked questions

Does every cleanroom need a recovery test?

Not always. It is most useful for non-unidirectional rooms and is expected for sterile manufacturing under GMP. For lower-risk ISO Class 8 support spaces, the owner's risk assessment may decide classification and pressure checks are sufficient.

What if recovery time is longer than expected?

Check the measured air change rate first, then look at the location of supply diffusers and low returns relative to the test point. Equipment layouts that block return grilles are a frequent cause.

Key takeaways

• Classification counts show cleanliness at a moment; recovery, leak and airflow tests show how robust that cleanliness is.
• ISO 14644-3 defines the 100:1 recovery time; in a well-mixed room it should be close to 4.6 divided by the ACH, in hours.
• Installed filter leak scans catch damaged media and bad seals that a factory rating cannot.
• Smoke studies are the key evidence for unidirectional zones and are expected under EU GMP Annex 1.
• Always record the occupancy state and repeat tests after any change that affects airflow.
• Independent building commissioning and testing ties these results back to the design intent.