Discussion for establishment of risk based approach on airflow visualization in cleanrooms

Smoke study or air flow visualization is the important factor in clean room qualification.
My query to some experts were as following

1. What will be the requalification frequency for Smoke study?
2. If there is no significant changes in the clean room design is it require to do periodic evaluation of smoke study?
3. Any Guidance available for video shooting?

I have received the following comments from  various experts (shared for public interest)  


Response by Expert 1:
1. After any addition of equipment or modification of process or HVAC/LAF balancing that may affect the airflows, if you follow ISO14644 then they recommend the test is executed every 24 months.
2. I would recommend 12-24 months for both Grade A and B environments and a risk assessment to cover lower grades unless significant changes have been made.
3. Short clips of both 'In operation' and 'at rest' around all equipment/doors/filters/extracts/operators. Look for dead spots, turbulence etc. 
 Finally, the value of the exercise is in the analysis of the data generated. Compare the data against your environmental monitoring locations and use the footage to train operators in clean room behavior.  
Response by  Expert 2:
I think what ISO 14644 actually says is that airflow visualization is an optional test on an ongoing basis that may be included within the testing plan used to prove continued compliance along with the mandatory/required components.  
So, call me lazy but I read that to mean that, once qualified with an appropriate smoke study, there is no need to repeat a smoke study.  It seems pointless to repeat studies of 1) air supplied at the same velocity travelling over the 2) same equipment under 3) the same operations.  Any training benefits could as well be achieved by having new staff review the original studies.  That said, changes in the equipment or operations would trigger the need for an evaluation. If airflow over and around critical spaces might be impacted by a change then a new smoke study would be required.  This leaves the air supply - if equipment and operations have not changed, could changes in the air as supplied impact the airflow? Obviously it could, however I'd argue that if your periodic testing shows no significant change in the face velocities delivered by air supply and the pressurization of the rooms and differential pressures between them is also monitored and shows no drift then this would represent excellent evidence that the air supplied continues to be supplied in manner substantially equivalent to the way it was supplied during the original study and that therefore no repeat study would be required.  
Alternately a change in the supply velocity, particularly a change in the relative velocities delivered across the face of the air supply could make a big difference to airflow. If it seemed like the performance of the supply and exhaust of the room might be performing differently over the years then new smoke studies would be needed.  I don't think any of this requires smoke studies to be repeated at any particular periodicity - new smoke studies should be gated by changes.  So, say you replaced the HEPA units in an old clean room and the new HEPAs delivered air velocities at the face that were different from the old ones.... then I think you'd need a new smoke study.
I have a couple of follow-on questions -
Does anyone execute airflow visualization for anything other than Grade A/B or <= ISO5? If so, what is gained by that? 
Does anyone have information/studies on the nature of failures in HEPA units in critical areas?  Is people’s experience that they eventually start to leak or that they eventually start to clog? I have seen some HEPA units (in spaces where a terminal HEPA filter is supplied with air that is already HEPA filtered) where performance of the unit appears unchanged after 10 years of operations. Procedure dictates that we replace these filters but I wonder how long they might function if unchanged.  How often are others replacing the HEPA units in critical areas and with what rationale?     
As for advice on smoke studies - use more than one camera, have one of your operators loudly and clearly announce what is being tested - preferably reading the language and test step directly from the pre-approved protocol (and then keep other talking to the minimum necessary to execute the study). Have a representative from Quality present during the execution and editing of the video footage. Keep all the original video but edit into clips to show the critical tests so that both the complete and edited versions can be presented to reviewers, auditors or regulators.    
Response by  Expert 3:

I agree it is optional ongoing test, but recommended, we have relatively small Grade A/B areas, so for us it’s not much of a drain in resources. If we noted no significant differences from routine testing (i.e. air changes/filter face velocity/particles/differentials) we would consider re-defining the interval in our VMP.

To clarify, I would not execute this task for new operators, I would hope any observations from the footage are transcribed in to applicable SOPs. I found it useful to demonstrate poor practice and its impact on flows. I would also expect small modifications, for example the addition of a piece of equipment within a VLAF, would be limited to that VLAF.

For the follow on questions:

We executed the initial study in all rooms up to ISO7, however the complexity/depth of the testing reduced. This was basically to show flows from transfer hatches and doors did not flow into higher classification areas. This was only tested at rest. 

For H14s we defined a 5 year life in Grade A and 7 Years in Grade B, some filter manufacturers recommend changing when the pressure drop doubles or at ten years... we based ours on the air supply quality, frequency of DOP testing and carry over from previous organizations. I would also be interested in any studies beyond the ten year life.

Response by Expert 4:

I would be interested in hearing from others, certainly not meaning to speak for everyone.

I would not let more than two years go by without performing the test, and would want to do it annually.  Why?  Something as simple as moving a table against the wall - even though it might only be moved a couple of inches, will make a big difference.  Also, Systems with as many 'moving parts' as an HVAC system for controlled environments will change, albeit in small increments, over time.  Perhaps the fan slows, perhaps pre-filter clogging, perhaps gaskets around doors become less pliable, perhaps ductwork begins to leak.

As to performing the testing in less stringent areas - we have done so periodically, but not on a regular basis.  It is done mostly because the 'coverage' of the HEPA filters is more dispersed and we just wanted to see what was happening.  Most times it was because the activities in the room changed.

As to clogging versus holes - I have found that for the HEPA themselves, the only meaningful clogging that occurs is when we patch holes.  We replace HEPA filters on a 5 year schedule, so we haven't seen clogging.  But holes and leaks happen.  For us, where the housing mates to the frame is the greatest source of leaks, but perhaps it is splashing during room sanitization or process operations that weakens the media that causes holes, or perhaps it is just a weak spot in the crease of the media that finally fails, we find leaks and do not find filter loading.

Response by Expert 5:

Smoke studies should be performed upon initial qualification and anytime modifications to the air handlers or rooms to which service is provided, i.e. adding or removing a wall or large piece of equipment, etc.  Emphasis should be given to hoods, bio safety cabinets, soft walls, isolators, doors, pass through, etc.  The intent is to verify that airflow is direct away from open or exposed product and toward the most "contaminated" environment as they relate to the product or product contact surfaces.

There is no requirement for the frequency smoke studies should be repeated.  Provided there have been no significant changes and environmental monitoring data suggests the "space" is adequately controlled, smoke studies do not have to be repeated each time a space is certified.  Smoke studies should, however be utilized as an investigative tool in the event of frequent environmental monitoring excursions. 

As far as videoing goes, the use of a combination of signage and voice description describing room to room relationship i.e. differential pressure, ISO classification, whether it’s a controlled space, operations performed in the room, etc. 

Attached is a very good article on Smoke Studies form the Journal of Validation Technology that I am sure will be helpful.

Keep the questions and supportive answers coming!

USCGMP