The October 2010 column discussed the design principles and practical considerations that govern the performance of containment barriers in BSL-3 laboratory and animal facilities. It provided the safety rationale and operational considerations used to assess the appropriate barrier system specifications for different scientific activities and facility operations. The article outlined: the performance criteria and justification for making containment barriers airtight, impervious, chemically and mechanically resilient for safety; maintainability, and cost. Within those considerations, special attention was given to the question of air leakage; all biocontainment guidelines and inspection documents require airtight containment boundaries. The reasons are partly obvious—the need to prevent the escape of airborne hazards—and partly not so obvious; the most likely dangerous airborne agent in a BSL-3 room and the biggest challenge to contain are decontamination gas or vapors (in most cases).

In theory, the air impermeability of the containment room system (barrier integrity) is straightforward. In practice, things can and often do go sideways in the transition from concept to implementation of this basic containment barrier requirement. The BMBL’s sparse language (amplified by a lack of verification criteria in our licensing inspection protocols) is without question open to interpretation. This lack of definition or direction has caused a history of problems and less than optimum results in the construction of containment laboratories—costing facility owners money, loss of research opportunity, and certification “hardship.”

How do we manage these risks effectively? As with most things, early prevention is the best cure.

The Status Quo: Open to Interpretation
The BMBL gives very little guidance to the appropriate level of BSL-3 containment barrier integrity, sparsely stating: (the laboratory) “surfaces should be sealed” and “openings should be capable of being sealed to facilitate space decontamination.” In addition, there are no criteria for verifying that these conditions have been achieved in the BMBL or in any official certification checklist for BSL-3. Official inspection criteria, if they say anything, only require a visual inspection to validate a containment boundary is sufficiently airtight. But it’s impossible to see microscopic and hidden air leakage pathways; this method of validation cannot provide accurate or adequate quality assurance for these important construction criteria. It often can produce the perfect storm: a lack of defined performance measures or validation criteria with huge variability in how much or how little we can build/spend to achieve what we think are the appropriate results. This often means owners not getting what they’ve paid for or constructors having to unfairly spend more time and effort than they reasonably budgeted for chasing undefined acceptance criteria. I won’t comment on which scenario is more common.

Before we present alternate strategies and recommendations for barrier integrity assurance, we should defend the status quo (ambiguity) on one level: open to interpretation can be a good thing when it means that the answer is context-based and derived from risk assessment. “Capable of being sealed to facilitate space decontamination” can vary from application to application and with operational parameters such as decontamination frequency, timing, the proximity of occupied zones to the decontamination zone, the pressure relationships during fumigation that could induce gases to migrate, and research programs that have to remain in operation while the fumigation process is occurring (i.e. an animal experiment in an adjacent suite). An airlock on the boundary of an operating BSL-3 suite where equipment will be frequently decontaminated should be capable of being sealed very gas tight, even if the procedure will occur during unoccupied (night) hours. Conversely, I have worked with owners who appropriately decided to construct certain containment zones less gas tight than normal after careful risk assessment that determined aerosol release or significant accident with infectious material was extremely unlikely (based on special protocols) and their acceptance that, in the unlikely event that this space had to ever be fumigated, the entire building would remain empty as long as required. This was not a cost driven decision but instead was based on the user requirements activity type for the space, which made it unfeasible to construct to normal containment integrity standards (and unnecessary).

Quantitative and Qualitative Validation
The question still remains: for BSL-3 containment, how sealed is sealed?

We aren’t sure, frankly.

Methods to validate the air leakage of a containment boundary can be classified into two main categories: qualitative and quantitative. Both approaches can be effective. Neither is necessarily better and often a validation strategy will include both qualitative and quantitative methods in a quality assurance program. Problems occur when no validation is done. We will briefly describe the most common methods and their pros/cons.

The only quantitative test prescribed in the U.S. and Canada is the Pressure Decay Test. It is performed by sealing all room penetrations and drawing very deep pressure differences across the containment barrier, shutting down all air supply/ exhaust to the space and witnessing the “decay” of the pressure difference as the room trends towards equalizing its pressure with its surrounding environment. A room is said to pass the test when: starting measurements at a 2” WC (500 Pa) pressure difference across the barrier; after 20 minutes in closed static mode, the pressure has decayed (equalized) less than 50% and is still more than 1” WC (250 Pa) difference. The test can be performed as either a positive or a negative pressure test.

The Pressure Decay Test is an extreme test in that it represents our highest standard for room integrity in the U.S. and Canada. It is designed for BSL-3Ag/BSL-4 containment construction validation, which is historically constructed with specialized and very expensive high performing containment barrier systems. Typical methods for constructing BSL-3 space will not pass its test criteria. As discussed in the previous article, the risks associated with aerosol hazards in BSL-3 space do not warrant constructing barriers that will pass this test.

Beyond certain failure to achieve the test—which should be enough of a deterent—the test can subject rooms to pressures outside their structural tolerances and can cause significant damage, especially if the test is performed as a negative pressure test. The room can go deep enough negative for a long enough period to implode walls and ceilings if they are constructed from drywall or panel construction and not designed for such deep pressures. This should not happen at -2” WC (-500 Pa) but room pressures can go much lower than that in the course of a decay test. The Pressure Decay Test is simply not well suited for BSL-3 performance validation.

Without a relevant quantitative test to use, the most common methods of testing BSL-3 barriers are not quantitative but are qualitative tests designed to find holes in the barrier and fix them. The idea being that if you eliminate all the leakage paths, the room will be tight enough, i.e. fit for purpose. That is pretty sound logic overall, but the devil is in the details and in how you record those details in the course of a construction project.

The two most common methods of identifying air leaks are smoke testing and soap bubble testing. Both tests rely on placing the containment room under negative pressure and scanning all potential leakage points within the room with an agent that will react to air infiltration. That agent can either be a gentle plume of smoke moved over the surface and witnessed for changes in the smoke plume (air turbulence at the leakage point), or a layer of soapy water applied to the barrier surface and checked for the formation of soap bubbles, which indicates that air is infiltrating into the room at that point—a leak! Both methods can be effective in assuring the room will be airtight enough to prevent unwanted escape of aerosols if done properly.

Regrettably, there is no guidance or consensus on what parameters should be used to perform these tests. You have to define them yourself on the basis of laboratory operations and failure scenarios. A common concept used is that the room should be tested for leaks at operating pressure. I cannot tell you how many hours of debate I have had with experts as to how it translates into an appropriate test parameter. Most BSL-3 labs in the US operate with a design value of --0.05” WC (-12 Pa) pressure between zones. Does that become the test parameter for each room tested? Rooms that are several pressure cascade steps into a BSL-3 facility (the most hazardous areas) can operate at much higher negative pressure—sometimes close to -0.5” WC (-125 Pa) or higher. Should these rooms be tested at this higher pressure? Should all rooms be tested at this higher pressure for consistency?

What constitutes pass or fail for these tests: zero smoke plume movement, which is very dependent on the hand that’s holding the smoke generating device as it is passed in front of a potential leakage point?No formation of soap bubbles for that test is pretty easy to understand and agree upon. For that reason, I like this test much better. However, covering all joints and penetrations in soapy water is messy and there’s a lot of cleaning up to do after a room soap bubble test.

In Part 2, I’ll discuss recording test data, defining integrity testing, and offer thoughts on developing a test standard using approaches developed in other countries as a starting point.

Randy Kray, AIA is a Senior Laboratory Planner and the Director of Laboratory Architecture for Merrick and Company, an international full-service laboratory planning, design, construction administration, commissioning, and lab operations planning firm. Randy has over 21 years of diversified experience in complex life science, health and technology driven facilities. He has focused his expertise for the past ten years exclusively on the planning, design, and construction of life science facilities world-wide. In addition to his ongoing project commitments, he is a regular contributor of journal articles and public presentations in an effort to learn and contribute to the understanding of laboratory design and biocontainment principles and best practices.