Genome engineering
Biosafety and laboratory biosecurity
Directional airflow and pressure cascades, why the cabinet is the primary barrier and the room the secondary one, how HEPA filtration actually works, and where human factors set the real limit.
Biosafety protects people and the environment from the organism; biosecurity protects the organism — or the information about it — from people. The two use overlapping hardware and answer to different threat models. This page is about the physical engineering of the first, because that is where there is a mechanism to explain.
Primary and secondary barriers
Containment is layered, and the layers have names because they fail differently. The primary barrier is what stands between the material and the worker at the moment of handling: the biological safety cabinet, a sealed centrifuge rotor, a closed vessel. The secondary barrier is the room and the building — the thing that contains a failure of the primary one.
A Class II cabinet is worth understanding in detail because it is the workhorse and it is routinely misused. It works by pulling room air down through a grille at the front opening, forming an air curtain that carries anything aerosolised at the work surface away from the operator’s face rather than out into the room. Separately, filtered air descends vertically over the work surface to protect the sample. The cabinet therefore protects operator, product and environment only while that inflow velocity is maintained — which is why blocking the front grille with equipment, or working with rapid arm movements through the air curtain, defeats it. A cabinet is not a fume hood and not a sterile box; it is a managed airflow.
Pressure cascades and filtration
At room scale, containment is directional airflow. The laboratory is held at negative pressure relative to the corridor, and the corridor relative to the outside, so that leakage is always inward. This requires exhaust to exceed supply by a designed margin and the room envelope to be sealed well enough that the differential can be maintained; the room is verified with airflow and, at higher levels, pressure-decay testing rather than assumed.
Exhaust passes through HEPA filters. The mechanism is not sieving, and this matters for understanding it: particles are captured by interception, inertial impaction and, for the smallest particles, diffusion as Brownian motion drives them into fibres. Because those mechanisms have opposite dependence on particle size, filter efficiency has a minimum at an intermediate diameter — around 0.3 micrometres — which is why the filter’s rating is quoted at that worst case. Particles both larger and smaller are captured more efficiently.
Where the highest containment actually fails
BSL-4 adds a positive-pressure suit on air supply, chemical shower egress, and effluent and waste decontamination. The engineering at that level is mature and its failures are overwhelmingly not filtration failures. The recurring causes in documented laboratory-acquired infections are procedural and human: sharps injuries, breaches during doffing, autoclave and inactivation procedures not verified, an animal bite, an unrecognised exposure reported late. This is why the operational system — training, competence, occupational health surveillance, incident reporting that is safe to use, inventory and access control on the biosecurity side — is not administrative overhead around the containment. It is the part of the containment with the highest failure rate.