Every biocontainment facility is rated by a Biosafety Level (BSL 1–4): each level up adds another physical/procedural barrier around the hazard and raises how well any single barrier blocks an agent. Here each active shell around the core is one such barrier. An agent released from the core drifts outward and, each time it reaches a shell, either bounces back (contained) or slips through, with a per-shell breach chance set by the pathogen's risk factor discounted by that barrier's strength. Only an agent that breaches every active shell in a row counts as escaped; a monitoring/response layer can still catch it afterwards, the way real biosurveillance intercepts a release after the fact.
p_shell = risk × (1 − 0.15·BSL) [per-barrier breach chance]
P(escape) = p_shell ^ BSL [must breach all N=BSL shells]
P(recapture | escaped) = 1 − (1 − monitorRate)^dt
- BSL containment level — number of active barrier shells (1–4) and, since each higher-rated barrier is also individually stronger, both factors compound: BSL-4 is far more than 4× safer than BSL-1.
- Pathogen risk factor — the agent's inherent hazard (transmissibility/virulence): how likely it is to get past any single barrier before containment strength is applied.
- Monitoring / response rate — biosurveillance intercepting agents that already escaped, independent of the physical barriers — a second line of defense once containment has failed.
- Agent spawn rate — how often the core releases a new agent; a busier facility gives the same risk formula more trials per second.
Real-world relevance: this is why biosafety regulation (Cartagena Protocol, WHO/FAO lab-biosafety guidance, EU GMO directives) mandates BSL ratings by hazard, not by intent — containment strength has to scale with what could go wrong, and monitoring is treated as a backstop, never a substitute for physical barriers.