Smoke Study Runbook

A smoke study runbook is the formal, controlled procedure that specifies how to design, execute, document, and approve airflow visualization studies in cleanrooms and aseptic processing areas. It operationalizes regulatory expectations by enumerating scenarios to film (routine setup, interventions, abnormal events), defining the environmental state (HVAC mode, grade, differential pressures), fixing camera viewpoints and lighting, and prescribing how smoke is generated, introduced, and observed to verify first-air protection and unidirectional flow around critical zones.

Beyond the filming plan, the runbook defines objective acceptance criteria, evidence capture (video, still frames, annotated screen-captures), metadata requirements (time, location, equipment ID, operator ID), contemporaneous notes, deviation thresholds, and approval routing. When executed in an MES, it enforces sequencing, qualifications, and electronic signatures with complete audit trails, enabling repeatability, comparability across campaigns, and readiness for regulatory inspection.

EU GMP Annex 1 (2022) requires airflow visualization studies to demonstrate that unidirectional airflow and first air are not compromised during dynamic operations and interventions; studies should be recorded, reviewed, and used to justify facility design, operator practices, and media fill interventions. FDA’s aseptic processing guidance similarly expects the use of smoke studies to verify laminar flow patterns, identify turbulence or reflux, and provide documented evidence to support aseptic line design and interventions. While 21 CFR 211 mandates appropriate aseptic facility design and documentary controls, it is the FDA guidance and Annex 1 that articulate how airflow visualization is expected to be planned, executed, and recorded.

Because smoke studies generate electronic evidence and approvals, Annex 11 and 21 CFR Part 11 apply to the computerized systems that create, store, and manage these records (video files, annotations, audit trails, approvals). GAMP 5 provides a lifecycle framework for specifying, validating, and governing the MES/evidence repository; MHRA data integrity guidance frames ALCOA+ expectations across metadata, time synchronization, and review processes. ISA‑95 scopes how such procedures and evidence flow across Level 2–4 systems.

A robust runbook is risk-based and process-specific. It must cover Grade A/ISO 5 critical zones and their Grade B background, and any Grade C/D interactions that could affect critical areas (e.g., door openings, pass-throughs). It should enumerate all routine operations and foreseeable non-routine interventions with credible risk to first air: component loading, needle/baffle/stopper bowl adjustments, operator gown repositioning, removal of fallen units, format part changes, line stops/restarts, and decontamination cycles transitioning to production.

The boundary conditions include HVAC mode, HEPA status (post-integrity test), line speed, equipment states (e.g., RABS doors latched, isolator gloves intact), environmental monitoring status, and worst-case load for components/tooling. Risk ranking guides prioritization of scenarios and determines the minimum set of camera angles and smoke vectors required to reveal potential entrainment, eddies, or backflow into the critical zone.

• In-scope: Grade A zones around open product pathways, aseptic connections, filling needles, stopper placement, lyophilizer loading/unloading, open vial exposure points.
• Out-of-scope: Areas with no credible pathway to impact Grade A first air (e.g., sealed conveyors post-closure), unless change control dictates otherwise.
• Change triggers: Equipment reconfiguration, airflow balance changes, HEPA relocation, new interventions, speed/throughput increases, or media fill redesign.

Core content the runbook must cover

• Objective and scope: facilities, units, critical zones; rationale linked to risk assessment.
• Preconditions: HVAC state, HEPA integrity verified, DP/temperature/humidity bands, RABS/isolator states, EM baseline status, equipment ID and revision.
• Instrumentation: smoke generator type (particle size/temperature), camera specifications (FPS, resolution), lighting plan, tripods/booms, calibration checks and time sync.
• Test matrix: routine and non-routine interventions, worst-case speeds/loads, door states, operator positions, number of replicates.
• Camera plan: fixed viewpoints that show first-air path, orthogonal angles to visualize eddies, and reference grids/scales.
• Acceptance criteria: no obstruction of first air over critical surfaces, no reflux from Grade B to A, dissipation time within predefined envelope, no stagnant zones around interventions.
• Evidence management: file naming, chain-of-custody, metadata schema, storage, retention, reviewer checklist, approval routing, deviations/CAPA links.
• Periodic review and triggers: requalification cadence and change control criteria to re-run scenarios.

1. Map process/line PFD and 3D layout to identify Grade A/B interfaces and first-air vectors.
2. Derive the intervention set from SOPs, media fill design, and risk analysis; include worst-case crew size and complexity.
3. Define smoke ingress points that minimally disturb flow while revealing entrainment/eddies; dry-run with cameras only.
4. Establish acceptance criteria and objective measures (e.g., dissipation timing markers) before filming to avoid bias.
5. Pilot one scenario to validate camera coverage, lighting, and metadata capture, then finalize the matrix.

Select smoke generators that create neutrally buoyant, visible aerosol with minimal thermal plumes that could artificially distort airflow; document particle characteristics and flow rates. Cameras should record at sufficient frame rate and resolution to discern laminar fronts and small eddies, with stable mounts and glare-controlled lighting. Time synchronization across devices is critical; all feeds must share a common timebase so operator actions, equipment states, and EM readings can be correlated during review.

Evidence governance must meet ALCOA+ principles. Define a metadata schema (area, equipment ID, scenario ID, operator ID, start/stop timestamps, environmental readings) and enforce it at capture time. Video files must be stored in controlled repositories with versioning, access control, and audit trails; annotate key frames non-destructively and maintain reasoned conclusions and approvals as Part 11/Annex 11-signature-controlled records. Reviewer checklists should explicitly confirm first-air protection, absence of reflux, and alignment with intervention SOPs/media fill design.

Treat the runbook as a master recipe (ISA‑88) instantiated into a batch of study runs. Unit procedures can represent major areas (e.g., vial infeed, filling, stoppering, lyophilizer load), operations are individual scenarios (e.g., needle adjustment), and phases are granular steps (prep, camera sync, smoke release, observation, shutdown). Procedural control enforces prerequisite checks (HVAC state, EM baseline captured), operator qualifications, and equipment status before allowing smoke release or recording to start.

Within the MES, force signatures at critical control points (e.g., camera ID/time sync verification, acceptance criteria confirmation), capture structured observations, and generate an immutable execution history tightly bound to the video artifacts. Exceptions should trigger deviation workflows with containment/impact assessment and corrective actions; completed studies should auto-link to media fill protocols, risk registers, and change records, enabling closed-loop compliance and rapid inspector recall.

Smoke study execution spans ISA‑95 levels. Level 2 systems (SCADA/PLC) provide equipment states (e.g., line speed, RABS door interlocks) that contextualize the video; Level 3 (MES) orchestrates the runbook, enforces signatures, and aggregates evidence; Level 4 (ERP/QMS) maintains change controls, training, deviations/CAPA, and archiving policies. Particle counters and EM systems feed measurement data that can be fused with video timestamps; CMMS contributes maintenance/HEPA certifications to prove readiness-state.

Define clear, observable acceptance criteria before execution. Typical criteria include: (1) uninterrupted first-air protection of exposed sterile surfaces and product pathways; (2) absence of visible turbulence, recirculation, or impingement that could carry contaminants from non-critical zones; (3) smoke dissipation times within predefined envelopes; and (4) no reflux from Grade B background into Grade A. Criteria should be specific to geometry, airflow velocities, and intervention posture, and linked to risk controls in SOPs and media fill design.

The review process should use multi-observer sign-off with pre-agreed checklists. Where borderline phenomena are observed, the reviewer documents rationale, risk impact, and any compensatory controls (e.g., modified intervention technique or additional barriers). Deviations initiate CAPA and may trigger re-study or design changes. Final approval notes whether EM during the study remained within limits, correlates observations with equipment states, and explicitly confirms continued validity of the media fill intervention set.

Annex 1 expects airflow visualization at initial qualification and after significant changes; sites typically adopt periodic requalification to confirm sustained state of control. The runbook should codify triggers that mandate re-study: layout changes, airflow rebalancing, HEPA relocations or failures, new equipment or robotics, new or modified interventions, or deviations suggesting compromised first-air protection. It should also define sampling of scenarios during periodic requalification to balance assurance with operational burden.

All changes must pass through formal change control with documented risk assessments tying proposed modifications to airflow implications, updated camera plans, and acceptance criteria. When media fill designs change (e.g., intervention list revisions), smoke study plans must be reviewed for alignment. The runbook should prescribe archiving/retention policies for videos and metadata, ensuring inspector-ready retrieval throughout the qualification lifecycle.

• Unrepresentative smoke: Thermal or momentum effects from the generator distort airflow; mitigate by validating particle/flow characteristics and using minimal energy injection.
• Insufficient camera coverage: Single angle misses eddies; mitigate with orthogonal fixed viewpoints and rehearsal dry-runs.
• No time-sync across streams: Inability to correlate actions, EM, and equipment events; mitigate with a single timebase and pre-run sync checks.
• Vague acceptance criteria: Reviewer bias and inconsistent decisions; mitigate with pre-defined, measurable criteria and structured checklists.
• Evidence sprawl: Videos on shared drives without metadata; mitigate with controlled repositories, mandatory metadata, and Part 11-compliant audit trails.
• Runbook drift: Informal additions not reflected in controlled documents; mitigate with change control and periodic review against Annex 1/FDA guidance.
• Media fill misalignment: Intervention sets diverge from observed practices; mitigate with cross-reviews and joint approvals between manufacturing, QC microbiology, and QA.

V5 executes smoke study runbooks as ISA‑88 recipes within MES, enforcing prerequisites (HVAC/HEPA/EM checks), operator qualifications, and forced-signature gates. It captures structured observations and binds them to tamper-evident video artifacts with synchronized timestamps. Interfaces to SCADA/PLC and EM instruments provide contextual event frames; LIMS data and calibration/CMMS certificates are linked to prove readiness-state; and QMS workflows manage deviations/CAPA and change control. Approval routing follows Part 11/Annex 11 controls with complete audit trails and reviewer checklists.