Annex 1 (2022) asks for continuous Grade A particle monitoring, ≥ 0.5 and ≥ 5 µm, through all critical processing at no less than 28 L/min. That decides where the probe sits, how short the sample tube must be, and that the counter clock matches the batch record. Fix it before the barrier is ordered.
ISO 14644-1 sets class limits at six particle sizes, and ISO 7 has limits at only three: 352,000 at ≥0.5 µm, 83,200 at ≥1 µm and 2,930 at ≥5 µm per m³. A URS that states the class without the size and occupancy state leaves both to the protocol writer.
ISO 14644-1 allows a room to be classified as-built: installation complete, services running, no equipment, materials or people. On a new suite it shows whether the HVAC, filters and envelope can reach the class alone. It says nothing about machines or people, so the GMP classification is still done at rest and in operation.
FDA's 2004 aseptic guidance asks for at least 10–15 Pa between rooms of different classification; Annex 1 gives at least 10 Pa as guidance. Setting each cascade step at 15 Pa meets both. Holding it is harder: seals wear and filters load, so trend the differentials instead of waiting for alarms.
In unidirectional airflow at 0.45 m/s, a 28.3 L/min particle counter needs an inlet about 37 mm across to sample isokinetically, and a 100 L/min counter about 69 mm. An undersized inlet draws air in too fast, large particles miss it by inertia, and the ≥5 µm count reads low.
When two OSD suites share a recirculating AHU, the return duct connects them whatever the layout shows. Source extraction, HEPA on recirculation, dedicated AHUs or once-through air each have a cost. The choice belongs in the contamination control strategy, with the AHU zoning drawing attached.
A contamination control strategy spread across dozens of documents is hard to present in one inspection day. A storyboard with one page per risk, carrying design controls, SOP references, monitoring trends, the owner and open CAPAs, gives the team a way to walk an inspector through it without rewriting the strategy.
EU GMP Annex 1 does not predetermine in-operation particle limits for Grade D. Copying the at-rest 3,520,000 per m³ into that column can put the action level 8.8 times above a room's worst day. Set alert and action levels per room from PQ and routine data, and justify them in the CCS.
Grade A allows 3,520 particles ≥ 0.5 µm per m³ at rest and in operation, and continuous monitoring shows it held. That only works if the sampling head sits near the point of fill facing the airflow, on short tubing that does not lose the ≥ 5.0 µm particles Annex 1 also limits.
Annex 1 (2022) expects an aseptic process simulation to imitate routine processing with worst-case conditions. An APS run right after a filter change and rebalance tests a different process. Schedule maintenance away from APS runs and record the HVAC state for each run like the interventions.
Mixing air lowers the average particle count but cannot sweep a particle off an open vial. Unidirectional airflow is checked by air speed, with an Annex 1 guidance range of 0.36–0.54 m/s, and a filmed smoke study; a mixed-flow room by its 100:1 recovery time (ISO 14644-3).
A microbiology lab protects media and samples while containing cultures, so pressure direction changes by zone: media prep and sterility testing positive, positive-control rooms negative with their own exhaust. A door straight between positive and negative rooms loses the difference whenever it opens; an airlock keeps one door shut.