A gowning room can pass every OQ test and still fail its first full shift change.
Pressures hold, recovery is inside the limit and particle counts are clean at rest. Then, say, 12 operators arrive within 15 minutes, the interlocked doors cycle faster than the cascade recovers, and the in-operation counts drift up. Nothing was wrong with the tests. The shift change was never written into the URS, so DQ never checked the design against it and OQ had nothing to simulate.
The V-model works in both directions. Each test stage proves a document on the left, and DQ checks the design against the URS before anything is bought:
URS states what the process needs, with numbers.
DQ confirms the design meets the URS, before anything is bought.
IQ proves what was installed matches the approved design.
OQ proves the system works across its operating range.
PQ proves it performs under routine production, over time.
A requirement missing on the left cannot be proven on the right. The fix is cheap if it happens early: write the operating scenarios into the URS as numbers. Operators per shift change, door openings per hour, the longest stop before a restart.
Topics




