IQ, OQ and PQ explained

Qualification builds evidence in layers. Each stage has a protocol with pre-defined acceptance criteria, an execution record, and a report approved before the next stage begins — or with a documented justification where stages overlap.

For a temperature-controlled chamber or room

• IQ: equipment identity, utilities, sensor positions, calibration certificates for controllers and monitoring sensors, drawings, manuals, software version.
• OQ: control at set points across the range, alarm function, door-opening recovery, power-failure behaviour, empty-chamber temperature distribution.
• PQ: loaded mapping under routine operating conditions, over a duration that captures normal use, against the product-derived acceptance criteria.

The mapping study is therefore not a separate activity bolted onto qualification; it is the substance of OQ and PQ for temperature-controlled spaces.

The three questions, and the evidence each one needs

Where teams actually get into trouble

Three errors account for most of it. The first is running PQ on an empty room because the product had not arrived. An empty room has unobstructed airflow and flatters the equipment; the loaded study is the one that supports a storage decision, and a PQ performed empty places no limit on how full the room may later be packed.

The second is using the room's own control sensor as the qualification sensor. Ask the instrument that is holding the set point whether the set point is being held and it will always say yes. Qualification sensors are independent, calibrated, and placed where product sits, not where the controller sits.

The third is treating the stages as paperwork to be signed in sequence rather than gates. Starting OQ before IQ is closed means testing operation of an installation nobody has confirmed; when OQ then fails, the investigation has to consider installation as a cause, and half the evidence is not yet recorded.

Overlapping stages, done properly

Annex 15 allows stages to be combined or to overlap where justified, and for small equipment a combined IQ/OQ is normal and sensible. What it does not allow is the justification being left implicit. Write down which activities were merged and why, and confirm that no test lost its prerequisite — a calibration certificate that IQ would have checked is still needed before an OQ test relies on that sensor.

The mapping study is not an extra

• Empty-room distribution is an OQ test: it characterises the equipment without the confounding effect of load.
• Loaded mapping is the PQ: it is the performance evidence, and the duration has to cover the operating cycle rather than a convenient weekend.
• The monitoring sensor positions come out of the mapping, so the monitoring system cannot be finally qualified before the mapping is analysed.
• The worst-case locations the study identifies become the reference points for every periodic review that follows.

A useful test of a qualification package: could a reader who has never seen the room tell, from the documents alone, where the warmest location is and why the monitoring sensor is where it is? If not, the mapping has not been connected to the rest.