Cold room temperature mapping

A 2–8 °C cold room is small enough to map thoroughly and busy enough to surprise you. Typical gradients run from the evaporator outlet (coldest, sometimes below 2 °C) to the door and the top shelves furthest from the airflow.

Placement

• Corners at low and high level, the geometric centre, directly in front of the evaporator, and beside the door.
• Product height, not the ceiling: sensors should sit where product actually sits.
• One sensor outside the room as an ambient reference helps explain door-related events.

What to expect in the data

Defrost cycles show as periodic rises at the evaporator sensor; door openings as short spikes near the door that recover within minutes. Neither is necessarily an excursion, but both must be assessed against the limits and durations in the protocol. Run the study for several days so that daily patterns and weekday deliveries are captured.

Reading the three signatures in the data

Almost everything interesting in a cold-room dataset is one of three patterns, and telling them apart is most of the analysis.

Defrost and door events are not automatically excursions, but they are not automatically excusable either. Both get assessed against the limits and durations the protocol set, and both need the ambient reference sensor to interpret: a rise at the door sensor that coincides with nothing outside is not a door opening.

The evaporator, which is the cold spot

In a 2–8 °C room the freezing risk is real and it is directly in the evaporator discharge, where air leaves several degrees below the room set point in order to average out to it. Product stacked in that airflow can sit below 0 °C while the room reads 5 °C, and for vaccines and biologicals that is an irreversible loss that no averaging will show.

So one sensor goes directly in the discharge, deliberately, and the report says what it found. If that position is below the lower limit, the answer is an exclusion zone — marked physically, not just in a document — or a deflector, not a wider acceptance criterion.

Duration and timing

• Run at least 72 hours so that several complete defrost cycles are captured. A 24-hour study may see one, or none, and cannot show whether they are consistent.
• Include a normal working day with real door traffic. A weekend study characterises a room nobody is using.
• Log every 1 to 5 minutes. Door events are short, and a 15-minute interval can put a whole excursion between two readings.
• Place one sensor outside the room. It costs one logger and explains most of the features above.
• Let sensors equilibrate before the study clock starts — a logger carried in from a 20 °C corridor takes time to settle, and the first hour will otherwise look like a warm spot.