Choosing monitoring locations from mapping results

The purpose of a mapping study is to find where a space behaves worst. Routine monitoring is then placed at those positions so that if the worst location is within limits, everything else is too. A monitoring sensor placed at the control probe's position typically shows the best-controlled point in the room — the least informative place to look.

Selection rules

• Hot spot (highest mean or maximum) and cold spot (lowest mean or minimum) from the loaded mapping.
• Positions with the slowest recovery after door openings or defrost.
• Each distinct zone if the space has more than one HVAC regime.
• Product-height positions, not ceiling or floor, unless product is stored there.

Record the rationale

A monitoring plan should show, for each sensor, which mapping position it corresponds to and why. When the space changes and is re-mapped, the plan is reviewed against the new results.

From mapping result to monitoring plan

The selection is mechanical once the study is done, which is the point of doing the study. Each monitoring sensor should trace to a mapping position and a reason.

Two sensors would be a defensible minimum here. Four is better and the marginal cost is small. What would not be defensible is one sensor at the centre, because the centre is the best-behaved point in the room and a monitoring record from it will say the room is fine while the door shelf runs at 10 °C.

Both extremes, not just the warm one

Monitoring plans routinely cover the hot spot and ignore the cold spot, because upper-limit excursions are the ones people think about. In a 2–8 °C room the freezing risk is the more serious of the two: it is irreversible for many biologicals, it happens in the evaporator discharge where nobody puts a sensor, and an averaged room reading will never show it. If the mapping found a position below the lower limit, that position needs either a monitoring sensor or a physically enforced exclusion zone.

Recording the rationale, and keeping it current

• The monitoring plan names, for each sensor, the mapping position it corresponds to and the reason. Without that link the two datasets are unrelated and neither supports the other.
• Where an existing sensor is not at a worst-case position, the mapping report has to say so and recommend the move — and somebody has to make it. A recommendation never implemented documents that you knew.
• When the space is re-mapped, review the plan against the new results. Racking changes move the worst case, and the sensor stays where it was bolted.
• Cover each distinct zone. A large space with separate airflow regions needs a sensor per region, not one sensor and an average.
• Keep one ambient reference. It costs a channel and explains most of what the others do.