Humidity monitoring in storage areas: when and how

Relative humidity matters for hygroscopic products, moisture-sensitive packaging, and in stability chambers where it is a controlled condition. Warehouses storing ambient products commonly monitor RH to demonstrate it stays below a defined ceiling, though the requirement is often less explicit than for temperature.

Practicalities

• Place RH sensors with temperature sensors: RH is temperature-dependent and a 1 °C error at 50 % RH corresponds to roughly 3 % RH.
• Avoid positions near doors, wet processes and direct airflow from HVAC outlets unless that is where the risk lies.
• Calibrate at points spanning the working range (e.g. 30, 50, 75 % RH) using salt solutions or a humidity generator.
• Set alarm limits with the sensor's uncertainty in mind; a ±3 % RH sensor cannot police a 60 % RH limit at 59 %.

Why RH uncertainty is larger than you expect

Relative humidity is a ratio of two vapour pressures, and one of them depends on temperature. That makes every RH reading partly a temperature measurement, and the temperature error propagates. Near room conditions the sensitivity is roughly 3 % RH per °C at 50 % RH: a sensor whose temperature channel is 1 °C out reports a humidity that is about 3 % RH out even if its humidity element is perfect.

Add the element itself. A good capacitive polymer sensor is specified at ± 2 % RH over part of its range and ± 3 % or worse at the extremes, drifts by 0.5 to 1 % RH per year, and shows hysteresis of a percent or two after a long spell at high humidity. A realistic combined uncertainty for a monitoring-grade RH sensor is 3 to 5 % RH, not the ± 2 % on the datasheet.

Calibrating an RH sensor

Two practical methods. Saturated salt solutions in a sealed chamber give fixed, well-characterised humidities at a known temperature — lithium chloride near 11 % RH, magnesium chloride near 33 %, sodium chloride near 75 % and potassium sulfate near 97 %, all at 25 °C. They are cheap and slow: allow hours for equilibrium, and keep the temperature stable, because the salt values shift with it. A two-pressure or divided-flow humidity generator is faster and covers the range continuously, and is what a calibration laboratory will use.

Calibrate at points that bracket the limits you actually police. Checking at 33 % and 75 % tells you little about a sensor guarding a 60 % ceiling if its error is nonlinear in between.

Placement, and when RH is worth monitoring at all

• Monitor RH where the label or the stability data says so, where packaging is moisture-sensitive, or where the product is hygroscopic. Monitoring it everywhere else generates alarms nobody can act on.
• Pair every RH sensor with a temperature sensor at the same point, because an RH figure without its temperature cannot be interpreted or corrected.
• Keep sensors away from doors, loading bays, wet cleaning routes and the direct discharge of an HVAC outlet — unless those are the risk, in which case put one there deliberately and say so.
• Expect a seasonal pattern. A warehouse that sits at 45 % RH in February and 68 % in August has not developed a fault; a single winter study simply did not see the summer.
• Set the alarm limit inside the specification by at least the sensor uncertainty. A ± 3 % RH sensor cannot credibly police a 60 % limit at 59 %.