Humidity calibration
Relative humidity is a ratio that changes with temperature, so the temperature measurement is half of the job. This page covers the common sensors, how they are checked with salt solutions and reference hygrometers, and the dew point maths.
Instruments and typical accuracy
| Instrument | Typical range | Typical accuracy | Where it is used |
|---|---|---|---|
| Capacitive polymer sensor | 0 to 100 %RH | ±1 to ±3 %RH | Transmitters, data loggers, HVAC, cleanrooms |
| Resistive sensor | 10 to 95 %RH | ±2 to ±5 %RH | Low cost loggers and older instruments |
| Chilled-mirror hygrometer | Dew point about -60 to +90 °C | ±0.1 to ±0.2 °C dew point | Reference in calibration labs |
| Psychrometer (wet and dry bulb) | About 5 to 100 %RH | ±2 to ±5 %RH, needs steady airflow | Field checks and weather work |
| Dew point transmitter | Down to -100 °C dew point | ±1 to ±3 °C dew point | Compressed air and gas dryers |
| Mechanical (hair) hygrometer | 20 to 90 %RH | ±5 %RH or worse | Indicators and museums |
Capacitive sensors dominate, with chilled-mirror units as the working reference. The accuracy figures are typical ranges to help you compare instruments. Always use the specification of the instrument in front of you.
Saturated salt fixed points at 25 °C
| Salt | Relative humidity | Notes |
|---|---|---|
| Lithium chloride (LiCl) | 11.3 %RH | Dry end, very slow to equilibrate |
| Magnesium chloride (MgCl₂) | 32.8 %RH | |
| Magnesium nitrate (Mg(NO₃)₂) | 52.9 %RH | Mid range, close to room humidity |
| Sodium chloride (NaCl) | 75.3 %RH | Most common check point |
| Potassium chloride (KCl) | 84.3 %RH | |
| Potassium sulfate (K₂SO₄) | 97.3 %RH | Prone to condensation errors |
Values from Greenspan, as used in OIML R 121 and ASTM E104. They shift with temperature, so keep the container at 25 °C or use the value for your temperature.
Reference standards and traceability
- Chilled-mirror hygrometer. Measures dew point directly from the mirror temperature at which condensation forms, so it traces back to temperature rather than to a humidity scale. Used as the reference in most labs.
- Humidity generators. National labs realise humidity with two-pressure, two-temperature and divided-flow generators, and calibrate chilled-mirror references against them.
- Saturated salt solutions. A sealed container with a saturated salt gives a known RH at a set temperature: about 11.3 %RH (LiCl), 32.8 (MgCl₂), 52.9 (Mg(NO₃)₂), 75.3 (NaCl), 84.3 (KCl) and 97.3 (K₂SO₄) at 25 °C. See OIML R 121 and ASTM E104. Cheap, but slow and sensitive to temperature.
- Quality system. ISO/IEC 17025 for the lab, with a stated uncertainty and a calibration chain to the national humidity standard.
How it is calibrated
Typical procedure
- Hold the chamber or container steady in temperature, within ±0.2 °C, and record the temperature at the sensor.
- Place the reference and the sensors as close together as possible, in the same airflow, and let everything stabilise (15 to 60 minutes per point, hours for salt solutions).
- Calibrate at several points, commonly near 10 to 20, 35, 50, 75 and 90 %RH, at one temperature such as 23 °C.
- Record reference and unit-under-test readings together, with temperature, after a stable period.
- Repeat the points in descending order to see hysteresis.
- Calculate the error at each point, estimate the uncertainty, and adjust or report as-found and as-left values.
Where it goes wrong
- Temperature gradients. At 50 %RH and 25 °C, a 0.5 °C difference between sensor and reference is about 1.5 %RH of error.
- Not enough settling time. Polymer sensors absorb water slowly, especially after a dry or wet excursion.
- Contaminated salts or low liquid volume. A solution that is not truly saturated gives the wrong RH.
- Condensation near 100 %RH. Cold spots wet the sensor and ruin the point.
- Pressure. Dew point changes with pressure, so state the pressure when comparing dew points.
What goes in the uncertainty budget
- Reference hygrometer calibration uncertainty and drift.
- Temperature measurement at the sensor and in the chamber, and the temperature uniformity (a large share of the RH uncertainty).
- Humidity stability and uniformity of the chamber or container.
- Resolution, repeatability and hysteresis of the unit under test.
Combine the contributions as in the uncertainty budget calculator, then decide pass or fail with a guard band as explained on the calculators page.
Calculators
Everything runs in your browser. Nothing you type is sent anywhere.