Logo of the Physikalisch-Technische Bundesanstalt

Humidity and Thermal State Behaviour

Working Group 3.41

Validation and traceability of gas humidity measurements

The national humidity scale is realized by several national standards. The range to be realized – from extreme trace humidity (volume concentration of water smaller than 20 nl/l) up to high humidity (volume fraction of water vapor clearly larger than 60 %) – cannot be covered by means of a single humidity generator, and the humidity generators used for this purpose work according to different principles. Also, different principles are used depending on the scope of application of the generator (whether they are portable or stationary). In the following, the different humidity generators used at PTB to realize the national gas humidity scale will be briefly presented.

  • The national standard for trace humidity is a coulometric trace humidity generator. It is based on the following sub-processes:

    • Generating a reference gas flow using UHP nitrogen (residual humidity concentration: approx. 0.5 nl/l volume fraction of water vapor),
    • Generating a defined quantity of hydrogen and oxygen in a partial gas flow of the reference gas by means of water electrolysis according to Faraday's law,
    • Subsequent drying of the electrolysis partial gas flow (that now consists of nitrogen with trace contents of hydrogen and oxygen) and then re-combining the hydrogen and the oxygen to water by means of catalysis,
    • Subsequent drying of the electrolysis partial gas flow (that now consists of nitrogen with trace contents of hydrogen and oxygen) and then re-combining the hydrogen and the oxygen to water by means of catalysis.

    In this way, the water vapor content of the generator gas is determined fundamentally by the electrolysis flow and the nitrogen flow, thus making it traceable to the SI units kilogram, second, ampere and mole.

  • In the medium and high humidity range, a two-pressure humidity generator is used where the generated humidity depends on the equilibration on the basis of the vapor pressure equation of water. The basis used here is filtered compressed air that is free of grease and dust, in the pressure range up to 15 bar. This filtered air flows through a humidifier containing water at a defined temperature and is thus saturated according to the vapor pressure equation to reach a state close to saturation at the given temperature. After that, the air is fed into a saturator. This is a sophistically designed metallic block with a large surface inside. The metallic block is kept at an exactly defined temperature in a high-precision thermostat. This temperature is lower by several degrees than that of the humidifier. The air thus cools down to the temperature of the saturator at the large inner surface, so that excess water condensates. The air at the outlet of the saturator is thus fully saturated, related to the temperature of the saturator. From the vapor pressure equation according to Sonntag [2], the saturated vapor pressure is calculated in the pure phase. By multiplying by the enhancement factor according to Greenspan [3], one obtains the saturated vapor pressure of the humid air. At a valve, the air is now reduced down to ambient pressure. In doing this, the partial pressure decreases in accordance with the quotient from the ambient pressure and the saturator pressure. Provided the vapor pressure equation is valid, it is thus possible to trace the partial vapor pressure – and thus the absolute humidity – to the measurement of the saturator temperature and of the saturator and ambient pressure. By varying the parameters saturator temperature and saturator pressure, it is thus possible to set a wide range of gas humidity.

  • Another type of generator is the permeation tube. In this generator, permeation of water molecules through certain synthetic materials, which only depends on the temperature and on a material constant, is exploited to generate a defined water content in the gas flow.

  • In addition, there is another type of humidity generator: the so-called gas mixture generator. Here, a low-humidity gas is generated from a gas with known humidity by admixing of dry gas with a known volume rate (often over several steps).


PTB's national standards for gas humidity are compared at regular intervals with those of other national institutes by means of transfer standards – at the European level within the scope of EURAMET and at the international level within the scope of key comparisons organized by the CCT of the BIPM. This ensures mutual equivalence of the humidity scales within the limits of the assigned measurement uncertainty.


In the following, a selection of important key comparisons is listed in which PTB took part. You will find further information as well as final reports on the individual comparisons in the Opens external link in new windowKCDB of the BIPM:

  • EURAMET.T-K6 (EURAMET P621): Comparison of the realizations of local dew/frost-point temperature scales in the range -50 °C to +20 °C
  • EURAMET.T-K8 (EURAMET P717): Comparison in dew-point temperature (high range)
  • CCT-K8: Comparison of realizations of local scales of dew-point temperature of humid gas - Dew-point Temperature: 30 °C to 95 °C
  • EURAMET P1002: International comparability in measurements of trace water vapour
  • EURAMET P1061: Comparison of air temperature calibrations

The humid gas flow supplied by the standard generators is directly used for the traceability of instruments measuring absolute humidity (dew point hygrometers, spectrometers, etc.). Sensors providing relative values are kept at the required measuring temperature in a special temperature-controlled chamber through which the reference value of relative humidity – the saturation degree at exactly this temperature – is defined. An overview of the different types of hygrometers and their functional principle can be found here Opens internal link in new windowHygrometers and their functional principle.

[1] Mackrodt, P., Int. J. Thermophys. 2012, 33: 1520-1535. DOI 10.1007/s10765-012-1348-0

[2] Sonntag, D., Zeitschrift für Meteorologie. 1990, 40 (5) (1990), 340-344.

[3] Greenspan, L., J. Res. Natl. Bur. Stand., Sect. A. 1976, No. 1 (Jan-Feb), 41-44.