The effect of water contamination on the dew-point temperature scale realization with humidity generators

The purpose of this paper is to study the effect of contaminated water in the context of humidity generators. Investigation of different methods to determine the drop in dew-point temperature due to contamination and experiments on actual contamination rates are reported. Different methods for calculating the dew-point temperature effect from electrical conductivity and density measurements are studied with high-purity water and aqueous solutions of NaCl and LiCl. The outcomes of the calculation methods are compared with the results of direct humidity measurements. The results show that the often applied Raoult's law based calculation method is in good agreement with other methods. For studying actual contamination, water samples were kept in glass, plastic, copper and stainless-steel vessels for up to 13 months to investigate natural ionic and organic contamination in vessels with different wall materials. The amount of ionic contamination was found to be higher in copper and glass vessels than in stainless-steel and plastic vessels. The amount of organic contamination was found to be highest in the plastic vessel. In all the cases, however, the corresponding drop in dew-point temperature due to natural contamination was found to be below 0.1 mK. The largest rate of change of dew-point temperature was 26 µK/month. Thus, if proper cleanness is maintained in a humidity generator the effect of contamination of water in the saturator is insignificant compared with the major uncertainty components even in the most accurate generators today.