Within the scope of Euromet Project 691, an intercomparison with a somewhat untypical character was started in 2003 and completed in March 2004. No specifications were made as regards the calibration procedure, each state institute could determine the volume of a 5 l volume standard with the measurement methods generally applied at the institute. This allowed the volume flown out of or filled into a dry or moist volume standard to be determined. A total of 18 countries participated in the intercomparison, the Swedish National Testing and Research Institute SP in Borås was responsible for coordination. Six of the participants had already investigated the same volume standard in 1989. Figure 1 shows the 5 l glass volume standard.
It was thus not the relation of the volume values determined at the different institutes to a value averaged over all values which stood in the fore, but rather the interest in the long-term stability of the volume standard, the reproducibility of previous results, the quality/controllability of the procedures used and the credibility of the entries in the CMC Tables.
Compared with the values obtained in 1989, the institutes determined in 2003 a slightly larger volume value (on average larger by 0.002%) which approximately corresponds to the volume of one drop of water.
Compared with 1989, improved measurement methods were applied. On average, the standard deviation of the measurements performed at the individual state institutes was reduced by 33 %. Moreover, there was an approximation of the measurement results furnished by the institutes involved in the comparisons 1989/2003, and the standard deviation of the six volume values was reduced by 40 %.
Compared with the results of other institutes which applied comparable methods of measurement, the PTB results are outstandingly good. On average, the relative standard deviation of all these state institutes was 2.3·10-5, that of PTB 0.7·10-5. The relative range (Max - Min) of the measurement values was 6.8·10-5, that of PTB 2.0·10-5. And the relative measurement uncertainty amounted on average to 8.2·10-5, in the case of PTB to 3.2·10-5.
In the case of almost all participants, the uncertainty values obtained within the scope of the intercomparison agreed with the entries in the CMC Tables. In a few institutes it will in future even be possible to perform a correction towards smaller values.
Summarizing it can be said that the intercomparison of the Euromet 691 Project was successfully performed, co-ordinated and analyzed.
In addition it is to be mentioned that the PTB test facility moved into the rooms of the hydrodynamic test field only last year and that this intercomparison offered the first possibility after this removal of checking the metrological performance and the measurement uncertainty calculated for the new conditions at the international level.
