Calibration Uncertainty of a Bell Prover

INRIM owns a bell piston prover for small/medium gas flow rates (BELLGAS) which provides flow rates up to about 100 L/min with uncertainties of the order of 0.12% due to several construction improvements which include the accurate balancing of the counterweights system, the reduction of the movement friction and the accurate measurement of the bell movement through a high-resolution encoder. The most important uncertainty contributions in the overall uncertainty estimate for this test rig are the associated to the volume variation measurement and to the dead volume (Par. 2.1). The BELLGAS dimensions were initially measured at the then-IMGC using the strapping method, which provided encouraging results, and later with the same method described here, whose results were compatible to the ones obtained with the other method. The procedure used during this first bottling calibration, although accurate, was quite cumbersome due to the fact that the oil vessel used was quite small and required to be filled twice for a complete calibration; a larger vessel was then built in order to simplify the procedure and avoid possible errors. In the present paper, the complete calibration procedure is described, including the secondary corrections performed on the raw data; results are presented together with the uncertainty budget of the calibration and of the test rig as a whole, describing therefore in detail the derivation process of the CMC claimed for this test rig.