Series of calibration tests at national standard loops and industrial application experiences of new type flow-metering system by ultrasonic pulse-doppler profile-velocimetry

Highly-accurate measurements are required to control the thermal outputs for reactor feedwater flow rates. Calibration tests for flowmeters such as time-of-flight (TOF) ultrasonic flow meters are conducted to assure their accuracy in measurements at standard loops, where flow profile factors, which adjust measurand to real flow rates, are determined usually by conducting calibration tests under ambient flow conditions. As nuclear power plants are aging, readings of flowmeters for reactor feedwater systems drift due to the changes of flow profiles. The causes of those deviations are affected by the change of wall roughness of inner surface of pipings, and those changes of flow patterns lead to large errors in measurements. Therefore, it is still in suspicion if its high accuracy on adjusting measurand by profile factors is traceable against the national standard, and moreover, we have to consider the difference of flow configuration between the flow standard and measurement position on-site. The difference of facilities for measurement conditions can be considered as the concept of 'Facility Factor' and the 'Profile Factor' is not sufficient for correcting the meter reading. It is so discussed that measurement of velocity profile on-site is essential. We proposed the new type of flowmeter called UdFlow/UDF, ultrasonic pulse-Doppler flowmeter, which can measure instantaneously-determined flow-velocity profiles and eliminate the effect of deviated flow profile from expected ideal one. Series of calibration tests of UdFlow were conducted at the national standard loops in Mexico, CENAM (The Centro National de Metrologia) and in USA, NIST (National Institute of Standard and Technology) in order to evaluate the accuracy of the flowmeter. Four ultrasonic transducers are mounted on the surface of stainless steel piping circumferentially with the diameters of 100mm and 200mm to measure four velocity profiles. The calibration tests found a deviation better than 0.3% between the two devices in terms of the average of the values recorded by six rounds of each measurement. From the results of measurement conducted with Reynolds number varied, it was found that the overall average deviation between the two devices was better than 0.3%. (author)