The consistency of pressure effects between three identical Coriolis flow meters

Abstract

Coriolis flow meters are one of the most popular flow measurement technologies in the world today for high accuracy measurement of single-phase liquids, gases and even slurries. They are capable of measuring both mass and density directly and can also infer the volume flow. They can be installed in challenging process environments and have been successfully deployed with non-Newtonian fluids, high viscosity fluids, pulsating flows and even at extreme temperatures and pressures.However, it is known that operating most Coriolis flow meters at a pressure which differs from the original calibration pressure requires compensation else significant measurement errors will occur. Pressure compensation coefficients appear to vary by manufacturer, meter geometry and sensor material. Furthermore, the manufacturer published pressure compensation coefficients are not fully traceable. To date, there has not been sufficient research exploring the consistency of the pressure compensation for identical Coriolis flow meters.This paper presents the findings of a research conducted at the TÜV SÜD National Engineering Laboratory (NEL) Elevated Pressure and Temperature (EPAT) oil flow facility to investigate the pressure effect uniformity for matching Coriolis devices. The first stage of the experimental programme calibrated three identical DN80 Coriolis flow meters at a range of pressures with no pressure compensation applied. A pressure compensation coefficient was then derived from the data and the Coriolis meters were then calibrated at two alternative pressures to ascertain the robustness of the coefficients and whether the compensation could be extrapolated successfully.From the experimental results, it can be concluded that the pressure effect for the three DN80 Coriolis flow meters was extremely repeatable and consistent with a discrepancy of less than 0.025% between the devices at 80 bar. Whilst the mass flow was significantly affected by fluid pressure, the fluid density did not appear to be influenced. The pressure corrected results were also well within the manufacturer specification of ±0.1%.