The effects of swirling flow on the measurement accuracy of a 152 mm (6 inch) orifice meter are presented in this paper. The test was conducted in a low-pressure air flow calibration facility with swirls generated by an axial vane-type swirler. The flow fields inside the orifice meter were characterized by a multiport pitot-static probe traversing across the meter run. The orifice meter performance was compared with reference critical flow nozzles.
The salient features of the observed flow fields compared quite closely with swirling flows generated by two close, out-of-plane elbows reported in the literature. The axial velocity profile was shown to be generally flatter than the fully-developed one. The swirl angle measurements delineated a secondary flow pattern of solid-body rotation. Without a flow conditioner in the line, the orifice meter was found to under-measure the true flow rate in a swirling flow. The deviation in flow measurements increased with increasing swirl angle. The tube bundle flow conditioner was very effective in removing swirls in the flow but not rectifying the deficient axial velocity profile. Based on the collected swirling flow data, a method was proposed to estimate possible swirl-induced flow rate errors for orifice meters in field use.
Measuring fluid flow rates by orifice meters is a well-established practice. The API 2530 standard, also known as AGA 3, generally practice. The API 2530 standard, also known as AGA 3, generally governs the installation and computational procedures for orifice meters used in natural gas operations. The standard assumes the flow condition at the orifice meter location is ideal, i.e., the velocity profile is fully developed and free from distortion and swirl. However, profile is fully developed and free from distortion and swirl. However, many devices in a field piping system can cause the flow condition to be non-ideal and thus influence the measurement accuracy of orifice meters. For example, two close out-of-plane elbows are known to produce swirl in addition to an asymmetric flow profile.
To condition the flow, AGA 3 prescribes an entry length of straight pipe that must precede the orifice meter. This length varies from 6 to 45 pipe diameters depending on the severity of upstream flow disturbances and the beta ratio, Beta, of the orifice meter. AGA 3 also allows straightening vanes of the 19-tube bundle design to be installed in the meter run with correspondingly shortened straight pipe entry lengths. However, recent results have suggested inadequacy of the entry lengths specified in AGA 3. Thus, a swirling flow could be present at an orifice meter even if the meter run conforms to the AGA 3 specification.
Over the years, the detrimental effects of swirling flow on orifice metering accuracy have been quantified by different investigators under varied operating conditions. McManus et al. reported large orifice meter under-measurements at Beta = 0.5 in a highly swirled flow exiting from a heat exchanger and an elbow. Blake et al. used a swirler to generate swirls and reported the resultant metering errors at three beta ratios. Brennan et al. generated swirls by elbows attached to a plate that was held between pipe flanges and reported 1.5% to 8% increases in the orifice meter discharge coefficient (i.e., resulting in flow rate under-measurement). There were also studies based on swirling flow generated by two out-of-plane elbows. Mottram and Rawat reported discharge coefficient increases by approximately 0.75% to 1% at mid-range beta ratios. Similar orifice discharge coefficient increases have also been reported for selected beta ratios by Sindt et al. and Mattingly and Yeh.