A Comparison Between Orifice and Flow Nozzle Laboratory Data and Published Coefficients

Abstract

Prediction of orifice and flow nozzle coefficients from their dimensions is a matter of major concern. In the mid 1930’s, laboratory flow coefficient data were obtained for flange top orifices and for the ASME flow nozzle. The working equations and tolerances developed from these data provide the basis for AGA Report Nos. (2,3), Fluid Meters, and ISO R541. Recent statistical evaluation of these data indicated that the tolerance should be at least twice the published values. This paper presents a considerable body of previously unpublished precision flow calibration data on orifices and flow nozzles. The significant difference between these data and the original 1980 data is that they were obtained on devices not constructed for a controlled laboratory test program. The majority of these devices were not manufactured by Foxboro but were purchased by the user from different manufacturers and were flow calibrated at Foxboro for high accuracy measurement applications. Data plate dimensions were used in all calculations, and conformity to ASME code requirements was the responsibility of the manufacturer. These data are compared point-by-point to the ASME and other proposed equations. These results indicate that a tolerance of ±0.7 percent for nozzles and ±0.5 percent or less for orifice plates is possible using a regression analysis equation. This compares to the ±2.0 percent and ±1 percent tolerance on previously published data. These data also point up the need for further work to bring the coefficient equations into line for pipe sizes less than 4 in. and for pipe sizes 24 in. and larger, as well as for beta ratios greater than 0.7.