Coriolis meter

Abstract

In accordance with example embodiments of the present disclosure, a method for determining parameters for, and application of, models that correct for the effects of fluid inhomogeneity and compressibility on the ability of Coriolis meters to accurately measure the mass flow and/or density of a process fluid on a continuous basis is disclosed. Example embodiments mitigate the effect of multiphase fluid conditions on a Coriolis meter.

Claims

1. A Coriolis flowmeter comprising: at least one flow tube configured to convey a process-fluid there through; a drive system configured to vibrate the at least one flow tube at a first natural frequency and a second natural frequency; electronics configured to determine a first measured process fluid density using the first natural frequency and a second measured process fluid density using the second natural frequency; a plurality of sensors positioned proximate the at least one flow tube configured to measure a speed of sound of the process fluid; an error model configured to use the speed of sound and the first measured process fluid density and the second measured process fluid density to quantify at least one effect of decoupling on the first measured process fluid density and to quantify at least one effect of decoupling on the second measured process fluid density; the electronics configured to determine a corrected density of the process fluid in real time; and a reporting device to report the corrected density of the process fluid.

2. The Coriolis flowmeter of claim 1 wherein the electronics are further configured to determine a liquid phase density from the corrected process fluid density.

3. The Coriolis flowmeter of claim 2 wherein the electronics are further configured to: determine a measured mass flow of the process fluid from at least one of the first natural frequency and the second natural frequency; determine a mass flow error using at least one of the at least one effect of decoupling on the first measured process fluid density and the at least one effect of decoupling on the second measured process fluid density; determine a corrected mass flow of the process fluid using the mass flow error and the measured mass flow of the process fluid; and the reporting device further configured to report the corrected mass flow of the process fluid.

4. The Coriolis flowmeter of claim 3 wherein the at least one effect of decoupling on the first measured process fluid density and the at least one effect of decoupling on the second measured process fluid density comprises a mass decoupling parameter.

5. The Coriolis flowmeter of claim 4 wherein said array of sensors responsive to pressure variations are strain-based sensors.

6. The Coriolis flowmeter of claim 1 wherein the electronics are further configured to determine a gas void fraction of the process fluid using the speed of sound of the process fluid.

7. The Coriolis flowmeter of claim 1 wherein the at least one effect of decoupling on the first measured process fluid density and the at least one effect of decoupling on the second measured process fluid density comprises a density decoupling parameter.

8. The Coriolis flowmeter of claim 1 wherein the system for measuring the sound speed of said process-fluid further comprises: at least one strain based sensor engaged with at least one conduit; and said at least one strain based sensor is electronically coupled with a central processor.

9. The Coriolis flowmeter of claim 1 wherein the system for measuring the vibration characteristics of said at least one conduit further comprises: at least one pick-off coil responsive to the vibration at least one conduit; and said at least one pick-off coil is electronically coupled with a central processor.

10. The Coriolis flowmeter of claim 1 wherein a central processor interprets said sound speed and vibrational characteristics of said at least one conduit vibrating at, at least two vibration frequencies, to provide a measurement of the process-fluid density.

11. The Coriolis flowmeter of claim 1 wherein a central processor interprets said sound speed and vibrational characteristics of said at least one conduit vibrating at, at least two vibration frequencies, to provide a measurement of the process-fluid mass flow.

12. The Coriolis flowmeter of claim 1 wherein said system that measures process-fluid sound speed is an array of sensors responsive the pressure variations within the process-fluid deployed on a conduit other than said at least one conduit.

13. The Coriolis flowmeter of claim 1 wherein said system that measures process-fluid sound speed determines a measure of gas void fraction of the process-fluid; and said system determines a reduced vibration frequency of more than one of the vibration frequencies of said at least one conduit.

14. The Coriolis flowmeter of claim 1 wherein said system that measures process-fluid sound speed determines a more than one reduced frequency of vibration of said at least one conduit.

15. A method for optimizing a process parameter of the Coriolis meter of claim 1 comprising: vibrating said at least one conduit at two or more frequencies; and said two or more frequencies being low or known reduced frequencies; and providing homogeneous flows through said at least one conduit; and measuring said process parameter at said two or more frequencies; and calibrating said Coriolis meter to operate on the effects of process-fluid variability; and measuring a process fluid sound speed; and calibrating said measurement of a process parameter interpreted by said sound speed, wherein an optimized process parameter is determined.

16. The method of claim 15 wherein: said process parameter is the density of said process-fluid.

17. The method of claim 15 wherein: said process parameter is the mass flow of said homogeneous flow.

18. The method of claim 15 wherein: the effects of process-fluid variability is fluid inhomogeneity and/or changes in fluid compressibility.

19. The method of claim 15 wherein calibrating said Coriolis meter to measure the effects of process-fluid variability further comprises the steps of: correcting measured process parameters; and minimizing the difference between corrected process parameters.

20. A flow metering system comprising: a process-fluid in one or more conduits; actuator(s) for vibrating said one or more conduits; a sensor engaged with said one or more conduits for determining a process-fluid gas void fraction and a reduced frequency; a model for interpreting process-fluid mass flow rate and/or density by the vibrational characteristics of said one or more conduits; a calibration of said vibrational characteristics that is applicable for homogeneous flows at a known or sufficiently low reduced frequency; a model to correct interpreted process-fluid mass flow rate and/or density based on said calibration, applicable for homogeneous flows at a known or sufficiently low reduced frequency; and wherein a plurality of correction terms for a plurality of effects of decoupling are determined substantially as a function of the process-fluid gas void fraction and the plurality of correction terms for a plurality of effects of compressibility are determined substantially as a function of the reduced frequency.

21. The system of claim 20, wherein said sensor is an array of strain-based sensors.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) To assist those of skill in the art in making and using the disclosed invention and associated methods, reference is made to the accompanying figures, wherein: Example figure descriptions follow:

(2) FIG. 1 is a schematic of an example Coriolis meter employing the measurement of the speed of sound through a process-fluid;

(3) FIG. 2 is an example embodiment showing a Coriolis meter having a plurality of strain based sensors and an exciter;

(4) FIG. 3A is a graph depicting mass flow and density error reading as a function of gas damping ratio and reduced frequency;

(5) FIG. 3B is a graph depicting mass flow and density error reading as a function of gas damping ratio and reduced frequency;

(6) FIG. 4A is a graph depicting dual frequency Coriolis optimization simulated with reduced order model;

(7) FIG. 4B is a graph depicting dual frequency Coriolis optimization simulated with reduced order model;

(8) FIG. 5 is a graph depicting optimization function over decoupling parameter using measured reduced frequency;

(9) FIG. 6 is a detailed cross section depicting decoupling parameters;

(10) FIG. 7 is a cross section depicting decoupling parameters

(11) FIG. 8 is an example Coriolis meter adapted to the embodiment;

(12) FIG. 9 is an example Coriolis meter adapted to the embodiment.

(13) FIG. 10 is a graph depicting reduced frequencies and the process-fluid sound speed plotted versus gas void fraction.

(14) FIG. 11 is a graph that shows the results of the optimization based on equating measured densities at two frequencies.

(15) FIG. 12 is a graph that shows densities normalized by the input liquid density versus gas void fraction.

(16) FIG. 13 is a graph that shows the results of a range of the randomly selected input parameters with 2% random noise added to the speed of sound measurement.

(17) FIG. 14 is a graph that shows and example of the Error function based on equating densities measured at two frequencies as a function of trial decoupling parameters.

(18) FIG. 15 is a graph that shows mass flow rate randomly varying between 1.5 and 1.8 kg/sec.

(19) FIG. 16 is a graph that shows the results of optimization based on equating measured mass flows at two frequencies.

(20) FIG. 17 is a graph that shows results of optimization based on equating measured mass flows at two frequencies.

(21) FIG. 18 is a graph that shows an example of an Error function based on equating mass flows measured at two frequencies.

DESCRIPTION

(22) Referring to FIG. 1, an example embodiment is depicted in the illustration. An array of strain based sensors 130 are in fluid communication with flow tubes of a Coriolis meter. The strain based sensors are used to calculate the speed of sound propagated through the process fluid 134. The flow tubes in the Coriolis meter 132 vibrate at two or more frequencies. The Coriolis meter electronics interpret vibrational characteristics in terms of apparent mass flow and apparent density 136. The calculated speed of sound and measured process-fluid pressure are used to determine a reduced frequency for each vibrational frequency and a gas void fraction of the process fluid 137. The ?.sub.apparent and {dot over (m)}.sub.apparent are combined with the speed of sound and reduced frequencies and are then sent to an algorithm that optimizes decoupling and/or compressibility parameters to minimize error between mass flows and/or densities measured at different frequencies 138. The optimized parameters are used to improve measured mass flow and/or density of the process fluid 140.

(23) Referring to FIG. 2, an example Coriolis meter 100 is shown with a number of strain based sensors 116 arrayed along one of a pair of flow tubes 110/112. Two pick-off coils (114) are shown to measure the natural frequency of the vibrations and twist of the vibrating flow tube. In other words, the pick-off coils responsive to the vibration and twist of a vibrating flow tube. An exciter 118 is supported by the framework surrounding the Coriolis meter.

(24) Referring to FIGS. 3A, 3B errors interpreting the mass flow and density errors for a Coriolis operating on a multiphase flow with an interpretation method applicable to homogeneous fluids operating at low reduced frequencies is depicted in the paired graphs. The errors were predicted using a simplified aeroelastic model of a Coriolis meter (ref Gysling) over a range of gas bubble damping parameters (decoupling parameter) and reduced frequency parameter (compressibility parameter). These predictions for the errors due to decoupling and compressibility were made based on a model that utilizes the speed of sound to calculate gas void fractions and reduced frequencies. Since gas void fraction and reduced frequency are strongly linked through the process fluid speed of sound, this formulation utilizes reduced frequency as the variable that captures both decoupling and compressibility effects within the model predictions.

(25) One skilled in the art understands that any empirical or computational model that characterizes the relationship between the measured vibrational characteristics of the fluid-conveying flow tube, i.e. tube phase shift and tube natural frequency, and the multiphase flow properties within the meters could be used in similar manner.

(26) In this example the reduced order model of Gysling was used to calculate the apparent mass flow and density measured by a dual frequency Coriolis meter operating on a bubbly mixture. The first in-vacuum bending frequency of the tube was set to 300 Hz, and the second was set at 1100 Hz. The tube diameter was 2 inches. The simulated operating conditions for the process fluid for this test case was bubbly mixture of air and water at ambient pressure with 2% gas void fraction. The actual mass flow through the meter was set at 4.0 kg/sec and the liquid density was set at 1000 kg/m{circumflex over ()}3. The reduced frequency of tube 1 is 0.57 and tube 2 is 2.09. The gas damping ratio, termed the decoupling parameter in the model, was set to 0.5 for both frequencies. The apparent mass flow and mixture density for tube 1 was 4.44 kg/sec and 1038 kg/m{circumflex over ()}3, and for tube 2, 14.18 kg/sec and 1927 kg/m{circumflex over ()}3.

(27) Referring to the aforementioned equation:
error??.sub.{dot over (m)}(({dot over (m)}.sub.f1.sub.trial?{dot over (m)}.sub.f2.sub.trial)/({dot over (m)}.sub.f1.sub.trial+{dot over (m)}.sub.f2.sub.trial)).sup.2+?.sub.?((?.sub.f1.sub.trial??.sub.f2.sub.trial)/(?.sub.f1.sub.trial+?.sub.f2.sub.trial)).sup.2

(28) The trial mass flows and densities are formed by correcting the measured, or apparent, mass flows and densities to actual mass flows and density using the over reading function shown as a surface in FIGS. 3A, 3B. for a given set of trial decoupling and compressibility parameters. An error is formed based on the trial mass flows and trial densities associated with the measured, or apparent, mass flows and densities at the two frequencies. The error is minimized when the corrected mass flows and corrected densities predicted at the two frequencies match, respectively for a given set of trial decoupling and compressibility parameters. To determine the parameters of the multiphase flow flowing through the meter, the error function is then minimized over a suitably wide range of decoupling (zsi.sub.gas) and compressibility parameter (f.sub.red) Once the parameters of the model are adjusted such that the error function is minimized, the optimized mass flow and mixture density are determined.

(29) Referring to FIGS. 4A, 4B the error function, for the example described in FIGS. 3A 3B versus the decoupling parameter and reduced frequency of the first tube frequency is described in the paired graphs. For this example, the weighting of the error contributions for the mass flow errors and the density errors are set to unity. The left figure shows the general surface shape, and the right shows the surface viewed from above with the color axis limited to highlight the existence of multiple solutions, for example, combinations of decoupling parameter, zsi_gas, and reduced frequency for which the error function approaches zero As shown, if the error function is evaluated over a range of reduced frequencies and coupling parameters, the optimization would be confounded, and unable to determine either the best reduced frequency or the best decoupling parameter, and therefore unable to report a unique mass flow or density.

(30) FIGS. 4A, 4B also shows the same optimization function using the measured process fluid sound speed and therefore having a known reduced frequency for each vibrational mode evaluated over a range of the decoupling parameters. As shown, this additionally-constrained optimization yields a unique solution for the multiphase parameters in this albeit simplified, yet representative, example, thereby enabling the meter to report a more accurate and robust measurement of the mass flow and density of the two phase mixture based on an optimization process equating the mass flow rates measured at two frequencies and/or the densities measured at two frequencies.

(31) Referring to FIG. 5 this self-consistency example, the decoupling parameter of 0.5 is identified by the optimization. Using this identified decoupling parameter, and the apparent mass flow and density at either tube frequency, enables meter to report accurate mass flow and mixture density. One skilled in the art understands that Coriolis based density measurements in multiphase flows may be more robust and repeatable than the Coriolis-based mass flow measurement under the same conditions. In these cases, it may be beneficial to increase the weighting of the density measurement error contribution in the error function compared to the weighting of the mass flow error contribution.

(32) Referring to FIG. 6 a horizontal cross section of two flow-tubes 110/112 is depicted in the illustration. The illustration demonstrates decoupling inhomogeneous flow. Flow tube 110 has a homogeneous flow 160 without entrained gas. The center of mass 161 is in the center of the flow-tube 110. The flow-tube 112 has an inhomogeneous flow 162 with entrained gas that is not homogeneous. The center of mass 163 is not in the center of the flow-tube.

(33) Referring to FIG. 7, a vertical cross section of a flow-tube is depicted in the illustration. An inhomogeneous gas-entrained flow 164 has a varying density of entrained gas about the flow path. Decoupling of an inhomogeneous fluid is said to occur within a fluid when one phase of a fluid vibrates differently than another phase. One skilled in the art understands that the effects of decoupling are determined substantially as a function of the measured gas void fraction.

(34) Referring to FIG. 8 an example Coriolis meter 200 is shown with a number of strain based sensors 216 arrayed along one of a pair of flow tubes 210/212. One skilled in the art understands that the example may include pick-off coils and an exciter as necessary to generate and measure the natural frequency of the vibrations and twist of the vibrating flow tube may also be installed on the example Coriolis meter.

(35) Referring to FIG. 9 an example Coriolis meter 300 is shown with a number of strain based sensors 316 arrayed along a flow tube 312. One skilled in the art understands that the example may include pick-off coils and an exciter as necessary to generate and measure the natural frequency of the vibrations and twist of the vibrating flow tube may also be installed on the example Coriolis meter.

(36) While example embodiments have been described herein, it is expressly noted that these embodiments should not be construed as limiting, but rather that additions and modifications to what is expressly described herein also are included within the scope of the invention. Moreover, it is to be understood that the features of the various embodiments described herein are not mutually exclusive and can exist in various combinations and permutations, even if such combinations or permutations are not made express herein, without departing from the spirit and scope of the invention.