(19)
(11)EP 2 561 320 B1

(12)EUROPEAN PATENT SPECIFICATION

(45)Mention of the grant of the patent:
10.06.2020 Bulletin 2020/24

(21)Application number: 11720154.1

(22)Date of filing:  18.04.2011
(51)International Patent Classification (IPC): 
G01F 1/712(2006.01)
G01F 1/60(2006.01)
G01F 1/58(2006.01)
G01F 1/74(2006.01)
(86)International application number:
PCT/GB2011/000600
(87)International publication number:
WO 2011/128656 (20.10.2011 Gazette  2011/42)

(54)

MEANS AND METHOD FOR MONITORING THE FLOW OF FLUID

VORRICHTUNG UND VERFAHREN ZUR ÜBERWACHUNG VON FLÜSSIGKEITSSTRÖMUNGEN

MOYEN ET PROCÉDÉ DE SURVEILLANCE DU DÉBIT D'UN FLUIDE


(84)Designated Contracting States:
AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

(30)Priority: 17.04.2010 GB 201006409

(43)Date of publication of application:
27.02.2013 Bulletin 2013/09

(73)Proprietor: University of Huddersfield
Queensgate, Huddersfield HD1 3DH (GB)

(72)Inventors:
  • LUCAS, Gary, P.
    Huddersfield HD1 3DH (GB)
  • LEEUNGCULSATIEN, Teerachai
    Huddersfield HD1 3DH (GB)

(74)Representative: HGF Limited 
4th Floor Merchant Exchange 17-19 Whitworth Street West
Manchester M1 5WG
Manchester M1 5WG (GB)


(56)References cited: : 
US-A- 5 693 891
  
  • Teerachai Leeungculsatien: "Novel Multi-Electrode Electromagnetic Flow Meter", Research Festival, 2 April 2009 (2009-04-02), pages 1-1, XP55002449, Retrieved from the Internet: URL:http://eprints.hud.ac.uk/4635/1/Novel_ Multi%2DElectrode_Electromagnetic_Flow_Met er_%2D_Teerachai_Leeungculsatien.pdf [retrieved on 2011-07-08]
  • T. KATOH ET AL: "3D flow tomography with electromagnetic flowmeter", PROCEEDINGS OF THE 41ST SICE ANNUAL CONFERENCE. SICE 2002., vol. 1, 1 January 2002 (2002-01-01), pages 535-538, XP55002878, DOI: 10.1109/SICE.2002.1195462 ISBN: 978-0-78-037631-1
  • WANG ET AL: "Relationship between velocity profile and distribution of induced potential for an electromagnetic flow meter", FLOW MEASUREMENT AND INSTRUMENTATION, BUTTERWORTH-HEINEMANN, OXFORD, GB, vol. 18, no. 2, 9 May 2007 (2007-05-09), pages 99-105, XP022066395, ISSN: 0955-5986, DOI: DOI:10.1016/J.FLOWMEASINST.2006.03.001
  • RAZZAK ET AL: "Electrical resistance tomography for flow characterization of a gas-liquid-solid three-phase circulating fluidized bed", CHEMICAL ENGINEERING SCIENCE, OXFORD, GB, vol. 62, no. 24, 22 November 2007 (2007-11-22), pages 7253-7263, XP022357542, ISSN: 0009-2509, DOI: DOI:10.1016/J.CES.2007.08.057
  • ISMAIL I ET AL: "Tomography for multi-phase flow measurement in the oil industry", FLOW MEASUREMENT AND INSTRUMENTATION, BUTTERWORTH-HEINEMANN, OXFORD, GB, vol. 16, no. 2-3, 1 April 2005 (2005-04-01), pages 145-155, XP027618955, ISSN: 0955-5986 [retrieved on 2005-04-01]
  • B Horner ET AL: "A multi-sensor induction flowmeter reducing errors due to non-axisymmetric flow profiles", Measurement Science and Technology, 1 March 1996 (1996-03-01), pages 354-360, XP055531853, DOI: 10.1088/0957-0233/7/3/016 Retrieved from the Internet: URL:http://iopscience.iop.org/article/10.1 088/0957-0233/7/3/016/meta [retrieved on 2018-12-07]
  
Note: Within nine months from the publication of the mention of the grant of the European patent, any person may give notice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in a written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art. 99(1) European Patent Convention).


Description

FIELD OF INVENTION



[0001] The invention relates to a means and method for monitoring the flow of a fluid. The invention particularly relates to a means and method that is suitable for monitoring the flow of a fluid having a non-uniform flow profile.

BACKGROUND TO THE INVENTION



[0002] Electromagnetic flow meters are used in a variety of industries to monitor the flow of conducting fluids. Electromagnetic flow meters utilise Faraday's law of electromagnetic induction to induce a voltage in the conducting fluid as it moves through a magnetic field. The flow rate of the conducting fluid is then derived from the measured induced voltage.

[0003] In a conventional electromagnetic flow meter the conducting fluid is directed to flow through a flow pipe, electromagnetic coils are located outside the flow pipe to create a magnetic field, two electrodes are mounted in the flow pipe wall to detect the induced voltage and processing means are configured to process the induced voltage data to determine the average flow rate. Although conventional electromagnetic flow meters are widely used it is recognised that they have a number of limitations. For example, conventional electromagnetic flow meters can only measure the average flow rate of a conducting fluid - they can not determine the axial velocity profile of a conducting fluid. Moreover, conventional electromagnetic flow meters are generally only effective when the conducting fluid has a uniform flow profile - they are unsuitable and/or inaccurate when the conducting fluid has a non-uniform flow profile.

[0004] Unfortunately, non-uniform flow conditions are often encountered. For example, a fluid may develop a non-uniform flow profile downstream of a pipe bend, at a partially open valve, in a blocked pipe and/or along an inclined pipe. A multiphase fluid may have a non-uniform flow profile if the component parts have different flow characteristics. The flow of a fluid is non-uniform when, for example, the fluid has a non-axisymmetric velocity profile.

[0005] One approach to accurate flow rate measurement of non-uniform single phase fluids has been proposed by HORNER in HORNER, B. (1998) A novel profile-insensitive multi-electrode induction flow meter suitable for industrial use. Meas. Sic. Technol., 24, 131-137. However, this type of flow meter does not provide information on the axial velocity profile of the fluid. This can be a major drawback, particularly in multiphase fluids where, for example, the volumetric flow rate of a particular phase can only be found by integrating the product of the local phase velocity and the local phase volume fraction in the flow cross section. As a result, the approach proposed by Horner can not be used to determine the flow rate of the conducting continuous phase of a multiphase fluid with a non-uniform flow profile.

[0006] Other known types of flow meters that are suitable for measuring the flow rate of a conducting phase of a multiphase fluid are constrained by high cost and the use of hazardous radioactive sources to monitor the flow.

[0007] Teerachai Leeungculsatien, "Novel Multi-Electrode Electromagnetic Flow Meter", Research Festival, 2009 discloses an electromagnetic flow meter using a uniform magnetic field in the pipe cross section. Induced voltages between N electrode pairs are measured to determine the liquid velocity in N distinct regions.

[0008] Wang et al, "Relationship between velocity profile and distribution of induced potential for an electromagnetic flow meter", Flow Measurement and Instrumentation, 2007 discloses an investigation into the relationship between the induced electric potential and the velocity distribution of the conductive continuous phase in two-phase flows in pipes to which an electromagnetic field is applied.

[0009] Horner et al, "A multi-sensor induction flowmeter reducing errors due to non-axisymmetric flow profiles", Measurement Science and Technology, 1 March 1996 discloses a method of reducing errors in profiles deviating from axial symmetry, using multiple electrode pairs and two magnetic field components.

SUMMARY OF THE INVENTION



[0010] The invention seeks to overcome or address the problems associated with the prior art as described above.

[0011] Embodiments of the invention are set forth in the appended claims.

[0012] The invention seeks to provide an electromagnetic flow meter and method that is suitable for monitoring the. flow of any suitable conducting fluid. The flow meter and method may be suitable for monitoring the flow of a conducting single phase fluid. The flow meter and method may be suitable for monitoring the flow of a conducting continuous phase of a multiphase fluid. The flow meter and method may also be suitable for monitoring the flow of the one or more dispersed phases of the multiphase fluid. For example, the flow meter and method may be suitable for monitoring the flow of extracted oil or gas mixtures, slurries, blood, nuclear waste or water. Moreover, the flow meter and method may be suitable for monitoring the flow of fluid in a number of different environments and technological applications such as in the oil, gas, medical, nuclear, chemical, food processing and mining industries.

[0013] The invention seeks to provide an electromagnetic flow meter and method that is suitable for monitoring the flow of a fluid with uniform or a non-uniform flow profile. For example, the flow meter and method may be suitable for monitoring the flow of a conducting single phase fluid with a non-uniform flow profile. The flow meter and method may be suitable for monitoring the flow of a conducting continuous phase of a multiphase flow with a non-uniform flow profile.

[0014] The invention seeks to provide an electromagnetic flow meter and method that are suitable for monitoring one or more flow characteristics of a conducting fluid. For example, the flow meter and method may be suitable for measuring the axial velocity profile of a conducting fluid (variation in axial velocity across the flow cross-section). Having derived the axial velocity profile, the flow meter and method may be suitable for determining the volumetric flow rate of the conducting fluid. The flow meter and method may also be suitable for measuring the axial velocity profile and subsequently the volumetric flow rates of each phase of a multiphase fluid. The flow meter and method may be suitable for measuring the velocity profile, and optionally the volumetric flow rate, in real time.

[0015] The invention seeks to provide an electromagnetic flow meter and method that is able to monitor the flow of a conducting fluid with a uniform or non-uniform flow profile more accurately than conventional measuring systems. For example, it has been found that the flow meter and method according to the present invention can measure the volumetric flow rate of a non-uniform conducting fluid with an error margin of approximately +/-0.5% in comparison to the error margin of approximately +/-3.5% of conventional flow meters.

[0016] The invention seeks to provide an electromagnetic flow meter and method for non-intrusively monitoring the flow of a conducting fluid.

[0017] The invention seeks to provide a low-cost electromagnetic flow meter and method for monitoring flow that is cheaper to manufacture and operate than conventional flow meters.

[0018] The invention seeks to provide an electromagnetic flow meter and method that does not require the use of a hazardous material for monitoring the flow of a fluid.

[0019] According to a first aspect of the present invention there is provided An electromagnetic flow meter for monitoring the flow of a conducting fluid comprising: a flow tube; a means for generating a magnetic field across the flow tube cross-section so that a voltage is induced in the conducting fluid as it flows through the flow tube; an array of voltage detection electrodes configured to divide the flow cross-section into multiple pixel regions and measure the induced voltage in each pixel region; and processing means for determining the axial velocity profile of the conducting fluid by calculating the local axial velocity of the conducting fluid in each pixel region.

[0020] In one embodiment the flow tube comprises a non-electrically conducting body. The flow tube may alternatively comprise: an outer body portion formed from a low magnetic permeability material; and an inner body portion formed from a non-electrically conducting material. In either case the flow tube may further comprise an annular liner having a conductivity that is generally the same as the conductivity of the conducting fluid.

[0021] Preferably the means for generating a magnetic field comprises a Helmholtz coil having a pair of coils arranged symmetrically on opposing sides of the flow tube. The means for generating a magnetic field may be configured to generate a substantially uniform magnetic field across the flow tube cross-section or may be configured to generate a non-uniform magnetic field across the flow tube cross-section.

[0022] In preferred embodiments of the invention the means for generating a magnetic field is configured to generate a magnetic field having a single magnetic field projection (P=1). The means for generating a magnetic field may further be configured to successively generate multiple magnetic fields, each magnetic field having a different magnetic field projections (P>1). Preferably the array of electrodes comprises E electrodes that are configured to divide the flow cross-section up to M pixel regions and measure up to M induced voltages in the conducting fluid when a single magnetic field projection is applied, whereby M= E-1. In particular the array of E electrodes may be configured to divide the flow cross-section into N pixel regions and measure up to N induced voltages in the conducting fluid when multiple magnetic field projections are applied, whereby N=PM.

[0023] In preferred embodiments of the invention the processing means is configured to calculate the local axial velocity of the conducting fluid in each said pixel region using the measured induced voltage for each said pixel region and predetermined weight functions for each said pixel region. Additionally the processing means may be configured to calculate the volumetric flow rate of the conducting fluid. Preferably, when the conducting fluid is a conducting single phase fluid, the processing means is configured to calculate the volumetric flow rate using the local axial velocity of the conducting fluid in each pixel region. When the conducting fluid is a conducting continuous phase of a multiphase fluid, the processing means is preferably configured to calculate the volumetric flow rate using the local axial velocity in each pixel region and local concentration distribution of the conducting fluid.

[0024] The flow meter may further comprise means for measuring the local concentration distribution of the conducting continuous phase of the multiphase fluid and optionally the local concentration distribution of the one or more dispersed phases of the multiphase fluid. The means for measuring the local concentration distribution may be configured to use an electrical resistance tomography technique or an impedance cross correlation technique. The flow meter may further comprise means for determining the mean density of the multiphase fluid and means for determining the density of each phase of a multiphase fluid. The flow meter may be configured to determine the axial velocity profile, and optionally the volumetric flow rate, of each phase of a multiphase fluid.

[0025] Preferably the processing means comprises means for controlling the operation of the means for generating the magnetic field. For example the means for controlling the operation of the means for generating the magnetic field may comprise a coil excitation circuit for controlling the flow of current to the Helmholtz coil.

[0026] Preferably the processing means comprises a temperature compensating circuit to compensate for the change in the resistance of the Helmholtz coil as the temperature varies.

[0027] Preferably the processing means comprises means for collating the induced voltages.

[0028] Preferably the processing means comprises a control circuit to compensate for the effects of any unwanted voltage components.

[0029] According to another aspect of the invention there is provided a method for monitoring the flow of a conducting fluid comprising: generating an induced voltage in the conducting fluid; measuring the induced voltage in multiple pixel regions across the flow cross-section; determining the axial velocity profile of the conducting fluid by calculating the local axial velocity of the conducting fluid in each pixel region.

[0030] Preferably calculating the local axial velocity of the conducting fluid in each pixel region comprises using the measured induced voltages and predetermined weight functions. The method may further comprise determining the weight functions for each pixel in the flow cross-section prior to monitoring the flow of the conducting fluid.

[0031] Preferably the method further comprises applying a magnetic field with a single magnetic field projection across the conducting fluid so as to induce a voltage in the conducting fluid, whereby the magnetic field is a uniform magnetic field or a non-uniform magnetic field. The method may also comprise successively applying multiple magnetic fields across the conducting fluid, whereby each magnetic field has a different magnetic field projection.

[0032] In preferred embodiments of the invention the method may comprise determining the volumetric flow rate of the conducting fluid.

[0033] The conducting fluid may be a conducting single phase fluid and the method may comprise determining the volumetric flow rate of the conducting fluid comprises using the axial velocity profile of the conducting fluid. The conducting fluid may alternatively be a conducting continuous phase of a multiphase fluid, and the method may comprise determining the volumetric flow rate of the conducting fluid comprises using the axial velocity profile and local concentration distribution of the conducting fluid. The method may further comprise measuring the local concentration distribution of the conducting fluid and optionally measuring the local concentration distribution of one or more dispersed phases.

[0034] Preferably measuring the local concentration distribution of the conducting fluid comprises using an electrical resistance tomography technique or an impedance cross correlation technique.

[0035] Preferably the method further comprises determining the axial velocity profile, and optionally the volumetric flow rate, of each phase of the multiphase fluid. The method may also further comprise controlling the magnetic field. For example the magnetic field may be generated by a Helmholtz coil, and the controlling of the magnetic field may comprise controlling the flow of current to the Helmholtz coil.

[0036] Preferably the method may further comprise compensating for a variation in the magnetic field due to temperature fluctuations.

[0037] The method may further comprise collating the induced voltages.

[0038] The method may further comprise compensating for the effects of any unwanted voltage components.

[0039] Viewed from another broad aspect the present invention provides a means for determining the local concentration distribution of a conducting continuous phase of a multiphase fluid using an impedance cross correlation technique comprising: a flow tube through which the multiphase fluid may flow; a first array of electrodes arranged uniformly around the circumference of the flow tube; a second array of electrodes arranged uniformly around the circumference of the flow tube and axially separated from the first array by a predetermined distance; means for applying a sequence of electrical potentials to the electrodes of the second array so as to generate an electrode potential rotational pattern; means for measuring the resistance of the multiphase fluid in a predetermined region of the flow cross section; means for determining the conductivity of the multiphase fluid using the resistance of the multiphase fluid; and means for determining the local concentration distribution of the conducting continuous phase using the conductivity of the multiphase fluid.

[0040] Viewed from a further broad aspect the present invention provides a means for determining the mean density of a multiphase fluid comprising: a flow tube through which the multiphase fluid may flow; means for determining a differential pressure in the multiphase fluid by measuring the pressure of the multiphase fluid at different points along the length of the flow tube; means for determining the mean fluid density of the multiphase fluid using the measured differential pressure.

BRIEF DESCRIPTION OF THE DRAWINGS



[0041] For a better understanding of the invention, and to show how the same may be carried into effect, reference will now be made, by way of example, to various specific embodiments of the different aspects of the invention as shown in the accompanying diagrammatic drawings, in which:

Figure 1 depicts a flow chart showing the operational steps of an embodiment of an electromagnetic flow mete according to the invention;

Figure 2a depicts a perspective view of a further embodiment of an electromagnetic flow meter according to the invention;

Figure 2b depicts a schematic cross-sectional view of the flow meter of Figure 2a;

Figure 2c depicts a schematic top view of the flow meter of Figure 2a;

Figure 2d depicts a schematic view of the electrode configuration and pixels of the flow cross-section of the flow meter of Figure 2a;

Figure 2e is a table showing how the pixels relate to the pairs of electrodes and potential different measurements for the flow meter of Figure 2a;

Figure 3a depicts a perspective view of an embodiment of a voltage detecting electrode according to the invention;

Figure 3b depicts the electrode of Figure 3a mounted in the wall of the flow tube of an electromagnetic flow meter;

Figure 4 depicts a housing enclosing at least part of the flow meter of Figure 2a;

Figure 5 depicts a combined coil excitation and temperature compensation circuit for an electromagnetic flow meter according to the invention;

Figure 6 depicts a graph showing how the coil current varies with time;

Figure 7 depicts a voltage measuring circuit for collating the voltage measured between the jth pair of electrodes and a control circuit for eliminating the effects of an unwanted voltage Uo.

Figure 8a depicts a 3-dimensional schematic diagram of an example of a flow meter where the flow tube defines the computing domain;

Figure 8b depicts a 2-dimensional schematic diagram of the flow meter of Figure 8a; Figure 9 depicts a schematic view of the flow pixels in the flow meter of Figure 8a;

Figure 10 depicts a table that lists the geometries of the flow meter of Figure 8a;

Figure 11a depicts a distribution of the Lorentz force per unit volume when simulating a flow meter to determine weight values;

Figure 11b depicts the electrical potential on the z-plane when simulating a flow meter to determine weight values;

Figure 12a depicts the induced voltages when simulating a flow meter to determine weight values;

Figure 12b depicts the weight values calculated from the induced voltages of Figure 12a;

Figure 13a depicts a distribution of the Lorentz force per unit volume on a conducting fluid with a uniform flow profile;

Figure 13b depicts the electrical potential on the z-plane of the conducting fluid of Figure 13a;

Figure 14a depicts a distribution of the Lorentz force per unit volume on a conducting fluid with a non-uniform flow profile;

Figure 14b depicts the electrical potential on the z-plane of the conducting fluid of Figure 14a;

Figure 15a depicts the axial velocity profile of a conducting fluid with a uniform flow profile;

Figure 15b depicts the axial velocity profile of a conducting fluid with a non-uniform flow profile;

Figure 16 depicts an electrode potential rotational pattern showing the resultant effective sensing region and centre of action for an 8 electrode system of an ICC measuring means.


DETAILED DESCRIPTION OF THE INVENTION


1. Determining The Axial Velocity Profile Of A Conducting Fluid



[0042] The invention relates to an electromagnetic flow meter and method for monitoring the flow of a conducting fluid. As part of the monitoring process, the flow meter and method determine the axial velocity profile of a conducting fluid. The conducting fluid may be a conducting single phase fluid or a conducting continuous phase of a multiphase fluid. The conducting fluid may have a uniform flow profile or a non-uniform flow profile.

[0043] In its simplest form, the flow meter comprises a flow tube, a means for generating a magnetic field across the flow tube cross-section, an array of voltage detection electrodes circumferentially arranged around the flow tube and processing means for determining the axial velocity profile of a conducting fluid.

[0044] The flow tube is a pipe along which fluid can flow when the flow meter is in use. As mentioned above, the fluid may be a conducting single phase fluid or a multiphase fluid comprising a conducting continuous phase and at least one dispersed phase.

[0045] The means for generating a magnetic field is configured to generate a magnetic field across the flow tube cross-section so that a voltage is induced in the conducting fluid as it flows through the flow tube. The means for generating a magnetic field may successively generate multiple magnetic fields, each magnetic field having a different magnetic flux density distribution. Each magnetic flux density distribution may be referred to as a magnetic field projection. In any given application, the number of projections generated is P, whereby P≥1.

[0046] The flow meter comprises an array of E electrodes. The electrodes are circumferentially mounted on an internal surface of the flow tube so as to detect the induced voltage between various points on the circumference of the flow tube. The electrodes are configured so as to divide the flow cross-section into pixels (discrete regions) and to measure the local induced voltage in each pixel.

[0047] The array of electrodes can divide the flow cross-section into a maximum of M pixels at any one time, where M=E-1. Hence, up to M independent induced voltage measurements can be made between E electrodes when a magnetic field is applied. Consequently, the total number of independent induced voltage measurements that can be made is N, where N=PM. In other words, the electrodes are able to measure the induced voltage of the conducting fluid in N pixels (discrete regions) in the flow cross-section.

[0048] As will be explained in detail below, the processing means is configured to determine the axial velocity profile of the conducting fluid using the measured induced voltages and predetermined weight functions.

[0049] In an electromagnetic flow meter, the charged particles of the conducting fluid experience a Lorentz force as they move in the magnetic field. The Lorentz force acts in a direction perpendicular to both the conducting fluid's motion and the applied magnetic field. The local current density j in the conducting fluid is governed by Ohm's law in the form of

where σ is the local fluid conductivity, E is the local electric field in the stationary coordinate system, v is the local fluid velocity, and B is the local magnetic flux density. The expression (v×B) represents the Lorentz force induced by the fluid motion, whereas E is principally due to charges distributed in and around the fluid.

[0050] For fluids of essentially constant conductivity, the local potential difference U generated in the conducting fluid is defined by the equation:

where v is the local fluid velocity and B is the local magnetic flux density.

[0051] In the present invention, where the flow cross section of the flow meter is divided into N pixels and where N potential difference measurements are made between the electrodes placed around the internal circumference of the flow tube a solution to equation 2 is of the form:

where Ai represents the cross sectional area of the ith of N pixels into which the flow cross section is divided, vi is the mean axial flow velocity in the ith pixel, Uj is the jth of N potential difference measurements made at the boundary of the flow, the term wij is a so called weight value which relates the flow velocity in the ith pixel to the jth potential difference measurement and a is the internal pipe radius and B is the mean magnetic flux density in the flow cross section.

[0052] Thus, the local axial velocity of the conducting fluid in each of the N pixels can be determined from predetermined weight values wij and N potential difference measurements Uj made on the boundary of the flow using the electrodes.

[0053] The N independent equations arising from equation 3 can be expressed by the following matrix equation:

in which V is a single column matrix containing the pixel velocities vi, W is a square matrix containing the relevant weight values wij, A is a square matrix containing information on the pixel areas Ai and U is a single column matrix containing the measured potential differences Uj.

[0054] Using a standard matrix inversion technique, such as the Tikhonov technique, equation 4 can be solved giving:



[0055] Thus, when a single magnetic field projection is applied, the flow meter can determine the axial velocity profile of the conducting fluid by dividing the flow cross section into N pixels and using equation 5 to derive the axial flow velocity in each of the N pixels.

[0056] It has been found that the accuracy of the flow meter is at least partially determined by its spatial resolution (i.e. the number of pixels, N) and thereby the number of potential difference measurements that can be taken. Since the number of electrodes is restricted by the circumferential size of the flow tube, the accuracy of the flow may thereby be improved by increasing the number of magnetic flux density distributions P.

[0057] For example, a flow meter may comprise 8 electrodes that are configured to measure the potential difference in 7 discrete locations on the boundary of the flow cross section. If a single magnetic field projection is applied (P=1) the flow cross section is divided into only 7 pixels because a total of only 7 independent potential difference measurements can be measured by the flow meter. However, if two different magnetic flux density distributions are applied (P=2), the flow cross section is divided into 14 pixels because a total of 14 independent potential difference measurements can be measured by the flow meter.

[0058] When multiple magnetic flux density distributions are applied, the N independent potential difference measurements can be related to the unknown axial flow velocity vi in each of N discrete regions, or pixels, in the flow cross section by N independent equations of the form

where Uj,p is the jth (of M) independent potential difference made using the pth (of P) projections, Ai is the area of the ith (of N) pixels, wi,j,p is a weight value relating the flow velocity in the ith pixel to the jth potential difference measurement using the pth magnetic field projection, Bp is the mean flux density in the flow cross section associated with the pth projection and a is the internal radius of the flow tube.

[0059] The N independent equations arising from equation 6 above can be written in matrix form as

where U is an (N×1) matrix containing the measured potential differences, W is an (N×N) matrix containing the known weight values, A is an (N×N) matrix containing information on the known pixel cross sectional areas, RB is an (N×N) matrix containing information on the reciprocals of the known mean flux densities in the flow cross section (associated with each of the P magnetic field projections) and V is an (N×1) matrix containing the unknown axial flow velocities in the N pixels.

[0060] By using standard matrix inversion techniques, such as the Tikhonov technique, equation 7 can be solved giving



[0061] Thus, when multiple magnetic field projections are applied, the flow meter can determine the axial velocity profile of the conducting fluid by dividing the flow cross section into N pixels and using equation 7 to derive the axial flow velocity in each of the N pixels.

[0062] The matrices used in equations 3 to 8 are defined below:
  1. (i) U is a (N×1) matrix where the term Uj,p represents the jth potential difference measurement associated with the pth magnetic field projection; (j = 1 to M and p = 1 to P).

  2. (ii) RB is an (N x N) diagonal matrix where Bp is the mean magnetic flux density in the flow cross section associated with the pth magnetic field projection; (p = 1 to P).

  3. (iii) W is an (N×N) matrix where wi,j,p is the weight value relating the axial flow velocity in the ith pixel to the jth potential difference measurement associated with the pth magnetic field projection (i = 1 to N , j = 1 to M and p = 1 to P).

  4. (iv) A is a (N×N) diagonal matrix where Ai is the cross sectional area of ith pixel; (i = 1 to N).

  5. (v) V is a (N×N) matrix where vi is the axial flow velocity in the ith pixel; (i = 1 to N).



[0063] It will be understood that when determining the axial velocity profile of a conducting single phase fluid, the flow meter calculates the mean velocity of the conducting fluid in each pixel. When determining the axial velocity profile of the conducting continuous phase of a multiphase fluid, the flow meter calculates the simple mean velocity of the conducting continuous phase which is calculated in each pixel. By simple mean it is implied that the calculated velocity is not weighted by the concentration, or local volume fraction, of the conducting continuous phase in the pixel.

[0064] As explained above, the axial velocity of the conducting fluid in each of the N pixels is determined using predetermined weight values wij, wi,j,p. The weight values represent the relative contribution of the fluid flow at a particular spatial location in the flow cross section to the measured potential difference. Thus, when determining the axial velocity in N pixels, N2 weight values are required. These weight values are calculated, from solutions of Maxwell's equations of electromagnetism. These N2 weight values are dependent upon the geometry of the electromagnetic flow meter, its materials of construction and also upon the magnetic field projections that are employed. The weight values need only be calculated once, prior to using the flow meter device. An example of how weight values can be calculated is described below.

[0065] It has been found that the axial flow velocity of each pixel as determined by the present invention are in agreement with the reference pixel velocities determined by other measuring means. Thus, the axial flow velocities enable reasonably accurate volumetric flow estimates to be made, even in the presence of highly non-uniform velocity profiles.

2. Determining Volumetric Flow Rate Of A Conducting Fluid



[0066] Once the axial velocity profile of the conducting fluid is determined, the flow meter and method may determine the volumetric flow rate of the conducting fluid.

[0067] For example when monitoring a conducting single phase fluid, the volumetric flow rate Qc of the conducting single phase fluid can be calculated from the determined axial velocity profile as follows;

in which Qc is the volumetric flow rate, Ai is the area of the ith (of N) pixel, and vi is the axial velocity in the ith pixel.

[0068] When monitoring a conducting continuous phase of a multiphase phase fluid, the volumetric flow rate of the conducting continuous phase Qc can be determined providing the local concentration distribution (also known as the local volume fraction distribution) of the conducting continuous phase αc in the flow cross section is known.

[0069] For example, in a two phase fluid the volumetric flow rate of the conducting continuous phase Qc can be calculated using the following equation:

where vc is the local axial velocity in the flow cross section, αc is the local volume fraction distribution of the conducting continuous phase and A is the flow tube cross sectional area.

[0070] In a three phase fluid, the volumetric flow rate of the conducting continuous phase Qc can be calculated using the following equation:

where Vic is the velocity of the conducting continuous phase in the jth pixel into which the flow cross section is divided, αic is the local volume fraction distribution of the conducting continuous phase in the ith pixel, and A i is the area of the ith pixel.

[0071] The flow meter preferably comprises means for measuring the local volume fraction distribution of the conducting continuous phase. The flow meter may comprise means for measuring the local volume fraction distribution using the well known technique of Electrical Resistance Tomography (ERT). Alternatively, the flow meter may comprise means for measuring the local volume faction distribution using an impedance cross correlation technique (ICC). An example of a flow meter comprising means for measuring the local volume fraction distribution using an impedance cross correlation technique is described below.

3.Monitoring The Flow Each Phase Of A Multiphase Fluid



[0072] In addition to monitoring the flow of a conducting continuous phase of a multiphase fluid, the flow meter and method may also monitor the flow of the one or more dispersed phases of the multiphase fluid. For example, the flow meter and method may determine the local velocity of a dispersed phase in the flow cross section vd and optionally the volumetric flow rate of a dispersed phase Qd.

[0073] The flow meter may comprise means for measuring the local velocity of the dispersed phase using an impedance cross correlation technique (ICC). An example of flow meter comprising means for measuring the local velocity of the dispersed phase using an impedance cross correlation technique is described below.

[0074] The volumetric flow rate of a dispersed phase can be determined providing the local concentration distribution (also known as the local volume fraction distribution) of the dispersed phase αd in the flow cross section is also known. (The flow meter may comprise means for measuring the local volume fraction distribution of the dispersed phase αd using an impedance cross correlation technique).

[0075] For example, in a two phase fluid the volumetric flow rate of the dispersed phase Qd can be calculated using the following equation:

where vd is the local axial velocity in the flow cross section, αd is the local volume fraction distribution of the dispersed phase and A is the flow tube cross sectional area.

[0076] In a three phase fluid, the volumetric flow rate of a particular dispersed phase Qd can be calculated using the following equation:

where λd is the means volume fraction of the dispersed phase in the flow cross section as measured using an impedance cross correlation technique, vd is the mean velocity of the dispersed phase and A is the flow tube cross sectional area.

[0077] The flow meter may also comprise density measuring means for determining the mean density of a multiphase fluid.

[0078] Examples of flow meters configured to determine the local velocity of a dispersed phase in the flow cross section vd the volumetric flow rate of a dispersed phase Qd and the means density of a multiphase fluid are described below.

Monitoring Of Conducting Fluid In A Partially Filled Or Partially Blocked Pipe



[0079] The flow meter and method can be used to make flow rate measurements of a conducting fluid in partially filled or partially blocked pipes, provided that a minimum of two electrodes are immersed in that part of the cross section of the pipe where flow still occurs.

Flow Diagram Depicting The Operation Of A Flow Meter And Method



[0080] Figure 1 depicts a flow diagram that shows the operational steps of an example of a flow meter and method according to the present invention. In this particular example, the conducting fluid is water:

Step 1 - Prior to starting the measuring procedure, the flow meter calculates the weight functions for each pixel.

Step 2 - On starting the measuring procedure, the flow meter generates a pth (of P) magnetic field projections so as to induce a voltage in the conducting fluid.

Step 3 and 4 - the flow meter takes M potential difference measurements at different points around the circumference of the flow tube.

Step 5 - if multiple magnetic field projections are to be used, the flow meter repeats steps 2 to 4 until P magnetic field projections have been applied.

Step 6 - the flow meter determine the total number of pixels/potential difference measurements.

Step 7 - the flow meter calculates the axial velocity of the conducting fluid in each of the pixels using equation 5 if P=1 or equation 8 if P>1 and optionally the flow meter stops/returns to the start if only the axial velocity profile of a conducting fluid is required.

Optional Step 8 - when monitoring a conducting single phase fluid, the flow meter may calculate the volumetric flow rate of the conducting fluid using equation 9.

Optional Step 9 - when monitoring a two phase fluid, the flow meter may combine the axial velocity profile of a conducting continuous phase with local volume fraction measurements of the conducting continuous phase and dispersed phase and also the local axial velocity of the dispersed phase (measured using an ERT or ICC technique) to determine the volumetric flow rate of both the conducting continuous phase and dispersed phase in a two phase fluid.

Optional Step 10 - when monitoring a three phase fluid, the flow meter may combine the axial velocity profile of the conducting continuous phase with local volume fraction measurements of the conducting continuous phase and dispersed phases, local axial velocity of the dispersed phases (measured using an ERT or ICC technique) and density measurements to determine flow characteristics of each phase of the three phase fluid.

Optional Step 11 - when monitoring a three phase fluid, the flow meter may combine the results determined in step 10 with venturi measurements to cross-reference the determined flow characteristics of each phase of the three phase fluid.


4. The Flow Meter



[0081] As mentioned above, the flow meter according to the present invention comprises a means for generating a magnetic field across the flow tube cross-section, an array of voltage detection electrodes circumferentially arranged around the flow tube and processing means for determining the flow characteristics of the conducting fluid.

4a The Flow Tube



[0082] The flow tube is a pipe along which fluid may flow when the flow meter is in use. The fluid may be a conducting single phase fluid or a multiphase fluid comprising a conducting continuous phase and at least one dispersed phase.

[0083] The flow tube preferably comprises a body formed from a non-electrically conducting material, such as PTFE. Alternatively, the flow tube may comprise an outer body portion formed from a low magnetic permeability material and an inner body portion (e.g. a liner) formed from a non-electrically conducting material, thereby ensuring the electrodes are electrically isolated from each other. In yet another embodiment, the flow tube may comprise an inner body portion (e.g. an annular liner) that is formed from a material having a conductivity that is at least similar to that of the conducting phase flow and deployed between the electrodes and the flow in order to improve the uniformity of weight function values.

[0084] The flow tube may have any suitable diameter and length. The diameter and length of the flow tube may be selected according to the type of fluid, volume of fluid and/or location of the flow tube.

[0085] When the flow meter is in use, an electrically conducting single phase fluid may flow through the flow tube, such as water. Alternatively, a multiphase fluid having an electrically conducting continuous phase and one or more dispersed phases may flow along the flow tube. Examples of a multiphase fluid include solids-in-water flows such as sludges and slurries, oil-in-water flows, gas-in-water flows and oils and gas-in-water flows.

[0086] Figures 2a-2d depict an embodiment of a flow meter according to the first aspect of the invention whereby the flow tube (1) is a PTFE pipe with an internal diameter of approximately 80mm, an external diameter of approximately 110mm and a length of approximately 410mm.

4b. The Flange



[0087] The flow meter may comprise a flange arranged at one or both ends of the flow tube. The flange may be configured so as to allow the flow meter to be coupled to a further apparatus, such as a pipe. The flange may comprise one or more apertures to receive securing means (e.g. bolts, screws, clips etc) suitable for securing the flow meter to a further apparatus.

[0088] In Figures 2a-2d, the electromagnetic flow meter comprises a first flange (2a) arranged at a first end of the flow tube and a second flange (2b) arranged at a second end of the flow tube. The flanges have a diameter of approximately 203mm and a thickness of approximately 24mm. The flanges are configured so that the flow meter can be connected to external pipe work. Each flange comprises a plurality of bolt holes (2c) with an internal diameter of approximately 16mm.

4c. Means for Generating a Magnetic Field



[0089] The means for generating a magnetic field in the flow meter is configured to generate a magnetic field so that a voltage is induced across the conducting fluid as it flows through the flow tube. As shown in Figures 2a-2d, the means for generating a magnetic field generate a magnetic field that is orthogonal to both the direction of the flowing fluid and plane of the array of electrodes so that the potential difference at the boundary of the flow tube can be detected by the electrodes.

[0090] The means for generating a magnetic field may optionally generate a generally uniform magnetic field across the flow tube. The means for generating a magnetic field is configured to generate a non-uniform magnetic field across the flow tube. The means for generating a magnetic field may be configured to generate a non-uniform magnetic field across each pixel (discrete region) of the flow tube cross-section. The non-uniform magnetic field may be applied to help distinguish between different axisymmetric or non-axisymmetric velocity profiles.

[0091] The means for generating a magnetic field comprises any suitable electromagnetic means for generating a magnetic field. For example, the means for generating a magnetic field may comprise a Helmholtz coil mounted around the periphery of the flow tube. The Helmholtz coil comprises a pair of identical coils (3a, 3b) arranged symmetrically on opposing sides of the flow tube as shown in Figures 2a to 2d. The Helmholtz coil may be securely mounted to the flow tube using coil supports/stiffeners (4) and coil mounting brackets (5).

[0092] The Helmholtz coil may comprise any suitable size and number of turns. The coils may have a mean diameter that is approximately twice the mean coil separation distance. In the embodiment depicted in Figures 2a-2d, the coils of the Helmholtz coil are approximately 30mm thick, approximately 29mm wide, have an internal diameter of approximately 202mm, have an external diameter of approximately 260mm, comprise approximately 1024 turns of 0.776mm diameter wire and have an approximately 5AMP capacity.

[0093] As mentioned above, the means for generating a magnetic field may successively generate multiple magnetic fields, each magnetic field having. a different magnetic flux density distribution. Each magnetic flux density distribution may be referred to as a magnetic field projection. In any given application, the number of projections generated is P, whereby P≥1. Each projection may be consecutively applied for a short time interval.

[0094] For each of the P magnetic field projections there is a 'complementary projection' in which the magnitude of the current in each coil remains the same but the direction of the current is reversed. If, immediately after applying a given projection, the 'complementary projection' is applied for an equivalent length of time this enables the direction of the magnetic field to be reversed which, in turn, minimises electrochemical effects at the interface between the electrodes and the flowing fluids. However application of the 'complementary projection' does not enable additional independent potential difference measurements to be made.

[0095] The magnetic flux density distribution across the flow tube is dependent on the magnitude and direction of the electrical current in the Helmholtz coil. Thus, different magnetic flux density distributions may be generated by varying the magnitude and/or direction of the electrical currents supplied to the Helmholtz coil. Alternatively or additionally, the different magnetic flux density distributions may be generated by arranging multiple pairs of Helmholtz coils in different planes around the flow tube.

[0096] The means for generating a magnetic field preferably generates a magnetic field having a rectangular waveform so as to minimise the effects of electrolysis at the electrodes. For example, the means for generating a magnetic field may generate a rectangular waveform magnetic field alternating between +/-40Gauss.

[0097] In the embodiment depicted in Figures 2a-2d, the Helmholtz coil is configured such that current flows through both coils in the same direction and each coil carries an equal amount of electric current. Hence, the Helmholtz coil generates a generally single and uniform magnetic flux density distribution in the flow cross section. The magnitude of the magnetic flux density in the y direction is relatively constant and the mean value B of the magnitude of the y component of the magnetic flux density in the flow cross section is 7.996Gauss (7.996 x 10-4 T).

4d. The Electrode Array



[0098] The flow meter comprises an array of E electrodes. The electrodes are circumferentially mounted on an internal surface of the flow tube so as to detect the induced voltage at various points on the internal circumference of the flow tube. The electrodes are arranged as opposing pairs around the circumference of the flow tube so as divide the flow cross-section into pixels (discrete regions) and to measure the local induced voltage in each of the pixels.

[0099] The array of E electrodes may be arranged so that up to M independent induced voltage measurements can be made between the electrodes, whereby M = E-1. Thus, the total number of independent induced voltage measurements that can be made is N, where N=PM. In other words, the electrodes are able to measure the induced voltage of the conducting fluid in N pixels (discrete regions) in the flow cross section.

[0100] In one embodiment, the electromagnetic flow meter may be configured to take seven independent potential difference measurements (M=7), provided that the number of electrodes E is greater than or equal to 8 (M+1). Thus, if a single magnetic field projection is used (P=1) the flow cross section is thereby divided into a total of 7 pixels (N=7) and a total of 7 independent potential difference measures can be taken over the measuring period. If three magnetic field projections (P=3) are successively applied then the flow cross section is divided into a total of 21 pixels and a total of 21 independent potential difference measurements can be taken over the measuring period.

[0101] As shown in Figure 2d, the pixels may be parallel, elongate regions extending between opposing sides of the flow tube. However, it should be noted that the pixels can be of any shape or size provided that, in total, they entirely cover the (normally circular) cross section of the flow tube.

[0102] The electromagnetic flow meter may comprise any suitable number of electrodes. In the embodiment depicted in Figures 2a-2d, the flow meter comprises 16 electrodes (e) arranged at the internal boundary of the flow pipe (at plane z=0) and in contact with the flowing fluid. The electrodes are placed at angular intervals of 22.5 degrees on the flow pipe boundary. In Figure 2d, the electrodes are denotated e1, e2, etc with electrodes e5 at the top of the flow cross section and electrode e13 at the bottom of the flow cross section. For this particular flow meter, the electrodes are configured such that the flow cross-section is divided into seven pixels. The geometry of these seven pixels is chosen such that the chords joining seven pairs of electrodes are located at the geometric centres (in the y direction) of the pixels. The fluid pixels are categorized as pixel 1 at the top of the flow cross section to pixel 7 at the bottom of the flow cross section. Seven potential difference measurements can be made between the seven electrode pairs. Since the jth potential difference measurement Uj is made between the jth electrode pair, the potential difference measurements are notated U1, U2 etc in the table depicted in Figure 2e. The local magnetic flux density B is perpendicular to both the flow direction and to the chords joining the electrode pairs.

[0103] The electrodes may be made from a low magnetic permeability material (e.g. stainless steel or brass) if the flow tube comprises a non-electrically conducting pipe wall. Alternatively, if the flow tube comprises an outer conducting wall and an inner non-conducting liner, the electrodes are configured to be electrically insulated from the outer metal wall. This may be achieved by enclosing each of the electrodes in a non-conducting liner.

[0104] Figure 3a depicts an embodiment of an electrode (10) for an electromagnetic flow meter when the flow tube comprises a non-electrically conducting body. The electrode comprises an elongate body portion (10a) and a head portion (10b). As shown in Figure 3b, the electrode is mounted on the flow tube such that the head portion is arranged on the inner side of the flow tube so it can form a contact with the flowing fluid and the body portion extends through an aperture in the wall of the flow tube (11) and protrudes beyond the flow tube. An O-ring seal (12) is arranged between the body portion and the internal surface of the aperture so as to form a seal between the electrode and the flow tube. The O-ring seal may be mounted in a circumferential groove formed on the body portion of the electrode. The electrode may be secured to the flow tube using any suitable securing means, such as a retaining nut (13). A rubber washer (14) may be arranged between the retaining nut and the external surface of the flow tube. The electrode may be electrically connected to an electrical cable using an suitable coupling means, such as a screw connector (15).

4e. The Housing



[0105] Figure 4 depicts an embodiment of a flow meter whereby the flow tube and array of electrodes may be located (encased) in a housing (100). The housing preferably acts as a Faraday cage so as to prevent external electrical fields from interfering with the induced voltages measure at the electrodes. The housing preferably acts as a magnetic shield so as to prevent external magnetic fields from interfering with the operation of the flow meter and also prevent "leakage" of the magnetic field generated by the means for generating a magnetic field into the surrounding environment.

[0106] The housing may comprise an aperture (101) through which an electrical cable may extend from the electrode array to the processing means.

4f. The Processing Means



[0107] The flow meter comprises processing means for determining at least the axial velocity profile of the conducting fluid.

[0108] The processing means is configured to determine the axial velocity profile by calculating the axial velocity of the conducting fluid in each pixel of the flow cross-section.

[0109] The processing means comprises means to calculate the axial velocity of the conducting fluid in each pixel of the flow cross section using an inverted matrix. The processing means comprises means to derive the axial velocity of the conducting fluid in each pixel of the flow cross section using equation 5 when a single magnetic field projection is applied. The processing means comprises means to derive the axial velocity of the conducting fluid in each pixel of the flow cross section using equation 8 when multiple magnetic field projections are applied.

[0110] The processing means may comprise means to determine the weight function values of the pixel prior to determining the axial velocity profile of the conducting fluid.

[0111] Having calculated the axial velocity of the conducting fluid in each pixel, the processing means is preferably further configured to determine the volumetric flow rate of the conducting fluid.

[0112] When the conducting fluid is a conducting single phase fluid, the processing means may comprise means to calculate the volumetric flow rate using the axial velocity profile of the conducting fluid. More specifically, the processing means may comprise means to calculate the volumetric flow rate using equation 9 when the conducting fluid is a conducting single phase fluid.

[0113] When the conducting fluid is a conducting continuous phase of a multiphase fluid, the processing means may be configured to derive the volumetric flow rate using the axial velocity profile and local volume fraction distribution of the conducting fluid.

[0114] The processing means may comprise means for measuring the local volume fraction distribution of the conducting fluid using an electrical resistance tomography technique or an impedance cross correlation technique.

[0115] The processing means may be configured to determine flow characteristics of each phase of a multiphase fluid. For example, the processing means may be configured to determined the axial velocity profiles, local volume fraction distributions and/or volumetric flow rates of the conducting continuous phase and the at least one dispersed phase.

[0116] The processing means may be configured to control the operation of the flow meter.

[0117] The processing means may be a microcontroller. One example of a suitable microcontroller is the Microrobotics VM-1 microcontroller. The processing means may comprise any suitable hardware, operating systems and/or software.

[0118] The processing means may comprise display means for displaying the calculated flow characteristics of the fluid.

[0119] The processing means may be configured to control the operation of the means for generating the magnetic field. For example, the processing means may comprise means for controlling the flow of current to the means for generating a magnetic field (e.g. the Helmholtz coil). In one embodiment, the means for controlling the flow of current supplies a sinusoidal current to the Helmholtz coil to achieve a time varying magnetic field. In another embodiment, the means for controlling the flow of current (such as the coil excitation circuit shown in Figure 5) may apply a hybrid square wave current (as shown in Figure 6) to the Helmholtz coil (Coil 1, Coil 2) to achieve a time varying magnetic field. The hybrid square wave current may complete a cycle in approximately 0.5 seconds or less.

[0120] With reference to the coil excitation circuit shown in Figure 5, the dc power supply unit (dc PSU) is connected to a network of solid state relays. The solid state relay network (SSRN) is controlled by the processing means in such a way that at any instant in time the voltages applied at points 'a' and 'b' in Figure 5 are as per the table below.
Relay PositionVoltage at 'a'Voltage at 'b'
RP1 Upsu 0
RP2 0 0
RP3 0 Upsu


[0121] When the voltage at a is Upsu and the voltage at b is 0 the maximum current ic,max flows to the coils (see Figure 6). When the voltage at a is 0 and the voltage at b is Upsu the minimum current ic,min flows to the coils (where ic,min = -ic,max). When the voltages at a and b are both 0, no current flows to the coils.

[0122] The two coils of the Helmholtz coil (Coil1, Coil 2) are closely matched and the resistance Rc of each coil has a known value of Rc,15 when the coils are at a temperature of 15°C. However, the ambient temperature variations and the heating of the coils due to the coil current cause the value of Rc, and thereby ic,max , to vary with time. As explained above, the flow induced voltages Uj from which the flow velocity profile is reconstructed are proportional to Bmax - the maximum value of the mean magnetic flux density in the flow cross section. In turn, Bmax is proportional to ic,max whereby

where K is a known constant.

[0123] Therefore, accurate velocity profile reconstruction relies upon knowing Bmax at all times and so it is necessary to know ic,max at all times. Accordingly, the processing means preferably comprises a temperature compensating circuit to determine ic,max and thereby Bmax .

[0124] Figure 5 depicts an example of a temperature compensating circuit whereby the coil current ic is passed through the high tolerance reference resistor with a known resistance Rref. The reference resistor has a very low temperature coefficient. A voltage Ur appears across Rref and is fed to the processing means via a differential amplifier (DA). Ur is then measured by the analogue to digital converters within the processing means. It is known that the maximum value of Ur is Ur,max where



[0125] Since Rref is known and Ur,max is measured by the processing means ic,max can be calculated from equation 15.

[0126] Since K is known Bmax can then be calculated from equation 14, thereby enabling the true value of the maximum mean magnetic flux density to be used at all times in the velocity reconstruction calculations.

[0127] The processing means may be further configured to collate the potential difference measurements in each pixel. The processing means may comprise means for collating the potential difference between each pair of electrodes. For example, Figure 7 depicts an example of a voltage measuring circuit suitable for collating the potential difference measurements between a pair of electrodes. In Figure 7, a voltage measurement is being made between the jth pair of electrodes. Note that a circuit such as that shown in Figure 7 is necessary for each pair of electrodes between which is required to make a voltage measurement. Hence, for a flow meter comprising 16 electrodes and requiring 15 voltage measurements, 15 such circuits are required.

[0128] It can be seen that in the voltage measuring circuit of Figure 7, a time dependent flow induced voltage U*j appears between the jth caused electrode pair. The induced voltage U*j is caused by the interaction of the flowing fluid and the imposed magnetic field.

[0129] A voltage Uj necessary for reconstructing the velocity profile is extracted from U*j at the processing means using appropriate signal processing techniques.

[0130] Because the applied magnetic field varies rapidly (it may be a sine wave or a hybrid square wave signal) U*j is also a rapidly varying signal. Typically the amplitude of U*j is only a few millivolts and so before being sampled by the analogue to digital converter in the processing means it must be amplified by a high gain differential amplifier (HGA) which has gain A , where A is typically equal to 1000. A voltage follower (VF) and a high pass filter (HPF) are used to condition the signals from the electrodes prior to being passed to HGA.

[0131] The high pass filters (HPF) are used to eliminate a very large dc offset which can appear on each electrode due to the effects of polarisation and accumulation of static charges. The output voltage Ux,j from the high gain amplifier (HGA) is fed to the processing means. However, despite the high pass filters (HPF), the differential voltage at the input to HGA consists of the sum of U*j and a residual, unwanted slowly varying dc voltage U0 due to the effects of polarisation and accumulation of static charges at the electrodes. U0 is generally larger than the amplitude of U*j. If the effects of U0 are not eliminated then the voltage Ux,j at point x in Figure 7 would be given by Ux.j =A(UO + U*j) and the resultant (slowly varying) dc component AU0 would make the value of Ux,j lie well outside of the range of operation of the analogue to digital converters on the processing means.

[0132] Accordingly, the processing means preferably comprises a control circuit to compensate for the effects of U0. Figure 7 depicts an example of a control circuit that is configured to compensate for the effects of U0 by applying a suitable offset (or reference) voltage to HGA at point y. In this way, the output voltage Ux,j is given by

where Usp is a 'set point' voltage provided by adjusting the 'set point adjust' circuit (SPA) in a once only operation. In many cases Usp will be chosen to have a value of 0V in which case the voltage Ux,j fed to the processing means will simply be given by



[0133] The control circuit depicted in Figure 7 comprises a low pass filter (LPF), a differential amplifier (DA), a set point adjustor (SPA) and an integrator (INT).

[0134] When the processing means comprises the coil excitation circuit, temperature compensation circuit, voltage measurement circuit and control circuit as depicted in Figures 5 and 7, the operation of the solid state relay network to control the current ic applied to the coils can be as follows:
  1. (i) When the SSRN is in position RP1 the maximum positive coil current ic,max flows in the Helmholtz coil resulting in a magnetic field of mean flux density +Bmax in the flow cross section, the positive sign indicating that the direction of the magnetic field is from Coil 2 to Coil 1 (Figure 5). This maximum coil current ic,max occurs at part s1 of the coil current cycle shown in Figure 6. With reference to Figure 7, the voltage Ux,j at the output from the jth 'voltage measurement and control' circuit, at part s1 of the coil current cycle, is denoted (Ux,j)1 where

    and where

    is the required 'positive' flow induced voltage between the jth electrode pair. The voltages (Ux,j)1 (where, for example, j = 1 to 15) are measured by the Analogue to Digital Converters (ADCs) of the processing means.
  2. (ii) When the SSRN is in position RP2 no coil current flows and so no magnetic field is present between Coil 1 and Coil 2. This corresponds to part s2 of the coil current cycle (Figure 6). The output voltage from the jth voltage measurement and control' circuit is denoted (Ux,j)2 where

    The voltages (Ux,j)2 are measured by the ADCs of the processing means.
  3. (iii) When the SSRN is in position RP3 the minimum coil current ic,min flows where ic,min = -ic,max. A magnetic field of mean flux density -Bmax now occurs between Coils 1 and 2 , the negative sign indicating that the direction of the magnetic field is from Coil 1 to Coil 2 (Figure 5). This corresponds to part s3 of the coil current cycle. The output from the jth 'voltage measurement and control' circuit is now (Ux,j)3 where

    where

    is the required 'negative' flow induced voltage between the jth electrode pair (and where

    if the flow velocity distribution has not changed significantly from s1 to s3). The voltages (Ux,j)3 are measured by the ADCs of the processing means.
  4. (iv) For part s4 of the coil current cycle the SSRN is again set to position RP2 so that no current flows in the coils. The output voltage from the jth 'voltage measurement and control' circuit is denoted (Ux,j)4 where

    The voltages (Ux,j)4 are measured by the ADCs of the processing means.
  5. (v) For a given coil current cycle, the jth potential difference measurement Uj required for the pixel velocity calculation is determined by the processing means using



[0135] From equations 19 and 20 it can be seen that Uj is given by



[0136] A summary of the relay positions, coil currents, magnetic flux densities and 'voltage measurement and control' circuit output voltages is given in the table below.
Relay PositionsTotal Coil CurrentMean Magnetic Flux DensityPart of coil current cycleUx,j
RP1 ic, max Bmax s1

RP2 0 0 s2 (Ux,j)2 = Usp
RP3 ic,min = -ic,max -Bmax s3

RP2 0 0 s4 (Ux,j)4 = Usp


[0137] The values of Uj given by equation 21 may be calculated for a single coil current cycle or they may be averaged using the processing means over G coil current cycles where G may take user specified values of, for example, 1, 2, 5 etc. The required value of G may be entered into the processing means software by the user via a touch screen display. Under steady state flow conditions, the larger the value of G the more accurate will be the values of Uj. Under transient flow conditions however the larger the value of G the slower will be the speed of response of the flow meter in calculating the time dependent flow velocities in the pixels.

[0138] During each coil current cycle (preferably once during each coil current cycle) the maximum voltage drop Ur,max across the reference resistance Rref is measured using the ADCs of the processing means. The maximum value Bmax of the mean magnetic flux density in the flow cross section is then determined by the processing means software, using equation 22 which was derived from equations 14 and 15



[0139] For the specific case where a single magnetic field projection is used, the mean magnetic flux density term B used in the calculation of the pixel velocities is set equal to Bmax in the processing means software.



[0140] Again for the specific case where a single magnetic field projection is used, the processing means software may now calculate the flow velocity vi in the ith pixel (i = 1 to 15, say) using

where v is the matrix containing the calculated pixel velocities vi, W is a matrix of the electromagnetic flow meter weight functions wij which are stored in the processing means software, A is a matrix of the pixel areas Ai which are stored in the processing means software and U is a matrix constructed from the measured potential differences Uj given by equation 20. The term a in equation 24 is the internal radius of the flow cross section of the flow meter body. [Note that for a single phase flow vi is the conducting fluid velocity in the ith pixel. For a multiphase flow vi is the velocity of the conducting continuous phase in that pixel].

[0141] The values of the pixel velocities vi may be displayed on the processing means display either graphically or 'alphanumerically'. In a single phase flow, and for a 15 pixel system, the total liquid flow rate Qc is calculated and displayed by the processing means using equation 25 below.



[0142] The processing means preferably operates continuously, updating the display of the pixel velocities vi and of Qc after every G coil current cycles

5. A First Example Of An Electromagnetic Flow Meter



[0143] The following description relates to an embodiment of an electromagnetic flow meter according to the present invention. The description describes how the weight function values of each pixel can be calculated for the flow meter. The description describes how the volumetric flow rate of a conducting single phase fluid (water) may be determined by the flow meter from the weight values and boundary voltage measurements. The description describes how the volumetric flow rate of a conducting single phase fluid having a uniform velocity profile across the cross-section of the flow tube may be determined using the flow meter. The description also describes how the volumetric flow rate of a conducting fluid having a non-uniform velocity profile across the cross-section of the flow tube may be determined using the flow meter.

5a. The Geometry Of The Electromagnetic Flow Meter



[0144] This particular flow meter consists of a PTFE (polytetrafluoroethylene) flow pipe (T) mounted within Helmholtz coils (C1, C2) . The flow meter contains 16 equispaced electrodes located at the plane z = 0. The inner diameter of the flow pipe is 0.08m, the outer diameter is 0.09m and its axial length is 0.3m. The inner and outer diameters of the two coils are 0.2048m and 0.2550m respectively. A cylindrical domain with a diameter of 0.32m and a length of 0.32m represents the boundary of the computing domain (refer to Figures 8a and 8b).

[0145] In order to measure the relevant potential differences electrode pairs are arranged at the internal boundary of the flow pipe (at the plane z = 0), these electrodes being in contact with the flowing medium. Sixteen electrodes are placed at angular intervals of 22.5 degrees on the flow pipe boundary (refer to Figure 9) the electrodes being denoted e1, e2 etc, with electrode e5 at the top of the flow cross section and electrode e13 at the bottom of the flow cross section. For this simple flow meter geometry the flow cross section is divided into seven pixels. The geometry of these seven pixels is chosen such that the chords joining seven pairs of electrodes are located at the geometric centres (in the y direction) of the pixels (refer to Figure 9). The fluid pixels are categorized as pixel 1 at the top of the flow cross section to pixel 7 at the bottom of the flow cross section. The pixel areas Ai are shown in the table depicted in Figure 10. Seven potential difference measurements can be made between the seven electrode pairs as shown in the table of Figure 10 (the jth potential difference measurement Uj was made between the jth electrode pair shown in the table of Figure 10 1). The local magnetic flux density B is perpendicular to both the flow direction and to the chords joining the electrode pairs.

5b. Magnetic Flux Density Distribution Of The Electromagnetic Flow Meter



[0146] In the embodiment of the flow meter a Helmholtz coil is used to produce a nearly uniform local magnetic flux density distribution.

[0147] The Helmholtz coil consists of two identical circular electromagnetic coils. In the flow meter design these coils are placed symmetrically on each side of the PTFE flow pipe as shown in Figure 8a and 8b. The system is designed such that the current flows through both coils in the same direction and each coil carries an equal amount of electric current giving rise to a relatively uniform magnetic flux density distribution in the flow cross section. The magnitude of the magnetic flux density in the y direction is relatively constant and has a maximum value of 7.757×10-4 T and minimum value of 8.044×10-4 T in the flow cross section. The mean value B of the magnitude of the y component of the magnetic flux density in the flow cross section is 7.996×10-4 T (7.996 gauss).

5c. Determining The Weight Function Values Of Each Pixel Of The Electromagnetic Flow Meter



[0148] The weight values wij of each pixel relate to the axial velocity vi in the ith pixel to the jth potential difference measurement Uj. The weight function values can be determined by simulating the flow meter using COMSOL Multiphysics software. As part of the simulation process, the flow channel is divided into seven pixels as described above (refer to Figure 9). The condition of the simulation is that the fluid in the pixel for which weight values are to be calculated is given a flow velocity of greater than zero in the z direction whilst the remaining pixels all have zero fluid velocity. Figures 11a and 11b show the distribution of the Lorentz forces and the induced electrical potentials when the fluid in pixel 4 has an imposed velocity in the z direction while the fluid in the remaining fluid pixels is at rest. Figure 11(a) illustrates the Lorentz force distribution arising from the imposed velocity in pixel 4. The magnetic field interacts with the charges carried in the water via these Lorentz forces causing the separation of charged ions (positive and negative) and giving rise to the electrical potential distribution shown in Figure 11(b). The arrows shown in Figure 11(a) also represent the direction of the local induced current density and it can be seen that for the (highly contrived) case in which flow occurs in pixel 4 only there is circulation of the electric current.

[0149] From the potential distribution given in Figure 11(b) the seven potential differences Uj between the 7 electrode pairs given in the Table of Figure 10 can be calculated allowing all of the weight values w4j associated with pixel 4 to be calculated according to equation 4 (with i = 4 and j = 1 to 7). The process is then repeated for each of the other six pixels in succession until all relevant 49 weight values have been calculated. Figure 12(a) shows the induced voltages plotted against electrode pairs for all of the seven simulations. [Note that the very large simulated pixel velocity of 500ms-1 was used to improve the accuracy of the weight values calculated using COMSOL]. Figure 12(b) shows the 49 weight values calculated from the induced voltages given in Figure 8(a) by using equation 26 .


5d. Determining The Velocity Profile Of A Conducting Fluid



[0150] Two different velocity profiles were investigated using the flow meter:- a uniform velocity distribution and a non-uniform (linear) velocity distribution as described below.

5e. The Uniform Velocity Profile



[0151] Figures 13a and 13b shows the effect of a uniform velocity distribution of 50ms-1 in the flow cross section on the Lorentz force distribution and the electrical potential distribution. The conducting fluid may be a single phase flow or the conducting continuous phase of a multiphase flow.

5f. A Non-Uniform Velocity Profile



[0152] Figures 14a and 14b shows the effect of a non-uniform velocity distribution in the flow cross section on the Lorentz force distribution and the electrical potential distribution. The flow velocity vz in the z direction is given by the expression

where y is the coordinate defined in figure 14a and 14b and a is the internal pipe radius. This results in v2 varying linearly from zero at y = -0.04m to 2ms-1 at y = 0.04m. This type of non-uniform velocity profile is non-axisymmetric and has a linear velocity distribution in the flow cross section. The non-uniform velocity profile can occur in single phase flows where flow is some how restricted (e.g. by a bend in the pipe or a partially opened valve.) Alternatively, this type of non-uniform velocity profile can occur in multiphase flows, for example, inclined multiphase flows.

[0153] The relevant induced voltages Uj for the uniform velocity profile and linear velocity profile were measured using the electrode pairs shown in the table of Figure 10. It should be noted that the electrical potential distribution for the uniform velocity profile and non-uniform velocity profile are entirely different from each other. For the uniform velocity distribution the induced voltages between pairs 1, 2 and 3 are the same as for pairs 7, 6, and 5 are respectively. For the non-uniform velocity profile the induced voltage between pair 1 is higher than that for pair 7. Similarly the induced voltages between pairs 2 and 3 are respectively higher than for pairs 6 and 5. Moreover, for the non-uniform velocity profile the highest induced potential is between electrode pair 3 while the maximum induced voltage for the uniform velocity profile is between electrode pair 4.

[0154] As mentioned earlier, the predetermined weight function values wij and measured induced voltages are used to determine the mean velocity vi in each pixel. For a conducting single phase fluid, the method can be expressed simply by the following matrix equation

in which V is a single column matrix containing the pixel velocities vi, W is a square matrix containing the relevant weight values wij, A is a square matrix containing information on the pixel areas Ai and U is a single column matrix containing the calculated potential differences Uj for a given imposed velocity profile.

[0155] The two velocity profiles of water as determined by the flow meter are shown in Figures 15(a) and 15(b). Also shown in Figures 15(a) and 15(b) are the imposed/reference velocity profiles that were measured using other detection means. The reference velocity profiles may be determined for example by using a laser Doppler anemometry device, hot wire anemometry device or a pitot-static tube.

[0156] With close inspection of Figure 15a and 15b it can be seen that the velocity profiles determined by the flow meter have excellent agreement with the reference velocity profiles for both the uniform and non-uniform velocity profiles. Figure 15(a) shows that for the uniform velocity profile the maximum (most overestimated) and minimum (most under estimated) errors occur in pixel 1 (+4.565%) and pixel 5 (-3.33%) respectively. The most accurate velocity is in pixel 4 with an error of only 0.722% . The non-uniform velocity profile has maximum and minimum errors in pixels 2 and 7 respectively. The most accurate velocities for the non-uniform velocity profile are in pixels 3 and 6 with errors of +0.912% and -0.797% respectively.

[0157] The total volumetric flow rate Qw of the water can be calculated from the determined velocity profile as follows;

in which Qw is the water volumetric flow rate, Ai is the area of the ith pixel, and vi is the axial velocity in the ith pixel. Let the true volumetric flow rate associated with the imposed (reference) uniform velocity profile be Qwiu and the volumetric flow rate associated with the calculated (determined) uniform velocity profile be Qwru. Also, let the true volumetric flow rate associated with the imposed (reference) non-uniform velocity profile is Qwil and the volumetric flow rate associated with the calculated (determined) non-uniform velocity profile is Qwrl.

[0158] For the uniform velocity profile Qwiu is calculated to be 2.509×10-1m3s-1 and Qwru is found to be 2.503×10-1m3s-1. There is thus an error of only -0.238% in the total volumetric flow rate obtained from the calculated uniform velocity profile.

[0159] For the non-uniform velocity profile Qwil is calculated to be 5.026×10-3m3s-1, and Qwrl is calculated to be 5.147×10-3m3s-1. There is thus an error of only +2.413% in the total volumetric flow rate obtained from the non-uniform velocity profile. It is believed that this error could be further reduced by improving the spatial resolution of the flow meter. This may be achieved by using a greater number of pixels. As discussed the flow tube cross-section may be divided into a greater number of pixels by increasing the number of electrodes and/or applying multiple magnetic flux density distributions.

6. A Second Example Of An Electromagnetic Flow Meter



[0160] The following description relates to a flow meter that is suitable for monitoring the flow of a two phase fluid. The fluid comprises a conducting continuous phase and a dispersed phase. The flow meter is suitable for determining the axial velocity profile of the conducting continuous phase as discussed above. The flow meter further comprises an impedance cross correlation measuring means (ICC) for measuring the distribution of the local velocity vd of the dispersed phase in the flow cross section. The ICC is also able to measure the distribution of the local volume fractions of the dispersed and continuous phase (αd and αc respectively) in the flow cross section. Thus, by combining the flow meter and the ICC the volumetric flow rates Qd and Qc of the dispersed and continuous phases respectively can be calculated as follows:



where A is the pipe cross sectional area.

6a. Measuring the Local Volume Fraction Measurement Using The ICC



[0161] The ICC measuring means consists of two arrays of electrodes denoted array 'Ã' and array 'B̃' spaced uniformly around the internal circumference of the pipe. One of these electrode arrays may, if required, be the same as the electrode array used in the flow meter for determining the axial flow velocity of the conducting continuous phase. is typically equal to 8 or 16. The axial separation L of the arrays is typically 50mm. The distributions of the local volume fractions αd and αc are measured using one of the arrays only (e.g. array B̃). A sequence / of electrical potentials is applied to the electrodes in array B̃, starting at electrode 'Elec1' (Figure 16), giving rise to an Electrode Potential Rotational Pattern (EPRP) denoted I1.

[0162] The potential applied to an electrode may be 'excitation' (denoted V+), 'virtual earth' (denoted ve) or 'true earth' (denoted E). For EPRP I1, by measuring the fluid resistance Rf between the V+ and ve electrodes (using an appropriate electronic circuit) the mean mixture conductivity σm in a small 'Effective Sensing Region' RI,1 of the flow cross section can be determined.

[0163] The 'Effective Sensing Region' RI,1 has a 'Centre of Action' denoted CI,1, with precisely defined coordinates. If the conductivity σw of the conducting continuous phase is known then the 'local' dispersed phase volume fraction αd in RI,1 (and hence at CI,1) can be determined from σm using the Maxwell relationship for mixtures of conducting materials



[0164] In equation a3 it is assumed that the dispersed phase has effectively zero conductivity such as would occur, for example, in an oil-in-water flow or an air-in-water flow. The value of αd thus calculated is taken as being the local dispersed phase volume fraction at the point CI,1. The local continuous phase volume fraction αc at CI,1 is given by



[0165] With reference to Figure 16 suppose that the electrical potential formerly applied to each electrode is now applied to the adjacent electrode in the clockwise direction (e.g. the potential formerly applied to electrode 'Elec1' is now applied to electrode 'Elec8', the potential formerly applied to electrode 'Elec2' is now applied to electrode 'Elec1', etc). The new EPRP is I2 and the new effective sensing region is RI,2 which has the same shape as RI,1 but which is rotated by

in the clockwise sense with respect to RI,1. The centre of action of RI,2 is denoted CI,2 and the local volume fractions αd and αc of the dispersed and continuous phases at CI,2 can be determined as described above. For an electrode system, by rotating the effective sensing region Ñ times, the values of αd and αc can be determined at the spatial locations CI,n (n = 1 to Ñ) in the flow cross section.

[0166] Let us now suppose that a different sequence of potentials denoted sequence II is used (e.g. ve, V+, ve, E, E, E, E, E). Using the methods outlined above this will enable αd and αc to be determined at an additional spatial locations CII,n (n = 1 to Ñ). The number of sequences of potentials can be increased until the values of αd and αc are known at the required number of spatial locations in the flow cross section. For example, if three sequences of potentials (I, II and III) are used the values of αd and αc will be known at the 3 locations CI,n, CII,n and CIII,n (n = 1 to ).

6b. Local Dispersed Phase Velocity Measurement using the ICC



[0167] In an array B̃ the instantaneous local dispersed phase volume fraction measured at point CI,1 at time t (using the methods outlined above) is denoted αd,(t). Let us suppose that at corresponding point CI,1 in array à (which is located upstream of array B̃) the instantaneous local dispersed phase volume fraction is αd,(t). A cross correlation function R(τ) can be defined as



[0168] The function R(τ) has a well defined peak at τ = τp representing the local transit time of particles of the dispersed phase between planes à and B̃ at the spatial location CI,1 in the flow cross section. The local dispersed phase velocity vd at CI,1 is then given by

where L is the axial separation of the electrode arrays. By repeating the process for points CI,n, CII,n and CIII,n (n = 1 to Ñ) (for example) the local dispersed phase velocity can be determined at the required number of spatial locations in the flow cross section.

[0169] Measurements of αd, αc and vd using the ICC device (as described above) are calculated using the same processing means as that used in the flow meter for determining the axial velocity profile.

[0170] It should be noted that the measurements that are made using the ICC device could alternatively be made using a high speed 'Dual-Plane Electrical Resistance Tomography' (DP-ERT) device.

7. A Third Example Of A Flow Meter



[0171] The following description relates to a flow meter that is suitable for monitoring the flow of each phase of a three phase fluid. The fluid comprises a conducting continuous phase and two dispersed phases. The flow meter further comprises an impedance cross correlation measuring means (ICC) and a density measuring means (DM) so as to create a three phase flow meter.

[0172] One specific (but not exclusive) embodiment of such a three phase flow meter would be for measuring the volumetric flow rates of oil, water and gas in a vertical pipeline, in which water is the continuous phase and where the oil and gas are dispersed phases. In this particular example, the density meter could simply consist of a vertical section of pipe (of typical length 1m) with pressure tapings separated by a vertical distance LDM. A differential pressure measurement ΔPDM made between the pressure tapings, compensated, for the effects of frictional pressure loss resulting from the motion of the multiphase mixture, enables the mean density ρm of the multiphase to be measured using

where is the acceleration of gravity. [Note that a number of alternative techniques are readily available to measure ρm]. The mean volume fraction λw of the water in the flow cross section can be determined using the ICC device described in the two phase flow meter using

where Ai is the area of the ith region (of N) into which the flow cross section is divided, A is the total pipe cross sectional area and

is the local volume fraction of the combined, non-conducting dispersed phases (oil and gas) in the ith region as measured by the ICC device. The mean volume fraction λo of the oil in the cross section can now be obtained using

where ρo, ρw and ρg respectively represent the densities of the oil, water and gas at the position of the three phase flow meter. It is necessary to calculate ρg using simple, auxiliary measurements of the absolute pressure and absolute temperature of the multiphase mixture at the position of the multiphase flow meter. The mean volume fraction λg of the gas in the cross section can now be obtained using



[0173] The volumetric flow rate Qw of the water can be obtained using

where vwIEF,i is the water velocity, as measured by the flow meter, in the ith region into which the flow cross section is divided. If it is now assumed that the oil and water are well mixed and flow homogeneously such that the mean oil velocity is the same as the mean water velocity then the mean oil velocity v0 can be calculated as follows;



[0174] The oil volumetric flow rate Qo can now be obtained using



[0175] The gas is not finely dispersed in the multiphase mixture in the way that the oil is finely dispersed in the water and so the mean gas velocity vg can be obtained as follows using the ICC device

where vgICC,i is the local gas velocity, in the ith region into which the flow cross section is divided, as measured by the ICC device using the cross correlation technique described in the two phase flow meter. The gas volumetric flow rate can now be obtained using



[0176] The features of the invention are applicable to all aspects of the invention and may be used in any combination.


Claims

1. An electromagnetic flow meter for monitoring the flow of a conducting continuous phase of a multiphase fluid comprising:

a flow tube (1) comprising a non-electrically conducting body with a flow tube cross-section and a circumference;

a means for generating a magnetic field (3a, 3b) across the flow tube cross-section so that a voltage is induced in the conducting fluid as it flows through the flow tube;

an array of voltage detection electrodes (10) arranged around the circumference of the flow tube configured to measure flow induced voltages at the boundary of the flow tube cross-section, the flow tube cross-section being divided into a plurality of pixel regions; and

processing means for determining an axial velocity profile of the conducting continuous phase of the multiphase fluid by calculating the local axial velocity of the conducting fluid in each of the pixel regions, wherein the processing means is configured to calculate the local axial velocity of the conducting continuous phase of the multiphase fluid in each pixel region using the measured induced voltages and predetermined weight functions for each pixel region,

characterized in that the electromagnetic flow meter further comprises a means for measuring a local concentration distribution of the conducting continuous phase of the multiphase fluid, and

wherein the electromagnetic flow meter is further characterised in that:

the processing means is configured to calculate the volumetric flow rate of the conducting continuous phase of the multiphase fluid using the local axial velocity of the conducting continuous phase of the multiphase fluid in each pixel region and the local concentration distribution of the conducting continuous phase of the multiphase fluid;

the means for generating a magnetic field is configured to generate a non-uniform magnetic field across the flow tube cross-section; and

the means for generating a magnetic field is further configured to successively generate multiple magnetic fields, each magnetic field having a different magnetic flux density distribution.


 
2. The flow meter of claim 1, wherein the flow tube (1) further comprises an annular liner having a conductivity that is generally the same as the conductivity of the conducting continuous phase of the multiphase fluid.
 
3. The flow meter of claim 1 or 2, wherein the means for generating a magnetic field (3a, 3b) is configured to generate a substantially uniform magnetic field across the flow tube cross-section.
 
4. The flow meter of any of claims 1 to 3, wherein the array of electrodes (10) comprises E electrodes, the flow cross-section is divided into N pixel regions and N induced voltages are measured on the boundary of the conducting continuous phase of the multiphase fluid when multiple magnetic flux density distributions are successively applied, wherein the number of magnetic flux density distributions generated is P, where N=PM and M = E-1.
 
5. The flow meter of claim 1, further comprising means for measuring the local concentration distribution of one or more dispersed phases of the multiphase fluid.
 
6. The flow meter of claim 1 or 5, wherein the means for measuring the local concentration distribution is configured to use an electrical resistance tomography technique or an impedance cross correlation technique.
 
7. The flow meter of any of claims 5 or 6, further comprising means for determining the mean density of the multiphase fluid and means for determining the density of each phase of a multiphase fluid.
 
8. The flow meter of claim 7 wherein the flow meter Is configured to determine the axial velocity profile and the volumetric flow rate of each phase of a multiphase fluid.
 
9. A method for monitoring the flow of a conducting fluid flowing in a flow tube (1) comprising a flow tube cross-section and a circumference, wherein said conducting fluid is the continuous phase of a multiphase fluid, the method comprising:

applying a non-uniform magnetic field across the conducting fluid;

successively applying multiple magnetic fields across the conducting fluid, whereby each magnetic field has a different magnetic flux density distribution;

generating an induced voltage in the conducting fluid;

measuring the induced voltages at an array of voltage detection electrodes (10) arranged around the circumference of the flow tube (1) on the boundary of the flow cross-section, the flow tube cross-section being divided into a plurality of pixel regions;

determining the axial velocity profile of the conducting fluid by calculating the local axial velocity of the conducting fluid in each pixel region using the measured induced voltages and predetermined weight functions, wherein said method further comprises determining the volumetric flow rate of the conducting fluid using the axial velocity profile and local concentration distribution of the conducting fluid, further comprising measuring the local concentration distribution of the conducting fluid and optionally measuring the local concentration distribution of one or more dispersed phases .


 
10. The method of claim 9, wherein measuring the local concentration distribution of the conducting fluid comprises using an electrical resistance tomography technique or an impedance cross correlation technique.
 
11. The method of claim 9 further comprising determining the axial velocity profile, and the volumetric flow rate, of each phase of the multiphase fluid.
 


Ansprüche

1. Elektromagnetischer Strömungsmesser zur Überwachung der Strömung einer leitenden kontinuierlichen Phase einer mehrphasigen Flüssigkeit, umfassend:

ein Strömungsrohr (1) umfassend einen nicht elektrisch leitenden Körper mit einem Strömungsrohrquerschnitt und einem Umfang;

ein Mittel zum Erzeugen eines Magnetfelds (3a, 3b) über den Strömungsrohrquerschnitt, sodass eine Spannung in der leitenden Flüssigkeit induziert wird, während sie durch das Strömungsrohr strömt;

ein Array von um den Umfang des Strömungsrohrs herum angeordneten Spannungserkennungselektroden (10), die dafür konfiguriert sind, strömungsinduzierte Spannungen an der Grenzfläche des Strömungsrohrquerschnitts zu messen, wobei der Strömungsrohrquerschnitt in eine Vielzahl von Pixelbereichen unterteilt ist; und

Verarbeitungsmittel zum Bestimmen eines Axialgeschwindigkeitsprofils der leitenden kontinuierlichen Phase der mehrphasigen Flüssigkeit durch Berechnen der lokalen Axialgeschwindigkeit der leitenden Flüssigkeit in jedem der Pixelbereiche, wobei das Verarbeitungsmittel dafür konfiguriert ist, die lokale Axialgeschwindigkeit der leitenden kontinuierlichen Phase der mehrphasigen Flüssigkeit in jedem Pixelbereich unter Verwendung der gemessenen induzierten Spannungen und vorbestimmten Gewichtsfunktionen für jeden Pixelbereich zu berechnen,

dadurch gekennzeichnet, dass der elektromagnetische Strömungsmesser ferner ein Mittel zum Messen einer lokalen Konzentrationsverteilung der leitenden kontinuierlichen Phase der mehrphasigen Flüssigkeit umfasst, und

wobei der elektromagnetische Strömungsmesser ferner dadurch gekennzeichnet ist, dass:

das Verarbeitungsmittel dafür konfiguriert ist, die volumetrische Strömungsrate der leitenden kontinuierlichen Phase der mehrphasigen Flüssigkeit unter Verwendung der lokalen Axialgeschwindigkeit der leitenden kontinuierlichen Phase der mehrphasigen Flüssigkeit in jedem Pixelbereich und der lokalen Konzentrationsverteilung der leitenden kontinuierlichen Phase der mehrphasigen Flüssigkeit zu berechnen;

das Mittel zum Erzeugen eines Magnetfelds dafür konfiguriert ist, ein nicht einheitliches Magnetfeld über den Strömungsrohrquerschnitt zu erzeugen;

das Mittel zum Erzeugen eines Magnetfelds ferner dafür konfiguriert ist, multiple Magnetfelder sukzessive zu erzeugen, wobei jedes Magnetfeld eine unterschiedliche magnetische Flussdichtenverteilung aufweist.


 
2. Strömungsmesser nach Anspruch 1, wobei das Strömungsrohr (1) ferner eine ringförmige Auskleidung mit einer Leitfähigkeit umfasst, die generell die gleiche wie die Leitfähigkeit der leitenden kontinuierlichen Phase der mehrphasigen Flüssigkeit ist.
 
3. Strömungsmesser nach Anspruch 1 oder 2, wobei das Mittel zum Erzeugen eines Magnetfelds (3a, 3b) dafür konfiguriert ist, ein im Wesentlichen einheitliches Magnetfeld über den Strömungsrohrquerschnitt zu erzeugen.
 
4. Strömungsmesser nach einem der Ansprüche 1 bis 3, wobei das Array von Elektroden (10) E Elektroden umfasst, der Strömungsquerschnitt in N Pixelbereiche unterteilt ist und N induzierte Spannungen an der Grenzfläche der leitenden kontinuierlichen Phase der mehrphasigen Flüssigkeit gemessen werden, wenn multiple magnetische Flussdichtenverteilungen sukzessive angewendet werden, wobei die Anzahl erzeugter magnetischer Flussdichtenverteilungen P ist, wobei N=PM und M = E-1.
 
5. Strömungsmesser nach Anspruch 1, ferner umfassend Mittel zum Messen der lokalen Konzentrationsverteilung von einer oder mehreren dispergierten Phasen der mehrphasigen Flüssigkeit.
 
6. Strömungsmesser nach Anspruch 1 oder 5, wobei das Mittel zum Messen der lokalen Konzentrationsverteilung dafür konfiguriert ist, eine elektrische Widerstandstomographietechnik oder eine Impedanz-Kreuzkorrelationstechnik anzuwenden.
 
7. Strömungsmesser nach einem der Ansprüche 5 oder 6, ferner umfassend Mittel zum Bestimmen der mittleren Dichte der mehrphasigen Flüssigkeit und Mittel zum Bestimmen der Dichte jeder Phase einer mehrphasigen Flüssigkeit.
 
8. Strömungsmesser nach Anspruch 7, wobei der Strömungsmesser dafür konfiguriert ist, das Axialgeschwindigkeitsprofil und die volumetrische Strömungsrate jeder Phase einer mehrphasigen Flüssigkeit zu bestimmen.
 
9. Verfahren zur Überwachung der Strömung einer leitenden Flüssigkeit, die in einem Strömungsrohr (1) umfassend einen Strömungsrohrquerschnitt und einen Umfang strömt, wobei die leitende Flüssigkeit die kontinuierliche Phase einer mehrphasigen Flüssigkeit ist, wobei das Verfahren umfasst:

Anlegen eines nicht einheitlichen Magnetfelds über die leitende Flüssigkeit;

sukzessives Anlegen multipler Magnetfelder über die leitende Flüssigkeit, wodurch jedes Magnetfeld eine unterschiedliche magnetische Flussdichtenverteilung aufweist;

Erzeugen einer induzierten Spannung in der leitenden Flüssigkeit;

Messen der induzierten Spannungen an einem Array von Spannungserkennungselektroden (10), die um den Umfang des Strömungsrohrs (1) herum an der Grenzfläche des Strömungsquerschnitts angeordnet sind, wobei der Strömungsrohrquerschnitt in eine Vielzahl von Pixelbereichen unterteilt ist;

Bestimmen des Axialgeschwindigkeitsprofils der leitenden Flüssigkeit durch Berechnen der lokalen Axialgeschwindigkeit der leitenden Flüssigkeit in jedem Pixelbereich unter Verwendung der gemessenen induzierten Spannungen und vorbestimmten Gewichtsfunktionen, wobei das Verfahren ferner das Bestimmen der volumetrischen Strömungsrate der leitenden Flüssigkeit unter Verwendung des Axialgeschwindigkeitsprofils und der lokalen Konzentrationsverteilung der leitenden Flüssigkeit umfasst, ferner umfassend das Messen der lokalen Konzentrationsverteilung der leitenden Flüssigkeit und optional das Messen der lokalen Konzentrationsverteilung von einer oder mehreren dispergierten Phasen.


 
10. Verfahren nach Anspruch 9, wobei das Messen der lokalen Konzentrationsverteilung der leitenden Flüssigkeit das Anwenden einer elektrischen Widerstandstomographietechnik oder einer Impedanz-Kreuzkorrelationstechnik umfasst.
 
11. Verfahren nach Anspruch 9, ferner umfassend das Bestimmen des Axialgeschwindigkeitsprofils und der volumetrischen Strömungsrate jeder Phase der mehrphasigen Flüssigkeit.
 


Revendications

1. Débitmètre électromagnétique pour surveiller l'écoulement d'une phase continue conductrice d'un fluide polyphasique comprenant :

un tube d'écoulement (1) comprenant un corps non conducteur électriquement avec une section transversale de tube d'écoulement et une circonférence ;

un moyen de génération d'un champ magnétique (3a, 3b) à travers la section transversale de tube d'écoulement de sorte qu'une tension soit induite dans le fluide conducteur lorsqu'il s'écoule à travers le tube d'écoulement ;

un réseau d'électrodes de détection de tension (10) disposées autour de la circonférence du tube d'écoulement configuré pour mesurer des tensions induites par l'écoulement à la limite de la section transversale de tube d'écoulement, la section transversale de tube d'écoulement étant divisée en une pluralité de régions de pixels ; et

un moyen de traitement pour déterminer un profil de vitesse axiale de la phase continue conductrice du fluide polyphasique en calculant la vitesse axiale locale du fluide conducteur dans chacune des régions de pixels, dans lequel le moyen de traitement est configuré pour calculer la vitesse axiale locale de la phase continue conductrice du fluide polyphasique dans chaque région de pixels en utilisant les tensions induites mesurées et des fonctions de poids prédéterminées pour chaque région de pixels,

caractérisé en ce que le débitmètre électromagnétique comprend en outre un moyen pour mesurer une distribution de concentration locale de la phase continue conductrice du fluide polyphasique, et

dans lequel le débitmètre électromagnétique est en outre caractérisé en ce que :

le moyen de traitement est configuré pour calculer le débit volumétrique de la phase continue conductrice du fluide polyphasique en utilisant la vitesse axiale locale de la phase continue conductrice du fluide polyphasique dans chaque région de pixels et la distribution de concentration locale de la phase continue conductrice du fluide polyphasique ;

le moyen de génération d'un champ magnétique est configuré pour générer un champ magnétique non uniforme à travers la section transversale de tube d'écoulement; et le moyen de génération d'un champ magnétique est en outre configuré pour générer successivement de multiples champs magnétiques, chaque champ magnétique ayant une distribution de densité de flux magnétique différente.


 
2. Débitmètre selon la revendication 1, dans lequel le tube d'écoulement (1) comprend en outre un revêtement annulaire ayant une conductivité qui est généralement la même que la conductivité de la phase continue conductrice du fluide polyphasique.
 
3. Débitmètre selon la revendication 1 ou 2, dans lequel le moyen de génération d'un champ magnétique (3a, 3b) est configuré pour générer un champ magnétique sensiblement uniforme à travers la section transversale de tube d'écoulement.
 
4. Débitmètre selon l'une quelconque des revendications 1 à 3, dans lequel le réseau d'électrodes (10) comprend E électrodes, la section transversale d'écoulement est divisée en N régions de pixels et N tensions induites sont mesurées sur la limite de la phase continue conductrice du fluide polyphasique lorsque de multiples distributions de densité de flux magnétique sont appliquées successivement, dans lequel le nombre de distributions de densité de flux magnétique générées est P, où N = PM et M = E-1.
 
5. Débitmètre selon la revendication 1, comprenant en outre un moyen de mesure de la distribution de concentration locale d'une ou plusieurs phases dispersées du fluide polyphasique.
 
6. Débitmètre selon la revendication 1 ou 5, dans lequel le moyen de mesure de la distribution de concentration locale est configuré pour utiliser une technique de tomographie par résistance électrique ou une technique de corrélation croisée d'impédance.
 
7. Débitmètre selon l'une quelconque des revendications 5 ou 6, comprenant en outre un moyen de détermination de la densité moyenne du fluide polyphasique et un moyen de détermination de la densité de chaque phase d'un fluide polyphasique.
 
8. Débitmètre selon la revendication 7, dans lequel le débitmètre est configuré pour déterminer le profil de vitesse axiale et le débit volumétrique de chaque phase d'un fluide polyphasique.
 
9. Procédé de surveillance de l'écoulement d'un fluide conducteur s'écoulant dans un tube d'écoulement (1) comprenant une section transversale de tube d'écoulement et une circonférence, dans lequel ledit fluide conducteur est la phase continue d'un fluide polyphasique, le procédé comprenant :

l'application d'un champ magnétique non uniforme à travers le fluide conducteur; l'application successive de multiples champs magnétiques à travers le fluide conducteur, moyennant quoi chaque champ magnétique a une distribution de densité de flux magnétique différente ;

la génération d'une tension induite dans le fluide conducteur ; la mesure des tensions induites à un réseau d'électrodes de détection de tension (10) disposées autour de la circonférence du tube d'écoulement (1) sur la limite de la section transversale d'écoulement, la section transversale de tube d'écoulement étant divisée en une pluralité de régions de pixels ;

la détermination du profil de vitesse axiale du fluide conducteur en calculant la vitesse axiale locale du fluide conducteur dans chaque région de pixels en utilisant les tensions induites mesurées et des fonctions de poids prédéterminées, dans lequel ledit procédé comprend en outre la détermination du débit volumétrique du fluide conducteur en utilisant le profil de vitesse axiale et la distribution de concentration locale du fluide conducteur, comprenant en outre la mesure de la distribution de concentration locale du fluide conducteur et éventuellement la mesure de la distribution de concentration locale d'une ou plusieurs phases dispersées.


 
10. Procédé selon la revendication 9, dans lequel la mesure de la distribution de concentration locale du fluide conducteur comprend l'utilisation d'une technique de tomographie à résistance électrique ou d'une technique de corrélation croisée d'impédance.
 
11. Procédé selon la revendication 9, comprenant en outre la détermination du profil de vitesse axiale et du débit volumétrique de chaque phase du fluide polyphasique.
 




Drawing





















































Cited references

REFERENCES CITED IN THE DESCRIPTION



This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.

Non-patent literature cited in the description