GB2292613A - Multiple electrode electromagnetic flowmeters - Google Patents

Abstract

An electromagnetic flowmeter comprises a flow tube, means such as a pair of permanent magnets for producing a magnetic field transverse to the flow of fluid through the tube and at least three electrodes 31, 32, 33 mounted in the wall of the tube. Potential differences between selected pairs of electrodes are sampled repetitively to provide an indication of flow rate within the tube. Also disclosed are flowmeters having more than three electrodes (figures 4, 5, 7, 8, 10) and further sets of electrodes upstream and downstream of the magnetic field source (figure 6). <IMAGE>

Description

MAGNETIC FLOWMETERS This invention relates to magnetic flowmeters in which the magnetic field is provided by permanent magnets or electromagnets.

The basic concept of. the magnetic flowmeter, identified by Michael Faraday in 1831, involves the measurement of the small potential developed between two electrodes positioned in a flowing conductive fluid perpendicular to a magnetic field as well as to the direction of flow. However, streaming potentials, polarisation and other electrochemical effects at the interface between the electrodes and the flowing fluid inhibited the development of flowmeters based on this concept for more than a century.

In the mid-1950's, electromagnetic flowmeters became available in which the effects of the spurious signals at the electrodes were largely overcome by using an electromagnet excited at mains frequencies, and extracting the flow signal from the ac component of the potential developed between the electrodes by the alternating magnetic field.

The principal difficulty with these systems was the lack of stability and interference from the mains, and this led to the development of systems in which the excitation of the magnetic field is switched repetitively from zero (1), to a steady value (2), which is maintained briefly, and then switched back to zero (1), as shown in Figure 1. The potential developed between the electrodes as a result of the interaction between the fluid flowing and the magnetic field is sampled during the periods (1) and (2). The former provides the reference potential and is subtracted from the latter to provide the flow signal.

Sampling of the electrode potential during the switching transient (3) is suppressed. The power for exciting the electromagnet is usually derived from the mains and switching of the field is synchronised with the mains frequency (4) so that the effect of interference from this source is minimised.

There are many variations of this method for exciting the electromagnet and recovering the flow signal from the potential developed between the electrodes due to the interaction of the flowing fluid and the magnetic field.

For all these measurements, it is customary to refer the electrode potentials to the mean potential of the flowing fluid. This is usually obtained by clamping circular metal plates (5) each with a central hole, the diameter of which is the same as the bore of the flowtube, between the adjacent pipework at both the inlet and outlet of the flowtube, as shown in Figure 2. Various other systems of electrodes are also used for the same purpose.

It is advantageous to mount the electrodes which sense the mean potential of the flowing fluid in the same transverse plane as the flow sensing electrodes, but in line with the magnetic field, as shown in Figure 3. With such an arrangement, the magnetic field does not induce a potential between these electrodes which is due to the interaction between the flowing fluid and the magnetic field. Therefore the fluid flow rate can be determined by subtracting the signal developed between this pair of electrodes from the signal developed between the pair of electrodes positioned perpendicular to the magnetic field.

The power required to energise the electromagnet in this way may be as much as 10 watts and, although this would seldom present a problem in most industrial sites, there are many individual applications where it is not feasible to provide this amount of power. One example is the metering in the distribution of domestic water supplies where it would be quite impractical to provide the necessary power at all the measurement points, which are likely to be at remote stations. Consequently, their usefulness for this purpose is limited in spite of the advantages which they offer from an operational viewpoint, namely an unobstructed passage through the meter and consequential low head loss.

According to one aspect of the present invention there is provided a magnetic flowmeter for determining the rate of flow of a liquid, the flowmeter comprising a flowtube, means for producing a magnetic field transverse to the liquid flow and at least three electrodes mounted in the wall of the flowtube and spaced (preferably substantially equidistantly around the periphery) so as to lie within said magnetic field and means for sampling repetitively potential differences between selected pairs of electrodes, whereby analysis of the sampled signals is indicative of flow rate within the flow tube.

Figure 4 represents a flowmeter in accordance with the invention. As shown, in the plane central to the magnetic field and perpendicular to the axis of the flowtube, a pair of electrodes (10.10) are mounted through the wall of the tube with their axis in line with the magnetic field. A second pair of electrodes (11.11) is mounted similarly through the wall of the tube but with their axis perpendicular to the magnetic field. The potential developed between the pair of electrodes (10.10) is due to the interaction of the flowing fluid and the electrodes themselves and as their axis is in line with the magnetic field there is no component due to the latter. The potential due to the interaction of the flowing fluid with the magnetic field plus the component due to the interaction between the flowing fluid and the electrodes is developed between the pair of electrodes (11.11).

Figure 5 provides further constructional details of the flowmeter incorporating a flowtube of circular cross section (7). The permanent magnets (6) are flat and circular, with central holes (12) which are aligned with the axis of the pair of electrodes (10.10). A further pair of electrodes (11.11) is located in the same plane but with the axis perpendicular to that of the electrodes (10.10). A soft iron body (13) completes the magnetic circuit.

In operation, the potential developed between the pair of electrodes (10.10) is principally determined by the interaction between the flowing fluid and the electrodes themselves, but it is also influenced by their size, shape and materials of construction. if they are perfectly aligned with the magnetic field, there is component due to an interaction between the flowing fluid and the magnetic field is zero.

The potential developed between the pair of electrodes (11.11) which are located perpendicular to the magnetic field, is determined by the same influences as electrodes (10.10) but, as their axis is perpendicular to both the magnetic field and the direction of fluid flow, there is a component which is proportional to the fluid flow rate.

If all the electrodes in positions (10.10) and (11.11) are the same shape and size, and made of the same material, the difference fV(f)) between the potential developed between the pair of electrodes (11.11) {(Vn+Vf)} and that developed between the pair of electrodes (10.10) {V(n)} is due to the interaction between the flowing fluid and the magnetic field and is proportional to the velocity of the flowing fluid.

In practice, V(n) is likely to be appreciably greater than V(f) and it is therefore important to achieve the best possible accuracy for all the measurements. It is also important to appreciate that the source impedance of the potentials developed at the electrodes in most practical applications is likely to be high - several megohms.

The uncertainty associated with the measurement of V(n) can be reduced by mounting four further pairs of electrodes through the wall of the flowtube, as shown in figure 6. Two pairs (14.14) and (15.15) are mounted upstream of the electrodes (10,10) and (11.11) in a plane perpendicular to the axis of the flowtube and at a distance where they are essentially outside the influence of the magnetic field. The pair (14.14) is located in the same axial plane as pair (10.10) and the pair (15.15) is located in the same axial plane as pair (11.11). The two remaining pairs of electrodes (16,16) and (17.17) are mounted the corresponding positions downstream of electrodes (10.10) and (11.11).

The value of V(n) is derived from the average value of the potential between the pairs of electrodes (10.10) (14.14) (15.15) (16.16) and (17.17). If an individual value differs significantly from the average value, it should be discarded and a warning given of the need for maintenance or corrective action. Other information regarding the status of both the measurement system and the associated process plant can be obtained by statistical, spectral and time series analyses of the measured potentials.

The uncertainty associated with the measurement of V(f) can be reduced by replacing the pair of electrodes (11.11) with two pairs of electrodes (18.18) and (19.19) in the same plane as electrodes (10.10) but with their axes positioned as shown in Figure 7. Again, the flow signal V(f) is derived from the average value of the two signals. if there is a significant difference between them, a warning should be given so that corrective action can be taken.

The description of the concept so far is based on a circular pipe and a system of small 'point' electrodes.

it is equally applicable for flowtubes which are square or rectangular, and to electrode systems of different shapes and sizes to that shown in Figure 4.

From the viewpoint of the interaction between the magnetic field and the flowing fluid, a flowtube with a rectangular cross section (9) has the advantage that, when compared with a circular cross section (7), it is possible to provide a more intense magnetic flux density and the system is less sensitive to the effect of distorted flow velocity profiles. However, a flowtube with a circular cross section is relatively stronger and introduces a smaller pressure drop. Also, it is probably easier to manufacture.

The use of large electrodes, whether they are circular, rectangular or some other shape has advantages in so far that they reduce the source impedance of the flow signal V(f). However, it is difficult to avoid leakage at the point where they are mounted in the wall of the flowtube, and the calibration of the flowmeter is likely to be affected when they become contaminated, because the contamination may displace the position of the effective axis of the pair of such electrodes (e.g.

(11,11)) with respect to the magnetic field, thereby changing the flow induced potential. For flowtubes with a rectangular or square cross section, line electrodes are preferable to point electrodes.

An alternative approach to the determination of the potential induced in a flowing fluid due to its interaction with a stationary magnetic field is to sample, repetitively and in sequence, the potential developed between individual pairs of electrodes arranged around the periphery of a flowtube. This is illustrated in Figure 8 which shows an arrangement based on a flowtube (20) having a circular cross section, a transverse stationary magnetic field created by the permanent magnets (21a.21b), electrodes (22) and (23) positioned with their axis perpendicular to the axis of the flowtube and in line with the magnetic field, and electrodes (24) and (25) positioned in the same plane but perpendicular to the magnetic field.

Assuming that: 1) the potential at electrode (22) is taken as a datum 2) the potential between any pair of electrodes is V(n) when there is no contribution due to the interaction between the flowing fluid and the magnetic field 3) the potential V(f) is developed between the electrodes (24) and (25) due to the interaction between the magnetic field and the flowing fluid 4) the potential developed between the electrodes (23) and (24), (23) and (25), (22) and (24), and (22) and (25) is V(f)/2 because the flowing fluid only cuts half the magnetic flux 5) the potential at (24) is negative with respect to the potential at electrode (25) 6) the fluid flow rate is constant during the period when the electrode potentials are sampled 7) the fluid velocity and the magnetic field are uniform across the section then the potential difference between the various pairs of electrodes is as follows: Sequence Electrode Pair Potential Difference 1 22/24 V(n) - V(f)/2 2 22/23 V(n) 3 22/25 V(n) + V(f)/2 4 24/23 V(n) + V(f)/2 5 24/25 V(n) + V(f) 6 24/22 V(n) + V(f)/2 7 23/25 V(n) + V(f)/2 8 23/22 V(n) 9 23/24 V(n) - V(f)/2 10 25/22 V(n) - V(f)/2 11 25/24 V(n) - V(f) 12 25/23 V(n) - V(f)/2 This time series is shown in Figure 9.The flow rate can be determined by the application of conventional signal analysis methods (including those developed for tomography) to the changes in the amplitude of successive sets of the (twelve) sampled electrode potential differences. Changes in the magnitude of the mean value as well as in the signals from individual electrodes provide a means for monitoring the operational status of both the flow measurement system and the associated process plant.

It is evident that the concept can be applied to a greater number of electrodes arranged around the perimeter of the flowtube, as shown in Figure 10 where the array incorporates five equally spaced electrodes (26) (27) (28) (29) and (30). However, in this case the number of sampled potentials during one complete sequence would increase from twelve to twenty, but analysis of the signals would provide even more detailed information regarding the status of the flow measurement system and the associated process plant.

It is interesting to note that, in the positions shown in Figure 10, none of the axes of the pairs of electrodes is in line with the magnetic field and therefore the reference potential can only be inferred from the measurements. It follows that this may provide a method for obtaining information regarding the velocity profile of the flowing fluid.

Because of the difficulties in securing the electrodes in the wall of the flowtube in such a manner that leakage of the flowing fluid does not occur, there are obvious advantages to be gained by reducing the number of electrodes to a minimum. Figure 11 shows a flowtube with three electrodes (31) (32) and (33) located in the same plane perpendicular to the axis of the flowtube and spaced equidistant around the wall. The potentials developed between the electrodes is sampled and processed in a similar manner to that described previously, leading to the time series being as shown in Figure 12.

Making the same assumptions as cited previously, the potential differences developed during a complete sampling sequence are substantially as follows: Sequence Electrode Potential Number Pair Difference 1 31/32 V(n)+0.47CV(f)) 2 31/33 V(n)-0.47CV(f)) 3 32/33 v(n)-V(f) 4 32/31 V(n)-0.47(V(f)) 5 33/31 V(n)+0.47CV(f)) 6 33/32 V(n)+V(f) The main invention aspects of my flowmeter are summarised as follows: 1. A magnetic flowmeter in which the flowtube is circular, the transverse magnetic field is provided by a permanent magnet and there are two pairs of point electrodes mounted through the wall of the flowtube in a plane perpendicular to the axis of the flowtube.The axis of one pair of electrodes is in line with the transverse magnetic field and the axis of the second pair is perpendicular to that of the former.

The flow signal is obtained by subtracting the potential developed between the pair of electrodes, the axis of which is in line with the magnetic flux, from the potential developed between the second pair of electrodes, the axis of which is perpendicular to both the magnetic flux and the direction of the fluid flow.

The flow rate is determined by multiplying the difference between these two potentials by a factor which is a constant for each flowmeter and which depends on the cross sectional area of the flowtube, the velocity profile of the flowing fluid, the shape of the magnetic field, and to a lesser extent on the shape and size of the electrodes as well as the interaction between their materials of construction and the flowing fluid.

2. A magnetic flowmeter as described above except that the pair of electrodes located perpendicular to the magnetic field are replaced by two pairs of electrodes, the axes of which are set at a small angle on either side of the axis of the original single pair of electrodes.

The average of the potential developed between these two pairs of electrodes is used in determining the flow rate.

3. A magnetic flowmeter as described above in which the mean potential between the individual flow sensing electrodes and the mean potential of the reference electrodes is monitored continuously to provide warning of an incipient fault.

4. A magnetic flowmeter in which additional electrodes are mounted through the wall of the flowtube and positioned either upstream or downstream (or both) of the plane of the flow sensing electrodes, but also outside the influence of the magnetic field. The average value of the potential sensed at these electrodes is used to enhance the quality of the reference potential.

5. A magnetic flowmeter as described previously in which the cross section of the flowtube is square or rectangular.

6. A magnetic flowtube as described previously in which circular, square, rectangular or line electrodes are used in place of point electrodes.

7. A magnetic flowmeter in which the cross section of the flowtube is circular and the transverse magnetic field is provided by a permanent magnet. Three electrodes are mounted through the wall of the flowtube in a plane centred in the magnetic field, perpendicular to the axis of the flowtube and spaced (preferably equally) around the periphery.

The potential developed between the individual pairs of electrodes as a result of the interaction between the flowing fluid and the magnetic field is measured in a repetitive sequence. The fluid flow rate is determined by applying established signal analysis methods to the sampled signals, whilst changes in the relative values of the individual potential differences between the electrodes provide warning of electrode contamination and a potentially unreliable measurement.

It should be noted that this method does not require either the measurement or determination of the fluid reference potential.

8. A magnetic flowmeter as described in 7 above in which there are four or more electrodes mounted through the wall of the flowtube in a plane perpendicular to the axis of the flowtube, and spaced (preferably equally) around the periphery. As before, the potential developed between the individual pairs of electrodes as a result of the interaction between the flowing fluid and the magnetic field is measured in a repetitive sequence. The fluid flow rate is determined by established signal analysis methods from alternating component of the sampled signals, whilst changes in the individual potential differences between the electrodes provide warning of electrode contamination and a potentially unreliable measurement.

As before, it should be noted that this method does not require either the measurement or identification of the fluid reference potential.

9. A magnetic flowmeter in which the flowtube is circular and the electrodes are regarded as point or large and either circular or rectangular in shape.

10. A magnetic flowmeter in which the flowtube is square or rectangular and the electrodes are regarded as either point, large, circular, line or rectangular.

11. The method and apparatus described herein is useful in principle for measuring flow rates of any conductive liquids. It is of more general use in the measurement of flow rates in fluids having. a conductivity of at least about 5 microsiemens. However, in theory, the invention would be applicable to measuring flow rates in liquids such as hydrocarbons, having a conductivity lower than 1 microsiemen. The main application currently envisaged is in measuring or monitoring flow rates in water supply pipes in which conductivities typically range from about 50 to about 400 microsiemens.

Claims (5)

CLAIMS:

1. A magnetic flowmeter for determining the rate of flow of a liquid, the flowmeter comprising a flowtube, means for producing a magnetic field transverse to the liquid flow and at least three electrodes mounted in the wall of the flowtube and spaced (preferably substantially equidistantly around the periphery so as to lie within said magnetic field and means for sampling repetitively potential differences between selected pairs of electrodes, whereby analysis of the sampled signals is indicative of flow rate within the flow tube.

2. A flowmeter as claimed in claim 1 wherein the electrodes are located in substantially the same transverse plane.

3. A flowmeter as claimed in claim 1 or 2 wherein the means for producing the magnetic field comprises at least one permanent magnet.

4. A magnetic flowmeter for determining the rate of flow of a liquid, the flowmeter comprising a flowtube, a permanent magnet for producing a magnetic field transversely of the liquid flow and at least two pairs of electrodes wherein a first pair of electrodes are aligned with the magnetic field and a second pair of electrodes lie on an axis which makes an angle between 0 and 900 with the magnetic field, the flow rate being related to the difference in the potentials developed between the two pairs of electrodes.

5. A flowmeter as claimed in claim 4 wherein the second pair of electrodes lie on an axis at right angles to said first pair.