GB2265009A - Flow measurement in medium of non-uniform electrical resistance - Google Patents

Abstract

The rate of flow of an electrically conductive non-uniform fluid is determined by cross-correlation of the resistance of the fluid measured at two spaced points 21 in the direction of flow. The method is particularly suitable for use in measuring the rate of flow from oil wells which produce an oil-in-water suspension with a high dissolved-salt content and replaces the correlation technique using capacitive measurements used on water-in-oil dispersions. Two spaced pairs 23 of conductivity electrodes are mounted on elongate former 22 and the former is mounted centrally in an oil well. Each electrode is preferably 5-20 mm in length with the electrodes of each pair spaced by 5-10 mm and the pairs spaced by 20-200 mm. Diameter of the former is 20-70 mm and the electrodes are preferably of titanium or plasticised titanium. The resistance is preferably measured using an AC voltage of 10KHz. <IMAGE>

Description

Flow Measurement This invention relates to flow measurement, and concerns in particular measurement of the rate of flow along a channel of a multiphase liquid the continuous phase of which is electrically conductive. More specifically still, the invention concerns the flow of the oil-in-water mixture commonly encountered in the bores of some oil wells.

Once completed, an oil well normally has a concrete-lined elongate borehole with 7 inch (17.5 cm) diameter tubular casing disposed therein. Formation fluids and treatment fluids pass through holes in the lining/casing and up the core of the borehole via whatever valves, sampling chambers and other mechanisms may be contained therein. One such mechanism may be a device for measuring the flow rate of the formation fluids up the borehole.

After a well has been drilled, lined and cased, and is producing, it may be desirable to measure, and log (record), the rate at which fluid is flowing out of the geological formations through which the bore has been drilled and is passing into and up the casing.

Moreover, because this fluid will often be issuing at two or more levels from the formations through which the bore passes, and thus can have a corresponding number of compositions and flow rates, it is desirable to be able to measure each flow rate in situ rather than as the combined rate when the fluids all reach the top of the casing as a mixture of the individual fluids. For this reason, then1 there is a need for a flow measuring device which can be lowered as part of a production logging tool string down the bore hole to a chosen depth, where there issues the fluid of interest.

As might be expected, one simple type of flow measurer is a propeller (or fan) acting as a turbine and driven round by the fluid passing "through" it. Such devices, which, coupled to a suitable generator, provide a flow-rate-dependent output signal (usually electrical in nature) that can be fed to metering equipment either included as part of the downhole measurer package or sited at the well head, are in use, and in many circumstances work well. However, they do have their limitations, and one of those is when required to measure the flow rate of multiphase fluids passing along a bore hole that is inclined rather than vertical. The problem can be explained as follows.

The formation fluids commonly encountered in oil wells are usually two- or three-phase fluids, being for the most part a mixture of water and oil or of water, gas and oil. One of the liquid phases - the water or oil - will be the continuous phase (new wells most often have oil as the continuous phase, whilst in older wells it is more often the water), and the other, and the gas (if present), will be the dispersed phase, being in the form of a myriad of bubbles or droplets, with quite a wide size spectrum, carried along by the continuous phase.Now, in a vertical bore hole all of the components of the fluid in the casing are carried along at much the same rate (and there is only a relatively minor velocity gradient across the casing), but in an inclined hole, whether the angle off the vertical be small (as 5 or 10 ) or large (as 700 or 800), this is not the case. Instead, because the components are of different densities - oil and gas are lighter than water - the multiphase fluid begins to separate out into two (or more) distinct streams, with the lighter oil and gas lying above the heavier water.Moreover, these streams move along the casing with significantly different velocities, so that a substantial velocity gradient can be set up across the bore, possibly even such as to cause an actual backs low in those regions immediately adjacent the underside of the casing (and because this situation is not unstable, there are effectively formed "waves" at the interface between the two streams, and these waves also travel along the string). The net result is that a turbine/propeller blade driven by the moving fluid experiences contradictory forces across its diameter, and the output of a flow measuring system relying on such a device becomes untrustworthy.

For this reason there has been developed a quite different variety of flow measuring device, namely one which measures the time taken for the multiphase fluid to travel between two spaced locations along the casing.

More specifically, this device in effect "measures" at each of the two spaced locations along the path of the fluid the dielectric nature of the fluid (at each point by using the fluid as the dielectric between the two plates of a capacitor, and "measuring" the resulting capacitance), and by comparing the capacitances for a portion of fluid seen at the first point at one time and then at the second point at a later time there may be calculated the speed - and thus the flow rate - of the fluid between the two.Identification of a particular portion of fluid is possible because the non-uniform nature of the fluid - its existence as a number of bubbles/droplets of one material randomly dispersed in another - means that it has a different dielectric constant at different positions along it, and because although the random dispersion itself also changes with time nevertheless over a short time, and thus over a short distance, it remains much the same, so that a short section of fluid measured at one location can actually be "recognised" by its dielectric "profile" as it passes the second location a second or so later.

Such a flow rate measuring method, and a logging tool using it, is the subject of our British Patent Application No: 2,227,841A (US Patent No: 4,975,645).

The idea of comparing the capacitance profile of the fluid passing one site with that seen as it passes a second site some time thereafter - the mathematical technique employed is known as "cross correlation", and a suitably-programmed computer or dedicated wave analyser is commonly utilised to find the best possible fit of the two profiles - works well in practice, but nevertheless has limits to its application.

Specifically, for this capacitance detection to work it is necessary that the fluid involved be a dielectric that is, a non-conductor, or insulator, of electricity.

Now, where it is to be applied to an oil well in which the formation fluid is in the form of a water-in-oil dispersion, with oil as the continuous phase, that is satisfactory, because oil is a non-conductor, as is required. However, in many wells, and especially in older wells, the formation fluid is instead obtained as an oil-in-water dispersion, with water as the continuous phase. Moreover, the water will commonly have a fairly high dissolved salt (electrolyte) content, and as a result is a good conductor of electricity - from which it follows that it has quite unsuitable properties as regards the use of capacitance detection flow rate measuring systems. The problem, then, is to find some other physical parameter that can be employed to provide data for the otherwise very satisfactory cross correlation technique.And the solution is to "measure" not the fluid's capacitance but instead its resistance, for that too will vary randomly - because of the random mixture of oil droplets dispersed in the water - to give a profile as the fluid passes one location that will hardly have degraded, and so can be recognised, as a short time later the same portion of fluid passes the second one spaced therefrom. By knowing the spacing of the two locations, and by performing the cross correlation to find a match of the two profiles - and by so finding the time taken for the fluid to travel from the first location to the second - there may be determined the speed of the fluid, and thus the fluid's flow rate, up the bore.The present invention proposes, accordingly, what is in essence a resistance version of the capacitance cross correlation flow rate method of the earlier invention, using the fluid's resistance rather than its dielectric nature. The invention also proposes apparatus - a logging tool - for accomplishing this method, the apparatus being two pairs of spaced electrodes themselves spaced along a mount (included within the tool string lowered down the bore hole) disposable within and along the bore and operatively connectable to means for measuring and recording the resistance of the fluid passing them, and for performing the desired cross correlation and subsequent flow rate calculation.

In one aspect, therefore, the invention provides a method of measuring the speed, and thus flow rate, of a fluid of non-uniform electrical resistance travelling along a channel between two locations spaced a known distance one from the other, in which method at each location there is "measured" the resistance of the fluid passing the location, to provide a time-dependent "profile" of the resistance of the fluid, and the two profiles are compared one with the other to find a time-dependent match indicative of the time taken for the fluid to travel between the locations, and thus of its speed therebetween and so of its flow rate.

In a second aspect, the invention provides apparatus for carrying out the method of the invention, and so measuring the speed, and thus flow rate, of a fluid of non-uniform electrical resistance travelling along a channel between two locations spaced a known distance one from the other, which apparatus comprises, operatively connected one to the next: for disposition at each location in contact with the fluid, electrical contact means through which can be detected the resistance of the fluid passing the location; detection means for detecting the resistance of the fluid passing each contact means; storage means for storing the resistance data provided by the detection means; and correlation means for using the stored data to provide two time-dependent "profiles" of the resistance of the fluid as it passes first one and then the other contact means, and for comparing the two profiles one with the other to find a time-dependent match indicative of the time taken for the fluid to travel between the locations, and thus of its speed therebetween and so of its flow rate.

The invention provides a method of measuring the speed, and thus flow rate, of a fluid travelling along a channel. The channel will in practice be the casing within the bore hole of an oil well, although in principle the method could find use in other circumstances, with a different type of channel - and, indeed, for different varieties of fluid.

The inventive method actually provides a measure of the speed of the fluid as it travels between the two points. However, by knowing the effective size of the channel - for oil well casing, the available crosssectional area within the casing - there can be calculated the volume flow rate, as required.

As explained above, the fluid is non-uniform in its resistive (and other) properties because it is composed of a random mix of oil (and possibly gas) droplets/ bubbles dispersed in an aqueous continuous phase. In addition, the fluid may very likely be turbulent, with quite large eddies formed and re-formed within it as it passes along the channel. These eddies add significantly to the fluid's non-uniform nature, and in fact it will be primarily the eddies which provide the matchable characteristics of the resistance profile.

The structure of any short length of fluid changes with time, but remains sufficiently constant over a short time to allow that length to be identified from its resistance profile as it reaches the second location spaced a short distance from the first. By a "short time" there is meant a period such that the dispersed structure of the portion of fluid, and thus its resistance profile, does not change significantly as it traverses at the relevant speed the distance between the two measurement locations, and so its resistance profile can be recognised at the second location.In any particular case a suitable short time can be found by simple experimentation; in the case of the formation fluid in an oil well borehole, however, by a "short time" there is generally meant a period of the order of a second or so - and preferably no more than a second (and advantageously not less than a twentieth of a second). Accordingly, for a fluid (such as the formation fluid passing into the casing of an oil well) travelling at a speed up to about 1.5 m/s, a suitable spacing for the two locations is around 100 mm. A range of possible spacings is from 20 to 200 mm; very small distances not only give poor time resolution but may cause undesirable cross-talk between the detectors, whilst large distances may make it difficult to find the match (because the profile has changed so much in the time taken).

As the fluid passes each location so its resistance is measured - continuously - to provide the desired time-dependent resistance profile. The measurement, which, as explained in more detail below, is achieved by providing electrical contacts in the fluid, and using some form of conventional resistance "meter" to measure the resistance through the fluid between them, does not need to be accurate (in an absolute sense), nor does it need to measure the resistance of all, or any particular part of, the fluid passing the contacts. All that is required is that there should be obtained at each location self-consistent figures for the relative resistance of the same portion of fluid; that will provide profiles that can meaningfully be correlated.

Using the preferred contact ring pairs discussed below, the resistance measured is in each location effectively that of the annular ring of fluid between the contact rings and barely more than a centimetre or so thick Once there has been obtained at each location the time-dependent resistance data - the resistance profile - the two can be compared - that is, cross correlated to find the match, and this match being time-spaced provides the time taken for the fluid to travel between the two locations. Though undoubtedly there are other ways of doing it, the storage of the data, and the cross correlation operation, is most conveniently handled by a suitably programmed computer.Most modern microcomputers are so fast that they are perfectly capable of doing this in real time - that is to say, as and while the data is being gathered - but it may be desirable to effect the correlation itself in a dedicated wave analyser.

In its second aspect the invention provides apparatus for carrying out the method of the invention, this apparatus comprising, operatively connected together, electrical contact means through which can be detected the resistance of the fluid passing the location, detection means for detecting that resistance, storage means for storing the resistance data provided by the detection means, and correlation means for using the stored data to find the required time factor, and so enable a calculation of the flow rate.

In essence the contact means are little more than conductive terminals or plates disposed suitably spaced in physical contact with the fluid. In practice, and for use in the casing of an oil well, they are conveniently short tubes - rings or collars - mounted flush with the exterior surface of a non-conducting rod member itself mountable in use within and preferably co-axially of the casing (so the fluid passes through the annular volume formed between the rod member and the inner walls of the casing; if the rod member is slightly off-axis this does not seem deleteriously to affect the results).The rod member - advantageously made of a relatively inert substance capable of withstanding the temperatures encountered downhole (up to 175"C), possibly a ceramic or a high melting point plastic, and held in place~centrally of the casing by suitable spacer elements or centralisers - is conveniently around 45 mm in diameter (a size in the range 20 to 70 mm will be acceptable), whilst each ring contact means is conveniently around 10 mm in length (though rings of from 5 to 20 mm would be suitable).

Whilst it might be possible to have a different arrangement - with a contact means common to both locations, for example - it is much preferred to employ a pair of contact means terminals or plates at each location. In the preferred electronic resistance measuring system described below, one such contact plate is used as the excitation electrode, while the other is a virtual earth electrode. The spacing of the two contact plates of each pair, which determines the portion of fluid the resistance of which is to be measured, should advantageously be small, else the definition of the system becomes poor, and the obtained profiles become less individual and identifiable (the peaks and troughs are smoothed out, making it much more difficult to find a match with confidence).A typical spacing is around 7 mm, though it depends to some extent on the size of the channel (and for most common oil well casings a spacing of from 5 to 10 mm should be satisfactory).

The contact means can be made of any suitable conductive material. A steel of some sort is possible, but the conditions down an oil well bore hole are rather fierce, and something somewhat more corrosion-resistant, such as titanium (possibly "platinised"), is preferred.

There are contact means at each location, so in the rod-member-mounted system they are spaced along the rod member a suitable distance. As discussed above, that distance is whatever is long enough to provide adequate time resolution and to avoid cross talk (and provide a reasonable profile) and short enough to ensure the profile stays matchable - from around 20 to 200 mm, preferably 100 mm.

The means for actually detecting/measuring the resistance of the fluid as it passes the contact means may be any suitable. For example, the contacts could be made the resistance component in one arm of a Wheatstones bridge. However, a particularly preferred system drives an alternating current through the fluid via the contact means, and then uses the output to bias an operational amplifier and a subsequent series of electronic filters, amplifiers, inverters and rectifiers to provide an analogue signal representative of the resistance, which signal can then be digitised (by an A-D converter) and stored in computer memory for subsequent manipulation.

The storage and the correlation means, and the subsequent means for calculating the actual flow rate given the known spacing and channel size and the determined time factor, is most conveniently a dedicated wave analyser, though a general purpose microcomputer suitably programmed, as intimated above, could be used and no more need be said about that here.

The method and apparatus of the invention has successfully provided useful oil-in-water flow rate information in pipe-like channels inclined at angles from 0 (vertical), through 150, 30 , 45" and 60 , and for flow rates varying from 1, through 20, 30 and 40, to 50 m3/hr.

A resistive, cross correlation logging tool made according to the invention is of a simple design, and relatively inexpensive to manufacture. It can be given a robust construction capable of withstanding the hostile environmental conditions encountered downhole.

Furthermore, it is usable as a continuous logging device, being logged up and down the well, whilst making velocity estimates, in the same way that the presentlyused spinner is utilised.

Various embodiments of the invention are now described, though by way of illustration only, with reference to the accompanying diagrammatic Drawings in which: Figure 1 shows an oil rig and a number of bores drilled therefrom down into the subterranean strata; Figure 2 shows a perspective view of the contact means arrangement used in the invention; Figure 3 shows in side view a length of oil well casing carrying a contact means arrangement as shown in Figure 2; Figures 4A & B show in block schematic form the electronic circuitry used with the Figure 2 contact means to provide a measurement of the resistance of the fluid flowing therepast; and Figure 5 shows two typical, matching resistance profiles obtained using the equipment of Figure 4.

Figure 1, which is not to scale, shows an oil rig (generally 10) on the surface above an oil-bearing stratum (11) many thousands of feet beneath it and sandwiched between other strata. From the rig have been drilled three different bore holes - one which is vertical (12), one which gradually deviates to the left as viewed (13), and one (14) which starts by gradually deviating to the right as viewed, and which then turns sharply to run nearly horizontal (at 15) to the right through the stratum 11.In that stratum, which may be several hundred feet thick, the vertical bore hole maintains, of course, a zero degrees angle off the vertical, the angled hole 13 is around 20 from the vertical at the top and around 30 at the bottom, while the nearly horizontal hole 14 is at 85" to the vertical in its portion 15.

Figures 2, 3 and 4A/B relate to an experimental cross correlation Production Logging tool designed and built for making velocity estimates in oil/water flows where water forms the continuous phase. The tool (generally 30 in Figure 3) consists of identical upper and lower contact means pairs, or flow sensors (generally 21u, 211), separated by an axial distance of 100 mm, mounted on a former (22) of 1.69 in (4.29 cm) outside diameter. Each flow sensor 21u, 211 comprises an excitation electrode (23ex) and a virtual earth electrode (23ve) which are of short circular cross section tubular structure and which fit flush with the outer surface of the former 22.

The tool is shown in place (in Figure 3) in the casing (31) of an oil well, located centrally thereof by centralisers (32). To provide some flow conditioning, an extension piece (33) matching the former 22 is fitted to the upstream end (and positioned at the middle of the working section using centralisers 32).

Variations in the resistance of the fluid (represented by arrows 24) in the shallow sleeve-like volume (that volume indicated in Figure 2 by the dashed lines 25) between each pair of excitation and virtual earth electrodes 23ex, 23ve, brought about by time dependent changes in the oil volume fraction in the vicinity of the flow sensors, are measured using the circuitry illustrated in Figures 4A/B and described below. These variations are then cross correlated to allow flow velocity estimates to be made.

On each flow sensor, the axial length of each tubular electrode 23ex, 23ve is 10 mm, whilst the gap between the two electrodes is 7 mm (see Figure 2). It is readily apparent that the sensing volume associated with this electrode configuration is relatively small, and would not extend out to the well casing.However this is not considered to be a disadvantage for the reasons given hereinbefore, namely that the tool preferentially measures the velocity of the large eddy structures that are present in deviated multiphase flows, which eddies generally extend across the entire pipe cross section and so are detected even though the sensing volume is relatively small, and small sensing volumes do not give rise, as do large ones, to spatial filtering effects which substantially reduce the frequency bandwidth of the signals produced by the flow sensors, resulting in a broadening of the peak of the resultant cross correlation function and so leading to reduced accuracy in the estimates of transit time.

For each flow sensor, the fluctuating resistance Rx between the excitation and virtual earth electrodes is measured using the circuitry shown in Figure 4A. As can be seen from the Figure, Rx forms the feedback resistor of a standard inverting amplifier circuit. A 10 kHz sinusoidal voltage V1 of amplitude Al was applied to the input of this inverting amplifier. The output voltage V2 from this stage was a 10 kHz sinusoidal signal of amplitude A2 where Rx A2 = -- Al R1 The voltage V2 is fed, via a non-inverting amplifier, into a demodulator circuit consisting of a precision diode rectifier and a low pass filter with a cut-off' frequency of 400 Hz.Thus, the output voltage V3 from the demodulator stage is a d.c. signal the amplitude of which is proportional to the instantaneous value of the resistance between the excitation and virtual earth electrodes of the corresponding flow sensor.

A circuit similar to that shown in Figure 4A is constructed for each of the two flow sensors, and the final output voltages V3 from both circuits are cross correlated using the Figure 4B circuit to yield flow velocity information.

Figure 5 shows graphically the correlation of two resistance profiles, the "solid line" one taken at time t at the upstream sensor and the lower one taken at the same time but (on correlation) showing a "match" at what is in effect a later time t+6t (in the Figure 6t is noted as 0.2 secs, so it has taken 0.2 secs for the portion of liquid having the upstream sensor solid line profile to travel into the same spatial relationship with the downstream sensor). The resistance data for each sensor from a significant time length of measurement has been stored, and then the two sets of data - forming the resistance profile - have been compared one with another and, notionally, moved along one another - this is actually carried out using a mathematical cross correlation technique - until there is a significant fit of one with the other. The correlation records the times to which each profile belongs, and from the "position" of the fit can thus calculate the time difference between the two, and so the time taken by the profile-represented portion of fluid to travel from the upstream sensor to the downstream one. With this speed information, and knowing the cross-sectional area of the pipeway between the two, there may be calculated the fluid's volume flow rate.

Claims (22)

1. A method of measuring the speed, and thus flow rate, of a fluid of non-uniform electrical resistance travelling along a channel between two locations spaced a known distance one from the other, in which method at each location there is "measured" the resistance of the fluid passing the location, to provide a time-dependent "profile" of the resistance of the fluid, and the two profiles are compared one with the other to find a time-dependent match indicative of the time taken for the fluid to travel between the locations, and thus of its speed therebetween and so of its flow rate.

2. A method as claimed in Claim 1, in which the channel is the casing within the bore hole of an oil well.

3. A method as claimed in either of the preceding Claims, in which the spacing for the two locations is from 20 to 200 mm.

4. A method as claimed in Claim 3, in which the spacing is around 100 mm.

5. A method as claimed in any of the preceding Claims, in which the two sets of time-dependent resistance data - the resistance profile - are cross correlated using a suitably-programmed computer and/or dedicated wave analyser.

6. A method as claimed in any of the preceding Claims and substantially as described hereinbefore.

7. Apparatus for carrying out the method claimed in any of the preceding Claims, and so measuring the speed, and thus flow rate, of a fluid of non-uniform electrical resistance travelling along a channel between two locations spaced a known distance one from the other, which apparatus comprises, operatively connected one to the next: for disposition at each location in contact with the fluid, electrical contact means through which can be detected the resistance of the fluid passing the location; detection means for detecting the resistance of the fluid passing each contact means; storage means for storing the resistance data provided by the detection means; and correlation means for using the stored data to provide two time-dependent "profiles" of the resistance of the fluid as it passes first one and then the other contact means, and for comparing the two profiles one with the other to find a time-dependent match indicative of the time taken for the fluid to travel between the locations, and thus of its speed therebetween and so of its flow rate.

8. Apparatus as claimed in Claim 7, wherein for use in the casing of an oil well the contact means are short tubes - rings or collars - mounted flush with the exterior surface of a non-conducting rod member itself mountable in use within and substantially co-axially of the casing.

9. Apparatus as claimed in Claim 8, wherein the rod member is held in place centrally of the casing by suitable spacer elements or centralisers.

10. Apparatus as claimed in either of Claims 8 and 9, wherein the rod member is from 20 to 70 mm diameter.

11. Apparatus as claimed in Claim 10, wherein the rod member is 45 mm diameter.

12. Apparatus as claimed in any of Claims 8 to 11, wherein each ring contact means is from 5 to 20 mm in length.

13. Apparatus as claimed in Claim 12, wherein each ring contact means is around 10 mm in length.

14. Apparatus as claimed in any of Claims 7 to 13, wherein there is employed a pair of contact means terminals or plates at each location.

15. Apparatus as claimed in Claim 14, wherein the spacing of the two contact plates of each pair is from 5 to 10 mm.

16. Apparatus as claimed in Claim 15, wherein the contact plate spacing is around 7 mm.

17. Apparatus as claimed in any of Claims 7 to 16, wherein the contact means are made of titanium or platinised titanium.

18. Apparatus as claimed in any of Claims 7 to 17, wherein the contact means locations are from 20 to 200 mm apart.

19. Apparatus as claimed in Claim 18, wherein the contact means locations are around 100 mm apart.

20. Apparatus as claimed in any of Claims 7 to 19, wherein the means for actually detecting/measuring the resistance of the fluid as it passes the contact means is one which drives an alternating current through the fluid via the contact means, and then uses the output to bias an operational amplifier and a subsequent series of electronic filters, amplifiers, inverters and rectifiers to provide an analogue signal representative of the resistance, which signal can then be digitised (by an A-D converter) and stored in computer memory for subsequent manipulation.

21. Apparatus as claimed in any of Claims 7 to 20, wherein the storage and the correlation means, and the subsequent means for calculating the actual flow rate given the known spacing and channel size and the determined time factor, is a dedicated wave analyser and/or a general purpose microcomputer suitably programmed.

22. Apparatus as claimed in any of Claims 7 to 21 and substantially as described hereinbefore.