GB2501165A - Interface detection using a vertical array of time domain reflectometry sensors - Google Patents

Abstract

The invention provides a method of and/or apparatus for detecting one or more interfaces in a fluid mixture. A vertical array of TDR sensor components are arranged within the mixture and each sensor is switched to a TDR transmit/receive circuit. Signals from each sensor in the array are then processed to identify the location of the interface(s). A time domain reflectometry (TDR) sensor is claimed comprising a carrier having a curved outer surface, and a pair of substantially parallel conductors mounted on said curved outer surface, said conductors being separated by insulation. An apparatus for determining an interface in a composition is claimed, said apparatus including: a plurality of TDR sensors mounted at known vertical spacings which, in use, is placed within said composition; a controller for activating each of said sensors individually; and a processing facility operable to identify, from signals from said sensors, a location of said interface, said apparatus being characterised in that each sensor includes a pair of substantially parallel conductors separated by insulation, said conductors being mounted on a curved outer surface of a carrier.

Description

IMPROVEMENTS IN OR RELATING TO INTERFACE DETECTION

Field of the Invention

This invention relates to interface detection and, more particularly, to a method of, and/or apparatus for, interface detection based on time domain reflectometry (rDR).

Background

There is a continuing need to accurately identify interfaces between different materials.

By way of example, crude oil separation on o±Thhore platforms is a necessary step in the oil production. Effective separation requires monitoring the interfaces between gas -foam, foam -oil, oil -emulsion, emulsion -water, and water -solids. Current methods and apparatus used to effect separation are based on the difference in density and viscosity of water and oil but may also include technologies that rely on differences in rclativc pcmilttivity, spccd of sound, and rclativc pcrmcability of air, watcr and oil.

Techniques to detect interfaces between materials include those based on ultrasonics, TDR, capacitive sensors, magnetic sensors, Gamma Ray sensors and RF intensity sensors. The dif&rent types of sensor can be broadly grouped into two categories: continuous measurement which includes ultrasonics and TDR, and discrete measurement which includes capacitive sensors, magnetic sensors and Gamma Ray sensors. Each has its own advantages and drawbacks.

Although ultrasound detectors are able to determine interface position continuously, the accuracy of determination may be limited because the ultrasound signal will not readily pass through foam, the speed of sound in different media is difficult to specify accurately, and the ultrasound signal is significantly attenuated at thc interface. As a consequence, the accuracy of the level indication may, at best, be uncertain.

TDR generally offers more accurate performance than ultrasound because the speed of the TDR pulse is not sensitive to temperature. 1('there is a distinct interface between a layer of oil and a layer of water, a conventional TDR probe can continuously provide a satisfactory indication of the oil-water interface. However there is often a dynamic layer of emulsion between the oil and water layers and, when the thickness of the oil layer is less than 25mm, the reflected echo of the TDR probe is significantly reduced and the resulting accuracy is not sufficient. Further, a conventional TDR probe is incapable of indicating the emulsion concentration change with tank depth.

In contrast to continuous level measurement, discrete techniques only provide level indications at specific, given, positions but any indication of change is generally reliable and can be accepted with a higher degree of confidence than is the case with continuous techniques. A further advantage is that discrete level sensors can be mounted in both top-down and bottom-up alignments in contrast to ultrasound sensors (bottom-up) and TDR (top down). Howcvcr capacitivc sensors suffer interference from environmental temperature changes, magnetic sensors are not able to determine the interfice between oil and gas, and Gamma Ray sensors are both expensive and have the potential to leak radiation leading to a safety concern.

In general, continuous level measurement techniques are more suitable for stable environments in which the interface is clearly defined whereas discrete level measurement techniques are more suited for use in complex, multi-interface environments.

As stated above, TDR is a mature and reliable technology fbr continuous level measurement but a conventional TDR probe has the following shortcomings: i) Significant signal strcngth is lost whcn thc thickness of the emulsion layer exceeds 25mm ii) To operate effectively, the oil layer must be at least 100mm in thickness iii) It has difficulty determining both foam and liquid levels reliably; and iv) It has difficulty determining emulsion levels.

v) It must be mounted at the top of a tank so as to extend down towards the bottom of thc tank.

US Patent 4,786,857 describes methods and apparatus for determining one or more interfaces between different fluids in a multiphase fluid system, particularly a multiphase fluid system forming part of a nuclear reactor. A number of TDR probes are spaced vertically within the multiphase fluid system and each is interrogated in turn to identify the locations at which a change of fluid phase occurs.

In systems such as nuclear reactors where the interfaces are water!steamiair, TDR on its own provides unsatisfactory results and it is therefore necessary, in addition, to measure both temperature and pressure and to integrate these measures into the TDR time traces.

This adds complexity and cost that would render such a solution impractical for many common industrial level sensing applications. Further, the sensors described in US 4,786,857 are unsuited for most commonly encountered industrial level sensing applications because the open co-axial arrangement of conductors will encourage shorting. Further, if used with high viscosity liquids, liquid residues will collect in the slots as well as between the outer tube wall and central conductor, leading to misbehaviour of the TDR signals. Ultimately, blockage of the slots could prevent a true interface level being present within the sensor.

It is an object of the invention to provide an interface detection method and apparatus which will go at least some way to addressing the aforementioned drawbacks; or which will at least provide a novel and useful choice.

Siuninary of the Invention Accordingly the invention provides apparatus for determining an interface in a composition, said apparatus including a plurality of TDR sensors mounted at known vertical spacings which, in use, is placed within said composition; a controller for actiyating each of said sensors individually; and a processing facility operable to identiI, from signals from said sensors, a location of said intcrfacc, said apparatus being characterised in that each sensor includes a pair of substantially parallel conductors separated by insulation, said conductors being mounted on a curved outer surface of a carrier such that the conductors lie substantially flush with said outcr surface.

Prcferably said carrier comprises a tube.

Preferably said TDR sensors are mounted in a common mount, there being a given separation between said sensors.

Preferably said controller is configured to activate adjacent sensors in sequence Preferably said processing facility includes a single TDR circuit placed in communication with each of said sensors by said controller.

Preferably said pmcessing facility is configured and operable to generate a visual indication of the location of said interface.

In a second aspect the invention pmvides a TDR sensor including a carrier having a curved outer surface, and a pair of substantially parallel conductors mounted on said curved outer surface such that the conductors lie substantially flush with said outer surface and are separated by insulation.

Preferably said conductors are substantially flush with said curved outer surface.

Preferably said carrier is tubular.

Said conductors may be tbllow the curvature of the outer surface of said carrier or may be aligned substantially perpendicular to the curvature of said outer surface.

Many variations in the way the invention may be perlbrmed will present themselves to those skilled in the art, upon reading the fbllowing description. The description should not be regarded as limiting but rather as an illustration, only, of one manner of perfomiing the invention. Where appropriate any element or component should be taken as including any or all equivalents thereof whether or not specifically mentioned.

BriefDescripdon of the Drawings One preferred method of; and apparatus for, reducing the present invention to practice will now be described with reference to the accompanying drawings in which: Figure 1: shows an isometric view of a sensor component for usc in a multi-level detection unit according to the invention; Figure 2: shows an elevational view of an alternative form of sensor coniponen 1; Figure 3: shows a schematic view of the sensor of Figure 1 incorporated into apparatus according to the invention; Figure 4: shows an operating circuit block diagram for use in the inventiow Figure 5: shows a schematic view of an interface detection system according to the invention in place in a tank;

Detailed Description of Working Embodiment

Referring firstly to the Figure 3, the invention provides a method of and/or apparatus for determining one or more interfaces in a fluid composition 5. In the particular embodiment depicted, the fluid composition comprises a layer 6 of air above a layer 7 of oil which, in turn, is above a body of water 8. In many industrial applications it is necessary or desirable to be able to locate the interface 9 between the oil and water layers; and the interface 10 between the air and oil layers. The present invention describes the use of a plurality of TDR sensor components 12 arranged in a vertical array within the body of fluid, the sensor units being arranged at known vertical spacings. The substantially identical sensor components 12 are switched individually, preferably in sequence, to a TDR drive/processing facility 13 which processes the signals from the individual sensing components. Given that the time interval between the initial TDR reference signal and signals reflected from the ends of the arcuate conductors (described below) will vary with the change of fluid in contact with the conductors, any significant change of time interval can be interpreted as indicating that adjacent sensor components are on opposite sides of a fluid interface. In this way the locations of both interfaces 9 and 10 can be determined.

Whilst two sensor components 12 arc shown in Figure 3, it is envisaged that a multiplicity of such components will be provided in a stack or vertically spaced on a common mount 14. The precise number and spacing of sensing components are not essential features of the invention and will depend on the number of interfaces being monitored as well as the desired accuracy with which the locations of those interfaces are to be located. In the example shown in Figure 3 there would need to be at least one further sensing component in what is shown as the oil layer 7 to allow the interface 9 to be detected.

The individual sensor components 12 are preferably switched in a linear sequence by multiplex controller 15 into a TDR circuit 16 starting with the lowermost unit and then moving up to the unit immcdiatcly thcrc-abovc. Data from cach scnsing componcnt 12 is processed by signal processor 17 and, when a change in signal is observed between adjacent sensing components 12, that is taken as indicating that those adjacent sensing components arc lying in different strata of the fluid mixture 5 and that an interface is located between those sensing components. The signal processor may also be programmed to present data from all the sensing components into a display such as a bar graph or other visual indication of where the interface(s) lie.

Tuning now to Figures 1 & 2, the sensor components must be configured to generate a clearly definable reflected signal from very compact physical structure. Hence specialist configurations have been developed. A particular problem that needs to be addressed is that a first reflection signal is generated from that end of the TDR sensor or connected to the drive/processing circuitty (the first end'). Because the length of the sensor is short, it can be extremely difficult to discriminate between this initial reflection signal generated at the first end and that the signal reflected from the other end of the sensor (the second end'). It is known that an input impedance at the first end of about 50 ohms suitably minimises the first reflection.

It will be appreciated that typical TDR sensors take the form of a co-axial cable but such a configuration is not suitable for the applications described herein. We have found that sensor components comprising two parallel conductors embedded in an insulating material, and driven simultaneously by the drive/processing facility 13, exhibit an easily controlled input impedance (and thus low initial reflection signal) together with a strong reflected signal. As a consequence the invention can operate effectively in difficult environments.

The conductors and the insulation in which they are embedded are mounted on a carrier having a curved outer surface. The conductors are preferably configured so that the outer surfaces thereof preferably lie substantially flush with the outer surface of the carrier and thus the conductors, themselves, may be curved.

A first form of sensing component 12 is shown in Figure 1 and comprises a round tubular stainless steel carrier body 20, 100 mm in diameter, having an areuate cavity 21 milled or otherwise formed in the outer surface thereof; although not through the full wall thickness of the body 20. The cavity is preferably occupies 1800 of the circumference of the steel body. Located within the cavity are two spaced arcuate conductors 22 which, in this particular embodiment are formed into arcs from 4mm x 4mm square section stainless steel stock. The conductors extend, in parallel, around the curved surface of the sensor body. In the form shown the conductors are held in parallel in a block 23 of PVC which fills the cavity 21 and locates the conductors 22 at a separation of about 6 mm. The lengths of the conductors are set so that each conductor occupies about 160° of the circumference of the body 20.

The PVC block 23 not only holds the conductors 22 at the desired spacing but also insulates the conductors 22 from one another and from the steel body 20. Coaxial cables 24 are provided, preferably passing down through the centre of the body 20 and connect through the wall of the body to adjacent, first, ends of the conductors 22 to allow transmission and receipt of the TDR signal for each component 12.

Referring now to Figure 2, another form of TDR sensor component 30 is shown which is perhaps more suitable for those applications in which there is a limited size of bore through which the sensing apparatus can be inserted into the fluid housing. In this form two stainless steel conductors 32 are mounted in parallel within an arcuate block 31 of PVC so as to lie substantially perpendicular to the curvature of the body. In cross-section (not shown) the arcuate block of PVC is semi-circular and is fitted into a cavity machined into a bar or tube 33 of stainless steel. Again, the PVC insulates the conductors 32 from one another as well as from the bar or tube 33. By way of example only, the conductors may be 120mm long, 10 mm wide and 3 mm thick.

The bar or tube may be 220mm in length and 50mm in diameter.

As alternatives to the use of PVC described above, other appropriate insulating materials such as, for example, PEEK or ceramics may be used and in yet a further alternative, the entire bar or tube could be formed from a suitable plastics or ceramic material with cavities formed therein into which the conductors could be cast. Further, the outside surfaces of the conductors 22 and 32 may be coated with a thin layer of Teflon or other insulating material to prevent condensation leading to short circuiting between the conductor pairs.

In both examples of sensor shown and described, the conductors are of uniform section and spacing. As alternatives the spacing between the conductors and/or the sections of the conductors could be varied along the lengths thereof to achieve enhanced impedance matching at the transmission ends of the conductors and enhanced reflection at the opposite ends of the conductors.

Whether in the form shown in Figure 1 or that shown in Figure 2, sensor components may be stacked vertically or installed on a long tube at a given separation between adjacent sensor components. With all sensors being connected to multiplex controller 15, a TDR signal is sent out and received from each sensor in turn. When all of the sensors have been scanned and the signals passed to signal processor 17, the location of the interface between different fluids can be determined and shown in display 18.

The multiplex controller 15 may simply comprise an array of ultra-wideband receivers where one receiver is provided for each of the TDR sensor components 12 or 30. Each of the receivers, which may for example be of the form described in US Patent 5,345,471, is switched in sequence to the respective sensor component.

Any form ofTDR circuit 16 deemed appropriate can be used in the performance of the invention, onc simplified version of a suitablc circuit being shown in Figurc 4. Rcfcrring to Figure 4, an oscillator 40 generates the time reference which is fed to microcontroller 41 to effect equivalent time sampling. The technique proposed herein is the dual ramping technique described in US Patent 3,010,071 (Carlson). The microcontroller 41 generates two control pulses, a short interval pulse and a long interval pulse. The short interval pulse is fed to a fast ramp generator 42 which produces a short, steep waveform whilst the long interval pulse is fed to a slow ramp generator 43 which produces a staircase waveform. The two wave forms arc fed to a comparator 44 which controls the function of the delayed receiving gate 45. The receiving gate 45 generates the receiving pulse using a combination of step recovery diodes and a fast logic switch.

The microcontroller 41 also provides a short interval pulse to transmit pulse generator 46 which, as with the receive pulse generator 45, generates the transmit pulse using a combination of step recovery diodes and a fast logic switch.

Transmit and receive signals are both applied to an array of fill diode bridge or semi-diode bridge couplers 47 which generates a signal representing the time interval between the transmit (reference) signal, and the reflected signal. The full diode bridge dc-coupler is described in greater detail in US Patent 3,597,633 Hirano) and the semi-diode bridge dc-coupler bridge is described in US Patent 5,345,471. The output of the dc-coupler 47 is amplified at 48 and then subjected to signal processing at 49. In particular the signal processing step involves comparing the measured time differences between two adjacent sensors, with given thresholds which are pre-defined according to the lengths of the conducting elements 22 or 32, to establish the existence of a change in medium between the sensors.

In use, TDR transmit signals are simultaneously applied to the first ends of the conducting elements 22 or 32 in any single sensor component and the time interval bctwccn thc initial TDR rcfcrcncc signal and the signals rdflcctcd from the sccond cnds of the conductors, is measured. A comparison between this time interval and the interval recorded from the previous sensor component will indicate whether or not the adjacent sensor components are in contact with different media and, accordingly, whether they lie on different sides of an interface.

Sensor components as described above were tested in different media and exhibited the time intervals summarised below, the Horizontal figures being obtained using a sensor component 20 while the Vertical figures were obtained using a sensor component 30.

Medium Air Sugar Oil Oil/Water Oil/Water Oil/Water Water (4:1) (1:1) (1:4) (+vcpcak) Time Interval Qis) 143 155 157 178 216 250 39 (Horizontal) Time Interval (jis) 144 154 158 224 280 348 36 (Vcrtical) Turning now to Figure 5, alternative modes of mounting sensor assemblies are shown so as to determine interface locations in a tank 50 containing a layer 51 of sand, a layer 52 of water, a layer 53 of oil/water emulsion, a layer 54 of oil, a thin layer 55 of foam above the oil and a body 56 of air above the foam layer 55. A combination of TDR assemblies according to the invention may be provided to determine the various interface locations.

In the form shown TDR sensor assembly 60a is mounted vertically and in a location suitable for determining the air/foam and foam/oil interfaces. TDR assemblies 60b and 60e are mounted obliquely and thus provide better vertical resolution using the same number of individual sensing components in contact with the contrasting media and thus providing a more accurate location of interface between foantloil and oil/emulsion. TDR assembly 60d is again mounted vertically to determine interface locations between sand/water and water/emulsion which, using a conventional TDR sensor, is difficult to determine.

Claims (11)

1. Cia juts 1. Apparatus for determining an interface in a composition, said apparatus including a plurality of TDR scnsors mounted at known vertical spacings which, in usc, is placed within said composition; a controller for activating each of said sensors individually; and a processing facility operable to identify, from signals from said sensors, a location of said interface, said apparatus being characterised in that each sensor includes a pair of substantially parallel conductors separated by insulation, said conductors being mounted on a curved outer surface of a carrier.
2. 2. Apparatus as claimed in claim 1 wherein said carrier comprises a tube.
3. 3. Apparatus as claimed in claim I or claim 2 wherein said TDR sensors are mounted in a common mount, there being a given separation between said sensors.
4. 4. Apparatus as claimed in any one of claims I to 3 wherein said controller is configured to activate adjacent sensors in sequence
5. 5. Apparatus as claimed in any one of the preceding claims wherein said processing facility includes a single TDR circuit placed in communication with each of said sensors by said controller.
6. 6. Apparatus as claimed in any one of the preceding claims wherein said pmcessing facility is configured and operable to generate a visual indication of the location of said interface.
7. 7. A TDR sensor including a carrier having a curved outer surface, and a -of substantially parallel conductors mounted on said curved outer surface, said conductors being separated by insulation.
8. 8. A sensor as claimed in claim 8 wherein said conductors lie substantially flush with said curved outer surface.
9. 9. A sensor as claimed in claim 7 or claim 8 whcrcin said carrier is tubular.
10. 10. A sensor as claimed in any one of claims 7 to 9 wherein conductors are aligned along the curvature of said curved outer surface.
11. 11. Asensoras claimed inany one of claims 7th 9 wherein said conductors are aligned substantially perpendicular to the direction of curvature.