WO1994016303A1 - Systeme et procede de detection et de localisation d'une fuite de liquide - Google Patents

Systeme et procede de detection et de localisation d'une fuite de liquide Download PDF

Info

Publication number
WO1994016303A1
WO1994016303A1 PCT/US1993/003499 US9303499W WO9416303A1 WO 1994016303 A1 WO1994016303 A1 WO 1994016303A1 US 9303499 W US9303499 W US 9303499W WO 9416303 A1 WO9416303 A1 WO 9416303A1
Authority
WO
WIPO (PCT)
Prior art keywords
cable
pulse
latch
leak detection
trip signal
Prior art date
Application number
PCT/US1993/003499
Other languages
English (en)
Inventor
William J. Reddy
Original Assignee
W. L. Gore & Associates, Inc.
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by W. L. Gore & Associates, Inc. filed Critical W. L. Gore & Associates, Inc.
Priority to AU42854/93A priority Critical patent/AU4285493A/en
Publication of WO1994016303A1 publication Critical patent/WO1994016303A1/fr

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
    • G01M3/00—Investigating fluid-tightness of structures
    • G01M3/02—Investigating fluid-tightness of structures by using fluid or vacuum
    • G01M3/04—Investigating fluid-tightness of structures by using fluid or vacuum by detecting the presence of fluid at the leakage point
    • G01M3/16—Investigating fluid-tightness of structures by using fluid or vacuum by detecting the presence of fluid at the leakage point using electric detection means
    • G01M3/165—Investigating fluid-tightness of structures by using fluid or vacuum by detecting the presence of fluid at the leakage point using electric detection means by means of cables or similar elongated devices, e.g. tapes

Definitions

  • This invention relates generally to the field of leak detection, wherein a leaking fluid causes an impedance change in a sensing cable and the impedance change is detected using time domain reflectometry (TDR) techniques.
  • TDR time domain reflectometry
  • a variety of chemicals e.g., oils, crude oil, refined petroleum products, gasoline, kerosene, organic solvents, and the like
  • chemicals are stored in both above-ground and underground tanks and are transported through above- ground and buried pipelines. Leakage from these tanks and pipelines can contaminate ground water and cause extensive environmental damage.
  • leaks are difficult to detect and often are not detected until extensive environmental damage has already occurred.
  • the cable includes a pair of conductors (e.g. , coaxial or twin lead) with a permeable insulation layer disposed therebetween.
  • a pair of conductors e.g. , coaxial or twin lead
  • the chemical will permeate the cable and will cause a change in the dielectric properties of the insulation layer such that the impedance of the cable is changed at the point of the leak.
  • TDR time domain reflectometry
  • the leak threshold must be set relatively high to accommodate variations in the characteristic impedance of the cable.
  • the relatively high threshold value reduces the sensitivity of the leak detection system.
  • U.S. Pat. No. 4,797,621 to Anderson et al. discloses a leak detector which overcomes this limitation.
  • Anderson et al. teach a leak detector which digitizes and stores the nominal reflected waveform from a cable in a memory to produce a stored representation of the cable. During monitoring, digitized samples of the reflection are compared to the stored representation. This allows the leak detector to maintain good leak detection sensitivity.
  • leak detection systems such as that taught by Anderson et al. , tend to be costly and consume large amounts of power (primarily because they require a high number of components to implement). It is an object of the present invention to overcome these and other shortcomings of prior leak detection systems.
  • the present invention is a leak detection system and method which uses time domain reflectometry to monitor one or more leak detection cables for impedance changes which are caused when a cable is contacted by a leaking liquid (or when a cable is broken or shorted).
  • a cable is monitored by transmitting a single electrical pulse down its length and then comparing the reflected signal to a predetermined threshold. If the threshold is exceeded, then a fault (leak, break or short) is indicated. For each transmitted pulse, only a selected portion of the cable is monitored. This selective monitoring is achieved by moving a time window along the cable. Each time window position may have unique leak detection and break and short thresholds. This allows the invention to account for signal attenuation and localized characteristic impedance variations of a cable so that sensitivity is preserved.
  • the system includes an electrical pulse generator for generating a single electrical launch pulse.
  • a plurality of switches are used to selectively couple a cable to the electrical pulse generator.
  • a window comparator circuit connected to the selected cable, compares a reflected launch pulse with the predetermined thresholds and produces a trip signal when the reflected launch pulse exceeds a threshold.
  • a latch circuit connected to the window comparator circuit, latches or sets upon receiving the trip signal. A latched state is indicative of a cable fault (leak or break).
  • An enable circuit allows a trip signal to set the latch circuit only during a selected time window.
  • a microcontroller causes the electrical pulse generator to generate the single electrical launch pulse; controls the switches, provides the predetermined thresholds to the window comparator circuit, and generates a fault status signal in response to a latched state of the latch circuit.
  • the controller also controls the enable circuit to step the time window along the cable. By moving the time window along the cable, the entire cable may be monitored one section at a time.
  • the system further includes a reset circuit and a counter circuit.
  • the counter circuit is stopped by a latched state of the latch circuit.
  • the reset circuit produces a reset signal (which is slightly delayed with respect to the launch pulse) which resets the counter circuit.
  • the counter circuit By computing the time between the reset signal and the latched stated of the latch circuit, the counter circuit provides an indication of the time between transmission of the launch pulse and receipt of the reflection. This time is used to compute the distance to the fault which caused the reflection.
  • the system is a four channel leak detection system. That is, four cables are monitored.
  • the invention monitors one cable at a time, but rapidly cycles through each of the four cables in serial fashion.
  • the microcontroller Prior to normal monitoring, the microcontroller autoreferences each cable to generate the predetermined thresholds. Break and short thresholds are set as a function of cable length. Leak thresholds are set as a function of cable length and the normal reflections which can be expected from a section of cable. That is, the leak thresholds for a particular section of cable are set to a value slightly greater than the greatest normal reflection.
  • a sensitive leak detection system is achieved with an elegant and relatively simple circuit. It is an additional advantage of the invention that multiple cables can be monitored by a single circuit. This feature reduces cost, complexity and power consumption of the TDR circuit. The reduced power consumption makes the leak detection system ideal for remote applications where battery and/or solar power is required. It is yet another advantage of the invention that the circuit is output versatile in that it provides several interface options allowing, for example, simple integration to other systems such as a computer.
  • Still another advantage of the invention is that different thresholds are used for detecting leaks, short circuits and broken cables. This allows each threshold to be specifically tailored to the type of fault being detected.
  • FIG. 1 is a high-level diagram showing the leak detection system of the present invention.
  • FIGS. 2 A and 2B show cutaway perspective views of two embodiments of leak detection cable 104.
  • FIG. 3 is a block/ schematic diagram of TDR circuit 102 of the present invention.
  • FIG. 4 is a schematic diagram detailing a portion of TDR circuit 102.
  • FIG. 5 is a high level flow chart illustrating autoreferencing of a leak detection cable in accordance with the present invention.
  • FIG. 6 is a graph illustrating the preferred leak and break thresholds used by the present invention.
  • FIG. 7 shows three scan windows superimposed over a sample portion of a reflection.
  • FIG. 8 is a detailed flow chart illustrating the method performed during autoreferencing for checking a leak detection cable for short circuits.
  • FIG. 9 is a detailed flow chart illustrating the method performed during autoreferencing for locating the end of a leak detection cable.
  • FIGS. 10A and 10B show a flow chart illustrating the method performed during autoreferencing for setting the leak detection thresholds for a leak detection cable.
  • FIG. 11 is a detailed flow chart illustrating the method performed during monitoring of a leak detection cable.
  • FIGS. 12A and 12B illustrate the menu driven operation of the system of the invention.
  • FIG. 13 shows, diagrammatically, the relationship of Figures 8, 9, 10A and 10B.
  • a leak detection system 100 is shown in the environment of a storage tank 105 and a pipeline 108.
  • System 100 includes a time domain reflectometry (TDR) circuit 102 and a leak detection cable 104.
  • Storage tank 106 and pipeline 108 contain a chemical (e.g.. fuel oil) for which leak detection is desired.
  • Cable 104 is laid adjacent to tank 106 and pipeline 108. One end of cable 104 is connected to TDR circuit 102. The other end of cable 104 is left unterminated.
  • Leak detection cable 104 is a conventional leak detection cable as is known in the art.
  • the specific cable chosen will depend on the prope ⁇ ies of the liquid to be detected.
  • a cable may be used such as that disclosed in U.S. Pat. No. 4,877,923 to Sahakian and available as part number LLP126 (water and liquid chemical cable) or LLP118
  • a first embodiment 104(a) of cable 104 is shown in detail in Figure
  • Cable 104(a) includes an inner conductor 202, a first insulation layer 204, a second insulation layer 206, a coaxial outer conductor 208, and an outer protective layer 210.
  • First insulation layer 204 is a hydrophobic, micro-porous insulation such as expanded, micro-porous polytetrafluoroethylene (EPTFE).
  • Second insulation layer 206 is a polyester braided filler which is permeable so as to pass chemicals, but is not hydrophobic.
  • Outer conductor 208 is of braided conductive metal construction and is fluid permeable.
  • Protective layer 210 is a permeable material such as polyethylene or polyester and can be woven or braided.
  • a second embodiment 104(b) of cable 104 is shown in detail in Figure 2B. This embodiment can be used for detecting liquid hydrocarbons.
  • Cable 104(b) includes an inner conductor 202, a first insulation layer 204, a coaxial outer conductor 208, an outer hydrophobic layer 209, and an outer protective layer 210.
  • First insulation layer 204 and hydrophobic layer 209 are both hydrophobic, micro-porous insulation such as EPTFE.
  • Outer conductor 208 is of braided conductive metal construction and is fluid permeable.
  • Protective layer 210 is a permeable material such as polyethylene or polyester and can be woven or braided.
  • cable 104(a) may be used to sense a variety of liquids (including water), while cable 104(b) is limited to liquids which are capable of permeating hydrophobic layer 209 (e.g. , hydrocarbons).
  • the operative difference between the cables is hydrophobic layer 209 of cable 104(b) which prevents water from reaching the conductors such that the impedance of cable 104(b) will not be affected by water.
  • Cable 104(a) allows water to be absorbed into second insulation layer 206 between the conductors such that water can be detected.
  • cable 104 While the preferred embodiments of cable 104 have been described, it should be noted that the present invention permits a leak detection cable of any characteristic impedance to be used so long as the cable impedance is substantially constant throughout its length. Note, however, that a cable with a low characteristic impedance will provide greater signal attenuation (due to loading effects) such that a shorter length of cable (as compared with a higher impedance cable) must be used for successful monitoring.
  • leak detection system 100 If a leak (e.g., leak 110) occurs in tank 106 or pipe 108, the leaking liquid will eventually come into contact with cable 104. In the case of cable 104(a), the liquid will pass through protective layer 210 and conductor 208, and will be absorbed into first insulation layer 204 and second insulation layer 206. The absorbed liquid will cause a change in the dielectric properties of insulation layers 204 and 206. This will result in a change in the impedance of the cable which can be sensed by TDR circuit 102 at a remote end of the cable.
  • a leak e.g., leak 110
  • the liquid will pass through protective layer 210 and contact hydrophobic layer 209. Only hydrocarbons will pass through layer 209 (i.e. , water will be excluded). The hydrocarbon will then pass through outer conductor 208 and will be absorbed into first insulation 204. As with cable 104(a), the absorbed liquid will cause a change in the dielectric properties of insulation layer 204.
  • TDR Circuit 102 Figure 3 shows a block diagram of TDR circuit 102 of the present invention.
  • Circuit 102 includes a microcontroller unit (MCU) 302, a memory 304, an electrical pulse generator 306, a driver circuit 312, relays 314, a delay circuit 346, a digital-to-analog converter (DAC) 316, a window comparator circuit 318, a control gate 320, a latch circuit 322, an oscillator 324, a counter 326, an enable circuit 328, serial input/output (I/O) circuits 336, an alarm
  • MCU microcontroller unit
  • memory 304 for a memory
  • an electrical pulse generator 306 a driver circuit 312, relays 314, a delay circuit 346, a digital-to-analog converter (DAC) 316, a window comparator circuit 318, a control gate 320, a latch circuit 322, an oscillator 324, a counter 326, an enable circuit 328, serial input/output (I/O) circuit
  • Electrical pulse generator 306 includes a programmable delay generator 308 and a programmable pulse width generator 310.
  • Enable circuit 328 includes a programmable delay generator 330 and a programmable pulse width generator 332.
  • Delay generator 330 is identical to delay generator 308.
  • Pulse width generator 332 is identical to pulse width generator 310.
  • Window comparator circuit 318 includes comparators 402 and 404.
  • Control gate 320 includes
  • Latch circuit 322 includes a first RS latch 409 and a second RS latch 413.
  • Latch 409 is formed form NAND gates 410,412.
  • Latch 413 is formed form NAND gates 414,416.
  • An ou ⁇ ut from each of latches 409,413 is input to the enable input of oscillator 324 through a NAND gate 418 and an inverter 420.
  • the output of inverter 420 is used to stop oscillator 324.
  • Another ou ⁇ ut from each of latches 409,413 is input to MCU 302 to indicate the type of fault detected.
  • TDR circuit 102 Operation of TDR circuit 102 proceeds as follows.
  • MCU 302 selects one of four leak detection cables for monitoring by closing one of relays 314A-314D.
  • MCU 302 then initiates production of a single launch pulse by pulse generator 306.
  • the launch pulse is transmitted down the selected leak detection cable. If the launch pulse encounters no impedance changes along the cable, then the pulse will be reflected only from the open end of the cable. If, however, the pulse encounters impedance changes along the cable, each change will cause a portion of the launch pulse to be reflected back to the proximal end of the cable.
  • Comparator 402 monitors the selected cable for negative amplitude reflections.
  • Comparator 404 monitors the selected cable for positive amplitude reflections.
  • the voltage threshold at which comparators 402,404 will trip is set by MCU 302 through DAC 316.
  • MCU 302 retrieves the detection thresholds from memory 304. Upon detecting a reflection which exceeds the respective detection threshold, comparator 402 or comparator 404 will output a signal indicative of the detection.
  • comparator 402 is connected to the input of latch 409 through NAND gate 406.
  • the output of comparator 404 is connected to the input of latch 413 through NAND gate 408.
  • NAND gates 406,408 allow the respective detection signals from comparators 402,404 to set latches 409,413 only when enabled by an ENABLE signal from enable circuit 328. Setting either of the latches in latch circuit 322 will disable oscillator 324 so that counter 326 will stop counting.
  • latch circuit 322 provides status output signals to MCU 302 indicating the type of reflection (positive or negative) which exceeded the threshold.
  • Counter 326 provides the propagation time data which MCU 302 uses to determine the position of the fault on the cable.
  • MCU 302 computes the position of the fault on the cable by multiplying the velocity of propagation (VOP) of the signal in the particular cable by one half of the elapsed time from counter 326.
  • VOP velocity of propagation
  • the factor of one half accounts for the fact that the signal must traverse the cable twice. That is, the incident pulse must travel out, and the reflected signal must travel back.
  • the size of the reflection which will set either of the latches is controlled by the detection thresholds.
  • the detection thresholds In order to detect small leaks (or other faults which cause only slight impedance changes), especially at a far end of a leak detection cable, it is important to set the detection thresholds very near zero. However, setting the detection thresholds near zero is not practical because even in the absence of leaks or breaks along the cable, some reflections will occur from inherent variations in the impedance of a cable. These variations will differ from cable to cable.
  • TDR circuit 102 implements a programmable time or scan window.
  • the scan window allows a leak detection cable to be monitored a segment at a time. For each launch pulse transmitted down the cable, only a portion of the reflected signal corresponding to the scan window is monitored for reflections.
  • Each segment of the cable (as defined by the scan window) will have unique negative and positive reflection thresholds which are stored in memory 304. Any reflections received during the scan window will be compared to the corresponding detection thresholds for that particular segment of cable. Monitoring of the entire cable is achieved by moving the scan window along the cable.
  • control gate 320 allows the detection signals from window comparator circuit 318 to set latches 409,413 only when enabled by an ENABLE pulse.
  • the ENABLE pulse controls the scan window.
  • the ENABLE pulse is produced by programmable delay generator 330 and programmable pulse width generator 332.
  • the MCU 302 initiates production of the ENABLE pulse simultaneously with initiation of the launch pulse.
  • the ENABLE pulse has a programmable delay and a programmable width.
  • the width is controlled by MCU 302 through programmable pulse width generator 332. Changing the width of the ENABLE pulse will alter the size of the scan window.
  • the ENABLE pulse delay (i.e. , the delay from the time the launch pulse is sent down the cable and the time the ENABLE pulse is provided to control gate 320) controls the position of the scan window along the cable.
  • the time of the delay is controlled by MCU 302 through programmable delay generator 330. Changing the delay value will alter the portion of the leak detection cable which is monitored. That is, the scan window will be moved or stepped along the cable.
  • the launch pulse is produced by electrical pulse generator 306 under control of MCU 302.
  • Electrical pulse generator 306 includes a programmable delay generator 308 and a programmable pulse width generator 310. Pulse width generator 310 allows generator 306 to produce a launch pulse of varying width.
  • a variable width launch pulse has the following important utility. Propagation down a leak detection cable will attenuate the amplitude of a launch pulse.
  • a variable width launch pulse will compensate for this attenuation and substantially prevent a loss of sensitivity at a far end of the cable.
  • a relatively narrow pulse may be transmitted down a cable when monitoring proximal portions of the cable.
  • a wider pulse (containing more energy) may be transmitted down a cable when monitoring more distant portions of the cable.
  • Programmable delay generator 308 is identical to programmable delay generator 330 of enable circuit 328. Delay generator 308 is normally programmed for a delay value of zero. It is required in the circuit, however, to match the intrinsic "zero" delay value of delay generator 330. That is, delay generator 330 has an internal delay even when programmed to ou ⁇ ut a zero delay signal. Using identical delay generators 308,330 in enable circuit 328 and in pulse generator 306 produces a "time zero baseline" common to both the launch pulse and the scan window. Delay generator 308 may be programmed for a non-zero delay value to eliminate a mismatch in the propagation delays through delay generators 308,330, if one should exist.
  • delay circuit 346 Upon transmission of a launch pulse by pulse generator 306, delay circuit 346 generates a RESET signal or pulse.
  • the RESET pulse is directed to latch circuit 322 and counter circuit 326.
  • the RESET pulse causes latches 409,413 to be reset from the previous measurement and causes counter 326 to be reset to zero at transmission of the launch pulse.
  • Relays 314A-314D are used to select one of the four leak detection cables for monitoring or referencing.
  • relays 314 are single pole, normally OPEN reed-type relays. Each relay is controlled by MCU 302 through a drive transistor (not shown).
  • Status relays 340A-340D provide status outputs for TDR circuit 102.
  • Each relay 340 corresponds to a channel of TDR circuit 102.
  • the relays may be used to activate a remote alarm, pump, etc., as an indication of the fault condition (e.g., a closed relay corresponds to a detection event on the corresponding cable).
  • relays 340A-340D are single pole, double throw, non-latching relays.
  • An alarm 338 is provided as a local indicator of a leak or break condition on a cable. Alarm 338 may be, for example, an audible or a visible indicator.
  • Serial I/O (input/output) ports 336A.336B provide normal serial data communications (e.g. , to a computer, printer, or other device).
  • Serial I/O ports 336A,336B also allow a plurality of TDR circuits 102 to be connected together over a network.
  • serial I/O ports 336A,336B are RS485 compatible transceivers.
  • Additional circuitry includes a real time clock and a battery power backup circuit for the clock.
  • TDR Circuit 102 has essentially two modes of operation: an autoreferencing (referencing) mode and a monitoring mode.
  • the referencing mode the leak detection thresholds are determined and stored in memory 304 for later use.
  • the monitoring mode comparison of reflections with the previously determined detection thresholds is performed.
  • circuit 102 During referencing of a cable, circuit 102 actually determines three thresholds for each scan window: a short (short circuit) threshold, a break (open circuit) threshold, and a leak threshold.
  • the short thresholds and the break thresholds are functions of the length of a cable only and are not dependent on the impedance of the cable.
  • Each leak threshold is a function of cable length and the impedance of the cable.
  • the normal reflections which can be expected from a segment of a cable i.e., the reflection occurring during a scan window
  • the leak threshold for that particular scan window is then set to a value slightly greater (i.e., slightly more negative) than the greatest normal reflection.
  • FIG. 5 is a high level flow chart illustrating the steps involved with referencing a leak detection cable.
  • a leak detection cable is selected for referencing.
  • TDR circuit 102 is a four channel device capable of monitoring four leak detection cables.
  • the selected cable is checked for short circuits. If. at step 506. any short circuits are detected, then referencing is aborted at step 508.
  • the end of the cable is located. If the end of the cable is left open or unterminated, a large amplitude reflection will occur from the abrupt low-to-high impedance change. This reflection is used to locate the end of the cable. If, at step 512, the end of the cable cannot be located, then referencing is aborted at step 508.
  • the cable is referenced at step 514 to determine the leak thresholds. After the leak thresholds have been determined for the entire cable, autoreferencing ends as indicated at step 516.
  • TDR circuit 102 Provided below are the preferred parameters for TDR circuit 102. These parameters include the size and number of scan windows used for monitoring; the manner in which a scan window is moved along a cable; launch pulse widths; and leak, break and short threshold values. These preferred values are design choices which are provided for purposes of illustration. The invention is not limited to these specific values.
  • the scan windows are used to divide a cable into three regions. For each region, the width of the scan window
  • window width the amount the window is moved along the cable during each cycle (scan step), the short, break and leak thresholds and the width of the launch pulse are unique.
  • the width of the launch pulse is varied to accommodate the requirements of different portions of the cable. As discussed above, it is desirable to use a wide launch pulse (e.g., 350 ns) when monitoring distant portions of a cable to compensate for attenuation which occurs as the pulse propagates through the cable. It is desirable to use a narrow launch pulse (e.g. , 50 ns) when monitoring the near end of the cable.
  • the narrow pulse reduces the "dead time" of TDR circuit 102. Dead time occurs because TDR circuit 102 is unable to monitor for reflections while the launch pulse is present at the near (transmitting) end of the cable (the wider the launch pulse. the longer the dead time).
  • the width of the scan window is varied for several reasons. Changes in temperature and changes in the impedance of the leak detection cable (such as would be caused by a leak) will change the propagation time of the launch pulse (and reflections) in the cable. Changes in propagation time will appear to TDR circuit 102 as changes in the length of the cable being monitored. This will in-turn cause the scan windows to shift position along the cable. If the detection (short, break and especially leak) thresholds between adjacent windows are vastly different, a shifting of the window positions along the cable may result in false alarms or missed detections. This is especially problematic at the distant end of a cable where the large magnitude end reflection is present. The end reflection rises slowly because of the distributed capacitance of the cable. This slow rise may traverse several scan windows.
  • the last scan window monitoring the cable is made very wide. This allows a single scan window to monitor the portion of the end reflection containing the sharpest slopes such that an abrupt change in threshold values between adjacent windows is prevented. By doing this, the possibility of false and missed detections is greatly reduced.
  • break threshold and the short threshold are absolute values.
  • the thresholds vary as a function of window number (which translates to distance along the cable). Contrast the leak detection threshold. In addition to being a function of window number, the leak threshold is a relative value.
  • the leak threshold is then set relative to this value. Thus, only the leak threshold is actually customized for each cable. Because shorts and breaks will produce significant impedance changes, highly sensitive thresholds are not required for detection.
  • 85 scan windows are used for monitoring a cable. Windows 1-6 are used to define region #1. Windows 7-24 are used to define region #2. Windows 25-85 are used to define region #3. To meet the requirements and specifications discussed above, the windows will have the following specifications:
  • the window step is less than the window width. This results in overlapping windows. The overlap helps to ensure that abrupt changes in threshold values will not occur between adjacent windows. The overlap also prevents the possibility of dead zones occurring between the windows. Note also that the detection thresholds become more sensitive as the windows become more distant from the proximal end of the cable. Further, the width of a window is made much wider if that window is found to be the last window monitoring the cable. As discussed above, this preserves leak detection sensitivity near the end reflection.
  • Figure 6 is a graph illustrating the preferred break and leak thresholds (voltage) as a function of window number (distance). Region #1 , region #2 and region #3 are also illustrated. Reference number 602 indicates the plot of the break threshold. Note that this threshold is shown in volts. Reference number 604 indicates the plot of the leak threshold. Note that the leak threshold is shown in millivolts (mV). Note also that the break threshold is an absolute threshold which will not vary from cable to cable. The leak threshold, on the other hand, is set relative to the most negative reflection in each window. For example, the leak threshold at window 30 is equal to -24 mV. If a leak detection cable had an inherent reflection of -5 mV occurring during this window, then the leak threshold for the cable at window 30 would be set at -29 mV.
  • the short threshold in not shown in Figure 6. However, if shown, the short threshold would appear as an identical mirror image of the break threshold.
  • Figure 7 is a graph showing several scan window positions (702,704 and 706) superimposed over a portion of a sample reflected waveform 701.
  • a 160ns wide window is moved in 120ns steps such that the windows overlap by 40ns.
  • Figures 8-10 illustrate the steps of Figure 4 in greater detail.
  • FIG 13 shows, diagrammatically, the relationship of Figures 8, 9, 10A and 10B.
  • Figure 8 is a flow chart detailing the steps involved in testing a cable for short circuits.
  • Figure 9 is a flow chart detailing the steps involved in locating the end of a cable.
  • Figure 10 is a flow chart detailing the steps involved in referencing a cable to determine the leak thresholds. Referring first to Figure 8, the method for testing a cable for short circuits is described.
  • the window number is set to 1.
  • the window number is checked to determine whether it is greater than 6. If the window number is not greater than 6, then the short threshold, the launch pulse width, and the window width are set, in step 805, to the values discussed above for region #1.
  • window number is greater than 6, then it is checked to see if it is less than 25 at step 806. If the window number is less than 25, then the short threshold, the logic pulse width, and the window width are set, in step 807, to the values discussed above for region #2.
  • step 808 If the window number is greater than 25, then the short threshold, the logic pulse width, and the window width are set, in step 808, to the values discussed above for region #3. After step 805, 807 or 808, the method proceeds to step 810.
  • variables x and y are set equal to zero. These variables are used as counters.
  • the variable x is used to keep count of the number of pulses transmitted down the cable.
  • the variable y is used to keep track of the number of pulses which produce a reflection which sets latch 409.
  • a variable "ACCUM" is also set to zero in step 810. ACCUM is used as an accumulator to sum the time values produced by counter 326.
  • step 811 a pulse is transmitted down the selected leak detection cable.
  • Latch 409 is then checked at step 812 to determine whether it is set. If the latch is set, then the variable y is incremented by 1 and the time value from counter 326 is added to ACCUM in step 813. If the latch was not set or after step 813 has completed, the method continues with step 814 where the variable x is incremented by 1. Variable x is then checked to determine whether it is greater than 30 at step 816. If x is not greater than 30, then steps 811 through 816 are repeated until x is greater than 30.
  • step 818 the variable y is checked to determine whether it is greater than 24. If y is greater than 24, then a short circuit is deemed to have been detected.
  • the distance to the short circuit is calculated at step 819 by dividing the value of ACCUM by the variable y. This yields an averaged distance to the short circuit.
  • the short circuit condition and the distance to the short are displayed on display 344 and archived in memory 304 (e.g. , non-volatile EEPROM) at step 820. Referencing is then aborted at step 821. Referencing is not possible until the short circuit condition is corrected.
  • step 822 the window number is increased by 1.
  • the window number is then checked at step 824 to determine whether it is greater than 85 (which is the maximum window number available). If the window number is not greater than 85, then the method returns to step 804, and steps 804 through 824 are repeated for the next window. If the window number is greater than 85, then no short circuit conditions were detected for the cable and referencing continues by finding the end of the cable as illustrated at step 510 of Figure 5. Referring now to Figure 9, the method for finding the end of a cable is discussed in detail.
  • Steps 902 through 916 for finding the end of the cable are substantially identical to steps 802 through 816 for testing for short circuits.
  • the break thresholds are used rather than the short thresholds and at step 912 latch 413 is monitored rather than latch 409 (as in step 813).
  • step 918 the variable y is checked to determine whether it is greater than 24. If y is not greater than 24, then the end of the cable is deemed to not be present in the selected time window. In that case, the method continues with step 922 where the window number is increased by 1. The window number is then checked at step 924 to determine whether it is greater than 85 (which is the maximum window number available).
  • steps 904 through 924 are repeated for the next scan window. If the window number is greater than 85, then steps 904 through 924 are repeated for the next scan window. If the window number is greater than 85, then referencing has not located an end of the cable. This condition is displayed at step 926 and referencing is aborted at step 928. Proper referencing requires that the end of the cable be found.
  • variable y will be greater than 24.
  • the distance to the end of the cable is calculated at step 919 by dividing the value of ACCUM by the variable y.
  • the length of the cable is then displayed on display 344 at step 920.
  • VOP velocity of propagation
  • TDR circuit 102 allows the VOP to be adjusted in tenth of a percentage increments. This allows a user to adjust the VOP until the computed cable length matches the actual cable length. Because the cable length is computed directly from the measured time and the VOP, the new cable length can be immediately computed and redisplayed as the
  • VOP is adjusted.
  • step 1002 the window number is set to 1.
  • variables x and y are set equal to zero.
  • the variable x is used to keep count of the number of pulses transmitted down the cable.
  • the variable y is used to keep track of -li ⁇
  • the leak threshold is also set to zero (mV) in step 1004.
  • step 1006 a pulse is transmitted down the selected leak detection cable.
  • Latch 409 is then checked at step 1008 to determine whether it is set. If the latch is set, then the variable y is incremented by 1 in step 1010. If the latch was not set or after step 1010 has completed, the method continues with step 1012, where the variable x is incremented by 1.
  • variable x is checked to determine whether it is greater than 30. If x is not greater than 30, then steps 1006 through 1014 are repeated until x is greater than 30.
  • step 1016 the variable y is checked to determine whether it is greater than 24. If y is greater than 24, then a variable z is set equal to + 1 at step 1018. If y is not greater than 24, then variable z is set equal to -1 at step 1020.
  • Variable z is used to indicate which direction the leak detection threshold must be adjusted for the selected scan window. A positive value indicates that the threshold needs to be made larger (i.e. , more negative). A negative value indicates that the threshold needs to be made smaller (i.e. , more positive).
  • step 1022 After a value of z has been set in one of steps 1018 or 1020, the method continues with step 1022.
  • variables x and y are reset to zero and the leak threshold is incremented by a preselected voltage amount (e.g., 2.5 mV) times z.
  • Steps 1024 through 1032 then send 30 pulses down the cable and count the number of times latch 409 is set.
  • variable y is compared to 24 and the value of variable z is checked. If y is greater than 24 and z is equal to -1 or if y is not greater than 24 and z is not equal to -1 , then the method returns to step 1022. Otherwise, the method proceeds to step 1040.
  • the window number is checked to determine whether it is greater than 6. If the window number is not greater than 6, then the leak threshold is set for region #1 at step 1402. In region #1 , the leak threshold is set equal to the leak threshold from step 1022 minus 25 mV (i.e. , the threshold is made more negative by 25 mV).
  • the leak threshold is set for region #2 in step 1046. In region #2, the leak threshold is set equal to the leak threshold from step 1022 plus [(.56 times the window number) minus 28.3] mV.
  • the leak threshold is set for region #3 in step 1048.
  • the leak threshold is set equal to the leak threshold from step 1022 minus (720 ⁇ the window number) mV.
  • step 1050 the window number is incremented by one.
  • the window number is then checked, in step 1052, to determine whether it exceeds the length of the cable. If it does, then "Reference Complete" is displayed on display 344 and referencing ends at step 1056. If, however, the end of the cable has not been reached, the method returns to 1004 and steps 1004 through 1052 are repeated for the next window.
  • TDR circuit 102 continuously monitors each leak detection cable for breaks and leaks.
  • the method is detailed in the flow chart of Figure 11. The method is a continuous loop.
  • window number 1 is selected for monitoring.
  • variables x, y,, y 2 , ACCUM, and ACCUM 2 are initialized to zero.
  • Variable x is used to count the number of pulse sent down the cable.
  • Variable y is used to count the number of times a negative reflection sets latch 409.
  • Variable y 2 is used to count the number of times a positive reflection sets latch 413.
  • ACCUM is used to accumulate the times from counter 326 each time latch 409 is set.
  • ACCUM 2 is used to accumulate the times from counter 326 each time latch 413 is set.
  • step 1106 the leak threshold and the break threshold for the current scan window of the current cable are retrieved from memory 304 and are loaded into DAC 316.
  • DAC 316 then provides the appropriate analog voltage thresholds to window comparator circuit 318.
  • a single launch pulse is transmitted down the leak detection cable.
  • latch 409 is checked to determine whether it is set. If latch 409 is set, then variable y, is incremented by 1 and the time value from counter 326 is added to ACCUM, at step 1112. If latch 409 is not set or after step 1112, the method proceeds to step 1114.
  • Latch 413 is checked at step 1114 to determine whether it is set. If latch 413 is set, then variable y 2 is incremented by 1 and the time value from counter 326 is added to ACCUM 2 at step 1116. If latch 413 is not set or after step 1116, the method proceeds to step 1118, where variable x is incremented by 1.
  • variable x is compared to 30 to determine whether 30 pulses have been sent down the cable for the currently selected scan window. If not, then steps 1108 through 1120 are repeated. If x is greater than 30. then the method proceeds at step 1121. At step 1121, y, is checked to determine whether is greater than 24.
  • y is greater than 24, then a leak is deemed to have been detected.
  • the distance to the leak is calculated at step 1122 by dividing the value of ACCUM, by the variable y,. This yields an average distance to the leak.
  • step 1124 After the distance to the leak is computed at step 1124, it is displayed at step 1124 on display 344. If, at step 1121 , y, was not greater than 24 (indicating no leak in the selected scan window) or after completion of step 1124, the method continues with step 1126.
  • y 2 is checked to determine whether is greater than 24. If y 2 is greater than 24, then a break is deemed to have been detected. The distance to the break is calculated at step 1128 by dividing the value of
  • step 1130 After the distance to the break is computed at step 1128, it is displayed at step 1130 on display 344. If, at step 1126, y 2 was not greater than 24 (indicating no break in the selected scan window) or after completion of step 1130, the method continues with step 1132.
  • the window number is increased by 1.
  • the window number is then checked at step 1134 to determine whether it is greater than 85 (which is the maximum window number available). If the window number is not greater than 85 than the method returns to step 1104 and steps 1104 through 1134 are repeated for the next window. If the window number is greater than 85, then a next channel (i.e., next leak detection cable) is selected for monitoring at step 1136. The method then returns to step 1102 and repeats for the selected cable.
  • TDR circuit 102 is capable of detecting and displaying a plurality of leaks for each cable without requiring re-referencing of the cable.
  • TDR circuit 102 controls operation of TDR circuit 102 via keypad 342 and display 344.
  • display 344 provides a user with a series of menu choices. The user will make selections via keypad 342.
  • FIG. 12A and 12B a sample menu setup for operation of TDR circuit 102 is described.
  • "Monitoring" will be displayed as indicated by screen 1202. Depressing a NEXT button will advance the display through five set up screens.
  • Screen 1204 displays configuration data (VOP and cable length) for channels one and two.
  • Screen 1206 displays configuration data (VOP and cable length) for channels three and four.
  • Screen 1208 is used to select system setup options.
  • Screen 1210 is used to select archival options.
  • Screen 1212 is used to select system tests.
  • test relays 340 note that password security is used to limit user access to this feature
  • test display 344
  • Archival data is stored in memory 304 by MCU 302.
  • the archival data includes the date and time when: a cable fault occurred, a cable was referenced, a relay was cleared, the date and time were changed, the power was turned on or off, the VOP was changed for a cable, or an alarm was silenced.
  • MCU 302 is an MC68331 microcontroller from Motorola Semiconductor Products, Inc. , Phoenix, Arizona.
  • Memory 304 includes two 256K EPROM's, a 256K EEPROM and a 256K RAM.
  • the software which controls operation of TDR circuit 102 is stored in the EPROM's. Thresholds and archival data are stored in the EEPROM.
  • DAC 308 is an AD7247 digital to analog converter from Analog Devices, Norwood, MA.
  • Delay generators 308,330 are available from Analog Devices under part number AD9401.
  • Pulse width generators 310,332 are available from Engineered Components Co., San Luis Obisto, CA, under part number PDL-ACT-8-50.
  • Driver 312 is a part number VP2210N3 MOSFET available from Supertex, Inc. , Sunnyvale, CA.
  • Counter 326 is a 14 stage binary counter available from General Electric Solid State, Somerville, New Jersey, as part number 74AC7061.
  • Oscillator 324 is a 100 MHz oscillator available as part number P1100-SY from Pletronics, Lynnwood, WA.
  • Comparators 402 and 404 are AD790 comparators from Analog
  • NAND gates All NAND gates are part number 74ACT00.
  • the NAND gates are available from numerous sources (e.g. , Harris, RCA, GE, Intersil, etc.).
  • Latches 409 and 413 are standard RS latches. Each latch includes two, two- input NAND gates.
  • Oscillator 324 provides a 100 MHz clock signal to counter 326. This gives counter 326 a resolution of 10 nanoseconds. At 0.8 feet per nanosecond, this yields a leak location sensitivity of ⁇ 4 feet (i.e. 10 ns x 0.8 ft/ns ⁇ 2).
  • a jumper cable may be used to connect circuit 102 to each leak detection cable. If the jumper cable is long enough, then the "dead zone" will be removed from the leak detection cable and will reside instead on the jumper cable.
  • the jumper cable should have the same characteristic impedance and velocity of signal propagation as the leak detection cable in order to ensure accurate propagation time/distance calculations and to maintain leak detection sensitivity.
  • the preferred jumper cable is part number LLP128 from W.L. Gore and Associates, Inc. In the preferred embodiment, a 100 foot jumper cable is used.
  • a trailer cable is normally connected to the distant end of the cable. Doing this effectively extends the apparent end of the cable away from the end of the detecting portion of the cable so that the end reflection will not interfere with monitoring.
  • a 25 foot trailer cable is used.
  • the preferred trailer cable is part number LLP128 from W.L. Gore and Associates, Inc.
  • the preferred embodiment of the invention was tested for the detection of unleaded gasoline (without an oxygenated additive) using five different lengths of cables.
  • the cables were formed by stinging together multiple reels of jumper cable to form the required lengths.
  • a W.L. Gore Model LLP1 18 hydrocarbon detection cable and a 25 foot trailer cable were attached to the jumper cable.
  • the results of the tests are presented in the table below. Each row indicates the results of between two and four tests which have been averaged together.
  • the first column contains the total length of the cable set-up (including jumper cable, detection cable and trailer cable).
  • the second column lists the VOP (as a percentage of the speed of light in a vacuum) used for testing the particular cable.
  • the third column indicates the length of cable which was immersed into the gasoline (i.e., wetted length).
  • the fourth column provides the time required for detection after immersion.
  • the fifth column lists the distance to the leak computed by TDR circuit 102.

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Examining Or Testing Airtightness (AREA)
  • Locating Faults (AREA)

Abstract

L'invention se rapporte à un procédé et un système de détection de fuites, dans lesquels un circuit électrique utilise la réflectométrie temporelle pour surveiller un ou plusieurs câbles de détection de fuite, afin de détecter des variations d'impédance qui sont causées lorsqu'un câble entre en contact avec un liquide qui fuit. On surveille un câble par transmission d'une impulsion électrique unique sur sa longueur puis par comparaison du signal réfléchi avec un seuil de détection de fuite prédéterminé. Si le seuil est dépassé, une défaillance est indiquée. Pour chaque impulsion transmise, seule une partie choisie du câble est surveillée afin d'y détecter des réflexions. On effectue cette surveillance sélective en déplaçant une fenêtre temporelle le long du câble. Chaque position de fenêtre temporelle peut présenter un seuil de détection de fuite unique.
PCT/US1993/003499 1993-01-15 1993-04-13 Systeme et procede de detection et de localisation d'une fuite de liquide WO1994016303A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
AU42854/93A AU4285493A (en) 1993-01-15 1993-04-13 System and method for detecting and locating a leaking liquid

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US284693A 1993-01-15 1993-01-15
US08/002,846 1993-01-15

Publications (1)

Publication Number Publication Date
WO1994016303A1 true WO1994016303A1 (fr) 1994-07-21

Family

ID=21702810

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US1993/003499 WO1994016303A1 (fr) 1993-01-15 1993-04-13 Systeme et procede de detection et de localisation d'une fuite de liquide

Country Status (2)

Country Link
AU (1) AU4285493A (fr)
WO (1) WO1994016303A1 (fr)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018086949A1 (fr) * 2016-11-11 2018-05-17 Leoni Kabel Gmbh Procédé et système de mesure servant à la surveillance d'une ligne
CN109186896A (zh) * 2018-09-21 2019-01-11 河海大学 一种用于检测隧道渗漏的长距离分布式监测系统
CN110823355A (zh) * 2019-11-19 2020-02-21 湖南国奥电力设备有限公司 基于位移数据的地下电缆故障判断方法和装置
CN112415323A (zh) * 2019-08-23 2021-02-26 微芯片技术股份有限公司 诊断网络内的电缆故障
CN113219310A (zh) * 2021-04-23 2021-08-06 深圳供电局有限公司 局部放电定位方法、装置、定位设备和存储介质

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO1991019206A1 (fr) * 1990-06-04 1991-12-12 W.L. Gore & Associates, Inc. Circuit electronique servant a detecter les fuites de substances chimiques

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO1991019206A1 (fr) * 1990-06-04 1991-12-12 W.L. Gore & Associates, Inc. Circuit electronique servant a detecter les fuites de substances chimiques

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
PATENT ABSTRACTS OF JAPAN vol. 7, no. 113 (P-197)(1258) 18 May 1983 *

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018086949A1 (fr) * 2016-11-11 2018-05-17 Leoni Kabel Gmbh Procédé et système de mesure servant à la surveillance d'une ligne
JP2019533821A (ja) * 2016-11-11 2019-11-21 レオニ カーベル ゲーエムベーハー 導線を監視するための方法と測定装置
US11041899B2 (en) 2016-11-11 2021-06-22 Leoni Kabel Gmbh Method and measuring assembly for monitoring a line
CN109186896A (zh) * 2018-09-21 2019-01-11 河海大学 一种用于检测隧道渗漏的长距离分布式监测系统
CN112415323A (zh) * 2019-08-23 2021-02-26 微芯片技术股份有限公司 诊断网络内的电缆故障
CN110823355A (zh) * 2019-11-19 2020-02-21 湖南国奥电力设备有限公司 基于位移数据的地下电缆故障判断方法和装置
CN113219310A (zh) * 2021-04-23 2021-08-06 深圳供电局有限公司 局部放电定位方法、装置、定位设备和存储介质

Also Published As

Publication number Publication date
AU4285493A (en) 1994-08-15

Similar Documents

Publication Publication Date Title
US5134377A (en) TDR system and method for detecting leakage of a liquid
EP0298479B1 (fr) Détecteur et localisateur de fuite utilisant la réflectométrie à domaine de temps et des techniques d'échantillonnage
US5159276A (en) Capacitance measuring circuit and method for liquid leak detection by measuring charging time
US4949076A (en) Apparatus for detecting and locating leakage in underwater structures
US4766386A (en) Time domain reflectometer for measuring impedance discontinuities on a powered transmission line
US7262607B2 (en) Characterizing substances with multistatic probes
US6121780A (en) Material interface level sensing
US7212008B1 (en) Surveillance device detection utilizing non linear junction detection and reflectometry
US4829284A (en) Procedure for monitoring of an object using a signal line, together with a pulse measuring apparatus to carry out this procedure
US20040066324A1 (en) Guided wave radar level transmitter
CA2742667A1 (fr) Systemes et procedes d'essai recourant a la reflectometrie temporelle
US7634945B2 (en) Apparatus to water detection in a storage tank
WO1994016303A1 (fr) Systeme et procede de detection et de localisation d'une fuite de liquide
EP3948181B1 (fr) Coupleur de système haute et/ou basse énergie
US10359322B2 (en) Method and device for detecting hot points in a facility, especially for detecting leaks in air ducts
NO803157L (no) Feildetekterende foeler.
US5179342A (en) Superconductor quench measuring device which evaluates reflected pulses
US3750012A (en) Cating system for waveguide transmission line utilizing spaced reflectors
EP0532616B1 (fr) Circuit electronique servant a detecter les fuites de substances chimiques
EP0160440A1 (fr) Dispositif pour détecter et recevoir des informations sur des changements de variables
Harney et al. Electromagnetic level indicating (EMLI) system using time domain reflectometry
Jeon et al. Cable fault and assessment using multiple frequency autoencoder regression-based reflectometry
RU2009476C1 (ru) Способ контроля протяженного трубопровода, заполненного жидким продуктом с диэлектрическими свойствами
EP0044431A1 (fr) Appareil pour détecter les variations locales de température d'un objet et les endroits des variations de température
KAYSER et al. 3.14 Radar, Contact Level Sensors

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): AU DE GB JP SE

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): AT BE CH DE DK ES FR GB GR IE IT LU MC NL PT SE

121 Ep: the epo has been informed by wipo that ep was designated in this application
DFPE Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101)
REG Reference to national code

Ref country code: DE

Ref legal event code: 8642

122 Ep: pct application non-entry in european phase