WO2014035719A1

WO2014035719A1 - Power line fault locating system

Info

Publication number: WO2014035719A1
Application number: PCT/US2013/055687
Authority: WO
Inventors: Peter REILLEY
Original assignee: Reilley Peter
Priority date: 2012-08-25
Filing date: 2013-08-20
Publication date: 2014-03-06

Abstract

The Power Line Fault Locating System is a means of locating low level faults in a utility power system by sampling the noise on the power line at two positions in the system and correlating the resulting data. When the correlated data produces a spike, a noise source has been detected. The position of the spike in the data will indicate the location of the noise source with great precision. The invention can operate on both continuous noise resulting from minor flaws in the power system as well as the noise burst resulting from a power system failure. Noise resulting from minor flaws may come from slack line fittings, electrical leakage in insulators or at poles. Burst noise may come from insulator flash over, lightning strike, or something contacting a power line.

Description

TITLE: POWER LINE FAULT LOCATING SYSTEM

CROSS-REFERENCE TO RELATED APPLICATIONS: Not Applicable

FEDERALLY SPONSORED RESEARCH: Not Applicable

SEQUENCE LISTING OR PROGRAM: Not Applicable

BACKGROUND OF THE INVENTION - FIELD OF INVENTION

This invention relates to finding the location of a fault on a power line.

BACKGROUND OF THE INVENTION - PRIOR ART

Electric utilities must monitor their transmission lines to insure their proper operation and prevent damage to their systems or customers equipment. Power outages can result from many different causes, some internal to the system such as equipment failure others from external causes such as a tree falling on a power line. In all cases the utility must find and repair the damage as rapidly as possible. Often the problem is rapidly locating the point of failure.

Utilities use many different devices to detect abnormal conditions in the power system. Devices currently in use measure the voltage and current on the power lines to insure if they are within normal limits. Should these devices detect a fault then a circuit breaker is opened and the line is protected from overload.

Some types of faults such as noise on the power lines are not detected by these devices. This noise may not indicate an immediate failure. The cause of the noise may not indicate a failure at all. The noise can go on for years. Such noise can emanate from arcing around mounting bolts and fixtures. These parts are very heavy and are not seriously damaged by the arcing. However, noise indicates a fault with the equipment and can result in an eventual failure.

This noise, even if it is not a problem for the utility, can be a problem for other people in the area. It can disrupt radio communication, television reception, and other equipment. Above certain limits this noise can violate government restrictions on unintended radiation emissions. This can result in fines.

Noise emissions can be located using special wide bandwidth RF receivers and directional antennas such as equipment from Exacter Inc or Radar Engineers. With these devices a person will drive around searching for the strongest reception on his receiver. The receiver allows the operator to listen to the noise. Sometimes the noise has audio characteristics that can give a clue as to the type of fault. This is tedious, time consuming, and cannot be automated.

Patent number 7,777,676 describes a system for locating lightning strikes by using the time difference as measured by receivers at multiple points. This system is synchronized using GPS receivers. However this system depends on the lightning strike having a known waveform. This known waveform must have large enough amplitude to stand above the background noise. It will not work on signals buried in the noise. This patent while not related to power lines does describe relevant prior art.

Patent number 2,717,992 discloses a system where the impulse resulting from a transient fault is detected at each end of the power line. The two detectors communicate and produce graphical output that indicates the position of the fault on the power line. The graph shows the relative timing of the impulse. The difference in the arrival times is proportional to the distance to the fault.

Patent number 3,462,681 discloses a system that detects an electrical fault by sensing the disturbance wave as it travels to each end of the cable and senses the disturbance as it passes detectors connected to the ends of the cables and calculates the time difference of the arrival of the disturbance at the sensing points and thereby determines the location of the fault. This is a similar method to patent "992 in that it times the arrival of the impulse resulting from a fault.

An early patent 2,493,800 discloses a technique where a high voltage is applied to the power line. This voltage is high enough to cause a voltage breakdown. The timing of the resulting wave from the breakdown is measured at each end of the line. The result is shown on a CRT display as a pattern that indicates the position of the breakdown.

Patent number 6,822,457 discloses a system for locating faults on a power line but includes transmitters that apply a test signal to the line to determine its characteristics. This measures the propagation delay of the power line. This is an important factor in the calculation of the exact position of a fault. However, it can be impractical to inject a test signal on an energized line. Additionally, such test signals will also radiate and potentially disrupt radio communications.

Patent 5,608,328 does attempt to locate an arcing fault. It does this be detecting the polarity of a high current fast rise time pulse. A portable detector is moved along the cable while the fault is occurring. Then the polarity of the pulse changes then the fault has been found. This system is not suitable for permanent installation and requires a fault pulse to work.

Patent 7,577,535 describes an apparatus capable of determining the direction of a fault relative to the apparatus. It is a system that uses Fourier analysis to extract the frequencies of interest. It is useful in locating arcing faults that have not yet resulted in a system outage. However it is not able to determine the distance to the fault rather it is a mobile device that locates the fault by moving along the path of the power line.

The present invention solves these problems because it does not depend on the fault signal having any recognizable characteristics. It need not be an impulse; the fault can be a noise like signal. When an autocorrelation is performed on a noise waveform the resulting correlation graph will have a peak only at zero offset. Noise will not correlate at any other offset. It is this characteristic of the autocorrelation function that allows the present invention to work. Further, the fault signal amplitude need not stand above the background noise; the correlation will extract the fault noise and its position even if it is embedded in the background noise.

BACKGROUND OF THE INVENTION - OBJECTS AND ADVANTAGES

The present invention addresses these issues;

1. Precise location of a noise source.

2. Able to operate at a long distance from the origin of the noise.

3. The invention continuously monitors the power line for faults.

4. Can be connected into the existing power line protection systems. 5. Completel y automated.

6. Does not depend on detecting an event but rather extracts location

information from a noisy set of data.

7. Able to locate faults that current systems cannot detect.

8. Will operate on AC or DC power lines.

SUMMARY

The Noise Source Locating System has significant advantages over the existing systems. It is able to locate an arcing fault with only two detectors, one on each end of a power line. It does this by correlating the data received from the two detectors. The

con-elation data along with precise timing data provides the physical distance to the fault. The data is not expected to have any recognizable pattern; rather it is expected to have the random characteristic of noise. Noise by definition is not predictable and has no pattern. Indeed it is one of the key facts that allows the present invention to work.

DRAWINGS

1. This shows a typical power line installation with two sensing devices

installed.

2. This shows the major elements of each sensing device.

3. This shows major elements of the correlation device and the presentation of the results to the user. DETAILED DESCRIPTION - LIST OF DRAWING ELEMENTS

10. Telephone poles

1 1. Telephone pole cross members.

12. Power line.

20. Monitoring unit A.

21. Monitoring unit B.

22. Connection of unit A to the power line.

23. Connection of unit B to the power line.

24. GPS antenna.

25. GPS antenna.

40. Filter/amplifier.

41. Sampling unit.

42. Memory.

43. Communication unit.

44. Global Positioning System (GPS) receiver.

45. Connection to power line.

46. GPS antenna.

47. Outgoing communication link.

60. Incoming data from monitoring unit A.

61 . Incoming data from monitoring unit B.

62. Communication unit.

63. Processing unit. 64. Display unit.

65. Correlation background noise.

66. Correlation peek.

67. Correlation graph.

68. Distance scale.

DETAILED DESCRIPTION - FIGURE 1

Figure 1 shows a typical power line installation. There are a series of telephone poles 10 with their cross-members 1 1. The power line being monitored is shown 12. There are two monitoring units; unit A 20 is at one end of the line and unit B 21 is at the other end of the line. The A monitoring unit 20 has a connection 22 to the power line 12. The B monitoring unit 21 also has a connection 23 to the power line 12. In this embodiment each monitoring unit has a global positioning system (GPS) antenna 24 and 25. There may be more than two monitoring units on a power line that work cooperatively but for this explanation we will consider only two.

The connections to the power line 22, 23 are shown as direct connections but with higher voltage installations a method of blocking the high voltage and only receiving the higher frequency noise is needed. One method would be to use a high voltage capacitor. One side of the capacitor would be connected to the power line and the other to the monitoring unit. The monitoring units are only interested frequencies above the electric distribution frequency so a capacitor would work well, blocking the high voltage but letting the high frequency noise through.

Another coupling method would be to use an antenna. This would be a conductor placed near the power line. It would be far enough away to prevent the high voltage jumping to it but close enough to couple to the radio frequency emissions of the noise on the line. This will work well because the input impendence of the monitoring unit can be made very high. Thus, it will be efficient at sensing the noise.

These two coupling methods are not shown for simplicity's sake and are well know in the industry.

Figure 1 shows only three telephone poles but an actual installation would likely have many more.

Even though telephone poles are shown in the drawings this invention is not limited to the relatively low voltages on conventional telephone poles. It will work equally well of the highest voltage power lines.

The present invention requires that time be synchronized between the sampling points represented by the two connections 22 and 23. In Figure 1 this is accomplished by using GPS receivers at both locations 24 and 25. One of the features present in many GPS receivers is an accurate time marker pulse. These receivers provide a pulse that occurs once a second and can be accurate to 20 nano seconds.

The means of synchronizing the two monitoring units can also be accomplished by other means such as a communication network with known timing characteristics. Power line installations often include a fiber optic cable that is used for communications with monitoring and protection equipment. This cable would have known timing characteristics and would be able to provide the necessary synchronizing information. It is the accuracy of timing that determines the precision with which the noise source can be located. This is not shown because the GPS timing method is a preferred embodiment.

DETAILED DESCRIPTION - FIGURE 2

Figure 2 shows the elements in each monitoring unit 20 and 21. The noise signal detected 45 at the power line is first filtered and amplified 40 then delivered to the sampling unit 41. The sampling unit 41 uses an analog to digital converter to sample the voltage of the noise signal at a high rate of speed. The sampling unit 41 synchronizes with timing information from the GPS unit 44. The resulting sampled data is saved in memory 42. When the sampling period is over, the communication unit 43 sends the data and timing information over the

communication channel 47 to the processing unit.

The filter and amplifier unit 40 limits the high frequency component of the noise signal detected 45 at the power line. It is necessary to limit the bandwidth of the noise to below the Nyquist limit of the sampling unit. Otherwise higher frequency components in the noise will reflect down into the bandwidth of interest and degrade the final correlated data. The amplification function adjusts the level of the noise signal to be compatible with the sampling unit input. The amplifier may be a logarithmic amplifier.

The rate of sampling is determined by the highest frequency of interest. That is, the sampling frequency must be at least double the frequency of the highest frequency of interest contained in the noise. Power lines do not propagate very J O

high frequencies for a great distance. The highest frequency of interest will likely be below 50 MHz. Therefore, removing higher frequencies removes very little useful information. If only noise of below 50 MHz is used then a sampling rate of 100 mega-samples per second is suitable. This results in a Nyquist limit of 50 MHz. The filter/amplifier 40 would be set to only pass frequencies below 50 MHz.

The timing of the sampling of the noise signal must be carefully controlled. The timing will be controlled by the Global Positioning System (GPS) receiver 44. The GPS receiver has an antenna 46 to receive the satellite signal. The GPS receiver 44 will provide precise timing signals to the sampling unit 41. GPS receivers are capable of generating a timing pulse at 1 pulse per second that is precise to 20 nano-seconds. They also provide the time of day at the same precision. The pulse is used to trigger the sampling unit which then begins sampling at 100 mega-samples per second until the sampling interval is over. The length of the sampling interval should be at least 2 times the propagation delay of the power line.

The sampled data and the timing information will be stored in memory 42. The memory 42 will hold the data until the sampling period is complete. The size of the memory is dependent on the length of the power line between the monitoring units 20 and 21. The memory should hold at least 2 times the propagation length. If, for example, the power line is 10 miles long and the sample rate is 100 mega- samples per second then the memory must hold at least 106,000 samples. This is assuming a propagation speed of about 1 foot per nano-second. I am neglecting the velocity factor of the power line in these rough calculations.

When the sampling process has completed the data and timing information from the memory 42 will be sent to the data analysis unit by communication device 43 using the outgoing communication link 47. The communication channel can be a radio link, a direct connection such as a fiber optic cable, or over the internet. The timing characteristics of this communication link 47 is not important since the critical timing data is stored in the memory 42 along with the sampled data.

DETAILED DESCRIPTION - FIGURE 3

Figure 3 shows the data analysis unit. This combines a communication receiver 62 with a data processing unit 63 and a display unit 64. The data produced by the monitoring units 20, 21 is received on the incoming

communication channels 60 and 61 . The communication receiver 62 forwards the data to the analysis unit 63. There can be multiple data analysis units but in this example we will describe one.

The analysis unit 63 takes the data from the two monitoring units A and B and performs a cross-correlation analysis. The range of the cross-correlation is from -T to +T. Where "T" is the length of the power line measured in terms of time. That is, the time for a signal to travel from one monitoring unit 20 to the other monitoring unit 21. For example, if the distance between monitoring units is 10 miles it will take about 53 microseconds for a signal to travel from monitoring unit A to monitoring unit B and the sample rate is 100 mega-samples per second then T = 53,000.

The length of the sample period should be at least 2 T. If the sampling period is longer, i.e. more samples, then the detected noise source will rise further out of the background noise. A system that ran non-stop where the sampling units ran continuously and the coirelation unit operated on the data as it streamed from the sampling units would be ideal.

The cross-correlation function will produce a dataset described by the discrete correlation function;

C(t) =∑A[t]B[T - f]

The data analysis unit 63 sends the cross-correlated data to the display unit 64. The display unit 64 plots the data as a graph 67. The X axis of the graph 67 represents the distance along the power line and is based on t. The X axis is demarked in distance 68 that represents the distance down the power line from the reference monitoring unit A.

The Y axis represents the amplitude of the noise detected. In this graph 67 there is a position where a high level of noise 66 is detected. This spike 66 rises above the noise floor 65. Such a graph 67 could indicate an arcing fault. The position of the spike 66 as measured on the scale 68 indicates the position on the power line where the fault is occurring.

Claims

I Claim;

1. A device for locating a fault on a power line comprising,

a. two or more data collection units at known locations, and b. coupling units between said data collection units and said power line, and

c. one or more communication devices, and

d. one or more synchronizing devices, and

e. one or more devices for producing correlated data, and

f. a device for presenting said correlated data to a user or monitoring system wherein the data from said data collection units is synchronized using said synchronizing device or devices and is communicated to said correlating device or devices and said correlated data indicating the distance to said fault is made available to said user or said monitoring system.

2. The device of claim 1 wherein said one or more synchronizing devices is one or more Global Positioning System receivers.

3. The device of claim 1 wherein said one or more synchronizing devices is one or more communication links.

4. The device of claim 1 wherein there is one or more memory units to store said data from said data collection units.

5. The device of claim 1 wherein said data collection devices are coupled to said power line through one or more capacitance devices.

6. The device of claim 1 wherein said data collection devices receive said

data by detecting the radiated electromagnetic emissions of said power line.

7. The device of claim 1 wherein there is a filter to prevent aliasing effects in said data collection units.

8. The device of claim 1 wherein the power line may be alternating current or direct current.

9. A method for finding the location of a fault on a wire comprising the steps of a. collecting sampled data from at least two points along said wire, and b. synchronizing said sampled data with one or more time

references to produce synchronized data, and

c. communicating said synchronized data to one or more

correlation processing units, and

d. con-elating said synchronized data to produce correlated data, and e. presenting said correlated data to one or more users or monitoring systems wherein said correlated data indicates the distance to said fault.

10. The method of claim 9 including the step of obtaining said time references from the Global Positioning System.

1 1. The method of claim 9 including the step of obtaining said time references from one or more communication links between said data collection points.

12. The method of claim 9 including the step of collecting data from said wire using analog to digital converters.

13. The method of claim 9 including the step of reducing aliasing in said

collected data by using a filter.

14. An apparatus for locating electrical faults on a wire comprising,

a. two or more analog to digital converters that sample the voltage on said wire, and

b. multiple memory devices to receive sampled data from said

analog to digital converters, and

c. a device or devices to synchronize said sampled data with a time reference signal to produce synchronized data, and

d. communication devices to send said synchronized data, and e. communication devices to receive said synchronized data for a processing device or devices, and

f. said processing device or devices to perform correlation on said synchronized data to produce correlated data, wherein said correlated data is presented to one or more users or monitoring systems and said correlated data contains an indication of the location of said faults.

15. The apparatus of claim 14 wherein said one or more synchronizing devices uses the Global Positioning System.

16. The apparatus of claim 14 wherein said one or more synchronizing devices is one or more communication links.

] 7. The apparatus of claim 14 wherein said analog to digital converters receive said voltage by detecting the electromagnetic emissions of said power line.

18. The apparatus of claim 14 wherein said analog to digital converters receive said voltage by using a capacitor connected to said power line.

19. The apparatus of claim 14 wherein there is a filter to prevent aliasing

effects in said analog to digital converters.