Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
TRAVELING WAVE BASED SINGLE END FAULT LOCATION
Document Type and Number:
WIPO Patent Application WO/2018/089598
Kind Code:
A1
Abstract:
Traveling wave information from a single end of an electric power delivery system is used to determine a fault location on a power line of the electric power delivery system. Hypotheses of which of a plurality of received traveling waves represents a first reflection from the fault are evaluated. A determination of an arrival time of the first reflection from the fault is used to calculate a distance from the single end of the power line to the fault location.

Inventors:
SCHWEITZER, Edmund O., III (330 NW Brandon Drive, Pullman, Washington, 99163, US)
GUZMAN-CASILLAS, Armando (305 SW Prairie Court, Pullman, Washington, 99163, US)
KASZTENNY, Bogdan Z. (26 Sir Galahad Place, Markham, Ontario L3P 2H6, 2H6, CA)
TONG, Yajian (1455 NE Brandi Way, Apt. FF303Pullman, Washington, 99163, US)
MYNAM, Mangapathirao Venkata (1400 Lehman Court, Pullman, Washington, 99163, US)
Application Number:
US2017/060800
Publication Date:
May 17, 2018
Filing Date:
November 09, 2017
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
SCHWEITZER ENGINEERING LABORATORIES, INC. (2350 NE Hopkins Court, Pullman, Washington, 99163, US)
International Classes:
G06F19/00; G01R13/04; G01R13/06; G01R13/08; G01R31/02
Foreign References:
US20130096854A12013-04-18
US20110227581A12011-09-22
US6798211B12004-09-28
Attorney, Agent or Firm:
EDGE, Richard M. (2350 NE Hopkins Court, Pullman, Washington, 99163, US)
Download PDF:
Claims:
1 . A system to determine a relative location of a fault in a power line of an electric power delivery system comprising:

a traveling wave detector to:

detect an arrival time (TPK) of each of a plurality of traveling wave amplitude peaks (VPKs) received at a first location on the power line, and

identify hypothesis TPKs as TPKs corresponding to VPKs detected after a first detected VPK that have the same polarity as the first detected VPK; and a fault location determination subsystem comprising at least one processor and a memory to:

calculate a fault reflection value, F(H), for each hypothesis TPK, calculate remote reflection value, R(H), associated with a remote location on the power line, for each hypothesis TPK,

compare the F(H) and the R(H) of each hypothesis TPK with a set of time difference values (DT set) that includes time difference values between the pair-combinations of detected TPKs associated with VPKs received after the first detected VPK,

determine a first number of matches (NM) as the number of time difference values in the DT set that match the F(H) of each hypothesis TPK, and determine a second number of matches (N1_M) as the number of time difference values in the DT set that match the R(H) of each hypothesis TPK; and a fault distance calculation subsystem to calculate a distance from the first location to a fault location based, at least in part, on:

(i) the length of the power line between the first location and the remote location,

(ii) a traveling wave line propagation time (TWLPT), and

(iii) one of the hypothesis TPKs selected based on a function of NM and NM_1 .

2. The system of claim 1 , wherein the traveling wave detector is configured to detect the TPKs of each of the VPKs received within an observation window.

3. The system of claim 1 , wherein the fault location determination subsystem is configured to calculate the F(H) of each hypothesis TPK as corresponding to a difference between the hypothesis TPK and the TPK of the first detected VPK. 4. The system of claim 1 , wherein the traveling wave detector is configured to detect the TPKs of each of the VPKs received after the first detected VPK relative to the TPK of the first detected VPK, such that the F(H) of a hypothesis TPK is equal to the hypothesis TPK.

5. The system of claim 1 , wherein the fault location determination subsystem is configured to calculate the R(H) of each hypothesis TPK as corresponding to twice the TWLPT less the F(H) value of the hypothesis, such that for each hypothesis TPK, R(H) = 2 * TWLPT - F(H). 6. The system of claim 1 , wherein the fault location determination subsystem is further configured to:

determine a set of expected traveling wave arrival times (ETWs) for each hypothesis TPK;

determine a third number of matches (NS), as the number of detected TPKs that match the ETWs for each hypothesis TPK;

assign a weight factor (WGHT) to each hypothesis TPK equal to:

the NS associated with each hypothesis TPK for which the associated R(H) corresponds to one detected TPK, and

zero (0) for each hypothesis TPK for which the associated R(H) value does not correspond to any detected TPK;

calculate a merit value (N) for each hypothesis TPK equal to the sum of the respective values of NM, N1_M, and WGHT, such that for each hypothesis TPK:

N = NM + N1_M + WGHT; and

wherein the fault distance calculation subsystem is configured to calculate the distance from the first location to the fault location based on the hypothesis TPK that has the highest N.

7. The system of claim 1 , wherein the fault location determination subsystem is configured to determine a match when two numbers are within a threshold range of one another. 8. The system of claim 1 , wherein the fault distance calculation subsystem is configured to calculate the distance from the first location to the fault location as one-half the length of the power line, multiplied by the selected hypothesis TPK, divided by the TWLPT. 9. The method of claim 1 , wherein the selected hypothesis TPK is the hypothesis TPK having the highest NM + N1_M.

10. A method of calculating a relative location of a fault in a power line of an electric power delivery system comprising:

detecting, via at least one sensor, an arrival time (TPK) of each of a plurality of traveling wave amplitude peaks (VPKs) at a first location on the power line;

identifying TPKs corresponding to VPKs detected after a first detected VPK that have the same polarity as the first detected VPK as hypotheses;

calculating, via a processor, a fault reflection value, F(H), for each hypothesis; calculating a remote reflection value, R(H), for each hypothesis;

determining a first number of matches (NM) of the F(H) of each hypothesis with a set of time differences (DT set) between pair-combinations of the detected TPKs received after the first detected VPK;

determining a second number of matches (N1_M) of the R(H) with the time differences in the DT set; and

calculating a location of the fault relative to the first location based on the length of the power line, the traveling wave line propagation time (TWLPT), and a selected hypothesis of the hypotheses.

1 1 . The method of claim 10, further comprising:

determining a set of expected traveling wave arrival times (ETWs) for each hypothesis;

determining a third number of matches (NS) of the ETWs of each hypothesis with the detected TPKs; and

assigning a weight factor (WGHT) to each hypothesis equal to:

the NS associated with each hypothesis for which the associated R(H) corresponds to one detected TPK, and

zero (0) for each hypothesis for which the associated R(H) value does not correspond to any detected TPK,

wherein the selected hypothesis is the hypothesis determined to have the highest merit value (N) equal to the sum of the respective values of NM, N1_M, and WGHT, such that for each hypothesis: N = NM + N1_M + WGHT. 12. The method of claim 10, wherein the selected hypothesis is the

hypothesis determined to have a highest merit value (N), where N is calculated for each hypothesis as a function of the associated NM and N1_M.

13. The method of claim 10, wherein calculating the F(H) of each hypothesis comprises determining a time difference between the hypothesis and the TPK of the first detected VPK.

14. The method of claim 10, wherein calculating the R(H) of each hypothesis comprises determining the difference between twice the TWLPT and the F(H) value of the respective hypothesis, such that for each hypothesis: R(H) = 2 * TWLPT - F(H).

15. The method of claim 10, wherein determining a first number of matches (NM) and determining the second number of matches (N1_M) comprises identifying numbers that are within a threshold range of one another.

16. The method of claim 10, calculating the distance from the first location to the location of the fault comprises multiplying one-half the length of the power line by the selected hypothesis divided by the TWLPT. 17. The method of claim 10, detecting, the TPKs of the VPKs at the first location on the power line comprises only detecting those VPKs with TPKs within an observation window.

18. The method of claim 17, wherein the observation window includes TPKs received before the TPK of the first detected VPK + (2.4 * TWLPT).

19. A computer-implemented method for calculating a relative location of a fault in a power line of an electric power delivery system comprising:

determining an initial guess of a relative location of a fault (m_ini) as being located within one of a first section of a power line, a second section of the power line, and a third section of a power line;

detecting, via at least one sensor, an arrival time (TPK) of each of a plurality of traveling wave amplitude peaks (VPKs) at a first location on the power line; and

identifying TPKs corresponding to VPKs detected after a first detected VPK that have the same polarity as the first detected VPK as hypotheses;

calculating, via a processor, a fault reflection value, F(H), for each hypothesis; calculating a remote reflection value, R(H), for each hypothesis;

determining a first number of matches (NM) of the F(H) of each hypothesis with a set of time differences (DT set) between pair-combinations of the detected TPKs received after the first detected VPK;

determining a second number of matches (N1_M) of the R(H) with the time differences in the DT set; determining a set of expected traveling wave arrival times (ETWs) for each hypothesis;

determining a third number of matches (NS) of the ETWs of each hypothesis with the detected TPKs;

assigning a weight factor (WGHT) to each hypothesis equal to:

the NS associated with each hypothesis for which the associated R(H) corresponds to one detected TPK, and

zero (0) for each hypothesis for which the associated R(H) value does not correspond to any detected TPK;

calculating a merit value (N) for each hypothesis that corresponds to the sum of the respective values of NM, N1_M, and WGHT, such that for each hypothesis:

N = NM + N1_M + WGHT;

sorting the hypothesis in a priority order based on:

NM from highest to lowest, where m_ini is determined to be within the first section of the power line,

N from highest to lowest, where m_ini is determined to be within the second section of the power line, and

N1_M from highest to lowest, where m_ini is determined to be within the third section of the power line; and

calculating a distance from the first location to the fault location based on a function of the length of the power line, the traveling wave line propagation time

(TWLPT), and the hypothesis sorted to have the highest priority.

20. The computer-implemented method of claim 19, wherein the m_ini is determined to be within one of the first section of a power line, the second section of the power line, and the third section of a power line based on information received from one of: a double-ended traveling wave fault location estimation system, a double-ended impedance fault location estimation system, and a single-ended impedance fault location estimation system.

Description:
TRAVELING WAVE BASED SINGLE END FAULT LOCATION

TECHNICAL FIELD

[0001] This disclosure relates to determining a location of a fault using single-end traveling wave information in an electric power delivery system.

BRIEF DESCRIPTION OF THE DRAWINGS

[0002] This disclosure includes illustrative embodiments that are non-limiting and non-exhaustive. Reference is made to certain of such illustrative embodiments that are depicted in the figures described below.

[0003] FIG. 1 A illustrates an example of an intelligent electronic device (IED), according to several of the embodiments described herein.

[0004] FIG. 1 B illustrates a block diagram of the inputs, parameters, and outputs of a single-ended traveling wave fault location system, such as may be implemented by the IED in FIG. 1A.

[0005] FIG. 2 illustrates an example of a Bewley lattice diagram that includes a substation, local terminal, fault location, and a remote terminal.

[0006] FIG. 3 illustrates a flowchart including inputs and parameters for performing operations of a method for calculating a fault location.

[0007] FIG. 4 illustrates a flowchart of an example of a method for calculating a fault location.

[0008] FIG. 5 illustrates a flowchart of portion of a method for calculating a fault location.

[0009] FIG. 6A illustrates a Bewley lattice diagram showing TPK(0), F(1 ), and R(1 ) based on a hypothesis associated with TPK(1 ), according to one embodiment.

[0010] FIG. 6B illustrates a Bewley diagram with traveling waves in dashed lines that originate from substation reflections, according to one embodiment.

[0011] FIG. 7 illustrates a portion of a method for calculating expected traveling waves for each hypothesis.

[0012] FIG. 8A illustrates an example of expected traveling waves from a

fault-local-fault-local traveling wave reflection pattern in a Bewley lattice diagram.

[0013] FIG. 8B illustrates an example of expected traveling waves from a

fault-remote-fault-local traveling wave reflection pattern in a Bewley lattice diagram. [0014] FIG. 8C illustrates an example of expected traveling waves from a local-fault-remote-fault-local traveling wave reflection pattern in a Bewley lattice diagram.

[0015] FIG. 9 a plurality of expected traveling waves at a local terminal for a vector of expected traveling wave arrival times, Th.

[0016] FIG. 10 illustrates a portion of a method for calculating a fault location based on evaluated hypotheses, according to one embodiment.

[0017] FIG. 1 1 provides a schematic diagram for a simulation to illustrate a specific example of the systems and methods described herein, according to one embodiment.

[0018] FIG. 12 illustrates the alpha currents, smoothed via a differentiator smoother, for the simulated fault with the corresponding VAL or FAL indications for each peak.

[0019] FIG. 13 illustrates a table of a calculated modal offset value, IMOFF.

[0020] FIG. 14 illustrates a table with the interpolated arrival time stamps (ITPIs), and their corresponding estimated signed magnitudes, IMAG, for the simulation.

[0021] FIG. 15 illustrates a table with interpolated arrival times (TPKs) referenced to the arrival time (TPK) of the first received peak.

[0022] FIG. 16 illustrates a table showing the sign, TPSIGN, of each peak with respect to the first VPK based on the signed value VPK of each peak.

[0023] FIG. 17 illustrates a table showing the number matches within a set of time differences of pair-combinations of TPK values for the F(H) and the R(H) of each hypothesis.

[0024] FIG. 18 illustrates a table showing calculated total errors for NM(H) and N1_M(H) within a TWTOL1 parameter.

[0025] FIG. 19 illustrates a table showing the first reflection from the fault underlined, the first reflection from remote station double underlined, and filtered values in strikethrough.

[0026] Figure 20 illustrates a graphical representation of the measured TPKs and the arrival times of expected traveling waves (ETWs) for three hypotheses with matches identified by circles.

[0027] FIG. 21 illustrates a table showing the scores, NS(H), based on TPK matches with the ETWs for each hypothesis.

[0028] FIG. 22 illustrates a table of the total errors for the scores, NS(H), within a 5 s threshold range. [0029] FIG. 23 is a table showing a weight, WGHT(H), assigned to each of three hypotheses based on an R(H) match with a received TPK value and the NS(H) of each hypothesis.

[0030] FIG. 24 includes a table with N values and the total per unit error, ERRpu, for each of the three hypotheses.

[0031] FIG. 25 shows the expected reflections for a fault at m=0.3 and the

corresponding TPK measurements as circles.

DETAILED DESCRIPTION

[0032] The systems and methods described herein relate to determining a location of a fault in an electric power delivery system based on measurements from a single end of the electric power delivery system. Any of a wide variety of systems and methods may be used to obtain current and/or voltage measurements from an electric power delivery system. The system may calculate traveling wave information, such as polarity, magnitude, and/or time of traveling waves observed at a local end (i.e., first location) of the electric power delivery system. The system may identify a plurality of the received traveling waves as being potentially the first received reflection from the fault location.

[0033] A traveling wave detector, comprising software, hardware, and/or a

combination thereof may be configured to detect arrival times (TPKs) of traveling wave peaks (VPKs). Each of the plurality of TPKs associated with the VPKs that match a polarity of the first received VPK may be considered and evaluated (i.e., considered as a hypothesis) to determine if it corresponds to the arrival time of the first received traveling wave peak reflection from the fault location.

[0034] The system may utilize one or more types of fault location information, including any one or combination of (i) multi-end traveling wave fault location

information, (ii) multi-end impedance-based fault location information, and (iii) single-end impedance-based fault location information to establish an initial guess, m_ini, of a location of the fault.

[0035] For example, if multi-end traveling wave fault location information is available, an initial estimate of the fault location, m_ini, may be made based on the multi-end traveling wave information; otherwise, if multi-end impedance-based fault location is available, an initial estimate of the fault location, m_ini, may be made based on the multi-end impedance-based fault location; otherwise, if single-end impedance-based fault location information is available, an initial estimate of the fault location, m_ini may be made based on the single-end impedance-based fault location information.

[0036] If none of these three types of fault location information types are available, an initial estimate of the fault location, m_ini, may be assigned by default. For example, m_ini may be set equal to 0.5 in per unit length. In some cases, the single-end traveling wave fault location method may use the initial estimate to determine a possible faulted section.

[0037] Generally speaking, traveling waves couple between faulted and healthy phases as they travel along the line in two sets of aerial modes (alpha and beta) and a ground mode. The alpha mode can be a good representation of the three phase traveling waves for ground faults and can be calculated for any faulted phase. The beta mode can be a good representation of the three phase traveling waves for phase to phase faults. The mode (alpha or beta) with the highest current magnitude (or alternatively, voltage in some embodiments) may be used. In some embodiments, a specific mode may be used to detect a specific fault type (e.g., if a fault type is known or identified by a fault detection system).

[0038] As described below, arrival times (TPKs) of traveling wave peaks (VPKs,) received at a local terminal may be used to calculate a fault location. The system may identify the reflection path of VPKs detected at the local terminal by matching patterns and/or expected arrival times based on an evaluation of hypotheses, where each hypothesis is considered as a potential TPK corresponding to the VPK from the first reflection from the fault location back to the local terminal. A hypothesis for the fault location that results in accurate predictions of traveling wave peak arrival times (TPKs of VPKs) can be used to calculate an exact or approximate fault location relative to the local terminal.

[0039] In various embodiments, a ruler evaluation (i.e., a repeating travel time "RTT" evaluation) can be used to evaluate each hypothesis based on the expected and received traveling waves. In other embodiments, a score evaluation (i.e., an expected traveling wave "ETW" evaluation) approach can be used to evaluate each hypothesis based on the expected and received traveling wave. In still other embodiments the ruler (RTT) and score (ETW) methods may be used in combination. [0040] In some embodiments, the hypotheses may be weighted and evaluated based on the degree of match (i.e. , an error rate) between calculated or expected traveling waves and measured or received traveling waves. Hypotheses may be rejected based on the number of matches being below a threshold (such as where the number of matches is zero). A fault location can be calculated based on the hypotheses with expected reflections that most closely match received reflections.

[0041] The systems and methods described and illustrated herein may be

implemented in hardware specifically configured for the monitoring and protection of electric power delivery systems. In one embodiment, the systems and methods herein may be implemented as intelligent electronic devices (lEDs) in communication with the electric power delivery system. An IED according to several embodiments herein may include inputs for receiving electric power system signals corresponding with the electric power delivery system. For example, the IED may include an input in electrical communication with a current transformer, potential transformer, or other similar device configured to obtain electric power system signals from a portion of the electric power delivery system and transmit the signals to the IED.

[0042] Additional understanding of the embodiments of the disclosure can be gained by reference to the drawings, wherein like parts are designated by like numerals throughout. It will be readily understood that the components of the disclosed

embodiments, as generally described and illustrated in the figures herein, could be arranged and designed in a wide variety of different configurations. Thus, the following description of the embodiments of the systems and methods of the disclosure is not intended to limit the scope of the disclosure, as claimed, but is merely representative of possible embodiments of the disclosure. In addition, the steps of a method do not necessarily need to be executed in any specific order, or even sequentially, nor need the steps be executed only once, unless otherwise specified or contextually required.

[0043] FIG. 1 A illustrates an example of an IED 100 according to several of the embodiments described herein. The IED 100 includes a voltage input for receiving electric power system signals from a potential transformer or the like, and a current input for receiving electric power system signals from a current transformer or the like. The signals may be received by a signal processing module 1 10 that may include transformers 102, 1 14, A/D converter(s) 1 18, filters, and the like to produce electric power system signals useful for the processor 124 via bus 122. The IED 100 may include a time input 1 12 configured to receive a common time signal among a certain subset of lEDs and other devices. The common time signal may be a global navigation satellite system (GNSS) time signal, a terrestrial common time signal, a network time signal, or the like.

[0044] The IED 100 may include a monitored equipment interface 108 for communication with power system equipment. The monitored equipment interface 108 may be configured to interface with a circuit breaker, recloser, capacitor bank, voltage regulator, reactor, or the like for controlling operation of the equipment. The IED 100 may include a communications interface 1 16 for communication with other lEDs, or other devices. In one embodiment, IED 100 may be in communication with local equipment at a local location of the electric power system via the voltage and current inputs, while being in communication via the communication interface 1 16 with another IED at a remote location of the electric power delivery system. Thus, IED 100 may obtain electric power system conditions, operation information, and other

communications related to a remote location of the electric power delivery system.

[0045] The IED 100 may include a computer-readable storage medium 126 for permanent and temporary storage of information. The computer-readable storage medium 126 may include a database 128, that may be a repository of settings, thresholds, and the like, useful for the operation of the IED 100. The IED 100 may further include a computer-readable storage medium 130 that includes several operating modules that when executed on the processor 124 cause the IED 100 to perform certain functions related to the monitoring and protection of the electric power delivery system. Each of the signal processing module 1 10, computer readable storage media 126, 130, time input 1 12, monitored equipment interface 108, communications interface 1 16, processor 124, and other modules may be in communication using a data bus 142.

[0046] The computer-readable storage medium 130 may include several modules operable on the processor 124. One such module may be a communication module 132 that includes instructions related to the transmission and receipt of

communications. The fault detector module 134 may include instructions related to detecting faults on the electric power delivery system using measurements obtained by the signal processing module 1 10 and/or from another IED via the communications interface 1 16. The fault type module 138 may be configured to determine a fault type (e.g. single-phase, multiple-phase, phase-to-phase, phase-to-ground, and the like) from available fault information. The data acquisition module 140 may include instructions related to treatment of the signals from the signal processing module 1 10 to produce data useful for other modules.

[0047] The traveling wave location module 144 may include instructions and/or electrical components to implement several of the systems, subsystems, and

operations described herein. For example, traveling wave location module may be configured to locate a fault using traveling waves from a single end of the electric power delivery system. The traveling wave location module 144 may be embodied within the computer-readable storage medium 130 and/or implemented as a stand-alone IED with subcomponents, modules, and inputs implemented in software, firmware, and/or hardware.

[0048] The protection action module 152 may include instructions for taking a protective action based on a detected fault and fault location. For example, for a fault within a selected zone of protection, the protection action module 152 may be configured to command a circuit breaker to open (via the monitored equipment interface 108) to remove power from the fault.

[0049] A fault locator system may utilize a traveling wave single-end fault location (TWSEFL) algorithm to estimate a fault location using single-end traveling wave measurements and evaluations. The TWSEFL algorithm may use fault location estimations from the multi-end traveling wave fault location (TWMEFL) information and/or impedance-based (ZFL) algorithms to develop an initial guess, m_ini. If the estimations of TWMEFL algorithm or ZFL algorithm are not available, the TWSEFL algorithm can still be used to estimate the fault location based on the traveling wave arrival times at a local terminal using an initial guess, m_ini, set to a default value (e.g., 0.5 in per unit length).

[0050] FIG. 1 B illustrates a block diagram 170 of the inputs 180, parameters 185, and outputs 190 of the TWSEFL algorithm 175 that may be implemented by a fault locator system (e.g., IED 100 in FIG. 1 A, as part of traveling wave location module 144 in FIG. 1A, and/or as an independent system). Inputs 180 may include any combination of: a vector of peak times (TPK or ILTPIC), a traveling wave based multi-end fault location (TWMEFL) estimate, an impedance-based multi-end fault location (ZMEFL) estimate, an impedance-based single-end fault location (ZSEFL) estimate, and a phase-to-phase fault (PPFLT) indicator.

[0051] Parameters 185 may include any combination of: traveling wave line propagation time (TWLPT), line length (LL), single-ended traveling wave fault location observation window (TWOBSW), single-ended traveling wave fault location observation window 1 (TWTOL1 ), and single-ended traveling wave fault location observation window 2 (TWTOL2). In some embodiments, some parameters may be user-input based on system conditions and some parameters may be dynamically calculated by the system instead of provided as an input. For example, the TWOBSW may be automatically set to 2.4, such that the observation window is equal to 2.4 times a TWLPT parameter. Similarly, the TWLPT parameter may be input as a parameter and/or may be automatically calculated based on an input LL and known system conditions/materials. Outputs 190 of the TWSEFL algorithm may include one or more traveling wave based single-end fault location estimations (TWSEFL1 , TWSEFL2, TWSEFL3, and TWSEFL4). Alternatively, a single traveling wave single-ended fault location, TWSEFL, may be output based on the hypothesis determined to be most accurate.

[0052] FIG. 2 illustrates a Bewley lattice diagram 200 that includes a local terminal, L, 210 and a remote terminal, R, 250 that are 1 line length unit apart. A fault at location 220 is at a relative distance m (per unit) from the local terminal 210 and a relative distance 1 -m from the remote terminal 250. A substation B 240 is also shown. The Bewley lattice diagram 200 illustrates the traveling current waves from the fault location 220 and the various reflections from the local terminal 210 and the remote terminal 250.

[0053] ti is the arrival time (i.e. , TPK) of the first received traveling wave from the fault at location 220. t2 is the arrival time of a traveling wave from the substation 240 behind the local terminal 210. t3 is the arrive time of a traveling wave that traveled from the fault at location 220 to the remote terminal 250 and was reflected back to the local terminal 210. t 4 is the arrival time of the first traveling wave after having been reflected by the local terminal 210, back to the fault at location 220, and then back to the local terminal 210. ts is related to reflections between the substation 240 the fault at location 220. Finally, ίβ is the arrival time of a traveling wave that traveled from the fault at location 220 to the local terminal 210, from the local terminal 210 to the remote terminal 250, and from the remote terminal 250 back to the local terminal 210. [0054] The arrival times of ti and t 4 are considered "companion traveling waves" because the relative time between their arrival times (TPKs) is calculatable based, at least in part, on the distance between the local terminal and the remote terminal 250. By identifying the origin or reflection path of received traveling waves and/or traveling wave patterns, the system can estimate an accurate distance between the local terminal 210 and the location of the fault 220.

[0055] The fault at location 220 is a distance m (in per unit length) from the local terminal 210, so the first traveling wave is received at time ti that is equal to m * TWLPT, where TWLPT is the traveling wave line propagation time. Another traveling wave should return from the remote terminal 350 at the time t3 that is equal to (2-m) * TWLPT. The difference between these two times is 2*(1 -m)*TWLPT. Thus, a

"companion traveling wave" is expected at ti + 2*(1 -m)*TWLPT, which is equal to t3. The first reflection from the fault is expected at time t 4 , which is equal to ti + 2 * m * TWLPT.

[0056] The timing and polarities of the traveling waves that are spaced by 2 * m * TWLPT and 2 * (1 -m) * TWLPT may be evaluated. Each reflection from a discontinuity behind the local terminal 210 (e.g. , from substation 240) may generate a traveling wave (a "test traveling wave") toward the fault. Consequently, the system may expect multiple pairs of traveling waves that are spaced by 2 *m * TWLPT. Multiple possible traveling wave pairs may be evaluated to identify the distance to the fault, m. The time distance between pairs of traveling waves that occurs most frequently may be determined to be equal to 2 * m * TWLPT.

[0057] The system may implement the TWSEFL algorithm by identifying a plurality of valid traveling waves at the local terminal 210 that are within an observation window (TWOBSW) that is at least two times the TWLPT (e.g. , 2.4 * TWLPT). The first TPK value at ti may be referred to as TPK(0).

[0058] FIG. 3 illustrates inputs 320 and parameters 330 received for performing operations of a method 300 to calculate a fault location. For example, a fault location system may implement the TWSEFL algorithm by identifying arrival times (TPKs) (i.e. , time stamps) of traveling wave amplitude peaks (VPKs) of traveling waves that correspond to the mode that includes the maximum amplitude within an observation window defined by a function of the TWOBSW parameter and the TWLPT (e.g. , TWOBSW * TWLPT). For example, the TWOBSW may be 2.4 such that the observation window is equal to 2.4 times the traveling wave line propagation time, TWLPT.

[0059] A first TPK value corresponding to the first received VPK within the observation window can be used as a reference traveling wave, and all other TPK values can be referenced, at 340, with respect to TPK(0), such that for each of the TPK values TPK(1 ) to TPK(n), a referenced TPK value can be calculated as follows for each value of X for X=1 to n where n is the number of reflections:

[0060] TPK Ref (X) = TPK(X) - TPK(0) Equation 1

[0061] A TPK vector (or set of TPK values) can be formed that includes the referenced TPK values below a threshold, tnmit. Input parameters 330 can used to establish the threshold, tnmit, as, for example:

[0062] t limit = m QTWOBSW * TWLPT, 10000 ^s) Equation 2

[0063] A set of hypotheses, Hi - H n , may be generated based the values in the TPK vector corresponding to VPKs that have the same polarity as the VPK corresponding to TPK(0). In some embodiments, up to 15 hypotheses (or other arbitrary maximum number of hypotheses) may be initially considered, at 350, as hypotheses for calculating the fault location. The system evaluates each of a plurality of hypothesis TPK values to determine if a hypothesis TPK value corresponds to the first VPK associated with a reflection from the fault. Identify a TPK value as corresponding to the arrival time of a VPK of the first traveling wave reflection from the fault allows for an accurate hypotheses that have a time difference relative to TPK(0) greater than a function of the TWLPT may be discarded. For example, hypotheses that have a time difference relative to TPK(0) greater than 2 TWLPT + 10 microseconds may be discarded. If no hypotheses exist, at 360, then no estimated TWSEFL may be generated, at 365.

[0064] If there are a plurality of hypotheses and inputs for TWMEFL, ZMEFL or ZSEFL are available, at 370, then one or more of the inputs for TWMEFL, ZMEFL or ZSEFL may be used to as an initial estimate of a fault location, m_ini. In various embodiments, the following priority may be utilized for determining, at 380, an initial estimate of the fault location, mjni: TWMEFL, ZMEFL, and then ZSEFL. In other embodiments, a different priority may be utilized and/or a weighted average of estimates based on one or more of TWMEFL, ZMEFL, and then ZSEFL. If none of TWMEFL, ZMEFL or ZSEFL is available, at 370, then a default initial estimate may be used, at 385 (e.g., the middle of the line).

[0065] FIG. 4 illustrates an example of a method 400 of one embodiment of calculating a fault location TWSEFL1 . As described above, if TWMEFL is available, at 410, then the hypothesis for the fault location (based on a hypothesis TPK as described herein) that is closest to the TWMEFL estimation is used, if it is within the TWTOL2 parameter, at 415. If a TWMEFL estimation is not available, at 410, but a ZMEFL estimate is available, at 420, then the system may utilize the hypothesis TPK that results in a fault location calculation that is within a predetermined range (e.g. , 3%) of the ZMEFL estimate, at 425. If both the TWMEFL and ZMEFL estimations are unavailable, but a ZSEFL estimation is available and a phase-to-phase fault is indicated, at 430, then the hypothesis that results in a fault location calculation within a predetermined range (e.g. , 5%) of the ZSEFL estimate may be utilized.

[0066] As long as the hypothesis are within allowable limits, at 450, based on input parameters, then the selected hypothesis is used to calculate the fault location

TWSEFL1 , at 460. If none of the three estimations (TWMEFL, ZMEFL, and ZSEFL) are available, at 445, and/or none of the hypotheses are not within allowable limits, at 455, then a fault location may be estimated based on a selected hypothesis TPK as described below.

[0067] In other embodiments, estimates and/or inputs from TWMEFL, ZMEFL, and/or ZSEFL may not be provided and/or considered in determining a fault location. In such instances, a fault location may be calculated based on a selected hypothesis as described below.

[0068] FIG. 5 illustrates a portion of a method 500 for calculating a fault location using a ruler evaluation approach, also referred to as a repeating travel time RTT evaluation approach. For each hypothesis TPK, a fault reflection value corresponding to 2 * (1 -m) may be calculated, at 502, as the difference between the hypothesis TPK value and TPK(0), where TPK(0) is the arrival time of the first received traveling wave peak (VPK). For a given hypothesis, H, the equation may be represented as:

[0069] F(H) = TPK(H) - TPK 0). Equation 3

[0070] For each hypothesis, H, a remote reflection value corresponding to the arrival time (TPK) of a traveling wave peak (VPK) from a remote location (e.g. , the second end of the power line) may be calculated, at 504, for a given hypothesis, H, that corresponds to 2 * (1 -m) as:

[0071] R(H) = 2 * TWLPT— F(H). Equation 4

[0072] FIG. 6A illustrates a Bewley lattice diagram showing TPK(0) 61 1 , F(1 ) 612, and R(1 ) 613, where F(1 ) 612 and R(1 ) 613 are based on the first hypothesis TPK(1 ). The time difference between TPK(1 ) and TPK(0), F(1 ) is equal to 2 * m * TWLPT, per Equation 1 above. R(1 ) is a calculated using Equation 4 as an expected reflection assuming the hypotheses TPK(1 ) is a correct hypothesis. The expected R(1 ) 613 reflection can be thought of as a "companion traveling wave" that, assuming the hypothesis TPK(1 ) is accurate, should correspond to one of the received TPK values. If no such TPK value is detected (i.e., no match is found), then a confidence level that they hypothesis is correct is decreased, as described below.

[0073] Returning to FIG. 5, the system may create a time difference vector, DT, that include all of the possible time differences using all TPKs in the observation window. That is, a set of time differences (DT set) or a time difference vector (DT) may be generated to include a plurality of difference elements. The difference elements included in the DT set (or DT vector) include all difference between all

pair-combinations of TPKs received after the TPK of the first received VPK. As a specific example: if the TPKs received during the observation window are A, B, C, D, E, and F, where A corresponds to the arrival time of the first traveling wave peak (VPK), then the DT set will include differences: (A-B), (A-C), (A-D), (A-E), (A-F), (B-C), (B-D), (B-E), (B-F), (C-D), (C-E), (C-F), (D-E), (D-F), and (E-F). The differences in the set may be expressed as absolute values and/or evaluated for matches independent of sign.

[0074] The system may determine, at 508, a count of the number of instances, NM(H), that the time differences in the vector DT match the calculated F(H). A match may be defined as a number that is within a predefined tolerance range, as defined by, for example, the TWTOL1 parameter (e.g. , TWTOL1 =10). A first error, ERRM(H), may be calculated, at 510, for each hypothesis that corresponds to the NM(H) set.

[0075] The fault location system may determine, at 512, the number of instances, N1_M(H), that the time differences in the vector DT match the calculated R(H). Again, a match may be considered valid only if it is an exact match or if numbers are within a predefined tolerance range. A second error, ERR1_M(H), may be calculated, at 514, for each hypothesis that corresponds to the N1_M(H) set. A total error, ERR(H), may be calculated, at 516, by adding ERRM(H) and ERR1_M(H).

[0076] FIG. 6B illustrates traveling waves in dashed lines that are caused by reflections for a substation 640 that the system may utilize to determine the number of instances NM(H) and N1_M(H). One benefit of the RRT method described above is that the approach leverages available reflections from the external elements behind the local terminal (e.g. , substation 240, FIG. 2).

[0077] Without exhaustively repeating the discussion above for the TPKs associated with the dashed lines in FIG. 6B, it is visually clear that the TPKs at the local terminal 610 corresponding to the dashed lines are similar to the solid lines discussed above for TPK(0) 61 1 , F(1 ) 612, and R(1 ) 613, except the dashed lines are shifted downward in time by an amount of time equal to two times the traveling wave propagation between the local terminal L 610 and the point B in the substation 640.

[0078] FIG. 7 illustrates a method for calculating expected traveling waves (ETW) for each hypothesis. FIG. 7 also includes operations for evaluating each hypothesis using weighted values from the RRT and ETW approaches. The system may implement the ETW approach by calculating, at 718, a vector of expected traveling wave arrival times, Th, for each hypothesis. A set of ETWs may be used as an equivalent or similar data structure for a number of ETW elements. The vector of expected traveling wave arrival times, Th, may be based on: the distance to the fault, m; the TWLPT parameter; and the threshold, tnmit, from Equation 2. A vector of expected traveling wave arrival times, Th, may be calculated for each hypothesis based on a number of traveling wave sets as, for example:

[0079] = [TLSET1, TLSET2, TLSET3, TLSET4] - TLSET1 (1 ) Equation 5

[0080] The vector of expected traveling wave arrival times, Th, may include only those expected traveling waves within each of TLSET1 ,TLSET2,TLSET3, and TLSET4 that are within an observation window. TLSET1 corresponds to a pattern of TPK values for reflections from the fault to the local terminal, back to the fault, and then back to the local terminal (i.e., fault-local-fault-local).

[0081] FIG. 8A illustrates a fault-local-fault-local traveling wave reflection pattern with the fault at location m=0.3 pu on a 100-mile line (i.e. m= 30 miles) with the TWLPT equal to 540 s.Thus, the system may calculate TLSET1 as follows: [0082] TLSET1 = k * m * TWLPT, k = 1, 3, 5, ... Equation 6

[0083] TLSET2 corresponds to the pattern of TPK values for reflections from the fault to the remote terminal, back to the fault, and then to the local terminal (i.e., fault-remote-fault-local).

[0084] FIG. 8B illustrates a fault-remote-fault-local traveling wave reflection pattern with similar parameters to the line in FIG. 8A. Thus, the system may calculate TLSET2 as follows:

[0085] TLSET2 = 2 * k * (1 - m) * TWLPT, k = 1, 2, 3, ... Equation 7

[0086] TLSET3 corresponds to the pattern of TPK values for reflections from the local terminal to the fault, to the remote terminal, back to the fault, and to the local terminal (i.e., local-fault-remote-fault-local).

[0087] FIG. 8C illustrates a local-fault-remote-fault-local traveling wave reflection pattern with similar parameters to the line in FIG. 8A. The solid line represents the first TLSET3 pattern. The dashed lines represent subsequent TLSET3 patterns. The bold lines with arrows show the reflections from the fault to the local terminal that are not part of TLSET3. Thus, the system may calculate TLSET3 as follows:

[0088] TLSET3 = [k * m + 2] * TWLPT, K = 1, 3, 5, ... Equation 8

[0089] TLSET4 corresponds to the pattern of TPK values for reflections from the local terminal to the fault, to the remote terminal, back to the fault, back to the remote terminal, back to the fault, and then to the local terminal (i.e., local-fault-remote-fault- remote-fault-local). Without belaboring the point with yet another drawing, the system may calculate TLSET4 as follows:

[0090] TLSET4 = [k * M + 2(2 - M)] * TWLPT, K = 1, 3, 5, ... Equation 9

[0091] Once the elements of TLSET1 , TLSET2, TLSET3, and TLSET4 have been identified, the vector of expected traveling wave arrival times, Th, (or a set of ETWs) may be generated by filtering out elements with the same value, and elements that have values outside of the observation window. The elements in the vector of expected traveling wave arrival times, Th, may be sorted in ascending order to facilitate comparisons and/or decrease computations processing times.

[0092] FIG. 9 illustrates all the expected TPK values in the vector of expected traveling wave arrival times, Th, for a hypothesis TPK associated with a fault location estimated at m=0.3. [0093] Turning again to FIG. 7, the fault location system may then determine, at 720, a count of the number of TPKs in the TPK vector that match the vector of expected traveling wave arrival times, Th. Again, a match may be defined as an exact match or as a match within a tolerance range, such as a tolerance ranged defined by the

TVVTOL2 parameter. The number of matches or scores may be referred to as NS(H). in some embodiments, an event error, ERR_S(H) may be calculated as well.

[0094] The results of the RTT and/or EWT approaches may be evaluated to select a hypothesis as correct. The system may select a correct hypothesis based on a weighted function of the results of the RTT and/or EWT approaches and may optionally consider the respective error calculations. In embodiments that use both the ruler and the score evaluation approaches, the system may assign, at 722, a weighting factor WGHT(H) equal to NS(H) if the R(H) of each respective hypothesis corresponds to a TPK value within the TPK vector. Otherwise, the WGHT factor for a given hypothesis may be assigned to zero/null

[0095] For each hypothesis, a sum N(H) may be calculated, at 724, as follows:

[0096] N(H) = NM(H) + N1_M(H) + WGHT(H). Equation 10

[0097] An error per event may be calculated, at 726, as follows:

[0098] ERRpu(H) = ( ERR W+™ R - S ^ Equation 11

[0099] In other embodiments that use only the score or ETW evaluation approach, Equation 10 may be modified to ignore NM(H) and N1_M(H) and the weighting factor WGHT(H) may be unrelated to the match of R(H) values with TPK values within the TPK vector. Thus, for each hypothesis being evaluated by only the score evaluation approach, an equation for N(H) may be expressed as:

[00100] N(H) = NS(H) Equation 10. 1

[00101] A fault location error of the initial estimate, FLERR, may be determined, at 728. For example, the FLERR may be determined as the greater of either 0.2 pu or three times the reciprocal of the line length. The power line may be conceptually divided, at 730, into three sections. The initial guess, m_ini, can be defined as being in one of the three sections as follows, for example:

Section 1 : m_ini < (1 /2-FLERR);

Section 2: (1 /2-FLERR) <= mjni <= (1 /2+FLERR); and

Section 3: m_ini > (1 /2+FLERR). [00102] FIG. 10 continues the method of FIG. 7 and branches, at 1032, based the section within which m_ini is located. For instance, if m_ini is in Section 1 , at 1034, then the hypothesis with the greatest NM is used. If there are equal values of NM, N1_M can be used to order the hypotheses in descending order. If there are an equal number of N1_M values, then the ERRpu can be used to order the hypothesis in ascending order.

[00103] If m_ini is in Section 3, at 1036, the hypothesis with the greatest N1_M is used. If there are equal values of N1_M, NM can be used to order the hypotheses in descending order. If there are an equal number of NM values, then the ERRpu can be used to order the hypothesis in ascending order.

[00104] If m_ini is in Section 2, at 1038, the hypothesis with the greatest N is used. If there are equal values of N, then the ERRpu can be used to order the hypothesis in ascending order.

[00105] The fault location system can then calculate, at 1040, TWSEFLn, based on the ordered hypotheses, where n=1 , 2, 3 and 4 as:

[00106] TWSEFLn = LL*DTn Equation 12

1 J 2*TWLPT ^

[00107] Using the F(H) notation described above, Equation 12 can alternatively be expressed as:

[00108] TWSEFLn = LL*F(H) Equation 12. 1

1 J 2*TWLPT ^

[00109] If m_ini is in Section 1 , the calculated distance to the fault, m, should be less than 0.3, or it is set to N/A. If m_ini is in section 3, the calculated distance to the fault should be between 0.7 and 1 .0, or it is set to N/A.

[00110] FIG. 1 1 provides a schematic diagram 1 100 for a simulation to illustrate a specific example of the systems and methods described herein. The description below uses actual data from a simulation of the schematic diagram 1 100 and is intended to provide a specific example of the systems and methods described above and is not intended to be limiting in any way.

[00111] The schematic diagram 1 100 simulates an AG fault at 30% of a 100-mile transmission line with a TWLPT of 540 s. The schematic diagram 1 100 also includes two 25-mile parallel lines connected to the two-line terminals (TWLPT=135us) and two sources behind the parallel lines.

[00112] FIG. 12 illustrates the alpha currents 1200 after being processed by a differentiator smoother for the simulated fault with the corresponding VAL or FAL indication for each peak. VAL or FAL value of logic one indicates that the traveling wave peak is a suitable input for the TWSEFL algorithm.

[00113] The data acquisition circuitry (including, for example, data acquisition module 140, FIG.1A) obtains the time stamps from alpha signal A 1201, alpha signal B 1202, and alpha signal C 1203. For this line, the observation window of the data buffer may be, for example, 2.4 * TWLPT (2.4 * 540ps =1296ps). For the signals in FIG.12, the data buffer provides the following outputs:

[00114] For alpha signal A 1201:

IATP = [51074, 51344, 51398, 51613, 51668, 51830, 51883, 51938, 51992, 52100, 52207, 52262].

IAVAL = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]

IAFAL = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]

[00115] For alpha signal B 1202:

IBTP = [51074, 51344, 51398, 51613, 51668, 51830, 51830, 51938,

51992, 52100, 52207, 52262].

IBVAL = [1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1]

IB FAL = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]

[00116] For alpha signal C 1203:

ICTP = [51074, 51344, 51398, 51613, 51668, 51830, 51830, 51938,

51992, 52100, 52207, 52262].

ICVAL = [1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1]

ICFAL = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]

[00117] The system may calculate the raw time stamps, ITP, in microseconds as follows:

IATPn-QAVALn+IAFALn)+IBTPn-(IBVALn+IBFALn)+ICTPn-(ICVALn+ICF ALn)

[00118] ITPn

AVALn+IAFALn)+ BVALn+IBFALri) + (ICVALn+ICFALn)

Equation 13

[00119] In Equation 13 above, n is the index referring to the corresponding record. In the illustrated embodiment, the system determines 12 records with the following time stamps: ITP = [51074, 51344, 51398, 51613, 51668, 51830, 51883, 51938, 51992, 52100, 52207, 52262].

[00120] The system identifies that the first traveling wave peak (VPK) is received at an arrival time (TPK) equal to 51074 s in mode 1, corresponding to the alpha mode. As described above, close-proximity peaks are removed by removing values within close proximity (e.g. , as defined by a TWDSW parameter, such as 10ps) of the identified first traveling wave. Because the next ITP value (51344) is more than 10ps from the identified first traveling wave, no ITP values are filtered.

[00121] The system may then replace all record modes with the mode number (i.e. , 1 ) of the identified first traveling wave. That is, ENUM = [1 , 1 , 1 , 1 , 1 , 1 , 1 , 1 , 1 , 1 , 1 , 1 ].

[00122] Then, the system may then calculate a modal offset value, IMOFF, according to ENUM and MODE as illustrated in FIG. 13. In this example, MODE = CLARKE. In this case the offset corresponds to IALOFF, where:

[00123] IMOFF = [-0.067, -0.019, 0.045, 0.016, 0.051 , -0.024, 0.037, 0.1 18, 0.120, 0.153, -0.047, -0.081 ]. The data acquisition circuitry may provide the offset data, IALOFF.

[00124] The resulting ENUM (modal number) is 1 , which indicates the fault is an AG fault. The system retrieves all three phase currents and reconstructs the alpha traveling wave signals. Using these signals, the algorithm interpolates the raw time stamps ITP, which are in microseconds, to obtain more precision with interpolated ITPI time stamps in nanoseconds. In some embodiments, if the interpolated ITPI data is different than the raw ITP data by more than a threshold amount (e.g., 2 s), then the ITP data may be used instead.

[00125] FIG. 14 illustrates a table with the interpolated time stamps ITPI, and their corresponding estimated magnitudes, IMAG, for the simulation. The system may reference the interpolated TPKs (ITPIs) to the first traveling wave time stamp as illustrated in FIG. 15.

[00126] The system may then determine the sign TPSIGN of each peak with respect to the first VPK using the signed VPK value of each peak, as illustrated in FIG. 16.

[00127] The system may then generate a time difference vector, DT, from the interpolated TPK vector. The time difference vector, DT, may be equivalently described as a set of time difference values, DT, instead of as a vector. The length of the time difference vector, DT, corresponds to the number of combinations of TPKs taken two at a time (i.e. , pair-combinations). Thus, the length of the time difference vector is: (1 1 +1 )

* 1 1/2 = 66). For this simulation, the time difference vector, DT, is:

[00128] DT = [53.348, 53.961 , 54.066, 54.082, 54.156, 54.191 , 107.453, 107.539,

108.152, 108.453, 161 .500, 161 .609, 162.414, 162.438, 215.785, 215.797, 215.906, 216.520, 216.605, 269.863, 269.867, 269.867,

269.879, 269.953, 269.977, 270.031 , 270.063, 323.938, 323.945, 324.023, 324.059, 324.059, 324.098, 377.406, 378.020, 378.215, 431 .563, 432.316, 432.391 , 485.664, 486.383, 486.473, 539.730, 539.844, 539.855, 539.895, 593.816, 593.922, 593.926, 593.977,

594.000, 647.883, 648.082, 702.270, 756.336. 756.414 809.723, 809.762, 863.789, 863.879, 863.953, 917.914, 917.945, 1026.367, 1 133.820, 1 187.977].

[00129] Based on the polarity check and the TWLPT, the first reflection from fault should be less than 2 * TWLPT + 10ps (i.e. , 2 * 540 + 10 = 1090MS), where the 10ps represents an example threshold for increasing inclusivity. In this case, the system identifies three hypotheses (underlined and bolded in FIG. 16).

[00130] The system may utilize the ruler evaluation approach and/or the score evaluation approach to detect valid reflections. Using the ruler evaluation (RTT evaluation) approach for each hypothesis, the system calculates two time references. A first reference corresponds to the reflection from the fault, F(H), and the second reference corresponds to the reflection from the remote terminal, R(H).

[00131] As illustrated in FIG. 17, the reflection from the fault F(H) for the first hypothesis is F(1 )=324.098 and the associated reflection from the remote terminal R(H) for the first hypothesis is R(1 )=755.902. The system may count the number of matches of F(H) with time differences in the time difference vector DT to determine NM(H), and count the number of matches of R(H) with time differences in the time difference vector DT to determine N1_M(H). As illustrated, in FIG. 17, there are six matches for F(H) in the time difference vector DT and two matches for R(H) in the time different vector DT for the first hypothesis. The six matches for F(H) are underlined in the DT vector above. The two matches for R(H) are double underlined in the DT vector above. The system may also calculate the total errors for NM(H) and N1_M(H) within TWTOL1 , as shown in FIG. 18.

[00132] In some embodiments, a hypothesis may be selected as "correct" or as the "best" hypothesis based on the hypothesis having the greatest number of NM matches, the greatest number of N1_M matches, and/or the greatest total number of matches NM + N1_M. In other embodiments, the score evaluation (ETW evaluation) approach may be used in addition to the ruler or RTT evaluation approach to select a "correct" or "best" hypothesis. In still other embodiments, the score evaluation (ETW evaluation) approach may be used by itself without the ruler or RTT evaluation approach. In such an embodiment, the weighting of the scores NS(H) based on a TPK match for R(H) would be irrelevant and so would be ignored and/or not calculated (See, e.g., Equation 10. 1).

[00133] Regardless, a selected hypothesis may be used to calculate a location of the fault relative to the local terminal. Whether the RRT method is used alone or in combination with the ETW evaluation approach, a hypothesis can be selected based on a function of the NM and NM_1 values for each of the evaluated hypotheses.

[00134] Using the score evaluation (ETW evaluation) approach, a set or vector of expected traveling wave arrival times, Th, is determined that includes a plurality of expected arrival times of traveling waves (ETWs) for each hypothesis. The ETWs are compared with the actual received TPK values in the TPK vector to identify matches that are within TWTOL2 (e.g. TWTOL2 = 5 s). In some embodiments, when the system creates the vector of expected traveling wave arrival times, Th, the system may remove close-proximity values.

[00135] FIG. 19 shows the number of TPK matches for each hypothesis per the associated vectors of expected traveling wave arrival times, Th. The expected traveling wave arrival times are in the third, fourth, and fifth columns corresponding to the vectors of expected traveling wave arrival times, Th, for each of the first, second and third hypotheses, respectively. The expected times corresponding to F(H) are underlined, the expected times corresponding to R(H) are double underlined. The strikethrough values are the values that are within TWDSW and may be removed or not considered. The boxed and bolded value can be used for supervision of R(H).

[00136] Figure 20 shows the measured TPKs and the expected traveling waves for each of the three hypotheses. Matches are identified between the TPK measurements and the calculated hypotheses of the vector of expected traveling wave arrival times, Th. The circles identify the matches between TPK and hypotheses within TWTOL2 (5 s in this simulation).

[00137] FIG. 21 illustrates a table showing the number of matches/scores NS(H) for each hypothesis.

[00138] FIG. 22 illustrates the total errors for NS(H) within TWTOL2 (e.g. 5μέ), as calculated by the system. [00139] The system may identify one or more matches between R(H) and F(H) within TVVTOL1 (e.g. 10ps) for a corresponding hypothesis. For example, if F(H) = 324.098, R(H) = 2*TWLPT - F = 2*540 - 324.098 = 755.902. The system may identify 756.414 in the TPK column (boxed and bolded in FIG. 14) and then NS(H) = 2(matches) * 1 (supervision logic) = 2.

[00140] FIG. 23 is a table showing the weight (WGHT) for each hypothesis based on a match of R(H) with a received TPK. In embodiments in which the score algorithm is used in combination with the ruler algorithm, the output of the score algorithm may be weighted based on a match of R(H) with a received TPK. In embodiments in which the score evaluation approach is used alone, no such weighting may be available or used.

[00141] The variable N can be defined as a figure of merit that indicates how closely the received TPKs match the expected TPKs for a fault location calculated based on a given hypothesis. The N values can be calculated using Equation 10 above. A total per unit error can also be calculated for each hypothesis using Equation 11 above.

[00142] FIG. 24 includes a table with the N values and the total per unit error, ERRpu, for each of the three hypotheses.

[00143] As previously described, the system may define three line sections as follows:

Section 1 : m_ini < 0.3

Section 2: 0.3 <= m_ini <= 0.7; and

Section 3: m_ini > 0.7,

[00144] As described above, the default initial guess when other data is unavailable is m_ini=0.5, and so the m_ini falls within Section 2. Therefore, the system assumes that the fault is in Section 2. For faults in Section 2, the algorithm uses N (and optionally ERRpu in the event of a tie) to order the hypotheses, as follows: TWSE = [0.3, 0.8, 0.95, NA].

[00145] The traveling wave single-ended fault location, TWSEFL, for hypothesis 1 can be calculated per Equation 12. 1 above as:

[00146] TWSEFLi = LL * F(1) = (100 * 324.098) / (2*540) = 30.009 miles.

1 J 1 2*TWLPT / \ /

[00147] FIG. 25 shows the expected reflections for a fault at m=0.3 and the TPK measurements shown with circles that match the received TPK values above for this simulation. [00148] In some cases, well-known features, structures or operations are not shown or described in detail. Furthermore, the described features, structures, or operations may be combined in any suitable manner in one or more embodiments. It will also be readily understood that the components of the embodiments as generally described and illustrated in the figures herein could be arranged and designed in a wide variety of different configurations.

[00149] Several aspects of the embodiments described may be implemented as software modules or components. As used herein, a software module or component may include any type of computer instruction or computer executable code located within a memory device and/or transmitted as electronic signals over a system bus or wired or wireless network. A software module or component may, for instance, comprise one or more physical or logical blocks of computer instructions, which may be organized as a routine, program, object, component, data structure, etc. , that performs one or more tasks or implements particular abstract data types.

[00150] In certain embodiments, a particular software module or component may comprise disparate instructions stored in different locations of a memory device, which together implement the described functionality of the module. Indeed, a module or component may comprise a single instruction or many instructions, and may be distributed over several different code segments, among different programs, and across several memory devices. Some embodiments may be practiced in a distributed computing environment where tasks are performed by a remote processing device linked through a communications network. In a distributed computing environment, software modules or components may be located in local and/or remote memory storage devices. In addition, data being tied or rendered together in a database record may be resident in the same memory device, or across several memory devices, and may be linked together in fields of a record in a database across a network.

[00151] Embodiments may be provided as a computer program product including a non-transitory computer and/or machine-readable medium having stored thereon instructions that may be used to program a computer (or other electronic device) to perform processes described herein. For example, a non-transitory computer-readable medium may store instructions that, when executed by a processor of a computer system, cause the processor to perform certain methods disclosed herein. The non- transitory computer-readable medium may include, but is not limited to, hard drives, floppy diskettes, optical disks, CD-ROMs, DVD-ROMs, ROMs, RAMs, EPROMs, EEPROMs, magnetic or optical cards, solid-state memory devices, or other types of machine-readable media suitable for storing electronic and/or processor executable instructions.

[00152] While specific embodiments and applications of the disclosure have been illustrated and described, it is to be understood that the disclosure is not limited to the precise configurations and components disclosed herein. For example, the systems and methods described herein may be applied to an industrial electric power delivery system or an electric power delivery system implemented in a boat or oil platform that may not include long-distance transmission of high-voltage power. Moreover, principles described herein may also be utilized for protecting an electric system from over- frequency conditions, wherein power generation would be shed rather than load to reduce effects on the system. Accordingly, many changes may be made to the details of the above-described embodiments without departing from the underlying principles of this disclosure. The scope of the present invention(s) should, therefore, be interpreted to encompass at least the following claims: