Systems and methods for detecting and isolating arc faults - Patents.com

Abstract

The arc fault detection system senses current flow in a power branch and one or more load branches of an electrical system. Over a frequency range divided into a predetermined number of frequency bins, the controller records and tallies, for each frequency bin, the branch having the highest magnitude of power spectral density. The branch with the highest total number is declared to be the branch in which an arc fault has occurred and can be safely isolated from the electrical system.

Description

Detailed Description of the Invention

[0001] [CROSS REFERENCE TO RELATED APPLICATIONS] This application claims priority to U.S. Provisional Application No. 63 / 232,468, filed August 12, 2021, which is incorporated by reference in its entirety. [Technical field] FIELD OF THE DISCLOSURE This disclosure relates to systems and methods for detecting the origin of an arc fault signal, and more particularly, to systems and methods for isolating an arc fault to a single load. [background] Arcing is a serious and potentially dangerous phenomenon that occurs in electrical wiring when a path or gap of sufficiently low resistance exists between exposed conductors of different potentials in the electrical wiring, resulting in a jump of electrical energy across the path or gap, i.e., an arc. Arcing can degrade or damage electrical circuitry, generate radio frequency noise, and ignite nearby combustible materials. The risk of arcing is particularly acute in older aircraft where wire insulation may have deteriorated or terminal connections may have loosened over time due to vibration. Bundling of wires due to space constraints increases the risk of arcing even more on aircraft.

[0002] With the increasing movement towards electric aircraft (MEA) and the use of 270VDC and higher DC distribution voltages on aircraft, there are growing concerns about the safety of these high power systems. Unlike AC power, DC power inherently has no zero crossings in voltage and current. Therefore, once an electric arc occurs across a short air gap, the air in the arc remains ionized and conductive, preventing the arc from extinguishing. A sustained arc in series with an electric load, i.e., a series arc fault, will generate significant amounts of heat that can damage electrical wiring, cause smoke and fire, and endanger the aircraft. Thus, there is a significant motivation to have techniques for detecting and isolating series arc faults in DC power systems. The detection is typically performed by analyzing the voltage and / or current waveforms of the arc signature, with the current signature being the most reliably detected. Detecting the current signature of an electric arc fault in a large system faces the usual challenges, such as dealing with high ambient electrical noise from a variety of common noise sources, complex system impedances, and low-level arc fault signatures. The primary challenge, however, is determining the source of the arc fault signature so that an appropriate response can be coordinated to isolate the arc fault without unnecessarily shutting down portions of the system that do not contain the arc fault. This disclosure discloses techniques for isolating the arc fault signature to a single electrical load so that only one load is shut down.

[0003] A typical arc detector is disclosed in U.S. Patent No. 6,625,550, issued Sep. 23, 2003 to Scott et al., entitled "Arc Fault Detection For Aircraft," which is incorporated herein by reference in its entirety. However, this detector is limited to detection only and lacks the ability to localize the source of the arc fault and isolate only the circuit branch in which the arc fault is detected.

[0004] Another arc detector is disclosed in U.S. Patent No. 7,834,637, issued Nov. 16, 2010, by Kojori et al., entitled "Method and Apparatus for Generalized AC and DC Arc Fault Detection and Protection," which is also incorporated herein by reference in its entirety. The disclosed method for detecting arc faults includes a running discrete Fourier series (RFDS) processor that senses changes in the magnitude of the fundamental component of the current and an i 2 As mentioned above, this method also lacks the ability to localize the source of the arc fault and isolate only the circuit branch in which the arc fault is detected.

[0005] Arcing in wiring is a potentially destructive condition that has proven difficult to reliably detect and isolate using existing methods. A method is needed to simply, quickly, and accurately detect arcing in electrical circuits. Furthermore, the method must robustly determine the source of the arc fault signature so that the single fault can be isolated without breaking non-faulty electrical connections.

[0006] The subject matter of this disclosure is directed to detecting the origin of an arc fault signal, and more particularly, to a system and method for isolating an arc fault to a single load or a single branch of a power system. [overview] The ability to isolate an arc fault to a single branch of a large power system is particularly important in systems that use an array of electronic breakers to provide arc fault detection and mitigation. If a series arc fault occurs, the fault must be detected and isolated to a single load, and power must be removed from that load, typically in less than 200 milliseconds, to prevent the propagation of electrical fault and / or damage to the aircraft. Detecting the presence of an arc fault is easy and has been shown to be possible in the related art. However, what continues to be difficult is isolating an electrical fault to a single electrical branch in which an arc fault exists. This capability is particularly important when an array of breakers performs arc fault detection and mitigation. Current induced in a single branch of a power system due to the presence of an arc fault will also flow through all other electrical branches connected to the faulted branch. [Brief description of the drawings]

[0007] The features and advantages of the present disclosure will be more fully understood with reference to the following more detailed description taken in conjunction with the accompanying drawings, in which: [Figure 1] FIG. 1 is a circuit diagram of an embodiment of an electronic circuit breaker controller (ECBU). [Diagram 2] FIG. 1 is a general circuit diagram of a shunt. [Diagram 3] FIG. 1 is a circuit diagram of an embodiment of an ECBU with an added capacitor according to the present disclosure. [Figure 4] 1 is a bar graph showing count sums for each load branch in an exemplary electrical system, with each count sum representing the number of frequency bins in which the relative magnitude of the sensed current in the corresponding load branch was greatest across the entire measurement frequency range. Like reference numbers and designations in the various drawings indicate like elements. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0008] [Detailed Description] Reference will now be made in detail to the exemplary embodiments of the present disclosure, which are illustrated in the accompanying drawings. The specific details disclosed herein should not be construed as limiting, but rather as a basis for the claims, and as teaching one of ordinary skill in the art how the present invention may be employed in any suitably detailed system, structure, or method. The same reference numerals will be used throughout the drawings to refer to the same or similar components, circuits, or functions.

[0009] FIG. 1 shows a conventional ECBU 100 with arc detection capability connected to four loads (load 1, load 2, load 3, and load 4) Z1, Z2, Z3, and Z4. An arc occurring in series with the first load (load 1) Z1 will generate an AC current arc signature that may be conducted to any of the other electrical circuits connected to the DC power bus. The path this current takes will depend on the relative AC impedances of the various electrical paths available in the electrical system. For example, if the second load (load 2) Z2 has a very low AC input impedance compared to the remaining loads and the power source, most of the arc fault current generated by the first load Z1 will pass through the second load Z2. Therefore, the current signatures measured at the first load Z1 and the second load Z2 will be very similar. In reality, this current will be shunted among all of the various electrical branches, but not evenly, and its signature will typically exceed the trip threshold of one or more of the remaining monitored loads, meaning that multiple loads will be turned off in the event of a single arcing fault. Alternatively, an arcing fault originating in an entirely different part of the electrical system may also conduct its signature to those loads, causing them to be erroneously turned off.

[0010] However, in the configuration shown in Figure 1, it will not be possible to detect the phase relationship of the arc fault noise because it is stochastic and wideband in nature. Any attempt to obtain the phase of the random noise will result in unusable phase information. A method must be developed to robustly determine the source of the arc fault signature so that a single fault can be isolated without breaking a fault-free electrical connection.

[0011] The main characteristic of an arc fault signature is the presence of broadband noise. To detect an arc fault signature, first the extracted block of current data is transformed into the frequency domain using a Fast Fourier Transform (FFT). After performing the FFT, the power spectral density (PSD) is calculated for small frequency intervals, or bins, across the entire measurement frequency range. The PSD of each frequency bin is compared to a predefined limit for that bin. An arc fault is determined to be present if a significant majority of the bin limits are exceeded for a sufficiently long period of time. In a real power system, there will always be some noise present, so that at any one time a small number of bins will exceed the PSD limit, but it is believed that only true broadband noise associated with an arc fault will exceed a majority of the bins for a long period of time.

[0012] The present disclosure takes advantage of the fact that arc fault current has a single source but its return path will split into multiple electrical paths in a branched power system. Kirchhoff's current law states that the current flowing into a node must be equal to the current flowing out of that node. As a result, the arc fault signature current generated by the first load Z1 in FIG. 1 will be divided in the return path by the impedance of the remaining connected loads and the impedance of the power source. Each of the illustrated electrical paths has a different impedance over frequency, and the current signal will be divided among the remaining paths according to the relative ratio of the impedances.

[0013] FIG. 2 illustrates an example shunt circuit, and Equation 1 below expresses Kirchhoff's current law according to FIG. 2, where N is the total number of electrical branches connected to node J1, and I(f) is the current generated by the electrical branch containing the series arc fault versus frequency.

[0014]

number

[0015] Additionally, Equation 2 below represents the current contained in each electrical branch connected to node J1 versus frequency. It is important to realize that because the impedance of each electrical branch varies over the measurement frequency range, the current ratios in each electrical branch will vary over frequency as well.

[0016]

number

[0017] In the case of a single source of arc fault current, if there are three or more electrical branches connected to node J1, the source of the current can be determined to be the current with the highest amplitude connected to node J1. If, in addition to the signal source, there are two or more other branches through which current can flow, the current will be divided between the two or more potential return paths in the ratio of the impedances of each return path. The signal source will always be the electrical branch with the highest amplitude.

[0018] In the special case where only two branches, one of which is an electrical load and the other of which is a power source, are connected to the connection point J1, the current signal from the arc and the power source may coincide, and no conclusion can be reached about the source of the arc signature. Therefore, in order to be able to identify the source of the arc signature in this case, another intentional electrical path must be added to the connection point. Figure 3 shows an embodiment of the ECBU configuration with an additional small capacitor C. Due to the inductance of the feeder and the inductance of the generator, the Z source has a relatively high impedance, especially at high frequencies. The additional capacitor C serves as a return path for a part of the arc-induced current. Now, by comparing the current magnitude between the load 1 current sensor LS1 and the source current sensor SS, it can be determined that the origin of the arc fault current is the load 1 connection, and not due to an arc fault elsewhere in the power system.

[0019] In a typical power system, there will be individual loads that send narrowband currents into the power system during normal operation. Also, certain types of equipment, such as motor drives and switch mode power converters, can produce narrowband emissions at very high amplitudes. When determining the source of an arc fault signature, the arc fault detection algorithm must not be influenced by normal emissions from individual electrical loads, so a measurement of the overall current amplitude, such as RMS current, is not useful for determining the source of the arc fault signature. To detect the source of an arc fault, it is necessary to compare the current amplitudes in each branch of the electrical circuit over the entire measurement frequency range. This comparison of the relative amplitudes between all of the current sensors for the individual loads and the power supply over all frequencies is important to detect the true source of the arc fault signature. The following algorithm can be used to determine the source of the arc fault signature.

[0020] 1) Assign a count variable (CNTn) to each monitored current. Initially, set all count variables to 0. 2) Once it is determined that an arc fault exists anywhere in the power system, compare the magnitude of the PSD for the lowest frequency bin among all of the sensed currents. The last extracted data set that resulted in the determination that an arc exists should be used for this purpose.

[0021] 3) Increment the count of the monitored current that has the highest PSD magnitude for the lowest frequency bin. 4) Repeat steps 2 and 3 for the next higher frequency bin. This process is repeated until the highest frequency bin is reached and the relative magnitudes are compared.

[0022] 5) Compare the magnitude of the count variables among all of the sensed currents. The monitored current with the highest count is determined to be the source of the arc fault signature.

[0023] 6) If the source of the arc fault signature is one of the connected loads, disconnect that load to isolate the series arc fault. If the source of the arc fault signature is the power supply, ignore the arc detection and continue monitoring.

[0024] It is not important which order is used, be it lowest frequency to highest frequency, highest frequency to lowest frequency, or any other order, but the intent is to determine which sensed current across the entire frequency range contains the strongest arc signature. The sensed current containing the highest magnitude across the entire frequency range is determined to be the source of the arc signature.

[0025] 4 shows the count totals for an exemplary system. CNT4 has the highest count, so the sensed current corresponding to CNT4 will be assigned as the source of the arc current signature. The remaining monitored currents may contain noisy loads with higher PSD for a narrow frequency range, but will not have the highest count values ​​if they do not contain an arc signature.

[0026] Although FIG. 3 shows an embodiment having four load branches, other embodiments having a different number of load branches are contemplated. While the disclosed subject matter has been described and illustrated with respect to embodiments thereof, it should be understood by those skilled in the art that features of the disclosed embodiments may be combined, rearranged, etc. to create additional embodiments within the scope of the present invention, and that various other modifications, omissions, and additions may be made to the present embodiments without departing from the spirit and scope of the present invention. For example, while the described embodiments relate to an aircraft and a particular type of ECBU and loads, it should be understood that various embodiments in other environmental settings and using other numbers and types of loads are within the scope of the present disclosure.

Claims

1. 1. A method for detecting an arc fault in an electrical circuit configured to supply power to a plurality of loads, the method comprising: monitoring a level of current being supplied to each of the plurality of loads at each of a plurality of time points; for each of said plurality of time points, converting each of said plurality of current levels into a corresponding frequency representation; for each component of said frequency representation of each current level, measuring a power spectral density for each frequency component; comparing each power spectral density with a predetermined limit corresponding to each frequency component; indicating that an arc fault has occurred if a predetermined minimum number of power spectral densities exceed corresponding limits for a predetermined minimum number of time points; A method comprising:

2. 2. The method of claim 1 , The method, wherein the step of transforming each of the plurality of current levels comprises performing a Fast Fourier Transform.

3. 2. The method of claim 1 , A method according to claim 1, wherein each component of the frequency representation comprises a corresponding range of a plurality of discrete frequencies.

4. 1. A method for identifying a source of an arcing fault in an electrical circuit configured to supply power to a plurality of loads, the method comprising: monitoring the level of current being supplied to each of the plurality of loads at a given time; converting each of the plurality of current levels to a corresponding frequency representation; for each component of said frequency representation of each current level, measuring a power spectral density for each frequency component; determining, for each frequency component, a maximum spectral density associated with the current level for the corresponding load; identifying the source of the arc fault as the load having the current level determined to have a maximum number of highest spectral densities; A method comprising:

5. 5. The method of claim 4, The method, wherein transforming each of the plurality of current levels comprises performing a Fast Fourier Transform.

6. 5. The method of claim 4, A method according to claim 1, wherein each component of the frequency representation comprises a corresponding range of a plurality of discrete frequencies.

7. The method of claim 4, further comprising: The method comprising isolating the source of the arc fault.

8. The method of claim 4, further comprising: assigning a count variable to each of the monitored currents corresponding to a load, and initially setting each count variable to zero; identifying the monitored current corresponding to a load having the highest spectral density for a particular range of discrete frequencies and incrementing the count variable for the identified current; repeating the identifying step for one or more other frequency ranges. Equipped with The method, wherein the step of identifying the source of the arc fault is determined based on the monitored current having a highest count variable.

9. 9. The method of claim 8, A method according to claim 1, wherein the frequency ranges are arranged from low frequency to high frequency, and the step of identifying the loads having the highest spectral density is performed from the low frequency range to the high frequency range.

10. 9. The method of claim 8, A method according to claim 1, wherein the frequency ranges are arranged from high to low frequencies and the step of identifying the loads having the highest spectral density is performed from the high frequency range to the low frequency range.

11. 1. A method for detecting and identifying a source of an arcing fault in an electrical circuit configured to supply power to a plurality of loads, the method comprising: monitoring a level of current being supplied to each of the plurality of loads at each of a plurality of time points; for each of said plurality of time points, converting each of said plurality of current levels into a corresponding frequency representation; for each component of said frequency representation of each current level, measuring a power spectral density for each frequency component; comparing each power spectral density with a predetermined limit corresponding to each frequency component; indicating that an arc fault has occurred if a predetermined minimum number of power spectral densities exceed corresponding limits for a predetermined minimum number of time points; determining, for each frequency component, a maximum spectral density associated with the current level for the corresponding load; identifying the source of the arc fault as the load having the current level determined to have a maximum number of highest spectral densities; A method comprising:

12. 12. The method of claim 11, The method, wherein transforming each of the plurality of current levels comprises performing a Fast Fourier Transform.

13. 12. The method of claim 11, A method according to claim 1, wherein each component of the frequency representation comprises a corresponding range of a plurality of discrete frequencies.

14. 12. The method of claim 11 further comprising: The method comprising isolating the source of the arc fault.

15. 1. A system for detecting and identifying a source of an arcing fault in an electrical circuit configured to supply power to a plurality of loads, the system comprising: A processor; When executed by the processor, the processor monitoring a level of current being supplied to each of the plurality of loads at each of a plurality of time points; converting, for each of said plurality of time points, each of said plurality of current levels into a corresponding frequency representation; determining, for each component of said frequency representation of each current level, a power spectral density for each frequency component; comparing each power spectral density with predetermined limits corresponding to each frequency component; indicating that an arc fault has occurred when a predetermined minimum number of power spectral densities exceed their corresponding limits for a predetermined minimum number of time points; determining, for each frequency component, a maximum spectral density associated with said current level for said corresponding load; Identifying the source of the arc fault as the load having the current level determined to have the greatest number of highest spectral densities. A memory that stores instructions and A system comprising:

16. 16. The system of claim 15, The system, wherein the processor is further programmed to transform each of the plurality of current levels by performing a Fast Fourier Transform.

17. 16. The system of claim 15, Each component of the frequency representation comprises a corresponding range of a plurality of discrete frequencies.

18. 16. The system of claim 15, The processor is further programmed to isolate the source of the arc fault.

19. 1. A system for detecting and identifying a source of an arcing fault in an electrical circuit configured to supply power to a plurality of loads, the system comprising: a plurality of current sensors, each current sensor capable of monitoring a level of current being supplied to a corresponding one of the plurality of loads at each of a plurality of time points; a signal processor operable to convert, for each of the plurality of time points, each of the plurality of current levels into a corresponding frequency representation; the signal processor further operable to measure, for each component of the frequency representation of each current level, a power spectral density for each frequency component; a first comparator operable to compare each power spectral density with a predetermined limit corresponding to each frequency component; an indicator capable of indicating that an arc fault has occurred when a predetermined minimum number of power spectral densities exceed corresponding limits for a predetermined minimum number of time points; a second comparator operable to determine, for each frequency component, a highest spectral density associated with the current level for the corresponding load; the second comparator further operable to identify the source of the arc fault as the load having the current level determined to have a maximum number of highest spectral densities; A system comprising:

20. 20. The system of claim 19, further comprising: The system includes an isolator operable to isolate the source of the arc fault.

Images