A method is disclosed for locating an earth fault in a DC network, to which multiple load zones are connectable. Each load zone has two symmetrically earthed load zone lines. In an embodiment of the method, for each load zone line, a line voltage is continuously measured between a line potential, at which the load zone line lies, and an earth potential. In the event that an earth fault is detected at a main line of the DC network, the main lines are separated from both load zone lines of a load zone and the earth fault is assigned to this load zone, if a magnitude difference between the magnitudes of the line voltages of the two load zone lines of this load zone are not significantly reduced after separating the main lines from the load zone lines.

The present invention relates to a method of identifying an electrical contact fault in a wind turbine blade electrical heating system, comprising inserting a predetermined enforced off cycle between or within a switching duty cycle, wherein the switching duty cycle controls the switching of one or more electrical contacts in the electrical heating system, measuring a current consumption during the predetermined enforced off cycle, determining if the measured current consumption exceeds a predetermined threshold, and wherein if the measured current consumption exceeds the predetermined threshold shutting down at least part of the electrical heating system. The present invention also relates to a wind turbine with one or more wind turbine blades that can identify an electrical contact fault.

A method is for locating a ground fault in a DC system to which a plurality of load zones can be connected. The method includes specifying a time window and, after the ground fault is detected, assigning the ground fault to a load zone which was connected to the DC system within the time window before the detection of the ground fault.

A method is for detecting a ground fault in a load zone that can be connected to a direct-current network. The load zone includes a first load zone line connectable to a first main line of the direct-current network via a circuit breaker, and a second load zone line connectable to a second main line of the direct-current network via an interrupter switch. In an embodiment of the method, at least one line voltage at a load zone line is continually measured. If a ground fault is not detected at any main line of the direct-current network before the load zone is connected, the interrupter switch is closed while the circuit breaker is open, and a ground fault in the load zone is inferred if at least one line voltage does not significantly change after the interrupter switch is closed.

Systems and methods may be used to determine fault types and/or directions even during a loss of potential by receiving, at one or more processors, an indication of a pre-fault power flow direction for a power delivery system. The one or more processors then determine a fault direction during a fault for the power delivery system using current vector angles and the pre-fault power flow direction.

A method for determining a location of a fault on a powerline, comprising: obtaining, using a measuring device at a measurement location on the powerline, time-dependent data of a voltage on the powerline and of a current through the powerline at the measurement location, wherein the time-dependent data is obtained during a time period after fault inception; removing fundamental frequency components from the time-dependent data; calculating, from the time-dependent data, a virtual fault inductance for at least two different potential fault locations, by inputting the time-dependent data into an equivalent circuit model, by means of time-domain calculations on the nonfundamental frequency waveform components; and deriving, from the virtual fault inductance for each of the potential fault locations, a location on the line at which the virtual fault inductance is substantially zero.

A method for fault location and isolation in a power distribution network, where the network includes a plurality of switching devices provided along a feeder, and at least one of the switching devices does not have voltage sensing capabilities. The method includes detecting an overcurrent event in the network from the fault and interrupting the overcurrent event by opening and then immediately locking out or subsequently reclosing and testing the fault. A count value is increased in each switching device that detected the overcurrent event. A message is sent from each of the switching devices that detected the overcurrent event and then detected the loss of voltage upstream to an upstream neighbor switching device. Current measurements in the messages, measured current by the devices and the counts values in the devices determine what devices are opened to isolate the fault.

A system and method for controlling microgrids composed of inverter-based distributed generation (IBDG) units. This includes a method using multiple IBDGs to inject impedance-modulated harmonic currents during fault conditions, with each IBDG injecting a unique, differentiable harmonic (i.e., non-fundamental) order from neighboring IBDGs. The method also involves using an inverse time-harmonic-current characteristic to detect faults by locally measuring the harmonic currents injected by IBDGs. A harmonic directional overcurrent relay is also used for fault detection.

A smart switch allows distributed generators to “ride through” non-three-phase faults by very quickly detecting a non-three-phase phase fault, locating the fault, identifying the “responsive sectionalizer switches” that will be involved in clearing or isolating the fault, and selecting one of the responsive sectionalizer switches to direct back-feed tie switch operations. The responsive sectionalizer switches trip only the faulted phase(s), and the selected sectionalizer switch instructs a back-feed tie switch to close to back-feed the distributed generators prior to conducting the typical fault response operation. This typically occurs within about three cycles, and is completed before the normal fault clearing and isolation procedures, which momentarily disconnect all three phases to the distributed generators from the normally connected feeder breaker. The looped connection to an alternate feeder breaker during these operations allows the distributed generators to “ride through” the normal fault clearing and isolation procedures.

A method for locating a ground fault in an IT network which has a converter with a rectifier connected to a power transmission network, a DC link and an inverter connected to an electrical machine includes measuring a common-mode voltage property of the converter or of the power transmission network and comparing the common-mode voltage property with an output voltage property of an output voltage of the inverter. When the IT network experiences a ground fault, the comparison of the common-mode voltage property with the output voltage property is used to determine whether a machine area of the IT network connected downstream of the converter, which includes the electrical machine and a connecting line between the electrical machine and the converter, causes the ground fault.