Passive harmonic filter power quality monitor and communications device

Abstract

A method and apparatus for detection of a failure of a rectifier connected to a passive harmonic filter with a tuned circuit reactor, or of the filter itself, and for generating system currents in the filter, using low cost voltage sensing, modeling of reactor resistance and saturable inductance, and a mathematical integration. The harmonic spectrum of the rectifier current is used to determine an estimate of the rectifier impedance. A template of expected rectifier current is calculated, and compared against a rectifier current calculated on the basis of sensed voltages, to generate a difference signal. The difference signal is compared against a predetermined fault threshold to determine if an error has occurred. The apparatus includes a DSP obtaining the voltages of the source and load, and the tuned circuit reactor voltage, calculating the template, and comparing the actual voltages with the template, to annunciate a fault.

What is claimed is: 
     
         1 . A method for detection of a failure of a rectifier connected to a passive harmonic filter, the rectifier passing a rectifier current having a harmonic spectrum, the method comprising:
 using the harmonic spectrum of the rectifier current to determine an estimate of a rectifier impedance of the rectifier;   calculating a template of expected rectifier current;   using a voltage measured in the filter to arrive at a calculated current into the rectifier;   comparing the expected rectifier current against the calculated current into the rectifier;   generating a difference signal based upon the difference between the expected rectifier current and the calculated current; and   determining whether an error has occurred based on the difference signal.   
     
     
         2 . The method of  claim 1 , wherein the voltage measured in the filter is measured at a line reactor in the filter. 
     
     
         3 . The method of  claim 1 , further comprising:
 passing the difference signal through a low pass filter to create a low-pass filtered signal; and   comparing the low-pass filtered signal against a fault threshold to determine whether an error has occurred.   
     
     
         4 . The method of  claim 2 , further comprising implementing multiple fault thresholds so as to enable detection of both a soft imminent failure and a hard failure. 
     
     
         5 . The method of  claim 4 , further comprising:
 parameterizing the thresholds for the detection of the failure to allow for adjustment based on unit size and power system differences such as power system voltage, frequency, and background voltage distortion.   
     
     
         6 . The method of  claim 1 , further comprising:
 using the calculated current to calculate load RMS and fundamental currents, to generate a measure of harmonic current distortion;   comparing pre-indexed harmonic table components and continuously calculated harmonic components in the load current and determining a difference; and   using that difference to determine whether an error has occurred so as to determine whether to annunciate an error or fault condition.   
     
     
         7 . The method of  claim 1 , further comprising:
 using the calculated current to calculate load RMS and fundamental currents to generate a measure of harmonic current distortion;   comparing the calculated load RMS and fundamental currents to overcurrent and overvoltage thresholds and determining a difference; and   using that difference to determine whether to annunciate an error or fault condition.   
     
     
         8 . The method of  claim 1 , further comprising
 indexing and storing harmonic spectrum components of a rectifier load for a VFD in a data table;   using the calculated current to calculate load RMS and fundamental currents to generate a measure of harmonic current distortion;   comparing the indexed harmonic components from the data table to continuously calculated harmonic components in the load current and determining a difference; and   using that difference to determine whether to annunciate an error or fault condition.   
     
     
         9 . The method of  claim 1 , wherein the passive harmonic filter includes at least one of a line reactor and a tuned circuit reactor, the method further comprising calculating the current by:
 measuring a voltage across at least one of the reactors to obtain a voltage signal;   conditioning the voltage signal to obtain a conditioned voltage signal;   converting the voltage signal to a converted signal in physical units;   integrating the converted signal to produce an initial calculated current;   applying a continuous offset calculation to the calculated current to remove any unwanted offsets; and   modeling an induction saturation effect by scaling the contribution of the reactor voltage integration to the reactor current estimation, by means of an inductive saturation look-up table.   
     
     
         10 . A method for detection of a failure of a rectifier connected to a passive harmonic filter, the rectifier passing a rectifier current having a harmonic spectrum, the method comprising:
 using the harmonic spectrum of the rectifier current to determine an estimate of the rectifier impedance;   calculating a template of expected rectifier current;   using a voltage measured in the filter to arrive at a calculated current into the rectifier;   comparing the expected rectifier current against the calculated current into the rectifier;   generating a difference signal based upon the difference between the expected rectifier current and the calculated current; and   determining whether an error has occurred based on the difference signal;   wherein the passive harmonic filter includes at least one line reactor having a tap at a number of turns in the reactor that is less than 100% of the turns, and wherein calculating the calculated current into the rectifier further comprises:   measuring a voltage across a tapped portion of a tapped line reactor to obtain a tapped reactor voltage signal;   conditioning the tapped reactor voltage signal to obtain a conditioned tapped reactor voltage signal;   converting the conditioned tuned circuit reactor voltage signal to a converted signal of physical units;   integrating the converted signal to produce an initial calculated current; and   applying a continuous offset calculation to the initial calculated current to remove any unwanted offsets and modeling a contribution of current injected at the line reactor tap to the initial calculated current by a factor of reactor tap turns to total turns of the line reactor, resulting in the calculated current.