Method for insulation monitoring of a converter-supplied power supply system

Abstract

The present invention relates to a method for determining an insulation resistance and for locating insulation faults in a power supply system whose active parts are ungrounded and which is supplied via a converter operated grounded and equipped with controlled power semiconductor switches. A common-mode voltage against ground is generated at the output of the converter and is superimposed on the ungrounded network as an active measuring voltage in order to measure the insulation resistance. The direct integration of the generation of the common-mode measuring voltage in the converter allows cost-effective implementation including similarly comprehensive insulation monitoring functions as those possible in fully ungrounded power supply systems. Furthermore, the method for determining the insulation resistance can be expanded into a method for locating insulation faults and thus for locating faulty system branches.

Claims

1. A method for determining an insulation resistance (R.sub.f) of a power supply system (20, 30, 40, 50) whose active parts are ungrounded and which is supplied via a converter (10, 14, 24, 34) operated grounded and equipped with controlled power semiconductor switches (SW10, SW20, SW30, SW40, SW50, SW60), the method comprising the following steps: generating a common-mode voltage against ground at the output of the converter (10, 14, 24, 34), wherein the common-mode voltage is generated by generating pulse patterns (sa, sb, sc) for controlling the power semiconductor switches (SW10, SW20, SW30, SW40, SW50, SW60), superimposing the common-mode voltage as a measuring voltage (U.sub.g) for determining the insulation resistance (R.sub.f) of the ungrounded power supply system (20, 30, 40, 50), detecting the measuring voltage (U.sub.g) in the ungrounded power supply system (20, 30, 40, 50), detecting a measuring current (I.sub.g) which is flowing in the ungrounded power supply system (20, 30, 40, 50) via the insulation resistance (R.sub.f) because of the measuring voltage (U.sub.g), and determining the insulation resistance (R.sub.f) by assessing the measuring current (I.sub.g).

2. The method according to claim 1, wherein the pulse patterns (sa, sb, sc) are generated by a voltage displacement (12) of a reference voltage (U.sub.L1, U.sub.L2, U.sub.L3) for each phase (L1, L2, L3) of the converter (10, 14, 24, 34) using a bias voltage (U.sub.m).

3. The method according to claim 2, wherein the bias voltage (U.sub.m) is set in such a manner that it has a square waveform with a fundamental frequency below a network frequency of the power supply system (20, 30, 40, 50), resulting in a measuring frequency of the measuring voltage (U.sub.g) at the level of the fundamental frequency of the bias voltage (U.sub.m).

4. The method according to claim 1, wherein an application in which the converter (10, 14, 24, 34) is configured as an inverter (14) or as a rectifier (24) or as a frequency converter (34).

5. A method according to claim 1, the method further comprising the steps of: detecting a residual current (I.sub.x) caused by the measuring voltage (U.sub.g) in a system branch (41, 42, 51, 52) to be monitored in the ungrounded power supply system (40, 50) by means of a residual current measuring device (6), assessing the residual current (I.sub.x) to detect a system branch (41, 51) exhibiting an insulation fault (R.sub.x).

6. The method according to claim 5, wherein determining a partial insulation resistance and/or a partial network leakage capacitance of the system branch (41, 42, 51, 52) to be monitored.

Description

BRIEF DESCRIPTION OF THE DRAWING FIGURES

(1) Other advantageous features are apparent from the following description and from the drawings, which show preferred embodiments of the invention by way of example.

(2) FIG. 1: is an illustration of the principle of a monitoring gap “symmetrical fault” of passive residual current measuring technology,

(3) FIG. 2: is an illustration of the principle of an impact of capacitive leakage currents on the measuring sensitivity of residual current measurements,

(4) FIG. 3: shows a simulation of a converter (inverter with STC modulation) with a voltage displacement for generating a common-mode voltage,

(5) FIG. 4a: shows simulation results (signal waveforms) of the converter simulation of FIG. 3 without voltage displacement,

(6) FIG. 4b: shows simulation results (signal waveforms) of the converter simulation of FIG. 3 with voltage displacement,

(7) FIG. 5a: shows insulation monitoring according to the invention in an ungrounded DC power supply system,

(8) FIG. 5b: shows insulation monitoring according to the invention in an ungrounded AC power supply system,

(9) FIG. 6a: shows a localization according to the invention of insulation faults in a branched ungrounded DC power supply system, and

(10) FIG. 6b: shows a localization according to the invention of insulation faults in a branched ungrounded AC power supply system.

DETAILED DESCRIPTION

(11) FIG. 1 is an illustration of a principle of a monitoring gap “symmetrical fault” which can occur when passive residual current measuring technology is used. What is shown is a power supply system 2 which comprises active conductors (phase conductors, phases) L1, L2 and L3 and which is grounded via a power source 4. A residual current measuring device 6 detects the vectorial sum of the currents flowing in conductors L1, L2 and L3. Residual current measuring device 6 may be part of a residual current protective device (RCD) or of a residual current monitor (RCM).

(12) Because of the typically symmetrical structures in electrical systems, a symmetrical fault case, i.e. an insulation fault that affects all three conductors L1, L2 and L3 in the same way, can typically be assumed when the insulation deteriorates. In FIG. 1, this insulation fault is illustrated by insulation resistance R.sub.iso—in the fault case also by R.sub.f (insulation fault or insulation fault resistance)—which is effective between conductors L1, L2, L3 (including all connected equipment) and ground and which models the complex-value leakage impedance of entire power supply system 2 as a concentrated element together with leakage capacitance C.sub.e.

(13) Because of the symmetrical phasing, the vectorial sum of residual currents I.sub.f1, I.sub.f2 and I.sub.f3 flowing via insulation resistance R.sub.iso in conductors L1, L2 and L3—and thus resulting residual current ΔI—is close to zero.

(14) Despite the significant magnitude of residual currents I.sub.f1, I.sub.f2 and I.sub.f3, they cannot be detected and cause a monitoring gap in connection with the occurrence of symmetrical faults when passive residual current measuring technology is used.

(15) FIG. 2 is an illustration of the principle of the impact of capacitive leakage currents on the measuring sensitivity of residual current measurements.

(16) One disadvantage that occurs when using residual current-based monitoring systems as compared to the use of insulation monitoring devices (IMD) installed according to standards is that the measuring behavior of residual current-based monitoring systems depends distinctly on network leakage capacitances present and on the leakage currents determined by them. If a significant capacitive leakage current I.sub.ab1 is flowing already, the vectorial addition with a smaller, yet still hazardous residual current I.sub.f results in only little increase of the residual current, which cannot be reliably detected in case of an insufficient sensitivity of the residual current measuring device. The numerical example in FIG. 2 illustrates an only marginal increase of 1.5 mA in residual current ΔI when a residual current I.sub.f of 30 mA occurs in addition to inherently present leakage current I.sub.ab1 (operational currents are not taken into account).

(17) FIG. 3 shows a simulation model of a converter 10 with a voltage displacement 12 for generating a common-mode voltage as a measuring voltage U.sub.g.

(18) In the simulation model, converter 10 is modelled as an inverter 14 whose power semiconductor switches SW10 to SW60 are controlled by a control circuit 16 by means of pulse patterns sa, sb, sc. In the model, the generation of pulse patterns sa, sb, sc for controlling power semiconductor switches SW10 to SW60 using the STC modulation method (sine-triangle comparison in the center aligned mode) is shown. The modification of control pulse patterns sa, sb, sc for generating an optimized measuring voltage U.sub.g as explained below is also possible using other known modulation methods (STC-THI, CSVM, DSVM type A, DSVM type B, DSVM type C).

(19) In the STC method described, control pulse patterns sa, sb, sc are basically modulated in a known way by modulation of a higher-frequency triangular carrier signal C.sub.r with standardized sine reference voltages U.sub.L1, U.sub.L2, U.sub.L3 in such a manner that output voltages U.sub.ab, U.sub.ac, U.sub.bc as sinusoidal as possible are generated.

(20) Without voltage displacement 12, i.e. when carrier signal C.sub.r is controlled symmetrically, no significant common-mode voltage would occur as a measuring voltage U.sub.g at the output of the inverter (cf. signal waveform of measuring voltage U.sub.g in FIG. 4a).

(21) However, if the control pulse pattern generation in control circuit 16 is modified by voltage displacement 12 to the effect that a bias voltage U.sub.m is superimposed on respective sine reference voltages U.sub.L1, U.sub.L2, U.sub.L3 for each phase L1, L2, L3, a common-mode voltage against ground that is measurable and exploitable as measuring voltage U.sub.g occurs at the output of inverter 14 (cf. signal waveform of measuring voltage U.sub.g in FIG. 4b).

(22) FIG. 4a shows signal waveforms that occur as simulation results from the converter simulation of FIG. 3 without voltage displacement 12.

(23) To illustrate the basic effect of an inverter operation, the three sine reference voltages U.sub.L1, U.sub.L2 and U.sub.L3 are shown, which are offset from each other by 120°, have a frequency of 50 Hz and each modulate higher-frequency (2 kHz) triangular carrier signal C.sub.r. As a superimposition of the center tapping of the bridge branches of inverter 14, resulting output voltages U.sub.ab, U.sub.ac and U.sub.bc do have largely square impulses, but they contain the frequency of sine reference voltages U.sub.L1, U.sub.L2 and U.sub.L3 as an undesired fundamental oscillation.

(24) The occurring common-mode voltage (measuring voltage U.sub.g) amounts to only a few 100 millivolt and is not usable, or usable to a limited degree only, as a measuring voltage U.sub.g for insulation monitoring or for locating insulation faults in power supply networks.

(25) FIG. 4b shows the signal waveforms as simulation results of the converter simulation of FIG. 3 with voltage displacement 12.

(26) Voltage displacement 12 superimposes a bias voltage U.sub.m having a square waveform, an amplitude of ±40 mV and a fundamental frequency of 10 Hz on each of sine reference voltages U.sub.L1, U.sub.L2 and U.sub.L3.

(27) As a result, a common-mode voltage against ground that has an amplitude of ±10 V and also a fundamental frequency of 10 Hz occurs as a measuring voltage U.sub.g (take note of the time scale compressed in comparison to FIG. 4a). Given that this measuring signal frequency of 10 Hz is significantly lower than the network frequency of 50 Hz, measuring voltage U.sub.g thus generated can be used to determine the insulation resistance even if great network leakage capacitances are present because capacitive leakage currents L.sub.ab1 (FIG. 2) have largely abated and do not distort the measurement result.

(28) FIGS. 5a and 5b show applications of the insulation monitoring according to the invention for an ungrounded DC power supply system (FIG. 5a) and for an ungrounded AC power supply system (FIG. 5b).

(29) FIG. 5a shows a converter 10 which is configured as a rectifier 24 and which is operated grounded via a power source 4. An ungrounded DC power supply system comprising active conductors L+ and L− is connected to the output side of rectifier 24. Leakage capacitances C.sub.e and insulation resistances R.sub.f of ungrounded power supply system 20 are illustrated as concentrated elements against ground PE (protective conductor PE).

(30) A monitor 26 measures measuring voltage U.sub.g provided according to the invention by rectifier 24 an the one hand and receives the value of a measuring current ((common-mode) residual current) I.sub.g caused by measuring voltage U.sub.g on the other hand, said measuring current I.sub.g being detected as a sum current on active conductors L+ and L− by a residual current measuring device 6. Insulation resistance R.sub.f is calculated from these two variables.

(31) By modifying control pulse patterns sa, sb, sc in controlled rectifier 24 (FIG. 3), measuring voltage U.sub.g can be adapted to ungrounded power supply system 20 in such a manner that—unlike in case of a residual current-based detection without an adapted measuring signal according to FIG. 2—leakage capacitances C.sub.e and accompanying leakage currents I.sub.ab1 do not have any substantial deteriorating effect on the measuring result.

(32) Thus, the method according to the invention for determining insulation resistance R.sub.f provides the same functionality with the same reliability as a standard insulation monitoring device (IMD), with the advantage that no external common-mode measuring voltage source requiring elaborate installation is required.

(33) Moreover, the magnitude of leakage capacitance C.sub.e can be determined in addition to insulation resistance R.sub.f in monitoring device 26.

(34) FIG. 5b shows an application in which converter 10 is configured as a frequency converter 34 which is operated grounded via a power source 4. An AC power supply system 30 is connected to the output of frequency converter 34. Here, too, insulation resistance R.sub.f and, if applicable, leakage capacitance C.sub.e are determined in a monitor 26, which assesses measuring voltage U.sub.g generated by the frequency converter and measuring current I.sub.g caused by said measuring voltage U.sub.g.

(35) FIGS. 6a and 6b show a localization according to the invention of insulation faults in a branched ungrounded DC power supply system (FIG. 6a) operated by grounded rectifier 24 and in a branched ungrounded AC power supply system (FIG. 6b) supplied by grounded frequency converter 34.

(36) Branched ungrounded DC power supply system 40 of FIG. 6a has two system branches 41, 42 to each of which a load 43, 44 is connected and each of which can be separated from the main system using a separating device 49.

(37) Analogously to the main system, system branches 41, 42 and system branches 51, 52 of FIG. 6b have—like any power supply system—inherently present (partial) insulation resistances and (partial) network leakage capacitances, neither of which are shown.

(38) In the event of an insulation fault R.sub.x in a system branch 41 due to faulty insulation on a load 43 as illustrated, for example, a closed current path is formed, which closes via insulation fault R.sub.x and grounded rectifier 24 and in which a residual current I.sub.x is flowing.

(39) An assessing device 46 is disposed in each of system branches 41, 42. Assessing devices 46 each have a residual current measuring device 6 for branch-selective detection of (potential) residual current I.sub.x which is driven as a common-mode current by measuring voltage U.sub.g (common-mode voltage) generated in rectifier 24. Assessing devices 46 each further comprise a microprocessor 47 and an interface 48 for transmitting detected residual current h.

(40) An enhanced monitor 45 additionally has a (receiving) interface 48 for receiving residual current I.sub.x transmitted by assessing devices 46.

(41) Based on residual current measuring devices 6 installed in the main system and in system branches 41, 42, the current path formed by residual current I.sub.x can be traced and, thus, faulty system branch 41 can be detected.

(42) In the event that the power supply system is fault-free, the (partial) insulation resistance and/or the (partial) network leakage capacitance of respective monitored system branches 41, 42 can be calculated in enhanced monitor 45 from measuring voltage U.sub.g measured by enhanced monitor 45 and from (common-mode) measuring currents flowing in respective monitored system branches 41, 42 and transmitted to enhanced monitor 45 by assessing devices 46.

(43) The assessment results can be locally stored, locally displayed or transmitted to a superordinate control center in order to initiate corresponding safety measures.

(44) Alternatively, the (partial) insulation resistance value of system branch 41, 42 to be monitored and, optionally, its (partial) network leakage capacitance can be determined directly in respective assessing device 46. In this case, enhanced monitor 45 distributes the information on measuring voltage U.sub.g to assessing devices 46 so as to allow branch-selective determination in respective assessing device 46 “on site”.

(45) System branches 41, 42 that are assessed as critical can be turned off via respective microprocessor 47 and separating devices 49.

(46) In terms of structure and function, the illustration in FIG. 6b basically corresponds to the illustration in FIG. 6a with the exception that instead of the branched ungrounded DC power supply system 40, a branched ungrounded AC power supply system 50 comprising two system branches 51, 52 is connected to the output of a frequency converter 34 operated grounded. Multi-phase loads 53, 54 are connected to system branches 51, 52, insulation fault R.sub.x having occurred in one load 51.

(47) Each system branch 51, 52 has an assessing device 56 comprising a microprocessor 47, a residual current measuring device 6, and a three-phase separating device 59 controlled via microprocessor 47.