WO2015087098A1

WO2015087098A1 - Method for measuring the insulation resistance of an ungrounded dc power network and for localizing the earth fault, and current injecting two-pole device

Info

Publication number: WO2015087098A1
Application number: PCT/HU2014/000125
Authority: WO
Inventors: József PRAUSE
Original assignee: Prause József
Priority date: 2013-12-13
Filing date: 2014-12-15
Publication date: 2015-06-18
Also published as: RU2666803C1

Abstract

A current injecting two-pole device (12) is inserted either between the network's negative terminal (2) and the ground terminal (8), or between the ground terminal (8) and the network's positive terminal (3). The two-pole device (12) has a capacitor (20) and a diode (15) injecting current into the network and restoring the network's voltage conditions relative to the ground terminal to near ground fault free conditions. The insulation resistance of the network is defined as the quotient of the DC voltage measurable between the two-pole device's two terminals (13,14), and the direct current flowing through the device. In networks with high resistance ground fault, the fault is located by maintaining the restored voltage and tracking the direct current injected into the network from the device.

Description

METHOD FOR MEASURING THE INSULATION RESISTANCE OF AN UNGROUNDED DC POWER NETWORK AND FOR LOCALIZING THE EARTH FAULT, AND CURRENT INJECTING TWO-POLE DEVICE

The invention relates to a method for measuring the

insulation resistance of an unground DC power network and for localizing of ground fault, which method is best

0 applicable to DC networks with large ground capacity, to localize of one or more simultaneously occuring high- resistance ground faults. Furthermore the invention relates to a current injection two-pole device, preferably for the execution of the method.

5

In the present specification, the use of the term „DC networks with large ground capacity" applies to networks with 1000 μΕ or larger ground capacity, whilst „high- resistance ground fault" refers to insulation faults in the0 50-100 kQ range.

One of the advantages of the ungrounded (ground insulated) DC power networks is, in a small-range network, allowable voltage and current conditions occur from a life-safety

5 aspect in the event that the high voltage conductor is

accidentally touched. Another is that a ground fault causing a disconnection to the power supply does not result in a power cut to the consumer. Yet another advantage in such networks is that, in the network's control and operating0 circuits, due to the network's positive polarity, faults occurring in open contact circuits can be detected. However, even with the ungrounded DC networks, with a second ground fault an excessively high contact voltage can occur and the second ground fault can cause a double ground fault, which results in a power cut at the consumers as the network is restored. Thus it is important to locate the first ground fault quickly and restore the network.

The use of well-known and widely used built-in devices to monitor the insulation conditions of the ungrounded DC networks will, if the deterioration of the network's insulation reaches a certain pre-defined value, provide information on whether the ground fault occurred in the positive terminal, usually the L+ bus, or the negative terminal, usually the L- bus. Some devices will also display a fault value in ohms.

There are other insulation monitoring devices, which can help locate the ground fault by selecting the branch of the ground fault and then sending a mobile device to the selected branch to find the fault. These devices create a square wave AC between one of the network terminals and the ground terminal and, by sensing the magnetic field of the AC, either through a CT coil or a clamp meter, the fault can be located. The intensity of the created current and the time it takes to measure are greatly affected by the

capacity of the network compared to the ground. In cases of large capacity, this can greatly increase measurement time. The size of the ground capacity depends on the nature of the network and its switch operating status, which can range from a few pF to several thousand μΓ . There are several known methods for measuring the

insulation resistance of the ungrounded DC networks. A common feature of one group of this methods is that, either using the voltage of the network to be examined, or by using an external voltage source, a square wave voltage jump is created between one of the network terminals and the ground terminal and this is usually connected to the network through an ohm resistor. Subsequently, by measuring the resulting currents and voltages and implementing complicated calculations, the insulation resistance of the network is defined .

This method is discussed in the insulation resistance measurement described in patent No. EP 0654673B1, in which, as a result of the power surge from the square wave voltage connected to the network, the network capacity is filled up by the connecting step voltage, to then be discharged via the insulation resistance. In DC networks, measurement needs enough time to ensure the attenuation of the emerging power surge because the value of the insulation resistance is calculated based on the stabilised values. The generator of the square wave voltage is placed between the ohm resistors interface midpoint and the ground terminal, which is

connected between the positive and negative terminals, thus, in one measurement cycle, alternating positive, then

negative voltage is supplied at specified intervals.

The disadvantage of this measurement method is that, in a network with large ground capacity, with a high-resistance ground fault, the higher current needed to separate the ground fault branch cannot be supplied without at the same time risking tripping the ground fault protective release relay, thus fault finding is restricted.

Another known method is described in patent No. JP

2007240426A, which can be applied to small ground capacity networks in vehicles. The voltage changes resulting from the vehicle's dynamic changes (accelerating, braking) adversely affect insulation measurement. In order to avoid this, the voltage from a charging and discharging capacitor is used. This method cannot be applied to power

distribution networks because their ground capacity is a great deal larger than the described circuit, which uses an insertable capacitor for insulation measurement.

The method described in patent No. US 2012/0119754A1 is for monitoring insulation conditions in DC energy distribution networks in which a dual-polarity pulsating voltage is applied between the ground terminal and another network terminal. In the method described in patent No. EP 1586910B1 the network is also monitored by connecting a pulsating voltage between a network terminal and the ground terminal, as well as an initial DC offset of a value fitting to the network. In both methods the value of the insulation

resistance is calculated using readings taken following the dying down of the pulsating voltage. The disadvantage of these methods is that the connected pulsating voltage jump dies down very slowly in large ground capacity networks with high insulation resistance, thus measuring takes a long time, whilst not enough perceivable current is supplied for fault finding, making it unsuitable for finding high resistance ground faults. In the method described in patent No. US 2012/0126839A1, a generator providing the sum of alternating DC and AC voltage is connected through a measuring resistance and other with it in series connected resistances to the ground terminal and the DC network's positive and negative terminals, or between the ground terminal and the positive terminal.

Insulation resistance is calculated using readings from the DC voltage and direct current connected to the network. The disadvantage of the method is that in the case of certain types of fault, eg. a ground fault occurring between the controlling positive end relay contacts of the network' s protective release relay and the coil of the release relay at the negative end, the resistance and generator connected to the positive terminal create such voltage on the release relay coil, which emerges in resistance to the fault, that then operates the ground fault protective release relay and provides a false measurement. If, on the other hand, in order to avoid a mistaken release, we restrict the current as the connecting resistance increases, the flow of current to the fault location will no longer be perceivable and the fault cannot be located.

In the other group of insulation measuring methods the value of insulation resistance is defined by the voltage and current changes of monitoring resistances alternately connected between the DC network's positive terminal and the ground terminal and the negative terminal and ground terminal .

Such a measuring method is described in patent No. RU

2281521, where insulation resistance is defined as the quotient of the voltage difference before and after

connecting a monitoring resistance, and the current that flows through the monitoring resistance. One disadvantage of the method is that fault finding cannot be carried out on a voltage-free part of the network. Another is that, in networks containing control circuits, during fault finding the ground fault in any protective release relay coil in the network will cause false tripping if the relay on the monitoring resistance attached to the positive terminal receives more than its response current, because that would form a closed circuit from the ground terminal, through the relay coil to the negative terminal, thus mistakenly

operating the release relay. The Schneider Electric Company's (France) recommended device for measuring insulation resistance in DC networks is known. The device is connected to the network's positive and negative terminals, as well as the ground terminal. The device, which is recommended for fault finding in DC

networks, includes a portable generator, receiver unit and a probe-type clamp meter. During fault finding the generator injects a 2.5 Hz, 2.5 mA alternating current in the DC network, which can be read from the portable receiver unit powered by the clamp meter. The applicability of the device is limited because, in the case of high-resistance ground fault, eg. ground fault incurred through a relay coil, the detectable alternating current on the faulty branch is considerably smaller than the current draining through the ground capacities of the other branches, thus the

determination of the fault location is dubious.

Another known method is that described in patent No. DE 3819880C2, in which the discovery of the ground faulting lead is proposed by inserting an electrical load between a selected terminal of the network and the ground terminal.

The current flowing through the load is split in the second or fifth sub-harmonic rhythm of the alternating voltage and, by sensing the pulsed magnetic field created by the current, the fault is found. The disadvantage of the solution is that, in the case of high resistance ground fault, the alternating current flowing down the faulty branch and its magnetic field are considerably smaller than the alternating current flowing through branches with large ground capacity and their magnetic field, thus the faulty branch cannot be selected.

Also known is the Bender Company's (Germany) insulation monitoring device and their sensor device to display

insulation faults. For fault finding, the path of the monitoring current injected by the insulation monitoring device is followed by AC rings or clamp meters. The

monitoring current consists of 40 V voltage, injected through 180 kQ resistance in a 2 s positive, 4 s break, 2 s negative, 4 s break rhythm alternating, ultimately 0.16 Hz current. The applicability of the device is limited because, with 150 F ground capacity, a max. 7 kQ fault location can be defined, because in faults larger than this, the

current, besides the fault location, also closes through any parallel ground capacities and gives a faulty signal.

A common disadvantage of the described methods is that they are limited when measuring the insulation resistance of large ground capacity, ungrounded DC networks voltage, requiring a long time for measurement and being unsuitable for localizing of high-resistance ground faults.

Thus it became our task to develop a method which, avoiding the listed deficiencies and disadvantages, can provide a solution for the quick measurement of the insulation

resistance of large ground capacity, ungrounded DC networks under voltage, and can define the location of a high

resistance fault or several, simultaneously occurring faults in such networks, for example in such cases where one of the ground faults occurs in the positive terminal of the

network, whilst the other ground fault occurs simultaneously in the negative terminal.

The basis for the invention is that, in ungrounded DC networks, with no ground fault, the ground capacities are essentially filled to half the network's voltage. When ground fault occurs, current leaks from the ground

capacities, as energy reservoirs, through the ground fault location leaking to the ground terminal, thus decreases the voltage in the network terminal burdened by ground fault.

The invention is based on the recognition that if, when high resistance ground fault occurs, the current draining from the ground capacities is constantly replaced by current injections, then, with the continuous refilling of the ground capacities, we can on the one hand restore the network's voltage conditions relative to the ground terminal to being close to those of fault-free conditions. On the other hand, we can generate and maintain the leakage current needed to locate the high resistance ground fault without risking false triggering of the network's protective relays. If we measure the direct current occurring from the current injecting device and the DC component of its current, then the fault resistance can be calculated from the readings and the fault can be located by tracing the leakage current.

When executing the task through the method according to the invention, the insulation resistance of an ungrounded DC networks is measured and, when ground fault occurs, the fault location is determined by inserting in the network a two-pole device, which device provides an alternating half- wave voltage, or a square wave voltage or a one-way DC voltage into the network, whereby the network' s insulation resistance is defined as the quotient of the DC voltage measurable between the two terminals of the two-pole device and the direct current flowing through the device.

The essence of the method is that

the operation of the network is continuously maintained under voltage,

there is applied a such current injection two-pole device that fills the DC network's capacities, which current injecting two-pole device has a capacitor and a diode connected in series between its terminals as well as a resistance connected in parallel with the capacitor, and where the capacitor is filled by rectification according to the peak voltage of an alternating voltage or a square wave voltage or to a DC voltage, and

the current injecting two-pole device is inserted

either between the network' s negative terminal and the ground terminal,

or between the ground terminal and the network' s positive terminal ,

performing the insertion either directly or through a switch alternatively or with a disconnectable connection and keeping the diode connected in the lead direction,

the DC voltage is measured between the inserted two-pole device's terminals and

the direct current is measured flowing through the two-pole device, and

the value of the insulation resistance between the ground terminal and the network's negative terminal, or between the ground terminal and the network' s positive terminal is determined from the measured values, furthermore

in the case of high resistance ground fault

either the ground capacities between the network' s negative terminal and the ground terminal are filled with current created by the current injecting two-pole device inserted between the network' s negative terminal and the ground terminal and restore the voltage on the network's negative terminal compared with the ground terminal to near fault- free conditions, or the ground capacities between the network' s positive terminal and the ground terminal are filled with current created by the current injecting two-pole device inserted between the network' s positive terminal and the ground terminal and restore the voltage on the network' s positive terminal compared with the ground terminal to near fault- free conditions,

and the fault is located by maintaining the restored voltage and tracking the direct current flowing through the current injecting two-pole device.

When executing the method of the invention, in the event that the asymmetric voltage suggesting ground fault does not occur between the network' s positive terminal and the ground terminal, or the network's negative terminal and the ground terminal, the possibility of a symmetrical ground fault affecting both terminals and the value of the insulation resistance are defined by inserting the current injecting two-pole device between the network' s negative terminal and the ground terminal and measuring the current flowing trough the device, whilst the resulting insulation resistance is defined as the quotient of the DC voltage change caused by the current injection from the

two-pole device and the measured direct current according to the formula below:

RisoPN — (Up-poli— ~ U_F-p₀i-) / 1 =

where

RisoPN is the value of the resulting insulation resistance U_F-poii= is the DC voltage measured between the network' s negative terminal and the ground terminal during inj ection,

U_F-poi= is the DC measured between the network' s negative

terminal and the ground terminal without injection, 1₌ is the current flowing from the dual-polarity

injecting device through the network's negative terminal to the ground terminal.

It is advantageous when executing the method of the

invention, the peak value of the alternating voltage, or the square wave voltage or the DC voltage issued from the current injecting dual-polarity device should be no more than 75% of the nominal voltage of the DC network.

Another beneficial execution of the method of the invention is when the current issued by the current injecting

two-pole device to fill the network' s ground capacities is split, preferably between 5-12 s intervals, particularly at 8 s intervals.

Yet another beneficial execution of the method is in the case of a network with a ground fault, when the current issued by the injecting two-pole device is followed using an instrument that detects the magnetic field created by its alternating voltage sub-harmonic frequency component.

A further advantageous execution of the method is when defining the location of a ground fault within the network; in certain branches of the network, the direct current flowing through the current injecting dual-polarity device is measured, ideally using a DC current transformer or a DC clamp meter and, in the branch showing a sharp difference in measured values, measuring is continued, ideally using a DC current transformer or a DC clamp meter, to uncover the location of the fault at the next point showing a sharply different value. During the execution of the method, within any ground faulting apparatus powered by DC DC network, the apparatus is de-energised to locate the ground fault in order to carry out the necessary repairs; following the de-energising of the apparatus, locating continues within, by connecting one of the terminals of the current injecting two-pole device to the ground terminal, and its other terminal to a terminal on the ground faulting apparatus under examination, locating the fault by tracking the path of the current . A further beneficial execution of the method is when in the branches of the network (44, 45, 46) using the current sensors inserted to read the faulty current, and the

readings, which are dependent on insulation resistance, are compared against each other and visually displayed on one or more screens.

The current injecting two-pole device according to the invention, mainly for executing of the method, has two terminals, between the terminals there are connected in series a diode, a limiting resistance, a capacitor and a load resistance, the diode' s cathode is connected to one of the terminals, a shorting switch is connected in

parallel with the limiting resistance, a resistance is connected in parallel with the capacitor and a second diode is connected in parallel with the load resistance, which second diode is connected in the same lead direction as the diode and furthermore the device has a voltage source servicing DC voltage, or AC voltage with or without DC voltage component, or square wave voltage or AC voltage sub- harmonic frequency component, one of the terminals of the voltage source is connected either directly or through a first switch, preferably a multi-contact pair switch, as well as through a rectifier diode, to the joint terminal of the capacitor and the load resistance, whilst the other terminal of the voltage source is connected either directly, or through a second switch, preferably the multi-contact pair switch, to the two-pole device's other terminal.

During the method we attempt to restore the voltage between the ground faulting terminal and the ground terminal, which is reduced due to the ground fault, with an auxiliary voltage having DC voltage component, to its fault-free value. We continuously maintain these elevated converging voltages for ground fault-free operations, by replacing the leakage current flowing through the ground fault using the current injecting two-pole device to fill the network from its ground capacities with continuous DC current, or with AC that has a DC component, and with this the leakage current, which is required to locate the high resistance ground fault, is maintained. By observing the current issued through the current

injecting two-pole device, which is inserted between the ground terminal and each polar terminal of the network, as well as the DC voltages between the network terminals and the ground terminal, we can detect a singular ground fault, or multiple ground faults occurring on either side of the network . In the event of ground fault, the value of the ground fault resistance is determined as the quotient of the DC voltage created through the current injection and the value of the direct current injected by the device. To select the ground faulting branch and locate the ground fault, we use sensors to detect the injected current and track it using known methods.

In our method, in order to track the path of the current, we use a direct DC meter, or a CT, ideally a DC clamp meter, sensor or device that detects magnetic field.

Following the principle of the procedure, if there is no ground fault, after the network's ground capacities have been filled up, the injected current stops, since there is no longer current leaking through the ground capacities towards the ground terminal.

In one execution of the method according to the invention we use a pulsed voltage in the rhythm of the generally available network AC voltage or its sub-harmonic, preferably 50Hz for AC voltage 25 Hz, 60 Hz for AC voltage 30 Hz, to fill the capacitor of the current injecting dual-polarity device in order to inject the network. The pulsed current makes tracking the path of the current easier.

Hereinafter we will present the essence of the method according to the invention in more details and with

reference to preferred embodiment and with reference to the attached drawings, where

Figure 1 shows a general representative circuit of an ungrounded DC network and its AC network parts connected via an inverter, along with the arrangement of its measuring circuit connecting to the DC network and

Figure 2 shows the schematic circuit diagram of a current injecting two-pole device arranged within the measuring circuits.

According to the the figure 1 an ungrounded DC network has one DC power source, that is, a battery, to the negative terminal of which is connected the DC network' s negative terminal 2, mostly L- bus, whilst the network's positive terminal 3, mostly L+ bus, is connected to its positive terminal. In the figure, the ground terminal, or earthed terminal or the ground environment, compared to which ground fault can occur, has been termed ground terminal 8.

The DC network' s terminals 2 and 3 are connected to various types of loads purporting to be groups of consumers; the first type being branch 44, the second branch 45 and the third type of consumer being branch 46. The consumer branches display various possible faults, as well as connecting to a 47 inverter, through which AC network part 48 or AC units, especially electric motors and/or generators are connected.

The DC network shows the network's positive terminal 3, mostly L+ bus, its Ri_SOp insulation resistance relative to the ground is resistance 29, its capacity relative to the ground is capacitor 31, whilst the network's negative terminal 2, mostly L - bus it has Ri_S0 insulation resistance relative to the ground resistance 30, its capacity relative to the ground is capacitor 32. For the overvoltage protection of the DC network,

protective diodes 27 and 28 are inserted respectively between terminals 3 and 2 and the ground terminal 8.

The figure 1 shows three typical consumer branches, 44, 45 and 46. In the branches, consumers 33, 40 and 43 are

connected respectively between the network's terminals 2 and 3. In the individual branches 44, 45 and 46, between the network's positive terminal 3 and its negative terminal 2 and the ground terminal 8, capacitors 35, 37 and 41 as well as 36, 38 and 42 are shown. In branch 44, a positive

terminal 3 ground fault is displayed through resistance 34, in branch 45, a negative 2 terminal ground fault is

displayed through resistance 39, which ground fault is on a section between open contact 58 and release relay coil 57. Besides this, the the figure shows resistance 49,

representing a ground fault affecting one of the phases of the AC network part 48.

For execution of the method according to the invention, we inserted a current injecting two-pole device 12 with measuring circuit arrangement 4 between the network' s two terminals 2 and 3. The current injecting two-pole device 12 has a terminal 13 and a terminal 14, and between the terminals there is a diode determining current direction, which makes only DC current injection possible. The

measuring circuit arrangement 4 includes the two-pole device 12 with an DC ammeter 11 connected in series to the two-pole device 12, and a DC voltmeter 10, as well as a dual- position, three-way switch 9.

Terminal 5 of the measuring circuit arrangement 4 is connected to the network's positive terminal 3, its

terminal 6 is connected to the network' s negative terminal 2 and its terminal 7 is connected to the ground terminal 8.

The measuring circuit arrangement 4 is configured to the dual-polarity device 12 and the ammeter 11 connected to it in series, in one of the positions of the switch 9, the right hand position according to the figure, between the negative terminal 2 and the ground terminal 8. In the other position of the switch 9, the left hand position according to the figure, it is inserted between the ground terminal 8 and the positive terminal 3, in such a way that one of the terminals 13 of the two-pole device 12, to which the diode directing current direction is connected, is connected to the network's positive terminal in both circuits. The voltmeter 10 is connected, regardless of the falling voltage on the ammeter 11, so as to always measure the terminal voltage of the two-pole device 12 in both positions of the switch 9, which in terms of absolute value, matches that of the voltage between the network's affected pole terminal 2 or 3 and the ground terminal 8.

The preferred use of the current injecting two-pole device

12 is shown in the figure 2. The two-pole device 12 has a diode, 15 a limiting resistance 16, a capacitor 20 and a load resistance 18 connected in series between the terminals

13 and 14 of the device 12. The cathode of diode 15 is connected to the terminal 13, connected in parallel with the limiting resistance 16 is shorting switch 19, connected in parallel with the capacitor 20 is a resistance 17 and connected in parallel with the load resistance 18 is a second diode 21, which diode 21 is connected in the same lead direction as diode 15.

The 12 dual-polarity device furthermore has a DC voltage, a component with DC voltage, or an AC voltage, servicing a square wave voltage or AC voltage sub-harmonic frequency component servicing resistance source 26. One of the

terminals 24 of resistance source 26 is connected either directly or through terminal 23, as well as through

rectifier diode 22, capacitor 20 and load resistance 18' s shared terminal, whilst the other terminal 25 is connected either directly or through a terminal, ideally a multi-way 23 terminal to the other terminal of the 12 dual-polarity device . Hereinafter we describe the method according to the

invention in more detail. It is known that single ground faults typically have significant asymmetric voltage between ground 8 terminal and the network's 2 and 3 terminals. It follows from the

recognition, which was the basis for the method that, when inserting the 12 dual-polarity device, current only flows between the switch positions of the terminal of the network showing lower voltage and the ground 8 terminal. It does not flow between switch positions of the terminal showing higher voltage and the ground terminal. It is further known that, when there is no ground fault in the network there is no significant voltage asymmetry in the network between terminals 2 and 3 and the ground terminal 8. It follows from the recognition, which was the basis for the method that no current flows through the 12 dual-polarity device, when it is inserted on either side.

If there is no voltage asymmetry between the network' s 2 and 3 terminals and the ground 8 terminal, yet current flows through the 12 dual-polarity device, whichever side it is inserted into, then there is a double ground fault. The existence of a double ground fault is reinforced by checking that the protective release relays on the negative side of the network have not been falsely triggered. This is done by inserting the 12 dual-polarity device between the network's negative 2 terminal, in or example the L- bus, and the ground 8 terminal. For various ground faults the value of the insulation resistance is the quotient of the voltage difference caused by the injection of the dual polarity device and its current, according to the following:

A./ In order to determine the existence of a double ground fault affecting the positive terminal 3, in this case the L+ bus, and the negative terminal 2, in this case the L- bus, or a ground fault at the mid terminal of the accumulator and measure it

1. According to the figure 1, switch 9 is switched to its right hand position and the two-pole device 12 is inserted between the L- bus and the ground terminal 8.

2. Using the voltmeter 10 we measure the U_F+poi= DC voltage between the L+ bus and the ground terminal 8.

3. Using the voltmeter 10 we measure the U_F-_pol= DC voltage between the L- bus and the ground terminal 8.

4. Using switch 23, we turn on the voltage source 26 of the current injecting two-pole device 12, thus injecting current into the network's negative side.

5. Using the voltmeter 10 we measure the U_F-_poii = DC voltage between the L- bus and the ground terminal 8, which equal with the voltage measurable on the terminals of the two- pole device 12 during the current injection. 6. Using the ammeter 11, we measure the 1= direct current flowing between the two-pole device 12 through the L- bus towards the ground terminal 8.

7. If the above measurements are almost identical voltage readings, that is, the asymmetry is small, and current is still flowing despite this, it means that there is a double ground fault.

8. The value of the resulting insulation resistance Ri_SOP _? the size of which can infer a double ground fault or a ground fault in the accumulator's mid terminal, is defined according to the following formula:

RisoPN

/ 1 =

B./ The measurement of faults occurring within the

accumulator, at the negative terminal 2, in this case the L- bus, or the insulation resistance of the connecting AC network part

2. Using the voltmeter 10 we measure the U_F__pol= DC voltage between the L- bus and the ground terminal 8. 3. Using switch 23, we turn on the voltage source 26 of the current injecting two-pole device 12, thus injecting current into the network's negative side.

4. Using the voltmeter 10 we measure the U_F-_poii= DC voltage between the L- bus and the ground terminal 8, which is egual with the voltage measurable on the terminals of the two-pole device 12 during the current injection.

5. Using the ammeter 11, we measure the 1= direct current flowing between the two-pole device 12 through the L- bus towards the ground terminal 8. 6. The value of the insulation resistance of the network's negative terminal 2, that is the L- bus, is calculated according to the following formula:

RisoN

7. For quick fault finding, the insulation resistance of the negative terminal 2, in this case the L- bus, can be defined with close approximation, when the switch 19 of the two-pole device 12 is closed, according to the following formula:

RisoN

C./ The measurement of the insulation resistance of the positive terminal 3, in this case the L+ bus 1. According to the figure 1, switch 9 is switched to the left hand position and the two-pole device 12 is inserted between the ground terminal 8 and the L+ bus.

3. Using switch 23, we turn on the voltage source 26 of the current injecting two-pole device 12, thus injecting current into the network's positive side.

4. Using the voltmeter 10 we measure the U_F+poii= DC voltage between the L+ bus and the ground terminal 8, which is equal with the voltage measurable on the terminals of the two-pole device 12 during the current injection.

5. Using the ammeter 11, we measure the 1₌ direct current flowing from the two-pole device 12 towards the L+ bus.

6. The value of the insulation resistance of the network's positive terminal 3, that is the L+ bus, is calculated according to the following formula: RisoP

7. For quick fault finding, the insulation resistance of the positive terminal 3, in this case the L+ bus, can be defined with close approximation, when the switch 19 of the two-pole 12 device is closed, according to the following formula : RisoP ^— Up+poii— / 1=

D./ Locating the ground fault

Once the value of the resistance at the ground fault is known, our method makes it possible to locate the fault. By closing the switch 19 of the two-pole device 12 we reduce the internal resistance of the two-pole device 12, thus increasing the amount of current injectable into the network. At full ground fault with a 220 V accumulator, the maximum injectable current is approx. 50 mA. Currents of such mA scale are easily detectable for example with a clamp meter.

When searching for low resistance faults it is an advantage that not only direct current flows towards the fault, but, if the current injection has a sub-harmonic rhythm, as a resulting alternating component of the filling up and discharging at sub-harmonic rhythm will mostly flow towards the fault location, because the low resistance fault will short the ground capacities of the other terminals. This makes it possible to detect alternating current as well, which can be through a clamp meter or for instance a 25 Hz selective magnetic field sensor.

The current injected by the device depends upon the high of the resistance at the fault location. If the network's insulation is good, there is no ground fault towards the ground terminal, after the ground capacities have been filled up, essentially no direct current flows. According to the figure 1, when switch 9 is in its left hand position, the L+ bus stabilises compared with the ground terminal 8 at the peak value of the injected voltage . If on the positive side ground fault occurs in branch 44, this essentially causes the fault location' s 34 resistance to short circuit the L+ bus with the ground terminal 8, it starts to discharge the capacities 31, 35, 37 and 41 between the L+ bus and the ground terminal 8, and the voltage between the L+ bus and the ground terminal 8, which has the same value as the voltage between the terminals 13 and 14 of the two-pole device 12, decreases.

The two-pole device 12 attempts to replace the amount of charge leaking from the ground capacity and injects current into the network. The current injected from the device 12 flows down the faulty branch to the fault location, and from there through the ground terminal 8 back to the two- pole device 12.

According to the figure 1, when switch 9 is in its right hand position, the voltage of the L- bus stabilises compared with the ground terminal 8 at the peak value of supplied voltage. If ground fault occurs on the negative side in branch 45, this essentially causes the fault location's 39 resistance to short circuit the L- bus with - li ^¬ the, ground terminal 8, it starts to discharge capacities 32, 36, 38 and 42 between the L- bus and the ground

terminal 8, and the voltage between the L- bus and the ground terminal 8, which has the same value as the voltage between the terminals 13 and 14 of the two-pole device 12, decreases .

The two-pole device 12 attempts to replace the amount of charge leaking from the ground capacity and injects current into the network. The current injected from the device 12 flows down the faulty branch to the fault location, and from there through the ground terminal 8 back to the two- pole device 12, the path of which is shown in the figure 1 with dotted line 59 and arrows.

In order to locate the ground fault, the path of the current flowing through the current injecting two-pole device 12 is measured by moving along the network's L+, L- bus in individual branches of the network, as marked on individual branches by DC clamp meters 50, 51, 52, 53 and 54. For any branches showing abrupt changes in readings, in this example at the branch 45, measuring continues along the branch, as indicated with the 56 DC clamp meter, locating the ground fault at the place where readings show abrupt changes again, in this example, resistance 39 indicates.

Measuring is done via a DC clamp meter, a magnetic field sensing ammeter, a DC current transformer or any other device capable of sensing the injected current. The method according to the invention and the two-pole device can be used for continuous monitoring of insulation, as well as determining and locating, ground fault and revealing it.

The greatest advantage of the method according to the invention is that it can quickly measure the insulation resistance of ungrounded, high voltage DC networks with large ground capacity and it can determine the location of the fault in a high resistance network with ground fault Thus, using the method, it is possible to determine the location for instance of ground faults of approx. 100 kQ occurring through relay coils, or even to determine the location of 100 kQ ground faults in 5000 xF ground capacity networks .

Another advantage of the method according to the invention is that its application makes it possible to recognise and locate double ground faults affecting the positive and negative side of DC networks, in which double ground fault event, there is no voltage asymmetry between the ground terminal and the network's positive and negative terminals. The method according to the invention makes it possible to get information on single or multiple ground faults

affecting both sides by observing the voltages and currents between the network's terminals, and by sensing the path of the short-circuit current, we can select the faulty branch, as well as finding the fault location. Another advantage of the method according to the invention is, when applied to fault finding, it significantly reduces the usual measuring time of insulation monitoring devices, thus the fault can be found more quickly. The measuring time of known insulation measuring devices in the case of C_E = 2000 pF ground capacity is a matter of hours, whilst the method according to the invention reduces measuring time to a matter of minutes, even with C_E = 5000 F ground capacity.

Another advantage of the method according to the invention is, with DC networks current sensing uses DC current transformer detection. When tracking a bus with ground fault, DC current transformer detection provides additional information, besides the measuring the intensity of the current, on its presage, since it shows the direction of the fault. DC current transformers installed at each branch speed up the fault finding process. The readings of

individual sensors can be transferred into a separate or joint evaluation system, and connected to real time and displayed visually.

Another advantage of the deevice according to the invention is, following the determination of ground fault, fault finding can be continued within the discovered ground faulting equipment, for instance ground fault can be located within a short-circuiting generator or a dismounted motor rotor or stator in order to make necessary repairs. For Solar power stations, faults can be found amongst the panels regardless of sunlight, it can determine whether the solar cells are generating electricity or not, or for an accumulator assembled of several hundred cells, it can determine which parallel unit have ground fault and where. Another advantage is that it is possible to direct current injection with a rhythm switch, eg. 8-10 s on-off switching. This helps select the faulty branch because in on-off mode, the presage of current flowing towards the fault location will not change in the faulty branch, whilst in branches containing large ground capacity, the ground capacities will change presages during filling and discharging, taking on or giving off direct current. After reaching a stabilized state no direct current will flow in the ground capacity. A limitation of the known fault finding devices is, with certain faults, eg. with ground faults occurring in the release relay controls between the positive side relay contact and the release relay coil, which appear on the negative side, the relay could false release due to the current supplied by the fault finding device through the relay coil. In the case of opto couplers, this danger is increased. The low value of allowable current makes it impossible to detect, whereas if a greater current is injected, there is a risk of malfunction. The fault finding application according to our invention reduces the risk of false tripping to a minimum, because in this case current is injected from the negative terminal and is so low, ideally at half the nominal voltage value, which does not reach the operating range of the network's control equipment, at the same time the injected current is a detectable size. With this solution, compared with known methods, we can achieve an additional result for fault finding in large ground capacity networks with a high resistance fault; the larger current needed to detect the faulting branch can be injected without the risk of unwanted false tripping, since the one-sided injection restores the reduced voltage due to the ground fault between the ground faulting terminal and the ground to near fault-free conditions by filling up the ground capacity, which does not occur with the known methods .

Using the method according to the invention, for high resistance ground fault occurring between a release relay's positive side control relay and the release relay coil appearing on the negative side, current injection does not increase the risk of false tripping because injection only occurs between the negative terminal and the ground terminal and with a voltage that does not reach the operating range of the network's control equipment eg. the network

protecting relays.

With voltage free AC equipment, the currant injecting two- pole device according to the invention can be inserted between any point along the AC part and the ground terminal in order to reveal a ground fault.

Claims

Claims :

1./ Method for measuring the insulation resistance of an ungrounded DC power network and, in the case of network with ground fault, locating the ground fault, wherein a two-pole device is inserted in the network, which device provides an alternating half-wave voltage, or square wave voltage, or one-way DC voltage, and the insulation resistance is determined as the quotient of the DC voltage measurable between the two terminals of the two-pole device and the direct current flowing through the device,

c h a r a c t e r i s e d by that

the operation of the network is continuously maintained under voltage,

there is applied a such current injection two-pole device (12) that fills the DC network's capacities, which current injecting two-pole device (12) has a capacitor (20) and a diode (15) connected in series between its terminals (13, 14), as well as a resistance (17) connected in parallel with the capacitor (20), and where the capacitor (20) is filled by rectification according to the peak voltage of an

alternating voltage or a square wave voltage or to a DC voltage, and

the current injecting two-pole device (12) is inserted either between the network's negative terminal (2) and the ground terminal (8),

or between the ground terminal (8) and the network's positive terminal (3) and

performing the insertion either directly or through a switch (9) alternatively or with a disconnectable connection and keeping the diode (15) connected in the lead direction, the DC voltage is measured between the inserted two-pole device's (12) terminals and

the direct current is measured flowing through the two-pole device ( 12 ) , and

the value of the insulation resistance between the ground terminal (8) and the network's negative terminal (2), or the ground terminal (8) and the network's positive terminal (3) is determined from the measured values, furthermore

in the case of high resistance ground faults

either the ground capacities between the network' s negative terminal (2) and the ground terminal (8) are filled with current created by the current injecting two-pole device (12) inserted between the network's negative terminal (2) and the ground terminal (8) and restore the voltage on the network's negative terminal (2) compared with the ground terminal (8) to near fault-free conditions,

or the ground capacities between the network' s positive terminal (3) and the ground terminal (8) are filled with current created by the current injecting two-pole device (12) inserted between the network's positive terminal (3) and the ground terminal (8) and restore the voltage on the network's positive terminal (3) compared with the ground terminal (8) to near fault-free conditions,

and the fault is located by maintaining the restored voltage and tracking the path of the direct current flowing through the current injecting two-pole device (12) .

2 . 1 The method according to claim 1

c h a r a c t e r i s e d by that

in the absence of voltage asymmetry which is suggestive of ground fault between the network's positive terminal (3) and the ground terminal (8), or the network's negative terminal (2) and the ground terminal (8), the existence of a possible double ground fault affecting both terminals and the value of the insulation resistance is establishable by inserting the current injecting two-pole device (12) between the network's negative terminal (2) and the ground terminal (8); and by measuring the current flowing through the current injecting two-pole device (12), and the resulting insulation resistance is determined as the quotient of the DC voltage change caused by the current injection from the two-pole device (12) and the reading of the direct current according to the formula below:

RisoPN

where

RisoP is the value of the resulting insulation resistance, is the DC voltage measured between the network' s negative terminal (2) and the ground terminal (8) during injection

is the DC voltage measured between the network's negative terminal (2) and the ground terminal (8) without injection, I₌ is the current flowing from the two-pole injecting device (12) through the network's negative terminal (2) to the ground terminal (8) .

3./ The method according to claim 1 or 2

c h a r a c t e r i s e d by that

the peak of the alternating voltage issued from the current injecting two-pole device (12) or the square wave voltage or the DC voltage is max 75% of the DC network' s nominal voltage .

4./ The method according to any of claims 1-3

c h a r a c t e r i s e d by that

the current issued by the current injecting two-pole device (12) to fill the network's ground capacities is interrupted, preferably between 5-12 s intervals, particularly at 8 s intervals.

5./ The method according to any of claims 1-4

c h a r a c t e r i s e d by that

for networks with ground fault, tracking the current issued from the current injecting two-pole device (12) by detecting the magnetic field created by its alternating voltage sub- harmonic frequency component using a tuned instrument.

6./ The method according to any of claims 1-4 c h a r a c t e r i s e d by that

by determining a ground fault location in the network the direct current flowing through the current injecting two- pole device (12) is measured in certain branches (44, 45, 46) of the network, especially with a DC current transformer or a DC clamp meter (50, 51 55), and the measuring is continued in the branch (45) displaying an abrupt change in readings, especially with a DC current transformer or a DC clamp meter (56), and the ground fault (39) is located at the next point displaying an abrupt change in readings.

7./ The method according to any of claims 1-6

c h a r a c t e r i s e d by that

for apparatus powered by the DC network, where ground fault has been discovered, locating the ground fault in order to carry out necessary repairs is continued within the

equipment after it has been de-energised, where one terminal of the current injecting two-pole device (12) is connected to the ground terminal (8), the other terminal is connected to one of the terminals of the ground faulting equipment being examined, and the fault is located by tracking the path of the current.

8./ The method according to any of claims 1-7

c h a r a c t e r i s e d by that

in the branches of the network (44, 45, 46) using the current sensors inserted to read the faulty current, readings, which are dependant on insulation resistance, are compared against each other and visually displayed on one or more screens .

9./ Current injecting two-pole device (12), mainly for executing of the method, according to any of claims 1-8 c h a r a c t e r i s e d by that

the device (12) has two terminals (113, 14), between the terminals (13, 14), there are connected in series a diode (15), a limiting resistance (16), a capacitor (20) and a load resistance (18), the diode's (15) cathode is connected to one of the terminals (13), a shorting switch (19) is connected in parallel with the limiting resistance (16), a resistance (17) is connected in parallel with the capacitor (20) and a second diode (21) is connected in parallel with the load resistance (18), which second diode (21) is connected in the same lead direction as the diode (15) and furthermore the device has a voltage source (26) servicing DC voltage, or AC voltage with or without DC voltage component, or square wave voltage or AC voltage sub- harmonic frequency component, one of the terminals (24) of the voltage source (26) is connected either directly or through a first switch, preferably a multi-contact pair switch (23), as well as through a rectifier diode (22), to the joint terminal of the capacitor (20) and the load resistance (18), whilst the other terminal (25) of the voltage source (26) is connected either directly, or through a second switch, preferably the multi-contact pair switch (23), to the two-pole device's (12) other terminal (14).