Detection of faults on transmission lines in a bipolar hiαh-voltaσe direct-current system
TECHNICAL FIELD
An HVDC installation comprises a plurality of protective functions. One important such function is the line fault protection. A line fault normally entails loss of transmission capacity and therefore needs to be corrected as quickly as possible. The interruption interval comprises time for detection of the occurrence of a fault, time for determining whether the fault is a line fault or an external fault, time for discharge of the transmission line, time for deionization of the arc occurring, time for elimination of the fault and time for start-up of the transmission. To minimize the interruption interval, it is important that all sub-times be kept as low as possible. The present invention relates to a method for detecting when a fault has occurred on transmission lines of bipolar high-voltage direct-current systems, and a device for carrying out the method.
BACKGROUND ART, THE PROBLEMS
To be able to place the invention in its proper context, a short description of high-voltage direct-current transmissions will first be given. Such transmissions will be referred to below as HVDC transmissions . An HVDC transmission is an electric power transmission using direct current and DC voltage. Since electric power generation takes place by means of alternating current/voltage (AC power) , current and voltage must be converted, for HVDC transmission, into direct current/voltage (DC power) at one end of the HVDC line and then be converted into alternating current/voltage at the other end of the line. In general terms, when it is a question of HVDC, the equivalence of an AC phase is referred to as a pole.
For HVDC transmission there are substantially two different systems which are described as monopolar and bipolar transmission systems, respectively.
Figure 1 shows in simplified form a diagram for a mono- polar system, often referred to as a single-pole HVDC transmission. From AC switchgear 1 the voltage is stepped up to the desired values. By using two transformers 2 and 3, one of them Y-Y-connected and the other Y-D-connected, a so-called 12-pulse-directed current/voltage may be obtained via a converter 4. The harmonics of the direct current and the direct voltage are limited by reactances and capacitances present in the converter station, in Figure 1 symbolized by a smoothing reactor 5 and a filter 6. The HVDC power is transmitted via a transmission line 7 or a "pole" which may be an overhead line, a ground cable or a submarine cable. On the receiver side there is a corresponding converter station with a filter 8, a smoothing reactor 9 and a converter 10 operating as an inverter. Via the transformers 11 and 12, the transmitted HVDC power may then be converted into ordinary three-phase alternating voltage for feeding AC switchgear 13. The direct current is returned through ground or sea and is connected co the respective station via ground electrodes 14 and 15. Power transmission may take place in both directions .
A bipolar transmission system is clear from Figure 2 and comprises substantially two balanced monopolar transmission systems. Such a system is used, inter alia, for increasing both the capacity and the availability and when the ground current must be limited or for various reasons is not allowed to occur. To simplify the description, the same reference numerals as in Figure 1 are used in Figure 2 for the various parts, supplemented by an "a" for one of the two monopolar transmission systems and by a "b" for the other system. As an example, the transmission line of one of the systems is designated 7a and the transmission line of the other system is designated 7b. Both poles have the same rated voltage with opposite signs. This implies that zero potential occurs between the two converters 4a and 4b, respectively, and between the two inverters 10a and 10b, respectively. These two zero-potential points are connected to ground via the electrodes 14 and 15 and thus form a possible ground line for a difference current, if any, between the two transmission lines. A difference current occurs in the case of unbalanced operation or when a fault has occurred on any of the transmission lines.
A summary of the prior art as regards detection of faults on HVDC transmission lines is described, inter alia, in a publication "Line Fault Detection for HVDC Overhead Lines" from TAMPERE UNIVERSITY OF TECHNOLOGY, Diploma thesis, by Jarvi, Seppo, published 31.05.1989, pp. 9-12 and pp. 72- 76.
The original technique comprised registering both current and voltage after the smoothing reactors in one of, or both of, the converter stations. A typical line fault, for example a short circuit, implies that a voltage drop and an increase in current occur. Problems which arise when operating in this manner are difficulties in discrimi- nating between a line fault, an external fault or another disturbance and that the detection takes a relatively long time .
The next step in the technical development for detecting faults on HVDC transmission lines was the introduction of a derivative criterion, a level criterion as well as certain time criteria as follows:
If the derivative of the line voltage exceeded a certain reference value for a given period of time and if the line voltage was below a certain reference value for a given period of time, this was interpreted as a fault on the transmission line.
When a fault occurs on a transmission line, so-called travelling waves arise which propagate in both directions, as viewed from the location of the fault. When these waves hit a station, the travelling waves are reflected and may in this way bounce back and forth. Such travelling waves arise both on AC and DC transmission lines and have been utilized for several decades for detection and fault location on power lines.
A method which utilizes travelling waves for this purpose on AC transmission lines is described in US 4,719,980, "Detection and Location of a Fault Point Based on a Travelling Wave Model of the High Voltage Transmission Line" .
In an article "Development and Field Data Evaluation of Single-end Fault Locator for Two-terminal HVDC Transmission Lines", published in IEEE Trans on Power Apparatus and Systems, Vol. PAS-104, No. 12, December 1985, pp. 3531-3537, a technique is described based on travelling waves for fault location in a bipolar HVDC system. The technique utilizes successive reflections from the fault and from the ends of the transmission lines.
An HVDC transmission line possesses both capacitive, inductive and resistive properties. When a fault occurs on such a line, the change in current and voltage will be dependent on the impedances of the line as well as on the travelling waves occurring. By studying the frequency and rate of propagation of the travelling waves, it may be determined that in order to measure current and voltage characteristics reasonably correctly in connection with the occurrence of a fault, measurement with a sampling frequency of about 15-20 kHz is required. Measurement and registration of current and voltage characteristics, measured with a relatively long sampling frequency, will thus not correctly indicate the actual characteristics comprising the travelling waves which will occur.
In the above-mentioned Diploma thesis by Jarvi, Seppo, a fault detection on a bipolar transmission system is described by means of a block diagram according to Figure 3. It is assumed here that the current and voltage measurement takes place in such a way that travelling waves occurring are comprised in the measurement. This in turn means that relevant values of current and voltage derivatives may be obtained. The general criterion that a fault has occurred is that
kx dl/dt - k. dU/dt > k3
where the constants are positive and dependent on the installation. To this are added, from a purely practical point of view, according to Figure 3, certain level discrimination, time delay and monitoring with respect to a non-faulted transmission line.
BRIEF DESCRIPTION OF THE DRAWINGS
Figure 1 shows a simplified diagram for a monopolar HVDC installation. Figure 2 shows a simplified diagram for a bipolar HVDC installation.
Figure 3 shows a block diagram for fault detection on a transmission line in a bipolar HVDC installation according to the prior art.
Figure 4 shows a block diagram for fault detection on a transmission line in a bipolar HVDC installation according to the invention.
Figure 5 shows a flow diagram for detection whether a fault has occurred on a transmission line in a bipolar HVDC installation according to the invention.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
A method for detecting a fault, according to the invention, on an HVDC transmission line in a bipolar HVDC installation will be described on the basis of a block diagram according to Figure 4. At the same time, the block diagram represents an embodiment of a fault detector for faults on HVDC transmission lines. To be able to explain the invention in the simplest manner, the block diagram is built up of logic components, comparison elements, summa- tor, multiplier, etc. The invention may also be in the form of a program implemented in a computer.
The invention is based on such measurement of the current and voltage of the poles that the travelling waves which arise in connection with a fault on any of the transmission lines may be detected. An indication that a fault has been detected is dependent on a number or criteria being fulfilled. As input signals to the fault detector, three so-called wave signals are formed, consisting of one positive pole- wave signal, one negative pole-wave signal and one ground- wave signal. With the aid of a high-frequency-sampling measurement, the positive pole current I+ and the positive pole voltage U+, the negative pole current I_ and the negative pole current U. are obtained. With knowledge of the impedance Z+ of the positive transmission line, the impedance Z_ of the negative transmission line and the impedance Z, of the ground line, the positive pole-wave signal is formed as
PPV = Lr - U+
and the negative pole-wave signal is formed as
NPV = I Z - U
and the ground-wave signal is formed as
JV = Z.(It + I.)/2-(U<. + U.)/2
The positive pole-wave signal is passed to a first input circuit 16 which consecutively forms a first difference signal d1 by comparison between two consecutive samples .
The negative pole-wave signal is passed in the same way to a second input circuit 17 which consecutively forms a second difference signal d2 by comparison between two consecutive samples .
A first criterion for being able to identify a fault on one of the transmission lines in a bipolar HVDC installation is that one or both of these difference signals is/are to have a value which exceeds a pre-set threshold value d0. Whether this is the case or not is determined by means of a first comparison circuit 18. If this is the case, the output signal of the first comparison circuit influences a closing element 19.
In a summator 20 the numerical value of the difference dα, between the difference signals is formed, dd = d1 - d2, which value is led to the above-mentioned closing element 19 and, if the first criterion is fulfilled, further to a second comparison circuit 21.
In a "MAX" value element 22 it is determined which of the two difference signals has the highest value. The highest value, ' dmax, ' is led to a multip-1-lier 23 where it is multi- plied by a pre-set factor fx . The product thus formed, p = fλ dmax, is led to the above-mentioned second comparison element 21.
A second criterion for being able to identify a fault on one of the transmission lines in a bipolar HVDC installation is now that the product p shall be greater than the difference between the difference signals, that is, p>dd, which gives the second comparison circuit 21 a logic "zero" as output signal, which signal is then inverted into a "one" in an inverter 24 which is then led to an "AND" element 25.
A third criterion for being able to identify a fault on one of the transmission lines in a bipolar HVDC installation is that a ground-wave signal JV has been detected. If this is the case, a "one" is obtained via a logic converter 26 and is supplied to the above-mentioned "AND" element 25.
Since the two inputs of the "AND" element are not set at "one" , the device will thus indicate a fault on one of the transmission lines in a bipolar HVDC installation. If the difference between the difference signals is greater than the above-mentioned product, that is, if dd>p, and if both the first and third criteria are fulfilled, this indicates that an external fault has occurred.
In summary, a fault on a transmission line in a bipolar HVDC installation can be determined if
- any of the difference signal d1 or d2 is greater than an assumed threshold value d0, and if
- the difference dd between the difference signals is smaller than the product p of the greatest of the difference signals and an assumed factor fx , and if
a ground-wave signal has been detected.
Figure 5 shows a program-related device in the form of a flow diagram for detection of a fault on a transmission line in a bipolar HVDC installation. The program is provided with information about the impedance of the two transmission lines and the ground line and is supplied with consecutively measured currents and voltages for the two lines. In accordance with the method described above:
the positive pole-wave signal PPV, the negative pole-wave signal NPV and the ground-wave signal JV are formed,
the two difference signals d1 and d2 are formed,
these two signals are compared with the threshold value d0,
the numerical value of the difference between the two difference signals dd is formed, it is determined which of the two difference signals has the hig —'hest value dmax,'
this value is multiplied by a factor f± l
the numerical value of the product p= f, d is formed,
an output signal is formed which indicates that a fault on any of the transmission lines has occurred if both p>dd and if a ground-wave signal has been detected.