Converter assembly for power transmission bv means of hiσh- voltaσe direct current
TECHNICAL FIELD
The present invention relates to a converter assembly for power transmission by means of high-voltage direct current. The assembly comprises a first, current-source, line-commuta- ted converter and a second, voltage-source, self-commutated converter. The converters are d.c. series-connected to each other and have a.c. terminals for connection to an alterna¬ ting-voltage network. The second converter is provided with control members for controlling the direct voltage of the converter by influencing the phase position of the internal alternating voltage of the converter and hence the flow of active power between the converter and the alternating-voltage network.
BACKGROUND ART
Converter systems for power transmission by means of high- voltage current are previously well-known, for example from Erich Uhlmann: "Power Transmission by Direct Current", Springer-Verlag, Berlin-Heidelberg-New York, 1975 (hereinafter referred to as "Uhlmann") , and from Ake Ekstrόm: "High Power Electronics, HVDC and SVC", Electric Power Research Center, The Royal Institute of Technology, Stockholm, June 1990 (hereinafter referred to as "Ekstrόm") .
Converter assemblies of the kind mentioned in the introduction are previously known from the Swedish published patent appli¬ cation with publication number 464 843. This publication describes an assembly with a first, current-source, line- commutated converter and a second, voltage-source, self- commutated converter. The converters are d.c. series-connected to each other and have a.c. terminals for connection to an alternating-voltage network. The assembly - and hence both converters - is primarily operating in inverter operation. In
the usual manner in connection with HVDC transmissions, the direct current of the assembly is determined by the other converter stations included in the d.c. network, whereas the direct voltage is determined by the converter assembly which operates in inverter operation. The direct voltage of the converter assembly is the sum of the direct voltages of the two converters. The first, current-source, line-commutated converter is controlled in a conventional manner and operates in inverter operation at minimum extinction angle. The second converter is provided with control members for controlling the direct voltage of the converter by influencing the phase position of the internal alternating voltage of the converter (in relation to the voltage of the alternating-voltage net¬ work) , and hence the flow of active power between the conver- ter and the alternating-voltage network. A control system is arranged for controlling the total direct voltage of the con¬ verter assembly by comparing the sensed value of this voltage with a reference value of the voltage and allowing the difference to influence the phase position of the internal alternating voltage of the second converter via a regulator.
An assembly of this kind offers several advantages, such as a possibility, with a small risk of commutating errors, of cooperating with weak alternating-voltage networks without the need of synchronous machines or extra shunt capacitors, a possibility of reduction of the reactive power consumed by the converter assembly, and of generation of reactive power, - a possibility of fast control of the flow of reactive power, for example for voltage control in the alterna¬ ting-voltage network, and a possibility, for example in case of a ground fault on the d.c. line, of reducing the direct voltage of the transmission to zero, hence limiting overcurrents, and this while maintaining the direct voltage of the voltage- source converter and hence without disturbances in the flow of reactive power.
However, it has proved that difficulties may arise in obtaining good control properties in an assembly of the kind referred to here. The two converters and their control systems affect each other, and this makes it difficult to obtain a fast voltage control in the alternating-voltage network.
Further, the voltage-current characteristic of the assembly deviates from the characteristic of a conventional converter station, which renders difficult the cooperation of the assembly with other curren -source converters.
SUMMARY OF THE INVENTION
The invention aims to provide a converter assembly of the kind described in the introductory part of the description, which possesses improved control properties, and the current-voltage characteristic of which makes possible a problem-free coopera¬ tion with other current-source converters.
What characterizes a converter assembly according to the invention will become clear from the appended claims. Accor¬ ding to the invention, the direct voltage of the voltage- source converter is controlled independently of the current- source converter. The current-source converter will hence operate in a conventional way and determine the current- voltage characteristic of the assembly. The characteristic is therefore of the same kind as in conventional current-source converters, which makes possible a problem-free cooperation with such converters. Further, with an assemblyt according to the invention, considerably improved control properties are obtained compared with prior art assemblies of the present kind.
According to one embodiment of the invention, a feed-forward control of the direct voltage of the voltage-source converter is arranged by setting the phase position of the internal alternating voltage of this converter in accordance with a preliminary value calculated from sensed operating quantities. The closed-loop control circuit then only need to correct any
minor deviations between the calculated preliminary value and the value which is needed to obtain the desired direct vol¬ tage. In this way, a very fast control may be obtained.
According to another embodiment of the invention, the refe¬ rence value of the direct voltage of the voltage-source con¬ verter is made current-dependent in such a way that the vol¬ tage-current characteristic of this converter is given a positive slope. By a suitable selection of this current depen- dence, the total voltage-current characteristic of the assembly may be made flat or with a certain positive slope, whereby the reason for instability, constituted by the nega¬ tive slope of the characteristic of a current-source conver¬ ter, may be eliminated.
According to a preferred embodiment of the invention, the voltage-source converter is designed such that the ratio between the internal alternating voltage and the direct vol¬ tage of the converter is controllable, which makes possible a fast control of the reactive power flow and hence an efficient feedback control of an operating quantity, preferably the voltage in the alternating-voltage network. Suitably, this converter is connected to the alternating-voltage network via a transformer with a tap changer, the control of the above- mentioned ratio then being arranged to cooperate, in an advantageous manner, with the control of the tap changer and possibly also with the control of the reference value of the direct-voltage of the converter.
BRIEF DESCRIPTION OF THE DRAWINGS
The invention will be described in more detail in the following with reference to the accompanying Figures 1-4, wherein
Figure 1 shows a converter assembly according to the inven¬ tion,
Figure 2 shows the design of the control system for the voltage-source converter in the assembly according to Figure 1,
Figure 3 shows the total voltage-current characteristic of the assembly and the characteristics of the two converters of the assembly,
Figure 4a and Figure 4b show how, according to one embodiment of the invention, the voltage-source converter can be con¬ trolled so as to further reduce the risk of commutating errors and/or so as to allow operation with lower extinction angles.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
Figure 1 shows a converter assembly according to the inven¬ tion. The assembly constitutes one pole in a station in a system for power transmission by means of high-voltage direct current (HVDC) . The assembly is primarily intended to operate in inverter operation. Temporarily or permanently, it may operate in rectifier operation, which, however, requires that one of the converters is pole-changed since the polarity of the current is given in SRA and the polarity of the voltage is given in SRB. The assembly comprises two converters SRA and SRB which, in series with each other and with a smoothing reactor DCR, are connected between a d.c. line DCL and ground. The direct current of the transmission is designated Id and the total direct voltage of the station is designated Ud-
The converter SRA is a current-source line-commutated thyristor converter. Both main circuits and control circuits for this converter are designed in accordance with known and generally applied principles for such converters which are described in Uhlmann and Ekstrόm. It operates in inverter operation in a conventional way, that is, it operates at the smallest possible value of the extinction angle γ in depen¬ dence on the present values of, inter alia, direct current and commutating voltage, that is, it always delivers the highest
possible direct voltage Uda- ar-d the converter and hence the station in which it is included become determining for the direct voltage of the transmission. In the usual manner, the converter has a converter transformer TRA provided with an on- load tap changer, and the converter is connected to a three- phase alternating-voltage network ACN via this converter transformer. The need of and the requirement for stepping and control range of the tap changer are, however, lower in an assembly according to the invention than in a conventional converter assembly since - as will be described below - the voltage-source converter SRB, series-connected to the con¬ verter SRA, has a current-voltage characteristic with the opposite slope in relation to the characteristic of the con¬ verter SRA.
The converter SRB is a voltage-source forced-commutated con¬ verter of the type described in Ekstrόm, the section entitled "Forced-commutated Voltage Convertor" on pages 11-17 - 11-32. In principle, it consists of a three-phase bridge with six branches, each branch having a controllable thyristor valve (capable of being both turned on and turned off) in anti- parallel with a diode valve. In parallel with the converter is a capacitor bank CB, which constitutes the low-impedance (stiff) direct-voltage source which is essential to the operation of the converter. The voltage across the capacitor bank, that is, the direct voltage of the converter, is designated Udb- Tne internal alternating voltage of the converter, that is, the fundamental tone of the alternating voltage directly generated by the converter bridge, is designated Uvb- τ-ne converter is connected to the 'network ACN via a converter transformer TRB which may be provided with an on-load tap changer. The voltage of the alternating-voltage network ACN is designated Uτ_,.
For controlling the converter SRB, a voltage transformer UMA is adapted to sense the line voltage UL and to deliver a signal ULm proportional to the voltage, a current-measuring device (e.g. a measuring transduc or) IM adapted to deliver a
signal Idm corresponding to the direct current Id/ and a measuring voltage divider UMB adapted to sense the direct voltage Udb °f the converter and to deliver a signal Udbm corresponding thereto.
Regarding the control of the converters, it is assumed in the following description that the converter SRA is operating at γ-min, which gives this converter the following current- voltage characteristic Ud = Udiocosγ - R • Id where
UdiO is the ideal open-circuit direct voltage γ is the extinction angle
Rx is a constant proportional to the leakage inductance of the transformer, that is, the characteristic has, in the known manner, a negative slope.
Figure 2 shows the control circuits for the converter SRB. The converter has a control pulse device SPD which delivers con¬ trol pulses SPi to the controllable valves of the converter for turning these on and off. The function of the control pulse device is controlled by the control signals δ and ku supplied to the control pulse device. The control signal δ controls the phase position of the voltage Uvb of the conver¬ ter in relation to the phase position of the voltage UL of the alternating-voltage network ACN such that the phase difference between these two voltages assumes the value δ. The measured signal U m is supplied to the control pulse device as a phase position reference. The control signal ku controls the ratio between the voltages Uvb (the internal alternating voltage of the converter) and Udb (the direct voltage of the converter) such that the ratio assumes the value ku, that is, UVb = ku • U b This control may be made in any of the ways described in the section from Ekstrόm cited above, for example by pulse-width modulation or by designing and controlling the converter as a so-called NPC converter (three-level converter) .
The direct current I is in the assumed operating case deter¬ mined externally. The direct voltage Udb °f the converter SRB is constant during steady state conditions, and the whole direct current I flows through the converter. The active power which is supplied to the converter from the d.c. network is
Pb = Udb • Id The active power flowing from the converter to the alterna¬ ting-voltage network (if the converter losses are neglected) is equally great and is given by the equation Pb = (Uvb • UL • sin δ ) / Xb where Xb is the impedance between the converter bridge and the network, that is, XD is practically equal to the reactance of the converter transformer. Thus, in steady state the following applies Udb • Id = (Uvb ' UL • sin δ ) / Xb that is, sin δ = (Udb • Id • Xb) / (Uvb • U ) The phase difference δ may thus, in steady state, be calcu- lated from the four operating quantities on the righthand side of the equal-sign in the latter equation. This calculation is performed by the circuit PAC in Figure 2. The circuit delivers a preliminary value δ' of the phase difference to a summator
S4. The calculation may be made more or less exact. In the example described, a simplified calculation is made while making use of the assumption that
UL = vb which applies approximately, and while utilizing Uvb = ku * b The circuit thus delivers the output signal δ' = arcsin ( (Id • Xb) / (Ku 2 • Udb) )
This signal constitutes an approximately correct value of the phase difference δ.
In a voltage control circuit UC, a basic reference value
UdbrefO of the direct voltage of the converter is formed in the manner described below. To this reference there is added in a summator S2 a current-dependent quantity Rb'Id' where Rfc>
is a constant . Rβ is chosen such that the positive slope of the characteristic of the converter SRB compensates for the negative slope of the characteristic of the converter SRA to such a great extent that the total characteristic of the assembly becomes flat or is given a positive slope. The output signal from the summator S2 becomes
Udbref = UdbrefO + b * Id It is compared in a summator S3 with the measured signal Udbm and the difference is supplied to a d.c. voltage regulator Ud~ reg with PI characteristic. The output signal Δδof the regu¬ lator is added in the summator S4 to the preliminary value δ', and the control pulse device SPD thus controls, the converter such that the phase position of the converter voltage in relation to the network becomes δ = 8' + Δδ
Thus, also in case of fast changes in the operating condi¬ tions, the calculating circuit PAC instantaneously provides an approximately correct phase position of the converter voltage. This entails exceedingly good control properties, and, for example, a stepwise change of the direct current may be made with a minimum of transient variations of the direct voltage of the converter SRB. The regulator Ud-reg need only correct the minor deviations which may remain as a result of inaccu¬ racies in calculation, measurement errors, transients, etc.
The quantity ku for control of the amplitude of the alterna¬ ting voltage of the converter SRB is obtained from the voltage control circuit UC, which in turn receives the output signal ku" from an alternating-voltage regulator U - reg. The regu- lator, which has PI characteristic, is supplied with the con¬ trol deviation, formed in the summator SI, between the line voltage UL and the reference value ULref °f this voltage. The output signal ku" of the regulator constitutes a basic value for the modulation factor ku. The voltage control circuit UC sets ku = ku" if ku" lies within a preferred operating range, which is determined by two predetermined limit values kumin and kumax. If
-■ j = -umax
the voltage control circuit delivers a control order NLb to the on-load tap changer of the transformer TRB to increase the ratio of the transformer, b Nvb- In a corresponding way, a control order is delivered to reduce the transformer ratio if ku" .≤. kurn n-
The voltage control circuit UC also generates the basic reference value UdbrefO for the direct voltage Udb- This is primarily a constant value, which is so chosen that Udb normally is lower than, for example 40% of, the maximum direct voltage of the converter SRA in inverter operation. This makes possible a fast reduction of the total direct voltage Ud to zero in case of, for example, a ground fault on the d.c. line while maintaining the direct voltage of the converter SRB and hence the desired reactive power flux. To make it possible to maintain the modulation factor ku within the preferred operating range even in case of major changes in the operating conditions, the circuit UC is adapted to adjust the reference UdbrefO i the on-load tap changer reaches one of its limit positions. If the on-load tap changer has stepped up to its upper limit position, UdbrefO increases, and if it has stepped down to its lower limit position, UdbrefO decreases. This change of UdbrefO may either be made as a slow continuous control or also be made step-by-step.
Figure 3 shows the total current-voltage characteristic ABCD of the assembly in stationary inverter operation. The converter SRB has the characteristic EFG, which, in the manner described above, has been given such a positive slope that the part CD of the total characteristic is flat or has a small positive or negative slope. The resultant characteristic of the other stations of the transmission consists of the curve HKFL. The working point of the converter SRB is thus, under normal conditions, the point F in the figure, and the working point of the station as a whole is the point M.
Normally, both the voltage difference Uvb ~ U and the phase difference δ are small.
The flow of reactive power from the converter SRB to the network is then (at the transformer ratio 1:1) approximately proportional to the voltage difference. Normally, the conver¬ ter operates such that Uvb > UL, which means that the conver¬ ter generates reactive power, which wholly or partially com¬ pensates (or possibly overcompenεates) the reactive power consumed by the converter SRA. The flow of reactive power is controlled by influencing the modulation factor ku of the converter, which directly influences the internal direct voltage Uvb of the converter. In the example shown in Figure 2, the reactive power flux is controlled such that the voltage UL in the network ACN is maintained constant.
In a converter assembly according to the invention, the voltage-source self-commutated converter SRB thus provides a counter direct voltage which, in principle, is constant. This entails considerable advantages. For one thing, the assembly operates, viewed from the d.c. link, as a conventional con¬ verter station provided with only current-source line-commuta- ted converters, which means a simplified cooperation with conventional converters and that no reverse direction current will be supplied from the voltage-source converter in case of a ground fault on the d.c. line. Further, by the principle of control according to the invention, a separation of the control of the two converters of the assembly is obtained, which entails considerably improved control properties, such as speed of operation and stability. The good control proper¬ ties may be further improved by providing the assembly, accor¬ ding to a preferred embodiment of the invention, with the above-described feed-forward control of the phase position of the voltage-source converter.
To reduce the risk of commutating errors in the line-commuta- ted converter, different transient interventions may be made both in the line-commutated converter and in the self-
commutated converter. At a measured transient in the line voltage, for example both γref in the line-commutated conver¬ ter and ku in the self-commutated converter may be given a transient addition. Another alternative is more directly to sense the probability of a commutating error and, in case of need, to give a maximum addition to the commutating voltage in the manner described in the following.
Figures 4a and 4b show the principle of such a control method by means of which, thus, the risk of commutating errors in the line-commutated converter SRA may be further reduced. This converter is provided with a sensing circuit which continuous¬ ly senses the commutating voltage ujζ(t) in question, the rate of change duk(t)/dt of this voltage, and the current ik(t) in the decommutating valve. Figure 4a shows the commutating voltage during the commutating interval. The commutation is assumed to be started at t = tl. The time Δt remaining until the commutation is completed at t = t2 (the current has dropped to zero) may be calculated with knowledge of the commutating inductance L per phase and based on the sensed quantities. The calculation is made continuously, and Δt is obtained by setting the voltage time area of the commutating voltage (dashed in the figure) up to the time of completion of the commutation equal to the product of commutating inductance and current. This gives the relationship
2 • Δt • uk(t) - Δt2 • duk(t)/dt = 4 • Lk • ik(t)
While assuming that the commutating voltage decreases at a constant rate, the time remaining till the zero crossing of the commutating voltage will be uk(t)/(duk(t)/dt) and the predicted extinction angle Ypred is calculated continuously from the relationship
Ypred/ω = u (t) / (duk(t) /dt) - Δt where ω is the angular frequency of the network. The value of γpred thus obtained is continuously compared with a predetermined minimum value Ycrit- If Ypred < Ycrit* this is interpreted such that a risk of a commutating error is present. In this case, an intervention is made in the control
LO LO t-o to H-- P> o LΠ σ o LΠ
α
O Ω
O 3 < Φ
Hi ft Φ Hi