GB2187576A - Optimisation of reactive power regulator operation - Google Patents

Abstract

In a method of operating a reactive power compensation regulator having n discrete capacitor stages, the optimum setting of the insensitivity area (C/K value, where C is capacitance and K is the transmission ratio of an associated transformer) is determined in dependence on state of the regulator by the automatic switching-on and -off of the capacitor stages. Repetition of a switching pattern after 2<n> directly following state-charger is used as a criterion for adjusting the C/K value and the procedure is controlled until the C/K value reaches a limiting value which is deposited in a first register as the present optimum value. The method may commence with a maximum C/K value which is reduced until the effect of the first capacitor stage is greater than the C/K value. An average value of a plurality of the determines optimum C/K values may be stored in a second register. A third register may store the number of averagings.

Description

SPECIFICATION Reactive power regulation The invention relates to reactive power regulation.

Reactive power regulators are used to control the switching-on or -off of capacitors for providing compensation of reactive power in an electrical system, thereby to improve the power factor of the system. The capacitors can be switched on or off in stages or steps, and a characteristic of such step-wise regulation is the fact that the reactive power regulator has an insensitivity area or zone the width of which is determined by the so-called C/K setting, where C is the capacitor power or capacitance, and K is the transmission ratio of an associated transformer or transducer. The C/K value represents the reactive power of the smallest capacitor stage, based on the input of the reactive power regulator.

With correct setting of the C/K value the regulator operates correctly as long as the switching-on or -off of the capacitors lies within the insensitivity area up to the resulting reactive power. If capacitors of different compensation powers are controlled by the reactive power regulator, then the transient recovery time of the regulating circuit is at its shortest when the compensation power of the capacitors is increasing. A smaller stepping of the compensation power compared with the previous compensation stage can lead to oscillation of the control circuit in certain operating states, so that usually the capacitor power of the respective next stage is always selected to be least the same size or greater than the previous compensation stage. In Fig.

1 there is shown an optimum C/K setting for a reactive power regulator with manual adjustment. This is the case, for example, when the insensitivity area predetermined by the C/K value is 10% to 20% greater than the effect of the first capacitor stage.

In practical use there are frequently problems with the C/K value setting. If the C/K setting is too low, then the control circuit oscillates which leads to combustion of the capacitor earthing contacts within a very short time. A C/K value that is set to be too high is indicated by the fact that the control procedure is only used with fairly great over- or under-compensation. The reactive power regulator works satisfactorily when the C/K value lies within a range 1.05 to 1.95 times the step power of the first capacitor stage.

According to one aspect of the present invention there is provided a method of operating a reactive power regulator by switching n discrete capacitor compensation stages for the optimum adjustment of an insensitivity zone dependent on the ratio C/K, where C is capacitance and K is the transmission ratio of an associated transformer, wherein dependent on the operating state of the regulator a switching state pattern of output switch elements for the capacitors is produced by automatic switching-on and -off of the n different compensation stages, the repetition period of which pattern is used as a criterion for adapting the C/K value, and this procedure is automatically controlled until the instantaneous insensitivity zone reaches a limiting value which is deposited as an instantaneous optimum value in a first register.

According to another aspect of the present invention there is provided a reactive power regulator adapted and arranged to operate in accordance with a method according to the immediately preceding paragraph.

An example of the invention may remove or reduce the aforementioned defects of a reactive power regulator and may provide an optimum working point adjustment of the regulator.

For a better understanding of the invention and to show how it may be put into effect reference will now be made by way of example to the accompanying drawing in which: Figure 1 is a diagram of an optimum C/K setting; Figure 2 is a diagram of an adaptatioh procedure in a reactive power regulator; Figure 3 is a diagram of a reactive power demand before switching-on the first stage; and Figure 4 is a table of the switching states for three output relays with stepping ratios 1:2:2.

As already described, the reactive power regulator oscillates when the C/K value is set too small. This fact is used in an algorithm for the adaptation of the C/K value in the control circuit of the reactive power regulator. As a starting point of the adaptation, the operating state used is that in which the first compensation stage is switched-off and compensation power is demanded. If, after the regulator has switched on the first compensation stage, it has to switch it off again immediately after, because the resulting reactive power lies outside the insensitivity area in the capacitive area (Fig. 2), then the insensitivity area is too small and has to be enlarged. This means that the C/K value has to be increased.Dependent on the reactive power demand of the system this procedure will be repeated often enough until, even in the limiting case, after the first stage has been switched-on this is not immediately subsequent switched-off again. The optimum C/K value is determined when the insensitivity area is exactly the same as, or slightly greater than, the effect of the first capacitor stage.

As a rule the regulator will not determine the optimum C/K value in one operation because for this purpose a specific compensation power demand of the system is neces sary (A in Fig. 1); however, one can proceed from the fact that once a day a compensation power is demanded which lies near this point A. In each case the regulator works with a C/K value which is always smaller than, or at the most the same as, the optimum value and therefore with good control properties.

If the demanded compensation power is very much greater than the half-stepped power, then this has no influence on the adaptation algorithm. Fig. 3 shows the reactive power demand of the system before switching-on the smallest stage. The capacitor powers of the switched-on stages in this example are in the ratio 1:1:2:2.

As already mentioned at the beginning, the fact that the capacitors are stepped other than increasing with regard to their compensation power should not be excluded. The described special case-oscillation of the first stages then no longer sufficient for assessing the stability of the control circuit.

The criterion for the oscillation of the control circuit is a cyclicai recovery of identical switching states of the output relays Rn (n= 1.2. . number of relays) after 2n directly following switching cycles. Recognition of an identical switching stage consequently means the increase in the C/K value, i.e. that an increase is required. Fig. 4 shows in this connection an example with three output relays R1, R2, R3 switching off and on three capacitor stages in the ratio of 1:2:2.

If there occurs an operating state whereby the resulting reactive power falls into the insensitivity area, then all the switching states stored up to that point are invalid. A new cycle begins with the next change in state of the output relays R1 .. . R3.

The optimum C/K value is determined when, in the most varied operating states of the system after 2n switchings, equality (repetition) in the switching states no longer occurs.

To determine the optimum C/K value the adaptation algorithm generally requires several switchings. An improvement can be achieved by the C/K value not being placed at the lowest but at the highest value during initialisation of the regulator. After the resulting reactive power lies in the insensitivity area either no reactive power is demanded or the regulator switches-on the necessary capacitors in advance-the C/K value is reduced to a value which corresponds to the resulting reactive power. This can be repeated at each control procedure until the point at which the effect of the smallest capacitor stage is greater than the insensitivity area. The initialising phase is thus concluded. From now on the optimum C/K value is determined as described above.

In operation of the control system, at the same time as the switching-on of the smallest capacitor stage a reactive power load may be switched off. If the C/K value was already previously set at optimum, then a greater value than necessary is now selected. To correct such an error the C/K value is therefore reduced daily by one or two stages.

An effective improvement in the behaviour in such cases of disturbance offers the determining of the C/K value by means of averaging. Thus, apart from a C/K register the regulator has two further registers. In the second register there is stored the mean value from the preceding C/K values determined in a period of respectively 24 hours, and in the third register the number N of the preceding averagings e.g. of a day or of all the averagings since the start-up of the reactive power regulator. At the beginning of a 24 hour cycle the sliding C/K value determined up to this point in time is read out, and if necessary is written into the C/K register more sensitive by one stage. The regulator corrects the C/K value if necessary in the course of a day.At the end of the cycle the new mean value is formed from the C/K value and the old mean value weighted with the number N, and is written into the mean value register.

In systems with high switching frequency the loading of the earthing contacts is great.

To prevent premature wear of the contactors, a switching-on delay time of e.g. 30 seconds in the case of more than 30 operating cycles within e.g. 30 minutes is increased to a delay time of e.g. 60 seconds. After a predetermined time, e.g. 3 hours, the delay time is then again reset to the original time, in the above example to 30 seconds.

In general terms there has been described a method of operating a reactive power regulator having discrete capacitor stages for the optimum setting of the insensitivity area (C/K value). It is proposed that the optimum insensitivity area (C/K value) is determined in dependence on the operating state of the control system or regulator by the automatic switch ing-on and -off of the compensation stages when the regulator is started up in the form of an oscillation procedure. This optimum value is deposited in a first register and is used as a starting value for the further automatic determining of the operating point of the regulator.

In other terms, there is provided a process for a reactive power regulator for use in control systems with discrete capacitor stages for the optimum adjustment of the insensitivity zone (C/K value), wherein a) dependent on the operating state of the control system a switching state-pattern of the output switch elements is produced by automatic switching-on and -off of the differ ent compensation stages n, the repetition of which after 2n directly following state-changes is used as a criterion for adapting the C/K value and b) this procedure is automatically controlled until the instantaneous insensitivity area (C/K value) reaches a limiting value which is deposi ted as an instantaneous optimum value in a first register.

It will be understood that the man skilled in the art will, with this information, be quite capable of adapting and arranging a reactive power regulator to operate in accordance with the method or methods described hereinbefore. The invention extends not only to a method of operating a reactive power regulator, but also to the reactive power regulator itself.

Claims (8)

1. A method of operating a reactive power regulator by switching n discrete capacitor compensation stages for the optimum adjustment of an insensitivity zone dependent on the ratio C/K, where C is capacitance and K is the transmission ratio of an associated transformer, wherein dependent on the operating state of the regulator a switching state pattern of output switch elements for the capacitors is produced by automatic switching-on and -off of the n different compensation stages, the repetition period of which pattern is used as a criterion for adapting the C/K value, and this procedure is automatically controlled until the instantaneous insensitivity zone reaches a limiting value which is deposited as an instantaneous optimum value in a first register.

2. A method according to claim 1, wherein it is the repetition of said pattern after 2" directly following state-changes which is used as a criterion for adapting the C/K value.

3. A method according to claim 1 or 2, wherein the method begins with a maximum insensitivity zone and this is reduced in dependence on the resulting reactive power until the effect of the first compensation stage is greater than the insensitivity zone.

4. A method according to claim 1,2 or 3, wherein a plurality of instantaneous optimum values of the insensitivity zone are determined at different periods of time, and an average value formed therefrom is deposited in a second register.

5. A method according to claim 4, wherein the number of the averagings is stored in a third register.

6. A method according to claim 4 or 5, wherein at the beginning of each new compensation time period the average value deposited in the second register is transferred to the first register and, if it is possible in view of the capacitor stages, is increased by one sensitivity stage.

7. A method of operating a reactive power regulator substantially as hereinbefore described with reference to Figs. 1 to 4 of the accompanying drawing.

8. A reactive power regulator adapted and arranged to operate in accordance with a method according to any one of claims 1 to 7.