STABILITY ISSUES OF THE OUTPUT PARAMETERS OF A SELF-EXCITED ESG DURING PROLONGED CONTINUOUS OPERATION

Yu. G. Leleko

(Presented by the scientific seminar of the Department of Theoretical Foundations of Electrical Engineering and by the ESG Department of the Scientific Research Institute of Nuclear Physics)

A small-size self-excited ESG, whose design and main characteristics are described in [1, 2], is intended for continuous operation for several hours with an active or capacitive load, the magnitude of which changes over a wide range during this time. However, the results described in [2] were obtained under relatively short-term operating conditions of the generator and give an idea only of the generator’s limiting capabilities.

In the mode of prolonged continuous operation, the principal characteristic determining the possibility of practical use of a self-excited generator is the stability of the output parameters, i.e., the degree of change in the load voltage and current as a function of the generator operating time.

The purpose of the present work is to determine the degree of stability of the output parameters in the mode of prolonged continuous operation and the causes that bring about their change, which will make it possible to refine recommendations for the design of self-excited ESGs and to determine their optimal operating mode.

To solve the questions posed, tests were carried out with several experimental specimens, in the design of which, as described in [1], certain minor changes were introduced. In particular, in order to reduce the magnitude of the induced charge on the transporters of the auxiliary rotor, their length was reduced and the depth of embedding in the dielectric was increased. At the same time, the width of the collector plates was increased to 2 mm, which is necessary for more reliable operation of the brushes.

In the generators tested, in order to create a uniform distribution of potential over the stator, instead of a glass cylinder a cylinder made of an epoxy compound was used, whose volume conductivity is comparable with the volume conductivity of glass. At the same time, the possibility of mechanically machining the stator made it possible to establish a uniform rotor–stator gap within 0.2–0.25 mm. During the tests, all generator specimens were provided with an additional point–brush in the auxiliary-rotor system, by means of which the output voltage was stabilized within either 30 kV or 40 kV.

Hydrogen and nitrogen under a pressure of 6 atm, monitored by a manometer, were used as the working medium in the experimental specimens. The motor was supplied both from a VSA-5 direct-current rectifier and from storage batteries.

The generators were tested under conditions as close as possible to the operating mode used in industrial service. The values of the output parameters were recorded both during continuous operation for 10–12 hours followed by shutdown of the generator and, without changing the operating conditions, by switching it on again after 14–12 hours, and during continuous operation without shutdowns for several tens of hours. The values of the output parameters were monitored using an S-100 kilovoltmeter and an M-194 microammeter, as well as with an N-373-3 recorder. Since active resistances of 16.6 GΩ and 41.4 GΩ were used as the generator load, it was sufficient to monitor one of the output quantities—voltage or current. Figure 1 presents the results of the first tests of experimental specimens with

Fig. 1

different gases as the working medium. Two generators operated continuously for 12 hours, after which they were switched off and carefully inspected (curves 2, 3). One of the generators was adjusted to an output voltage of 40 kV, the second to 34 kV, and both operated into a load resistance of 16.6 GΩ. The generator adjusted to 40 kV operated in hydrogen at a pressure of 6 atm and had average motor parameters \(U_{\mathrm{dv}} = 16\ \mathrm{V}\), \(I_{\mathrm{dv}} = 75\ \mathrm{mA}\). The second generator operated in nitrogen at a pressure of 6 atm, with \(U_{\mathrm{dv}} = 16\ \mathrm{V}\), \(I_{\mathrm{dv}} = 75\ \mathrm{mA}\). Two other specimens were not disassembled after shutdown and, after 12 hours, were switched on again. In this case the load voltage was restored to the values at which shutdown had occurred during the first 1–2 min. The generators operated in a hydrogen atmosphere (curve 1) and nitrogen (curve 4) at a pressure of 6 atm into a load resistance \(R_{\mathrm{n}} = 41.4\ \mathrm{G}\Omega\). The motor supply voltage was 16 V; the current consumed by the motor under load averaged 65 mA for the first generator and 55 mA for the second.

As these dependences show, the mean value of the generators’ output voltage increases during the first hours of operation by an average of 10% relative to the initial value, and only after 10–12 hours of operation does its growth cease. External inspection of the generator parts after 12 and 28–30 hours of operation showed that in the system of the auxiliary rotor, especially in the region where the auxiliary point-brush is located, there is a layer of dust on the rotor and on the collector plates, apparently formed by friction of the brushes against the rotor compound. A layer of dust was also observed on the elements of the main rotor, but in a significantly smaller amount. As is known from the literature [3], the appearance of dust in the generator volume reduces its output parameters, but this was not observed in the tests performed. To verify the results obtained, the generators were tested in longer-duration operating modes. Figures 2 and 3 present, by way of example, the results of tests of two specimens

generators with parameters \(U_{\mathrm{n}} = 30\ \text{kV}\), \(U_{\mathrm{dv}} = 16\ \text{V}\), \(I_{\mathrm{dv}} = 65\ \text{mA}\) and \(U_{\mathrm{n}} = 40\ \text{kV}\), \(U_{\mathrm{dv}} = 16\ \text{V}\), \(I_{\mathrm{dv}} = 75\ \text{mA}\), operating in a hydrogen atmosphere at a pressure of \(6\ \text{ati}\) into an active load \(R_{\mathrm{n}} = 41.4\ \text{G}\Omega\). In carrying out this series of experiments, some of the results were recorded using a self-recording microammeter, which made it possible to refine the oscillations of voltage and current at each instant of time.

Fig. 2

Fig. 3

As the dependences presented show, the oscillations of the output voltage relative to the mean value, recorded from the kilovoltmeter readings, are approximately \(\pm 3\text{–}4\%\). However, in all cases the mean value of the output voltage decreased during the tests by \(30\text{–}40\%\) of the initial value. At the same time, the rate of decrease of the mean voltage increases with the number of hours of generator operation. The maximum voltage value occurs after \(8\text{–}12\) hours of operation and subsequently falls to values amounting to only \(60\text{–}70\%\) of the initial value. As the tests showed, the stability of the output parameters at each instant of time is maintained within \(\pm 2\%\).

Thus, the tests carried out showed that the output parameters of the generator, while remaining practically constant at each instant of time, decrease in their mean values with increasing number of hours of operation as compared with the initial values. Only in the first hours of operation do the generator output parameters increase in magnitude.

External factors—which may include the magnitude of the resistance into which the generator operates, the pressure inside the generator casing and the chemical composition of the gas used as the working medium, and the motor supply voltage, which affects the rotational speed of the rotor—when there is an additional needle–brush and constant pressure in the generator casing, have practically no effect on changes in the generator output parameters. In this connection it may be assumed that the stability of the output voltage is significantly affected by changes in the gas discharge in the commutating units and at the additional needle–brush of the generator. This change may be due to a number of causes, the principal ones of which, apparently, are the following:

1. A change in the geometry of the electrodes under the action of discharges during operation of the generator. In this case, strengthening of the gaps is possible in the process of “conditioning” the electrodes during the initial period of their operation, with subsequent weakening due to erosion under the action of discharges during prolonged operation.

2. Dusting of the generator volume by dust from the brush-collector system.

1. The appearance of products of decomposition of the rotor material under the action of a gas discharge in the brush–commutator system.

Observations of the operation of experimental specimens and an external inspection of the main elements of the generator after various numbers of hours of operation confirm the above assumptions concerning the influence of the instability of the gas discharge on the magnitude of the average output voltage. Thus, for example, an external inspection of a generator that had operated for 180 hours in a hydrogen atmosphere showed that on the additional rotor, on the outer side of the commutator plates, there was a layer of dark-gray dust 1.5–2 mm wide. On the remaining part of the additional rotor, slight dusting was observed. On the flange of the additional rotor, dust of the same color was concentrated mainly around the additional brush tip. On the side of the main rotor, both on the commutator and on the adjoining parts of the stators, there was a slight coating of light-gray dust. In this test the generator operated properly for 140 hours with an external active load resistance equal to 16.6 GΩ. After this, the load resistance was increased to 41.4 GΩ, and the generator operated for another 40 hours continuously, but no substantial increase in voltage, as compared with the voltage at the moment of switching, occurred. After disassembly of the generator and external inspection, all elements of the generator were wiped with alcohol, the generator was assembled and restarted. The load voltage rose to its initial values, i.e., to 40 kV, and subsequently repeated the previously obtained characteristic.

Thus, the investigations carried out make it possible to conclude that the decrease in the average value of the voltage during prolonged continuous operation of the generator occurs because the working volume of the generator becomes dusted from the brush–commutator system; moreover, this occurs mainly on account of the rotor material, since after prolonged operation the generator brushes were practically not deformed or destroyed. The dust that forms decomposes as a result of ionization processes at the generator brushes and worsens the insulating properties of the generator’s working medium. It should be noted that the ionization processes occur most intensively in the system of the additional rotor, which operates under the most severe conditions both in current and in voltage.

In connection with the foregoing, in order to eliminate the indicated shortcomings and, consequently, to increase the number of hours of generator operation with nominal output parameters, the following changes in the generator design may be recommended:

1. Increase the additional rotor–stator gap and reduce the length of the additional-rotor conveyors, which will lead to a reduction in the parameters generated by the additional rotor.

2. Modify the design of the commutator of the additional rotor so that, during friction, the brushes contact the commutator plates over a longer path, and commutation of the conveyors with the brushes takes place within metallic surfaces.

3. At least around the commutator plates, the rotor material should be heat-resistant and should not wear away under brush friction.

4. Improve the ventilation system of the additional rotor.

LITERATURE

1. V. V. Patsevich, Yu. G. Leleko. A compact self-excited ESG. Izv. TPI, No. 210. In press.

2. Yu. G. Leleko. Problems in the study of a compact self-excited electrostatic generator. Izv. TPI, No. 232. In press.

3. I. Zh. Felitsi. “Electricity,” No. 1, 61, 1962.