A study on the relationship between the source frequency and the radio interference

of the transmission line

Shuai Zhang1 Jinliang He2

A State Key Lab of Power Systems, and Department of Electrical Engineering, Tsinghua University, Beijing 100084, P.

R. China and email: zhangshuai94@gmail.com, hejl@tsinghua.edu.cn

Abstract With the increasing use of EHV and UHV transmission lines, the electromagnetic pollution caused by corona discharge is becoming a serious problem. One of the most important electromagnetic effects of corona activity of the transmission line is the radio interference. As the negative and positive corona are very different, the RI generated by them is closely related to the source. To find out the relationship between the source frequency and the RI, it is measured as the frequency of source changes. And through the analysis of the results, some regular patterns are found out. In this paper the RI is not studied at a particular frequency, the spectrum is from 1 kHz to 5 MHz. This work provides useful knowledge of the RI on transmission lines and find out some connections between the RI and the frequency of applied voltage.

1. Introduction Generally, the quasi-peak values of RI in the experiment are measured at a specific frequency. The parameters of the measure instrument are either the bandwidth is 9 kHz, measuring frequency of 0.5 MHz (CISPR) or the bandwidth is 5 kHz, measuring frequency of 1 MHz (ANSI).The empirical and semi-empirical results are only reflect the RI at a particular frequency, such as 0.5 MHz and 1 MHz, while the RI is existent in a wide frequency range. This paper carried out a series of experiments to study the RI in a wide spectrum. In the experiments, the frequency of the voltage applied to the wire was systemically changed. Based on the experimental data, some characteristics of the RI spectrum were found out. For the first time, the influence of the source frequency on the RI was studied and the result would be useful for the research on the RI of the DC, AC or hybrid DC/AC transmission lines.

2. Test method and Facilities The experimental facility of this paper consists of a reduced scale transmission line, a shielding cage, a sampling resistor, a data acquisition card (DAQ), a photoelectric conversion device, a signal generator, a power amplifier and the PC. The signal generator creates the voltage wave of a certain frequency and the power amplifier turns it into the high voltage and then applies it to the line. The sampling resistor is to convert the corona current to voltage signal which is gathered by the DAQ. The photoelectric conversion device is used to transport the signals. Both the sampling resistor, DAQ and photoelectric conversion are put in the shielding cage. The shielding cage here is to protect the devices from the high voltage.

Fig.1 The experimental settings

978-1-4673-5225-3/14/$31.00 ©2014 IEEE

3. III. Experimental results and Analysis

After Fourier transform, the measuring results of the corona current turn into the frequency spectrum that can represent the characteristics of the RI for the test line. The typical spectrum of the RI is shown in Figure 2, and the voltage applied to the line is 50 Hz AC voltage. The spectrum can be divided into three parts according to its features. The first part is low frequency band, ranging from 1 kHz to about 100 kHz. In this band, there are several peaks and the results have a certain randomness. Unusually, the maximum of the RI is in this part. In the previous studies, there were less attention paid to this band of RI. The next part is about from 0.1MHz to 1 MHz. The RI changes much slower along the frequency spectrum, so there are no apparent peaks or valleys in this band. The amplitude of the RI decreases gradually in the latter part. The RI we generally referred to (e.g. The CISPR and ANSI) is in this part. The last part is above 1 MHz, because of the limit of the sampling frequency of the DAQ, the results of this part may be not very accurate. This paper focused on the RI under 5 MHz.

Fig. 2. The spectrum of the RI for the test line.

As we know, the negative and positive corona are different, so the frequency of the voltage applied to the line may have influence on the RI. Figure 3 shows the measuring results of the RI under the voltage of different frequency. When the applied voltage changes from 10 Hz to 50 Hz, there are some obvious difference of the RI. Firstly, the amplitude of the peaks in the first part varies from each other, although the position basically remains the same. Because of the randomness, there is no strict changing rules, but we can tell that as the source frequency increases, the RI peaks become weaker. Secondly, the total energy of the RI is clearly stronger when the voltage is 10 Hz than that is 50 Hz. Finally, in the second part of the RI spectrum, the slope of the waveform turns to be flat as the frequency increases from 10 Hz to 50 Hz.

Fig. 3. The measure results of the RI under the 10 Hz to 50 Hz AC voltage.

Unusually, in the engineering measurement, the RI is a sum within a certain bandwidth. So sum up the Fourier transform results above every 1 kHz and the results are shown in Figure 4. Starting from 1 kHz, the strength of RI grows and reaches the peak between 0.01 MHz and 0.1 MHz, then it decreases and there is another peak between 0.1 MHz and 1 MHz. Figure 4 reflects the distribution of the energy alone the RI frequency spectrum. As the applied voltage changes, the distribution does not change much, but the value of the RI decreases.

Fig. 4. The summed up results of the RI with the bandwidth of 1 kHz under the 10 Hz to 50 Hz AC voltage.

To further support the argument above, the value of the RI at some specific frequencies (i.e. 0.1 MHz, 0.5 MHz, 1 MHz, 2 MHz and 3 MHz) are studied. As it shown in Figure 5, the curves of the RI are decreasing. That is to say, when the frequency of the voltage raises, the RI level of the transmission line comes down, even the changes are not apparent. As it is known, under the positive and negative voltage of the same amplitude, the RI is much stronger when the polarity is positive. When it comes to the AC voltage, the positive half cycle contributions more to the RI of the transmission line. Because the creating of positive corona pluses needs longer time and more energy than the negative corona pules, so the length of the positive half cycle influences the positive corona much more. Under the AC voltage of same amplitude, the shorter the positive half cycle is, the harder the positive corona forms. Therefore, the rise of the source frequency can suppress the RI.

Fig. 5. The trend of RI at some specific frequency as the source frequency changes

Apart from the distinction of the RI intensity, the energy distribution of the spectrum is changing with the source frequency. Divide the spectrum into three parts: under 0.1 MHz, 0.1 to 1 MHz and over 1 MHz. and the percentages of the total energy are shown in Figure 6. Under 0.1 MHz, it is about ten percent. And as the frequency of applied voltage rises, the percentage has a slight incensement. From 0.1 to 1 MHz, the percentage is about 40% and it only has a small

fluctuations when the source frequency changes. Last, for the RI over 1 MHz, it takes about 50% and the changing rule just contrasts to that under 0.1 MHz.

Fig. 6. The percentage of the energy in the three frequency bands (under 0.1 MHz, 0.1 to 1 MHz and over 1

MHz)

4. Discussion and Extension

In section III, we have studied the impact of the voltage frequency on the RI under 100 Hz. In order to find out more, the frequency then rises from 100 Hz to 1000 Hz and some interesting phenomena are observed: (a). When the frequency exceeds a specific value, like 500 Hz in this experiment, the RI intensity dose not decrease. On the contrary, the RI becomes a little stronger sometimes as the frequency rises. (b). The RI between 0.1 MHz and 1 MHz is relatively stable and can represent the RI level of the transmission line. So it is reasonable to use the RI at 0.5 MHz or 1 MHz to predict the RI of the line. (c).Unusually, under the AC voltage, the positive corona is the primary source for the RI. But if the voltage frequency is high enough, the negative will contribute more to the RI. On one hand, the variation of frequency changes the movement of the ions around the line which is the fundamental of the RI. On the other hand, the frequency can influence the inductance and capacitance of the line. So the RI is closely related to the source frequency. The conclusions and phenomena above may help us understand the RI better, and provide a new perspective on the study of RI.

5. Summary

In summary, the RI spectrum of the transmission line can be divided into three parts by its characteristic and the demarcation points are about 0.1 MHz and 1 MHz. The RI in the second part, namely between 0.1 MHz and 1 MHz, is relatively stable. The intensity of the RI decreases as the source frequency increases, but when the frequency reaches a certain level, the RI does not go down any more. It will stay the same or become a little stronger. As the frequency rises, some energy will transfer from high band to low band of the RI spectrum. This study may contribute to the analyzing the generation of the RI and help us to understand the RI from new perspectives.