Surface Charge Accumulation on Basin-Shape Insulator in Various Eco-Friendly Gases with Metal Particle Under AC Voltage

Abstract

Surface charge accumulation is considered one of the key factors that lead to unexpected insulator flashover failures in gas-insulated switchgear (GIS). With the existence of metal particles, charge accumulation characteristics on insulator surfaces become intricate in eco-friendly gases under AC voltage. In this study, the surface charge behavior on a down-scaled 252 kV AC GIS basin insulator model with a linear metal particle adhered to the HV electrode on the convex surface in compressed air (80%N₂/20%O₂) and C₄F₇N/CO₂ mixtures was investigated. After applying an AC voltage of 40 kV for 5 min, the charge densities on both surfaces were measured, and the effect of the metal particle and gas parameters was discussed. The results showed that charge spots were induced by metal particles on the insulator surfaces, and the polarities of which varied with the gas atmosphere. A decrease in maximum charge density was detected with an increase in C₄F₇N proportion at 0.1 MPa, and soar of which was observed at 0.5 MPa. With an increase in gas pressure, the maximum charge density increased in both atmospheres. The total quantity of charges showed similar behavior to the charge densities. It is indicated that the high electronegativity of C₄F₇N molecules presents a competing relationship in charge accumulation as the pressure increases.

1. Introduction

Gas-insulated switchgear (GIS), with the advantage of higher reliability, lower land occupation, and fewer maintenance requirements, has been widely employed in power systems in recent years [1]. As a vital insulating component of GIS, insulators play a critical role in ensuring its safe and steady operation [2]. With the long-term application of high voltage (HV) and possible pollution of metal particles, charges tend to deposit on the gas-solid surface of epoxy insulators [3,4], which may distort the local electric field and lead to a lurking flashover [5]. Therefore, understanding the characteristics of charge accumulation behavior on the insulator has significant importance in benefiting the power system. Typical metal particle shapes which can be observed in GIS are spindly linear, screw linear, schistose, spherical, and dust-like [6].

Under the action of the electric field, free charges migrate and accumulate on polymer insulator surfaces, which are affected by multiple factors, including applied voltage characteristics, attachment of metal particles, gas state parameters, etc. B. Qi et al. [7] found that in SF₆, the charge accumulation on insulators under DC voltage showed an obvious polarity effect, and the charge density accumulated under AC voltage was smaller than the result under DC voltage. J. Wang et al. [8] reported that the accumulation of the surface charge surge might be caused by the metal particles attached to the surface of the insulator, and the surge caused by the particles attached to the middle part was more remarkable. H. Wang et al. [9] found that the total amount of surface charge in 20% SF₆/N₂ was far more than that in pure SF₆ under DC voltage, but the maximum surface charge density in 20%SF₆/N₂ was approximately equal to that in pure SF₆. Apart from the proportion, charge behaviors are influenced by pressure as well. W. Du et al. [10] pointed out that when the gas pressure ranged from 0.1 MPa to 0.7 MPa, the charge density on the surface of insulator increased with a trend of gradual saturation. Overall, plentiful efforts have been made to elucidate the charge behavior in SF₆ related atmospheres; charge accumulation under DC and AC voltages differs distinctly and could be affected by metal particles.

As equally high dielectric strength and relatively low global warming potential (GWP), eco-friendly gases, such as compressed air and the C₄F₇N/CO₂ mixture, are believed to be promising alternatives to SF₆. T. Rokunohe et al. [11] reported that the dielectric strength for air was roughly equal to the 10%SF₆/N₂ mixture when the gas pressure was 0.6 MPa. Z. C. Li et al. [12,13] revealed that small quantities of C₄F₇N mixed in CO₂ could increase the surface flashover voltage, and as the gas pressure and the fraction of C₄F₇N increased, the surface flashover voltage increased. The flashover voltage of insulators in the 9%C₄F₇N/91%CO₂ gas mixture at 0.6 MPa was approximately equal to that in SF₆ at 0.5 MPa. Scholars have also done some research on charge accumulation in C₄F₇N/CO₂ mixtures in recent years. J. Dong et al. [14], for example, pointed out that compared to SF₆, the C₄F₇N/CO₂ mixture was more sensitive to the electric field, enhancing gas ionization and gas conduction, resulting in more charges accumulating in C₄F₇N/CO₂ mixtures. Previous studies have generally focused on the surface charge accumulation characteristics and the effect of metal particles in SF₆ under DC voltage. However, under AC voltage, charge accumulation characteristics on insulator surfaces become more intricate in eco-friendly gases with the existence of metal particles, which have not been studied thoroughly.

In this study, the surface charge behavior on a down-scaled 252 kV AC GIS basin insulator model attached with a linear metal particle in compressed air and C₄F₇N/CO₂ mixtures was investigated. Various gas pressures and proportions were considered and an instance free of metal particles was carried out for comparison. The results indicated that surface charge spots would appear at both the convex surface and the corresponding area of the concave surface where the metal particle was placed. The maximum charge density within the charge spot increased with the decrease of C₄F₇N proportion and the increase of gas pressure.

2. Materials and Methods

2.1. Sample and Experiment Platform

In this paper, a down-scaled basin insulator, whose prototype could be utilized in a 252 kV AC GIS, was selected as the sample. This down-scaled insulator was produced by Taikai Group Co. Ltd., Taian, China, and the diagram of the sample is illustrated in Figure 1.

An enclosed multifunctional testing platform was used to charge and measure the surface potential of the sample. A sectional view of the chamber is demonstrated in Figure 2. The HV electrode was used to charge the insulator, which was placed at the bottom of the chamber. Valves and the pressure gauge were used to create certain gas atmospheres during the experiment. The accuracy of the pressure gauge is 0.0001 MPa, which was sufficient for forming gas mixtures of a certain proportion. The measurement of the potential was undertaken via a Kelvin probe (3455ET, Trek, New York, NY, USA) connected to an electrostatic voltmeter (P0865, Trek). The measuring system is comprised of a shifting platform and a rotating platform (MTS and MRS Series, Becic, Beijing, China), whose detailed demonstration could be found in our previous publication [15]. 648 points of surface potential were sampled within 252 s, then the charge density could be obtained through an inversion algorithm based on COMSOL Multiphysics 6.2 [15].

2.2. Test Procedure

Prior to the surface charge accumulation test, the sample was dried at 60 °C for 12 h to remove the absorbed moisture. Due to the fact that the metal particles in GIS are not always spherical, more particles are linear particles and needle particles [16]. To focus on the effect of metal particles on surface charge distribution, an aluminum linear metal particle of 5 mm was chosen in this experiment. The length of 5 mm would sufficiently distort the electric field and strengthen the gas-side ionization, while a larger size would facilitate the experiment. As shown in Figure 3, the metal particle was adhered to the HV electrode of the convex surface of the sample, and an instance of no metal particle was set up for comparison. The position of the metal particle was selected to trigger the most severe electric field distortion and gaseous ionization according to our previous work [17].

Before each test, the sample was carefully wiped with absolute alcohol and dried in the drying oven for 1 h to guarantee that there were no residual charges on the surface. The charging and testing process was conducted at room temperature. To simulate a realistic electric field distribution in the real-sized 252 kV GIS, a coaxial electrode was designed, and the amplitude of the AC voltage was set to 40 kV. According to our previous study [18], after applying HVAC of 2, 5, 30, and 60 min, the characteristics of the charge distribution were almost unchanged, and the charge accumulation tended to a dynamic balance. Thus, the charging time here was set to be 5 min in forming a relatively large charge accumulation and facilitating multiple repetitions of the experiment.

The eco-friendly gases adopted in the experiment were divided into two parts in general: C₄F₇N/CO₂ mixtures and compressed air. For the C₄F₇N/CO₂ mixtures, the proportion of C₄F₇N was set to be 4%, 6%, 8%, and 10%, and the pressure was set to be 0.1, 0.3 and 0.5 MPa. For the compressed air, the proportion was set to be 80%N₂/20%O₂, and the pressure was set to be 0.3, 0.5, and 0.7 MPa. After the charging process, the surface potential was measured, and the corresponding charge density of the sample was obtained. Each test was repeated at least 5 times, and the typical results were presented and discussed in this paper.

3. Results

The top view of the typical charge distribution on the sample after being charged by an AC voltage of 40 kV for 5 min is displayed below. The locations and the values of the maximum positive and negative charge densities are marked in the figures. The boundary between the positive and the negative charge densities is separated by a contour line.

3.1. Effect of Metal Particles on Surface Charge Distribution

Figure 4 and Figure 5 display the charge accumulation on both surfaces of the sample after being charged in C₄F₇N/CO₂ mixture and compressed air respectively. The results show a significant accumulation of charges when the metal particle was attached to the sample. As shown in Figure 4b, a distinct negative charge spot with a maximum density of −4.08 pC/mm² appears beneath the metal particle. As for Figure 5b, a distinct positive charge spot with a maximum density of 3.08 pC/mm² appears at the same place. The charge accumulation characteristics on the concave surface are similar, but the density of the charges is lower than that on the convex surface. As shown in Figure 4d and Figure 5d, the maximum charge densities accumulated are −1.35 pC/mm² and 0.37 pC/mm² respectively.

When there is no metal particle, few surface charges are accumulated on the sample surfaces in both atmospheres, as shown in Figure 4a,c and Figure 5a,c, neither the maximum charge densities in both atmospheres are larger than −1 pC/mm². To sum up, charge spots will appear at both the convex surface and the corresponding area of the concave surface where the metal particle is placed under AC voltage, and the densities on the concave surface are lower. The charge spots detected in C₄F₇N/CO₂ mixtures are negative, and those in compressed air are positive.

3.2. Effect of Proportion on Surface Charge Distribution with Metal Particle

Figure 6 and Figure 7 display the charge accumulation on both surfaces of the sample after being charged in C₄F₇N/CO₂ mixtures of various proportions at 0.1 MPa.

It can be seen that the maximum charge density decreases with the increase of C₄F₇N proportions at 0.1 MPa. As shown in Figure 6a and Figure 7a, at a proportion of 4%C₄F₇N/96%CO₂, the maximum charge density on the convex surface and the concave surface is −4.94 pC/mm² and −1.37 pC/mm² respectively. When it comes to 10%C₄F₇N/90%CO₂, the maximum charge density decreased to −3.18 pC/mm² and −0.92 pC/mm² respectively, which is approximately 65% of that in 4%C₄F₇N/96%CO₂.

To illustrate the variation of the quantity of the charge accumulated on the sample more specifically, the absolute values of the total positive and the total negative surface charges on both surfaces were calculated respectively. Figure 8 displays the trend in the quantity of the charges on both surfaces as the proportion increases, and it appears that the charge quantity tends to increase initially and decrease subsequently. Additionally, in the atmosphere of C₄F₇N/CO₂ mixtures, the total quantity of negative charges is significantly greater than the positive charges, and the quantity on the convex surface is higher than that on the concave surface. To summarize, in C₄F₇N/CO₂ mixtures, with an increase in C₄F₇N proportion, the maximum surface charge density decreases, and the total quantity of the charges increases initially and decreases subsequently.

Figure 9 and Figure 10 display the charge accumulation on both surfaces of the sample after being charged in C₄F₇N/CO₂ mixtures of various proportions at 0.5 MPa.

It can be seen that the maximum charge density increases with the increase of C₄F₇N proportions at 0.5 MPa. As shown in Figure 9a and Figure 10a, at a proportion of 4%C₄F₇N/96%CO₂, the maximum charge density on the convex surface and the concave surface is −6.82 pC/mm² and −2.26 pC/mm² respectively. When it comes to 10%C₄F₇N/90%CO₂, the maximum charge density increased sharply to −23.70 pC/mm² and −6.12 pC/mm² respectively, which is approximately 3 times that in 4%C₄F₇N/96%CO₂. Furthermore, as shown in Figure 11, the total charge quantity increased with the proportion of C₄F₇N significantly as well.

To summarize, at low pressure, with an increase in C₄F₇N proportion, the maximum surface charge density decreases, and the total quantity of the charges increases initially and decreases subsequently. When it comes to high pressure, with an increase in C₄F₇N proportion, both the maximum surface charge density and the total quantity of the charges increase sharply.

3.3. Effect of Pressure on Surface Charge Distribution with Metal Particle

Figure 12 and Figure 13 display the charge accumulation on both surfaces of the sample after being charged in C₄F₇N/CO₂ mixtures of various pressures.

As the pressure goes up, the maximum density of charges on both surfaces appears to increase sharply, in Figure 12a and Figure 13a, at the pressure of 0.1 MPa, the maximum charge density is −4.08 pC/mm² and −1.35 pC/mm² respectively. When it comes to 0.5 MPa in Figure 12c and Figure 13c, the maximum charge density soars sharply to −14.95 pC/mm² and −3.90 pC/mm² respectively, about 2–3 times as much as that at 0.1 MPa. Furthermore, as shown in Figure 14, the total charge quantity increases with the pressure significantly as well.

Compared with C₄F₇N/CO₂ mixtures, the charge accumulation characteristics in compressed air are generally similar but with differences. Figure 15 and Figure 16 show the charge accumulation on both surfaces of the sample after being charged in compressed air of various pressures. As well as the results in C₄F₇N/CO₂ mixtures, the maximum density of charges on both surfaces also increases with pressure, but the charges accumulated in the charge spots are positive and the pace of the increase is relatively slow. As shown in Figure 15a,c and Figure 16a,c, the maximum charge density at 0.7 MPa is only about 1–2 times as much as that at 0.3 MPa.

Different from the results in C₄F₇N/CO₂ mixtures, both the negative charges and the positive charges are accumulated fairly much on the surfaces in compressed air. Figure 17 shows that with an increase in pressure, the quantity of positive charges accumulated on the surfaces increases while the quantity of negative charges decreases. To sum up, with an increase in gas pressure, the maximum surface charge density in both atmospheres increases. A decrease in the total quantity of negative charges is observed with the increase of pressure in compressed air.

4. Discussion

The surface charge can be accumulated on the epoxy insulator through the following three ways: deposition of charges from gas ionization on the insulator surface under the external electric field, the charge injection from the electrode, and the charge transportation in the bulk or along the surface [19,20]. In this paper, the charge distribution behavior on the insulator surface is analyzed from the perspective of partial discharge and gas ionization primarily because of the absence of continuous and stable electric field forces under AC voltage.

When there is no metal particle, the electric field on the sample is relatively even, and the partial discharge and gas ionization are weak, thus the amount of charge accumulation is particularly rare. With the presence of a metal particle, similar to a needle electrode connected to the HV, the electric field around it will be severely distorted; the partial discharge and gas ionization are strong, generating numerous charged particles. Moreover, the considerable distortion caused by the metal particle will lead to a great increase in the normal component of the electric field on the surface of the insulator, which will also contribute to the accumulation of the charges.

The distorted electric field prompts a swift migration of electrons from the tip of the metal particle to the surface of the sample and the gas side ionization. Because of the strong electronegativity of C₄F₇N, the electrons can be easily adsorbed to form negative ions. As shown in Figure 18, in the atmosphere of C₄F₇N/CO₂ mixtures, the negative ions migrate towards the surface of the insulator under the electric field force during the negative half cycle of AC voltage. Due to their large mass and slow speed, these negative ions near the insulator cannot completely migrate away or be fully neutralized before the next negative half cycle. Over time, the negative ions are settled on the sample and the negative charge spots are formed. In the atmosphere of compressed air, the electronegativity of molecules is not that strong, and fewer electrons are absorbed. Due to the severe distortion of the electric field, positive ions are generated by the gas ionization and settle onto the sample [21,22]. During the positive half cycle, positive ions are driven to the surface of the insulator by the electric field. After the voltage polarity is reversed, these positive ions near the insulator cannot completely migrate away or be fully neutralized before the next polarity reversal as well. As time goes on, positive charge spots are formed at last.

At low pressure, the number of gas molecules is relatively small, the mean free path is longer, and the migration of ions is easier. With an increase in C₄F₇N proportion, the number of C₄F₇N molecules increases accordingly. More electrons are absorbed by C₄F₇N molecules, which moderates the gas ionization. Therefore, the number of electrons and ions settled on the sample will be reduced, leading to a decrease in charge density and total quantity.

As the pressure increases, however, the number of gas molecules increases rapidly. In the atmosphere of C₄F₇N/CO₂ mixtures, despite the mitigation of ionization caused by C₄F₇N, the distortion of the electric field induced by the metal particle is so severe. The strongly distorted electric field and the large quantity of gas molecules still strengthen the gas ionization, which contributes to the increase in surface charges. The behavior of charges in compressed air is alike. When the pressure increases, the number of gas molecules increases, generating more positive ions and resulting in an increase in surface charges.

At high pressure, the number of gas molecules is relatively large, the mean free path is shorter, and the migration of ions is harder. The negative C₄F₇N ions generated by the absorption of electrons cannot completely migrate away. With the increase of C₄F₇N proportion, the number of C₄F₇N molecules within per unit increases. Large quantities of negative C₄F₇N ions accumulate around the metal particle, forming a layer of space charge. As shown in Figure 19, during the half cycle of the AC voltage, the space charge layer aggravates the distortion of the electric field, leading to a second discharge and finally an increase in surface charge accumulation.

To summarize, a metal particle attached to the electrode will distort the electric field, which will result in a strong gas ionization and a great increase in the normal component of the electric field, leading to a significant surface charge accumulation. The increase of C₄F₇N proportion will mitigate the gas ionization at low pressure. When it comes to an increase in pressure, the quantity of gas molecules increases. The increase of C₄F₇N proportion will form a layer of negative ions around the metal particle, which aggravates the distortion of the electric field and leads to a second discharge. Numerous ions formed by the gas ionization and the absorption of C₄F₇N molecules settle on the sample, resulting in a surge of charge accumulation.

5. Conclusions

In this article, the surface charge accumulation characteristics on a down-scaled basin insulator in C₄F₇N/CO₂ mixtures and compressed air under AC voltage have been investigated, and the main conclusions can be summarized as follows.

• Few surface charges are accumulated on metal-particle-free samples, while charge spots will appear at both the convex surface and the corresponding area of the concave surface where the metal particle is placed under AC voltage, and the densities on the concave surface are lower. The charge spots detected in C₄F₇N/CO₂ mixtures are negative, and those in compressed air are positive.
• At low pressure, with an increase in C₄F₇N proportion, the maximum surface charge density decreases, and the total quantity of the charges increases initially and decreases subsequently. When it comes to high pressure, with an increase in C₄F₇N proportion, both the maximum surface charge density and the total quantity of the charges increase sharply.
• With an increase in gas pressure, the maximum surface charge density in both atmospheres increases. The total quantity of both charges increases with pressure in C₄F₇N/CO₂ mixtures, while a decrease in the total quantity of negative charges is observed with the increase of pressure in compressed air.

In short, it is found that metal particles adhered to the HV electrode of the convex surface of insulators could induce an accumulation of charges, which is affected by different parameters of the operating atmosphere. Such a finding deepens the understanding of surface charge accumulation on epoxy insulators under AC voltage and provides valuable information for the gas selection of an eco-friendly GIS.