Research on Radiated Disturbance to Secondary Cable Caused by Disconnector Switching Operation

Abstract

With the development of smart grids, the application of localized relay protection devices has greatly reduced the distance between the secondary equipment and the primary equipment. The secondary equipment will be in a more complex electromagnetic environment during the operation of the GIS disconnector. The present study takes the multi-path electromagnetic disturbance on the secondary cable caused by the disconnector switching operation of the domestic 1000 kV ultra-high voltage GIS test circuit as the research background, solves the field-line coupling problem based on the finite integral technique, and combines the multi-conductor transmission line theory to solve the radiation disturbance and obtains its influencing factors. The results demonstrate that the radiated disturbance accounted for 16% of the overall electromagnetic disturbance when both ends of the shielding layer are grounded. The use of grounding at both ends of the shielding layer, reducing the height of the secondary cable wiring, avoiding the parallel arrangement of the secondary cables and the GIS pipe mother, and installing a low-pass filter, have different levels of suppression effects on electromagnetic disturbances. The research results will guide the reasonable arrangement of secondary cables in GIS substations to some extent and have reference significance for the protection of secondary equipment.

1. Introduction

The switching operation in gas-insulated switchgear (GIS) can generate very fast transient overvoltage (VFTO), which may generate transient enclosure voltage (TEV) during its propagation. Both of them have high-frequency and high-amplitude characteristics [1,2]. VFTO and TEV may not only cause damage to the insulation of primary equipment, but also cause interference to secondary equipment. Electromagnetic disturbance couples to the secondary equipment in conducted and radiated ways. As the proportion of radiated disturbance is much smaller than that of conducted disturbance, it is usually ignored in the research of some scholars [3].

With the rapid development of smart grids, a large number of smart substation pilot projects have been put into construction. As shown in Figure 1, a large number of localized relay protection devices are used in smart substations. The secondary equipment is arranged on-site in the switchyard, and the distance between the primary equipment and the primary equipment is greatly reduced. This demonstrates that the electromagnetic environment where the secondary cable is located is worse, and the radiated disturbance will be more serious. So we need to pay enough attention to it.

At present, most of the calculation methods for the radiated disturbance problem adopt the classical field-line coupling model, including the Taylor model [4], Agrawal model [5] and Rachidi model [6]. Although the total induced voltage calculated by these three formulas is the same, the contribution of each component of the excitation electromagnetic field to the total induced voltage depends on the coupling formula used. The classical transmission line theory is based on the quasi-transverse electromagnetic wave (TEM) approximation, which can convert the problem of solving the “field” into the problem of solving the “circuit”. But this only applies to the case of low frequency, when the frequency is increased to the corresponding wavelength close to the lateral dimension of the transmission line, a larger model error will be generated. In order to solve this problem, Zhao Xiang [7,8] used the transmission-line super theory (TLST) model to calculate the distribution parameters of the non-uniform multi-conductor transmission line under high frequency conditions to solve the high-frequency field-line coupling problem. The calculation results are consistent with the full-wave analysis method.

Using the full wave method to directly solve the full Maxwell equations can obtain accurate calculation results without being affected by TEM, and the calculation accuracy will be greatly improved. Tattematsu et al. [9] studied the electromagnetic transient phenomenon caused by lightning through the finite difference time domain (FDTD) program, obtained the induced voltage of secondary cable and compared it with the measured value. Guo Jun [10,11] proposed a time-advancing method with time-domain and frequency-domain hybrids for the difficult-to-consider non-linear problems in frequency domain calculations and considered the lumped source excitation and the high-frequency response of the transmission line with non-linear load. Compared with the full wave method, it has good consistency and fast calculation speed.

Any electromagnetic disturbance problem can be divided into three parts: the disturbance source, the disturbance path and the disturbed device, and the suppression of the electromagnetic disturbance can also be carried out separately for the above three parts. At present, the protection of electromagnetic disturbance of secondary equipment is mainly carried out from the suppression of disturbance sources. Commonly used methods include the use of damping resistors and the installation of ferrite magnetic rings and arresters [12,13,14,15]. However, there are few studies on the suppression and protection of electromagnetic disturbance from the perspective of the disturbance path.

In view of the above-mentioned problems, the present study focuses on the multi-path coupling problem of the secondary equipment by operating the disconnector, and thoroughly studied the common influence of different disturbance paths on the secondary equipment. Through simulation calculation, the present study analyzed the level of overall disturbance voltage on secondary cables caused by the disconnector switching operation and the influencing factors and proposed the electromagnetic disturbance suppression measures for the secondary cable. The research results of the present study will guide the reasonable arrangement of secondary cables in smart substations and have guiding significance to the establishment of electromagnetic disturbance suppression measures for secondary equipment in smart substations.

2. Radiated Disturbance Coupling Mechanism

The VFTO on the GIS bus will radiate electromagnetic fields around the bus during the propagation process. Due to the shielding effect of the GIS shell, this part of the electromagnetic field only exists inside the GIS and is not reflected in the external space. The TEV propagating on the GIS shell will radiate electromagnetic fields to the space around the switchyard, and the radiation capacity is related to the location. For the GIS parallel to the ground, the GIS and the ground form a transmission line structure, so the ability of electromagnetic radiation is weak. However, the GIS casing is generally perpendicular to the ground and forms an antenna structure with the ground, so it has a strong ability to radiate electromagnetic fields and will cause harm to the electromagnetic environment in the power station.

The radiated disturbance problem of the secondary cable can be decomposed into two parts. The first part is that when the space electromagnetic field propagates to the secondary cable, the distributed current and distributed voltage will be induced on the armor layer of the secondary cable. In this part, the distributed disturbance sources along the armor layer can be obtained by solving the field-line coupling problem.

The second part is the crosstalk problem of the distributed current and distributed voltage induced intrusion into the core wire of the secondary cable on the armor. The secondary cable is arranged in parallel with the ground, which can be regarded as a multi-conductor transmission line system, including a “core wire-shielding layer”, “shielding layer-armor”, and “armor-earth” [12]. As shown in Figure 2, when the electromagnetic field radiates to the secondary cable, it will first propagate to the “armor-earth” transmission line system, and the distributed voltage U_ar and distributed current I_ar will be generated on the armor. The transient current I_a and transient voltage U_a flowing on the armor layer will be coupled to the shielding layer through the transfer impedance Z_t^’ and transfer admittance Y_t^’ between the armor and the shielding layer, and transient voltage U_s and transient current I_s will be generated on the shielding layer. They are further coupled to the core wire of the secondary cable through the transfer impedance Z_t and the transfer admittance Y_t between the core wire and the shielding layer, and finally a disturbance voltage U and a disturbance current I are generated on the core wire.

In summary, the radiated disturbance problem can be decomposed into two parts: the field-line coupling problem and the crosstalk problem. For the field-line coupling problem, as the electromagnetic field in space also has high-frequency characteristics, the assumption of electrical small size may no longer be satisfied between the armor and the ground. The classic transmission line method used in the past will inevitably produce errors, so it is necessary to use the full-wave method to solve this problem. As for the crosstalk problem, as the cross-sectional size of the secondary cable is generally relatively small, the assumption of electrical small size can be satisfied. Therefore, a multi-conductor transmission line model with the shielding layer as a reference can be established for the internal structure, and the shielding layer distributed current and distributed voltage can be used as excitation, combined with transfer impedance or transfer admittance to finally obtain the disturbance voltage level on the core under the excitation of the spatial electromagnetic field.

3. Simulation Calculation of Radiation Disturbance

3.1. Finite Integral Theory

The present study used the finite integration technique (FIT), one of the full-wave methods, to solve the field-line coupling problem. This technique is an algorithm derived from the FDTD algorithm. FIT discretizes Maxwell curl equations in time and space, using a hexahedral grid, and places the electric and magnetic fields on the edges of the grid and the center of the surface, using the second-order precision central difference to transform the differential equation into a difference form, and can directly perform numerical calculations [16,17].

Assuming that M_ε, M_μ, M_σ, and M_κ are the matrix forms of relative permittivity, permeability, conductivity and magnetic loss in the constitutive relationship, according to the derivation, the iterative formula of electric and magnetic field can be obtained as:

$$\begin{array}{l}{\overline{\mathbf{e}}}^{n+1}={\left(\frac{{\mathbf{M}}_{\sigma }}{2}+\frac{{\mathbf{M}}_{\epsilon }}{\Delta t}\right)}^{-1}\left[\left(\frac{{\mathbf{M}}_{\epsilon }}{\Delta t}-\frac{{\mathbf{M}}_{\sigma }}{2}\right){\overline{\mathbf{e}}}^n+\tilde{\mathbf{C}}{\overline{\mathbf{h}}}^{n+\frac{1}{2}}-{\overline{\overline{\mathbf{i}}}}_{e,i}{}^{n+\frac{1}{2}}\right]\\ {\overline{\mathbf{h}}}^{n+\frac{1}{2}}={\left(\frac{{\mathbf{M}}_{\kappa }}{2}+\frac{{\mathbf{M}}_{\mu }}{\Delta t}\right)}^{-1}\left[\left(\frac{{\mathbf{M}}_{\mu }}{\Delta t}-\frac{{\mathbf{M}}_{\kappa }}{2}\right){\overline{\mathbf{h}}}^{n-\frac{1}{2}}-\mathbf{C}{\overline{\mathbf{e}}}^n-{\overline{\overline{\mathbf{i}}}}_{m,i}^n\right]\end{array}$$

(1)

Among them, for any physical quantity s, $\overline{s}$ represents the line integral of the physical quantity s on the edge, and $\overline{\overline{s}}$ represents the area fraction of the physical quantity s on the grid surface. $\overline{\mathbf{e}}$ and $\overline{\mathbf{h}}$ respectively represent the line integrals of the electric field and magnetic field strength on the grid. ${\overline{\overline{\mathbf{i}}}}_{e,i}$ and ${\overline{\overline{\mathbf{i}}}}_{m,i}$ represent the area fraction of the current on the grid. C and $\tilde{\mathbf{C}}$ are the discrete curl operators on the primary and dual grids, respectively. Solving the above formula can obtain the distributed current and distributed voltage generated by the space electromagnetic field on the armor.

It can be found that FIT is a differential calculation of a matrix of discrete media materials. As the matrix is a diagonal matrix, there is no problem of matrix inversion difficulties.

3.2. Radiated Disturbance Simulation Calculation Results

In this paper, three-dimensional electromagnetic field simulation software based on finite integral method FIT is used to simulate the radiation disturbance of the secondary cable. As shown in Figure 3, a 3 m × 30 m thicknessless PEC plate was selected in the study to simulate the ideal earth, creating a KVVP2-22 quad cable with a length of 20 m above the PEC board, parallel to the x-axis, with the cable cross-section set, as shown in Figure 4. The cable-shielding layer is set to ground at both ends. In the model, the plane wave is used as the excitation source of the space electromagnetic field. The propagation direction of the electromagnetic wave is perpendicular to the earth, and the direction of the electric field is parallel to the x-axis and the direction of the magnetic field is parallel to the y-axis.

Then, the finite integral method is used to solve the distributed voltage and distributed current on the armor. In the process of solving, the boundary condition is set to the perfect matching layer (PML), which means that the electromagnetic waves are completely absorbed when they reach the boundary of the solution domain, in order to simulate the electromagnetic field problem in an infinite space and avoid the interference of the calculation caused by the reflected electromagnetic waves generated on the boundary.

The excitation source adopts the measured spatial electric field waveform of the literature [18], as shown in Figure 5. It is assumed that the propagation direction of the electromagnetic field points to the negative direction of the z-axis, the electric field direction is parallel to the x-axis, and the magnetic field direction is parallel to the y-axis. It can be observed that the disturbance source has been basically attenuated at 3 μs. Therefore, this study selected the total calculation time as 5 μs to obtain the distribution current and distribution voltage changes at each point on the armor, and the distribution voltage and distributed current at the end of the armor.

As the internal lateral dimension of the secondary cable is only millimeters, which is much smaller than the wavelength of the spatial electromagnetic field, the multi-conductor transmission line theory can be used to solve the crosstalk problem. The substituted calculated results of the distributed current and distributed voltage on the armor into the spice model of the secondary cable are shown in Figure 4b. The electromagnetic coupling problem can be solved by solving the multi-conductor transmission line equation; the disturbance voltage on the core wire of secondary cable can be obtained. Single_1 to Single_4 are the core wires of the cable, and Screen_1 and Screen_2 represent the shielding layer and armor of the cable, respectively. In the calculation, it is set that both ends of the cable core wire are connected with a matching resistance of 50 Ω, and both ends of the shielding layer are directly grounded.

Use the calculated distributed current and distributed voltage on the armor as the excitation to bring into the multi-conductor transmission line model of the secondary cable. Finally, the disturbance voltage level on the cable core can be obtained, as shown in Figure 6.

From the calculation results, it can be observed that the maximum disturbance voltage of the secondary cable terminal under the excitation of a typical space electromagnetic field is about 276 V, and the main frequencies are 5.8 MHz and 7.8 MHz, which are close to the main frequency of the disturbance source.

4. Influencing Factors of Radiation Disturbance

4.1. The Influence of Electromagnetic Wave Incident Direction on Radiation Disturbance

The propagation of electromagnetic waves can be described by three quantities, the incident angle Ψ, the azimuth angle φ, and the polarization angle θ, and the representative meaning is shown in Figure 7. In the calculation in the previous section, the electric field direction was assumed, but the actual electromagnetic field direction may be different from the calculation conditions. Therefore, this section studies the influence of different electromagnetic field directions on radiated disturbances. During the calculation process, the cable-shielding layer is kept grounded at both ends.

4.1.1. The Influence of Magnetic Field Direction on Radiation Disturbance

The azimuth and polarization angles are kept at 0° unchanged, and the disturbance voltage level when the incident angles are 0°, 30°, 60°, and 90° are calculated, respectively. At this time, the direction of the electric field is always parallel to the x-axis, and the direction of the magnetic field gradually changes from the y-axis to the z-axis. The obtained radiation disturbance voltage level is shown in

Table 1. The influence of the horizontal component of the magnetic field on the radiation disturbance.

Angle of Incidence (°)	Disturbance Voltage Amplitude (V)	Main Frequency (MHz)
0	276	5.8, 7.8
30	250	5.8, 7.8
60	181	5.8, 7.8
90	100	5.8, 7.8

.

The calculation results demonstrate that as the incident angle increases, the radiation disturbance voltage shows a downward trend. When the incident angle is 0°, the direction of the magnetic field points to the y-axis, which means it is perpendicular to the cable-earth plane. In addition, an induced current is generated in the circuit formed by the shielding layer and the ground, and then a disturbance voltage is generated on the core wire. As the incident angle increases, this part of the magnetic field component gradually decreases. When the incident angle is 90°, the direction of the magnetic field points to the z-axis, which means it is completely parallel to the cable-earth plane. At this time, the magnetic field part no longer disturbs the core wire, indicating that the radiation disturbance is related to the horizontal excitation field H_y.

4.1.2. The Influence of Electric Field Direction on Radiation Disturbance

The azimuth angle is kept at 0° and the polarization angle is kept at 90° unchanged, and the disturbance voltage level when the incident angles are 0°, 30°, 60°, and 90° are calculated, respectively. At this time, the direction of the magnetic field is always parallel to the x-axis, and the simulation results are shown in

Table 2. The influence of the vertical component of electric field on radiated disturbance.

Angle of Incidence (°)	Disturbance Voltage Amplitude (V)	Main Frequency (MHz)
0	3	7.2, 10.7, 21.5
30	240	7.2, 10.7, 21.5
60	419	7.2, 10.7, 21.5
90	484	7.2, 10.7, 21.5

. It can be found from the calculation results that as the incident angle increases, the disturbance voltage gradually increases. During this process, the direction of the electric field gradually changes from the y-axis to the z-axis, until the direction of the electric field is completely parallel to the z-axis when the incident angle is 90°, which demonstrates that the radiation disturbance is related to the vertical excitation field, E_z.

Table 3. The influence of the horizontal component of the magnetic field on the radiation disturbance.

Angle of Incidence (°)	Disturbance Voltage Amplitude (v)	Main Frequency (mhz)
0	100	5.8, 7.8
30	87	5.8, 7.8, 10.7
60	61	5.8, 7.8, 10.7
90	4	7.2, 10.7, 21.6

shows the disturbance voltage level when the incident angle is 90° and the polarization angle is 0° unchanged, and the azimuth angle is 0°, 30°, 60°, and 90°. At this time, the direction of the magnetic field is always parallel to the z-axis, which basically has no effect on the disturbance voltage, and the direction of the electric field gradually changes from the x-axis to the y-axis. As the azimuth angle increases, the disturbance voltage amplitude gradually decreases. When the azimuth angle is 90°, the electric field direction is parallel to the y-axis and perpendicular to the cable direction. At this time, the disturbance voltage is almost 0, indicating that the electromagnetic disturbance is related to the horizontal component E_x of the electric field.

Based on the above calculation and analysis, it can be found that the direction of the electromagnetic field has a greater impact on the electromagnetic disturbance level of the secondary cable. For the secondary cable laid parallel to the x-axis, the disturbance voltage at the core terminal is mainly related to the horizontal component E_x and the vertical component E_z in the electric field parallel to the cable. It is also related to the horizontal component H_y in the magnetic field perpendicular to the cable.

4.2. The Influence of Shielding Layer Grounding Method on Radiated Disturbance

This section compares the disturbance voltage levels on the core wire when the shielding layer of the secondary cable is not grounded, the switch station side is grounded at one end, the single-ended grounding measured by the control cabinet, and both ends are grounded. Keeping the incident angle of the electromagnetic wave constant, the core wire disturbance voltage under the four grounding modes is shown in Figure 8.

The calculation results demonstrate that when both ends of the secondary cable shielding layer are grounded or single-ended grounded on the side of the control cabinet, the disturbance voltage level on the core wire is the lowest, with amplitudes of 276 V and 359 V. When the shielding layer is grounded at the switchyard side or ungrounded, the radiated disturbance is serious, and the amplitude is 513 V and 1270 V.

Analyzing the reason, it is believed that when both ends of the shielding layer are grounded, the shielding layer and the ground will form a closed loop through two grounding wires. The disturbance current can circulate in this closed loop and generate an induced electromagnetic field opposite to the direction of the external excitation field, which can shield the core wire and reduce the disturbance voltage on the core wire. When the shielding layer is grounded at the single end on the side of the control cabinet, although no closed loop is formed, the potential at this position is reduced, and the disturbance wave on the shielding layer can flow into the ground network through this point, which also reduces the disturbance of the secondary cable voltage. When the shielding layer is grounded at a single end on the switchyard side, as the end of the cable shielding layer is suspended, a large induced voltage will be generated at the end of the shielding layer, which reduces the electromagnetic shielding effect of the shielding layer. When both ends of the shielding layer are not grounded, the secondary cable cannot be shielded, and the external excitation field will induce a serious overvoltage on the core wire of the cable.

4.3. The Influence of Cable Height above the Ground on Radiated Disturbance

This section studies the influence of wire height on the level of radiated disturbance. The shielding layer of the secondary cable is grounded at both ends to keep the direction of the electromagnetic field unchanged. Figure 9 shows the change of the secondary cable disturbance voltage when the cable height above the ground is 20 cm, 40 cm, 60 cm, and 80 cm.

It can be observed from the calculation results that the disturbance voltage level shows a linear increasing trend with the increase of the cable height above the ground. This is because the main influencing factors of radiated disturbance are the horizontal component of the magnetic field H_y, the horizontal component of the electric field E_x and the vertical component E_z. When the cable height increases, the horizontal component E_x of the electric field basically remains unchanged. For the horizontal component H_y of the magnetic field, the passing area increases in proportion to the cable height h. For the vertical component E_z of the electric field, the cable height h is the length of the integral path. As the excitation field component refers to the spatial electromagnetic field when the wire structure is not considered, assuming that the excitation field in the space is equal everywhere, as the height of the cable from the ground increases, the disturbance voltage will increase proportionally.

5. Analysis of Overall Electromagnetic Disturbance and Suppression Measures

5.1. Overall Disturbance Voltage

The conducted disturbance voltage caused by the disconnecting switch to secondary cable can be divided into two parts: capacitive component and resistive component, which are related to the stray capacitance of the transformer and the increase of the potential of the ground grid, respectively. According to our previous researching results [19], in the simulation calculation of conducted disturbance level, the broadband equivalent circuit model of transformer and ground grid is established by the vector fitting method and impedance synthesis method, and then the simulation model of conducted disturbance is built in EMTP/ATP combined with the measured voltage waveform of the conducted disturbance source in existing literatures. The simulation model of conducted disturbance is shown in Figure 10.

Combined with the calculation results of conducted disturbance, the overall disturbance voltage level generated by the operation of the disconnector on the secondary cable core wire can be obtained, as shown in Figure 11. Under the conditions of this study, the disturbance voltage amplitude of the secondary cable terminal is about 2.7 kV, the maximum peak-to-peak value is about 3.4 kV, and the main frequency is around 1 MHz and 6.7 MHz. In addition, it contains high frequency components of 20–40 MHz, corresponding to the main frequency of the disturbance source.

The IEC 61000-4-18 standard defines the damped oscillating wave generated by the operation of the GIS isolating switch on the secondary cable and gives the immunity test level, as shown in

Table 4. Fast damping oscillating wave test level.

Grade	Common-Mode Voltage (kV)
1	0.5
2	1
3	2
4	4
Xa 1	Special

¹ X^a can be any level and should be specified in the special equipment specification.

.

As the end conditions of the four-core cable are the same, the voltage on each core wire is the common mode voltage. Referring to Table 2 to evaluate the disturbance voltage obtained by the simulation calculation finds that the overall electromagnetic disturbance level does not exceed the level 4 standard. Analyze each component in the overall disturbance voltage and use the area ratio method to calculate the proportion of each component. It can be obtained that capacitive conduction disturbance accounted for 53% of the overall disturbance level, resistive conduction disturbance accounted for 31%, and radiated disturbance accounted for 16% (proportion of each disturbance voltage = voltage area corresponding to each disturbance path/total disturbance voltage area × 100%). It can be observed that although radiated disturbance accounts for less than conduction disturbance, it also exceeds 15%. Therefore, the radiated disturbance cannot be ignored in the simulation calculation of the overall electromagnetic disturbance.

5.2. Electromagnetic Disturbance Suppression Measures

5.2.1. Shielding Layer Is Properly Grounded

Internationally, there is a consensus that shielded cables should be used for secondary cables, but opinions on the grounding method of the shielding layer are not uniform. IEEE believes that the shielding layer should be grounded at one end on the side of the control cabinet, while IEC and China’s National Grid believe that the cable-shielding layer should be grounded at both ends of the switchyard and the control cabinet. By comparing the changes of the disturbance voltage when the shielding layer of the secondary cable is grounded at the switchyard side, single-ended grounded at the control cabinet side, and grounded at both ends, it can be found that the shielding layer grounding method has a greater impact on the disturbance voltage. The shielding effect of the secondary cable is better when single-ended grounding on the side of the control cabinet or both ends grounding, and the amplitudes are 1.5 kV and 2.7 kV. The single-ended grounding disturbance voltage at the switchyard side is serious, with the maximum amplitude reaching 8 kV, far exceeding the IEC’s immunity test level.

Table 5. Variation of disturbance voltage under different grounding methods.

Grounding Method	Ground at Both Ends	Single-Ended Grounding on the Side of the Control Cabinet	Single-Ended Grounding on Switch Station Side
Capacitive conduction amplitude (V)	1564	1306	5587
Capacitive conductivity specific gravity (%)	53	61	68
Resistive conduction amplitude (V)	1045	89	2818
Resistive conduction specific gravity (%)	31	11	20
Radiated disturbance amplitude (V)	273	358	513
Proportion of radiated disturbance (%)	16	28	12

shows the variation results of each component in the overall disturbance voltage.

The shielding layer grounding method has a greater impact on the conducted disturbance, and the resistive conducted disturbance amplitude is only 89 V when single-ended grounding on the side of the control cabinet. However, it reaches 2.8 kV when single-ended grounding at the switch station side, and capacitive conduction and resistive conduction have the same changing trend. Grounding at both ends of the shielding layer has the best suppression effect on radiated disturbance, and the radiated disturbance reaches the maximum when the switch station side is grounded at one end. But the amplitude is much smaller than conducted disturbance. Therefore, it is recommended that the secondary cable-shielding layer must be grounded on the side of the control cabinet.

In reference [12], based on the same test circuit in this paper, the overall disturbance voltage of the secondary cable was measured under the same grounding modes: grounding at both ends of the secondary cable shield and single-ended grounding at the side of the control cabinet or switchyard, which are 395 V, 295 V and 4557 V, respectively. Compared with the results of 2.7 kV, 1.5 kV, and 8.0 kV obtained in this paper, we can draw the following two conclusions:

• The results in this paper have the same laws as that in reference [12], which means that when the shield layer is only grounded at the switchyard side, the disturbance voltage is significantly higher than the results of the other two grounding methods.
• There are two reasons for the difference between the simulation value and the experimental value: (a) the length of the secondary cable is 20 m in simulation and 100 m in experimental measurement. Theoretically, the longer the cable length, the lower the disturbance voltage at the end of the core wire; (b) The VFTO and TEV used as the excitation source in the calculation are the actual measured values under the most severe conditions of the test loop, so all the simulation calculation results are strictly considered.

Above all, although there are some differences between the simulation results and the measured values in this paper, the laws they presented are the same, and the numerical differences can be explained by analyzing the different simulation conditions.

5.2.2. Reasonable Wiring of Secondary Cables

The wiring height of the secondary cable is the main factor in the wiring. Assuming that the shielding layer of the secondary cable is always grounded at both ends, the wiring height of the secondary cable is changed.

Table 6. Disturbance voltage changes under different cable heights above the ground.

Ground Clearance (cm)	20	40	60	80
Capacitive conduction amplitude (V)	1441	1345	1261	1172
Capacitive conductivity specific gravity (%)	46	38	32	29
Resistive conduction amplitude (V)	916	808	728	651
Resistive conduction specific gravity (%)	24	21	19	17
Radiation disturbance amplitude (V)	546	949	1331	1669
Proportion of radiated disturbance (%)	30	41	49	54

shows the variation results of each component in the overall disturbance voltage. The overall disturbance voltage level gradually increases with the increase of wiring height, from 2.5 kV at a height of 20 cm from the ground to 2.9 kV at a height of 80 cm from the ground, but it does not exceed the IEC level 4 standard.

The calculation results demonstrate that as the wiring height increases, the conducted disturbance amplitude shows a gradual decrease, but the radiated disturbance amplitude rises rapidly. When the height of the cable from the ground increases, the overall electromagnetic disturbance is more severely affected by the radiated disturbance. When the height of the cable from the ground increased from 20 cm to 80 cm, the proportion of radiated disturbance increased from 30% to 54%, which exceeded the conducted disturbance and became the main influencing factor of overall electromagnetic disturbance. Therefore, when the height of the cable from the ground is high, it is recommended that the shielding layer is also grounded on the switch station side. Grounding at both ends can make the transient current generate a circulating loop, and the induced magnetic field generated by it will cancel out the external excitation field, thereby reducing the overall electromagnetic disturbance level.

Considering the wiring direction of another element of the wiring of the secondary cable, assuming that the secondary cable is arranged directly below the GIS bus, the grounding point positions at both ends will always remain unchanged during the process of changing the direction of the secondary cable wiring. At this time, the conducted disturbance will remain at a constant value, and only the radiated disturbance is changing. With the increase of the angle between the secondary cable and the horizontal projection of the GIS bus, the radiated disturbance changes are shown in Figure 12.

When the included angle changes from 0° to 90°, the radiated disturbance voltage decreases, accordingly. When the included angle is 90°, the radiated disturbance is almost zero. This is because when the included angle is 0°, the secondary cable is arranged directly below and parallel to the GIS bus. At this time, the magnetic field generated by the transient current flowing through the GIS shell will be orthogonal to the cable-earth plane, so a high-amplitude distributed current will be induced on the secondary cable shielding layer, and then a higher amplitude disturbance voltage will be generated on the core wire.

Based on the above analysis, it can be concluded that the layout height of the secondary cables should be reduced as much as possible, and the loop should be as far away from the disturbance source as possible and avoid parallel arrangement with the GIS bus.

5.2.3. Install Low-Pass Filter

A filter is an electronic device that filters useless frequency signals and allows useful frequency signals to pass. Installing a filter between the signal circuit and the power circuit can effectively prevent conducted disturbances, as the high frequency component of disturbance voltage studied in this paper will disturb the normal operation of secondary equipment. Therefore, installing a low-pass filter to filter the high-frequency components in the disturbance voltage, and allowing the low-frequency signal to pass normally, can be considered

This text adopts the simplest RC filter and obtains the change situation of the overall disturbance voltage after the low-pass filter of different cut-off frequency is shown, as in Figure 13.

The calculation results demonstrate that after the low-pass filter is installed, the content of high-frequency components exceeding the cut-off frequency is significantly reduced, and the amplitude is also reduced. Moreover, the lower the cut-off frequency, the more obvious the suppression effect. When the cutoff frequency is reduced from 10 MHz to 5 MHz, the filter filters most of the high-frequency components in the disturbance voltage, and the amplitude of the disturbance voltage is significantly reduced at this time. When the cut-off frequency of the low-pass filter is 1 MHz, the overall disturbance voltage level is lower than 1 kV, which is smaller than the second level of the IEC standard.

However, it can be found from Figure 13 that although the low-pass filter is installed, the high-frequency components exceeding the cut-off frequency are greatly suppressed. However, the frequency components below the cutoff frequency are also partially attenuated. Therefore, when using this method to suppress the disturbance voltage of the secondary cable in practice, the cut-off frequency of the low-pass filter must be selected reasonably to prevent the high-frequency signals that need to be collected by the secondary equipment from being filtered out.

6. Conclusions

This study focuses on the study of the electromagnetic disturbance to the secondary cable caused by the disconnector switching operation in GIS. Firstly, a simulation model of radiation disturbance is established based on a finite integral method and multi-conductor transmission line theory. Then, the radiation disturbance levels under different influencing factors are discussed in detail based on the model. Finally, the suppression measures for the overall disturbance on the secondary cable are proposed under the condition of considering the conducted disturbance. The simulation results demonstrate that the incidence direction of the electromagnetic wave, grounding mode of secondary cable shielding layer and cable height have great influence on the level of radiation disturbance. Through simulation calculation, it can be found that although the conducted disturbance is the main components of the overall disturbance, in some cases the radiation disturbance proportion also cannot be ignored. Reasonable grounding through the secondary cable-shielding layer, changing the way of secondary cable wiring, and installing a low-pass filter can effectively restrain the level of overall disturbance. The specific conclusions are as follows:

• By combining the finite integral method and the theory of multi-conductor transmission lines, a simulation calculation model of the radiated disturbance voltage is established. The influence of electromagnetic wave direction, shielding layer grounding method, and the height of cable above ground on radiation disturbance are compared. The results demonstrate that for cables laid parallel to the x-axis, the radiation disturbance is mainly related to the horizontal component Ex and the vertical component E_z of the electric field and is related to the horizontal component H_y in the magnetic field perpendicular to the cable. The shielding layer grounding method also has a great impact on the radiation disturbance. The shielding layer is not grounded, or the single-ended grounded on switch station side will produce serious electromagnetic disturbance, and the disturbance voltage level is the lowest when the two ends are grounded. Besides, the height of the cable from the ground has a great effect on the radiation disturbance. With the increase of height above ground, the disturbance voltage increases approximately and proportionally.
• Conducted disturbance is the main component of the overall disturbance voltage generated by the disconnector operation. However, when both ends of the shielding layer are grounded, the proportion of radiated disturbance has reached 16%; thus, it should not be ignored in the simulation calculation of the overall disturbance voltage.
• The shielding layer grounding method has a greater impact on electromagnetic disturbance. For conducted disturbance, the single-ended grounding effect on the side of the control cabinet is the best, while for radiated disturbance, the shielding effect at both ends is the best. Single-ended grounding at the switchyard side will produce serious disturbance voltage. Therefore, the secondary cable-shielding layer must be grounded on the side of the control cabinet.
• The radiated disturbance will increase significantly, as the height of the cable from the ground increases. It is recommended that the height of the secondary cable should be reduced as much as possible. When the cable height is high, it is recommended that both ends of the shielding layer be grounded. The loop should be as far away as possible from the disturbance source and avoid parallel arrangement with the GIS bus.
• Installing a low-pass filter can significantly reduce the disturbance voltage level, and the lower the cut-off frequency, the better the suppression effect. However, the cut-off frequency cannot be too low to prevent the high-frequency signals that need to be collected by the secondary equipment from being filtered out.

In this paper, the radiation disturbance of the secondary cable caused by the disconnecting switch operation in GIS is studied in depth and some achievements have been made in the study of electromagnetic disturbance influencing factors and suppression measures. There are still some problems in this paper that need to be further studied in the later stage, and the prospect of the follow-up work is put forward here:

• This paper mainly studies the simulation calculation of secondary cable electromagnetic disturbance. The typical value of the disturbance source provided in literature is used as excitation to calculate the disturbance source, but the complete waveform of disturbance source can not be obtained and there are certain errors. The simulation calculation of electromagnetic disturbance source (VFTO and TEV) should be further carried out.
• This paper takes GIS test circuit of Wuhan UHV AC test base as the research object, which has a simple layout and a small number of equipment. For the actual GIS power station, the on-site protection device and other equipment makes the secondary loop more complex, and the high-frequency electromagnetic transient characteristics of different secondary equipment terminals need to be considered, which greatly increases the workload and solving time of simulation modeling. How to properly simplify the modeling calculation considering the dual requirements of computational efficiency and engineering accuracy is the main problem for subsequent engineering applications.