X International Symposium on Lightning Protection

9 -13 November, 2009 Curitiba, Brazil

SIMULATION EXPERIMENTS ON IMPULSE CHARACTERISTICS OF GROUNDING GRIDS IN DIFFERENT SOIL STRUCTURES

State Key Laboratory of Power Transmission Equipments & System Security and New Technology, Chongqing University, Chongqing 400030, China, yuantao_cq@cqu.edu.cn 2 Chongqing Electric Power Companies, Chongqing 400014, China, zhouyinch@yahoo.com.cn the parameters such as the structure of the grounding grid, the current injection point, the resistivity and structure of the soil and so on, which affect the impulse characteristics. As a consequence, it is an effective and important method for researching the transient response of the grounding grid. In this paper, simulation experiments are performed to analyze the impulse characteristics of the grounding grid in homogeneous and two-layer soils by considering the factors such as current injection location, grid buried-depth and soil structure. The procedures of the test are commonly as follows: at first, the grounding device models are made by scaling down the practical ones; then the models are buried in medium like sand or water.; by measuring the current and potential on the grounding grid models when lightning impulse currents are injected into, the regulations of the influence factors of the impulse characteristics of grounding grids are presented. Researches at home and abroad show that the result is accurate and reliable as long as the simulation test is carried out in strict accordance with the simulated test theory. It is believed that the simulation experiment has the same physical process as the full-scale experiment. 2 EXPERIMENTAL DEVICES AND METHODS

Yuan Tao1, Sima Wenxia1, Yang Caiwei1, Yang Qing1, Zhou Yinchun2

Abstract - In order to find the correct impulse characteristics of the grounding grids, impulse grounding simulation experiments are carried out by employing impulse current generator in uniform and two-layer soils which is based on principle of dimension similarity. This test mainly measured impulse grounding resistance in uniform soil with five different embedment depths and three different current-injected points, ground potential rise (GPR) in homogenous soil and impulse grounding resistance in two-layer soils with three different kinds of upper layer soil thickness and three different current-injected points. The experimental results indicate that the factors such as current injection location, grid buried-depth and soil structure have influences on the impulse characteristics of the grounding grids, and the influence regulations are described in paper. Key words - grounding grid; simulation impulse experiment; impulse characteristics; influence factors; soil structure; influence regulation

INTRODUCTION

When lightning strike took place in or near the substation, and short-circuit faults happened in GIS substation, high-frequency impulse currents with big peak value may inject to the grounding systems. This situation makes the comprehension of the transient phenomena and safety performance of the large-scale grounding grid extremely complicated. Therefore, it is of great significance to analyze of the impulse characteristics of grounding grid for the optimization design. Currently, three methods are usually used to evaluate the performance of substation grounding grid [1]: field measurement, numerical calculation and simulation experiment. The field measurement may cause high cost, and the operation is difficult. Moreover, this method can only be carried out after the grounding grid has been built instead of during the process of designing. The models of the grounding grids are more or less based on certain assumptions when using numerical calculation method, so it cannot fully represent the actual grounding grids. The simulation experiment method may overcome shortcomings above, and furthermore it allows changing

2.1 Test Set-up The double exponential test circuit used in the investigation that the grounding grid buried in homogeneous soil is shown in figure 1. The capacitance of the charging capacitor bank C by paralleling eight low-inductance capacitors is 2F; the front resistor R is 6 and the adjustable wave inductance L is 32.97L. The impulse current generator shown in the dotted line box in figure 1 was charged up by 380V power frequency voltage through the isolated transformer and voltage transformer. Then the impulse current can be formed and injected into the grounding grid model in the water pool by triggering the ignition ball-gap. Current measurement is achieved with a tubular current shunt that is rated at

347

10kA with the resistor value of 0.019357 . A capacitive

voltage divider with a ratio of 100:1 is used for voltage

Fig.1 Wiring diagram of impulse characteristic experiment in homogenous soil T1: Isolated Transformer (rated at 3kVA 0); T2: 3000 /100V Transformer; D: Silicon Stack; R: Water Protection Resistance; C: Capacitor Bank; G: Ignition Ball-gap; L, R: Waveform Adjustable Resistance and Inductance; F: capacitive voltage divider; S: Tubular Resistance Shunt; DSO: Digital Storage Oscilloscopes.

Fig 2 Wiring diagram of impulse characteristic experiment in two-layer soils K1: Main Contactor Switch; FU: Fuse; C1, R3: Overheating Protection Device; ICM: Intelligent control module; R4, R5: Current-Limiting Resistance for Transformer; T: 60kV Transformer; R: Charging Resistance; D: Silicon Stack; C: Capacitor Bank; G: Ignition Ball-gap; R1, L1: Waveform Adjustable Resistance and Inductance; FYQ: Resistance Voltage Divider; FLQ: Tubular Current Shunt; DSO: Digital Storage Oscilloscope; S: Grounding Grid Model; P: Sand Pond.

measurement. The voltage and current waveform signals are captured on a Tektronix TDS3012C, 100MHz Digital Storage Oscilloscope (DSO) by cable. The wiring diagram of impulse characteristic experiment in two-layer soils is shown in figure 2. The impulse current generator as a big HV test platform named ICGS was constructed by Chongqing University recently. It can produce some representative waveform impulse currents such as 1.2/50s, 8/20s, 10/350s and 10/1000s and the maximum peak value of the output short-circuit current is up to 200kA. This test platform with intelligent control system using optical fiber for the transmission of measurement and control information has good human-machine conversation function interface. By operating through the touch screen, the process of charging and discharging and signal recording can be accomplished automatically. On the other hand, the measures enhancing the system security is adopted to prevent weak current equipment from suffering harm of transient surge. The waveform of the impulse current in this study is 8/20s, with peak for 10 kA.

2.2 Test Methods The simulation experiment in this paper is based on dimensional similarity principle[2], which ensures that the grounding device geometry size, embedding depth, soil resistivity, injection current and so on in the simulation experiment are proportional to the ones in the practical situation. In this way, the test results are able to reflect the impulse characteristics accurately. In homogeneous soil, not only the impulse resistance of the grounding grid but also the ground potential rise (GPR) was analyzed when the grounding grid were buried at five different depths through three different current injection point locations. In two-layer soils, the impulse grounding resistance was studied with three different thicknesses of topsoil and three different current injection points. The impulse grounding resistance is a ratio of the peak value of the voltage to current recorded by the digital storage oscilloscope. In order to reduce accidental measurement error, each kind of experiment was repeated for six times under the same test conditions, and the average value of four recorded data excluding the maximum or the minimum data was used as effective measurement value.

348

The interval of every two impulse tests should be not less than five minutes [3]. 3 RESULTS AND ANALYSIS

3.1 The impulse grounding resistance of the grounding grid in homogenous soil The simulation experiments in homogenous soil were carried out in a pool with the dimension of 10m10m2.5m by using water with the resistivity of 25 m as uniform soil. The test object is a square grid model with four meshes made of iron wire conductors with the radius of 0.002m. The length of each side of the grid is 1.0m. The impulse current has been injected from the center point, corner point of the grid and middle point of a side during the experiment (see figure 3). The grounding grid is placed at the depths of 0.06m, 0.08m, 0.10m, 0.12m and 0.14m under the water surface respectively. The peak value of the 8/20us impulse currents is 300A. The scale coefficient N of the full-scale grounding grid made by conductors with the radius of 0.02m to the grid model is 10:1. As a result, the length of the side of the corresponding full-scale grounding grid is 10m and the corresponding actual impulse wave is 80/200us scaled. On the other hand, the soil resistivitives and relative permeabilities of the the full-scale experiment system and the simulation system should be the same. The resistivity and relative permeability of iron are 1.710-7 m and 636 respectively. By using the grounding calculation software CDEGS and the parameters above, the impulse grounding resistance of the corresponding full-scale grounding grid can be computed. The computational results are shown in table 1. The relationship of impulse grounding resistance between the full-scale and simulating grounding devices is
Fig 3 The injection points of impulse current Table 1: Impulse grounding resistances in uniform soils

Buried-depth

Injection points Center Middle Corner Center Middle Corner Center Middle Corner Center Middle Corner Center Middle Corner Test 1.44 1.49 1.58 1.35 1.45 1.56 1.32 1.35 1.53 1.29 1.33 1.42 1.30 1.34 1.40

Rch1( ) CDEGS 1.45 1.50 1.61 1.36 1.47 1.58 1.34 1.37 1.54 1.30 1.36 1.48 1.31 1.35 1.42

0.6m

0.8m

1.0m

1.2m

1.4m

Rch1 =

Where Rch impulse grounding resistance; the subscript 1 presents the full-scale experiment system; and subscript 2 presents the simulation system[4]. So the impulse grounding resistance Rch1 of respective full-scale grounding grid can be obtained by converting from the simulation experimental results based on simulation principle as shown in table 1. From table 1, it can be found that the value of Rch1 computed by CDEGS and gained from the simulation test is nearly equal. It can be concluded that the simultion test in this work is rational.

1 Rch 2 N

By the comparison study of the impulse test results in different condition from Table 1, we can get some variation regularity as follows: 1) When the impulse current is injected at the same point, the impulse grounding resistance Rch1 decreases with the buried-depth of the grounding grid; when the buried-depth exceeds a certain value, even if the buried-depth increases, but the Rch1 decreases slowly or even has an increase trend. This is probably a consequence of multi-direction dispersal process of the impulse current from the grounding grid into the soil. The soil parts over and under the grid plane can also dissipate current. When the buried-depth increases, the dispersal performance of the soil around the grounding grid improves because the effective region for current dispersing expands in upper soil parts. On the other hand, the action sphere of the impulse current in the ground is finite. After the buried-depth reaches a certain value, the effective region for current dispersing will not expand. If the buried-depth goes on increasing, on the contrary, the connection wire will become longer so that it will go against the dispersal process [5].

(1)

349

2) When the buried-depth keeps unchanged, the impulse grounding resistance of the grid is influenced by the location of the current injection obviously. The value of Rch is the least and the biggest when the currents are injected at the center and the corner of the grid respectively. The results may be due to the inductance effect of the conductor. When an impulse current is injected into a grounding device, the inductive impedance of grounding conductor is very large due to the very high frequency of impulse current, which obstructs the current flowing to the far position of the grounding device. Consequently, the grounding grid has an effective geometrical dimension and the impulse current mainly disperses in this area. So the effective dispersal area is the largest when the injection location is at the center point. As a result, some large grounding grids in substation with strong inductance effect should be connected at the center as far as possible with the electrical equipments on the ground [6]. 3) When the current injection points are different, the decrement of differs from each other if the buried-depth increases per unit. When the injection locations are at the corner and the center of the grid and at the middle of one side, the decreasing trends of Rch1 are the most obviously at the buried-depths of 1.0m, 0.8m and 0.6m respectively. Therefore, in the engineering design, we can choose the most economical and efficient buried-depth according to different current injection points. 3.2 The ground potential rise (GPR) in homogenous soil The grounding grid used in the GPR measurement experiment is a square grid model with 49 meshes. The side length of the grid is 0.7m. The conductors with the radius of 0.002m are arranged with the same distance from each other. Figure 4 illustrates three directions of observation lines a, b and c and the numbering nodes. The numbering rule is that the first and second numbers of the node name stand for the row and column numbers of the corresponding node respectively. The impulse current is injected at the node NO.44. Table 2 shows the measured GPR values on grounding grid surface. From Table 2, it can be concluded that the grounding potential rises are extremely uneven at different nodes when an impulse current is injected. The GPR is the highest at the current injection point. On the same direction of observation lines, the farther the nodes are away, the lower the GPR is. It is because that when the high frequency impulse current is injected into the grid, the current can not disperse from the all the grid area into the ground rapidly due to the inductance effect; instead,

Fig 4 Observation profile and node of grounding grid Table 2 The measured potential values ongrounding grid surface

Node 11 22 33 44 55 66 77 88

GPR (V) 2.48 3.00 3.97 5.14 3.73 3.02 2.63 2.13

Node 14 24 34 44 54 64 74 84

GPR (V) 2.65 3.11 4.02 5.14 3.94 3.13 2.72 2.31

Node 41 42 43 44 45 46 47 48

GPR (V) 2.62 3.19 4.01 5.14 3.92 3.15 2.70 2.35

the impulse current just disperses in a finite region near the injection point. Moreover, the leakage current density decreases with the distance from the injection point. Thus the GPRs of the edge and corner nodes of the grid are lower than those near the injection point. On different direction of observaion lines, although the distance from the mesuring nodes to the impulse current injection point is the same, the GPRs of the nodes may be different. For instance, the nodes NO.14, NO.41 and NO.22 have equidistance, but the GPR at node NO.22 is higher than these of the other two points. This is mainly because that point 22 locates at the inner part of the grounding grid, where the self-inductance and multi-indcuance of the node are greater. Certainly, for the structural symmetry, the ponits 14 and 41 nearly have the same GPR. Similar situation is also present among the nodes NO.48, NO.84 and NO.77. 3.3 The impulse grounding resistance of the grounding grid in two-layer soils The simulation experiments for measuring the impulse grounding resistance Rch are carried out in the Grounding Technology Laboratory in Chongqing University. The diameter of the semi-sphere sand pond constructed by steel is 5 m. The soil is simulated by fine sand. The resistivity of the sand can be changed by heating the sand and by spraying tap water or salt water. The upper soil thickness of the two-layer soils is set as 1.2m, 0.9m and 0.6m in the sand pond. The resistivity of the upper and lower soils is 500 m and 170 m respectively. The peak value of the 8/20us impulse current used in the simulation

350

test is 10kA. The grounding grid has the same structure as the one shown in figure 3. But its geometrical dimension changes to be 0.5m0.5m. The radius of the conductors is 0.0007m and scale coefficient N of the full-scale grounding grid made of conductors with the radius of 0.021m to the grid model is 30:1. Therefore the length of the side of the corresponding full-scale grounding grid is 15m and the corresponding actual impulse wave is 240/600us scaled Table 3 gives the value of Rch gained by the simultion test that has been converted to the full-scale grounding grid. The results computed by CDEGS of the impulse grounding resistance of the corresponding full-scale grounding grid are also given in Table 3.
Table 3 Impulse grounding resistances in two-layer soil

much more evident in two-layer soils. Eventually, this leads to the values of Rch1 varying a lot. 4 CONCLUSION

In our study, impulse grounding simulation experiments are carried out in homogenous and two-layer soils and the impulse characteristics of grounding grid are analyzed. From the comparison the value of impulse grounding resistance computed by CDEGS and gained from the simulation test, it can be found that the results are almost equal. This illustrated that the simultion test in this work is rational. a) The closer the current injection points are to center of the grounding grid, the smaller the impulse grounding resistances are; when the soil resistivity of the upper layer soil is much more than that of the lower layer soil, the impulse grounding resistances are influenced by the current injection locations more obviously in two-layer soils than that in homogenous soil. b) The trends of the impulse grounding resistance decreasing with buried-depth are inconsistent when impulse currents are injected into the grounding grid at different points. When the injection locations are at the corner and the center of the grid and at the middle of one side, the decreasing trends of Rch are the most obviously at the buried-depths of 1.0m, 0.8m and 0.6m respectively. c) The stratified soil structure strongly influences the impulse grounding resistance. The experimental results indicate that when the upper layer soil thickness increases from 0.85 to 1.7 times of the diagonal length of the ground grid, the corresponding Rch increases to 2~3 times of its former value if the upper layer soil resistivity is bigger than that of the lower soil. 5
[1] [2] [3] [4]

Thickness of upper soil 18m

Injection points Center Middle Corner Center Middle Corner Center Middle Corner

Rch1( ) Test 5.92 6.94 7.69 7.83 8.88 10.31 13.29 18.32 23.21 CDEGS 5.96 7.02 7.85 8.04 8.94 10.37 13.33 18.38 23.29

27m

36m

By comparing with the values of Rch1 from the impulse test results in different condition, we can get the conclusions: 1) The value of Rch1 increases with the upper soil thickness in two-layer soils. The diagonal length of the corresponding full-scale grounding grid is 21m. When the upper soil thickness increases from 0.85 to 1.7 times of the diagonal length, the corresponding Rch increases to 2~3 times of its former value. This is resulted from that the upper soil has bigger resistivity than that of the lower soil. Therefore impulse currents prefer dispersing into the lower soil to dispersing in the upper soil. If the upper soil gets thicker, the dispersal effect of the grounding grid will become worse and the value of Rch will rise significantly [7-8]. 2) From the data in Table 1 and Table 2, we can find that the values of Rch1 are influenced by the injection points more obviously in two-layer soils than in homogenous soil at the same buried-depth (0.6m). It is also caused by the different soil structures. The impulse currents tend to disperse into the lower layer soil rather than the upper layer soil with higher resistivity, which make the difference of effective grid area with different injection

REFERENCES

[5] [6]

[7]

Zhang Renyu, Chen Changyu, Wang Changchang. High-voltage Testing Technology, Beijing, China: Tsinghua University Press, 2003. A. B. Kopcyntsef, Simulation of impulse property of grounding device, Elektpitestvo, no. 5, pp. 3135, 1958. Russian. Zhang Jinyu, Guan Zhicheng, High-voltage Test, Beijin, China: China Electric Power Press, 1996. J. L. He, R. Zeng, Y. P. Tu, et al, Laboratory investigation of impulse characteristics of transmission tower grounding devices, IEEE Transaction on Power Delivery, vol.18, no.3, pp.994-1001, 2003. He Jinliang, Zeng Rong, Grounding Technology in Power System , Beijin, China: Science Press, 2007 F. P. Dawalibi, Wei Xiong, Jinxi Ma, Transient performance of substation structures and associated grounding systems, IEEE Transaction on Industry Applications, vol.31, no.3, pp.520-526, 1995. F. P. Dawalibi, Jinxi Ma, R.D.Southey, Behavior of grounding systems in multilayer soils: a parametric

351

[8]

analysis, IEEE Transaction on Power Delivery, vol.9, no.1, pp. 334-342, 1994. Alan Selby, F. P. Dawalibi, Determination of current distribution in energized conductors for the computation of

electromagnetic fields, IEEE Transaction on Power Delivery, vol.9, no.2, pp. 1069-1075, 1994.

352