An EMTP – RV Based Analysis of the Line Surge Arrester Energy Duty Due to Lightning Discharges S. Grebovic, S. Pack, S. Sadovic
Abstract—This paper describes modelling procedure for calculating energy stresses of transmission line surge arresters. A 110 kV shielded transmission line is modelled using ElectroMagnetic Transient Program – Restructured Version (EMTP – RV). Some parts of transmission line such as transmission line towers and stations on both ends of line are modelled thanks to EMTP – RV possibility of grouping electrical elements into subcircuits. Also it was possible to create subcircuit consists of number of other subcircuit. This possibility was very important for dividing transmission line into number of sections that consists of number of towers. Arrester current and voltage, energy duty, insulator voltage and voltage on tower top due to bipolar lightning are calculated and presented. Keywords: Bipolar Lightning Stroke, Line Surge Arrester, Energy duty, EMTP – RV, Transmission line, Modelling, Simulation.
I. INTRODUCTION
T
HERE are several methods used for the improvement of lightning performance of the existing transmission lines, such as: tower footing resistance reduction, increase of line insulation level, installation of additional ground wires, etc. Line surge arresters (LSA) are mainly used for line lightning performance improvement. LSA may fail in service if it is not correctly selected. In the selection of the line surge arrester for the line lightning performance improvement it is very important to determine arrester energy duty [1]. Energy stresses of transmission line surge arresters (LSA) are commonly analyzed due to unipolar flashes that transport to ground charge of one polarity. Also, it is very important to analyze energy stresses on transmission line surge arresters due to bipolar flashes that can cause serious problems in electrical power systems. Modelling is of great importance in power system transient analysis. Modelling can be used to describe system, to analyze the behavior of a system under conditions that can occur in nature. The main goal is to achieve precise simulation with all required modelling data and within minimal computer timings. So, this paper presents S. Grebovic is PhD student at Graz University of Technology and with Faculty of Electrical Engineering University of Sarajevo (e-mail:
[email protected]). S. Pack is with Institute of High Voltage Engineering and System Management, Graz University of Technology (e-mail:
[email protected]). S. Sadovic is with the Department of Electrical Engineering, Faculty of Electrical Engineering University of Sarajevo (e-mail:
[email protected]). Paper submitted to the International Conference on Power Systems Transients (IPST2015) in Cavtat, Croatia June 15-18, 2015
modelling of complete transmission line with large number input and output data. II. MODELLING PROCEDURE Transmission line is modelled into several parts: towers, insulators, phase conductors, shield wire, line surge arresters and tower footing resistance. Lightning stroke is modelled as a current source (so called 'I – point by point source'). The studied 110 kV transmission line is 42 km long, span or line length between two towers is considered as a mean value of 300 meters. It is taken that line has a different footing resistance along the line route, so tower footing resistance value is varied. Soil resistivity is 1200 Ωm. In the case when tower is equipped with one LSA it is installed on bottom phase. For installation configuration with two LSA, LSA are installed on bottom and middle phase [2]. A. Tower model Transmission line tower model, used in simulation is presented in Figure 1.
Fig. 1. Tower representation.
Each tower is divided in four parts. First part is section of the tower from bottom crossarm to the ground and it is represented as the propagation element, which is defined by the surge impedance and the propagation length. In EMTP – RV software this part is modelled as constant parameter (CP) and line model (1 – phase version). Second part is sections between tower top and top crossarm. Third and fourth part
T1
are sections between crossarms. Sections on the tower top (between tower top and top crossarm and between crossarms) are modelled as inductance branches. On this way it was possible to calculate transient on tower top. Tower surge impedance is calculated according tower shape theory. Branch inductance is determined according to the section length, tower surge impedance and propagation velocity. Wave propagation speed on the tower was taken to be equal to the velocity of light [1]. Figure 2 shows described tower model in EMTP – RV with insulators and pins that enable electrical connection with other elements. On this way it is possible to connect line surge arresters in parallel to the phase conductor insulators and model tower footing resistance.
c
C1
0 L1
2.8uH
+
Pin for connection with tower top
504kV
A2
A1
?vi>S + Air1
+ L2
1.37uH
Air2
B2
C3
c
0
L3
1.37uH
+
0
C2
c
504kV
504kV
C2
C1
CP
20
TLM1
+
?vi>S + Air3
T2
Pin for connection with tower footing resistance
Fig. 2. Tower model in EMTP – RV. Pin for connection with tower top
DEV1
Tower top
Phase A
B2 Pins for connection with phase B
Phase B
B1
Pin for connection with tower footing resistance
where: vgap – voltage across insulator string V0 – reference voltage K=1 D=0,2045d d – length of gap (m) t – time to flashover (µs). CFO – critical flashover voltage can be represented by equation (2) [4].
710 V0 0, 9CFO 0, 9 400 0,75 d t
(2)
C. Line surge arrester
+ ?vi>S
B1
B. Line insulation flashover model Insulators are modeled as Air Gap elements. It operates based on equal area flashover model. Flashover occurs when the following integral becomes greater or equal to D [3,4]: K t dt D (1) v gap t V0 t0
Phase C
A2 Pins for connection with phase A A1
C2 Pins for connection with phase C C1
Grounding
Fig. 3. EMTP – RV symbol for tower subcircuit.
All elements shown on Figure 2 are grouped in subcircuit. This subcircuit is illustrated on Figure 3 and can be use many times for modelling complete transmission line.
Used arrester has the following characteristics [2]: Rated voltage: 108 (kVrms) MCOV: 86 (kV) IEC class: II Nominal discharge current 10 (kA) The non – linear behavior of line surge arrester is represented by the U – I characteristic [2,3]. D. Lightning current model Bipolar lightning flashes transport to ground both negative and positive charges [5, 6]. A possible explanation of observed bipolar lightning currents is given by Narita et al. They suggested that, in a bipolar discharge, currents of both polarities follow the same channel to ground [7]. There are basically three types of bipolar lightning discharges, although some events may belong to more than one category [6, 8]. Lightning current (bipolar) shown on Figure 4 was measure on Corsica. The negative charge transfer is determined by integrating the absolute value of the current over the negative current component duration. The positive charge transfer is determined by integrating the current over the positive current component duration. The trapezoidal rule is used for integration. The total transfer is summary of the negative and positive charge transfers. The negative part has peak current of -52,86 kA and positive part has peak current of +26,3 kA, with total charge transfers being -3,3025 As and +1,2414 As, respectively. Lightning stroke is modelled as a current source (so called 'I – point by point source') because it is possible to import measured data point by point. The current shapes and values can be changed in order to study different lightning stroke currents.
Bipolar Lightning Current 40 t =55,2 (s) 1
30
t2 =78,5 (s)
SADOVIC measurements
t =81,2 (s) 3
20
Current (kA)
10 0 -10 -20 -30 -40 -50 -60 0
100
200
300 Time (s)
400
500
600
Fig. 4. Bipolar Lightning Current.
E. Transmission line sections Shield wires and phase conductors are subdivided into number of segments. Each segment is presented with constant parameter (CP) line model (multiphase). Middle of the span for phase A Middle of the span for shield wire
Middle of the span for phase B
Connection with tower top
Middle of the span for phase C
T1 T_1_2_s T_1_2_a T_1_2_b T_1_2_c T2 T_2_3_s T_2_3_a T_2_3_b T_2_3_c T3 T_3_4_s T_3_4_a T_3_4_b T_3_4_c T4 T_4_5_s T_4_5_a T_4_5_b T_4_5_c T5 T_5_6_s T_5_6_a T_5_6_b T_5_6_c T6 T_6_7_s T_6_7_a T_6_7_b T_6_7_c T7 T_7_8_s T_7_8_a T_7_8_b T_7_8_c T8 T_8_9_s T_8_9_a T_8_9_b T_8_9_c T9 T_9_10_s T_9_10_a T_9_10_b T_9_10_c T10
DEV1
S_1 A_1 B_1 C_1
S_10 A_10 B_10 C_10
Section_from_1_to_10
Fig. 5. EMTP – RV section of 110 kV transmission line.
Each span of 300 m is divided in two segments to enable connection between current source and middle of the span for shield wire and phase conductors. Certain number of tower subcircuiet elements, spans, line surge arresters, grounding footing resistances and corresponding pins (for connection with other elements) are grouped into subcircuit element that presents transmission line section. Example for section that consists of 10 towers is given on Figure 5. It is possible to model sections with 50 or more towers. In this paper some sections are consists of 50 towers and some of them are consists of 10 towers. Stations are modelled on both ends of 110 kV transmission line. On Figure 5 is illustrated the main idea for EMTP – RV modelling in this work. The idea was grouping subcircuits inside other subcircuits with possible connections between all elements in model. On both ends of section shown on Figure 5 are pin connections for shield wire and phase conductors. This means that is possible to make connection between two or more section subcircuit elements to get complete
transmission line. On the top of section subcircuit element shown on Figure 5 are pins for lightning current source connection with tower top and middle of the span. The position of the lightning stroke is varied from hitting the shield wire or the phase conductor in the middle of the span or directly at tower top. Waveform shapes that can be observed in this simulation are: currents for all line surge arresters, energy duties for all line surge arresters, flashover currents and voltages for all insulators and tower footing resistance voltage and current. III. CPU TIMERS To the author’s best knowledge, it is first time that complete transmission line is created and solved in EMTP – RV for purposes of transient analyses. CPU time is the amount of time for which a central processing unit (CPU) was used for processing instructions of a computer program. Amount of input and output data is large, so it is important to show CPU timers. Simulation time was 600 (µs). Time – step
conductor induces a voltage on the phase conductor. Voltages on the tower top and at the bottom phase are shown on figure 10. Energy duty 9 Energy duty for arrester installed on bottom phase 8 7 6 Energy (kJ)
was 50 (ns). Simulation statistics are: total number of network nodes 1742, size of the main system equations 2175, total number of solution points in time-domain 15441, total number of iterations 47972 and number of iterations per timepoint 3,11. The CPU timings are presented in Table 1. CPU timings are based on a Quad – core i7 processor (logical processors: 8 and 2,2 GHz) with 8 GB of RAM. There are no parallel computation in this simulation. TABLE I CPU TIMINGS (S) CPU timers (s)
5 4 3
Prepare data Read data Time-domain A matrix formulation Time-domain B vector updating Time-domain Ax=B solution Solution of control systems Time-domain history updating Time-domain solution Total
0,23438 0,01562 0,0 1,23438 22,29688 0,15625 2,20312 26,68750 27,18750
2 1 0 0
100
200
300 Time (s)
400
500
600
Fig. 8. Energy duty for arrester – case 1. Voltage across insulator strings 800 Voltage across insulator strings - top phase Voltage across insulator strings - middle phase Voltage across insulator strings - bottom phase 600
In this section are given simulation results in the case when lightning source is bipolar lightning current from Figure 4. Three cases are considered. The first case is that lightning hitting tower top and on tower is installed one line surge arrester and tower footing resistance is 35 Ω. On Figures 6 and 7 are presented respectively current and voltage for arrester installed on bottom phase.
Voltage(kV)
400
IV. SIMULATION RESULTS AND DISCUSSION
200
0
-200
-400 0
100
200
300 Time (s)
400
500
600
Fig. 9. Insulator voltages – case 1. Voltage on tower top and at bottom phase 600 Voltage on tower top Voltage at bottom phase
400
Arrester current 3
200
Arrester current - bottom phase
0
Voltage(kV)
2.5
Current (kA)
2
-200 -400 -600
1.5
-800
1
-1000 -1200
0.5
-1400 0
100
200
0
300 Time ( s)
400
500
600
Fig. 10. Voltages on the tower top and at the bottom phase – case 1. -0.5 0
100
200
300 Time (s)
400
500
600
Fig. 6. Current through arrester – case 1. Arrester voltage 300 Arrester voltage - bottom phase
200
Second case is that lightning hitting tower top and on tower are installed two line surge arrester and tower footing resistance is 50 Ω. On Figures 11 and 12 are presented respectively currents and voltages for arresters installed on bottom and middle phase. Arrester current 4
Voltage (kV)
100
Arrester current - bottom phase Arrester current - middle phase
3.5 0
3 2.5
Current (kA)
-100
-200
-300 0
100
200
300 Time (s)
400
500
600
Fig. 7. Arrester voltage – case 1.
On Figure 8 is shown energy duty for arrester installed on bottom phase. Voltage waves across insulators are presented on Figure 9. Coupling between the shield wire and the phase
2 1.5 1 0.5 0 -0.5 -1 0
100
200
300 Time (s)
400
Fig. 11. Currents through arresters – case 2.
500
600
Arrester voltage 300 Arrester voltage - middle phase Arrester voltage - bottom phase
respectively currents and voltages for arresters installed on bottom, middle and top phase.
200
Arrester current 5 Arrester current - top phase Arrester current - middle phase Arrester current - bottom phase
100 Voltage (kV)
4 0
Current (kA)
3 -100
2
-200
1 -300 0
100
200
300 Time (s)
400
500
600
0
Fig. 12. Voltages on arresters – case 2.
On Figure 13 is shown energy duty for arresters installed on bottom and middle phase. Voltage waves across insulators are presented on Figure 14. Voltages on the tower top and at the bottom phase are shown on figure 15.
-1 0
200
300 Time (s)
400
500
600
Fig. 16. Currents through arresters – case 3. Arrester voltage 300 Arrester voltage - top phase Arrester voltage - middle phase Arrester voltage - bottom phase
Energy duty
200
12
100 Voltage (kV)
10
8 Energy (kJ)
100
0
-100
6
-200
4
-300 0
2
100
Energy duty for arrester installed on middle phase Energy duty for arrester installed on bottom phase 0 0
100
200
300 Time (s)
400
500
600
Fig. 13. Energy duties for arresters – case 2. Voltage across insulator strings 800 Voltage across insulator strings - top phase Voltage across insulator strings - middle phase Voltage across insulator strings - bottom phase
200
300 Time (s)
400
500
600
Fig. 17. Voltages on arresters – case 3.
On Figure 18 is shown energy duty for arresters installed on bottom, middle and top phase. Voltage waves across insulators are presented on Figure 19. Voltages on the tower top and at the bottom phase are shown on figure 20.
600
Energy duty 14
400
200
10
0
8
Energy (kJ)
Voltage(kV)
12
-200
6
4 -400 0
100
200
300 Time (s)
400
500
600
2
Energy duty for arrester installed on top phase Energy duty for arrester installed on middle phase Energy duty for arrester installed on bottom phase
Fig. 14. Insulator voltages – case 2. Voltage on tower top and at bottom phase
0 0
1000
100
200
Voltage on tower top Voltage at bottom phase
300 Time (s)
400
500
600
Fig. 18. Energy duties for arresters – case 3.
500
Voltage across insulator strings 300 Voltage across insulator strings - top phase Voltage across insulator strings - middle phase Voltage across insulator strings - bottom phase 200 -500
100 Voltage(kV)
Voltage(kV)
0
-1000
-1500
0
-100 -2000 0
100
200
300 Time (s)
400
500
600
-200
Fig. 15. Voltages on the tower top and at the bottom phase – case 2.
Third case is that lightning hitting tower top and on tower are installed three line surge arrester and tower footing resistance is 60 Ω. On Figures 16 and 17 are presented
-300 0
100
200
300 Time (s)
400
Fig. 19. Insulator voltages – case 3.
500
600
Voltage on tower top and at bottom phase 1000 Voltage on tower top Voltage at bottom phase 500
Voltage(kV)
0
-500
-1000
-1500
-2000 0
100
200
300 Time (s)
400
500
subcircuits inside other subcircuits is very useful for modelling large systems. It has been shown that is possible to model complete transmission line and provide plenty of output data. Despite of fact that there is a plenty of input and output data CPU time is not long. Line surge arrester energy duty depends of arrester current shape. So, a knowledge of the occurrence and characteristics of bipolar lightning is needed for the selection of the arrester class (arrester block size) as well as for designing adequate lightning protection schemes for various objects and systems.
600
Fig. 20. Voltages on the tower top and at the bottom phase – case 3.
In Table 2 are presented LSA current and voltage peaks and energies for all three cases.
VI. REFERENCES [1]
TABLE II LSA DATA CASE
Phase Bottom Middle Top Bottom Middle Top Bottom Middle Top
1 LSA
2 LSA
3 LSA
U (kV) 256,27 263,15 263,27 269,28 261,18 250,67
I (kA) 2,86 3,52 3,54 4,23 3,32 2,28
W (kJ) 8,24 11,93 9,73 12,38 9,68 6,15
In table 3 are shown peak values of insulator voltages. TABLE III INSULATOR VOLTAGES CASE 1 LSA
2 LSA
3 LSA
Phase Bottom Middle (no LSA) Top (no LSA) Bottom Middle Top (no LSA) Bottom Middle Top
[2]
[3]
[4] [5] [6] [7] [8]
Ui (kV) 256,35 771,38 786,59 263,15 263,53 793,29 269,49 261,63 251,37
From presented arrester current shapes it is possible to see that the arrester current peaks are not so high as peak of injected lightning stroke. The arrester voltage shape for all cases is similar but with different peak values. It depends of LSA position (is it installed on bottom, middle or top phase) and tower footing resistance. From energy duties consideration we can see that arrester energy rise up as bipolar lightning current arriving at arrester. The corresponding energy duties are not so high. Reasons for that are short durations of arrester currents and fact that considering transmission line is shielded. Insulator voltages are high, especially for phases without LSA installed (for example 771,38 kV). Line critical flashover voltage for 110 kV is 550 kV. V. CONCLUSIONS This paper presents powerful and effective EMTP – RV modelling. Grouping of elements into subcircuits and creating
S. Sadovic, M. Babuder, M. Hrast, D. Bokal, M. Marnisek and T. Sadovic, "Line Surge Arrester Application on 123 kV Double Circuit Line", International Symposium on High Voltage Engineering, Johannesburg, South Africa, 2009. A. Xemard, S. Sadovic, T. Sadovic, M. Mesic, A. Guerrier, "Lightning current measurement on an overhead line equipped with line arresters", CIGRE Colloquium, 2010. T. Sadovic, S. Sadovic, "EMTP – RV modelling for the transmission line lightning performance computation", User Group Meeting, 20 April 2009, Dubrovnik, 2009. IEEE Guide for Improving the Lightning Performance of Transmission Lines, IEEE Standards Department, June. 1996. CIGRE WG C4.407, "Lightning Parameters for Engineering Applications", CIGRE Technical brochure no 549, August 2013. V.A. Rakov, M.A. Uman, "Lightning Physics and Effects, Cambridge Univ. Press, Cambridge, U. K, 2003, pp. 232-240. K. Narita, Y. Goto, H. Komuro, and S. Sawada, "Bipolar lightning in winter at Maki, Japan" J. Geophys. Res., vol. 94, 13,191-13,195, 1989. V.A. Rakov, "A Review of Positive and Bipolar Lightning Discharges", American Meteorological Society, 2003.
