UPEC2010 31st Aug - 3rd Sept 2010

Effects of Simulation Models of Overhead Transmission Line Basic Components on Backflashover Surges Impinging on GIS Substations
Pantelis N. Mikropoulos1 , Thomas E. Tsovilis, Zacharias G. Datsios and Nikos C. Mavrikakis High Voltage Laboratory, School of Electrical & Computer Engineering, Faculty of Engineering, Aristotle University of Thessaloniki, Thessaloniki 541 24, Greece 1 [email protected] resistance; the latter models consider the reduction in the tower grounding resistance associated with soil ionization [19]-[25]. Flashover of insulator strings is represented by voltage-dependent switches, considering either the voltagetime characteristic of the insulator strings [26], [27] or leader progression models [28], [29]. In the present study the overvoltages arising at the entrance of 150 kV and 400 kV GIS substations due to backflashover of the incoming overhead transmission lines, normally associated with higher amplitude and steepness than shielding failure overvoltages, are evaluated with the aid of ATPEMTP simulations, by considering several simulation models of the basic components of the lines. The computed overvoltages vary significantly among tower grounding system models and among insulator string flashover models; the effect of tower simulation models is rather insignificant. The present analysis is important for insulation coordination of substations since the computed peak overvoltages are used for the evaluation of the substation outage rate as well as for the selection of the necessary protection measures. II. SIMULATION MODELS OF THE BASIC COMPONENTS OF OVERHEAD TRANSMISSION LINES A. Tower Models Transmission line towers are divided in segments between the crossarms. Each segment is represented by either a vertical lossless single-phase frequency-independent distributed parameter line or by a combination of the latter and lumped circuit elements [3]-[17] (Fig. 1). Thus, tower models can be classified into three categories.

Abstract- Overvoltages arising in 150 kV and 400 kV GIS substations due to backflashover of the incoming overhead transmission lines were computed with the aid of ATP-EMTP simulations, by considering the effects of several simulation models of the basic transmission line components. The protection offered against impinging surges by surge arresters operating at the substation entrance is evaluated with respect to the basic insulation level of the GIS system. The computed overvoltages vary considerably among tower grounding system models and among insulator string flashover models whereas rather insignificantly among tower simulation models. There is no systematic variation in computed overvoltages among insulator string flashover models. Single vertical lossless line models and a constant rather than a current dependent resistance are considered, in terms of simulation simplicity and safe design, as satisfactory for simulating transmission line tower and its grounding resistance, respectively, in insulation coordination studies of substations. Index TermsATP-EMTP, backflashover, fast-front overvoltages, GIS substations, insulation coordination, lightning, overhead transmission lines.

I. INTRODUCTION Lightning surges may impinge on substations due to either backflashover in the connected overhead transmission lines, that is, flashover of line insulation caused by lightning flash to shield wire, or shielding failure in the incoming overhead lines, that is, lightning flash to phase conductors. The impinging surges may cause substation outages, thus, consequently, interruptions in power supply, which result in economic losses and affect reliability of power systems. Thus, insulation coordination of substations necessitates the estimation of the arising backflashover and shielding failure overvoltages. This can be accomplished analytically [1] or through computer simulations [2]. For modeling of the basic components of overhead transmission lines, that is, the towers and their grounding system and flashover of insulator strings, several simulation models have been proposed in literature. Specifically, tower segments are modeled either as single-phase vertical lossless lines with the same [3]-[10] or different surge impedance (multiconductor models) [10]-[13] or as a combination of the latter and lumped circuit elements (multistory models) [14][17]. The tower grounding system is represented by models of either constant [6], [18] or current dependent grounding

Fig. 1. Tower models: (a) single vertical lossless line models [3]-[10], (b) multiconductor models [11], [12] (c) Hara et al. multiconductor model [10], (d) multistory models [14]-[16], (e) Baba & Ishii multistory model [17].

i. Single vertical lossless line tower models By following a simplified approach, towers can be represented by a simple geometric shape, such as a cylinder or a cone, with surge impedance calculated according to expressions derived theoretically or through scale model experiments (Table I). Thus, all tower segments between the crossarms are represented by using an equal surge impedance, ZT () [Fig. 1(a)]. The surge propagation velocity is assumed equal to 85% of the speed of light [6]. The surge impedance of the 150 kV and 400 kV line towers under study (Fig. 2) according to different tower models is given in Table I.

TABLE I TOWER SURGE IMPEDANCE FOR SINGLE VERTICAL LOSSLESS LINE MODELS

TABLE II SURGE IMPEDANCE ACCORDING TO MULTICONDUCTOR TOWER MODELS

Fig. 2. Typical towers of (a) 150 kV and (b) 400 kV double-circuit lines of the Hellenic transmission system; the length of insulator strings is 1.86 m and 3.62 m for the 150 kV and 400 kV line, respectively.

ii. Multiconductor tower models By representing tower segments by multiconductor vertical lossless lines [11] or, for simplicity, by single lossless lines [11], [12], the surge impedance of each segment can be calculated through expressions derived from electromagnetic field analysis. Thus, according to multiconductor models the tower segments are simulated as single vertical lossless lines with different surge impedance, ZTi [Fig. 1(b)]. Hara et al. [10] based on scale model experimental results proposed a simulation model, which takes also into account the effect of bracings and crossarms by representing the former by lossless lines, ZLi, in parallel to the lines of the tower segments, ZTi, and the latter by lossless horizontal lines, ZAi [Fig. 1(c)]. The surge propagation velocity is assumed equal to the speed of light. The surge impedance of the segments of the towers of the 150 kV and 400 kV lines (Fig. 2) calculated according to models [10]-[12] is given in Table II. iii. Multistory tower models Each tower segment is modeled as a lossless line in series with an R-L parallel circuit [Figs. 1(d) and 1(e)]. The values of the surge impedance of tower segments, ZTi, and the attenuation coefficient (Table III) are determined by a trialand-error process so as to fit measured [14]-[16] or accurately

computed [17] voltage waveforms at several points along Japanese large-scale towers. he damping resistances and inductances, Ri and Li respectively, can be estimated for models [14]-[16] [Fig. 1(d)] as

Ri =

2 Z Ti ln h1 + h2 + h3

hi , i = 1 3 and R4 = 2 Z T 4 ln (1)
ki h vT , i = 1 4 (2)

Li = Ri

where h is the tower height, h1, h2, h3 (m) are defined in Fig. 1, ki = 2, and vT is the surge propagation velocity equal to the speed of light. Baba and Ishii model [17] incorporates fixed

TABLE III SURGE IMPEDANCE AND ATTENUATION COEFFICIENT ACCORDING TO

TABLE VI CURRENT DEPENDENT TOWER GROUNDING RESISTANCE MODELS

TABLE IV DAMPING RESISTANCE AND ki ACCORDING TO BABA & ISHII [17]

TABLE V DAMPING RESISTANCE AND INDUCTANCE FOR MULTISTORY TOWER MODELS

C. Insulator String Flashover Models According to [26], [27], flashover of line insulation occurs when the voltage across the insulator strings becomes equal to flashover strength VFO(kV) determined by the voltage-time characteristic of the insulator strings. VFO(kV) is given as [27] VFO = ( 400 + 710 t 0.75 ) D (3)

values of Ri (Table IV) and inductances, Li, calculated according to (2) by using the values of ki given in Table IV. The values of the damping resistances and inductances of the towers of the 150 kV and 400 kV lines (Fig. 2), calculated according to models [14]-[17], are listed in Table V. B. Tower Grounding System Models Following a simplified approach, a concentrated tower grounding system can be represented as a constant resistance equal to the low current and low frequency grounding resistance, R0 [6], [18]. Alternatively, according to models [19]-[25], a concentrated tower grounding system can be modeled by a current dependent grounding resistance, R(I), so as to consider the reduction due to soil ionization in the grounding resistance from the initial value of R0. Soil ionization occurs when the lightning current, I, flowing through the grounding system, results in the electric field strength at the surface of the grounding electrodes to attain a value equal to soil ionization gradient, E0. The tower grounding system can be modeled in ATP-EMTP on the basis of the expressions shown in Table VI with the aid of MODELS.

where D(m) is the insulator string length and t(s) the elapsed time after lightning stroke. According to leader progression models [28], [29] line insulation flashover occurs when the insulator string length is bridged by a leader. The leader progresses when the average electric field strength in the unabridged gap becomes equal to or higher than a critical value E0. Leader progression models employ differential equations (Table VII) to compute the leader length, L, at each time instant; flashover occurs when L becomes equal to the gap length, D, that is, the insulator string length. Thus, the 150 kV and 400 kV line insulator strings were represented by voltage-dependent flashover switches with the aid of MODELS by incorporating (3) according to [26], [27] and the differential equations shown in Table VII according to leader progression models [28], [29].
TABLE VII LEADER PROGRESSION MODELS

III. OVERVOLTAGES ARISING IN GIS SUBSTATIONS ATP-EMTP simulations were performed for two 150 kV and 400 kV GIS substation configurations (Fig. 3). For both substations, the last section of 1.75 km of the incoming overhead transmission lines was represented by a sequence of J.Marti frequency dependent models, considering the line span (350 m) and the tower geometries shown in Fig. 2. The underground XLPE power cables were represented by the Bergeron model. Surge arresters were modeled as nonlinear current dependent resistors taking into account their characteristics (Table VIII). GIS bays were represented as lossless stub lines with a surge impedance of 75 [2]. The step-up transformers were represented by a capacitance picircuit together with BCTRAN model. Cable connections and the surge arrester lead lengths shorter than 3 m were modeled by a lumped parameter inductance of 1H/m [2]. The earth resistivity was assumed 200 m.

Fig. 4. Computed peak overvoltages at the entrance of the (a), (c) 150 kV and (b), (d) 400 kV GIS substations; models are numbered according to Tables I-III, vertical bars indicate the variation of peak overvoltages among tower models, red line corresponds to safety margin of BIL/1.15.

Fig. 3. Schematic diagrams of the evaluated systems (a) 150 kV and (b) 400 kV GIS substations. TABLE VIII SURGE ARRESTER CHARACTERISTICS

The effects of simulation models of towers and their grounding system and of insulator string flashover on the arising overvoltages in the GIS substations were investigated for the following worst case scenario: lightning is assumed to strike to the first tower close to the substation, at the time instant of positive power-frequency voltage peak of the upper phase of the overhead transmission line. Lightning stroke was represented by a current source of negative polarity producing a current waveshape 7/77.5 s with front upwardly concave [21], [30], [31]. The lightning current had an amplitude of 200 kA and a maximum steepness calculated according to [31]. Finally, simulations were performed with and without surge arresters operating at the substation entrance so as to evaluate the protection offered against impinging surges with respect to the basic insulation level, BIL, of the GIS system; the latter is 750 kV and 1425 kV for the 150 kV and 400 kV substations, respectively. A. Effect of Tower Models Fig. 4 shows the computed peak overvoltages arising at the entrance of the 150 kV and 400 kV GIS substations by

considering in simulations the tower models presented in Section II.A; insulator string flashover was modeled according to [27]. From Figs. 4(a) and 4(b) it is obvious that the overvoltages, obtained by using a fixed tower grounding resistance of 10 , do not vary significantly among tower models. Actually, as can be deduced for Figs. 4(c) and 4(d), the overvoltages, increasing with tower grounding resistance, vary among models within 10% over the range of tower grounding resistance of 5-20 ; generally, the differences in overvoltages among tower models are less pronounced for higher tower grounding resistance. It is noteworthy that increasing tower surge impedance within the range given in Table I results in a small increase in overvoltage peak, especially for lower tower grounding resistance; however, this is insignificant when surge arresters are operating at the substation entrance. The increase of overvoltage peak due to an increase in either tower surge impedance or grounding resistance can be explained by the corresponding increase in the instantaneous flashover voltage of line insulation. It is important that when surge arresters are operating at the substation entrance the peak overvoltages are limited to values lower than the safety margin of BIL/1.15 [32] (Fig. 4). B. Effect of Tower Grounding System Models Fig. 5 shows the computed peak overvoltages at the entrance of the substations under study for constant and current dependent tower grounding resistance by assuming a low current low frequency grounding resistance, R0, of 10 ; towers and insulator string flashover were modeled according to [6] and [27], respectively. From Fig. 5 it can be deduced that the computed overvoltages vary a little among tower grounding system models. However, when a higher value of R0 is employed in simulations (Fig. 6), differences in overvoltages among simulation models increase up to 19%. The overvoltages, increasing with R0, are highest when the tower grounding system is modeled by a constant grounding resistance equal to R0, as suggested by [6] and [18], or according to model [25]; this is more pronounced for

Fig. 5. Computed peak overvoltages at the entrance of the (a) 150 kV and (b) 400 kV GIS substations; low current and low frequency grounding resistance 10 , red line corresponds to safety margin of BIL/1.15.

Fig. 7. Computed peak overvoltages at the entrance of the 150 kV and 400 kV GIS substations; tower grounding resistance 10 ; red line corresponds to safety margin of BIL/1.15.

Fig. 8. Computed peak overvoltages at the entrance of the (a) 150 kV and (b) 400 kV GIS substations as a function of tower grounding resistance; red line corresponds to safety margin of BIL/1.15.

Fig. 6. Computed peak overvoltages at the entrance of the (a) 150 kV and (b) 400 kV GIS substations as a function of low current and low frequency grounding resistance; red line corresponds to safety margin of BIL/1.15.

relatively higher R0. However, when surge arresters are operating at the substation entrance the peak overvoltages are limited to values lower than the safety margin of BIL/1.15 [32] (Figs. 5 and 6) and do not vary significantly, less than 6%, among tower grounding system models. Summarizing, the use in simulations of a constant tower grounding resistance, yielding the highest computed overvoltages in the substations, may result in a safer design. C. Effect of Insulator String Flashover Models Fig. 7 shows the computed peak overvoltages at the entrance of the 150 kV and 400 kV GIS substations by considering in simulations the insulator string flashover models presented in Section II.C; towers were represented

according to [6] and were terminated by a constant grounding resistance of 10 . The computed overvoltages vary notably among models, up to 18%, especially for the 400 kV system. They increase with increasing tower grounding resistance (Fig. 8), especially when surge arresters are not operating at the substation entrance. It is important to note that for the 400 kV system when leader progression models are employed in simulations the overvoltages, even when surge arresters are installed, are higher than the safety margin of BIL/1.15 for tower grounding resistance higher than about 15 . This, contrary to common practice [33], indicates the necessity of using a higher insulation level for the GIS equipment, and is considered as unrealistic since care is normally taken to reduce the last tower grounding resistance to values lower than 10 . As a general result, due to the non-systematic variation in overvoltages among insulation string flashover models conscious selection of the latter is needed for insulation coordination of substations. IV. CONCLUSIONS The effect of the simulation models of the basic components of overhead transmission lines on the arising overvoltages at the entrance of 150 kV and 400 kV GIS substations due to backflashover of the incoming lines has been investigated with the aid of ATP-EMTP simulations. Tower simulation model does not affect significantly the computed overvoltages, especially with increasing tower

grounding resistance. Thus, single vertical lossless line models are considered as satisfactory for simulating transmission line towers, due to their simplicity, in insulation coordination studies of substations. Tower grounding system simulation model affects considerably the computed overvoltages, which increase with tower grounding resistance; however this is insignificant when surge arresters are operating at the substation entrance. A constant rather than current dependent resistance, resulting in higher computed overvoltages, is considered as satisfactory in terms of safer design for simulating tower grounding resistance in insulation coordination studies of substations. The computed overvoltages vary significantly however non-systematically among insulator string flashover models. Thus, conscious selection of the overhead line insulator string flashover model is needed for insulation coordination of substations. ACKNOWLEDGEMENTS T. E. Tsovilis wishes to thank the Research Committee of Aristotle University of Thessaloniki for the support provided by a merit scholarship. REFERENCES
A. R. Hileman, The incoming surge and open breaker protection, in Insulation Coordination for Power Systems. Boca Raton, FL: CRC Press, Taylor & Francis Group, New York, 1999. [2] IEEE Task Force, Modeling guidelines for fast front transients, IEEE Trans. Power Delivery, vol. 11, no. 1, pp. 493506, Jan. 1996. [3] C. A. Jordan, Lightning computations for transmission lines with overhead ground wires, Gen. Elec. Rev., vol. 37, 1934. [4] C. F. Wagner and A. R. Hileman, A new approach to calculation of lightning performance of transmission lines III A simplified method: Stroke to tower, AIEE Trans. Power App. Syst., vol. 79, pp. 589603, Oct. 1960. [5] M. A. Sargent and M. Darveniza, Tower surge impedance, IEEE Trans. Power App. Syst., vol. PAS88, no. 5, pp. 680687, May 1969. [6] IEEE Working Group, A Simplified method for estimating lightning performance of transmission lines, IEEE Trans. Power App. Syst., vol. PAS-104, no. 4, pp. 919-932, Apr. 1985. [7] C. Menemenlis and Z. T. Chun, Wave propagation on nonuniform lines, IEEE Trans. Power App. Syst., vol. PAS101, no. 4, pp 833 839, Apr. 1982. [8] W. A. Chisholm, Y. L. Chow, and K. D. Srivastava Travel time of transmission towers, IEEE Trans. Power App. Syst., vol. PAS104, no. 10, pp. 29222928, Oct. 1985. [9] IEEE Working Group on Estimating Lightning Performance of Transmission Lines, IEEE guide for improving the lightning performance of transmission lines, Proposed IEEE P1243 draft, 1996. [10] T. Hara and O. Yamamoto, Modelling of a transmission tower for lightningsurge analysis, IEE Gener. Transm. Distrib., vol. 143, no. 3, pp. 283289, May 1996. [11] A. Ametani, Y. Kasai, J. Sawada, . Mochizuki, and . Yamada, Frequencydependent impedance of vertical conductors and a multiconductor tower model, IEE Gener. Transm. Distrib., vol. 141, no. 4, pp 339345, Jul. 1994. [1]

[12] D. Rondon, M. Vargas, J. Herrera, J. Montana, H. Torres, M. Camargo, D. Jimenez, and A. Delgadillo, Influence of grounding system modelling on transient analysis of transmission lines due to direct lightning strike, in Proc. Power Tech, St. Petersburg, Russia, 2005, pp. 1-6. [13] J. A. Gutierrez, P. Moreno, L. Guardado, J. L. Naredo, Comparison of transmission tower models for evaluating lightning performance, in Proc. Power Tech, Bologna, Italy, 2003, pp. 1-6. [14] M. Ishii, T. Kawamura, T. Kouno, E. Ohsaki, K. Shiokawa, K. Murotani, and T. Higuchi, Multistory transmission tower model for lightning surge analysis, IEEE Trans. Power Delivery, vol. 6, no. 3, pp. 13271335, Jul. 1991. [15] T. Yamada, A. Mochizuki, J. Sawada, E. Zaima, T. Kawamura, A. Ametani, M. Ishii, and S. Kato, Experimental evaluation of a UHV tower model for lightning surge analysis, IEEE Trans. Power Delivery, vol. 10, no. 1, pp 393402, Jan. 1995. [16] H. Motoyama, K. Shinjo, Y. Matsumoto, and N. Itamoto, Observation and analysis of multiphase back flashover on the Okushishiku test transmission line caused by winter lightning, IEEE Trans. Power Delivery, vol. 13, no. 4, pp. 13911398, Oct. 1998. [17] Y. Baba and M. Ishii, Numerical electromagnetic field analysis on lightning surge response of tower with shield wire, IEEE Trans. Power Delivery, vol. 15, no. 3, pp 10101015, Jul. 2000. [18] A. Ametani and T. Kawamura, A method of a lightning surge analysis recommended in Japan using EMTP, IEEE Trans. Power Delivery, vol. 20, no. 2, pp. 867875, Apr. 2005. [19] E. E. Oettle, A new general estimation curve for predicting the impulse impedance of concentrated earth electrodes, IEEE Trans. Power Delivery, vol. 3, no. 4, pp 20202029, Oct. 1988. [20] W. A. Chisholm and W. Janischewskyj, Lightning surge response of ground electrodes, IEEE Trans. Power Delivery, vol. 4, no. 2, pp. 13291337, Apr. 1989. [21] CIGRE Working Group 33.01, Guide to procedures for estimating the lightning performance of transmission lines, Technical Bulletin 63, Oct. 1991. [22] K. H. Weck, The current dependence of tower footing resistance, CIGRE 33-88 (WG01), 14 IWD, 1988. [23] P. Chowdhuri, Grounding for protection against lightning, in Electromagnetic transients in power systems. Research Studies Press Ltd., John Wiley & sons inc., New York, 1996, pp. 104-113. [24] A. V. Korsuntchev, Application of the theory of similitude to the calculation of concentrated earth electrodes, Electrichestvo, no.5, pp. 31-35, May 1958. [25] Y. Yasuda, Y. Hirakawa, K. Shiraishi, and T. Hara, Sensitivity analysis on grounding models for 500kV transmission lines, Trans. IEE Japan B, vol. 121, no. 10, pp. 13861393, 2001. [26] M. Darveniza, F. Popolansky, and E. R. Whitehead Lightning protection of UHV lines, Electra, no. 41, pp. 3969, Jul. 1975. [27] IEEE Working Group, Estimating lightning performance of transmission lines II Updates to analytical models, IEEE Trans. Power Delivery, vol. 8, no. 3, pp. 1254-1267, Jul. 1993. [28] K. H. Weck, Lightning performance of substations, CIGRE SC 33, Rio de Janeiro, Brazil, 1981. [29] A. Pigini, G. Rizzi, E. Garbagnati, A. Porrino, G. Baldo, and G. Pesavento, Performance of large air gaps under lightning overvoltages: Experimental study and analysis of accuracy of predetermination methods, IEEE Trans. Power Delivery, vol. 4, no. 2, pp 13791392, Apr. 1989. [30] Lightning and Insulator Subcommittee of the T&D Committee, Parameters of lightning strokes: A review, IEEE Trans. Power Delivery, vol. 20, no. 1, pp. 346-358, Jan. 2005. [31] R. B. Anderson and A. J. Eriksson, Lightning parameters for engineering application, Electra, no. 69, pp. 65102, 1980. [32] IEC 60071-2, Insulation Coordination, 1996. [33] Joint Working Group 33/23.12, Insulation co-ordination of GIS: Return of experience, on site tests and diagnostic techniques, Electra, no. 176, pp. 67-97, Feb. 1998.