The Use of a Special Grounding Arrangement to Improve the Lightning Performance of Transmission Line

Transcription

1 1 The Use of a Special Grounding Arrangement to Improve the Lightning Performance of Transmission Line Alexander B. Lima, José Osvaldo S. Paulino, Wallace C. Boaventura Abstract -- This paper presents a new technique aimed at lightning performance of transmission line improvements. It is based on increasing the grounding capability to inject current into the soil. As a consequence, the grounding impedance decreases leading to a lower overvoltage on insulator strings. Overvoltages developed across insulator strings of existing 138- kv line due to direct strikes to the towers are obtained using PSPICE simulation program. The grounding is modeled as an impedance, not as a resistance as frequently occurs. The effectiveness of the technique is evaluated by verifying the occurrence of flashover on the insulation. The results show that the proposed technique is quite appropriate when it is applied on towers installed in regions of unfavorable soil. Also, it is applicable in a wide range of values for soil resistivity becoming evident its high potential in solve flashover problems. transmission lines. A. Standard Settings II. MODELING METHOD In order to obtain realistic results, the evaluations are developed considering an existing 138-kV transmission line structure shown in Fig. 1, when it is struck by lightning. The incident lightning current is approached by a triangular wave (2/50s) with 50-kA peak current. The peak current and crest time have been properly chosen to represent a severe condition associated with the first return-stroke current pulse. Index Terms Transmission lines, Backflashover, Lightning, Grounding. U I. INTRODUCTION NPLANNED transmission line outage is an important parameter in defining its performance and the lightning is the main reason of such occurrences [1], [2]. Thus, some devices and techniques are adopted to minimize this unintended occurrence. A widely used technique is the use of shielding wires to avoid direct lightning incidence on the phase cables. However, the voltage developed along the tower can be high enough to cause insulation disruption through backflashover mechanism even if the lightning is captured by this protection. Low values for towers grounding impedance must be guaranteed to increase the effectiveness of this technique as indicated in specialized literature [3], [4], [5], [6]. However, it sometimes is not achieved, particularly when the tower is installed at places of unfavorable soil conditions exhibiting very high resistivity values. In this perspective, this paper addresses an unconventional technique for towerfooting grounding design. The significant reduction in grounding impedance achieved with this technique may lead to potential improvements of the lightning performance of This work was supported by CNPq (Brazilian Eletricity National Council for Scientific and Technological Development) and FAPEMIG (Research Foundation of the State of Minas Gerais). A. B. Lima (alexlima@cpdee.ufmg.br), J. O. S. Paulino (josvaldo@cpdee.ufmg.br), W. C. Boaventura (wventura@cpdee. ufmg.br) are with Department of Electrical Engineering of Federal University of Minas Gerais, Belo Horizonte, MG , Brazil. Fig. 1. Grounding system arrangement The length of insulator string is defined by the number of insulator elements. Each one is comprised by 9 elements. The insulator elements are m each. Thus, the upper and lower phase conductors are m and m high, respectively. The phase conductor radius is 1.15-cm and the shield wire radius is 0.5 cm. The grounding arrangement comprises four-leg radial counterpoise with varying lengths depending on soil resistivity, buried in the ground at a depth of 0.5-m and the

2 2 radius is 0.5-cm. The pairs of electrodes 1-2 e 3-4 are 35-m spaced and these electrodes depart from the base of the tower. The considered setting is used to obtain reference values to be compared with those obtained from the new technique presented in this paper. B. Describing the Technique The grounding impedance is characterized by the opposition of current input into the soil and the spreading into it as well. Also, increasing soil resistivity leads to higher impedance. This problem is usually solved by increasing the length of the counterpoises. However, as it is well known, only a certain electrode length is effective in controlling the grounding impedance, which is referred as effective length, L eff [7], [8], [9]. The proposed technique overcomes this limitation by increasing the length of the counterpoises in a different way. According to [10], additional counterpoises are placed at a distance from the tower and after conventional grounding electrodes. This new grounding arrangement is connected to the tower by means of steel cables sustained by 6-m height poles, as show Fig. 2. This technique is particularly important and becomes very attractive in those situations where the effective length of the electrodes has been reached, but the grounding impedance is not yet sufficiently low to avoid backflashover. computes the resistance, inductance, and capacitance matrices of an arbitrary arrangement of conductors of an overhead transmission line. The frequency of 500-kHz was considered in calculation procedures. The schematic diagram used in the simulations is shown in Fig. 3. The transmission line theory is used to modeling the tower. Moreover, the tower is divided into a certain number of segments, depending on the height of the conductors, where a, b, c, and d are tower parts, as show the schematic diagram in Fig. 3. Each part of it is represented by a lossless distributed parameter model. Thus, it is only necessary two input parameters: time delay, TD, and characteristic impedance, Z 0. In the first case, the propagation surge velocity is considered as 90% of the speed of light. The travelling time or the time delay in each tower segment is shown in Table I. TABLE I TRAVELLING TIME ALONG THE TOWER Tower segment Segment length Time delay (ns) a b c d Approximate tower impedance is calculated by (1) [3]. 2 h r Zt 30ln 2 r 2 2 where, h is the tower height and r is half the width of the tower base. Thus, the calculated tower impedance is 163. The e and f segments are related to cables that connect the tower to additional grounding, referred as auxiliary cables in this paper. These one are also modeled by a lossless distributed parameter model and the LC per-unit-length parameters are calculated by [11]. (1) L cable 0 2 hc ln 2 a (2) Fig. 2. Proposed grounding arrangement The efficiency of the proposed grounding arrangement is assessed by systematic simulations in PSPICE computing simulation environment. C. Overhead line, tower and connecting cables The overhead line is represented by multi-phase model considering the distributed nature of the line parameters due to the range of frequencies involved. The mutual effects between all phase conductors and shield wire are considered, whereas this is not in relation to the cables that connect the tower to additional groundings. All mutual and self-parameters are obtained from power_lineparam function available in MatLab. This function C cable hc ln a where, able and C cable are per-unit-length inductance and capacitance, respectively, of the auxiliary cables, h c is the cable height, a is the cable radius, 0 is the vacuum permittivity and 0 is the vacuum permeability. Thus, the calculated characteristic impedance of the cable is 467-Ω. The TD parameter is, as in the preceding case, calculated for 90% of the speed of light. It is worth mentioning that the auxiliary cable length depends on conventional grounding electrode length and that e and f-parts are not present in the simulations that only consider the conventional grounding. (3)

3 j j LG' j C ' (6) Z j Z 0 coth. (7) where, Z 0 is the characteristic impedance, is the propagation constant, Z is the input impedance of the line, is the angular frequency, l is the line length and j is 1. It is worth noting that R, L and C are per-unit-length parameters, hence L' L/, C' C/ and G' 1/ ( R ). The electrical parameters, RLC, are calculated using a set of equations [15]: 3 R 2 ln 1 2rh (8a) C (8b) R Fig. 3. Schematic diagram used in the simulations D. Tower-footing Impedance As previously stated, the grounding impedance, Z g, has a fundamental role in backflashover process. Basically, this impedance is related with the amplitude of lightning overvoltages developed across insulator strings due to direct strikes to the tower or to the shield wires at tower vicinities, with this amplitude diminishing with decreasing impedance [12]. A very simple model to calculate the grounding impedance of the conventional arrangement shown in Fig. 1 is presented in detail in [13]. This model considers the lightning current passing through the counterpoises including the mutual couplings and evaluations are based on transmission lines theory. Thus, the grounding is modeled as an impedance rather than a resistance and it is defined as: Z s Z Z g m (4) 4 where, Z g is the grounding impedance, Z s is the self-impedance of one electrode, whatever it is, and Z m is the mutual impedance between a par of electrodes. Each of the impedances, Z s e Z m, is calculated using the transmission line theory, as shown in Fig L ln 1 2 r (8c) where, is the resistivity, the electric permittivity and the magnetic permeability of the soil. With respect to the electrode l is the length, r is the radius and h is the depth which is buried. The line parameters calculation for modeling the mutual coupling is done by replacing in (8) the radius r by distance d between the electrodes and the depth h by average depth of the electrodes [15]. Although simple, this approach requires two models to define the impedance of the grounding, which could become a limitation when it intends to include other parts of the system in a more complex simulation. So, it is desirable that the self and mutual parameters might be represented by only one model in order to use it in a SPICE-based program and in a more complex system. Based on electromagnetic theory, Sunde demonstrates that n equally spaced parallel grounding electrodes with the same geometry may be represented as only one electrode with an equivalent radius, r eq. This can be used in a proximate way to model each pair of grounding electrodes as show Fig. 5a. Thus, the conventional grounding might be represented by two electrodes with equivalent radius. It is worth noting that these one are in a parallel circuit, as show Fig. 5b. Z s (j) l + Z m (j) l Z g = (Z s + Z m ) / 4 Fig. 4. Self and mutual impedances calculated individually by two transmission lines models and the combination of both defining the grounding impedance, Z g. The transient problem is first solved by a formulation in the frequency domain given by [14], [11]: l Z 0 j jl' G ' jc ' (5)

4 4 2r 2r d d 2 r eq (a) 2r d Fig. 6 - Voltage developed in the grounding current input point. I I / 2 I / 2 Fig. 5 Parallel electrodes modeled by one electrode with equivalent radius (a) and the grounding arrangement modeled by two electrodes with equivalent radius in a parallel circuit (b). As before, RLC parameters are calculated by (8) and L' L/, C' C/, G' 1/ ( R ), but considering the equivalent radius that is defined by [15]: r eq (b) d n r 2 1 n1 n where, r is the electrodes radius, d is the distance between them and n is the number of electrodes. The results obtained using the model proposed in [13] based on transmission line theory (TLT) are in a good agreement with those obtained using the electromagnetic theory (EMT) approach. Hence, as a straightforward verification, the TLT model is used to verify the closeness in results obtained from the model that considers the equivalent radius (ER). It is considered the configuration presented in Fig. 5, comprised by two 50-m long electrodes, 5-mm radius, 0.5-m buried and spaced by 35-m. The soil resistivity is 2400m and the relative electric permittivity is 10. In this case, the calculated equivalent radius is mm. Fig. 6 shows the voltage developed in the grounding input point when subjected to a current surge (double exponential 1.2/20s). It becomes clear that both models produce very similar results. Hence, the ER model will be used in this paper in order to evaluate the new proposed grounding arrangement performance. 2 r eq (9) III. RESULTS Initially, it is worth mentioning that the following assessments do not focus on the absolute values of the overvoltages but on the relative decrease of this quantity and even more important, this paper mainly concerns the applicability of the proposed technique. In order to establish references to define the improvement achieved using the proposed unconventional technique, the overvoltage developed under insulator strings considering the conventional grounding of Fig. 1 is initially obtained. The results are compared with the limits for insulation supportability. For comparative purposes only, the voltage x time curve is used to establish the performance of the proposed method. It is define by [3], [4] 0.7 w v( t) 0.4 w (10) t 0.65 where, w is the insulator string length and t is the time. This technique is virtually feasible when effective length of the electrodes has been reached. Thus, for conventional grounding the actual length of the electrode,, is slightly smaller than effective length in all situations and the maximum length considered is 90m, as show Table II. These values were obtained using the model presented in [13]. TABLE II EFFECTIVE LENGTH AS A FUNCTION OF SOIL RESISTIVITY. (.m) Fig. 7 shows the resulting overvoltage waves across the upper and lower insulator strings due to a strike to the tower top with all of the assumptions indicated in the previous section. The curves indicate that surge voltages are approximately the same regardless of the insulator string position. However, the voltages are slightly higher in the lower insulators. Also, it is observed that there exists a soil resistivity value from which flashover occurs.

5 observing that when soil resistivity is 2400.m and 3200.m the developed voltage is considerably lower than that necessary for the absence of the flashover. Thus, electrodes in the additional grounding may have variable lengths. 5 Fig. 9. Overvoltage across lower insulator strings. Fig. 10 shows the overvoltage developed across lower insulator strings considering a fixed value of 90-m long to conventional grounding electrodes. Electrode lengths of the additional grounding vary depending on condition for flashover occurrence as indicated in Table III. In all situations the flashover problem was solved. Fig. 7. Overvoltage across (a) upper and (b) lower insulator strings for different soil resistivity values (2/50-s triangular current wave). Fig. 8 shows that towers installed at places that exhibit soil resistivity above about 1700.m, would be subject to the occurrence of flashover. Fig Overvoltage across lower insulator strings varying the length of the additional grounding electrodes. TABLE III ELECTRODE LENGTHS OF THE ADDITIONAL GROUNDINGS. Fig. 8 - Overvoltage across upper and lower insulator string, = 1700.m. Fig. 9 shows the effect of using the technique proposed in this paper. In this case, the additional grounding electrodes have the same length of the conventional grounding which approximately corresponds to their effective length. As might be observed, the towers installed on soils with resistivity ranging from 2400.m to 5600.m no longer would disrupt the insulation. It is worth noting that it is not necessary using additional grounding electrodes with effective length in all situations because it could be a waste. In Fig. 9, it becomes clear (.m) additional electrode length IV. FINAL CONSIDERATIONS AND CONCLUSIONS In this paper a method to improve the lightning performance of transmission line is proposed. Assessments performed in previous section considered a limit length of

6 6 90-m for conventional electrodes since this is a common practice adopted among the energy companies in Brazil. However, this length is not sufficient to diminish the grounding impedance to ensure non-occurrence of flashover in some situations. On the other hand, increasing the electrode beyond the effective length does not provide any benefit, besides being a waste. In this paper a new technique was proposed to overcome these limitations. However it is important to note that the results indicate the existence of a range for soil resistivity value in which this technique is technically feasible. It was found values ranging from 1700.m up to 5600.m above which the counterpoise lengths exceed 90-m. This range of values may represent the variability of values for soil resistivity normally existing in different locations. So, this kind of solution is particularly interesting in regions that exhibit high soil resistivity similar to that of Minas Gerais which has apparent value of 2400.m. It is also possible to associate this range with the variations of soil resistivity due to alteration of soil moisture throughout the year seasons, particularly in acid soils, although additional studies are necessary to determine this condition in a more precise way. For more reliable results it is prudent to observe the need of including some characteristics in the simulated system, notably a more adequate model in defining the occurrence of flashover. However, in spite of this conditions, the results reveal that the proposed technique exhibit high potential in solve flashover problems due to lightings. V. REFERENCES [1] V. Martinez, E. Uzcatagui e P. Jimenez, Study of Lightning Overvoltages in Valcor-Guanta II 230kV Overhead Transmission Line, Transmission & Distribution Conference and Exposition: Latin America, TDC '06. IEEE/PES, pp. 1-5, Aug [2] B. Shen, D. Koval, W. Xu, J. Salmon e S. Shen, An analysis of extremeweather-related transmission line outages, Electrical and Computer Engineering, IEEE Canadian Conference on, pp , May [3] General Electric Company, General Electric Company. Project UHV, Electric Power Research Institute, Transmission line reference book, 345 kv and above, Electric Power Research Institute, [4] A. R. Hileman, Insulation coordination for power systems, New York: Marcel Dekker, Inc, [5] IEEE Guide for Improving the Lightning Performance of Transmission Lines, IEEE Std , p. 0_1, [6] Work Group on Lightning Performance of Overhead Lines, IEEE Power & Energy Society. [7] Y. Liu, M. Zitnik e R. Thottappillil, An improved transmission-line model of grounding system, IEEE Trans. Electromagn. Compatibility, vol. 43, n. 3, pp , Aug [8] M. Lorentzou, N. Hatziargyriou e B. Papadias, Time domain analysis of grounding electrodes impulse response, IEEE Transactions on Power Delivery, vol. 18, n. 2, pp , abr [9] L. Grcev, Impulse efficiency of simple grounding electrode arrangements, Electromagnetic Compatibility, EMC Zurich th International Zurich Symposium on, pp , Set [10] J. O. S. Paulino, I. J. S. Lopes, E. N. Cardoso, A. B. Lima, T. C. Dias, J. R. S. Machado e M. A. M. Rennó, Melhoria do desempenho de linhas de transmis-são de alta tensão através da utilização de malhas de aterramento de baixo valor de impedância, VI Congresso de Inovação Tecnológica em Energia Elétrica (VI CITENEL), Ago [11] C. Paul, Analysis of Multiconductor Transmission Lines, New York, [12] W. Chisholm, Y. Chow e K. Srivastava, Lighting Surge Response Of Transmission Towers, Power Apparatus and Systems, IEEE Transactions on, vol. PAS_102, n. 9, pp , Sept [13] A. Lima, J. Paulino, I. Lopes e T. Dias, Modelo para Malhas de Aterramento de Torres de Linhas de Transmissão Submetidas a Descargas Atmosféricas, IEEE Power and Energy Society - T&D 2010 Latin America, 8-10 Nov [14] J. D. KRAUS e D. Fleisch, Electromagnetics, 5a ed., McGraw Hill Higher Education, [15] E. D. Sunde, Earth conduction effects in transmission systems, New York: Dover Publications, VI. BIOGRAPHIES Alexander Barros Lima received the B. Sc. Degree in Electronics and Telecommunication Engineering from the Pontifical Catholic University of Minas Gerais in 2007and the M. Sc. Degree in 2010 from the Federal University of Minas Gerais, where he is currently a Dr. Sc. student. His current research interests include lightning protection, electromagnetic transients and electromagnetic compatibility. José Osvaldo Saldanha Paulino received the B.Sc. and M.Sc. degrees from the Federal University of Minas Gerais, Belo Horizonte, Brazil, in 1979 and 1985, respectively, and the Dr. Sc. degree from the State University of Campinas, Campinas, Brazil, in 1993, all in electrical engineering. In 1980, he joined the Department of Electrical Engineering, Federal University of MinasGerais, as a Professor. His current research interests include high voltage and electromagnetic compatibility. Wallace do Couto Boaventura (M 94) was born in Brazil in He received the B.Sc. and M.Sc. degrees in electrical engineering from the Federal University of Minas Gerais (UFMG), Belo Horizonte, Brazil, in 1988 and 1990, respectively, and the Dr. Sc. degree in electrical engineering from the State University of Campinas, Campinas, Brazil, in Since 1992, he has been with the Department of Electrical Engineering, UFMG. His current research interests include electromagnetic compatibility and signal processing applications in power systems.