Impulse Resistance of Concentrated Tower Grounding Systems Simulated by an ATPDraw Object

Transcription

1 mpulse Resistance of Concentrated Tower Groundin Systems Simulated by an ATPDraw Object Z. G. Datsios, P. N. Mikropoulos, T. E. Tsovilis Abstract--Tower roundin system accurate modelin is very important in evaluatin the backflashover sures arisin at overhead transmission lines and impinin on the connected substations. A new ATPDraw object, called TGR, has been developed with the aid of which a concentrated tower roundin system can be represented on the basis of several tower roundin system models. The TGR object was employed in ATP-EMTP simulations of a 15 kv GS substation. The computed backflashover sures impinin on the substation vary considerably amon the tower roundin system models employed in simulations, as a result of the variability in the roundin impulse resistance. The TGR object is a useful tool within the ATP-EMTP environment for insulation co-ordination studies; the effects of tower roundin system modelin on backflashover sures arisin at overhead transmission lines and impinin on the connected substations can be easily quantified. Keywords: ATP-EMTP, ATPDraw, concentrated tower roundin system, GS substation, insulation co-ordination, lihtnin sures, MODELS, overhead transmission lines.. NTRODUCTON ACKFLASHOVER, that is line insulation flashover due B to lihtnin strokes to shield wires, is one of the main causes of transmission line outaes and may also result in substation outaes, caused by incomin sures with amplitude exceedin the insulation level of substation equipment. Backflashover is associated with the overvoltaes arisin across the transmission line insulator strins owin to the increase of the tower potential when lihtnin strokes are intercepted by shield wires. These overvoltaes are reatly determined by the transmission line tower roundin system. Hence, modelin of the latter is important for evaluatin the backflashover sures arisin at the overhead transmission line, thus also, incomin to the connected substations. Dependin on the roundin electrode dimensions, tower roundin systems can be cateorized as concentrated or extended. Most commonly, for fast-front transient studies, a concentrated tower roundin system is modeled as either a Z. G. Datsios, P. N. Mikropoulos and T. E. Tsovilis are with the Hih Voltae Laboratory, Department of Electrical Enery, School of Electrical and Computer Enineerin, Faculty of Enineerin, Aristotle University of Thessaloniki, Thessaloniki Greece. ( zdatsios@auth.r, pnm@en.auth.r, tsovilis@ee.auth.r). Paper submitted to the nternational Conference on Power Systems Transients (PST211) in Delft, the Netherlands June 14-17, 211 constant resistance [1]-[3] or a current-dependent resistance, when considerin the reduction of the tower footin resistance due to soil ionization [4]-[11]. n this study, a new ATPDraw [12] object is presented yieldin the roundin impulse resistance of a concentrated tower roundin system on the basis of several models reported in literature. The new object, called TGR, has been developed by usin MODELS lanuae [13], [14]. The TGR object has been applied to simulate the roundin system of a typical 15 kv tower of the Hellenic transmission system and to evaluate the computed backflashover sures impinin on a 15 kv GS substation with respect to the tower roundin system model adopted. The computed overvoltaes vary sinificantly in terms of both amplitude and waveshape amon the tower roundin system models employed in simulations. This results from the variability in the roundin impulse resistance amon models and has been easily quantified with the aid of the new ATPDraw object.. MODELS OF CONCENTRATED TOWER GROUNDNG SYSTEMS Tower roundin systems can be considered as concentrated when roundin electrodes with relatively small dimensions cover distances shorter than 3 m from the tower base [9], [15]. For fast-front transient studies, a concentrated tower roundin system can be modeled as a constant resistance with value equal to either the power frequency resistance [1] or 1 Ω as recommended in [2] for system voltae hiher than 77 kv or a sure-reduced constant resistance takin into account a sure reduction curve in order to consider soil ionization [3]. n addition, a concentrated tower roundin system can be modeled more accurately as a current-dependent resistance [4]-[11], considerin, thus, the decrease of the tower footin impedance to values lower than the initial low current and low frequency roundin resistance due to soil ionization. Accordin to the concentrated tower roundin system models [4]-[8], which are all based on the similarity theory, the tower footin impedance is described in terms of two dimensionless parameters, Π 1 and Π 2 as iven by (1) and (2), respectively. Several curves correlatin Π 1 and Π 2 have been proposed based on a lare number of experimental results for roundin electrodes with different dimensions and shape and for different soil conditions (Table ). n (1) and (2) R() is the current-dependent tower footin resistance in Ω, s (m) is the characteristic dimension of the

2 TABLE DMENSONLESS PARAMETERS Π 1 AND Π 2 CGRE [9] adopted Weck s simplified concentrated tower roundin system model [1]; accordin to the latter, the tower roundin system is represented by a current-dependent tower footin resistance iven as R R 1 (6) where R (Ω) is the low current and low frequency resistance, (ka) is the current flowin throuh the roundin system and (ka) is the limitin current to initiate sufficient soil ionization expressed as E 2 R 2 (7) R s 1 (1) s E 2 2 roundin electrode, which is defined as the distance from the the eometric center of the electrode on the round surface to the outermost point of the electrode, ρ (Ωm) is the soil resistivity, E (kv/m) is the critical soil ionization radient and (ka) is the current flowin throuh the roundin system. Oettle [6], to account for three-dimensional types of roundin electrodes defined the characteristic dimension as (2) s d d d (3) where d 1 (m) is larest horizontal distance of the electrode, d 2 (m) is the horizontal dimension which lies perpendicular to the larest horizontal distance and d (m) is the burial depth. Chisholm et al. model [7] incorporates the experimental curve relatin Π 1 and Π 2 from Popolansky [5] to estimate the resistive response of the roundin system, R(), and considers also the sure response of a round plane. Thus, the tower footin impedance, R f (), is iven in Ω as R ( ) R( ) L t (4) f f f where L f (μh) is the tower footin inductance, iven by (5), correspondin to the inductive component of the roundin electrode sure response and t f (μs) is the front time of the lihtnin current. n (5) T t (μs) is the tower travel time. L 6 T ln f T t (5) f t t Accordin to Yasuda et al. model [11], the currentdependent tower footin resistance, R(), is iven by (8). n the latter, the limitin current (ka) is iven by (9), where r (m) is the equivalent radius of the tower footin and n is the number of footins per tower. R R, R, 2 (8) 2 rne (9). TGR OBJECT A new ATPDraw [12] object, called Tower Groundin mpulse Resistance (TGR), has been developed by usin MODELS lanuae [13], [14] within the ATP-EMTP [16] environment. The TGR object implements the concentrated tower roundin system models detailed above by incorporatin a MODEL that controls a TACS Type 91 timedependent resistor. Fi. 1 shows the ATPDraw dialo box of the TGR object; the user enters the input data, namely values for the low current and low frequency resistance (R), soil resistivity (SR), critical soil ionization radient (E), characteristic dimension of the roundin system (S), number of footins per tower (N), equivalent radius of the tower footin (ER), front time of the lihtnin current (TF) and tower heiht (H). The user also selects the tower roundin system model to be used in simulations, by assinin a value to the parameter model selection (MS), ranin from 1 to 7 correspondin to the adopted roundin system model as numbered in Table. This information on the input parameters is also provided in the help viewer window of the object. The TGR object was applied to simulate the roundin system of a typical 15 kv tower of the Hellenic transmission system. Fi. 2 shows the tower roundin impulse resistance

3 based on several tower roundin system models with parameters as iven in Table. n these simulations the lihtnin current flowin throuh the roundin system had an amplitude of 1 ka and a waveshape of 6/77.5 μs with front upwardly concave and maximum steepness accordin to [9]. From Fi. 2(a) it can be deduced that for relatively low values of power frequency resistance the tower roundin impulse resistance varies a little amon models. However, this is not the case for relatively hih values of power frequency resistance [Fi. 2(b)] and this may affect considerably the computed backflashover sures arisin at the overhead transmission lines, thus also impinin on the connected substations. The latter is demonstrated in what follows. V. APPLCATON OF THE TGR OBJECT FOR THE EVALUATON OF BACKFLASHOVER SURGES MPNGNG ON SUBSTATONS Fi. 1. ATPDraw input dialo box of the developed TGR object. TABLE CONCENTRATED GROUNDNG SYSTEM MODELS PARAMETERS The TGR object was employed in ATP-EMTP [16] simulations for the evaluation of the backflashover sures impinin on a 15 kv GS substation (Fi. 3). Simulations were performed for the followin worst case scenario: neative lihtnin is assumed to strike to the top of the first tower (Fi. 4) close to the substation, at the time instant of positive power-frequency voltae peak of the upper phase of the overhead transmission line. Lihtnin stroke was represented by a current source producin a current with an amplitude of 2 ka and a waveshape 8/77.5 μs with front upwardly concave and maximum steepness calculated accordin to [9]. The last section of the incomin overhead transmission line, 1.75 km in lenth, was represented by a sequence of J.Marti frequencydependent models, considerin the line span (35 m) and tower eometry (Fi. 4). Towers were modeled as vertical lossless sinle-phase frequency-independent distributed parameter lines with a sure impedance of 167 Ω calculated accordin to [1], [18]. Towers were terminated with the TGR Fi. 3. Schematic diaram of the evaluated 15 kv GS substation. Fi. 2. Groundin system impulse resistance of a typical 15 kv tower of the Hellenic transmission system; lihtnin stroke current 1 ka, 6/77.5 μs, low current and low frequency roundin resistance (a) 1 Ω and (b) 2 Ω. Fi. 4. Tower of a typical 15 kv double circuit overhead line of the Hellenic transmission system and lihtnin stroke location considered in simulations.

4 object by usin a low current and low frequency resistance of 2 Ω and input values for the parameters required as shown in Table. Transmission line insulator strins, with standard lihtnin impulse withstand voltae level of 75 kv and lenth of 1.86 m, were represented by voltae-dependent flashover switches controlled by a MODEL implementin Weck s leader development model [9], [15]. The underround XLPE power cables were represented by the Bereron model with parameters calculated at 5 khz. Sure arresters were represented by the Pinceti and Giannettoni frequencydependent model [19] as shown in Fi. 5, with parameters calculated based on the sure arrester characteristics iven in Table. GS bays were represented as lossless stub lines with a sure impedance of 75 Ω [15]. The step-up transformer was represented by a capacitance pi-circuit toether with a BCTRAN model. Cable connections and the sure arrester lead lenths shorter than 3 m were modeled by a lumped parameter inductance of 1 μh/m [15]. Finally, simulations were performed with and without sure arresters operatin at the line-cable junction so as to evaluate the protection offered aainst impinin sures with respect to the basic insulation level, BL, of the GS system (75 kv), considerin also a safety factor of 1.15 [2]. Fi. 6. Overvoltae at the entrance of the 15 kv GS substation due to backflashover of the incomin line; dashed line depicts the safety marin of BL/1.15, (a) and (b) without and with sure arresters operatin at the linecable junction, respectively. Fi. 5. Frequency-dependent sure arrester model [19]; parameters calculated based on the sure arrester characteristics iven in Table. TABLE SURGE ARRESTER CHARACTERSTCS Fi. 6(a) shows the computed overvoltaes arisin at the 15 kv GS entrance usin the tower roundin impulse resistance yielded by the TGR object for several concentrated tower roundin system models, without sure arresters operatin at the line-cable junction. The overvoltae, bein dependent upon tower roundin impulse resistance, varies notably in terms of both peak and waveshape, amon the tower roundin system models implemented in the TGR object. However, from Fi. 6(b) it is obvious that this is less pronounced when sure arresters are operatin at the linecable junction accordin to common practice [21]. Fi. 7 summarizes the computed peak overvoltaes arisin at the 15 kv GS entrance, obtained usin the tower roundin impulse resistance yielded by the TGR object. t is obvious that when sure arresters are not operatin at the linecable junction, the peak overvoltae varies sinificantly within the rane of about 15% to 23% of the BL of the GS Fi. 7. Peak overvoltaes arisin at the entrance of the 15 kv GS substation due to backflashover of the incomin line, with and without sure arresters operatin at the line-cable junction; 1 p.u. = 75 kv, dashed line depicts the safety marin of BL/1.15. system (75 kv) amon tower roundin system models. This is not the case when sure arresters are operatin at the linecable junction; the computed peak overvoltae varies lesser takin values lower than the BL (<83%) of the 15 kv GS system. t must be noted that differences in peak overvoltae amon tower roundin system models are less pronounced for lower power frequency resistance [22]. Furthermore, from Fi. 7 it can be seen that representin the tower roundin system as a constant resistance with value equal to the power frequency tower footin resistance yields the hihest peak overvoltae. This results in a safer desin of the GS substation in terms of the protection measures required aainst incomin backflashover sures. Finally, in the present study, the TGR object was used in ATP-EMTP to compute the backflashover sures impinin

5 on the entrance of a GS substation. t is well known that dependin on substation layout, hiher overvoltaes may arise at other locations within GS or alon cables. These overvoltaes would be also affected by tower roundin system modellin and this effect can be easily quantified with the aid of the TGR object. V. CONCLUSONS A new ATPDraw object, called TGR, has been developed by usin MODELS lanuae. The TGR object yields the roundin impulse resistance of a concentrated tower roundin system by considerin the dimensions and power frequency resistance of the roundin system and the soil resistivity on the basis of several models reported in literature. The TGR object enables the easy quantification of the differences in tower roundin impulse resistance amon models; these differences are sinificant for relatively hih values of power frequency tower footin resistance. The TGR object was employed in ATP-EMTP simulations of a 15 kv GS substation. The computed backflashover sures impinin on the substation, bein dependent upon tower roundin impulse resistance, vary in terms of both amplitude and waveshape amon tower roundin system models. This is less pronounced when sure arresters are operatin close to the substation entrance. The TGR object is a useful tool within the ATP-EMTP environment for utilities in assessin the backflashover sures arisin at overhead transmission lines and impinin on the connected substations, as well as in selectin the necessary protection measures. The TGR object can also be used for educational purposes in hih voltae enineerin courses and it is available at V. REFERENCES [1] EEE Workin Group, A Simplified method for estimatin lihtnin performance of transmission lines, EEE Trans. Power App. Syst., vol. PAS-14, no. 4, pp , Jul [2] A. Ametani and T. Kawamura, A method of a lihtnin sure analysis recommended in Japan usin EMTP, EEE Trans. Power Del., vol. 2, no. 2, pp , Apr. 25. [3] M. Darveniza, M. A. Sarent, G. J. Limbourn, Liew Ah Choy, R. O. Caldwell, J. R. Currie, B. C. Holcombe, R. H. Stillman, and R. Frowd, Modellin for lihtnin performance calculations, EEE Trans. Power App. Syst., vol. PAS-98, no. 6, pp , Nov./Dec [4] A. V. Korsuntsev, Application of the theory of similitude to the calculation of concentrated earth electrodes, Electrichestvo, no.5, pp , May 1958 (in Russian). [5] F. Popolansky, Determination of impulse characteristics of concentrated electrodes, CGRE SC (WG 1) WD 22, [6] E. E. Oettle, A new eneral estimation curve for predictin the impulse impedance of concentrated earth electrodes, EEE Trans. Power Del., vol. 3, no. 4, pp , Oct [7] W. A. Chisholm and W. Janischewskyj, Lihtnin sure response of round electrodes, EEE Trans. Power Del., vol. 4, no. 2, pp , Apr [8] P. Chowdhuri, Groundin for protection aainst lihtnin, in Electromanetic transients in power systems. Research Studies Press Ltd., John Wiley & sons inc., New York, 1996, pp [9] CGRE Workin Group 33.1, Guide to procedures for estimatin the lihtnin performance of transmission lines, Technical Bulletin 63, Oct [1] K. H. Weck, The current dependence of tower footin resistance, CGRE (WG1), 14 WD, [11] Y. Yasuda, Y. Hirakawa, K. Shiraishi, and T. Hara, Sensitivity analysis on roundin models for 5kV transmission lines, Trans. EE Japan B, vol. 121, no. 1, pp , 21. [12] L. Prikler and H. K. Høidalen, ATPDRAW version 3.5 Users Manual, SNTEF Enery Research, Norway, 22. [13] L. Dubé, MODELS in ATP, lanuae manual, Feb [14] L. Dubé, Users uide to MODELS in ATP, Apr [15] EEE Task Force, Modelin uidelines for fast front transients, EEE Trans. Power Del., vol. 11, no. 1, pp , Jan [16] Canadian-American EMTP Users Group, ATP Rule Book, [17] A. M. Mousa, The soil ionization radient associated with dischare of hih currents into concentrated electrodes, EEE Trans. Power Del., vol. 9, no. 3, pp , Jul [18] M. A. Sarent and M. Darveniza, Tower sure impedance, EEE Trans. Power App. Syst., vol. PAS-88, no. 5, pp , May [19] P. Pinceti and M. Giannettoni, A simplified model for zinc oxide sure arresters, EEE Trans. Power Del., vol. 14, no. 2, pp , Apr [2] EC 671-2, nsulation co-ordination Part 2: Application uide, [21] Joint Workin Group 33/23.12, nsulation co-ordination of GS: Return of experience, on site tests and dianostic techniques, Electra, no. 176, pp , Feb [22] P. N. Mikropoulos, T. E. Tsovilis, Z. G. Datsios, and N. C. Mavrikakis, Effects of simulation models of overhead transmission line basic components on backflashover sures impinin on GS substations, in Proc. 45th UPEC, Cardiff, Wales, 21, paper no. 72.