Produção Nacional
Revisado por pares
Parametric analysis of the travelling wave‐based differential protection TW87
2018; Institution of Engineering and Technology; Volume: 2018; Issue: 15 Linguagem: Inglês
10.1049/joe.2018.0230
ISSN2051-3305
AutoresFelipe V. Lopes, João Paulo G. Ribeiro, Eduardo Jorge Silva Leite, Kleber. M. Silva,
Tópico(s)HVDC Systems and Fault Protection
ResumoThe Journal of EngineeringVolume 2018, Issue 15 p. 1297-1302 The 14th International Conference on Developments in Power System Protection (DPSP 2018)Open Access Parametric analysis of the travelling wave-based differential protection TW87 Felipe V. Lopes, Corresponding Author Felipe V. Lopes felipevlopes@ene.unb.br University of Brasília (UnB), Electrical Engineering Department, Brasília-DF, BrazilSearch for more papers by this authorJoão Paulo G. Ribeiro, João Paulo G. Ribeiro University of Brasília (UnB), Electrical Engineering Department, Brasília-DF, BrazilSearch for more papers by this authorEduardo J.S. Leite, Eduardo J.S. Leite University of Brasília (UnB), Electrical Engineering Department, Brasília-DF, BrazilSearch for more papers by this authorKleber. M. Silva, Kleber. M. Silva University of Brasília (UnB), Electrical Engineering Department, Brasília-DF, BrazilSearch for more papers by this author Felipe V. Lopes, Corresponding Author Felipe V. Lopes felipevlopes@ene.unb.br University of Brasília (UnB), Electrical Engineering Department, Brasília-DF, BrazilSearch for more papers by this authorJoão Paulo G. Ribeiro, João Paulo G. Ribeiro University of Brasília (UnB), Electrical Engineering Department, Brasília-DF, BrazilSearch for more papers by this authorEduardo J.S. Leite, Eduardo J.S. Leite University of Brasília (UnB), Electrical Engineering Department, Brasília-DF, BrazilSearch for more papers by this authorKleber. M. Silva, Kleber. M. Silva University of Brasília (UnB), Electrical Engineering Department, Brasília-DF, BrazilSearch for more papers by this author First published: 23 August 2018 https://doi.org/10.1049/joe.2018.0230Citations: 1AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat Abstract Time-domain protection relays have increasingly attracted the attention of electrical power utilities. These relays speed up the tripping process of traditional protection schemes, increasing the stability margins of high-loaded power grids. Among the existing time-domain protection functions, the travelling wave (TW)-based differential protection TW87 has shown to be promising. However, as such a function is relatively new in the relay market, studies on the TW87 performance have been of interest to protection engineers. Thus, this study presents a parametric analysis of the TW87 performance. To do so, massive fault simulations are carried out, considering different fault scenarios and protection settings. From the obtained results, important issues related to the TW87 performance are addressed. 1 Introduction The growing demand for electric energy has driven utilities to study strategies to increase the stability margins of high-loaded power networks [1]. Indeed, the greater the power transmitted in a given line, the smaller the system stability margins, being the protection operation time a decisive factor during the determination of the amount of power to be transmitted in transmission grids in a safe way [2-4]. Fundamental component-based protection schemes are still the most used in electrical power systems. However, in this kind of application, the phasor estimation process results in tripping delays in the order of 0.5–1.5 power cycles due to the windowing process of the monitored signals [3]. Hence, to guarantee the system stability, protection operation delays must be taken into account when the amount of power to be transmitted in a given transmission line is defined, which is usually limited to values smaller than the line rated transmission capacity [1]. In this context, the use of time-domain protection functions (which do not require the phasor estimation process) has shown to be a good solution to speed up the protection operation [2-5], although it has been ignored for decades due to technological limitations of older protective relays. Among the existing time-domain protection functions, the travelling wave (TW)-based differential protection TW87 has shown to be promising [3]. By evaluating current TWs, it is able to identify internal faults in a few milliseconds, without loss of security for external faults [3]. Nevertheless, as the TW87 involves concepts different from those applied in traditional phasor-based differential relays, studies on the TW87 performance and sensitivity limits have been of great interest to utilities and protection engineers. In this paper, the TW87 performance is investigated from the analysis of a wide variety of fault scenarios, in which different relay setting groups are taken into account. The Alternative Transients Program (ATP) was used to carry out massive fault simulations in a 500 kV/60 Hz test power system, considering different fault features and system operating conditions. The obtained results attest the TW87 function is reliable and fast. 2 TW87 protection function The TW87 main concepts are derived from the classical theory of TWs. It considers that transmission line faults launch TWs on the monitored power system, which travel from the fault point towards the line terminals with velocities that depend on the transmission line parameters [3-5]. The shape and polarity of measured fault-induced current TWs in transmission lines have a physical reason, which depend mainly on the fault features (location, inception angle and resistance), line parameters (surge impedance and length) and polarities of current transformers (CTs) [6]. As soon as a TW reach the monitored terminal, distortions on voltage and current signals show up, which typically look like steps [3]. Therefore, the TW87 analyses these transients to extract information about the event. 2.1 Extracting TW information It is well-known that TW-based functions require high sampling frequencies to properly represent the signal high-frequency content [3]. Moreover, reliable filtering techniques are needed to estimate arrival time and amplitude of fault-induced TWs. Thereby, the TW87 analyses current signals sampled at a frequency equal to 1 MHz [6], and extracts TW information by using the differentiator-smoother filter [3], called here DS filter. To allow the analysis of TW amplitudes, the DS filter responds to ideal step changes with triangle-shaped outputs, maintaining a unity gain [3]. Even in cases of attenuated transients, when ramp changes show up in the monitored signals instead of step changes, the DS filter responds with parabola-shaped outputs, whose peak value is associated with the TW arrival time [3]. Fig. 1 illustrates the DS filter application in a current signal with fault-induced transients. In this paper, local and remote DS filter outputs are referred to as i TWL and i TWR, respectively. Fig. 1Open in figure viewerPowerPoint Using the DS filter to extract TW information 2.2 Operating and restraining signals To better explain the TW87 operating and restraining signals, Fig. 2 depicts current TWs measured via DS filter at both terminals of a given line in external and internal fault cases. In the figure, P represents the time lag between the arrival times of the first TWs at local and remote terminals (Bus L and Bus R, respectively), and τ stands for the line propagation time, which consists in the time that TWs launched at a given line terminal spend to propagate until the opposite line end, i.e. τ = L /Vp, where L is the line length and vp is the TW propagation velocity [5, 7]. From Fig. 2, it is observed that, in internal fault cases, P <τ, whereas in external fault cases, P = τ [3]. Based on that, operating and restraining current TWs called i OP and i RT are obtained using [3] (1) (2) being (3) (4) Fig. 2Open in figure viewerPowerPoint Fault-induced TWs during transmission line external and internal faults In external fault cases, as P ≃τ, i TWL (t −P)≃i TWL (t −τ)≃−i TWR (t). Hence, i OP is ideally zero, whereas i RT tends to present high values. On the other hand, in internal fault cases, the first incident TWs at both line ends have the same polarity, thereby i OP tends to be greater than i RT. For instance, considering the internal fault case illustrated in Fig. 2, since P i RT is obtained. Thus, by evaluating (1) and (2), line faults can be distinguished between internal and external events [5]. 2.3 Detection of the incident and exit TWs The TW87 starts its internal fault identification procedures as soon as the first incident TWs are detected at the line ends. Here, to filter out the TWs of interest, the DS filter outputs are compared with an adaptive threshold, whose value is set based on the DS filter results during the pre-fault period. If the threshold is exceeded, a search peak procedure in the DS filter outputs is carried out to identify the arrival times of the first incident TWs at both local and remote line ends, which are represented here by t LI and t R1, respectively. After the detection of t LI and t RI, the TW87 analyses the transient content within a search window with length 2Δτ (in this paper, Δτ = 10 μs is used) positioned at the nominal line propagation time t following the first TW detection at the opposite terminal [3]. Then, the peak value found within the search window is assumed to be the exit time of the TW that leaves the monitored transmission line, called here 'exit TW'. Such a procedure is depicted in Fig. 3, where t Lexit and t Rexit represent the instant at which the peak values of the exit TWs are detected at local and remote terminals, respectively. Fig. 3Open in figure viewerPowerPoint Using search windows to detect exit TWs 2.4 Tripping logic For the sake of security, as soon as t L1, t R1, t Lexit and t Rexit are detected, the TW87 estimates the amplitudes of the first incident TWs at local and remote terminals, called here I L and I R, respectively, and the amplitudes of i OP and i RT, represented here by I OP and I RT, respectively. To do so, first, the sample indexes NL FIRST, NR FIRSt, NL EXIT and NR EXIT associated to the time instants t LI, t RI, t Lexit and t Rexit are derived, and then, I L, I R, I OP and I RT are computed considering M samples (in this paper, M is equal to the number of samples in half of the DS filter window) around the referred four indexes using [3, 5] (5) (6) (7) (8) being (9) (10) where C is a scaling factor adjusted to maintain a unity gain of the estimated amplitudes [3]. Besides the analysis of I OP, I RT, I L and I R, to provide security during switching events at the line terminals and external fault cases, the TW87 also evaluates the per unit fault distance m 87 obtained from the classical TW-based double-ended fault location method [3] (11) where T L is τ given in a number of samples. Fig. 4 illustrates the simplified TW87 tripping logic divided into three main conditions, where K TW defines the minimum TW value to operate, SLP is the operating slope and i pickup is the TW87 pickup value. In Fig. 4, Condition 1 represents the minimum TW current level supervision logic. It enables the TW87 operation only if the analysed TWs have energy enough to operate reliably. Condition 2 stands for the restraining logic, which is the one responsible to distinguish faults between internal and external events. Finally, Condition 3 represents the fault location logic, which verifies if the fault is located within the protected line [5]. In this paper, the local and remote line terminals are defined as the two blocking regions of the TW87 scheme, thereby Condition 3 blocks the TW87 operation if m 87 is ∼0 or 1 pu. Fig. 4Open in figure viewerPowerPoint Simplified TW87 tripping logic It is worth emphasising that the TW87 function applies additional security layers to guarantee reliable operations in practical cases of faults in parallel lines, lightning strokes, among other scenarios [3]. Among these security conditions, the overcurrent supervision (OC87) and polarising voltage (VPOL) supervision deserve attention [5]. The OC87 verifies if the detected event is a fault and no other low-energy event [5]. In addition, the VPOL supervision compares the polarities of I OP and the polarising voltage (which consists of an estimate of the pre-fault voltage at the fault point), in such a way that the TW87 function restrains if both signals have different polarities [5]. By considering these security layers in conjunction to those shown in Fig. 4 (Conditions 1–3), the TW87 guarantees reliable operations for internal faults, but maintaining itself secure during external events. In this paper, both OC87 and VPOL supervisions were implemented following the instructions reported in [5], but, due to space limitations, the operation of these elements in the evaluated fault scenarios are not analysed. 3 TW87 performance evaluation 3.1 Test power system and simulations The 500 kV/60 Hz power system shown in Fig. 5 was modelled using the ATP in order to carry out the proposed parametric analysis of the TW87 performance. It consists of a transmission line 200 km long (taken as the protected line), which connects two Thévenin equivalent circuits that represent the power systems around the line. The electrical parameters of the modelled line were taken from the Brazilian power grid, and a 1 MHz sampling frequency was simulated. To evaluate the TW87 performance itself, CTs and the communication channel were intentionally modelled as ideal. However, for the simulated line, a communication latency of about 2 ms would be realistic for optical fibre-based links [8], including relay transmit, receive and processing delays. Fig. 5Open in figure viewerPowerPoint Test power system To analyse the TW87 performance under a wide variety of fault scenarios, the fault distance d, fault inception angle θ, fault resistance R f, the source-to-line impedance ratio (SIR) at the remote terminal, and the loading angle δ (angle between the voltages at buses L and R) were varied. In all cases, the fault location was varied in conjunction to θ, R f, SIR or δ one pair of variables at a time. As a result, four sets of simulation parameters were obtained, through which the effects of d, θ, R f, SIR and δ on the TW87 performance were evaluated. Details about the simulation parameters are shown in Table 1. Table 1. ATP simulation parameters Variable Parameter simulation set 1: analysing θ d, pu 0.05, 0.10,…, 0.90, 0.95 θ, ° 0, 10, 20,…, 340, 350 R f, Ω ≃ 0 δ, ° −15 SIRRO,RI 0.1 simulation set 2: analysing δ d, pu 0.05, 0.10,…, 0.90, 0.95 θ, ° 90 R f, Ω ≃ 0 δ, ° −60, −55,…, 55, 60 SIRRO,RI 0.1 simulation set 3: analysing R f d, pu 0.05, 0.10,…, 0.90, 0.95 θ, ° 90 R f, Ω 0, 20, 40,…, 980, 1000 δ, ° −15 SIRRO,RI 0.1 simulation set 4: analysing SIRRO,RI d, pu 0.05, 0.10,…, 0.90, 0.95 θ, ° 90 R f, Ω ≃ 0 δ, ° −15 SIRRO,RI 0.1, 0.2,…, 1.9, 2.0 In all simulations, the OC87 and the VPOL supervisions were set following instructions reported in [5, 7]. Nevertheless, since in this paper only faults scenarios in a single-circuit transmission line are analysed, both OC87 and VPOL were set to be sensitive enough to allow the TW87 operation if Conditions 1–3 shown in Fig. 4 are satisfied. By doing so, the TW87 performance irrespectively of the additional security layers could be evaluated, providing a better understanding of the TW87 behaviour for internal faults when d, θ, R f, SIR and δ are varied. 3.2 Parametric analysis of the TW87 variables Figs. 6-9 illustrate the behaviour of the TW87 tripping logic variables (I L, I R, |I OP |, I RT and m 87) for the simulation sets shown in Table 1. Surfaces are used to analyse the values of each TW87 variable, highlighting their relation with the cross effect between d and θ, R f, SIR and δ. Fig. 6Open in figure viewerPowerPoint Simulation set 1 (variation of θ) Fig. 7Open in figure viewerPowerPoint Simulation set 2 (variation of δ) Fig. 8Open in figure viewerPowerPoint Simulation set 3 (variation of Rf) Fig. 9Open in figure viewerPowerPoint Simulation set 4 (variation of SIR) It is observed that the fault inception angle θ, the fault resistance R f and SIR are those which most critically affect the TW87 operation among the evaluated simulation parameters. Indeed, if θ = 0° or 180°, there is no voltage step at the fault point, thereby TWs are not launched in the line [3]. Also, the greater the R f the smaller the amplitude of fault-induced TWs [6]. Finally, regarding the simulation set 4 in which different SIR values were analysed by varying the terminal impedance connected to the remote line end, it should be noted that if the terminal impedance is greater than the line surge impedance, negative current TW reflection coefficients are obtained. Thus, as the measured TWs are given by the superposition of incident and reflected wavefronts, the amplitude of measured TWs is reduced, leading |I OP |, I L and I R to present small values. In these scenarios, the TW87 may be blocked by Condition 1 of the protection tripping logic shown in Fig. 4, depending on the used settings i pickup and K TW . On the other hand, considering the simulated δ variations, there are no regions in the obtained surfaces where |I OP |, I L and I R are close to zero, in such a way that Condition 1 tends to be met in all cases, attesting that the TW87 is not significantly affected by the system load flow. In all evaluated cases, m 87 was properly estimated, resulting in 0 < m 87 SLP.I RT) indicates the occurrence of an internal fault on the protected line. Finally, one observes that the maximum operation times did not exceed 1.4 ms. Even considering a fibre-based link communication latency delay of 2 ms, which is realistic for the simulated line [8], operation times of a few milliseconds would be verified, attesting that the TW87 is fast and reliable. 4 Conclusions A parametric analysis of the TW-based differential protection function TW87 was presented. Initially, the TW87 main concepts and some important issues related to its tripping logic were addressed. Then, massive ATP fault simulations were carried out, considering different fault features and system operational conditions, such as fault inception angle, resistance and location, as well as system loading and SIR. The TW87 was implemented using four setting groups and the impacts of these settings on the TW87 performance were addressed. The obtained results reveal that the more secure the TW87 is set to be, the less sensitive it becomes. Fault resistance, fault inception angle and SIR (representing variations in the terminal impedance) showed to be more critical to the TW87 operation than the system loading. Also, it was observed that the TW87 operation time depends on the fault distance and line propagation time only, resulting in well-defined operation times. In fact, faster operations are verified for faults close to the middle of the transmission line, and the slower ones for faults close to the line terminals. Finally, considering the evaluated internal fault cases, the minimum pickup value has shown to be the most critical setting in the TW87 tripping logic, since it determines the protection sensitivity limits in cases of over-damped transients. 5 Acknowledgments The authors thank the Federal District Research Support Foundation (FAP-DF) and the Coordination for the Improvement of Higher Education Personnel (CAPES) for the financial support. 6 References 1Anderson P.M., Fouad A.A.: ' Power system control and stability' ( Wiley-IEEE Press, Piscataway, NJ 08854, 2003, 2nd edn.) 2Eastvedt R.B.: 'The need for ultra-fast fault clearing'. Third Annual Western Protective Relay Conf., Spokane, 1976 3Schweitzer E., Kasztenny B., Mynam M.: 'Performance of time-domain line protection elements on real- world faults'. 42nd Annual Western Protective Relay Conf., Spokane, 2015 4Kasztenny B., Voloh I., Udren E.: 'Rebirth of the phase comparison line protection principle'. 59th Annual Conf. for Protective Relay Engineers, Texas, 2000 5SEL T400L – Ultra-High-Speed Transmission Line Relay, Traveling Wave Fault Locator, High Resolution Event Recorder, Instruction Manual, 2017 6Greenwood A.: ' Electrical transients in power systems' ( John Wiley and Sons, New York, 1991, 2nd edn.) 7Kasztenny B., Guzmán A., Fischer N. et al.: 'Practical setting considerations for protective relays that use incremental quantities and traveling waves'. 43rd Annual Western Protective Relay Conf., Spokane, Washington, 2016 8Altuve H.J., Schweitzer E.O.: ' Modern solutions for protection, control and monitoring of electric power systems' ( Schweitzer Engineering Laboratories, Inc., Pullman, USA, 2010) Citing Literature Volume2018, Issue15October 2018Pages 1297-1302 FiguresReferencesRelatedInformation
Referência(s)