METHOD FOR IDENTIFYING FAULTS IN TRANSMISSION SYSTEMS

Abstract

METHOD FOR IDENTIFYING FAULTS IN TRANSMISSION SYSTEMS The present invention describes a method capable of identifying (or selecting) the various types of faults that can occur in a three-phase transmission system. The method falls within the field of Electrical Engineering, especially in the transmission of electrical energy and related matters. The identification of the fault type is used in distance protection, single-pole reclosing or three-pole reclosing. The method has application from short lines to very long lines, especially those called "half-wavelength" lines, with lengths greater than 2500 km or very long compensated lines that behave as half-wavelength lines.

Description

1 / 9 METHOD FOR IDENTIFYING FAULTS IN TRANSMISSION SYSTEMS Field of invention [1] The present invention describes a method capable of identifying (or selecting) the various types of faults that may occur in a three-phase transmission system. The use of the method is a further object. The method has application in the field of Electrical Engineering, especially in the transmission of electrical energy and related matters. [2] Fault type identification is used in distance protection, single-pole reclosing or three-pole reclosing. The method is applicable from short lines to very long lines, especially those called half-wavelength lines, with lengths greater than 2500 km or very long compensated lines that behave as half-wavelength lines, where traditional methods do not correctly identify the fault type. FUNDAMENTALS OF THE INVENTION [3] In power transmission lines, the ability to identify and clear a fault is essential for their reliability. There are several types of faults that can occur, such as single-phase, two-phase without ground, three-phase without ground, two-phase with ground, and three-phase with ground. Identifying the type of fault is necessary for distance protection to function correctly. Furthermore, each type of fault requires its own reclosing strategy, so there are several methods that address the problem of fault type identification. [4] One of the most widely used algorithms is based on the relationship between negative and zero sequence currents, specifically between the angles of these currents. These methods are quite satisfactory in identifying faults in short lines, such as those from a few kilometers up to 600 km. However, in very long lines such as half-wavelength lines, or in lines longer than 600 km, these traditional methods used in commercial relays do not perform satisfactorily, because their variables exhibit non-monotonic behavior, making it impossible to identify the type of fault. [5] Document US8791704, for example, uses the difference in the angles of the symmetrical components, while document US8045307 describes a method that uses the variation of voltages and currents. Document US8791704, in addition to using the difference (under fault conditions) of the angles of the negative sequence currents and 2 / 9 zero also uses the variation of line currents to identify two-phase faults with ground. For this reason, neither of them applies to half-wavelength lines. [6] Benmouyal's work (2004), in turn, uses three quantities defined as the inner product of line voltage variations by line current variations, however, its application in half-wavelength lines has not been verified. Samantaray (2009) uses voltage and current measurements under fault conditions, but focuses only on compensated lines, whose electrical length is reduced. [7] Document US8045307 uses current variations or voltage variations depending on which variation is greater, and its application to half-wavelength lines has not been verified. Document US8717725 is based on monitoring the magnitudes and phases of the currents. Especially when faults occur in the central region of very large lines, the voltages and currents at the transmitter terminal hardly change, either in magnitude and / or phase. The problem persists if we consider document US20120095707. [8] In contrast, document US8655608 describes a method that uses current measurements from a remote terminal that are sent to the reference terminal and with the measurements from both the type of fault is determined. It should be noted that, on very long lines, identifying the type of fault would be very time-consuming, in addition to requiring measurements at both terminals of the line. [9] Even document US6405184, which uses a more differentiated proposal in the selection of the type of fault in transmission lines, using neural networks, is unable to deal with the observed problem, since the behavior of a half-wavelength line is quite different from conventional lines.

[10] Thus, it can be concluded that no prior art technology sufficiently solves the problem of fault type selection in half-wavelength lines, which supports the impact of the present patent application in this application. The invention developed is robust enough to be applied from short lines to very long lines, and with or without compensation, as it is based on variations in apparent impedances and powers from current and voltage measured instantaneously, before and during the fault condition. 3 / 9 BRIEF DESCRIPTION OF THE INVENTION

[11] The present invention describes a method capable of identifying (or selecting) the various types of faults that can occur in a three-phase transmission system.

[12] The method for identifying the type of fault in a transmission system comprises the following steps: (A) Initialize the phasor variables of voltage [y / re, Vbpre, / cpre], current 1Γ.rÊ] and the complex impedance variable [Zpre, Zpr\ Zpre.Zparbe, Z^e, zp™] before the fault (pre-fault); (B) Receive the voltages and currents in the voltage and current phasor variables [Va,Vb, Vc, Ia>Ib, / c]; (C) Calculate the zero-sequence, positive-sequence, and negative-sequence currents [ / 0, and / 2]; (D) Calculate the parameters Λ01 and R21 that relate the zero and negative sequence currents to the positive sequence current, according to equations 1 and 2; (E) Test if R014- R21> Umbralp. - if the result is 0.15, there is an asymmetrical fault and it proceeds to stage (I); - if the result is not 0.15, there is a symmetrical fault or normal operating condition and it proceeds to step (F); (F) Detect if there is an occurrence of symmetrical (three-phase) faults and select type ABC or ABCG using traditional methods; (G) Update and store the pre-fault voltage and current variable values ​​with the current voltage and current values, according to equation 3 if there is no fault (normal operating condition); (H) Calculate, update, and store the pre-fault impedance values ​​using equation 4, and return to step (B) for reading the input variables; (I) Store voltage and current values ​​under fault conditions if a fault condition is identified with equation 5; (J) Calculate the fault impedances according to equation 6; (K) Calculate the variables of variation according to equation 7; (L) Test if R01> Umbraí2: - If the result is 0.1, it indicates an unbalanced foul. 4 / 9 involving the earth, selected from the type group AG, BG, CG, ABG, B-CG, CAG, and proceeds to stage (M); - if the result is not 0.1, there is a fault that does not involve ground, and it proceeds to stage (O); (M)Test the variation of impedances ^a' %b> %c>Z^b,Zb^,Z^^ p^ra a Tolerance = 5 %: • If there was a change in Zae, there was no change in Zbc, that is, ΔZα> Tol.e ÂZbc< ToL, then the fault is AG. • If there was a change in Zbe there was no change in Zca, that is, AZb > Tol. and ΔZ^ < Tol., then the fault is BG. • If there was a change in Zce there was no change in Zab, that is, AZC > Tol. and AZab < Tol., then the fault is CG. (N) Test the variation of impedances Za,Zb,Zc,Zab,Zbc,Zcapara a Tolerance - 35%: • If there was a change in Zabe, there was no change in Zc, that is, ΔZα1}> Tol. and AZC< Tol., then the fault is ABG. • If there was a change in Zbce there was no change in Za, that is, AZbc > Tol. and ΔZα < Tol., then the deficiency is BCG. • If there was a change in Zcae and no change in Zb, that is, ΔZεα> Tol. and ΔZ0< Tol., then the fault is CAG. (O) Test the variation of impedances Za,Zb,Zc,Zab,Zbc,Zcapara a T tolerance = 5%: • If there was a change in Zabe there was no change in Zc, that is, ΔZα1}> Tol. and ΔZ€< Tol., then the fault is AB. • If there was a change in Zbce there was no change in Za, that is, ΔZ^ > Tol. and ΔZα < Tol., then the fault is BC. • If there was a change in Zcae and no change in Zb, that is, ΔZ^ > Tol. and ΔZ^ < Tol., then the fault is CA.

[13] An additional object is the use of the method to be applied in identifying the type of fault in power transmission lines of any length, preferably very long lines such as half-wavelength lines. 5 / 9 BRIEF DESCRIPTION OF THE FIGURES Figure 1 - Flowchart of the fault selection method for half-wavelength lines. Figure 2 - Graph showing AG faults along the 2600 km line. Figure 3 - Graph showing BCG faults along the 2600 km line. Figure 4 - Graph showing BC faults along the 2600 km line. DETAILED DESCRIPTION OF THE INVENTION

[14] The present invention is a method capable of identifying (or selecting) the various types of faults that can occur in a three-phase transmission system. The proposed method has as input variables the phase-to-ground voltage and current phasors Vb> Vo la> and as output the type of fault, which can be: AG, BG, CG, A- B, BC, CA, ABG, BCG, CAG. The method is illustrated by the flowchart shown in Figure 1 and can be described by the following steps: (A) Initialize the voltage phasor variables [I4pre, Vbre, Vpre], the current Ua™' and the complex impedance variable [Ζ^,Ζ^,Ζ^,Ζ^,Ζ^,Ζ^] before the fault (pre-fault); (B) Receive the voltages and currents in the voltage and current phasor variables [Va,Vb,VcJaJb>kl; (C) Calculate the zero-sequence, positive-sequence, and negative-sequence currents [ / 0, 4, and / 2]; (D) Calculate the parameters β01 and R21 that relate the zero-sequence and negative-sequence currents to the positive-sequence current, according to the following equations: 01I / 1 I (Equation 1) R21iqi (Equation 2) (E) Test if ^01+ R21> Threshold^ which in this case was set to 0.15, as this value ensures that there is a significant imbalance in the transmission line, indicating that there is an asymmetrical fault condition, and the method proceeds to step (I). If it is not met, it means that there is a 6 / 9 symmetrical fault or normal operating condition, and the method should proceed to step (F). (F) Detect whether a symmetrical (three-phase) fault occurs and what type it is, which can be ABC or ABCG, using traditional methods. Traditional methods can be used to identify three-phase faults, involving or not the ground, for lines longer than 600 km, including half-wavelength lines. (G) If there is no fault (normal operating condition), update and store the pre-fault voltage and current variable values ​​with the current voltage and current values, according to equation 3; vapre= va, Vbpre= ypre = vl?re= / 'aJa jpre _ i - 'b lvre= / 'c 'c (Equation 3) (H) Calculate, update and store the pre-fault impedance values, using equation 4, and return to step (B) of reading the input variables; ypre _ va ipre ' a vPre ypre _ b ^“b .pre ' b vPre 7 pre _ C jpre ' *c 7pre ^ab 7pre ^bc 7pre ^ca \fPre — ab .pre 'ab ypre _ bc (Equation 4) — .pre 'bc ypre _ rca “ jPre fca Where ι / pre _ ..pre __ ..pre vab ~ va vb ' iPre 'ab _ iP^ _ 'a rPre 'b uPre _ i / P™ _ vpre vbc vb vc > .pre 'bc — fPre _ — 'b / Pre 'c vpre = Vpre — l / ^re kca Yc Ya * iPre lca _ jpre _ ipre (Equation 4a) (I) Once a fault condition is identified, the values ​​of the fault voltages and currents must be stored as indicated in equation 5; 7 / 9 yfaita = V va va> .falta _ r 'a ~ b yfalta _ ,, Vb “ Vb> .falta _ r 'b - hv / alta = vc, .falta _ , B.C b (Equation 5) (J) Calculate the fault impedances according to equation 6; ..missing}7missing ymissing _ 7 missing £ab _ Vab jmissing ! 'a .missing ab , missing ..missing ymissing _ Vb y missing £bc _ Vbc .missing ' b .missing bc vmissing vc ..missing f ci l ta __ ymissing ^ca _ *ca rmissing ' j missing 'ca (Equation 6) (K) Calculate the variation variables according to equation 7; ΔΖα= Z^alta— Z^re, AZab AZb = ZÍalta- Z^, AZbc AZC= z[a{ta- Zfe, AZca_ yfalta _ yfalta “ ^bc _ yfalta ^ca _ 7Vre^ab _ 7^re^ca (Equation 7) (L) Test if R01> Threshold2, in this case it was defined as 0.1, since this value guarantees a significant zero sequence component exists in the system, which indicates that there is an unbalanced fault involving the earth, which may be of type AG, BG, CG, ABG, BCG, CAG, and the method proceeds to step (M). (M)Test the variation of impedances Za,Zb,Zc,Zab,Zbc,Zcapara a Tolerance = 5%. 5% was chosen because in tests carried out with different systems it was found that this value is sufficient to guarantee that there was variation in the impedances for phase-to-ground faults. • If there was a change in Zae, there was no change in Zbc, that is, ΔZα > Tol. and AZbc < Tol., then the fault is AG. 8 / 9 • If there was a change in Zbe there was no change in Zca, that is, AZb > Tol. and AZca < Tol., then the fault is BG. • If there was a change in Zce there was no change in Zab, that is, AZC > Tol. and AZab < Tol., then the fault is CG. (N) Test the variation of impedances Za,Zb,Zc,Zab,Zbc,Zcapara a Tolerance = 35%. 35% was chosen because in tests carried out with different systems it was found that this value is adequate to ensure that there was variation in the impedances for phase-phase-to-ground faults. • If there was a change in Zabe, there was no change in Zc, that is, AZab > Tol. and AZC < Tol., then the fault is ABG. • If there was a change in Zbce there was no change in Za, that is, AZbc > Tol. and ΔZα < Tol., then the deficiency is BCG. • If there was a change in Zcae and no change in Zb, that is, AZca > Tol. and AZb < Tol., then the fault is CAG. (0) Test the variation of impedances Za,Zb,Zc,Zab,Zbc,Zcapara a Tolerance = 5%. 5% was chosen because in tests carried out with different systems it was found that this value is adequate to ensure that there was variation in the impedances for phase-to-phase faults. • If there was a change in Zabe, there was no change in Zc, that is, AZab > Tol. and AZC < Tol., then the fault is AB. • If there was a change in Zbce and no change in Za, that is, AZbc > Tol. and ΔZα < Tol., then the fault is BC. • If there was a change in Zcae and no change in Zb, that is, AZca > Tol. and AZb < Tol., then the fault is CA.

[15] A general matching scheme can be summarized by the table below, according to the checks performed. All possible fault combinations were covered by the proposed method, there being no other combination that can identify the faults. Table 1 - Correspondences of fault types according to the checks performed for the different variables involved in the method above. Type of offense Verifications 9 / 9 R01+R21 R01 AZa AZb AZc 1 AZbc I AZca AZab Normal Condition N / A < Umbrall Traditional identification methods Under Fault Symmetrical fault ABC < Umbrall ABC- G < Umbrall Phase(s)- Earth AG > Umbrall > Umbral2 > 5% < 5% BG > Umbrall > Umbral2 > 5% <5% CG > Umbrall > Umbral2 >5% < 5% BCG > Umbrall > Umbral2 < 35% - - > 35% - - CAG > Umbrall > Umbral2 - <35% - - > 35% - ABG > Umbrall > Umbral2 - - < 35% - - >35% Phase-phase BC > Umbrall < Umbral2 <5% >5% CA > Umbrall < Umbral2 <5% >5% AB > Umbrall < Umbral2 < 5% >5% Sampling of signals

[16] The voltage and current phasors used as input for the proposed method are calculated at the same sampling rate as commercial relays and the phasor calculation is the same as for conventional relays. Io, li ek

[17] The parameters Io, li and eh correspond to the phasors of the zero, positive and negative sequence components of the current and are calculated in commercial relays using traditional methods.

[18] Figure 2 refers to graphs showing the AG faults along the 2600 km line. The dZ parameters, which are the variations in impedances as a percentage, can be followed. It is possible to see that ΔZα shows a large variation, while ΔZ^ varies less than 5%.

[19] The same parameters are plotted in Figure 3, where BC-G faults along the line are analyzed. It can be seen that AZbca shows a large variation, while ΔZα changes less than 35%.

[20] Finally, there are BC faults along the line in Figure 4. It can be seen that ΔZ^ shows a large variation, while ΔZα changes less than 5% which is the 0 threshold for fault adjustment between two phases.

Claims

1 / 3 CLAIMS 1. A method for identifying the type of fault in a transmission system characterized by being applicable to identifying the type of fault in a power transmission line of any length, preferably very long lines such as half-wavelength lines, and comprising the following steps: Initialize the voltage phasor variables [Vpre, VpPe, Vpre], current [Ipre, f™, ^τε] abc JL abc J and the complex impedance variable [Zppe, Zppe, Zppe, Zp™, Zppe, Zpre] before the fault L abc ab bc ca J (pre-fault); (A) Receive the voltages and currents in the voltage and current phasor variables [V , VV,1,11 ]; abcabc (B) Calculate the zero, positive and negative sequence currents [ / 0, / 1 and / 2]; (C) Calculate the parameters Λ and Λ that relate the zero-sequence and negative-sequence currents to the positive-sequence current, according to the following equations 1 and 2;(D) Test if Λ + Λ > Threshold1: - if the result is 0.15, there is an asymmetrical fault and proceed to step (I); - if the result is not 0.15, there is a symmetrical fault or normal operating condition and proceed to step (F); (E) Detect if there is a symmetrical fault (three-phase) and type selected between ABC or ABCG by traditional methods; (F) Update and store the values ​​of the pre-fault voltage and current variables, with the current values ​​of voltages and currents, according to equation 3 if there is no fault (normal operating condition); (G) Calculate, update and store the pre-fault impedance values, with equation 4, and return to step (B) of reading the input variables; (H) Store the values ​​of the voltages and currents under fault if a fault condition is identified with equation 5; (I) Calculate the impedances under fault according to equation 6; Petition 870230010234, dated 03 / 02 / 2023, page 5 / 10; 2 / 3 (J) Calculate the variation variables according to equation 7; (K) Test if R > Threshold2: - if the result is 0.1, there is an unbalanced fault involving ground, selected from the type group AG, BG, CG, ABG, BCG, CAG, and proceed to step (M); - if the result is not 0.1, there is a fault that does not involve ground, and proceed to step (O); (L) Test the variation of impedances Z, ZZ, ZZZ for a Tolerance = 5%: • If there was variation in Za and no variation in Zbc, i.e., ΔZ > Tol. and ΔZ < Tol., then the fault is AG. • If there was variation in Zb and no variation in Zca, i.e., ΔZ > Tol. and ΔZ < Tol., then the fault is BG. b ca • If there was a variation in Z, and there was no variation in Z,b, that is, ΔZ > Tol. and ΔZ < Tol., then the fault is CG. c ab (M) Test the variation of the impedances Z, Zb, Z, Z^, Z^Z for a Tolerance = 35%: • If there was a variation in Zab and there was no variation in Z^, that is, ΔZ > Tol. and ΔZ < Tol., then the fault is ABG.abc • If there was a variation in Zbc and no variation in Za, i.e., ΔZ > Tol. and ΔZ < Tol., then the fault is BCG. bca • If there was a variation in Z and no variation in Zb, i.e., ΔZ > Tol. and ΔZ < Tol., then the fault is CAG. cab (N) Test the variation of impedances Z, Zb, Z, Z^, Z ,Z for a Tolerance = 5%: • If there was a variation in Zab and no variation in Z^, i.e., ΔZ > Tol. and ΔZ < Tol., then the fault is AB. ab c Petition 870230010234, dated 03 / 02 / 2023, page 6 / 10. 3 / 3 • If there was a change in Zbe and no change in Zb, that is, ΔZ > Tol. and ΔZ < Tol., then the fault is BC. be a • If there was a change in Zb and no change in Zb, that is, ΔZ > Tol. and ΔZ < Tol., then the fault is CA. ea b Petition 870230010234, dated 03 / 02 / 2023, page 7 / 10

Images