IR ees Chapter 4 mass, Synchronous Machine Parameters The synchronous machine equations developed in Chapter 3 have the inductances and resistances of the stator and rotor circuits as parameters. These are referred to as fundamental or hasic parameters and are identified by the elements of the d- and q-axis equivalent circuits shown in Figure 3.18. While the fundamental parameters completely specify the machine electrical characteristics, they cannot be directly determined from measured responses of the machine. Therefore, the traditional approach to assigning machine data has been to express them in terms of derived parameters that are related to observed behaviour as viewed from the terminals under suitable test conditions. In this chapter, we will define these derived parameters and develop their relationships to the fundamental parameters. 4.1 OPERATIONAL PARAMETERS A convenient method of identifying the machine electrical characteristics is in tecms of operational parameters relating the armature and field terminal quantities. Referring to Figure 4.1, the relationship between the incremental values of terminal quantities may be expressed in the operational form as follows: AY, (s) = Gis)Aels)-LyAi,(s) 4.1) AWS) = -L(s)Ai,(9) 4.2) 139140 Synchronous Machine Parameters Chap. 4 ; d-axis ; q-axis Ai, eg Ay, Ata network network AW, Figure 4.1. The d- and q-axis networks identifying terminal quantities where G(s) is the stator to field transfer function Ls) is the d-axis operational inductance L,{8) is the q-axis operational inductance In the above equations, s is the familiar Laplace operator and the prefix A denotes incremental or perturbed values. Equations 4.1 and 4.2 are true for any number of rotor circuits. In fact, RH. Park in his original paper [1] expressed the stator flux equations without specifying the number of rotor circuits. With the equations in operational form, the rotor can be considered as a distributed parameter system. The operational parameters may be determined cither from design calculations or more readily from frequency response measurements. When a finite number of rotor circuits are assumed, the operational parameters can be expressed as a ratio of polynomials in s. The orders of the numerator and denominator polynomials of L,(s) and L,(s) are equal to the number of rotor circuits assumed in the respective axes, and G(s) has the same denominator as L,{s), but a different numerator of order one less than the denominator. We will develop here the expressions for the operational parameters of the model represented by the equivalent circuits of Figure 4.2. This model structure is generally considered adequate for stability studies and is widely used in large scale stability programs. The rotor characteristics are represented by the field winding and a damper winding in the d-axis and two damper windings in the g-axis. The mutual inductances Ly and Lg are assumed to be equal; this makes all mutual inductances in the d-axis equal. In the next section we will consider the effect of not making this simplifying assumption.Sec. 4.1 Operational Parameters 141 Figure 4.2 Structure of commonly used model With equal mutual inductances, Equations 3.127, 3.130 and 3.131 for d-axis flux linkages in the operational form become Wyls) = Lil) Lagi) Ladi aS) (4.3) Wg) = ~LygiglS)*L gpl) Laging() (44) Wy) = —Lagig() Lygig (DL igi g (4.5) The operational forms of Equations 3.123 and 3.124 for rotor voltages are eal) = SWeg(S)-Wea(O)+RysigalS) (4.6) 0 = s¥i4(5)-V,O) "Raia an where w,{0), w,A0) and w,A0) denote initial values of the flux linkages. It is preferable to express the above equations in terms of incremental values about the initial operating condition so that the initial values drop out; this makes it more convenient to manipulate the operational equations. Substituting for the flux linkages in terms of the currents, the rotor voltage equations in incremental form become Megs) = SAVy(5)+*RyAig(S) = SL ,yAig(S)+(Rey*Shg) Aig(9) +SL,gAisg(S) (4.8)142 Synchronous Machine Parameters Chap. 4 0 = SAW 4(8)+RigAig(s) as) = SL yy Mig(9) +L Mig 9) (Ryy*SL 1) ii g(S) Our objective is to express the d-axis equations in the form of Equation 4.1, and this can be achieved by eliminating the rotor currents in terms of the terminal quantities ey and i,. Accordingly, solution of Equations 4.8 and 4.9 gives Aig(s) = Dol gPSly, NG g(S) +5Ly(Ryy*Sl,,)i(9)] (4.10) Ai,g(s) = Del tadeal +L gy (Ryt hg Ai,(s)] (4.11, (s where DS) = Syrah ga Led *8 Cy aRa*hgaR ia) RaRa 2) Given that Lgl; La = Laat lya Lig = Laathyg substitution of Equations 4.10 and 4.11 in the incremental form of Equation 4.3 then gives the relationship between the d-axis quantities in the desired form: AY,(S) = G@)Aeg(s)-L{(9) Ai) The expressions for the d-axis operational parameters are given by 1H(T+T,)s+T,T,s? aay L) = L, “ [047,547,757sec. 4.1 Operational Parameters 143 (1+sT,) Gs) = G,___-# __ (4.14) 1+(T,+T,)s+7,T,s? where L, L, G, - 7, = Ra Ra 1, Lygtlyy n- Ligtlyg Ru Ra (4.15) Lal, Lgl, 1, - af gt eet T, = nent Ral Leathe Ra © Lath Legh Lghygl 1 = aye taste 1, = Ete Lesa Ral Lyly RA Llp Laghy Lgl Equations 4.13 and 4.14 can be expressed in the factored form: LesT/)+sTy! typ) = 1g Lets «ais CesT, d-sT,) 1+sT, (6) = G, Us) _ (4.17) (1+sT jg) +8Tyg ) The expression for the q-axis operational inductance may be written by inspection and the similarities between d- and q-axis equivalent circuits, In the factored form, it is given by 1,0 = t, (losTy ya 7) (4.18) (1+sTj.)(1 +875) The time constants associated with the expressions for Ls), L,(s) and G(s) in the factored form as scen in the following section represent important machine Parameters.144 Synchronous Machine Parameters Chap. 4 4.2 STANDARD PARAMETERS Following a disturbance, currents are induced in the machine rotor circuits. As shown in Section 3.7 for a short-circuit, some of these induced rotor currents decay more rapidly than others. Machine parameters that influence rapidly decaying components are called the swbtransient parameters, while those influencing the slowly decaying components are called the fansient parameters and those influencing sustained components are the synchronous parameters. The synchronous machine characteristics of interest are the effective inductances (or reactances) as seen from the terminals of the machine and associated with the fundamental frequency currents during sustained, transient and subtransient conditions. In addition to these inductances, the corresponding time constants which determine the rate of decay of currents and voltages form the standard parameters used in specifying synchronous machine electrical characteristics. These standard parameters, as discussed below, can be determined from the expressions for the operational parameters 1,{s), G(s) and L,(s). The constants Tj, Tj, Tj and T,' are the four principal d-axis time constants of the machine. Their relationships to fundamental parameters are determined by equating the respective numerators and denominators of Equations 4.13 and 4.16. Thus, (+sTjg\(l+sT yg) = 1+8(T,+T,)+8°(2,7,) (4.19) (l+sTj)+sTj) = 1+8(0,+T,)+s°TyT,) (4.20) The expressions for the four time constants can be determined accurately by solving the above equations. However, such expressions would be very complex. Simpler expressions can be developed by making some reasonable approximations. Parameters based on classical definitions The solution of Equations 4.19 and 4.20 is considerably simplified by recognizing that the value of Rj, is very much larger than Ry This makes 7, and T; very much smaller than 7,, and 7; and 7, very much smaller than T,. Hence (+sTi)(l+sTj,) = (1+sT,(1+8T;) (4.21) (l+sT/)(+sT/) = (1+sT, +57.) 4.22)Sec. 4.2 Standard Parameters 145 We thus have the following approximate relationships Typ * T, Ta = T, (4.23) Tj =T%, Ty = T, The expressions for 7, to Ty in terms of the fundamental parameters are given by Equation 4.15. These time constants are in per unit (radians). They have to be divided by rated angular frequency (cy) =2n/) to be converted to seconds. For a 60 Hz system, @=376.991 (usually approximated to be 377) radians per second. With the stator terminals open (Ai,=0), from Equations 4.1 and 4.17, we have sha a, (4.24) (+sT al +sTs, AY,(S) = Gy fd The above indicates that, for open-circuit conditions, the d-axis stator flux and hence the terminal voltage respond to a change in field voltage with time constants Tj) and Tj. Since Ryq is much larger than Ry, Tj is much smaller than Tj Thus Tp is associated with the initial change and is referred to as the d-axis open-circuit subtransient time constant. The time constant Tj) represents a slow change corresponding to the transient period and is referred to as the d-axis open-circuit transient time constant. The time constants Tj and 7 represent the transient and subtransient short-circuit time constants, respectively. Let us now examine the effective values of 1s) under steady-state, transient, and subtransient conditions. Under steady-state conditions, with s=0 Equation 4.16 becomes 1,0) = L, (4.25) This represents the d-axis synchronous inductance. During a rapid transient, as s tends to infinity, the limiting value of Ls) is given by Li Ll) . of rity (4.26) TyoTa0146 Synchronous Machine Parameters Chap, 4 This represents the effective inductance Aw,/Ai, immediately following a sudden change and is called the d-axis subtransient inductance. In the absence of the damper winding, the limiting value of inductance is Lj = Ll) T - a : uz] 40, This is referred to as the d-axis transient inductance. Substitution of the expressions of the time constants from Equations 4.15 and 4.23 in Equations 4,26 and 4.27 gives the following alternative expressions for L'' and Lj in terms of the mutual and leakage inductances. (427 Lj = 1»_—_Latttng (4.28) Leaky baglig*hylig and Ll, Ly = b+ (4.29) Lagtlyg The above expressions for the subtransient and transient inductances can also be derived from the principle of constant flux linkages, which states that the flux linking an inductive circuit with a finite resistance and emf cannot change instantly. Following a disturbance, the rotor flux linkages do not therefore change instantly. For conditions immediately following a disturbance, the equivalent circuit of Figure 3.17, with incremental rotor flux linkages (Ay,,, Ay),) set to zero, reduces to that shown in Figure 4.3. Ay,=0 AW, Liq AWg=0 Figure 4.3 Equivalent circuit for incremental values, immediately following a disturbancegec. 4.2 Standard Parameters 147 From Figure 4.3, the effective inductance Aw,/Ai, representing Lj is seen to be the came as that given by Equation 4.28. With the damper winding absent (L,.=2), the effective inductance corresponds to Lj given by Equation 4.29. The expression for the g-axis parameters may be readily written by recognizing the similarities in the structure of d- and g-axis equivalent circuits. Thus, the g-axis open-circuit transient and subtransient time constants are given by L,, +L, a Mag * 1 Tj = 8 (4.30) Ig Tyo = ra | 431) and the subtransient and transient inductances are given by L's Laakrgi-ag (4.32) Lt 0 TT Liga lade Li = Ly Lob (4.33) a he Ly*lhig ‘The g-axis synchronous inductance is given by the steady-state value of L,(s), which is equal to Ly. ‘The expressions derived above for the machine standard parameters are based on the assumptions that during the subtransient period Ryj=R,,=0 and that during the transient period R,4=R,,=2. These assumptions have been used in the classical theory on synchronous machines [2,3]. However, in recent years there has been some concern [4,5] that significant errors could occur between the values of parameters calculated using the above assumptions and those derived from test measurements such as those described in IEEE Standard 115-1983. Expressions which more closely reflect the definition of the standard parameters are derived below. Accurate expressions for standard parameters The exact values of T'jg and Ty, are given by the poles of L,{s) and those of Ty and 7’ by the zeros of L,(s). In other words, we need to use the exact solutions of Equations 4.19 and 4.20. From Equation 4.19, the poles of L,{s) are given by Th ste — TT, NT; =0 (4.34)148 Synchronous Machine Parameters Chap. 4 However, the exact expressions for the poles of 1,48) become cumbersome and difficult to handle, The expressions can be simplified considerably without much loss of accuracy if it is recognized that 47,7; is much less than (7,+7,)*. With this simplification, the roots of Equation 4.34 reduce to _ 1 Sp=- T+T; 7,47, s, - 4th qT, The open-circuit time constants are equal to the negatives of the reciprocals of the roots, Ty = T,+T, (4.35) TT, pe = 2s (4.36) ©” TT, Similarly, by solving for the roots of the numerator of L,4s), we have Tj = TT, (437) ry = 2s (4.38) TT, The transient and subtransient inductances are given by substituting for the above time constants in Equations 4.26 and 4.27. Thus, Lj = L, TyTs (4.39) TT, yr, ur =p le (4.40) ‘a TT, The accurate and approximate (classical) expressions for the standard parameters are summarized in Table 4.1. These expressions apply to a synchronous machine model represented by the equivalent circuit of Figure 4.2 which considers two rotor circuits in each axis with equal mutual inductances.Sec. 4.2 Standard Parameters 149 It should be noted that most stability programs assume that the input data in terms of transient and subtransient parameters are based on the simplifying assumptions from classical theory. However the data provided by some generator manufacturers and those obtained by standard test procedures correspond to the exact values of these parameters. There could thus be some inconsistency associated with the definition and use of standard parameters. As illustrated in Example 4.1, the discrepancies between the two definitions are likely to be significant primarily for the q-axis parameters. In this book, unless otherwise specified, we will assume the classical definition. Parameters including unequal mutual effects In deriving the expressions for the standard parameters above, it is assumed that all mutual inductances in the d-axis are equal. The reciprocal per unit system chosen makes the mutual inductances between the armature and field and between the armature and damper equal. The mutual inductance between field and damper, however, could be different from these mutual inductances. Although the assumption of all d-axis mutual inductances being equal gives good results in calculating armature quantities, it could lead to significant errors in calculating the field current during transient conditions [6]. The d-axis equivalent circuit including unequal mutual effects and the various mutual and leakage fluxes involved are shown in Figure 4.4. Ly SYN | Ly PY, 3 La | Ra (a) Equivalent circuit (b) Flux paths Figure 4.4 Unequal mutual effects in d-axis150 Synchronous Machine Parameters Chap. 4 Table 4.1 Expressions for Standard Parameters of Synchronous Machine Parameter Classical Expression Accurate Expression Tio T, T,+T, T; T, TT, Ty tr; TAT T,+T)] Tj Ts TIT IT *T5)] Lj LAC JT) LETT -T) Ly LATIMES) LT IMT) with Lak t Laat Lath, Ll, T, = 4 L, ets __ Rul Lgl Leaky lal, Notes: 1. Similar expressions apply to g-axis parameters. 2. All parameters are in per unit. 3. Time constants in seconds are obtained by dividing the per unit values given in the table by @)=2a/ 4. All mutual inductances in d-axis assumed equal.sec. 4.2 Standard Parameters 151 The series inductance, L,,=Lyjq-L,q. corresponds to the peripheral leakage flux (6p) which links the field and damper, but not the armature. Assuming that Ry=0 during the subtransient period and that Ry =:0 during the transient period, following the approach used previously, the expressions for the standard parameters may be written by inspection of the equivalent circuit as follows: a L,qth, L, Lj = —— pcos aethed “Ey ta, Laat hath y Ly Lyatly Lghjah aa Lal glhag bealyalins Lf = Lp 1 head pad ed fp __ Lea hig bealrg bahia taki Lak Lagtlytl, Tj «en aan Ry 1 ae Ty = —|L, yr = Barta 1 i] Ty = —|LytLyt R, oe Lath pe Dp 4 Peale ga boa bya a Re Leabieg yah yh yh ea E ply Lgaly The above expressions are based on the approximations associated with the classical definition of the parameters. Accurate expressions applicable to a more complex model structure consisting of these rotor circuits in each axis and unequal d-axis mutual inductances are given in reference 7. Parameters of salient pole machines: In the discussion of standard parameters so far, we have considered a model structure with two rotor circuits in each axis. This is applicable to a round rotor machine. However, for a laminated salient pole machine, the damper winding is the only rotor circuit in the q-axis; therefore, only one q-axis rotor circuit (denoted by the subscript 1g) is applicable. The parameters of this rotor circuit are such that it tepresents rapidly decaying subtransient effects. The second rotor circuit (denoted by subscript 2g) is ignored and no distinction is made between transient and synchronous152 Synchronous Machine Parameters — Chap. 4 (steady-state) conditions. Hence, the expressions for the q-axis parameters of a salient pole machine are as follows: Ly = Lythyy Ly = 1 cat. (4.42) Lag*Lig ” da +h, To" Rt Ig The transient parameters Z and T,', are not applicable in this case. In the d-axis, it is appropriate to consider two rotor circuits (field and damper) and the expressions previously derived are applicable to salient pole machines. Reactances: In per unit, the subtransient, transient and synchronous reactances are equal to the corresponding inductances. Hence, common practice is to identify synchronous machine parameters in terms of the reactances, instead of the inductances. Typical values of standard parameters: Table 4.2 gives ranges within which generator parameters normally lie. From the expressions for machine parameters summarized in Table 4.1, it is readily apparent that X,2X,>Xp2Xp> Xf 2x7 (4.43) Tyy > Ty > Tyy > Ty! > Trg (4.44) Typ > Ty > Ty > Ty (445)sec. 4.2 Standard Parameters 153 Table 4.2 Parameter Hydraulic Units | Thermal Units Synchronous Xg 0.6 - 1.5 1.0 - 2.3 Reactance X, 04-10 10-23 Transient | 02-05 0.15 -04 Reactance e - wie 4 3-1 Subtransient xy 0.15 - 0.35 0.12 - 0.25 Reactance x; 0.2 - 0.45 0.12 - 0.25 Transient OC To 15-908 3.0- 10.08 Time Constant Tyo z Ds.20s Subtransient OC Ti 0.01-0.05s | 0.02-0.05s Time Constant ‘f 0.01-0.09s | 0.02-0.05s Stator Leakage Inductance x 0.1 - 0.2 0.1 - 0.2 Stator Renan ce R, 0.002 - 0.02 | 0.0015 - 0.005 Notes: 1. Reactance values are in per unit with stator base values equal to the corresponding machine rated values. 2. Time constants are in seconds, Example 4.1 The following are the per unit values of the standard parameters of the 555 MVA, 0.9 p.£,, 60 Hz turbine generator considered in the examples of Chapter 3: 1421.81 L,=1.16 1,=0.15 R,=0.003 14703 Lj =0.23 1 =0.25 Tj=8.0 s Tj-0.03 s | T,,=0.07 sSynchronous Machine Parameters Chap. 4 The transient and subtransient parameters are based on the classical definitions. The parameters given correspond to unsaturated values of Lag and Lay. (a) (b) Assuming that Ly,=Laj determine the per unit values of the fundamental parameters, i.e., elements of the d- and q-axis equivalent circuits. Using the fundamental parameters computed in (a) above, and without making the simplifying assumptions of the classical definitions, calculate the accurate values of the transient and subtransient parameters. How do they compare with the values based on the classical definitions? If, at the rated output conditions, the effect of magnetic saturation is to reduce Tq and Ly, to 83.5% of their unsaturated values, find the corresponding values of the standard parameters based on the classical definitions. Solution (a) We first compute the unsaturated mutual inductances 1,81-0.15 = 1.66 pu 1.76-0.15 = 1.61 pu We then compute the rotor leakage inductances from the expressions for transient and subtransient inductances. From Equation 4.29, the expression for Lj based on the classical definition is Laka Lg = Ly 6 Latha Substituting the respective numerical values gives 1.66%Ly 1.66+L, 0.3 = O.15+ Solving for Ly yields = 0.165. pu Similarly, from the expression for Z{ given by Equation 4.33, we obtain 161xL,, 0.65 = 0.15+ 4 161+L, Solving for Z,, givesSec. 4.2 Standard Parameters 155 Iy, = 0.7252 pu From Equation 4.28 Lylgl, | a Leakya* Laaliat Lalya Substituting the respective numerical values gives 1.66%0.165%L,4 1.660.165 +1.66xL,,°0.165%L, 0.23 = 0.15+ Solving for Lig, we have Lyg = 0.1713 pu Similarly, from the expression for Z," given by Equation 4.32, we obtain 1.61x0.7252xL,, 0.25 = 0.15+ 1.61%0.7252+(1.61+0.7252)Ly, and solving for L2, yields = 0.125 pu Lag Next, we compute the rotor resistances from the expressions for the open-circuit time constants. From Equations 4,15 and 4.23, the expressions for Tj and Tj, based on the classical definitions, are and Leak we Dp ye batt ve - alt it Substituting the numerical values and noting that the time constant in per unit is equal to 377 times the time constant in seconds, we obtain = +0.165 ff" 8.0%377 = 0.000605 pu156 Synchronous Machine Parameters Chap. 4 and 1 1.66x0.165 Ry = (0.1713 +O 1d 0.03x377 Lets) = 0.0284 pu Similarly, using Equations 4.30 and 4.31 to compute the g-axis rotor circuit resistances, we obtain Rj, = 0.0062 pu and Ry, = 0.0237 pu The following is a summary of the per unit values of the fundamental parameters: R, = 0.003 1,-0.15 Lyg7 1.66 Lyg= 161 R= 0.0006 14g 0.1713 R,4= 0.0284 R= 0.0062 L, =0.125 R,,=0.0237 (b) The accurate expressions for the d-axis transient and subtransient constants are summarized in Table 4.1. Substituting the fundamental parameters computed above in the expressions for the time constants T; to J, and dividing by 377 to convert to seconds, we obtain T, = 80 s T,=O171 s 7, =003 s T, = 1326 s 7, = 0.0288 s T, = 0.023 s The accurate value of the d-axis transient inductance is TAT, , atts 14 Lan itty = 181 1.326+0.0288 8.0+0.171 =03 pu The accurate values of transient and subtransient open-circuit time constants aresec. 4.2 Standard Parameters 157 Typ = T,4T, = 8.040.171 = 8.171 s TT, T*T, Tyg = 8.0x0.03 8.0+0.03 = 0.0294 s Similarly, for the q-axis, we obtain T,=10 s T, = 0.1943 s 7, = 007 s T, = 0.3693 s 7, = 0.0294 s T, = 0.0269 s The accurate value of the g-axis transient inductance is pny teh «TST, 6 0.3693 +0.0294 1.0+0.1943 = 0.5875 pu = 17 The accurate values of the g-axis open-circuit time constants are Tj, = T,+T, = 1.0+0.1943 = 1.1943 s The ge vt, ~ 007 08 = 0.0586 s There are no approximations associated with the classical definitions of the subtransient inductances; hence, accurate values of L4 and 1," are the same as the given values. The following is a comparison of the accurate and approximate values of the parameters.158 Synchronous Machine Parameters Chap. 4 Parameter uit based on | Accurate Value 7 03 03 Ti 80 8.171 Tip 0.03 0.0294 Ly 0.65 0.5875 Tyo 10 1.1943 Ts 0.07 0.0586 The differences are significant only for the q-axis. (©) At the rated output condition, the saturated values of mutual inductances are 0.835xL,,, = 0.835x1.66 = 1.386 pu 0.835xL 4, = 0.835x1.61 = 1.344 pu ‘The corresponding values of the standard parameters, based on the classical definitions, are Lg = 1.386+0.15 = 1.536 pu L,= 1.344+0.15 = 1.494 pu 1.386x0,165x0.1713 Lf = 0.15 +O OO OT 4 "38630165 +1.386x0.1713+0.165x0.1713 = 0.2292 pu 1,344 x0.7252 Lj = 015+ 1.344 +0.7252 = 0.621 pusec. 4.3 Frequency-Response Characteristics 159 1,344 0.7252 0.125 Ly = 0.15+— 1 34400-7252 0.125 ___ 1.344 x0.7252+1.344 «0,125 +0.7252x0.125, = 0.2488 pu 1.386+0.165 Ty = —————— = 6.86 4 °0,0006%377 . Ty = 1 (0.1713 + 1.386x0.165° 0.028437 1.386 +0.165 = 0.0298 s 1, = 1344007252. 9 g95 5 0.0062 377 T - 1 (0.125 + 1.344x0.7252 0.0237 377 1.344+0.7252, = 0.0667 s : 4.3 FREQUENCY-RESPONSE CHARACTERISTICS It is of interest to examine the frequency-response characteristics of the operational parameters and relate them to standard parameters. Such characteristics provide useful insight into the dynamic characteristics of the machine and may be readily sketched using the asymptotic approximation. With the rotor effects represented by two circuits in the d-axis, L,{s) and G(s) are given by Equations 4.16 and 4.17, respectively. Figure 4.5 shows the magnitude of L.{s) as a function of frequency. The transient and subtransient time constants and inductances have been used to identify the corner points for the asymptotic approximation. The plot shown is for the generator considered in Example 4.1. The general shape of the frequency response characteristic, however, is applicable to any synchronous machine. The effective inductance is equal to the synchronous inductance L, at frequencies below 0.02 Hz, the transient inductance 1} in the range 0.2 Hz to 2 Hz, and the subtransient inductance L4 beyond 10 Hz. In stability studies, the frequency range of interest is that corresponding to Lj. Figure 4.6 shows the magnitude of G(s) as a function of frequency for the same machine, with Gy normalized so that it is equal to 1.0. From the plot we see that the effective gain drops off considerably at high frequencies. This indicates that high- frequency variations in field voltage are not reflected in the stator flux linkage and hence other stator quantities.160 Synchronous Machine Parameters Chap. 4 (les; )0+sT/) 19) ~ Lager dosTQ) Magnitude in pu ° ri 1 1 Ty Te o in rad/s 3 aS Figure 4.5. Variation of magnitude of L,(s) Gyo) 20 dB/decade 3-20 3 40 dB/decade 2 0 z gf . = G— 2st = 60 oo) Ta esT iy | 100 © in rad/s a Tyg Figure 4.6 Variation of magnitude of G(s) with frequencysec. 4.4 Determination of Synchronous Machine Parameters 161 ‘Armature time constant As discussed in Section 3.7.2, the armature time constant, T,,, gives the rate of decay of the unidirectional component (dc offset) of armature phase currents following a three-phase short-circuit at the terminals. The unidirectional components of currents, which have unequal magnitudes in the three phases, produce a stationary mmf wave. This induces fundamental frequency currents in the rotor circuits, keeping the flux linking with these circuits constant. The flux path due to the direct current in the armature is similar to those corresponding to subtransient inductances L and L/", Since the rotor moves at synchronous speed with respect to the mmf wave produced by the direct currents in the stator, the mmf wave alternately sees conditions corresponding to Lj and L,’. Hence, the effective inductance seen by the direct currents in the armature lies between Lj and Z,’. It will be shown later that this is also the inductance Z,, seen by the negative sequence currents applied to the stator windings. Therefore, Tr (4.46) “R ‘a A(Lirhy 2 per unit The value of 7, lies between 0.03 and 0.35 s. 4.4 DETERMINATION OF SYNCHRONOUS MACHINE PARAMETERS The conventional method of determining synchronous machine parameters is from short-circuit tests on unloaded machines. The test procedures are specified in IEE Standard 115-1983 [8]. These tests provide Xj, X,. Xj, Xj! Tfo Ths Tf and TY. They do not, however, provide g-axis transient and subtransient constants. In addition, they do not include measurement of the field circuit during the short-circuit tests, and consequently the field circuit is not specifically identified. The limitations of these procedures for providing data suitable for stability studies have been recognized for some time. Several alternative testing and analytical methods have been proposed and used to obtain better models: . Enhanced sudden short-circuit tests; * Stator decrement test; . Frequency-response tests — Standstill frequency response, — Open-circuit frequency response,162 Synchronous Machine Parameters Chap. 4 — On-line frequency response; . Analysis of design data; and . Quadrature axis saturation measurements. The special features of theses methods are briefly discussed below. A detailed description of these methods is, however, beyond the scope of this book. Readers may refer to the references provided for additional information. Enhanced short-circuit tests Improved methods of utilizing results from sudden short-circuit tests to determine more accurate d-axis parameters are described in references 9 and 10. The most important feature of these methods is the utilization of rotor current measurements during the short-circuit tests to identify the field-circuit characteristics more accurately. Among the disadvantages of the short-circuit approach are the inability to provide q-axis parameters accurately and the necessity of exposing the machine to a severe shock. Decrement tests These tests are similar in approach to the sudden short-circuit tests in that the time responses of machine variables following a sudden disturbance are [Link] identify machine characteristics. References 10, 11 and 12 use this approach. In this approach, with operating conditions arranged such that current is flowing only in the direct axis (i,=0), the unit is tripped and the resulting terminal voltage and ficld current decay are used to extract parameters for a model in much the same way as with sudden short-circuit data. A similar test is performed with current flowing only in the quadrature axis (j,=0) to obtain q-axis data. To maintain unsaturated conditions, these tests are conducted at partial load and reduced voltage. These tests provide both d- and q-axis data, However, they are somewhat difficult and expensive to conduct. For most machines, it is difficult or impossible to attain unsaturated conditions which unfortunately complicates the testing and the analysis of results. The tests require rescheduling of generation and several unit trips, which is often not practical. Frequency-response tests a. Standstill frequency response (SSFR) In the SSFR technique, all tests are conducted with the unit at rest (rotor stationary) and disconnected from the system. The rotor must be aligned to two particular positions with respect to the stator during the tests. With the stator excitedsec. 4.4 Determination of Synchronous Machine Parameters 163 ty a low level (460 A, +20 V) source over the range of frequencies from 1 mHz to 1 kHz, the following responses are measured: With the stator winding excited and field shorted Ai, (8) sG(s) = —# (4.47) Ai,(s) and Ae,(s) Z,(s) = —# (4.48) a Ais) With the stator winding excited (field condition immaterial) e(s) Zs) = -—# (4.49 a) Ais) ) The latter two, Zs) and Z,(s), are the d- and q-axis operational impedances as viewed from the armature terminals. The operational inductances are computed by subtracting the armature resistance from these impedances: L,(s) = 2K, (4.50) s and Ls) = 2f-Ra (4.51) s where R, is the de resistance of one armature phase and s=jo. The parameters of the d-axis equivalent circuit are obtained using transfer function approximations for the functions Ls) and sG(s). The function sG(s) rather ‘han G(s) is used because the former can be measured at the same time as Zs), with he field shorted. The g-axis parameters are obtained using the transfer function for £,(s). The order of the transfer function depends on the number of rotor circuits assumed in the respective axes. Because the tests are conducted at very low flux levels, the results must be sorrected to bring them from the “toe” of the saturation curve to normal unsaturated evels. This is done by minor adjustment of the mutual reactance in each axis. This technique has been used in references 7, 13, 14, and 15. In addition, it is164 Synchronous Machine Parameters Chap, 4 now a “trial use” standard of the IEEE [16]. SSFR testing is relatively easy to perform either in the factory or during a maintenance outage on a unit. Besides the required adjustment for unsaturated conditions, there are two other limitations to this type of testing, both of which involve rotational effects. Where damper windings are employed, they are often just overlapped and may not form a good connection at standstill. In addition, the extent to which rotation causes the slot wedges to form a low impedance path to the rotor is largely unknown. At low and high frequencies, the data obtained are expected to be good. In the mid-frequency range it is expected to be good for most machines with no specific damper windings or the equivalent, short dampers and non-magnetic retaining rings (17). Details of models developed from SSFR tests on three large generators are given in references 7 and 14. The measured L{s), sG(s) and L,(s) characteristics for one of these generators (500 MW, 3800 RPM generator at Lambton GS in Ontario) are shown in Figures 4.7, 4.8 and 4.9. Second- and third-order transfer function approximations to the measured characteristics are also shown in the figures. For this generator, second-order transfer functions are seen to be good approximations for both d- and q-axis. However, this may not always be the case; depending on the rotor construction, third-order transfer functions may be more appropriate in some cases [7,14]. b. Open-circuit frequency response (OCFR) Open-circuit frequency-response testing allows confirmation of some 6f the SSFR data in the middle of the frequency range for the d-axis only [7]. For this test, the unit is operated on open-circuit at reduced voltage. The field is excited at various frequencies and the field-to-stator frequency-response measured. The difference between this response and the equivalent one from the standstill tests gives some indication of rotational effects. The test is normally done at more than one voltage to examine saturation effects. By conducting the test with various signal amplitudes, slot wedge conduction effects can be assessed. c. On-line frequency response (OLFR): In many respects, on-line frequency response (OLFR) testing is the “proof in the pudding” as far as small-signal verification of machine models is concerned. Here the machine is tested under the same conditions as those under which the model is expected to perform, although over a restricted operating range. For this test, the machine is operated near rated (or at reduced) load preferably over a substantial impedance to the system. The excitation is modulated by either sinusoidal or random noise. Components are resolved on the two axes and data similar to those of the SSFR tests are used to derive a model. The frequency range of usable data in this case is more limited than that of the SSFR tests, but the SSFR data aresec. 4.4 Determination of Synchronous Machine Parameters 165 ade in Phase in degrees Magni : sort? 149189 5535)(1o80.41)(1 650.0053) 20 as con Ly 1p (8077)ss0.0038) (55:06) (150.0089) 20 is 40 ww LH > Magni 4 . oxen 60 Figure 4.7. Lambton GS - Variation of 1,66) (teased) With frequency at standstill with field closed Magnitude in dB Phase in degrees _ i so “5646x0425 —(1280:646)(1 450.003) ° GOS) = 84.25 5 95)(1 50.41) (1 20.0083) « a2 5G(9) =54.25 (180.0028) _ 0 © (@) -34 257 55.06)(1-s0.0089) a 20 400 ° > Magnitud “ ' 2 agnine 20 Figure 4.8 Lambton GS - Variation of sG) (Measured) with frequency at standstill with field closed Magvitude in pu Phase in deerces ° (1+50.562)(1+50.052)(1 480.0022) 30 —— 1,6) =1.9 1280:552)(1+80.052)(1+50.0022) (Tes1.020)(1os0.118)(1+s0.0040) 2s 20 = yong 180428) (12200005) wi - (1+s0417)(150.0086) 1s 40 © Magnitude © Phase | Measured) 6.001 0.001 a 10 10 100 J (ha) Figure 4.9 Lambton GS - Vari with frequency at stand: of 49)166 Synchronous Machine Parameters Chap. 4 expected to be good at the frequency extremes. The OLFR testing allows the middle of the frequency range to be filled in with data for both axes which include rotational effects. The disadvantage of this test is that it requires testing on an operating unit connected to the system, possibly under special system or unit conditions. In addition, it does not provide large signal response information. Details of OLFR measurement techniques and model identification procedures are given in reference 18. Results of OLFR tests on two large units and improved models developed from the tests are given in references 7 and 19. In reference 20, models for three large generators based on SSFR and OLFR tests are validated by comparing the results of simulations with measured responses involving line switching. For one generator, models derived from short-circuit tests and decrement tests are also validated. Calculation of machine parameters from design data Improved generator models developed from design information are described in reference 21. Reference 22 presents more recent work using finite element analysis. Work done on two and three dimensional finite element models in the steady state has shown good agreement with measured results. In addition, two dimensional finite element models have been developed which provide frequency response data that compare well with measured results. The degree to which conduction occurs across slot wedges appears to have a significant effect on the model. at higher frequencies. The significance of rotational effects and the effect of disturbance amplitude on the model require additional investigation REFERENCES [1] R.H. Park, “Two-Reaction Theory of Synchronous Machines,” AJEE Trans., Part I, Vol. 48, pp. 716-730, 1929; Part II, Vol. 52, pp. 352-355, 1933. [2] C. Concordia, Synchronous Machines - Theory and Performance, John Wiley & Sons, 1952. [3] B. Adkins, The General Theory of Electric Machines, Chapman and Hall, 1964. [4] G. Shackshaft, “New Approach to the Determination of Synchronous Machine Parameters from Tests,” Proc. IEE (London), Vol. 121, No. 11, pp. 1385- 1392, 1974. [5] IEE Committee Report, “Supplementary Definitions and Associated Test Methods for Obtaining Parameters for Synchronous Machine Stability Simulations,” /EEE Trans., Vol. PAS-99, pp. 1625-1633, July/August 1980.References (6 (7) (8] (I (10) [i] [12] {13] (4) U5] [6] (17) 167 1M. Canay, “Causes of Discrepancies on Calculation of Rotor Quantities and Exact Equivalent Diagram of Synchronous Machines,” JEEE Trans., Vol. PAS- 88, pp. 1114-1120, July 1969. EPRI Report EL-1424, “Determination of Synchronous Machine Stability Constants,” Vol. 2, prepared by Ontario Hydro, December 1980. ANSI/IEEE Standard 115-1983, IEEE Guide: Test Procedures for Synchronous Machines. Y. Takeda and B. Adkins, “Determination of Synchronous Machine Parameters Allowing for Unequal Mutual Inductances,” Proc. JEE (London), Vol. 121, No. 12, pp. 1501-1504, December 1974. G. Shackshaft and A.T. Poray, “Implementation of New Approach to Determination of Synchronous Machine Parameters from Tests,” Proc. [EE (London), Vol. 124, No. 12, pp. 1170-1178, 1977. FP. deMello and J.R. Ribeiro, “Derivation of Synchronous Machine Parameters from Tests,” JEEE Trans. Vol. PAS-96,. pp. 1211-1218, July/August 1977. FP. deMello and L.N. Hannett, “Validation of Synchronous Machine Models and Determination of Model Parameters from Tests,” JEEE Trans., Vol. PAS-~ 100, pp. 662-672, February 1981. ME. Coultes and W. Watson, “Synchronous Machine Models by Standstill Frequency Response Tests,” IEEE Trans., Vol. PAS-100, pp. 1480-1489, April 1981. P.L. Dandeno and A.T. Poray, “Development of Detailed Turbogenerator Equivalent Circuits from Standstill Frequency Response Measurements,” JEEE Trans., Vol. PAS-100, pp. 1646-1653, April 1981. M.E. Coultes, “Standstill Frequency Response Tests,” Presented at the IEEE Symposium on Synchronous Machine Modelling for Power System Studies, IEEE Publication 83THO101-6-PWR, pp. 26-30. IEEE Standard 115A-1984, [EEE Trial Use Standard Procedures for Obtaining Synchronous Machine Parameters by Standstill Frequency Response Testing. AG. Jack and T.J. Bedford, “A Study of the Frequency Response of Turbogenerators with Special Reference to Nanticoke GS,” IEEE Trans., Vol.