Вы находитесь на странице: 1из 3

Table 1.

Comparison of the simulated results Simulation of Power Transformers


Method C (mF) PF (%) I, (A) tl (%) TI. (kW) Using State Variables
Presented CO = 0.9307 91.78 774.925 99.40 10.350
Reference [1] C1 = 0.9095 91.82* 772.951 99.40 10.364 0. Ozgonenel, G. Onbilgin
C2 = 0.9302 91.77 774.875 99.40 10.349* Author Affiliations: Ondokuz Mayis University, Amasya Techni-
_ _ = 0.9565 91.56 778.520 99.40* 10.370* cal and Vocational College, Amasya, Turkey; Ondokuz Mayis Univer-
* Those values indicate which criterion is used. sity, Electric & Electronic Engineering Faculty, Samsun, Turkey.
Abstract: This letter describes a new approach for power trans-
former modeling that uses state variables in time domain.
guarantees convergence and obtains the optimal capacitor in a unified Introduction: Dolinar, Pihler, and Grcar identifed a simulation
manner. model of a three-phase, three-limb, core-type power transformer. The
Economic Aspects: Assuming the specific cost of the installed ca- behavior of the transformer model during switching-in and fault tran-
pacitor to be KC [$/mF], the cost of energy K [$/kWh], the annual losses sients was studied [1]. Bertrand and Meunier introduced a method of
Wi,, the interest rate i, and the expected capacitor life N years, then the modeling internal faults (turn to ground and turn to turn) in power
total annual cost is [5] transformers. The method leads to a model that is compatible with the
EMTP software [2]. Leibfried and Feser introduced the transfer func-
tion concept for power transformers. Another application for this
method is monitoring of power transformers in service [3]. Wilcox in-
AC =K*Wa +(A/ P)*Kc *C (10) troduced a time-domain modeling and modal analysis that described
how a new form of transformer model, based on modal transformer
where analysis, might be converted from the frequency domain into the time
domain for EMTP implementation [4].
In this letter, a power transformer (with two windings) is modelled.
In this simulation, leakage inductors and magnetizing inductor are ana-
A/P = [i(i + 1)N] [(i +1) 1]. lyzed and assumed as state variables.
Development of the Model: In this section, we will describe an ar-
(11) rangement by which the state equations of a two-winding transformer
can be implemented in a computer simulation. In this work, currents are
For this case, it is clear that the most economical capacitor from the chosen as state variables. The equivalent circuit representation of Fig-
results given in Table 1 is not necessarily yielding minimum energy ure 1 has a cut set of three inductors. Since their currents obey
losses. Kirchhoff's current law at the common node, all three inductor currents
Conclusions: For nonsinusoidal sources, maximum transmission can not be independent.
efficiency, minimum transmission loss, and maximum power factor do v1 is chosen as an input variable, whilev2 is and output variable. Core
not lead to the same shunt compensator values. When the three criteria loses are included by approximating them as losses proportional to the
are combined into one model and solved by the penalty function ap- square of the flux density in the core, or the square of the internal volt-
proach, the results do satisfy the three criteria by one value of capacitor. age em shown in Figure 1; then an appropriate core loss resistance is
A simple theoretical engineering economic analysis shows that the best connected across em, in parallel with the magnetizing inductance, Lm 1-
capacitor may not render minimum energy losses. The resultant equivalent circuit would then be the same as that derived
Ongoing research effort consists of the modification and application from steady-state considerations. From Figure 1, Li I is represented as
of this method to time variations of the load and system impedance. x1, x2 for L12 and x3 for LmiP
References: The general form of state equations is defined as
[1] M.M. Abdel Aziz, E.E. Abou Elzahab, A.M. Ibrahim, and A.F.
Zobaa, "Skin effect on optimal power factor," accepted paper for x =Ax+Bu
LESCOPE2002, June 2002. y = Cx + Du. (1)
[2] S.S. Rao, Optimization Theory and Applications. New York:
Wiley, 1984. The Case of Magnetizing Inrush: Since there is no load connected
[3] P.F. Riberio, "Tutorial on harmonics modeling and simulation," to the secondary winding, the secondary current i2 will be zero. For sim-
IEEE PES Winter Meeting, 1998. plicity, the ratio of terminal voltages is selected 1, (vl / v2 =1).
[4] I.M. El-Amin, S.O. Duffuaa, and A.U. Bawah, "Optimal shunt
compensators at nonsinusoidal bus bars," IEEE Trans. Power Syst., vol.
10, pp. 716-723, May 1995.
[5] D.G. Newnan, Engineering Economic Analysis. San Jose, CA:
L Idi, v= - x1 (RI + Lx1 + x3)Rfe

Engineering Press, 1976. 1 1 Rfe Rfe


[6] IEEE Guide for Harmonic Control and Reactive Compensation
of Static Power Converters, IEEE Standard 519, 1981.
x L -R1xt
Lx + + X3
Appendix: Data for Test System: [6]
A three-phase load of 5100 kW and 4965 kVAR is connected to a
supply bus with voltage 4160 V (2400 line-to-ground), 60 Hz fre-
quency and 80 MVA short circuit capacity. The resistance to reactance
ratio of the power system impedance is assumed to be 10%. The voltage
is distorted; it contains 5% fifth harmonic, 3% seventh harmonic, 2%
eleventh harmonic, and 1% thirteenth harmonic. The system data for
equivalent single-phase mode are:
Transmission system impedance = 0.02163 + j 0.2163 Q
Load impedance = 1.7421 + 1 1.696 Q2.
Copyright Statement: ISSN 0282-1724/02/$17.00 C 2002 IEEE.
Manuscript received 21 April 2002; revised 28 May 2002. This paper
is published herein in its entiret Figure 1. Full equivalent circuit of a two-winding transformer (unloaded)

48 0272-17241021$17.0002002 IEEE IEEE Power Engineering Review, October 2002


xi RfeR 1x1x ++ReX3 + -VI RI Lll L12 FR2
(2)
Ifel 5

il = io =
if, + im vl ~~~~~~~Rfef
Xm sie old"llry $O
Lmm dm -=ifRfe Ar w
dt o

Lm3 =(x1 x3)Rfe


Figure 2. Loaded transformer
Rfe Rfe
X3 =- x1 - X3
L,n Lm (3)
Xi -X2 =fe +X3
Rfe - RI Rfe Ife X1 X2 X3

[3= > Rfe [ + Li vil LI didt R


=V I -X RI±Ijf +ioRfe
[ii Rf Rfe
(4)
The output function is then described as
LpXI =V - R + (X1 -x2 -x3)Rfe
R1 +Rfe Rfex2 Rfe 1
v2 =Rfe(xI -X3) XI =-.
L
x1
I L 2 -x3±
L +-V1
L 3 I
(7)
y =(Rfe -Rfe)LXXI (5) L2di2 =E-x2R2
dt
From (4) and (5), [A], [B], [C], and [D] matrices are obtained for the
case of magnetizing inrush L22 =(X -X2 -x3)Rfe x2R2
Rfe- R1 Rfe Rfe -Rfe - R2 Rfe
x2- 1± x2 x3
[A]= Rfe Rfe -2 -2 L2 (8)
L Lm Lm j
Lm d=
dt
Rfe(x 1 -X2-X)
[B] = LA
O Lm3=Rfe. x 1 fe.X2 - Rfe-X3
[C ] = (Rfe - Rfe)
[D] = [0]. Rfe Rfe Rfe
x3 Lm
x1 -X2
Lm Lm x3.
The transfer function can be obtained by using [A], [B], [C], and [D] (9)
matrices Again v2 is chosen as the output variable, then (10) is obtained
T(s) = C* inv(S * eye(2) -A) *B + D (6)
V2 = i2RIoad = X2Rload
T(s) =(Rfe*(S *Lm+Rfe)
*L1/ (S^2 *L1 *Lm + S *L1 *Rfe y =(O Rload 0) X21

-Rfe* S *Lm+RI* S *Lm+Rl*Rfe) (10)


-Rfe2 / (S2 L1*Lm + S*L *Rfe As a general form, (1 1) can be obtained
-Rfe * S *Lm + R1* S *Lm + R1 *Rfe)*Ll) / L1.
Rfe -R Rfe Rfe 1
T(s) function is calculated in Matlab environment by using symbolic
operators such as Rfe, R1, LI, ... etc.
~1 L 1_
L L

Rfe -Rfe R, Rfe jx *lv1.


The Case of Loaded Transformer:
-~ Rfe Rfe Rfe Lx3
i =f±+ m
Lm Lm Lm
(1 1)
tum ratio = a = 1

From (10) and (11), [A], [B], [C], and [D] matrices are obtained for the
jo X1 X2 case of loaded transformer

IEEE Power Engineering Review, October 2002 49


Step Response

T(s) =Rload *Rfe*SI(R1*Rfe 2*L2+R1*R2*Rfe


14
-S*Ll*Rfe 2+5 3*L1*L2*Lm+S 2*L1*L2*Rfe
12 -7 -S*L1*Rfe 2*L2+S 2*Ll*R2*Lm+S*Ll*R2*Rfe
±Rfe 2*L2*S*Lm-Rfe*S 2*L2*Lm
1 0- -----------------I
-Rfe*R2* S*Lm+Rl* S A2*L2*Lm+Rl* S*L2*Rfe
------------------
±Rl*R2*S*Lm Rl*Rfe 2+Rfe 2*S*Lm.
aU ~-
_SA 2*Ll*Rfe*L2*Lm Rl*Rfe*L2*S*Lm)*Lm
C
-

The function T(s) is calculated in the Matlab environment by using


symbolic operators.
Case Studies: Analyzing the Case of Magnetizing Inrush Condi-
tion: Using state equations of the transformer, the following matrices
are obtained:
0 ~~~~ ~ ~~0.5 1 15
Time (sac) x 1Ii [A
[Al106 1.7252 -1.7264]
Figure 3. Step response of the system
0.0004 -0.0004j

Nyquist Diagram [B] L1.0482]


0.5

0.4
K---
[C] [1 647 -1647]
0.3

0.2
[DI] 0.
.8 0.1

0 Eigenvalues of the system =1.7248. 106.


Controllability Matrix (CM) of the System:
-0.

-0.3

-0.4
+ 04& 40.41494]

-0.5
If we provide an estimate of the number of linearly independent rows or

-1 -0.9 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0


columns of a matrix CM, we get 2 (rank of the system). Thus the system
Real Axis during magnetizing inrush is controllable.
Observability Matrix (OM) of the System:
Figure 4. Nyquist response of the system
16M 7
1647]
Rfe -RI Rfe _Rfe
2.849 -2.8429j
L Li Li
Rfe -Rfe -R Rfe The rank of the observability matrix of the system is 2. The system dur-
[l
ing magnetizing inrush is observable.
L2 L
Conclusions: This letter has addressed the definitions of state equa-
Rfe Rfe Rfe
tions of a two-winding transformer. As an example study, magnetizing
La Lm La inrush condition is analyzed and monitored in laboratory environment.
It is clear that the whole system is controllable and observable during
magnetizing inrush. If these conditions change, there is an intemal fault

during energizing.
References:
Li
[1] D. Dolinar, J. Pihler, and B. Grcar, "Dynamic model of three
[B] 0
phase power transformer," IEEE Trans. Power Delivery, vol. 8, Oct.
0 1993.

[2] P. Bastard, P. Bertrand, and M. Mevnier, "A transformer model


for winding faults studies," IEEE Trans. Power Delivery, vol. 9, Apr.
1994.

[C] (0 Rload 0) [3] T. Leibfried and K. Feser, "Monitoring of power transformers


using the transfer function method," IEEE Trans. Power Delivery, vol.

14, Oct. 1999.


[DI] [0]. [4] D.J. Wilcox, "Time-domain modeling of power transformers
using modal analysis," IEE Proc. Electric Power Applicat., vol. 144,
Mar. 1997.
The transfer function of the system is defined in (12)
Copyright Statement: ISSN 0282 1724/02/$ 17.00 2002 IEEE.

Manuscript received 23 October 200 1. This paper is published herein in


T(s)= C *inv(S *eye(3) -A) *B±+D (12) its entirety.

50 ~~~~~~~~~~~~~~~~~~~~~~~~~~IEEE Power Engineering Review, October 2002

Вам также может понравиться