Symmetrical Components 2 Sequence impedances Although the following focuses on loads, the results apply equally well to lines,

or lines and loads. Read these notes together with sections 12.6 and 12.9 of text. Consider the Y-connected balanced load, Fig. 1. It is grounded through an impedance Zn. Because of this, the neutral point may not be at the same potential as the ground.
Ia Vag Ic ZY Vcg ZY In Zn

Ib

ZY

Vbg

Fig. 1
1

Lets use KVL to write the voltage equations for the three phase-to-ground voltages Vag, Vbg, and Vcg as a function of the line currents and the load impedances. Assume no mutual coupling between phases. Vag = Z Y I a + Z n I n

Vbg = Z Y I b + Z n I n Vcg = Z Y I c + Z n I n

(1)

We can replace In if we apply KCL at the junction node at the center of the Y. I n = Ia + Ib + Ic (2) Substitution of (2) into (1) yields: Vag = Z Y I a + Z n ( I a + I b + I c )
Vcg = Z Y I c + Z n ( I a + I b + I c )
Vbg = ZY I b + Z n ( I a + I b + I c )

(3)

Expanding the Zn through and then collecting terms with like currents yields: Vag = ( Z Y + Z n ) I a + Z n I a + Z n I b
Vcg = Z n I a + Z n I b + ( Z Y + Z n ) I c
2

Vbg = Z n I a + ( Z Y + Z n ) I b + Z n I c

(4)

Now lets write these voltage equations in matrix form.

Vag Z Y + Z n Vbg = Z n Vcg Z n Zn ZY + Z n Zn Z n I a Z n I b (5) ZY + Z n I c

In compact notation, eq. (5) is:

V abc = Z abc I abc

(6)

Now I pose the following question. If we can somehow find a way to transform eq. (5) into an equation that relates sequence voltages VS on the LHS to sequence currents IS on the RHS, what will the impedance matrix look like? To answer this question, we need to derive eq. (7). VS = ZS IS (7)

We refer to the impedance matrix, ZS, that relates sequence voltages to sequence currents, as the sequence impedance matrix. To derive eq. (7), consider what we have: eq. (6). Recall that Vabc=AVS and Substituting into eq. (6) yields: Iabc=AIS.

(8) Now pre-multiply both sides by A-1. This is:

AV S = Z abc A I S
1 1

A AV S = A Z abc AI S
The left hand side is just VS.

(9) (10)

V S = A Z abc A I S

Comparison of (10) with (7) indicates that the sequence impedance matrix, ZS, is given by

Z S = A Z abc A

(11)

So what does ZS look like? We know all three elements of eq. (11) so why dont we do the matrix math and find out
Z S = A Z abc A 1 ZY + Z n 1 1 1 = 1 2 Z n 3 2 1 Zn Zn ZY + Z n Zn 1 1 1 Z n 1 2 2 ZY + Z n 1 Zn ZY Z Y 2 ZY 0 ZY 0 0 0 ZY
1

Multiplying the two right-hand matrices:

0 ZS 1 Z Y + 3Z n ZY 1 1 + 1 Z S = Z S = 1 2 ZY + 3Z n 2 ZY 3 2 Z S 1 Z Y + 3Z n Z Y

Now multiply the remaining matrices:

0 ZS 3ZY + 9Z n + 1 Z S = Z S = 0 3 Z S 0

0 3ZY 0

0 Z Y + 3Z n 0 = 0 3ZY 0

Plugging this expression into eq. (10)

V S = A Z abc A I S = Z S I S

0 Vag ZY + 3Z n + 0 Vag = Vag 0

0 ZY 0

0 0 Ia I + 0 a (12) ZY I a

Now this is an amazing thing all off-diagonal terms are zero! What does this mean? It means that the only current that determines the zero sequence voltage is the zero sequence current. the only current that determines the positive sequence voltage is the positive sequence current. the only current that determines the negative sequence voltage is the negative sequence current. This is the case whenever the impedance matrix is diagonal, with off-diagonals all 0.
6

We say that the three equations represented by the matrix relation are uncoupled in that no variable (current) appears in more than one equation. So these 3 uncoupled equations are:

V = ( ZY + 3Z n ) I
0 ag + ag ag

0 a

V =Z I V =Z I

+ Y a Y a

(13)

The really nice thing about these 3 equations is that they represent 3 separate and distinct SINGLE PHASE CIRCUITS!!!! Therefore we can just apply EE 303 perphase analysis to analyze them. Fig. 2 illustrates the single phase circuits.

Va0

0 Ia

ZY 3Zn

Z0=ZY+3Zn

Zero sequence network

+ a

+ Ia
Positive sequence network

ZY

+ Z S = ZY

Va

Ia
Negative sequence network

ZY

Z S = ZY

Fig. 2

Some questions: 1. Why doesnt the neutral impedance appear in the positive & negative sequence networks? Because the positive and negative sequence networks contain balanced currents only, and balanced currents sum to 0 and therefore do not contribute to flow in the neutral.
2.

Why do we have 3Zn in the zero sequence network instead of just Zn? Recall IA+IB+IC=In. We defined
0 Ia =

1 In . 3

So I n = 3I a0 actually flows in the Zn. But our zero sequence network has only I a0 flowing. Therefore, to obtain the correct voltage drop seen in the neutral conductor with a flow of only I a0 , we model the zerosequence impedance as 3Zn. Then the voltage drop is 3I a0 Z n , as it should be .
9

3.

What do these three networks look like if the neutral is solidly grounded (no neutral impedance)? Positive and negative sequence networks are the same. Zero sequence is the same except Zn=0.

4. What do these three networks look like if the neutral is ungrounded (floating)? Positive and negative sequence networks are the same. Zero sequence has an open circuit, which means I n = I a0 = 0 .
5.

What is benefit of the SC transformation?

Answer: If the load (or line, or load and line) is symmetric (so that ZS is diagonal), then the three networks will decouple and we can analyze an unbalanced situation with three separate per-phase analyses.
10

What if the load (or line, or load and line) is not symmetric? Look at this case closely.
6.

Consider a general a-b-c impedance matrix as given below (consistent with the matrix of eq. 3.40, pg. 473, in text)
Z abc Z aa = Z ab Z ac Z ab Z bb Z bc Z ac Z bc Z cc

(14)

This is a general impedance matrix in that The self-terms Zaa, Zbb, and Zcc differ. There exist non-0 mutuals Zab, Zac, and Zbc The mutuals differ The above occurs when the phase geometry is not equilateral & transposition is not used. Recall eq. (11), convenience:
1

repeated

here

for (11)

Z S = A Z abc A

Then we compute ZS using the a-b-c impedance matrix of eq. (14):

11

0 ZS + 1 Z S = A Z abc A = Z S 0 Z S 0

0 ZS+ + ZS ZS +

0 ZS + ZS ZS

1 Z aa Z ab Z ac 1 1 1 1 1 1 = 1 2 Z ab Z bb Z bc 1 2 3 2 2 1 Z ac Z bc Z cc 1 We will not go through the detailed matrix multiplication here but instead will just provide the expressions for each of the 9 terms in the ZS matrix, as follows:
1 ( Z aa + Zbb + Z cc + 2Z ab + 2Z ac + 2Z bc ) 3 1 + Z S = Z S = ( Z aa + Z bb + Z cc Z ab Z ac Z bc ) 3
0 ZS =
0 ZS+ = ZS 0 =
0 + ZS = ZS 0

(13) (14)

+ ZS = ZS +

1 ( Z aa + 2 Zbb + Z cc + 2Z ab + 2 2 Z ac + 2Zbc ) (17) 3 1 = Z aa + Z bb + 2 Z cc + 2 2 Z ab + 2Z ac + 2 Z bc (18) 3

1 Z aa + 2 Z bb + Z cc Z ab 2 Z ac Z bc (15) 3 1 = ( Z aa + Z bb + 2 Z cc 2 Z ab Z ac Z bc ) (16) 3

12

Note eq. (15) indicates Z S0+ = Z S0 and from eq. (16) that Z S0 = Z S+0 . The Zs matrix is, in general, not symmetric. For our 0+- circuits to be decoupled (and thus obtain the advantage of symmetrical component decomposition see question 5 above), the off-diagonal elements of ZS must be 0. So what are the conditions for the offdiagonal elements of ZS to be 0? We can obtain these conditions by setting eqs. (15)-(18) to 0 and solving them simultaneously. We will not go through this pain here. Rather, you should be able to inspect eqs. (15)-(18) and notice that for them to be 0, it must be true that

13

(19) (Diagonal phase impedances must be equal) Z ab = Z ac = Z bc (20) (Offdiagonal phase impedances must be equal). Under the conditions imposed by eqs. (19) and (20), it will be the case that the 0+impedances are given by 0 + 0 0 + Z S + = Z S 0 = Z S = Z S = Z S = Z S + = 0 (21)

Z aa = Z bb = Z cc

Z = Z aa + 2 Z ab
0 S

(22) (23)

Z = Z = Z aa Z ab

+ S

Check it for the example we worked above corresponding to Fig. 1:

Z abc ZY + Z n = Zn Zn Zn ZY + Z n Zn ZY + Z n Zn Zn

14

 ZY + 3Z n ZS = 0 0

0 ZY 0

0 0 ZY

Some additional observations from this work: By eq. (14), the positive and negative sequence impedances are always equal, independent of whether the load is symmetric or not. This is true for transmission lines, cables, and transformers1. 0 The 0-sequence impedance Z S is not equal to the positive and negative sequence impedances of a symmetrical load unless the mutual phase impedances Zab=Zac=Zbc are 0.

It is not true for rotating machines because positive sequence currents, rotating in the same direction as the rotor, produce fluxes in the rotor iron differently than the negative sequence currents which rotate in the opposite direction as the rotor.
1

15