ECEN 667

Power System Stability

Lecture 7: Synchronous Machine Modeling

Prof. Tom Overbye

Dept. of Electrical and Computer Engineering
Texas A&M University
[email protected]
 Announcements
• Read Chapter 3
• Homework 2 is due on Thursday September 19

2
 Sinusoidal Steady-State
Va  2Vs cos  st   vs 
Here we consider the
2
Vb  2Vs cos  st   vs   application to balanced,
 3  sinusoidal conditions
2 
Vc  2Vs cos  st   vs  
 3 
I a  2I s cos  st   is 
 2 
Ib  2I s cos  st   is  
 3 
 2 
I c  2I s cos  st   is  
 3 

3
 Simplifying Using d
• Define 
P
 shaft   st If we know d, then
2 we can easily relate
the phase to the dq
• Hence Vd  Vs sin     vs  values!
Vq  Vs cos    vs 
I d  I s sin     is 
I q  I s cos    is 
• These algebraic
equations can be  Vd  jVq  e j     / 2
 Vse
j vs

written as complex
equations  Id  jIq  e j     / 2
 I se
j is

4
 Summary So Far
• The model as developed so far has been derived
using the following assumptions
– The stator has three coils in a balanced configuration,
spaced 120 electrical degrees apart
– Rotor has four coils in a balanced configuration located
90 electrical degrees apart
– Relationship between the flux linkages and currents must
reflect a conservative coupling field
– The relationships between the flux linkages and currents
must be independent of qshaft when expressed in the dq0
coordinate system

5
Assuming a Linear Magnetic Circuit
• From the book we have
Note that the
first three
matrices depend
upon qshaft; the
rotor self-
inductance
matrix Lrr is
independent of
qshaft

6
Assuming a Linear Magnetic Circuit
• With this assumption of a linear magnetic circuit
then we can write
 a     ia 
     ib 

 b   Lss  shaft  
Lsr  shaft   
 c     ic 
    i 
 fd      fd 
 1d     i1d 

   Lrs  shaft  
Lrr  shaft   i 
 1q     1q 
2 q    i2 q 
7
 Conversion to dq0 for Angle
Independence

 d   id 
  i 
 q   q 
 o   1   io 
Tdqo Lsr 
  Tdqo LssTdqo i 
 fd      fd 
 1d     i1d 
1
   LrsTdqo Lrr   
 1q     i1q 
2 q  i2 q 

8
 Conversion to dq0 for Angle
Independence
d   Ls  Lmd  id  Lsfd i fd  Ls1d i1d 3
Lmd   LA  LB  ,
3 2
 fd  Lsfd id  L fdfd i fd  L fd 1d i1d 3
2 Lmq   LA  LB 
3 2
1d  Ls1d id  L fd 1d i fd  L1d 1d i1d
2
For a round rotor
q   Ls  Lmq  iq  Ls1qi1q  Ls 2q i2q machine LB is small
3 and hence Lmd is
1q  Ls1qiq  L1q1qi1q  L1q 2 qi2 q
2 close to Lmq. For a
3 salient pole machine
2 q  Ls 2 qiq  L1q 2 qi1q  L2 q 2 q i2 q
2 Lmd is substantially
larger
o  Lsio
9
 Convert to Normalized at f = ws

• Convert to per unit, and assume frequency of w s

• Then define new per unit reactance variables
s Ls s Lmd s Lmq
X s  , X md  , X mq 
Z BDQ Z BDQ Z BDQ
s L fdfd s L1d 1d  s L fd 1d Lsfd
X fd  , X 1d  , X fd 1d 
Z BFD Z B1D Z BFD Ls1d
s L1q1q s L2q2q  s L1q2q Ls1q
X 1q  , X 2q  , X 1q2q 
Z B1Q Z B2Q Z B1Q Ls 2q
X fd  X fd  X md , X 1d  X 1d  X md
X 1q  X 1q  X mq , X 2q  X 2 q  X mq
X d  X s  X md , X q  X s  X mq
10
 Key Simulation Parameters

• The key parameters that occur in most models can

then be defined as
1 2
X md These values
X d  X s   Xd  will be used in
1 1 X fd
 all the
X md X fd
synchronous
1
2
X mq machine models
X q  X s   Xq 
1 1 X1q

X mq X 1q
In a salient rotor machine
Xfd X 1q Xmq is small so Xq = X'q;
 
Tdo  
, Tqo
s R fd s R1q also X1q is small so
T'q0 is small
11
Key Simulation Parameters

12
Example Xd/Xq Ratios for a
WECC Case

13
Example X'q/Xq Ratios for a
WECC Case

About 75% are Clearly Salient Pole Machines! 14

 Internal Variables
• Define the following variables, which are quite
important in subsequent models
X md Hence E'q and E'd are
Eq   fd scaled flux linkages
X fd and Efd is the scaled
field voltage
X mq
Ed  1q
X1q
X md
E fd  V fd
R fd

15
 Dynamic Model Development
• In developing the dynamic model not all of the
currents and fluxes are independent
– In this formulation only seven out of fourteen are
independent
• Approach is to eliminate the rotor currents,
retaining the terminal currents (Id, Iq, I0) for
matching the network boundary conditions

16
 Rotor Currents

• Use new variables to solve for the rotor currents

 d   X d I d 
 X d  X s 
Eq 
 X d  X d 
1d
 X d  X s   X d  X s 
1
I fd   Eq   X d  X d   I d  I1d  
X md
X d  X d
I1d     X d  X s  I d  Eq 
2  1d
 X d  X s 

17
 Rotor Currents

 q   X q I q 
 X q  X s 
Ed 
 X q  X q 
 2q
 X q  X s   X q  X s 
1 
I1q   E    X q  X q   I q  I 2 q  
X  
d
mq

X q  X q
I 2q  2   
 2 q  X q  X s I q  Ed 

 X q  X s 
 o  X s   I o 

18
 Final Complete Model

These first three equations

define what are known
as the stator transients; we
will shortly approximate
them as algebraic constraints

dEq  X d  X d 

Tdo   Eq   X d  X d   I d  
1d   X d  X s  Id  Eq    E fd
dt   X d  X s  2

 X q  X q 
dEd

Tqo   
  Ed  X q  X q I q   
 X q  X s  
I q  Ed 
 
2q
dt  X q  X s
2 
 

19
 Final Complete Model
d 1d

Tdo   1d  Eq   X d  X s  I d
dt
d 2 q TFW is the friction

Tqo
dt
 
  2 q  Ed  X q  X s I q
and windage
d component
   s
dt
2 H d
 s dt
 
 TM   d I q   q I d  TFW

 d   X d I d 
 X d  X s 
Eq 
 X d  X s 
1d
 X d  X s   X d  X s 
 q   X qI q 
 X q  X s 
Ed 
 X q  X q 
 2q
 X q  X s   X q  X s 
 o   X s I o
20
 Single-Machine Steady-State
0  Rs I d   q  Vd   s   d  Eq  X d I d
0  Rs I q  d  Vq
 q   X qI q  Ed
0  Rs I o  Vo
0   Eq   X d  X d  I d  E fd
 o   X s I o

0   1d  Eq   X d  X s  I d The key variable

we need to
 
0   Ed  X q  X q I q determine the
initial conditions

0   2 q  Ed  X q  X s I q  is actually d, which
0    s doesn't appear
explicitly in these
 
0  Tm   d I q   q I d  TFW equations!
21
 Field Current

• The field current, Ifd, is defined in steady-state as

I fd  E fd / X md

• However, what is usually used in transient stability

simulations for the field current is the product

• So the value of Xmd is not needed

22
 Single-Machine Steady-State

• Previous derivation was done assuming a linear

magnetic circuit
• We'll consider the nonlinear magnetic circuit later but
will first do the steady-state condition (3.6)
• In steady-state the speed is constant (equal to ws), d is
constant, and all the derivatives are zero
• Initial values are determined from the terminal
conditions: voltage magnitude, voltage angle, real and
reactive power injection

23
 Determining d without Saturation

• In order to get the initial values for the variables we

need to determine d
• We'll eventually consider two approaches: the simple
one when there is no saturation, and then later a
general approach for models with saturation
• To derive the simple approach we have

Vd  Rs I d Ed  X q I q
Vq   Rs I q Eq  X d I d

24
 Determining d without Saturation

Since j  e  
j  /2

 
E   X q  X d I d  Eq  e j
 
• In terms of the terminal values

E  Vas  ( Rs  jX q ) Ias
The angle on E  

25
 D-q Reference Frame

• Machine voltage and current are “transformed” into

the d-q reference frame using the rotor angle, 
• Terminal voltage in network (power flow) reference frame
are VS = Vt = Vr +jVi

Vr   sin  cos   Vd 

V     cos   V 
sin    q 
 i 
Vd   sin   cos   Vr 
V      
 q  cos  sin   Vi 

26
 A Steady-State Example

• Assume a generator is supplying 1.0 pu real power at

0.95 pf lagging into an infinite bus at 1.0 pu voltage
through the below network. Generator pu values are
Rs=0, Xd=2.1, Xq=2.0, X'd=0.3, X'q=0.5

Transient Stability Data Not Transferred

Bus 1 Bus 2
X12 = 0.20
Bus 4
XTR = 0.10 Infinite Bus

slack

1.095 pu
Bus 3
X13 = 0.10 X23 = 0.20
Delta (Deg): 52.08
Speed (Hz): 60.00
P: 100.00 MW Angle = 6.59 Deg Angle = 0.00 Deg
Eqp: 1.130
Edp: 0.533

27
 A Steady-State Example, cont.

• First determine the current out of the generator from

the initial conditions, then the terminal voltage

I 1.0526  18.20  1  j 0.3288

Vs  1.00   j 0.22   1.0526  18.20 
 1.094611.59  1.0723  j 0.220

28
 A Steady-State Example, cont.
• We can then get the initial angle and initial dq values
E  1.094611.59   j 2.0   1.052  18.2   2.81452.1
   52.1
Vd  0.7889 0.6146  1.0723 0.7107 
V        
 q  0.6146 0.7889  0.220   0.8326 
 I d  0.7889 0.6146   1.000  0.9909
I        
 q  0.6146 0.7889  0.3287   0.3553 

Vd  jVq  Vs e j e j ( /2 )  1.0945(11.6  90  52.1)

 1.094549.5  0.710  j 0.832

29
 A Steady-State Example, cont
• The initial state variable are determined by solving
with the differential equations equal to zero.
Eq'  Vq  Rs I q  X d I d  0.8326   0.3  0.9909   1.1299
Ed'  Vd  Rs I d  X q I q  0.7107  (0.5)(0.3553)  0.5330
E fd  Eq'  ( X d  X d ) I d  1.1299  (2.1  0.3)(0.9909)  2.9135

30
 Single Machine, Infinite Bus
System (SMIB)

Usually infinite bus

angle, q vs, is zero

This example can be simplified by combining machine

values with line values     
de d ed
X de  X d  X ep
etc
Rse  Rs  Re
31
 Introduce New Constants

t  Ts     s  “Transient Speed”

2H Mechanical time
Ts 
s constant

1
 A small parameter
s

We are ignoring the exciter and governor for now;

they will be covered in detail later
32
 Stator Flux Differential Equations
d de   
  Rse I d  1  t  qe  Vs sin     vs 
dt  Ts 

d qe   
  Rse I q   1  t  de  Vs cos     vs 
dt  Ts 

d oe
  Rse I o
dt

33
 Elimination of Stator Transients
• If we assume the stator flux equations are much faster
than the remaining equations, then letting e go to zero
allows us to replace the differential equations with
algebraic equations
This assumption
0  Rse I d   qe  Vs sin     vs  might not be valid if
we are considering
faster dynamics on
0  Rse I q   de  Vs cos     vs 
other devices (such as
converter dynamics)
0  Rse I o

34
 Impact on Studies

Stator transients are not usually considered

in transient stability studies
Image Source: P. Kundur, Power System Stability and Control, EPRI, McGraw-Hill, 1994

35
 Machine Variable Summary
• Three fast dynamic states, now eliminated
 de ,  qe ,  oe
We'll get
• Seven not so fast dynamic states to the
Eq , 1d , Ed , 2 q ,  , t E fd exciter
and
• Eight algebraic states governor
I d , I q , I o ,Vd ,Vq ,Vt , ed , eq shortly

Vt  Vd2  Vq2
Vd  Re I d  X ep I q  Vs sin     vs 
Vq  Re I q  X ep I d  Vs cos     vs 
36
 Network Expressions

Vd  Re I d  X ep I q  Vs sin     vs 
Vq  Re I q  X ep I d  Vs cos     vs 
These two equations can be written as one complex
equation.

Vd  jVq e j    2    Re  jX ep  I d  jI q e j    2 
 Vs e j vs

37
 Machine Variable Summary

Three fast dynamic states, now eliminated

 de ,  qe ,  oe

Seven not so fast dynamic states

We'll get
Eq , 1d , Ed , 2 q ,  , t E fd to the
exciter
and
Eight algebraic states
governor
shortly
I d , I q , I o ,Vd ,Vq ,Vt , ed , eq

38
 Stator Flux Expressions

 I d 
 de   X de
 X d  X s 
Eq 
 X d  X d 
1d
 X d  X s   X d  X s 

 I q 
 qe   X qe
 X q  X s 
Ed 
 X q  X q 
 2q
 X q  X s   X q  X s 
 oe   X oe I o

39
Subtransient Algebraic Circuit


 X q  X s  
X q  X q  
 Ed  
 2 q  X q  X d I q  

 X q  X s
  
X q  X s  


  X d  X s   X d  X d    j    2 
 j Eq  1d   e
  X d  X s   X d  X s   
40
 Network Reference Frame
• In transient stability the initial generator values are
set from a power flow solution, which has the
terminal voltage and power injection
– Current injection is just conjugate of Power/Voltage
• These values are on the network reference frame,
with the angle given by the slack bus angle
V j  Vr , j  jVi , j or V j  VDj  jVQj
• Voltages at bus j converted to d-q reference by
Vd , j   sin   cos   Vr , j  Vr , j   sin  cos   Vd , j 
V      V     
 q , j  cos  sin   Vi , j   i , j    cos  sin   Vq , j 

41
 Network Reference Frame
• Issue of calculating d, which is key, will be
considered for each model
• Starting point is the per unit stator voltages
Vd   q  Rs I d
Vq   d   Rs I q
Equivalently,  Vd +jVq   Rs  I d +jI q      q  j d 
• Sometimes the scaling of the flux by the speed is
neglected, but this can have a major solution
impact
• In per unit the initial speed is unity

42
 Simplified Machine Models
• Often more simplified models were used to
represent synchronous machines
• These simplifications are becoming much less
common but they are still used in some situations
and can be helpful for understanding generator
behavior
• Next several slides go through how these models
can be simplified, then we'll cover the standard
industrial models

43
 Two-Axis Model
• If we assume the damper winding dynamics are
sufficiently fast, then T"do and T"qo go to zero, so there
is an integral manifold for their dynamic states
 1d  Eq   X d  X s  I d
 2 q   Ed   X q  X s  I q

44
 Two-Axis Model
d 1d

Tdo   1d  Eq   X d  X s  I d  0
dt
dEq Note this entire

Tdo   Eq   X d  X d   term becomes zero
dt
 X d  X d 
 Id   
 1d   X d  X s  I d  Eq   E fd
  X d  X s  2

Which can be simplified to
dEq

Tdo   Eq   X d  X d  I d  E fd
dt
45
 Two-Axis Model
d 2 q

Tqo
dt
 
  2q  Ed  X q  X s I q  0

dEd Likewise this entire


Tqo
dt

  Ed  X q  X q   term becomes zero
 q  X q 
X
I   X   X 
I  E 
 
 q
 
2q q s q d
X q  X s
2 
 
Which can simplified to
dEd

Tqo
dt

  Ed  I q X q  X q 
46
 Two-Axis Model

0   Rs  Re  I d   X q  X ep  I q  Ed  Vs sin     vs 

0   Rs  Re  I q   X d  X ep  I d  Eq  Vs cos    vs 

47
 Two-Axis Model
dEq

Tdo   Eq   X d  X d  I d  E fd No saturation
dt
dEd effects are

Tqo
dt

  Ed  X q  X q I q  included with
d this model
   s
dt
2 H d
s dt
 
 TM  Ed I d  Eq I q  X q  X d I d I q  TFW

 
0   Rs  Re  I d  X q  X ep I q  Ed  Vs sin     vs 

0   Rs  Re  I q   X d  X ep  I d  Eq  Vs cos     vs  \
Vd  Re I d  X ep I q  Vs sin     vs 
Vq  Re I q  X ep I d  Vs cos     vs 

Vt  Vd2  Vq2 48
 Example (Used for All Models)
• Below example will be used with all models. Assume
a 100 MVA base, with gen supplying 1.0+j0.3286
power into infinite bus with unity voltage through
network impedance of j0.22
– Gives current of 1.0 - j0.3286 = 1.0526-18.19 
– Generator terminal voltage of 1.072+j0.22 = 1.0946 11.59 
Sign convention
Bus 1 Bus 2
Bus 4
XTR = 0.10
X12 = 0.20
Infinite Bus
on current is out
slack
of the generator is
Bus 3
positive
100.00 MW 1.0946 pu -100.00 MW
X13 = 0.10 X23 = 0.20
57.24 Mvar 11.59 Deg -32.86 Mvar

1.0463 pu 1.0000 pu
6.59 Deg 0.00 Deg

49
 Two-Axis Example
• For the two-axis model assume H = 3.0 per unit-
seconds, Rs=0, Xd = 2.1, Xq = 2.0, X'd= 0.3, X'q = 0.5,
T'do = 7.0, T'qo = 0.75 per unit using the 100 MVA base.
• Solving we get
E  1.094611.59   j 2.0   1.0526  18.19   2.8152.1
   52.1
Vd  0.7889 0.6146  1.0723 0.7107 
V        
 q  0.6146 0.7889  0.220   0.8326 
 I d   0.7889 0.6146   1.000  0.9909 
I        
 q  0.6146 0.7889  0.3287   0.3553 
50
 Two-Axis Example
• And Eq  0.8326   0.3  0.9909   1.130
Ed  0.7107  (0.5)(0.3553)  0.533
E fd  1.1299  (2.1  0.3)(0.9909)  2.913
Saved as case B4_TwoAxis
• Assume a fault at bus 3 at time t=1.0, cleared by
opening both lines into bus 3 at time t=1.1 seconds
Gen Bus 4 #1 Rotor Angle
74

72

70
68

66

64
Gen Bus 4 #1 Rotor Angle

62

60

58

56
54

52

50

48

46
44

0 1 2 3 4 5
Time
Gen Bus 4 #1 Ro to r An g le
51
 Two-Axis Example
• PowerWorld allows the gen states to be easily stored

Gen Bus 4 #1 Machine State\Edp

0.56
Gen Bus 4 #1 Machine State\Edp

Graph shows
0.54
0.52
0.5
0.48
0.46
variation in
0.44
0.42 Ed’
0 1 2 3 4 5
Time
Gen Bus 4 #1 Machin e State\Ed p

52
 Flux Decay Model

• If we assume T'qo is sufficiently fast that its

equation becomes an algebraic constraint
dEd
Tqo   Ed   X q  X q  I q  0 This model
dt assumes that
dEq Ed’ stays constant.
Tdo   Eq   X d  X d  I d  E fd
dt In previous example
d Tq0’=0.75
   s
dt
2H d 
 TM  Ed I d  Eq I q   X q  X d  I d I q  TFW
s dt
 TM   X q  X q  I q I d  Eq I q   X q  X d  I d I q  TFW
 TM  Eq I q   X q  X d  I d I q  TFW
53
 Rotor Angle Sensitivity to Tqop
• Graph shows variation in the rotor angle as Tqop is
varied, showing the flux decay is the same as Tqop = 0

54
 Classical Model
The classical
model had
been widely
used because
it is simple.
At best it
d can only
   s
dt approximate
0
a very short
2 H d E  Vs
0
 TM  sin     vs   TFW term response.
 0 dt X d  X ep It is no longer
common.
This is a pendulum model

55
 Classical Model Justification
• It is difficult to justify. One approach would be to
go from the flux decay model and assume
X q  X d  
Tdo
E   Eq  0  0
• Or go back to the two-axis model and assume
X q  X d  
Tdo  
Tqo
( Eq  const Ed  const)
02 02
E  Eq  Ed

0 1  E 0 
   tan  q
0   2
 Ed 
56
 Classical Model Response
• Rotor angle variation for same fault as before
Gen Bus 4 #1 Rotor Angle Notice that
even though
48
46
44
42
40
the rotor
angle is
38
36
Gen Bus 4 #1 Rotor Angle

34
32
30
quite different,
28
26 its initial increase
24
22
20
(of about 24
18
16 degrees) is
similar. However
14
12
10
8
0 1 2 3 4 5
there is no
Time
Gen Bus 4 #1 Rotor Angle damping.

Saved as case B4_GENCLS

57