EE4031 5 Transient Stability
EE4031 5 Transient Stability
It is necessary that they should maintain synchronism despite the ever present
1 disturbances induced by nature and by man load changes, switching actions, faults and so on.
When the system is subject to some form of disturbance, there is a tendency for
the system to develop forces to bring it to a normal or stable condition. The ability of a system to reach a normal or stable condition after being disturbed is call stability.
Synchronous stability may be divided into two main categories depending upon & '
the magnitude of the disturbance steady-state, dynamic and transient stability.
% $
However in practice the dynamics of rotating machines will be effected even the
increase in load is gradual. Dynamic stability is an extension of steady-state stability where the dynamic effects of synchronous machines and automatic 2 control devices such as governors and voltage regulators are included. The dynamic stability is concerned with small disturbances lasting for a long time and in dynamic stability studies non-linearities are neglected.
KWCn v2.2
&
% 1
When one of the strings is cut, representing the loss of a transmission line, the masses undergo transient oscillations and the forces on the strings uctuate. The system will then either settle down to a new steady-state operating point or additional strings will break, resulting in an even weaker network and eventual
& '
system collapse.
2 Stability Limits
The stability limit is the maximum power that can be transferred in a network
between sources and loads without loss of synchronism.
% $
The steady-state stability limit is the maximum power that can be transferred
without the system becoming unstable under steady-state conditions. 4
The dynamic stability limit is the maximum power that can be transferred without
the system becoming unstable under small disturbances lasting for a long time.
The transient stability limit is the maximum power that can be transferred without
the system becoming unstable when a sudden or large disturbance occurs.
In general, the transient stability limit is the lowest whilst the steady-state limit is
the largest. i.e. transient stability limit < dynamic stability limit < steady-state stability limit
KWCn v2.2
&
% 2
First consider case of two identical unloaded machines that are subsequently
disturbed.
E1 Ic E2 E1 Ic E1 E2 Ic= E1 - E 2 E2
E1 - E 2 2jX
& '
Suppose the machines were initially loaded and equally sharing a load. Apart
from losses, both machines would have a mechanical input equal to the electrical output.
E1 I1 X Ic X I 1 = I2 I2 E1 Ic I2 E2 E1 - E 2 E2 I1
% $
E1
E2
Once again say the same disturbance causing generator 1 to accelerate with
respect to generator 2 were to occur. The effects of this are shown in the above gure. As can be seen, the main effect is to increase the electrical burden on the faster machine and to reduce it on the slower machine. The resulting power unbalance with respect to mechanical input tends to correct the disturbance.
KWCn v2.2
&
The power ow between the generators is called the synchronous power ow.
% 3
In a power system, normally more than two generators operate in parallel. The
machines may be located at different places. A group of machines located at one place may be treated as a single large machine. Also, the machines not connected to the same bus but separated by lines of low reactance, may be grouped into one large machine. 7
The operation of one machine connected in parallel with such a large system
comprising many other machines is of great interest. The capacity of the system is so large that its voltage and frequency may be taken constant. The connection or disconnection of a single small machine on such a system would not affect the magnitude and phase of the voltage and frequency.
& '
% $
jX E
8 where
Infinite Bus
V0
V = phase voltage of the innite bus E = excitation phase voltage of the generator X = total synchronous reactance from the source to innite bus = load angle, i.e. phase angle between V and E EV sin X
and the power transfer P from the generator to the innite bus is given by
KWCn v2.2
&
P =
% 4
= 90o , i.e.
Pmax =
EV X
(3.3.1)
P = Pmax sin
9
Equation (3.3.1) represents the steady-state stability limit of the power system. It is clean from the equation that steady-state stability limit Pmax can be increased by 1. Increasing system voltages E or V : by increasing the excitation 2. Decreasing system reactance X by : use of double-circuit line use of bundled conductors series compensation of the reactance use of machines of low impedances In fact, increasing steady-state stability limit by decreasing reactance is economical
& '
% $
The graphical representation of power received P and the load angle is called the power-angle diagram as shown below. 1. Magnitude of magnitude of . 2. There is a maximum limit to the power 10 that can be extracted from a generator for given E , V and X . 3. The sign of the output power depends on the sign of (Generating/Motoring). 4. There is a stable range of operation from
-Pmax
-180 -90
P (output) depends on
P Pmax
Generator
90 180
Motor
KWCn v2.2
&
transmitted power also increases, i.e. the dP d should be positive. The rate dP d is called the synchronizing power coefcient and is taken as the measure of the stability of a system, i.e. 11 synchronizing power coefcient,
ps =
dP = Pmax cos d
ps > 0, i.e. the synchronizing power coefcient is positive. The steady-state stability limit is reached when ps = 0 and if ps < 0, then system is unstable.
If 0 is the synchronous speed, then synchronizing torque coefcient,
ts =
& '
Pmax ps = cos 0 0
% $
Most of these assumptions are rarely true. The most unsatisfactory one is
regarding slow changes. This assumption is not acceptable for transient and dynamic stability studies.
KWCn v2.2
&
V 0o
% 6
R = Rl X = Xg + X l
and Z
EV sin( + ) P= |Z | EV = sin( + ) |Z |
line.
jXg E
V2 sin |Z | V2 R |Z |
where
sin =
R |Z |
If the line is long enough, it may be necessary to incorporate a T or model for the
y12 V 0o E y11 y22 V 0o
& '
By well known methods of network reduction the circuit can be reduced to a single equivalent as shown in the above gure with all circuit quantities being shown in admittance form. Now, since
% $
I1 I2
Y11 = y12 + y11 ; Y12 = Y21 = y12 and Y22 = y12 + y22 P = EV |Y12 | sin( + ) + E 2 G12 Y12 = |Y12 |/90o and G12 = Re(Y12 )
Y11 Y21
Y12 Y22
E/ V /0
o
KWCn v2.2
&
the ux linkages of the damper circuits and solid rotor iron also.
% 7
Hence for transient and dynamic stability studies a more sophisticated machine model is required.
X
The equivalent circuit and phasor diagram for a cylindrical pole rotor machine are shown. 15
Since Xd
R V
EV P = sin Xd
o / I
E V
IR
IX
& '
5 Transient Stability
The steady-state stability limit does not provide a useful indication because the
actual disturbance that occur in power systems can be very large (faults, loss of generators, line switching, major load changes, etc.) and the system should be operated to be stable through these events.
% $
Different utiilies may specify different criterion disturbances. For example the
criterion disturbance could be a 3-phase short circuit which is cleared by circuit breaker operation in a time corresponding to the system fault clearing time.
KWCn v2.2
&
% 8
3. control loops such as AVRs and governors 4. mechanical factors such as boiler and hydraulic time constants 5. load characteristics
Under normal operations, the relative position of the rotor axis and stator
magnetic eld axis is xed. The angle between the two is known as the load or rotor angle and depends upon the loading of the machine. 17
The problem of stability revolves around the determination of whether or not the torque and rotor angle will stabilize after a sudden disturbance. In case continues to increase after a disturbance the machine will lose synchronism. In order to determine the angular displacement between the generating units of
a power system following a major disturbance, it is necessary to solve a set of differential equations, called the swing equations, describing the motion of the
& '
rotors.
% $
output is the electrical torque Te . If the losses are neglected the difference between the mechanical torque and the electrical torque is equal to the accelerating or decelerating torque Ta , i.e.
Ta = Tm Te
18 If J is the moment of inertia of the rotor and is the angular acceleration, the motion of the rotor can be expressed as:
Ta = Tm Te = J
Let
be the angular position of the rotor at any instant t, be the corresponding angular displacement of the rotor and , s be the angular velocity and synchronous velocity respectively. = s t + d d = s + = dt dt d2 d2 = 2 = 2 dt dt % 9
KWCn v2.2
&
Tm Te = J d2 or Pm Pe = M = M 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (5.1.1) dt where Pm = Tm the mechanical power Pe = Te the electrical power M = J the angular momentum
Equation (5.1.1) is known as the swing equation and describes the behaviour of a synchronous machine during transients. As the angular momentum M is proportional to the angular velocity, it is not constant but varies somewhat during the swings due to variation in . In practice, the change in from the normal system angular velocity s is not much during the swing and it is commonly assumed that M is constant and is equal to Js . This
19
& '
Another important constant denoted by H and is dened as the ratio of the kinetic stored energy in megajoules rating in MVA
% $
energy at rated speed to the rated apparent power of the machine, i.e.
H=
MWs/MVA or MJ/MVA
H is also sometimes called as inertia constant. A relation between M and H is derived as follows: 20
1 GH = energy stored = 1 2 M = 2 M 2f
or
M=
GH f
where G is the rating of the machine in MVA and f is the system frequency. Note M depends upon the size of the machine as well as on its type whereas
H does not vary widely with size and has a characteristic value or set of values
for each class of machines. Substituting M in Equation (5.1.1): or in per unit
KWCn v2.2
&
% 10
where f0 is the rated (synchronous) frequency. Hence for a N multi-machine system, the swing equations are
for i
= 1, 2, ... , N
. . . . . . . . . . . . . . (5.1.3)
Damping has been ignored in the derivation of the above relation. The solution of 21 Equation (5.1.3) provides the values of with respect to time t which when plotted gives the swing curve. Thus, for a multi-machine system, there will be a swing curve for each generator and the transient stability of the system is assessed by observing the angular difference between each pair of generators. If the difference is found to be decreasing after all switching operations are over, the system will be generally stable and the oscillation will be successively damped out. However, it is possible that the units may go out of synchronism after withstanding
& '
Swing curves provide information regarding stability and show any tendency of to oscillate and/or increase beyond the point of return.
Unstable Unstable
% $
22
Stable
First-swing stability
Swing curves are also useful in determining the adequacy of relay protection on power system with regard to the clearing of faults before one or more machines become unstable and fall out of synchronism. The critical clearing time is found to
KWCn v2.2
&
% 11
Example 5.1.1
A 50Hz, 20 MVA, 13.2kV generator has an inertia constant H = 9 MWs/MVA. Determine : a) the K.E. stored in the rotor at synchronous speed. b) the acceleration if the net mechanical input power is 18 MW and the electrical power developed is 15 MW.
23
c) the change in rotor angle for a period of 15 cycles with rotor acceleration being constant. Solution: a) K.E. stored in the rotor in megajoules
= Pm Pe = 18 15 = 3 MW
% $
d2 Pa = where dt2 M
M=
GH 20 9 = = 1.146 MJsec/rad fo 50
KWCn v2.2
&
d = = 2.618 t = 1.309 t2 rad dt 15 = 0.3 sec Period of 15 cycles = 50 Hence, = 1.309 0.32 = 0.11781 rad = 6.75 deg
% 12
25
at standstill with respect to each other or not. This method is known as Equal Area Criterion for stability. The use of this method eliminates partially or wholly the calculation of swing curves which thus saves a considerable amount of work. The Equal Area Criterion method is applicable to any two machine systems which satises the above assumption. It is not applicable to a multi-machine system directly or if the critical clearing time is required to be determined. In such case, step-by-step (or point-by-point) method may be used for numerical solution of swing
& '
equations.
% $
d2 = Pa = Pm Pe dt2 d2 d dt = dt2 dt
2
2M
2(Pm Pe ) (Pm Pe )d
d dt dt
or
d dt
=2
0
or
d = dt
2 M
(Pm Pe )d + constant
0
KWCn v2.2
&
% 13
Also when t = 0, d dt
is zero. The angle will stop changing and the machine will again be operating at synchronous speed after a disturbance when d dt
(Pm 0
Pe )d = 0 =
Pa d
This means that the area under the curve Pa should be zero. This condition is possible only when Pa is both accelerating and decelerating powers. 27 For a generator action
Ps > Pe for positive area A1 and Pe > Ps for negative area A2 for stable
P -ve Ps A2 A1 +ve Pe
operation. Hence the name equal area criterion. The area A1 represents the kinetic energy stored by the rotor during acceleration and the area
A2
0
represents the kinetic energy given up by the rotor to the system and when it is all given up, the
& '
Consider the following sequence of events. A generator working initially and stably at an input = output =
% $
P0 operating point with torque angle 0 is subjected to a sudden increase of input to level P1 . The generator should settle down to new stable operation at P1 , 1 .
1. Rotor begins to accelerate 2. Point
28
balance but the large rotor mass is overspeed 3. The rotor runs past P1 , 1 4. The output commences
P P1 P0
X
a
0 1
5a. At some point such as X , speed has fallen again to synchronous speed s but there is no power balance (P1
> Pe ) % 14
KWCn v2.2
&
6a. Speed falls below s and the rotor comes back downwards P1 , 1
7a. The next time P1 , 1 is reached though power balance and exits again with the rotor mass being underspeed
8a. The rotor runs below P1 , 1 but acceleration commences again 9a. If damping is taken into account, the rotor oscillates about P1 , 1 and settles down to a stable operation Alterative scenario unstable case: 29 5b. The rotor runs past the steady-state stability limit before speed falls to synchronous value 6b. The rotor reaches point Y before speed falls to synchronous value 7b. Output falls below input again 8b. The overspeed rotor begins to accelerate again
& '
In summary, the equal area criterion makes the following assumptions or approximations: 1. Neglects electro-magnetic, electrical (I 2 R) and mechanical damping 2. Neglects control loops such as AVRs and governors 3. Assumes constant power input
% $
None of these are generally true or small. Hence the equal area criterion is 30 pessimistic i.e. over-estimates the extent of the expected swing.
= A2 . A3 .
P P1 A1 P0
A3 A2
KWCn v2.2
&
0 1
% 15
Now what is the maximum value of the P1 such that the system is critically stable, i.e. any attempt to increase P1 beyond this value the system becomes unstable ? Referring to the power-angle curve plot
m = ( 1 ) = 1 .
The value of 1 with this condition is known as critical torque angle c , i.e. c 31 Also the following condition should be satised for A1
m
= A3 :
P1 (m 0 ) =
0
Pmax sin d
& '
may be greater than 90o as long as the equal area criterion is satised.
5.4.2 Sustained network fault Line switching Consider a system represent by a synchronous generator connected to an innite busbar by a double-circuit line. The two circuits are connected in parallel.
Generator Line 1
Line 2
32 If one of the circuits is switched out the system may become unstable in spite of the fact that the load could be supplied over by the other circuit under steady-state conditions. Applying the equal area criterion: Limit of rst swing given by A1
P
A1
A3
P0 P1
A2
= A2 . A3 .
0 1
KWCn v2.2
&
% 16
5.4.3 Network fault subsequently cleared Faulted line Suppose that a fault occurs on one of the two parallel circuits connecting to an innite busbar system.
Generator
Line 1
fault
33
Line 2
The faults produces a transient change which may render the system unstable. However, if the circuit breakers clear the fault in time, it is possible to maintain stability. The maximum value of time allowed for protective gear to operate without loss of stability is called the critical clearing time. The torque angle corresponding to this time is called critical clearing angle. Let cl be the value of at which the fault is cleared.
& '
Observe the inuence of both severity and duration of fault on stability. This also shows the need for fast protection relaying and circuit breaker clearing of faults. On important e.h.v. or h.v. systems, 0.5 to 1 cycle operation of protection and 1 to 3 cycle circuit breaker clearing is now possible to achieve. Let the three power angle curves (PACs) be represented as
% $
Pre-fault PAC
A3
Post-fault PAC Faulted PAC
34
= A3
m cl
0 cl m
r2 Pmax sin d
P0 (m 0 ) =
0
r1 Pmax sin d +
KWCn v2.2
&
% 17
Substituting P0
= Pmax sin 0 and rearranging the above equation into: (m 0 ) sin 0 r1 cos 0 + r2 cos m r2 r1 P0 = Pmax sin 0 = r2 Pmax sin( m ) sin 0 r2
cos cl =
m = sin1
Example 5.4.1
1.2pu j0.3 j0.15 1pu j0.25 j0.15 1pu j0.15
P
Generator j0.15
fault
j0.25
j0.15
Infinite System
All impedances are given in ohms. Determine the critical clearing angle for the
& '
generator for a 3- fault at the point P when the generator is delivering 1 pu power.
% $
Xpre = 0.3 +
0.55 2
+ 0.15 = 0.725 pu
Xf lt = ZA B = = = 3.133 pu
36
before the fault : during the fault : & after the fault :
EV Xpre EV Xf lt
= =
EV Xpost
1.21.0 1.0
KWCn v2.2
0.383 Pmax1 = = 0.23 Pmax 1.667 Pmax2 1.2 r2 = = = 0.72 Pmax 1.667 The rotor angle, 0 , when the system is operating normally r1 = Pmax sin 0 = 1 pu 0 = sin1
1 1.667
37
Pmax2 sin m = 1 pu
m = sin1
cos cl = =
(2.156 0.643) 0.6 0.23 0.8 0.72 0.553 0.72 0.23 cl = 48.35o & '
5.5 Step-by-Step (Point-by-Point) Method
% $
The equal area criterion (EAC) method is useful in determining the critical
clearing angle, i.e. the condition when the system will be stable provided the fault is cleared before the rotor angle exceeds the critical clearing angle.
However the EAC method provides little information on the critical clearing time
which is important for setting up the operation times of the relay and circuit 38 breaker such that the total time taken by them would less than the critical clearing time for stable operation of the system.
Because this method involves the calculations of the rotor angles as time is &
incremented, it is also referred to as the time domain simulation method.
KWCn v2.2
% 19
Numerical integration techniques are used to solve the swing equations. Each swing curve ( t) is then plotted to check its tendency to remain stable.
As the step-by-step method calculates the change in the angular position of the
rotor during a short interval of time (milliseconds), the following assumptions are commonly made during the computational procedure.
1. The accelerating power Pa and the angular acceleration ( d dt2 ) computed at
2
the beginning of an interval is assumed to be constant from the middle of the preceding interval to the middle of the interval under consideration. 2. The angular velecity computed at the middle of an interval remains
& '
Pa Pa(n-2) Pa(n-1) Pa(n)
calculation point
% $
Assumed Actual
w(n)
(n) (n-1)
40
n-2
n-1
time step
n-2
n-1
time step
n-2
n-1
time step
The above assumptions are not strictly correct, since is changing continuously and both Pa and depend upon . However, they are acceptable if time increment t is made small.
5.5.1 Simple Step-by-Step Numerical Integation
KWCn v2.2
&
% 20
The procedure for the n-th interval is outlined below: 1. Evaluate the accelerating power
1 d2 = = Pa(n1) 2. From the swing equation (n) dt2 M where is the acceleration, evaluate (n) .
3. The change in angular velocity for the n-th interval is 41
(n) = (n) t
Hence,
Here is the relative angular velocity and is zero at t 4. The change in rotor angle for the n-th interval is
(n) = (n) t = [(n1) + (n) ]t = (n1) + (n) t = (n1) + (n) (t)2 & '
Hence,
Equation (5.5.1) forms the basis of the numerical solution by step-by-step method. The accelerating power (based on the powers calculated for last interval) is calculated at the beginning of each new interval. In other words, once the value of (n) obtained from this equation, it forms one point on the swing curve and will be used for evaluating the accelerating power Pa(n+1) and the procedure will be repeated during the subsequent intervals.
% $
42 where and
Once the swing curve is plotted and the critical clearing time, corresponding to the critical clearing angle as obtained by using the equal area criterion, can be determined and a suitable protection scheme can be designed to avoid the system
KWCn v2.2
&
% 21
Example 5.5.1
Three-phase fault takes place at B as shown in the gure below with all line impedances being given in . The breakers at A and B operate simultaneously. The generator has an inertia constant H = 4.0 and it is delivering 1.0 pu power before the fault takes place. Use a time step of 0.05 sec.
Generator j0.25 j0.12 1.1pu j0.3 Infinite Bus
43
fault A B
j0.035 1pu
1. Determine the critical fault clearing angle. 2. Determine the critical clearing time. 3. Plot the swing curves both under sustained fault and when the breakers have
& '
Solution : 1. Critical clearing angle calculation: Pre-fault reactance between the buses
% $
when the system is operating normally EV .11.0 Pmax = X =1 0.29136 = 3.775 pu pre during the fault Pmax1 = 0 pu after the breakers operate EV 11.0 Pmax2 = X = 1. 0.405 = 2.716 pu post & r1 =
Pmax1 Pmax
= 0 & r2 =
Pmax2 Pmax
2.716 3.775
= 0.7194
KWCn v2.2
% 22
Pmax sin 0 = 1 pu
Pmax2 sin m = 1 pu
cos cl = =
& '
M=
First interval: Since the power transmitted is 0 once the fault applied while the shaft power is 1, the average value of the power in the rst interval is (case 2 discontinuity):
% $
P a0 =
Or 46
Pa0 + Pa0+ (1.0 1.0) + (1.0 0.0) = = 0.5 pu 2 2 Pa 0 = 0 = 0.5 2250 = 1125 deg/sec2 M 1 = 0 + 1 = 0.0 + 56.25 = 56.25 deg/sec
where 0 is the angular velocity of the rotor with respect to the 50 Hz reference axis and is therefore zero before the fault takes place.
&
KWCn v2.2
Second interval:
Or
Pa1 = 1.0 0.0 = 1.0 pu Pa 1 = 1 = 1.0 2250 = 2250 deg/sec2 M 2 = 1 t = 2250 0.05 = 112.5 deg/sec 2 = 1 + 2 = 56.25 + 112.5 = 168.75 deg/sec
47
Successive intervals:
Pa , and are the same as the second interval. = 281.25 deg/sec = 393.75 deg/sec = 506.25 deg/sec = 618.75 deg/sec = 731.25 deg/sec 3 4 5 6 7 = 40.66 deg = 60.34 deg = 85.65 deg = 116.58 deg = 153.14 deg
3 4 5 6 7 & '
= 2 + 3 = 3 + 4 = 4 + 5 = 5 + 6 = 6 + 7
% $
48
0.2582 0.05 0.1 0.15 0.2 Time (s) 0.25 0.3 0.35
% 24
Hence, taken from the swing plot, the critical clearing time is 0.2582 sec.
KWCn v2.2
3. Swing curve plots: Once the circuit breaker is operated at the end of 0.2 s, the faulty section of the system will be isolate and the system will operate on the power characteristic:
49
Or
Pa4 + Pa4+ 2 (1.0 0.0) + (1.0 2.716 sin 60.34) = = 0.18 pu 2 Pa 4 = 4 = 0.18 2250 = 405 deg/sec2 M 5 = 1 t = 405 0.05 = 20.25 deg/sec P a4 = 5 = 4 + 5 = 393.75 20.25 = 373.5 deg/sec 5 = 5 t = 373.5 0.05 = 18.67 deg 5 = 4 + 5 = 60.34 + 18.67 = 79.01 deg % $
& '
Similarly,
6 7 8 9
160
6 7 8 9
Rotor Angle 140 120 Delta (deg) 100 80 60 40 20 0 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 Time (s)
50
KWCn v2.2
&
% 25
5.5.2 Discontinuities due to Network Changes A change in operating conditions causes a change in the value of Pa hence discontinuity occurs. Possible changes can be the removal of a fault or any line switching operations. Depending on the exact ocurring moment of the discontinuity, there are 3 possible cases. 1. The discontinuity occurs at the middle of an interval this is simplest as no
51
n-2
n-1
& '
Case 1
time step
n-2
n-1
Case 2
time step
2. The discontinuity occurs at the beginning of the interval an average values of the accelerating powers Pa(n) and Pa(n)+ will be used instead, i.e.
1 Pa(n) = 2 (Pa(n) + Pa(n)+ )
% $
where Pa(n) and Pa(n)+ is the accelerating power immediately before and after the change, e.g. clearing the fault. 3. The discontinuity occurs at some other than the beginning or the middle of an 52 interval in this case, a weighted average value of Pa before and after the discontinuity may be used. In practice, a precise evaluation of Pa is not required in general as the time interval used in calculation is so short that it is sufciently accurate to consider the discontinuity as either in case (1) or (2). Alternatively, a variable time step approach can be used to adjust the the time interval at which the discontinuity occurs such that the discontinuity locates exactly
KWCn v2.2
&
% 26
From the swing equation, the acceleration of the rotor is inversely proportional to the inertia constant M of the machine when accelerating power is constant.
This means higher the inertia constant, the slower will be the change in the rotor angle of the machine and thus large the critical clearing time. However, it is uneconomical to improve the transient stability by increasing the inertia constant 53 and is normally not used.
The methods normally used for improving the transient stability are:
1. Higher system voltage an increase in system voltage results in higher value of the steady-state stability limit (Pmax ). The higher the Pmax value, the smaller will be the transmission angle reqired to transfer a given amount of power. This means the greater is the margin between the steady-state transmission angle and the critical clearing angle.
& '
reactance again reducing the series reactance by using parallel lines and/or series capacitors can achieve the same objective of improving the
% $
Pmax .
3. Use of high speed circuit breakers and auto-reclosing breakers the quicker a breaker can operate, the faster the fault will be removed from the system and the smaller will be the change in rotor position. Hence, the better the 54 tendency of the system to restore to normal operating conditions. 5.6.1 Automatic Voltage Regulator (AVR)
A voltage regulator is the heart of the excitation system. The output voltage of
the generator changes only when the voltage regulator instructs the excitation system to do so irrespective of the speed of response of the exciter.
A regulator senses changes in the output voltage and/or current and causes
corrective action to take place. If the regulator is slow, the system will be a poor
KWCn v2.2
&
one.
% 27
The settings and physical limits on the AVR will have a direct impact on the
system performance. With a good setting, both the steady-state and transient stability limits can be improved with the use of AVR.
55
& '
% $
In the short period ( 1s) after the disturbance, the governor and turbine will be
too slow to have any signicant effect on the generator rotor response.
However, the governing effects will kick in and improve the system response as
56 excess mechanical power coming the steam turbine has been reduced by the closing the main steam controlling and interceptor valves.
KWCn v2.2
&
% 28
Fig.1 2. A 50Hz generator with a reactance of 0.3 pu has an inertia constant H = 5. It delivers power of 0.8 pu to an infinite bus through a transmission network with transfer reactance of 0.2 pu. The internal voltage of the generator and the voltage of the infinite bus are maintained at 1.05 pu and 1 pu, respectively. a) Determine the load angle of the generator. b) Determine the steady-state stability limit. c) Determine the synchronizing power coefficient. d) Due to a fault on the transmission network, the transfer reactance is suddenly doubled. Plot the swing curve of the generator for the first 0.2 s after the fault using a time step of 0.05 s. 3. A generator operating at 50 Hz with an inertia constant H of 5 sec delivers half-load (i.e. Po = 0.5 pu) to an infinite bus of voltage 1.0 pu through a transmission link with total reactance XL of 0.4 pu. The internal voltage E and reactance XG of the generator is 1.5 pu and 0.8 pu, respectively. a) Find the power angle o and the steady-state stability limit Pmax. b) A three-phase fault occurs at the generator terminal, find the critical clearing angle cr and the corresponding critical clearing time tcr.
4.
A generator with 0.2 pu reactance is connected to a busbar via a transformer having a reactance of 0.15 pu. The busbar is then connected to a large power system by a doublecircuit line, each circuit of which has a reactance of 1.2 pu. The voltages at both the generator terminal and the large power system is 1.0 pu. The generator has an inertia constant of 5 sec and delivers 0.9 pu real power to the large power system.
_
a) Find the internal voltage and the power angle of the generator. b) Due to malfunction of a relay, one of the circuits in the double circuit line was tripped suddenly and then reclosed after 0.2 sec. Calculate the power angle of the generator when the tripped circuit was reclosed, hence plot its value for the next 0.1 sec using time steps of 0.05 sec. 5. Fig.5 shows a synchronous generator connected to an infinite bus via a double circuit line with reactances and voltages given on a common system base. A transient threephase fault occurs at the mid-point of one of the lines. At the instant preceding the fault, the generator is delivering a steady-state power Pg = 1.0 pu. a) Determine the pre-fault steady-state rotor angle o of the generator. b) The fault is first isolated by switching out the faulty line when the rotor angle increases to 1 = 90o, and is subsequently cleared by the time when the line is reclosed at rotor angle 2 = 120o. Apply the equal area criterion, and determine the transient stability of the post-fault system.
Fig.5 6. A 50Hz generator with a reactance of 0.25 pu delivers a steady-state power of 0.8 pu over a transmission system to an infinite busbar. The steady-state voltages at both the generator terminal and the infinite bus are 1.0 pu. The generator has an inertia constant H of 5 sec, and the effective transfer reactance of the transmission system is 0.4 pu. a) Find the internal voltage and the power angle of the generator. b) A solid three phase fault occurs in the transmission system and is subsequently cleared after 0.15 sec. The effective transfer reactances of the transmission system during and after the fault are 1.0 and 0.6 pu, respectively. Plot the power angle of the generator for the first 0.1 sec after the fault using time steps of 0.05 sec, and hence determine the power angle and power output of the generator after the fault has been cleared for 0.1 sec.
X pre = 0.2 + 0.4 // 0.4 + 0.05 = 0.45 pu The maximum power output:
EV 1.2 8 = = pu X pre 0.45 3 and their ratio P 24 /13 9 r2 = max 2 = = Pmax 8/3 13 The initial rotor angle Pmax =
Pmax 2 =
3 = 0.3844 rad 8
13 = 2.5692 rad 24
cr = cos1
b) power-angle curves
c) Install a braking resistor close to the generator terminals, the amount of energy available to accelerate the rotor can then be significantly reduced, and hence the system stability can be improved.
EE4031, KWCn, 23 Apr 2010 3
2. a)
Pmax =
c)
M=
d)
t (s)
00+ 0avg 0.05 0.10 0.15 0.20
Pmax =
2.1 1.5 1.5 1.5 1.5 1.5
EV X
Pe = Pmax sin
0.8 0.5714 0.5837 0.6192 0.6733 0.7393
Pa = 0.8 Pe
0 0.2286 0.1143 0.2163 0.1808 0.1267 0.0607