Lecture-12

Automatic Load Frequency Control (ALFC)

1. Introduction
 The different nomenclatures are used in literature such as Load Frequency
Control, Power Frequency Control, Real Power Frequency Control or
Automatic Generation Control.
 The basic role of ALFC is:
o To maintain the desired megawatt output power of a generator
matching with the changing load.
o To assist in controlling the frequency of larger interconnection.
o To keep the net interchange power between pool members, at
the predetermined values.
 In power systems, the active and reactive power demand change
continuously with increase or decrease in load.
 In order to meet the active power demand, input to the generators must
be regulated to maintain the speed and hence frequency for which governor
controller is used.
 In case of turbo-generators steam is regulated andin case of hydro
generators water is regulated.
 In order to meet the reactive power demand, excitation of generators
must be continuously regulated to maintain the voltage within limits. The
excitation controller is used to maintain the voltage within limits by
controlling the excitation.
 Thus, generator controllers are:
o Governor Systemfor real power control and hence
speed/frequency control
o Excitation System for voltage and reactive power control
 The excitation control is fast-actingas its major time constant is
encountered due to generator field.
 Governor control is slow actingits major time constant is contributed by
turbine and generator moment of inertia.
 Two controllers are analyzed by decoupling them because of high difference
in time constant.
2. P-f and Q-V Control Loops
In order to perform voltage and frequency control, a basic generator will have two
control loops namely:

1) Automatic voltage regulator loop and

2) Automatic load frequency control loop.
 The automatic voltage regulator (AVR) loop controls the magnitude of terminal
voltage, |V|. Terminal voltage is continuously sensed, rectified and smoothed.
The strength of this dc signal, being proportional to |V|, is compa
compared with a dc
reference |V|ref. The resulting “error voltage” after amplification and signal
shaping, serves as input to the exciter, which applies the required voltage to the
generator field winding, so the generator terminal voltage |V| reaches the value
|V|ref.

The automatic load frequency control (ALFC) loop regulates the real power output
of the generator and its frequency (speed). This loop is not a single one as in the
case of AVR. It has two lops.

1) A relatively fast primary loop responds to a frequency

frequency (speed) changes via
the speed governor and the steam (or hydro) flow is regulated with the aim
of matching the real power generation to relatively fast load fluctuations.
By “fast” mean changes that take place in one to several seconds. Thus,
aiming to maintain a megawatt balance, this primary loop performs a course
speed or frequency control.
2) A slower secondary loop maintains the fine adjustment of the frequency,
and also maintains proper real power interchange with other pool members.
This loop is insensitive
ensitive to rapid load and frequency changes, but focuses
drift-like
like changes which take place over periods of minutes.

Fig. 1 ALFC and AVR Loops

3. Real Power Control of Generator

Fig. 2 Speed Governor Control of Generating Unit

The realpower control mechanism

m of a generator having four parts as shown
below.

1) Speed changer
2) Speed governor
3) Hydraulic amplifier
4) Control valve.

Fig.1: Functional diagram of real power control mechanism of generator

Movement of Linkages corresponding to raise/lower or speed
drops/increase

 Corresponding to “raise” command, linkage movements will be:

o “A” moves downwards; “C” moves downward; “D” moves
upwards; “E” moves downwards.
o This allows more steam or water flow into the turbine resulting
incremental increase in generator output power.
 When the speed drops, linkage point “B” moves upwards and again
generator output power will increase.

 Corresponding to “lower” command, linkage movements will be:

o “A” moves upwards; “C” moves upwards; “D” moves downward;
“E” moves upwards.
o This allows lesser steam or water flow into the turbine resulting
incremental decrease in generator output power.
 When the speed increases, linkage point “B” moves downwards and
again generator output power will decrease.

4. Modelling of speed governor

The movements are assumed positive in the directions of arrows.
Output : The output commend of speed governor is ΔPg which
corresponds to movement ΔxC.
Input: The speed governor has two inputs:
1) Change in the reference power setting, ΔPref
2) Change in the speed of the generator, Δf, as measured by
ΔxB.
It is to be noted that a positive ΔPref will result in positive ΔPg
A positive Δf will result in linkage points B and C to come down causing negative
ΔPg. Thus
ΔPg = ΔPref– (1/R)Δf (1)
Where, the constant R has dimension hertz per MW and is referred as
speed regulation of the governor.
Taking Laplace transform of eq. 1 yields
ΔPg (s) = ΔPref (s) - (1/R)Δf (s) (2)
The block diagram corresponding to the above equation is as follows:

Fig.2: Block diagram of speed governor

5. Modelling of Hydraulic Valve Actuator

Output: The output of the hydraulic actuator is ΔPv. This depends

on the position of main piston, which in turn depends on
the quantity of oil flow in the piston.
For a small change ΔxD in the pilot valve position, we have

 PV  k H  X D dt (3)

Where, the constant “kH” depends on the orifice, cylinder geometries and fluid
pressure.
Input:The input to ΔxD are ΔPg and ΔPv. It is to be noted that for a
positive ΔPg, the change ΔxD is positive. Further, for a positive ΔPv, more fuel is
admitted, speed increases, linkage point B moves downwards causing linkage
points C and D to move downwards resulting the change ΔxD as negative.
Thus
ΔxD = ΔPg- ΔPv (4)

Laplace transformation of the last (3) and (4) are as follows:

ΔPv(s) =skHΔxD (s)
ΔxD(s) = ΔPg(s) -ΔPv(s)
Eliminating ΔxD and writing ΔPv (s) in terms of ΔPg(s), we get
1
PV ( s )  Pg ( s ) (5)
1  s TH
Where, THis the hydraulic time constant given by
TH= 1/kH (6)
In terms of the hydraulic valve actuator’s transfer function G H(s), eq. 5
can bewritten as
PV ( s) 1
GH ( S )   (7)
Pg ( s ) 1  s TH

Hydraulic time constant TH typically assumes values around 0.1 sec.

The block diagram of the speed governor with the hydraulic valve actuator
is as follows:

Fig. 3: Block diagram of speed governor together with hydraulic

valve actuator
6. Modelling of Turbine – Generator
In normal steady state, the turbine power PT keeps balance with the
electromechanical air-gap power PG resulting in zero acceleration and a constant
speed and frequency.
During transient state, let the change in turbine power be ΔPT and the
corresponding change in generator power be ΔPG .
The accelerating power in turbine generator unit = ΔPT- ΔPG
Thus, the accelerating power can be expressed in Laplace Transform as
Pa = ΔPT(s) - ΔPG (s) (8)
If ΔPT- ΔPG is negative, it will decelerate.
The turbine power increment ΔPT depends entirely upon the valve power
increment ΔPv and the characteristic of the turbine as different type of turbines
will have different characteristics.
Taking transfer function with single time constant for the turbine, we can write
1
PT ( s)  GT PV ( s)  PV ( s) (9)
1  s TT
The generator power increment ΔPG depends entirely upon the change ΔPD in
theload PDbeing fed from the generator. The generator always adjusts its output so
as to meet the demand changes ΔPD.
These adjustments are essentially instantaneous, certainly in comparison with
the slow changes in PT.
We can therefore set
ΔPG= ΔPD, i.e., ΔPG (s) = ΔPD(s)(10)

In view of equation 8,9,10, the block diagram is updated as follows:

This corresponds to the linear model of primary ALFC loop excluding the
Generator response.
Fig 4: Block diagram corresponding to primary loop of ALFC
excluding Generator
7. Automatic Generation Control
 Electric power is generated by converting mechanical energy into electrical
energy.
 The rotor mass, which contains turbine and generator units, stores kinetic
energy due to its rotation.
 This stored kinetic energy accounts for sudden increase in the load.
 Let us denote the mechanical torque input by Tm and the output electrical
torque by Te . Neglecting the rotational losses, a generator unit is said to be
operating in the steady state at a constant speed when the difference between
these two elements of torque is zero. In this case we say that the accelerating
torque is zero. The accelerating torque can be expressed as

(11)

 When the electric power demand increases suddenly, the electric torque
increases.
 However, without any feedback mechanism to alter the mechanical
torque, Tm remains constant.
 Therefore the accelerating torque given by (11) becomes negative causing a
deceleration of the rotor mass.
 As the rotor decelerates, kinetic energy is released to supply the increase in
the load. Also note that during this time, the system frequency, which is
proportional to the rotor speed, also decreases.
 We can thus infer that any deviation in the frequency for its nominal value
of 50 or 60 Hz is indicative of the imbalance between Tm and Te.
 The frequency drops when Tm < Te and rises when Tm > Te .
The steady state power-frequency relation is shown in Fig. 5. In this figure the
slope of the ΔPref line is negative and is given by

(12)

Where, R is called the regulating constant. From this figure we can write the
steady state power frequency relation as

(13)
 Fig. 5 A typical steady-state power-frequency curve.

Suppose an interconnected power system contains N turbine-generator units. Then

the steady-state power-frequency relation is given by the summation of (13) for
each of these units as

(14)

In the above equation,

 ΔPm is the total change in turbine-generator mechanical power and
 ΔPref is the total change in the reference power settings in the power
system.
 Also note that since all the generators are supposed to work in synchronism,
the change is frequency of each of the units is the same and is denoted by
Δf. Then the frequency response characteristics is defined as

(15)

We can therefore modify (13) as

Pref  Pm
Or, f  (16)

Example 1
Consider an interconnected 50-Hz power system that contains four turbine-
generator units rated 750 MW, 500 MW, 220 MW and 110 MW. The
regulating constant of each unit is 0.05 per unit based on its own rating.
Each unit is operating on 75% of its own rating when the load is suddenly
dropped by 250 MW. We shall choose a common base of 500 MW and
calculate the rise in frequency and drop in the mechanical power output of
each unit.

Solution
The first step in the process is to convert the regulating constant, which is
given in per unit in the base of each generator, to a common base. This is
given as

new
Sbase
Rnew  Rold  old
Sbase

We can therefore write

Therefore

per unit

We can therefore calculate the total change in the frequency from (16)
while assuming ΔPref = 0, i.e., for no change in the reference setting. Since
the per unit change in load - 250/500 = - 0.5 with the negative sign
accounting for load reduction, the change in frequency is given by
Then the change in the mechanical power of each unit is calculated from
(13) and multiplied by 500 to get values in MW as

It is to be noted that once ΔPm2 is calculated to be - 79.11 MW, we can also

calculate the changes in the mechanical power of the other turbine-
generators units as

This implies that each turbine-generator unit shares the load change in
accordance with its own rating.
8. Static Performance of Speed Governor

Consider the Fig.4 for analyzing different cases

Fig 4: Block diagram corresponding to primary loop of ALFC

excluding Generator

The present control loop shown in Figure is open. We can

nevertheless obtain some interesting information about the static
performance of the speed governor. The relationship between the
static signals (subscript “0”) is obtained by letting.
S 0
As GH(0) = GT(0) = 1 we obtain directly from above figure

1
PT 0  Pref 0  f 0 (17)
R
Case A

The generator is synchronized to a network of very large size, so large in fact,

that its frequency will be essentially independent of any changes in the power
output of this individual generator (“infinite” network). Since Δf0 = 0, the above
eq. becomes

PT 0  Pref 0 (18)

Thus for a generator operating at constant speed, (or frequency) there exists a
direct proportionality between turbine power and reference power setting.
ΔPT 0 = ΔPref0, i.e., when the generator is operating at constant frequency, if the
speed changer setting is INCREASED,(DECREASED) turbine output power will
increase (decrease) to that extent.

Case B
Now we consider the network as “finite”. i.e. its frequency is variable. We do,
however, keep the speed changer at constant setting. i.e. ΔPref = 0. From eq. (17),
i.e.,
1
PT 0  Pref 0  f 0 we obtain
R

1
PT 0   f 0 (19)
R

1. The above eq. shows that for a constant speed changer setting, the static
increase in turbine power output is directly proportional to the static
frequency drop.
2. The above eq. (19) can be rewritten as Δf0 = - R ΔPT0. This means that the
plot of f 0 with respect to PT0 (or PG0) will be a straight line with slope of
– R.
We remember that the unit for R is hertz per MW. In practice, both the frequency
and the power can be expressed in per unit.

Case C

In general case, changes may occur in both the speed changer setting and
frequency in which case the relationship ΔPT 0 = ΔPref 0–(1/R)Δf0 applies.

For a given speed changer setting, ΔPref 0 = 0 and hence Δf0 = - R ΔPT0. In a
frequency-generation power graph, this represents a straight line with a
slope = - R.
For a given frequency, Δf0 = 0 and hence ΔPT0 = ΔPref 0. This means that for a
given frequency, generation power can be increased or decreased by suitable raise
or lower command.

Thus, the relationship ΔPT0 = ΔPref0–(1/R)Δf0 represents a family of sloping lines

as depicted in Fig. 5, each line corresponding to a specific speed changer setting.
The thick line shows that corresponding to 100% rated frequency, the output
power is 100 % of rated output. But for the new speed changer setting as shown by
the dotted line, for the same 100% rated frequency, output power is 50 % of rated
output. Hence the power output of the generator at a given frequency can be
adjusted at will, by suitable speed changer setting. Such adjustment will be
extreme importance for implementing the load division as decided by the optimal
policy.

Let the governor characteristic of two units be 10 and 20. Let the operating
frequency be fr. Then load shared by unit 1 and 2 are P10 and P20. If the economic
division of load dictates the load sharing as P1’ and P2’, the governor
characteristics should be shifted to 1’ and 2’ as shown in Fig. below.
 9. Closing of ALFC Loop

As observed earlier that the loop in Fig. 4 is “open”. We now proceed now to
“close” it by finding a mathematical link between ΔPT and Δf.
As our generator is supplying power to a conglomeration of loads in its service
area, it is necessary in our following analysis to make reasonable assumptions
about the “lumped” area behaviour. We make these assumptions:

1) The system is originally running in its normal state with complete power
balance, that is, PG0 = PD0 + losses. The frequency is at normal value f 0. All
rotating equipment represents a total kinetic energy of W0kin MW sec.

2) By connecting additional load, load demand increases by ΔPD which we

shall refer to as “new” load. (If load demand is decreased new load is
negative). The generation immediately increases by ΔPG to match the new
load, that is ΔPG = ΔPD.
3) It will take some time for the control valve in the speed governing system to
act and increase the turbine power. Until the next steady state is reached, the
increase in turbine power will not be equal to ΔPG.
Thus, there will be power imbalance in the area that equals ΔPT - ΔPG i.e.
ΔPT – ΔPD. As a result, the speed and frequency change. This change will be
assumed uniform throughout the area.
The above said power imbalance gets absorbed in two ways.
 By the change in the total kinetic K.E.
 By the change in the load, due to change in frequency.

Since the K.E. is proportional to the square of the speed, the area K.E. is
2
0  f 
W kin  W kin  0  MW Sec (20)
 f 

The “old” load is a function of voltage magnitude and frequency.

The Frequency dependency of load can be written as

 PD
D (21)
f
d
Thus, PT  PD  (Wkin )  D . f (22)
dt

Noting that f  f 0  f
 f 0  f 2  2 f  f  2 
0 0 0  f 
Wkin W (
kin )  Wkin 1       Wkin 1  2 0  (23)
f0  f0  f0    f 


Differentiate above equation with respect to time and get as

0
d 2Wkin
Wkin   0 d f 
dt f dt

Substitute the above in equation (22) and get

0
2Wkin d
PT  PD  0
f   D f MW (24)
f dt

By dividing this equation by the generator rating Pr and by introducing

per unit inertia constant
0
Wkin
H  MW Sec / MW (or Sec)
Pr

It takes the form

2H d
PT  PD  f   D f pu (25)
f 0 dt

The ΔP’s are now measured in per unit (on base Pr) and D in pu per Hz.
Typical H values lie in the range 2 – 8 sec. Laplace transformation of the
above equation yields

2H
PT (s)  PD (s)  s f (s)  D f (s) (26)
f0

2 H 
  0 s  D f (s)
 f 

1
Or, f ( s ) 
2H
PT (s)  PD (s)
sD
f0

Or, f (s)  Gp (s) PT (s)  PD (s) (27)

1
Where, G p ( s ) 
2H
sD
f0
 1D

2H
1 s 0
f D

Kp

1  s Tp

1 2H
KP  Tp 
D f0D

Equation (27) represents the missing link in the control loop of Fig. 4. By
adding this, block diagram of the primary ALFC loop is obtained as shown
in Fig. 6.

9.1 Primary ALFC Loop for Uncontrolled case, i.e., for constant
reference input
The ALFC loop of Fig. 6 with ΔPref = 0 is as follows
 This can be simplified as shown below

Fig. 7 Simplified Diagram

Kp
Where,  Gp
1  sT p

From this simplified diagram, we can write

GP
 f ( s)    PD ( s) (28)
1
1 GH GT GP
R

9.2 Static Frequency Drop due to Step Load Change

For a step load change of constant magnitude ΔPD = M,

M
we have,  PD ( s ) 
s

Using the final value theorem, we readily obtain from eq. (28), i.e.,
GP
 f ( s)    PD ( s)
1
1 GH GT GP
R

The static frequency drop as

   
liim  Gp   Gp M
f 0  s  0 sf ( s )  s   PD ( s )   s   
 1  1 G H GT G p   1  1 G H GT G P s 
 R   R 

Kp M M M
 M   Hz   Hz (29)
Kp 1 1 1 
1  D
R Kp R R

1
Where,   D  pu / Hz , Which is called “Area frequency Response
R
Characteristic (AFRC)”
M
Thus, the static frequency drop is given by f 0   Hz

 9.3 Dynamic Response of Primary ALFC Loop
The transfer function of primary ALFC loop is

GP
 f ( s)    PD ( s)
1
1 GH GT GP
R

It shows that the characteristic equation is of third order. To reduce the

complexity and make the analysis simple, the assumption is made that action of
speed governor plus the turbine generator is “instantaneous” compared with the
rest of the power system.The approximate analysis can be performed by setting TH
= TT = 0.

So withTH = TT = 0, from above equation, we getfor step load

GP M
 f (s)  
1 s
1 . GP (30)
RT

KP
1  sT p M
 f ( s)  
1 KP s
Or, 1 .
R 1  sT p

RKp M
 f (s)  
Or, R1  sT   K P s (31)

Dividing numerator and denominator of (31) by RTp , we get

Kp 1
 f (s)   M (32)
Tp  R  Kp 
s s  
 RT 
 p 
Using the following fact in a above equation

A A 1 1 
  
s( s   )   s s   

A K P RTP
We have  M
 TP R  K P
So, Equation (32) becomes

 
RK p 1  1 
(33)  f ( s)   M   
R  K p  s s  R  KP 
 RTp 

For time response take Laplace inverse and get

R K p
RK p   t
 f (t )   M 1  e RTp 
R Kp  
 
Thus, approximate time response is purely exponential

Time response of primary loop of ALFC corresponding to typical example is as

follows:

Fig. 8 Response of primary loop of ALFC

10. Proportional plus integral Control ( Secondary ALFC loop)
It is seen that with the speed governing system installed in each area, for a given
speed changer setting, there is considerable frequency drop for increased system
load.

System frequency specification is rather stringent and therefore, so much

change in frequency cannot be tolerated. In fact, it is expected that the steady state
frequency change must be zero. In order to maintain the frequency at the scheduled
value, the speed changer setting must be adjusted automatically by monitoring the
frequency changes.

For this purpose, INTEGRAL CONTROLLER is included. In the integral

controller the frequency error is first amplified and then integrated. Further, a
negative polarity is also included so that a NEGATIVE frequency deviation will
give rise to RAISE command.

The signal fed into the integrator is referred as Area Controlled Error (ACE).
For this case ACE =Δf. Thus

Pref  KI   f dt (35)

Taking Laplace transform, we get

KI
Pref (s)    f (s) (36)
s
The gain constant KI controls the rate of integration and thus the speed of response
of the loop.

The signal Δf (s) is fed to an integrator whose output controls the speed changer
position resulting in the block diagram configuration shown in Fig. 9.

Fig. 9 Block diagram corresponding to complete ALFC

The Integral controller will give rise to ZERO STEADY STATE FREQUENCY
ERROR following a step load change.

The input to GHT is

KI 1
   f ( s)   f ( s)
s R

K 1
   I    f ( s)
 s R

Using above equation, the block diagram of Fig. 9 can be reduced as Fig. 10 and
11

Fig.10 Reduced Block

Fig.11 Reduced Block

GP
f ( s )   PD ( s )
 KI 1 
1   GHT . GP
 s R

GP
Or, f ( s )   PD ( s ) (37)
K 1
1  I GHT G P  G HT GP
s R
The, time response Δf (t) can be determined by taking Laplace inverse

10.1Static Frequency Drop Following A Step Load Change

Let the step load change be ΔPD, which is equal to M. Then

M
PD ( s) 
s
Using final value theorem,

f0  lim
S0 s  f ( s )  (38)

GP M

KI 1
1 GHT GP  GHT GP s 0
s R

KP M

KI 1
1 KP  KP s 0
s R

0 (39)

Thus, static frequency drop due to step load change becomes zero, which is a
desired feature we were looking. This is made possible because of the integral
controller that has been introduced.

10.2 Dynamic Analysis

Consider equation corresponding to secondary ALFC loop, i.e., equation (37) as

follows:

GP
f ( s)   PD ( s)
K 1
1  I GHT GP  GHT GP
s R

Let us assume time constants TH and TT as zero. Then GHT = 1.

For step load change of ΔPD = M, from above equation we get

GP M
f ( s )   .
KI 1 s
1 G P  GP
s R
 KP
1  s TP M
Or, f ( s )   .
K KP 1 KP s
1 I 
s 1  sTP R 1  sTP

Multiplying the numerator and denominator by s R( 1 + s Tp ) we get

R KP M
f ( s)  
sR1  sTP   RK I K P  sK P (40)

Dividing the numerator and denominator by R Tp the above equation becomes

KP M
f ( s)  
TP R  KP K K
s2  s I P
RTP TP

KP M

TP KP (41)
1
s2  R s  KI KP
TP TP

The time response Δf(t) can be obtained by taking Laplace inverse .

Since response depends upon the poles of the denominator polynomial, and
system stability can be accessed by examining poles and zeros of characteristic
equation, i.e., as follows:

KP
1
s2  R s  KI KP  0 (42)
TP TP

A second order system of type as2 + bs + c = 0 will be stable if the coefficients a, b

and c are greater than zero. Since this condition is met with, the system under
consideration is STABLE.

For a second order equation s2 + bs + c = 0, the roots are

b b2
s1 , s2    c
2 4
b2
And the nature of roots depends on c and form different cases
4
 1) Critical case
b2
 c  0 , i.e., b 2  4c
4

Now, both the roots are equal and negative. For this critical case

2
 K 
1 P  4 KI KP
 R s 
 TP  TP
 
 

2
1  KP 
K I crit  1  
Thus, 4 K P TP  R  (44)

2) Supper Critical Case

When b2 < 4 c the roots are complex conjugate and the solution is exponentially
damped sinusoidal. For this case the integral gain constant KI is obtained from
2
 K 
1 P  4 KI KP
 R s  , i.e., K I  K I crit
 TP  TP (45)
 
 
When KI> KIcrit the system is said to be super-critical. Due to damping present, the
final solution will tend to zero. However, the solution will be oscillatory type.

3) Sub-Critical Case

When b2 > 4 c the roots are real and negative. For this case, the integral gain
constant KI is given by

2
 K 
1 P  4KI KP
 R s  , i.e., K I  K I crit (46)
 TP  TP
 
 
This case is referred as sub-critical integral control. In this case the solution
contains terms of the type e – α1t and e –α2t and it is non-oscillatory. However,
finally the solution will tend to zero.

Thus the integral controller system is STABLE and ISOCHRONOUS i.e.

following a step load change, the frequency error always returns to zero. The
dynamic response for different values of KI are shown in Fig. 12.

Fig. 12 Dynamic response for different KI

Assignment 5

1. Primary ALFC loop parameters for a control area are: Total rated capacity
Pr = 2000 MW Normal operating load P0D = 1000 MW Inertia constant H =
5.0 sec. Regulation = 2.0 Hz / p.u. Assume that the load-frequency
dependency is linear, meaning that the load would increase one percent for
one percent frequency increase.
Obtain the power system transfer function and Find the static
frequency drop for load increase of 1% of system rating.

2. Determine Laplace
M inverse of following transfer function for step load M
,i.e., ΔPD(s) = s and give time response Δf(t).

GP
f ( s)   PD ( s)
K 1
1  I GHT GP  GHT GP
s R

3. Determine Laplace inverse of following equationand give time response

Δf(t)

KP M
f ( s)  
TP R  K K K
s2  P
s I P
RTP TP