Professional Documents
Culture Documents
Boiler Control PDF
Boiler Control PDF
CHAPTER 4
4.1 INTRODUCTION
Thus, in the present work, the manipulated variables are air and fuel
flow to the boiler with respect to the set value. In the following sections, the
design and closed loop simulation with conventional PID, cascaded PID and
adaptive fuel set point based cascaded PID schemes are presented.
Steam pressure to the turbine inlet is the key variable that indicates the
state of balance between steam supply and demand. If supply exceeds demand,
then the pressure rises. Conversely, when demand exceeds the supply pressure
will drop.
50
The boiler master modulates the Firing Rate Demand (FRD) in order to
maintain a constant steam pressure. The air-fuel control modulates the air and
fuel supply in order to meet the firing rate demand.
There are many conventional control strategies for controlling the air
and fuel namely, single point positioning control, parallel positioning control,
ratio control and full-metered cross-limited control.
With single-point positioning control, the air damper and fuel control
valve are connected to a common master actuator through a jackshaft or
mechanical linkage. The combustion controller modulates the master fuel
actuator, with the jackshaft connections maintaining a minimum air to fuel ratio.
Single point positioning control is used where neither air nor fuel flow
measurements are available.
In ratio control, the air /fuel ratio is maintained with respect to FRD.
The ratio can be adjusted and provides a more constant air/fuel ratio across the
span of the boiler-firing-rate capacity.
The unit operation modes of thermal power plant are namely, turbine
following mode (TFM), boiler following mode (BFM), unit coordinated mode
(UCM) and frequency compensation mode (FCM).
The load controller generates the demand to the steam throttle valve
from the unit load demand and the measured generated power. The throttle
53
pressure controller generates heat rate to the combustion controller from the
measured throttle pressure and the turbine inlet pressure set point. Finally
combustion controller activates the fuel valve and air dampers. Oxygen trimming
fine-tunes the air-fuel ratio. Block diagram for combustion control of boiler is
shown in Figure 4.1.
The function generator on the air flow measurement scales the air flow
signal relative to the fuel flow signal to provide optimum air /fuel ratio. The
function generator values are determined by adjusting the fuel flow relative to
air flow at each test load. This allows air and fuel flow set points to be driven
by the same firing rate demand signal. If an increase in the firing rate is needed
due to increase in steam demand, the airflow leads the fuel. As the load curtails,
the fuel decreases first, followed by a decrease in combustion airflow. This
assures an air rich mixture always. A lower limiter is used in the air controller
to prevent the airflow set point reduction less than 25%of full span.
The arrangements for conventional PID controller scheme for air and
fuel control is shown in Figure 4.2.
modified to
TgS
U (s) = Kp 1 + - - +
T,s aTDs +1 (4.2)
54
f(x)
Legends
f(x)=function generator.
FT=flow transmitter.
T=transfer switch
A=auto switch
Fuel Air \
Valve Damper'
Figure 4.2 Arrangements for conventional PID controller scheme for air
and fuel control
Deshpande, 1981) as the transfer function of a first-order system with unity gain
and a time constant equal to ‘TD’ is referred as a filter which will not affect the
performance of the controller because the typical values of ‘a’ range between
T„s +1
Controller Output = KP 1 + E(s) (4.3)
T,s Ti>cis +1
where
KdS +1
PID controller output = KKp 1-* E(s) (4.4)
s Kd
5+1
Ka
Where
K = Gain multiplier
Kp = Proportional gain
K, = Integral gain
Kd — Derivative gain
The overall optimum PID controller parameters for air and fuel flow
controllers for combustion process are obtained from the behavior modeling
approach as discussed in chapter 2. The controller parameters for fuel are as
follows: Gain multiplier K=2, Proportional gain KPF= 1.75, Integral gain
Kif =1.0, Derivative gain KDF=0.2, and Derivative lag constant KAF=10. The
controller parameters for air are as follows: Gain multiplier K=2, Proportional
57
gain Kpa =1.5, Integral gain KIA =0.8, Derivative gain KDA=0.2, and Derivative
lag constant KAA=10. These values are considered as optimum because of
satisfactory agreement with the real time behavior responses obtained from
210 MW thermal power plant.
After designing PID controller for air and fuel, several experiments
were conducted on the experimental set-up and the performances for both
positive and negative changes in the set point as well as in the load perturbation
are studied. The responses obtained are presented in Figures 4.3(a) - 4.3 (1).
PID Fuel
Load 21MW-42MW
F 30
u
e
1 20
f
I 10
o
w
0
t/hr
Time in seconds
A
500
r
400
f
1 300
0
w 200
t/hr 100
Time in seconds
PID Fuel
Load 21MW-63MW
KP=1.75, K^l, Kd= 0.2
F
u 30
e
ii
20
f
1i
0 10
w
t/hr 0
Time in seconds
PID Air
600
A 500
r 400
f 300
1
o 200
w
100
t/hr
0
Time in seconds
Figure 4.3 (d) PID air flow response for 21 to 63 MW change in load
F
u
e
1
f
1
o
w
t/hr
PID Air
Load 21MW-84MW
600 KP=1.5, K]=0.8, Kd= 0.2
A 500
r 400
f 300
1
o 200
w
100
t/hr
0
Time in seconds
Figure 4.3 (f) PID air flow response for 21 to 84 MW change in load
PID Fuel
Load 110MW-42MW
F 30
u
e
1
20
f
1
o 10
w
t/hr
0
Time in seconds
Figure 4.3(g) PID fuel flow response for 110 to 42 MW change in load
61
PID Air
600
A 500
i
400
r
300
f
1 200
0
w 100
t/hr 0
Time in seconds
Figure 4.3 (h) PID air flow response for 110 to 42 MW change in load
30
F
u
e
1 20
f
o 10
w
t/hr
0
0 30 60 90 120 150 180
Time in seconds
Figure 4.3(i) PID fuel flow response for 110 to 63 MW change in load
62
PID air
600
A 500
i
400
r
300
f
1 200
o
w 100
t/hr 0
Figure 4.3 (j) PID air flow response for 110 to 63 MW change in load
40
F
u 30
e
ii
20
f
1i
0 10
w
t/hr 0
Figure 4.3 (k) PID fuel flow response for 110 to 84 MW change in load
63
PID Air
600
500
A
i 400
r
300
f
1 200
o
w 100
t/hr 0
Time in seconds
Figure 4.3 (1) PID air flow response for 110 to 84 MW change in load
uses steam pressure as controlled process variable and allocates set point for fuel
and air flow. Perfect stoichiometric air-fuel ratio cannot be achieved by the set
point for airflow, derived from master controller. The quantity of air required
depends on the quantity of fuel actually supplied. In order to make an
improvement to the existing system, a new methodology has been proposed and
designed using cascaded PID scheme for fuel and air control for combustion of
utility boiler. In this new approach, fuel controller will get the set point with
respect to the load and the air controller will get the set point as a function of
the fuel flow. Thus required air-fuel ratio will be maintained. Further, oxygen
controller carries out fine-tuning for air-fuel ratio, which is not included in the
cascaded loop.
The difference between the set point and the actual fuel flow is
computed as the error signal. The control signal from the fuel controller is
applied to the fuel control valve. Same signal is applied to a function generator,
which develops a set point for the airflow loop with respect to the fuel flow.
The actual value of oxygen is measured from the flue gas analyzer for
the oxygen contents of the flue gas at the economizer outlet of the boiler. Since
real time combustor is not available, the set point for the oxygen controller is
derived as a function of steam flow through combustion simulator. The actual
value with respect to air/fuel ratio is collected from power plant and stored in
the computer database. The computer simulated error signal is given as input to
the oxygen controller. The output signal from the oxygen controller is added
with the airflow controller output, which trims the airflow to optimize the
combustion process.
The overall optimum cascaded PID controller parameters for air and
fuel flow controllers for combustion process is obtained from the behavior
modeling approach as per the procedure discussed in chapter 2. The controller
66
parameters for fuel are as follows: Gain multiplier K=2, Proportional gain
KPF=2.75, Integral gain K1F =1.5, Derivative gain KDF=0.4, and Derivative lag
constant KAF=10. The controller parameters for air are as follows: Gain
multiplier K=2, Proportional gain KPA =2.0, Integral gain KIA =1.0, Derivative
gain Kda=0.4, and Derivative lag constant KAA=10. These values are considered
as optimum because of satisfactory agreement with the real time behavior
response obtained from thermal power plant.
0 15 30 45 60 75 90 105%
After designing the cascaded PID controllers for air and fuel, several
simulations were conducted on the lab scale experimental set-up and the
performance for both positive and negative changes in the set point as well as in
the load perturbation were studied. The responses obtained are presented in
Figures 4.7 (a) - 4.7 (f).
68
Cascaded PID
LOAD 21MW-42MW
F 500
u
e A 30 400
1
▲
300
f 20
1 200
o
w 10 100
t/hr 0
0
Time in seconds
Cascaded PID
Load 21MW-63MW
KP=2.75, K,= 1.5, KD= 0.4
40
F 500
A
u
A 30 400 I
e
300
f 20 F
1 200 L
o O
w 10 100 w
t/hr 0
Time in seconds
Figure 4.7 (b) Cascaded PID air/fuel flow response for 21 to 63 MW change in
load
69
Cascaded PID
Load 21MW-84MW 700
KP=2.75, K,= 1.5, Kd= 0.4
600
F
500 A
u 30
e 4 400
20 300 f
1
200 o
o 10
w w
100
t/hr 0 t/hr
0
Time in seconds
700
Cascaded PID
Load 110MW-42MW 600
F 500 A
u
e 4 30 i
400
1 r
300 f
f 20
1
200
o o
w 10 100 w
t/hr t/hr
0
0
Time in seconds
Cascaded PID
Load 110MW-63MW 700
KP=2.75, K,= 1.5, Kd= 0.4
40 600
F
500 A
u
30 i
e 400
1 r
f 20 300 f
1 1
200 o
o
w 10 w
100
t/hr t/hr
0
Time in seconds
Cascaded PID
Load 110MW-84MW 700
KP=2.75, K,= 1.5, KD= 0.4
600
F A
500
u 30 i
e r
400
▲
20 300 f
f 1
1
200 o
o 10 w
w
100
t/hr
t/hr 0 0
Time in seconds
4.5.1 Design of adaptive fuel set point based cascaded PID scheme
p
Energy demand (for any value of throttle pressure) =-^-xJ°s (4.5)
PT
O
where
EF(Ada) is the required adaptive fuel, Pi is the turbine first stage pressure
PT is the turbine throttle pressure, Ps is the throttle pressure set point
O
The measuring circuit for calculating the adaptive set point is shown in
Figure 4.8. The response for load vs. steam flow, load vs. first stage pressure
and steam flow vs. oxygen were obtained from 210 MW thermal power plant
during real time operation and stored in the computer for simulation are shown
in Figures (4.9-4.11).
Figure 4.8 Measuring system for adaptive fuel set value calculation
74
10 20 30 40 50 60 70 80 90 100%
Load in MW
0 15 30 45 60 75 90 105%
The proposed conventional adaptive fuel set point based cascaded PID
scheme for fuel and air control is shown in the Figure 4.12.
Since real time combustor is not available, the set point for the oxygen
controller is derived as a function of steam flow through combustion simulator.
The actual value with respect to air/fuel ratio is collected from power plant and
stored in the computer database. The computer simulated error signal is given as
input to the oxygen controller. The output signal from the oxygen controller is
added with the airflow controller output, which trims the airflow to optimize the
combustion process.
76
Figure 4.12 Arrangement for the adaptive fuel set point based cascaded PID
controller scheme
77
The overall optimum cascaded PID controller parameters for air and
fuel flow controllers for combustion process is obtained from the behavior
modeling approach as discussed in chapter 2.
The controller parameters for fuel are as follows: Gain multiplier K=2,
Proportional gain KPF = 1.5, Integral gain KIF =2.0, Derivative gain KDF=0.5, and
Derivative lag constant KAF=10. The controller parameters for air are as follows:
Gain multiplier K=2, Proportional gain KPA =1.5, Integral gain KJA =1.2,
Derivative gain KDA=0.5 and Derivative lag constant KAA=10. The controller
parameters for oxygen are as follows: Gain multiplier K=l, Proportional gain
KP0 =4, Integral gain KI0 =2.0, Derivative gain KDO=1.0, and Derivative lag
constant KA0=5.
4.5.2 Simulation studies with adaptive fuel set point based cascaded PID
scheme
After designing the adaptive fuel set point based cascaded PID
controllers for fuel, several simulations were conducted on the lab scale
experimental setup and the performance for both positive and negative changes
in the set point as well as in the load perturbation are studied. The responses
obtained are presented in Figures 4.13 (a)-4.13 (f).
78
f
1 10 200 {
o
w
----------------------------------------------------------------- 0
t/hr 0
0 30 60 90 120 150 180
Time in seconds
Figure 4.13(a) Adaptive fuel SP based Cascaded PID air/fuel flow response
for 21 to 42 MW change in load
f 20 300 f
1 1
o 200 o
w 10 w
100
t/hr t/hr
0 0
0 30 60 90 120 1 50 180
Time in seconds
Figure 4.13(b) Adaptive fuel SP based Cascaded PID air/fuel flow response
for 21 to 63 MW change in load
79
Figure 4.13(c) Adaptive fuel SP based Cascaded PID air/fuel flow response
for 21 to 84 MW change in load
Time in seconds
Figure 4.13(d) Adaptive fuel SP based Cascaded PID air/fuel flow response
for 110 to 42 MW change in load
80
Time in seconds
Figure 4.13(e) Adaptive fuel SP based Cascaded PID air/fuel flow response
for 110 to 63 MW change in load
Figure 4.13(f) Adaptive fuel SP based Cascaded PID air/fuel flow response
for 110 to 84 MW change in load
81
PID Air
19 22 73 21 26 83 29 34 90
flow
Fuel
28 36 69 32 39 74 36 42 84
flow
Cascaded Air
16 20 50 23 27 45 26 30 72
PID flow
Fuel
18 31 42 26 36 48 32 37 66
flow
Adaptive Air
14 18 41 20 25 41 18 24 39
fuel SP flow
based
cascaded Fuel
14 28 34 22 23 41 26 31 52
flow
PID
O
O
OO
Cascaded PID Air flow 8792 9667 8076 9025 8129 8445 8654 7288 9596 9076 9544
00
r-
r-
Fuel flow 9043 8971 8669 8277 8071 8112 8437 7723 8821 8659 8452
00
o
o
05
Adaptive fuel Air flow 8072 8975 7221 8493 7659 6989 i 8759 8856
7387 7541 8997
i
1
SP based ----------
Fuel flow 8817 8637 8432 7665 7165 7062 7667 7503 7045 8640 8329 8199
cascaded PID i
1
82
83
Figures 4.14(a) - 4.14(1) show the bar charts for time domain and
performance criteria to various positive & negative step changes in loads and for
several set point variations.
Settling tim e in seco n d s
□ PID
■ CASCADED PID
ilADAPTIVE CASCADED
PID
Figure 4.14(a) Comparison of air flow settling time for positive step change
in loads with conventional schemes
Settling time in seconds
□ PID
■ CASCADED PID
S3 ADAPTIVE CASCADED PID
21-84 MW
Figure 4.14(b) Comparison of fuel flow settling time for positive step
change in loads with conventional schemes
84
□ PID
■ CASCADED PID
a ADAPTIVE CASCADED
PID
Figure 4.14(c) Comparison of air flow settling time for negative step change
in loads with conventional schemes
o o- * Mo Uo J oi 0 1o C oD oN lo0 0 o( f loO
Settling time in seconds
□ PID
■ CASCADED PID
B ADAPTIVE CASCADED
PID
ON
C k)
oO
N)
s3
Figure 4.14(d) Comparison of fuel flow settling time for negative step
change in loads with conventional schemes
85
10000
9000 -
8000 - —
7000 -
6000 - □ PID
LU
C/3 5000 - ■ CASCADED PID
4000 - 0 ADAPTIVE CASCADED PID
3000 —
2000 -
1000 —
0 -
Figure 4.14(e) Comparison of ISE (air) for positive step change in loads
with conventional schemes
Figure 4.14(f) Comparison of ISE (fuel) for positive step change in loads
with conventional schemes
86
10000
9000
8000
7000
6000 □ P!D
in
< 5000 ■ CASCADED RD
4000 0 ADAPTIVE CASCADED PID
3000
2000
1000
0
Figure 4.14(g) Comparison of IAE (air) for positive step change in loads
with conventional schemes
Figure 4.14(h) Comparison of IAE (fuel) for positive step change in loads
with conventional schemes
87
10000
9000 -
8000 -
7000
6000 - □ PID
M 5000 - ■ CASCADED PID
4000 B ADAPTIVE CASCADED PID
3000 -
2000 -
1000 -
0 -
Figure 4.14(i) Comparison of ISE (air) for negative step change in loads
with conventional schemes
10000
Figure 4.14(k) Comparison of IAE (air) for negative step change in loads
with conventional schemes
10000
9000
8000
7000
6000
< 5000
4000
3000
2000
1000
0
110-84 MW 110-63 MW 110-42 MW
Figure 4.14(1) Comparison of IAE (fuel) for negative step change in loads
with conventional schemes
89
Air flow
PID to 44 51 54 35 39 36
AFSCPED
Fuel flow
50 45 39 36 37 35
ISE IAE ISE IAE ISE IAE ISE IAE ISE IAE ISE IAE
PID to Air flow 9.8 3 6 9.4 6.8 15 5 6 5.3 3.8 8.7 2.8
Cascaded
PID Fuel
8.4 4.6 3.1 8.1 4.2 9.2 2.5 5 9.8 2.4 4.7 6.8
flow
4.6 CONCLUSION
Also in all cases air flow follows the fuel flow to match the required
air/fuel ratio for complete combustion. It has been observed from the responses,
that in the entire negative step changes in the boiler demand the fuel decreases
first and then air decreases and for all the positive step changes in boiler
demand, the air increases first and then the fuel increases to maintain optimal
air/fuel at all loads.