CHAPTER 3

CLOSED LOOP PERFORMANCE OF THE BOILER

DURING LOAD VARIATION

3.1 INTRODUCTION

The direct application of classical tuning methods in determining the PID

controller parameters warrants the determination of the intercept values in X-
axis (designated as “ L” ) and Y-axis (designated as “ a” ) from the open loop
step response of the output (to be controlled) with respect to input signal (to be
regulated). However, for complex processes, the above procedure is
cumbersome and hence as an alternative step, the researchers usually opt for the
heuristic algorithms for the determination of PID controller parameters. In this
work, the boiler demand signal as the input signal has been regulated to control
the main steam pressure of the 500 MW steam generator during load variations.
Using the proposed CHR method of PID controller tuning, the PID controller
parameters were determined. Further, it has been demonstrated that with the use
of these controller parameters, the transient responses obtained during load
disturbances are much better than the earlier published results.

3.2 BOILER SYSTEM DESCRIPTION

A thermal power plant consists of steam generator (also known as

boiler), turbine and generator. Boiler consists of furnace, heat exchangers like
super heaters, reheaters, economizer and circulation system. Circulation system
in turn comprises of drum, down comers and water walls which constitute the

41
walls of the furnace. Coal and air are burnt in the furnace and the products of
combustion become the flue gas. The flue gas formed in the furnace transfers
the heat to various heat exchangers and leaves the furnace. The location of
various heat exchangers in the furnace is shown in Fig. 3.1.

Fig. 3.1 Schematic diagram of boiler furnace

Fig. 3.2 schematically represents the coal, primary air (PA) and
secondary air (SA) path of the boiler system. The furnace is a tangentially fired
system and the number of coal burner elevations depends on the capacity of the
boiler. For a typical 500 MW unit, there will be nine elevations out of which six
or seven will be functioning during full load conditions.

42
 Fig. 3.2 Schematic diagram of Coal, Primary Air and Secondary Air path

The heat transfer mechanism in furnace is quite complex and the

furnace walls receive heat by direct radiation. The water inside the water walls
(WW) get converted into steam and the two phase mixture goes to drum where
the mixture gets separated as water and steam. Panel and platen super heaters
and reheaters also receive heat partly by direct radiation. The coal burners can
be tilted upwards and downwards. An upward tilt decreases the heat flow to
water walls and increases the heat flow to radiant super heaters and reheaters
(RH) while a downward tilt increases the heat flow to water walls and decreases
the heat flow to radiant super heaters and reheaters. The feed water after being
preheated in high pressure heaters and economizer enters the drum. The water
level in the drum is usually kept slightly below the central axis of the drum.

The high pressure steam flow to the turbine is controlled to meet the
set load demand in three different ways. These are known as Boiler following
mode, Turbine following mode and coordinated mode. Depending upon the
mode of selection, the boiler demand signal which ultimately regulates the
combustion process is derived from the master pressure control of Boiler-

43
Turbine-Generator unit. The method of deriving the boiler demand signal
corresponding to different modes of operation is discussed in next section.

3.3. BOILER DEMAND SIGNAL

Fig. 3.3 schematically represents the steam and water path of the
boiler system. The main steam (MS) pressure at the boiler outlet remains
constant if the steam production in circulation system is equal to the demand of
steam at the outlet of the boiler.

Fig. 3.3 Schematic diagram of steam and water path

The boiler demand signal which regulates the combustion process is

identified as the signal to be manipulated to control the main steam pressure.
This boiler demand signal is derived based on three modes of the operation of
the power plant. The modes of operation and the generation of the boiler
demand signal are given in Table 3.1.

44
 Table 3.1 Generation of Boiler Demand Signal (BDS) for different modes of operation

Mode of Schematic Diagram Explanation

Operation
Boiler
Following Boiler follows the actions of turbine to meet the requested
Mode generation. The load demand error immediately controls the
throttle valve. Boiler demand (derived as the difference between
the actual and desired main steam pressure) is used to adjust fuel
Boiler Following Mode and air into the furnace to modify the steam production resulting
faster response to meet the load requirement.
Turbine Turbine follows the actions of boiler to meet the requested
Following generation. Boiler Demand (derived as the difference between the
Mode actual and desired MW) is used to adjust fuel & air into the
furnace to modify the steam production. The throttle valve is
regulated to keep the main steam pressure (MSP) at the set point
value resulting sluggish response to meet the load requirement
but stable response to load changes with minimal steam pressure
Turbine Following Mode fluctuations.
Coordinated
Mode Boiler demand signal is derived combining the stable response
feature of T- following mode and fast response characteristic of
B- following mode.

Coordinated Mode

45
3.4 BASIC REQUIREMENTS FOR THE EVALUATION OF PID
CONTROLLER CONSTANTS

PID controller constants for any complex systems / processes can be

calculated using classical tuning methods such as Ziegler-Nichols, Chien-
Hrones-Reswick etc., provided the system response to unit step variation in
input variable is available. A typical unit step input response of the system G(s)
is shown in Fig. 3.4. According to the classical tuning methods, the values of
‘a’ and ‘L’ shown in Fig. 3.4 are only required to calculate the PID controller
tuning parameters, as employed by Astrom et al. (2006).

Fig. 3.4 System response for unit step input

Once these two values are obtained, PID controller constants can be
calculated using any method of interest by referring to Table 3.2. In the case of
controlling the main steam pressure of a boiler, the input (manipulated) variable
is boiler demand and the main steam pressure is the system output (response).
It can be seen from Fig. 3.2 that manipulated variable (Boiler demand) controls
the fuel flow and the hot primary air which carries the powdered coal in the
mills to furnace.

46
 Table 3.2 PID controller parameters for different tuning methods

Ziegler-Nichols CHR Method with CHR Method

Controller Method 0% Overshoot with 20%
Overshoot
Kp Ti Td Kp Ti Td Kp Ti Td
P 1/a 0.3/a 0.7/a
2.3
PI 0.9/a 3L 0.6/a 4L 0.7/a
L

PID 1.2/a 2L 0.5L 0.95/a 2.4L 0.42L 1.2/a 2L 0.42L

3.5 GENERATION OF STEP RESPONSE FOR A 500 MW BOILER

In this thesis, a nonlinear mathematical model developed and

validated for a 500 MW unit is available at Centre of Excellence for Simulators,
Corporate R&D, BHEL, Hyderabad. This model has been used in some of the
earlier studies by Dharmalingam et al. (2010, 2011a, b, c). The same model has
been used for the present study. This model has been initialized to operate at
500 MW load level with all control loops in auto mode. After having ensured
that the plant is running at steady state, all control loops have been put into
manual mode. Then a step change in boiler demand signal is given and the open
loop response of main steam pressure is documented. The variation of the main
steam pressure with respect to time is given in Fig. 3.5.
The values of ‘a’ and ‘L’ required for calculating the PID controller
constants are extracted from Fig. 3.6 and the values of ‘a’ and ‘L’ have been
found to be “a” = 0.1 ata and “ L”= 40 seconds. Based on these values, the PID
controller constants calculated due to different methods are given in Table 3.3.

47
 Fig. 3.5 MSP variation to step change in Boiler Demand Signal

Table 3.3 Calculated PID controller constants

Tuning Methods
CHR method
PID Values
Ziegler Nichols CHR method with with 20%
method 0% overshoot Overshoot

P 12 9.5 12
I 80 96 80
D 20 16.8 16.8

48
 The ProControl P13 platform of ABB has been used as distributed
control system in the 500 MW unit and the same has been emulated in the
simulator model. The PID structure of ProControl P13 in frequency domain is
given by Equation (3.1).

1
𝐺 (𝑠) = 𝐾𝑝 (1 + ( 𝑠 ) + (𝑇𝑉. 𝑠/(1 + 𝑇𝑠. 𝑠))) (3.1)
𝑇𝑁

where each of the four parameters (KP, TN, TV, TS) can be varied at will within
the valid range. Accordingly, the calculated values of KP, TN, and TV are
indicated in Table 3.4. One can choose the KP, TN and TV corresponding to
either Ziegler Nichols method or CHR method with 0% overshoot or 20%
overshoot as shown in Table 3.4. In this work, the controller parameters -
calculated based on 0% overshoot is used as the gain and derivative time are
smaller and integral time is larger for investigating the transient performance
behavior of main steam pressure and main steam temperature due to different
disturbances.

Table 3.4 Values of KP, TN and TV used in 500 MW unit simulation

Tuning Methods
PID CHR method
Ziegler Nichols CHR method with with 20%
Values method 0% overshoot Overshoot

KP 12 9.5 12
TN 0.15 0.098 0.15
TV 1.66 1.768 1.40

49
3.5.1 Main Steam Temperature Control

The purpose of the super heater is to raise the temperature of the

steam generated by the circulation system in the boiler to above saturation level.
In many applications, it is essential that the main steam temperature remains
constant over a wide load range. Steam temperature may be regulated by
changing the flue gas temperature or mass flow rate, or by removing part of the
heat absorbed by the steam in the super heater. An example of the former is
control by tilting burners, which varies the effective furnace size and amount of
heat absorbing surface with load changes. Response is immediate and there is
no hazard of steam contamination. Steam temperature may also be controlled
through by-passing a part of the flue gas around the super heater at high loads.
By-pass dampers are preferably located near the outlet of the second pass of the
boiler to avoid exposure to high flue gas temperature. A popular method of
steam temperature control is desuperheating by spray water. Water is usually
introduced between sections of the super heater to minimize outlet tubing
temperature. Desuperheating water must have a purity comparable to that of the
steam if contamination is to be avoided.

One of the critical control loops in a thermal power plant is main

steam temperature control. The temperature is affected by variations in boiler
load. The temperature of steam from boiler drum is increased in super heaters.
This temperature if it goes beyond a particular limit, reduces the life time of the
boiler and turbine components. In this control, the manipulated variable is the
spray water flow to the desuperheater. Through the attemperator, the spray
water is injected in the steam pipelines. It is essential that the control loop should
be tuned properly to compensate for the variations in main steam temperature.
This research describes the application of classical tuning of Proportional

50
Integral Derivative (PID) controller to control the main steam temperature in a
500 MW Utility Boiler.

When the super heater outlet temperature increases above the

required value, the spray water flow reduces the temperature which in turn
produces an increased volume of steam from the boiler to the turbine. The set
point for main steam temperature for most of the units is 540° C.

Steam Temperature can be easily controlled by varying the spray

water valve position for correcting the deviations from the steam temperature
set point and this is shown in Fig. 3.6. Because of the time lag and the long
dead time of the super heater, this temperature loop responds very slowly even
when the control loop is tuned for the quickest possible response without
overshoot. The disturbance rejection can be improved by employing an inner
(secondary) loop at the desuperheater. The desuperheater outlet temperature is
measured and the control valve position is operated so that the deviation
between the desuperheater outlet temperature and the desired desuperheater
outlet temperature is eliminated.

The spray water comes from upstream of the feedwater control

valves, and changes in feedwater control valve position will cause changes in
spray water pressure, and therefore disturb the spray water flow rate. The
desuperheater outlet temperature control loop will provide a gradual recovery
when this happens. If the spray water flow rate to the attemperator is measured,
a flow control loop can be implemented as a tertiary inner loop to provide very
fast disturbance rejection. However, in many cases spray water flow rate is not
measured at the individual attemperators and hence this flow loop cannot be
implemented.

51
 Fig. 3.6 Super heater Steam Temperature Control

3.6. BOILER CONTROLS – PERFORMANCE REQUIREMENTS

Original Equipment Manufacturers (OEM) of Control and

Instrumentation Package has to fulfill certain specified performance
requirements for the closed loop of controls. The important control loops
include main steam pressure, main steam temperature, Drum level, Furnace
pressure and hot reheat steam temperature.

A typical Performance Requirements of Auto Controls for 500 MW

boilers for different load changes is given in Table 3.5.

52
 Table 3.5 Maximum deviation of MS pressure from set point

Allowable
S. No Load Change (% of TMCR per deviations in
minute) MS pressure
(kg/cm2)
1 ±
1. ± 3% for five minutes ±2
• Load is decreased from 500
MW to 425 MW in 5 minutes
• Load is increased from 425
MW to 500 MW in 5 minutes
2 ±
2. ± 5% for five minutes ±3
• Load is decreased from 500
MW to 375 MW in 5 minutes
• Load is increased from 375
MW to 500 MW in 5 minutes
3 ±
3. ± 10% for three minutes ±4
• Load is decreased from 500
MW to 350 MW in 3 minutes
• Load is increased from 350
MW to 500 MW in 3 minutes

3.6.1 Main Steam Pressure and Temperature Control – Simulation results

and Discussion
a) Tuning Map method
Tuning map is a procedure which has been developed to direct
manual tuning. In order to provide perception about how variations in the
controller parameters affect the behavior of the closed loop system, tuning maps
were developed. The transient responses or frequency responses are arranged in
53
two-dimensional arrays in an organized way. The figures in the map
demonstrate how the response is varied by changes in proportional, integral and
derivative gain. In this method, the controller parameters are adjusted for
optimum operation.

b) Embedded Control Concept

For more than last one decade or so, IT based software packages are
extensively implemented as part of Distributed Control System (DCS)
envisaged for power plants. These software packages mainly take care of
Performance Analysis, Diagnostics and Optimization (PADO) of the entire
plant (Sivakumar et al. 2006). In this method, an embedded controller is
envisaged as part of the Distributed Control System.

c) CHR tuning method

The Chien, Hrones and Reswick(CHR) method is a modification of

Ziegler Nichols method. To provide damped closed-loop systems, the step
response was modified by CHR. The “ quickest response without overshoot” or
“quickest response with 20% overshoot” was considered as the design criteria
for this method of tuning. They also observed and projected that tuning formulas
for set point response and load disturbance response were not the same. The
parameters ‘a’ and ‘L’ has to be determined for tuning the PID controller using
CHR method.
The main steam pressure and temperature are critical parameters to
be controlled at the inlet of steam turbine. The normal set values for these
parameters are 17.25 MPa (170.3 ata) and 540○ C respectively. The efficacy of
the controller is judged based on the transient behavior of pressure and
temperature during various disturbances. The control of main steam pressure is
quite complex and researchers have resorted to different control philosophies.
While the proportional, integral and derivative (PID) control is simple to be

54
implemented, the nature of transient response depends on the choice of the
controller parameters. In this work, the researcher has demonstrated the efficacy
of controller parameters obtained through classical tuning algorithm - in
yielding better transient performance of main steam pressure during load
variations.

3.6.2 Transient response of Main steam pressure and temperature control

during load variation
To prove the efficacy of the controller parameters obtained by CHR
method is better than tuning map concept and embedded controller concept, the
simulations were carried out with the following details:
The plant was initialized to 500 MW. The steady state values for full
load conditions are:
Main Steam Pressure = 17.25 MPa (170.3 ata)
Main Steam temperature = 540° C
Steam Flow =1495 tons/hour
HP Bypass flow = 0.0 tons/hour
Once the steady state had been achieved, the following disturbances were
introduced and the analysis done.
Disturbance Simulation : Boiler Load decreased from 500 MW to 350 MW
in 3 minutes (10% load variation per minute for 3 minutes)

3.6.3 Simulation results due to Tuning Map method

Using the tuning map concept discussed by Astrom and Hagglund,

Dharmalingam et al. (2011) has obtained a set of PID controller parameters for
the worst disturbance of 10% ramp for 3 minutes. The Kp, Ki and Kd values
obtained are 2.0, 60, 120 respectively. The response of main steam pressure
variation to a disturbance of -10% load variation over duration of 3 minutes
from 500 MW load is shown in Fig. 3.7 to 3.9. Fig. 3.7 shows the load variation,

55
Fig. 3.8 shows the boiler demand signal to correct the pressure variation to its
set value and Fig. 3.9 (blue colour) denotes the actual pressure variation with
respect to time. It is observed that undershoot and overshoot are well within the
desired values specified in the performance requirements in Table 3.5.

Fig. 3.7 Load variation Fig.3.8 Boiler demand

signal

Fig. 3.9 MS pressure variation

The manipulated variable which is the spray water flow used to

control the super heater temperature is shown in Fig. 3.10 and the variation in
main steam temperature is shown in Fig. 3.11(blue colour).

3.6.4 Simulation results due to Embedded Controller method

Dharmalingam et al. investigated the main steam pressure and
temperature variation with an embedded controller with the set of PID controller

56
parameters as kp=5, ki=60.0 and kd=120. These parameters improved the
transient performance of the main steam pressure and temperature during load
variation when compared to the tuning map method. Fig. 3.9 and 3.11 (red
colour) denotes the actual pressure and temperature variation with respect to
time.

Fig. 3.10 Variation in Spray Water Flow Fig. 3.11 Variation in Super
Heater steam Temperature

3.6.5 Simulation results due to proposed Classical CHR tuning Method

In this research, based on the step response test, the values of ‘a’ and
‘L’ were extracted and CHR tuning method has been applied. Computer
simulations for load variations were carried out (with the controller parameters
as kp = 9.5, ki = 10.0 and kd = 16.8). The variations in main steam pressure
and temperature due to CHR tuning method is given in Fig. 3.9 and 3.11 (green
colour).

57
3.7 CONCLUSION

In this research, the ability of classical PID controller to meet the desired
transient performance requirements of complex systems during different
disturbances has been demonstrated. Towards this, the main steam pressure and
temperature control of a 500 MW unit has been considered as a case study. The
step response method enables the calculation of controller parameters through
Ziegler-Nichols and Chien-Hrones-Reswick (CHR) methods. Controller
parameters obtained based on CHR method has been used in the computer
simulations to obtain the transient characteristics of main steam pressure and
temperature during a worst load disturbance. An analysis on the transient
performance characteristics prove the applicability of classical tuning methods
to complex systems / processes in determining the PID controller parameters
and in turn providing due credit to PID controllers amidst upcoming modern
controllers. The thesis uses the PID controller parameters obtained using CHR
method for further simulation studies during fuel switching in the following
Chapter 4.

58