UNIVERSITI TEKNOLOGI MARA

FAKULTI KEJURUTERAAN KIMIA

CHEMICAL PROCESS CONTROL
(CPE501)

NAME : NURLINA SYAHIIRAH BINTI MD TAHIR

STUDENT NO : 2017632214
GROUP : EH2205I
EXPERIMENT : CONTROL LOOP SYSTEM (PID CONTROLLER)
SEMESTER :5
PROGRAMME / CODE : CHEMICAL ENGINEERING / EH220
SUBMIT TO : ABDUL AZIZ BIN ISHAK
DATE SUBMITTED : 2nd NOVEMBER 2018

No. Title Allocated Marks (%) Marks

1 Abstract/Summary
2 Introduction
3 Aims
4 Methodology/Procedure
5 Results
6 Discussion
7 Conclusion
8 Recommendations
9 Reference
TOTAL MARKS 100

Remarks:

Checked by: Rechecked by:

--------------------------- ---------------------------

Date: Date:
 TABLE OF CONTENT

1.0 ABSTRACT ......................................................................................................................3

2.0 INTRODUCTION............................................................................................................4

3.0 OBJECTIVES ..................................................................................................................7

4.0 METHODOLOGY ..........................................................................................................8

5.0 RESULTS .......................................................................................................................13

6.0 DISCUSSION .................................................................................................................21

7.0 CONCLUSION ..............................................................................................................25

8.0 RECOMMENDATIONS...............................................................................................26

9.0 REFERENCES ...............................................................................................................27

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 2

1.0 ABSTRACT

The Simulink Lab is done to investigate the effect of PID’s parameter (Proportional, Integral and
Derivative) and Deadtime of the process towards the process controllability. 3 set of varying
parameter value with the other two parameter is set at constant for Lab 1, Lab 2 and Lab 3. For
Lab 4, 3 varying value of time delay with fixed value of PID is consider. Larger value of P will
reduce the time taken for the process to reached the set point since the controller action becomes
slower and slows down the process response. Larger P value also reduce the number of oscillates.
Larger value of I increase the time taken for the process to reach the set point as the controller
actions becomes faster and fasten the process response. Larger I value also increase the number
of oscillates. Increasing the D value will makes the controller action becomes faster and the
process becomes faster, dampening the overshoot. However, no changes in the number of
oscillates present. Increasing the value of time delay, will increases the time taken for the process
to reached the set point. The number of oscillation increases as the time delay increases due to
the longer time taken for the response to reach half of its final value from the zero instant.

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 3

2.0 INTRODUCTION

In order to stabilise the industrial processes due to the changes in operating conditions and
present of process disturbances, controllers come in handy in an automatic control system. PID
controller using the mathematical function consisting of three parameters which are Proportiona,
Integral and Derivatives. All the lab is done in closed loop system since the controller is set to
automatic.

Effect of Proportional of PID Controller

Figure 1 - Effect of P in PID Controller at fixed Integral and Derivative Time

The controller action becomes slower and slows down the process response by
increasing the value of P. Proportional control eliminate the error due to the differences in
magnitude between the set point and the process variable. Many control system work well with
P-controller, however this could bring to a large offset, PID-controller brings a more precise and
accurate determination of error and corrective action.

Effect of Integral Time of PID Controller

Figure 2 - Effect of I in PID Controller at fixed Proportional and Derivative Time

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 4

 The controller action becomes slower and slows down the process response by increasing
the value of I. In the PID – controller, the integral time eliminate the residual error after the
proportional done its job. This control has small offset and lead to slow response time, though it
always return to steady state. The slow response is due to the error must have to accumulate
before any significant response can be detected from the output of the controller.

Effect of Derivatives Time of PID Controller

Figure 3 - Effect of D in PID Controller at fixed Proportional and Integral Time

The controller action becomes faster and the process response becomes faster by
increasing the derivatives value, dampening the graph of the process response curve. The
derivative control look out at the rate of change of the process variable and make corrective
action to the output variable for any error.

Effect of Increasing and Decreasing Value of P, I and D toward Process Response

Manual parameter tuning that be done using trial and error if the changes in operating condition
and present of disturbance in the process.

Table 1 - Theoretical Effect for The Process Response

Parameter Speed of Response Stability Accuracy

Increasing Kp Increases Deteriorate Improves
Increasing Ki Decreases Deteriorate Improves
Increasing Kd Increases Improves No Effect

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 5

Dead Time : The time at the first instance of change in measurement. The delay from when
the controller output signal is issued until the measured process variable first
begins to respond.

Delay Time : The time required for the process to reach at 50% of its final value by a time
response signal during its first cycle of oscillation.

Settling Time : The time at which the PV reaches ±5% of the total change in the process variable.

Overshoot : Most notably associated with P-only controller. The difference between the SP
to where the PV settles out at a steady state value.

Decay ratio :The size of the second peak above the new steady state divide by the size of the
first peal above the same steady state level.

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 6

3.0 OBJECTIVES

The following are the objectives for the Simulink Labs:

1) To investigate the effect of controller gain.
2) To investigate the effect of integral time.
3) To investigate the effect of derivative time.
4) To investigate the effect of deadtime.

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 7

4.0 METHODOLOGY

4.1 General Start Up Procedure

1) The mathlab software is opened and the Simulink icon located inside the software is
clicked to start the Simulink.
2) The New Window icon is clicked to open up a new empty model window, named
Untitled, the platform for the Simulink model is build.
3) The Simulink Browser icon is clicked to open up the Simulink Browser to find all
the blocks for the Simulink model.
4) The Simulink simulation parameters is set up to 600.

4.2 LAB 1 – Effect of Controller Gain

Figure 4 - Process Flow Diagram for Varying Proportional value in PID

Controller
1) ‘Analog Clock’ block and ‘To Workspace’ block entitled Time is linked together.
2) ‘Constant’, ‘PID(s) Controller’, ‘Transfer Function’, ‘Sum’, ‘Scope’ and ‘Display’
block is drag from the Simulink browser to the model window.
3) The blocks are then connected by placing the cursor at the output part of a block,
symbolized as > on the right edge of the block and then drag to the input port
symbolized < on the left edge of the other block. (The signal line is created through
this. The signal line has an arrowhead indicating the direction of signal flow)
4) The process set point is set up to 1 at the ‘Constant’ block entitled Set Point (SP).

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 8

 5) P, I and D controller’s parameters is set up to P = 0.05, I = 0.01 and D = 0.00,
respectively at the ‘PID(s) Controller’ block.
6) Process transfer function is set up as following at the ‘Transfer Function’ block.
5s
s2 + 10s
7) All the process response is set up in array-2D dimension.
8) After all the blocks has been connected to each other, the simulation is then run and
the ‘Scope’ block is clicked for the output, generating the process response
graphically.
9) The PV vs Time is the plotted via the mathlab window command as the following.
>>plot(time,PV)
10) All the blocks in Step 2 is copied and paste for the second set of PID’s value, with
different value of P. For the second set, P = 0.10. Then, the simulation is run.
11) The second process response is then generated via the mathlab command as the
following.
>>figure(2)*plot(time,PV1)
12) All the blocks in Step 2 is copied and paste for the third set of PID’s value, with
different value of P. For the third set, P = 0.15. Then, the simulation is run.
13) The third process response is then generated via the mathlab command as the
following.
>>figure(3)*plot(time,PV2)
14) All the three response is then combined by dragging and linking all the three set of
blocks using the ‘Mux’ block. All the respective ‘Scope’ block is removed and
replaced with only one ‘Scope’ block connected to the ‘Mux’ block for all the process
response to be in a figure.
15) The simulation is then run. The PV vs Time graph is then generated via the mathlab
command as the following.
>>plot(time,PV)
>>hold
Current plot held
>>plot(time,PV1)
>>plot(time,PV2)

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 9

4.3 LAB 2 – Effect of Integral Time

Figure 5 - Process Flow Diagram for Varying Integral value in PID Controller
1) The same step as in Lab 1 but with varying value of I and constant value of P and D.
1 0.05 0.01 0.00
2 0.05 0.05 0.00
3 0.05 0.10 0.00

4.4 LAB 3 – Effect of Derivative Time

Figure 6 - Process Flow Diagram for Varying Derivative value in PID Controller
1) The same step as in Lab 1 but with varying value of I and constant value of P and D.
1 0.05 0.01 0.00
2 0.05 0.01 0.05
3 0.05 0.01 0.10

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 10

4.5 LAB 4 – Effect of Deadtime

1) ‘Analog Clock’ block and ‘To Workspace’ block entitled Time is linked together.
2) ‘Constant’, ‘PID(s) Controller’, ‘Transfer Function’, ‘Sum’, ‘Transport Delay’,
‘Scope’ and ‘Display’ block is drag from the Simulink browser to the model window.
The blocks are then connected.
3) The process set point is set up to 1 at the ‘Constant’ block entitled Set Point (SP).
4) P, I and D controller’s parameters is set up to P = 0.2, I = 0.01 and D = 0.00,
respectively at the ‘PID(s) Controller’ block.
5) Process transfer function is set up as following at the ‘Transfer Function’ block.
5s
s2 + 10s
6) The Time Delay is the set up to 5 at the ‘Transport Delay’ block.
7) All the process response is set up in array-2D dimension.
8) After all the blocks has been connected to each other, the simulation is then run and
the ‘Scope’ block is clicked for the output, generating the process response
graphically.
9) The PV vs Time is the plotted via the mathlab window command as the following.
>>plot(time,PV)
10) All the blocks in Step 2 is copied and paste for the second set with different value of
Time Delay. For the second set, Time Delay = 7. Then, the simulation is run.
11) The second process response is then generated via the mathlab command as the
following.
>>figure(2)*plot(time,PV1)
12) All the blocks in Step 2 is copied and paste for the third set with different value of
Time Delay. For the third set, Time Delay = 9. Then, the simulation is run.
13) The third process response is then generated via the mathlab command as the
following.
>>figure(3)*plot(time,PV2)
14) All the three response is then combined by dragging and linking all the three set of
blocks using the ‘Mux’ block. All the respective ‘Scope’ block is removed and
replaced with only one ‘Scope’ block connected to the ‘Mux’ block for all the process
response to be in a figure.

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 11

 15) The simulation is then run. The PV vs Time graph is then generated via the mathlab
command as the following.
>>plot(time,PV)
>>hold
Current plot held
>>plot(time,PV1)
>>plot(time,PV2)

4.6 General Shut Down Procedure

1) The Simulink browser and Model window are closed.

2) The mathlab software is then closed.

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 12

5.0 RESULTS

Table 2 – Simulink Lab' Parameter

Proportional Integral Derivative

Lab / Set
(P) (I) (D)
Lab 1 1 0.05 0.01 0.00
Effect of Controller 2 0.10 0.01 0.00
Gain 3 0.15 0.01 0.00
1 0.05 0.01 0.00
Lab 2
2 0.05 0.05 0.00
Effect of Integral Time
3 0.05 0.10 0.00
Lab 3 1 0.05 0.01 0.00
Effect of Derivative 2 0.05 0.01 0.05
Time 3 0.05 0.01 0.10
Time Delay
1 5 P = 0.20
Lab 4
2 7 I = 0.01
Effect of Dead Time
3 9 D = 0.00
Simulation Parameters = 600, Response = Array 2D dimension

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 13

 5.1 Lab 1 – Effect of Controller Gain

(a)

(b)

(c)

Figure 7 – Process Flow Diagram for (a) Set 1(P = 0.05), (b) Set 2 (P=0.10), (c) Set 3 (P=0.15)

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 14

Figure 8 – Process Flow Diagram for Combined Response for Varying Proportional Value

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 15

5.2 Lab 2 – Effect of Integral Time

(a)

(b)

(c)

Figure 9 – Process Flow Diagram for (a) Set 1, I = 0.01, (b) Set 2, I = 0.05 . (c) Set 3, I = 0.10.

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 16

Figure 10 - Process Flow Diagram for Combined Response for Varying Integral Value

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 17

5.3 Lab 3 – Effect of Derivative Time

(a)

(b)

(c)

Figure 11 - Process Flow Diagram for (a) Set 1, D = 0.00, (b) Set 2, D = 0.05, (c) Set 3, D =
0.10

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 18

Figure 12 - Process Flow Diagram for Combined Response for Varying Derivative Value

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 19

5.4 Lab 4 – Effect of Deadtime

Figure 13 - Process Flow Diagram for Combined Response for Varying Time Delay
Value

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 20

 6.0 DISCUSSION

6.1 Lab 1: Effect of Controller Gain

Process Variable, %

Time

Figure 14 - Combination of Process Response for varying proportional value

The figure shows the combine process response at varying P of P = 0.05, P = 0.10 and P
= 0.15 for Set 1, Set 2 and Set 3, respectively. At P = 0.05, the max process variable is at 1.6%
compared to at P = 0.10 and P = 0.15 where the process variables is at 1.4% and approximately
1.25%, respectively. The proportional control applies corrective action appropriate to eliminate
error. From the graph, larger value of controller gain will further decreasing the error from the
process.

Other than reducing the oscillates number, larger value of P ensure the stability of the
process. As seen in figure, for P = 0.15 at approximate time = 100, the process already reached
stabilization compared to at P = 0.10 and P = 0.05 where the process reached stabilization at time
= 200 and time = 500, respectively. Larger value of P will reduce the time taken for the process
to reached the set point since the controller action becomes slower and slows down the process
response.

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 21

6.2 Lab 2: Effect of Integral Time
Process Variable, %

Time

Figure 15 - Combined Process Response for Varying Integral Value

The figure shows the combined process response at varying Integral value of I = 0.01, I
= 0.05 and I = 0.10 for Set 1, Set 2 and Set 3, respectively. At I = 0.10, the max process variables
is approximately 1.85%, whereas at I = 0.05 and I = 0.01, the process variables is approximately
at 1.8% and 1.6% respectively. Integral control eliminates the residual error after the proportional
control by adding a control effect to the process. From the graph, larger value of I will increase
the number of the oscillates.

Other than increase the number of oscillates, larger value of I also reduce the stability of
the process. As seen in figure, for I = 0.10 the time taken for the process to stabilized is
approximately at time = 550, compared to I = 0.05 and I = 0.01 at time = 500 and time = 420,
respectively. Larger value of I increase the time taken for the process to reach the set point as the
controller actions becomes faster and fasten the process response.

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 22

6.3
Process Variable, % Lab 3: Effect of Derivative Time

Time

Figure 16 - Combine Process Response for Varying Derivative Value

The figure shows the combine process response at varying Derivative value of D = 0.00,
D = 0.05 and D = 0.10 for Set 1, Set 2 and Set 3, respectively. At D = 0.00, the max process
variable is at 1.6% compared to at D= 0.05 and D = 0.10 where the process variables is at 1.58%
and approximately 1.56%, respectively. The derivative control depends on the rate of error
change. From the graph, smaller value of D, increase the overshoot of the graph but do not have
any effects on the number of oscillates.

The derivative action only dampening the graph. Increasing the D value will makes the
controller action becomes faster and the process becomes faster. Besides, in order to enhance the
effects of this controller, larger value is needed to be consider. Thus, this control is rather suitable
for those process that is slow and taking longer time constant.

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 23

 Lab 4: Effect of Deadtime
Process Variable, %

Time

Figure 17 - Combine Process Response for Varying Time Delay Value

The figure shows the combine process response at varying Time Delay value of TD = 5,
TD = 7 and TD = 9 for Set 1, Set 2 and Set 3, respectively. At TD = 5, the max process variable
is at 2.2% compared to at TD= 7 and TD = 9 where the process variables is at approximately
1.8% and 1.75%, respectively. From the graph, increasing value of time delay, increases the time
taken for the process to reached the set point. The number of oscillation increases as the time
delay increases. The size of the oscillate is vary though, with TD = 9 as the largest follows by
TD = 7 and TD = 5, respectively. This is due to the longer time taken for the response to reach
half of its final value from the zero instant.

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 24

7.0 CONCLUSION

The process response towards the controller action varies according to the different value of the
PID’s parameter value. Larger value of P will reduce the time taken for the process to reached
the set point since the controller action becomes slower and slows down the process response.
Larger P value also reduce the number of oscillates. Larger value of I increase the time taken for
the process to reach the set point as the controller actions becomes faster and fasten the process
response. Larger I value also increase the number of oscillates. Increasing the D value will makes
the controller action becomes faster and the process becomes faster, dampening the overshoot.
However, no changes in the number of oscillates present. Increasing the value of time delay, will
increases the time taken for the process to reached the set point. The number of oscillation
increases as the time delay increases due to the longer time taken for the response to reach half
of its final value from the zero instant.

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 25

8.0 RECOMMENDATIONS

There are several recommendations can be made to improve the process response for the lab
based on the results obtained. Reduced the value of the integral action and increase the value of
the proportional action and the derivative action. This is to reduce the number of oscillates and
time taken for the process to reached the set point. Besides, the optimum P, I and D values should
be calculated using tuning rules to choose the right value of PID to control the process response
for lab 4. The tuning rules that can be consider is Tuning Rules by Ziegler – Nichols, Cohen –
Coon and CHR (Chien, Hrones & Reswick).

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 26

9.0 REFERENCES

Abdul Aziz Ishak, Z. A. (2013). PID Tuning Fundamental Concepts and Applications. Shah
Alam: Penerbit Press UiTM.

H. Bischoff, D. H. (1997). Process Control System. Festo Didactic GmbH & Co.

Hanafi, S. M. (5 December, 2015). Control Loop Simulation.

Lab 1 - Effect of Controller Gain. (9 October, 2018). Retrieved from I-Learn (CPE501 -
Chemical Process Control).

Lab 2 - Effect of Integral Time. (9 October, 2018). Retrieved from I-Learn (CPE501 - Chemical
Process Control).

Lab 3 - Effect of Derivative Time. (9 October, 2018). Retrieved from I-Learn (CPE501 -
Chemical Process Control).

Lab 4 - Effect of Deadtime. (9 October, 2018). Retrieved from I-Learn (CPE501 - Chemical
Process Control).

SIMULINK LAB REPORT – CONTROL LOOP SYSTEM 27