PROCESS INSTRUMENTATION & CONTROL 302265

Experiment 1

Essentials of Process Control Level Control Process

Name: Calvin Tse-Liang Chin

Student ID: 17098128

Date: 28/10/2014
Calvin Tse-Liang Chin 17098128

Table of contents:

1.0 Objectives

1.1 Experiment 1 2
1.2 Experiment 2 2

2.0 Background and Equipment 2

3.0 The PID Controller 2

4.0 Experimental Procedure

4.1 Experiment 1: Proportional Control of Flow Control (Closed loop,

P only) 3
4.2 Experiment 2: Proportional Control of Flow Control (Closed loop,
P + I) 3

5.0 Discussion

5.1 Introduction 4

5.2 P Only Controller

5.2.1 P Only Method of Control Graph 4

5.2.2 Analysis of the P Only Method of Control 5

5.3 P and I Controller

5.3.1 P and I Method of Control Graph 5

5.3.2 Analysis of the P and I Method of Control 6

6.0 Conclusion 6

1
Calvin Tse-Liang Chin 17098128
1.0 Objectives
1.1 Experiment 1: To demonstrate the effectiveness of a P only controller to
vary and control level
1.2 Experiment 2: To demonstrate the effectiveness of a P+I controller to
vary and control level

2.0 Background and Equipment

PCT50 Level Control Process is a process unit that provides an insight into the
fundamentals of process engineering. It consists of a sump tank, variable speed
centrifugal pump and a quick release connector. In this laboratory we had to carry out
2 exercises to demonstrate flow control by varying pump speed. These involved
operating a proportional (P only) controller and a Proportional and integral (P+I)
controller.
Disturbances can be added as external variables, not controlled in any way by the
operator or the system that impact the process. The main objective of a feedback
controller is to respond to changes in the system which are usually either set point
tracking or disturbance mitigation. The controller attempts to force the process
variable back towards the desired set point when a disturbance or other load on the
process causes a deviation.

3.0 The PID Controller

The Proportional Integral Derivative Controller calculates the error (offset) between
measured values and the set point value. The Integral control eliminates the offset
created by the proportional control. This, however, produces an oscillatory response.
Derivative control anticipates the future error and stabilizes the system by adjusting
the process accordingly. Therefore, it can be observed that PID controllers respond to
all aspects mentioned and regulated the process to achieve the perfect system.

2
Calvin Tse-Liang Chin 17098128
4.0 Experimental Procedure
4.1 Experiment 1: Proportional Control of Level (Closed loop, P
only and P + I)
For the P only controller:
1. Click on the PID box on the diagram and set the Proportional Band (P) to 100%.
Ensure everything else is 0 (This ensures that the Integral time which is I and
Derivative time which is D are not factors of the controller making it P only) except
the Set Point that should be set to 75mm.
2. Ensure once again that the mode of operation is set to automatic and click apply.
3. Click the green GO icon and observe the graph.
4. Once a clear trend is visible for L1, change the Proportional Band to 50.
5. Set the Set Point value back to 75mm and observe the trend as L1 drops to match the
Set Point.
6. Change the Proportional Band to 25.
7. Once the trend is clear, change the Set Point back to 1.5 L/min observe the trend.
8. When the trend once again becomes clear on the graph, click on the red STOP icon.

For the PI controller:

1. Click on the PID box and set the Proportional band to 50 and the Integral Time to 100.
Set Point which should be set to 75mm.
2. After the trend becomes clear, Integral Time to 50.
3. When the trend can be easily observed, change the Integral Time to 0.1.
4. Click the red STOP icon to finish the experiment after a trend can be clearly seen.

3
Calvin Tse-Liang Chin 17098128
5.0 Discussion
5.1 Introduction
The two different controllers known as the P Only Controller and the P and I
controller all have different ways of functioning and working. However, it is
clear that the most efficient controller to use in level control would in fact be the
P and I controller and the reason why will be discussed below.

P Only Controller
140

120

100

80

60

40

20

0
00:00 02:52 05:45 08:38 11:31 14:24 17:16 20:09 23:02

Level L1 _x000d_[mm] Pump Speed _x000d_[%] Proportional Band _x000d_[%] Set Point _x000d_

6.2 P Only Controller:

6.2.1 P Only Method of Control Graph

4
Calvin Tse-Liang Chin 17098128
6.2.2 Analysis of the P Only Method of Control
Simply scanning over the graph to the P Only Method of Control system
reveals that it has a high degree of offset between the set point line and
the L1 line. In comparison to a On/Off system, the pump is under much
less strain making it much more efficient that that respect. The large
offset is due to the Proportional Band, which can be simply defined as the
inverse of the gain. When the Proportional Band is large, the offset will
be large too but as it decreases (as seen when P=10%) the offset
decreases too. This method of level control can be considered too
inaccurate to be used in a realistic environment due to the high amount of
offset present.

6.3 P and I Controller:

P + I Controller
120

100

80

60

40

20

0
00:00 02:52 05:45 08:38 11:31 14:24 17:16 20:09

Level L1 _x000d_[mm] Pump Speed _x000d_[%] Proportional Band _x000d_[%] Integral Time _x000d_[s]
Set Point _x000d_

6.3.1 P and I Method of Control Graph

5
Calvin Tse-Liang Chin 17098128
6.3.2 Analysis of the P and I Method of Control
From the graph, it is clear that this is by far the best of the two options in
level control. The contribution of the Integral Time can be clearly seen by
the low offset when the Integral time is set to 100s. The combination of
the Proportional Band and the Integral Time allow the level to have a
very low offset and, in addition, allows the pump to maintain a steady
level, preventing possible damage. At the end the high oscillation and
offset is a clear example of the effects including an Integral Time has.

7.0 Conclusion
In conclusion, the best method of level control is in fact the P and I controller.
By having a high value of P, the offset in a P controller is large and it is slow to
react however the system is stable. By reducing the P value the offset is reduced
and the controller is more sensitive but the system can become unstable.
By reducing integral time offset can be reduced faster however the system can
become unstable. By a larger integral time the system is stable but less
sensitive, the offset is reduced slowly.
A controller with P is not desirable because of offset from the set point and a I
only controller is not desirable because it oscillates and doesn’t settle on the set
point so to achieve optimal control a combination of P+I should be used.