FACULTY OF CHEMICAL ENGINEERING

PROCESS CONTROL PRACTICES

(CPE 622)

LAB REPORT ON PID TUNING

GROUP: EH2206A

NAME STUDENT ID

MUHAMMAD SYAHSAIFULLAH BIN OSMAN 2016577903

NUR FADILAH BINTI HARUN 2016706453

NORANIERAH BINTI NOHO 2016728775

LECTURER’S NAME: MADAM SYAZANA MOHAMAD PAUZI

DATE OF SUBMISSION: 20/12/2019

TABLE OF CONTENT
NO. TITLE PAGES
1.0 SUMMARY 1
2.0 INTRODUCTION 2
2.1 THEORY ON LINEARITY, OPEN LOOP TEST, PID SETTING, 2-3
CLOSED LOOP TEST AND PROCESS STABILITY
2.2 PID CONTROLLER DESCRIPTION 3-4
2.3 P, I & D DESCRIPTION 5-6
2.4 PID TUNING EFFECT ON PROCESS OUTPUT 7-9
2.5 PID APPLICATION IN REAL PROCESS 10
3.0 OBJECTIVES 11
4.0 PROCEDURES 12
4.1 DCS-DELTA-V EMERSON 12-13
4.2 DCS FOXBORO 14-15
5.0 PROCESS LINEARITY LAB 16
5.1 FLOW CONTROL (FIC31) - FOXBORO 16-17
5.2 FLOW CONTROL (FIC21) - EMERSON 18-20
6.0 TANGENT METHOD LAB 21
6.1 RESULT 21
6.2 CALCULATION 22-25
6.3 DISCUSSION 25-26
7.0 REFORMULATED TANGENT METHOD LAB 27
7.1 RESULT 27
7.2 CALCULATION 28-31
7.3 DISCUSSION 32
8.0 NUMERICAL METHOD LAB 33
8.1 RESULT 33-34
8.2 CALCULATION 35-37
8.3 DISCUSSION 38-39
9.0 DISCUSSION 39-40
10.0 CONCLUSION 41-42
10.1 RECOMMENDATIONS 42
11.0 REFERENCES 43
1.0 SUMMARY

PID (proportional integral derivative) controllers are the most accurate and stable controller
using a control loop feedback mechanism to control process variables. A PID controller is an
instrument for regulating temperature, flow, pressure, speed and other process variables used in
industrial control applications. The experiment's main objective is to conduct open loop testing
and closed loop testing to verify the optimal condition for the process control loop system. Two
systems used for testing were DCS FOXBORO and DCS DELTA – V Emerson. FOXBORO and
EMERSON flow control loop show both a slightly nonlinear process response relationship as
well as a manipulated variable. A reduction in the Controller gain, the Kc value, slows the
system response, PV since PV is directly proportional to MV and MV is directly proportional to
Kc and shifts the process response curve to the right. A rise in PB delays the system response,
PV and moves the process response curve to the right. An increase in I will lead to a slower
process response, while PV will move the process response curve to the right. Numerical
approach provides the most precise data compared to the Tangent method and the Reformulated
Tangent method.

1
2.0 INTRODUCTION

2.1 Theory

Process Linearity

Process linearity is a study of process behaviour on changing manipulated variable. Process

linearity can be either linear or nonlinear. A linear process changes proportionally with changing
manipulated variables, whereas for a nonlinear process the change in process is not proportional
and not easily predictable.

Open Loop Test

Open-loop test is test that act solely based on the input and do not use feedback from the output
to self-correct while the test is running. Open-loop systems tend to be simple and economical as
they do not provide feedback from the device movement to the controller. The main difference
between an open-loop test and a closed-loop test is that the closed-loop system could self-correct
while the open-loop system doesn’t. Open loop test also known as non – feedback controls
system.

PID Setting

A proportional–integral–derivative controller (also known as PID controller. or three-term

controller) is a control loop mechanism employing feedback that is widely used in industrial
control systems and a variety of other applications requiring continuously modulated control.

Closed Loop Test

A closed loop test is also referred as a feedback control system. These tests record the output
instead of input and modify it according to the need. It generates preferred condition of the
output as compared to the original one. It doesn’t encounter any external or internal disturbances.

2
Process Stability

Process Stability refers to the consistency of the process with respect to important process
characteristics such as the average value of a key dimension or the variation in that key
dimension. If the process behaves consistently over time, then we say that the process is stable or
in control. The process distribution remains consistent over time.

2.2 PID Controller

Proportional-Integral-Derivative (PID) control is the most common control algorithm used in

industry. The acceptance of PID controllers can be attributed partly to their strong performance
in a wide range of operating conditions and partly to their functional simplicity, which allows
engineers to operate them in a simple, straightforward manner. The PID algorithm consists of
three basic coefficients; proportional, integral and derivative which are varied to get optimal
response.

1
( )= ( )+ ( ) + ( )

Equation 1

Proportional term: ( )

Integral term: (∫ ( ) )

Derivative term: ( )

Output term: ( )

3
The standard form of the PID algorithm computes its output CO (t) according to the formula
shown in equation 1. PV (t) is the process variable measured at time t, and the error e(t) is the
difference between the process variable and the set point. The PID formula weights the
proportional term by a factor of P, the integral term by a factor of P/TI, and the derivative term
by a factor of P.TD where P is the controller gain, TI is the integral time, and TD is the
derivative time.

The first feedback controllers included just the proportional term. For mathematical reasons that
only became apparent later, a P-only controller tends to drive the error downward to a small, but
non-zero, value and then quit. Operators observing this phenomenon would manually increase
the controller’s output until the last vestiges of the error were eliminated. They called this
operation "resetting" the controller.

As the integral term was introduced, operators observed that it would tend to perform the reset
operation automatically. That is, the controller would augment its proportional action just enough
to eliminate the error entirely. Hence, integral action was originally called "automatic reset" and
remains labelled that way on some PID controllers to this day. The derivative term was invented
shortly thereafter and was described, accurately enough, as "rate control’.

4
2.3 Actions of the P, I & D

Proportional Response

Proportional or P- controller gives output which is proportional to current error e (t). It compares
desired or set point with actual value or feedback process value. The resulting error is multiplied
with proportional constant to get the output. If the error value is zero, then this controller output
is zero.

This controller requires biasing or manual reset when used alone. This is because it never reaches
the steady state condition. It provides stable operation but always maintains the steady state
error. Speed of the response is increased when the proportional constant Kc increases.

Integral Response

Due to limitation of p-controller where there always exists an offset between the process variable
and set point, I-controller is needed, which provides necessary action to eliminate the steady state
error. It integrates the error over a period until error value reaches to zero. It holds the value to
final control device at which error becomes zero. Integrals control decreases its output when
negative error takes place. It limits the speed of response and affects stability of the system.
Speed of the response is increased by decreasing integral gain Ki.

5
While using the PI controller, I-controller output is limited to somewhat range to overcome the
integral wind up conditions where integral output goes on increasing even at zero error state, due
to nonlinearities in the plant.

Derivative Response

I-controller doesn’t have the capability to predict the future behaviour of error. So, it reacts
normally once the set point is changed. D-controller overcomes this problem by predicting future
behaviour of the error. Its output depends on rate of change of error with respect to time,
multiplied by derivative constant. It gives the kick start for the output thereby increasing system
response.

6
2.4 PID tuning effect on process output

Type of process response:

1. Underdamped

The system oscillates (at reduced frequency compared to the undamped case) with the amplitude
gradually decreasing to zero.

2. Overdamped

The system returns to equilibrium without oscillating.

3. Critically damped

The system returns to equilibrium as quickly as possible without oscillating.

Case Damping ratio Type of response

1 >1 Overdamped
2 =1 Critically damped
3 <1 Underdamped

7
Damping ratio: Dimensionless measure describing how oscillations in a system decay after a
disturbance.

Table of tuning rules

Tuning rules by Ziegler – Nicholas

The Ziegler – Nicholas rule is a heuristic PID tuning rule that attempts to produce good values
for the three PID gain parameters.

Kp: the controller gain

Td: the controller’s derivative time constant

Ti: the controller’s integral time constant

Settling criteria – QAD; performance tests – set point and disturbance in load variable.

Mode PB (%) I (TIME) D (TIME)

P 100 RR Td

PI 111.1 RR Td 3.33 Td

PID 83.3 RR Td 2 Td 0.5 Td

Tuning rules by Cohen – Coon

The Cohen-Coon tuning method isn’t suitable for every application. It can be used only on self-
regulating processes. Most control loops, for example, flow, temperature, pressure, speed, and
composition, are, at least to some extent, self-regulating processes. (On the other hand, the most
common integrating process is a level control loop).

8
A self-regulating process always stabilizes at some point of equilibrium, which depends on the
process design and the controller output. If the controller output is set to a different value, the
process will respond and stabilize at a new point of equilibrium.

Settling criteria – QAD; performance test – disturbance in load variables.

Mode PB (%) I (TIME) D (TIME)

100
P
1+

100 1+
PI 3.33
1+ 1+

100 1+ 0.37
PID 2.5
1.35 1 + 1+ 1+

Where, =

Selection for mode of control

Mode Process

P All processes. Process may not stabilise at the set point. (Off-set type)

PI Flow, pressure and liquid level

PID Temperature

9
2.5 PID in Real Process

The process control applications are now enormous in the industrial sector. In industrial
applications such as food processing, agriculture, filtration, effluent treatment, nuclear power
plants, pharmaceutical industries, water purification systems, industrial chemical production and
spray coating and boilers, liquid level control play an important role in all sectors. Typical
actuators used in liquid level control systems include pumps, motorized valves, on - off valves
and level sensors such as displacement float, capacitance probe and pressure sensor for feedback
control purposes so that the fluids can be regulated according to the process requirements.
(Farhad A., 2011).

In the level control system, the controller's aim is essentially to retain a level set point at a given
value and to be able to dynamically accept new set point values. Usually conventional PID is
used to control the level. In a Simulink model with a PID controller block, a PID gain is
automatically tuned. This allows a reasonable balance between efficiency and robustness to be
achieved by tuning the PID. It will automatically determine the plant's linear model. PID tuner
should find the plant as the combination of all blocks between the input and output of the PID
controller. The plant contains all the components in the control system, other than the controller
itself.

PID controllers are optimized to make the plant more general, achieve the minimum stable state
error and boost the other dynamic behavior. In a wide range of operating conditions, PID
controllers attributed performance to controls, their functional simplicity, and the ease with
which engineers can implement it using current computer technology. Therefore, PID control
algorithms are popular and give many advantages as ease of use, new development helps to
introduce other variants of PID controllers and control for specific application in the industry.

10
3.0 OBJECTIVES

1. To determine either the response curve is linear, slightly non-linear or roughly non-linear.
2. To perform open loop test using DCS Delta A-V Emerson and DCS FOXBORO.
3. To find the Td, Tc and RR value using tangent method, reformulated tangent method and
numerical tangent method.
4. To perform closed loop test using DCS Delta A-V Emerson and DCS FOXBORO.
5. To compute the PB and I value using Ziegler Nichol’s and Cohen-Coon’s method.
6. To undergo fine tuning to achieve the optimum PI controller setting.
7. To execute the correct method in order to do open loop test and closed loop test.

11
4.0 PROCEDURE
4.1 DCS-Delta-V Emerson

Open Loop Test

1. The control loops has been selected.

i. Boiler Drum Heat Exchanger Plant (LIC11)
ii. Liquid Flow Plant (FIC21)
2. The faceplate has been open by double clicked at the Controller example as
below, PIC92.

3. To view the trends, the process history view has been clicked.
4. The stabilized the process, the manual or auto mode has been selected.
5. The data of manipulated variables (MV in %) was been recorded for the initial value.
6. Once stabilized, the manual mode has been chosen.

12
 7. The manipulated variables have been changed in the range of 5-20%.
8. For self-regulating system: Once steady state reached, the result was
recorded by printed the response curve.
9. For non-self-regulating: Once the response curve can be calculated, that
response curved been printed.
10. Tangent and Reformulated tangent method has been used to determine the
time delay, response rate and time constant.
11. PI controller setting has been calculated using Ziegler Nichol’s and Cohen Coon’s
method.
Closed Loop Test

1. Controller has been set into auto mode.

2. Details icon and Faceplate has been clicked to set the controller setting.
3. PI value has been inserted as Kc (100/P) and I.
4. Fine tuning been done to reduce the oscillation and overshoot.
5. The optimum value of P and I have been recorded.
Load Disturbance Test

1. The controller has been set into manual mode.

2. MV has been change about 10% of current MV value.
3. Time taken was 3 seconds.
4. Then, the controller was set to automatic mode.
5. The result was recorded until the response become stable.
6. The result was recorded in the graph.
Set Point Test

1. Auto mode has been selected.

2. The set point has been change higher 10% from the current operating
process value.
3. The result was recorded until the response become stable.
4. The result has been combining with the load disturbance test.
5. The graph was printed.

13
4.2 DCS- FOXBORO

Open Loop Test

1) Water Flow and Level Control Plant has been selected.

2) The selected controller been activated
 FIC31
 LIC31
3) Stabilized the process in manual or auto mode.
4) The initial value of manipulated variables (MV in %) has been recorded.
5) The auto mode then changed into manual mode to perform the open loop test.
6) The manipulated variables have been change in range 5-20% of the initial manipulated
variables.
7) For self-regulating system: Once steady state reached, the result was recorded by printed
the response curve.
8) For non-self-regulating: Once the response curve can be calculated, that response curved
been printed.
9) The data has been collected from AIM Historian Data Display at desktop.
10) Numerical method has been chosen to calculate the Response rate, time delay, and time
constant.
11) PI controller setting has been calculated using Ziegler Nichol’s and Cohen Coon’s
method.

Closed Loop Test

1) The controller set to auto mode.

2) PB and I value has been inserted in the PBAND and INT.
3) Fine tuning been done to reduce the oscillation and overshoot.
4) The optimum value of P and I have been recorded.

14
Load Disturbance Test

1) The controller has been set into manual mode.

2) MV has been change about 10% of current MV value.
3) Time taken was 3 seconds.
4) The result was recorded in the graph.

Set Point Test

1) Auto mode has been selected

2) The set point has been change higher 10% from the current operating process value.
3) The result was recorded until the response become stable.
4) The result has been combining with the load disturbance test.
5) The graph was printed.

Process Span

I. EMERSON
 Flow = 6 m3/h
 Level = 1000 mm
II. FOXBORO
 Flow = 3 m3/h

15
5.0 PROCESS LINEARITY LAB

5.1 Flow Control (FIC31) – FOXBORO

Table shows process linearity against manipulated variable using FOXBORO system in manual
mode. The experiment was set to flow controller which is FIC 31 with a constant set point of 1.5
m3/h. Then, constantly changing the manipulated variable of process with increment 10%
starting at 0% and ending at 100%. From this experiment, the data below is obtained.

Table 1 - Process Steady State Values for each Manipulated Variable

MV, % PV, m3/h

10 0
20 0
30 0.56279
40 0.99334
50 1.4091
60 1.9008
70 2.221
80 2.4261
90 2.4562
100 2.4562

16
 Process Response Vs Manipulated Variable

2.5

2
PV, M3/H

1.5

0.5

0
0 20 40 60 80 100 120

MV, %

Figure 1 - Plot of PV versus MV (FOXBORO)

Process linearity was determined by plotting a graph of PV against MV. Based on graph, it can
conclude that this process is slightly nonlinear against the manipulated variable, since PV is
directly proportional to MV.

17
5.2 Flow Control (FIC21) – EMERSON

Table shows a process linearity against manipulated variable using EMERSON system in manual
mode. The experiment was set to flow controller which is FIC 21 with a constant set point of 3.0
m3/h. Then, constantly changing the manipulated variable of process with increment 10%
starting at 0% and ending at 100%. The process span is 6 m3/h. From this experiment, data PV
versus MV is obtained from Figure 2.

Figure 2 - Multiple Steps Test of Flow Control Loop

18
 Table 2 - Process Steady State Values for each Manipulated Variable

MV (%) PV (m3/h) PV (%)

10 0.6 10

20 0.8 13.33
30 1.2 20

40 1.7 28.33
50 2.4 40

60 3.4 56.67

70 4.6 76.67
80 5.1 85

90 5.3 88.33
100 5.4 90

Change the unit of PV from m3/h to % using the following conversion:

PV
PV (%) = × 100%
process span

At MV = 10%,

0.6 /ℎ
PV (%) = × 100% = 10%
6 /ℎ

19
 Process Response vs Manipulated Variable
100

90

80

70

60
PV (%)

50

40

30

20

10

0
0 20 40 60 80 100 120

MV (%)

Figure 3 - Plot of PV versus MV (EMERSON)

Process linearity was determined by plotting a graph of PV against MV. Based on graph, it can
conclude that this process is slightly nonlinear against the manipulated variable, since PV is
directly proportional to MV.

20
6.0 TANGENT METHOD LAB
6.1 Results
Flow control (FIC21) – Emerson

MV = 10%

Process span = 6.0 m3/h

Parameter Response rate, RR Dead time, Td Time constant, Tc

Value 0.6745/s 0.9232 s 3.004 s

Optimum controller setting

Tuning rule by Cohen – Coon

Mode of control: Proportional – Integral mode

Process: liquid flow

Mode 1st trial 2nd trial

Proportional band (%) 60.5754 121.21

Integral (time) 1.8845 3.769

Kc 1.6508 0.825

21
6.2 Calculations

An open loop response was done where a change in manipulated variable was made from 50% to
60%. The process dead time (Td), response rate (RR) and time constant (Tc) was obtained from
open loop test by using tangent method. The process span is 6.0 m3/h.

Data calculated from the graph

MV initial (%) MV final (%) MV (%)

50 60 10

Measure the y – axis scaling factor (a)

.
ℎ
= × 100% /14

.
ℎ
.
× 100%
ℎ
=
14

= 0.5952 %

Measure the x -axis scaling factor (b)

15
= = 0.2308 ⁄
65

Measure the distance between the horizontal lines

Δ = 34

Measure the distance between the vertical lines

Δ = 13

22
The process dead time, Td

Td (length) = 4 mm

Td (time) = Td (length) x b

= 4 mm x 0.2308 sec/mm

= 0.9232 sec

Response rate (RR)

Δ /Δ
=
Δ

34/13 0.5952
=
10 0.2308

= 0.6745/

The process time constant, Tc

Tc (length) = 13 mm

Tc (time) = Tc (length) x b

= 13 mm x 0.2308 sec/mm

= 3.004 sec

23
Tuning rule: Cohen – Coon

Controller mode: Proportional – Integral mode (PI)

= 0.9232 3.004

= 0.3077

Proportional band, PB (%)

PB=

PB= . (0.6745)(0.9232)

PB = 60.5754%

Integral (Time)

I= 3.33

I= 3.33 ( . ) (0.9232)

I = 1.8845 sec

Gain, Kc

Kc =

Kc =
.

Kc = 1.6508

24
 Figure 4 - Process Response for Closed Loop Test, Load Disturbance Test and Set
Point Test.

6.3 Discussions

From the open loop test, we can obtain the data on response rate (RR), dead time (Td) and time
constant (Tc) of the process control. We use tangent method to gather the information for the
DCS – Emerson (FIC21).

In order to run the experiment, first we change the manipulated variable, by increasing 10% from
the process lifespan. The increment goes up from 50% to 60%. As the results, the response rate,
dead time and time constant were 0.6745/sec, 0.9232 sec, and 3.004 sec respectively.

These data were used for setting the optimum controller for the FIC21. Cohen – Coon was used
as the tuning rule in finding the values for proportional band (PB) and integral time (I). From the
calculated data, the PB and I value were 60.5754% and 1.8845 sec, respectively. The control
mode for the FIC21 is PI due to the process involved is the regulation of liquid flow. This
present of integral time mode will reduce the off – set resulting from the process. Since the

25
FIC21 use Kc value, thus this value must to be determine, resulting from the PB value divided by
100.

In the first attempt, the result is shown as in the figure, an oscillation was observed on the graph.
This shown that the inserted value of Kc and I are far from the optimum condition. Thus,
according to the PID tuning method, we are dividing the Kc value by 2 and multiplying the
integral value by 2. As the result, the oscillation reduced and the MV linear to the SP line. From
this result, we can say that, a decrease in Kc resulted in the slower action of MV since MV is
directly proportional to Kc. A slower action of MV will result in a slower process response (PV)
since PV is directly proportional to the MV. A slower process response will shift the graph to the
right.

The load disturbance test was continued since the optimum condition for the loop test had
obtained. The load disturbance test was resulted from the change of 10% from the current
MV.As the result, the initial MV is 45.2% and the final value is 55.2%. From the graph, it shown
that the MV line increase and the valve line also increase. But after a few second, the MV line
reach to the SP as it is self - regulated process mode.

The last experiment which is set point test was carried out. A change was made in set point by
increasing about 10%from current operating process value. By using the same value from the 2nd
calculated Cohen – Coon for Kc and I and the result for set point test was collected. From the
graph, it showed a small oscillation for few second, but then the graph tuning to the new set point
without changing the Kc and I values.

26
7.0 REFORMULATED TANGENT METHOD LAB

7.1 Result

Level Control (LIC11) – EMERSON

MV = 10%

Process span = 1000 mm

Parameter Response rate, RR Dead time, Td

Value 0.0769/s 7.0588 s

Optimum controller setting

Tuning rule by Ziegler-Nichols

Mode of control: Proportional – Integral mode

Process: Liquid level

Mode 1st trial

Proportional band (%) 60.31

Integral (time) 23.51

Kc 1.66

27
7.2 Calculations

An open loop response was done where a change in manipulated variable was made from 50% to
60%. The process dead time (Td), response rate (RR) and time constant (Tc) was obtained from
open loop test by using tangent method. The process span is 1000 mm.

Figure 5 - An Open Loop Process Response of Liquid Level Control

Data calculated from the graph

MV initial (%) MV final (%) MV (%)

50 60 10

Measure the y – axis scaling factor (a)

100%
=
17

=5.8824 %

28
Measure the x -axis scaling factor (b)

60
= = 1.7647 ⁄
34

Measure the angle ( )

= 13o

The process dead time, Td

Td (length) = 4 mm

Td (time) = Td (length) x b

= 4 mm x 1.7647 sec/mm

= 7.0588 sec

Response rate (RR)

tan θ
=
Δ

tan 13 5.8824
=
10 1.7647

= 0.0769/

29
Tuning rule: Ziegler-Nichols

Controller mode: Proportional – Integral mode (PI)

Proportional band, PB (%)

PB (%) = 111.1

= 111.1 (0.0769/ ) (7.0588 s)

= 60.3075 %

Integral (Time)

I (Time) = 3.33

= 3.33 (7.0588 s)

= 23.5058 s

Since Emerson system used Kc value instead of PB value,

Gain, Kc

Kc =

Kc =
.

Kc = 1.6582

30
The value Kc and I is then used for closed loop testing with load disturbance test and set-point
test for open loop process response. Figure 4 shows the three test results.

Figure 6 - Process Response for Closed Loop Test, Load Disturbance Test and Set
Point Test.

31
7.3 Discussion

Information on response rate (RR), dead time (Td) and time constant (Tc) of process control can
be obtained from the open loop test. We use reformulated tangent method approach to collect
information for DCS – Emerson (LIC11).

First, we change the manipulated variable to run the experiment by increasing 10 percent from
the lifespan of the process. The rise will go up from 50% to 60%. As the results 0.0769/ and
7.0588 sec respectively were the response rate and the dead time.

These data have been used to set the optimum LIC11 controller. Ziegler-Nichols was used to find
values for proportional band (PB) and integral time (I) as the tuning rule. The original value of
PB and I for the experiment using the reformulated tangent method is 60.3075 % and 23.5058 s,
respectively, from the calculated part. The control mode for the LIC11 is PI since liquid level is
involved in the process. First of all, this equation, Kc = 100/P, transforms the value of P to Kc
since LIC11 use Kc value instead of P value which then Kc = 1.6582.We then performed a
closed loop test and performance test to ensure that these values were acceptable for the system.
The result is shown in the first attempt as the figure shows a slight oscillation on the graph but
after a while the oscillation decreased and the MV linear to the SP line.

Since the optimum condition for the loop test was reached, the load disturbance test was
continued. The load disturbance test was the result of a 10% change from the existing MV.
Correspondingly, the initial MV is 68% and the final value is 78%. It showed from the graph that
the MV line is increasing and that the valve line is also increasing. But after a few seconds, as it
is self-regulated process mode, the MV line reaches the SP.

The last test to be performed was the set point test. A change was made at the set point by
increasing the current operating process value by about 10%. In which the initial SP is 400 and
the final value is 500.By using the same value from the calculated Ziegler-Nichols for Kc and I,
the result was collected for the set point test. It showed for a few seconds from the graph a small
oscillation, but then the graph tuning to the new set point without changing the values of Kc and
I.

32
8.0 NUMERICAL METHOD LAB

8.1 Result
Flow Control (FIC31) –FOXBORO

Table shows a discrete open loop process response of water flowrate control after a change in
manipulated variable was made from 60% to 70%. Result from data was tabulated based on
experiment. The process response rate (RR), dead time (Td) and time constant (Tc) are calculated
based on Tangent Method, numerically. The process span is 3m3/h.

Time (s) PV (m3/h) PV(%) RR (1/s)

0 1.9155 63.85 0.0000

1 1.9155 63.85 0.0185

2 1.9265 64.22 0.2610

3 2.0721 69.07 0.3790

4 2.1541 71.80 0.2030

5 2.1940 73.13 0.0995

6 2.2137 73.79 0.0500

7 2.2240 74.13 0.0170

8 2.2240 74.13 0.0000

9 2.2240 74.13 0.0000

∆h = 1.0 s, ∆MV = 10 %, PVi = 63.85 %, PVf = 74.13 %

33
 Parameters Values
Dead Time (Td), min 0.01504
Time Constant (Tc), min 0.0452
Response Rate (RR), 1/min 22.74

Optimum controller setting

Tuning Rules by Ziegler Nichols’s (ZN)

Controller mode: Flow → Proportional - Integral (PI)

Test Closed Loop Load Disturbance Set Point

PB (%) 37.997 151.9 151.9
I (min) 0.02 0.02 0.02
MVi = 51.34 % SPi = 1.5 m3/h
MVf = 61.34 % SPf = 1.8 m3/h

34
8.2 Calculation

Change the unit of PV from m3/h to % using the following conversion:

PV
PV (%) = × 100%
process span

At time = 1 s,

1.9155
PV (%) = × 100% = 63.85 %
3

All the data for PV in % is determined for each respective time.

Response rate, RR

PV − PV
RR =
2∆h ∆MV

At time = 3 s,

PV − PV 71.80 − 64.22
RR = = = 0.379/
2∆h ∆MV 2(1.0)(10)

All the data for RR is determined for each respective time. RRmax is determined from the
highest value of RR presence in the calculated data of RRs.

0.379 60
RR = = 22.74 /
1

35
Process Dead Time, Td

PV − PV
= − 2∆ℎ
PV − PV

The dead time is determined at maximum RR. Therefore, at time = 3s,

PV − PV 71.80 − 63.85
= − 2∆ℎ = 3 − 2(1) = 0.9024
PV − PV 71.80 − 64.22

For FOXBORO system, unit for time must be converted to min. Therefore,

1min
= 0.9024 = 0.01504
60 s

Time Constant, Tc – (Tangent Method)

PV − PV
= 2∆ℎ
PV − PV

The time constant is determined at maximum RR using tangent method. Therefore, at time = 3s,

PV − PV 74.13 − 63.85
= 2∆ℎ = 2(1) = 2.7124
PV − PV 71.80 − 64.22

For FOXBORO system, unit for time must be converted to min. Therefore,

1min
= 2.7124 = 0.0452
60 s

Tuning Rules by Ziegler Nichols’s (ZN)

Controller mode: Flow → Proportional - Integral (PI)

36
Proportional band, PB (%)

PB (%) = 111.1 RR Td

PB (%) = 111.1 (22.74 /min )(0.01504 min ) = 37.997 %

Integral (Time)

I (min) = 1.33 Td

I (min) = 1.33 (0.01504 min ) = 0.02 min

The PB and I value are then used for closed loop test proceeding with load disturbance test and
set point test for closed loop process response. Figure below shows the result for three tests that
has been conducted.

Figure 7 - Process Response for Closed Loop Test, Load Disturbance Test and Set Point Test.

37
8.3 Discussion

An open loop test has been performed to control the flowrate of water by using FIC 31
equipment. The process response should be in self-regulating process which is manual mode as
making the manipulated variable changed from 60% to 70% with 10% of increment. The test is
done as the process response reaching the steady state without oscillates. The graph will shows
oscillate, if the process response is not steady state. Then, the value of PV and MV were obtained
from graph. These value used to calculate the response rate (RR), dead time (Td) and time
constant (Tc) by using Tangent Method, numerically. The value of PV and MV obtained should
be in %.

Besides, numerical method can be calculated simply using Microsoft Excel. For numerical
method, there are various types of RR but the highest RR will be considered and being used to
calculate Td and Tc. From the calculation, RR value was 22.74 min-1, Td was 0.01504 min and
Tc was 0.0452 min. These values are used in closed loop test to calculate the PI controller setting
by using Ziegler Nichols’s method. For closed loop test, the value of the optimum proportional
band, PB is 37.997 % while the integral time, I is 0.02 min. There is no value for derivative time,
D since the value of PB, I and D is considered based on the types of process. The mode of
control for this experiment is PI mode as it in flow control process, so the tuning test only
considered on PB and I value only. Close loop test should be in automatic mode.

Based on experiment, the response oscillates after insert PB and I, do fine tuning to achieve
stable condition. The value for PB was change after tuning which is 151.9 % while the value of I
does not change. The graph shows no oscillates, this means that the value is optimum value for
this process. Then, the load disturbance test was conducted in manual mode by changing the
value of manipulated variable from 51.34 % to 61.34%. After 3s, set the controller to automatic
mode. The graph shows that the process response is simply quiet fast to response during the test.
This means that no need to tuning because PB and I value inserted was the optimum value for
this process.

Lastly continue with set point test by changed the set point about 10% of current operating
process value which is 1.5 m3/h to 1.8 m3/h. The purpose of set point test is to see how fast the
processes can response to the change of set point. The system must be in automatic mode. The

38
graph shows that the process response was quiet fast so there is no oscillates. This shows that,
the value was in optimum value for this process as the process in stable condition.

9.0 DISCUSSION
Table 3- Summary of Process Linearity
System Type of Control Process Linearity
FOXBORO Flow Slightly nonlinear
EMERSON Flow Slightly nonlinear

Process linearity is a process behavior study on changing the manipulated variable in which can
be linear and nonlinear. The flow control loop undergoes multiple stage tests in the process
linearity laboratory by constantly changing the controlled variable by 10 % from 0 % to 100 %.
From the graph it can be seen that both processes move slightly away from the linear thread line
of the process response to prove that the process response to the manipulated variable is slightly
nonlinear. For a nonlinear process the change in process is not proportional and not easily
predictable.

Table 4- Summary of Proportional – Integral (PI) Mode for the Respective Lab

Method Controller Before Optimization After Optimization

PB I (Time) PB I (Time)
(%) (s) (%) (s)
Tangent FIC21 60.5754 1.8845 121.21 3.769
Method (Emerson)
(Cohen–
Coon)
Reformulated LIC11 60.3075 23.5058 60.3075 23.5058
Tangent (Emerson)
Method (Ziegler –
Nichols)
Numerical FIC31 37.997 0.02 151.9 0.02
Method (Foxboro)
(Ziegler –
Nichols)

39
The data tabulated in Table 4 for Integral Time is considered in second, s, in order to see a
contrast between the data value for the entire lab involved in view of the fact that the time unit
used for closed loop test, load disturbance test and set-point test is in minute, min for FOXBORO
system.

A decrease in controller gain, Kc value, resulting in a slower manipulated variable, MV action

since, Proportional Band (PB) with Kc is inversely proportional, the PB value increases as Kc is
reduced. As PV is directly proportional to MV, this in effect slows down the process response.
The response curve of the process will be moved to the right. As I is inversely proportional to
MV, a change in I makes the controller action slower. A slower response of MV would result in
a slower response of the process, because MV is directly proportional to PV. A slower response
to the process would change the response curve to the right. The effect is a reduction in the
number of oscillations and a stable response to the process that reaches the set point.

40
10.0 CONCLUSION

As the conclusion, this experiment has achieved its objective. For process linearity lab, both flow
control loop system which is FOXBORO (FIC 31) and EMERSON (FIC 21) shows slightly
nonlinear relation of process response and manipulated variable. It shows that the process
response is not proportional to manipulated variable. The open loop test must be performed using
DCS Delta A-V Emerson (FIC 21) and DCS FOXBORO (FIC 31), in order to determine the
value of Response rate (RR), Dead Time (Td) and Time Constant (Tc) using tangent method,
reformulated tangent method and numerical tangent method. These values of RR, Td and Tc are
substituted into Tuning Rules by Ziegler Nichol’s and Cohen Coon’s method to be used for
closed loop test in order to achieved optimum control value process. Once the process
characteristics have been identified, the PI mode is selected because it used for fast flow process
response. As the result for open loop test, the numerical method is the most accurate data
corresponding to the tuning rules compare to tangent method and reformulated tangent method,

For closed loop test, the test was performed using DCS Delta A-V Emerson (FIC 21) and DCS
FOXBORO (FIC 31). For FIC 21 loop used the value of Kc and I while for FIC 31 loop used the
value of PB and I, these value are inserted onto gain and reset section respectively. If the process
oscillates and not stable the value of Kc will reduce by divide to 4 and increase value of I by
times to 4. The test is finished when the graph shows no oscillates, this means that the value
inserted is the optimum value for this process. Then continue to the load disturbance test, this test
making a change of MV about 10% from the current MV until the process become stable. To
make a change of MV, the system should be in manual mode, after 3s changes the mode into
automatic mode. Lastly fro set point test, this test changes the set point about 10% of current
operating process value. The system for this test must be in automatic mode, if the response
oscillates, does fine tuning to achieve the stable condition. This test is finished as the process
response is stable. Hence, the closed loop test, load disturbance test and set point were done as
the process stable in the optimum condition for the process control loop. As the result, an
increase in PB value makes MV action slower, a slower action of MV result in a slower of
process response and the new process response curve shift to the right. A decrease in Kc results

41
in slower action of MV, as slower action of MV result in a slower of process response and the
new process response curve shift to the right. Lastly, an increase in value makes MV action
slower, a slower action of MV result in a slower of process response and the new process
response curve shift to the right

10.1 Recommendations

There are several recommendations on process output optimization and improvement to ensure
the result is accurate. Firstly, minimize the error over time for linearity test by adjustment of
manipulated variable onto system correctly, in order to determine the process behaviour on
changing the manipulated variable which is linear or nonlinear. Secondly for tangent method and
reformulated tangent method, when drawing a horizontal line or vertical line, ensure that the
horizontal or vertical line is parallel to a reference line such as x-axis, y-axis or grid line because
the skill of line drawing and measurement is very important towards calculation accuracy. For
the numerical method, make sure to keep at least three significant figures when rounding-up
number, in order to used it for later calculation and to avoid inconstant value due to decimal
places. Then, the unit for all calculation should be consistent and fit in equation to avoid wrong
final result. Inconsistent of unit will lead many disastrous end result such as wrong controller
setting and either oscillatory, sluggish or runaway. Other than that, careful improved the tuning
by inserted the right value of PB, I and Kc on the system because the performance may be
unacceptable with poor tuning. Tuning should be doing when the response shows oscillates and
it done when the process response is in stable condition due to the value inserted is the optimum
value for the process response. Lastly, the system mode for every test must be in right mode such
as the open loop test must be in manual mode and close loop test in automatic mode. Setting the
wrong system mode will lead the performance of process response incorrectly.

42
11.0 REFERENCES

Aslam, F., & Haider, M. Z. (2011). An Implementation and Comparative Analysis of PID
Controller and their Auto tuning method for Three Tank Liquid Level
Control. International Journal of Computer Applications, 21(8), 42–45. doi:
10.5120/2529-3443

Engineering, O. (2019, April 17). What is a PID Controller? Retrieved from

https://www.omega.com/en-us/resources/pid-controllers.

Open-Loop vs Closed-Loop Systems in the Materials Testing Industry. Admet. Retrieved from

https://www.admet.com/open-loop-vs-closed-loop-systems-materials-testing-industry/

PID tuning improves process efficiency. (2013, December 12). Retrieved from

https://www.controleng.com/articles/pid-tuning-improves-process-efficiency/.

Understanding PID control and loop tuning fundamentals. (July 26,2016). Retrieved from

https://www.controleng.com/articles/understanding-pid-control-and-loop-tuning-
fundamentals/

what is process linearity? Control engineering. (October 29,2010). Retrieved from

https://www.controleng.com/articles/what-is-process-linearity/

43