Addis Ababa Science and Technology University

College of Biological and Chemical Engineering

Department of Chemical Engineering

Process control lab

1
Table of Content

s
CHAPTER 1: INTRODUCTION OF CONTROL SYSTEM................................................................4

1.1. Elements of A Control System.........................................................................................................6

1.2 Ziegler-Nichols Closed-Loop Method (Ultimate Gain).....................................................................8

CHAPTER 2: TYPES OF CONTROLLER..........................................................................................10

2.1. On-Off Control...............................................................................................................................10

2.2. The Proportional Action.................................................................................................................11

2.3. Proportional-Plus-Integral Control.................................................................................................13

2.4. Proportional-Plus-Integral-Plus-Derivative Control.......................................................................17

2
CHAPTER ONE

INTRODUCTION OF CONTROL SYSTEM

Suppose to keep a constant level of liquid in a vessel: the liquid is continuously drawn from this
container which can be filled up through a feed cock. When the level lowers, the cock must be
opened some more, and vice versa. If the operator would accidentally open the cock too much,
the level would rise and the vessel could overflow. In the opposite case, the vessel could empty.
In other words, the cock could be kept open although the liquid overflows, and in the opposite
case it would be kept closed although the vessel is empty. This is an example of an open-loop
control system. The word "open" means that the final result of the control - the level - cannot
directly affect the control element represented by the cock. In fact the relationship between the
height of level and the opening or closing of the cock is carried out by the operator who
measures the level, compares this value with that desired and consequently turns the cock wheel
in the right direction. Now provide the vessel with a float connected to the shutter of the cock
through a compound lever. In this case, when the level lowers for a considerable drawing of
liquid or an excessive closing of the feed cock, the float too lowers opening the cock some more.
Therefore, the level tends to rise again or, however it will lower more slowly. Thanks to the float
connected to the cock the final object of the control (the level of liquid) directly affects the
control element (the cock) without external interventions. In fact, this new mechanism can carry
out the measurement of level corresponding to the position of the float in the compound lever
and correct the position of the control element varying the opening of the same cock. This is a
closed-loop control system. In fact, the automatic control consists in carrying out this type of
closed loop (or feedback) control systems. The elements of an automatic control system are
clearly shown in a block diagram below.

3
 Figure. 1.1 elements of an automatic control system

In this diagram the process is represented by the whole set of equipment and physical actions
forming the operation under control; the controlled variable is one of the various characteristics
(temperature, pressure, level, pH, etc.) that specify a state of a point or a zone of the process: it is
the object of the control system; the measuring instrument is the apparatus that measures the
value of the controlled variable and sends it to the control instrument; the control instrument is
the equipment that receives the measure of the controlled variable, compares it with the value of
the same variable having to be kept and being pre-fixed through an external operation, and
transmits a corrective action to the actuator (the intensity and direction of this action depend on
the result of the comparison); the actuator receives and transforms this corrective signal into a
variation of the controlled variable. Refer to the example of the control of level mentioned
above: the process is the system including the vessel, the liquid and its movements; the
controlled variable is represented by the height of liquid in the vessel; the measuring instrument
is the float: in fact, it sends the measuring signal corresponding to the buoyancy received by the
liquid, to the control instrument (the compound lever connecting the float to the shutter of the
cock). The actuator is the cock: varying its opening under the action of the control instrument
provokes a variation of the manipulated process variable (the flow-rate of the liquid, in this case)
and the flow-rate directly affects the controlled variable, that is the level.

4
1.1. Elements of A Control System
The main components of a control system are the sensing element (sensor): it measures the
instantaneous value of the controlled variable; the control instrument: it compares the
instantaneous value of the controlled variable which has been measured by the sensor, with the
value desired and sends a corrective signal to the actuator; this signa depends on the difference
between these two quantities; the actuator: it carries out the corrections sent by the control
instrument; the actuator varies a characteristic quantity of the control fluid called manipulated
variable.
In this case, the manipulated variable is the flow rate of the control fluid. The overall actions of
these three elements make that the value of the controlled variable coincides with the value
desired. This last value is fixed during the definition of the operational conditions of the process
and is the final target of the control system. The desired value of the controlled variable is called
"SET-POINT". Fig 1.1 shows an example of trend of a controlled variable recorded by a control
instrument in a generic process with automatic control.

Fig 1.2 Recording curves of a control instrument

As it is shown in the figure, the reaction of the control system to eventual deviations of the
controlled variable consists in resetting the set-point. The difference between the actual value of
the variable and the set-point is defined error. Sometimes the control system can keep the
variable around a constant value differing from the set-point. In this case, the control point
differs from the set-point for a quantity called offset. This condition is shown in the fig. 1.2
A control system has been conceived to keep a condition of dynamic equilibrium constant; it is
difficult that recording the value of the controlled variable in time generates a straight line.

5
Generally, in steady state its trend will be slightly wavelike. On the other hand, this very slight
undulation is the clearest evidence that the system is actually operating.

Fig. 1.3 Set-point, offset and error

In a process, being all the other conditions equal, each value of the controlled variable depends
on a value of flow rate of the control fluid. Then, for any process it is possible to draw a curve
corresponding to the flow rate of control fluid required to maintain any value of the controlled
variable. This is called load curve (fig.1 .4).

Fig 1.4 Load curve of a process

6
In an ideal process in steady state, the value of the controlled variable would be kept strictly
constant without any intervention of the control system. The flow-rate of the control fluid
required to keep the controlled variable at a certain set-point would be constant too in time and it
would only vary versus the set-point. Therefore, in this process the load curve would be constant
in time.
On the contrary, actually industrial processes are unavoidably affected by some disturbances
tending to provoke deviations of the controlled variable from the set-point. The causes of these
disturbances are different, such as:
 unexpected variations in the flow rates of the fluids of process;
 variations of the nature or composition of these fluids involving changes of their
characteristics;
 variations of their inlet temperatures or of their thermal contents;
 variations of the quantity of heat lost for changes in ambient temperature;
 Variations of the quantity of heat generated or absorbed by a chemical reaction.
Whatever the cause may be, any disturbance affecting a process under control provokes the
intervention of the control system: it0s action is aimed to correct the deviation occurred in the
process through a variation of the flow rate of the control fluid. It ensures that the flow rate
of the control fluid required to keep the controlled variable coinciding with a certain set-point
undergoes some variations in time for the disturbances mentioned above. For example, if the
fluid to be heated enters a heat exchanger at a higher temperature, there is a reduction of
the flow rate of the heating fluid necessary to maintain a certain outlet temperature of the fluid
heated.
Actually, the load curve of a process undergoes continuous changes that are called load
variations.
A load variation can actually be converted into a translation of the load curve in the direction of
the axis of abscissas, or into a variation of slope, or into a variation of its shape, according to the
various cases. This means that the same value of the controlled variable can be obtained through
a different flow rate of the control fluid.

1.2 Ziegler-Nichols Closed-Loop Method (Ultimate Gain)

7
In closed feedback loop, if the total amount of signal amplification provided by the instruments
is too much, the feedback loop will self-oscillate at the system’s natural (resonant) frequency.
While oscillation is almost always considered undesirable in a control system, it may be used as
an exploratory test of process dynamics if the controller acts purely on proportional action (no
integral or derivative action): providing data useful for calculating effective PID controller
settings.
Thus, a “closed-loop” PID tuning procedure entails disabling any integral or derivative actions in
the controller, then raising the gain value of the controller just far enough that self-sustaining
oscillations ensue. The minimum amount of controller gain necessary to sustain sinusoidal
oscillations is called the ultimate sensitivity (Su) or ultimate gain (Ku/Gu) of the process, while
the time (period) between successive oscillation peaks is called the ultimate period (Pu/Tu) of the
process. We may then use the measured values of Ku and Pu to calculate reasonable controller
tuning parameter values (Kp/G, τi/I, and/or τd/Td).

When performing such a test on a process loop, it is important to ensure the oscillation peaks do
not reach the limits of the instrumentation, either measurement or final control element. In other
words, in order for the oscillation to accurately reveal the process characteristics of ultimate
sensitivity and ultimate period, the oscillations must be naturally limited and not artificially
limited by either the transmitter or the control valve saturating. Oscillations characterized by
either the transmitter or the final control element reaching their range limits should be avoided in
order to obtain the best closed-loop oscillatory test results.

8
CHAPTER TWO

TYPES OF CONTROLLER

The relationship existing between the output signal and the input signal states the type of control
carried out by the instrument.
The main types of control carried out by the instruments of industrial use are:
1) on-off control
2) Proportional control
3) Integral control
4) Derivative control.
The terms "on-off", "proportional", "integral" and "derivative" are referred to a particular form of
the function linking the output and input signals of the instrument.
The various controls can also be combined with each other in the same instrument: in this case
the output signal is the result of the sum of the signals corresponding to each control.
The various controls will be examined either from a theoretical point of view and as regards their
actual application, in this and in the following chapters: the peculiar properties of each control
will be exhaustively explained so that the most suitable combination can be chosen for each
process.
2.1. On-Off Control

This is the simplest type of control: the output signal can take only two values corresponding to
only two positions of the final control element according to whether the value of the controlled
variable is lower or higher than the set-point.

9
A very common application of automatic on-off control concerns electrical water heaters.
2.2. The Proportional Action
A control instrument carries out a proportional control when there is a biunivocal proportion
between the output and input signals. The characteristic equation of a proportional controller can
be written as follows: y = y0 ± k(x - x0)
where y and y0 are the instantaneous and initial values of opening of the control valve, x and x0
are the instantaneous and initial values of the variable, whereas k is the factor of proportionality.
The sign ± depends on the type of process under control: it is plus when the valve opens more as
the variable increases, whereas it will be minus in the opposite case, that is, when the valve
opening is reduced as the variable increases.
Proportional band of an instrument results from the ratio between the difference of the
values of the variable corresponding to the extreme positions of the control valve and the
difference of the values of the variable limiting the operating range of the instrument.
Figure below shows some operating lines of a proportional controller corresponding to various
values of the proportional band.

10
 Fig.2.2.l a. Variation of the proportional band
As it is shown, the proportional band can also exceed 100%; in this case the control valve never
attains the position of total closing and opening and its range is included between two
intermediate positions. As wider the proportional band is, as lower the variation of valve opening
(corresponding to a certain variation of the variable) is. Therefore, the sensitivity of the
instrument also depends on the amplitude of the proportional band. As narrower the proportional
band is, as higher the sensitivity is. In instruments with very narrow proportional bands, even a
slight variation of the variable is sufficient to make the control valve covers a very wide range.
When the proportional band is null, this eventuality falls within the case of the on-off control
already considered. Also offset increases as the proportional band increases.

The fig. 2.2.2 shows the characteristics of a process with automatic proportional control for a
step variation of load.
Example: a proportional controller having an operating range of 30 to 130 °C must keep the
temperature of a liquid at the set-point of 80 °C. The corresponding opening of the control valve
is 50%, whereas the proportional band of the instrument is set at 20%. Which are the values of
valve opening corresponding to a temperature of 74 °C, and those of the opening speed of the
same valve corresponding to a temperature drop of the fluid of 6 °C per minute?
percentage error corresponds to e = (x0 - x) / (x2 - x1)

where p, e and b are respectively the opening degree of the control valve (in percentage) and the
percent error and proportional band.

11
 p0 = 0,5
b = (90 - 70) / (130 - 30) = 0,2
e = (80 - 74) / (130 - 30) = 0,06
from the equation (4.6) it results:
p = 0.5 + 0.06 / 0.2 = 0.8

During the design phase any effort is made to fix the intermediate position of the control valve
corresponding to the set-point of the controlled variable.
In this way it has equal possibilities of action in both directions, widening or reducing its
opening according to the direction of deviation of the variable.
The gain of a control system, considered as ratio between effect and cause, depends on the
proportional band of the instrument. It increases as the proportional band decreases.
A system controlled starts oscillating as easier as longer the delays are and as narrower the
proportional band is. Therefore, when in a process there are long delays, a narrow proportional
band cannot be adopted and the possibilities of application of proportional control are limited
very much.
Summarizing, the proportional control can be applied to a system only at the following
conditions: as longer the delays are, as wider the proportional band must be to avoid oscillations;
as wider the proportional band is, as longer the offset due to a load variation is.
2.3. Proportional-Plus-Integral Control
The integral control has the advantage of eliminating the offset of the variable due to lasting load
variations in the process. On the other hand, the effects of this control become effective with a
certain delay. When the variable undergoes a sudden and strong unbalance, the integral signal
generated by the instrument starts varying the opening of the control valve progressively. This
valve moves as faster as higher the unbalance is; nevertheless, in the first moments its
intervention could not be able to correct considerable unbalances. The integral control is fully
effective only after a certain time and goes on carrying out its corrective action until the offset is
completely annulled. But, if a sudden cause of unbalance must be compensated at its starting, the
integral action must be supported by a prompter control such as the proportional action.

12
In the proportional control the speed of displacement of the control valve is proportional to the
speed of variation of the error. Then, if the variable deviates quickly from the set-point, the
corrective action must be fast too. On the contrary, in the integral control, as it was already
explained, the speed of displacement of the valve is directly proportional to the value of error,
consequently its corrective action becomes effective only when there is an error of a certain
importance.
When both these control actions are added, the resulting control is characterized by the quickness
and stability of the proportional control, besides guaranteeing the coincidence of the control
point with the setpoint (integral control).

Example: - In a system of proportional-plus-integral control the proportional band and the reset
rate of the instrument are respectively equal to 20% and to 0.5 repetition/min.
In normal operational conditions the opening of control valve is 50%. Calculate the opening
degree of the control valve after two minutes from the time when the controlled variable
undergoes a permanent deviation from the set-point of 3% of the operational range of the
instrument.
Solution: the data of the problem are:
p0 = 0,5
b = 0,2
r = 0,5 repetition/min
e = 0,03.
If these values are substituted in the equation (5.10), the result is:

If the equation above is resolved for t = 2, the result is: p = 0,80

Therefore, it can be concluded that, when an error e = 0.03 lasts for 2 minutes, the control
instrument varies the opening degree of the control valve from 50% to 80%
The valve opening due to the only proportional control would be equal to 65%, as in case of
proportional control last term is omitted.

13
The integral action is added to the proportional control mainly for its property of eliminating the
offset. This automatically provokes the reset of the variable to the set-point.
The proportional-plus-integral controller is also called automatic-reset controller. As the
response of a system of proportional-plus-integral control is the result of the combined actions of
these two types of control, it depends on the pre-set values of the proportional band and of the
reset rate.
The figures 3.2.1 and 3.2.2 show the effects suffered by the response characteristic when each of
these two quantities is varied.

Fig. 2.3.1 - Effect of the reset rate on the response of a system of proportional-plus-integral
control

14
Fig. 2.3.2 - Effect of the amplitude of the proportional band on the response of a system of
proportional-plus-integral control

As it is shown, both increasing the reset rate and reducing the proportional band provoke an
increase of oscillations. The integral control repeats the effect of the proportional action for all
the time when the error lasts. If the proportional band is narrow, as soon as a certain error
appears, the position of the control valve undergoes a strong variation due to the proportional
control. In this case the controlled variable starts oscillating easily and the integral control opens
the valve more and more, increasing the oscillations. On the contrary, if the proportional band is
wide, the position of the control valve undergoes a slight variation due to the proportional control
and the controlled variable is slowly reset to the set-point. In this case the integral control
accelerates the return of the variable and this reset is sometimes so strong to provoke oscillations
that would not be generated with the proportional action.
In a proportional-plus-integral controller the position of the final control element depends not
only on the amplitude, but also on the duration of the deviation of the variable from the set-point;
therefore, the final corrective action may be excessive when the reset of the variable is slackened
by delays or dead times of the system.

15
2.4. Proportional-Plus-Integral-Plus-Derivative Control

Generally, the proportional and integral controls cannot be easily carried out in processes with
long delays and dead times. As it has been already explained in the previous paragraphs, when
there are some delays, wide proportional bands and slow reset rates must be adopted to avoid
excessive oscillations in the system. In these conditions, when some load variations occur in the
process, a wide proportional band provokes considerable deviations of the variable from the set-
point: these offsets cannot often be tolerated in the production process. However, these
deviations are eliminated only after a certain time due to the slow reset rate. The problems
concerning the control in these situations are resolved through a further mode of operation of the
automatic controllers, that is, through the derivative action.
The main characteristic of this type of action consists in the linear relation of proportionality
between the output signal and the speed of variation of the input error carried out through the
control instrument. The derivative action is often combined with the proportional and integral
controls. The result is the most complete control which can be carried out with the instruments
commonly used in industry.
In this type of control, an enduring load variation in the process makes the controlled variable
deviates from the set-point; at the end of the corrective action of the instrument the variable
becomes equal to the set-point again because the integral action makes that the control valve
offers a proper final opening degree corresponding to the new value of load. The trend of the
curve describing the opening of the control valve up to the final value depends not only on the
deviation of the variable, but also on the values of the proportional band, of the reset rate and of
the rate time set in the instrument. Therefore, the output signal resulting from the combination of
these three controls depends on the offset of the variable, on its duration and on the speed of
variation. The three corrective actions start operating simultaneously from the instant when the
variable starts differing from the set-point.
Generally, P-only controller does offer the advantage of faster response time, it produces
deviation from the set point. The key advantage of adding a I-control to your controller is that it
will eliminate the offset. The disadvantages are that it can destabilize the controller, and there is
an integrator windup, which increases the time it takes for the controller to make changes. The

16
primary benefit of D controllers is to resist change in the system, the most important of these
being oscillations.

17