Chapter Five

Advanced Process control

Advanced Process control

➢Cascade control
➢Over ride control
➢Split range control
➢Selective control
5.1-Cascade control

• As the processes requirements tighten, in processes with slow

dynamics and processes with too many, or frequently occurring,
upsets, the control performance provided by feedback control often
becomes unacceptable. It is necessary to use other strategies to provide
the required performance. The strategies presented complement
feedback control; they do not replace it. Remember that it is always
necessary to provide some feedback from the controlled variable
Cascade control

• Cascade control is a strategy that improves, in some applications

significantly, the performance provided by feedback control. This strategy
has long been well known.
• Computers provide a simpler, safer, and less costly implementation of
cascade control than is obtained by the use of analog instrumentation.
• Feed Forward and Cascade control Both require additional
instrumentation and engineering time in return for a controller better able
to reject disturbances
• Cascade control is an Architectures for improved disturbance rejection.
 Two level Cascade control architecture

• Where:
FCE-final
control
element
PV1-process
variable 1
PV2-process
variable 2
Cascade control

• A cascade control is comprised of two ordinary PID controllers The

inner secondary loop has a traditional feedback structure, and it is
nested inside the outer primary loop.
• Cascade architectures seek to improve the disturbance rejection
performance of PV1.
Cascade control requires an early warning process variable ---
PV2
How to select PV2—early warning process variable

• Success in a cascade design depends on the measurement and control

of an "early warning" process variable PV2.
• Characteristics for selecting early warning PV2 include:
➢it must be measurable with a sensor.
➢the same final control element(actuator) (e.g., valve) used to
manipulate PV1 also manipulates PV2.
➢the same disturbances that are of concern for PV1 also disrupt PV2.
➢PV2 responds before PV1 to disturbances of concern and to final
control element( actuator) manipulations.
A two level cascade control design requires:

➢two sensors
➢two controllers
➢One final control element (FCE)
➢The output of the outer primary controller, rather than going to a
valve, becomes the set point of the inner secondary controller Because
of this nested architecture:
➢Cascade requires that the settling time of the inner secondary inner loop
is significantly faster than that of the outer primary outer loop.
 Example: jacketed heat exchanger-Feedback control

• Consider the following

feedback control of Fig.3
jacketed heat exchanger
• Variation in steam
temperature is assumed
as disturbance.
• It is obvious that
variation of steam
temperature affects both
the Jacket and process
liquid temperature.
Feedback control block diagram of jacketed heat exchanger

Fig.4
jacketed heat exchanger-Feedback control

• The system consists of a tank through which process liquid, entering

at temperature 𝑻𝒊 , is heated to 𝑻𝒐 by the flow of steam through a
jacket surrounding the tank.
• The temperature of the water in the tank is measured and transmitted
to the controller, which in turn adjusts the flow of mass of steam
through the jacket.
• This control system is satisfactory for controlling the tank
temperature.
jacketed heat exchanger-Feedback control

• however, if the temperature of the Steam supply drop, the tank

temperature can undergo a large prolonged deviation from the set
point before control is again established.
• The reason is that the controller does not take corrective action until
the effect of the drop in Steam supply temperature results drop in
process liquid outlet temperature.
Jacketed heat exchanger– Cascade control

• To prevent the sluggish response of tank temperature to a disturbance

in Steam supply temperature, the control system shown in Fig.4 is
proposed.
• In this system, which includes two controllers and two measuring
elements, the output of the primary controller is used to adjust the set
point of a secondary controller, which is used to control the jacket
temperature.
Jacketed heat exchanger– Cascade control

• Under these conditions, the primary controller indirectly adjusts the

jacket temperature. If the Steam temperature drop, the secondary
control loop will act quickly to maintain the jacket temperature close to
the value determined by the set point that is adjusted by the primary
controller. This system shown in Fig. 5 is called a cascade control system.
Jacketed heat exchanger– Cascade control

• The primary controller

is also referred to as
the master controller,
and the secondary
controller is referred to
as the slave controller.

Fig.4
Jacketed heat exchanger– Cascade control

Fig.5
Cascade control

• The following must be stressed:

In designing cascade control strategies, the most important consideration
is that the inner loop must be faster than the outer loop and the faster the
better. This requirement makes sense, and it is extended to any number
of cascade loops.
• Note that the inner most controller is the one that sends its output to
the valve. The outputs of all other controllers are used as set points to
other controllers; the final control element of these controllers is the
set point of another controller.
Implementation and Tuning of Cascade controller

• Two important questions still remain:

• how to put the cascade strategy into full automatic operation and how
to tune the controllers.
• The answer to both questions is the same:
➢That is, the inner controller is the first to be tuned and put into auto
state while the other loops are in manual.
Tuning of Cascade controller

• For the process shown in Fig.5, 𝑮𝒄𝟐 is tuned first and set in cascade
while 𝑮𝒄𝟏 is in manual.
• The control performance of 𝑮𝒄𝟐 is then checked before proceeding to
𝑮𝒄𝟏 . This checking can usually be done very simply by varying the set
point to 𝑮𝒄𝟐 .
• Remember, it is desired to make 𝑮𝒄𝟐 as fast as possible, even if it
oscillates a bit, to minimize the effect of the upsets. Once this is done,
𝑮𝒄𝟏 can be tuned and set in automatic.
• However, before 𝑮𝒄𝟏 is set in automatic, 𝑮𝒄𝟐 must be set to the cascade
state, which means that it will accept the output from 𝑮𝒄𝟏 as its set
point.
Example 2: Preheater/reactor process

Fig.6
Preheater/reactor process

• Consider the furnace/preheater and reactor process shown in Fig. 6 In

this process, reaction A B occurs in the reactor. Reactant A is
usually available at a low temperature, so it must be heated somewhat
before being fed to the reactor. The reaction is exothermic, and to
remove the heat of reaction, a cooling jacket surrounds the reactor.
• The important controlled variable is the temperature in the reactor,
𝑻𝑹 .The original control strategy called for controlling this
temperature by manipulating the flow of coolant to the jacket.
Preheater/reactor process

• The inlet reactant temperature to the reactor was controlled by

manipulating the fuel valve.
• It was noticed during the start-up of this process that the cooling
jacket could not provide the cooling capacity required; the cooling
valve was open almost all the time. Thus it was decided to open the
cooling valve completely and control the reactor temperature by
manipulating the fuel to the preheater, as shown in the fig. 4. This
strategy worked well enough, providing automatic control during
start-up.
Preheater/reactor process

• Once the process was “lined-out,” the process engineer noticed that so
often the reactor temperature would move from set point enough to make
off-spec product. After checking the feedback controller tuning to be sure
that the performance obtained was the best possible, the engineer started
to look for possible process disturbances.
• Several upsets were found around the reactor itself-cooling fluid
temperature and fluid flow variations, etc. -and others around the
furnace-variations in inlet temperature of reactant A, in heating value of
fuel, in inlet temperature of combustion air, etc. Furthermore, the
engineer noticed that every once in a while the inlet reactant temperature
to the heater would vary by as much as 25°C certainly a major upset.
Numerical Example– Preheater/reactor process feedback
control block diagram

Fig.7
Numerical Example– Preheater/reactor process feedback
control block diagram

• The characteristic equation for the simple feedback control system,

Fig. 7, is:
𝟏. 𝟓 𝑲𝑪
𝟏+ 𝟐 𝟐
=𝟎
𝟎. 𝟐𝒔 + 𝟏 𝟑𝒔 + 𝟒𝒔 + 𝟏 𝟒𝒔 + 𝟓𝒔 + 𝟏
• Applying the direct substitution method or the frequency response
technique to the feedback control system characteristic equation, we
can calculate the ultimate gain and ultimate frequency.
𝑲𝒄𝒖 = 𝟒. 𝟑𝟑 𝒂𝒏𝒅 𝝎𝒖 = 𝟎. 𝟓𝟎𝟕
𝟐𝚷 𝟐𝚷
𝑻𝒖 = = = 𝟏𝟐. 𝟒
𝝎𝒖 𝟎. 𝟓𝟎𝟕
The Quarter Decay Ratio Response by Ultimate Gain is a
tuning method introduced by Ziegler and Nichols (1942) for
adjusting PID controllers. It involves two steps:
• Determining the loop’s dynamic characteristics (ultimate gain and ultimate
period).
• Estimating controller tuning parameters to achieve the desired response.
• In this method the dynamic characteristics of the process are represented by
the ultimate gain of a proportional controller and the ultimate period of
oscillation of the loop.
• It can be determined by the direct substitution method if the transfer
functions of all the components of the loop are known quantitatively.
• But because this is not usually the case, we must often experimentally
determine the ultimate gain and period from the actual process
Procedure:
• Set the controller to proportional mode only (disable integral and
derivative actions).
• Increase proportional gain until sustained oscillations occur, recording
Kcu.
• Measure and record the oscillation period (Tu) from the response.
• These values help in setting optimal controller parameters for improved
system performance.
Numerical Example– Preheater/reactor process feedback
control block diagram

• Using Z-N
𝑲𝒄𝒖 𝟒. 𝟑𝟑
𝑲𝒄 = = = 2.547
𝟏. 𝟕 𝟏. 𝟕
𝑻𝒖 𝟏𝟐. 𝟒
𝝉𝑰 = = = 𝟔. 𝟐
𝟐 𝟐
𝑻𝒖 𝟏𝟐. 𝟒
𝝉𝑫 = = = 𝟏. 𝟓𝟓
𝟖 𝟖
Numerical Example– Preheater/reactor process Feedback
control

• It is fairly simple to realize that the effect of an upset in the furnace

results first in a change in the reactant exit temperature from the
furnace, 𝑻𝑯 and that this then affects the reactor temperature, 𝑻𝑹 .
• Once the controller senses the error in 𝑻𝑹 , it manipulates the signal to
the fuel valve. However, with so many lags in the process, furnace plus
reactor, it will take a considerable amount of time to bring the reactor
temperature back to set point.
• Because of these lags, the simple feedback control shown in the Fig. 6
will result in cycling and in general sluggish control.
Numerical Example– Preheater/reactor process Cascade
control

• A superior control strategy can be designed by making use of the fact

that the upsets in the furnace first affect 𝑻𝑯 . Thus it is logical to start
manipulating the fuel valve as soon as a variation in 𝑻𝑯 is sensed,
before 𝑻𝑹 starts to change.
• The idea is not to wait for an error in 𝑻𝑹 to start changing the
manipulated variable. This control strategy uses an intermediate
variable, 𝑻𝑯 in this case, to reduce the effect of some dynamics in the
process.
• This is the idea behind cascade control, and it is shown in Fig. 8.
Fig. 8
Numerical Example– Preheater/reactor process Cascade
control

• This strategy consists of two sensors, two transmitters, two controllers,

and one control valve. One sensor measures the intermediate, or
secondary, variable, 𝑻𝑯 in this case, and the other sensor measures the
primary controlled variable, 𝑻𝑹 .
• Thus this strategy results in two control loops:
• One loop controlling 𝑻𝑯 , and the other loop controlling 𝑻𝑹 .
• To repeat, the furnace exit temperature is used only as an intermediate
variable to improve the control of the reactor temperature, which is
the important control objective (controlled variable).
Numerical Example– Preheater/reactor process Cascade
control: The strategy works as follows

➢Controller TC-101 looks at the reactor temperature and decides how

to manipulate the furnace outlet temperature to satisfy its set point.
➢This decision is passed on to TC-102 in the form of a set point. TC102
in turn manipulates the signal to the fuel valve to maintain 𝑻𝑯 at the
set point required by TC-101.
➢If one of the upsets mentioned earlier enters the furnace, 𝑻𝑯 deviates
from set point and TC-102 takes corrective action right away, before
𝑻𝑹 changes. Thus the dynamic elements of the process have been
separated to compensate for upsets in the furnace before they affect
the primary controlled variable.
Numerical Example– Preheater/reactor process Cascade
control

• The controller that keeps the primary variable at set point is referred
to as the master controller, outer controller, or primary controller.
• The controller used to maintain the secondary variable at the set point
required by the master controller is usually referred to as the slave
controller, inner controller, or secondary controller.
 Numerical Example– Preheater/reactor process Cascade
control block diagrme

𝑻𝑹𝒔𝒑 𝑻𝑯𝒔𝒑 𝑻𝑯 𝑻𝑹

Fig. 9
Numerical Example– Preheater/reactor process Cascade
control
• characteristic equation for the cascade control strategy, Fig. 9, is:
𝟏. 𝟓𝑮𝑪𝟏 𝑮𝑪𝟐
𝟎. 𝟐𝒔 + 𝟏 𝟑𝒔𝟐 + 𝟒𝒔 + 𝟏 𝟒𝒔𝟐 + 𝟓𝒔 + 𝟏
𝟏+ =𝟎
𝟏. 𝟓𝑮𝑪𝟐
𝟏+
𝟎. 𝟐𝒔 + 𝟏 𝟑𝒔𝟐 + 𝟒𝒔 + 𝟏
• To determine the ultimate gain and frequency of the primary controller
of the cascade strategy, the tuning of the secondary controller must first
be obtained by determining the ultimate gain of the secondary loop in
Fig. 9. The characteristic equation for this secondary process is:
𝟏. 𝟓𝑲𝑪𝟐
𝟏+ 𝟐
=𝟎
𝟎. 𝟐𝒔 + 𝟏 𝟑𝒔 + 𝟒𝒔 + 𝟏
Numerical Example– Preheater/reactor process Cascade
control

𝟏. 𝟓𝑲𝑪𝟐
𝟏+ 𝟐
=𝟎
𝟎. 𝟐𝒔 + 𝟏 𝟑𝒔 + 𝟒𝒔 + 𝟏

• Applying the direct substitution method or the frequency response

technique (Chapter 3) to the feedback control system characteristic
equation, we can calculate the ultimate gain and ultimate frequency.
𝑲𝒄𝒖𝟐 = 𝟏𝟕. 𝟎𝟔
• Assuming a proportional-only controller and using the Ziegler-Nichols
suggestion, we find that the tuning for the secondary controller
becomes.
𝑲𝒄𝟐 = 𝟎. 𝟓 𝑲𝒄𝒖𝟐 = 𝟎. 𝟓 ∗ 𝟏𝟕. 𝟎𝟔 = 𝟖. 𝟓𝟑
Numerical Example– Preheater/reactor process Cascade
control
• When we use this tuning value for the secondary controller, the
characteristic equation for the cascade strategy yields, for the primary
controller:
𝟏. 𝟓 𝑲𝑪𝟏 𝑲𝑪𝟐
𝟎. 𝟐𝒔 + 𝟏 𝟑𝒔𝟐 + 𝟒𝒔 + 𝟏 𝟒𝒔𝟐 + 𝟓𝒔 + 𝟏
𝟏+ =𝟎
𝟏. 𝟓𝑲𝑪𝟐
𝟎. 𝟐𝒔 + 𝟏 𝟑𝒔𝟐 + 𝟒𝒔 + 𝟏
𝟏. 𝟓 𝑲𝑪𝟏 ∗ 𝟖. 𝟓𝟑
𝟎. 𝟐𝒔 + 𝟏 𝟑𝒔𝟐 + 𝟒𝒔 + 𝟏 𝟒𝒔𝟐 + 𝟓𝒔 + 𝟏
𝟏+ =𝟎
𝟏. 𝟓 ∗ 𝟖. 𝟓𝟑
𝟎. 𝟐𝒔 + 𝟏 𝟑𝒔𝟐 + 𝟒𝒔 + 𝟏

𝑲𝒄𝒖𝟏 = 𝟕. 𝟐 𝒂𝒏𝒅 𝝎𝒖𝟏 = 𝟏. 𝟓𝟒

Numerical Example– Preheater/reactor process Cascade
control

𝟐𝚷 𝟐𝚷
𝑻𝒖 = = = 𝟒. 𝟎𝟕
𝝎𝒖 𝟏. 𝟓𝟒
• Using Z-N
𝑲𝒄𝒖 𝟕. 𝟐
𝑲𝒄 = = = 2.547
𝟏. 𝟕 𝟏. 𝟕
𝑻𝒖 𝟒. 𝟎𝟕
𝝉𝑰 = = = 𝟐. 𝟎𝟑𝟓
𝟐 𝟐
𝑻𝒖 𝟒. 𝟎𝟕
𝝉𝑫 = = = 𝟎. 𝟓𝟎𝟖𝟕𝟓
𝟖 𝟖
Numerical Example– Preheater/reactor process Cascade
control Vs Feedback control
5.2. Selective control

• Involves selectin of one signal between different signals.

•Low selector:

•High selector:
 5.2. Selective control

• If there are more controlled variables than number of manipulated

variables or actuators available, signal selectors are used to select the
highest, lowest or median of controlled variables or controller outputs.
➢They are applied in three basic areas
1. Auto-actioneering of multiple inputs
2. Redundancy to minimize system failures
3. Override control to protect personnel and equipment
5.2.1. Auto-actioneering

• Auto-actioneering is the selection of the highest or lowest of an

independent input signals as representing the limiting case
• As an example, control of the highest of several temperatures
measured at different locations in a chemical reactor.
• the “ hot spot” – whose location is subject to change with fouling,
production rate, catalyst activity, etc.
Auto-actioneering
5.2.2. Redundancy to minimize system failures

• Redundancy applies multiple measurements to the same variable. To

protect against the possibility of an instrument failure creating an
unsafe condition.
• As an example, when oxidizing hydrocarbons to form ethylene oxide
and similar products, the oxygen content of the reaction mixture must
be controlled below the explosive limit. It is customary to employ
redundant oxygen analyzers in this service, selecting the higher
indicated oxygen content to be controlled. If one analyzer then fail, the
other will be selected for control and the plant will continue to operate
safely. So that, the problem may be corrected before a second failure
occurs.
5.2.3 Override Control

• Override control, or constraint control, is a powerful yet simple control

strategy generally used as a protective strategy to maintain process
variables within limits that must be enforced to ensure the safety of
personnel, equipment, and product quality and an optimization
strategy that permits smooth transition between controllers to obtain
maximum benefit.
• To explain the concept of override control consider the following
saturated solution level and flow control.
• Where the level controller override the flow controller.
Example: Level controller overriding Flow controller
Example: Level controller overriding Flow controller

• A hot saturated liquid enters a tank and from there pumped under
flow control. Under normal operation, the level in the tank is at height
𝒉𝟏 . If under any circumstances the liquid level drops to height 𝒉𝟐 , the
liquid will not have enough net positive suction head (NPSH), and
cavitation at the pump will result. It is therefore necessary to design a
control scheme that avoids this condition.
Example: Level controller overriding Flow controller
 Example: Level controller overriding Flow controller

• The level in the tank is now measured and controlled. It is important to note
the action of the controllers and final control element.
• When the saturated liquid flow rate increases beyond set point the flow
controller should decrease pumper speed and if flow rate decrease beyond
set point the flow controller should increase pumper speed. Therefore flow
controller is reverse acting.
• When the saturated liquid level in the tank increases beyond set point the
level controller should increase pumper speed to decrease the over level
condition and if level in the tank decrease beyond set point the level
controller should decrease pumper speed to increase the under level
condition. Therefore level controller is direct acting.
Example: Level controller overriding Flow controller

• Let us now suppose that the flow of hot saturated liquid (process
inlet) slows down and the level in the tank starts to drop. As soon as
the level drops below the set point, level controller will reduce its
output.
• When the level controller’s output drops below the output of the flow
controller, the low selector selects the output of the level controller to
manipulate the pump. It can be said that the level controller
“overrides” the flow controller.
Example: Level controller overriding Flow controller

• When the flow of hot saturated liquid (process inlet) returns to its
normal value, and the level increases above the set point, the level
controller increases its output to speed up the pump. Once the output
of the level controller increases above the output from the flow
controller, the low selector selects the flow controller, and the
operation returns to its normal condition.
• An important consideration in designing an override control system is
that of reset windup protection on any controller that has integration.
The output of the controller not selected must stop at 100%, not at a
higher value, or at 0%, not at a lower value.
Example: Level controller overriding Flow controller

• The reset feedback signal to the controller(s) comes from a low

selector, sometimes we refer to this signal as external reset feedback.
The output from the low selector is the one that goes to the pump; it is
used as the reset feedback signal to the controllers.
Example: Level controller overriding Flow controller
Numerical Example: Level controller overriding Flow
controller

• Assume that at steady state the flow controller outputs a 75% signal to
maintain its set point and the level controller is below 75%, therefore
the level controller is in action.
• Assume the level in the tank is above its set point.
• In this case the level controller increases its output, to speed up the
pump, and the level controller output become grater than 75%.
• Thus the low selector selects the 75% signal from the flow controller;
this is the normal operating condition.
Numerical Example: Level controller overriding Flow
controller

• The output signal from the selector is the RFB signal to the
controllers, and the corresponding selector output signal will also be
75%. At this steady state, the error in the flow controller is zero, and
the proportional calculation of this controller is also zero.
• Because the level in the tank is above the set point, remember level
controller is direct acting, so the output of level controller
proportional calculation will have a certain output, depending on the
error(positive) and controller gain (positive), say 10%. The low
selector signal is 75%, so the output from the level controller to the
low selector is 85% (75% + 10%).
Numerical Example: Level controller overriding Flow
controller

• Now suppose that the input flow to the tank decreases and the level in
the tank starts to drop. As this happens, the proportional calculation
in the level controller also starts to decrease from 10% down, and the
controller output from 85% starts to decrease. As the level in the tank
drops below the set point, the error in the controller, and the
proportional calculation, become negative due to negative error, which
results in an output less than 75%-say 74%.
• At this moment the selector selects this signal, and the level controller
overrides the flow controller and sends it to the pump to slow it down;
this new value is also the RFB signal to the controllers.
Split rang control

• A split-range strategy is often used in situations where two or more

final control element (actuator) are used.
• An example application is a batch reactor. Batch reactors must often
be heated from ambient temperature to a desired operating
temperature. Particularly if the reaction is exothermic, then cooling
will be needed to maintain the desired reactor temperature.
Example: Split range control of batch reactor

• Batch reactor temperature

control. The jacket
temperature controller has
a split range output, where
the cold glycol valve is
open during "cooling
mode“ and the hot glycol
valve is open during
"heating mode."
Example: Split range control of batch reactor
Illustration of the split-range controller action
Example: Split range control of batch reactor

• In the split-range control strategy shown in Figure, if the jacket

temperature controller output is between 0 and 50%, the cold glycol
valve is open.
• If the jacket temperature controller output is between 50 and 100%,
the hot glycol valve is open.
• For safety reasons we wish the cold glycol valve to fail-open and the
hot glycol valve to fail-closed. The diagram in Figure more clearly
shows the relationship between valve positions and the jacket
temperature controller output
Example: Implementation of Split range control for batch
reactor

•Let 𝑢 represent the output of the jacket temperature

controller (0–100%).
•let 𝑣𝑐 represent the cold valve positions.
•and 𝑣ℎ represent hot glycol valve positions.

The split-range algorithm can be stated as:

𝟎 ≤ 𝒖 ≤ 𝟓𝟎% ∶ 𝒗𝒄 = 𝟏𝟎𝟎 − 𝟐𝒖, 𝒗𝒉 = 𝟎
𝟓𝟎 < 𝒖 ≤ 𝟏𝟎𝟎% ∶ 𝒗𝒄 = 𝟎, 𝒗𝒉 = 𝟐(𝒖 − 𝟓𝟎)