Reactor Control
1. Cascade control of CSTR
Consider a CSTR. The reaction is exothermic and the heat generated is removed by a coolant
which flows in the jacket around the tank.
Objective: To keep the temperature of the reacting mixture constant at T.
Possible disturbances include the feed temperature (Ti) and the coolant temperature (TC).
The only variable that can be manipulated in this case is the coolant flow rate.
To keep the temperature of the reacting mixture constant, we have two options for
temperature monitoring and control. The feed control must either assess and control the
changes in temperature of the incoming feed or it must control the changes in temperature
of the coolant. The former directly affects the temperature of the reacting mixture while the
latter is less effective. i.e., Controlling Ti is more effective than controlling Tc for maintaining
the reaction temperature, T.
We have two different approaches: Simple Feedback control or the Cascade control.
In simple feedback (Fig 20.1a and Fig
20.2b), we attempt to control the
reaction temperature, T by managing
either Ti or Tc. As we saw earlier, the
latter is more effective and therefore, a
better option in this case. So, a simple
feedback loop would work by
monitoring the temperature inside the
reacting mixture, compare it with the set
point. When there is a deviation, the
flowrate of the coolant is adjusted such
that the temperature in the reaction
mixture is manipulated. Here, the
temperature of the coolant is not of
concern although it affects the heat
transfer.
In cascade control (Fig 20.1b and Fig
20.2c), the temperature of the reacting
mixture, T and the coolant, Tc is
measured. Thus, we have two loops
using different measured temperatures,
T and Tc sharing a common manipulated
variable, the flow rate of the coolant.
The two loops are:
1. The loop that measures T (controlled variable) is the dominant, or primary, or master
control loop and uses a set point supplied
2. The loop that measures Tc uses the output from the primary controller as its set
point and is called the secondary or slave loop.
Let the whole system be divided into two parts: Process I and II. Process-I (primary) has the
variable we want to control as the output while the Process-II controls a variable that we do
not necessarily wish to control but that which affects the primary process anyway.
In a cascade control, the secondary controller corrects the disturbances in the secondary
loop before it can affect the value of the primary output. This provides a faster response and
better control of the system.
Other than Cascade control, another type of multiple loop control system is the “Selective
Control Systems”, which is further divided as “Override control” and “Auctioneering control”.
An example for Auctioneering Control
2. Auctioneering Control in Reactor
An auctioneering control system considers the maximum value measured for a variable and
sends that to the controller, i.e., it possesses several measured outputs but only one
manipulated input.
Consider a tubular catalytic reactor where the highly exothermic reactions can lead to hot
spots at some point inside the reactor. The temperature along the length of the reactor is
measured by using thermocouples at multiple locations. Hot spots need to be controlled and
therefore, these control systems help limit the temperature below an upper limit. Therefore,
in this controller, the coolant flow rate is controlled by the highest value for temperature
measured.
3. Split-Range Control
Split range control configurations have only one measured variable and more than one
manipulated variable, i.e., the desired output is only one but that can be affected by many
variables. Such systems are rare in chemical processes but when they are present, they are
quite useful.
Split Range Control of a Chemical Reactor
Consider a reactor where a gas phase reaction is taking place. The reactor has two valves V1
and V2 regulating the flow of the feed and the reaction products respectively. The position
of the valves controls the pressure inside the reactor and therefore, cannot act
independently.
Now, in (a), the reactor pressure is
controlled/regulated by the position of the two
valves. Look at Table 20.1 for the pressure or
output signal and the corresponding position of
the valve.
Let’s say the controller’s output signal correspond to a value of 6psig. Look at the graph
below. For 6 psig, the values V2 is completely open and V1 is partially open. (Note that the
trend for V2 is represented in dash lines and V1 in solid lines). If the reactor pressure
suddenly increases, then to manage the pressure, the output signal also increases. So let the
signal increase to 9 psig. As per the graph, both valves are now open. Further increase in
output signal should lead to partial closing of V1 and complete opening of V2.
