Process Simulation
Process Simulation
Reference book: Process Modeling, Simulation, and Control for Chemical Engineers (2n ed.) by
Luyben, W. L.
Note: The text in this pdf should only be used as helping outline to understand the chapter 3 of the
text book. This pdf should not be used as class notes.
3.1 Introduction
In various chemical engineering systems we deal with physical and chemical phenomena.
Theoretically we can appreciate theses phenomena but this is of no use at practical grounds. For
controlling the system we need to convert them into mathematical equations. This is called
modeling of the system.
Modeling a system: For modeling a system there are following basic steps:
After writing all the above equations for a system, check for the degree of freedom is performed.
Check for degree of freedom: The number of equations should be equal to the number of unknown
variables. If the number of unknown variables are more than the equations, then we try to find out
the extra equations that are applicable to the system. If number of unknown variables are equal to
the number of equations then the solution of the system is possible. It means that we can calculate
the numerical values of the unknown variables.
CSTR: This reactor has a feed input and product output. This has a stirrer for mixing the contents.
For example some of the variables: Volumetric flow rates in and out of the three CSTRs (F 0, F1, F2,
F3), Volumes of the three CSTRs (V1, V2, V3), Molar concentration of reactant A in the feed (CA0),
Concentrations of A in the three CSTRs (CA1, CA2, CA3) etc.
2. State all the assumptions that have been applied to make the system easy to handle
For example: the temperatures in the three CSTRs are assumed constant etc.
3. Formulate the model: (Various relationships and laws are applied to get equations)
For example in this system the volumes of the reactors (V1, V2, V3), temperatures of the reactors (T1,
T2, T3) etc. are already known to us. Some variables like F and C A0 are unknown but they can be
specified (called forcing functions). In other terms the numerical values of F and C A0 can be
assumed by us. So they are also known.
Ultimately we are left with three unknown variables i.e. CA1, CA2, and CA3
This way there are three equations and three unknowns or dependent variables. If number of
equations are equal to the unknown variables, then only we can solve the equations.
Same as the last section 3.2. Only difference is that the volume of the reactors are not constant.
They can vary with time. keeping this in mind the mass balances are performed.
It gives six equations and nine unknowns. So there should be three more equations. It may be
possible that during the analysis of the system, some relationships are left due to our ignorance. We
need to find these missing relationships. After more analysis we find three more relationships as
follows:
The flow rate out of the tank is related to the quantity of liquid in the tank. If there is more liquid in
the tank then there will be more flow rate. This way we find three more relationships.
So now we have nine equations and nine unknowns variables and the system can be solved.
Here temperature of the CSTR can change with time. So energy balance is also to be taken into
account
There is a cooling jacket surrounding the reactor. Cooling water is flowing into the jacked.
Now in this system there can be assumed many types of flow patterns for cooling water jacket as
follows:
The temperature everywhere in the jacket is the same. Apply the heat transfer equation for this.
In this model we assume that the cooling water is not perfectly mixed. The temperature of the
entering water and the exiting water are different. Here average of these two temperatures are taken
to define the jacket temperature.
C. Lumped jacket model: Under this assumption the jacket volume is divided into many perfectly
mixed lumps. For our system let’s assume there are four lumps of equal volume. So there will be
four energy balance equations for the jacket in this case.
D. Significant metal wall capacitance: The mass of the metal walls is also taken into
consideration. Metal wall is also included in the heat transfer relationship.
If we select cooling water jacket of type A (see above), then our formulation of model will be this
way:
We have five equations and five unknowns: V, F, CA, T and TJ. So solution is possible.
The same way formulation of model can be performed for other cases of cooling water jacket (B, C,
D)
This is the simplest model. Neglect the dynamics of both vapor and liquid phase. (Dynamics means
the relationship between liquid volume and flow rate etc.)
Formulation:
1. Only the total continuity equation for the liquid phase is need.
2. Controller relationship between pressure and heat input
3. Controller relationship between liquid level and feed flow rate
4. Energy balance
5. State equation of vapour (PV = RT)
6. Vapor pressure equation. (Relationship between temperature and vapour pressure)
Check degree of freedom: This way it gives six equations and there are six unknowns. So solution is
possible.
If there is a large volume of vapor then the dynamics of the vapor phase is also taken into
consideration.
Formulation:
1. Relationship between liquid boiling rate, vapour pressure and pressure in the vapour phase.
For vapor:
For liquid:
Check degree of freedom: There are nine equations and ten variables. So we need one more
equation. In we analyze the system more we can find this tenth equation. This one more extra
equation comes when we define a relationship for the total volume of the tank. This way we have
total ten equations and ten variables. So the solution is possible.
2. Relationship between liquid volume (holdup) in a tray and the outlet flow rate of liquid.
3. Condenser and reflux drum equations:
4. Top tray:
6. nth tray:
7. Feed tray:
8. First tray:
So we need two more equations. If we analyze the system more then we can find that we have
missed the following two relationships:
1. Reboiler: the controller relationships between vapour out rate from reboiler and composition (xB)
2. Condenser: the controller relationships between between reflux flow rate (R) and composition
(xD)
This way we have number of equations = 4N T + 9, which is equal to the number of unknown
variables. This way this system is completely solvable.