CHE 452, 3 credits Chemical Process Control Spring 2010
Instructor: Dr. Srinivas Palanki Oce: Room 242 Phone: 460-6160 Email: spalanki@usouthal.edu
Text: Lecture: No textbook is necessary. The instructor will provide all the lecture material on the course
Introduction to Process Control
Srinivas Palanki
University of South Alabama
Srinivas Palanki (USA)
Introduction to Process Control
1/5
Introduction
What does a control system do?
A control system maintains desired conditions in a physical system by adjus
Home Assignment
Note: All the problems below have MULTIPLE steady-state solutions. Make sure you get all of them. For example, Problem 1 has steady states (+0.781, +1.603) and (-0.781, +1.603). It is a good idea to plug in your solutions into the original
Home Assignment
Problem 1: Consider the following dynamic system that represents an isothermal, unsteady state CSTR: F F dCA 2 = CAin CA k1 CA dt V V dCB F 2 = CB + k1 CA k2 CB dt V dCC F = CC + k2 CB dt V F = 1, k1 = 2, k2 = 1.5, simulate the above syste
Home Assignment
Problem 1: Consider the following matrices: 1. A= 2. B= 3. 1 2 3 C = 4 5 7 6 3 2 4. 200 D= 5 2 0 842 5. E= 6. 2 F = 5 8 Compute the determinant for each matrix. Compute A.B , C.D, B.A, D.C and A.E . Compute C.F . Is it possible to compute
Home Assignment
Problem 1: Consider the following dynamic system in deviation form: dX = AX + BU dt If the input U = 0, compute the unforced response for the following A matrices assuming that all elements of X (0) are equal to 1. 1. A= 2. A= 3. 1 2 3 A =
Home Assignment
Problem 1: An autocatalytic reaction A R takes place in a constant volume isothermal reactor. Component mass balances give the following reactor model: F dCA = (CAin CA ) kCA CR dt V dCR F = (CRin CR ) + kCA CR dt V In the above model CAin
Home Assignment
Problem 1: Consider the following linearized model in deviation form: dX = AX + BU dt where X= B= X1 X2 1 2 (2) (3) (1)
and U is a scalar. Suppose the input undergoes the following step change: U= 0 t<0 3 t0 (4)
Compute how the X vector ch
Interconnected Systems
Problem 1: Consider the following interconnected system:
Y sp + Y
e
System 1
U
System 2
Y
Suppose System 1 is modeled as: d dt U and System 2 is modeled as: d dt Y X3 X4 = A2 = C2 X3 X4 X3 X4 + B2 U (2) + D2 U X1 X2 = A1 = C1 X1 X2
tq (q 7 X ( $ 7 o P ( 7 3 7 A h1 C 3 $ ( h 0 q 0 A A& |)WY7XX6Y$X6qY()'&f2HlSq7g"58og"n8$)YW`(2C04(itS")"8`AYC2CgW`7XG8C8C)n eS2v2cfw_ @t@t x 0 ( Vw)r#o8$A 2Cghf h co23f0$y)Y7C(hmVx ( !r ( Spv2vStu@t@t@t hgR)4"Q7Q7R4P4P`(VV8$A A h)g6g|`0r r 2r#o2&2X8$)
ECH 452: Chemical Process Control SIMULATION PROJECT Propylene glycol is produced by the hydrolysis of propylene oxide in a CSTR. The reaction is normally conducted using excess water.
H2SO4
CH2-CH-CH3 + H2O O
CH2-CH-CH3 OH OH
A
+
B
C
The objective of thi
Tuning Guidelines
Srinivas Palanki
University of South Alabama
Srinivas Palanki (USA)
Tuning Guidelines
1 / 11
Tuning Commercial Controllers
Two Approaches to Dynamic Analysis
We have considered two dierent approaches to dynamic analysis: Analytical Appro
Introduction to Process Control
Srinivas Palanki
University of South Alabama
Srinivas Palanki (USA)
Introduction to Process Control
1/5
Introduction
What does a control system do?
A control system maintains desired conditions in a physical system by adjus
Process Models
Srinivas Palanki
University of South Alabama
Srinivas Palanki (USA)
Process Models
1/8
Introduction to Process Models
What is a Process Model?
A process model is a set of equations, including the necessary input data to solve the equations,
Examples of Dynamic Models
Srinivas Palanki
University of South Alabama
Srinivas Palanki (USA)
Examples of Dynamic Models
1 / 19
Examples
Liquid Storage Process
A typical liquid storage process is shown in the gure below.
qin
q
Filling Process
How does th
General Form of Dynamic Models
Srinivas Palanki
University of South Alabama
Srinivas Palanki (USA)
General Form of Dynamic Models
1 / 20
General Form of Dynamic Process Models
Mathematical Representation
A general representation of the dynamic process mod
Solution of Linear Dierential Equations
Srinivas Palanki
University of South Alabama
Srinivas Palanki (USA)
Solution of Linear Dierential Equations
1 / 15
Exponential of a Matrix
Exponential of a Matrix
The quantity e at where a and t are scalar is an inn
Analysis of Linear System Dynamics
Srinivas Palanki
University of South Alabama
Srinivas Palanki (USA)
Analysis of Linear System Dynamics
1 / 16
Dynamics of a Linear System
Recap
A linear dynamical system in deviation form is represented as dX = AX + BU d
Forced Dynamics of a Linear System
Srinivas Palanki
University of South Alabama
Srinivas Palanki (USA)
Forced Dynamics of a Linear System
1 / 12
Dynamics of a Linear System
A linear dynamical system in deviation form is represented as dX = AX + BU dt X (0
Example Calculation of Forced Dynamics
Srinivas Palanki
University of South Alabama
Srinivas Palanki (USA)
Example Calculation of Forced Dynamics
1/8
Forced Dynamics
Illustrative Example
Consider the following system in deviation form: d dt X1 X2 = 1 0 1
Process Outputs
Srinivas Palanki
University of South Alabama
Srinivas Palanki (USA)
Process Outputs
1/7
Process Outputs
Denition
Very often, we are not interested in calculating the response for the entire state vector, x (t ); we may care about only a fe
Interconnected Systems
Srinivas Palanki
University of South Alabama
Srinivas Palanki (USA)
Interconnected Systems
1 / 12
Interconnected Systems
Larger Systems by Connecting Sub-Systems
So far in this course, we have done the following: Developed dynamic m
Dynamics of Commercial Controllers
Srinivas Palanki
University of South Alabama
Srinivas Palanki (USA)
Dynamics of Commercial Controllers
1 / 18
Commercial Controllers
Controller
A controller is an electronic device that helps to:
1
regulate a process at