Process Optimization

Course code: 2CHDE15

 Course Objectives
• The course will provide the students with the ability to
formulate, solve and interpretate the meaningful
optimization problem in engineering and Science.

• Emphasis will be given to formulate the mathematical

models for the use of commonly available techniques

• Software packages will be used to solve the realistic

engineering problems
 This course is designed to give the students :

• A view of Optimization as a tool for design and control

• An opportunity to learn chemical engineering process

and solve it using computer based tools such as Excel,
MATLAB and Using optimization Techniques.

• A functional introduction to optimization techniques such

as linear programming and Non linear programming
 Course Outcomes
At the end of the course the students will be able to
• Understand the difference between different optimization techniques
• Formulate an optimization problems
• Define the objective function
• Understand the fundamental of linear programing and non linear
programing
• Solve linear programing graphically
• Solve linear programing using simplex methods and using excel
• Using Matlab to solve basic chemical Engineering problems
• Understand how to formulate non linear programing problems
• Solve non linear programing by excel
Example 1 We need to enclose a rectangular field with a fence. We have 500
feet of fencing material and a building is on one side of the field and so won’t
need any fencing. Determine the dimensions of the field that will enclose the
largest area.

Solution: Maximize : A=xy

Constraint : 500=x+2y

Substituting this into the area function gives a function of y.

A(y)=(500−2y)y=500y−2y2
A(y)=(500−2y)y=500y−2y2

Now we want to find the largest value this will have on the interval [0,250][0,250]. The limits
in this interval corresponds to taking y=0y=0 (i.e. no sides to the fence)
and y=250y=250 (i.e. only two sides and no width, also if there are two sides each must be
250 ft to use the whole 500ft).
A′(y)=500−4y

Setting this equal to zero and solving gives a lone critical point of y=125

A(125)=31250 ft2

x=500−2(125)=250
Method 1 : Finding Absolute Extrema.

Method 2 : First Derivative Test.

Example 2 We want to construct a box whose base length is 3 times the base width. The material used to build
the top and bottom cost $10/ft2 and the material used to build the sides cost $6/ft2. If the box must have a
volume of 50ft3 determine the dimensions that will minimize the cost to build the box.
Plugging this into the cost gives,
C(w)=60w2+48w(50/3w2)=60w2+800/w
Now, let’s get the first and second (we’ll be needing this later…) derivatives,
C′(w)=120w−800w-2
=(120w3−800)/w2
C′′(w)=120+1600w−3

Now we need the critical point(s) for the cost function. First, notice
that w=0 is not a critical point. Clearly the derivative does not exist at w=0
but then neither does the function and remember that values of w will only
be critical points if the function also exists at that point. Note that there is
also a physical reason to avoid w=0. We are constructing a box and it
would make no sense to have a zero width of the box.
Example 3 We want to construct a box with a square base and we only have 10
m2 of material to use in construction of the box. Assuming that all the material is
used in the construction process determine the maximum volume that the box can
have.
 V′(w)=12(5−3w2)
V′′(w)=−3w
Note as well here that provided w>0w>0, which from a physical
standpoint we know must be true for the width of the box, then the
volume function will be concave down and so if we get a single
critical point then we know that it will have to be the value that gives
the absolute maximum.
Setting the first derivative equal to zero and solving gives us the two
critical points,
w=±√5/3=±1.2910
l=w=1.2910
h=(5−1.291022)/(1.2910)=1.2910
Example 4 A manufacturer needs to make a cylindrical can that will hold 1.5 liters
of liquid. Determine the dimensions of the can that will minimize the amount of
material used in its construction.
Example 5 :We have a piece of cardboard that is 14 inches by 10 inches and
we’re going to cut out the corners as shown below and fold up the sides to form a
box, also shown below. Determine the height of the box that will give a maximum
volume.
Lecture #2
 Classification of optimization problems
(1) Classification Based on the Existence of Constraints:
h (X) >= 0 Inequality Constraints
h (X) = 0 Equality Constraints
(2) Classification Based on the Nature of equations involved:
Linear Programming – If both objective function and constraints are linear.
Non-Linear Programming - If any of the functions among the objective and
constraint functions is nonlinear, the problem is called a nonlinear programming
(NLP) problem.
Quadratic Programming - If any of the functions among the objective and
constraint functions is quadratic, the problem is called a quadratic programming
problem.
 Classification of optimization problems
(3) Classification based on the nature of the decision variables:
Continuous Optimization or Discrete optimization:
⚫ Some models only make sense if the variables take on values from a discrete set, often
a subset of integers, whereas other models contain variables that can take on any real
value.
⚫ Models with discrete variables are discrete optimization problems; models with
continuous variables are continuous optimization problems.
⚫ Continuous optimization problems tend to be easier to solve than discrete optimization
problems; the smoothness of the functions means that the objective function and
constraint function values at a point x can be used to deduce information about points in
a neighborhood of x.
 Classification of optimization problems
(4) Classification based on search space:

Local search methods, Global search methods

 Classification of optimization problems
(5) Classification based on number of objective functions:
No ojective: Feasibility problems are problems in which the goal is to find values for the
variables that satisfy the constraints of a model with no particular objective to optimize.
Single objective optimization: The goal is to find a solution that satisfies the complementarity
conditions.
Multi-objective optimization: when optimal decisions need to be taken in the presence of
trade-offs between two or more conflicting objectives. For example, developing a new
component might involve minimizing weight while maximizing strength or choosing a portfolio
might involve maximizing the expected return while minimizing the risk.
 Classification of optimization problems
(6) Deterministic optimization, Stochastic optimization
Deterministic optimization: In deterministic optimization, it is assumed that the data for the
given problem are known accurately.
Stochastic optimization: For many actual problems, the data cannot be known accurately
for a variety of reasons. The first reason is due to simple measurement error. The second
and more fundamental reason is that some data represent information about the future (e.
g., product demand or price for a future time period) and simply cannot be known with
certainty. In optimization under uncertainty, or stochastic optimization, the uncertainty is
incorporated into the model.
Stochastic optimization models take advantage of the fact that probability distributions
governing the data are known or can be estimated.
 GENERAL PROCEDURE FOR SOLVING OPTIMIZATION
PROBLEMS
➢Analyze the process itself so that the process variables and
specific characteristics of interest are defined; that is, make a
list of all of the variables.
➢Determine the criterion for optimization, and specify the
objective function in terms of the variables defined.
➢Using mathematical expressions, develop a valid process or
equipment model that relates the input-output variables of the
process and associated coefficients. Include both equality and
inequality constraints.. Use well-known physical principles
(mass balances, energy balances), empirical relations,
implicit concepts, and external restrictions
➢Identify the independent and dependent variables to get the
number of degrees of freedom.
➢If the problem formulation is too large in scope:
(a) break it up into manageable parts or
(b) simplify the objective function and model
➢Apply a suitable optimization technique to the
mathematical statement of the problem.
➢Check the answers, and examine the sensitivity of the
result to changes in the coefficients in the problem and the
assumptions
 OBSTACLES TO OPTIMIZATION
➢The objective function or the constraint functions may have
finite discontinuities in the continuous parameter values.
➢The objective function or the constraint functions may be
nonlinear functions of the variables.
➢The objective function or the constraint functions may be
defined in terms of complicated interactions of the variables
➢The objective function or the constraint functions may
exhibit nearly "flat“ behavior for some ranges of variables
or exponential behavior for other ranges.
➢The objective function may exhibit many local optima,
whereas the global optimum is sought.
 CLASSIFICATION OF MODELS
1. Those based on physical theory.
2. Those based on strictly empirical descriptions (so-called black
box models).

In addition to classifying models as theoretically based versus

empirical, we can generally group models according to the
following types:
• Linear versus Nonlinear.
• Steady state versus unsteady state.
ESP collection efficiency versus specific collection area for a linear
model η = 0.129A + 85.7 and a nonlinear model η = 100{ 1 - [e-0.0264A/(4.082-3.15 X l0-6A )] ) .
 • Lumped parameter versus distributed parameter.

• Lumped parameter representation means that spatial variations

are ignored and that the various properties and the state of the
system can be considered homogeneous throughout the entire
volume.

• A distributed parameter representation, on the other hand, takes

into account detailed variations in behavior from point to point
throughout the system.
Flow patterns in a stirred tank (lumped parameter system) and a
tubular reactor (distributed parameter system).
 • Continuous versus discrete variables.

Continuous variables can assume any value within an interval; discrete

variables can take only distinct values. An example of a discrete variable
is one that assumes integer values only.

Often in chemical engineering discrete variables and continuous variables

occur simultaneously in a problem. If you wish to optimize a compressor
system, for example, you must select the number of compressor stages (an
integer) in addition to the suction and production pressure of each stage
(positive continuous variables)
Model building can be divided into
four phases:
(1)Problem definition and formulation
(2) Preliminary and detailed analysis
(3) Evaluation, and
(4) Interpretation application.
I. Problem definition and formulation phase
Problem is defined and its solution are identified:
i. The number of independent variables to be included
in the model.
ii. The number of independent equations required to
describe the system (sometimes called the "order" of
the model).
iii. The number of unknown parameters in the model
 [Link] phase
It includes specification of the information content, general description of the
programming logic and algorithms necessary to develop and employ a useful
model, formulation of the mathematical description and simulation of the
model.
i. Define the input and output variables, and determine what the
"system" and the "environment”.
ii. Select the specific mathematical representation(s) to be used in the model, as
well as the assumptions and limitations of the model resulting from its
translation into computer code.
[Link] computer input-output media, develop program logic and
flowsheets, and define program modules and their structural
relationships.
[Link] of existing subroutines and databases will saves your time
III. Evaluation phase
• This Phase is basically final check of the model as a whole
• Evaluation of the model is carried out according to the evaluation
criteria and test plan established in the problem definition phase.
• Carry out sensitivity testing of the model inputs and parameters, and
determine if the apparent relationships are physically meaningful

IV. Application Phase

Regression
Regression is a statistical procedure that determines the
equation for the line/curve that best fits a specific set of data.
 Fitting Models by
Least Squares Error
y=mx+b,
 SSE =

SSE

SSE

SSE
m 0.428571
b 1
Example 1: By the method of least squares find the straight
line to the data given below:
Least Squares Fitting--Polynomial