FLOW SHEET OPTIMIZATION

Compiled by BATANIDZO TENDAI, MUDUZU

PANASHE AND DENGURE KUDAKWASHE.
1

PRESENTATION OUTLINE
Introduction
Optimization basics.
Classification of Optimization problems
Development of process(mathematical models) constrains in
optimization problems.
Components of an optimization model
Types of optimization models.
Steps for optimization
Steps used to solve optimization problems.
Optimization example.(Non-Linear Programming)
Introduction to LINGO optimization software
Merits and demerits of optimization.
Safety and environment.

INTRODUCTION

PROCESS ENGINEERING is a practical, creative and innovative application of the

chemical engineering principles to a process so that existing processes can be
improved and new process designs can be developed. The new designs will be aimed
at producing cost effective, green and more profitable processes.

PROCESS FLOWSHEET
Also known as process flow diagram(PFD) is a diagram commonly used in
chemical and process engineering to indicate the general flow of plant processes
and equipment.
It displays the relationship between major equipment of a plant facility and does
not show minor details such as piping details and designations .
3

OPTIMIZATION BASICS
WHAT IS OPTIMIZATION?

A process of making a design, system or decision as fully functional or effective as possible.

It is a fundamental and frequently applied task for most engineering activities. However this task is done by
trial and error through case study.
Research in optimization can be observed at a number of different levels that necessarily need to overlap but
are often considered by separate communities:

At the mathematical programming1 level, research focuses on understanding fundamental properties of

optimization problems and algorithms. Key issues include existence of solutions, convergence of algorithms, and
related issues such as stability and convergence rates.
The scientific computing level is strongly influenced by mathematical properties as well as the implementation
of the optimization method for efficient and practical use. Here research questions include numerical stability,
ill-conditioning of algorithmic steps, and computational complexity and performance.
At the level of operations research, attention is focused on formulation of the optimization problem and
development of solution strategies, often by using well-established solution methods. Many of the problems
encountered at this level consider well structured models with linear and discrete elements.
At the engineering level, optimization strategies are applied to challenging, and often poorly defined, real-world
4
problems. Knowledge of optimization at this level is engaged with the efficiency and reliability of applicable
methods, analysis of the solution, and diagnosis and recovery from failure of the solution method.

OPTIMIZATION BASICS
Common objectives in optimization of an industrial
process are:
1.Achieve lower capital cost design.
2.Increase process production.
3.Deduct unit operation cost.
4.Reduce environmental impact.
5.Deduct energy consumption.
5

Classification of Optimization Problems

Optimization is a key enabling tool for decision making in chemical engineering. Optimization
algorithms form the core tools for :
(a) experimental design, parameter estimation, model development, and statistical
analysis;
(b) process synthesis, analysis, design;
(c) model predictive control and real-time optimization; and
(d) planning, scheduling, and the integration of process operations into the supply chain for
manufacturing and distribution.
Optimization problems that arise in chemical engineering can be classified in terms of continuous
and discrete variables. When represented in algebraic form, the formulation of discrete/continuous
optimization problems can be written as mixed integer optimization problems. The most general
of these is the MIXED INTEGER NONLINEAR PROGRAM (MINLP) of the form:
min x,y
f (x,y)
s.t. h(x,y) = 0,
g(x,y) 0,
x Rn, y {0,1}t

where f (x,y) is the objective function (e.g., cost, energy consumption, etc.), h(x,y) = 0 are the
equations that describe the performance of the system (e.g., material balances, production rates),
and the inequality constraints g(x,y) 0 can define process specifications or constraints for feasible
plans and schedules. Note that the operator max f (x,y) is equivalent to min f (x,y). We define the
real n-vector x to represent the continuous variables while the t-vector y represents the discrete
variables, which, without loss of generality, are often restricted to take 0/1 values to define logical
or discrete decisions, such as assignment of equipment and sequencing of tasks. (These variables
can also be formulated to take on other integer values as well.)

FROM PREVIOUS PAGE WE HAVE SEEN:

The solution of optimization problems involves the use of various tools of
mathematics hence, the formulation of an optimization problem requires the use of
mathematical expressions.
Every optimization problem contains three essential categories:
1.

A function to be optimized (revenue function, cost function, etc.)

2.

Equality constraints (equations)

3.

Inequality constraints (inequalities)

No single method or algorithm of optimization exists that can be applied efficiently

to all problems. The method chosen for any particular case will depend primarily on
4.

the character of the objective function.

5.

the nature of the constraints.

6.

the number of independent and dependent variables. (Edgar and Himmelblau,

Optimization of Chemical Processes, McGraw-Hill, New York, 1988).
8

DEVELOPMENT OF PROCESS (MATHEMATICAL) MODELS

CONSTRAINTS IN OPTIMIZATION PROBLEMS
Arise from physical bounds on the variables, empirical relations, physical laws, etc.
Two general categories of models exist:
1. Those based on physical theory
2. Those based on empirical descriptions
Models based on physical and chemical laws (e.g., mass and energy balances,
thermodynamics, chemical reaction kinetics) are frequently employed in
optimization applications.
These models are conceptually attractive because a general model for any system
size can be developed before the system is constructed. On the other hand, an
empirical model can be devised that simply correlates input output data without
any physiochemical analysis of the process.

DEGREES OF FREEDOM
These are the number of input variables you need to specify
A degree of freedom analysis is incorporated in the development of each
subroutine that simulate a process unit.
1. Over specified problem: too much information and it is either redundant
or inconsistent.
Nvariables >> Nequations
Fitting data
2. Equally specified problem: proceed solving the problem
Units in flow sheet
Nvariables= Nequations
3. Under specified problem: some values are underdetermined-can be
manipulated to optimize the process.
Optimization
11
Nvariables<< Nequations

COMPONENTS OF AN OPTIMIZATION
MODEL
An optimization model has three main components:
An objective function. This is a function that needs to be optimised.
A collection of decision variables. The solution to the optimization
problem is the set of values of decision variables for which the objective
function reaches its optimal value.
A collection of constraints that restrict the value of decision variables.
The mathematical approach is represented as follows:

Any optimization problem can be represented as, min or max f(x), c(x)=0 , g(x)

f(x) is the objective function

c(x) is the set of m equations in n variables x. the equality constraints.

g(x) is the set of r inequality constrains. These are bound to the

feasible region of operation

TYPES OF OPTIMIZATION MODELS

1. Unconstrained optimization problem: optimization problem where the objective function can
be of any kind, linear or non-linear and there are no constraints.
2. Linear program: optimization problem with an objective function that is linear in the variables,
and or constraints are also linear. Linear programs are implemented by the Linear-Program
class.
3. Quadratic program: optimization problem with an objective function that is quadratic in the
variables and all constraints are equal. Quadratic programs are implemented by the
Quadratic-Program class.
4. Non-linear program: optimization problem with an objective function that is arbitrary nonlinear function of the decision variables, and the constraints can be linear or non-linear. Nonlinear programs are implemented by the Non-Linear program class.
.Decision Variables
The goal of an optimization model is finding the optimal values of the decision variables. In
the optimization framework, variables are implemented by the DecisionVariable class.
. Constraints
Constraints limit the possible values for the decision variables in an optimization model

4. Number of Decision Variables

ND=Nvariables-Nequations
5.Objective Function is optimized with respect to ND Variables
Minimize Cost
Maximize Investor Rate of Return
6. Subject To Constraints
Equality Constraints
Mole fractions add to 1
Inequality Constraints
Reflux ratio is larger than Rmin
Upper and Lower Bounds
Mole fraction is larger than zero and smaller than 1
14

Steps for Optimization Problems

1.
2.
3.
4.
5.

6.

Draw and label a picture. Optimization problems often involve geometry. Draw a picture
of the situation. Include any information you are given in the problem. Identify the quantities
under your control and assign variables to them.
Find the objective function. What is the quantity you want to maximize or minimize?
Write a formula for it in terms of the variables in your picture.
Identify the constraints. If your objective function has more than one variable, you will
need to use one or more constraints in the problem to write equations that relate the variables
together.
Reduce the objective function to one variable. Solve each of the constraint equations for
one of the variables and substitute this into the objective function. At the end, the objective
function should contain just one variable.
Identify the domain of the objective function. Now that your objective function has a
single variable, what values of that variable make sense in the problem? Think about what the
variable means in the problemfor example, lengths cannot be negative. There may be both
a lower bound and an upper bound for meaningful values of the variable; if so, make sure you
identify both of them. If the domain has both a lower and an upper bound, and the endpoints of
this interval do not cause any mathematical problems in the objective function (such as division
by zero), consider including the endpoints even if they dont quite make sense in the problem;
this will allow you to use the Closed Interval Method in step 8.
Differentiate the objective function.

7. Find the critical numbers. Critical numbers are values of the variable that cause the derivative
to equal zero or
be undefined. These numbers are potential local maxima or minima. Ignore
any critical numbers that are outside the domain of the objective function.
8. Test the critical numbers (and possibly the endpoints of the domain). The critical numbers
are not automatically the answer. You need to test each one to see if its a local minimum, a
local maximum, or neither. Usually you can use one of the following tests:
i.
Closed Interval Method. Requirements: The domain of the objective function must
be a closed, bounded interval (that is, an interval that has endpoints on both ends, and
includes those endpoints). Test: Evaluate the objective function at all critical numbers
and at the endpoints of the domain, and choose the largest value (if you are maximizing)
or the smallest value (if you are minimizing).
ii.
First Derivative Test. Requirements: There must be only one critical number in the
domain. Test: Evaluate the first derivative of the objective function at two numbers in
the domain, one of which is less than the critical number and the other of which is greater
than the critical number. If the first derivative is positive to the left of the critical number
and negative to the right, then the critical number represents the absolute maximum. If
the first derivative is negative to the left of the critical number and positive to the right,
then the critical number represents the absolute minimum.
iii. Second Derivative Test. Requirements: There must be only one critical number in
the domain. Test: Find the second derivative of the objective function and evaluate it
at the critical number. If this value is negative, then the critical number represents the
absolute maximum. If the value is positive, then the critical number represents the
absolute
minimum.

9.

Answer the question. The critical number itself may not be the answer to the question.
Reread the question to see exactly what it is asking for. If it is asking for the dimensions of a
rectangle, for example, make sure you give the width and height of the rectangle, not just the
area. And dont forget units!

STEPS USED TO SOLVE OPTIMIZATION PROBLEMS

1. Analyze the process itself so that the process variables and specific characteristics of

interest are defined (i.e., make a list of all of the variables).

2. Determine the criterion for optimization and specify the objective function in terms of
the above variables together with coefficients. This step provides the performance model
(sometimes called the economic model when appropriate).
3. Develop via mathematical expressions 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 (number of degrees of freedom).
4. If the problem formulation is too large in scope, (a) break it up into manageable parts
and/or (b) simplify the objective function and model.
5. Apply a suitable optimization technique to the mathematical statement of the problem.
6. Check the answers and examine the sensitivity of the result to changes in the
coefficients in the problem and the assumptions.

Classification of Optimization Problems

Continued

Linear Programs (LPs)

A mathematical program is linear if

f(x1,x2,,xN) and gi(x1,x2,,xN)0 are linear in each of

their arguments:
f(x1,x2,,xN) = c1x1 + c2x2 + . cNxN
gi(x1,x2,,xN) = ai1x1 + ai2x2 + . aiNxN
where ci and aij are known constants.

The accepted procedure, linear programming (LP), has become quite popular, solving a
wide range of industrial problems. It is increasingly being used for on-line optimization. For
processing plants, there are several different kinds of linear constraints that may arise,
making the LP method of great utility.
1. Production limitation due to equipment throughput restrictions, storage limits, or market
constraints.
2. Raw material (feedstock) limitation.
3 . Safety restrictions on allowable operating temperatures and pressures.
4. Physical property specifications placed on the composition of the final product.
5. Material and energy balances of the steady-state model.

 Non-Linear Programs (NLPs)

A mathematical program is non-linear if any of the arguments are non-linear. For example:
min 3x + 6y2
s.t. 5x + xy 0
general case for optimization that occurs when both the objective function and constraints are
nonlinear
Other established types of constrained optimization methods include the following types of
algorithms:
1. Penalty functions with augmented Lagrangian method (an enhancement of the classical
Lagrange multiplier method)
2. Successive quadratic programming
Integer Programming
Optimization programs in which ALL the variables must assume integer values. The most
commonly used integer variables are the zero/one binary integer variables.
Integer variables are often used as decision variables, e.g. to choose between two reactor
types.

 Mixed Integer Linear Programs (MILPs)

Linear programs in which SOME of the variables are real and other variables are
integers
Can be solved as individual LPs by fixing the integer variables, thus a global
optimum can be identified.
Mixed Integer Non-Linear Programs (MINLPs)
Non-linear programs in which SOME of the variables are real and other variables
are integers
Can be solved as individual NLPs by fixing the integer variables, but depending on
the nature of the NLPs it may not be possible to find a global optimum.

Design of a Small Heat Exchanger Network

Consider the optimization of the small process network shown in Figure below with two
process
streams and three heat exchangers. Using temperatures defined by Tin > Tout and tout > tin,
the hot stream with a fixed flow rate F and heat capacity Cp needs to be cooled from Tin
to Tout, while the cold stream with fixed flow rate f and heat capacity cp needs to be
heated from tin to tout. This is accomplished by two heat exchangers; the heater uses
steam
at temperature Ts and has a heat duty Qh, while the cooler uses cold water at temperature
Tw
and has a heat duty Qc. However, considerable energy can be saved by exchanging heat
between the hot and cold streams through the third heat exchanger with heat duty Qm and
hot
and
cold
exit
temperatures,
Tm
and
tm,
respectively.
The model for this system is given as follows:
The energy balance for this system is given by:

Qc = F Cp(Tm Tout),

INTRODUCTION TO LINGO
WHAT IS LINGO ???????

LINGO is a simple tool for utilizing the power of linear and

nonlinear optimization to formulate large problems concisely,
solve them, and analyze the solution.
Optimization helps you find the answer that yields the best result;
attains the highest profit, output, or happiness; or achieves the
lowest cost, waste, or discomfort. Often these problems involve
making the most efficient use of your resources including money,
time, machinery, staff, inventory, and more.
Optimization problems are often classified as linear or nonlinear,
depending on whether the relationships in the problem are linear
with respect to the variables.

LINGO includes several features, including:

A new global solver to confirm that the solution found is the global
optimum,
Multi-start capability to solve problems more quickly,
Quadratic recognition and solver to identify quadratic programming (QP)
problems,
A faster and more robust Dual Simplex solver,
An improved integer solver to enhance performance in solving many types
of problems,
Linearization capability to transform common non smooth functions to a
series of linear
functions,
Infeasible and unbounded analytical tools to help identify model definition
problems,
A decomposition feature to identify if a model contains independent sub
models,
A thread safe DLL for various classes of models, and
More fun than ever before

Creating a LINGO Model

An optimization model consists of three parts:
Objective function This is single formula that describes exactly what the
model should optimize. A general manufacturing example of an objective
function would be to minimize the cycle time for a given product.
Variables These are the quantities that can be changed to produce the
optimal value of the objective function. For example, when driving a car, the
duration of the trip (t) and the speed at which it is taken (v) determine the
distance (d) that can be travelled.
Constraints These are formulas that define the limits on the values of the
variables. If an ice cream store is determining how many flavours it should offer,
only a positive number of flavours is feasible. This constraint could be expressed
as
Flavours >= 0;
A sample model for cookie production by two bakers at a bakery is given by:
A cookie store can produce drop cookies and decorated cookies, which sell
for $1 and $1.50 a piece, respectively. The two bakers each work 8 hours per
day and can produce up to 400 drop cookies and 200 decorated cookies. It
takes 1 minute to produce each drop cookie and 3 minutes to produce each
decorated cookie. What combination of cookies produced will maximize the
baker's profit?

Several set looping functions are also available for use in LINGO. These functions are as follows:
@FOR generates constraints over members of a set.
@SUM sums an expression over all members of the set.
@MIN computes the minimum of an expression over all members of the set.
@MAX computes the maximum of an expression over all members of the set.

SHOW CASE ON WINDOW

MODEL:

OBJECTIVE
FUNCTION

MAX = 400 * drop + 200 *decorated;

drop <= 400;
decorated <= 200 ;
1/60*drop+ 3/60*decorated <= 16;
END:
VARIABLES

CONSTRAINTS

Several other items must be noted about this model:

Comments in the model are initiated with an exclamation point
(!) and appear in green text.
LINGO specified operators and functions appear in blue text.
All other text is shown in black.
Each LINGO statement must end in a semi-colon (;).
Variable names are not case-sensitive and must begin with a
letter (A-Z). Other characters in the variable name may be
letters, numbers (0-9), or the underscore character (_). Variable
names can be up to 32 characters in length.

Solving a LINGO Model

Once the LINGO model has been entered into the LINGO Model window, the model can be solved
by clicking the Solve button on the toolbar, by selecting LINGO | Solve from the menus, or by
using the ctrl + s keyboard shortcut. LINGO will notify you of any errors it has encountered. The
best way to get information about these errors is to consult the Error Messages section in the
softwares proprietary tutorial. If no errors are found, then the LINGO Solver Status window
appears.

LINGO Programming Example

A good programming model example is the transportation problem. Transportation problems
deal with transporting goods from one location to another at minimal cost. This example
shows how to model a simple transportation problem.
MODEL:
! A 6 Warehouse 8 Vendor Transportation Problem;
SETS:
WAREHOUSES / WH1 WH2 WH3 WH4 WH5 WH6/: CAPACITY;
VENDORS / V1 V2 V3 V4 V5 V6 V7 V8/ : DEMAND;
LINKS( WAREHOUSES, VENDORS): COST, VOLUME;
ENDSETS
! The objective;
MIN = @SUM( LINKS( I, J): COST( I, J) * VOLUME( I, J));
! The demand constraints;
@FOR( VENDORS( J):
@SUM( WAREHOUSES( I): VOLUME( I, J)) = DEMAND( J));

! The capacity constraints;

@FOR( WAREHOUSES( I): @SUM( VENDORS( J): VOLUME( I, J))
<= CAPACITY( I));
! Here is the data;
DATA: CAPACITY = 60 55 51 43 41 52;
DEMAND = 35 44 34 42 39 30 38;
COST = 6 2 6 7 4 2 5 9
49538582
52197433
76739271
23957265
55228143
4 4 5 3 1 1 7 2;
ENDDATA
END

ADVANTAGES OF OPTIMIZATION
1.

improved plant performance.

2.

improved yields of valuable products (or reduced yields of contaminants).

3. reduced energy consumption.

4.

higher processing rates.

5.

longer times between shutdowns.

6.

lead to reduced maintenance costs

7.

less equipment wear

8. better staff utilization.

9. helpful to systematically identify the objective, constraints, and degrees of
freedom in a process plant.
10. improved quality of designs, faster and more reliable troubleshooting, and
faster decision making is achieved

DISADVANTAGES OF OPTIMIZATION
1.Application to systems is vexatious
2.Involve complex mathematical models
3.Optimization at large scale is dependent on high quality expensive
machines
4.Involve high computational input of data onto computer packages such
as ASPEN, MATLAB and LINGO
5.Requires more time to simulate changes.
6.Produce a compromising product conversion that sometimes produce
high toxins

SAFETY, HEALTH AND ENVIRONMENT

Optimization transforms existing engineering disciplines and
practices to those that lead to sustainability. It incorporates
development and implementation of products, processes, and
systems that meet technical and cost objectives while protecting
human health and welfare and elevates the protection of the
biosphere as a criterion in engineering solutions.
Optimization of flow sheets of chemical manufacturing process is
described as inherently safer if it reduces or eliminates hazards
associated with materials used and operations, and this reduction
or elimination is a permanent and inseparable part of the process
technology. (Kletz, 1991; Hendershot, 1997a, b)