EMCQ 4244 Chemical Process Optimization 48 hrs 1.

0 unit
Background Information:
Common misconceptions, Estimating problem difficulty, Top-down/Bottom-up strategies,
Communication of optimization results.
Strategies: Base case, Objective functions, Analysis of the base costs, Identifying and
prioritizing key decision variables.
Topological Optimization: Introduction, Elimination of non-hazardous and hazardous waste
streams, Elimination and rearrangement of equipment, Alternative separation schemes and
reactor configurations, Improvements in heat integration.
Parametric Optimization: Single variable optimization (e.g. case study of a dimethyl ether
separation column), Two variable optimization (e.g. effect of pressure and reflux ratio of a
dimethyl ether separation column), Flowsheet optimization using key decision variables.
Lattice Search techniques vs. Experimental Design.
Process Flexibility and the Sensitivity of the Optimum.

Reference Books
1. Smith R, 2016, Chemical Process Design and Integration, 2nd Edition, John Wiley,
ISBN 9781119990147
2. Optimization of Chemical Processes, Edgar Himmeblau and Lasdon, 2nd Edition
3. Gandhe Pandu Rangaiah. MULTI-OBJECTIVE OPTIMIZATION: Techniques &
Applications in Chemical Engineering.

1
Introduction to Chemical Process Optimization

Optimization refers to finding the values of decision (or free) variables,

which correspond to and provide the maximum or minimum of one or
more desired objectives.

Optimization is concerned with selecting the best feasible solution from an

infinite number using efficient quantitave methods.

Process optimization involves the determination of process parameters

(temperature, pressure, pH, time etc.) that provides us maximum output.
In the chemical industry, proper selection of batch time gives the
maximum selectivity for a batch reactor. Maximum amount of heat
recovery is possible by the optimization of HEN.

Process modeling and optimization along with control characterizes the

area of process systems engineering (PSE), important in chemical
engineering with a wide range of applications.

Chemical process industries consist of several “unit operations” and “unit

processes” e.g., heat exchanger, distillation column, batch reactor,
packed bed reactor, etc. It is the responsibility of a process engineer to
run the plant at an optimum condition to obtain the maximum profit with
minimum environmental impact.

As a process engineer, it is our primary concern to utilize resources

carefully with minimum loss. An optimized process that uses minimum
input (raw material, energy, labor) and gives maximum output (product
quality and quantity, most environmentally friendly) is always favorable.

Reasons for Process Optimization

 Improved plant performance such as improved reactor yields of
valuable products, reduced energy consumption and longer lead
times between plant maintenance or shutdowns.
 Competitive advantage: Chemical process design and optimization
can give manufacturers a competitive advantage by improving their
efficiency, quality, and sustainability.
 Attain improved designs; maximize profits; reduce cost of
productions.
 Improved product quality: By optimizing the process, manufacturers
can improve the quality of the product. The process can be designed
to produce products that meet specific quality standards, resulting
in fewer defects and customer complaints.

2
  Innovation: Chemical process design and optimization can drive
innovation in the industry by promoting the use of new technologies
and approaches. This can lead to the development of new products
and processes that are more efficient and sustainable.

Common Misconceptions during Process Optimization

Problem formulation: A common process optimization pitfall is wrong
formulation of the problem. Typical issues include; omitting some key
decision variables, oversimplified objective function, ignoring constraints.
Formulation is more important than solving. If we formulate the wrong
problems, we get wrong solutions.

Failure for the objective function to converge: The objective function

may fail to converge due to a number of reasons namely;

(i) poorly formulated objective function which yields unfeasible

solutions
(ii) Infeasible or conflicting constraints
(iii) Poor initial guesses for decision variables
(iv) Discontinuous or highly nonlinear models
(v) A trivial problem with many variables e.g 1000 variables, way
much larger than our computers will allow us to solve at this
time. The implicit solution is that we need only wait, and larger
computers will solve the problem.

Confusing optimization with simulation: Many students think

simulation automatically optimizes a process.

However:

Simulation → predicts system behavior

Optimization → searches for the best design

For example:

Aspen Plus/ChemCAD/Hysys can simulate a distillation column.

Optimization in Excel Solver, Matlab or Python determines reflux ratio,

pressure, and number of trays that minimize cost.

3
Ignoring sensitivity and uncertainty: Optimization results are often presented
as a single optimal point, but real chemical processes operate under uncertain conditions
such as:

 feed composition fluctuations

 equipment fouling
 changes in market prices

Therefore, good optimization practice includes:

 sensitivity analysis
 process flexibility analysis

Key Terms in Process Optimization

Objective function:

Objective function is the mathematical expression that we need to

optimize. This describes the correlation between the decision variables
and process parameters.

In most of the chemical engineering optimization, the objective functions

are either profit from the process or cost of production. However, there
are many types of objective functions like error during curve
fitting/parameter estimation, minimization of environmental impact.

Optimization problems must contain an objective function which we need

to optimize and can be represented in some standard form as follows:

Min/ Max=f ( X )

Subject to g ( X )=0 ; h ( X )=0

()
x1
The solution X = x 2 , this gives the optimum value of f ( X )
xn

Decision Variable

Decision variables are the variables within a model that one can control.
Decision variables are usually input to the model that can be changed by
the decision maker with the aim of revising the response of the system.
For example, a decision variable might be operating temperature of a
reactor, diameter of pipe, number of plates in a distillation column.
Constraint
Constraints are additional relations between the decision variables and
process

4
parameters other than the objective function. For instance, during the
optimization of any

blending process, summation of all components should be equal to unity

i.e.
n
¿ ∑ x i=1
i=1

Limit of decision variables

Every decision variables have some upper and lower limit, e.g mass/mole
fraction of any component in a mixture must have the value 0 ≤ X ≤1

Classification of Optimization Problems

Classification of optimization methods into different categories has been
made based on the physical structure of the problem, type of constraints,
nature of design variables, nature of algorithms and area of applications.

a) Topological optimization and parametric optimization

Topological optimization deals with the arrangement of the process
equipment during the plant design. Topological optimization should be
considered first because the topological changes usually have a large
impact on the overall profitability of the plant.

Topological arrangement examples include elimination of unwanted by-

product, elimination/rearrangement of equipment, alternative separation
methods, and alternative reactor configurations may be employed for the
improvement of heat integration. For instance, addition of one heat
exchanger can change the scenario of the whole heat integration of the
process plant.

Parametric optimization consists of the optimization of operating

conditions like temperature, pressure, concentration, flow rate etc.
Parametric optimization is easier to interpret when the topology of the
flow sheet is fixed.

b) Single objective optimization (SOO) and Multi-objective

optimization
The main focus of optimization of chemical processes so far has been
optimization for one objective at a time.

SOO focuses on a single, scalar objective function usually economic

related, such as minimizing total cost or maximizing profit.

5
Examples: Optimization of reactor temperature to maximize yield,
reducing distillation column energy consumption, or minimizing the heat
exchanger surface area.

However, practical applications involve several objectives to be

considered simultaneously e.g optimizing capital cost/investment,
selectivity, quality and/or recovery of the product, conversion, energy
required, efficiency, process safety and/or complexity, operation time,
robustness.

Multi-objective optimization (MOO), also known as multi-criteria

optimization, particularly outside engineering, refers to finding values of
decision variables which correspond to and provide the optimum of more
than one objective.

The appropriate objectives are in most cases conflicting, a tradeoff has to

be made during this optimization e.g tradeoff between energy costs,
capital costs and optimal insulation thickness. The conflicting objectives
can be handled by combining them to one single objective.

Multi-objective optimization problem can be represented as

Min(F 1 ( X ) , F 2 ( X ) … … … … F n ( X ) )
X ∈Rm
h (X ) = 0
g (X ) ≥ 0

Figure 1 illustrates the trade-offs between efficiency and emissions, suggesting that more
than one performance criterion may exist: We are forced to consider maximizing efficiency
versus minimizing emissions, resulting in some compromise of the two objectives.

6
Applications of Optimization to Chemical Processes

Optimization can be applied in numerous ways to chemical processes and

plants.

Typical projects in which optimization has been used include

i. optimization of heat exchanger network

ii. optimization of chemical reactor and reactor network
iii. application of optimization in thermodynamics, determination of chemical
equilibrium
iv. optimization of distillation system
v. optimization of water storage tank
vi. optimization of water pumping network
vii. optimization of biological wastewater treatment plant
viii. tuning of PID controller
ix. optimization of air pollution control system
x. optimization of adsorption process