CL 615 Optimization Techniques

Dr. Senthilmurugan S.

Course objective
• To provide in-depth knowledge towards optimization
techniques and its implementation techniques for industrial
applications.
• The chosen topics include linear programming, nonlinear
programming etc. Eventually, the course provides a
thorough understanding towards various optimization
techniques adopted in academic and industrial application
• Eventually, various basic approaches adopted for the
formulation of optimization models and its
implementation in process systems are summarized.

Sunday, April 10, 2022 CL615 Optimization 1

CL 615 Optimization Techniques
Course contents (3 - 0 - 0 - 6)
Up to mid Semester
• Optimization basics and convexity;
• Multidimensional constrained
optimization:
• Gradient, Secant and Newton methods;
• KarshKuhnTucker optimality conditions;
• Linear programming: Simplex
method
• Nonlinear programming:
• Sequential Quadratic Programming
(SQP),
• generalized reduced gradient method
(GRG) and
• penalty function methods;
Sunday, April 10, 2022 CL615 Optimization
After mid Semester
• Evolutionary optimization
techniques:
• Genetic Algorithm, Simulated
Annealing, particle swarm
optimization, differential evolution,
self organizing migrating algorithm
and scatter search; formulation of
optimization models in process
systems.
• Mixed integer linear
programming (MILP),
• Mixed integer nonlinear
programming (MINLP),
2
Texts/References:
1. T. F. Edgar, D. M. Himmelblau and L. S. Lasdon, Optimization
of Chemical Processes, McGraw Hill, 2001.
2. S. S. Rao, Engineering Optimization: Theory and Practice, 4th
Ed., John Wiley & Sons, 2009.
3. G. C. Onwubolu, Emerging Optimization Techniques in
Production Planning and Control, ImperialCollege Press, 2002.
4. G.V. Reklatis, A. Ravindran and K.M. Ragsdell, Engineering
Optimization Methods andApplications, John Wiley, 1983
5. L.T. Biegler, I.E. Grossmann and A.W. Westerberg, Systematic
Methods of Chemical Process Design, Prentice Hall
International Series, 1997.

Sunday, April 10, 2022 CL615 Optimization 3

 Exam Weightage
Exam Marks
Mid Semester 15
End Semester 15
Assignment 10
Viva 10
Project - Problem formulation and report 15
Project – Modelling and simulation report 15
Project - Optimization report 20
Total 100

Sunday, April 10, 2022 CL615 Optimization 4

 Optimization for
Chemical Engineering -
Introduction
Dr. senthilmurugan S,
Chemical Engineering Department,
IIT Guwahati.
Course Overview
• Chapter 1 Introduction to optimization – (2 L)
• Chapter 2 Process Modeling & Problem formulation
for optimization – (5 L)
• Chapter 3 Optimization Techniques – (7 L)
• Chapter 4 Optimization in Industrial Practice – (6 L)

Sunday, April 10, 2022 CL615 Optimization 6

Introduction to Optimization
• Optimization
• Finding an best alternative with the most cost effective
or highest achievable performance under the given
constraints, by maximizing desired factors and
minimizing undesired ones.

Optimal point Cost objective = cost insulation + cost energy loss

Cost $/year

Cost of insulation

Cost of energy loss

Insulating material thickness

Sunday, April 10, 2022 CL615 Optimization 7

Optimization of chemical process
Waste energy

Raw material Product 1

Process 1 Process 2 Product 2
Energy By Product

Waste products

• Design optimization
• Optimization of process equipment size, operating condition considering long term energy consumption, waste
and By product reduction and enhanced life of equipment, etc., for given target plant capacity
• Plant operation optimization
• After commissioning plant scenarios like, the raw material quality variations, design and fabrication mismatch,
wear tears leads to plant operating sub optimal condition then the initial design condition. Optimization of
plant operating condition for given plant design can given improved yield of valuable products and reduced
yield of by products.
• Plant level production Optimization
• Depending on market demand, the management can optimize production value products at higher rate and
others lower rate.
Sunday, April 10, 2022 CL615 Optimization 8
Hierarchy of Optimization – Based
on time horizon for operating plant
• Enterprise level resource optimization
• Longer time horizon ( day to month/year)
Level1

• Plant Level resource optimization

• Medium time horizon ( hrs to day)
Level2

• Process/equipment level optimization

• Small time horizon ( min to hrs)
Level 3

Sunday, April 10, 2022 CL615 Optimization 9

Hierarchy of Optimization –
chemical engineering examples
Raw material Market Requirement
cost variation Increase product 1
production Q1
Increase product 2
Level 1 Production planning for individual units production Q2

Production target for individual units

Level 2 Unit 1 optimization Unit 2 optimization Unit 4 optimization

Production target for individual process equipment

Process Process Process Process Process Process

Level 3 equipment 1 equipment 2 equipment 1 equipment 2 equipment 1 equipment 2

Sunday, April 10, 2022 CL615 Optimization 10

 Chapter
Co 1
n
tro
la
nd
op
tim
iza
tio
nH
i er
ar
ch
y

11
Classes of optimization problems
• Dynamic / steady state
• Optimal plant start-up and/or shut down (Minimum time problems)
• Constrained / unconstrained
• Single / multi objective
• Single decision variable / Multiple decision variable
• Deterministic / Stochastic
• Global / local optimization
• Iterative or successive / grid
• Mixed integer / continuous
• Gradient / function value
• Offline / real time or online optimization

Sunday, April 10, 2022 CL615 Optimization 12

Classes of optimization problems

Sunday, April 10, 2022 CL615 Optimization 13

Examples of applications of
optimization at different levels
• Solution for Decision support system
• Determining the best sites for plant location.
• Routing tankers for the distribution of crude and refined products.
• Sizing and layout of a pipeline.
• Designing equipment and an entire plant.
• Scheduling maintenance and equipment replacement.
• Operating equipment, such as tubular reactors, columns, and absorbers.
• Evaluating plant data to construct a model of a process.
• Minimizing inventory charges.
• Allocating resources or services among several processes.
• Planning and scheduling construction.

Sunday, April 10, 2022 CL615 Optimization 14

Sunday, April 10, 2022 CL615 Optimization 15
Optimal design of pumping
station Objective: Minimum capital cost, operating cost
By manipulating pump size distance between pumping station etc.
Subject to pump size available in market and safety precaution etc.

Storage tank

Pumping station 1 Pumping station 2

Sunday, April 10, 2022 CL615 Optimization 16

Defining Optimization problem
1. objective function: we have a goal which can be expressed in
mathematical form (a measure of how close we are to the
goal)
min f (x)
2. decision variables: means (actions) how to reach the goal
(what we can alter in order to improve the measure; degrees
of freedom)
min f(x)
x – decision variable
3. constraints: given restrictions must be fulfilled (various
requirements, facing real-life conditions)
g(x) ≤ 0
h(x) = 0

Sunday, April 10, 2022 CL615 Optimization 17

Three ingredients in optimization problem

1.  Objective  function                         economic  model

Unique solution

Underdetermined
Overdetermined

Sunday, April 10, 2022 CL615 Optimization 18

Example: Underdetermined

19
 Example: Overdetermined case-
Data Reconciliation: data preprocessing

(a) 81.4 kg /hr

(b) 84.3 Kg /hr
(c) 83.0 kg/hr

Objective function
2
𝑃∑
min ∑ ( 𝑚 𝐴 ,𝑖 mod 𝑒𝑙
+𝑚 𝐴, 𝑖 () )
exp
𝑖=1
Decision variables
mA i , mb, i , mc I, where I = 1 to number of data sets

Constraints
LB  mA i , mb, i , mc I  UB 𝑚 𝐴, 𝑖 =𝑚𝐶 , 𝑖exp 𝐵,𝑖
mod 𝑒𝑙 exp

Sunday, April 10, 2022 CL615 Optimization 20

THE SIX 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).

Sunday, April 10, 2022 CL615 Optimization 21

THE SIX STEPS USED TO
SOLVE OPTIMIZATION
PROBLEMS
3. 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 (number
of degrees of freedom).

Sunday, April 10, 2022 CL615 Optimization 22

THE SIX STEPS USED TO
SOLVE OPTIMIZATION
PROBLEMS
4. If the problem formulation is too large in scope:
Break it up into manageable parts and/or
Simplify the objective function
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.

Sunday, April 10, 2022 CL615 Optimization 23

EXAMPLES – SIX STEPS OF OPTIMIZATION
Produce specialty chemical 100,000 no/yr.
Minimize the production cost.
Total production cost consist of two components
(i) inventory or storage 8000

(ii) production cost (operation) 6000

unit/run
4000
How many no of units produced 2000
per run?
0
Step 1: Analyze the process
define variables Run number
One year
Q = total # no units produced/yr (100,000)
D = # no units produced per run One run 100, 000 units
n = # runs/yr 𝑄 One run 1 units
𝑛= One run 6667 units * 15 =100005
𝐷
Sunday, April 10, 2022 CL615 Optimization 24
Step 2: Formulate objective function and constraints

𝑀𝑖𝑛𝐶

𝐶=𝑘1 𝐷+𝑛(𝑘2+𝑘3 𝐷) evaluate constraints

⟨
𝑛 ¿ integer
¿continuous
(0)LB<D>UB (100,000)
k1=Cost of working capital for inventory and storage (Rs/ inventory size )
k2= Fixed cost to setup the factory equipment and building for production (Rs)
K3= Operating cost of plant ( Rs/unit)
Sunday, April 10, 2022 CL615 Optimization 25
Step 3: Model development
Develop objective function

Inventory, storage cost = k1D ( Storage requirement between each run)

production cost = k2 + k3 D
per run (set up operating
cost) cost per unit
(could be nonlinear then = k2 + k4 D1/2

Sunday, April 10, 2022 CL615 Optimization 26

Step 4: Simplification
simplification – converting nonlinear to linear
function

Sunday, April 10, 2022 CL615 Optimization 27

Step 5: Apply suitable optimization technique
computation of the optimum

analytical vs. numerical solution

𝑑𝐶
𝑑𝐷
𝑘2 𝑄
𝐷
𝑜𝑝𝑡
=𝑘1 − 2 =0 → 𝐷 =
𝑘2 𝑄
𝑘1 √

1/2
suppose  cost  per  run =𝑘2+𝑘4 𝐷
Sunday, April 10, 2022 CL615 Optimization 28
Step 6: Sensitivity Analysis
Sensitivity of the optimum
subst Dopt into C

𝐶 =2√𝑘1𝑘2𝑄+𝑘3𝑄
𝑜𝑝𝑡
Sunday, April 10, 2022 CL615 Optimization 29
Step 6

Sunday, April 10, 2022

𝑜𝑝𝑡
𝐷 =
𝑘 2
𝑘1
𝑄
√
CL615 Optimization 30
Example for multi-pump
optimization
• Multi-pump operation
in series / parallel
• Achieve desired
pressure head and flow
• Objective minimize
energy required for
pumping
• Vary the Pump speed?
• Any impact in energy
consumption

Sunday, April 10, 2022 CL615 Optimization 31

Step 1: Analyze the process
• Case1: If pumps are
equipped with variable
speed drive then speed
can be varied
• Case2:If pumps are not
equipped with Variable
speed drive
• Use control valve to
control the individual
pump pumping capacity

Sunday, April 10, 2022 CL615 Optimization 32

Step 2 : Formulate objective
function and constraints
• Total power = P1+P2
• Decision variables
• Pump speed n1& n2
• Constraints
• q-(q1+q2) =0,H1=H2 For
parallel connection
• H-(H1+H2) = 0, Q1=Q2 for
series
• LB<n1>UB, LB<n2>UB
Sunday, April 10, 2022 CL615 Optimization
• P= q ρ g h / (*3.6 *106)
P= power (kW)
q = flow capacity (m3/h)
ρ = density of fluid (kg/m3)
g = gravity (9.81 m/s2)
h = differential head (m) 
• h=f(q,n) =a1q2+a2(ns/n)q+a3(ns/n)2
• =f(q,n)= a4q2(n/ns)2+a5q(n/ns)+a6
• ns=refence pump speed
• a1,a2,..a6= model constants
33
Step 3: Model development
• Develop the pump energy Pump curve

consumption model as function 60

60

of speed, flow and head 50

50
• P= q ρ g h / (*3.6 *106) 40

40
30

Efficency
Head (m)
 P= power (kW) 30
20
 q = flow capacity (m3/h) 20 10
 ρ = density of fluid (kg/m3)
10
 g = gravity (9.81 m/s2) 0

 h = differential head (m)  0

15 25 35 45 55 65
-10

• h=f(q,n) =a1q2+a2(ns/n)q+a3(ns/n)2 Flow rate (m3/hr)

• =f(q,n)= a4q2(n/ns)2+a5q(n/ns) 1200 rpm H Vs Q

1500 rpm H Vs Q
1350 rpm H Vs Q
1650 rpm H Vs Q
1800 rpm H Vs Q 1200 rpm Eff vs Q
+a6 1350 rpm Eff vs Q 1500 rpm H Vs Q
1650 rpm H Vs Q 1800 rpm H Vs Q
Qn
Head  n2
Sunday, April 10, 2022 CL615 Optimization 34
Power n3
Step 3: Model development
The characteristics curve linking power to mass flow rate and pressure head
to mass flow rate may be approximated by 3rd and 4th order polynomial
respectively:

For non-nominal speeds values, all

polynomial coefficients must be
recalculated using an affinity law
according to the equations shown
below:

Sunday, April 10, 2022 CL615 Optimization 35

Step 4: Simplification

• One can derive direct power model from pump

curve leads less number of model equation
• P=f(q,n) =(a1q2(ns/n)+a2(ns/n)2q+a3(ns/n)3)/
• =f(q,n)= a4q2(n/ns)2+a5q(n/ns)+a6

Sunday, April 10, 2022 CL615 Optimization 36

Step 5: Apply suitable
optimization technique
• In order to solve the optimization problem, the
Lagrange multipliers algorithm can be been used.
This algorithm provides a strategy for finding the
maximum/minimum of a function subject to
constraints.
• For a given objective function 𝑓(𝑥,𝑦) and
constraints𝑔(𝑥,𝑦)=𝑐 a new variable (λ) is introduced
called the “Lagrange multiplier”, therefore the
optimization function takes the form: Λ(𝑥,𝑦,𝜆)=𝑓(𝑥,𝑦)
+ 𝜆∙(𝑔(𝑥,𝑦)−𝑐)

Sunday, April 10, 2022 CL615 Optimization 37

Multi pump optimization - example
n1 = 1850 rpm n1 = 1514 rpm
n2 = 1400 rpm n2 = 2042 rpm Q= 0.0028 m3/h
Pinput = 4899 W Pinput = 5489 W H = 45m
Efficiency = 22% Efficiency = 26% Poutput = 1260 W
Power (W)

rpm
Pum d
p 1 spe s pee
Pumps: ed rp 2
m ump
Tofama 3KAN25 P
Tofama 4KAN20
Step 6: Sensitivity Analysis

• Impact of optimal pump speed with respect to total

head and flow
• If pumps are identical what will be solution ?
• Impact using 3rd order polynomials model for pump
curve?
• Pump rotating sequence possibilities

Sunday, April 10, 2022 CL615 Optimization 39

To Calculate the speed of a induction motor

Srpm = 120 x F
            P
Srpm = synchronous revolutions per minute.
120   = constant
F       = supply frequency (in cycles/sec)
P       = number of motor winding poles
Example: 
What is the synchronous of a motor having 4 poles connected to a 50 hz power supply?
Srpm = 120 x F
            P
Srpm = 120 x 50
            4
Srpm = 1500 rpm

4/10/22 | Slide 40