Krishna's

O ptimization
T E C H N IQ U E S
As per U.P. UNIFIED Syllabus (w.e.f. 2012–2013)
( for B.C.A. IVth Sem. Students of all Colleges Affiliated to Universities in Uttar Pradesh)

By

J.P. Chauhan
M.Sc. (Mathematics)

C.C.S. University, Meerut

KRISHNA Prakashan Media (P) Ltd.

KRISHNA HOUSE, 11, Shivaji Road, Meerut-250 001 (U.P.), India
 Jai Shri Radhey Shyam

Dedicated
to

Lord

Krishna
Author & Publishers
 A bout the book
F The book has been written in a simple, logical and systematic way.

F The basic concept of each topic is carefully and clearly explained.

F Each chapter includes a problem set containing routine, intermediate and challenging
exercises.

F Solutions to all problem sets are given at the end of each chapter for learners to
practice.

F The goal of the book is to present a wide variety of applications of Optimization

Techniques in a flexible and accessible format.

F The book has been written to meet the requirements of students pursuing B.C.A. in
colleges affiliated to U.P. UNIFIED.

F The book also contains illustrative examples and theorems along with their proof.
"

— J.P. Chauhan

(iv)
 P reface
I t gives me pleasure to place the book ‘Optimization Techniques’ (All Universities
in U.P.) which has been completely revised and made upto date in the light of
suggestions received from the students and the learned teachers of various universities.

The following are the main features of third edition:

> More than 150 new problems and examples have been added.
> Includes recent elegant and elementary proof.
> All chapters are arranged according to the latest U.P. UNIFIED Syllabus
implemented from 2012 – 2013, keeping in view the requirement of the students of
B.C.A. IVth semester of all Universities in Uttar Pradesh (U.P.).
> Errors and omissions have been corrected.
> Up-to-date yearwise references from various universities examination papers have
been given throughout the book.
> Each step is made clear by giving reasons with clarification.

We hope that the book in its present form will prove beneficial to the students. We shall
also forward to receive your suggestions and comments about the book so that it could be
further improved.
"

— J.P. Chauhan

(v)
 A cknowledgement
T
his book has been written in simple and lucid style for better understanding and

appreciation of study. It has been compiled systematically to enable sure one to

evaluate detail of the book.

Needless to say, I am deeply indebted to the publisher Shri Satyendra Rastogi

(Managing Director) and Shri Sugam Rastogi (Executive Director) for providing

continuous support and enthusiasm.

This acknowledgement would fall short of completion if I fail to mention the entire team

of Krishna Prakashan. I would like to thank the contributors for their creativity and

especially their excellent work.

Finally, we shall be very grateful to all instructors and students who send us their helpful

suggestions.

Meerut (U.P.)
"

—J.P. Chauhan

(vi)
 S yllabus
Optimization Techniques
U.P. UNIFIED (w.e.f. 2012-2013)
B.C.A. – S 209 B.C.A. IVth Semester

Un i t - 1
Linear Programming: Central problem of linear programming various definitions
included statements of basic theorem and also their properties, simplex methods, primal
and dual simplex method, transport problem, tic-tac problem, and its solution. Assignment
problem and its solution. Graphical method formulation, Linear programming problem.

Un i t - 2
Queuing Theory: Characteristics of queuing system, Classification of queuing model
single channel queuing theory, Generalization of steady state M/M/1 queuing models
(Model-I, Model-II).

Un i t - 3
Replacement Theory: Replacement of item that deteriorates replacement of items that
fail. Group replacement and individual replacement.

Un i t - 4
Inventory Theory: Cost involved in inventory problem-single item deterministic model
economics long size model without shortage and with shorter having production rate
infinite and finite.

Un i t - 5
Job Sequencing: Introduction, solution of sequencing problem johnson s algorithm for n
jobs through 2 machines.

(vii)
 Metrization Theorems and Paracompactness 1

B rief Contents
Dedication ........................................................................................................................................... (iii)
About the Book......................................................................................................................................(iv)
Preface .................................................................................................................................................. (v)
Acknowledgement...................................................................................................................................(vi)
Syllabus (B.C.A. – U.P. UNIFIED)....................................................................................................(vii)
Brief Contents.....................................................................................................................................(viii)
Detailed Contents.............................................................................................................................(ix-xii)

Chapter 1: Basic Concepts of Optimization...............................................................................................(01-08)

Chapter 2: Simplex Methods..........................................................................................................................(09-74)

Chapter 3: Transportation Problem...........................................................................................................(75-124)

Chapter 4: Assignment Problem................................................................................................................(125-164)

Chapter 5: Linear Programming and Graphical Methods................................................................(165-230)

Chapter 6: Queuing Theory.........................................................................................................................(231-270)

Chapter 7: Replacement Problems............................................................................................................(271-314)

Chapter 8: Inventory Theory......................................................................................................................(315-344)

Chapter 9: Job Sequencing.........................................................................................................................(345-384)

(viii)
 D etailed Contents
Chapter 1: Basic Concepts of Optimization.....................................................(01-08)
1.1 Historical Background of Operations Research 01
1.2 What is Operations Research? 02
1.3 The Essential Characteristic of Operations Research 02
1.4 Modelling in Operations Research 03
1.4.1 Advantages and Limitations of a Model 03
1.4.2 Classification Schemes of Models 04
1.5 Limitations of Operations Research 05
1.6 Management Applications of Operations Research 05
1.7 Techniques Used in Operations Research 07
1.8 The Importance of Operations Research in Decision Theory 07
❖ Problem Set 08

Chapter 2: Simplex Methods..............................................................................(09-74)

2.1 Linear Programming Problems (Simplex Method) 09
2.1.1 Standard form of a Linear Programming Problem 09
2.1.2 Conversion of a L.P.P. into Standard Form 09
¤ Solved Examples 11
2.1.3 Some Important Definitions Regarding the Solution of L.P.P. Using Standard
Form/Canonical Form 13
2.1.4 Canonical Form of a General L.P.P. 14
❖ Problem Set 14
❖ Answers 15
2.2 Procedure of the Simplex Method 15
2.2.1 Brief Explanation of Simplex Tableau 1 and the Various Factors Connected With
its Construction 16
❖ Problem Set 39
❖ Answers 42
2.3 Linear Programming Problems Big 'M' Method and Two Phase Method 42
2.3.1 Use of Artificial Variables 42
2.3.2 Method of Penalties 43
2.3.3 Two Phase Method (Due to Dantzing, Orden and Wolfe) 55
❖ Problem Set 63
❖ Answers 64
(ix)
 2 Topology

2.4 Duality in Linear Programming 65

2.4.1 Symmetric Primal Dual Problems 65
2.4.2 Matrix form of Symmetric Primal Dual Problem 66
2.5 Unsymmetric Primal Dual Problems 67
2.5.1 Dual of an L.P.P. with Mixed Restrictions 67
❖ Problem Set 73
❖ Answers 74

Chapter 3: Transportation Problem................................................................(75-124)

3.1 Transportation Problem 75
3.2 Mathematical Formulation of Transportation Problem 76
3.2.1 Definition 76
3.2.2 Existence of Feasible Solution 77
3.2.3 Existence of an Optimal Solution 79
3.3 Methods to Solve Transportation Problem 79
3.3.1 Algorithm for North-West Corner Cell Method 80
3.3.2 Transportation Problem as L.P. Problem 80
3.3.3 Objective of Transportation Problem 80
¤ Solved Examples 81
3.3.4 Algorithm for Least Cost Cell Method 84
3.3.5 Algorithm for Vogal's Approximation Method 87
3.4 Transportation Algorithm (MODI) Method 100
❖ Problem Set 121
❖ Answers 124

Chapter 4: Assignment Problem ...................................................................(125-164)

4.1 Assignment Problem 125
4.1.1 Mathematical Formulation of Assignment Problem 126
4.1.2 Difference between Assignment and Transportation Problem 126
4.1.3 Fundamental Theorems 127
4.2 Assignment Algorithm or Hungarian Method or Reduced Matrix Method 128
¤ Solved Examples 130
4.3 Unbalanced Assignment Problem 145
4.4 Maximal Assignment Problem 146
4.5 Restrictions on Assignment Problem 146
4.6 Application of Assignment Problem 146
4.7 Objective of Assignment Problem 146
❖ Problem Set 159
❖ Answers 164

Chapter 5: Linear Programming and Graphical Methods ........................(165-230)

5.1 Linear Programming 165
5.1.1 Important Applications of Linear Programming 165
5.1.2 Objectives of Using Linear Programming 166
5.1.3 Main Characteristics of Linear Programming 166
5.1.4 Mathematical Form of Linear Programming Problem 167
5.1.5 Formulation of Linear Programming Problem 168
5.1.6 Limitations of Linear Programming 168

(x)
 ¤ Solved Examples 169
❖ Problem Set 182
❖ Answers 185
5.2 Solution of a Linear Programming Problem 186
5.2.1 Methods of Solution of a Linear Programming Problem 187
❖ Problem Set 225
❖ Answers 229

Chapter 6: Queuing Theory............................................................................(231-270)

6.1 Introduction 231
6.2 The Basic Queuing Process and its Characteristics 232
6.2.1 Characteristics of Queuing System 232
6.2.2 Customers Behavior in a Queue 233
6.3 Important Definitions in Queuing Problem 233
6.4 Poisson Process 234
6.5 Some Distributions 235
6.6 Kendall's Notation for Representing Queuing Models 236
6.7 A List of Symbols 237
6.8 Model 1 (M / M /1) : (∞ / FCFS) (Birth and Death Model) 238
6.8.1 Inter-relationship between L s , Lq , Ws , Wq 239
¤ Solved Examples 239
6.9 Model II (M/M/1): (N/ FCFS) Finite Queue Length Model 250
6.10 Multi-Channel Queuing Theory Model III (M / M / C : ∞ / FCFS) 253
6.11 Model IV (M / E k /1) : (∞ / FCFS) 260
❖ Problem Set 266
❖ Answers 270

Chapter 7: Replacement Problems ...............................................................(271-314)

7.1 The Replacement Problems 271
7.2 Replacement of Items that Deteriorate with Time 272
7.2.1 Theorem-1 272
¤ Solved Examples 274
❖ Problem Set 282
❖ Answers 282
7.3 Money Value Present Worth Factor (P.W.F.) and Discount Rate 283
7.3.1 Money Value 283
7.3.2 Present Value or Present Worth Factor 283
7.3.3 Discount Rate (Depreciation Value) 283
7.4 Replacement of Items whose Maintenance Costs Increase with Time and Value
of Money also Change with Time 285
7.4.1 Theorem 285
7.4.2 Theorem 288
7.4.3 Procedure for Finding best Machine 290
❖ Problem Set 294
❖ Answers 294
7.5 Replacement of Items that Fail Suddenly 294
7.5.1 Individual Replacement Policy : Mortality Theorem 295
7.5.2 Group Replacement Policy 295
(xi)
 4 Topology

7.6 Recruitment and Promotion Problems 306

❖ Problem Set 313
❖ Answers 314

Chapter 8: Inventory Theory..........................................................................(315-344)

8.1 Introduction 315
8.2 Necessity for Maintaining Inventory 315
8.3 Inventory Costs 316
8.3.1 Inventory Carrying Costs or Stock Holding Costs 316
8.3.2 Ordering or Set-up Cost 317
8.3.3 Shortage or Stock Out and Customer-service Cost 317
8.3.4 Total Inventory Cost 317
8.4 Replenishment Lead Time 318
8.5 Inventory Control Problem 318
8.6 Concept of Economic Ordering Quantity (E.O.Q. ) 318
8.7 List of Symbols Used 319
8.8 Deterministic Models 320
8.8.1 Single Item Inventory Control Models without Shortages 320
¤ Solved Examples 322
8.9 Economic Lot Size with Different Rates of Demand in Different Cycles 326
8.9.1 Economic Lot Size with Finite Rate of Replenishment or Production
Lot-Size Model (PLS) 327
8.9.2 Limitations of E.O.Q. Formula 331
8.10 II nd Model the E.O.Q. Model with Shortage 332
8.10.1 I st Form Single Item Inventory Control Models with Shortages 332
nd
8.10.2 II Form The E.O.Q. with Constant Rate of Demand Scheduling
Time Variable 334
rd
8.10.3 III Form the Production Lot Size Model with Shortages 336
❖ Problem Set 341
❖ Answers 344

Chapter 9: Job Sequencing ............................................................................(345-384)

9.1 Sequencing Problem 345
9.2 General Assumptions 346
9.2.1 Terminology and Notations 346
9.3 Sequencing Decision Problem for n-jobs on Two Machines 347
¤ Solved Examples 348
9.4 Sequence Decision Problem for n-jobs on three Machines 356
9.5 Sequence Decision Problem for n-jobs on m Machines 363
9.6 Processing Two Jobs Through m Machines 366
❖ Problem Set 371
❖ Answers 376
9.7 The Travelling Salesman Problem 377
❖ Problem Set 383
❖ Answers 384
(xii)
 1

C HAPTER
1

Basic Concepts of Optimization

1.1 Historical Background of Operations Research

No science has ever been born on a specific day. Operations Research is no exception. Its
roots are as old as science and society. It was in 1917 when A.K. Erlang, a Danish
mathematician, published his work on the problem of congestion of telephone traffic.
The difficulty was that during busy periods, telephone operators were unable to handle
the calls the moment they were made, resulting in delayed calls. A few years after the
British Post Office as the basis for calculating circuit facilities. During the 1930, H.C.
Levinson applied scientific analysis to the problems of merchandising. His work
included scientific study of customers 'buying habits, response to advertising and
relation of environment to the type of article sold.

Operations Research come to the fore and become established as a subject in its own
right during world war II. It was necessary at that time to deploy resources in the most
economical and efficient ways.

Past-World War Development

Following the end of world war II, the success of military teams attracted the attention of
industrial managers of U.K., who were seeking solutions to their complex executive type
problems.
 2

In recent years, operation research had an increasingly great impact on the management
of organizations.

Many industries including air craft and missile, automobile, communication, computer,
electronics, mining paper, petroleum and transportation have made wide spread use of
Operations Research in determining their tactical and strategical decisions more
scientifically.

1.2 What is Operations Research?

[B.C.A. (Agra) 2004, 2006, 2009; B.B.A. (Meerut) 2006, 2008, 2008 (Dec)]
Operations Research (O.R.) has been variously described as the “Science of Use”,
“quantitative commonsense”, “Scientific approach to decision making problems”. But
only a few are commonly used and widely accepted, namely.

“O.R. is the art of giving bad answers to problems which otherwise have worse answers”.
− T.L. Saaty

“O.R. is the application of scientific methods, techniques and tools to problems involving the
operations of a system so as to provide those in control of the system with optimum solutions to the
problem”.
− C.W. Churchman, R.L. Ackoff

“O.R. is a scientific approach to problems solving executive management”.

− H.M. Wagner

“O.R. is a scientific method of providing executive departments with quantitative basis for decisions
regarding the operations under their control.”

− Morse and Kimball.

1.3 The Essential Characteristics of Operations Research

[B.C.A. (Kanpur) 2007; B.B.A. (Meerut) 2003]

The significant features of O.R. are given below:

1. Decision Making: A major premise of O.R. is that decision-making, irrespective of
the situation involved, can be considered as a general systematic process.
2. Scientific Approach: O.R. employs scientific methods for the purpose of solving
problems. It is a formalised process of reasoning.
3. Objective: O.R. attempts to locate the best or optimal solution to the problem
under consideration.
 3

4. Inter-disciplinary Team Approach: O.R. is inter-disciplinary in nature and

requires a team approach to a solution of the problem. Managerial problems have
economic, physical, biological and engineering aspects. This requires a blend of
people with expertise in the areas of mathematics, statistics, engineering, economics,
management, computer science and so on.
5. Digital Computer: Use of digital computer has become an integral part of O.R.,
approach to decision-making. The computer may be required due to complexity of
the model, volume of data required and the computations to be made.

1.4 Modelling in Operations Research

[B.C.A. (Meerut) 2009]
By building a model, the uncertainties and complexities of a decision-making problem
can be changed to a logical structure that is amenable to formal analysis. Such as model
specifies the decision alternatives and their anticipated consequences for all possible
events that may occur, indicates the relevant data for analysing the alternatives and leads
to meaningful and informative managerial conclusions. In short we may say, modelling is
a means of providing a clear structural framework to the problem for purposes of
understanding and dealing with reality. Following are the main characteristics of good
modelling.
1. A good model should be capable of taking into account new formulations without
having any significant change in its frame.
2. Assumptions made in the model should be as small as possible.
3. The number of variables must be less.
4. The good model should be simple and coherent.
5. It should not take much time in its construction for any problem.

1.4.1 Advantages and Limitations of a Model

[B.C.A. (Meerut) 2010]
1.4.1.1 Advantages of a Model
1. Through a model, the position under consideration becomes controllable.
2. It help in finding evenues for few research and improvements in a system.
3. It indicates the limitations and scope of an activity.
4. It provide some logical and systematic approach to the problem.

1.4.1.2 Limitations of a Model

1. Models are only an attempt in understanding operations and should never be
considered as absolute in any sense.
2. Validity of any model with regard to corresponding operation can only be verified
by carrying the experiment and relevant data characteristics.
 4

1.4.2 Classification Schemes of Models

Although the classification of models is a subjective problem they may be distinguished
as follows:
1. Models by Degree of Abstraction: These models based on past data or information
of problem under consideration and can be classified into
(i) Language models
(ii) Case studies
2. Models by Function: These models consist of
(i) Descriptive models
(ii) Predictive models
(iii) Normative models
F Descriptive models describe, explain and predict facts and relationships among the
various activities of the problem. These are used to describe mathematically some
particular aspects of the system being modelled.
F Predictive models indicate ‘if this occurs, then what will follow’. They relate
dependent and independent variables and permit trying out, ‘what if ’ questions. We
can say that these models are used to predict the out comes due to a given set of
alternatives for the problem. These models do not have an objective function as a
part of the model to evaluate decision alternatives.
F Normative or optimization models are prescriptive in nature and develop
objective decision rules or criteria of optimum solutions.
3. Models by Structure: These modes are represented by
(i) Iconic models
(ii) Analogue models
(iii) Symbolic models
F Iconic or physical models are pictorial representation of real systems and have the
appearance of the real thing, example of such models are:
City maps, houses blue prints, globe and so on.
An iconic model is said to be scaled-down or scaled up' according as the dimensions
of the model are smaller or greater than those of the real item. Iconic models are
easy to observe, build and describe, but are difficult to manipulate and not very
useful for the purpose of prediction. Hence we can say these models represent a
static event.
F Analogue models are more abstract than the iconic ones for there is no look alike
correspondence between these models and real life items. They are built by utilizing
one set of properties to represent another set of properties.
F Mathematical or symbolic models are more abstract in nature. They employ a set
of mathematical symbol to represent the components of the real system. These
models are most general and precise. However it is not possible to depict a real
system in mathematical formulation. Sometime it is easier to used to mathematical
symbol for describing the relationship of the component.
 5

4. Model by Nature of the Environment: These model can be classified into

(i) Deterministic models, and
(ii) Probabilistic models.
In deterministic models, all the parameters and functional relationships are
assumed to be known with certainly when the decision is to be made. Linear
programming and break even models are the examples of deterministic models.

1.5 Limitations of Operations Research

[B.C.A. (Agra) 2006; B.B.A. (Kanpur) 2007]
1. Mathematical models which are essences of O.R. do not take into account
quantitative factors or emotional factors which are quite real. All influencing
factors which cannot be quantified find no place in mathematical models.
2. Mathematical models are applicable to only specific categories of problems.
3. Being the new field there is resistance from the employees to the new proposals.
4. Management, who has to implement the advised proposals, may itself offer a lot of
resistance due to conventional thinking.
5. Young enthusiasts, overtaken by its advantages and exactness generally forget that
O.R. is meant for men and not that men are meant for it.

1.6 Management Applications of Operations Research

[B.C.A. (Rohilkhand) 2010; B.B.A. (Meerut) 2003, 2004, 2006, 2008]
Some of the areas of management decision making, where the ‘tools’ and ‘techniques’ of
O.R. are applied, can be outlined as follows:
1. Finance, Budgeting and Investments
(i) Cash-flow analysis, long range capital requirements, dividend policies,
investment portfolios.
(ii) Credit policies, credit risks and delinquent account procedures.
(iii) Claim and complaint procedure.
2. Purchasing, Procurement and Exploration
(i) Rules for buying, supplies and stable or varying prices.
(ii) Determination of quantities and timing of purchases.
(iii) Bidding policies.
(iv) Strategies for exploration and exploitation of raw material sources.
(v) Replacement policies.
 6

3. Production Management

(i) Physical Distribution

(a) Location and size of warehouses distribution centres and retail outlets.
(b) Distribution policy.

(ii) Facilities Planning

(a) Numbers and location of factories, warehouses, hospitals etc.
(b) Loading and unloading facilities railroads and tracks determining the
transport schedule.

(iii) Manufacturing
(a) Production scheduling and sequencing.
(b) Stabilization of production and employment training, layoffs and
optimum product mix.

(iv) Maintenance and Project Scheduling

(a) Maintenance policies and preventive maintenance.
(b) Maintenance crew siges.
(c) Project scheduling and allocation of resources.

4. Marketing

(i) Product selection, timing, competitive actions.

(ii) Number of salesman, frequency of calling on accounts percent of time spent

on prospects.

(iii) Advertising media with respect to cost and time.

5. Personnel Management

(i) Selection of suitable personnel on minimum salary.

(ii) Mixes of age and skills.

(iii) Recruitment policies and assignment of jobs.

6. Research and Development

(i) Determination of the areas of concentration of research and development.

(ii) Project Selection

(iii) Determination of time cost trade-off and control of development projects.

(iv) Reliability and alternative design.

From all above areas of application we may conclude that O.R. has replaced
management by personality.
 7

1.7 Techniques Used in Operations Research

[B.C.A. (Rohilkhand) 2008; B.B.A. (Meerut) 2008]
The following techniques used in Operations Research.

1. Linear Programming: It is used in the solution of problems concerned with

assignment of personal, blending of materials, distribution and transportation and
investment properties.

2. Dynamic Programming: It is used in such area as planning advertising

expenditures distributing sales effort and production scheduling etc.

3. Queuing Theory: It is used in solving problem concern with traffic, servicing

machines subject to break down, air traffic scheduling, production scheduling,
hospital operations, determining optimum number of replacement for a group of
machines.

4. Inventory Theory: In determing when and how much a production or purchase.

5. Game Theory: The primary objective of game theory is to develop rational criteria
for selecting a strategy.

6. Simulations: The technique of simulation is an important tools of the designer in

stimulating airplane flight in a wind tunnel, simulating lines of communication with
an organisation chart. With the advent of the high speed digital computer with
which to conduct simulated experiments, this technique has become experimental
arm of researches.

1.8 The Importance of Operations Research in Decision

Theory
[B.B.A. (Meerut) 2008]
Operations Research uses the method of science to understand and explain the
phenomena of operating systems. It devises the theories (models) to explain these
phenomena. Uses these theories to describe what takes place under altered conditions,
and checks these predictions against new observations. Thus, operations research may be
regarded as a tool employed to increase the effectiveness of managerial decisions as an
objective supplement to the subjective feelings of the decision maker. O.R. may suggest
alternative courses of action when a problem is analysed and solution is attempted.
However, the study of complex problems by O.R. techniques becomes useful only when a
choice between two or more courses of action is possible. O.R. may be regarded as a tool
that enables the decision-maker to be objective in choosing an alternative from among
 8

many that he can conceive of. Following are the salient advantages of an operations
research study approach in decision-making:
1. Better Decisions: O.R. models frequently yield actions that do improve on
intuitive decision making. A situation may be so complex that the human mind can
never hope to assimilate all the significant factors without the aid of O.R. guided
computer analysis.
2. Better Coordination: Sometimes, operations research has been instrumental in
bringing order out of chaos. For instance, an O.R.-oriented planning model becomes
a vehicle for coordinating marketing decisions within the limitations imposed on
manufacturing capabilities.
3. Better Control: The management of large organizations recognize that it is
extremely costly to provide continuous executive to devote his attention to more
pressing matters. The most frequently adopted application in this category deal
with production scheduling and inventory replenishment.
4. Better Systems: Often an O.R. study is initiated to analyse a particular decision-
problem, such as whether to open a new warehouse. Afterwards, the approach is
further developed into a system to be employed repeatedly. Thus, the cost of
undertaking the first application may produce benefits.

P roblem S et
1. Define O.R. and discuss its scope.
[B.C.A. (Agra) 2010; B.B.A. (Meerut) 2006, 2008]

2. What is operations research? Give main characteristics of O.R. also discuss the
importance of O.R. in a decision making.
[B.C.A. (Kanpur) 2009; B.B.A. (Meerut) 2008]

3. What is O.R.? Describe briefly its applications.

[M.B.A. (Garhwal) 2008]

4. Discuss the significance and scope of operations research in modern management.

[B.B.A. (Meerut) 2002; B.B.A. (Delhi) 2007]

5. Discuss scientific method in O.R. [B.B.A. (Kurukshetra) 2007, 2008]

❍❍❍
 9

C HAPTER
2

Simplex Methods

2.1 Linear Programming Problems (Simplex Method)

2.1.1 Standard Form of a Linear Programming Problem

A linear programming problem is said to be in standard form if,
1. all the constraints are in the form of equations except for non-negativityrestrictions
which remain as inequalities ( ≥ 0);
2. the right hand side constant (element) of each constraint equation is non-negative;
3. all the decision variables have non-negative values, and
4. the objective function is in maximization form (or minimization form).
Remark: To write a L.P.P. in standard form is often called Reformulation of L.P.P.

2.1.2 Conversion of a L.P.P. into Standard Form

[B.C.A. (Meerut) 2006, 2008, 2012]
Step 1: Change the linear constraints of inequality type into equations:
This is done with the help of slack or surplus variables. These non-negative variables are
added to (or subtracted from) the left hand side of each such constraint. These new
variables are added if the constraint is (≤) and are called slack variables. On the other
hand, these new variables are subtracted if the constraint is (≥) and are called negative
slack variables or surplus variables.
 10

Thus, the non-negative variables which are introduced to represent the slack between the
left hand side and right hand side of each inequality (constraints) are known as slack
variables.

For example:
1. The linear constraint 3 x1 + 4 x2 ≤ 20 is converted to an equation with slack
variables S1 {or x3 } as

3 x1 + 4 x2 + S1 = 20, where S1 ≥ 0

{or 3 x1 + 4 x2 + x3 = 20, where x3 ≥ 0} ;

2. The linear constraint 3 x1 + 7 x2 ≥ 28 is converted to an equation with surplus

variable S2 {or x4 }

3 x1 + 7 x2 − S2 = 28, where S2 ≥ 0
{or 3 x1 + 7 x2 − x4 = 28, where x4 ≥ 0};

In the objective function the coefficients of slack and surplus variables are shown as
zero.
The objective function Z = 2 x1 + 3 x2 is written as

Z = 2 x1 + 3 x2 + 0 . S1 + 0 . S2

{or Z = 2 x1 + 3 x2 + 0 . x3 + 0 . x4 }
as their contribution is nil.

Hence, the cost of slack or surplus variable is taking zero in objective function.

Step 2: Make the right hand element of each constraint non-negative.

This is done by multiplying both sides of the resulting constraint by (−1).

For example: Consider the constraint 2 x1 − 3 x2 + 4 x3 = − 17

This constraint is changed (in standard form) to

− (2 x1 − 3 x2 + 4 x3 ) = (− 1) (− 17)

⇒ − 2 x1 + 3 x2 − 4 x3 = 17

Step 3: Make the unrestricted variables as non-negative :

This is done by replacing the unrestricted variable by a difference of two non-negative

variables. For example, if x2 is unrestricted in sign, it can be replaced by

( x2 ′ − x2 ′ ′ ), where x2 ′ ≥ 0, x2 ′ ′ ≥ 0
or by ( x3 − x4 ), where x3 ≥ 0, x4 ≥ 0
or by ( y1 − y2 ), where y1 ≥ 0, y2 ≥ 0
 11

Step 4: Convert the objective function in maximization form: [Although it is not

necessary]. This is done by changing the sign of the objective function.

For example: Minimize Z = c1 x1 + c2 x2 + … + cn x n is equivalent to the expression

Maximize (− Z ) = − c1 x2 − c2 x2 − … − cn x n

After going through the steps 1 to 4 a given L.P.P. in standard form as:
Optimize Z = c1 x1 + c2 x2 + … + cn x n + 0 . S1 + … + 0 . Sm

Subject to a11 x1 + a12 x2 + … + a1n x n ± S1 = b1

a21 x1 + a22 x2 + … + a2 n x n ± S2 = b2
: : : : :

ai1 x1 + ai2 x2 + … + ain x n ± Si = bi

: : : : :

am1 x1 + am2 x2 + … + amn x n ± Sm = bm

and x1, x2 , x3 , … , x n, S1, ... , Sm ≥ 0
where b1, b2 , … , bm ≥ 0

Since standard form involves only equations, we can write it in matrix form as follows:

Optimize Z = CX

Such that AX = b

and X ≥0

S olved E xamples
Example 1: Express the following linear programming problem in standard form:
Maximize Z = 3 x1 + 4 x2 + 7 x3
subject to 2 x1 + 3 x2 − 2 x3 ≤ 30
4 x1 − 2 x2 + x3 ≤ 22
x1 − 5 x2 − 6 x3 ≥ 4
x1 ≥ 0, x2 , x3 unrestricted.
[B.C.A. (Kanpur) 2004, 2007; B.C.A. (Agra) 2003]

Solution: Since x2 , x3 are unrestricted, we express x1, x2 and x3 as

x1 = y1 y1 ≥ 0

x2 = y2 − y3 , y2 , y3 ≥ 0

x3 = y4 − y5 , y4 , y5 ≥ 0
 12

Thus, the given L.P.P. is

Maximize Z = 3 y1 + 4 ( y2 − y3 ) + 7 ( y4 − y5 )

such that 2 y1 + 3 ( y2 − y3 ) − 2 ( y4 − y5 ) ≤ 30

4 y1 − 2 ( y2 − y3) + ( y4 − y5 ) ≤ 22

y1 − 5 ( y2 − y3 ) − 6 ( y4 − y5 ) ≥ 4

y1, y2 , y3 , y4 , y5 ≥ 0

or Maximize Z = 3 y1 + 4 y2 − 4 y3 + 7 y4 − 7 y5

such that 2 y1 + 3 y2 − 3 y3 − 2 y4 + 2 y5 ≤ 30

4 y1 − 2 y2 + 2 y3 + y4 − y5 ≤ 22

y1 − 5 y2 + 5 y3 − 6 y4 + 6 y5 ≥ 4
Introducing slack variables S1, S2 and surplus variables S3 , the required standard form is

Maximize Z = 3 y1 + 4 y2 − 4 y3 + 7 y4 − 7 y5

2 y1 + 3 y2 − 3 y3 − 2 y4 + 2 y5 + S1 = 30

4 y1 − 2 y2 + 2 y3 + y4 − y5 + S2 = 22

y1 − 5 y2 + 5 y3 − 6 y4 + 6 y5 − S3 = 4

and y1, y2 , y3 , y4 , y5 ≥ 0

Example 2: Reformulate the following L.P.P. into standard form:

Minimize Z = 2 x1 + 5 x2 + 4 x3

subject to the constraints − 2 x1 + 4 x2 ≤ 4

x1 + 2 x2 + x3 ≥ 5

2 x1 + 3 x3 ≤ 3

x1, x2 ≥ 0 and x3 unrestricted is sign.

[B.C.A. (Lucknow) 2009; B.C.A. (Agra) 2005, 2009]

Solution: Since x3 is unrestricted we can express x3 as x3 = x3 ′ − x3 ′ ′.

Then the given L.P.P. is

Minimize, Z = 2 x1 + 5 x2 + 4 x3 ′ − 4 x3 ′ ′

such that − 2 x1 + 4 x2 ≤ 4

x1 + 2 x2 + x3 ′ − x3 ′ ′ ≥ 5

2 x1 + 3 x3 ′ − 3 x3 ′ ′ ≤ 3

x1, x2 , x3 ′ , x3 ′ ′ ≥ 0
 13

Introducing the slack S1, S3 and surplus variable S2 , we have the required standard form.

Minimize Z = 2 x1 + 5 x2 + 4 x3 ′ − 4 x3 ′ ′ + 0 . S1 − 0 . S2 + 0 . S3

subject to the constraints

− 2 x1 + 4 x2 + S1 = 4

x1 + 2 x2 + x3 ′ − x3 ′ ′ − S2 = 5

2 x1 + 3 x3 ′ − 3 x3 ′ ′ + S3 = 3

With non-negative conditions:

x1 ≥ 0, x2 ≥ 0, x3 ′ ≥ 0, x3 ′ ′ ≥ 0, S1 ≥ 0, S2 ≥ 0, S3 ≥ 0.

However, if the objective function is converted into maximization, then we have

Maximize (− Z ) = Z * = − 2 x1 − 5 x2 − 4 x3 ′ + 4 x3 ′ ′ − 0 . S1 + 0 . S2 − 0 . S3

2.1.3 Some Important Definitions Regarding the Solution of L.P.P.

Using Standard Form/Canonical Form
1. The solution of an L.P.P. obtained from a canonical system by setting the
non-basic variables to zero and solving for the basic variables is called a Basic
Solution.
For a general linear programming problem with n decision variables and m
constraints, a basic solution may be found by setting (n − m) of the decision
variables equal to zero and solving for the remaining m variables. If a unique
solution exists it is a basic solution.

2. A basic solution is said to be basic feasible solution (BFS) if all basic variables are
non-negative.

3. A basic feasible solution is said to be non-degenerate basic feasible solution (N-d

BFS) if the value of all basic variables are positive.

4. A basic feasible solution is said to be degenerate basic feasible solution (DBFS) if

the values of one or more basic variables are zero.

5. A solution of L.P.P. is called a Feasible solution if it satisfies all the constraints

and the non-negativity restrictions.

6. A basic feasible solutions is said to be optimum if it optimizes (maximizes or

minimizes) the objective function.

7. If the value of the objective function Z can be increased or decreased indefinitely,

then such solutions are called unbounded solutions.
 14

2.1.4 Canonical Form of a General L.P.P.

A general L.P.P. can always be put in following form called the canonical form:
n
Maximize Z = ∑ cjx j
j =1

n
subject to ∑ aij x j ≤ bi, i = 1, 2 , … , m
j =1

and x j ≥ 0, for j = 1, 2 , … , n

The characteristics of canonical form are :

1. All decision variables are non-negative.

2. All constraints are of the (≤) type.

An inequality in directions (≥) can be changed to an inequality in the direction (≤)

by multiplying with sides of the inequality by −1.

3. Objective function is of maximization type.

The minimization of objective function Z is equivalent to the maximization of
(− Z ).

P roblem S et
1. What do you mean by a Linear Programming problem? Explain the procedure to
convert a L.P.P. into standard form. [B.C.A. (Purvanchal) 2005, 2007]

Express the following L.P.P. in standard form:

2. Minimize Z = 2 x1 + 3 x2
subject to 2 x1 − 3 x2 ≤ 5

− x1 + 3 x2 ≥ − 21

x1 ≥ 0, x2 is unrestricted.

3. Maximize Z = 2 x1 + x2 − 5 x3 + 3 x4

subject to 3 x1 + 2 x2 ≤ 15

4 x1 + 5 x2 ≥ 20

x1 + x2 − x3 + 2 x4 = 10

2 ≤ 2 x1 + 4 x2 − x3 ≤ 30

x i ≥ 0, i = 1, 2, 3, 4 . [B.C.A. (Meerut) 2007]

 15

A nswers
2. Minimize Z = 2 x1 + 3 ( x3 − x4 ) + 0 . S1 + 0 . S2
subject to 2 x1 − 3 ( x3 − x4 ) + S1 = 5
x1 − 3 ( x3 − x4 ) + S2 = 21 (using x1 − 3 x2 ≤ 21)
x1, x3 , x4 , S1, S2 ≥ 0

3. Maximize Z = 2 x1 + x2 − 5 x3 + 3 x4 + 0 . S1 + 0 . S2 + 0 . S3 + 0 . S4
subject to 3 x1 + 2 x2 + S1 = 15
4 x1 + 5 x2 − S2 = 20
x1 + x2 − x3 + 2 x4 = 10
2 x1 + 4 x2 − x3 + S3 = 30
2 x1 + 4 x2 − x3 − S4 = 2

x j ≥ 0, S j ≥ 0, j = 1, 2, 3, 4

2.2 Procedure of the Simplex Method

[B.C.A. (Meerut) 2006]
The steps taken in the simplex method are as follows:

Step 1: Formulate the L.P.P. if given otherwise.

Step 2: Convert the L.P.P. in maximization problem if it is a minimization problem.

Step 3: Express the L.P.P. in standard form and then in canonical form. We use slack or
surplus variables to convert the inequality to an equality.

Step 4: Start with an initial basic feasible solution by constructing a simplex table often
called Simplex tableau.

When slack or surplus variables cannot provide the initial basic feasible solution, we use
artificial variables.

Step 5: Check weather the initial basic feasible solution is optimum or not. If optimum
stop, otherwise go to step 6.

Step 6: Decide the non-basic variable to enter the basic and the basic variable to be
replaced by this non-basic variable using minimum ration rule.

Step 7: Obtain the current basic feasible solution. An adjacent basic feasible solution
differ from the present basic feasible solution only in exactly one basic variable.

Test it for optimality. If optimum stop, otherwise go to step 8.

 16

Step 8: Continue to find basic feasible solutions and test for optimality. When a
particular basic feasible solution is found to be optimal the simplex method terminates.

2.2.1 Brief Explanation of Simplex Tableau 1 and the Various

Factors Connected with its Construction
The form a11 x1 + a12 x2 + 1 . x3 + 0 . x4 = b1

a21 x1 + a22 x2 + 0 . x3 + 1 . x4 = b2

is often known as tableau form; its matrix contains two column form identity matrix.
This is written prior to setting up the initial simplex tableau or simplex tableau 1.

Simplex Tableau 1

CB Cj c1 c2 0 0 xB Ratio

Basis x1 x2 x3 x4

. . . a11 a12 1 0 b1 minimum

. . . a21 a22 0 1 b2 (key row)

Body Matrix Identity Matrix ...

X1 X2 X3 X4 ↓

Zj Z1 Z2 Z3 Z4

∆j = ∆1 ∆2 ∆3 ∆4 Z =…

Cj − Z j maximum

(key column)

Remark: After the initial simplex tableau or (including this) the elements of coefficient
matrix A are often denoted by yij which are identical to aij in the initial simplex tableau
and obtained by transformation in the consecutive simplex tableau. The elements of the
column X B are often denoted as X B .
i

1. C j Row: The C j row at the top of the simplex tableau is known as the objective row.

It represents the coefficients of the decision variables, slack variables, surplus

variables in the objective function. The row just below it represents the respective
variables of the objective function.
 17

2. CB Column: The CB column to the left of the simplex tableau is known as the
objective column. It represents the contribution per unit of the feasible variables,
i.e., coefficients of those variables in the objective function which form the identity
matrix.

3. Basis Column: The basis column represents the variables which are feasible for
solution (i.e. basic variables whose coefficients are written in the column CB . The
variables in this column are so arranged that the row headed by a variable and the
column headed by the same variables cross in the identity, i.e. where the 1 occurs.

4. Body Matrix: It is made up of the coefficients of the variable (which are not in the
identity matrix) on LHS of the constraints. The elements of body matrix can be
positive, negative and 0 as well.

5. Identity Matrix : It represents the matrix formed by the coefficients of feasible

variables in LHS of the constraints which has 1 at the diagonal and 0 at other
places.

6. X B Column: This column is also known as quantity column which gives the
values of the corresponding variables in basis column.

7. Vector Row: This row represents the vector corresponding to the variables in the
Body Matrix and Identity Matrix.

8. Index Row or Net Evaluation Row or ∆ j Row: This row represents the relative
profit coefficient ∆ j , ∆ j is obtained by a simplex formula :

∆ j = Cj − Z j

= C j − CB ′ X j‘, where CB ′ = transpose of CB

= C j − (inner product of CB and X j in the canonical system).

Note that ∆ j for basic variables are zero. Optimality is tested on the basic of ∆ j.

(i) If none of the columns in the index row shows a positive value (i.e., the values
of ∆ j are either 0 or negative) then the solution is optimal.
If none of ∆ j is positive, but any are zero, then other optimal solution exists
with the same value of Z.
If all of ∆ j are negative, the solution under consideration is unique optimal
solution.

(ii) If ∆ j > 0 for some j, the solution under consideration is not optimal, so we
proceed to next interation.

(iii) If corresponding to maximum positive ∆ j all elements in the column X j are

negative or zero, the solution under consideration will be unbounded.
 18

9. Key Column: The column showing maximum positive ∆ j is indicated as the key
column (or pivot column). The vector X j corresponding to this column enters into
the subsequent simplex tableau. This is marked with an upward arrow (↑ ).

10. Index Column: This column (the last column) is also known as ratio column. The
various ratios in this column are found out by dividing each quantity in quantity
column by their corresponding value in the key column.

11. Key Row: The row with the smallest (non-negative) ratio shown in the index
column is indicated as the key row (or pivot row). The vector corresponding to the
variable for which this ratio is minimum becomes out going vector and marked with
a downward arrow (↓).

12. Key Factor (or Key Element): The number that lies at the intersection of the key
column and the key row of a simplex tableau is known as key factor (or pivot
factor). This number is reduced to unity to obtain the row values of the main row
relating to entering vector of the next simplex tableau. The row of a subsequent
tableau that replaces the key row of its immediately proceeding table is known as
the main row.

Example 3: A firm produces two types of products A1 and A2 through two machines P1 and P2 . A1
needs 2 hours of P1 and 2 hours of P2 . A2 needs 3 hours of P1 and 1 hour of P2 . Machine P1 can run
at least for 24 hours per day. Machine P2 can run at least 16 hours per day. If profit from
A1 and A2 are ` 4 and ` 5 per unit respectively, ascertain by simplex method how many units of A1
and A2 should be produced to maximize the profit? [B.C.A. (Bhopal) 2006, 2011]

Solution: Step 1: (Formulation of L.P.P.):

Let x1 = number of units of Product A1

x2 = number of units of Product A2

Then we have the L.P.P.:

Maximize Z = 4 x1 + 5 x2

subject to 2 x1 + 3 x2 ≤ 24

2 x1 + x2 ≤ 16

and x1, x2 ≥ 0

Step 2: (Conversion of L.P.P. into Standard Form): Convert the constraint inequalities
into equations with the introduction of slack variables S1 and S2 as follows:

2 x1 + 3 x2 + S1 = 24
Constraints 
2 x1 + x2 + S2 = 16

2 x1 + 3 x2 + S1 + 0 . S2 = 24

2 x1 + x2 + 0 . S1 + S2 = 16
 19

Consequently, the objective function becomes

Z = 4 x1 + 5 x2 + 0 . S1 + 0 . S2

Step 3: (Construction of Sim plex Tab leau 1): The pre sen ta tion in this form is
of ten known as tab leau form, its matrix contains two columns to form identity matrix.

2 3 1 0 
 
2 1 0 1 

Simplex Tableau 1

CB Cj 4 5 0 0 xB Ratio

Basis x1 x2 S1 S2

0 S1 2 3 1 0 24 24
= 8 (min) key row
3

0 S2 2 1 0 1 16 16
= 16
1

................................................

X1 X2 ↑ S1 ↓ S2

Zj 0 0 0 0

∆j 4 5 0 0 Z =0
(maximum
= Cj − Z j key column)

Initial basic feasible solution is x1 = 0, x2 = 0, S1 = 24, S2 = 16

24 
Z = CB ′ X B = [0 0]   = 0 × 24 + 0 × 16 = 0
16 

Step 4: (Test for Optimal Solution):

2 
Z1 = CB ′ X1 = [0 0]   = 0
2 

3 
Z2 = CB ′ X2 = [0 0]   = 0
1

1
Z3 = CB ′ S1 = [0 0]   = 0
0 
 20

0 
Z4 = CB ′ S2 = [0 0]   = 0
1

∆ j = C j − Z j = C j − CB ′ X j

= C j − [0 0] X j

= C j for all j

Since there are some positive values in the ∆ j row, the initial basic feasible solution is not
optimal.

Step 5: (To Find Incoming and Outgoing Vector): ∆ j is maximum for x2 . Hence, X2 is

the entering vector.

XB
The ratio is minimum for S1. Hence, S1 is the outgoing
Corresponding element in X 2

vector.

Step 6: (Construction of Tableau 2): The element 3 in X2 is the pivot number or key
element; hence it is to be reduced to 1 and the other element 1 in X2 is to be reduced to
zero.

The elements of the key row will become

x1 x2 S1 S2 XB

2 3 1 0 24
3 3 3 3 3

i.e. R1 2 1 1 0 8
3 3

The element of the other row will be

x1 x2 S1 S2 XB

R2 = R2 − R1 2 1 −1 1 1−0 16 − 8
2− 0−
3 3

i.e. 4 0 1 1 8
−
3 3
 21

Simplex Tableau 2

CB Cj 4 5 0 0 XB Ratio

Basis x1 x2 S1 S2

5 x2 2 1 1 0 8 2
8÷ = 12
3 3 3

0 S2 4 0 1 1 8 4
− 8÷ =6
3 3 3
(min) key
row

X1 ↑ X2 S1 S2 ↓

Zj 10 5 5 0
3 3

∆j 2 0 5 0 Z = 40
−
3 3
(maximum
= Cj − Z j
key column)

The current basic feasible solution is x1 = 0, x2 = 8, S1 = 0, S2 = 8

8
Z = CB ′ X B = [5 0]   = 40
8

Step 7: (Test for Optimal Solution):

∆ j = C j − Z j = C j − CB ′ X j

2 
  10 2
∆1 = C1 − CB′ X1 = 4 − [5 0] 3  = 4 − =
4 3 3
 
3 

 1
  5 5
∆3 = C3 − CB′ X3 = 0 − [5 0]  3  = 0 − = −
1 3 3
− 
 3

Since ∆1 is positive, the current feasible solution is not optimal.

Step 8: (To Find Entering and Outgoing Vectors): Since ∆1 is maximum, the entering
vector is X1.

Since the ratio is minimum for S2 , the outgoing vector is S2 .

 22

4
Step 9: (Construction of Tableau 3): The element in vector X1 is the pivot number.
3
Hence, it is to be reduced to zero.

The elements of the key row will become

x1 x2 S1 S2 XB

R2 4 4 4 1 4 4 4
÷ 0÷ − ÷ 1÷ 8÷
3 3 3 3 3 3 3

R2 i.e. 1 0 1 3 6
−
4 4

The elements of the other row will be

x1 x2 S1 S2 XB

2 2 2 2 1  1 2 3 2 2
R1 = R1 − R2 , −1 × 1−0 × − − ×  0− × 8−6 ×
3 3 3 3 3  4 3 4 3 3

i.e. 0 1 1 1 4
−
2 2

Simplex Tableau 3

CB Cj 4 5 0 0 XB Ratio

Basis x1 x2 S1 S2

5 x2 0 1 1 1 4 No
−
2 2 need

4 x1 1 0 1 3 6
−
4 4

-------------------------------------------------

X1 X2 S1 S2

Zj 4 5 3 1
2 2

S1 = 0 ∆ j = Cj − Z j 0 0 3 1 Z = 44
− −
2 2
S2 = 0

The current basic feasible solution is x1 = 6, x2 = 4, S1 = 0, S2 = 0

4 
Z = CB ′ X B = [5 4]   = 20 + 24 = 44
6 
 23

Step 10: (Test for Optical Solution):

∆ j = C j − Z j = C j − CB ′ X j

 1 
 2  = 0 −  5 − 1 = − 3
∆3 = 0 − [5 4] 
−1  2  2
 
 4 

 −1 
   5  1
∆4 = 0 − [5 4]  2  = 0 − − + 3  = −
3  2  2
 
 4 

Since all ∆ j ≤ 0, hence the current basic feasible solution is optimal with x1 = 6, x2 = 4 and
Z max = ` 44.

Example 4: Solve the following linear programming problem by simplex method

Maximize Z = 3 x1 + 2 x2

subject to x1 + x2 ≤ 4

x1 − x2 ≤ 2

and x1, x2 ≥ 0. [B.C.A. (Meerut) 2011; B.C.A. (Agra) 2002, 2005]

Solution: Step 1: (Write L.P.P. in Standard Form): For this all the right hand side
constants should be positive. The inequality constraints are converted to equations by
introducing the non-negative slack or surplus variables. The coefficients of slack and
surplus variables are taken zero in the objective function. The given L.P.P. in standard
form is
Maximize Z = 3 x1 + 2 x2 + 0 . x3 + 0 . x4

subject to x1 + x2 + x3 = 4 ...(1)

x1 − x2 + x4 = 2 ...(2)
and x1, x2 , x3 , x4 ≥ 0

The tableau form of the L.P.P. is

x1 + x2 + x3 + 0 . x4 = 4

x1 − x2 + 0 . x3 + x4 = 2
which forms identity matrix
1 0
 
0 1

Step 2: The system of equations (1) and (2) is in canonical form, x3 is basic variable in
equation (1) and x4 is basic variable in equation (2).
 24

Step 3: Let C j = coefficient of x j in the objective function.

CB = column vector of the coefficients of basic variables in the objective function.

X j = column vector of the coefficients of x j in the linear constraints.

X B = Basic feasible solution.

0 
For initial basic feasible solution, CB =  
0 

and X B = vector of the right hand side constants of the linear constraints = B.
Using the method of detached coefficients the Simplex Tableau 1 is as follows:

Simplex Tableau 1

CB Cj 3 2 0 0 XB Ratio (to be computed

Basis x1 x2 x3 x4 in step 5)

0 x3 1 1 1 0 4 4
=4
1

0 x4 1 −1 0 1 2 2
= 2 (minimum key row)
1

↑ X1 X2 X3 ↓ X4

Zj 0 0 0 0

∆ j = Cj − Z j 3 2 0 0 Z = CB ′ X B = 0

To be computed (maximum
in step 4 key row)

From the above table the initial basic feasible solution is

x1 = 0, x2 = 0, x3 = 4, x4 = 2
The value of objective function is given by
Z =3 ×0 + 2 ×0 + 0 ×4 + 0 ×2 =0
or Z = The inner product of the vectors CB and X B
4 
= CB ′ X B = [0 0]   = 0 × 4 + 0 × 2 = 0
2 

Step 4: (Test for Optimal Solution): To check if the initial basic feasible solution is
optimal, calculate the relative profit coefficient ∆ j by a simple formula known as the
inner product rule.
 25

Relative profit coefficient of the variable x j, denoted by ∆ j is given by

C j − Z j = C j − (inner product of CB and X j in the canonical system)

= C j − CB ′ X j

The relative profit coefficient ∆ j corresponding to the column of basic variables is always
zero. So we calculate for non-basic variables. Thus

1
∆1 = Ci − CB ′ X1 = 3 − [0, 0]   = 3 − 0 = 3
1

 1
∆2 = C2 − CB ′ X2 = 2 − [0, 0]   = 2,
−1

similarly ∆3 = 0, ∆4 = 0.

The initial B.F. solution is optimal (in case of maximization problem) if the relative
profit coefficients of its non-basic variables are all negative or zero.

Since, there are some positive values in the ∆ j row, the initial basic feasible solution is not
optimal.

Step 5: (Determine the Incoming Vector and Outgoing Vector): The non-basic
variable x1 gives the greatest value of ∆ j hence it will be chosen as the new basic variable
to enter in the basis (i.e. vector X1 is incoming vector).

In order to decide with basic variable is going to be replaced (i.e. which X j. In

mathematical form, we calculate the ratio

 Element of X B 
 
Corresponding element of the Incoming Vector X j

For each constraint row (in the last column of tableau) as follows :

Row Number Basic Variable Ratio (over X1)

1 x3 4
=4
1

2 x4 2
= 2 (minimum)
1

Since the minimum ratio is for x4 , so x4 is the outgoing vector (i.e. second row is pivot
row).

The element 1 denoted as [1] is the key number or pivot number.

 26

Step 6: (Construction of Tableau 2): To have the vector X1 in the basis, apply the
0 
operation R1 → R1 − R2 in tableau1 to make the vector X1 as   to construct the tableau 2.
1
The key row in Tableau 1 becomes.

x1 x2 x3 x4 XB

1 −1 0 1 2

in tableau 2.

The other row in tableau 2 is

x1 x2 x3 x4 XB

1 −1 1 − (− 1) 1−0 0 −1 4 −2

i.e. 0 2 1 −1 2

Now, the system of equations in row 1 and row 2 is in canonical form, where x3 is basic
variable in row 1 and x1 is basic variable in row 2.

Simplex Tableau 2

CB CJ 3 2 0 0 XB Ratio (to be
computed in
Basis x1 x2 x3 x4
step 8)

0 x3 0 2 1 −1 2 2
= 1 (min.
2
key row)

3 x1 1 −1 0 1 2 2
neglect
−1

X1 X2 ↑ X3 ↓ X4

Zj 3 −3 0 3

0 5 0 −3
(maximum
∆ j To be computed Z = CB ′ X B = 6
key column)
in step 7
 27

Here, the current basic feasible solution is

0 . x1 + 1 . x3 = 2 ⇒ x3 = 2
x2 = 0, x4 = 0 and 
1 . x1 + 0 . x3 = 2 ⇒ x1 = 2
(non-basic variable)

The value of the objective function is given by the inner product of the vectors CB and
XB :
2 
CB ′ X B = [0 3]   = 0 + 6 = 6
2 
[or Z = 3 × 2 + 2 × 0 + 0 × 2 + 0 × 0 = 6]

Step 7: (Test for Optimal Solution):

∆1 = 0
 2
∆2 = C2 − Z2 = 2 − [0 3]   = 2 + 3 = 5
− 1
∆3 = 0
− 1
∆4 = C4 − Z4 = 0 − [0 3]   = 0 − 3 = − 3
 1

Since there is one positive value of ∆ j the current basic feasible solution is not optimal.

Step 8: (Determine Incoming Vector and Outgoing Vector): The non-basic variable
x2 gives the greatest value of ∆ j hence it will be chosen as the new basic variable to enter
the basis (i.e. X2 is the incoming vector).

In order to decide which basic variables is going to be replaced (i.e. which X j, j = 1, 4 is

outgoing vector) we use minimum ratio rule. In mathematical form we calculate the ratio:
 Element of X B 
 
Corresponding element of the Incoming Vector X j 

For each constraint row as follows:

Row Number Basic Variable Ratio (over X2 )

1 x3 2
= 1 (minimum)
2

2 x1 2
= −2
−1
(negative value is never considered)

The ratio is minimum for x3 so the vector X3 is the outgoing vector. The element 2
denoted as 2 is the key number of pivot number.
 28

Step 9: (Construction of Tableau 3): To have the vector X2 in the basis, divide first
1
row by 2 in tableau 2 and apply the operation R2 = R2 + R1 to make the vector X2 =  
0 
The key row in tableau 2 becomes

x1 x2 x3 x4 XB

0÷2 2÷2 1÷ 2 − 1÷ 2 2÷2

i.e. 0 1 1 1 1
−
2 2

in tableau 3.

The other row in tableau 3 will be

x1 x2 x3 x4 XB

1+ 0 −1 + 1 1 1 2 +1
0+ 1−
2 2

i.e. 1 0 1/2 1/2 3

Simplex Tableau 3

CB Cj 3 2 0 0 XB Ratio

Basis x1 x2 x3 x4

2 x2 0 1 1/2 –1/2 1 —

3 x1 1 0 1/2 1/2 3 —

X1 X2 X3 X4

Zj 3 2 5/2 1/2

∆ j = Cj − Z j 0 0 –5/2 –1/2 Z = CB ′ X B = 11
(to be computed in step 10)

Form this table the current basic feasible solution is x3 = 0 = x4 and

0 . x1 + 1 . x2 = 1 ⇒ x2 = 1.

(non-basic variable) 1 . x1 + 0 . x3 = 3 ⇒ x1 = 3

The values of objective function is given by the inner product of the vectors CB and X B .
1
CB ′ X B = [2 3]   = 2 + 9 = 11
3 
[or Z = 3 × 3 + 2 × 1 + 0 × 0 × 0 + 0 = 9 + 2 = 11]
 29

Step 10: (Test for Optimal Solution):

∆1 = 0

∆2 = 0
1 / 2  5 5
∆3 = C3 − Z3 = 0 − [2 3]   =0 − = −
1 / 2  2 2

 −1 / 2  1 1
∆4 = C4 − Z4 = 0 − [2 3]   =0 − = −
 1 / 2  2 2

Since there is no positive value in the ∆ j row, the current basic feasible solution is optimal.

Thus, the optimal solution is x1 = 3, x2 = 1

Remark 1: The computation in simplex method may be shown in a single table as given
below (without giving an expression in words):

CB Cj 3 2 0 0 XB Ratio

Basis x1 x2 x3 x4

0 x3 1 1 1 0 4 4
=4
1

0 x4 1 −1 0 1 2 2
= 2 (min.)
1

X1 ↑ X2 X3 ↓ X4

∆ j = Cj − Z j 3 2 0 0 x1 = 0 Z = CB ′ X B = 0
= C j − CB ′ X j (max.) x2 = 0

0 x3 0 2 1 −1 2 2
= 1 (min)
2

3 x1 1 −1 1 1 2 2
= −1
−1
(neglected)

∆j 0 5↑ 0↓ −3 x3 = 0 Z = CB ′ X B = 6
(max.) x4 = 0

2 x2 0 1 1 1 1 No need
−
2 2

3 x1 1 0 1 1 3
2 2

∆j 0 0 5 1 x3 = 0 Z = CB ′ X B = 11
− −
2 2
x4 = 0
 30

Thus, optimum solution: x1 = 3, x2 = 1, with maximum Z = 11.

Remark 2: Some authors use ∆ j = Z j − C j = CB ′ X j − C j

To check whether the basic feasible solution is optimal (maximum), we proceed as

follows:

1. If all ∆ j ≥ 0, the basic feasible solution is optimal.

2. If at least one ∆ j is negative, the basic feasible solution is not optimal then proceed

the next step.

3. If corresponding to most negative ∆ j, all elements of the column X j are negative or

zero (≤ 0), then basic feasible solution will be unbounded.

4. To determine incoming vector use the minimum ∆ j.

5. To find outgoing vector use the minimum ratio.

Remark 3: However, some author use

∆ j = C j − Z j = C j − CB ′ X j

So, the students are advised to be careful while going through various notations.

Example 5: Maximize Z = 3 x1 + 2 x2

subject to 2 x1 + x2 ≤ 5

x2 + x2 ≤ 3

and x1, x2 ≥ 0.
[B.C.A. (Rohilkhand) 2006, 2010, 2012]

Solution: Introducing x3 , x4 as slack variables, we have

Maximize Z = 2 x1 + 2 x2 + 0 . x3 + 0 . x4

subject to 2 x1 + x2 + x3 + 0 . x4 = 5

x1 + x2 + 0. x3 + x4 = 3

and x1, x2 , x3 , x4 ≥ 0.

With usual notation, the computation in a single table without any explanation is as
follows:
 31

CB Cj 3 2 0 0 XB Ratio

Basis x1 x2 x3 x4

0 x3 2 1 1 0 5 5
(min)
2

0 x4 1 1 0 1 3 3

↑ X1 X2 X3 ↓ X4

Zj 0 0 0 0

∆ j = Z j − Cj −3 −2 0 0 Z = CB ′ X B = 0

(most negative)

3 x1 1 1/2 1/2 0 5/2 1

5/2 ÷ =5
2

0 x4 0 1/2 −1/2 1 1/2 1 1

÷ = 1 (Min.)
2 2

X1 ↑ X2 X3 ↓ X4

Zj 3 3/2 3/2 0

∆ j = Z j − Cj 0 −1/2 3/2 0 15
Z = CB ′ X B =
2

(most negative)

3 x1 1 0 1 −1 2

2 x2 0 1 −1 2 1

X1 X2 X3 X4 No need

Zj 3 2 1 1

Dj = Z j − Cj 0 0 1 1 Z = CB ′ X B = 8

the required optimal solution is x1 = 2 x2 = 1, Z max = 8.

 32

Example 6: Solve the following problem by simplex method:

Max (Z ) = 3 x1 + 5 x2 + 4 x3
subject to 2 x1 + 3 x2 ≤ 8
0 x1 + 2 x2 + 5 x3 ≤ 10
3 x1 + 2 x2 + 4 x3 ≤ 15
and x1, x2 , x3 ≥ 0 . [B.C.A. (Meerut) 2004, 2009, 2012]

Solution: L.P.P. in standard form.

Max (Z ) = 3 x1 + 5 x2 + 4 x3 + 0 x4 + 0 x5 + 0 x6
subject to 2 x1 + 3 x2 + 0 x3 + x4 + 0 x5 + 0 x6 = 8

0 x1 + 2 x2 + 5 x3 + 0 x4 + x5 + 0 x6 = 10

3 x1 + 2 x2 + 4 x3 + 0 x4 + 0 x5 + x6 = 15

x1, x2 , x3 , x4 , x5 , x6 ≥ 0

Simplex Tableau 1

CB Cj 3 5 4 0 0 0 XB Ratio

Basis x1 x2 x3 x4 x5 x6

0 x4 2 3 0 1 0 0 8 8
= 2 .6 (min.)
3

0 x5 0 2 5 0 1 0 10 10
=5
2

0 x6 3 2 4 0 0 1 15 15
= 75
.
2

X1 X2 ↑ X3 X4 ↓ X5 X6 x1 = 0

∆ j = Z j− C j −3 −5 −4 0 0 0 x2 = 0 Z = CB ′X B = 0
x3 = 0
(min.)

∆1 = CB ′ X1 − C1 = 0 − 3 = − 3

∆2 = CB ′ X2 − C2 = 0 − 5 = − 5

∆3 = CB ′ X3 − C3 = 0 − 4 = − 4

Since all the ∆ j are not ≥ 0, the initial basic feasible solution ( x1 = 0, x2 = 0, x3 = 0,
x4 = 8, x5 = 10, x6 = 15) is not optimal.

Since ∆2 is minimum, hence X2 is the entering or incoming vector.

 33

XB
Since the ratio is minimum for x4 , the departing or outgoing vector is X4 .
X2
1
 
To convert the vector X2 as 0  , we apply operations
0 
 
1
R1 → R1, R2 → R2 − 2 R1, R3 → R3 − 2 R1 successively and have.
3

Simplex Tableau 2

CB Cj 3 5 4 0 0 0 XB Ratio

Basis x1 x2 x3 x4 x5 x6

5 x2 2 1 0 1 0 0 8/3 8 1
× =∞
3 3 3 0

0 x5 4 0 5 2 1 0 14/3 14 1 14
− − × =
3 3 3 5 15
(min.)

0 x6 5 0 4 2 0 1 29/3 29 1 29
− × =
3 3 3 4 12

X1 X2 X3 ↑ X4 X5 ↓ X6 x1 = 0

∆ j = Z j − Cj 1/3 0 −4 5/3 0 0 x2 = 0 Z = CB ′X B
x3 = 0 40
=
3
10 1
∆ j = CB ′ X1 − C1 = −3 =
3 3
∆3 = CB ′ X3 − C3 = 0 − 4 = − 4
5 5
∆4 = CB ′ X4 − C4 = − 0 =
3 3
8
Since ∆3 is negative, the current basic feasible solution x1 = 0, x2 = , x3 = 0, x4 = 0,
3
14 29
x5 = x6 = is not optimal.
3 3
Since ∆3 is the minimum net evaluation, hence X3 is the entering or incoming vector.
x
Since the ratio B is minimum for x5 , the departing or outgoing vector is x5 .
X3
0 
 
To convert the vector X3 as 1 , we apply operations
0 
 
1
R2 → R2 R1 → R1, R3 → R3 − 4 R2 successively and have.
5
 34

Simplex Tableau 3

CB Cj 3 5 4 0 0 0 XB Ratio

Basis x1 x2 x3 x4 x5 x6

5 x2 2 1 0 1 0 0 8/3 8 3
× =4
3 3 3 2

4 x3 4 0 1 2 1 0 14/15 negative
− −
15 15 5

0 x6 41 0 0 2 4 1 89/15 89 15 89
− − × =
15 15 5 15 41 41

X1 ↑ X2 X3 X4 X5 X6 ↓ x1 = 0 (min.)

∆ j = Z j − Cj –11/15 0 0 17/15 4/5 0 x2 = 0 Z = CB ′

(min.) 256
XB =
15

x3 = 0

10 16  24 11
∆1 = CB ′ X1 − C1 =  −  −3 = −3 = −
 3 15  15 15
5 8 17 17
∆4 = CB ′ X4 − C4 =  −  − 0 = −0 =
 3 15  15 15
4 4
∆5 = CB ′ X5 − C5 = −0 =
5 5

Since ∆1 is negative, the current basic feasible solution

 8 14 89 
 x1 = 0, x4 = 0, x5 = 0, x2 = 3 , x3 = 15 , x6 = 15 

is not optimal since ∆1 is the minimum net evaluation, hence X1 is the entering or
incoming vector.
XB
Since the ratio is minimum for X6 , the departing or outgoing vector is X6 .
X1
0 
 
To convert the vector X1 as 0  we apply operations
1
 
15 2 4
R3 → R1 → R1 − R3 , R2 → R2 + R3
41 3 15
successively and have.
 35

Simplex Tableau 4

CB Cj 3 5 4 0 0 0 XB Ratio

Basis x1 x2 x3 x4 x5 x6

5 x2 0 1 0 15 8 10 50 No need
−
41 41 41 41

4 x3 0 0 1 6 5 4 62
−
41 41 41 41

3 x1 1 0 0 2 12 15 89
− −
41 41 41 41

∆ j = Z j − Cj X1 X2 X3 X4 X5 X6 x4 = 0 Z = CB ′ X B

0 0 0 45 24 11 x5 = 0 765
=
41 41 41 41

x6 = 0

 15 
 41
 6 75 24 60 45
∆4 = CB ′ X4 − C4 = [543] −  − 0 = − − =
 41 41 41 41 41
 −2 
 41 

 8
 41
 5 40 20 30 24
∆5 = CB ′ X5 − C5 = [543]   −0 = + − =
 41  41 41 41 41
− 12 
 41

 10 
− 41
 4 50 16 45 11
∆6 = CB ′ X6 − C6 = [543]   −0 = − + + =
 41  41 41 41 41
 15 
 41

Since all ∆ j are ≥ 0, the current basic feasible solution

 89 50 62 
 x1 = 41 , x2 = 41 , x3 = 41 , x4 = 0, x5 = 0, x6 = 0 

is optimal with maximum Z = CB ′ X B

 36

50 
 41 
62  250 248 267 765
= [5 4 3]   = + + =
 41  41 41 41 41
84 
 41 

Example 7: Solve the following linear programming problem by simplex method.

Minimize Z = x1 − 3 x2 + 2 x3
subject to 3 x1 − x2 + 3 x3 ≤ 7
− 2 x1 + 4 x2 ≤ 12
− 4 x1 + 3 x2 + 8 x3 ≤ 10
and x1, x2 , x3 ≥ 0. [B.C.A. (Meerut) 2002, 2009]

Solution: Convert the objective function into maximization form:

Maximize Z * = − x1 + 3 x2 − 2 x3 , where Z * = − Z

The L.P.P. in standard form (adding slack variables x4 , x5 and x6 ).

Maximize Z * = − x1 + 3 x2 − 2 x3 + 0 . x4 + 0 . x5 + 0 . x6

subject to 3 x1 − x2 + 3 x3 + x4 + 0 . x5 + 0 . x6 = 7

− 2 x1 + 4 x2 + 0 . x3 + 0 . x4 + x5 + 0 . x6 = 12

− 4 x1 + 3 x2 + 8 x3 + 0 . x4 + 0 . x5 + x6 = 10

x1, x2 , x3 , x4 , x5 , x6 ≥ 0
with usual notation, we have
Simplex Tableau 1

CB Cj −1 3 −2 0 0 0 XB Ratio

Basis x1 x2 x3 x4 x5 x6

0 x4 3 −1 3 1 0 0 7 7
= − 7 (neg.)
−1
0 x5 −2 4 0 0 1 0 12 12
= 3 (min.)
4
0 x6 −4 3 8 0 0 1 10 10
= 3.3
3
X1 ↑ X2 X3 X4 ↓ X5 X6 x1 = 0

∆ j = Z j − Cj −1 3 −2 0 0 0 x2 = 0 Z* = 0, Z = 0
(max.)
x3 = 0
 37

Since there are some positive values of ∆ j the initial basic feasible solution
( x1 = 0, x2 = 0, x3 = 0, x4 = 7, x5 = 12, x6 = 10) is not optimal.

Since C2 is maximum so, X2 is the incoming vector

Since ratio is minimum in second row so X5 is the outgoing vector.

0 
  1
To make the vector X2 as 1 apply the operations R2 → R2 , R1 → R1 + R2 ,
0  4
 
R3 → R3 − 3 R2 successively, then we have

Simplex Tableau 2

CB Cj −1 3 −2 0 0 0 XB Ratio

Basis x1 x2 x3 x4 x5 x6

0 x4 5/2 0 3 1 1/4 0 10 5
10 ÷ =4
2
(minimum)

3 x2 − 1/2 1 0 0 1/4 0 3  1
3 ÷ −  = − 6
 2
(negative)

0 x6 −5/2 0 8 0 −3/4 1 1  5 5
1÷  −  = −
 2 2
(negative)

∆ j = Cj − Z j X1 ↑ X2 X3 ↓ X4 X5 X6 x1 = 0

1/2 0 −2 0 −3/4 0 x3 = 0 Z * = 9, Z = − 9
(max.) x5 = 0

1
Since ∆1 = , the current basic feasible solution
2
[ x1 = 0, x3 = 0, x5 = 0, x4 = 10,x2 = 3, x6 = 1] is not optimum.

Since ∆1 is maximum, so X1 is the incoming vector, since ratio is minimum in first row so
X4 is the outgoing vector.
1
 
To make the vector X1 as 0  apply the operations
0 
 
2 1 5
R1 → R1, R2 → R2 + R1, R3 → R3 + R1 successively; then we have
5 2 2
 38

Simplex Tableau 3

CB Cj −1 3 −2 0 0 0 XB Ratio

Basis x1 x2 x3 x4 x5 x6

–1 x1 1 0 6/5 2/5 1/10 0 4 No Need

3 x2 0 1 3/5 1/5 3/10 0 5

0 x6 0 0 11 1 −1/2 1 11

X1 X2 X3 X4 X5 X6 x3 = 0

∆ j = Cj − Z j 0 0 −
13
−
1
−
8 0 x4 = 0 Z * = 11,
5 5 10
x5 = 0 Z = − 11

6 / 5 
   6 9   3 13
∆3 = − 2 − [− 1 3 0] 3 / 5  = − 2 −  − + + 0 = − 2 −   = −
 5 5   5 5
 11 
 

2 / 5 
   2 3  1
∆4 = 0 − [− 1 3 0] 1 / 5  = 0 −  − + + 0 = −
 5 5  5
 1 
 

1 / 10 
   1 9  8
∆5 = 0 − [− 1 3 0] 3 / 10  = 0 −  − + + 0 = −
 10 10  10
 −1 / 2 
 

Since all the ∆j are ≤ 0, hence the current basic feasible solution
( x1 = 4, x2 = 5, x6 = 11, x3 = 0, x5 = 0, x4 = 0) is optimal. Thus, x1 = 4, x2 = 5 and x3 = 0
with minimum Z = − 11.
 39

P roblem S et
Solve the following problems using simplex method:

1. Maximize Z = 5 x1 + 3 x2

subject to 3 x1 + 5 x2 ≤ 15

5 x1 + 2 x2 ≤ 10

and x1, x2 ≥ 0 [B.C.A. (Rohilkhand) 2003]

2. Maximize Z = 3 x1 + 2 x2 + 5 x3 + 1000

subject to x1 + 2 x2 + x3 ≤ 430
3 x1 + 2 x2 ≤ 460

x1 + 4 x2 ≤ 420

x1, x2 , x3 ≥ 0

3. Maximize Z = 3 x1 + 5 x2 + 4 x3

subject to 2 x1 − x2 ≤ 8

2 x2 + 5 x3 ≤ 10

3 x1 + 2 x2 + 4 x3 ≤ 15

x1 ≥ 2, x2 ≥ 4, x3 ≥ 0 [B.C.A. (Lucknow) 2010]

4. Maximize Z = 2 x1 + 5 x2 + 7 x3

subject to 3 x1 + 2 x2 + 4 x3 ≤ 100

x1 + 4 x2 + 2 x3 ≤ 100

x1 + x2 + 3 x3 ≤ 100

x1, x2 , x3 ≥ 0

5. Maximize Z = 4 x1 + x2 + 3 x3 + 5 x4

subject to 4 x1 − 6 x2 − 5 x3 + 4 x4 ≥ − 20

3 x1 − 2 x2 + 4 x3 + x4 ≤ 10

8 x1 − 3 x2 + 3 x3 + 2 x4 ≤ 20

and x1, x2 , x3 , x4 ≥ 0
 40

6. Minimize Z = x2 − 3 x3 + 2 x5

subject to 3 x2 − x3 + 2 x5 ≤ 7

− 2 x2 + 4 x3 ≤ 12

− 4 x2 + 3 x3 + 8 x5 ≤ 10
and x2 , x3 , x5 ≥ 0 [B.C.A. (Kanpur) 2009]

7. Maximize Z = 2 x + 5 y

subject to x + y ≤ 600
0 ≤ x ≤ 400
0 ≤ y ≤ 30 [B.C.A. (Purvanchal) 2009; B.C.A. (Kanpur) 2008]

8. Min (Z ) = x1 − 3 x2 + 2 x3

subject to 3 x1 − x2 + 2 x3 ≤ 7

− 2 x1 + 4 x2 ≤ 12

− 4 x1 + 3 x2 + 4 x3 ≤ 10

x1, x2 , x3 ≥ 0

9. Min (Z ) = 2 x1 + 4 x2 + x3 + x4

subject to x1 + 3 x2 + x4 ≤ 4

2 x1 + x2 ≤ 3

x2 + 4 x3 + x4 ≤ 3

x1, x2 , x3 , x4 ≥ 0 [B.C.A. (Aligarh) 2008, 2012]

10. Max (Z ) = 5 x1 + 3 x2

subject to 3 x1 + 5 x2 ≤ 15

5 x1 + 2 x2 ≤ 10

x1, x2 ≥ 0 [B.C.A. (Avadh) 2011, B.C.A. (Kanpur) 2007]

11. Max (Z ) = 3 x1 + 5 x2

subject to 3 x1 + 2 x2 ≤ 18

x1 ≤ 4

x2 ≤ 6

x1, x2 ≥ 0 [B.C.A. (Kashi) 2004, 2012]

 41

12. Max (Z ) = 4 x1 + 10 x2

subject to 2 x1 + x2 ≤ 50

2 x1 + 5 x2 ≤ 100

2 x1 + 3 x2 ≤ 90

x1, x2 ≥ 0 [B.C.A. (Agra) 2003, 2007]

13. Minimize (Z ) = x1 + x2 + 3 x3

subject to 3 x1 + 2 x2 + x3 ≤ 3

2 x1 + x2 + 2 x3 ≤ 2

x1, x2 , x3 ≥ 0

14. Write short notes on the simplex method. [B.C.A. (Meerut) 2007, 2011]

15. Give the brief explanation of simplex method. [B.C.A. (Bhopal) 2008, 2010, 2012]

16. How do the graphical and simplex methods of solving a L.P.P.?

[B.C.A. (Meerut) 2008, 2009, 2011]

17. Define the following :

(i) Optimal solution

(ii) Surplus variable

(iii) Basic feasible solution [B.C.A. (Meerut) 2008, 2011]

(iv) Slack and surplus variable [B.C.A. (Meerut) 2004, 2006]

18. Write notes on the following terms as used in a L.P.P.:

(i) Objective function

(ii) Surplus variables

(iii) Basic feasible solution [B.C.A. (Meerut) 2006, 2010]

19. Give an account of linear programming problem with its important components.
What does the non-negativity restriction mean?

[B.C.A. (Kanpur) 2006; B.C.A. (Meerut) 2003, 2005, 2010]

 42

A nswers
1. 20 45 235
x1 = , x2 = , Z max =
19 19 19

2. x1 = 0, x2 = 100, x3 = 230, Z max = 1350 + 1000 = 2350

89 50 62 765
3. x1 = , x2 = , x3 = , Max (Z ) =
41 41 41 41

50 50
4. x1 = 0, x2 = , x3 = , Z max = 200
3 3

5. The solution is unbound.

6. x2 = 4, x3 = 5, x5 = 0, Z min = − 11

7. x = 300, y = 300, Z max = 21

8. x1 = 4, x2 = 5, x3 = 0, Min (Z ) = − 11

9. 1 13
x1 = 1, x2 = 1, x3 = , x4 = 0, Max (Z ) =
2 2
20 45 235
10. x1 = , x2 = , Max (Z ) =
19 19 19

11. x1 = 2, x2 = 6, Max (Z ) = 36

12. x1 = 0, x2 = 20, Max (Z ) = 200

75 25
or x1 = , x2 = , Max (Z ) = 200
4 2

13. x1 = x2 = x3 = 0, Min (Z ) = 0

2.3 Linear Programming Problems Big 'M' Method and

Two Phase Method

2.3.1 Use of Artificial Variables

The solution in case of constraints of the (= or ≥) type we introduce non-negative
variables to the left hand side of all such constraints. The variables are called artificial
variables. The artificial variables play the same role as that of slack variables in
providing an initial basic feasible solution. Use of artificial variables is a systematic way
of getting a canonical system with a basic feasible solution when none is available by
inspection.
 43

There are two methods to solve such problems:

1. The M-methods or Big M-method or Charne's method of penalty
2. The two phase method.

2.3.2 Method of Penalties

When slack variables do not provide the initial basic feasible solution, we use artificial
variables. That is in order to complete the identity matrix, we introduce artificial
variables along with slack and/or surplus variables.

As the coefficients of slack variables are taken as 0 in the objective function, we assign a
very high penalty cost (say − M, M ≥ 0) to the artificial variables in the objective
function (that is their coefficients are taken as − M in case of maximization) and we
assign a very high penalty cost (say, M, M ≥ 0) in case of minimization so that they may
be driven to zero while reaching optimality. Then L.P.P. is solved as usual by simplex
method. The method of solving a linear programming problem in which a high penalty
cost has been assigned to the artificial variables is known as the Charne's Method of
Penalties of Big -M Method. If the artificial variable introduced in L.P.P. is present in
the final tableau at zero level the solution so obtained is optimal. Otherwise the solution
is known as Pseudo-Optimal solution.

(Big − 'M' Method)

Example 8: Solve the following L.P.P. by simplex method.
Minimize Z = 3 x1 + 2 .5 x2
such that x1 + 2 x2 ≥ 20
3 x1 + 2 x2 ≥ 50
x1, x2 ≥ 0.

Solution: Standard form

Minimize Z = 3 x1 + 2 . 5 x2 + 0 . S1 + 0 . S2
such that x1 + 2 x2 − 1. S1 + 0. S2 = 20
3 x1 + 2 x2 + 0. S1 − 1. S2 = 50
and x1, x2 , S1. S2 ≥ 0

Here, identity matrix is not complete so we use artificial variables a1, a2 . The coefficients of
a1 and a2 are taken as M in the objective function Z. Thus, the L.P.P. is
Minimize Z = 3 x1 + 2.5 x2 + 0. S1 + 0. S2 + Ma1 + Ma2
such that x1 + 2 x2 − 1. S1 + 0. S2 + 1. a1 + 0. a2 = 20
3 x1 + 2 x2 + 0. S1 − 1. S2 + 0. a1 + 1. a2 = 50
and x1, x2 , S1, S2 , a1, a2 ≥ 0
 44

Simplex Tableau 1

CB Cj 3 2.5 0 0 M M XB Ratio

Basis x1 x2 S1 S2 a1 a2

M a1 1 2 −1 0 1 0 20 20
= 10
2
(min)

M a2 3 2 0 −1 0 1 50 50
= 25
2

X1 X2 ↑ S1 S2 A1↓ A2

Zj 4M 4M −M −M M M x1 = 0

x2 = 0

∆ j = C j− Z j 3 − 4M 2.5 − 4 M M M 0 0 S1 = 0 Z = 70 M
(key S2 = 0
column)

Conclusion: Since some ∆ j are negative, hence the initial basic feasible solution is not

optimal. The highest negative value of ∆ j is ∆2 = 2.5 − 4 M, hence vector x2 will enter.

Since the corresponding ratio is minimum in first row so vector A1 is outgoing vector.

Key element is 2

Simplex Table 2

First row x1 x2 S1 S2 a1 a2 XB
(First row ÷ 2)
1 1 1 0 1 0 10
−
2 2 2

Second row (old second row −2 × new first row)

x1 x2 S1 S2 a1 a2 XB

3 −2 ×
1 2 − 2 ×1  1 −1 − 2 × 0 0 − 2 × 1 1−2 ×0 50 − 2 × 10
0 − 2 × − 
2  2 2

i.e. 2 0 1 −1 −1 1 30
 45

Simplex (Revised) Tableau 2

CB Cj 3 2.5 0 0 M M XB Ratio

x1 x2 S1 S2 a1 a2
Basis

2.5 x2 1 1 1 0 1 0 10 10
− = 20
2 2 2 1/2

M a2 2 0 1 −1 −1 1 30 30
= 15
2
(min.)

X1 ↑ X2 S1 S2 A1 ↓ A2

Zj 5 2.5 − 5 + 2 M − M 5 M
+ 2M −M
4 4 4

∆ j = Cj − Z j 7 0 5 M 5 0 Z = 25 + 30 M
− 2M −M − + M
4 4 4
(key
column)

Conclusion: Some ∆ j are negative, hence this solution is not optimal. The largest
7
negative value of ∆1 is ∆ j = − 2 M, hence vector X1 will enter. The ratio is minimum for
4
second row, hence, vector A2 is outgoing vector. The key element is 2

For Simplex (Revised) Table 3

Main row (Second old row ÷ 2), Second new row

x1 x2 S1 S2 a1 a2 XB

2 0 1 1 1 1 30
− −
2 2 2 2 2 2 2

i.e. 1 0 1 1 1 1 15
− −
2 2 2 2

1
First row (old first row − main first row)
2

x1 x2 S1 S2 a1 a2 XB

1 1 1 1 1 1 1  1 1 1  1 1 1 15
− 1− ×0 − − × 0− −  −   0− × 10 −
2 2 2 2 2 2 2  2 2 2  2 2 2 2

i.e. 0 1 3 1 3 1 5
− −
4 4 4 4 2
 46

CB Cj 3 2.5 0 0 M M XB Ratio
x1 x2 s1 s2 a1 a2
Basis
2.5 x2 0 1 3 1 3 1 5 No Need
− −
4 4 4 4 2
3 x1 1 0 1 1 1 1 15
− −
2 2 2 2
X1 X2 S1 S2 A1 A2
Zj 3 2.5 3 7 3 7 25
− − Z = + 45
8 8 8 8 4
∆ j = Cj − Z j 0 0 3 7 3 7 = 51.25
M− M−
8 8 8 8

Conclusion: Since all ∆ j are positive, hence present solution is optimal.

5
Where, x1 = 15, x2 =
2
5
and Z = 3 × 15 + 2 .5 ×
2
= 45 + 6 .25 = 51.25

Example 9: Use penalty (or Big-M) method due to A. Charnes to solve the following L.P.P. .
Minimize Z = 4 x1 + 3 x2
subject to 2 x1 + x2 ≥ 10
− 3 x1 + 2 x2 ≤ 6
x1 + x2 ≥ 6
and x1, x2 ≥ 0

Solution: Convert the L.P.P. in maximization form. Introduce surplus variables x3 ≥ 0

and x5 ≥ 0 and slack variable x4 ≥ 0 in the constraints inequalities. Then the problem in
standard form becomes
Maximize (− Z ) = Z * = − 4 x1 − 3 x2 + 0 . x3 − 0 . x4 + 0 . x5
subject to 2 x1 + x2 − x3 = 10
− 3 x1 + 2 x2 + x4 = 6
x1 + x2 − x5 = 6
clearly we do not have ready initial basic feasible solution. The surplus variables carry
negative coefficients (− 1). So, introduce two artificial variables a1 ≥ 0 and a2 ≥ 0 in first
and third equations respectively. Then we have
Maximize Z * = − 4 x1 − 3 x2 + 0. x3 + 0. x4 + 0. x5 − M. a1 − M. a2

subject to 2 x1 + x2 − x3 + a1 = 10
− 3 x1 + 2 x2 + x4 = 6
x1 + x2 − x5 + a2 = 6
with usual notations, we have.
 47

Simplex Tableau 1

CB Cj −4 −3 0 0 0 −M −M XB Ratio

x1 x2 x3 x4 x5 a1 a2
Basis

−M a1 2 1 −1 0 0 1 0 10 10
= 5 (min)
2
0 x4 −3 2 0 1 0 0 0 6 negative
−M a2 1 1 0 0 −1 0 1 6 6
=6
1

∆ j = Z j − Cj ↑ X1 X2 X3 X4 X5 ↓ A1 A2 x1 = 0 Z * = − 16 M

 − 3 M  − 2 M M 0 M 0 0 x2 = 0
 +4   
   +3 
↓ x3 = 0 Z = 16 M
most
negative x5 = 0

 10 
*   
Z = CB ′ X B =  − M 0 M  6  = − 16 M
 6 
  

Test for Optimality:

 2
 
∆1 = CB ′ X1 − C1 = [− M 0 − M] −3  − (− 4)
 1
 
= (− 2 M + 0 − M) + 4 = (− 3 M + 4)
1
 
∆2 = CB ′ X2 − C2 = (− M 0 − M) 2  − (− 3)
1
 
= (− M + 0 − M) + 3 = (− 2 M + 3)
− 1
 
∆3 = CB ′ X2 − C3 = [− M 0 − M]  0  − 0 = M
 0
 
∆4 = 0
 0
 
∆5 = CB ′ X5 − C5 = [− M 0 − M]  0  − 0 = M
−1
 
∆6 = 0
∆7 = 0
 48

Since some of the net evaluations ∆ j are negative, the initial basic feasible solution
[ x1 = 0, x2 = 0, x3 = 0, x5 = 0, a1 = 10, x4 = 6, a2 = 6] is not optimal. The most negative ∆ j
is ∆1 = (− 3 M + 4) . The corresponding column vector X1, therefore enters the basis.

Further the ratio is minimum for a1, therefore A1 is the outgoing vector, (we drop A1 forever,
because we do not like it to re-enter in any subsequent iteration).

1
  1
To convert the vector X1 as 0  we apply the transformation R1 → R1 , R2 → R3 + 3 R1,
2
0 
 
R3 → R3 − R1 successively.

Simplex Tableau 2

CB Cj −4 −3 0 0 0 −M XB xB
Ratio =
X2
x1 x2 x3 x4 x5 x6
Basis

−4 x1 1 1 1 0 0 0 5 5
− = 10
2 2 1/2

0 x4 0 7 3 1 0 0 21 21 42
− = =6
2 2 7 /2 7

−M a2 0 1 1 0 −1 1 1 1
= 2 (min.)
2 2 1/2

X1 ↑ X2 X3 X4 X5 ↓ A2 x1 = 0

∆ j = Z j − Cj 0  M   M  0 M 0 x2 = 0 Z * = − 20 − M
− 2 + 1 − 2 + 1

most negative x5 = 0 Z = 20 + M

5 
*  
Z = CB ′ X B = [− 4 0 M] 21 = − 20 − M
1
 

Test for Optimality:

∆1 = 0
1 / 2 
 
∆2 = CB ′ X2 − C2 = [− 4 0 − M] 7 / 2  − (− 3)
1 / 2 
 
 4 M M
= − + 0 − +3=− +1
 2 2  2
 49

 − 1 / 2
 
∆3 = CB ′ X3 − C3 = (− 4 0 − M) − 3 / 2  − 0
 − 1 / 2
 
4 M M
= +0−  =− +2
 2 2  2
∆4 = 0
 0
 
∆5 = CB ′ X5 − C5 = [− 4 0 − M]  0  − 0 = M
− 1
 
∆6 = 0

Since some of the net evaluation ∆ j are negative, the current basic feasible solution
[ x1 = 5, x2 = 0, x3 = 0, x4 = 21, x5 = 0, a2 = 1] is not optimal.
M
The most negative ∆ j is ∆2 = − + 1. The corresponding column vector x2 , therefore
2
enter the basis.

Further the ratio minimum for a2 , therefore, A2 is the outgoing vector we drop A2
forever.

0 
 
To convert the vector X2 as 0  , we apply the transformation
1
 
1 7
R3 → 2 R3 , R1 → R1 − R3 , R2 → R2 − R3 successively
2 2

Simplex Tableau 3

CB Cj − 4 −3 0 0 0 XB Ratio

Basis x1 x2 x3 x4 x5

−4 x1 1 0 −1 0 1 4 No need

0 x4 0 0 −5 1 7 14

−3 x2 0 1 1 0 −2 2

∆ j = Z j − Cj X1 X2 X3 X4 X5 x3 = 0 Z * = CB ′ X B = − 22

0 0 1 0 2 x5 = 0 Z = 22

Test for Optimality

∆1 = 0
∆2 = 0
 50

 − 1
 
∆3 = CB ′ X3 − C3 = [− 4 0 − 3] − 5  − 0 = 4 − 3 = 1
 1
 
∆4 = 0
 1
 
∆5 = CB ′ X5 − C5 = [− 4 0 − 3]  7 − 0 = − 4 + 6 = 2
− 2 
 
Since all net evaluation ∆ j are positive. The current basic feasible solution
( x1 = 4, x2 = 2, x3 = 0, x4 = 14, x5 = 0) is optimal. Thus, the required solution is x1 = 4,
x2 = 2 with minimum Z = 22.

Example 10: Solve the following linear programming problem by 'Big M' method:
Maximize Z = x1 + 2 x2 + 3 x3 − x4
subject to x1 + 2 x2 + 3 x3 = 15
2 x1 + x2 + 5 x3 = 20
x1 + 2 x2 + x3 + x4 = 10
and x1, x2 , x3 , x4 ≥ 0 [B.C.A. (Lucknow) 2010]
Solution: Introducing the artificial variables a1 ≥ 0 and a2 ≥ 0, the L.P.P. in standard
form is as follows:
Maximize Z = x1 + 2 x2 + 3 x3 − x4 − Ma1 − Ma2
subject to x1 + 2 x2 + 3 x3 + 0. x4 + a1 = 15
2 x1 + x2 + 5 x3 + 0. x4 + a2 = 20
x1 + 2 x2 + x3 + x4 = 10
and x1, x2 , x3 , x4 , a2 ≥ 0
with usual notations, we have
Simplex Tableau 1
CB Cj 1 2 3 −1 − M −M CB Ratio
x1 x2 x3 x4 a1 a2
Basis
−M a1 1 2 3 0 1 0 15 15
=5
3
−M a2 2 1 5 0 0 1 20 20
=4
5
(min.)
−1 x4 1 2 1 1 0 0 10 10
= 10
1
X1 X2 X3 ↑ X4 A1 A2 ↓ x1= 0
∆ j = Z j − Cj (−3 M − 2) (−3 M − 4) (−8 M − 4) 0 0 0 x2 = 0 Z = − 35
(min)
x3 = 0 M −10
 51

1
 
∆1 = CB ′ X1 − C1 = [− M − M − 1] 2  − 1
1
 
= [ − M − 2 M − 1] − 1 = − 3 M − 2
2 
 
∆2 = CB ′ X2 − C2 = [− M − M − 1] 1 − 2
2 
 
= [− 2 M − M − 2] − 2 = − 3 M − 4
3 
 
∆3 = CB ′ X3 − C3 = [− M − M − 1] 5  − 3
1
 
= [− 3 M − 5 M − 1] − 3 = − 8 M − 4
15 
 
Z = CB ′ X B = − M − M − 1] 20  = − 35 M − 10
10 
 
Since some net evaluations ∆ j are negative, the initial basic feasible solution
( x1 = 0, x2 = 0, x3 = 0, a1 = 15, a2 = 20, x4 = 10) is not optimal.
Since ∆3 is most negative, hence vector x3 is the incoming vector.
Since the ratio is minimum for a2 , so vector a2 is dropped forever (i.e. A2 is the outgoing
vector).
0 
 
To convert the vector X3 as 1 , we apply the transformations.
0 
 
1
R2 → R2 , R1 − 3 R2 , R3 → R3 − R1 respectively.
5

Simplex Tableau 2
CB Cj 1 2 3 −1 −M XB Ratio
Basis x1 x2 x3 x4 a2
−M a1 1 7 0 0 1 3 7 15
− 3÷ =
5 5 5 7
(min.)
3 x3 2 1 1 0 0 4 1
4 ÷ = 20
5 5 5
−1 x4 3 9 0 1 0 6 9 10
6÷ =
5 5 5 3
X1 ↑ X2 X3 X4 ↓ A1 x4 = 0 Z = − 3 M + 6
∆ j = Z j − Cj  M 2  −7 M 16  0 0 0 x2 = 0
 −   15 − 5 
 5 5
 52

− 1 / 5 
 
∆1 = CB ′ X1 − C1 = [− M 3 − 1]  2 / 5  − 1
 3 / 5
 
M 6 3 M 2
= + −  −1 = −
 5 5 5  5 5
7 / 5 
 
∆2 = CB ′ X2 − C2 = [− M 3 − 1] 1 / 5  − 2
9 / 5 
 
 7M 3 9  7 M 16
= − + −  −2 = − −
 5 5 5 5 5

Since ∆2 is negative, the current basic feasible solution ( x1 = 0, x2 = 0, a1 = 3, x3 = 4, x4 = 6)

is not optimal. So X2 is the incoming vector. Since the ratio is minimum for a1 so vector
1
 
A1 is the outgoing vector. To convert the vector X2 as 0  we apply the transformation
0 
 
5 1 9
R1 → R1, R2 → R2 − R1, R3 → R3 − R1 successively.
7 5 5

Simplex Tableau 3

CB Cj 1 2 3 −1 XB Ratio

x1 x2 x3 x4
Basis

2 x2 1 1 0 0 15/7 Negative
−
7

3 x3 3 0 1 0 25/7 25 3 25
÷ =
7 7 7 3

−1 x4 6 0 0 1 15/7 15 6 5
÷ =
7 7 7 2
(min.)

↑ X1 X2 X3 X4 ↓ x1 = 0

∆ j = Z j − Cj 6 0 0 0
−
7
Key Column

−1 / 7
 
∆1 = CB ′ X1 − C1 = [2 3 − 1]  3 / 7 − 1
 6 / 7
 
 2 9 6 1 6
= − + −  − 1 = − 1 = −
 7 7 7 7 7
 53

Since ∆1 is negative, the current basic feasible solution

 15 25 15 
 x1 = 0, x2 = , x3 = , x4 = 
 7 7 7

is not optimal. Hence, the vector X1 is the incoming vector, since the ratio is minimum
for X4 so the vector X4 is the outgoing vector.
0 
 
To convert the vector X1 as 0  , we apply the transformation
1
 
7 1 3
R3 → R3 , R1 → R1 + R3 , R2 → R2 − R3 successively.
6 7 7

Simplex Tableau 4
CB Cj 1 2 3 −1 xB Ratio
Basis x1 x2 x3 x4
2 x2 0 1 0 1/6 5/2 No need
3 x3 0 0 1 −1/2 5/2
1 x1 1 0 0 7/6 5/2
X1 X2 X3 X4 x4 = 0
∆ j = Z j − Cj 0 0 0 1 Z = CB ′ X B = 15

 1 / 6
   2 3 7
∆4 = [2 3 1] − 1 / 2  − (− 1) =  − +  + 1 = 1
 6 2 6
 7 / 6
 
Since all ∆ j ≥ 0 the current basic feasible solution is optimal.
5 5 5
Hence, x1 = , x2 = , x3 = , x4 = 0 with maximum Z = 15.
2 2 2

Example 11: Maximize Z = 4 x1 + 5 x2 − 3 x3

subject to x1 + x2 + x3 = 10
x1 − x2 ≥ 1
2 x1 + 3 x2 + x3 ≤ 40
and x1, x2 , x3 ≥ 0 [B.C.A. (Bhopal) 2012]

Solution: Using a1 ≥ 0, a2 ≥ 0 artificial variables and x4 ≥ 0, x5 ≥ 0 as slack variables, the

given L.P.P. is of the form:
Maximize Z = 4 x1 + 5 x2 − 3 x3 − 0. x4 − Ma1 − Ma2 + 0 . x5
subject to x1 + x2 + x3 + a1 = 10
x1 − x2 − x4 + a2 =1
2 x1 + 3 x2 + x3 + x5 = 40
and x1, x2 , x3 , x4 , x5 , a1, a2 ≥ 0
with usual notations, we have
 54

Simplex Tableau 1

CB Cj 4 5 −3 0 −M −M 0 XB Ratio

Basis x1 x2 x3 x4 a1 a2 x5

−M a1 1 1 1 0 1 0 0 10 10

−M a2 1 −1 0 −1 0 1 0 1 1 (min.)

0 x5 2 3 1 0 0 0 1 40 20

X1 ↑ X2 X3 X4 A1 ↓ A2 X5 x1 = 0

∆ j = Z j − Cj (−2 M − 4) −5 (− M + 3) M 0 0 0 x2 = 0
(Most x3 = 0 Z = − 11M
Negative) x4 = 0

0 
 
To convert the vector X1 as 1 , we apply the transformation R2 → R2 , R1 → R1 − R2 ,
0 
 
R3 → R3 − 2 R2 successively. Drop vector A2 forever.

Simplex Tableau 2

CB Cj 4 5 −3 0 −M 0 XB Ratio

x1 x2 x3 x4 a1 x5
Basis

−M a1 0 2 1 1 1 0 9 9
= 4 .5
2
(min.)

4 x1 1 −1 0 −1 0 0 1 Negative

0 x5 0 5 1 2 0 1 38 28
= 7 .6
5

X1 X2 ↑ X3 X4 ↓ A1 X5 x2 = 0

∆ j = Z j − Cj 0 (−2 M − 9) (− M + 3) (− M − 4) 0 0 x3 = 0 Z =
−9 M + 4
x4 = 0

1
  1
To convert the vector X2 as 1 , we apply the transformation R1 → R1, R2 → R2 + R1,
2
0 
 
R3 → R3 − 5 R1 successively. Drop vector A1 forever.
 55

Simplex Tableau 3
CB Cj 4 5 −3 0 0 XB Ratio
Basis x1 x2 x3 x4 x5

5 x2 0 1 1 1 0 9 No
2 2 2 need
4 x1 1 0 1 1 0 11
−
2 2 2
0 x5 0 0 3 1 1 31
− −
2 2 2
X1 X2 X3 X4 X5 x3 = 0 89
Z =
15 1 2
∆ j = Z j − Cj 0 0 0 x4 = 0
2 2
5 4 15
∆3 = + − (− 3) =
2 2 2
5 4 1
∆4 = − −0 =
2 2 2

Since the net evaluation for not basic variables x3 and x4 are positive, hence the current
 11 9 31
basic feasible solution  x1 = , x2 = , x3 = 0, x4 = 0, x5 =  is optimum with maximum
 2 2 2
11 5 × 9 89
Z =4 × + −3 ×0 = .
2 2 2

2.3.3 Two Phase Method (Due to Dantzing, Orden and Wolfe)

In the two phase method the L.P.P. is solved in two phase using the artificial
variables.

Phase I: In this phase the artificial variables are eliminated and a basic feasible solution
of the original linear programming problem is found. For this an artificial objective
function is created which is the sum of all the artificial variables. Assign a cost ‘ − 1’ to
each artificial variables and cost 0 to all other variables in the new objective function. The
artificial objective function is then minimized using the simplex method. If the minimum
value of the artificial objective function is zero, then all the artificial variables will be zero
and we have a basic feasible solution to the original problem. Then we go to phase II.

In case the minimum value of the artificial objective function is positive then at least one
of the artificial variables is positive. This means that the original L.P.P. without the
artificial variables is feasible and we terminate.

Phase II: In this phase, the basic feasible solution found at the end of Phase I is
optimized (i.e. improved) with respect to original objective function. That is the final
tableau of Phase I becomes the initial tableau for Phase II after changing the objective
function.
 56

On Two Phase Method

Example 12: Solve the following L.P.P. by using two phase method.
Minimize Z = x1 + x2
subject to 2 x1 + x2 ≥ 4
x1 + 7 x2 ≥ 7
and x1, x2 ≥ 0

Solution: Convert the problem of minimization to maximization.

Maximization (− Z ) = − x1 − x2

Introduce surplus variables x3 ≥ 0, x4 ≥ 0 in the constraints:

2 x1 + x2 − x3 = 4
x1 + 7 x2 − x4 = 7

The initial basic feasible solution cannot be obtained (i.e. the basic variables do not
exist). Introduce artificial variables a1 ≥ 0, a2 ≥ 0 then the constraints are:
2 x1 + x2 − x3 + a1 + 0. a2 = 4
x1 + 7 x2 − x4 + 0 . a1 + a2 = 7
where a1 < x3 and a2 < x4 otherwise the constraints of the problem will not hold.

1 0 
Now, the basic variables are a1 and a2 which complete the identity matrix I =   . So
0 1
the basic initial basic feasible solution is
[ x1 = 0, x2 = 0, x3 = 0, x4 = 0 , a1 = 4, a2 = 7] .

Phase I: Assign a cost (− 1) to artificial variables a1 and a2 and a cost 0 to all other
variables, the new objective function is:

Z * = 0 . x1 + 0 . x2 + 0 . x3 + 0 . x4 − a1 − a2

Simplex Tableau 1
CB Cj 0 0 0 0 −1 −1 XB Ratio
Basis x1 x2 x3 x4 a1 a2

−1 a1 2 1 −1 0 1 0 4 4
=4
1
−1 a2 1 7 0 −1 0 1 7 7
=1
7
(min.)
X1 ↑ X2 X3 X4 A1 A2 ↓ x1 = 0
∆ j = Z j − Cj −3 −8 1 1 0 0 x2 = 0
(min.) x3 = 0 Z * = − 11
x4 = 0
 57

Now, the vector X2 is incoming vector and A2 is the outgoing vector (drop it forever). To
0  1
make the vector X2 as   , apply the transformation R2 = R2 and R1 → R1 − R2
1  7
successively.

Simplex Tableau 2

CB Cj 0 0 0 0 −1 XB Ratio

x1 x2 x3 x4 a1
Basis

−1 a1 13 0 −1 1 1 3 21
(min.)
7 7 13

0 x2 1 1 0 1 0 1 7
−
7 7

↑ X1 X2 X3 X4 ↓ A1 x1 = 0 Z* = − 3

∆ j = Z j − Cj 13 0 1 1 0 x3 = 0
− −
7 7
x4 = 0
(min.)

Now, the vector X1 is incoming vector and A1 is the outgoing vector (drop it forever). To
1  7 1
make the vector X1 as   apply the transformation. R1 → R1 and R2 → R2 − R1
0
  13 7
successively.

Simplex Tableau 3

CB Cj 0 0 0 0 XB Ratio

x1 x2 x3 x4
Basis

0 x1 1 0 7 1 21 No need
−
13 13 13

0 x2 0 1 1 2 10
−
13 13 13

X1 X2 X3 X4 x3 = 0

∆ j = Z j − Cj 0 0 0 0 x4 = 0 Z *= 0

Since all ∆ j ≥ 0, the current basic feasible solution is optimum.

Phase II: Consider the original objective function, − Z = x1 − x2 and the constraints from
simplex tableau 3, we have:
 58

Simplex Tableau 4

CB Cj −1 −1 0 0 XB Ratio

Basis x1 x2 x3 x4

−1 x1 1 0 7 1 21 No need
−
13 13 13
−1 x2 0 1 1 2 10
−
13 13 13
X1 X2 X3 X4 x3 = 0
∆ j = Z j − Cj 0 0 6 1 x4 = 0 21 10 31
−Z = − − =−
13 13 13 13 13

Since all ∆ j ≥ 0, hence the current basic feasible solution is optimal.

21 10 31 31
Thus, x1 = , x2 = with maximum (− Z ) = − of minimum Z = .
13 13 13 13

Example 13: Solve the following L.P.P. by two phase method.

Maximize Z = 2 x1 − x2 + x3
subject to x1 + x2 − 3 x3 ≤ 8
4 x1 − x2 + x3 ≥ 2
2 x1 + 3 x2 − x3 ≥ 4
and x1, x2 , x3 ≥ 0 [B.C.A. (Rohtak) 2012]

Solution: Introducing x4 ≥ 0 as slack variables, x5 ≥ 0 and x6 ≥ 0 as surplus variable the

given L.P.P. becomes
Maximize Z = 2 x1 − x2 + x3 + 0 . x4 + 0 . x5 + 0 . x6
subject to x1 + x2 − 3 x3 + x4 + 0 . x5 + 0 . x6 = 8
4 x1 − x2 + x3 + 0 . x4 − x5 + 0 . x6 = 2
2 x1 + 3 x2 − x3 + 0 . x4 + 0 . x5 − x6 = 4
and x1, x2 , x3 , x4 , x5 , x6 ≥ 0

Since no identity matrix exists, we introduces artificial variable a1 ≥ 0 and a2 ≥ 0 . Then

the constraints becomes
x1 + x2 − 3 x3 + x4 + 0 . x5 + 0 . x6 + 0 . a1 + 0 . a2 = 8
4 x1 − x2 + x3 + 0 . x4 − x5 − 0 . x6 + a1 + 0 . a2 = 2
2 x1 + 3 x2 − x3 + 0 . x4 + 0 . x5 − x6 + 0 . a1 + a2 = 4
and x1, x2 , x3 , x4 , x5 , x6 , a1, a2 ≥ 0

Phase I: Assign a cost (−1) to each artificial variable and a cost 0 to all other variables to
form a new objective function.
Let Z * = 0 . x1 + 0 . x2 + 0 . x3 + 0 . x4 + 0 . x5 + 0 . x6 − a1 − a2
 59

Simplex Tableau 1

CB Cj 0 0 0 0 0 0 −1 −1 XB Ratio

x1 x2 x3 x4 x5 x6 a1 a2
Basis

0 x4 1 1 −3 1 0 0 0 0 8 8

−1 a1 4 −1 1 0 −1 0 1 0 2 1
(min.)
2

−1 a2 2 3 −1 0 0 −1 0 1 4 2

↑ X1 X2 X3 X4 X5 X6 A1 ↓ A2 x1 = 0

∆ j = Z j − Cj −6 −2 0 0 1 1 0 0 x2 = 0
(min.) x3 = 0

x5 = 0 Z* = − 6

x6 = 0

The incoming vector is X1 and the outgoing vector is A1 (drop it forever). To convert the
0 
  1
vector X1 as 1 , apply the transformation R2 → R2 , R1 → R1 − R2 , R3 → R3 − 2 R1
4
0 
 

successively.
Simplex Tableau 2

CB Cj 0 0 0 0 0 0 −1 XB Ratio
x1 x2 x3 x4 x5 x6 a2
Basis

0 x4 0 5 13 1 1 0 0 15 15 5 3
− + =
4 4 4 2 2 4 2
0 x1 1 1 1 0 1 0 0 1 Negative
− −
4 4 4 2

−1 a2 0 7 3 0 1 −1 1 3 7 6
− 3+ =
2 2 2 2 7
(min.)

X1 ↑ X2 X3 X4 X5 X6 ↓ A2 x2 = 0

∆ j = Z j − Cj 0 7 3 0 1 1 0 x3 = 0 Z = −3
− −
2 2 2
(Key Column) x5 = 0
x6 = 0
 60

The incoming vector is X2 and the outgoing vector is A2 (drop it forever). To convert
0 
  2 5
the vector X2 as 0  apply the transformation. R3 → R3 , R1 → R1 − R3 and
7 4
1
 
1
R2 → R2 + R3 successively.
4
Simplex Tableau 3

CB Cj 0 0 0 0 0 0 XB Ratio
x1 x2 x3 x4 x5 x6
Basis

0 x4 0 0 19 1 1 5 45 No need
−
7 14 14 7
0 x1 1 0 1 0 3 1 5
− −
7 14 14 7
0 x2 0 1 3 0 1 2 6
− −
7 7 7 7

X1 X2 X3 X4 X5 X6 x3 = 0
∆ j = Z j − Cj 0 0 0 0 0 0 x5 = 0 Z* = 0
x6 = 0

Since ∆ j ≥ 0 and none of the artificial vector appears in the basis, the current basic
 5 6 45 
feasible solution  x1 = , x2 = , x4 = , x5 = 0, x6 = 0  will serve a BFS for phase II.
 7 7 7 

Phase II: Consider Z = 2 x1 − x2 + x3 + 0 . x4 + 0 . x5 + 0 . x6 and the constraints as in

simplex tableau 3, we have
Simplex Tableau 4

CB Cj 2 −1 1 0 0 0 XB Ratio
x1 x2 x3 x4 x5 x6
Basis

0 x4 0 0 19 1 1 5 45 90
−
7 14 14 7
2 x1 1 0 1 0 3 1 5 Nega-
− −
7 14 14 7
tive
−1 x2 0 1 3 0 1 2 6 6
− −
7 7 7 7 (min.)

X1 ↓ X2 X3 X4 ↑ X5 X6 x3 = 0
∆ j = Z j − Cj 0 0 2 0 4 1 x5 = 0 4
− − Z =
7 7 7 x6 = 0 7
 61

0 
 
The incoming vector is X5 and the outgoing vector is X2 . To convert the vector X5 as 0  ,
1
 
R2 3
apply the transformation R3 → 7R3 , R1 → R1 − , R3 → R2 + R3 successively.
14 14

Simplex Tableau 5

CB Cj 2 −1 1 0 0 0 XB Ratio

Basis x1 x2 x3 x4 x5 x6

0 x4 0 1 5 1 0 1 6 Nega-
− −
2 2 2 tive
2 x1 1 3 1 0 0 1 2 ”
− −
2 2 2
0 x5 0 7 −3 0 1 −2 6 ”

X1 X2 X3 X4 X5 X6 x2 = 0

∆ j = Z j − Cj 0 4 −2 0 0 −1 x3 = 0 Z =4
x6 = 0

Since the most negative value of net evaluation ∆ j is for ∆3 but the elements of vector X3
are all negative, the given L.P.P. admits an unbounded solution.

Example 14: Use two phase method to solve the following L.P.P.:
Maximize Z = 5 x1 + 3 x2
subject to 2 x1 + x2 ≤ 1
x1 + 4 x2 ≥ 6
and x1, x2 ≥ 0 [B.C.A. (Kurukshetra) 2010]

Solution: Introducing slack variables x3 ≥ 0 and surplus variable x4 ≥ 0 we have the given
L.P.P. in standard form.
Maximize Z = 5 x1 + 3 x2 + 0. x3 + 0. x4
subject to 2 x1 + x2 + x3 = 1
x1 + 4 x2 − x4 = 6
and x1, x2 , x3 , x4 ≥ 0

Here is no basic variable in second constraint, so we introduce an artificial variable a1 ≥ 0.

Then the constraints are:
2 x1 + x2 + x3 + 0. x4 + 0. a1 = 1
x1 + 4 x2 + 0. x3 − x4 + a1 = 6
and x1, x2 , x3 , x4 , a1 ≥ 0
 62

1 0
Here the basic variables are x3 and a1, which complete the identity matrix I =  .
0 1

So, the initial basic feasible solution is x1 = 0, x2 = 0, x4 = 0, x3 = 1 and a1 = 6 .

Phase I: Assign a cost ‘−1’ to artificial variable a1 and a cost 0 to all other variables, the

new objective function is:

Z * = 0 . x1 + 0 . x2 + 0 . x3 + 0 . x4 − a1

Simplex Tableau 1

CB Cj 0 0 0 0 −1 XB Ratio

Basis x1 x2 x3 x4 a1

0 x3 2 1 1 0 0 1 1
(min.)
1

−1 a1 1 4 0 −1 1 6 6
4

X1 ↑ X2 ↓ X3 X4 A1 x1 = 0

∆ j = Z j − Cj −1 −4 0 1 0 x2 = 0 Z* = − 6

(min.) x4 = 0

Now, the vector X2 is incoming vector and X3 is the outgoing vector. Apply the
transformation R1 → R1 and R → R2 − 4 R1.

Simplex Tableau 2

CB Cj 0 0 0 0 −1 XB Ratio

x1 x2 x3 x4 a1
Basis

0 x2 2 1 1 0 0 1 No need

−1 a1 −7 0 −4 −1 1 2

X1 X2 X3 X4 A1

∆ j = Z j − Cj 7 0 4 1 0 Z* = − 2

Since all ∆ j ≥ 0 an optimum basic feasible solution to the new L.P.P. is obtained. But
maximum Z * is negative and an artificial vector A1 is in the basis at a positive level.
Therefore, the original L.P.P. does not possess any feasible solution.
 63

P roblem S et
Solve the Following L.P.P. Using two Phase Method:

1. Min (Z ) = x1 + x2

subject to 2 x1 + x2 ≥ 4
x1 + 7 x2 ≥ 7
x1, x2 ≥ 0 [B.C.A. (Meerut) 2004, 2007]

2. Max (Z ) = 5 x1 + 8 x2

subject to 3 x1 + 2 x2 ≥ 3
x1 + 4 x2 ≥ 4
x1 + x2 ≤ 5
x1, x2 ≥ 0 [B.C.A. (Aligarh) 2009]

3. Max (Z ) = 3 x1 + 2 x2 + x3 + 4 x4

subject to 4 x1 + 5 x2 + x3 + 5 x4 = 5
2 x1 − 3 x2 − 4 x3 + 5 x4 = 7
x1 + 4 x2 + 5 x3 − 4 x4 = 6
x1, x2 , x3 , x4 ≥ 0 [B.C.A. (Agra) 2009]

4. Maximum (Z ) = 5 x1 − 2 x2 + 3 x3

subject to 2 x1 + 2 x2 − x3 ≥ 2
3 x1 − 4 x2 ≤ 3
x2 + 3 x3 ≤ 5
x1, x2 , x3 , x4 ≥ 0 [B.C.A. (Kashi) 2006, 2008]

Apply Big-M Method to Solve the Following Problems:

1. Max (Z ) = − 2 x1 − x2

subject to 3 x1 + x2 = 3
4 x1 + 3 x2 ≥ 6
x1 + 2 x2 ≤ 4
x1, x2 ≥ 0 [B.C.A. (Kanpur) 2009]

2. Max (Z ) = 4 x1 + 5 x2 − 3 x3

subject to x1 + x2 + x3 = 10
x1 − x2 ≥ 1
2 x1 + 3 x2 + x3 ≤ 40
x1, x2 , x3 ≥ 0 [B.C.A. (Bundelkhand) 2007]
 64

3. Min (Z ) = 5 x1 + 6 x2

subject to 2 x1 + 5 x2 ≥ 1500
3 x1 + x2 ≥ 1200
x1, x2 ≥ 0 [B.C.A. (Rohilkhand) 2009]

4. Min (Z ) = 2 y1 + 3 y2

subject to y1 + y2 ≥ 5
y1 + 2 y2 ≥ 6
y1, y2 ≥ 0 [B.C.A. (Indore) 2010]

5. Min (Z ) = x1 + x2 + 3 x3

subject to 3 x1 + 2 x2 + x3 ≤ 3
2 x1 + x2 + 2 x3 ≥ 3
x1, x2 , x3 ≥ 0 [B.C.A. (Rohilkhand) 2002, 2006]

A nswers
Two Phase Method:

21 10 31
1. x1 = , x2 = , Min (Z ) =
13 13 13

2. x1 = 0, x2 = 5, Max (Z ) = 40

3. No feasible solution

23 85
4. x1 = , x2 = 5, x3 = 0, Max (Z ) =
3 3

Big-M Method:

1. x1 = 3 5 , x2 = 6 5 , Max (Z ) = − 12 5

2. x1 = 11 2 , x2 = 9 2 , Max (Z ) = 89 2

3. x1 = 4500 13 , x2 = 2100 13 , Min (Z ) = 2700

4. y1 = 4, y2 = 1, Min (Z ) = 11

3 3
5. x1 = , x2 = 0, x3 = , min Z = 3
4 4
 65

2.4 Duality in Linear Programming

[B.C.A. (Lucknow) 2010]

The concept of duality was one of the most important discoveries in the early development
of linear programming. It explains that corresponding to every linear programming
problem there is another linear programming problem. The original problem is known as
primal and the other, the related problem is known as the dual. The two problems are
replicas of each other. If the primal problem is of maximization type, the dual will be of
the minimization type. If the optimal solution to one problem is known to us, we can
easily find the optimal solution of the other. This fact is important because at times it is
easier to solve the dual than the primal.

2.4.1 Symmetric Primal Dual Problems

Consider the following linear programming problem, we may call it the symmetric primal
problem.

Primal L.P.P.: Find the variables x1, x2 , … , x n which maximize

Z x = c1 x1 + c2 x2 + … + cn x n

subject to

a11 x1 + a12 x2 + … + a1n x n ≤ b1

a21 x1 + a22 x2 + … + a2 n x n ≤ b2
... ... ... ... ...
... ... ... ... ...

am1 x1 + am2 x2 + … + amn x n ≤ bm

and x1, x2 , … , x n ≥ 0

the signs of the parameters a, b, c are arbitrary. To obtain the dual of the above primal
following steps are required:

1. Minimize the objective function in place of maximizing it.

2. Interchange the role of constant terms and the coefficients of the objective
function.

3. Find A′ where A′ denotes the transpose of the coefficient matrix A.

4. Reverse the direction of inequalities.

 66

Thus, the dual problem is:

Find the variables w1, w2 , … , wm which minimize

Z w = b1w1 + b2 w2 + … + bm wm

subject to a11 w1 + a21 w2 + … + am1 wm ≥ c1

a12 w1 + a22 w2 + … + am2 w2 ≥ c2

... ... ... ... ... ... ... ... ...

... ... ... ... ... ... ... ... ...

a1n w1 + a2 n w2 + … + amn wm ≥ cn

and w1, w2 , … , wn ≥ 0

2.4.2 Matrix form of Symmetric Primal Dual Problem

[B.C.A. (Meerut) 2011]
n
Primal Problem: Find a column vector x ∈ R which maximizes

Z x = cx, c ∈ R n

subject to Ax ≤ b, b ∈ R m

x ≥ 0 and A is an m × n real matrix.

Dual Problem: Find a column vector w ∈ R m which minimizes

Z w = b′ w
subject to A′ w ≥ c′
w ≥ 0, A′ , b′ , c′ are the transposes of A, b and c respectively.

Example 15: Consider the symmetric primal problem

Max Z x = 5 x1 + 9 x2
subject to x1 ≤ 6
x1 + x2 ≤ 13
x2 ≤ 8
x1, x2 ≥ 0

Solution: The corresponding dual problem is

Min Z w = 6 w1 + 13 w2 + 8w3

subject to w1 + w2 ≥ 5
w2 + w3 ≥ 9
w1, w2 , w3 ≥ 0
 67

2.5 Unsymmetric Primal Dual Problems

[B.C.A. (Delhi) 2010, 2012]
n
Primal Problem: Find a column vector x ∈ R which maximizes.

Z x = cx, c ∈ R n

subject to Ax = b, b ∈ R m

x ≥ 0 and A is an m × n real matrix.

Dual Problem: Find a column vector w ∈ R m which minimizes

Z w = b′ w
subject to A′ w ≥ c′
In this case the dual variables are unrestricted in sign.

2.5.1 Dual of an L.P.P. with Mixed Restrictions

[B.C.A. (Lucknow) 2011]

Sometimes a given L.P.P. contains a mixture of inequalities (≥, ≤) equations; non-negative

variables and unrestricted variables, then to obtain its dual we proceed in the following
manner:

1. If a constraint is an equation (has = sign) replace it by two constraints involving the

inequalities going in the opposite directions.
For Example, the equation x1 + 2 x2 = 3 is replaced by

x1 + 2 x2 ≤ 3 ...(1)
and x1 + 2 x2 ≥ 3 ...(2)

2. If the given problem is of maximization, all constraints should have ≤ sign. If some
constraints has ≥ sign, multiply both sides by −1 and make the sign ≤.
Thus, in the above example multiply both sides of (2) by −1 so that it becomes
− x1 − 2 x2 ≤ − 3
Similarly, if the problem is of minimization, all constraints should have ≥ sign.

3. If there is some unrestricted variable, replace it by the difference of two variables.

4. Now to find the dual problem.

Standard Primal Form: A linear programming problem is said to be in standard primal

form if:
1. For a maximization problem all the constraints have ≤ sign
2. For a minimization problem all the constraints have ≥ sign.
 68

Example 16: Find the dual of the following L.P.P.

Min Z = 10 x1 + 20 x2

subject to 3 x1 + x2 ≥ 18

x1 + 3 x2 ≥ 8

2 x1 − x2 ≤ 6,

x1, x2 ≥ 0 [B.C.A. (Meerut) 2004]

Solution: The given L.P.P. in the standard primal form is

Min Z = 10 x1 + 20 x2

subject to 3 x1 + 2 x2 ≥ 18

x1 + 3 x2 ≥ 8

− 2 x1 + x2 ≥ − 6

and x1, x2 ≥ 0

The matrix form of this problem is

Min Z = 10 x1 + 20 x2 = (10, 20) [ x1, x2 ] = cx

 3 2  18
   x1   
subject to  1 3    ≥  8
 −2  x2   
 1 − 6 

or AX ≥ b, x1, x2 ≥ 0

∴ the dual of this problem is

Max. Z D = b′ y = (18, 8, − 6) ( y1, y2 , y3 )

= 18 y1 + 8 y2 − 6 y3

 y1 
3 1 −2    10 
Subject to A′ y ≤ c′ or   y2 ≤  
2 3 1   20 
y 
 3

3 y1 + y2 − 2 y3  10 
or  ≤ 
2 y1 + 3 y2 + y3  20 

or 3 y1 + y2 − 2 y3 ≤ 10, 2 y1 + 3 y2 + y3 ≤ 20, y1, y2 , y3 ≥ 0

Hence, dual of the given problem is

Max Z D = 18 y1 + 8 y2 − 6 y3

Subject to 3 y1 + y2 − 2 y3 ≤ 10, 2 y1 + 3 y2 + y3 ≤ 20, y1, y2 , y3 ≥ 0

 69

Example 17: Write the dual of the following L.P.P.

Max Z = 2 x1 + 3 x2 + x3
subject to 4 x1 + 3 x2 + x3 = 6
x1 + 2 x2 + 5 x3 = 4
x1, x2 , x3 ≥ 0. [B.C.A. (Meerut) 2005, 2010]

Solution: Now to convert the given L.P.P. into standard primal form.

Here, both constraints are qualities, so replacing each by two inequalities, we get the
constraints

4 x1 + 3 x2 + x3 ≤ 6 and 4 x1 + 3 x2 + x3 ≥ 6

x1 + 2 x2 + 5 x3 ≤ 4

and x1 + 2 x2 + 5 x3 ≥ 4

Since, the given problem is of maximization, so all the constraints should have the sign ≤ .

The standard primal form of the given L.P.P. is

Max Z = 2 x1 + 3 x2 + x3

subject to 4 x1 + 3 x2 + x3 ≤ 6

− 4 x1 − 3 x2 − x3 ≤ − 6

x1 + 2 x2 + 5 x3 ≤ 4

− x1 − 2 x2 − 5 x3 ≤ − 4

x1, x2 , x3 ≥ 0

The matrix form of the above problem is

Max Z = 2 x1 + 3 x2 + x3 = (2, 3, 1) [ x1, x2 , x3 ] = c . x

subject to

 4 3 1  6
   x1   
 −4 −3 −1    −6 
 1 2 5  x2  ≤  4 
  x   
 −1 −2 −5   3   −4 
 

or AX ≤ b

x1, x2 , x3 ≥ 0

∴ the dual of the given primal is

Min Z D = b′ . y = (6 , − 6 , 4 , − 4) [ y1′ , y1′ ′ , y2 ′ , y2 ′ ′ ]

= 6 ( y1′ − y1′ ′ ) + 4 ( y2 ′ − y2 ′ ′ )

Subject to A′ y ≥ c′
 70

 y1′ 
4 −4 1 −1   2 
or

−3 −2 
  y1′ ′  ≥ 3 
3 2
 y ′  
1 −1 5 −5   2  1
  y2 ′ ′   

4 y1′ − 4 y1′ ′ + y2 ′ − y2 ′ ′  2 
   
or 3 y1′ − 3 y1′ ′ + 2 y2 ′ − 2 y2 ′ ′  ≥ 3 
 y ′ − y ′ ′ + 5 y ′ − 5 y ′ ′  1
 1 1 2 2   
or 4 ( y1′ − y1′ ′ ) + y2 ′ − y2 ′ ′ ≥ 2
3 ( y1′ − y1′ ′ ) + 2 ( y2 ′ − y2 ′ ′ ) ≥ 3
( y1′ − y1′′ ) + 5 ( y2′ − y2′′ ) ≥ 1

y1′ , y1′ ′ , y2 ′ , y2 ′ ′ ≥ 0
Substituting y1 = y1′ − y1′ ′ , y2 = y2 ′ − y2 ′ ′, the required dual is
Min. Z D = 6 y1 + 4 y2
subject to 4 y1 + y2 ≥ 2, 3 y1 + 2 y2 ≥ 3, y1 + 5 y2 ≥ 1
where y1, y2 are unrestricted in sign.

Example 18: Write the dual of the following problem:

Min Z = x1 + x2 + x3
subject to x1 − 3 x2 + 4 x3 = 5,
x1 − 2 x2 ≤ 3,
2 x2 − x3 ≥ 4,
x1, x2 ≥ 0, x3 is unrestricted. [B.C.A. (Meerut) 2002, 2008]

Solution: Write the given L.P.P. in the standard primal form, substituting x3 = x3 ′ − x3 ′′ ,
x3 ′ ≥ 0 , x3 ′′ ≥ 0. The given problem can be written in the standard primal form as
Min Z = x1 + x2 + x3 ′ − x3 ′ ′
subject to − x1 + 3 x2 − 4 ( x3 ′ − x3 ′ ′ ) ≥ − 5
x1 − 3 x2 + 4 ( x3 ′ − x3 ′ ′ ) ≥ 5
− x1 + 2 x2 ≥ − 3
2 x2 − ( x3 ′ − x3 ′ ′ ) ≥ 4
x1, x2 , x3 ′ , x3 ′ ′ ≥ 0

The matrix form of the above problem is

Min Z = (1, 1, 1, − 1) [ x1, x2 , x3 ′ , x3 ′ ′ ] = c. x
subject to
 −1 3 −4 4  x1  −5 
     
 1 −3 4 −4   x2  ≥  5 
 −1 2 0 0  x3 ′  −3 
     
 0 2 −1 1  x3 ′ ′   4 
 71

or AX ≥ b, x1, x2 , x3 ′ , x3 ′′ ≥ 0

Now, the dual of the given primal is

Max Z D = (− 5, 5, − 3, 4) [ y1′ , y1′ ′ , y2 , y3 ]
= − 5 ( y1′ − y1′ ′ ) − 3 y2 + 4 y3
subject to A′ y ≤ c′
 −1 1 −1 0  y1′   1
     
 3 −3 2 2  y1′ ′  ≤  1
or
 −4 4 0 −1  y   1
   2   
 4 −4 0 1  y3  −1

or − ( y1′ − y1′ ′ ) − y2 ≤ 1
3 ( y1′ − y1′ ′ ) + 2 y2 + 2 y3 ≤ 1
− 4 ( y1′ − y1′ ′ ) − y3 ≤ 1
4 ( y1′ − y1′ ′ ) + y3 ≤ − 1
y1′ , y1′ ′ , y2 , y3 ≥ 0

Substituting y1 = y1′ − y1′ ′, the required dual is

Max Z D = − 5 y1 − 3 y2 + 4 y3
subject to − y1 − y2 ≤ 1, 3 y1 + 2 y2 + 2 y3 ≤ 1, − 4 y1 − y3 ≤ 1
4 y1 + y3 ≤ − 1
or − 4 y1 − y3 ≥ 1
y2 , y3 ≥ 0 and y1 is unrestricted.

Hence, the required dual is

Max Z D = − 5 y1 − 3 y2 + 4 y3
subject to − y1 − y2 ≤ 1, 3 y1 + 2 y2 + 2 y3 ≤ 1, − 4 y1 − y3 = 1
y2 , y3 ≥ 0, y1 is unrestricted in sign.

Example 19: Write the dual of the following problem:

Min Z = 2 x2 + 5 x3
subject to x1 + x2 ≥ 2
2 x1 + x2 + 6 x3 ≤ 6
x1 − x2 + 3 x3 = 4
x1, x2 , x3 ≥ 0 [B.C.A. (Bhopal) 2011]

Solution: First we shall convert the given problem to standard primal form:
1. Since the problems of minimization, therefore all the constraints should have the
sign ≥.

2. Multiplying the second constraint by −1, it becomes

− 2 x1 − x2 − 6 x3 ≥ − 6
 72

3. Since, the third constraint is an equality, so replacing it by the following two

constraints
x1 − x2 + 3 x3 ≥ 4
and x1 − x2 + 3 x3 ≤ 4
or x1 − x2 + 3 x3 ≥ 4
and − x1 + x2 − 3 x3 ≥ − 4
∴ the given problem is standard primal form is
Min Z = 0 x1 + 2 x2 + 5 x3
subject to x1 + x2 ≥ 2
− 2 x1 − x2 − 6 x3 ≥ − 6
x1 − x2 + 3 x3 ≥ 4
− x1 + x2 − 3 x3 ≥ − 4
x1, x2 , x3 ≥ 0
The matrix form of the above problem is
Min Z = (0, 2, 5) [ x1, x2 , x3 ] = c. x
 1 1 0  2
   x1   
−2 −1 −6     −6 
subject to   x2  ≥  4  or AX ≥ b, x1, x2 , x3 ≥ 0
 1 −1 3
  x   
 −1 1 −3   3 −4 

∴ the dual of the given primal is

Max Z D = b′ y = (2, − 6, 4, − 4) [ y1, y2 , y3 ′ , y3 ′ ′ ]
= 2 y1 − 6 y2 + 4 [ y3 ′ − y3 ′ ′ ]
subject to b′ y ≤ c ′
 y1 
1 −2 1 –1   0 

−1 −1
  y2  ≤ 2 
1 1
 y3 ′   
0 −6 3 –3    5 
  
 y3 ′ ′ 

 y1 − 2 y2 + y3 ′ − y3 ′ ′  0 
   
or  y1 − y2 − y3 ′ + y3 ′ ′  ≤ 2 
0. y − 6 y + 3 y ′ − 3 y ′ ′  5 
 1 2 3 3   

y1, y2 , y3 ′ , y3 ′ ′ ≥ 0
or Max Z D = 2 y1 − 6 y2 − 4 ( y3 ′ − y3 ′ ′ )
subject to y1 − 2 y2 + ( y3 ′ − y3 ′ ′ ) ≤ 0
y1 − y2 − ( y3 ′ − y3 ′ ′ ) ≤ 2
− 6 y2 + 3 ( y3 ′ − y3 ′ ′ ) ≤ 5
y1, y2 , y3 ′ , y3 ′ ′ > 0
 73

Substituting y3 = y3 ′ − y3 ′ ′, the required dual is

Max Z D = 2 y1 − 6 y2 + 4 y3
subject to y1 − 2 y2 + y3 ≤ 0, y1 − y2 − y3 ≤ 2, − 6 y2 + 3 y3 ≤ 5
y1, y2 ≥ 0 and y3 is unrestricted in sign.

Note:
The variable corresponding to the equality equation (here, third equation) in the
constraint will be unrestricted in sign.

P roblem S et
Find the Dual of the Following Linear Programming Problems:

1. Max Z = x1 − x2 + 3 x3 2. Min Z = 3 x1 + x2
subject to x1 + x2 + x3 ≤ 10 subject to 2 x1 + 3 x2 ≥ 2
2 x1 − x3 ≤ 2 x1 + x2 ≥ 1
2 x1 − 2 x2 + 3 x3 ≤ 6 x1, x2 ≥ 0
x1, x2 , x3 ≥ 0

3. Min Z = 4 x1 + 6 x2 + 18 x3 4. Min Z = 2 x1 + 3 x2 + 4 x3
subject to x1 + 3 x3 ≥ 3 subject to 2 x1 + 3 x2 + 5 x3 ≥ 2
x2 + 2 x3 ≥ 5 3 x1 + x2 + 7 x3 = 3
x1, x2 , x3 ≥ 0 x1 + 4 x2 + 6 x3 ≤ 5
x1, x2 ≥ 0, x3 is unrestricted
[B.C.A. (I.G.N.O.U.) 2006]

5. Min Z = x1 − 3 x2 − 2 x3 6. Max Z = 3 x1 + x2 + 4 x3 + x4 + 9 x5
subject to 3 x1 − x2 + 2 x3 ≤ 7 subject to 4 x1 − 5 x2 − 9 x3 + x4 − 2 x5 ≤ 6
2 x1 − 4 x2 ≥ 12 2 x1 + 3 x2 + 4 x3 − 5 x4 + x5 ≤ 9
− 4 x1 + 3 x2 + 8 x3 = 10 x1 + x2 − 5 x3 − 7 x4 + 11x5 ≤ 10,
x1, x2 ≥ 0, x3 is unrestricted. x1, x2 , x3 , x4 , x5 ≥ 0
[B.B.A. (Meerut) 2006]

7. How do the graphical and simplex methods of solving as L.P.P. differ?

[B.B.A. (Meerut) 2004]

8. Explain in brief duality in L.P.P. [B.B.A. (Meerut) 2006]

 74

A nswers
1. Min Z D = 10 y1 + 2 y2 + 6 y3 2. Max Z D = 2 y1 + y2

subject to y1 + 2 y2 + 2 y3 ≥ 1 subject to 2 y1 + y2 ≤ 3

y1 − 2 y3 ≥ − 1 3 y1 + y2 ≤ 1

y1 − y2 + 3 y3 ≥ 3 y1, y2 ≥ 0

y1, y2 , y3 ≥ 0

3. Max Z D = 3 y1 + 5 y2 4. Max Z D = 5 y1 − 3 y2 − 5 y3

subject to y1 ≤ 4 subject to 2 y1 − 3 y2 − y3 ≤ 2

y2 ≤ 6 3 y1 − y2 − 4 y3 ≤ 3

3 y1 + 2 y2 ≤ 18 5 y1 − 7 y2 − 6 y3 = 4

y1, y2 ≥ 0 y1, y3 ≥ 0, y2 is unrestricted.

5. Max Z D = − 7 y1 + 12 y2 + 10 y3 6. Min Z D = 6 y1 + 9 y2 + 10 y3

subject to − 3 y1 + 2 y2 − 4 y3 ≤ 1 subject to 4 y1 + 2 y2 + y3 ≥ 3

− y1 + 4 y2 − 3 y3 ≥ 3 − 5 y1 + 3 y2 + y3 ≥ 1

2 y1 − 8 y3 = 2 − 9 y1 + 4 y2 − 5 y3 ≥ 4

y1, y2 ≥ 0, y3 is unrestricted. y1 − 5 y2 − 7 y3 ≥ 1

− 2 y1 + y2 + 11 y3 ≥ 9

y1, y2 , y3 ≥ 0

❍❍❍
 75

C HAPTER
3

Transportation Problem

3.1 Transportation Problem

[B.C.A. (Meerut) 2004, 2007, 2011; B.C.A. (Agra) 2006, 2010; B.C.A. (Rohilkhand) 2011;
B.C.A. (Kanpur) 2006; B.B.A. (Meerut) 2002, 2003, 2004, 2007, 2009]

Transportation problem is a special kind of linear programming problem in which goods

are transported from a set of sources to a set of destination subject to the supply and
demand of the source and destination, respectively, such that the total cost of transportation
is minimized.

Let m = the number of sources

n = the number of destinations

ai = the supply at the source i

bj = the demand at the destination j

cij = the cost of transportation per unit from ith source to jth destination

x ij = the number of units to be transported from the source ith to the jth the
destination
 76

3.2 Mathematical Formulation of Transportation Problem

[B.C.A. (Kashi) 2011; B.C.A. (Lucknow) 2008; B.C.A. (Agra) 2002, 2004, 2009;
B.B.A. (Meerut) 2002]
Mathematically, the problem may be stated as follows:
m n
Minimize (Z ) = ∑ ∑ cij xij
i =1 j =1

Subject to the constraints:

n

∑ x ij = ai for i = 1, 2, 3 , … , m
j =1
m
and ∑=1 x ij = bj for j = 1, 2, 3 , … , n
i

where x ij ≥ 0 for all i and j

For a feasible solution to exist, it is necessary that total supply equals total requirement
i.e.
m n

∑=1 ai = ∑=1 bj [Rim Condition]

i j

The above information can be put in the form of a general matrix shown below:
Destination
w1 w2 ... wj ... wn Supply

F1 c11 c12 ... c1 j ... c1n a1

F2 c21 c22 ... c2 j ... c2 n a2

Origin ⋮ ⋮ ⋮ ⋮ ⋮ ⋮

Fi c i1 c i2 ... ci j ... ci n ai

⋮ ⋮ ⋮ ⋮ ⋮ ⋮

Fm cm1 cm2 ... cmj ... cmn am

m n
Demand b1 b2 ... bj ... bn
∑
i=
ai = ∑
j=
bj
1 1

3.2.1 Definition
1. Feasible Solution: A set of non-negative values x ij (i = 1, 2 , … m, j = 1, 2, 3, … n) that
satisfies the constraints is called a feasible solution to the transportation problem.
[B.C.A. (Meerut) 2009, 2010]
 77

2. Basic Feasible Solution: A feasible solution that contains no more than m + n − 1

non-negative allocations is called a basic feasible solution to the transportation
problem. [B.C.A. (Meerut) 2009]

3. Optimal Solution: A feasible solution is said to be optimal if it minimizes the

total transportation cost. [B.C.A. (Meerut) 2009]

4. Non-degenerate Basic Feasible Solution: A basic feasible solution to (m × n)

transportation problem that contains exactly m + n − 1 allocations in independent
positions.
5. Degenerate Basic Feasible Solution: A basic feasible solution that contains less
than m + n − 1 non-negative allocations.
6. Balanced and Unbalanced Transportation Problems: A transportation problem
is said to be balanced if the total supply from all the sources equals the total
demand in all the destinations and is called unbalanced otherwise. Thus,
m n
For balance problem ∑=1 ai = ∑=1 bj
i j

m n
For unbalance problem ∑=1 ai ≠ ∑=1 bj.
i j

3.2.2 Existence of Feasible Solution

Theorem 1: A necessary and sufficient condition for the existence of feasible
solution of a transportation problem is:
Σai = Σb j (i = 1, 2, … , m; j = 1, 2, … , n).
[B.C.A. (Lucknow) 2010; B.C.A. (I.G.N.O.U.) 2001, 2004, 2012]

Proof: The Condition is Necessary: Let there exist a feasible solution to the transportation
problem. Then,
n

∑=1 xij = ai, i = 1, 2 , … , m

j
m
and ∑=1 x ij = bj, j = 1, 2, … , n
i

Summing over all i and j respectively, we get

m n m

∑=1 ∑=1 xij = ∑=1 ai

i j i
n m n
and ∑=1 ∑=1 x ij = ∑=1 bj
j i j
 78

m n
⇒ ∑ ai = ∑ bj
i =1 j =1

m n
The Condition is Sufficient: Let ∑ ai = ∑ bj = k (say).
i=1 j =1

If x ij = λ i bj for all i and j, where λ i ≠ 0 is any real number, then

n n n

∑ x ij = ∑ λ ib j = λ i ∑ b j = k λ i
j =1 j =1 j =1
n

∑=
1 a
⇒ λi = x ij = i
k k
j 1

ai bj
Thus, x ij = λ i bj = , for all i and j.
k
As ai ≥ 0, bj ≥ 0 so x ij ≥ 0 for all i and j.
Hence, a feasible solution exists.

Theorem 2: Out of (m + n) equations, there are only m + n − 1 independent equations

in a transportation problem, m and n being the number of origins and destinations
and any one equation can be dropped as the redundant equation.
Proof: Consider m row equations and n − 1 column equations of the transportation problem
as:
n

∑
j =1
x ij = ai, i = 1, 2, … , m ...(1)

m
and ∑=1 x ij = bj, j = 1, 2, … , n − 1 ...(2)
i

Now, adding m origin constraints given in (1), we get

m n m

∑=1 ∑=1 x ij = ∑=1 ai ...(3)

i j i

Also, adding (n − 1) destination constraints given in (2), we get

n −1 m n −1

∑=1 ∑=1 x ij = ∑=1 bj ...(4)

j i j

Subtracting (4) from (3), we get

m n n −1 m m n −1

∑=1 ∑=1 x ij − ∑=1 ∑=1 x ij = ∑=1 ai − ∑=1 bj

i j j i i j
 79

m  n n −1  n n −1
 
or ∑ ∑ 

x ij − ∑
x ij =
 j =1
∑bj − ∑ bj [∵ Σai = Σbj]
i =1  j =1 j =1  j =1
m
or ∑ x in = bn , which is the nth destination-constraint.
=1
i

It follows that if m + n − 1 constraints are satisfied then the (m + n)th constraint will be
automatically satisfied due to the condition Σai = Σbj. Thus we have only (m + n − 1)
linearly independent equations. Out of (m + n) equations, one (any) is redundant.
Hence, the theorem is proved.

Note:
It indicates that a B.F.S. will contain atmost m + n − 1 positive variables, others being
zero.

3.2.3 Existence of an Optimal Solution

Theorem 3: There always exists an optimal solution to a balanced transportation
problem.
m n
Proof: We have ∑=1 ai = ∑=1 bj
i j

⇒ feasible solution exists of the problem i.e., x ij ≥ 0 for all i and j.

From the constraints of the problem each x ij ≤ min. (ai, bj).
Thus, 0 ≤ x ij ≤ min (ai, bj) i.e., the feasible region of the problem is non-empty, closed and
bounded.
Hence, there exists an optimal solution.

3.3 Methods to Solve Transportation Problem

[B.B.A. (Meerut) 2004]
The solution procedure for the transportation problem consists of two parts:
1. Finding the initial basic feasible solution.
2. Optimization of the initial basic feasible solution obtained in part (i) There are
three techniques to find the initial basic feasible solution. The solution from these
techniques may not be optimal. The three techniques are:
(i) North-west corner cell method
(ii) Least cost entry method
(iii) Vogal’s approximation method (VAM) or penalty method.
These techniques are mentioned in order of their solution accuracy i.e., the VAM
method gives the most accurate solution than the above two methods.
 80

3.3.1 Algorithm for North-West Corner Cell Method

[B.B.A. (Meerut) 2008]
Step 1: Find the minimum of the supply and demand values with respect to the current
north-west corner cell of the cost matrix.

Step 2: Allocate this minimum value to the current north-west cell and subtract this
minimum from the supply and demand value of the current north-west corner cell.

Step 3: Check whether exactly one of the row or column correspondence to north-west
corner cell has zero supply or demand, respectively. If so, go to step 4 else go to step 5.

Step 4: Delete that row or column of the current north-west cell which has zero supply or
demand and go to step 6.

Step 5: Delete both the row and column with respect to the current north-west corner
cell.

Step 6: Check whether exactly one row or column is left out. If yes, go to step 7, else go to
step 1.

Step 7: Match the supply or demand of that row or column with the remaining demands
or supplies of the undeleted columns or rows.

Step 8: Go to phase 2 for optimization of solution obtained above.

3.3.2 Transportation Problem as L.P. Problem

[B.B.A. (Meerut) 2006]
Simplex algorithm can be used to solve any linear programming model. But this
algorithm is lengthy and laborious, due to their specialized structure wherever possible,
we try to simplify the calculations. One such model requiring simplified calculations is
called transportation model. The model is applicable to the sub-class of linear programming
problems in which resources and requirements are expressed in terms of only one kind of
unit. The name of this model is derived from transport to which it was first applied. This
model can be used for machine assignment, plant location, product mixed problems and
many others, so that the model is really not confined to transportation or distribution only.

3.3.3 Objective of Transportation Problem

[B.B.A. (Meerut) 2002, 2004, 2006]
The transportation problem deals with the transportation of a single product from
several sources to several destinations. Transportation models which involve a number of
shipping sources and a number of destinations, each shipping source has a certain
capacity and each destination has a certain requirement associated with a certain cost of
shipping from the sources to the destinations. The object is to minimize the cost of
transportation while meeting the requirements at the destinations.
The distinct feature of transportation problems is that sources and jobs must be expressed
in terms of only one kind of unit.
 81

S olved E xamples
Example 1: Consider the following transportation problem and obtain basic feasible solution using
north-west corner cell method.

Destination

1 2 3 4 Supply

1 3 1 7 4 300

Source 2 2 6 5 9 400

3 8 3 3 2 500

Demand 250 350 400 200 1200

[B.C.A. (Avadh) 2008]

Solution: In the given table, the sum of supplies and the sum of demands are equal.
Hence, the given transportation problem is a balanced one. We see that the supply and
demand values corresponding to the north-west corner cell (1, 1) are 300 and 250
respectively. The minimum of these value is 250. Hence, we allocate 250 units to the cell
(1, 1) and subtract the same from the supply and demand of the cell (1, 1). Now, the
supply to the destination 1 is fully met.

Hence, delete this column and the resultant data is shown in table:

Destination

1 2 3 4 Supply

1 250

3 1 7 4 50

Source 2 6 5 9 400

8 3 3 2 500

Demand 0 350 400 200 950

 82

Now, deleting column 1, then

Destination

2 3 4 Supply

50

1 7 4 0

Source 6 5 9 400

3 3 2 500

Demand 300 400 200 900

The supply and demand values corresponding to the north-west corner cell are 50 and
350 respectively. The minimum of these is 50. Hence, we allocate 50 units to the cell (1, 2)
and subtract the same from the supply and demand values of the cell (1, 2). In this
process, the supply of row 1 is fully exhausted. Hence, we delete this row and the
resultant data.

Destination

2 3 4 Supply

2 300

Source 6 5 9 100

3 3 2 500

Demand 0 400 200 600

The supply and demand values corresponding the north-west corner cell are 400 and 300
respectively. We allocate 300 to the cell (2, 2) and subtract the same from the supply and
demand values of the cell (2, 2). Now, the demand of destination 2 is fully met. Hence,
we delete this column and the resultant data.
 83

Again, the supply and demand values to the north-west corner cell are 100 and 400. We
allocate 100 to the cell (2, 3) and subtract the same from the supply and demand values
of the cell (2, 3).
Destination

3 4 Supply

2 100

Source 5 9 0

3 2 500

Demand 300 200 500

In this process, the supply of source 2 is fully exhausted. Hence, we delete this row and
resultant data.
Destination

3 4

300 200

Source 3 2 0

0 0

In this table, only one source is left out. Hence, the demands of the destination 3 and 4
need to be matched with the supply of source 3.

The initial basic feasible solution for the given problem is shown in the following table:

Destination

1 2 3 4 Supply

1 250 50

3 1 7 4 300

Source 2 300 100

2 6 5 9 400

3 300 200

8 3 3 2 500

Demand 250 350 400 200 1200

 84

A basic cell is the one which has a positive allocation. Thus, total cost
= 3 × 250 + 1 × 50 + 6 × 300 + 5 × 100 + 3 × 300 + 2 × 200
= 750 + 50 + 1800 + 500 + 900 + 400
= ` 4,400 .

Example 2: Find the initial basic feasible solution of the following transportation problem.
To
w1 w2 w3 Available
F1 2 7 4 5
From F2 3 3 1 8
F3 5 4 7 7
F4 1 6 2 14
Requirement 7 9 18 34
[B.C.A. (Bhopal) 2006, 2012]

Solution:
w1 w2 w3 Available
F1 5
2 7 4 5
F2 2 6
3 3 1 8
F3 3 4
5 4 7 7
F4 14
1 6 2 14
Requirement 7 9 18 34

Thus, Total cost = 2 × 5 + 3 × 2 + 3 × 6 + 4 × 3 + 7 × 4 + 2 × 14 = 102.

= 10 + 6 + 18 + 12 + 28 + 28
= ` 102
The method of this example is explained in example (1) and in algorithm for north-west
corner cell.

3.3.4 Algorithm for Least Cost Cell Method [B.B.A. (Meerut) 2008]

Step 1: Find the minimum of the values or matrix minimum in the cost matrix.

Step 2: Find the minimum of the supply and demand values with respect to the cell
corresponding to the minimum value obtained in step 1.
 85

Step 3: Allocate the minimum of the supply and demand to the cell having matrix
minimum. Also subtract the same from the supply and demand values corresponding to
the cell with matrix minimum.
Step 4: Check whether exactly one of the row or column corresponding to the cell with
matrix minimum has zero supply or zero demand respectively. If yes, go to step 5 else go
to step 6.
Step 5: Delete the row or column with respect to the cell with matrix minimum which
has the zero supply or zero demand and go to step 7.
Step 6: Delete both the row and the column with respect to the cell with the matrix
minimum.
Step 7: Check whether exactly one row or column is left out. If yes, go to step 8, else go to
step 1.
Step 8: Match the supply or demand of that row or column with the remaining demands
or supplies of the undeleted columns or rows.
Step 9: Go to phase 2 for optimization of the solution obtained above.

Example 3: Consider the following transportation problem and obtain the initial basic feasible
solution using least cost cell method.
Destination
1 2 3 4 Supply
1 3 1 7 4 300
Source 2 2 6 5 9 400
3 8 3 3 2 500
Demand 250 350 400 200 1200

Solution: The matrix minimum belongs to the cell (1, 2) and the corresponding supply
and demand values are 300 and 350 respectively. We allocate minimum of these two in
the cell (1, 2) and subtract the same from the supply and demand corresponding to the
cell (1, 2), so we have
Destination
1 2 3 4 Supply
1 300
3 1 7 4 0
Source 2
2 6 5 9 400
3
8 3 3 2 500
Demand 250 50 400 200 900
 86

In this process, the supply of source 1 is completely exhausted. Hence, row 1 is deleted. Then,
Destination

1 2 3 4 Supply

2 250

2 6 5 9 150

8 3 3 2 500

Demand 0 50 400 200 650

Again, the matrix minimum is 2 which occur at the cell (2, 1) and (3, 4). The cell (2, 1) is
randomly selected for allocation. The supply and demand values corresponding to cell
(2, 1) are 400 and 250 respectively. The minimum of these values is 250. Hence allocate
250 to cell (2, 1) and subtract the same from the supply and demand of the corresponding
cell. In this process the demand of destination is completely met. Therefore, we delete
thus column from the above table. Then,
Destination

2 3 4 Supply

Source 6 5 9 150

3 200

8 3 2 300

Demand 50 400 0 450

Again, the matrix minimum is 2 at the cell (3, 4). The supply and the demand values
corresponding to this cell are 500 and 200 respectively. We allocate the minimum of the
two i.e., 200 to the cell (3, 4) and subtract the same from the supply and demand values
of the cell (3, 4). In this process the demand of destination 4 is fully met. After deleting
this column, we get
Destination

2 3 Supply

Source 6 5 150

3 300

3 3 0

Demand 50 100 150

 87

Now, the matrix minimum is 3 which occurs at cells (3, 2) and (3, 3). The cell (3, 3) is
randomly chosen for allocation. The supply and demand values corresponding to this cell
are 300 and 400 respectively. The minimum of these two is 300. Hence, we allocate 300
to the cell (3, 3) and subtract the same from the supply and demand values of the cell
(3, 3). In this process the supply of source 3 is fully exhausted. Hence, this row is deleted.
Then,
2 3 Supply
50 100
2
6 5 0

Demand 0 0 0

In this table only one source is felt out. Therefore, we should match the demands of
destination 2 and 3 with supply of source 2. We allocate 100 and 50 to the cells (2, 3)
and (2, 2) respectively. Hence, initial basic feasible solution using L.C. cell method.
Destination
1 2 3 4 Supply
1 300
3 1 7 4 300
Source 2 250 50 100
2 6 5 9 400
3 300 200
8 3 3 2 500
Demand 250 350 400 200 1200

The total cost is calculated by adding the products of the cost of transportation per unit
in each cell and the corresponding number of units allocated to it shown below:
Total cost = 1 × 300 + 2 × 250 + 6 × 50 + 5 × 100 + 3 × 300 + 2 × 200
= 300 + 500 + 300 + 500 + 900 + 400
= ` 2900

3.3.5 Algorithm for Vogal's Approximation Method

[B.B.A. (Meerut) 2008]
Step 1: Find row penalties i.e., the difference between the first minimum and second
minimum in each row. If the two minimum values are equal, then the row penalty is zero.
Step 2: Find column penalties i.e., the difference between first minimum and second
minimum in each column. If two minimum values are equal, then the column penalty is
zero.
 88

Step 3: Find the maximum amongst row penalties and column penalties and identify
whether it occurs in a row or in column. If the maximum penalty is in row go to step 4
otherwise go to step 7.

Step 4: Identify the cell for allocation which has the least cost in that row.

Step 5: Find the minimum of the supply and demand values with respect to the selected
cell.

Step 6: Allocate this minimum values to that cell and subtract this minimum from the
supply and demand values with respect to the selected cell and go to step 10.

Step 7: Identify the cell for allocation which has the least cost in that column.

Step 8: Find the minimum of the supply and demand values with respect to the selected
cell.

Step 9: Allocate this minimum value to the selected cell and subtract this minimum from
the supply and demand values with respect to the selected cell.

Step 10: Check whether exactly one of the rows and the columns corresponding to the
selected cell has zero supply or demand, respectively if yes, go to step 11, otherwise go to
step 12.

Step 11: Delete the row or column which has zero supply or demand and revise the
corresponding row or column penalties. Then go to step 13.

Step 12: Delete both the row and the column with respect to the selected cell. Then,
revise the row and the column penalties.

Step 13: Check whether exactly one row or column is left out, if yes, go to step 14,
otherwise go to step 3.

Step 14: Match the supply or demand of the left out row or column with remaining
demand or supplies of the undeleted columns or rows.

Step 15: Go to phase 2 for the optimization of the solution obtained above.

Example 4: Consider the following transportation problem and obtain the initial basic feasible
solution using Vogal’s approximation method (VAM).

Destination

1 2 3 4 Supply

1 3 1 7 4 300

Source 2 2 6 5 9 400

3 8 3 3 2 500

Demand 250 350 400 200 1200

 89

Solution: First, find row penalties and column penalties. The maximum of these is 3
which is in row 2. Hence, the cell with the least cost in the row 2 is to be identified. This
occurs at the cell (2, 1). The supply and demand values corresponding to the cell (2, 1)
are 400 and 250 respectively. The minimum of these two is 250. Therefore, 250 units are
allocated to the cell (2, 1) and the same is subtracted from the supply and demand values
of the cell (2, 1).

Destination

1 2 3 4 Supply Penalty

3 1 7 4 300 2

Source 2 250

2 6 5 9 150 3

8 3 3 2 500 1

Demand 0 350 400 200 950

Penalty 1 2 2 2

In this process, the demand at destination 1 is fully met. Hence, this column is deleted
and the resultant data is shown in next table.

Destination

2 3 4 Supply Penalty

1 300

1 7 4 0 3

Source 2

6 5 9 150 1

3 3 2 500 1

Demand 50 400 200 650

Penalty 2 2 2
 90

As the column 1 deleted, we again revise the penalty. The maximum of these penalties is
3 which is in row 1. The cell with least cost is identified in row 1 as (1, 2). The supply and
demand value corresponding to cell (1, 2) are 300 and 350 respectively. The minimum of
these values is 300. Hence, we allocate 300 units to cell (1, 2). Thus, the supply at the
row 1 is fully met. Hence, we delete row 1 and the resultant data is shown is next table.

2 3 4 Supply Penalty

6 5 9 150 1

3 200

3 3 2 300 1

Demand 50 400 0 450

Penalty 3 2 7

Similarly, working in the same way, we find that the column 4 is deleted and the resultant
data is shown in next table.

2 3 Supply Penalty

6 5 150 1

3 50

3 3 250 0

Demand 0 400 400

Penalty 3 2

Similarly, we find that column 2 is deleted and only one column is left out. The supplies
of the source 2 and 3 are matched with the demand of the destination 3 as follows:

3 Supply

150
2
5 150

250
3
3 250

Demand 400 400

 91

The initial basic feasible solution of the problem using VAM is as follows:
Destination
1 2 3 4 Supply
1 300
3 1 7 4 300
Source 2 250 150
2 6 5 9 400
3 50 250 200
8 3 3 2 500
Demand 250 350 400 200 1200

The total cost of the solution is obtained by adding the products of the cost of transportation
per unit in the each and every basic all and the corresponding units allocated to it. Thus,

Total cost = 1 × 300 + 2 × 250 + 3 × 50 + 5 × 150 + 3 × 250 + 2 × 200

= 300 + 500 + 150 + 750 + 750 + 400

= ` 2850

Thus, we can see that VAM provides the optimal transportation cost when the same is
compared with other two methods.

Example 5: Explain the transportation problem. What is the difference between a transportation
and assignment problem?
Four factories ( A, B, C, D) supply the requirements of three warehouses ( E, F, G) . The availability
at the factories, the requirement of the warehouses and the unit transportation costs are presented in
the following table:
Warehouse
E F G Available
A 10 8 9 15
B 5 2 3 20
Factory C 6 7 4 30
D 7 6 8 35
Required 25 26 49 100

Find an initial basic feasible solution of the transportation problem by using:

(i) North-West corner rule
(ii) Matrix minimum method

(iii) Vogal’s Approximation method. [B.B.A. (Meerut) 2004]

 92

Solution: (i) By North-West Corner Rule.

E F G Available

A 15
10 8 9 15

B 10 10
5 2 3 20

C 16 14
6 7 4 30

D 35
7 6 8 35
Required 25 26 49 100

Total transportation cost = 10 × 15 + 5 × 10 + 2 × 20 + 7 × 16 + 4 × 14 + 8 × 35

= 150 + 50 + 40 + 112 + 56 + 280

= ` 688

Note:
The process is explained in above example and algorithm.

(ii) By Matrix-Minimum Method.

E F G Available

A 15

10 8 9 15

B 20

5 2 3 20

C 30

6 7 4 30

D 25 6 4

7 6 8 35

Required 25 26 49 100

Total transportation cost = 2 × 20 + 9 × 15 + 4 × 30 + 7 × 25 + 6 × 6 + 8 × 4

= 40 + 135 + 120 + 175 + 36 + 32

= ` 538
 93

Note:
The process is explained in above example and algorithm.

(iii) By Vogal’s Approximation Method.

E F G Available Penalty
A
10 8 9 15 1
B 20
5 2 3 20 1
C
6 7 4 30 2
D
7 6 8 35 1
Required 25 26/6 49 100
Penalty 1 4 1

Reduced Matrix is
E F G Available Penalty
A
10 8 9 15 1
C 30
6 7 4 30 2
D
7 6 8 35 1
Required 25 6 49/19 80
Penalty 1 1 4

Reduced Matrix is
E F G Available Penalty
A
10 8 9 15 1
D 25
7 6 8 35/10 1
Required 25 6 19 50
Penalty 3 2 1
 94

Reduced Matrix is
F G Available Penalty

8 9 15 1

D 6

6 8 10/4 2

Required 6 19 25

Penalty 2 1

Again reduced Matrix is

G Available Penalty

A 15 15 9

D 4

8 4 8

Penalty 1

Hence,
E F G Available

A 15

10 8 9 15

B 20

5 2 3 20

C 30

6 7 4 30

D 25 6 4

7 6 8 35

25 26 49 100

The total transportation cost = 9 × 15 + 2 × 20 + 4 × 30 + 7 × 25 + 6 × 6 + 8 × 4

= 135 + 40 + 120 + 175 + 36 + 32

= ` 538
 95

Example 6: Determine an initial B.F.S. to the following transportation table using Vogel’s
approximation method.
Destination

D1 D2 D3 D4 Supply

O1 1 2 1 4 30

Origin O2 3 3 2 1 50

O3 4 2 5 9 20

Demand 20 40 30 10 100
[B.C.A. (Lucknow) 2009; B.C.A. (Indore) 2010; B.B.A. (Delhi) 2004; (Meerut) 2008]

Solution: First, we write cost and requirement matrix and compute the penalties which
is the difference between lowest and second lowest element in each row and each column

D1 D2 D3 D4 Penalties (P)

O1 1 2 1 4 30 1

O2 3 3 2 10 (1) 50 − 10 1

O3 4 2 5 9 20 2

20 40 30 10 − 10 90

Penalties (P) 2 1 1 3
↑

Since maximum penalty 3 is in column 4th, then allocate maximum possible amount 10
in the cell (1, 4) having minimum cost i.e., x14 = 10. Thus the requirements of 4th column
are completed, so leaving column 4th then, reduced matrix as follows:

D1 D2 D3 Penalties (P)

O1 1 2 1 30 1

O2 3 3 2 40 1

O3 4 20 (2) 5 20 − 20 2 ←

20 40 − 20 30 70

Penalties (P) 2 1 1

Since maximum penalty 2 is in Ist column and IIIrd row, so we select one of them, say
row 3rd and allocate maximum possible amount 20 in the cell (3, 2) having minimum cost
i.e., x32 = 20.
 96

Thus, the requirements of 3rd row are completed, so leaving row 3rd then reduced matrix
as follows:

D1 D2 D3 Penalties (P)

O1 20 (1) 2 1 30 − 20 1

O2 3 3 2 40 1

20 − 20 20 30 50

Penalties (P) 2 1 1
↑

Since maximum penalty 2 is Ist column and allocate maximum possible amount 20 in the
cell (1, 1) having minimum cost i.e., x11 = 20.

Thus, the requirements of Ist column are completed, so leaving column Ist then reduced
matrix as follows:
D2 D3 Penalties (P)

O1 2 10 (1) 10 − 10 1 ←

O2 20 (3) 20 (2) 40 1

20 30 40

Penalties (P) 1 1

The maximum penalty 1 which is in Ist, IInd rows and I, II column then select any i.e. we
select Ist row and allocating maximum possible amount 10 in row Ist at (1, 3), we complete
the remaining allocation.

Thus, the final table giving initial B.F.S. is as follows:

Destination

D1 D2 D3 D4

O1 20 (1) 2 10 (1) 4 30

Origin O2 3 20 (3) 20 (2) 10 (1) 50

O3 4 20 (2) 5 9 20

20 40 30 10 100

The total minimum cost according to above transportation problem

= 20 × 1 + 10 × 1 + 20 × 3 + 20 × 2 + 10 × 1 + 20 × 2
= 20 + 10 + 60 + 40 + 10 + 40
= ` 180
 97

Example 7: Use north-west corner rule to determine an initial basic feasible solution to the following
transportation problem:

To

I II III IV Supply

A 13 11 15 20 2

From B 17 14 12 13 6

C 18 18 15 12 7

Demand 3 3 4 5 15
[B.B.A. (Rohilkhand) 2006; M.C.A. (Meerut) 2007]

Solution: This is the balance transportation problem as Σ ai = Σ bi = 15.

Start with cell (1, 1) and allocate it the maximum amount thus x11 = 2. Since minimum of
a1 = 2 and b1 = 3 is 2. Now no amount is left at source A, so we move vertically downwards
to the cell (2, 1) and allocate it as much as possible. Since column I still needs amount 1
and the amount 6 is available at source B in row 2, so we allocate the maximum amount 1
to the cell (2, 1) . Thus, allocation 3 for column 1 is complete. Then move to right of the
cell (2, 1) i.e., at cell (2, 2) and allocate here as much as possible. Since the amount 5 is still
available in row 2 and amount 3 is available in column 2, so we can allocate maximum
amount 3 in the cell (2, 2) i.e., x22 = 3. Since the allocation 3 in column 2 is also completed.
So we move to the cell (2, 3) on right of the cell (2, 2) . Since amount 2 is still available in
row 2 and amount 4 is available in column 3, so we allocate maximum amount 2 in cell
(2, 3) i.e., x23 = 2. Now no amount is available in row 2, so we move downwards to the cell
(3, 3) and allocate there are much amount as possible. Since the amount 2 is still required
in column 3 and amount 7 is available in row 3, so the amount 2 is allocated in the cell
(3, 3) i.e., x33 = 2 . Thus, the allocation 4 in the column 3 is complete. Now the amount 5
is still available in the row 3 and also the amount 5 is required in column 5, so the
amount 5 is allocated in the cell (3, 4) i.e., x34 = 5 which complete the allocation.
I II III IV bj

A 2 (13) 2

B 1 (17) 3 (14) 2 (12) 6

C 2 (15) 5 (12) 7

ai 3 3 4 5 15

The total minimum cost according to above transportation problem

= 2 × 13 + 1 × 17 + 3 × 14 + 2 × 12 + 2 × 15 + 5 × 12
= 26 + 17 + 42 + 24 + 30 + 60
= 199 units
 98

Example 8: Determine an initial basic feasible solution to the following transportation problem by
using north-west corner rule.

Destination

D1 D2 D3 D4 Supply

S1 21 16 15 3 11

Source S2 17 18 14 23 13

S3 32 27 18 41 19

Demand 6 10 12 15 43

[B.C.A. (Agra) 2008; B.B.A. (Meerut) 2002]

Solution:

Destination

D1 D2 D3 D4 Supply

S1 6 (21) 5 (16) 15 3 11

Source S2 17 5 (18) 8 (14) 23 13

S3 32 27 4 (18) 15 (41) 19

Demand 6 10 12 15 43

We start from N-W. corner rule which is corresponding to cell (1, 1), we allocate maximum
to units to this cell. Thus, allocation of 6 units leaves the surplus amount of 5 units for
the first row. Now 5 units are allocated to the cell C (1, 2). Now, the allocation of first row
and first column is complete. Now, we move to cell (2, 2) and allocate 5 units to this cell.
Now, move to cell (2, 3) and allocate 8 units to this cell. So that demand of second row is
satisfied. Now 4 units are required by third column, so we allocate 4 units to cell (3, 3)
and 15 units to cell (3, 4).

The total minimum cost according to above problem

= 6 × 21 + 5 × 16 + 5 × 18 + 8 × 14 + 4 × 18 + 15 × 41

= 126 + 80 + 90 + 112 + 72 + 615

= ` 1095.
 99

Example 9: Find the initial basic feasible solution by least cost method.

Warehouse

W1 W2 W3 W4 Supply

F1 48 60 56 58 140

Factories F2 45 55 53 60 260

F3 50 65 60 62 360

F4 52 64 55 61 220

Demand 200 320 250 210 980

Solution: Since the total demand is equal to total supply so it is balanced transportation
problem.

W1 W2 W3 W4 Supply

F1 48 60 56 140 (58) 140

F2 200 (45) 55 60 (53) 60 260

F3 50 320 (65) 60 40 (62) 360

F4 52 64 190 (55) 30 (61) 220

Demand 200 320 250 210 980

The least cost is 45 to the cell (2, 1), thus we allocate maximum of 200 units to this cell.
We proceed for next least cost which is 53 corresponding to cell (2, 3), we allocate only
60 units to this cell as 200 units have already been allocated to (2, 1) cell. Now the
allocation of IInd row is complete. We proceed for next least cost which is 55
corresponding to (4, 3) cell, we allocate 190 units to this cell as 60 units have already been
allocated to cell (2, 3) . We proceed to next least cost 58 corresponding to (1, 4) cell, we
allocate maximum 140 units to this cell, the allocation of the F1 row is complete. This
process is repeated again and again till all demands and supply is not complete.

Total minimum cost according to above problem

= 45 × 200 + 58 × 140 + 65 × 320 + 62 × 40 + 55 × 190 + 61 × 30 + 60 × 53

= 9000 + 8120 + 20, 800 + 2, 480 + 10450 + 1830 + 3180

= ` 55, 860
 100

3.4 Transportation Algorithm (MODI) Method

Various steps involved in solving any transportation problem may be summarized in the
following iteration procedure.

Step 1: Find the initial basic feasible solution by using any of the three methods.

Step 2: Check the number of occupied cells. If these are less than m + n − 1, there exists
degeneracy and we introduce a very small positive assignment of ∈ (∈→ 0) in suitable
independent positions, so that the number of occupied cells is exactly equal to m + n − 1.

Step 3: For each occupied cell in the current solution, solve the system of equations

ui + v j = cij

Starting initially with some ui = 0 or v j = 0 and entering successively the values of ui and v j
in the transportation table margins.

Step 4: Compute the net evaluations dij = (ui + v j) − (cij) for all unoccupied basic cells and
enter them in the upper right corner of the corresponding cells.

Step 5: Examine the sign of each dij if all dij ≤ 0, then the current basic feasible solution is
an optimum one. If at least one dij ≥ 0 select the unoccupied cell having the largest
positive net evaluation to enter the basis.

Step 6: Let the unoccupied cell (r , s) enter the basis. Allocate an unknown quantity say θ
to the cell (r , s) . Identify a loop that starts and enters at the cell (r , s) and connects some of
the basic cells. Add and subtract interchangeably θ to and from the transition cells of the
loop in such a way that the rim requirements remain satisfied.

Step 7: Assign a maximum value θ in such a way that the value of one basic variable
becomes zero and other basic variables remain non-negative. The basic cell whose
allocation has been reduced to zero leaves the basis.

Step 8: Return the step 3 and repeat the process untill an optimum basic feasible solution
has been obtained.

Note:
1. If at least one dij = 0, there exists an alternative optimum solution.
2. For maximization transportation problems, replace each element of the
transportation table by its difference from the maximum element of the table.
Then apply the steps of minimization transportation problem on the revised
transportation table.
 101

Example 10: Solve the following transportation problem. [B.B.A. (Meerut) 2002]

To

A B C Available

I 50 30 220 1

From II 90 45 170 3

III 250 200 50 4

Requirement 4 2 2

Solution: Using Vogel's approximation method and initial basic feasible solution is
obtained as follows:

A B C Available

I 1

50 30 220 1

II 3

90 45 170 3

III 2 2

250 200 50 4

Requirement 4 2 2 8

The Basic feasible solution (B.F.S.) = 1 × 50 + 3 × 90 + 2 × 200 + 2 × 50

= 50 + 270 + 400 + 100

= 820.

Now, to test the optimality by Modi method.

The number of basic cells = m + n − 1.

where m = number of rows, n = number of columns.

 102

But m = 3, n = 3, m + n − 1 = 3 + 3 − 1 = 5 ≠ not equal to number of basic cells which is 4.

Then, it is the problem of degeneracy, we have allocated a small quantity ∈ (> 0) in the
cell (3, 1) to overcome degeneracy as the number of occupied cells was only 4. Using
Modi method an optimal solution is obtained as:

ui

1 (−ve) (−ve)

50 30 220 −200

3 (−ve) (−ve)

90 45 170 −160

∈ 2 2

250 200 50 0

vj 250 200 50

where ui and v j is obtained by ui + v j = cij

Let u3 = 0 then, u3 + v3 = 50

0 + v3 = 50, v3 = 50

u3 + v1 = 250, u3 + v2 = 200, but u3 = 0

v1 = 250, v2 = 200, u1 + v1 = 50, u2 + v1 = 90

u1 + 250 = 50, u2 + 250 = 90

⇒ u1 = − 200 u2 = − 160

The net evaluation for each of the unoccupied cells are now determined
d12 = u1 + v2 − C12 = − 200 + 200 − 30 = − 30
d13 = u1 + v3 − C13 = − 200 + 50 − 220 = − 370
d22 = u2 + v2 − C22 = − 160 + 200 − 45 = − 5
d23 = u2 + v3 − C23 = − 160 + 50 − 170 = − 280

These are obtained by dij = (ui + v j − cij)

some author may use dij = cij − (ui + v j) (then all the net evaluation are positive)

Since all the net evaluations are non-positive, the current solution is an optimum one.
The optimum allocation is given by x11 = 1, x22 = 3, x31 = ∈, x32 = 2 and x33 = 2.

The transportation cost according to the above route is given by

= 1 × 50 + 3 × 90 + 2 × 200 + 2 × 50 + ∈ × 250
= 50 + 270 + 400 + 100 + 250 ∈
= ` 820 + 250 ∈ = ` 820 as ∈→ 0
 103

Example 11: Find the optimal solution of transportation problem.

Destination

1 2 3 4 Supply

1 3 1 7 4 300

Source 2 2 6 5 9 400

3 8 3 3 2 500

Demand 250 350 400 200 1200

Solution: First, find the B.F.S. by north-west corner cell method.

First, find the B.F.S. by north-west cell method.

Destination

1 2 3 4 Supply

1 250 50 300

3 1 7 4

Source 2 300 100 400

2 6 5 9

3 300 200 500

8 3 3 2

Demand 250 350 400 200 1200

The B.F.S. by north-west corner rule

= 3 × 250 + 1 × 50 + 6 × 300 + 5 × 100 + 3 × 300 + 2 × 200

= 750 + 50 + 1800 + 500 + 900 + 400

= ` 4, 400

Now, to test the optimality by Modi method. The number of basic cells is equal to
m + n − 1, where m is the number of sources and n is the number of destinations column.
Hence, m = 3, n = 4.

∴ m + n − 1 = 3 + 4 − 1 = 6 = number of basic cells entry.

 104

Now, to find ui and v j by cij = ui + v j

i.e. u1 + v1 = 3, u1 + v2 = 1, u2 + v2 = 6, u2 + v3 = 5

u3 + v3 = 3, u3 + v4 = 2, let u1 = 0 then

v1 = 3, u2 = 5, v3 = 0, v4 = − 1, v2 = 1, u3 = 3

Then, the penalty for each of the non-basic cells is computed using the following formula
and summarized within the squares at the bottom left corners of the respective cell

dij = ui + v j − cij

Destination
ui
250 50

3 − + 7 4 0 u1
2 − ve − ve

Source 300 100

2 + − 6 5 9 5 u2

6 − ve

8 3 300 200 3 u3

− ve 1 3 2

3 1 0 −1
vi v1 v2 v3 v4

d13 = u1 + v3 − C13 = 0 + 0 − 7 = − 7 = − ve

d14 = u1 + v4 − C14 = 0 − 1 − 4 = − 5 = − ve

d24 = u2 + v4 − C24 = 5 + (− 1) − 9 = − 5 = − ve

In the above table if all the dij ≤ 0 then optimality is reached. Otherwise, select the cell
which has most positive penalty. Here, the cell (2, 1) has most positive penalty. So, the
solution can be improved. This non-basic cell is to be converted into a basic cell without
effecting the supply and demand restrictions. Hence, we construct a closed loop starting
from this new basic cell and passing through the basic cell (2, 2), (1, 2) and (1, 1). Then
alternately, the + sign and − sign are assigned in the basic cell on the closed loop
commencing from the new basic cell. The minimum of the existing allocation amongst
the negatively signed cells on the loop is identified, which is 250 in this case. This
 105

minimum allocation is now added to all the positively signed cells on the closed loop but
is subtracted from all the negatively signed cells. The improve table is as follows:

Destination

300 ui

3 1 7 4 0

−ve −ve −ve

250 50 100

2 6− + 5 9 5

Source −ve

300 200
+ −
8 3 3 2 3

−ve +1

vj 3 1 0 −1

Repeated the above process we get another improved table since (3, 2) has positive
penalty.

300 Supply

3 1 7 4 300

−ve −ve −ve

250 150

2 6 5 9 400

−ve −ve

50 250 200 500

8 3 3 2

−ve

Demand 250 350 400 200 1200

 106

It is found that all these penalties are less than or equal to zero. (Written as −ve in block).
Hence, the optimality is reached. Therefore, the total optimal cost
= 1 × 300 + 2 × 250 + 5 × 150 + 3 × 50 + 2 × 200 + 3 × 250

= 300 + 500 + 750 + 150 + 400 + 750

= ` 2850

Example 12: Obtain an initial basic solution to the following transportation problem. Is this solution
on optimal solution? If not, obtain the optimal solution. [B.B.A. (Meerut) 2003, 2004, 2009]

W1 W2 W3 W4 ai

F1 19 30 50 10 7

F2 70 30 40 60 9

F3 40 8 70 20 18

bj 5 8 7 14 34

Solution: The different necessary tables in order to obtain basic feasible solution by
Vogal’s approximation method are as follows. First write cost and required matrix and
compute the penalties which is the difference between lowest and second lowest element
in each row and each column.

W1 W2 W3 W4 ai Penalties (P)

F1 19 30 50 10 7 9

F2 70 30 40 60 9 10

F3 40 8 (8) 70 20 18 − 8 12

bj 5 8 −8 7 14 26

Penalties (P) 21 22 10 10
↑
Since maximum penalties is 22 in column w2 , then allocate maximum possible amount 8
in the cell (3, 2) . Thus, the requirements of w2 column are completed, so leaving the
column w2 then reduce matrix as follows:

W1 W3 W4 P

F1 5 (19) 50 10 7 −5 9

F2 70 40 60 9 20

F3 40 70 20 10 20

5 −5 7 14 21

P 21 10 10
↑
 107

Since maximum penalty is 21 which is in column w1 and allocate maximum possible

amount 5 in cell (1, 1) having maximum cost i.e., x11 = 5. Thus, the requirements of w1 column
are completed so leave the column w1 then reduce matrix as follows:
W3 W4 P
F1 50 10 2 40 W3 W4 P
F2 40 60 9 20 F1 50 2 (10) 2 40
F3 70 10 (20) 10 − 10 50 ← F2 7 (40) 2 (60) 9 20
7 14 − 10 7 4
P 10 10 P 10 50
↑
Thus, the final giving the initial B.F.S. by VAM is as follows:
W1 W2 W3 W4
F1 5 (19) 30 50 2 (10) 7
F2 70 30 7 (40) 2 (60) 9
F3 40 8 (8) 70 10 (20) 18
5 8 7 14 34

The total minimum cost according to above transportation problem

= 5 × 19 + 2 × 10 + 7 × 40 + 2 × 60 + 8 × 8 + 10 × 20
= 95 + 20 + 280 + 120 + 64 + 200
= ` 779

To Test the Optimality: Since m + n − 1 = 3 + 4 − 1 = 6 = number of independent positions

Where m = number of rows
n = number of columns
Now, we determine a set of ui and v j for each occupied cell (r, s) such that
crs = ur + vs
ui
(19) (30) (− 2) (50) (– 10) (10) ↓

5 2 10 (u1)
(32) (60)
(70) (69) (30) (48) (40) (60)
7 2 60 (u2 )
(1) (− 18)
(40) (29) (8) (70) 0 (20)
8 10 20 (u3 )
(11) (70)
vj → 9 − 12 − 20 0
(v1) (v2 ) (v3 ) (v4 )
 108

Take any ui and v j to be zero which has maximum number of allocations i.e., v4 = 0 and
other entry is obtain by crs = ur + ν s .
c14 = 10 = u1 + v4 = u1 + 0, c24 = 60 = u2 + v4 = u2 + 0
c34 = 20 = u3 + v4 = u3 + 0
⇒ u1 = 10, u2 = 60, u3 = 20
c11 = 19 = u1 + v1 = 10 + v1, c23 = 40 = u2 + v3 = 60 + v3
c32 = 8 = u3 + v2 = 20 + v2
⇒ v1 = 9, v3 = − 20, v2 = − 12
Now, we have to find ui + v j for each unoccupied cell and enter at the upper right corner
of the corresponding cell then find dij = cij − (ui + v j) for each unoccupied cell and enter at
the lower right corner of the corresponding cell. By completing all entries, we find
d22 = − 18 < 0 .
Therefore this solution is not optimal. So we proceed to improve this solution. Since
minimum dij = d22 − 18 < 0, we shall allocate to this cell (2, 2) as much as possible. This
allotment is shown in table (a) and table (b).

5 2

+2 7 2 −2

8 10
−2 +2

ui
↓
(19) (30) (− 2) (50) (8) (10)
5 2 0 (u1)
(32) (42)
(70) (51) (30) (40) (60) 42
2 7 32 (u2 )
(19) 18
(40) (29) (8) (70) (18) (20)
6 12 10 (u3 )
(11) (52)
vj → 19 −2 8 10
(v1) (v2 ) (v3 ) (v4 )
 109

For this solution, total transportation cost

= 5 × 19 + 2 × 10 + 2 × 30 + 7 × 40 + 6 × 8 + 12 × 20
= 95 + 20 + 60 + 48 + 280 + 240
= ` 743
For this solution is optimal we again test the optimality till all dij > 0. Therefore, this
solution is optimal as all dij > 0.

Example 13: Solve the transportation problem.

D1 D2 D3 D4 Available
O1 1 2 1 4 30
O2 3 3 2 1 50
O3 4 2 5 9 20
Required 20 40 30 10 100
[B.C.A. (Purvanchal) 2010; B.C.A. (Meerut) 2008, 2010]

Solution: By "Lowest cost entry method', we get by usual process the following B.F.S. of
this problem.
D1 D2 D3 D4
(1) (2) (1) (4)
O1 20 10 30

(3) (3) (2) (1)

O2 20 20 10 50

(4) (2) (5) 9

O3 20 20

20 40 30 10 100

Total transportation cost

= 1 × 20 + 1 × 10 + 3 × 20 + 2 × 20 + 1 × 10 + 2 × 20
= 20 + 10 + 60 + 40 + 10 + 40
= ` 180

To Test the Optimality: Since m + n − 1 = 3 + 4 − 1 = 6 = number of independent positions,

then applying Modi method.

Now, we have to find set of ui and v j by

crs = ur + vs where crs is occupied cell
 110

Taking u2 = 0 as row IInd has maximum number of allocations.

ui
↓
(1) (2) (2) (1) (4) (0)
20 10 − 1 (u1)
(0) (4)

(3) (2) (3) (2) (1)

20 20 10 0 (u2 )
(1)

(4) (1) (2) (5) (1) (9) (0)

20 − 1 (u3 )
(3) (4) (9)

vj → 2 3 2 1
(v1) (v2 ) (v3 ) (v4 )

As c22 = 3 = u2 + v2 = 0 + v2 ⇒ v2 = 3
c23 = 2 = u2 + v3 = 0 + v3 ⇒ v3 = 2
c24 = 1 = u2 + v4 = 0 + v4 ⇒ v4 = 1
c32 = 2 = u3 + v2 = u3 + 3 ⇒ u3 = − 1
c13 = 1 = u1 + v3 = u1 + 2 ⇒ u1 = − 1
c11 = 1 = u1 + v1 = − 1 + u1 ⇒ v1 = 2
For each unoccupied cell, we find cell evaluations ui + v j and enter at the upper right
corner of the corresponding cell. Then we find cell dij = cij − (ui + v j) which is the
difference of upper right corner entry from the upper left corner entry for all unoccupied
cells and enter at the lower right corner of the corresponding cell. We see all dij ≥ 0 then
B.F.S. is also optimal solution i.e., minimum transportation cost = ` 180.

Example 14: The cost-requirement table for the transportation problem is given below:
To

W1 W2 W3 W4 W5 Available

F1 4 3 1 2 6 40

From F2 5 2 3 4 5 30

F3 3 5 6 3 2 20

F4 2 4 4 5 3 10

Required 30 30 15 20 5

Obtain the optimal solution of the problem. [B.B.A. (Delhi) 2004, 2007; (Gwalior) 2008]
 111

Solution: By Vogel's approximation method, an initial B.F.S. of the problem is given in the
following table:
(4) (3) (1) (2) (6)
5 15 20 40

(5) (2) (3) (4) (5)

30 30

(3) (5) (6) (3) (2)

15 5 20

(2) (4) (4) (5) (3)

10 10

30 30 15 20 5

For this solution total transportation cost

= 5 × 4 + 15 × 1 + 20 × 2 + 30 × 2 + 3 × 15 + 5 × 2 + 10 × 2
= 20 + 15 + 40 + 60 + 45 + 10 + 20
= ` 210
Here number of allocations is 7 which is one less than the number m + n − 1 = 8.
So the solution is degenerate solution.

To Test the Solution for Optimality: To make the number of allocations equal to 6 we
allocate a small positive amount ∈ to the cell (2, 3) having minimum cost in empty cells.
To test the solution for optimality, we find the set of ui and v j so that for each occupied
cell (r , s), crs = ur + vs , in the usual manner, by considering u1 = 0.
ui ↓
(4) (3) (0) (1) (2) (6) (3)
5 15 20 0 (u1)
(3) (3)
(5) (6) (2) (3) (4) (4) (5) (5)
30 ∈ 2 (u2 )
(− 1) (0) (0)
(3) (5) (− 1) (6) (0) (3) (1) (2)
15 5 −1 (u3 )
(6) (6) (2)
(2) (4) (− 2) (4) (−1) (5) (0) (3) (1)
10 − 2 (u4 )
(6) (5) (5) (2)
vj → 4 0 1 2 3
v1 v2 v3 v4 v5
 112

Now, in each unoccupied cell we enter the values of ui + v j and dij = cij − (ui + v j) at the
appropriate corners and find that d21 < 0. So this solution is not optimal. Taking the cell
evaluation ∈from cell (2, 3) to (2, 1) and proceeding similarly, we get the following table for
the test of optimality of this solution:

ui ↓

(4) (3) (1) (1) (2) (6) (3)

5 15 20 0 (u1)

(2) (3)

(5) (2) (3) (2) (4) (3) (5) (4)

∈ 30 1 (u2 )

(1) (1) (1)

(3) (5) (0) (6) (0) (3) (1) (2)

15 5 −1 (u3 )

(5) (6) (2)

(2) (4) (−1) (4) (−1) (5) (0) (3) (1)

10 − 2 (u4 )

(5) (5) (5) (2)

vj → 4 1 1 2 3

In this table all dij ≥ 0.

Hence, this solution is optimal. Hence the optimal solution is B.F.S. by VAM
and minimum transportation cost = ` 210.

Example 15: A company has factories A, B and C which supply warehouses at D, E, F and G.
Monthly factory capacities are 160, 150 and 190 units respectively. Monthly warehouse
requirements are 80, 90, 110 and 160 units respectively. Unit shipping costs (in rupees) are as follows:

To

D E F G

A 42 48 38 37

From B 40 49 52 51

C 39 38 40 43

Determine the optimum distribution for company to minimize shipping costs.

[B.C.A. (Kurukshetra) 2010; B.C.A. (Meerut) 2007, 2011]
 113

Solution: Here, total monthly capacities = 160 + 150 + 190 = 500 units and total monthly
requirements = 80 + 90 + 110 + 160 = 440 units.
Since two are not equal, so it is an unbalanced T.P. To convert the problem to a balanced
one, we introduce a fictitious warehouse H with requirement of 60 units and having all
the transportation costs to zero. Thus, the balanced transportation problem is as follows:

To
D E F G H Capacities

A (42) (48) 100 (38) 60 (37) 0 160

From B 80 (40) (49) 10 (52) (51) 60 (0) 150

C (39) 90 (38) (40) 100 (43) (0) 190

Requirements 80 90 110 160 60 500

By Vogel’s approximation method, an initial B.F.S. of the problem is given in the above
table.

For this solution, total transportation cost

= 100 × 38 + 60 × 37 + 80 × 40 + 10 × 52 + 60 × 0 + 90 × 38 + 100 × 43

= 3800 + 2220 + 3200 + 520 + 0 + 3420 + 4300

= ` 17460.

Test of the Solution for Optimality: Here number of allocations is 7 = m + n − 1 (= 7)

and are at independent positions. So, to test this solution for optimality, all necessary
entries are made in the following table. In this table d33 = − 4 < 0, so this solution is not
optimal. Therefore we proceed to modify this solution.

ui ↓
(42) (26) (48) (32) (38) (37) (0) (− 14)
100 − θ 60 + θ − 14 (u1)
(16) (16) (14)

(40) (49) (46) (52) (51) (51) (0)

80 10 60 0 (u2 )
(3) (0)

(39) (32) (38) (40) (44) (43) (0) (− 8)

90 θ 100 − θ − 8 (u3 )
(7) (− 4) (8)

vj → 40 46 52 51 0
v1 v2 v3 v4 v5
 114

Modification of the Solution (First): Since minimum dij = d33 = − 4 < 0, so we allocate
maximum possible allocation θ to this cell (3, 3). See the above table.
Here we observe that minimum allocation containing − θ is 100 − θ. Equating it to zero,
we get θ = 100 .
Thus, making necessary changes in the allocations, the new solution is given in the
following table, in which the number of allocations is 6 which is one less than the number
m + n − 1 (= 7).
So the solution is degenerate. To make this solution non-degenerate, we allocate a small
amount ∈ to the cell (1, 3) .

Total transportation cost for this solution

= 160 × 37 + 80 × 40 + 10 × 52 + 90 × 38 + 100 × 40
= 5920 + 3200 + 520 + 3420 + 4000
= ` 17060

which is less than the cost for previous solution.

ui ↓

(42) (26) (48) (34) (38) (37) (0) (− 14)

∈ 160 − 14 (u1)

(16) (14) (14)

(40) (49) (50) (52) (51) (51) (0)

80 θ 10 − θ 60 0 (u2 )

(− 1) (0)

(39) (28) (38) (40) (43) (39) (0) (− 12)

90 − θ 100 + θ − 12 (u3 )

(11) (4) (12)

vj → 40 50 52 51 0
v1 v2 v3 v4 v5

Now, to test the optimality of this solution all necessary entries are made in the above
table and we observe that d22 = − 1 < 0.
So this solution is also not optimal.
Thus, we proceed to modify this solution also.

Modification of the Solution (Second): Since minimum dij = d22 = − 160, so we allocate
as much as possible to this cell (2, 2). Allocating θ to this cell and making necessary
changes in the other allocations, the minimum allocation containing − θ is 10 − θ,
equating it to zero, we get θ = 10.
 115

Thus, making necessary changes in the allocations, the new solution is given in the table
given below in which the allocations are at independent positions.

In this case the total transportation cost

= 160 × 37 + 80 × 40 + 10 × 49 + 80 × 38 + 110 × 40
= 5920 + 3200 + 490 + 3040 + 4400
= ` 17050
which is less than the cost after first modification.

Now, to test the optimality of this solution all necessary entries are made in this table
and we observe that all dij ≥ 0. Therefore this solution is optimal.

D E F G H ui ↓
(42) (27) (48) (36) (38) (37) (0) (− 13)
A ∈ 160 − 13 (u1)
(15) (12) (13)
(40) (49) (52) (51) (51) (50) (0)
B 80 10 60 0 (u2 )
(1) (1)
(39) (29) (38) (40) (43) (39) (0) (− 11)
C 80 110 − 11 (u3 )
(10) (4) (11)
vj → 40 49 51 50 0
v1 v2 v3 v4 v5

Hence, the optimal solution to the problem is

x14 = 160, x21 = 80, x22 = 10, x32 = 80, x33 = 110.
60 units are left with factory B which cannot be despatched as all requirements have been
fulfilled. In this case minimum transportation cost = ` 17050.

Example 16: Solve the following transportation problem (cell entries represent unit cost).

To Available

5 3 7 3 8 5 3

5 6 12 5 7 11 4
From
2 1 2 4 8 2 2

9 6 10 5 10 9 8

Requirement 3 3 6 2 1 2 17
[B.C.A. (Rohtak) 2011; B.B.A. (Raipur) 2008]
 116

Solution: By Vogel's approximation method, an initial B.F.S. of the problem is given in

the following table:
(5) (3) (7) (3) (8) (5)
1 2 3

(5) (6) (12) (5) (7) (11)

3 1 4

(2) (1) (2) (4) (8) (2)

2 2

(9) (6) (10) (5) (10) (9)

2 4 2 8

3 3 6 2 1 2

The total minimum transportation cost of given problem

= 1 × 3 + 2 × 5 + 3 × 5 + 1 × 7 + 2 × 2 + 2 × 6 + 4 × 10 + 2 × 5
= 3 + 10 + 15 + 7 + 4 + 12 + 40 + 10
= ` 101

To Test the Solution for Optimality: Here number of allocations is 8 which is one less
than m + n − 1 = 4 + 6 − 1 = 9. m = number of rows, n = number of columns.
So this solution is degenerate solution.
∴ To resolve degeneracy we allocate a small amount ∈ to the cell (3, 6) at independent
position. We cannot take ∈in cell (3, 2) having minimum cost among unoccupied cells, as
it is not the independent position.
ui ↓
(5) (7) (3) (7) (7) (3) (2) (8) (9) (5)
1 2 − 3 (u1)
(− 2) (0) (1) (− 1)
(5) (6) (1) (12) (5) (5) (0) (7) (11) (3)
3 1 − 5 (u2 )
(5) (7) (5) (8)
(2) (1) (− 2) (2) (4) (− 3) (8) (4) (2) (0)
∈ 2 − 8 (u3 )
(3) (7) (4) (2)
(9) (10) (6) (10) (5) (10) (12) (9) (8)
2 4 2 0 (u4 )
(−1) (− 2) (1)
vj → 10 6 10 5 12 8
 117

Now, to test the optimality of the solution we find the set of ui and v j so that for each
occupied cell (r , s), crs = ur + vs , in the usual manner, taking u4 = 0.

Now in each unoccupied cell, we enter the values of ui + v j and dij = cij − (ui + v j) at the
appropriate corner and observe that c11, c15 , c41, c45 < 0. So this solution is not optimal.

Since minimum d11 = − 2 < 0, so we allocate ∈ in this cell (1, 1) in place of c31.

ui ↓
(5) (3) (7) (7) (3) (2) (8) (7) (5)
∈ 1 2 0
(0) (1) (1)
(5) (6) (3) (12) (7) (5) (2) (7) (11) (5)
3 1 0
(3) (5) (3) (6)
(2) (0) (1) (− 2) (2) (4) (− 3) (8) (2) (2) (0)
2 −5
(2) (3) (7) (6) (2)
(9) (8) (6) (10) (5) (10) (10) (9) (8)
2 4 2 3
(1) (0) (1)
vj → 5 3 7 2 7 5

Let u1 = 0, we find the set of ui and vj and enter in all unoccupied ui + v j and
dij = cij − (ui + v j) and see that all dij > 0.

Thus, this solution is optimal.

ui ↓
(4) 2 (3) (2) 0 (0)
5 5 0
2 2
(2) (5) 3 (0) (0)
8 4 1 0
2
(3) 2 (8) 3 (6) 0 (0)
12 0
1 5 6
vj → 2 3 0 0
 118

Since all dij ≥ 0, so the solution under test is optimal.

Thus, the solution of the given problem is x12 = 5, x21 = 8 , x23 = 4 and minimum
transportation cost = 3 × 5 + 0 × 5 + 2 × 8 + 0 × 4 + 0 × 1 + 12 × 0
= 15 + 0 + 16 + 0 + 0 + 0
= ` 31.

Example 17: Consider the following unbalanced transportation problem:

To
1 2 3 Supply
1 5 1 7 10
From 2 6 4 6 80
3 3 2 5 15
Demand 75 20 50

Since there is not enough supply, some of the demands at these destinations may not be satisfied.
Suppose there are penalty costs for every unsatisfied demand unit which are given by 5, 3 and 2 for
destinations 1, 2 and 3 respectively. Find the optimal solution. [B.B.A. (Meerut) 2005, 2012]

Solution: In this problem, demand = 75 + 20 + 50 = 145

and supply = 10 + 80 + 15 = 105.

i.e., demand > supply, so we introduce a dummy source 4 with supply 145 − 105 = 40 and
costs 5, 3, 2 respectively.

By VAM an initial B.F.S. to the problem is shown in the table.

1 2 3
(5) (1) (7)
1 10 10

(6) (4) (6)

2 60 10 10 80

(3) (2) (5)

3 15 15

(5) (3) (2)

4 40 40

75 20 50 145
 119

Here, number of allocations = 6 = m + n − 1 = (4 + 3 − 1)

and they are at independent positions.
Total transportation cost
= 10 × 1 + 60 × 6 + 10 × 4 + 10 × 6 + 15 × 3 + 40 × 2
= 10 + 360 + 40 + 60 + 45 + 80
= ` 595.

To Test the Solution for Optimality: To test the optimality of the solution we find the
set of ui and v j for each occupied cell (r , s) s.t. crs = ur + vs , by considering u2 = 0, in the
usual manner.
ui ↓
(5) (3) (1) (7) (3)
10 − 3 (u1)
(2) (4)
(6) (4) (6)
60 10 10 0 (u2 )

(3) (2) (1) (5) (3)

15 − 3 (u3 )
(1) (2)
(5) (2) (3) (0) (2)
40 − 4 (u4 )
(3) (3)
vj → 6 4 6
v1 v2 v3

Now in each unoccupied cell, we enter ui + v j and dij = cij − (ui + v j) at the appropriate
corners and see that all dij ≥ 0.
Thus, this solution is optimal.
Hence, the optimal solution is B.F.S. i.e., ` 595.

Example 18: A steel company has three open hearth furnaces and five rolling mills. Transportation
costs (rupees per quintal) for shipping steel from furnaces to rolling mills are shown in the following table:
Mills
M1 M2 M3 M4 M5 Available
F1 4 2 3 2 6 8
Furnaces F2 5 4 5 2 1 12
F3 6 5 4 7 3 14
Required 4 4 6 8 8

What is an optimal shipping schedule? [B.C.A. (Meerut) 2000; (Kanpur) 2007]

Solution: Here requirement = 4 + 4 + 6 + 8 + 8 = 30
and availability = 8 + 12 + 14 = 34,
 120

i.e., total requirement is less than availability.

To convert the problem to a balanced one, we introduce a fictitious mill M6 with
requirement 34 − 30 = 4 and having all the transportation costs equal to zero. Thus, the
balanced transportation problem is as follows:

Mills
M1 M2 M3 M4 M5 M6 Available

F1 4 (4) 4 (2) 3 2 6 0 8
Furnaces F2 5 4 5 8 (2) 4 (1) 0 12
F3 6 5 6 (4) 7 4 (3) 4 (0) 14

Required 4 4 6 8 8 4 34

By Vogel's approximation method, an initial B.F.S. of the problem is given in the above
table, for which total transportation cost
= 4 × 4 + 4 × 2 + 8 × 2 + 4 ×1 + 6 × 4 + 4 × 3 + 4 × 0
= 16 + 8 + 16 + 4 + 24 + 12 + 0
= ` 80.

To Test the Solution for Optimality: Here total number of allocations is 7 which is one
less than the number m + n − 1 (= 8). So the solution is degenerate.

ui ↓
(4) (2) (3) (2) (2) (6) (1) (0) (− 2)
F1 4 4 ∈ − 2 (u1)
(1) (5) (2)
(5) (4) (4) (2) (5) (2) (2) (1) (0) (− 2)
F2 8 4 − 2 (u2 )
(1) (2) (3) (2)
(6) (6) (5) (4) (4) (7) (4) (3) (0)
F3 6 4 4 0 (u3 )
(0) (1) (3)
6 4 4 4 3 0
vj → v1 v2 v3 v4 v5 v6

To convert it to non-degenerate we introduce a small amount ∈ in the cell (1, 4), so that
the allocations remain at independent positions.

To test the optimality of the solution, we find the set of ui and vj s.t. for each occupied cell
(r , s), crs = ur + vs and enter the values of ui + vj and dij = cij − (ui + vj ) in each unoccupied
cell at the appropriate corner, and observe that all dij ≥ 0 .
Thus, this solution is optimal.
Hence, the optimal solution is B.F.S. by VAM = ` 80.
 121

P roblem S et
1. A company has three plants and four warehouses. The supply and demand in units
and the transportation costs are given. The solution of the problem is given below:
To
w1 w2 w3 w4 Supply

O1 5 10 4 5 10
From O2 6 8 7 5 25

O3 4 2 5 7 20

Demand 25 10 15 5

Answer the following questions:

(i) Is the solution feasible?
(ii) Is the solution degenerate?
(iii) Is this solution optimal?
(iv) Does this problem have more than one optimum solution? If so, show all of
them.
(v) If the cost of route O2 w3 is reduced from ` 7 to ` 6 per unit. What will be
optimum solution? [B.C.A. (Meerut) 2006, 2008, 2012]

2. Use north-west corner cell to determine an initial basic feasible solution to the
following transportation problem.
(i)
To
I II III IV Supply
A 13 11 15 20 2
From B 17 14 12 13 6
C 18 18 15 12 7
Demand 3 3 4 5 15
[B.C.A. (Indraprastha) 2012]

(ii)
Destination
D1 D2 D3 D4 Supply
O1 6 4 1 5 14
Origin O2 8 9 2 7 16
O3 4 3 6 2 5
Demand 6 10 15 4 35
 122

3. Determine an initial B.F.S. to the following transportation table using (i) matrix
minima method (ii) VAM:
Destination
D1 D2 D3 D4 Supply
O1 1 2 1 4 30
Origin O2 3 3 2 1 50
O3 4 2 5 9 20
Demand 20 40 30 10 100
[B.C.A. (Meerut) 2008]

4. Find the initial B.F.S. of the following transportation problem using (i) north-west
corner rule (ii) matrix minima method (iii) VAM:
Warehouse
w1 w2 w3 w4 Capacity
F1 19 30 50 10 7
Factory F2 70 30 40 60 9
F3 40 8 70 20 18
Requirement 5 8 7 14

5. Determine the optimal B.F.S. of the following (T.P.):

To
D1 D2 D3 D4 ai
O1 5 3 6 2 19
From O2 4 7 9 1 37
O3 3 4 7 5 34
bj 16 18 31 25

6. Solve the following Transportation Problem (T.P.):

To
D1 D2 D3 ai
O1 7 4 0 5
From O2 6 8 0 15
O3 3 9 0 9
bj 15 6 8 29

7. Find the optimal solution of the following (T.P.):

(i) To
w1 w2 w3 w4 ai
F1 19 30 50 10 7
From F2 70 30 40 60 9
F3 40 8 70 20 18
bj 5 8 7 14 34
[B.C.A. (Meerut) 2003, 2004]
 123

(ii) To
Available
7 3 4 2
From 2 1 3 3
3 4 6 5
Demand 4 1 5

(iii) To
D1 D2 D3 D4 D5 Available
O1 5 5 6 4 2 9
From O2 6 9 7 8 5 13
O3 5 6 4 6 3 9
Demand 3 7 8 5 8

(iv) To

D1 D2 D3 D4 D5 ai
O1 73 40 9 79 20 8

O2 62 93 96 8 13 7
From O3 96 65 80 50 65 9

O4 57 58 29 12 87 3

O5 56 23 87 18 12 5

bj 6 8 10 4 4

8. Define transportation problem. [B.C.A. (Meerut) 2004, 2008, 2012]

9. A company has three plants at locations A, B and C which supply to warehouses

located of D, E, F, G and H. Monthly plant capacities are 800, 500 and 900 units
respectively. Monthly warehouse requirements are 400, 400, 500, 400 and
800 units respectively. Unit transportation cost (in Rupees) are given below:

To

D E F G H

A 5 8 6 6 3

From B 4 7 7 6 5

C 8 4 6 6 4

Determine an optimum distribution for the company in order to minimize the total
transportation cost. [B.C.A. (Meerut) 2007, 2009]
 124

A nswers
1. (i) Yes, solution is feasible. (ii) Solution is non-degenerate.

(iii) Solution is optimum. (iv) No.

(v) The solution is not feasible.

2. (i) x11 = 2, x21 = 1, x22 = 3, x23 = 2, x33 = 2, x34 = 5.

(ii) x11 = 6, x12 = 8, x22 = 2, x23 = 14, x33 = 1, x34 = 4.

3. For (i) and (ii) both: x11 = 20, x13 = 10, x22 = 20, x23 = 20, x24 = 10, x32 = 20.

4. (i) x11 = 5, x12 = 2, x22 = 6, x23 = 3, x33 = 4, x34 = 14 .

(ii) x14 = 7, x21 = 2, x23 = 7, x31 = 3, x32 = 8, x34 = 7.

(iii) x11 = 5, x14 = 2, x23 = 7, x24 = 2, x32 = 8, x34 = 10.

5. x12 = 18, x13 = 1, x21 = 12, x24 = 25, x31 = 4, x33 = 30.

6. x12 = 5, x21 = 6, x22 = 1, x23 = 8, x31 = 9.

7. (i) x11 = 5, x14 = 2, x22 = 2, x23 = 7, x32 = 6, x34 = 12 .

(ii) x13 = 2, x22 = 1, x23 = 2, x31 = 4, x33 = 1.

(iii) x12 = 4, x14 = 5, x21 = 3, x22 = 2, x25 = 8, x32 = 1, x33 = 8.

(iv) x13 = 8, x24 = 4, x25 = 3, x31 = 5, x32 = 4, x41 = 1, x43 = 2, x52 = 4, x55 = 1.

9. Total cost = ` 141.00.

❍❍❍
 125

C HAPTER
4

Assignment Problem

4.1 Assignment Problem

[B.C.A. (Lucknow) 2010; B.C.A. (Avadh) 2008; B.C.A. (Agra) 2008; B.C.A. (Meerut) 2002, 2006]
The assignment problem can be stated in the form of (n × n) square cost matrix [c ij ] of real

number. Suppose there are n- jobs to be performed and n- persons are available for doing
these jobs. Assume that each person can do each job at a time, through with varying
degree of efficiency. Let ci j be the cost (payment) if the i th person is assigned the j th job,
the problem is to find an assignment i.e. which job should be assigned to which person, so
that the total cost for performing all jobs is minimum.

Jobs
1 2 3 ... j ... n

1 c11 c12 c13 ... c1 j ... c1n

2 c21 c22 c23 ... c2 j ... c2 n
3 c31 c32 c33 ... c3 j ... c3 n
Persons ⋮ : : : : :
i ci1 ci2 ci3 ... cij ... cin
⋮
n cn1 cn2 cn3 ... cnj ... cnn
 126

4.1.1 Mathematical Formulation of Assignment Problem

[B.C.A. (Meerut) 2009, 2010; B.C.A. (Agra) 2007; B.C.A. (Lucknow) 2007;
B.B.A. (Meerut) 2002, 2006]

Mathematically, the assignment problem can be stated as:

Minimize the total cost:

n n
Z = ∑ ∑ cij x ij , ...(1)
i=1
j =1

i = 1, 2 , … , n
where 
 j = 1, 2 , … , n

Subject to restrictions of the form

1, if ith person is assigned jth job

x ij = 
0, if not
n

∑=1 xij = 1 (one job is done by the ith person i = 1, 2, 3 … n)

j

∑=1 x ij = 1 (only one person should be assigned the j th job, j = 1, 2, 3 , … n)

i

where x ij denotes the j th job is to be assigned to the ith person.

4.1.2 Difference between Assignment and Transportation Problem

[B.C.A (Agra) 2004, 2006, 2010; B.C.A. (Kanpur ) 2008; B.C.A. (Lucknow) 2005, 2009;
B.C.A. (I.G.N.O.U.) 2012; B.C.A. (Meerut) 2008, 2011]

1. Assignment is the square matrix but transportation is not necessary square matrix if
we put m = n in transportation then it becomes the problem of assignment. Hence,
we can say that assignment problem is the special case of transportation problem.

2. The numerical evaluations of such association are called ‘effectiveness’ instead of

‘transportation costs’.

3. Mathematically, all ai and bj are unity and each x ij is limited to one of the two values
0 and 1. In such circumstances, exactly n of the x ij can be non-zero. One of each
origin and one for each destination.

Remark. Give the mathematical formulation of assignment. How does it differ from the
transportation problem? [B.C.A. (Meerut) 2009, 2011]
 127

4.1.3 Fundamental Theorems

Theorem 1: (Reduction Theorem). In an assignment problem if we add (or
subtract) a constant to every element of a row (or column) of the cost matrix [c ij ],

then an assignment which minimizes the total cost for one matrix also minimizes
the total cost for the other matrix.

or

Mathematical Statement of Reduction Theorem: If xij = X ij

n n
minimizes Z = ∑ ∑ cij xij over all xij = 0 or 1 such that
i =1 j =1

n n

∑= 1 xij = 1, ∑= 1 xij = 1, then xij = X ij also

i j

n n
minimizes Z′ = ∑1 ∑1 cij′ xij where cij ′ = cij ± ai ± bj
i= j=

ai, b j are some real numbers for i, j = 1, 2 , … , n.

[B.B.A. (Delhi) 2006, 2008, 2012]

n n n n
Proof: We have Z ′ = ∑=1 ∑=1 c′ij x ij = ∑=1 ∑=1 (cij ± ai ± bj) xij
i j i j

n n n n n n
= ∑ ∑
i =1 j =1
cij x ij ± ∑ ∑
i =1 j =1
ai x ij ± ∑ ∑ bj xij
i =1 j =1

n n n n  n n 
=Z ± ∑ ∑ ai x ij ± ∑ ∑ bj x ij ∵ Z =
 ∑ ∑ cij x ij 

i =1 j =1 j =1 i =1  i =1 j =1 

n n  n n 
=Z ± ∑ ai . 1 ± ∑ bj . 1 ∵
 ∑ x ij = 1 = ∑ x ij 

i =1 j =1  i = 1 j =1 

n n
=Z ± ∑= ai ± ∑= bj
i 1 j 1
 128

n n
Since, terms ∑ ai, ∑ bj are independent of xij it follows that Z′ is minimized whenever Z
i =1 j =1

is minimized and conversely.

Hence, x ij = X ij which minimizes Z will also minimize Z′.

Theorem 2: If all c ij ≥ 0 and there exists a solution xij = X ij which satisfies

n n

∑ ∑ cij xij = 0
i =1 j =1

then this solution is an optimal solution for the problem (i.e. minimizes the
objective function).

Proof: Since all cij ≥ 0 and all x ij ≥ 0, the objective function

n n
Z = ∑=1 ∑=1 cij xij cannot be negative.
i j

The minimum possible value that Z can attain is 0.

n n
Hence, the solution x ij = X ij for which ∑ ∑ cij xij = 0 is an optimal solution.
i =1 j =1

4.2 Assignment Algorithm or Hungarian Method

or
Reduced Matrix Method
[B.C.A. (Meerut), 2005, 2009; B.C.A. (Rohilkhand) 2007, 2009]

Various steps of the computational procedure for obtaining an optimum assignment may
be summarized as follows:

Step 1: Check whether the number of rows and columns in the cost matrix are equal. If
not, add dummy rows (columns) to form a square matrix.

Step 2: In the square cost matrix:

(i) Reduce each row element the lowest element of that row

(ii) Reduce each column element of the lowest element of that column.
 129

Step 3: In the reduced matrix, search for optimum solution as follows:

(i) Examine the rows successively until a row with exactly single zero is found. Mark
this zero by enrectangling (• ) and cross out (×) all others zeros of the
corresponding column. Proceed in this manner until all rows have been examined.

(ii) Examine the columns successively until a column with exactly single zero is found.
Mark this zero by enrectangling (• ) and cross out (×) all other zero of the
corresponding row. Proceed in this manner until all columns have been examined.

(iii) If each row and each column has one and only one marked zero, the optimum
allocation is attained which is indicated by the marked positions, otherwise go to
next step.

Step 4: Draw the minimum number of lines passing through all the zeros as follows:

(i) Tick (P) rows that do not have assignments.

(ii) Tick (P) columns that have zeros in ticked rows.

(iii) Tick (P) rows that have assignment in ticked columns.

(iv) Repeat (ii) and (iii) until the chain is completed.

(v) Draw straight lines through all unticked rows and ticked columns.

Step 5: If the minimum number of lines passing through all the zeros is equal to the
number of rows or columns, the optimum solution is attained by an arbitrary allocation in
the positions of the zeros not crossed in step 3. Otherwise go to next step.

Step 6: Revise the costs matrix as follows:

(i) Find the smallest element not covered by any of the lines of step 4.

(ii) Subtract this from all the uncrossed elements and add the same at the point of
intersection of the two lines.

(iii) Other elements crossed by the line remain unchanged.

Step 7: Go to step 4 and repeat the procedure till an optimum solution is attained.
 130

S olved E xamples
Example 1: Solve the following assignment problem:

Machines

M1 M2 M3

J1 8 7 6

Jobs J2 5 7 8

J3 6 8 7

[B.C.A. (Meerut) 2004, 2006, 2010]

Solution:
Step 1: Subtract the smallest element of each row from every element of the
corresponding row, we get
M1 M2 M3

J1 2 1 0

J2 0 2 3

J3 0 2 1

Step 2: Subtract the smallest element of each column from every element of the
corresponding column, we get
M1 M2 M3

J1 2 0 0

J2 0 1 3

J3 0 1 1

Step 3: Reduced matrix, make assignments in rows and columns having single zeros.

M1 M2 M3

J1 2 0 ×0
J2 0 1 3
P
J3 ×0 1 1
P
P
 131

Step 4: Draw the minimum number of lines to cover all the zeros of the reduced matrix
now subtract, from uncovered elements and adding the same at the intersection of two
lines. The lowest element from uncovered.
M1 M2 M3

J1 3 ×0 0

J2 0 ×0 2

J3 ×0 0 ×0
Since, the number of assignments is equal to the order of the matrix, we have the
optimum assignment schedule.
J1 → M3 , J2 → M1, J3 → M2
Minimum cost = 6 + 5 + 8
= 19

Example 2: Solve the minimal assignment problem whose effectiveness matrix is:

I II III IV

A 2 3 4 5

B 4 5 6 7

C 7 8 9 8

D 3 5 8 4

[B.C.A. (Lucknow) 2007; B.C.A.(Meerut) 2002, 2011; B.C.A. (Delhi) 2007]

Solution:

Step 1: Subtracting the minimum element of each row from every element of the
corresponding row, the matrix reduces to

I II III IV

A 0 1 2 3

B 0 1 2 3

C 0 1 2 1

D 0 2 5 1
 132

Step 2: Now subtracting the minimum element of each column from every element of
the corresponding column, the matrix reduces to

I II III IV

A 0 0 0 2

B 0 0 0 2

C 0 0 0 0

D 0 1 3 0

Step 3: Now to test whether it is possible to make an assignment using only zeros.

Here none of the rows or columns contain exactly one zero, therefore we start with row,
searching two zeros. While examining rows successively, it is observed that row 4 has two
zeros. Now, arbitrarily make an assignment (• ) one of these two zeros, say zero in the
column 1 and cross other zeros in row 4 and column 1. Now we examine the columns and
first column 4 which contains only one unmarked zero in row 3. We make assignment
(• ) at this zero and cross all other zeros of this row. Now again we check the rows and
columns for one unmarked zero. There is no such row or column. So we start with row 1
searching two unmarked zeros and find the row 1 containing two such zeros. We mark
(• ) at any one of these zeros. Let zero at column 2 and cross other zeros of row 1 and
column 2. Now the second row contains only one unmarked zero in third column where
we can make an assignment.

At this stage all zeros have been either assigned or cross out. We observe that every row
and column have one assignment, so we have the complete zero assignment.

Table (a), (b), (c) show the necessary steps for reaching the optimal assignment.

(a) I II III IV

A ×0 0 0 2

B ×0 0 0 2

C ×0 0 0 0

D 0 1 3 ×0
 133

I II III IV
(b)
A ×0 0 0 2

B ×0 0 0 2

C ×0 ×0 ×0 0

D 0 1 3 ×0
I II III IV
(c)
A ×0 0 ×0 2

B ×0 ×0 0 2

C ×0 ×0 ×0 0

D 0 1 3 ×0
Thus, we get the following optimal assignment A → II , B → III , C → IV , D → I .

Minimum cost = 3 + 6 + 8 + 3 = 20

Note:
The other optimal solutions are also possible. Each will have the cost 20.

Example 3: A departmental head has four subordinates and four tasks to be performed. The
subordinates differ in efficiency and tasks differ in their intrinsic difficulty, this estimate of the time
each man would take to perform each task is given in the matrix below.
Subordinates

I II III IV

A 18 26 17 11

B 13 28 14 26
Tasks
C 38 19 18 15

D 19 26 24 10

How should the tasks be allocated, one to a man, so as to minimize the total man hours?

[B.C.A.(Agra ) 2005; B.C.A. (Avadh) 2007; B.C.A. (Rohilkhand) 2004, 2006, 2008, 2011]
 134

Solution:

Step 1: Subtract the smallest element of each row from every element of the
corresponding row, we get the reduced matrix :

I II III IV

A 7 15 6 0

B 0 15 1 13

C 23 4 3 0

D 9 16 14 0

Step 2: Subtracting the smallest element of each column of the reduced matrix from
every element of the corresponding column. We get the following reduced matrix :

I II III IV

A 7 11 5 0

B 0 11 0 13

C 23 0 2 0

D 9 12 13 0

Step 3: Starting with row 1 we enrectangled (• ) to a single zero, if any and cross (×) all
other zeros in the column of zero so marked. Thus, we get

I II III IV

A 7 11 5 0

B 0 11 ×0 13

C 23 0 2 ×0
D 9 12 13 ×0
In the above matrix, we arbitrarily enrectangled a zero in column 1, because row 2 had
two zeros.

It may be noted that column 3 and row 4 do not have any assignment. So we move on to
the next step.
 135

Step 4:

(i) Since row 4 does not have any assignment, we tick this row (P).

(ii) Now there is a zero in the fourth column of the ticked row. So we tick fourth
column (P).

(iii) Further there is an assignment in the first row of the ticked column. So we tick first
row (P).

(iv) Draw straight lines through all unticked rows and ticked columns. Thus, we have

I II III IV

A 7 11 5 0 P

B 0 11 ×0 13

C 23 0 2 ×0
D 9 12 13 ×0 P

Step 5: In step 4, we observe that the minimum number of lines so drawn is 3, which is less
than the order of the cost matrix, indicating that the current assignment is not optimum.

To increase the minimum number of fives, we generate new zeros in the modified matrix.

Step 6: The smallest element not covered by the lines is 5 subtracting this element from
all the uncovered elements and adding the same to all the elements lying at the
intersection of lines, we obtain the following new modified cost matrix.

I II III IV

A 2 6 0 0

B 0 11 0 18

C 23 0 2 5

D 4 7 8 0
 136

Step 7: Repeating step 4 on the modified matrix, we get

I II III IV

A 2 6 0 ×0
B 0 11 ×0 18

C 23 0 2 5

D 4 7 8 0

Now, since each row and each column has one and only one assignment, an optimum
solution is reached.

The optimum assignment is

A → III , B → I , C → II , D → IV

The minimum total time for this assignment is 17 + 13 + 19 + 10 = 59 man hours.

Example 4: Consider the problem of assigning five jobs to five persons. The assignment costs are
given as follows:

Job

1 2 3 4 5

A 8 4 2 6 1

B 0 9 5 5 4

Persons C 3 8 9 2 6

D 4 3 1 0 3

E 9 5 8 9 5

Determine the optimum assignment schedule.

[B.C.A. (Bundelkhand) 2008; B.C.A. (Kanpur) 2005, 2007]

Solution:

Step 1: To minimize the row

 137

1 2 3 4 5

A 7 3 1 5 0

B 0 9 5 5 4

C 1 6 7 0 4

D 4 3 1 0 3

E 4 0 3 4 0

Step 2: To minimize the column, in row and column that have single zeros, we get the
optimum assignment table.

1 2 3 4 5

A 7 3 ×0 5 0

B 0 9 4 5 4

C 1 6 6 0 4

D 4 3 0 ×0 3

E 4 0 2 4 ×0
Optimum assignment schedule is A→ 5, B→1, C → 4, D → 3 and E → 2

The minimum cost of assignment is 1 + 0 + 2 + 1 + 5 = 9

Example 5 : A car hire company has one car at each of five depots a, b, c, d and e. A customer requires a
car in each town namely A, B, C, D and E. Distance (in kms) between depots (origins) and towns
(destinations) are given in the following distance matrix :

a b c d e

A 160 130 175 190 200

B 135 120 130 160 175

C 140 110 155 170 185

D 50 50 80 80 110

E 55 35 70 80 105

[B.C.A. (Bhopal) 2006, 2012; B.C.A. (Rohilkhand) 2004, 2007]

How should cars be assigned to customers so as to minimize the distance travelled?

 138

Solution:

Step 1: Subtracting the minimum element of each row from every element of the
corresponding row, the matrix reduces to

a b c d e

A 30 0 45 60 70

B 15 0 10 40 55

C 30 0 45 60 75

D 0 0 30 30 60

E 20 0 35 45 70

Now subtracting the minimum element of each column from every element of a
corresponding column, the matrix reduces to

a b c d e

A 30 0 35 30 15

B 15 0 0 10 0

C 30 0 35 30 20

D 0 0 20 0 5

E 20 0 25 15 15

Step 2: Now we give the zero assignments in our usual manner. Row 1 has a single zero
in column 2. Make an assignment by marking ‘• ’ around it and delete other zeros (if
any) in column 2 by marking ‘×’. Examining the set of rows completely, an identical
procedure is applied successively to columns.

a b c d e

A 30 0 35 30 15

B 15 ×0 0 10 ×0
C 30 ×0 35 30 20

D 0 ×0 20 ×0 5

E 20 ×0 25 15 15
 139

Now column 1 has a single zero in row 4. Make an assignment by marking ‘• ’ at this zero
and cross the other zero of row 4 which is not yet crossed. Column 3 has a single zero in
row 2, make an assignment at this zero by putting ‘• ’ and cross the other zero of row
which is not yet crossed. At this stage all zeros have been either assigned or crossed out. It
is observed that row 3, row 5, column 4 and column 5 each has no assignment. Hence,
the required solution cannot be obtained at this stage. So we proceed to the next step.

Step 3: In this step we draw minimum number of lines to cover all zeros at least once.
For this we proceed as follows :

(i) Tick (P) row 3 and row 5 as they have no assignments.

(ii) Tick (P) column 2 as having zeros in the marked rows 3 and 5.

(iii) Tick (P) row 1 as it contains assignment in the marked column 2.

No further rows or columns will be required to mark during this procedure.

(iv) Now draw line L1 through marked column 2. Then draw lines L2 and L3 through
unmarked rows 2 and 4.

The required lines will be L1, L2 and L3 . No zero is left uncovered.

a b L1 c d e

A 30 0 35 30 15 P (4)

L2 B 15 ×0 0 10 ×0
C 30 ×0 35 30 20 P (1)

L3 D 0 ×0 20 ×0 5

E 20 ×0 25 15 15 P (2)

(3)

Step 4: In this step we select the smallest element among all uncovered elements of the
matrix of step 3.

Here, this element is 15. Subtracting this element 15 from all the elements that do not
have a line through them and adding to every element that lies at the intersection of two
lines and leaving the remaining elements unchanged we get the following matrix.
 140

a b c d e

A 15 0 20 15 0

B 15 15 0 10 0

C 15 0 20 15 5

D 0 15 20 0 5

E 5 0 10 0 0

Step 5: Now again performing the step 2 we make the zero assignments. It is observed
that there are no remaining zeros and every row (column) has an assignment as shown in
the table.

a b c d e

A 15 ×0 20 15 0

B 15 15 0 10 ×0
C 15 0 20 15 5

D 0 15 20 ×0 5

E 5 ×0 10 0 ×0
Thus, the complete optimal assignment plan is given by

A → e, B → c, C → b, D → a, E → d

From the original matrix, the minimum cost (distance travelled)

= (200 + 130 + 110 + 50 + 80) kms. = 570 kms .

Example 6: Solve the assignment problem represented by the following matrix :

I II III IV V VI

A 9 22 58 11 19 27

B 43 78 72 50 63 48

C 41 28 91 37 45 33

D 74 42 27 49 39 32

E 36 11 57 22 25 18

F 3 56 53 31 17 28

[B.C.A. (Indore) 2012; B.C.A. (Agra) 2001, 2004, 2006,2009]

 141

Solution:

Step 1: Subtracting the minimum element of each row from every element of the
corresponding row and then subtracting the minimum element of each column from
every element of the corresponding column, the matrix reduces to

I II III IV V VI

A 0 13 49 0 0 13

B 0 35 29 5 10 0

C 13 0 63 7 7 0

D 47 15 0 20 2 0

E 25 0 46 9 4 2

F 0 53 50 26 4 20

Step 2: Make the ‘zero assignments’ in usual manner. The illustration is shown in the
table. Since row 3 and column 5 have no assignments so we proceed to the next step.

I II III IV V VI

A ×0 13 49 0 ×0 13

B ×0 35 29 5 10 0

C 13 ×0 63 7 7 ×0
D 47 15 0 20 2 ×0
E 25 0 46 9 4 2

F 0 53 50 26 4 20

Step 3: Draw minimum number of lines to cover all zeros at least once. For this we
proceed as follows:

(i) Tick (P) row 3 as having no assignment.

(ii) Tick (P) columns 2 and 6 as having zeros in marked row 3.

(iii) Tick (P) rows 5 and 2 as having assignments in the marked columns 2 and 6.

(iv) Tick (P) column 1 (not already marked) as having zero in the marked row 2.

(v) Then tick (P) row 6 as having assignment in the marked column 1.
 142

L1 L2 L3

I II III IV V VI

L4 A ×0 13 49 0 ×0 13

B ×0 35 29 5 10 0 P (5)

C 13 ×0 63 7 7 ×0 P (1)

L5 D 47 15 0 20 2 ×0
E 25 0 46 9 4 2 P (4)

F 0 53 50 26 4 20 P (7)

P P P

(6) (2) (3)

Now draw lines L1, L2 , L3 through marked columns 1, 2, 6 respectively and L4 , L5

through unmarked rows 1, 4 respectively. This way minimum set of five lines (5 < 6) to
cover all the zeros is obtained.

Step 4: Now the smallest element among all uncovered elements is 4. Subtracting this
element 4 from all the uncovered elements, adding to every element that lies at the
intersection of two lines and leaving the remaining elements unchanged the matrix of step 3
reduces to the new form as shown in the table.

I II III IV V VI

A 4 17 49 0 0 17

B 0 35 25 1 6 0

C 13 0 59 3 3 0

D 51 19 0 20 2 4

E 25 0 42 5 0 2

F 0 53 46 22 0 20

Step 5: Repeating the step 2, make the ‘zero assignments as shown in the following
table. Thus, exactly one marked ‘• ’ zero in each row and each column of the matrix is
obtained.
 143

I II III IV V VI
A 4 17 49 0 ×0 17
B 0 35 25 1 6 ×0
C 13 ×0 59 3 3 0
D 51 19 0 20 2 4
E 25 0 42 5 ×0 2
F ×0 53 46 22 0 20

Thus, the optimal assignment is

A → IV , B → I , C → VI , D → III , E → II , F → V

From the original matrix, minimum cost = 11 + 43 + 33 + 27 + 11 + 17 = 142.

Note:
Another optimal solution of this assignment problem is shown in the following table i.e.
A → IV , B → VI , C → II , D → III , E → V , F → I .

I II III IV V VI
A 4 17 49 0 ×0 17
B ×0 35 25 1 6 0
C 13 0 59 3 3 ×0
D 51 19 0 20 2 4
E 25 ×0 42 5 0 2
F 0 53 46 22 ×0 20

From the original matrix, minimum cost = 11 + 48 + 28 + 27 + 25 + 3 = 142.

Example 7: Solve the fol low ing as sign ment prob lem hav ing the fol low ing cost el e ments.

1 2 3 4 5 6 7
A 35 22 60 41 27 52 44
B 51 39 42 33 65 47 58
C 25 32 53 41 50 36 43
D 32 28 40 46 3 55 49
E 43 36 45 63 57 49 42
F 27 18 31 46 35 42 34
G 48 50 72 59 43 64 58

[B.C.A. (Agra) 2004; B.C.A. (Lucknow) 2008; B.B.A. (I.G.N.O.U.) 2004, 2008, 2009]
 144

Solution:

Step 1: Subtract the smallest element of each row (column) from all elements of the
respective row (column).

1 2 3 4 5 6 7

A 13 0 38 19 5 30 22

B 18 6 9 0 32 14 25

C 0 7 28 16 25 11 18

D 29 25 37 43 0 52 46

E 7 0 9 27 21 13 6

F 9 0 13 28 17 24 16

G 5 7 29 16 0 21 15

and

1 2 3 4 5 6 7

A 13 0 29 19 5 19 16

B 18 6 0 0 32 3 19

C 0 7 19 16 25 0 12

D 29 25 28 43 0 41 40

E 7 0 0 27 21 2 0

F 9 0 4 28 17 13 10

G 5 7 20 16 0 10 9

Step 2: To make assignment in each row and each column with usual method.

1 2 3 4 5 6 7

A 13 0 29 19 5 19 16 P

B 18 6 ×0 0 32 3 19

C 0 7 19 16 25 ×0 12

D 29 25 28 43 0 41 40 P

E 7 ×0 ×0 27 21 2 0

F 9 ×0 4 28 17 13 10 P

G 5 7 20 16 ×0 10 9 P

P P
 145

Now subtracting the element 4 from all the elements not covered by lines and adding the
same at the intersection of two lines.

1 2 3 4 5 6 7

A 9 0 25 15 5 15 12

B 18 10 ×0 0 36 3 19

C 0 11 19 16 29 ×0 12

D 25 25 24 39 0 37 36 P

E 7 4 ×0 27 25 2 0

F 5 ×0 0 24 17 9 6

G 1 7 16 12 ×0 6 5 P

Final Iteration: Subtracting 1 from all the elements not covered by lines and add the
same at the intersection of two lines.

1 2 3 4 5 6 7

A 9 0 25 15 6 15 12

B 18 10 ×0 0 37 3 19

C ×0 11 19 16 30 0 12

D 24 24 23 38 0 36 35

E 7 4 ×0 27 26 2 0

F 5 ×0 0 24 18 9 6

G 0 6 15 11 ×0 5 4

The optimum assignment is A → 2, B → 4, C → 6, D → 5, E → 7, F → 3 and G → 1. The

minimum cost 22 + 33 + 36 + 3 + 42 + 31 + 48 = 215.

4.3 Unbalanced Assignment Problem

[B.C.A. (Meerut) 2005, 2008]

If the cost matrix of an assignment problem is not square matrix, the assignment problem
is called unbalanced assignment problem. In such cases add dummy rows or columns with
zero costs. Then the usual assignment algorithm can be applied to this resulting balanced
problem.
 146

4.4 Maximal Assignment Problem

[B.B.A. (Kanpur) 2007]
Sometimes, the assignment problem deals with the maximization of an objective function
rather than to minimize it. For example, it may be required to assign persons to jobs in
such a way that the expected profit is maximum. Such problem may be solved easily by
first converting it to a minimization problem and then applying the usual procedure of
assignment algorithm. This conversion can be very easily done by subtracting from the
highest element, all the elements of the given profit matrix or by placing minus sign
before each element of the profit matrix in order to make it cost matrix.

4.5 Restrictions on Assignment Problem

[B.B.A. (Delhi) 2008]
Sometimes due to some restrictions the assignment of a particular facility to a particular
job is not permitted. To overcome such difficulty we assign a very high cost (∞) to the
corresponding cell, so that the activity will be automatically excluded from the optimal
solution.

4.6 Application of Assignment Problem

[B.B.A. (Meerut) 2004, 2005]
The assignment problem which finds many applications in allocation and scheduling, for
example, in assigning planes or crews to commercial airline flights, trucks or drivers to
different routes, man to offices and space to departments. If there are more jobs to do
than can be done, we can decide either which job to leave undone or what resources to
add. There are more number of decision making situations where assignment models can
be used. For example, assignment of machines to jobs; assignment of salesman to sales
territories; assignment of workers to various tasks, vehicles to routes; contracts to bidder
and so on.

4.7 Objective of Assignment Problem

[B.B.A. (Meerut) 2002, 2003, 2004, 2006]
The assignment problem is a special type of allocation problem. In both cases the
objective is to fulfil the targets by means of available resources which are available in
specified amounts. But the operating conditions are different. While in the allocation
problem each target can be obtained in one way only. In assignment problem the
individual targets can be attained in different ways, we find that a transportation problem
is degenerated, but in assignment problem each resource can be assigned to only one job
and each job requires only one resource. Hence, the assignment problem is completely
degenerated from the transportation problem.
 147

Example 8: Five jobs are to be processed and five machines are available any machine can process
any job with the resultant profit (in rupees) as follows:
A B C D E
1 32 38 40 28 40
2 40 24 28 21 36
3 41 27 33 30 37
4 22 38 41 36 36
5 29 33 40 35 39
What is the maximum profit that may be expected if an optimum assignment is made?
Solution: It is maximization problem. To convert it into a minimization one, we multiply
each element of the given matrix by −1. Thus, the resulting matrix is :
A B C D E
1 − 32 − 38 − 40 − 28 − 40
2 − 40 − 24 − 28 − 21 − 36
3 − 41 − 27 − 33 − 30 − 37
4 − 22 − 38 − 41 − 36 − 36
5 − 29 − 33 − 40 − 35 − 39

We shall solve this minimization problem by usual assignment algorithm.

Step 1: Subtract the smallest element of each row from every element of the corresponding
row then subtracting the smallest element of each column from each element of the
corresponding column, we get the following table.
A B C D E
1 8 0 0 7 0
2 0 14 12 14 4
3 0 12 8 6 4
4 19 1 0 0 5
5 11 5 0 0 1

Step 2: To make assignment in each row and each column by usual manner.
A B C D E
1 8 0 ×0 7 ×0
2 0 14 12 14 4 P
3 ×0 12 8 6 4 P
4 19 1 0 ×0 5
5 11 5 ×0 0 1

P
Step 3: Now, to draw the minimum number of lines to cover all the zeros at least once.
 148

Step 4: Select the smallest element from uncovered elements i.e. 4. Subtract this element
4 from all the uncovered elements, adding to every element that lies at the intersection of
two lines and leaving remaining elements unchanged.
A B C D E
1 12 0 0 7 0
2 0 10 8 10 0
3 0 8 4 2 0
4 23 1 0 0 5
5 15 5 0 0 1

Step 5: Given zero assignments in the usual manner we see that each row and each
column have an assignment.
A B C D E
1 12 0 ×0 7 ×0
2 0 10 8 10 ×0
3 ×0 8 4 2 0
4 23 1 0 ×0 5
5 15 5 ×0 0 1

Hence the optimal solution for maximum sale to the problem is

1 → B, 2 → A, 3 → E, 4 → C, 5 → D
i.e., Maximum sales = 38 + 40 + 37 + 41 + 35
= 191 i.e., ` 19,100

Example 9: A company has 4 machines to do 3 jobs. Each can be assigned to one and only one
machine. The cost of each job on each machine is given in the following table.
Machine
W X Y Z
A 18 24 28 32
Job B 8 13 17 19
C 10 15 19 22

What are the job assignments which will minimize the cost? [B.C.A. (Meerut) 2003, 2008, 2010]

Solution: Since the matrix is not square. It is an unbalanced assignment problem, then
adding dummy row D to get the square matrix.
W X Y Z
A 18 24 28 32
B 8 13 17 19
C 10 15 19 22
D 0 0 0 0
 149

Now the problem can be solved by usual method.

Step 1: Subtracting the smallest element in each row from every element of the
corresponding row and then subtracting the smallest element in each column from every
element of the corresponding column, the matrix reduces to

W X Y Z
A 0 6 10 14

B 0 5 9 11

C 0 5 9 12

D 0 0 0 0

Step 2: To make assignment in each row and each column by usual manner.

W X Y Z
A 0 6 10 14 P
B ×0 5 9 11 P
C ×0 5 9 12 P
D ×0 0 ×0 ×0
P
Step 3: Here, we draw minimum number of lines to cover all the zeros at least once the
smallest elements among all the uncovered elements is 5. Subtracting this elements from
all uncovered elements, adding to every element that lies at the intersection of two lines
and leaving remaining elements unchanged the table thus, reduces as follows :
W X Y Z

A 0 1 5 9

B 0 0 4 6

C 0 0 4 7

D 5 0 0 0

Step 4: To make assignments in the usual manner.

W X Y Z
A 0 1 5 9 P
B ×0 0 4 6 P
C ×0 ×0 4 7 P
D 5 ×0 0 ×0
P P
 150

Step 5: Repeat the step 4 i.e.,to select the smallest element from uncovered elements i.e.,
4 then subtract to uncovered, covered remain same and add to all elements that lies at the
intersection of two lines:

W X Y Z

A 0 1 1 5

B 0 0 0 2

C 0 0 0 3

D 9 4 0 0

Step 6: To make assignments in each rows and each columns.

W X Y Z W X Y Z

A 0 1 1 5 A 0 1 1 5

B ×0 0 ×0 2 B ×0 ×0 0 2

C ×0 ×0 0 3 C ×0 0 ×0 3

D 9 4 ×0 0 D 9 4 ×0 0

Hence, the optimal assignments of jobs to machine for minimum cost are as

Job A → W , B → X , C → Y , D → Z

and Job A → W , B → Y , C → X , D → Z

The minimize cost (Z ) = 18 + 13 + 19 + 0 = 50.

Example 10: Four engineers are available to design four projects. Engineer 2 is not competent to
design the project B. Given the following time estimates needed to each engineer to design a given
project, find how should be engineers be assigned to projects so as to minimize the total design time of
four projects? [B.C.A. (Rohilkhand) 2003, 2007]

Projects

A B C D

1 12 10 10 8

2 14 Not suitable 15 11
Engineers
3 6 10 16 4

4 8 10 9 7
 151

Solution: To avoid the assignment 2 → B, we take its time to be very large (say ∞). Then
the cost matrix of the resulting assignment problem is shown in the following table :

A B C D
1 12 10 10 8
2 14 ∞ 15 11
3 6 10 16 4
4 8 10 9 7

Now we apply the assignment technique in the usual manner.

Step 1: Subtracting the minimum element of each row from every element of the
corresponding row and then subtracting minimum element of each column from every
element of the corresponding column, the reduced matrix is

A B C D
1 3 0 0 0
2 2 ∞ 2 0
3 1 4 10 0
4 0 1 0 0

Step 2: To make assignments in each rows and each columns by usual manner.

A B C D

L2 1 3 0 ×0 ×0
2 2 ∞ 2 0 P (3)
3 1 4 10 ×0 P (1)
L3 4 0 1 ×0 ×0
P
L1

Step 3: In the above table, the smallest of the uncovered elements is 1. Subtracting this
element 1 from all uncovered elements, adding to each element that lies at the
intersection of two lines and leaving remaining elements unchanged, we get the following
matrix

A B C D
1 3 0 0 1
2 1 ∞ 1 0
3 0 3 9 0
4 0 1 0 1
 152

Step 4: Giving zero assignments in the usual manner, we observe that each row and each
column have a zero assignment.

A B C D

1 3 0 ×0 1

2 1 ∞ 1 0

3 0 3 9 ×0
4 ×0 1 0 1

Hence, the optimal assignment is

Engineer → Project: 1 → B, 2 → D, 3 → A, 4 → C

From the given matrix total minimum time = 10 + 11 + 6 + 9 = 36.

Example 11: Find the optimal solution for the assignment problem with the following cost matrix :

I II III IV V

A 11 17 8 16 20

B 9 7 12 6 15

C 13 16 15 12 16

D 21 24 17 28 26

E 14 10 12 11 15

[B.C.A. (Meerut) 2002, 2008]

Solution: Step 1: Subtracting the smallest element of each row from every element of
the corresponding row and then in the resulting matrix subtracting the smallest element
of each column from every element of the corresponding column, we get the following
matrix:

I II III IV V

A 2 9 0 8 8

B 2 1 6 0 5

C 0 4 3 0 0

D 3 7 0 11 5

E 3 0 2 1 1
 153

Step 2: Reduced matrix, make assignments in row and column having single zero

I II III IV V
A 2 9 0 8 8
B 2 1 6 0 5
C 0 4 3 ×0 ×0
D 3 7 ×0 11 5

E 3 0 2 1 1

Since the 4th row and 5th column have no assignments, then applying tick rule.

Step 3: Draw the minimum number of lines to cover all zeros at least one. For this
applying the following rules:
(i) Mark (P) row 4 having no assignment
(ii) Mark (P) column 3 having zero in marked row 4
(iii) Mark (P) row 1 having assignment in the marked column

I II III IV V
A 2 9 0 8 8 P
B 2 1 6 0 5
C 0 4 3 ×0 ×0
D 3 7 ×0 11 5 P

E 3 0 2 1 1

Draw the line through unmarked row 2, 3 and 5 and marked column.
Step 4: Select the smallest element among all uncovered elements is 2. Subtract this
element from all uncovered elements, adding to every element that lies at the
intersection of two lines and leaving remaining elements unchanged, the matrix of step 3
reduces to the form.

I II III IV V

A 0 7 0 6 6

B 2 1 8 0 5

C 0 4 5 0 0

D 1 5 0 9 3

E 3 0 4 1 1
 154

Step 5: Repeating the step 2, make "zero assignment" we get the following table:

I II III IV V
A 0 7 ×0 6 6

B 2 1 8 0 5
C ×0 4 5 ×0 0

D 1 5 0 9 3
E 3 0 4 1 1

Thus, we get assignment 0 in each row and each column of the matrix.
Hence, the optimal assignment is
A → I , B → IV , C → V , D → III , E → II
From the original matrix
Minimum cost = 11 + 6 + 16 + 17 + 10 = 60

Example 12: The number of man-hours needed to complete a job for each job-man combination are
given below

Jobs

Men A B C D

1 5 3 1 8

2 7 9 2 6

3 6 4 5 7

4 5 7 7 6

Find the optimal assignment that will result in minimum man-hours needed.
[B.C.A. (Kanpur) 2004; B.C.A. (Meerut) 2006, 2011]

Solution: Step 1: Subtracting the smallest element each row from every element of the
corresponding row and then subtracting the smallest element of each column from every
element of the corresponding column, we get the following matrix :

Jobs

A B C D

1 4 2 0 6

Men 2 5 7 0 3

3 2 0 1 2

4 0 2 2 0
 155

Step 2: Make assignments 0 in each row and each column by usual manner, we get the
following matrix:

Job
A B C D
1 4 2 0 6
Men 2 5 7 ×0 3

3 2 0 1 2
4 0 2 2 ×0
Since row 2 and column 4 have no zero assignment, then applying the following marked
or tick (P) rule.

Step 3: We draw the minimum number of lines to cover all the zeros at least once by
usual manner.

A B C D
1 4 2 0 6 P
2 5 7 ×0 3 P
3 2 0 1 2
4 0 2 2 ×0
P

Step 4: The smallest element among all uncovered element is 2. Then subtract this
element from all the uncovered elements, adding to every element that lies at the
intersection of two lines and leaving remaining elements unchanged, the matrix in step 3
reduces to

A B C D
1 2 0 0 4
2 3 5 0 1
3 2 0 3 2
4 0 2 4 0

Step 5: Again repeat the step 2. We get the following matrix:

A B C D
1 2 ×0 ×0 4

2 3 5 0 1
3 2 0 3 2
4 0 2 4 ×0
 156

Step 6: Again repeat step 3 of drawing minimum number of lines to cover all zero at least
once then we find the following matrix:

A B C D
1 2 ×0 ×0 4 P
2 3 5 0 1 P
3 2 0 3 2 P
4 0 2 4 ×0
P P

Step 7: Now repeat the step 4 where smallest element from uncovered element is 1 we
get the following matrix :
A B C D
1 1 0 0 3
2 2 5 0 0
3 1 0 3 1
4 0 3 5 0

Step 8: Make zero assignments in each row and column by usual manner, we get the
following matrix:

A B C D
1 1 ×0 0 3
2 2 5 ×0 0
3 1 0 3 1
4 0 3 5 ×0
Hence the optimal assignment for minimum cost of the problem is
1 → C, 2 → D, 3 → B, 4 → A
∵ Minimum cost = 1 + 6 + 4 + 5 = 16.

Example 13: Explain the objective of an assignment problem. Five men are available to do five
different jobs. From the past records, the time (in hours) that each man takes to do each job is known
and given in the following table:
I II III IV V

A 2 9 2 7 1
Men B 6 8 7 6 1
C 4 6 5 3 1
D 4 2 7 3 1
E 5 3 9 5 1

Find the assignment of men to jobs that will minimize the total time taken.
[B.C.A. (Lucknow) 2002, 2004, 2006]
 157

Solution: Step 1: Subtracting the smallest element each row from every element of the
corresponding row, we get the following matrix :

I II III IV V

A 1 8 1 6 0

B 5 7 6 5 0

C 3 5 4 2 0

D 3 1 6 2 0

E 4 2 8 4 0

Step 2: Subtracting the smallest element each column from every element of the
corresponding column, we get the following matrix :

I II III IV V

A 0 7 0 4 0

B 4 6 5 3 0

C 2 4 3 0 0

D 2 0 5 0 0

E 3 1 7 2 0

Step 3: Reduced matrix, make assignments in row and column having single zero.

I II III IV V
A 0 7 ×0 4 ×0
B 4 6 5 3 0
C 2 4 3 0 ×0
D 2 0 5 ×0 ×0
E 3 1 7 2 ×0
Since row 5 th and column 3 rd having no assignments. Then applying tick (P) rule.

Step 4: Draw minimum number of lines to cover all zeros at least once. For this we apply
following rules:

(i) Tick (P) the row 5 having no assignment.

(ii) Tick (P) the column 5 th having zero in tick row.

(iii) Tick (P) row 2 nd having assignment in tick (P) column.

 158

I II III IV V
A 0 7 ×0 4 ×0
B 4 6 5 3 0 P
C 2 4 3 0 ×0
D 2 0 5 ×0 ×0
E 3 1 7 2 ×0 P
P

Draw the line through unticked row and tick column 5 th .

Step 5: The smallest element among all uncovered element is 1. Then subtract this
element from all the uncovered elements, adding to every element that lies at the
intersection of two lines and leaving remaining elements unchanged the matrix in step 4
reduces to.

I II III IV V

A 0 7 0 4 1

B 3 5 4 2 0

C 2 4 3 0 1

D 2 0 5 0 1

E 2 0 6 1 0

Step 6: Repeat the step 3 in step 6.

I II III IV V
A 0 7 ×0 4 1
B 3 5 4 2 0
C 2 4 3 0 1
D 2 0 5 ×0 1
E 2 ×0 6 1 ×0
Step 7: Repeat the step 4

I II III IV V
A 0 7 ×0 4 1
B 3 5 4 2 0 P
C 2 4 3 0 1 P
D 2 0 5 ×0 1 P
E 2 ×0 6 1 ×0 P
P P P
 159

Step 8: Select the smallest element 2 from uncovered then repeat step 5.

I II III IV V
A ×0 9 0 6 3
B 1 5 2 2 0
C ×0 4 1 0 1
D 0 ×0 3 ×0 1
E ×0 0 4 1 ×0
Hence the optimal assignment for minimum cost of the problem is
A → III , B → V , C → IV , D → I , E → II
Minimum cost = 2 + 1 + 3 + 4 + 3 = 13

P roblem S et
1. Consider the problem of assigning five operators to five machines. The assignment
costs are given below :
Operators
I II III IV V
A 10 5 13 15 16
B 3 9 18 3 6
Machines C 10 7 2 2 2
D 5 11 9 7 12
E 7 9 10 4 12

Assign the operators to different machines so that total cost is minimized.

[B.C.A.(Indore) 2011; B.C.A. (Meerut) 2010; B.B.A. (Meerut) 2002]

2. A national car rental service has a surplus of one car in each of the cities 1, 2, 3, 4, 5, 6
and a deficit of one car in each of the cities 7, 8, 9, 10, 11, 12. The distances in miles
between cities with a surplus and cities with a deficit are displayed in matrix below.
To
7 8 9 10 11 12
1 41 72 39 52 25 51
2 22 29 49 65 81 50
From 3 27 39 60 51 32 32
4 45 50 48 52 37 43
5 29 40 39 26 30 33
6 82 40 40 60 51 30
[B.C.A. (Bhopal) 2005, 2007]
 160

3. Six machines M1, M2 , M3 , M4 , M5 and M6 are to be located in six places

P1, P2 , P3 , P4 , P5 and P6 . Cij the cost of locating machine Mi at place Pj is given in the
matrix below :
P1 P2 P3 P4 P5 P6

M1 20 23 18 10 16 20

M2 50 20 17 16 15 11

M3 60 30 40 55 8 7

M4 6 7 10 20 25 9

M5 18 19 28 17 60 70

M6 9 10 20 30 40 55

Formulate the L.P. model to determine an optimal assignment. Write the objective
function and the constraints in detail. Define any symbol used. Find an optimal
layout by assignment technique of linear programming.
[B.C.A. (Kashividhyapeeth) 2008]

4. Solve the following assignment problems.

Jobs

1 2 3 4 5

1 2.5 5 1 6 2

2 2 5 1.5 7 3

(a) Machines 3 3 6.5 2 8 3

4 3.5 7 2 9 4.5

5 4 7 3 9 6

6 6 9 5 10 6

Place

1 2 3 4 5

1 15 10 25 25 10

2 1 8 10 20 2

(b) Machines 3 8 9 17 20 10

4 14 10 25 27 15

5 10 8 25 27 12
 161

I II III IV V
1 11 17 8 16 20
(c) 2 9 7 12 6 15
3 13 16 15 12 16

4 21 24 17 28 26
5 14 10 12 11 15

[B.B.A. (Delhi) 2007; B.B.A. (Meerut) 2008]

Rider

R1 R2 R3 R4 R5

H1 5 3 4 7 1

H2 2 3 7 6 5

(d) Horse H3 4 1 5 2 4

H4 6 8 1 2 3

H5 4 2 5 7 1

1 2 3 4 5
1 −2 −4 −8 −6 −1
2 0 −9 −5 −5 −4
(e) 3 −3 −8 −9 −2 −6
4 −4 −3 −1 0 −3
5 −9 −5 −8 −9 −5

5. Assign four trucks 1, 2, 3 and 4 to vacant spaces 7, 8, 9, 10, 11 and 12 so that the
distance travelled is minimized. The matrix below shows the distance.

1 2 3 4

7 4 7 3 7

8 8 2 5 5

9 4 9 6 9

10 7 5 4 8

11 6 3 5 4

12 6 8 7 3

[B.C.A. (Bundelkhand) 2003]

 162

6. A company has 5 jobs to be done. The following matrix shows the return in rupees
on assigning ith (i = 1, 2, 3, 4, 5) machines to the j th job ( j = A, B, C, D, E). Assign the
five jobs to the five machines so as to maximize, the total expected profit.

A B C D E

1 5 11 10 12 4
2 2 4 6 3 5
3 3 12 5 14 6
4 6 14 4 11 7
5 7 9 8 12 5

7. Alpha Corporation has four plants each of which can manufacture any of the four
products. Production costs differ from plant to plant as do sales revenue. From the
following data, obtain which product each plant should produce to maximize profit?
Sales Revenue (` 1000) Production Cost (` 1000)
Product Product

Plant ↓ 1 2 3 4 Plant 1 2 3 4
A 50 68 49 62 A 49 60 45 61
B 60 70 51 74 B 55 63 45 69
C 55 67 53 70 C 52 62 49 68
D 58 65 54 69 D 55 64 48 66

8. There are 3 persons P1, P2 and P3 and 5 jobs J1, J2 , … , J5 . Each person can do only
one job and a job is to be done by one person only using Hungarian method. Find
which 2 jobs should be left undone in the following cost minimizing assignment
problem.

(a) J1 J2 J3 J4 J5

P1 7 8 6 5 9
P2 9 6 7 6 10
P3 8 7 9 5 6

(b) J1 J2 J3 J4 J5 J6

P1 10 9 11 12 8 5

P2 12 10 9 11 9 4

P3 8 11 10 7 12 6

P4 10 7 8 10 10 5
 163

9. A company is faced with the problem of assigning 4 machines to 6 different jobs.

The profits are estimated as follows:

Machines

A B C D

1 3 6 2 6

2 7 1 4 4

Jobs 3 3 8 5 8

4 6 4 3 7

5 5 2 4 3

6 5 7 6 4

Solve the maximize problem. [B.B.A. (I.G.N.O.U.) 2001, 2003, 2006, 2007]

10. Define assignment problem. or What is the assignment problem?

[B.C.A. (Avadh) 2008; B.B.A. (Meerut) 2002, 2006]

11. What is an unbalanced assignment problem?

[B.C.A. (Kanpur) 2006; B.B.A. (Garhwal) 2001, 2005; B.B.A. (Agra) 2002, 2005, 2006]

12. Explain the objective of an assignment problem.

[B.C.A. (Lucknow) 2006; B.B.A. (Meerut) 2002, 2003]

13. Discuss practical application of an assignment problem. [B.B.A. (Meerut) 2004]

14. What do you mean by an assignment problem ? What is the objective in each
problem ? [B.B.A. (Meerut) 2006]

15. Explain the difference between a transportation problem and an assignment

problem.
[B.C.A. (Agra) 2004, 2006, 2011; B.C.A. (Rohilkhand) 2008;
B.C.A. (Lucknow) 2005, 2010; B.B.A. (Meerut) 2008]
 164

A nswers
1. A → II , B → V , C → III , D − I , E → IV .

2. 1 → 11, 2 → 8, 3 → 7, 4 → 9, 5 → 10, 6 → 12, Min. z = 185 miles.

3. M1 → P4 , M2 → P6 , M3 → P5 , M4 → P3 , M5 → P1, M6 → P2 , Min (z) = 67.

4. (a) 1 → 4, 2 → 1, 3 → 5, 4 → 3, 5 → 2, 6-Dummy, Min (z) = ` 2,000.

(b) 1 → 5, 2 → 3, 3 → 4, 4 → 2, 5 → 1, Min. (z) = 60.

(c) 1 → I , 2 → IV , 3 → V , 4 → III , 5 → II , Min (z) = 60.

(d) H1 → R5 , H2 → R1, H3 → R4 , H4 → R3 , H5 → R2 , Min. (z) = 8.

(e) 1 → 3, 2 → 2, 3 → 5, 4 → 1, 5 → 4, Min (z) = − 36.

5. 7 → 3, 8 − 2, 9 → 1, 12 → 4, Min (z) = 12.

6. 1 → C, 2 → E, 3 → D, 4 → B, 5 → A, Max. (z) = ` 50.

7. A → 2, B → 4, C → 1, D → 3 , Max. (z) = 22,000 .

8. (a) (i) P1 → J3 , P2 → J2 , P3 → J4 : Jobs J1 and J5 left undone.

(ii) P1 → J4 , P2 → J2 , P3 → J5 : Jobs J1 and J3 left undone.

In both cases minimum cost is 17.

(b) P1 → J5 , P2 → J6 , P3 → J4 , P4 → J2 ;

Jobs J1 and J3 left undone.

9. 2 → A, 3 → B, 4 → D, 6 → C, Job 1 and 5 are left undone, Max (z) = 28.

mmm
 165

C HAPTER
5
Linear Programming and
Graphical Methods

5.1 Linear Programming

[B.B.A. (Meerut) 2003]
The terms ‘Linear programming’ has been defined by various authors as given below:

According to Dantzig “Linear programming is a programming of inter-dependent activities in a

linear structure”.

According to Richard H. Leftwich “Linear programming is the simplest and most widely used
mathematical technique that has come into vague since world war II”.

5.1.1 Important Applications of Linear Programming

[B.C.A. (Agra) 2005, 2008; B.B.A. (Meerut) 2003]

The main areas of the applications of linear programming technique are as follows:
1. Linear programming approach is being used in five year plans in the area such as
foodgrain storage, transportation, multi-level planning etc.
2. Air lines are also using linear programming in the selection of routes and allocation
of aircrafts to various routes.
3. Linear programming is also being used in Production Management, Marketing
Management, Human Resources etc.
 166

4. Indian Railways have applied linear programming in a number of areas.

5. The oil refineries, steel industry use programming technique for blending decisions
and for the moment of finished products.
6. Linear programming approach is being used in agriculture for crop rotation, mix of
cash crops and food crops and fertilizer mix.
7. The main problem of linear programming are: Diet problem, Production Line
Problem, Transportation Problem, Investment Problem, Staffing Problem, Product
Mix Problem, Advertising Media Selection Problem, Production Plan Problem,
Agricultural Problem, Food Management Problem etc.

5.1.2 Objectives of Using Linear Programming

[B.B.A. (Meerut) 2005]

The primary reason for using linear programming is to achieve the best allocation of some
resources. For example:
1. To select an investment portfolio from a variety at stock and bond investment
alternatives that maximize the return of investment.
2. To allocate the available machine, time and labour hours in each department along
with the raw material to the activities of producing different products which have
been scheduled and to determine the number of units of each of the products so as
to maximize the profit.
3. To determine which warehouse should ship how much product to which market so
that the total transportation costs are minimized.
In linear programming terminology, the maximization or minimization of a
quantity is referred to as the objective of the problem.

5.1.3 Main Characteristics of Linear Programming

1. Decision Variable: The decision variables refer to limited resources and are
denoted by x, y etc.
2. Objective Function: The linear function between decision variables which is to be
maximized or minimized is known as objective function of linear programming
problem.
[B.B.A. (Meerut) 2006]

3. Constraints: The restrictions or limits which are determined to make the solution
optimum (maximum or minimum) are called constraints.
[B.B.A. (Meerut) 2006]
4. Linearity Proportionality and Non-Negativity Conditions: The objective
function and all the constraints must be linear and the values of all the decision
variables should be non-negative. If any of the decision variables is unrestricted in
sign, a trick can be employed which will enforce the non-negativity condition
without changing the original information of the problem.
 167

5. Additivity: Two or more variables under consideration are additive. For example,
if one unit of product A required 12 kg of material in its manufacture and product B
requires 19 kg of material, then one unit of A and one unit of B require (12+19) kg
of material.
6. Divisibility: Divisibility means solutions may take any fractional value.
7. Deterministic or Certainly: All the coefficients used in a L.P.P. are completely
known. For example, profit per unit, quantity of resource, available etc.
8. Finiteness: The number of decision variables and constraints is finite.

5.1.4 Mathematical Form of Linear Programming Problem

[B.C.A. (Meerut) 2009,2011; B.B.A. (Meerut) 2003, 2009]

Let x1, x2 , x3 , … , x n be n decision variables. Then the mathematical form of linear

programming problem or so called general linear programming is as follows :

Determine the values of x1, x2 , … , x n which optimize

z = c1 x1 + c2 x2 + … + cn x n ...(1)
Subject to the constraints

a11 x1 + a12 x2 + a13 x3 + … + a1n x n (≤, =, ≥) b1


a21 x1 + a22 x2 + a23 x3 + … + a2 n x n (≤, =, ≥) b2

............................................................ ...(2)
............................................................

am1 x1 + am2 x2 + am3 x3 + … + amn x n (≤, =, ≥) bm

and non-negative restrictions

x1, x2 , … , x n ≥ 0 ...(3)

where all c's b's, a's are known constants.

The equation (1) is the objective function, the equation (2) are constraints and
equation (3) is non-negativity conditions (≤, =, ≥) means that any one of the three signs
may be there. The constraints which do not affect the possible solutions are known as
redundant constraints.
Or
A general linear programming problem can be stated as
n
Optimize (z) = ∑=1 c j x j
j

Subject to Σaij x j (≤, = , ≥) bi, i = 1, 2, … , m

and x j ≥ 0, j = 1, 2, … , n
Or
 168

The general linear programming problem can be written in matrix form as

Optimize z = CX
such that AX (≤ , = , ≥) B
and X ≥0
where C = [c1, c2 , … , cn] a row vector
X ′ = ( x1, x2 , … , x n) a row vector and X is a column vector

A = [aij]m × n be the matrix of the coefficient aij of order m × n,

i = 1, 2, 3, … m, j = 1, 2, 3, .. , n
′
B = [b1, b2 , … , bn] a row vector and B is column vector.

5.1.5 Formulation of Linear Programming Problem

While formulating a linear programming problem we proceed as follows:

Step 1: Identify the decision variables (the variables whose values are to be found out by
solving the L.P.P.) and assign the symbols (x1, x2 , x3 , … or x, y, z, …) to them.

Step 2: Identify the objective function (to be optimized) and represent it as a linear
function of the decision variables.
z = c1 x1 + c2 x2 + c3 x3 + …

Step 3: Identify all the restrictions or constraints in the problem regarding the resources
which are available upto or beyond certain limit and express them as linear equations
and/or inequalities in terms of decision variables in the following manner.
a11 x1 + a12 x2 + … + a1n x n (≤ , = , ≥) b1

Step 4: Since the negative values of decision variables do not have any valid physical
interpretation.
x i ≥ 0 ∀ i = 1, 2, 3, … .

Step 5: In the end write the objective function, linear constraints and non-negativity
conditions together to form a model of linear programming problem.

5.1.6 Limitations of Linear Programming

[B.B.A. (Meerut) 2005]

1. The basic assumption that objective function and constraints are linear may not
hold good in many practical situations.

2. When number of variables in given problem is large, then procedure become

complicated, then use of computer is necessary.

3. Sometime it may be possible that the solutions are in terms of fractions.

4. The technique of L.P.P. does not consider the problems related to uncertainity.
 169

S olved E xamples
Example 1: A company manufactures two types of bulbs A and B by using two machines
M1 and M2 . One bulb of type A requires 2 hours at machine M1 and 1 hour at machine M2 .One
bulb of type B requires 1 hour at machine M1 and 2 hours at M2 . The profit contributions from each
bulb of type A is ` 3 and from each bulb of type B is ` 2. The number of hours available per week on
machine M1 and M2 are 20 hours and 30 hours respectively. Formulate the above problem as a
linear programming problem, the aim of the company is to earn maximum profit.

Solution:

Step 1: The aim of the company is to earn maximum profit so the company desires to
know the number of bulbs to be produced per week to achieve this aim.

Let
x1 = The number of bulbs of type A to be produced per week

x2 = The number of bulbs of type B to be produced per week

Then total profit z = 3 x1 + 2 x2

Step 2: Objective Function: The objective function is

z = 3 x1 + 2 x2 which is to be maximized. ...(1)

Step 3: Constraints: Let us arrange the given information in tabular form to get the
constraints:

Time in Production of one Time available per week

Machine bulb (in hours) (in hours)

A B

M1 2 1 20

M2 1 2 30

Constraints for Machine M1: The time consumed on machine M1 for x1 bulbs of type A
and x2 bulbs of type B is 2 x1 + x2 .
This time cannot exceed the time available on machine M1.
Hence, 2 x1 + x2 ≤ 20 ...(2)
Constraints for Machine M2 : The time consumed on machine M2 for x1 bulbs of type
A and x2 bulbs of type B is x1 + 2 x2 .

This time cannot exceed the time available on machine M2 .

Hence, x1 + 2 x2 ≤ 30 ...(3)
 170

Step 4: Condition of Non-Negativity: Since x1 and x2 are the numbers of bulbs to be

produced these values can not be negative, hence x1 ≥ 0, x2 ≥ 0 ... (4)

Step 5: Thus, the mathematical form of the linear programming problem is

Maximize z = 3 x1 + 2 x2
such that 2 x1 + x2 ≤ 20
x1 + 2 x2 ≤ 30
x1, x2 ≥ 0

Example 2: A goldsmith manufactures necklace and bracelets. The total number of necklaces and
bracelets that he can handle per day is at most 24. It takes one hour to make a bracelet and half an
hour to make a necklace. It is assumed that he can work for a maximum of 16 hours a day. Further
the profit on a bracelet is ` 300 and the profit on a necklace is ` 100. Formulate this problem as a
linear programming problem so as to maximize the profit.

Solution:

Step 1: The aim of a goldsmith is to earn maximum profit so that the goldsmith desires to
know the number of necklaces and bracelets to be handled per day.

Let x1 = The number of necklace to be manufactured per day

x2 = The number of bracelet to be manufactured per day.

Then total profit z = 100 x1 + 300 x2

Step 2: Objective Function: The objective function is

z = 100 x1 + 300 x2 , which is to be maximized ...(1)

Step 3: Constraints: Since it takes half an hour to make one necklace, so the time
1
required to make x1 necklaces = x1 hours.
2

Again it takes one hour to make one bracelet, so the time required to make x2 bracelets
= x2 hours.

Therefore, total time required to make x1 necklaces and x2 bracelets

1 
=  x1 + x2  hours ...(2)
2 

Since total available per day is 16 hours, therefore

1
x1 + x2 ≤ 16
2
or x1 + 2 x2 ≤ 32 ...(3)
The total number of necklaces and bracelets that the goldsmith can manufactures in a
day is at most 24
∴ x1 + x2 ≤ 24 ...(4)
 171

Step 4: Condition of Non-Negativity: Also the number of necklaces and bracelets

manufactured can never be negative, therefore

x1 ≥ 0, x2 ≥ 0 ...(5)

Step 5: Thus, the linear programming problem formulated from the given problem is:
Maximize z = 100 x1 + 300 x2

Subject to the constraints

x1 + 2 x2 ≤ 32
x1 + x2 ≤ 24
x1 ≥ 0, x2 ≥ 0

Example 3: A manufacturer of medicines is proposed to prepare a production plan for medicines A

and B. There are sufficient ingredients available to make 20,000 bottles of medicine A and 40,000
bottles of medicines B, but there are only 45,000 bottles into which either of the medicine can be
filled. Further, it takes three hours to prepare, enough material to fill, 1,000 bottles of medicine A
and one hour for medicine B and there are 66 hours available for this operation. The profit is ` 8 per
bottles of medicine A and ` 7 per bottle of medicine B. Formulate this problem as L.P.P.

Solution:
Step 1: The decision variables are:
x1 = number of bottles of medicines A
and x2 = number of bottles of medicines B.
produced by the manufacturer.
The profit on medicine A = ` 8 per bottle.
The profit on medicine B = ` 7 per bottle.
Hence, the total profit is z = 8 x1 + 7 x2 .

Step 2: Objective Function: The objective function is z = 8 x1 + 7 x2 which is to be

maximized.
Step 3: Constraint
(i) Time Constraint: Since it takes 3 hours to prepare 1,000 bottles of medicine.
3
A, hence the time required for x1 bottles of medicines A is x1 hours.
1000
1
Similarly, time required for x2 bottles of medicines B is x2 hours.
1000
3 1
∴ The total time required = x1 + x2
1000 1000
But the total time available is 66 hours.
3 1
Hence, the constraint is x1 + x2 ≤ 66
1000 1000
 172

or 3 x1 + x2 ≤ 66,000

(ii) Number Constraints: There are only 45,000 bottles available for filling medicines
A and B. Hence, the constraints is
x1 + x2 ≤ 45,000

(iii) Ingredients Constraints: Since there are sufficient ingredients available to make
20,000 bottles of medicines A and 40,000 bottles of B, hence the constraints are
x1 ≤ 20,000, x2 ≤ 40,000

Step 4: The condition of non-negativity:

x1, x2 ≥ 0

Step 5: Thus, the formulation of L.P.P. is:

Maximize z = 8 x1 + 7 x2

subject to 3 x1 + x2 ≤ 66,000

x1 + x2 ≤ 45,000

x1 ≤ 20,000

x2 ≤ 40,000
and x1, x2 ≥ 0

Remark: We may also take the number of bottles produced as x1 thousands and x2
thousand respectively, then we have:
Maximize z = 8 x1 + 7 x2

subject to 3 x1 + x2 ≤ 66

x1 + x2 ≤ 45

x1 ≤ 20

x2 ≤ 40
and x1, x2 ≥ 0

Example 4: A firm can produce three types A, B and C of cloths using three kinds, red, green and
blue wool. One unit length of type A cloth need 2 metres of red wool and 3 metres of blue wool. One unit
length of type B cloth need 3 metres of red wool, 2 metres of green wool and 2 metres of blue wool. One
unit of type C cloth need 5 metres of green wool and 4 metres of blue wool. The firm has only a stock of 8
metres of red wool, 10 metres of green wool and 15 metres of blue wool. It is assumed that the income
obtained from unit length of type A cloth is ` 3, of type B cloth is ` 5 and of type C cloth is ` 4.
Formulate this problem as a L.P.P. to have maximum income from, the cloth produced.
 173

Solution:

Step 1: Construct a table for clear understanding of the problem

Requirement of wool (in metres)

Wool available
Kinds of wool for cloth of type (in metres)
A B C
Red 2 3 0 8
Green 0 2 5 10

Blue 3 2 4 15
Income from one 3 5 4
unit length of cloth

Step 2: Objective Function: Let the decision variable be

x1 = the quantity produced of type A cloth (in metres)

x2 = the quantity produced of type B cloth (in metres)

x3 = the quantity produced of type C cloth (in metres)

one unit length is taken as one metre

Hence, income (in rupees) of the firm, (i.e., objective function) is

z = 3 x1 + 5 x2 + 4 x3
Step 3: Constraint
(i) Constraint for Red Wool: The quantity of red wool required to prepare x1 metres
of type A, x2 metres of type B, x3 metres of type C is
2 x1 + 3 x2 + 0 x3
Since, the stock of red wool is 8 metres so the constraint is
2 x1 + 3 x2 + 0 x3 ≤ 8
(ii) Constraint of Green Wool: The quantity of green wool required to prepare x1
metres of type A, x2 metres of type B and x3 metres of type C is
0. x1 + 2 x2 + 5 x3
Since the stock of green wool is 10 metres so the constraint is
0. x1 + 2 x2 + 5 x3 ≤ 10
(iii) Constraint for Blue Wool: The quantity of blue wool required to prepare x1
metres of type A, x2 metres of type B and x3 metres of type C is
3 x1 + 2 x2 + 4 x3
Since the stock of blue wool is 15 metres so the constraint is
3 x1 + 2 x2 + 4 x3 ≤ 15
 174

Step 4: The Condition of Non-Negativity: x1, x2 , x3 ≥ 0

Since firm can not produce negative quantity

Step 5: The Mathematical form of L.P.P.

Maximize z = 3 x1 + 5 x2 + 4 x3
such that 2 x1 + 3 x2 ≤ 8
2 x2 + 5 x3 ≤ 10
3 x1 + 2 x2 + 4 x3 ≤ 15
and x1, x2 , x3 ≥ 0

Example 5: A tyre factory produces three types of tyres T1, T2 , T3 . Three different types of
chemicals say C1, C2 , C3 are required for production. One T1 tyre needs 2 units of C1, 3 units of C3 ;
one T2 tyre needs 3 units of C1, 2 units of C2 and 2 units of C3 ; and one T3 tyre needs 5 units of C2
and 4 units of C3 . The factory has only stock of 20 units of C1, 25 units of C2 and 30 units of C3 .
Further the profit from the sale of one tyre T1 is ` 6, one tyre T2 is `10 and of one tyre T3 is ` 8 .
Assuming that the factory can sell all that it produces, formulate a linear programming problem to
maximize its profits. [B.C.A. (Rohilkhand) 2007]

Solution:

Step 1: Let the factory produce x1 tyres of type T1, x2 tyres of type T2 and x3 tyres of type T3 .

The given information can be put in a tabular form as given below:

Tyre T1 Tyre T2 Tyre T3 Total chemical available

Chemical C1 2 3 0 20

Chemical C2 0 2 5 25

Chemical C3 3 2 4 30

Profit in ` (per tyre) 6 10 8

Step 2: Objective Function: The total profit Z in ` is given by

Z = 6 x1 + 10 x2 + 8 x3 ...(1)

Step 3: Constraints: The total quantity of chemical C1 required

= (2 x1 + 3 x2 ) units.

Since the factory has a stock of 20 units of chemical C1, therefore we have

2 x1 + 3 x2 ≤ 20 ...(2)
 175

Similarly, considering the total quantity of the chemicals C2 and C3 required, we have
2 x2 + 5 x3 ≤ 25 ...(3)
and 3 x1 + 2 x2 + 4 x3 ≤ 30 ...(4)

Step 4: Condition of Non-Negativity:

x1 ≥ 0, x2 ≥ 0, x3 ≥ 0 ...(5)

Since the factory cannot produce negative quantities.

Step 5: Hence, the linear programming problem formulated from the given problem is as
follows:
Maximize Z = 6 x1 + 10 x2 + 8 x3
s.t. 2 x1 + 3 x2 ≤ 20, 2 x2 + 5 x3 ≤ 25,
3 x1 + 2 x2 + 4 x3 ≤ 30
and x1 ≥ 0, x2 ≥ 0

Example 6: The objective of a diet problem is to ascertain the quantities of certain foods that should
be eaten to meet certain nutritional requirement at a minimum cost. The consideration is limited to
milk, green vegetables, eggs, and to vitamins A, B, C. The number of milligrams of each of these
vitamins contained within a unit of each food and their daily minimum requirement along with the
cost of each food is given in the table below:

Litres of Kg. of Dozen of eggs Minimum daily

Vitamin
milk vegetables requirement

A 1 1 10 1 mg.

B 100 10 10 50 mg.

C 10 100 10 10 mg.

Cost in ` 20 10 8

Formulate a linear programming problem for this diet problem. [B.C.A. (Kanpur) 2009]

Solution: Let the daily diet consist of x1 litres of milk, x2 kg. of vegetables and x3 dozens
of eggs.

∴ the total cost Z per day in ` is given by

Z = 20 x1 + 10 x2 + 8 x3 ...(1)

Total amount of vitamin A in the daily diet is

( x1 + x2 + 10 x3 ) mg

which should be at least equal to 1 mg., therefore

x1 + x2 + 10 x3 ≥ 1 ...(2)
 176

Similarly, considering the total amounts of vitamins B and C in the daily diet, we have
100 x1 + 10 x2 + 10 x3 ≥ 50 ...(3)
and 10 x1 + 100 x2 + 10 x3 ≥ 10 ...(4)
Since the quantities of different food items to be consumed cannot be negative, therefore
we have
x1 ≥ 0, x2 ≥ 0, x3 ≥ 0
Hence, the linear programming problem formulated from the given diet problem is:
Minimize Z = 20 x1 + 10 x2 + 8 x3
subject to the constraints
x1 + x2 + 10 x3 ≥ 1
100 x1 + 10 x2 + 10 x3 ≥ 50
10 x1 + 100 x2 + 10 x3 ≥ 10
and the non-negative restrictions
x1 ≥ 0, x2 ≥ 0, x3 ≥ 0

Example 7: According to the medical experts it is necessary for an adult to consume at least 75 gm
of proteins, 85 gm of fats and 300 gm of carbohydrates daily. The following table gives the analysis
of the food items readily available in the market with their respective costs.

Food value (in gm.) per 100 gm

Food Type Costs in per kg.
Proteins Fats Carbohydrates

A 18.0 15.0 — 3.0

B 16.0 4.0 7.0 4.0

C 4.0 20.0 2.5 2.0

D 5.0 8.0 40.0 1.5

Minimum daily 75.0 85.0 300.0

requirement

Formulate a linear programming problem for an optimum diet.

Solution: Let the daily diet consist of x1 kg of food A, x2 kg of food B, x3 kg of food C and
x4 kg of food D.
Then the total cost per day in ` is
Z = 3 x1 + 4 x2 + 2 x3 + 1.5 x4 ...(1)

Total amount of proteins in the daily diet is

(180 x1 + 160 x2 + 40 x3 + 50 x4 )
 177

Since the minimum daily requirement of proteins is 75 gm, therefore we have

180 x1 + 160 x2 + 40 x3 + 50 x4 ≥ 75 ...(2)

Similarly, considering the total amounts of fats and carbohydrates in the diet, we have

150 x1 + 40 x2 + 200 x3 + 80 x4 ≥ 85 ...(3)

and 70 x2 + 25 x3 + 400 x4 ≥ 300 ...(4)

Since, the daily diet cannot contain quantities with negative values of any food item,
therefore

x1 ≥ 0, x2 ≥ 0, x3 ≥ 0, x4 ≥ 0 ...(5)

Hence, the linear programming problem formulated for the given diet problem is :

Minimize Z = 3 x1 + 4 x2 + 2 x3 + 1.5 x4

subject to the constraints

180 x1 + 160 x2 + 40 x3 + 50 x4 ≥ 75

150 x1 + 40 x2 + 200 x3 + 80 x4 ≥ 85

70 x2 + 25 x3 + 400 x4 ≥ 300

and the non-negative restrictions

x1 ≥ 0, x2 ≥ 0, x3 ≥ 0, x4 ≥ 0

Example 8: A firm produces three products. These products are processed on three different
machines. The time required to manufacture one unit of each of the three products and the daily
capacity of the three machines are given in the table below:

Time per unit (minutes)

Machine capacity
Machine
Product 1 Product 2 Product 3 (minutes/day)

M1 2 3 2 440

M2 4 - 3 470

M3 2 5 - 430

It is required to determine the daily number of units to be manufactured for each product. The profit
per unit for product 1, 2 and 3 is ` 4, ` 3 and ` 6 respectively. It is assumed that all the amounts
produced are consumed in the market. [B.C.A. (Meerut) 2006]

Solution: Formulation of Linear Programming Model

Step 1: From the study of the situation find the key-decision to be made. In this
connection looking for variables helps considerably. In the given situation key decision is
to decide the extent of products 1, 2 and 3 as the extents are permitted to vary.
 178

Step 2: Assume symbols for variable quantities noticed in step 1. Let the extent
(amounts) of products 1, 2 and 3 manufactured daily be x1, x2 and x3 respectively.

Step 3: Express the feasible alternatives mathematically in terms of variables. Feasible

alternatives are those which are physically, economically and financially possible. In the
given situation feasible alternatives are sets of values of x1, x2 and x3 .
Where x1, x2 , x3 ≥ 0 ...(1)

Since negative production has no meaning, is not feasible.

Step 4: Mention the objective quantitatively and express it as a linear function of

variables. In the present situation objective is to maximize the profit
i.e., maximize Z = 4 x1 + 3 x2 + 6 x3 ...(2)

Step 5: Put into the words the influencing factors or constrains. These occur generally
because of constraints on availability (resources) or requirements (demands). Express
these constraints also as linear equalities/inequalities in terms of variables.

Here, constraints are on the machine capacities and can be mathematically expressed as

2 x1 + 3 x2 + 2 x3 ≤ 440

4 x1 + 0 x2 + 3 x3 ≤ 470

2 x1 + 5 x2 + 0 x3 ≤ 430 ...(3)

Example 9: A company has two grades of inspectors, 1 and 2 to undertake quality control
inspection. At least 1,500 pieces must be inspected in an 8-hour day. Grade 1 inspector can check
20 pieces in an hour with an accuracy of 96%.

Grade 2 inspector checks 14 pieces in an hour with an accuracy of 92%.

The daily wages of grade 1 inspector are ` 5 per hour while those of grade 2 inspector are ` 4 per
hour. Any error made by an inspector costs ` 3 to the company. If there are, in all, 10 grade 1
inspector and 15 grade 2 inspectors in the company, find the optimal assignment of inspectors that
minimize the daily inspection cost.

Solution: Formulation of L.P. Model.

Step 1: The key decision to be made is to determine the number of grade 1 and grade 2
inspectors for assignments.

Step 2: Let x1 and x2 represent the number of these inspectors.

Step 3: Feasible alternatives are sets of values of x1 and x2 , where x1 ≥ 0, x2 ≥ 0.

Step 4: The objective to minimize the daily cost of inspection. Now the company has to
incur two types of costs wages paid to the inspectors and the cost of their inspection
errors. The cost of grade 1 inspector/hour is
` (5 + 3 × 0.04 × 20) = ` 7.40
 179

Similarly, cost of grade 2 inspector/hour is

` (4 + 3 × 0 .08 × 14) = ` 7.36

∴ The objective function is minimize

z = 8 (7 .40 x1 + 7 .36 x2 ) = 59 .20 x1 + 58 .88 x2

Step 5: Constraints are

on the number of grade 1 inspectors : x1 ≤ 10,

on the number of grade 2 inspectors : x2 ≤ 15,

on the number of pieces to be inspected daily:

20 × 8 x1 + 14 × 8 x2 ≥ 1500
,

or 160 x1 + 112 x2 ≥ 1,500

Example 10: A person wants to decide the constituents of a diet which will fulfil his daily
requirements of proteins, fats and carbohydrates at the minimum cost. The Choice is to be made from
four different types of foods. The yield per unit of these foods are given in table:

Yield per unit Cost per

Food Type
unit (`)
Proteins Fats Carbohydrates

1 3 2 6 45
2 4 2 4 40
3 8 7 7 85
4 6 5 4 65

Minimum 800 200 700

requirement

Formulate linear programming model for the problem.

Solution: Formulation of L.P. Model:

Step 1: Key decision is to determine the number of units of food of type 1, 2, 3 and 4 to
be used.

Step 2: Let these units be x1, x2 , x3 and x4 respectively.

Step 3: Feasible alternatives are sets of values of x j

where x j ≥ 0, j = 1, 2, 3, 4 ...(1)
Step 4: Objective is to minimize the cost i.e.,

minimize z = ` (45 x1 + 40 x2 + 85 x3 + 65 x4 ) ...(2)

 180

Step 5: Constraints are on the fulfilment of the daily requirements of the various
constituents:

i.e. , for proteins 3 x1 + 4 x2 + 8 x3 + 6 x4 ≥ 800 


for fats 2 x1 + 2 x2 + 7 x3 + 5 x4 ≥ 200  ...(3)
and for Carbohydrates, 6 x1 + 4 x2 + 7 x3 + 4 x4 ≥ 700 

Thus, the L.P. Model is to determine the number of units of x1, x2 , x3 and x4 that
minimize equation (2) subject to constraints (3) and non-negativity restrictions (1).

Example 11: A firm manufactures two types of products A and B and sells them at a profit of ` 2
on type A and ` 3 on type B. Each product is processed on two machines E and F. Type A requires
one minutes of processing time on E and two minutes on F. Type B requires one minute on E and one
minute on F. The machine E is available for not more than 6 hours 40 minutes while machine F is
available for 10 hours during any working day. Formulate the problem as a linear programming
problem. [B.C.A. (Meerut) 2010]

Solution: Let the manufacturer produce x1 units of the product of type A and x2 units of
the product of type B.
The given information can be systematically arranged in the form of the following table:

Time of processing (minutes per unit)

Time available
Machine
(minutes)
Type A (x1 units) Type B (x2 units)

E 1 1 400

F 2 1 600

Profit per unit `2 `3

Since, the machine E takes 1 minute time for processing a unit of type A and 1 minute
time for processing a unit of type B, therefore the total time required on machine E is
( x1 + x2 ) minutes.

But the machine E is available for not more than 6 hours and 40 minutes = 400 minutes,
therefore we have

x1 + x2 ≤ 400

Similarly, the total time (in minutes) required on machine F is 2 x1 + x2 .

Since, the machine F is available for not more than 10 hours = 600 minutes, therefore we
have
2 x1 + x2 ≤ 600

Since, it is not possible to produce negative quantities, so we have x1 ≥ 0, x2 ≥ 0.

 181

The profit on type A is ` 2 per unit so the profit on selling x1 units of type A will be ` 2 x1.
Similarly, the profit on selling x2 units of type B will be ` 3 x2 .

Therefore, the total profit (in `) on selling x1 units of type A and x2 units of type B is

Z = 2 x1 + 3 x2

Hence, the required linear programming problem formulated for the given problem is

Maximize Z = 2 x1 + 3 x2

subject to the conditions

x1 + x2 ≤ 400,

2 x1 + x2 ≤ 600

and x1 ≥ 0, x2 ≥ 0

Example 12: Sun chemical company is producing two products A and B. The processing times are
3 and 4 hours per unit for A an operations one and two respectively and 4 hours and 5 hours per unit
of B an operations on one and two respectively. The available time is 18 hours and 21 hours for
operation one and two respectively. The product A can be sold for ` 3 profit per unit and B of ` 8
profit per unit, solve for maximum program only formulate the problem.
[B.B.A. (Meerut) 2006]

Solution: We have from given problem.

Time

A B Available time

Operation I 3 4 18 hrs

Operation II 4 5 21 hrs

Profit per unit ` 3 8

Let x1 and x2 be the product of type A and B. Then objective function will be

Max (Z ) = 3 x1 + 8 x2

subject to the constraints:

3 x1 + 4 x2 ≤ 18

4 x1 + 5 x2 ≤ 21

∀ x1, x2 ≥ 0
 182

P roblem S et
1. A firm manufactures two types of products A and B and sells them at a profit of ` 2
on type A and ` 3 of type B. Each product is produced on two machines M and N.
Type A requires one minute of processing time of M and 2 minutes of processing
time of N. Type B requires one minute of processing time of M and one minute of
processing time of N. The machine M is available for not more than 400 minutes
whereas machine N is available for 10 hours. What is the linear programming
formulation for this problem?
2. One kind of hens can be bought for ` 4 .00 each but another kind cost `10 .00. First
kind of hens lay 3 eggs per week and another 5 eggs per week each egg worth being
60 paise. A hen cost ` 2 .00 per week to feed. A person has only `160 to spend for
hens and desire to gain a profit of more than `12 per week, while he cannot house
more than 20 hens. Formulate this problem as a L.P.P to determine the number of
each kind of hens the person should buy.
3. A company has undertaken a contract to supply a customer with at least 260 units
in total of two products X and Y during the next month. At least 50% of the total
output must be units of X. The products are each made by 2 grades of labour as
follows:

‘X’ ‘Y ’

Grade A labour 4 hours 6 hours

Grade B labour 4 hours 2 hours

Although additional labour can be made available at short notice, the company
wishes to make use of 1200 hours of Grade A labour and 800 hours of Grade B
labour which has already been assigned to working on the contract next month.
The total variable cost per unit is `120 for X and `100 per Y . The company wishes
to minimize expenditure on the contract next month. How much of X and Y should
be supplied in order to meet the terms of the contract? Formulate linear
programming model for this problem.
4. A seller buys some tables and chairs. He has ` 5,000 to invest and a space to store at
most 60 pieces. A table costs him ` 250 and a chair ` 50. He can sell a table at a
profit of ` 50 and a chair at a profit of `15. Assuming that he can sell all the pieces
that he buys, prepare a mathematical formulation of this linear programming
problem to determine the number of pieces of each type to gain maximum profit.
5. A toy company manufactures two types of toys A and B, say. Each toy of type B
takes twice as long to produce as one of type A. Company can produce a maximum
of 2,000 toys per day if it produces only A type toys. The supply of plastic is
sufficient to produce 1,500 toys per day of both type. Type B toy requires a dress of
which there only 600 per day available. If the company makes a profit of ` 3 .00 and
 183

` 5 .00 per toy respectively on toy of type A and B, then how many of each of toy
should be produced per day in order to maximize the profit? Formulate this
problem as linear programming problem.
6. A man makes two types of furniture, chairs and tables. Profits are ` 20 per chair and
` 30 per table. Both products are processed on three machines M1, M2 , M3 . The
time required for each product in hours and total time available in hours for each
machine are as follows:

Machine Time required to Time required to Available time

produce a chair produce a table (hour)
(hour) (hour)

M1 3 3 36

M2 5 2 50

M3 2 6 60

Formulate this problem as a linear programming problem to maximize the profit.

7. A dairy has two plants each of which produces and supplies two products : milk and
butter. Each plant can work upto 16 hours a day. In plant-I, it takes 3 hours to
prepare from powder and pack 1,000 litres of milk and 1 hour to prepare and pack
100 kg of butter. In plant-II, it takes 2 hours to prepare and pack 1,000 litres of
milk and 1.5 hours to prepare and pack 100 kg of butter. In plant-I it costs `15,000
to prepare and pack 100 litres of milk and ` 28000
, to prepare and pack a 100 kg of
butter, whereas these costs are `18000, and ` 26,000 respectively in plant-II. The
dairy is obliged to produce daily at least 10,000 litres of milk and 800 kg of butter.
Formulate this as a L.P.P. to find out as to how should the company organise its
production so that the required amount of the products be obtained at minimum
cost.
8. A furniture dealer deals in two items, viz. tables and chairs. He has `10,000 to
invest and a space to store at most 60 pieces (including both tables and chairs) A
table cost to him ` 500 and a chair `100. He can sell all the items that he buys,
earning a profit of ` 50 for each table and `15 for each chair. Formulate this
problem as a L.P.P. so that he maximizes the profit.
9. A dietician decides a certain minimum intake of vitamins A, B and C for a family.
The minimum daily needs of the vitamins A, B and C are 30, 20, 16 units
respectively. For the supply of these, the dietician depends on two types of foods X
and Y . The first one gives 7, 5, 2 units per gram of vitamins A, B and C respectively.
The second one gives 2, 4, 8 units per gram of these vitamins respectively. The first
food costs ` 2 per gram and the second ` 1 per gram.
How many grams of each food stuff should the family buy everyday to keep the
food expense at a minimum? Formulate a linear programming problem for this
problem.
 184

10. A furniture firm manufactures chairs and tables, each requiring the use of three
machines A, B and C. Production of one chair requires 2 hours on machine A,1 hour
on machine B and 1 hour on machine C. Each table requires 1 hour each on
machines A and B and 3 hours on machine C. The profit realized by selling one
chair is ` 30 while for a table the figure is ` 60. The total time available per week on
machine A is 70 hours, on machine B is 40 hours and on machine C is 90 hours.
How many chairs and tables should be made per week so as to maximize the profit?
Formulate a mathematical model for the problem.
11. A diet is to contain at least 4000 units of carbohydrates, 500 units of fat and 300
units of protein. Two foods A and B are available. Food A costs ` 2 per unit and food
B costs ` 4 per unit. A unit of food A contains 10 units of carbohydrates, 20 units of
fat and 15 units of protein. A unit of food B contains 25 units of carbohydrates, 10
units of fat and 20 units of protein. Formulate the problem as a L.P.P. so as to find
the minimized cost for a diet consists of a mixture of these two foods and also meet
the minimum nutrition requirements.
12. A company manufactures two kinds of leather purses, A and B. A is a high quality
purse and B is lower quality. The sales of each of these purses A and B earn profit of
` 4 and ` 3 respectively. Each purse of type A requires twice as much time as a purse
of type B, and if all purses are of type B, the company could make 1000 purses per
day. The supply of leather is sufficient for only 800 purses per day (both A and B
combined). Purse A requires a fancy buckle and only 400 buckle per day are
available. There are only 700 buckles available for purse B. What should be the
daily production of each type of purse to get the maximum profit? Formulate the
problem as a L.P.P.
[B.C.A. (Meerut) 2006]
13. A city hospital has the following minimal daily requirements for nurses:

Clock time Minimum number of

Period
(24 hr day) nurses required

1 6 A. M. — 10 A. M. 2

2 10 A. M. — 2 P. M. 7

3 2 P. M. — 6 P. M. 15

4 6 P. M. — 10 P. M. 8

5 10 P. M. — 2 A. M. 20

6 2 A. M. — 6 A. M. 6

Nurses report to the hospital at the beginning of each period and work for 8
consecutive hours. The hospital wants to determine the minimum number of
nurses available for each period. Formulate this problem as a L.P.P. by setting up
appropriate constraints and objective function.
14. Write note on the use of linear programming problem in the business.
[B.C.A. (Rohilkhand) 2002, 2004, 2007]
 185

A nswers
1. Maximize z = 2 x1 + 3 x2 2. Maximize z = − 0 .2 x + y
such that x1 + x2 ≤ 400 subject to 4 x + 10 y ≤ 160
2 x1 + x2 ≤ 600 x + y ≤ 20
and x1, x2 ≥ 0 − 0 .2 x + y ≥ 12
and x > 0, y ≥ 0

3. Minimize z = 120 x + 100 y 4. Maximize z = 50 x1 + 15 x2

subject to 4 x + 6 y ≥ 1200
, such that 250 x1 + 50 x2 ≤ 5,000
4 x + 2 y ≥ 800 x1 + x2 ≤ 60
x + y ≥ 260 and x1, x2 ≥ 0
and x, y ≥0

5. Maximize z = 3 x1 + 5 x2 6. Maximize z = 20 x + 30 y
subject to x1 + 2 x2 ≤ 2000 subject to 3 x + 3 y ≤ 36
x1 + x2 ≤ 1500 5 x + 2 y ≤ 50
x2 ≤ 600 2 x + 6 y ≤ 60
and x1 ≥ 0, x2 ≥ 0 x, y ≥ 0

7.  x11 = Number of times 1,000 litre of milk is prepared

Plant-I 
 x12 = Number of times 100 kg of butter is prepared
 x21 = Number of times 1,000 litre of milk is prepared
Plant-II 
 x22 = Number of times 100 kg of butter is prepared
Minimize z = 15000 x11 + 28000 x12 + 18000 x21 + 26000 x22
subject to 3 x11 + x12 ≤ 16
2 x21 + 5 x22 ≤ 16
x11 + x21 ≥ 10
x21 + x22 ≥ 8
and x11, x12 , x21, x22 ≤ 0

8. Maximize z = 50 x + 15 y,
subject to the constraints 5 x + y ≤ 100
x + y ≤ 60
and the non-negative restrictions x ≥ 0, y ≥ 0.

9. Minimize z = 2 x + y
subject to the constraints 7 x + 2 y ≥ 30
5 x + 4 y ≥ 20
2 x + 8 y ≥ 16
and the non-negative restrictions x ≥ 0, y ≥ 0.
 186

10. Maximize z = 30 x + 60 y
subject to the constraints 2 x + y ≤ 70
x + y ≤ 40
x + 3 y ≤ 90
and the non-negative restrictions x ≥ 0, y ≥ 0.

11. Minimize z = 2 x + 4 y
subject to the constraints 10 x + 25 y ≥ 4000
20 x + 10 y ≥ 500
15 x + 20 y ≥ 300
and the non-negative restrictions x, y ≥ 0

12. Maximize z = 4 x + 3 y
subject to the constraints 2 x + y ≤ 1000
x + y ≤ 800
x ≤ 400
y ≤ 700
and the non-negative restrictions x ≥ 0, y ≥ 0.

13. Minimize z = x1 + x2 + x3 + x4 + x5 + x6
subject to the constraints x1 + x2 ≥ 7, x2 + x3 ≥ 15
x3 + x4 ≥ 8, x4 + x5 ≥ 20
x5 + x6 ≥ 6, x6 + x1 ≥ 2
with non-negativity restrictions : x1, x2 , x3 , x4 , x5 , x6 ≥ 0.

5.2 Solution of a Linear Programming Problem

To obtain a solution of a L.P.P. means to determine the values of the different decision
variables, viz. x1, x2 , x3 … depicted in its model.

Definition
“A set of values of decision variables which satisfies the linear constraints of a L.P.P. is known as its
solution”.

“A set of values of decision variables which satisfies the linear constraints and the non-negativity
conditions of a L.P.P. is called its Feasible Solution (F.S.)”. [B.C.A. (Meerut) 2009,2011]

Or
“A solution of a L.P.P. is called feasible solution if it satisfies the non-negativity conditions. A
solution which does not satisfy the non-negativity condition of L.P.P. is called a non-feasible
solution”.
 187

“Any feasible solution which optimizes (minimize or maximize) the objective function of a L.P.P. is
called its optimum solution”. [B.C.A. (Meerut) 2009]

“If a feasible solution minimize the objective function it is called minimum feasible solution. If the
feasible solution maximizes the objective function it is called maximum feasible solution”.

“If the value of objective function z can be increased or decreased indefinitely such solutions are called
unbounded solutions”.

Note:
In solving a linear programming problem the solutions may be non-feasible, multiple
feasible and unbounded feasible.

5.2.1 Methods of Solution of a Linear Programming Problem

There are two methods of solving a linear programming problem, viz., (1) Graphic
method and (2) Simplex method.

5.2.1.1 Graphic Method

A linear programming problem which involves only two decision (or decisive) variables,
viz., x1 and x2 can be easily solved by graphic method using a cartesian space (i.e.
rectangular co-ordinate system). One decision variable is taken along the x-axis and
another along the y-axis of the co-ordinate system. Due to non-negativity conditions the
graph shows only zero and positive values of the decision variables in the first quadrant
only.
1. Procedure of Graphic Method: To solve a L.P.P. by graphic method the following
steps are to be taken up one after another:

Step 1: Formulate the Appropriate L.P.P.: (If data is given textual and/or tabular form).

Step 2: Convert the Linear Constraints: (If they are given in the form of inequalities)
and the non-negativity conditions into equations corresponding to each equation draw
to points arbitrary.

The values of each of the decision variables under each equation may be obtained by
assuming the other variable to be zero. The two points may be obtained as the intercept
of the equation with x-axis and y-axis.

Step 3: Draw the Graph: Consider a two-dimensional co-ordinate system by taking one
decision variable along x-axis another along y-axis with a suitable scale. Draw the lines
corresponding to each equation by joining the points obtained in step 2.

Step 4: Identify the Region in which the points satisfy the linear constraints and negativity
conditions by making an arrow (→) on the corresponding line drawn on the graph.
 188

(i) Let us consider a linear constraint as

2 x + 5 y ≥ 40
The straight line corresponding to it, is

2 x + 5 y = 40
x y
or + =1
20 8

Put x = 0, then y = 8, so one point is (0, 8). Again put y = 0, then x = 20, so another
point is (20, 0).
y

12
10 ,8) The region which satisfies
(0
A 2x+5y>40
8
6
4
B (20,0)
2
0 2 4 6 8 10 12 14 16 18 20 X

Fig. 5.1

Plot these two points on the graph with a suitable scale and draw a line AB, joining
the points A (0, 8) and B (20, 0). To make an arrow indicating the region
corresponding to the inequality 2 x + 5 y > 40, put x = 0, y = 0, which gives 0 > 40,
which is not true. This means the origin 0 (0, 0) does not satisfy this inequality,
that is origin is not in the region represented by 2 x + 5 y > 40. Thus, the region
corresponding to the inequality 2 x + 5 y > 40 is opposite the origin.

The region corresponding to the inequality 2 x + 5 y ≥ 40 includes the lines

2 x + 5 y = 40 also.
(ii) Let us consider the linear constraint

4 x + 3 y ≤ 24
The equation corresponding to this inequality is

4 x + 3 y = 24
x y
or + =1
6 8
Put x = 0, then y = 8, so one point is (0, 8)
 189

Again put y = 0, then x = 6, so one point is (6, 0).

8 A (0,8)
7
6

4x
5

+3
Th

y>
4
e

24
re tisf
gio ies
sa
3
2 n
wh
ich
1 B (6,0)
0 1 2 3 4 5 6

Fig. 5.2

Draw a line AB, joining the points A (0, 8) and B (6, 0). To make an arrow indicating
the region corresponding to the inequality 4 x + 3 y < 24, put x = 0, y = 0 which
gives 0 < 24 which is true. This means origin 0 (0, 0) satisfies this inequality. That is
origin is, in the region represented by 4 x + 3 y < 24. Thus, the region corresponding
to the inequality, 4 x + 3 y > 24 is towards the origin. The region corresponding to
the inequality 4 x + 3 y ≤ 24 includes the line 4 x + 3 y = 24 also.

(iii) Corresponding to the non-negativity conditions x ≥ 0 and y ≥ 0, the straight lines

are x = 0 and y = 0 respectively, x = 0 represents y axis and y = 0 represents x-axis.
The values of x and y are only in quadrant first.

satisfies x > 0

x=0

satisfies y > 0

X
y=0

Fig. 5.3
 190

Step 5: Identify the Feasible Region: The area containing all the feasible solutions to
the L.P.P. is called a feasible region in a graph. In other words the area bounded by the
lines drawn in step 3 is called a feasible region. The points in the feasible region satisfy all
the linear constraints and the non-negatively conditions the feasible region is generally,
shaded by drawing lines or otherwise.

Step 6: Find the Optimum Solution in the Feasible Region: For this make use of the
theorem, “At least one extreme point (corner point) of the feasible region will give the optimum
(maximum or minimum) value of the objective function”.

In fact, optimum solution to an L.P.P. always at least on the two vertices of the feasible
region (one for maximization and another for minimization).

It is possible for the objective function value of an L.P.P. to be the same at two distinct
extreme points.

To determine optimum point (or corner point or extreme point) there are two methods:
(i) Locate the corner points on the graph or find them by solving the concerned linear
equations taken two at a time. Calculate the value of the objective function at each
corner point. The corner at which the objective function is optimum (minimum or
maximum), will give the optimum solution of the L.P.P.
(ii) Draw the line z = 0 (the objective function line). Continue drawing lines parallel to it
(i.e. draw the lines on the graph whose slopes are equal to the slope of the objective
function line) such lines are often called iso-profit (or iso-cost) lines.
In case of maximization, the drawing of parallel lines will stop farthest from the
origin and passes through a corner point of the feasible region. Find the co-ordinate
of the point (farthest from or nearest to the origin), put these values in the objective
function and obtain the optimum value of the objective function (if required).
Remark 1: When the feasible region is unbounded then solution to the L.P.P. is
unbounded solution.
Remark 2: When there is no common region which satisfies all the constraints
then, there is no solution to the L.P.P.
Remark 3: If there is more than one solution (i.e., there are more than one pairs of x
and y for which the objective function is optimum) then at least two must
correspond to the vertices of the feasible region.
2. Indication of Common Region/Feasible Region

Example 13: Draw the graph of the following linear inequalities and indicate the common region:
4 x + 5 y ≤ 20
2 ≤ x ≤4
and y ≥ 1. [B.C.A. (Meerut) 2001]

Solution: Draw x-axis and y-axis and take a suitable scale (suppose 1 cm = 1 unit)
 191

(i) For 4 x + 5 y ≤ 20, the straight line is

4 x + 5 y = 20
x y
+ =1
5 4
Hence, join the points (5, 0) and (0, 4) to draw this line. The region corresponding
to the inequality is towards the origin.
(ii) For x ≤ 4, the straight line is x = 4, which is parallel to y-axis at a distance of 4 units
from origin on the positive x-axis.
The region corresponding to the inequality x ≤ 4 is towards the origin.
(iii) For x ≥ 2, the straight line is x = 2, which is parallel to y-axis at a distance of 2 units
from origin on the positive x-axis.
The region corresponding to the inequality x ≥ 2 is opposite the origin.
(iv) For y ≥ 1, the straight line is y = 1, which is parallel to x-axis at distance of 1 unit
above the origin on the positive y-axis.
The region corresponding to the inequality y ≥ 1 is opposite the origin. The
required graph with shaded region as common region (which is a bound region) is
as follows:
5 x=2

4 x=4

2
y=1
1

1 2 3 4 5 4x+
5y
=
20

Fig. 5.4

Example 14: Indicate the common region which satisfied.

x1 − 2 x2 ≤ 1
x1 + 2 x2 ≥ 3
and x1 ≥ 0, x2 ≥ 0

Solution: Take x1 along x-axis and x2 along y-axis with a suitable scale on a graph paper.

Consider the line

x1 − 2 x2 = 1
1
Put x1 = 0, then x2 = − , Put x2 = 0, then x1 = 1
2
 192

 1
Plot the points 0, −  and (1, 0) and join them to draw the line on the graph.
 2

x2

x 1=0

0, 3
2

x1
0 (1,0) (3,0) x 2=0

0, –1
2
Fig. 5.5

Put x1 = 0, x2 = 0 in x1 − 2 x2 ≤ 1 ⇒ 0 − 0 ≤ 1 or 0 ≤ 1

which is true, hence origin is in the region, make the arrow accordingly.
Again consider the line x1 + 2 x2 = 3
3
Put x1 = 0, then x2 = , Put x2 = 0, then x1 = 3
2

 3
Plot the point 0,  and (3, 0) and join them to draw the line on the graph.
 2

Put x1 = 0, x2 = 0 in x1 + 2 x2 ≥ 3 ⇒ 0 + 0 ≥ 3 or 0 ≥ 3

which is not true, hence origin is not in the region. Make the arrow accordingly.

The required graph is as follows:

Here the common region (shaded area) is unbounded.

Example 15: Indicate the region satisfied by the inequalities.

−2 x1 + 2 x2 ≤ 9

3 x1 − 2 x2 ≤ − 20
and x1 ≥ 0, x2 ≥ 0

Solution: Take x1 along x-axis and x2 along y-axis with a suitable scale on a graph paper
x1 = 0 is the y-axis and x2 = 0 is the x-axis.
 193

(i) Consider the line −2 x1 + 3 x2 = 9

9
Put x1 = 0, then x2 = =3
3
9 1
Put x2 = 0, then x1 = − = −4
2 2
 1 
Plot the points (0, 3) and  − 4 , 0 and join to draw the line on the graph.
 2 

Put x1 = 0, x2 = 0 in −2 x1 + 3 x2 ≤ 9 ⇒ 0 ≤ 9 which is true, origin is in the region.

Make the arrow accordingly.

(ii) Consider the 3 x1 − 2 x2 = − 20

−20
Put x1 = 0, then x2 = = 10
−2
−20 + 2
Put x2 = 1, then x1 = = −6
3
Plots the points (0, 10) and (− 6, 1) and join them to draw the line on the graph. Put
x1 = 0, x2 = 0 in 3 x1 − 2 x2 ≤ − 20 ⇒ 0 − 0 ≤ − 20 or 0 ≤ − 20 which is not true, hence
the origin is not in the region. Make the arrow accordingly.

The required graph (with no common region) is as follows:

x2

x1
x1=0

(0,10)

(0,3)

x 2=0

(–6, 0)
x1
0
– 4 1 ,0
2

Fig. 5.6
 194

Example 16: Solve the following L.P. problem using graphical method.

Minimize z = 2 x1 + 3 x2

subject to x1 + x2 ≥ 6

7 x1 + x2 ≥ 14

and x1, x2 ≥ 0 [B.C.A. (Meerut) 2007,2010; B.B.A. (Meerut) 2004, 2006]

Solution: Scale x-axis and y-axis : 1 cm = 1

(i) Obtain two points to draw the equation x1 + x2 = 6 on the graph corresponding to
the inequality x1 + x2 ≥ 6

If x1 = 0 then x2 = 6 point (0, 6)

If x2 = 0 then x1 = 6 point (6, 0)

Line l1 represents x1 + x2 = 6 on the graph. To mark the arrow put x1 = 0, x2 = 0 in

the inequality x1 + x2 ≥ 6 which gives 0 > 6 which is not true. This means origin
(0, 0) does not satisfies the inequality x1 + x2 > 6. So make the arrow accordingly.
x2
14 A
13
12
11
10
9
8
7
6 B
5
4
3
2 l2
1 C
x1
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
l1

Fig. 5.7

(ii) The equation corresponding to the inequality

7 x1 + x2 ≥ 14 is 7 x1 + x2 = 14
on the graph corresponding to the inequality 7 x1 + x2 ≥ 14 if x1 = 0 then x2 = 14
point (0, 14)
 195

If x2 = 0 then x1 = 2 point (2, 0)

Line l2 represents 7 x1 + x2 = 14 on the graph.

To mark the arrow put x1 = 0, x2 = 0 in the inequality 7 x1 + x2 ≥ 14 ⇒ 0 ≥ 14 which

is not true. This means origin (0, 0) does not satisfies the inequality
7 x1 + x2 ≥ 14

So make the arrow accordingly.

Thus, the area bounded by these lines corresponding to the two inequalities is ABC
which is shaded. This shaded area is the feasible region. The co-ordinate of corner
point A are (0, 14). The co-ordinate of corner point B are (4/3, 14/3).

(Intersection of the lines x1 + x2 = 6, 7 x1 + x2 = 14)

The co-ordinate of corner point C are (6, 0)

The values of objective function at the corner points of the feasible region are:

Corner points Co-ordinates z = 2 x1 + 3 x2

A 0 14 2 × 0 + 3 × 14 = 42

B 4 14 4 14 8 42 50
2× +3× = + =
3 3 3 3 3 3 3

C 6 0 2 × 6 + 3 × 0 = 12

The minimum value of z is 12 at the corner point C (6, 0). The optimal solution to
the given L.P.P. is x1 = 6, x2 = 0, Min (z) = 12 .

Example 17: Solve the following L.P.P. using graphical method.

Min (z) = 3 x + 5 y
subject to constraints
− 2x + y ≤ 4
x + y ≥3
x −2 y ≤2
x, y ≥ 0 [B.C.A. (Meerut) 2006; B.B.A. (Meerut) 2005]

Solution: Scale x-axis and y-axis. 1 cm = 1 unit

(i) Obtain two points to draw the equation

− 2x + y = 4
on the graph corresponding to the inequality
−2 x + y ≤ 4
 196

If x = 0 then y = 4 point (0, 4)

If y = 0 then x = − 2 point (−2, 0)

Line l1 represents 2 x + y = 4 on the graph. To mark the arrow, put x = 0, y = 0 in the

inequality −2 x + y ≤ 4 which gives 0 ≤ 4 which is true. This means origin (0, 0)
satisfies the inequality

−2 x + y ≤ 4

So make the arrow accordingly.

(ii) To obtain two points to draw the equation

x + y =3
on the graph corresponding the inequality
x + y ≥3
If x = 0 then y = 3 point (0, 3)
If y = 0 then x = 3 point (3, 0)
Line l2 represents x + y = 3 on the graph. To mark the arrow.
(iii) To obtain two points to draw the equation
x −2 y =2
on the graph corresponding to the inequality
x −2 y ≤2

l1

A l3
B
2 C
1
-5 -4 -3 -2 -1 0 1 2 3 4 5

-1
l2
-2
-3
-4

Fig. 5.8

If x = 0 then y = − 1 point (0, − 1)

If y = 0 then x = 2 point (2, 0)

 197

Line l3 represents x − 2 y = 2 on the graph. To mark the arrow.

Thus, the area bounded by these three lines is ABC.

The co-ordinate of corner point A are (0, 4)

The co-ordinate of corner point B are (0, 3)

 8 1
The co-ordinate of corner point C are  , 
 3 3

(Intersection of the lines x + y = 3, x − 2 y = 2)

The values of objective function at the corner points of feasible region are:

Co-ordinates

Corner points x y z = 3x + 5 y

A 0 4 3 × 0 + 5 × 4 = 20

B 0 3 3 × 0 + 5 × 3 = 15

C 8 1 8 1 5 29
3× + 5 × =8+ =
3 3 3 3 3 3

29  8 1
The minimum value of z = at the corner point C  ,  . The optimal solution of the
3  3 3
8 1 29
given L.P.P. is x = , y = Min. (z) = .
3 3 3

Example 18: Solve by graphical method, the linear programming problem:

Minimize Z = 20 x1 + 10 x2

subject to the constraints, x1 + 2 x2 ≤ 40

3 x1 + x2 ≥ 30

4 x1 + 3 x2 ≥ 60

and the non-negative restrictions x1, x2 ≥ 0. [B.C.A. (Delhi) 2004,2007 ]

Solution: Step 1: Considering the constraints as equations, we get the following

equations

x1 + 2 x2 = 40, 3 x1 + x2 = 30, 4 x1 + 3 x2 = 60

Step 2: Draw lines corresponding to each equation.

 198

x2

30

25 3x1+x2=30

20 Q (4,18)

15 Line for mini.Z

x1
+2
x2
=4
10 0

5 (40,0)
2k
P
Z=20x1+10x2=0
x1
x 0
⇒ x1 = -1 -k 5 10 15 20 25 30 35 40
2 2 4x 1
+
3x 2
=60

Fig. 5.9

Step 3: The shaded region PQRSP is the permissible region of the problem.

Step 4: By Corner-Point Method: Solving simultaneously the equations of the

corresponding intersecting lines, the co-ordinates of the vertices of the convex polygon
are

P (40, 0), Q (4, 18), R (6, 12) and S (15, 0).

Now the values of the objective function Z at these vertices (corner point) are as given in
the table below:

Point ( x, y ) Value of the objective function

Z = 20 x1 + 10 x2

P (40, 0) Z = 20 × 40 + 10 × 0 = 800

Q (4 , 18) Z = 20 × 4 + 10 × 18 = 260

R (6 , 12) Z = 20 × 6 + 10 × 12 = 240 (Mini)

S (15 , 0) Z = 20 × 15 + 0 = 300

Obviously, Z is minimum at R (6, 12).

 199

Hence, the optimal solution of the given L.P.P. is

x1 = 6, x2 = 12
and minimum Z = 240.

Step 5: By Iso-Profit Method: Here we draw the line through the origin corresponding
to Z = 0, which is parallel to iso-profit line.
x (−1)
Z = 0 ⇒ 20 x1 + 10 x2 = 0 ⇒ 1 =
x2 2

The dotted line through the origin is shown in the figure. Drawing parallel lines away
from the origin (note) O, we see that the nearest line (since it is minimization problem) in
the permissible region passes through the vertex R (6, 12).
Hence, the optimal solution is
x1 = 6, x2 = 12
and minimum Z = 20 × 6 + 10 × 12 = 240

Example 19: A farm is engaged in breeding hens. In view of the need to ensure certain nutrients
(say x1, x2 , x3 ), it is necessary to buy two types of food, say A and B. One unit of food A contains 36
units of x1, 3 units of x2 and 20 units of x3 . One unit of food B contains 6 unit of
x1,12 unit of x2 and 10 unit of x3 . The minimum daily requirement of x1, x2 and x3 is 108, 36
and 100 units respectively. The cost of food A is ` 20 per unit whereas food B costs ` 40 per unit.
Find the minimum food cost so as to meet the minimum daily requirement of nutrients.
[B.C.A. (Bundelkhand) 2010]

Solution: Formulation of the Problem as L.P.P.

Let x units of food A and y units of food B be bought to fulfill the minimum requirement
of the nutrients x1, x2 , x3 and to minimize the cost.

If Z be the cost in ` of the two types of food bought, then Z = 20 x + 40 y.

According to the minimum daily requirements of x1, x2 , x3 , we have

36 x + 6 y ≥ 108, 3 x + 12 y ≥ 36, 20 x + 10 y ≥ 100

Also, since the quantity of each type of food bought cannot be negative, therefore
x ≥ 0, y ≥ 0.

Hence, the mathematical form of the given problem as a L.P.P. is as follows:

Minimize Z = 20 x + 40 y

subject to the constraints 36 x + 6 y ≥ 108, 3 x + 12 y ≥ 36, 20 x + 10 y ≥ 100

and the non-negative restrictions x ≥ 0, y ≥ 0.

Solution of the Problem: Proceeding stepwise, the permissible region (the set of all
points satisfying all the constraints and the non-negative restrictions) consists of the
shaded region yLSTPx which is unbounded.
 200

To find the minimum value of Z, we draw the dotted line through the origin
y

18

20x+10y=100 16

14

12

10
(0,8) A

6
3x+12x=36
4
(0,3) N
k 2
(3,0)
-2k
0 2 4 6 8 12 14 16 x
Z=20x+40y=0
x
⇒ y =–2 Line for mini.Z
1
36x+6y=108

Fig. 5.10

corresponding to Z = 20 x + 40 y which is parallel to iso-cost line. We now move this line

parallel to itself so that it passes through only one point [here the corner point T (4 , 2)] of
the feasible region. This line is an iso-cost line having only one point viz. T in the feasible
region which will give the minimum value of Z.

Therefore, Z is minimum for x = 4 and y = 2 and the minimum value of Z is

20 × 4 + 40 × 2 = `160.

Hence, 4 units of food A and 2 units of food B should be bought to fulfil the minimum
requirements of x1, x2 , x3 at a minimum cost of `160.

Example 20: A car manufacturing company manufactures cars of two models A and B. Model A
requires 150 man hours for assembling 50 man hours for painting and 10 man hours for checking
and testing. Model B requires 60 man hours for assembling 40 man hours for painting and 20 man
hours for checking and testing. There are available 30 thousand man hours for assembling 13
thousand man hours for painting and 5 thousand man hours for checking and testing. Express this
using linear inequalities. Draw graphs of these inequalities and then mark the feasible region.
 201

Solution: Suppose the company manufactures x1 cars of model A and x2 cars of model B,
then decision variables are x1 and x2 .

Inequality for man hours for assembling:

150 x1 + 60 x2 ≤ 30,000

Inequality for man hours for painting:

50 x1 + 40 x2 ≤ 13,000

Inequality for man hours for checking and testing:

10 x1 + 20 x2 ≤ 5,000
clearly, x1, x2 ≥ 0

To obtain the feasible region we proceed as follows:

(i) Consider the lines x1 = 0 as y axis and x2 = 0 as x-axis and a suitable scale.
(ii) Consider the line
150 x1 + 60 x2 = 30,000
x1 x
+ 2 =1
200 500

(0,500)

400

(0,325)
300
A
(0,250)
B
200

100

D
0
100 200 (260,0) 300 400 (500,0)

Fig. 5.11
 202

The intercept on the x-axis is 200 and on the y-axis is 500. Draw this line l1 and
make the arrow to indicate the region given by 150 x1 + 6 x2 ≤ 30000.
(iii) Consider the line
50 x1 + 40 x2 = 13,000
x1 x
+ 2 =1
260 325
The intercept on the x-axis is 260 and on the y-axis is 325. Draw this line l2 and
make the arrow to indicate the region given by 50 x1 + 40 x2 ≤ 13,000.
(iv) Consider the line
10 x1 + 20 x2 = 5,000
x1 x
⇒ + 2 =1
500 250
The intercept on the x-axis is 500 and on the y-axis is 250. Draw the line l3 and
make the arrow to indicate the region given by 10 x1 + 20 x2 ≤ 5000.
The required graph, marked with feasible region OABCDO (Shaded Area) is as
shown in figure 5.11.

Example 21: Draw the graph of inequality 3 x + y ≥ 9. [B.C.A. (Meerut) 2003]

Solution: Give the value of x find the value of y in the equation 3 x + y = 9.

x 0 1 2 3

y 9 6 3 0

The bounded region of this line is A, B, C, D.

12
11 Feasible Region
10
9
8 A
7
6
B
5
4
3 C
2
1 D
1 2 3 4 5 6 7 8 9 10 11 12

Fig. 5.12

The co-ordinate of A (0, 9), B (1, 6), C (2, 3), D (3, 0).
 203

Example 22: Find two number x1 and x2 such that

4 x1 + 2 x2 ≥ 80
2 x1 + 5 x2 ≤ 180
x1 ≥ 0, x2 ≥ 0
and maximize objective function
F = 3 x1 + 4 x2

Solution: Scale x-axis and y-axis : 5 cm=1

(i) Obtain two points to draw the line 4 x1 + 2 x2 = 80 on the given graph corresponding
to the inequality
4 x1 + 2 x2 ≥ 80
If x1 = 0 then x2 = 40 point (0, 40)
If x2 = 0 then x1 = 20 point (20, 0)
Line l1 represents, 4 x1 + 2 x2 = 80 on the graph. To mark the arrow accordingly.
(ii) Obtain two points to draw the line
2 x1 + 5 x2 = 180
on the given graph corresponding to the inequality
2 x1 + 5 x2 ≤ 180
180
If x1 = 0 then x2 = = 36 point (0, 36)
5
If x2 = 0 then x1 = 90 point (90, 0)
Line l2 represents, 2 x1 + 5 x2 = 180 on the graph. To mark the arrow accordingly.
The bounded region by these lines are ABC
The co-ordinate of A (90, 0)
5 
The co-ordinate of B  , 35
2 

The co-ordinate of C (20, 0)

The values of objective function at the corner points of feasible region are:

Corner points Co-ordinates

F = 3 x1 + 4 x2
x1 x2

A 90 0 F = 270
5 35 15 295
B F= + 140 =
2 2 2

C 20 0 F = 60
 204

295 5 
The maximum value of F is at the corner point B  , 35 . The optimal
2 2 
5
solution to the given L.P.P. is x1 = , x2 = 35
2

50
45
(0, 36)
40
35
B
30
25
20 Fea
s ibl
15 e Re
gio
10 n
5 A
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
C
Fig. 5.13

295
Max ( F) =
2

Example 23: Solve the following L.P.P. graphically.

Maximize (z) = 10 x + 30 y such that
x + 2 y ≤ 20, x + 5 y ≤ 35, x + 4 y ≤ 48, x, y ≥ 0
[B.B.A. (Meerut) 2002]

Solution: Step 1: To convert the inequalities in equations:

The equation corresponding to x + 2 y ≤ 20 is x + 2 y = 20.

The equation corresponding to x + 5 y ≤ 35 is x + 5 y = 35.

The equation corresponding to x + 4 y ≤ 48 is x + 4 y = 48.

The equation corresponding to x ≥ 0 is x = 0.

The equation corresponding to y ≥ 0 is y = 0.

Step 2: To draw straight lines:

Scale : x-axis : 1 cm = 5
y-axis : 1 cm = 2
(i) Equation x + 2 y = 20
If x = 0 then y = 10
If y = 0 then x = 20 .
 205

Plot the points M1 (20, 0) and N1 (0, 10) where OM1 = 20 and ON1 = 10. Draw a
straight line M1N1 joining the two points M1 and N1 which represents the
equations x + 2 y = 20 and various sets of ( x, y) . Each points below M1N1 will
satisfy the inequality x + 2 y < 20 (and including this line will satisfy the inequality
x + 2 y ≤ 20) which has been indicated by the arrow.
(ii) To represent the equation x + 5 y = 35 on the graph,
if x = 0 then y = 7, point (0, 7)
if y = 0, then x = 35, point (35, 0)
Plot the points M2 (35, 0) and N2 (0, 7), where OM2 = 35 and ON2 = 7. Draw
straight line M2 N2 joining the two points M2 and N2 which represents the
equation x + 5 y = 35 and the various sets of ( x, y). Each points below M2 N2 will
satisfy the inequality x + 5 y < 35 (and including this line will satisfy the inequality
x + 5 y ≤ 35) which has been indicated by the arrow.
(iii) To represent the equation x + 4 y = 48 on the graph,
if x = 0, then y = 12
if y = 0, then x = 48
Plot the points M3 (48, 0) and N3 (0, 12), where OM3 = 48 and ON3 = 12 .

12 N3 (0, 12)

10 N1
8
7
N2 E
6

2 M1 M2 M3 (48, 0)

0 5 10 15 20 25 30 35 40 45 50

Fig. 5.14

Draw straight line M3 N3 joining the two points M3 and N3 which represents the
equation x + 4 y = 48 and the various sets of ( x, y). Each points below M3 N3 will
satisfy the inequality x + 4 y < 48 (and including this line will satisfy the inequality
x + 4 y ≤ 48) which has been indicated by the arrow.
It is obvious that the effect of inequality x + 4 y < 48 is a redundant constraint.
(iv) The area between the line OX (on x-axis) and the line OY (on y-axis) satisfied the
non-negativity conditions ( x ≥ 0, y ≥ 0). The area OM1 EN2 bounded by the lines
corresponding to the inequalities is the region of feasible solutions, where E (10, 5)
is the intersection of the lines M1N1 and M2 N2 .
 206

Step 3: To find optimum solution. The optimal solution is any corner point of the
feasible region OM1 EN2 . The values of objective function at the corner points are as
follows:

Co-ordinates
Corner point Z = 10 x + 30 y
x y

O 0 0 Z = 10 × 0 + 30 × 0 = 0

M1 20 0 Z = 10 × 20 + 30 × 0 = 200

E 10 5 Z = 10 × 10 + 30 × 5 = 250

N2 0 7 Z = 10 × 0 + 30 × 7 = 210

It is clear from the above table that the value of Z is maximum at the corner point (10, 5).
Hence, the optimum solution of the linear programming problem is x = 10, y = 5 and the
maximum value of Z is 250.

Example 24: Solve the following L.P.P. by graphical method:

Minimize Z = 3 x + 2 y
such that x + y ≤ 5
3x + y ≥ 6
x + 4 y ≥4
0 ≤ x ≤3
and 0 ≤ y ≤3 [B.B.A. (Meerut) 2005]

Solution: Step 1: Draw two perpendicular lines ox and oy with a suitable scale (x-axis : 1
cm=1, y-axis : 1 cm=1) which represent y = 0 and x = 0. Since x ≥ 0, y ≥ 0 hence the
feasible region will be in the first quadrant.

Step 2: The equation corresponding to x + y ≤ 5 is x + y = 5.

The equation corresponding to 3 x + y ≥ 6 is 3 x + y = 6.

The equation corresponding to x + 4 y ≥ 4 is x + 4 y = 4.

The equation corresponding to x ≤ 3 is x = 3.

The equation corresponding to y ≤ 3 is y = 3.
x y
The line M1N1 represents the equation x + y = 5 or + = 1 which passes through the
5 5
points M1 (5, 0) and N1 (0, 5). The line M2 N2 represents the equation 3 x + y = 6 or
x y
+ = 1 which passes through the points M2 (2, 0) and N2 (0, 6).
2 6
 207

x y
The line M3 N3 represents the equation x + 4 y = 4 or + = 1 which passes through the
4 1
points M3 (4, 0) and N3 (0, 1).
The line M4 N4 represents the equation x = 3.
The line M5 N5 represents the equation y = 3.

Step 3: To find feasible region:

(i) Since the co-ordinates of the origin O (0, 0) satisfy the inequality x + y < 5, the
feasible region will be in that side of M1N1 in which side the origin falls. This is
indicated by the arrow on it.
(ii) Since the co-ordinates of the origin O (0, 0) do not satisfy the inequality 3 x + y > 6,
the feasible region will be in that side of M2 N2 in which side the origin does not
fall. This is indicated by the arrow on it.
(iii) Since the co-ordinates of the origin O (0, 0) do not satisfy the inequality x + 4 y > 4
the feasible region will be in that side of M3 N3 in which side the origin does not
fall. This is indicated by the arrow on it.
(iv) Since the co-ordinates of the origin O (0, 0) satisfy the inequality x < 3, the feasible
region will be in that side of M4 N4 in which side the origin falls. This is indicated
by the arrow on it.
(v) Since the co-ordinates of the origin O (0, 0) satisfy the inequality y < 3, the feasible
region will be in that side of M5 N5 in which side the origin falls. This is indicated
by the arrow on it.
Thus, the required feasible region (shaded area) is the polygon ABCDE.
y

6 N2

N4
5
N1 x=3
4

M5 E D y=3 N5
3

C
2
N3
1
A B M1
x
0 1 2 3 4 5 6
M2 M4 M3

Fig. 5.15
 208

Corner Co-ordinates Objective function

Remark
points
x y Z = 3x + 2 y

A 20 6 20 6 72 point of intersection of
Z =3 × +2× =
11 11 11 11 11 3 x + y = 6 and
x + 4 y =4

B 3 1 1 19 point of intersection of
Z =3 ×3 + 2 × =
4 4 2 x = 3 and x + 4 y = 4

C 3 2 Z = 3 × 3 + 2 × 2 = 13 point of intersection of
x = 3 and x + y = 5

D 2 3 Z = 3 × 2 + 2 × 3 = 12 point of intersection
y = 3 and x + y = 5

E 1 3 Z = 3 ×1 + 2 × 3 = 9 point of intersection of
y = 3 and 3 x + y = 6

72
It is clear from the above table that the minimum value of Z is corresponding to the
11
 20 6  20
point A  ,  . Hence, the optimal solution of linear programming problem is x = ,
 11 11 11
6
y= .
11

Example 25: Old hens can be bought at ` 2 each and young ones at ` 5 each. Old hens lay 3 eggs
per week and the young ones lay 5 eggs per week, each egg being worth 30 paise. A hen (young or old)
cost `1 per week to feed. I have only ` 80 to spend for hens. How many of each kind should I buy to
give me a maximum profit, assuming that I can not house more than 20 hens ?
[B.C.A. (Meerut) 2003,2011]

Solution: Formulation of the Problem as L.P.P.

Let, x1 = Total numbers of old hens

x2 = Total numbers of young hens.

Since old hens lay 3 eggs per week and the young hens lay 5 eggs per week,

Total number of eggs I have per week = 3 x1 + 5 x2

The cost of each egg = 30 paise = ` 0 .3

Total income per week = ` 0 .3 (3 x1 + 5 x2 )

= ` ( .9 x1 + 1 .5 x2 )
Total number of hens = x1 + x2
 209

Expenditure on feeding = ` 1 ( x1 + x2 )

= ` ( x1 + x2 )

Total profit per week = [(.9 x1 + 1.5 x2 ) − ( x1 + x2 )] = − 0 .1x1 + 0 .5 x2

Since the cost of old hen is ` 2 and the cost of one young hen is ` 5 and I have only ` 80 to
spend for hens

∴ 2 x1 + 5 x2 ≤ 80

Since I cannot house more than 20 hens

∴ x1 + x2 ≤ 20

But hens cannot be negative

∴ x1 ≥ 0, x2 ≥ 0

Hence general L.P.P. of given example is

Maximize z = − 0 .1x1 + 0 .5 x2

subject to condition 2 x1 + 5 x2 ≤ 80

x1 + x2 ≤ 20

x1 ≥ 0, x2 ≥ 0

Graphical Solution: Scale x-axis and y-axis 4 cm=1

(i) Obtain two points to draw the equation

2 x1 + 5 x2 = 80
on the graph corresponding to the inequality

2 x1 + 5 x2 ≤ 80
If x1 = 0 then x2 = 16 point (0, 16)
If x2 = 0 then x1 = 40 point (40, 0)
Line l1 represents 2 x1 + 5 x2 = 80 on the graph. To mark the arrow accordingly.
(ii) Obtain two points to draw the equation

x1 + x2 = 20
on the graph corresponding to the inequality

x1 + x2 ≤ 20

If x1 = 0 then x2 = 20 point (0, 20)

If x2 = 0 then x1 = 20 point (20, 0)

Line l2 represents x1 + x2 = 20 on the graph to mark the arrow accordingly.
210

x2

40
36
32
28
24
20
16 B
C
12
8 l1
4 A
x1
0 4 8 12 16 20 24 28 32 36 40
l2

Fig. 5.16

The common feasible region in all lines is OABC.

The co-ordinate of corner point are O (0, 0)
The co-ordinate of corner point A are (20, 0)
 20 40 
The co-ordinate of corner point B are  , 
 3 3

(Intersection of the lines 2 x1 + 5 x2 = 80, x1 + x2 = 20)

The co-ordinate of corner point C are (0, 16)
The value of objective function at the corner points of the feasible region are:

Corner points Co-ordinates Z = − 0 .1x1 + 0 . 5 x2

x1 x2

O 0 0 − 0 .1 × 0 + .5 × 0 = 0

A 20 0 − 0 .1 × 20 + . 5 × 0 = − 2

B 20 40 20 40
− 0 .1 × + .5 × =6
3 3 3 3

C 0 16 − 0 .1 × 0 + . 5 × 16 = 8

The maximum Z = 8 at x1 = 0, x2 = 16
 211

Example 26: Solve the following L.P.P. graphically

Maximize Z = 3 x1 + 5 x2

such that x1 + 2 x2 ≤ 20

x1 + x2 ≤ 15

x2 ≤ 6

and x1, x2 ≥ 0

Solution: Scale x-axis and y-axis : 1 cm = 5

(i) Obtain two points to draw the equation

x1 + 2 x2 = 20
on the graph corresponding to the inequality

x1 + 2 x2 ≤ 20
If x1 = 0 then x2 = 10 point (0, 10)
If x2 = 0 then x1 = 20 point (20, 0)

Line l1 represents x1 + 2 x2 = 20 on the graph. To make the arrow put x1 = 0, x2 = 0

in the inequality x1 + 2 x2 ≤ 20 which gives = 0 + 2 × 0 ≤ 20 i.e., 0 < 20 which is true.
This means origin (0, 0) satisfies the inequality x1 + 2 x2 ≤ 20. So make the arrow
accordingly.

(ii) The equation corresponding to the inequality

x1 + x2 ≤ 15 is x1 + x2 = 15

If x1 = 0 then x2 = 15 point (0, 15)

If x2 = 0 then x1 = 15 point (15, 0)

Line l2 represents x1 + x2 = 15 on the graph. To make the arrow put x1 = 0, x2 = 0 in

x1 + x2 ≤ 15 ⇒ 0 ≤ 15 which is true. This means origin (0, 0) satisfies the inequality
x1 + x2 ≤ 15. So make the arrow accordingly.

(iii) The equation corresponding to the inequality x2 ≤ 6 is x2 = 6.

The line l3 represents this equation on the graph. The arrow on it indicates the
region in which all points satisfy the inequality x2 ≤ 6.

(iv) x1 = 0 is y-axis and x2 = 0 is x-axis. In the first quadrant x1 ≥ 0, x2 ≥ 0.

212

20

15

l2

10

C l3
6 D
B
5

l1
A
0
5 10 15 20

Fig. 5.17

Thus, the area bounded by five lines corresponding to the five inequalities is
OA BCD which is shaded. This shaded area is the feasible region. The co-ordinates
of corner points O (0, 0), A (15, 0), B (10, 5), C (8, 6), D (0, 6).

The values of objective function at corner points of the feasible region are:

Corner points Co-ordinates Z = 3 x1 + 5 x2

x1 x2

O 0 0 Z =3 ×0 + 5 ×0 =0

A 15 0 Z = 3 × 15 + 5 × 0 = 45

B 10 5 Z = 3 × 10 + 5 × 5 = 55

C 8 6 Z = 3 × 8 + 5 × 6 = 54

D 0 6 Z = 3 × 0 + 5 × 6 = 30

The maximum value of z is 55 at the corner point B (10, 5) . Hence, Z is maximum

for x1 = 10 and x2 = 5 . This is optimal solution.
 213

Example 27: A company owns two flour mills, A and B, which have different production capacities
for high, medium and low grade flour. This company has entered a contract to supply flour to a firm
every week with at least 12, 8 and 24 quintals of high, medium and low grade respectively. It costs
the company ` 1,000 and ` 800 per day to run mills A and B respectively. On a day, mill A
produces 6, 2 and 4 quintals of high, medium and low grade flour respectively. Mill B produces 2, 2
and 12 quintals of high, medium and low grade flour respectively. How many days per week should
each mill be operated in order to meet the contract order most economically? Solve graphically to
L.P.P. [B.C.A. (Meerut) 2006]

Solution:

Formulation of the Problem as L.P.P.

Let x1 = Total product of mill A per day
x2 = Total product of mill B per day
Min (Z ) = 1000 x1 + 800 x2

subject to condition 6 x1 + 2 x2 ≥ 12
2 x1 + 2 x2 ≥ 8
4 x1 + 12 x2 ≥ 24
x1, x2 ≥ 0

Graphical Solution: Scale x-axis and y-axis, 1 cm = 1

(i) Obtain two points to draw the equation

6 x1 + 2 x2 = 12
on the graph corresponding to the inequality
6 x1 + 2 x2 ≥ 12
If x1 = 0 then x2 = 6 point (0, 6)
If x2 = 0 then x1 = 2 point (2, 0)
Line l1 represent 6 x1 + 2 x2 = 12 on the graph. To mark the arrow accordingly.
(ii) Obtain two points to draw the equation
2 x1 + 2 x2 = 8
on the graph corresponding to the inequality
2 x1 + 2 x2 ≥ 8
If x1 = 0 then x2 = 4 point (0, 4)
If x2 = 0 then x1 = 4 point (4, 0)
Line l2 represent 2 x1 + 2 x2 = 8 on the graph. To mark the arrow accordingly.
(iii) Obtain two points to draw the equation
4 x1 + 12 x2 = 24
214

on the graph corresponding to the inequality

4 x1 + 12 x2 ≥ 24
If x1 = 0 then x2 = 2 point (0, 2)
If x2 = 0 then x1 = 6 point (6, 0)
Line l3 represent 4 x1 + 12 x2 = 24 on the graph. To mark the arrow accordingly.
The common feasible region is ABCD
The co-ordinate of corner point A are (6, 0)
The co-ordinate of corner point B are (3, 1)
The co-ordinate of corner point C are (1, 3)
The co-ordinate of corner D are (0, 6).

10
9
8
7
6 D

5
4
C
3
l3
2
1 B A

1 2 3 4 5 6 7 8 9 10

l1 l2

Fig. 5.18

The value of objective function at corner points of the feasible region are:

Corner points Co-ordinates Z = 1000 x1 + 800 x2

x1 x2

A 6 0 6000 + 0 = 6000
B 3 1 1000 × 3 + 800 = 3800
C 1 3 1000 × 1 + 800 × 3 = 3400
D 0 6 1000 × 0 + 800 × 6 = 4800

The minimum value of the objective function Z = 3400 at the corner point C (1, 3)
where x1 = 1, x2 = 3.
 215

Example 28: Multiple Optimal Solutions. Solve the following problem by graphical method:
Maximize Z = x1 + 1.5 x2 + 30
subject to 2 x1 + 3 x2 ≤ 6
x1 + 4 x2 ≤ 4
and x1, x2 ≥ 0

Solution: The constant (i.e., 30) included in the objective function is deleted in the
beginning and finally adjusted in the optimum value of Z.
(i) Draw the line x1 = 0 as y-axis and mark the region represented by these lines.
(ii) x x
Draw the line 2 x1 + 3 x2 = 6 taking the points (0, 2) and (3, 0) or 1 + 2 = 1, taking
3 2
the intercept 3 and 2 on x-axis and y-axis respectively and mark the region
represented by the inequality
2 x1 + 3 x2 ≤ 6

(1, 0) C
1 . ...... .. .. . . . . . . . . . . . ( B
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .... ..
12 , 2
5 5 )
0 1 2 3 4
A (3,0)

Fig. 5.19

(iii) Draw the line x1 + 4 x2 = 4 taking the points (0, 1) and (4, 0)
x1 x2
or + =1
4 1
taking the intercept 4 and 1 on x-axis and y-axis respectively and mark the region
represented by the inequality
x1 + 4 x2 ≤ 4
(iv) Shade the feasible region OABCO.
x −1.5
Draw the line Z = x1 + 1.5 x2 = 0, i.e., x1 = − 1.5 x2 i.e., 1 = .
x2 1
 216

Draw lines parallel to this line. One of these lines is passing through the points B
and A, i.e., is one edge of the feasible region (OABCO). This meant there is no
unique points for which the value of Z is maximum. There are many points on the
line AB for which the same maximum value of Z is obtained.
So this problem has infinitely many solutions with maximum Z.
(v) To find optimum solution we proceed as follows:

Corner points Co-ordinates Z = x1 + 1. 5 x2

O (0, 0) Z = 0 + 1.5 × 0 = 0
A (3, 0) Z = 3 + 1.5 × 0 = 3

B 12 2  12 2
  Z = + 1.5 × = 3
 5, 5  5 5

C (0, 1) Z = 0 + 1.5 × 1 = 1.5

12 2 
Z is maximum of A (3, 0) and at B  ,  . That is Z max = 3. Finally,
 5 5
Z max = 3 + 30 = 33.

Example 29: A firm produces three different products viz. R , S and T through two different plants
viz., P1 and P2 , the capacities of which in number of products per day are as follows:

Product
Plant
R S T
P1 3,000 1,000 2,000
P2 1,000 1,000 6,000

The operating cost per day of running the plants P1 and P2 are ` 600 and ` 400 respectively. The
expected minimum demands during any month for the products R , S and T are 24000 units,
16000 units and 48000 units respectively.

Show by graphic method how many days should the firm run each plant during a month so that the
production cost is minimized while still meeting the market demand. [B.C.A. (Lucknow) 2008]

Solution: The formulation of L.P.P. : Let

x1 = number of days plant P1 runs
and x2 = number of days plant P2 runs
obviously, x1, x2 ≥ 0.
∴ The cost of running the plants P1 and P2 is 600 x1 + 400 x2
Thus, objective function is:
Minimize Z = 600 x1 + 400 x2

Constraints for Product R: The product of R is (3000 x1 + 1000 x2 ) units and the
minimum extracted demand is 24,000.
 217

Hence, 3000 x1 + 1000 x2 ≥ 24000 units

Constraints for Product S: The product of S is (1000 x1 + 1000 x2 ) units and the
expected minimum demand is 16000 units.
Hence, 1000 x1 + 1000 x2 ≥ 16000

Constraint for Product T : The production of T is (2000 x1 + 6000 x2 ) units and

expected minimum demand is 48000 units.
Hence, 2000 x1 + 6000 x2 ≥ 48000

Thus, the required L.P.P. is as follows:

Minimize Z = 600 x1 + 400 x2
subject to 3000 x1 + 1000 x2 ≥ 24000
1000 x1 + 1000 x2 ≥ 16000
2000 x1 + 6000 x2 ≥ 48000
and x1, x2 ≥ 0

Construction of graph
(i) The equation corresponding to 3000 x1 + 1000 x2 ≥ 24000 is:
3000 x1 + 1000 x2 = 24000
3 x1 + x2 = 24
x1 x
+ 2 =1
8 24
l1 ...... .
................
............
24 .. A.....
......... ... ..
... .
.... ........
20
... .. .......
... ....... ..
16 .......... .. 1 cm = 4 units
. . ... .
. . . .......
. .. . ..
12 B .. .. .. .. .. .
P2 . ......... .
. ..
. .... .. ...
8 . ..... . ...
. . .. .
.
........ . ........ . .
P3 ... .. .. ... ... .. . . .. ... .. . . . . . ..
4 C . . .. ... .. ..... .. .... ...... ... .. .......
. .. ..... D....... ....... ...........
.
4 8 12 16 20 24 l3
l2

Fig. 5.20

Draw this line l1 by joining the points (8, 0) and (0, 24). The region corresponding
to this in equally is opposite the origin.
 218

(ii) The equation corresponding to 1000 x1 + 1000 x2 ≥ 16,000 is:

1000 x1 + 1000 x2 = 16000  when x1 = 0, then x2 = 16

or x1 + x2 = 16 when x2 = 0, then x1 = 16
Draw this line l2 by joining the points (16, 0) and (0, 16). The region corresponding
to this inequality is opposite the origin.
(iii) The equation corresponding to the equally 2000 x1 + 6000 x2 ≥ 48000 is
2000 x1 + 6000 x2 = 48000
or x1 + 3 x2 = 24
x1 x
or + 2 =1
24 8
Draw the line l3 by joining the points (24, 0) and (0, 8). The region corresponding
to this inequality is opposite the origin.
According to the non-negativity conditions x ≥ 0, y ≥ 0, the region is in the first
quadrant.
The required graph and the feasible region (shaded area) is as given above.
Computation of Production at Corner Points of the Feasible Region
(i) Co-ordinates of A are (0, 24).
(ii) Co-ordinates of B are obtained by solving the equations l1 and l2 , i.e.,
3 x1 + x2 = 24
and x1 + x2 = 16 (given in table)
(iii) Co-ordinates of C are obtained by solving the equations l2 and l3 , i.e.,
x1 + x2 = 16
and x1 + 3 x2 = 24 (given in table)
(iv) The co-ordinates of D are (24, 0).

Co-ordinates Objective function

Corner points
x1 x2 Z = 600 x1 + 400 x2

A 0 24 600 × 0 + 400 × 24 = 9600

B 4 12 600 × 4 + 400 × 12 = 7200

C 12 4 600 × 12 + 400 × 4 = 8800

D 24 0 600 × 24 + 400 × 0 = 14400

It is clear from the above table that the value of Z, the objective function is
minimum at the corner point B (4, 12). Hence, the firm should run the plant P1 for 4
days and plant P2 for 12 days to minimize the cost while still meeting the market
demand.
 219

Example 30: The optimal solution to the following linear programming problem

Maximize (Z ) = 3 x1 + 2 x2

subject to the constraints

x1 + x2 ≤ 20

x1 ≤ 15

x1 + 3 x2 ≤ 45

− 3 x1 + 5 x2 ≤ 60

and x1, x2 ≥ 0

is Z = … and is attained at x1 = .... x2 = … you may use graphical method to fill in the blanks.
[B.C.A. (Agra) 2002, 2004, 2006, 2010]

Solution: Draw the lines

x1 + x2 = 20 points (0, 20), (20, 0)

x1 = 15 point (15, 0)

x1 + 3 x2 = 45 points (0, 15), (45, 0)

− 3 x1 + 5 x2 = 60 points (0, 12), (− 20, 0)

x2
x1=15
(0,20)
x1 + 60
3x x1+x2=20
2 =4 5x 2=
5 +
x
–3 1
(0,
15) D

)E C
12
(0,
B

x1
(– 20, 0) 0 A (2 (45
0, , 0)
(15, 0) 0)

Fig. 5.21
 220

The common feasible region OABCDEO

Corner points Co-ordinates Z = 3 x1 + 2 x2

x1 x2

O 0 0 Z =3 ×0 + 2 ×0 =0

A 15 0 Z = 3 × 15 + 2 × 0 = 45

B 15 5 Z = 3 × 15 + 2 × 5 = 55

C 15 25 15 25
Z =3 × +2× = 47.5
2 2 2 2

D 3 14 Z = 3 × 3 + 2 × 14 = 37

E 0 12 Z = 3 × 0 + 2 × 12 = 24

Hence maximize (z) = 55 at corner point B (15, 5) where x1 = 15, x2 = 5.

Example 31: The manager of a courier company wishes to hire extra helpers during the Christmas
season, because of a large increase in the volume of mail handling and delivery. Keeping in view the
limited office space and the budgetary conditions, the number of temporary helpers must not exceed
10. According to past experience, a man can handle 300 letters and 80 packages per day and a
woman can handle 400 letters and 50 packages per day. It is believed that the daily volume of extra
mail and packages will be no less than 3400 and 680 respectively. A man receives ` 25 a day and a
woman receives ` 22 a day. How many men and women helpers should be hired to keep the pay-roll
at a minimum ?

Solution: Formulation of the Problem as L.P.P.

Let x1 men and x2 women be hired by the company to keep the pay-roll at a minimum.
If Z be the daily pay-roll in ` of the extra helpers, then
Z = 25 x1 + 22 x2

Since the maximum number of helpers can be only 10, therefore x1 + x2 ≤ 10.

Given that a man can handle 300 letters daily and a woman can handle 400 letters daily
and that the number of extra letters expected daily is not less than 3400, we must have
300 x1 + 400 x2 ≥ 3400
Similarly 80 x1 + 50 x2 ≥ 680

Since the number of men and women hired cannot be negative, we have
x1 ≥ 0, x2 ≥ 0

Hence the L.P.P. formulated for the given problem is as follows:

Minimize Z = 25 x1 + 22 x2
subject to the constraints x1 + x2 ≤ 10
300 x1 + 400 x2 ≥ 3400
 221

80 x1 + 50 x2 ≥ 680
and the non-negative restrictions x1 ≥ 0, x2 ≥ 0.

Solution of the Problem: Proceeding stepwise, the permissible region (the set of all
points satisfying all the constraints and the non-negative restrictions) consists of the
point G (6, 4) only.
x2
80x1+50x2=68014 E(0,68/5)
13
x1+x2=10 12
11
300x1+400x2=3400 10 A
(0,10)
9
(0,17/2) C
8
7
6
5
G (6,4)
4
3
2 B (10,0)
1 D (34/3,0)
x1
0 1 2 3 4 5 6 7 8 9 10 11 12 13

F (17/2,0)

Fig. 5.22

Therefore the optimal solution is x1 = 6, x2 = 4 and the optimal value of

Z = 25 × 6 + 22 × 4 = 238

Hence, 6 men and 4 women helpers should be hired by the courier company to meet its
seasonal requirements and keep the pay-roll at a minimum of ` 238.

Example 32: Solve by graphical method the L.P.P.

Minimize Z = 5 x1 + 6 x2
subject to the constraints
x1 + x2 ≥ 50
x1 + 2 x2 ≤ 40
3 x1 + 4 x2 ≤ 100
and the non-negative restrictions x1 ≥ 0 and x2 ≥ 0. [B.C.A. (Lucknow) 2004, 2010]

Solution: Solving simultaneously the inequations of the constraints and then

non-negative restrictions by graphical method, we see that there exist no values of x1 and
x2 that simultaneously satisfy all the constraints and the non-negative restrictions.
 222

Hence, the given problem does not have any feasible solution.
x2

60
x1+x2=50
50
x1+x2 ≥ 50
40

30

25

20
x1+2x2=40

10
x1
0 10 20 30 40 50 60
3x1+4x2=100
Fig. 5.23

Problems Having Unbounded Solution

Example 33: Solve graphically the following L.P.P.
Maximize Z = 3 x1 + 2 x2
subject to x1 − x2 ≤ 1
x1 + x2 ≥ 3
and x1, x2 ≥ 0
x2

3
1
=
x1

x2

2
+
+
x2

x1
=
3

–2k 0 1 2 3 x1
Z=3x1+2x2=0
–1
x
⇒ x1 =–2
2 3
Fig. 5.24
 223

Solution: Proceeding stepwise, the permissible region is shaded in the figure which is
unbounded. From the figure it is clear that the dotted line through the origin
representing Z = 0 can be moved parallel to itself in the direction of Z increasing and still
have some points in the permissible region.

Thus, Z can be made arbitrarily large and so the problem has no finite maximum value of
Z. Hence, the problem has unbounded solution.

Example 34: Solve the following L.P.P. graphically .

Maximize (z) = 40 x1 + 35 x2

subject to 2 x1 + 3 x2 ≤ 60

4 x1 + 3 x2 ≤ 96

x1, x2 ≥ 0 [B.B.A. (Meerut) 2003]

Solution: Change inequality to equations

2 x1 + 3 x2 = 60 ...(1)

4 x1 + 3 x2 = 96 ...(2)

The co-ordinate of line (1) is (0, 20), (30, 0)

The co-ordinate of line (2) is (0, 32), (24, 0)

Plotting these two lines on graph with corresponding region

(0, 32)

A
(0, 20) ......... .
.. ..... .........
................... ....... . ..... ... . . B (18, 8)
.
. . .. . . . . . . . ..
..................... .................. .... ... ... .. .
.................. ................ ... ............. ...... . ....
.. . ... . .. . . . . . . . . ... ... .
.. . . ... .. . .. .
........ .. . .....

.......... ....................... .. ..... ....... .......... . . .

....................

....... . . .
. .. . ....
.... .. .

. . .. .. . . . . . . .. . . .. . .. .. .
.. .
. ..

.. ............ ... . ........ .. .. .. .. . . ....................................................

. . . . .. . .. .
................................. ................ ............. ...........................................
. . .

O C (24, 0) (30, 0)

Fig. 5.25

we get common feasible region OABC. The co-ordinate of O = (0, 0) A = (0, 20),
C = (24, 0), B = point of intersection of line (1) and (2)

i.e., B = (18, 8)
 224

Points Value of objective function

Z = 40 x1 + 35 x2

O (0, 0) Z =0

A (0, 20) Z = 40 × 0 + 35 × 20 = 700

B (18, 8) Z = 40 × 18 + 35 × 8 = 1000

C (24, 0) Z = 40 × 24 + 0 = 960

The maximum value of

Z = 1000 at point B
where x1 = 18, x2 = 8.
 225

P roblem S et
Solve Graphically the Following L.P.P.
1. Find the values of x1 and x2 which satisfy the constraint x1 + 2 x2 ≤ 2000,
x1 + x2 ≤ 15000, x2 ≤ 600, x1, x2 ≥ 0 and which maximize Z = 3 x1 + 5 x2 .
[B.C.A. (Rohilkhand) 2007]
2. Minimize Z = 1.5 x + 2 .5 y
subject to x + 3 y ≥3
x + y ≥2
and x, y ≥ 0 [B.B.A. (Meerut) 2008]

3. Maximize Z = 5 x1 + 3 x2
subject to 3 x1 + 5 x2 ≤ 15
5 x1 + 2 x2 ≤ 10
and x1, x2 ≥ 0 [B.C.A. (Kashi) 2002, 2008]

4. Minimize Z = 2 x1 + x2
subject to 5 x1 + 10 x2 ≤ 53
x1 + x2 ≥ 1
x1 ≤ 4, and x1, x2 > 0
5. Maximize Z = 7 x1 + 5 x2
subject to x1 + 2 x2 ≤ 6
4 x1 + 3 x2 ≤ 12
and x1, x2 ≥ 0 [B.C.A. (Aligarh) 2007, 2012]
6. Minimize Z = 4 x1 + 2 x2
subject to x1 + 2 x2 ≥ 2
3 x1 + x2 ≥ 3
4 x1 + 3 x2 ≥ 6
and x1, x2 ≥ 0 [B.C.A. (Delhi) 2006, 2012]
7. Maximize Z = 3 x1 + 2 x2
subject to 2 x1 − x2 ≥ 2
x1 + x2 ≤ 8
and x1, x2 ≥ 0
[B.C.A. (Kanpur) 2007]
8. Maximize Z = 3 x1 + 2 x2
subject to − 2 x1 + x2 ≤ 1
x1 ≤ 2 x1 + x2 ≤ 3
and x1, x2 ≥ 0 [B.C.A. (Rohilkhand) 2010]
 226

9. Maximize Z = 5 x1 + 7 x2
subject to x1 + x2 ≤ 4
3 x1 + 8 x2 ≤ 24,
10 x1 + 7 x2 ≤ 35
x1, x2 ≥ 0 [B.B.A. (Meerut) 2008]

10. Minimize Z = 3 x + 5 y
subject to − 3 x + 4 y ≤ 12
x ≤4
2x − y ≥ − 2
y ≥2
2 x + 3 y ≥ 12
x, y ≥ 0 [B.C.A. (Kurukshetra) 2012]

11. Maximize Z = 3 x1 + 4 x2
subject to 5 x1 + 4 x2 ≤ 200
3 x1 − 5 x2 ≤ 150
5 x1 + 4 x2 ≥ 100
3 x1 + 4 x2 ≥ 80
and x1, x2 ≥ 0 [B.C.A. (Agra) 2005, B.C.A. (I.G.N.O.U) 2012]

12. A manufacturer makes two types of products (A and B). The two products are
manufactured by using two machines (I and II). It requires two hours on each
machine to manufacture one unit of product A whereas one unit of B requires 3
hours on machine I and 1 hour on machine II, 12 and 8 machine hours are available
respectively on machine I and II. The profit contributions from product A and B are
` 6 and ` 7 respectively. Find the optimal product mix and optimal profit using
graphical method.
13. A home decorator manufactures two types of lamps A and B. Both lamps go
through two technicians (i) a cutter and (ii) a finisher. Lamp A requires 2 hours of
the cutters time and 1 hour of the finishers time. Lamp B requires 1 hour of cutters
time and 2 hours of finishers time. The cutter has 104 hours and the finisher has 76
hours available time each month. Profit on one lamp A is ` 6 and on one lamp B is
`11. Formulate a mathematical model and solve by graphic method.
14. A manufacturer produces two types of products A and B. Each unit of a requires 4
hours on the machine G and 2 hours on the machine P. Each unit of B requires 2
hours on the machine G and 4 hours on the machine P. The manufacturer has 2
machines G and 3 machines P. Each machine G works for 40 hours a week and each
machine P works for 60 hours a week. Profit on product A is ` 30,000 and on product
B is ` 40,000. The manufacturer wants to make the maximum profit in a week.
Formulate the above problem as a L.P.P. and solve it by the graphical method.
 227

15. Max (Z ) = 5 x1 + 7 x2
subject to x1 + x2 ≤ 4,
3 x1 + 8 x2 ≤ 24
10 x1 + 7 x2 ≤ 35
x1, x2 ≥ 0
16. Minimize Z = 3 x1 + 5 x2
subject to − 3 x1 + 4 x2 ≤ 12, 2 x1 − x2 ≥ − 2
2 x1 + 3 x2 ≥ 12, x1 ≤ 4, x2 ≥ 2
and x1, x2 ≥ 0
17. Maximize Z = 3 x1 + 2 x2
subject to x1 + x2 ≤ 4, x1 − x2 ≤ 2, x1, x2 ≥ 0
18. Maximize (Z ) = 3 x1 + 4 x2
subject to 5 x1 + 4 x2 ≤ 200, 3 x1 + 5 x2 ≤ 150, 5 x1 + 4 x2 ≥ 100
8 x1 + 4 x2 ≥ 80, x1, x2 ≥ 0
19. A company produces two types of leather belts say type A and B. Belt A is of superior
quality and belt B is of a lower quality. Profits on the two types of belts are 40 and 30
paise per belt, respectively. Each belt of type A requires twice as much time as
required by a belt of type B. If all belts were of type B, the company would produce
1000 belts per day. But the supply of leather is sufficient only for 800 per day. Belt A
requires a fancy buckle and 400 fancy buckles are available for this per day. For belt
of type B, only 700 buckles are available per day. How should the company
manufacture of two types of belts in order to have maximum over all profit?
[B.C.A. (Rohilkhand) 2010]

20. Give an account of linear programming problem with its important components.
What does the non-negativity restriction mean ?
[B.B.A. (Meerut) 2002, 2003]

21. Solve the following L.P.P. graphically.

Maximize Z = 40 x1 + 35 x2
subject to 2 x1 + 3 x2 ≤ 60
4 x1 + 3 x2 ≤ 96
x1, x2 ≥ 0 [B.B.A. (Meerut) 2003]

22. Write a note on the following terms used in a L.P.P.

(i) Objective function coefficient
(ii) Objective function
(iii) Set of constraints
(iv) Slack and surplus variables in L.P.P. [B.B.A. (Meerut) 2004, 2006]
 228

23. Solve the following L.P.P. by using graphical method.

Min (z) = 2 x1 + 3 x2

subject to x1 + x2 ≥ 6

7 x1 + x2 ≥ 14

x1, x2 ≥ 0
[B.C.A. (Meerut) 2005]

24. Use the graphical method to solve the following L.P.P. :

Max (z) = 3 x1 + 6 x2

subject to 3 x1 + 4 x2 ≥ 12

− 2 x1 + x2 ≤ 4

x1, x2 ≥ 0 [B.B.A. (Meerut) 2007]

25. Solve the following using graphical method.

Maximize z = 9 x + 10 y
subject to 11x + 9 y ≤ 9900
7 x + 12 y ≤ 8400
6 x + 16 y ≤ 9600
x ≥ 0, y ≥ 0 [B.B.A. (Meerut) 2008]

26. Explain the following terms in relation to L.P.P.

(i) Feasible solution (ii) Infeasible solution
(iii) Optimum solution (iv) Alternative solution
(v) Unbounded solution
[B.B.A. (Meerut) 2007]
 229

A nswers

1. x1 = 1000, x2 = 500, Z max = 5500. 2. 3 1

Z min = 3 .5, x = , y= .
2 2

3. 20 45 235 4. x1 = 0, x2 = 1, Z min. = 1.
x1 = , x2 = , Z max = .
19 19 19

5. x1 = 3, x2 = 0, Z max. = 21. 6. x1 = 0 .6, x2 = 1.2, Z min. = 4 .8

7. L.P.P. has no solution. 8. x1 = 2, x2 = 1, Z max. = 8.

9. x1 = 1.6, x2 = 2 .4, Z max. = 24 .8. 10. x = 3, y = 2, Z min. = 19.

11. 400 150 1800 12. Maximize Z = 6 x + 7 y

x1 = , x2 = , Z max. = .
13 13 13
subject to 2 x + 3 y ≤ 12

2x + y ≤ 8

x, y ≥ 0

Z max. = ` 32 at x = 3, y = 2

13. Maximize Z = 6 x + 11 y 14. Maximize Z = 30,000 x1 + 10,000 x2

subject to 2 x + y , = 104 subject to 4 x1 + 2 x2 ≤ 80

x + 2 y ≤ 76 2 x1 + 4 x2 ≤ 180

x, y ≥ 0 and x1, x2 ≥ 0.

Z max. = 440 at x = 44, y = 16

15. x1 = 16
. , x2 = 2.4, Z = 24.8 16. x1 = 3, x2 = 2, Z = 19.

17. x1 = 3, x2 = 1, Z = 11. 18. 400 150 1800

x1 = , x2 = ,Z = .
13 13 13
 230

19. 200 belts of type A, 600 belts of type B, Profit = ` 260.

21. Maxi (z) = 1000

, at x1 = 18, x2 = 8

23. Min (z) = 12 at x1 = 6, x2 = 0

24. Min (z) = 12 at x1 = 4, x2 = 0

25. Max (z) = 898 .63 at x = 626 .67, y = 334 .8

mmm
 231

C HAPTER
6

Queuing Theory

6.1 Introduction
[B.B.A. (Meerut) 2002, 2004]

Waiting lines or queues are omnipresent. Businesses of all types, industries, schools,
hospitals, book stores, libraries, bank, post offices, petrol pump, all have queuing problems.

Waiting line problems arise either because:

1. There is too much demands on the facilities so that we say that there is an excess of
waiting time or inadequate number of service facilities.
2. There is too less demand, in which case there is too much idle facility time or too
many facilities.

In either case, the problem is to either schedule arrivals or provide facilities or both so
as to obtain an optimum balance between the costs associated with waiting time and idle
time.

A group of items waiting to receive service,including.Those receiving the service is known as a waiting
line or a queue.
[B.C.A. (Lucknow) 2011]
 232

6.2 The Basic Queuing Process and its Characteristics

[B.C.A. (I.G.N.O.U.) 2012; B.B.A. (Meerut) 2002, 2003, 2004, 2005,2007,2011]
The basic queuing process can be described as a process in which the customers arrive for
service at a service counter or station, wait for their turn in the queue if the server is busy
in the service of the customer and are served when the server gets free and the customer
leave the system as soon as he is served.

6.2.1 Characteristics of Queuing System

[B.C.A. (Agra) 2004, 2006,2009; B.B.A. (Meerut) 2002, 2003, 2004]
A queuing modal or system is specified completely by six main characteristics:
1. Arrival Distribution: Its represents the pattern in which the number of customers
arrive at the system. Arrivals may also be represented by the inter-arrival time,
which is the period between two successive arrivals.
The rate at which customers arrive to be serviced, i.e. number of customers arriving
per unit of time is called arrival rate. When the arrival rate is random, the customers
arrive in no logical pattern. This represents most cases in the business world.
2. Service Distribution: It represents the pattern in which the number of customers
leave the system. Departures may also be represented by the service time which is
the time between two successive services. Service time may be constant or variable
but known or random.
3. Service Channels: The queuing system may have a single service channel. Arriving
customers may form one line and get serviced, as in a doctor’s clinic. The system
may have a number of service channels which may be arranged in parallel or in
series or complex combination of both. In case of parallel channels, several
customers may be serviced simultaneously as in a barber shop. For series channels,
a customer must pass successively through all the channels before service is
completed. A queuing model is called one service model when the system has one
server only and a multi-server model when the system has a number of parallel
channels each with one server.
4. Service Discipline: The service discipline refers to the manner in which the
members in the queue are chosen for service. The following service disciplines are
seen in common practice.
(i) First Come, First Served (FCFS): According to this discipline the
customers are served in the order of their arrival. This service discipline may
be seen at a railway ticket window etc.
(ii) Last Come, First Served (LCFS): According to this discipline the items
arriving last are taken out first. This discipline may be seen in big godowns
where the units (items) which come last are taken out first.
(iii) Service in Random Order.
(iv) Service on Some Priority-Procedure: Some customers are served before the
others without considering their order of arrival.
 233

5. Maximum Number of Customers Allowed in the System: Maximum number of

customers in the system can be either finite or infinite. In some facilities, only a limited
number of customers are allowed in the system and new arriving customers are not
allowed to join the system unless the number becomes less than the limiting value.
6. Service Mechanism: The service mechanism refers to
(i) the pattern according to which the customers are served
(ii) facilities given to the customers:
(a) Single-channel: Here the customers are served by one counter only.
(b) Multi-channel: Here the customers are served by more than one counter.

6.2.2 Customers Behavior in a Queue

Customers generally behave in the following ways:
1. Balking: A customer may not like to wait in a queue due to lack of time or
space or otherwise.
2. Reneging: A customer may leave the queue due to impatience.
3. Collusion: Some customers may collaborate and only one of them may join the queue.
4. Jockeying: If there are more than one queue then one customer may leave one
queue and joining the other. This occurs generally in the super market.

6.3 Important Definitions in Queuing Problem

Here we give the definitions of various terms used in this chapter.
1. Queue Length: Queue length is defined by the number of persons (customers)
waiting in the line at any time.
2. Average Length of Line: Average length of line (or Queue) is defined by the
number of customers in the queue per unit time.
3. Waiting Time: It is time upto which a unit has to wait in the queue before it is
taken into service.
4. Servicing Time: The time taken for servicing of a unit is called its servicing time.
5. Busy Period: Busy-period of a server is the time during which he remains busy in
servicing. Thus, it is the time between the start of service of the first unit to the end
of service of the last unit in the queue.
6. Idle Period: When all the units in the queue are served. The idle period of the
server begins and it continues upto the time of arrival of the unit (customer). The
ideal period of a server is the time during which he remains free because there is no
customers present in the system.
7. Mean Arrival Rate: The mean arrival rate in a waiting-line situation is defined as
the expected number of arrivals occurring in a time interval of length unity.
 234

8. Mean Servicing Rate: The mean servicing rate for a particular servicing station is
defined as the expected number of services completed in a time interval of length
unity, given that the servicing is going on throughout the entire time unit.
9. Transient State: A system is said to be transient state when its operating characteristics
are dependent on time. Thus, a queuing system is in transient state when the probability
distributions of arrivals, waiting time and servicing time of the customers are dependent
on time. This state occurs at the beginning of the operation of the system.
10. Steady State: A system is said to be in steady state when its operating
characteristics becomes independent of time.
11. Traffic Intensity: [B.B.A. (Meerut) 2002]

In case of a simple queue the traffic intensity is the ratio of mean arrival rate and
the mean servicing rate.
Mean arrival rate
i.e. Traffic intensity =
Mean servicing rate

6.4 Poisson Process

Theorem 1: In Poisson process the probability of n arrivals during time intervals
of length t is given by.
( λ t) n e − λ t
Pn (t) = ...(1)
n!
where λ, t is the parameter.
[B.C.A. (Kanpur) 2009; B.C.A. (Avadh) 2008; B.C.A. (Lucknow) 2010]

Proof : Case I: When n = 0 then from (1), the probability of no arrival in time ∆ t is given by

( λ ∆ t) ° e − λ ∆ t
P0 (∆ t) = = e− λ ∆ t
0!
λ2
=1 − λ ∆ t + (∆ t)2 + … = 1 − λ ∆ t + O (∆ t)
2!
where O (∆ t) = smaller order of magnitude then ∆ t
If ∆ t is very small then O (∆ t) → 0
∴ P0 (∆ t) = 1 − λ ∆ t
i.e., the probability of no arrival in time ∆ t = 1 − λ ∆ t

Case II: When n = 1, then from (1)

(λ ∆ t)1 e − λ ∆ t
P1 (∆ t) =
1!
 λ2 
= λ ∆ t 1 − λ ∆ t + (∆ t)2 + ...
 2! 
 235

= λ ∆ t + O ( ∆ t)

∴ P1 (∆ t) = λ ∆ t

i.e., the probability of one arrival in time ∆ t = λ ∆ t

Case III: When n = m > 1, then from (1), we have

( λ ∆ t) m e − λ ∆ t
Pm (∆ t) =
m!

λ m (∆ t)m  λ 2 (∆ t)2 
= 1 − λ ∆ t + + …
m!  2! 
λm
= {(∆ t)m − λ (∆ t)m +1 + … }
m!

=0

Thus, the probability of more than one arrival in time ∆ t = 0

6.5 Some Distributions

Here, we give some distributions which we are used in queuing theory.
1. Exponential Distribution: If T is a random variable representing the inter-arrival
time (the time between consecutive arrivals), then T obey on exponential distribution
i.e. if f (T ) denotes the probability function of T, then

f (T ) = λ e − T

where λ is parameter.

2. Regular Distribution: In regular distribution the inter-arrival time is taken as

constant and defined as
0 for T < a
f (T ) = 
1 for T ≥ a
a is constant.

3. Erlang Service Time Distribution with k Phases ( Ek ): The probability density

function for the Erlang distribution is given by

(µ k)k k −1 − k µ t
f (t, µ, k) = t e
(k − 1) !

t ≥ 0, k ≥ 1
where µ and k are positive parameters of the distribution.
 236

6.6 Kendall's Notation for Representing Queuing Models

D.G. Kendall and later A. Lee introduced useful notation for queuing models. The
complete notation can be expressed as

(a / b / c) : (d / e / f )

where a = arrival or inter arrival distribution

b = departure or service time distribution

c = number of parallel service channels in the system

d = service discipline

e = maximum number of customers allowed in the system

f = calling source or population.

The following conventional codes are generally used to replace the symbol a, b and d

Symbols for a and b

M = Markovian (Poisson) arrival or departure distribution (or exponential inter
arrival or service time distribution).

Ek = Erlangian or gamma inter arrival or service time distribution with

parameter k.

G I = general independent arrival distribution.

G = general departure distribution.

D = deterministic inter arrival or service times.

Symbol for d

E C F S = first come, first served

L C F S = last come, first served

S I R O = service in random order

G D = general service discipline

The symbols e and f represent a finite (N) or infinite (∞) number of customers in the
system and calling source respectively.

( M / Ek / 1) = Poisson arrival, Erlangian departure, single server.

( F C F S / N / ∞) = first come, first served, discipline, maximum allowable customer N in

the system and infinite population model.
 237

6.7 A List of Symbols

The following notations will be used throughout the chapter :

T = The inter arrival time between two successive customers.

n = Number of customers in the queuing system.

Pn(t) = Transient state-probability that there are exactly n units in the system at
any time t.
 
Pn = Steady state probability of having n units in the system  lim Pn(t) = Pn .
 t→ ∞ 

λ n = Mean arrival rate of customers when there are n units presents in the
system.

µ n = Mean service rate when there are n units in the system.

λ = Mean arrival rate of customer (independent of n).

µ = Mean service rate (independent of n).

s or c = Number of parallel service stations.

λ
ρ= = Traffic intensity.
µ

O (∆ t) =A quantity which is of smaller order of magnitude than ∆ t.

FCFS = First come first served (service discipline).

ψ (ω) = Probability density function of waiting time in the system.

E( Ls ) or Ls = Expected number of customers in the system (waiting and in service) i.e.,

expected or mean line length.

E( Lq ) or Lq = Expected number of customers in the queue i.e., expected or mean queue

length (excluding the number of units in service).

E(Ws ) or Ws = Expected waiting time per customer in the system (including service
time).

E(Wq ) or Wq = Expected waiting time per customer in the queue. (excluding service
time).

E( L / L > 0) = Expected length of the non-empty queue.

E(W / W > 0) = Expected waiting time of a customer who has to wait.

 m m m!
  = Cn =
n n !(m − n ) !

The value of ρ = λ / µ must always be less than 1, so that the steady state condition may
be obtained. If ρ > 1 i.e., λ > µ, then extremely large queues will be obtained.
 238

The Queuing Problems:

6.8 Model 1 ( M / M /1) : ( ∞ / FCFS) (Birth and Death Model)

[B.B.A. (Meerut) 2002, 2004, 2005, 2006, 2008, 2009, 2012]
λ
The basic characteristics of this queuing problem (valid only when < 1) are:
µ

1. Probability of no customer in the system is

P0 = 1 − ρ where ρ = λ /µ

2. Probability of n customers in the system is

λ
Pn = (1 − ρ) ρ n where ρ= and n ≥ 0
µ

3. Probability that there are more than n customers in the system

P (> n) = ρ(n +1)

4. Probability that are more than n customers in the queue

P (> n + 1) = ρ(n + 2)

5. The average (expected) number of customers in the system is

λ ρ
L or Ls or E (n) = =
µ − λ 1−ρ

6. Average (expected) queue length is

λ λ2 ρ2
Lq or E (m) = E (n) − = =
µ µ (µ − λ ) 1 − ρ

λ ρ2
or Lq = Ls − =
µ 1−ρ

7. Average (expected) waiting time of customer in queue is

1 λ
Wq or W or E (W ) = E (m) =
λ µ (µ − λ )

8. Average (expected) waiting time that a customer spends in the system is

1 1 1
Ws or E (v) = E (n) or ( Lq ) =
λ λ µ−λ

9. Average (expected) waiting time of an arrival who has to wait is

1
E (W / W > 0) =
µ−λ

10. Average (expected) length of a non-empty queue is

µ 1
( L / L > 0) or E (m / m > 0) = =
µ − λ 1−ρ
 239

11. The probability of arrivals during the service time of any given customer.
r
 λ  µ
Kr =  
 λ + µ λ+µ

where r the probability of r arrivals.

12. The variance of queue length.

ρ λ
Var (n) = where ρ= <1
(1 − ρ)2 µ

N
 λ
13. Probability [queue size ≥ N] =   .
µ

6.8.1 Inter-relationship between L s , L q , Ws , Wq

Ls = λ Ws ...(1)

Lq = λ Wq ...(2)
1
Wq = Ws − ...(3)
µ

λ
Lq = Ls − ...(4)
µ

S olved E xamples
Example 1: A TV repairman finds that time spend on his jobs has an exponential distribution
with mean 30 minutes. If he repairs sets in the order in which they come in, and if the arrival of sets is
approximately Poisson with an average rate 10 per 8-hour day, what is repairman’s expected idle
time each day? How many jobs are ahead of the average set just brought in?
[B.C.A. (Kashi) 2008, 2011; B.C.A. (Delhi) 2006, 2010; B.C.A. (Rohilkhand) 2007, 2009]

1 10 1
Solution: Here, µ = ,λ = =
30 8 × 60 48

Therefore, expected number of jobs are

λ /µ λ 1 / 48 5
Ls = = = =
1 − λ / µ µ − λ 1 / 30 − 1 / 48 3

Since the fraction of the time the repairman is busy is equal to λ / µ, the number of hours
for which the repairman remains busy in a 8 hour day is

 λ 30
=8  =8× = 5 hours.
µ 48
 240

Example 2: At what average rate must a clerk at a supermarket work in order to ensure a
probability of 0 .90 that the customer will not have to wait longer than 12 minutes? It is assume that
there is only one counter at which costumers arrive in a Poisson fashion at an average rate of 15 per
hour. The length of service by the clerk has an exponential distribution.
[B.C.A. (Agra) 2003; B.B.A. (Lucknow) 2007; B.C.A.(Garhwal) 2008, 2012; B.C.A. (I.G.N.O.U.) 2011]

15 1
Solution: Here λ = = customer/minute, µ = ?
60 4
Prob [Waiting time ≥ 12] = 1 − 0 .90 = 0 .10
∞  λ
Therefore,
∫12 λ 1 − µ  e−(µ − λ) w dw = 0 .10
∞
 λ   e − (µ − λ ) w 
or λ 1 −    = 0 .10
 µ   − (µ − λ ) 
12

or e(3 − 12 µ) = 0 .4 µ
1
or = 2 .48 minute per service.
µ

Example 3: In super market the average arrival rate of customers is 5 every 30 minutes. The
average time it takes to list and calculate the customers purchases at the cash desk is 4 .5 minutes,
and this time is exponentially distributed.
(i) How long will the customer expect to wait for service at the cash desk?
(ii) What is the chance that the queue length will exceeds 5 ?
(iii) What is the probability that the cashier is working? [B.C.A. (I.G.N.O.U.) 2004, 2012]

Solution: Here
1
λ = 5 every 30 minutes or per minute.
6
2
µ= per minute
9
λ 1 9 3
∴ ρ= = × =
µ 6 2 4

1 1
λ 6 6 9 × 9 × 6 27
(i) Wq = = = = = = 13 .5
µ (µ − λ ) 2  2 1  2
×
3 2 ×3 ×6 2
 − 
9  9 6 9 9 × 6

(ii) P [> n + 1] = ρ5 + 2 = (0 .75)7 or .133, since n = 5

(iii) Probability that cashier is working

= Probability of one or more customers in the system
= 1 − probability of no customer in the system
= 1 − P0 = ρ = 0 .75, Since P0 = 1 − ρ
 241

Example 4: On an average 96 patients per 24 -hour day require the service of an emergency clinic.
Also on average, a patient requires 10 minutes of active attention. Assume that facility can handle
only one emergency at a time. Suppose that it costs the clinic ` 100 per patient treated to obtain an
average servicing time of 10 minutes, and that each minute of decrease in this average time would cost
` 10 per patient treated. How much would to be budgeted by the clinic to decrease the average size of
1 1
the queue from 1 patients to patient ?
3 2 [B.C.A. (Lucknow) 2006]
96 1
Solution: Here λ= = patient/minute
24 × 60 15
1
µ= patient/minute
10
∴ ρ = λ /µ = 2 /3

Expected number of patients in the waiting line

2
 1
 
λ2 15  4
Lq = = =
µ (µ − λ)  1   1 1 3
   − 
10  10 15 

Fraction of the time for which there are no patients is given by

2 1
P0 = 1 − ρ = 1 − =
3 3
4 1
Now, when the average queue size is decreased from patient to patient, we are to
3 2
determine the value of µ . So, we have
2
 1
2  
λ 1 15 
Lq = ⇒ =
µ (µ − λ ) 2  1
µ µ − 
 15 
2
⇒ µ= patients per minutes
15
1 15
∴ Average rate of treatment required = = = 7 .5 minutes.
µ 2
Decrease in the average rate of treatment = (10 − 7 .5) = 2 .5 minutes

Budget per patient = ` (100 + 2 .5 × 10) = ` 125

Hence, in order to get the required size of the queue, the budget should be increase from
` 100 per patient to ` 125 per patient.

Example 5: In a railway marshalling yard, goods trains arrive at a rate of 30 trains per day.
Assuming that the inter-arrival time follows an exponential distribution and the service time
distribution is also exponential with an average 36 minutes. Calculate the following:
(i) The mean queue size (line length)
(ii) The probability that the queue size exceeds 10 .
 242

If the input of trains increase to an average 33 per day, what will be the change in (i) and (ii)?

Solution: Here
30 1
λ= =
60 × 24 48
1
µ= trains per minutes
36
1
λ 48 36 3
∴ ρ= = = = = 0 .75
µ 1 48 4
36

ρ 0 .75
(i) Lq = = = 3 trains
1 − ρ 1 − 0 .75

(ii) P [≥ 10] = ρ10 = (0 .75)10 = 0 .06

when the input increases to 33 trains per day, we have

33 11 1
λ= = and µ = trains per minutes
60 × 24 480 36
λ 11
∴ ρ= = × 36 = 0 .83, hence
µ 480

0 .83
(i) Ls = = 5 trains
0 .16

(ii) Prob (queue size ≥ 10) = (0 .83)10 = 0 .2 (approximation)

Example 6: A self-service store employs one cashier at its counter. Nine customers arrive on an
average every 5 minutes while the cashier can serve 10 customers in 5 minutes. Assuming Poisson
distribution for arrival rate and exponential distribution for service rate, find
(i) Average number of customers in the system.
(ii) Average number of customers in queue or average queue length.
(iii) Average time a customer spends in the system.
(iv) Average time a customer waits before being served.

Solution: Arrival rate λ = 9 / 5 = 1.8 customers/minute

Service rate µ = 10 / 5 = 2 customers/minute

(i) Average number of customers in the system
λ 1.8
Ls = = =9
µ − λ 2 − 1.8
(ii) Average number of customers in the queue
λ2 λ λ 1.8 1.8
Lq = = . = × = 8 .1
µ (µ − λ) µ µ − λ 2 2 − 1.8
 243

(iii) Average time a customer spends in the system

1 1
Ws = = = 5 minutes.
µ − λ 2 − 1.8

(iv) Average time a customer spends in the queue

λ  1  1.8  1 
Wq =   =   = 4 .5 minutes.
µ µ − λ 2  2 − 1.8

Example 7: A person repairing radios finds that the time spend on the radio sets has been
exponential distribution with mean 20 minutes. If the radios are repaired in the order in which they
come in and their arrival is approximately Poisson with an average rate of 15 for 8-hour day, what is
the repairman’s expected idle time each day? How many jobs are ahead of the average set just brought
in? [B.B.A. (Meerut) 2004]

15 1
Solution: Arrival rate λ = = units/minute service rate
8 × 60 32
1
µ= units/minute
20
Number of jobs ahead to the set brought in = Average number of jobs in the system
1
λ 32
Ls = = = 5 /3
µ−λ 1
–
1
20 32
Number of hours for which the repairman remains busy in an 8-hours day
λ 8 × 1 / 32 20
=8 = =8× = 5 hours.
µ 1 / 20 32
∴ Time for which repairman remains idle in an 8-hours day = 8 − 5 = 3 hours.

Example 8: A branch of Punjab National Bank has only one typist. Since the typing work varies
in length (Number of pages to be typed), the typing rate is randomly distributed approximating a
Poisson distribution with mean service rate of 8 letters per hour. The letters arrive at a rate of 5 per
hour during the entire 8 -hour work day. If the typewriter is valued at ` 1.50 per hour, determine
(i) Equipment utilization.
(ii) The percent time that an arriving letter has to wait.
(iii) Average system time.
(iv) Average cost due to waiting on the part of typewriter. [B.B.A. (Rohilkhand) 2005, 2007]

Solution: Arrival rate, λ = 5 per hour

Service rate, µ = 8 per hour

λ 5
(i) Equipment utilization, ρ = = = 0 .625
µ 8

(ii) The percent time an arriving letter has to wait = percent time type writer remains
busy = 62 .5%
 244

1 1 1
(iii) Average system time, Ws = = = hr = 20 minutes
µ − λ 8−5 3

(iv) Average cost due to waiting on the part of the type writer per day
 5
= 8 × 1 −  × ` 1.50 = ` 4 .50
 8

Example 9: Arrival rate of telephone calls at a telephone booth are according to Poisson
distribution, with an average time of 9 minutes between two consecutive arrivals. The length of
telephone call is assumed to be exponentially distributed, with mean 3 minutes.
(i) Determine the probability that a person arriving at the booth will have to wait.
(ii) Find the average queue length that is formed from time to time.
[B.B.A. (Delhi) 2005, 2008]

(iii) The telephone company will install a second booth when convinced that an arrival would
expect to have to wait at least four minutes for a phone. Find the increase in the flow of arrivals
which will justify a second booth.
(iv) What is the probability that an arrival will have to wait for more than 10 minutes before the
phone is free? [B.C.A. (I.G.N.O.U.) 2010]
(v) Find the fraction of a day that the phone will be in use.
1
Solution: Here, arrival rate λ = per minute
9
1
Service rate µ = per minute
3
λ 1/9 1
(i) Probability that a person will have to wait = = = = 0 .33
µ 1/3 3

µ 1/3
(ii) Average queue length = = = 1.5 persons
µ − λ 1 /3 −1 /9

λ1
(iii) Average waiting time in the queue =
µ (µ − λ1 )
λ1
4=
1 1 
 − λ1
3 3 
1 λ1 λ1 7 1
or − = or λ1 × =
9 3 4 12 9
4
or λ= arrivals/minute
21
4 1 5
∴ increase in the flow of arrivals = − = per minute
21 9 63

∞  λ
(iv) Probability [waiting time ≥ 10] =
∫10 1 − µ  λ e(λ − µ) t dt
 245

∞
 λ   e(λ −µ) t 
= λ 1 −   
 µ   λ − µ 
10

λ (µ − λ)  1 (λ −µ)10 
= 0 − λ − µ e 
µ  
λ (λ −µ)10
= e
µ
 1 1
1 / 9  9 − 3  10 1 − 20 /9 1
= e = e =
1/3 3 30

(v) The expected fraction of a day that the phone will be in use = λ / µ = 0 .33

Example 10: Customers arrive at a one window drive in bank according as a Poisson distribution
with mean 10 per hour. Service time per customer is exponential with mean 5 minutes. The space in
front of the window, including that for the serviced car can accommodate a maximum of three cars.
Others cars can wait outside this space.
(i) What is the probability that an arriving customer can derive directly to the space in front of
the window?
(ii) What is the probability that an arriving customer will have to wait outside the indicated
space?
(iii) How long is an arriving customer expected to wait before starting service?
[B.C.A. (Rohtak) 2006, 2009, 2011]

60
Solution: Here, λ = 10 per hours, µ = = 12 per hour
5
(i) The probability that an arriving customer can derive directly to the space in front of
the window.
= p0 + p1 + p2
2
 λ λ  λ  λ  λ
= 1 −  + 1 −  +   1 − 
 µ µ  µ µ  µ

 λ  λ λ2 
= 1 −  1 + + 2 
 µ   µ µ 

 10   10 100 
= 1 −  1 + 12 + 144  = 0 .42
 12 

(ii) The probability that an arriving customer has to wait outside the indicated space
= 1 − 0 .42 = 0 .58 .
(iii) Average waiting time of a customer in the queue
1 λ 10
= = = 0 .417 hours.
µ µ − λ 12 (12 − 10)
 246

Example 11: A repairman is to be hired to repair machine which breakdown at an average rate of
6 per hour. The breakdown follow Poisson distribution. The non-productive time of a machine is
considered to cost ` 20 per hour. Two repairman, Mr. X and Mr. Y have been interviewed for this
purpose. Mr. X charge ` 10 per hour and he services breakdown machines at the rate of 8 per hour.
Mr. Y demands ` 14 per hour and he services at an average rate of 12 per hour. Which repairman
should be hired? (Assume 8 hours shift per day).

Solution: We have, average down time (non-productive time) of a machine per hour
= Average waiting time of a breakdown machine in the system per hour
= Ws = 1 / (µ − λ )

Here mean arrival rate (breakdown rate), λ = 6 machines per hour.

For repairman Mr. X : Mean service rate, µ = 8 machines per hour.

1 1 1
∴ Ws = = = hour.
µ − λ 8−6 2

So the total cost per hour = Down time cost of 6 breakdown machines per hour
+ Charges paid to Mr. X per hour
= ` [20 × 6 × (1 / 2) + 10] = ` 70 .
∴ the total cost in 8 hours shift per day.
= ` 70 × 8 = ` 560.

For repairman Mr. Y : Mean service rate.

µ = 12 machines per hour.
1 1 1
∴ Ws = = = hour.
µ − λ 12 − 6 6

So, the total cost per hour

= down time cost of 6 breakdown machine per hour
+ charges paid to Mr. Y per hour
= ` [20 × 6 × (1 / 6) + 14] = ` 34

∴ the total cost in 8 hours shift per day = ` 34 × 8 = ` 272.

Since, per day cost of Mr. Y is less than that of Mr. X. Hence, the repairman Mr. Y should
be hired.
Example 12: A warehouse has only one loading dock manned by a three persons crew. Trucks
arrive at the loading dock at an average rate of 4 trucks per hour and the arrival rate is Poisson
distributed. The loading of a truck takes 10 minutes on an average and can be assumed to be
exponentially distributed. The operating cost of a truck is ` 20 per hour and the members of the
loading crew are paid ` 6 each per hour. Would you advise the truck owner to add another crew of
three persons ? [B.C.A. (Delhi) 2012]
 247

Solution: It is ( M / M / 1) : (∞ / FCFS) problem

Here, mean arrival rate,

λ = 4 trucks per hour.

For present crew of 3 persons.

Mean service rate, µ = 60 / 10 = 6 trucks per hour.
∵ Ws = Average waiting time per truck, per hour
1 1 1
= = = hour.
µ − λ 6 −4 2

∴ Total cost per hour

= Loading crew cost + waiting time cost
= ` 6 × 3 + Ws × number of trucks arrive in one hour
× waiting time cost per hour
= ` 18 + ` {(1/2) × 4 × 20} = ` 58.

After the addition of another crew of 3 persons.

Mean service rate µ = 2 × 6 = 12 trucks per hour.
1 1 1
∴ Ws = = = hour
µ − λ 12 − 4 8

∴ Total cost per hour, (after the addition of another crew)

= Loading crew cost + waiting time cost
= ` 6 × 6 + ` {(1/8) × 4 × 20} = ` 46.

Since total cost per hour is the second case is less than that in the previous case, hence we
advice the truck owner to add another crew of three persons.

Example 13: Problems arrive at a computing centre in a Poisson fashion at an average rate of five
per day. The rules of the computing centre are that man waiting to get his problem solved must aid the
man whose problem is being solved. If the time to solve a problem with one man has an exponential
distribution with mean time of 1/3 day, and if the average solving time is inversely proportional to
the number of peoples working on the problem, approximate the expected time in the centre of a person
entering the line. [B.C.A. (I.G.N.O.U.) 2010]

Solution: This problem is base on ( M / M / 1) : (∞ / FCFS) model.

Here, λ = 5 problems/day
and µ (mean service rate with one unsolved problem)

= 3 problems/day.

e − λ /µ  λ 
n
and Pn = .  .
n! µ
 248

∴ Expected number of persons working on the problem at any instant

∞
Ls = ∑ n. Pn
n =0

∞
e − λ /µ  λ 
n
= ∑ n.
n!
. 
µ
n =0

∞
 λ (λ / µ)n −1 
 
= e − λ /µ .  
µ ∑
n =1
 
 (n − 1) ! 

 λ λ 5
= e − λ /µ .   . e λ /µ = = persons
µ µ 3

Average solving time (which is inversely proportional to the number of peoples working
on the problem)

= 1/5 days/problems.

∴ Expected time in the centre for a person entering the line

1 1 5 1
= Ls = × = days = 8 hours.
5 5 3 3

Example 14: Customers arrive at a box window, being manned by single individual, according to
Poisson input process with a mean rate of 30 per hour. The time required to serve a customer has an
exponential distribution with a mean of 90 seconds. Find the average waiting time of a customer.
Also determine the average number of customers in the system and average queue length.
[B.C.A. (Kurukshetra) 2007, 2011]

Solution: It is ( M / M / 1) : (∞ FCFS) problem queuing model

30
Here, λ= = 0 .5
60
60
and µ= = 0 .67 per minute
90

∴ Average waiting time of a customer is given by

λ
Wq = E (w) =
µ (µ − λ )

0 .5
= = 4 .5 minute per customer
0 .67 (0 .67 − 0 .5)

Average number of customers in the system is given by

λ 0 .50 0 .50
= = = =3
µ − λ 0 .67 − 0 .50 0 .17
 249

Average queue length is given by

λ2 (0 .5)2
E (m) = Lq = = = 2 .27
µ (µ − λ) (0 .67) (0 .67 − 0 .5)

Example 15: A departmental store has a single cashier. During the rush hours, customers arrive at
a rate of 20 customers per hour. The average number of customers that can be processed by the cashier
is 24 per hour. Assume that the conditions for use of the single-channel queuing model apply.

(i) What is the probability that the cashier is idle?

(ii) What is the average number of customers in the queuing system?

(iii) What is the average time a customer spends in the system?
(iv) What is the average number of customers in the queue?
(v) What is the average time a customer spends in the queue waiting for service?
[B.C.A. (Kanpur) 2007]

Solution: We are given λ = 20 and µ = 24 customers per hour

(i) Probability that the cashier is idle is

λ 20 1
P0 = 1 − =1 − = = 0 .67
µ 24 6

(ii) Average number of customers in the system is

λ 20 1
E (n) = Lq = L = = = = 0 .25
µ − λ 24 − 20 4

(iii) Average time a customer spends in the system

1 1 1
E (v) = Wq = = = = 0 .25
µ−λ 24 − 20 4

(iv) Average number of customers in the queue is

λ2 20 × 20
E (m) = Lq = = = 4 .17
µ (µ − λ) 24 (24 − 20)

(v) Average time a customer spends in the queue is

λ 5
E (W ) = Wq = W = E (W ) = = = 0 .21
µ (µ − λ) 24
 250

6.9 Model II ( M / M /1) :(N/FCFS) Finite Queue Length Model

[B.B.A. (Meerut) 2008]
In this queuing system, the number of customers is limited to N

1. The probability of no customer in the system is

1−ρ
P0 =
1 − ρ N +1
λ
where ρ=
µ

λ = mean arrival rate or number of arrivals per unit of time.

µ = mean service rate per busy server or number of
customers served. per unit of time

2. The probability of n customers in the system

1−ρ
Pn = . ρn
1 − ρ N +1

3. Average number of customers in the system

ρ [1 − (1 + N ) ρ N + N ρ N +1
Ls =
(1 − ρ) (1 − ρ N +1)
λ
where ρ=
µ

4. Average number of customers in the queue.

1 − N ρ N −1 + (N − 1) ρ N 2
N
Lq =
(1 − ρ) (1 − ρ N +1)
ρ = ∑
n Pn
n =0

5. Average time a customer spends in the system

L
Ws = s where λ′ = λ (1 − ρ N )
λ′

6. Average waiting time in the queue

Lq
Wq = , where λ′ = λ (1 − ρ N )
λ′

Example 16: Trains arrive at the yard every 15 minutes and the service time is 33 minutes. If the
line capacity of the yard is limited to 4 trains, find

(i) The probability that the yard is empty.

(ii) The average number of trains in the system. [B.C.A. (Agra) 2009]
 251

Solution: Here
1
λ= per minute
15
1
µ= per minute
33
λ 33
∴ ρ= = = 2 .2
µ 15

N =4

1−ρ ρ −1 2 .2 − 1 1.1
(i) P0 = = N +1 = =
1 − ρ N +1 ρ − 1 (2 .2)5 − 1 51.15 − 1

1.1
= = 0 .021934 = 0 .237 (Appro.)
50 .15

(ii) Average number of trains in the system

N
Ls = ∑ n Pn = 0 + P1 + 2 P2 + 3 P3 + 4 P4 , N = 4
n =0

= P0 (ρ + 2 ρ2 + 3 ρ3 + 4 ρ4 )

= 0 .237 (2 .2 + 2 × (2 .2)2 + 3 (2 .2)3 + 4 (2 .2)4

= 0 .237 (2 .2 + 9 .68 + 31.94 + 93 .70)

= 32 . 6

Example 17: Consider a single server queuing system with a Poisson input, exponential service
times. Suppose the mean arrival rate is 3 calling units per hour the expected service time is
0 .25 hours, and the maximum number permissible number calling units in the system is two. Derive
the steady state probability distribution of the number of calling units in the system, and then
calculate the expected number in the system. [B.C.A. (Garhwal) 2008]

Solution: Here

λ = 3 units per hour

1
and = . 25 or µ = 4 units per hours
µ

λ 3
∴ ρ= = = 0 .75 which implies that
µ 4

1 − ρ = 1 − 0 .75 = 0 .25

Also, we are given N = 2

 252

The steady-state probability distribution of the number of calling units in the system is
given by
(1 − ρ) ρ n (0 .25) (0 .75)n
Pn = N +1
= = 0 .43 (0 .75)n
1−ρ 1 − (0 .75)3
1−ρ (0 .25)
and P0 = 2 +1
= = 0 .43
1−ρ 1 − (0 .75)3

Expected number in the system is given by

N
E (n) or E ( Lq ) or Lq = ∑ n Pn
n=0
2 2
= ∑= n (0 .43) (0 .75)n = (0 .43) ∑= n(0 .75)2
n 0 n 0

= (0 .43) [1 × (0 .75) + 2 × (0 .75)2 ] = 0 .81

Example 18: Assume that the goods trains are coming in a yard at the rate of 30 trains per day
and suppose that the inter-arrival times follow the exponential distribution. The service time for each
train is assumed to be exponential with an average of 36 minutes. If the yard can admit 9 trains at a
time (there being 10 lines, one of which is reserved for shunting purpose) calculate the probability that
the yard is empty and find the average queue length. [B.C.A. (I.G.N.O.U.) 2009]

Solution: Here
30 1
λ= = ,N =9
60 × 24 48
1
and µ= trains per minute
16
λ 36
∴ ρ= = = 0 .75
µ 48

The probability that the yard is empty is given by

1−ρ 1 − 0 .75
P0 = =
1 − ρ N +1 1 − (0 .75)10
0 .25
= = 0 .28
0 .90

Average queue length is given by

ρ2 [1 − N ρ N −1 + (N − 1) ρ N ]
E (m) or E ( Lq ) or Lq =
(1 − ρ) (1 − ρ N +1)

(0 .75)2 [1 − 9 × (0 .75)8 + 8 (0 .75)9 ]

=
0 .25 (1 − (0 .75)10 )
[1 − 0 .303]
= 2 .22 = 1.55
[1 − 0 .005]
 253

Example 19: The railway marshalling yard is sufficient only for trains (there being 10 lines, one
of which earmarked for the shunting engine to reverse itself from the crest of the hump to the rear of
train). Trains arrive at the rate 30 trains per day, inter-arrival time follows on exponential
distribution and service time distribution is also exponential with an average of 36 minutes calculate
the following:
(i) The probability that the yard is empty
(ii) The average line length

Solution: Here
1
λ = 30 trains per day or trains per minute
48
1
µ= trains per minute
36

N =9
λ 36
∴ ρ= = = 0 .75
µ 48

1−ρ 1 − 0 .75
(i) P0 = = = 0 .28
1 − ρN +1 1 − (0 .75)10

N 9
(ii) Lq = P0 ∑
n=0
n ρ n = 0 .28 × ∑
n=0
n (0 .75)n

= 0 .28 × 9 .58 = 2 .68 or 3 trains

6.10 Multi-Channel Queuing Theory Model III

( M / M / C : ∞ / FCFS)
Multi-channel queuing theory treats the condition in which there are several service
stations in parallel and each element in the waiting line can be served by more than one
station. The arrival rate λ and service rate µ are mean values from Poisson distribution
and exponential distribution respectively, service discipline is first come, first served and
customers are taken for single queue

Let n = number of customers in the system

Pn = probability of n customers in the system

c = number of parallel service channels (c > 1)

λ = arrival rate of customers

µ = service rate of individual channel

 254

 1  λ n
   P0 for 1 ≤ n < c − 1
 n!  µ 
1. Pn = 
n
 1  λ
 c n − c c !  µ  P0 for n ≥ c


−1
 c −1 n c 
1  λ 1  λ µc 
2. P0 = 
 ∑  
n!  µ 
+  
c! µ µ c − λ 
n = 0 
3. Expected (average) number of customers in the system
c
 λ
λ .µ  
µ λ
Ls = P0 +
(c − 1) !(c µ − λ )2 µ

4. Expected (average) number of customers waiting in the queue

c
 λ
λµ  
µ
Lq = P0
(c − 1) !(c µ − λ )2

5. Average time a customer spends in the system

c
 λ
µ 
L µ 1
Ws = s = P0 +
λ (c − 1) !(c µ − λ )2 µ

6. Average waiting time of a customer in the queue

c
 λ
µ 
Lq µ
Wq = = P0
λ (c − 1) !(c µ − λ )2

7. Probability that a customer has to wait

c
 λ
µ 
µ
P (n ≥ c) = P0
(c − 1) !(c µ − λ )

8. Probability that a customer enters the service without waiting

c
 λ
µ 
µ
1 − P (n ≥ c) = 1 − P0
(c − 1) !(c µ − λ )

9. Average number of idle servers

= c − (average number of customers served)
λ
10. Utilisation rate ρ =
cµ
 255

Example 20: A tax consulting firm has 3 counters in its office to receive people who have problems
concerning their income, wealth and sales taxes. On the average 48 persons arrive in an 8-hour day.
Each tax adviser spends 15 minutes on an average on an arrival. If the arrivals are Poissonly
distributed and service time are according to exponential distribution, find
(i) The average number of customers in the system.
(ii) Average number of customers waiting to be served.
(iii) Average time a customers spends in the system.
(iv) Average waiting time for a customer.
(v) The number of hours each week a tax adviser spends performing his job.
(vi) The probability that a customer has to wait before he gets service.

Solution: It is ( M / M / C) : (∞ FCFS) problem.

48
Here λ= = 6 per hour, c = 3
8

1
µ= × 60 = 4 per hour
15

λ 6 3
= =
µ 4 2

Now to find the probability P0 which is the probability of having no customer in the
system.

1
P0 = n c
 λ  λ
c −1    
µ µ cµ
∑
n=0
n!
+
c!
.
cµ −λ

1
= 2
( λ / µ)n
(λ / µ)3 3µ
∑ n!
+
3!
×
3µ − λ
n=0

1
= 3
 λ
2  
λ 1  λ µ 3µ
1+ +   +
µ 2 µ 6 3µ − λ

1 1 8 4
= = = =
 3 9  27 12 29 9 38 19
+
1 + +  + .
 2 8 48 12 − 6 8 8

= 0 .21
 256

(i) Average number of customers in the system.

c
 λ
λµ  
µ λ
Ls = 2
P0 +
(c − 1) !(c µ − λ ) µ
3
 3
6 ×4 ×  
 2 3
= × 0 .021 + = 1.74
2 !(12 − 6) 2

(ii) Average number of customers waiting to be served.

λ 3
Lq = Ls − = 1.74 − = 0 .24
µ 2

(iii) Average time a customer spends in the system.

L 1.74
Ws = s = = 0 .29 hour
λ 6
= 17 .4 minutes.

(iv) Average waiting time for a customer.

Lq 0 .24
Wq = = = 0 .04 hour = 2 .4 minutes
λ 6
λ 6 1
(v) Utilisation factor ρ = = =
cµ 3 ×4 2
Therefore number of hours each day a tax, adviser spends doing his job
1
= × 8 = 4 hours
2
∴ on an average, a tax adviser is busy 4 × 5 = 20 hours based on 5-working day week.

(vi) Probability that a customer has to wait,

c
 λ
µ 
µ
P (n > c) = P0
(c − 1) !(µ c − λ )
3
 3
4× 
 2
= × 0 .21
2 !(12 − 6)

= 0 .236

Example 21: A telephone exchange has two long distance operators. The telephone company finds that,
during the peak load, long distance calls arrive in Poisson fashion at an average rate of 15 per hour. The
length of service on these calls is approximately exponentially distributed with mean length 5 minutes.
(i) What is the probability that a subscriber will have to wait for his long distance call during the
peak hours of the day?
(ii) If the subscribers will wait and serviced in turn, what is the expected waiting time?
[B.C.A. (Bhopal) 2006, 2008, 2012]
 257

Solution: It is ( M / M / C) : (∞ / FCFS) problem

15 1
Here, λ= = calls/minute
60 4

1
µ= calls/minute, c = 2
5

λ 1/4 5
ρ= = =
µ c (1 / 5) × 2 8

Now to find P0 by

1
P0 = c
 λ
c −1  
( λ / µ) n  µ  µc
∑
n=0
n!
+
c!
.
cµ −λ

1
= n 2
 5  5 1
1     ×2
 4  4
∑ n!
+ .
2! 
5
1 1
n=0 2 × − 
 5 4

1 3
= =
5 25 × 8 13
1+ +
4 32 × 3
1!

(i) Probability that a subscriber will have to wait for his long distance call
∞
= Probability (n ≥ 2) = ∑
n=
Pn
2

∞ 1
= ∑
n=
Pn − ∑ Pn = 1 − P0 − P1
n=
0 0

3  1  5
1   1  λ
n 
=1 − −    P0  , Pn =   P0 , n ≤ c 
13 1 !  4    n! µ 

3 5 3
=1 − − ×
13 4 13

25
=
52

= 0 .48
 258

(ii) Expected waiting time = wq

Lq
=
λ
c
 λ
µ 
µ
= P0
(c − 1) !(c µ − λ )2

2
1  5
 
5  4
= 2
 1 1
1 ! 2 × − 
 5 4

125
= = 3 .2 minutes.
39

Example 22: Four counters are being run on the frontier of a country to check the passports and
necessary papers of the tourists. The tourists choose counter at random. If the arrivals at frontier is
λ
Poisson at the rate λ and the service time is exponential with parameter , what is the steady-stage
2
average queue at each counter? [B.C.A. (Indore) 2010; B.B.A. (Meerut) 2005]

Solution: This problem is of ( M / M / C) : (∞ / FCFS) model

λ λ 1
we have λ = λ, µ = , c = 4, ρ = =
2 µc 2

1
P0 = n c
 λ  λ
c −1    
µ µ µc
∑
n=0
n!
+ .
c! µc −λ

1
=
λ
3 n 4 ×4
2
∑
(2) (2)
+
n! 4! λ ×4 − λ
n=0
2
1 3
= 2 3
=
2 2 2 4 23
1+ + + +
1! 2 ! 3 ! 3

Average queue length = Lq

c
 λ
λµ  
µ
= P0
(c − 1) !(c µ − λ )2
 259

λ
λ× (2)4
2 3
= 2
×
 λ  23
3 ! 4 × − λ
 2 

λ 2 × 23 3 4
= × =
6 × (λ 2 ) 23 23

Example 23: A super market has two girls ringing sales at the counters. If the service time for each
customer is exponential with mean 4 minutes, and if people arrive in a Poisson fashion at the counter
at the rate of 10 an hour:
(i) What is the probability of having to wait for service?
[B.C.A. (I.G.N.O.U.) 2008; B.C.A. (Agra) 2008]

(ii) What is the expected percentage of idle time for each girl?

Solution: This problem is base on ( M / M / C) : (∞ / FCFS) model.

10 1
Here, λ= = people/minute
60 6

1
µ= people/minute
4

λ 1
ρ= = ,c =2
µc 3

Now we have to find P0

1
P0 = n c
 λ  λ
c −1    
µ µ µc
∑
n=0
n!
+
c!
×
(µ c − λ)

1
or P0 = c − 1
n
(c ρ) (c ρ)c
∑
n=0
n!
+
c !(1 − ρ)

1
=
  2
n
 2
2 
 1     
 3  3

∑ n!
+

−
1 

n = 0 2 ! 1 

 3  

1 1
= =
(2 / 3) 1  2
1+ + 
1! 3
 260

(i) Probability to having to wait for service.

∞
= P (n ≥ 2) = ∑ Pn = 1 − P0 − P1
n=2

1 (2 / 3)  (c ρ)n 
=1 − − × P0 Pn = P0 , s ≤ n
2 1!  n! 

1 2 1 1
= − . = = 0 .167
2 3 2 6
1
(ii) Fraction of time the service remains busy = ρ (traffic intensity) =
3

The fraction of the time the service remains idle

1 2
=1 − =
3 3

i.e. expected percentage of idle time for each girl

2
= × 100% = 67%
3

6.11 Model IV ( M / E k / 1) : ( ∞ / FCFS)

This is a queuing model with Poisson arrivals, Erlang service time with k phase, single
channel, first-come-first served discipline and infinite production. In this model one unit
is served in k phases. The arrival or departure of one unit, therefore, means an increase or
decrease of k phases in the system and the completion of service of one phase of a unit will
mean the decrease of one phase in the system.

If at any time there are m units waiting in the queue with one unit in service which has to
wait pass through s phase, then the total number of phases n in the system at that time
will be given by

n = nk + s

Now, µ = the number of units served per unit time.

µ k = the number of phases served per unit time.

∴ λ n = λ, µ n = µ
The characteristic properties of this model are as follows:
1. Average number of units in the system
 261

(k + 1)  λ   λ  λ
= Ls =     +
2 k  µ µ − λ µ

2. Average number of units in the queue

(k + 1)  λ   λ 
Lq = .  . 
2k  µ µ − λ

3. Average time spend by a unit in the system

(k + 1) λ 1
Ws = . +
2k µ (µ − λ) µ

4. Average waiting time of a unit in the queue

(k + 1) λ
Wq = .
2k µ (µ − λ)

5. When k → ∞ then
1 λ2 λ
Ls = . +
2 µ (µ − λ) µ

1  λ λ
Lq =  .
2  µ  (µ − λ )
1 λ 1
Ws = . +
2 µ (µ − λ) µ
1 λ
Wq = .
2 µ (µ − λ)

When k = 1 then the Erlang service time distribution reduces to exponential distribution
and values of Ls , Lq , Ws and Wq are same as Model I.

6. Probability that there are no unit in the system

P0 = (1 − ρ k)
λ
where ρ= , k = phase
µ

Example 24: In a factory cafeteria the customers have to pass through three counters. The
customers busy coupons at the first counter, select and collect the snacks at the second counter and
collect tea at the third. The server at each counter takes on an average 1.5 minutes although the
distribution of service time is approximately exponential. If the arrival of customers to the cafeteria is
approximately Poisson at an average rate of 6 per hour, calculate
(i) The average time a customer spends waiting in the cafeteria.
(ii) The average time of getting the service.
(iii) The most probable time in getting the service.

Solution: This problem is of ( M / Ek / 1) : (∞ / FCFS) model.

 262

Here, number of phase k = 3

Service time per phase = 1.5 minutes

∴ Service time per customer = 1.5 × 3 = 4 .5 minutes

1 40
∵ µ= customers/minute = customer/hours
4 .5 3

λ = 6 customers/hour

(i) Average waiting time

 k + 1  λ   1 
Wq =      
 2 k   µ µ − λ
 
 3 + 1  6 × 3  1 
=     
 2 × 3  40   40
− 6
 3 
9 27
= hours = = 2 .45 minutes
220 11

(ii) Average time of getting the service i.e. in collecting coupons, snacks etc is the mean
of t when it is following the IIIrd member of Erlang family.
1 3
∴ Average time spent = = hours = 4 .5 minutes
µ 40

(iii) The most probable time spend in getting the service is the model value of t for the
IIIrd member of Erlang family.
k −1
∴ Most probable time spend =
µk
3 −1
=
40
×3
3
1
= hour = 3 minutes.
20

Example 25: A hospital clinic has a doctor examining patients brought in for a general check up.
The doctor averages 4 minutes on each phase of the check up although the distribution of time spent on
each phase is approximately exponential. If each patient goes through four phases in the check up and
if the arrivals of the patients is the doctors office are approximately Poisson at an average rate of 3 per
hour, what is the average time spent by a patient waiting in the doctor’s office? What is the average
time spent in the examination? What is the most probable time spent in the examination?
[B.C.A. (Kurukshetra) 2006]

Solution: This problem is base on ( M / Ek / 1) : (∞ / FCFS) model.

Here, we have
 263

3 1
λ = mean arrival rate = = patients /minutes
60 20
k = number of phase = 4

µ = mean service time per phase = 4 × 4 = 16 minutes

(i) Average time spent by a patient waiting in the doctor’s office

 
 k + 1  λ  4 +1 1 / 20

Wq =     =   = 40 minutes
 2 k   µ (µ − λ) 8  1  1 1 
  − 
 16 16 20  

(ii) Average time spent in the examination

1
= = 16 minutes
µ

(iii) Most probable time spent in the examination

k −1 4 −1
= = = 12 minutes
µk  
1
  ×4
16 

Example 26: Repairing a certain type of machine which breaks down in a given factory consists
of 5 basic steps that must be performed sequentially. The time taken to perform each of the 5 steps is
found to have an exponential distribution with mean 5 minutes and is independent of other steps. If
these machines breakdown in a Poisson fashion at an average rate of two per hour and if there is only
one repairman, what is the average idle time for each machine that has broken down?

Solution: This problem is base on ( M / Ek / 1) : (∞ / FCFS) model

Here, we have
k = number of phase = 5

Service time per phase = 5 × 5 = 25 minutes

1 12
∵ µ= minutes/minute = units/hour
25 5
λ = 2 units/hour
Average ideal time of the machine = Average time spent by the machine in the system
 k + 1  λ   1  1
Ws =       +
 2 k   µ µ − λ µ

   
5 +1  2   1  1 5
Ws =     + = hours
2 × 5  12   12 − 2  12 3
 5   5  5
= 100 minutes.
 264

Example 27: A colliery working one shift per day uses a large number of locomotives which
breakdown at random intervals, an average one failing per 8-hour shift. The fitter carries out a
standard maintenance schedule on each faulty locomotive. Each of the 5 main parts of this schedule
1
takes an average of hour but the time varies widely. How much time will be fitter have for other
2
tasks and what is the average time a locomotive is out of service?

Solution: This problem is base on ( M / Ek / 1) : (∞ / FCFS) model.

Here, we have
1
k = 5, λ = per hours
8
1
Service time per part = hour
2
5
∴ Service time per locomotive = hours
2
2
∴ µ= per hours
5

Fraction of time the fitter will have for other tasks

= Fraction of time for which the fitter is idle

λ 1/8 5 11
=1 − =1 − =1 − =
µ 2 /5 16 16

11
∴ Time the fitter will have for other tasks in a day = × 8 = 5 .5
16

Average time a locomotive is out of service

= Average time spent by the locomotive in the system
 
 k + 1  λ   1  1 5 + 1 (1 / 8)  1 
=      + = × .  + 5 /2
 2 k   µ   µ − λ  µ 2 × 5 (2 / 5)  2 − 1 
 5 8
6 5 40 5
= × × + = 3 .18 hours
10 16 11 2
Example 28: A barber with one-man taken exactly 25 minutes to complete one hair cut. If
customers arrive in a Poisson fashion at an average rate of one every 40 minutes, how long on the
average must a customer wait for service? [B.B.A. (Rohilkhand) 2007]

Solution: This problem is based on ( M / Ek / 1) : (∞ / FCFS) model.

Here, we have
1
λ = mean arrival rate = customers/minutes
40
 265

Since service rate is constant then k = ∞

1
µ = mean service rate = customers/minute
25
Average waiting time of a customer in queue
 k + 1  λ 
Wq = lim    
k → ∞  2 k   µ (µ − λ )

λ
=
2 µ (µ − λ )
1 / 40 126
= = = 20 .8 minutes.
1 1 1 6
2×  − 
25  2 40 

Example 29: In a car manufacturing plant, a loading crane takes exactly 10 minutes to load a car
into wagon and again come back to the position of loading another car. If the arrivals of a cars is a
Poisson stream at an average of one every 20 minutes. Calculate the average waiting time of a car.

Solution: Here, we have k = ∞, λ = 3, µ = 6

Average waiting time of a car is given by

1 λ 1 1 3 1
Wq = . = × ×
2 µ µ − λ 2 6 6 −3
1
= hour = 5 minutes.
12

Example 30: At a certain airport it takes exactly 5 minutes to land an air plane, once it is given the
signal to land. Although in coming planes have scheduled, arrival times, the wide variability in
arrival times products an effect which makes the incoming planes appear to arrive in a Poisson
fashion at an average rate of six per hour. This produces occasional stock-ups at the airport which can
be dangerous and costly. Under these circumstances, how much time will a pilot expect to spent
circling the field waiting to land?

Solution: This problem is base on ( M / Ek / 1) : (∞ / FCFS) model.

Here, we have
6 1
λ = mean arrival rate = = airplanes/minute
60 10
1
µ = mean service rate = airplanes/minutes
5
Since service rate is constant then k = ∞
λ
Average waiting time (Wq ) =
2 µ (µ − λ )
1 / 10
= = 2 .5 minutes.
1 1 1 
2×  − 
5  5 10 
 266

P roblem S et
1. Explain, what you mean by a Poisson process ?
[B.C.A. (Avadh) 2008]
2. Give some important properties of the Poisson process.
3. Write short note on queuing models.
[B.B.A. (Meerut) 2002, 2003, 2004]
4. Give the essential characteristics of queuing process.
[B.B.A. (Meerut) 2005]
5. In a public telephone booth the arrivals are on the average 15 per hour. A call on the
average takes 3 minutes. If there is just one phone.
(i) What is the expected number of callers in the booth at any time?
(ii) For what proportion of time is the booth expected to be idle?
[B.C.A. (Lucknow) 2005]
6. Customers arrive at a box office window, being manned by single individual,
according to a Poisson input process with a mean rate of 20 per hour. The time
required to serve a customer per on exponential distribution with mean of 90 seconds
find the average waiting time of a customer. Also determine the average number of
customers in the system and average queue length.
[Hint: λ = 0 .5, µ = 0 .67 then used Wq , Lq and Ls ]

7. At a one-man barber shop, customers arrive according to Poisson distribution with

a mean arrival rate of 5 per hour and his hair cutting time was exponentially
distributed with an average hair cut taking 10 minutes. It is assumed that because
of his excellent reputation, customers were always willing to wait. Calculate the
following:
(i) Average number of customers in the shop and the average number of customers
waiting for a hair cut.
(ii) The percentage of time an arrival can walk right in without having to wait.
(iii) The percentage of customers who have to wait prior to getting into the
barber’s chair.
[B.C.A. (Rohilkhand) 2006]
1
[Hint: λ = 1 / 2 customers/minutes, µ = customers per minutes ρ = 5 / 6
10
λ λ2 26
(i) Ls = = 5, Lq = =
µ−λ µ (µ − λ) 5
1 1
(ii) P0 = 1 − ρ = Required % = × 100 = 16 .7%
6 6
1 5
(iii) P (n ≥ 0) = 1 − P0 = 1 − = = . 833
6 6
Required % = 83 .3%]
 267

8. Customers arrive at a first class ticket counter of a theater in a Poisson distributed

arrival rate of 25 per hour. Service time is constant at 2 minutes. Calculate
(i) the mean number in the waiting line.
(ii) the mean waiting time.
25 5 1 5
[Hint: λ = = customers per minutes, µ = cost/min, ρ =
60 12 2 6
25
(i) Ls = 5 customers, Lq = = 4 customers
6
λ λ
(ii) Wq = = 10 min, Ws = = 12 min]
µ (µ − λ) µ−λ
9. In a bank, cheques are cashed at a single teller counter. Customers arrive at the
counter in a Poisson manner at an average rate of 30 customers per hour. The teller
takes, on an average a minute and a half to cash cheque. The service time has been
shown to be exponentially distributed.
(i) Calculate the percentage of time the teller is busy.
(ii) Calculate the average time a customer is expected to wait.
[Hint: λ = 30 cus/h, µ = 40 cus/h

1 
(i) 1 − P0 = ρ (ii) Ws = 
µ − λ

10. Data have been accumulated at a banking facility regarding a waiting time for
delivery trucks to be loaded. The data show that the average arrival rate for the
trucks at the loading dock is 2/hour. The average time to load a truck, using
2 loaders is 10 minutes so that the service rate is 3 trucks per hour. The
management is considering hiring another loader at ` 5 per hour to reduce the
loading time. Drivers are paids ` 4 per hour and truck utilisation is valued at ` 3 per
hour. Should the additional loader be hired if an increase in the service rate to
4 trucks per hour would result?
λ 2
[Hint: Without additional loader Ls = = =2
µ − λ 3 −2

∴ cost/hour = ` 2 (3 + 4) = ` 14
2
with additional loader Ls = =1
4 −2

∴ cost/hour = ` [1 (3 + 4) + 5] = ` = 12]
11. In a railway marshalling yard, goods trains arrive at a rate of 30 trains per day.
Assuming that the inter-arrival time follow as exponential distribution and the
service time distribution is also exponential with an average 36 minutes, calculate
the following:
(i) Average number of customers in the system.
(ii) The probability that the queue size exceeds 10.
(iii) If the input of trains increase to an average 33 per day. What will change in
(i) and (ii)?
 268

12. The tool room company’s quality control department is manned by a single clerk
who takes an average of 5 minutes in checking parts of each of the machine coming
for inspection. The machines arrive once in every 8 minutes on the average. One
hour of the machine is valued at ` 15 and a clerk’s time is valued at ` 4 per hour.
What are the average hourly queuing system costs associated with the quality
control department?
[B.C.A. (Kashi) 2007]
60 60
[Hint: λ = = 7 .5 / hr, µ = = 12 / hr
8 2
1 1 2
Ws = = = hr
µ − λ 12 − 7 .5 9
 2 10
Average queuing cost/machine = ` 15 ×  = `
 9 3
10
∴ Average queuing cost/hr = ` × 7 .5 = ` 25
3
Average cost of a clerk/hr = ` 4
∴ Total cost for the department/hr = ` 29]
13. Cars arrive at a toll gate on a frequency according to Poisson distribution with
mean 90/hour. Average time for passing through the gate is 38 seconds. Drivers
complain of long waiting time. Authorities are willing to decrease the passing time
through the gate to 30 seconds by introducing new automatic devices. This can be
justified only if under the old system, the number of waiting cars exceeds 5. In
addition, the percentage of the gate’s idle time under the new system should not
exceed 10% can the new device be justified?
[B.C.A. (Kanpur) 2004]
60 30
[Hint: λ = 90 / hr = 1.5 minutes, µ = = per minutes
38 19
λ λ 1.5 × 19 15
∴ λq = . = × = 18 (> 5)
µ µ−λ 30 30
− 1.5
19

λ 15 
P0 = 1 − =1 − = 0 .25 > 0 .11
µ′ 60

30 
14. Problems arrive at computer centre in a Poisson fashion at an average rate of 5/day.
The rules of the computer centre are that any man waiting to get his problem solved
must aid the man whose problem is being solved. If the time to solve a problem with
1
one man has an exponential distribution with mean time of day and if the average
3
solving time is inversely proportional to the number of people working on the
problem, find the expected time for person entering the line.
∞ ∞ ∞
ρ n −ρ ρ n −1 [B.B.A. (Meerut) 2011]
[Hint: Ls = ∑ n Pn = ∑ n
n!
e = e −ρ ρ ∑ (n − 1) !
=
n 0 n=0 n =1
 269

5 1 1 
= e −ρ ρ eρ = ρ = 5 / 3 , Ws = × = day = 8 hrs 
3 5 3 
15. A super marks has two girls ringing up sales at the counters. If the service time for
each customer is exponential with mean 4 minutes and if people arrive in a Poisson
fashion at the rate of 10 per hours
(i) What is the probability of having to wait for the service?
(ii) What is the expected percentage of idle time for each girl?
(iii) Find the average queue length and the average number of units in the system.
λ 10
[Hint: ρ = = = 1 / 3,
cµ 2 × 5
2
Fraction of time the service channels remain idle = 1 − ρ =
3
2 
∴ Expected percentage of idle time for each girl = × 100 = 67% 
3 
16. A warehouse in a small state receives orders for a certain item and sends them by a
truck as soon as possible to the customer. The orders arrive in a Poisson fashion at a
mean rate of 0.9 per day. Only one item at a time can be shipped by a truck from
the warehouse, which is located in the control part of a state. Because the customers
are located in various places in the state, the distribution of service time in days has
a distribution with probability density 4 t e − 2 t . What is the expected delay between
the arrival of an order and the arrival of the item to the customer? Service time here
implies the time the truck takes to load, gets to the customer, unloads and returns
to the warehouse loading and unloading times are small as compared to the travel
time. [B.C.A. (I.G.N.O.U.) 2007]
k +1 λ
[Hint: µ = 1, λ = 0 .9, k = 2, wq = = 6 .75 days
2 k µ (µ − λ)

k +1 λ 1 
Ws = + = 7 .75 days 
2 k µ (µ − λ) µ 
17. In a Bhawan cafeteria it was observed that there is only one bearer who takes
exactly 4 minutes to serve a cup of coffee once the order has been placed with him.
If the students arrive in the cafeteria at an average rate 10 per hour, how must time
one is expected to spend waiting for his tern to place the order?
18. A petrol pump station has two pumps. The service time follow the exponential
distribution with a mean of 4 minutes and cars arrive for service in a Poisson
process at the rate 10% per hour. Find the probability that a customer have to wait
for service. What proportion of time the pumps remain idle?
19. A barber with one man shop takes exactly 40 minutes to complete one hair cut. If
the inter-arrival time of the customer follows in exponential distribution with
average one every 50 minutes, how long a customer must wait for service?
20. Explain the different types of queuing models. Why must the service rate be greater
than the arrival rate in a single channel queuing system? [B.B.A. (Meerut) 2008]
 270

21. Discuss M / M /1 queue system with its important properties.

[B.B.A. (Meerut) 2002, 2004, 2005, 2006, 2008]
22. Discuss the basics of a queuing model.
[B.B.A. (Meerut) 2006]
23. Explain the objectives of a queuing model. Explain properties of M / M / 1:( FCFS)
system.
[B.B.A. (Meerut) 2002, 2003, 2011]

13.
A nswers
5. 1
(i) 3 (ii)
4

6. Wq = 4 .5 min/customer, Lq = 2 .25, Ls = 3

7. 1
(i) 5 customers in shop and 4 customers waiting for hair cut
6

(ii) 16.7% (iii) 83.3%

8. (i) Ls = 5 customers, Lq = 4 customers

(ii) Wq = 10 min, Ws = 12 minutes

9. (i) 75 % (ii) 6 minutes

10. ` 2 per hour saving with new loader

11. (i) 3 (ii) (0 .75)10 = 0 .06

(iii) 4.8, (0 .83)10 = 0 .2

12. Ws = 2 / 9 hr, Total cost for deptt/hr = ` 29

13. Lq = 18, P0 = 0 .25 (> 0 .10)

14. Ls = 5 / 3, Ws = 8 hours

15. (i) 0.167 (ii) 67%

16. Wq = 6 .75 days Ws = 7 .75 days.

17. 4 minutes.

18. 0.167, 67% for each pump.

19. 1 hour and 20 minutes.

❍❍❍
 271

C HAPTER
7

Replacement Problems

7.1 The Replacement Problems

[B.C.A. (Bhopal) 2008, 2012; B.C.A. (Lucknow) 2011; B.C.A. (Agra) 2004,2012]

The efficiency of all industrial and military equipments deteriorate with time. Some
times the equipment fails completely and effects the whole system. For example, a
machine requires higher operating cost, a transport vehicle such as car or air plane
requires more and more maintenance cost. The ever increasing repair and maintenance
cost necessitates the replacement of equipment.

The replacement problems are concerned with situations that arise when some items
such as men, car, truck, air plane etc need replacement due to their decreased efficiency
or working capacities.
There are four type of replacement problem
1. Replacement of old items has become in bad condition and their working efficiency
decrease with time or require expensive maintenance.
2. Replacement of items when the system is completely fail due to accident or otherwise.
3. Problem in mortality or staffing.
4. Replacement of an equipment, when a better or more efficient design of machine or
equipment has become available in the market. For example, an equipment may
have an economic life of 20 years, yet it may become obsolete after 10 year because
of better technical developments.
 272

7.2 Replacement of Items that Deteriorate with Time

It is found that repair and maintenance costs of items increase with time and a stage may
become when these costs become so high that it is more economical to replace the item
by a new one. Thus, the problem of replacement in this case is to find the best time at
which the old machine should be replaced by new one.

Now, we shall consider a few cases of items that deteriorate with time and it will be
assumed that suitable expressions for maintenance costs are available.
To find the best replacement age (time) of a machine when
1. Its maintenance cost is given by a function increase with time.
2. Its scrap value is constant and
3. The money value is not considered.

7.2.1 Theorem
The cost of maintenance of a machine is given as a function increasing with time
and its scrap value is constant.
1. If time is measured continuously, then the average annual cost will be
minimized by replacing the machine when the average cost becomes equal to
the maintenance cost.
2. If time is measured in discrete units, then the average annual cost will be
minimized by replacing the machine when the next periods, maintenance
cost become greater than the current average cost.
[B.C.A.(Kanpur) 2006; B.C.A. (Rohilkhand) 2011]

Proof: 1. When time ‘t’ is a continuous variables.

Let C = Capital cost of the item or cost of the machine
S = Scrap value of the item
Cm (t) = Maintenance cost of the machine at the time
A (n) = Average annual total cost of the item
n = Number of years the item is to be in use
Annual cost of the item at any time
= capital cost − scrap value + maintenance cost at time t
n
Now total maintenance cost incurred during n years =
∫0 Cm (t) dt
n
Total cost incurred during n years = C − S +
∫0 Cm (t) dt
The average cost of the item per year during n year is given by
1 n 
A (n) = C − S +
n 0∫ Cm (t) dt 

 273

C−S 1 n
or A (n) =
n
+
n 0 ∫
Cm (t) dt ...(1)

Now we shall find that value of n for which A (n) is minimum, differentiating eq. (1)
w.r.t.n, we find
d − (C − S) 1 n 1
dn
( A (n)) =
n2
− 2
n ∫0 Cm (t) dt + n Cm (t)
d
For A (n) = 0, we have
dn
n
(C − S) +
∫0 Cm (t) dt
C m ( t) = = A (n) by (1)
n

i.e., Maintenance cost of item at time t = Average cost of the item per year
d2 A (n) 2 (C − S) 2 n
Also
dn 2
=
n 3
+ 3
n ∫0 Cm (t) dt > 0
Thus, the item should be replace when the average annual cost to date becomes equal to
the current maintenance cost.
2. When time ‘t’ is a discrete variable

In this case, the total maintenance cost of the machine during n years
n
= ∑ C m ( t)
m =1
...(1)

∴ Average annual cost of item during n years is given by

n
C−S+ ∑ C m ( t)
m =1
A (n) = ...(2)
n

But A (n) is minimum for that value of n for which

∆ A (n − 1) < 0 < ∆ A (n)

Now ∆ A (n) = A (n + 1) − A (n) ...(3)
 n +1   n 
C−S C−S 1
∑ ∑
1
= + C m ( t)  −  + C m ( t) 
 n+1 n+1   n n 
 m =1   m =1 

 n  n
 1 1  1  
∑ ∑
1
= (C − S)  −  +  Cm (t) + Cn +1 (t) − C m ( t)
 n + 1 n  n + 1   n
m = 1  m =1
n
(C − S)  1 1
∑ C m ( t) + n + 1 C n + 1 ( t)
1
=− + − 
n (n + 1)  n + 1 n
=
m 1
 274

 n 
(C − S) +
 m =1
∑
C m ( t) 
 1
=− + Cn + 1 (t)
 n ( n + 1)  n +1
 
 
A (n) 1
=− + C n + 1 ( t) (from 2)
n+1 n+1
A (n − 1) 1
Similarly, ∆ A (n − 1) = − + C n ( t)
n n

∴ A (n) is minimum for value of n for which

∆ A (n) > 0 and ∆ A (n − 1) < 0

A (n) 1
i.e., − + C n + 1 ( t) > 0
n+1 n+1

A (n − 1) 1
and − + C n ( t) < 0
n n

or Cn + 1 (t) > A (n) and Cn (t) < A (n − 1) ...(4)

Hence, we can say that

1. Do not replace if the next years maintenance cost is less than the previous years
average total cost.
2. Replace if the next years maintenance cost is greater than the previous years
average total cost.

This completes the proof.

S olved E xamples
Example 1: The cost of a machine in ` 6100 and its scrap value (resale value) is only ` 100. The
maintenance costs are found from experience to be under

Year 1 2 3 4 5 6 7 8

Maintenance cost 100 250 400 600 900 1250 1600 2000
in `

when should the machine be replaced?

[B.C.A. (Agra) 2004; B.C.A. (Lucknow) 2006; B.B.A. (Rohilkhand) 2006;
B.C.A. (Bhopal) 2008, 2009, 2011]

Solution: Given, C = ` 6100 and S = ` 100

Optimum replacement period is determine as below:
 275

Year (n) Maintenance Total C−S Total cost Average

cost Cm(t ) Maintenance =C−S cost
cost or + Σ Cm (t )  5
A (n) =  
cumulative  n
cost Σ Cm (t )

(1) (2) (3) (4) (5)=(3)+(4) (6)

1 100 100 6000 6100 6100.00

2 250 350 6000 6350 3175.00
3 400 750 6000 6750 2250.00
4 600 1350 6000 7350 1837.50
5 900 2250 6000 8250 1650.00
6 1250 3500 6000 9500 1583.33←
7 1600 5100 6000 11100 1585.71
8 2000 7100 6000 13100 1637.50

This table shows that the value A (n) during the 6 th year is minimum. Hence, the
machine should be replaced after every 6 th year.

Example 2: The cost of a machine is ` 6100 and its scrap value is ` 100. The maintenance costs
found from experience are as follows:

Year 1 2 3 4 5 6 7 8
Maintenance cost (in `) 100 250 400 600 900 1200 1600 2000

When should the machine be replaced ? [B.C.A. (Indore) 2010; B.B.A. (Delhi) 2008, 2009]

Solution: Since the scrap value of the machine is ` 100. Then resale value of machine
after one year become constant throughout.
The optimum replacement period is determine as below:
Year (n) Maintenance Total C−S Total cost Average
cost Cm (t ) Maintenance =C−S cost
cost or + Σ Cm (t )  5
A (n) =  
cumulative  n
cost Σ Cm (t )
(1) (2) (3) (4) (5) = (3) + 4 (6)
1 100 100 6000 6100 6100
2 250 350 6000 6350 3175
3 400 750 6000 6750 2250
4 600 1350 6000 7350 1837
5 900 2250 6000 8250 1650
6 1200 3450 6000 9450 1575 ←
7 1600 5050 6000 11050 1579
8 2000 7050 6000 13050 1631
 276

This table shows that the value A (n) during the 6 th year is minimum. Hence, the machine
should be replaced after every 6 th year.

Example 3: Following table give the running cost per year and resale price of a certain equipment
whose purchase price is ` 5000

Year 1 2 3 4 5 6 7 8
Running cost (in `) 1500 1600 1800 2100 2500 2900 3400 4000
Resale value (in `) 3500 2500 1700 1200 800 500 500 500

At what year is the replacement due ? [B.B.A. (Meerut) 2007]

Solution: Given C = ` 5000

The optimum replacement period is determine as below:

Year (n) Running Total Resale C − S Total cost = Average

cost running cost S (C − S ) + cost
Cm (t ) or Σ Cm (t )  5
A (n) =  
cumulative  n
cost Σ Cm (t )
(1) (2) (3) (4) (5) = (3) + (4) (6)
1 1500 1500 3500 1500 3000 3000
2 1600 3100 2500 2500 5600 2800
3 1800 4900 1700 3300 8200 2733.3
4 2100 7000 1200 3800 10800 2700 ←
5 2500 9500 800 4200 13700 2740
6 2900 12400 500 4500 16900 2816.6
7 3400 15800 500 4500 20300 2900
8 4000 19800 500 4500 24300 3037.5

This table shows that the value of A (n) during the 4 th year is a minimum. Hence, the
equipment should be replaced every 4 th year.

Example 4: A fleet owner finds his past records that the costs per year of running a vehicle whose
purchase price is ` 50,000, are as under:

Year 1 2 3 4 5 6 7

Running cost (in `) 5000 6000 7000 9000 11500 16000 18000

Resale value (in `) 30000 15000 7500 3750 2000 2000 2000

Thereafter, running cost increases by ` 2000 but resale value remains constant at ` 2000.
At what age replacement due ? [B.C.A. (Agra) 2003]

Solution: Given C = 50000

 277

Year (n) Running Total running cost C−S Total cost Average
cost Cm (t ) or cumulative cost =C−S cost
Σ Cm (t ) + Σ Cm (t )  5
A (n) =  
 n
(1) (2) (3) (4) (5) = (3) + (4) (6)

1 5000 5000 20000 25000 25000

2 6000 11000 35000 46000 23000
3 7000 18000 42500 60500 20167
4 9000 27000 46250 73250 18312
5 11500 38500 48000 86500 17300
6 16000 54500 48000 102500 17083 ←
7 18000 72500 48000 120500 17214

This table shows that the value of A (n) during the 6 th year is a minimum. Hence, the
vehicle should be replaced every 6 th year.
Example 5: Fleet cars have increased their costs as they continue in service due to increased direct
operating cost (gas and oil) and increased maintenance (repairs, tyres, batteries etc.) The initial cost
is ` 3,500 and trade in value drops as time passes until it reaches a constant value of ` 500. Given
cost of operating, maintaining and trade in value determine the proper length of service before cars
should be replaced.

Years of service 1 2 3 4 5
Year end trade in value 1900 1050 600 500 500
Annual operating cost 1500 1800 2100 2400 2700
Annual maintaining cost 300 400 600 800 1000
[B.B.A. (Garhwal) 2008]
Solution: Given C = ` 3500
S = year end trade in value
Cm (t) = operating cost + maintenance cost
The optimum replacement period is determined in the table below:

Year (n) Cm (t ) Σ Cm (t ) (C − S ) Total cost = Average cost

(C − S ) + Σ Cm (t )  5
A (n) =  
 n

(1) (2) (3) (4) (5) = (2) + (3) (6)

1 1800 1800 1600 3400 3400
2 2200 4000 2450 6450 3225
3 2700 6700 2900 9600 3200 ←
4 3200 9900 3000 12,900 3225
5 3700 13600 3000 16, 600 3320
 278

This table shows that the value of A (n) during the third year is minimum. Hence, the car
should be replaced every third year.

Example 6: (i) Machine A cost ` 9000. Annual operating costs are ` 200 for the first year and
then increase by ` 2000 every year. Determine the best age at which to replace the machine. If the
optimum replacement policy is followed. What will be the average yearly cost of owning and
operating the machine ?
(ii) Machine B costs `10000. Annual operating costs are ` 400 for the first year and then increase
by `. 800 every year. You now have a machine of type A which is one year old. Should you replace it
with B if, so when ? [B.C.A. (Kurukshetra) 2007, 2011]

Solution: (i) It is given that machine A has no resale value, when replaced. The average
annual cost is computed as below:

Year (n) Maintenance Total S (C − S ) Total cost Average

cost Cm(t ) maintenance cost
cost Σ Cm(t ) A (n)
= (5 / n)
(1) (2) (3) (4) (5) =(4) + (3) (6)

1 200 200 Nil 9000 9200 9200

2 2200 2400 Nil 9000 11400 5700
3 4200 6600 Nil 9000 15600 5200 ←
4 6200 12800 Nil 9000 21800 5450
5 8200 21000 Nil 9000 30000 6000

From this table we find that machine A should be replaced at the end of 3 year and average
yearly cost of owning and operating the machine at this time replacement is ` 5200.
(ii) For machine B, the average cost per year can be calculated as follows:

Year (n) Maintenance Total S C−S Total cost Average

cost Cm (t ) maintenance cost A (n)
cost Σ Cm (t ) =  
6
 n
(1) (2) (3) (4) (5) (6 ) =(5 ) + (3 )

1 400 400 Nil 10000 10400 10400

2 1200 1600 Nil 10000 11600 5800

3 2000 3600 Nil 10000 13600 4533

4 2800 6400 Nil 10000 16400 4100

5 3600 10000 Nil 10000 20000 4000 ←

6 4400 14400 Nil 10000 24400 4066

 279

Since the minimum average cost for machine B is lower than that for machine A, machine
should be replaced by machine A.
Now we have to determine as to when machine A should be replaced. Machine A should
be replaced when the cost for next year of running this machine becomes more than the
average yearly cost for machine B.
Now, total cost of machine A in the first year = ` 9200
total cost of machine A in the second year = ` 11400 − ` 9200
= ` 2200
total cost of machine A in the third year = ` 4200
total cost of machine A in the fourth year = 6200
As the cost of running machine A in third year (` 4200) is more than the average yearly
cost for machine B (` 4000): machine A should be replaced at the end of two years. i.e.,
one year after it is one year old (one year hence).

Example 7: For a machine, for the following data are available:

Year 0 1 2 3 4 5 6

Cost of spares (`) − 200 400 700 1000 1400 1600

Salary maintenance staff (`) − 1200 1200 1400 1600 2000 2600

Loss due to break- down (`) − 600 800 700 1000 1200 1600

Resale value (`) 12000 6000 3000 1500 800 400 400

Determine the optimum period for replacement of the above machine.

Solution: The optimum period for replacement is determine in the table below:

Year Cost of Salary of Losses due C−S Total Average

(n) spares maintenance to break- depreciation cost cost
(`) staff (`) down (`) (1) A (n) =
(1 / n)

1 200 1200 600 6000 8000 8000

2 600 2400 1400 9000 13400 6700

3 1300 3800 2100 10500 17700 5900

4 2300 5400 3100 11200 22000 5500

5 3700 7400 4300 11600 27000 5400 ←

6 5300 10000 5900 11600 32800 5466.66

This table shows that the value of A (n) during 5 th year is minimum. Hence, the machine
should be replaced every 5 th year.
 280

Example 8: (i) An auto-rickshaw driver finds from his pervious records that the cost per year of
running an auto-rickshaw whose purchase prices is ` 7000 is as given below:

Year 1 2 3 4 5 6 7 8

Running cost (`) 1100 1300 1500 1900 2400 2900 3500 4100

Resale Price (`) 3100 1600 850 475 300 300 300 300

At what age is the replacement due?

(ii) Another person has three auto rickshaws of the same purchase price and cost of running each as
import (i). Two of these vehicles are 2 years old and the third one is 1 year old. He is considering a new
type of auto-rickshaw with 50% more capacity than one of the old ones and at a unit price of ` 9000.
He estimates that the running costs and resale price for the new vehicle will be as follows:

Year 1 2 3 4 5 6 7 8

Running cost (`) 1300 1600 1900 2500 3200 4100 5100 6200

Resale Price (`) 4100 2100 1100 600 400 400 400 400

Assuming that the loss of flexibility due to fewer vehicles is of no importance, and that he will continue
to have sufficient work for three of the old vehicles, what should be his policy?
[M.B.A. (Meerut) 2003]

Solution: (i) The average annual cost for old auto-rickshaw is determine as follows:

Year Annual Total Resale Purchase Total cost Average

(n) maintenance maintenance value price- = [(C − S ) cost
cost Cm (t ) cost or S ` resale + (ΣCm (t )] ` annual
` cumulative value cost
cost Σ Cm (t ) ` (C − S ) `  6
A (n) =  
 n
(1) (2) (3) (4) (5 ) (6) = (5)+ (3)
1 1100 1100 3100 3900 5000 5000
2 1300 2400 1600 5400 7800 3900
3 1500 3900 850 6150 10050 3350
4 1900 5800 475 6525 12325 3081
5 2400 8200 300 6700 14900 2980
6 2900 11100 300 6700 17800 2967 ←
7 3500 14600 300 6700 21300 3043
8 4100 18700 300 6700 25400 3175

This table shows that the value of A (n) during 6 th year is minimum. Hence, the old
auto-rickshaw should be replace at the end of every 6 th year.
 281

(ii) Now let us determine the average annual cost of the new auto-rickshaw of larger
capacity.

Year Annual Total Purchase Resale Total cost Average

(n) maintenance maintenance price- value (S ) = (C − S ) annual
cost Cm (t ) cost or resale + (Σ Cm (t ) cost
cumulative value = (6 )
A (n) =
cost Σ Cm (t ) (C − S ) n

(1) (2) (3) (4) (5) (6) = (4) + (3)

1 1300 1300 4900 4100 6200 6200

2 1600 2900 6900 2100 9800 4900
3 1900 4800 7900 1100 12700 4233
4 2500 7300 8400 600 15700 3925
5 3200 10500 8600 400 19100 3820 ←
6 4100 14600 8600 400 23200 3867
7 5100 19700 8600 400 28300 4043
8 6200 25900 8600 400 34500 4312

As the new auto-rickshaw has 50% more capacity than the old one, the minimum average
annual cost of ` 3820 for the former is equivalent to
2
` 3820 × = ` 2547 for the latter. Since this amount is less than ` 2967 for it, the latter
3
will be replaced by new auto-rickshaw.

The new vehicles will be purchased when the cost for the next year of running the three
old vehicles becomes more than the average annual cost of the two new ones.
Total annual cost of one smaller auto-rickshaw during first year = ` 5000
annual cost of one smaller auto-rickshaw during second year = ` 7800 − ` 5000 = ` 2800
annual cost of one smaller auto-rickshaw during third year = ` 2250
annual cost of one smaller auto-rickshaw during fourth year = ` 2275
annual cost of one smaller auto-rickshaw during fifth year = ` 2575
annual cost of one smaller auto-rickshaw during sixth year = ` 2900 and so on.

Total cost during next first year for two smaller vehicles aged two years and one vehicle
aged one year = 2 × 2250 + 2800 = ` 7300
Similarly, total cost during next second year = 2 × 2275 + 2250 = ` 6800
total cost during next third year = 2 × 2575 + 2275 = ` 7425
total cost during next fourth year = 2 × 2900 + 2575 = ` 8375
and so on.
But minimum average cost for two new vehicles = 2 × 3820 = ` 7640
 282

As the total cost of old vehicles during next third year is less than the minimum average
cost of the new vehicles and becomes more only in the next fourth year. Hence, old
auto-rickshaws should be replaced by the new larger ones after the next third year of this
life.

P roblem S et
1. Write short note on replacement problems.
2. Discuss some important replacement situations. [B.B.A. (Meerut) 2012]

3. What is replacement? Describe some important replacement situations.

4. A firm is considering when to replace its machine whose price is ` 12200. The scrap
value of the machine is ` 200 only. From past experience the maintenance costs of
the machine are as under:

Year 1 2 3 4 5 6 7 8

Maintenance cost in (`) 200 500 800 1200 1800 2500 3200 4000

Find when the new machine should be purchased. [B.B.A. (Delhi) 2007]
5. The cost of a truck is ` 10000. The salvage value and the running costs are given
below. Find the most economical age for replacement.

Year 1 2 3 4 5 6 7

Running cost (`) 3000 3200 3600 4200 5000 5800 6800

Resale value (`) 7000 5000 3400 2400 1600 1000 1000

6. Machine A costs ` 45000 and the operating costs are estimated at ` 1000 for the
first year increasing by ` 10,000 per year in the second and subsequent years.
Machine B costs ` 50000 and operating cost are ` 2000 for the first year increasing
by ` 4000 per year in the second and subsequent years. If you now have a machine
of type A, should you replace it with B ? If so, when ? Assume that both the
machines have no resale value, and that the future costs are not discounted.

7. Explain with examples the failure mechanism of items. [B.C.A. (I.G.N.O.U) 2012]

8. What are the situations which makes the replacement of items necessary ?

A nswers
4. After 6 th years
5. After 4 th years
6. Machine A should be replaced by B, when its age is 2 years.
 283

7.3 Money Value Present Worth Factor (P.W.F.) and

Discount Rate

7.3.1 Money Value

The value of money change with time. Since has value over time. This can be explain with
the help of following example:

If we borrow ` 100 at the rate of interest 10% per year and spend this amount today, then
we have to pay ` 110 after one year

Therefore
` 110 after one year = ` 100 today
100
∴ ` 1 after one year =`
110
10
= . )− 1
= (11
(11)
. )− n today
So, ` 1.00, n year after from now = ` (11

at the rate of 10%.

7.3.2 Present Value or Present Worth Factor

The quantity (1.1)− n is called the present worth factor (pwf) or present value of one
rupee spend n years from now.
If r is the rate of interest then (1 + r )− n is called the present worth factor (pwf) of one
rupee spent in n years.

7.3.3 Discount Rate (Depreciation Value)

1
The present worth factor of unit amount to be spent after one year is given by v =
(1 + r )
where r the rate of interest.
Then v is called discount rate or depreciation value.

Example 9: The yearly cost of two machines A and B, when money value is neglected is shown
following table:

Year 1 2 3
Machine A (`) 1800 1200 1400
Machine B (`) 2800 200 1400

[B.C.A. (I.G.N.O.U.) 2012]

Solution: The total expenditure for three years for machine A
= 1800 + 1200 + 1400 = ` 4400
The total expenditure for three years for machine B
 284

= 2800 + 200 + 1400 = ` 4400

Thus, the two machines are equally good if the money has no value over time.
Now consider the money value at the rate of 10% per year, the discount rate
1 1 1
v= = = = 0 .9091
1+
10 1 + 0 .10 1.1
100
The discount cost patterns for machines A and B for three year is shown in table:

Year 1 2 3 Total cost

(`)

Machine A (Discounted 1800 1200 × 0 .9091 1400 × (0 .9091)2 404794

cost in `) = 1090 .90 = 1157 .04

Machine B (Discounted 2800 200 × 0 .9091 1400 × (0 .9091)2 4138.86

cost in `) = 181.82 = 1157 .04

The data show that the cost for machine A is less than that for machine B, machine A is
more economical.

Example 10: Let the value of money be assumed to be 10% per year and suppose that machine A
is replaced after every 3 years where as machine B is replaced after every 6 years. The yearly costs of
both the machines are given as under:

Year 1 2 3 4 5 6

Machine A (`) 1000 200 400 1000 200 400

Machine B (`) 1700 100 200 300 400 500

Determine which machine should be purchased.

Solution: The present worth factor is given by

1 1 100 100 100 10
v= = = = = =
1 + r 1 + 10 100 + 10 100 + 10 110 11
100

∴ Total discount cost of A for three years

 10 10  
2
= ` 1000 + 200 × + 400 ×   
 11  11 

= ` 1512 (approximately)

∴ Total discount cost of B for six years

 10 10 
2
10 
3
= ` 1700 + 100 × + 200 ×   + 300 ×  
 11  11  11
 285

4 5
10  10 
+ 400 ×   + 500 ×   
 11  11 

= ` 2765
1512
Average annual cost of A= = ` 504 ...(1)
3
2765
Average annual cost of B= = ` 461 ...(2)
6

From (1) and (2), we find advantage with B but this comparison is unfair because the
period of both machines are different.

If we consider 6 th year period for machine A, then total discount of

2 3
10 10  10 
A = 1000 + 200 × + 400 ×   + 1000 ×  
11  11  11
4 5
10  10 
+ 200 ×   + 400 ×   = ` 2647
 11  11
Which is ` 118 less than machine B over the same period so machine A should be purchased.

7.4 Replacement of Items whose Maintenance Costs

Increase with Time and Value of Money also Change
with Time

7.4.1 Theorem
The maintenance cost increases with time and the money value decreases with
constant rate i.e., depreciation value is given. Then replacement policy will be
1. Replace if the next period’s cost is greater than the weighted average of previous
costs.
2. Do not replace if the next period’s cost is less than the weighted average of
previous costs. [B.C.A. (Rohilkhand) 2009; B.C.A. (Rohtak) 2008]
Proof: Suppose that the item (may be a machine or equipment etc.) is available for use
over a series of time periods of equal interval say one year.
Let A = purchase cost of a machine
Cr (r = 1, 2 , 3 … n) be maintenance costs in year
i.e., C1, C2 , … Cn be maintenance costs in years 1, 2, 3 … n respectively.
r = rate of interest
1
v= is the present worth of a rupee to be spent a year
1+ r
The present value of maintenance costs C1, C2 … Cn in the years 1, 2, 3 … n are
C1, C2 v, C3 v2 , … Cn v n − 1 respectively
The present value of the total expenditure on the machine in n years will be given by
 286

n
P (n) = A + C1 + C2 v + C3 v2 + … + Cn v n −1 = A + ∑ Cr v r −1
r=1

If the machine is replaced after n years from now

Then, the present value of the total expenditures on the machines with a replacement
policy of n years is given by
C (n) = [ A + C1 + C2 v + C3 v2 + … + Cn v n −1]
+ [( A + C1) v n + C2 v n +1 + C3 v n + 2 + … + Cn v2 n −1]
+ [( A + C1) v2 n + C2 v2 n +1 + … + Cn v3 n −1] + ...
 n   n   n 
= A +
 ∑ Cr v r −1  + v n  A +
  ∑
Cr v r −1  + v2 n  A +
  ∑
Cr v r −1  + …

 r =1   r =1   r =1 
 n 
 1 
= A +
 ∑ Cr v r −1  [1 + v n + v2 n + ....]

n 2n
∵1 + v + v + ..... = 1 – v n 
 r =1 
n
C(n) = A + ∑ Cr v r − 1
=
r 1
n
A+ ∑
r= 1
Cr v r − 1
P (n)
or C(n) = n
= ...(1)
1− v 1 − vn
Now we are interested to find that value of n for which C(n) is minimum.
C(n) is minimum for that value of n for which
∆ C(n − 1) < 0 < ∆ C(n) ...(2)
Now,
∆ C(n) = C(n + 1) − C(n)
n +1 n
A+ ∑= Cr v r −1 A+ ∑= Cr vr −1
r 1 r 1
= n +1
−
1− v 1 − vn

 n +1   n 
 r −1  r −1
 A + ∑ C r v  (1 − v n
) −  A + C r v ∑ n +1
 (1 − v )
 r =1   r =1 
=
(1 − v n +1) (1 − v n)

 n +1   n 
   
A (v n +1 − v n) +  ∑
Cr v r −1 (1 − v n) −  ∑
Cr v r −1 (1 − v n +1)
 r =1   r =1 
=
(1 − v n +1) (1 − v n)
 287

n
− A (1 − v) v n + ∑ Cr vr −1 {(1 − vn) − (1 − vn+1)} + Cn+1 vn (1 − vn)
r =1
=
(1 − v n +1) (1 − v n)

  n 
vn  r −1 
= .
(1 − v n +1) (1 − v n) 
(1 − v n) C
n +1 − (1 − v) .  A + C∑r v 
  r =1  

  n 
v n (1 − v) n
(1 − v ) C  r −1 
or ∆ C(n) = .
(1 − v n +1) (1 − v n)  (1 − v)
−
n +1  A + ∑C r v 
  r =1  

Similarly,

  n −1 
v n −1 (1 − v) n −1
(1 − v ) C −  A + r −1 
∆ C (n − 1) = .
(1 − v n) (1 − v n −1)  (1 − v)
n  C r∑v 
  r =1  

Now, from (2), C(n) is minimum for that value of n for which

  n −1 
v n −1 (1 − v) 1 − v
n −1
 r −1 
(1 − v n) (1 − v n −1)  1 − v
. C −
n  A + C r∑v  <0
  r =1  

  n 
v n (1 − v) n
1 − v C  r −1 
< .
(1 − v n +1) (1 − v n)  1 − v
−
n +1  A + C r v ∑ 
  r =1  

 n −1   
1 − v n −1
n
  1 − vn  
or
1− v
Cn −  A + ∑
Cr v r −1 < 0 <
1− v
. Cn +1 −  A + ∑
Cr v r −1
 r =1   r =1 

 v n −1 (1 − v) v n (1 − v) 
Since n n −1
and n +1 n
are both positive as v < 1
 (1 − v ) (1 − v ) (1 − v ) (1 − v ) 

1 – vn − 1 1 − vn
or . Cn − P (n −1) < 0 < . Cn +1 − P (n) ...(3)
1– v 1− v
Thus, the best replacement age of the machine is n years for which inequality (3) holds.
Again from (3), we find

 n 
1 − vn  
1− v
Cn +1 −  A + ∑
Cr . v r −1 > 0
 r =1 

 n −1 
1 − v n −1  
and
1− v
Cn −  A + ∑
Cr . v r −1 < 0
 r =1 
 288

n n −1
A+ ∑
r=
Cr . v r −1 A+ ∑ Cr . v r −1
=1
1 r
or Cn +1 > and Cn <
1 − vn 1 − v n −1
1− v 1− v
n n −1
A+ ∑ Cr . v r −1
A+ ∑ Cr . v r −1
r =1 r= 1
or Cn +1 > and Cn <
1 + v + v2 + … v n −1 1 + v + v2 + … v n −1

1 − vn
Since = 1 + v + v2 + … + v n −1
1− v

or Cn +1 > R(n) and Cn < R (n −1) ...(4)

n
A+ ∑
r=
Cr . v r −1
1
where R (n) = = weighted average of costs in n years.
1 + v + v + v3 + … + v n −1
2

Hence, we conclude,
1. Do not replace if the operating cost of the next period (year) is less than the
weighted average of previous costs.
2. Replace if the operating cost of the next period (year) is greater than the weighted
average of the previous costs.

7.4.2 Theorem
A discounted cost P (n) is invested by taking loan at the interest rate r; and the loan is
repaid by fixed annual payments say x, throughout the life of the machine. To find
the minimum value of x for optimum period n at which to replace the machine
[B.C.A. (Lucknow) 2004]

Proof: The present worth of fixed annual payments x for n years must be equal to P (n) the
sum borrowed.
∴ P (n) = x + vx + v2 x + ..... + v n −1 . x ...(1)
1
[where v = . (Discount rate)]
1+ r

= (1 + v + v2 + … + v n −1) x

1 − vn
= .x
1− v
(1 − v)
or x= . P (n) ...(2)
1 − vn

Since (1 − v) is constant, independent of n.

 289

∴ x is minimum when
P (n)
C (n) = ...(3)
1 − vn

is minimum.
C (n) given by (3) is minimum, when

∆ C (n − 1) < 0 < ∆ C (n) ...(4)

From (3), ∆ C (n) = C (n + 1) − C (n)

P (n + 1) P (n) (1 − v n) P (n + 1) − (1 − v n +1) P (n)
= − =
1 − v n +1 1 − v n (1 − v n +1) (1 − v n)

P (n + 1) − P (n) − v n [P (n + 1) − v. P (n)]
= ...(5)
(1 − v n +1) (1 − v n)

But if A is the cost of the machine and C1, C2 , ..., Cn are the running costs in years
1, 2, ..., n then the present value of the total expenditure in n years must be equal to P (n).
∴ P (n) = A + C1 + C2 v + C3 v2 + … + Cn v n −1

∴ P (n + 1) = A + C1 + C2 v + C3 v2 + … + Cn v n −1 + Cn +1 v n

= P (n) + Cn +1 v n.

∴ From (5), we get

Cn +1 v n − v n [P (n) + Cn +1 v n − v. P (n)]
∆ C (n) =
(1 − v n +1) (1 − v n)

(1 − v n) Cn +1 v n − (1 − v) P (n) . v n
=
(1 − v n +1) (1 − v n)

(1 − v) v n 1 − v n 
= n +1 n 1− v
. Cn +1 − P (n)
(1 − v ) (1 − v )  

∴ ∆ C (n) > 0, gives

(1 − v) v n 1 − v n 
n +1 n 1− v
. Cn +1 − P (n) > 0
(1 − v ) (1 − v )  

1 − vn (1 − v) v n
or Cn +1 − P (n) > 0 since is positive as 0 < v < 1
1− v (1 − v n +1) (1 − v n)

P (n) A + C1 + C2 v + C3 v2 + … + Cnv n −1
or Cn +1 > or Cn +1 > ...(6)
1 − v n  1 + v + v2 + … + v n −1
 
 1− v 

Similarly, put n = n − 1 in C (n) , we get

∆ C (n − 1) and ∆ C (n − 1) < 0
 290

1 − v n −1
∴ Cn − P (n − 1) < 0
1− v

A + C1 + C2 v + C3 v2 + … + Cn −1v n − 2
or Cn < ...(7)
1 + v + v2 + … + v n − 2

Hence, we conclude from (6) and (7) that

1 − v n −1 1 − vn
Cn − P (n − 1) < 0 < Cn +1 − P (n) ...(8)
1− v 1− v

From this, we find best replacement age n of machine for which C (n) or P (n) is minimum
and minimum of x is obtain from (2) equation.

7.4.3 Procedure for Finding best Machine

Step 1: Find the present value of the maintenance cost for each of the years

∑ Cn vn−1 where n = 1, 2 , 3 … and v = 1 + r , r = rate of interest

1
i.e.,

Step 2: Calculate cost plus the accumulated present values obtained in step 1

i.e., A+ ∑ Cn vn−1 where A = purchase cost

Step 3: Find the cumulative present value factors up to each of the year n = 1, 2 , 3 … i.e.,

∑ vn−1
Step 4: Determine annualized cost w (n) by adding the entries obtained in step 2 by the
corresponding entries obtained in step 3 i.e.,

A +
 ∑ Cn vn−1 
∑ vn−1
Once the annualized costs or weighted average costs for different years is obtained, the
following sale are followed to decide on the replacement

1. Do not replace the equipments if the next period’s cost is less than the weighted
average of previous costs.

2. Replace the equipments if the next period’s cost is greater than the weighted
average of previous costs.
 291

Example 11: Let v = 0 .9 and initial price is ` 5000. Running cost varies as follows:

Year : 1 2 3 4 5 6 7
Running cost (`) : 400 500 700 1000 1300 1700 2100

[B.C.A. (Avadh) 2010]

What would be the optimum replacement interval?

Solution: Given v = 0 .9, A = Initial price = ` 5000, Cn = maintenance cost

The optimum replacement age n is obtained in the following table :

Year
(n)
Maintenance
cost Cn
v n −1 Cn v n −1 A + ∑ Cnv n −1 ∑ v n −1 Weighted
average
w (n) (7)
(1) (2) (3) (4) (5) (6) =(5) / (6)

1 400 1.0 400 5400 1.00 5400

2 500 0.90 450 5850 1.90 3079

3 700 0.81 567 6417 2.710 2368

4 1000 0.729 729 7146 3.439 2083

5 1300 0.6556 853 7999 4.096 1953

6 1700 0.599 1016 9015 4.694 1921 ←

7 2100 0.540 1134 10149 5.234 1939

Since C5 < w (6) < C7 i.e., 1300 < 1921 < 2100.

Optimum replacement is just under 6 years.

Example 12: Assume that present value of one rupee to be spent in a year’s time is ` 0 .9 and A =
` 3000 capital cost of equipment and the running cost are given in the table below. When should the
machine be replaced?

Year (n) : 1 2 3 4 5 6 7

Running cost (`) : 500 600 800 1000 1300 1600 2000

Solution: We have v = 0 .9, A = ` 3000, v n−1 = discount factor and Cn (n = 1, 2, 3, … 7)

where Cn v n−1 = discounted cost, Cn = maintenance cost for n = 1, 2, 3, 4, 5, 6, 7 years.
 292

The optimum replacement age n is obtained in the following table:

Year
(n)
Maintenance
cost Cn
v n −1 Cn v n −1 A+ ∑ Cnv n −1 ∑ v n −1 w (n) =  
5
6

(1) (2) (3) (4) (5) (6) (7)

1 500 1.00 500 3500 1.00

3500
2 600 0.90 540 3540 1.90
1863
3 800 0.81 648 3648 2.71
1346
4 1000 0.73 730 3730 3.44
1084 ←
5 1300 0.66 858 3858 4.10
941
6 1600 0.59 944 3944 4.69
841
7 2000 0.53 1060 4060 5.22
778

Since C3 < W (4) < C5 . Optimum replacement period is 4 th year.

Example 13: The cost of a new machine is ` 5000. The maintenance cost of nth year is given by
Cn = 500 (n − 1) : n = 1, 2 , 3 … . . Suppose that the discount rate per year is 0 . 05 . After how many
years it will be economical to replace the machine by a new one?
[B.C.A. (Kashi) 2009; B.B.A. (Indore) 2008]

Solution: We are given A=`5000 and Cn = 500 (n − 1) , n = 1, 2, 3 … . Since the discount

rate of money per year is 0.05, the present worth of the money to be spent over a period
1
of one year is v = = 0 .9523
1 + 0 .05

The optimum replacement line is determined in the following table :

Year
(n)
Cn
Maintenance
v n −1 Cn v n −1 A + ∑ Cnv n −1 ∑ v n −1 w (n) =
(5 )
(6 )
cost
(1) (2) (3) (4) (5) (6) (7)
1 0 1.000 0 5000 1.00 5000
2 500 0.9523 476 6476 1.9523 3317
3 1000 0.9070 907 6383 2.8593 2232
4 1500 0.8638 1296 7679 3.7231 2062
5 2000 0.8227 1645 9924 4.5458 2083 ←
6 2500 0.7835 1959 11283 5.3293 2117

Since, w (n) is minimum for n = 5 and C4 = 1500 < w(5) as well as w (5) > C6 = 2500. It is
economical to replace the machine by new one at the end of five years.
 293

Example 14: A machine costs ` 10000 operating costs are ` 100 per year for the first five year. In
the sixth and succeeding years operating cost increases by ` 100 per year. Assuming a 10% discount
rate of money per year, find the optimum length of time to hold the machine before we replace it.

Solution: The discount rate of money per year is given as 10%. Therefore, the present
worth of the money to be spent a period of one year is
1 1 100
v= = = = 0 .9091
1+ r 1+ 10 110
100
The optimal replacement time is determined in the following table:

Year
n
Cn v n −1 Cn v n −1 A + ∑ Cnv n −1 ∑ vn −1 w (n)

(1) (2) (3) (4) (5) (6) (7) = (5) / (6)

1 500 1.000 500 10500 1.000 10500

2 500 0.9091 456 10956 1.9091 5731

3 500 0.8264 413 11369 2.7355 4156

4 500 0.7513 376 11745 3.4868 3368

5 500 0.6830 342 12087 4.1698 2898

6 600 0.62090 373 12460 4.7907 2600

7 700 0.5645 395 12885 5.3552 2400

8 800 0.5133 411 13266 5.8684 2260

9 900 0.4665 420 13686 6.3349 2160

10 1000 0.4241 424 14110 6.7590 2087

11 1100 0.3856 424 14534 7.1446 2034

12 1200 0.3506 421 14955 7.4952 1995

13 1300 0.3187 414 15369 7.8139 1966

14 1400 0.2897 406 15775 8.1036 1946

15 1500 0.2637 396 16171 8.3673 1932

16 1600 0.2397 384 16555 8.6070 1923

17 1700 0.2179 370 16925 8.8249 1917

18 1800 0.1981 357 17282 9.0230 1915

19 1900 0.1801 342 17624 9.2031 1915 ←

20 2000 0.1637 327 17951 9.3668 1914

The machine should be replaced after 19 years of service.

 294

P roblem S et
1. Purchase price of a machine in ` 3000 and its running cost is given in the table
below. If the discount rate is 0.90, find at what age the machine should be replaced.

Year 1 2 3 4 5 6 7

Running cost (`) 500 600 800 1000 1300 1600 2000

[B.B.A. (Bhopal) 2007]

2. If you wish to have a return of 10% per annum on your investment, which of the
following plans would your prefer ?

Plan A (`) Plan B (`)

Ist cost : 2,00,000 2,50,000

Scrap value after 15 years : 1,50,000 1,80,000

Excess of annual revenue : 25,000 30,000

Over annual disbursement

3. A truck is priced at ` 60000 and running costs are estimated at ` 6000 for each of
the first four years, increasing by ` 2000 per year in the fifth and subsequent years.
If money is worth 10% per year, when should the truck be replaced? Assume that
the truck will eventually be sold for scrap at a negligible price.
[B.B.A. (Delhi) 2008]

A nswers
1. After 3 years
2. Plan A
3. After 9 years

7.5 Replacement of Items that Fail Suddenly

There are many real life situations in which items do not deteriorate with time but fail
suddenly. A system usually consists of a large number of low cost items that increasingly
liable to failure with age some times, the failure of an item may cause a complete
breakdown of the system. The costs of failure, in such a case will be quite higher than the
cost of item itself. It is, therefore, to know as to when the failure is likely to take place so
that item can be replaced before it actually fails. The problem, then is to find the optimal
value of time t which minimizes the total cost involved in the system.
 295

The following two policies are followed:

1. Individual Replacement Policy: In that case an item is replaced immediately
after it fails.
2. Group Replacement Policy: In that case all items are replaced, irrespective of
whether they have failed or not, with a provision that if any item fails before the
optimal time, it may be individually replaced.

7.5.1 Individual Replacement Policy: Mortality Theorem

These problems are the special cases of the problems in industry where the failure of any
item can be treated as death and the replacement of any item on failure can be taken to be
a birth in the human populations. To determine the probability distribution of failure its
mortality tables are used.

7.5.1.1 Mortality Theorem (Only Statement)

A large population is subject to a given mortality law for a very long period of time. All
deaths are immediately replaced by births and there are no other entries or exits. Show
that the age distribution ultimately becomes stables and that the number of deaths per
unit time becomes constant. Which is equal to the size of total population divided by the
mean age at death.
The size of population (N)
i.e., A0 =
Mean age at death

7.5.1.2 Mortality Tables

Mortality tables for any item can be used to obtain the probability distribution of its life
span.
Let N = The total number of items in the system in the beginning
N (t) = Number of survivors at any time t.
Then the probability that any item will fail in the time interval (t − 1, t) is given by
N (t − 1) − N (t)
...(1)
N
and the probability that any item that survived upto the age (t − 1) will die in the next year
is given by
N (t − 1) − N (t)
...(2)
N (t − 1)

7.5.2 Group Replacement Policy

Group replacement policy is defined in the following theorem and later it is explained in
numerical problems.
 296

7.5.2.1 Theorem (Group Replacement Policy)

1. Group replacement should be made at the end of n th period if the cost of
individual replacement for the n th period is greater than the average cost per
period by the end of n periods.

2. Group replacement should not be made at the end of n th period if the cost of
individual replacement at the end of (n − 1) th period is not less than the

average cost per period by the end of (n − 1) periods.

Proof: Here it is proposed to replace all items at fixed interval ‘t’ whether they have
failed or not, and continue replacing failed items as and when they fail.

Let N = Total number of items in the system

N ( X ) = Number of items failed during Xth period, X = 1, 2, 3 … (n − 1)

C g = The cost per item when all the items are replaced as a group

Ci = The cost of replacing an individual item on its failure

C (n) = Total cost in the interval t consisting of n periods.

= N . C g + Ci [total number of failures in the periods 1, 2, 3, … (n − 1)]

n −1
= N . C g + Ci ∑ N ( X)
X =1
...(1)

C (n)
Average cost per period A (n) = ...(2)
n

Now in order to determine the replacement age ‘t’, the average cost per period should be
minimum.

The condition for minimum of A (n) is

∆ A (n − 1) < 0 < ∆ A (n) ...(3)

Now ∆ A (n) = A (n + 1) − A (n)

C (n + 1) C (n)
= −
n+1 n

C (n) + Ci N (n) C (n)

= −
n+1 n

C (n)
n Ci N (n) − C (n) Ci N (n) − n
= =
n (n + 1) (n + 1)
 297

C (n − 1)
Ci N (n − 1) −
Similarly, ∆ A (n − 1) = n −1
n

Put ∆ A (n) and ∆ A (n − 1) in (3), we get

C (n − 1) C (n)
Ci N (n − 1) − Ci N (n) −
n −1 n
<0 <
n n+1

C (n − 1) C (n)
or Ci N (n − 1) − < 0 < Ci N (n) − ...(4)
(n − 1) n

C (n)
or Ci N (n) > ...(5)
n
C (n − 1)
and Ci N (n − 1) < ...(6)
(n − 1)

Thus, from (5) and (6) the group replacement policy is completely established.

Example 15: The following mortality rates have been observed for a certain type of light bulb:

Week 1 2 3 4 5

Percent failing by end of week 10 25 50 80 100

There are 1000 bulbs in use and it costs ` 1.00 to replace an individual bulb which has burnt out. If
all bulbs were replaced simultaneously it would cost 25 paise per bulb. It is proposed to replace all
bulbs at fixed intervals, whether or not they have burnt out and to continue replacing burnt out bulbs
as they fail. At what intervals should all the bulbs be replaced? [M.C.A. (Meerut) 2008]

Solution: Let Pi be the probability that a light bulb, which was now when placed in
position for use, fails during the ith week of its life. Thus, following frequency
distribution is obtained assuming to replace burnt out bulbs as and when they fail.
10
P1 = the probability of failure in the Ist week = = 0 .10
100
25 − 10
P2 = the probability of failure in the IInd week = = 0 .15
100
50 − 25
P3 = the probability of failure in the IIIrd week = = 0 .25
100
80 − 50
P4 = the probability of failure in IVth week = = 0 .30
100
100 − 80
P5 = the probability of failure in Vth week = = 0 .20
100

Here, P1 + P2 + P3 + P4 + P5 = 1
 298

Since the sum of all probabilities can never be greater than unity, therefore all further
probabilities P6 , P7, P8 and soon will be zero.

Therefore, a bulb can not survive for more than five weeks i.e., a bulb which has survived
for four weeks is sure to fail in the fifth week. Further, we assume that the burnt out bulbs
in any week are replaced just at the end of that week.

Let N i = the number of replacement made at the end of i th week, while all 1000 bulbs
were now initially.

Then, we have

N0 = Number of bulbs in the beginning = 1000

N1 = Number of burnt out bulbs replaced at the end of first week

= N0 P1 = 1000 × 0 .10 = 100

N2 = Number of burnt out bulbs replaced at the end of second week

= N0 P2 + N1 P1 = 1000 × 0 .15 + 100 × 0 .10 = 160

N3 = Number of burnt out bulbs replaced at the end of third week

= N0 P3 + N1 P2 + N2 P1

= 1000 × 0 .25 + 100 × 0 .15 + 160 × 0 .1 = 281

N4 = Number of burnt out bulbs replaced at the end of fourth week

= N0 P4 + N1 P3 + N2 P2 + N3 P1

= 1000 × 0 .30 + 100 × 0 .25 + 160 × 0 .15 + 281 × 0 .10 = 377

N5 = Number of burnt out bulbs replaced at the end of the fifth week

= N0 P5 + N1 P4 + N2 P3 + N3 P2 + N4 P1

= 1000 × 0 .20 + 100 × 0 .30 + 160 × 0 .25 + 281 × 0 .15 + 377 × 0 .10 = 350

N6 = Number of burnt out bulbs replaced at the end of the sixth week

= N0 P6 + N1 P5 + N2 P4 + N3 P3 + N4 P2 + N5 P1

= 0 + 100 × 0 .20 + 160 × 0 .15 + 281 × 0 .25 + 377 × 0 .15 + 350 × 0 .10 = 230

N7 = Number of burnt out bulbs replaced at the end of the seventh week

= N0 P7 + N1 P6 + N2 P5 + N3 P4 + N4 P3 + N5 P2 + N6 P1

= 0 + 0 + 160 × 0 .20 + 281 × 0 .30 + 377 × 0 .25 + 350 × 0 .15 + 230 × 0 .10 = 286

N8 = Number of burnt out bulbs replaced at the end of eighth week

= N0 P8 + N1 P7 + N2 P6 + N3 P5 + N4 P4 + N5 P3 + N6 P2 + N7 P1
 299

= 0 + 0 + 0 + 281 × 0 .20 + 377 × 0 .30 + 350 × 0 .25 + 230 × 0 .15 + 289 × 0 .10 = 320

and soon.
Thus, we find that the number of bulbs failing each week increases till the fourth week,
then decreases and again increasing from seventh week. Thus, N i will continue to
oscillate till the system attains a steady state.
The average life of a bulb
= Σ X i Pi, where X i = week, Pi = Probability of failure in the i-th week
= X1 P1 + X2 P2 + X3 P3 + X4 P4 + X5 P5 + X6 P6 + …
= 1 P1 + 2 P2 + 3 P3 + 4 P4 + 5 P5 + 0
= 0 .10 + 2 × 0 .15 + 3 × 0 . 25 + 4 × 0 . 30 + 5 × 0 . 20 = 3 . 35
N 1000
∴ Average number of replacement per week = = = 299
(Mean Age) 3 . 35

∴ Average cost of weekly individual policy = ` 299

Now we consider the case of group replacement.

End of Total cost of group replacement in ` Average cost per

Week week in `

1 1000 × 0 .25 + 100 × 1 = 350 350

2 1000 × 0 .25 + (100 + 160) × 1 = 510 255

3 1000 × 0 . 25 + (100 + 160 + 281) × 1 = 791 263.66

Thus, the minimum cost per week is ` 255.00 if the bulbs are replaced as a group after
every two weeks and this cost is also less than the average cost of weekly individual
replacement policy. It is optimal to have a group replacement after every two weeks.

Example 16: The following mortality rates have been observed for a certain type of light bulbs:

End of week : 1 2 3 4 5 6

Probability of failure to date : 0.09 0.25 0.49 0.85 0.97 1.00

There are large number of such bulbs which are to be kept in working order. If a bulb fails in service,
its costs ` 3 to replace but if all the bulbs are replaced in the same operation, it can be done for only
` 0 .70 a bulb. It is proposed to replace all bulbs at fixed intervals, whether or not they have burnt
out, and to continue replacing burnt out bulbs as they fail.
(i) What is the best interval between group replacements?
(ii) At what group replacement price per bulb, would a policy of strictly individual replacement
become preferable to the adopted policy ?
 300

Solution: Let the total number of bulbs in use be 1000. Let Pi be the probability that a
new light bulb fails during the ith of its life.

Thus, we have

P1 = 0 . 09

P2 = 0 . 25 − 0 . 09 = 0 .16

P3 = 0 . 49 − 0 . 25 = 0 . 24

P4 = 0 . 85 − 0 . 49 = 0 . 36

P5 = 0 . 97 − 0 . 85 = 0 .12

P6 = 1. 00 − 0 . 97 = 0 . 03
Since the sum of all the probabilities is unity, all probabilities higher than P6 must be
zero i.e., P7 = P8 = P9 , etc. = 0. Thus all light bulbs are sure to burn out by the 6 th week.

Further we assume
(i) That light bulbs which fail during a week are replaced just before the end of that
week.
(ii) That the actual percentage of the failures during a week for a sub-population of the
bulbs with the same age is the same as the expected percentage of failures during
the week for that sub-population.
Let N i represent the number of replacements made at the end of ith week when all the
1000 bulbs are new initially. Then we have

N0 = N0 = 1000

N1 = N0 P1 = 1000 × 0 . 09 = 90

N2 = N0 P2 + N1 P1 = 1000 × 0 .16 + 90 × 0 . 09 = 168

N3 = N0 P3 + N1 P2 + N2 P1

= 1000 × 0 . 24 + 90 × 0 .16 + 168 × 0 . 09 = 269

N4 = N0 P4 + N1 P3 + N2 P2 + N3 P1

= 1000 × 0 . 36 + 90 × 0 . 24 + 168 × 0 .16 + 269 × 0 . 09 = 432,

N5 = N0 P5 + N1 P4 + N2 P3 + N3 P2 + N4 P1

= 1000 × 0 .12 + 90 × 0 . 36 + 168 × 0 . 24 + 269 × 0 .16 + 432 × 0 . 09 = 274

N6 = N0 P6 + N1 P5 + N2 P4 + N3 P3 + N4 P2 + N5 P1

= 1000 × 0 . 03 + 90 × 0 .12 + 168 × 0 . 36 + 269 × 0 . 24

 301

+ 432 × 0 .16 + 274 × 0 . 09 = 260

N7 = 0 + N1 P6 + N2 P5 + N3 P4 + N4 P3 + N5 P2 + N6 P1

= 90 × 0 . 03 + 168 × 0 .12 + 269 × 0 . 36 + 432 × 0 . 24

+ 274 × 0 .16 + 260 × 0 . 09 = 291

and so on.
Thus, we find that the number of bulbs failing each week increases till the 4 th week, then
decreases and again increases from 7th week. Thus N i, will continue to oscillate till the
system attains a steady state.
6
Average (expected) life of light bulbs = ∑= Xi Pi
i 1

= X1 P1 + X2 P2 + X3 P3 + X4 P4 + X5 P5 + X6 P6
= 1 × 0 . 09 + 2 × 0 .16 + 3 × 0 . 24 + 4 × 0 . 36 + 5 × 0 .12 + 6 × 0 . 03 = 3 . 35
1000
∴ Average number of failures per week = = 299.
3 . 35

∴ Cost individual replacement of bulbs = ` 3 × 299 = ` 897

Since the replacement of all the 1,000 bulbs in one operation costs ` 0.70 per bulb and
replacement of an individual bulb cost ` 3. The total cost of replacement is

End of Week Total cost of group replacement (`) Average cost per
week (`)

1. 1000 × 0. 70 + 90 × 3 = 970 970.00

2. 1000 × 0. 70 + 3 (90 + 168) = 1474 737.00

3. 1000 × 0. 70 + 3 (90 + 168 + 269) = 2281 760.33

(i) As the average minimum cost is in the 2 nd week, it is optimal to have a group
replacement after every two weeks.
(ii) Let ` x be the group replacement price for bulb. Then

1, 000 x + 3 (90 + 168)

` 897 <
2
∴ x > ` 1.02.

Example 17: It has been suggested by a data processing firm that they adopt a policy of
periodically replacing all the tubes in a certain piece of equipment. A given type of tube is known to
have the mortality distribution shown in the table:
 302

Tube failures/week : 1 2 3 4 5

Probability of failure : 0.3 0.1 0.1 0.2 0.3

There are approximately 1000 tubes of this type in all the combined equipment. The cost of replacing
the tubes on an individual basis is estimated to be ` 1.00 per tube and the cost of a group replacement
policy average ` 0 .30 per tube. Compare the cost of preventive replacement with that of remedial
replacement.

Solution: Let Pi be the probability that a tube, which was new when placed in the
position for use, fails during the i th week of its life. Then, we have

P1 = 0 . 3, P2 = 0 .1, P3 = 0 .1, P4 = 0 . 2, and P5 = 0 . 3

Let N i = the number of replacements at the end of the i th week.

N0 = number of tubes in the beginning = 1000

N1 = N0 P1 = 1000 × 0 . 3 = 300

N2 = N0 P2 + N1 P1 = 1000 × 0 .1 + 300 × 0 . 3 = 190

N3 = N0 P3 + N1 P2 + N2 P1

= 1000 × 0 .1 + 300 × 0 .1 + 190 × 0 . 3 = 187

N4 = N0 P4 + N1 P3 + N2 P2 + N3 P1

= 1000 × 0 . 2 + 300 × 0 .1 + 190 × 0 .1 + 187 × 0 .3 = 305

N5 = N0 P5 + N1 P4 + N2 P3 + N3 P2 + N4 P1

= 1000 × 0 . 3 + 300 × 0 . 2 + 190 × 0 .1 + 187 × 0 .1 + 305 × 0 . 3 = 489

and so on
Thus, we find that the number of tubes failing each week decrease till the third week, then
increase in 4 th and 5 th week, Thus, N i, will continue to oscillate till the system attains a
steady state.
5
The expected life of each tubes = ∑= Pi Xi
i 1

= P1 X1 + P2 X2 + P3 X3 + P4 X4 + P5 X5
= 0 . 3 × 1 + 0 .1 × 2 + 0 .1 × 3 + 0 . 2 × 4 + 0 . 3 × 5 = 3 .10
1000
Average number of failures per week =
3 .10

= 323 (approx.)
 303

Therefore, the cost of individual replacement

= ` (323 × 1) = ` 323

Now, since the replacement of all the 1000 tubes simultaneously cost ` 0.30 per tube and
the replacement of an individual tube on failure costs ` 1.

Now consider the cost of group replacement

End of Total (Group + individual) cost of group Average

Week replacement in ` cost (`)

1 1000 × 0 . 30 + 300 × 1 = 600 600

2 1000 × 0 . 30 + (300 + 190) × 1 = 790 395

3 1000 × 0 . 30 + (300 + 190 + 187) × 1 = 977 326

4 1000 × 0 . 30 + (300 + 190 + 187 + 305) × 1 = 1282 321 ←

5 1000 × 0 . 30 + (300 × 190 + 187 + 305 + 489) × 1 = 1771 354

Thus, the minimum cost per week is ` 321 if the tubes are replaced as a group after every
4 week and this cost is also less than the average cost of weekly individual replacement
policy. It is optimal to have a group replacement after every four weeks.

Example 18: A large computer installation contains 2000 components of identical nature which
are subject to failure as per probability distribution given below:

Week end 1 2 3 4 5

Percentage failure to date 10 25 50 80 100

Components which fail have to be replaced for efficient functioning of the system. If they are replaced
as and when failure occur, the cost of replacement per unit is ` 3 . Alternatively if all components are
replaced in one lot at periodical intervals and individually replaced only as such failures occur
between group replacement, the cost of component replaced is ` 1
(i) Assess which policy of replacement would be economical.
(ii) If group replacement is economical at current costs, then assess at what cost of individual
replacement would group replacement be uneconomical.
(iii) How high can the cost per unit in group replacement be to make a preference for individual
replacement policy?
Solution: (i) Assume Pi is the probability that a component fails during the ith week of its
life. Then, we have
10
P1 = = 0 .10
100
 304

25 − 10
P2 = = 0 .15
100
50 − 25
P3 = = 0 . 25
100
80 − 50
P4 = = 0 . 30
100
100 − 80
P5 = = 0 . 20
100

Let N i = the number of replacements made at the end of i th week. Then, we have

initial number of components.

N0 = initial number of components. = 2000

N i = N0 P1 = 2000 × 0 .10 = 200

N2 = N0 P2 + N1 P1 = 2000 × 0 .15 + 200 × 0 .10 = 320

N3 = N0 P3 + N1 P2 + N2 P1

= 2000 × 0 . 25 + 200 × 0 .15 + 320 × 0 .10 = 562

N4 = N0 P4 + N1 P3 + N2 P2 + N3 P1

= 2000 × 0 . 30 + 200 × 0 . 25 + 320 × 0 .15 + 562 × 0 .10 = 754

N5 = N0 P5 + N1 P4 + N2 P3 + N3 P2 + N4 P1

= 2000 × 0 . 20 + 200 × 0 . 30 + 320 × 0 . 25 + 562 × 0 .15 + 754 × 0 .10

= 700

Thus, we find that number of components which fail, increases till the fourth week and
then start decreasing.

Thus, N i will continue to oscillate till the system attains a steady state.
5
Expected life of each components = ∑= Xi Pi
i 1

= X1 P1 + X2 P2 + X3 P3 + X4 P4 + X5 P5

= 1 × 0 .10 + 2 × 0.15 + 3 × 0 .25 + 4 × 0 . 30 + 5 × 0 . 20

= 3 . 35 weeks
2000
Average number of failures per week =
3 . 35

= 597 per week

 305

Since, the cost of individual replacement is ` 3, the average cost of individual replacement
= ` 597 × 3 = ` 1791

Again, since the cost of 2000 components simultaneously is ` 1 per component, the
average cost for different group replacement policies is obtain as below:

End of Total (Group + individual) cost of group Average

Week replacement in ` cost (`)

1 2000 × 1 + 200 × 3 = 2600 2600

2 2000 × 1 + (200 + 320) × 3 = 3560 1780

3 2000 × 1 + (200 + 320 + 562) × 3 = 5246 1748

4 2000 × 1 + (200 + 320 + 562 + 754) × 3 = 7508 1877 ←

5 2000 × 1 + (200 + 320 + 562 + 754 + 700) × 3 = 9605 1921

Since the average cost is lowest against month 3, the optimal interval between group
replacements is 3 months. Also, since the average cost is less than ` 1.791 for individual
replacement policy, the policy of group replacement is better.

(ii) For a group replacement policy to be uneconomical, the average cost per month of
this policy should be greater than the average cost per month of the policy of individual
replacement. If k be the cost of an individual replacement (IR), we can determine its
value in respect of each of the group replacement policies and then decide on the basis of
the least of all the k value obtained.

Group Average cost of Average k value

replacement by group replacement cost of IR
month (I)

1 (2000 + 200 k) > 597 ⇒ k < 5 . 04

2 [(2000 + 520 k) / 2] > 597 ⇒ k < 2 .97

3 [(2000 + 1082 k) / 3] > 597 ⇒ k < 2 . 82 ←

4 [(2000 + 1836 k) / 4] > 597 ⇒ k < 3 . 62

5 [(2000 + 2535 k) / 5] > 597 ⇒ k < 4 .45

 306

Since the least of the k values is 2.82, it follows that if the cost of the individual
replacement is anything smaller than ` 2.82, the group replacement policy would be
uneconomical.

(iii) As in (ii) a preference for individual replacement policy implies that the average cost
per month of a group replacement policy should be greater then the average monthly case
under this policy. With p as the per unit cost of item replacement under group
replacement policy, we have

Group Replacement Average cost per month p-value

Group replacement IR

1 month (2000 p + 600) > 1791 ⇒ p > 0 .60

2 month [(2000 p + 1560) / 2] > 1791 ⇒ p > 1.01

3 month [(2000 p + 3246) / 3] > 1791 ⇒ p > 1.06 *

4 month [(2000 p + 5508 .6) / 4] > 1791 ⇒ p > 0 .83

5 month [(2000 p + 7607 .76) / 5] > 1791 ⇒ p > 0 .67

Since the largest of the p value is 1.06, it follows that per unit cost (of replacement on a
group basis) of any value greater than ` 1.06 would imply that an Individual
Replacement Policy would be preferred to a Group Replacement Policy.

7.6 Recruitment and Promotion Problems

Problems concerning recruitment and promotion of staff can sometimes be analysed in a
manner similar to that used in replacement problems in industry. In staffing problems,
with fixed total staff and fixed size of staff groups, the proportion of staff in each group
determines the promotion age, and conversely.

Example 19: An airline requires 250 assistant hostesses, 350 hostesses and 50 supervisors. Girls
are recruited at age 21 and, if in service, they retire at age 60 . Given the life table, determine:
(i) How many girls should be recruited each year?

(ii) At what age promotions should take place?

 307

Age (Years) No. in service Age (Year) No. in service

21 1000 40 125

22 700 41 120

23 500 42 112

24 400 43 105

25 300 44 100

26 260 45 92

27 230 46 88

28 210 47 80

29 195 48 72

30 180 49 65

31 170 50 60

32 165 51 53

33 160 52 45

34 155 53 40

35 150 54 32

36 145 55 26

37 140 56 20

38 135 57 18

39 130 58 15

59 10

60 −

Solution: If 1000 girls had been recruited each year for the past 39 years, the total
number of them serving upto the age of 59 years = 6603.
Total number of girls required in the airline
= 250 + 350 + 50 = 650
 308

(i) ∴ Number of girls to be recruited every year in order to maintain a strength of 650
1000
= × 650 = 98 (approximately)
6603
(ii) Let the assistant hostesses be promoted at the age x. Then upto age x-one year,
number of assistant hostesses required = 250

Now out of 650 girls, 250 are assistant hostesses, therefore out of 1000,
250
their number = × 1000 = 385 (approx.)
650
From table, this number is available upto the age of 24 years.

∴ Promotion of assistant hostesses is due in the 25 th year.

Now out of 650 girls, 350 are hostesses. Therefore, if we recruit 1000 girls.
350
The number of hostesses = × 1000 = 538 (approx.)
650

∴ Total number of assistant hostesses and hostesses in a recruitment of 1000

= 385 + 538 = 923

∴ Number of supervisors required = 1000 − 923 = 77
From the table, this number is available upto the age of 47 years.
∴ Promotion of hostesses is due in the 48th year.

Example 20: A faculty in a college is planned to rise to a strength of 50 staff members and then to
remain at that level. The wastage of recruits depends upon their length of service and is as follows:

Year 1 2 3 4 5 6 7 8 9 10

Total % who left up to 5 35 56 65 70 76 80 86 95 100

the end of year

[B.C.A. (Kurukshetra) 2010]

(i) Find the number of staff member to be recruited every year.

(ii) If there are seven posts of Head of department for which length of service is the only criterion of
promotion. What will be average length of service after which a new entrant should expect
promotion.

Solution: Let us consider the recruitment per year is 100. Then it is clear that the 100
who join in the 1st year will become zero in the 10 th year, the 100 who join in the 2 nd year
will (serve for 9 years and) become 5 at the end of the 10 th year and the 100 who join in
the 3 rd year will (serve for 8 years) and become 14 at the end of the 10 th year and so on.
Thus, when the equilibrium is attained, the distribution of length of service of staff
members will be as follows:
 309

Year No. of staff members

0 100

1 95

2 65

3 44

4 35

5 30

6 24

7 20

8 14

9 5

10 0

Total = 432

(i) Thus, if 100 staff members are recruited every year, the total member of staff
members after 10 years of service = 432

To maintain a strength of 50, the number to be recruited every year

100
= × 50 = 11.6
432

If it is assumed that they left immediately after completing x years service, the total
number will become = 432 − 100 = 332
100
And required intake = 50 × = 15
332

But in practice they may leave at any time in the year so that reasonable number of
11.6 + 15
recruitments per year = = 13 (approx.)
2
(ii) If we recruit 13 persons every year then we want 7 seniors. Hence, if we recruit 100
7
every year then we shall require × 100 = 54 (approx.) seniors.
13

It can be soon from table and 54 seniors will available if we promote them during 6 th
year of their service. (∴ 0 + 5 + 14 + 20 + 24 = 63 > 54)

Therefore, the promotion of a newly recruited staff member will be due after completing
5 years and before putting in 6 years of service.
 310

Example 21: Calculate the probability of staff resignation in each year from the following table:

Year (n) 0 1 2 3 4 5 6 7 8 9 10

No. of original staff 1000 940 820 580 400 28 190 130 70 30 0
in service at the end of
the year

[B.C.A. (Purvanchal) 2009]

Solution: Let Pt denote the probability of a staff resignation (failure) in the t-th year.

N (t − 1) − N (t)
Thus, we have Pt =
N

where N = total number of staff members is the system

N (t) = total number of staff members in the system at the end of t-th year

Here, N = 1000
N (0) = 1000, N (1) = 940, N (2) = 820, N (3) = 580,
N (4) = 400, N (5) = 280, N (6) = 190, N (7) = 130,
N (8) = 70, N (9) = 30, N (10) = 0.

∴ P1 = Probability that a staff will resign in 1st year

N (0) − N (1) 1000 − 940

= = = 0 . 06
N 1000

P2 = Probability that a staff will resign in the 2 nd year

N (1) − N (2) 940 − 820

= = = 0 .12
N 1000

P3 = Probability that a staff will resign in the 3 rd year

N (2) − N (3) 820 − 580

= = = 0 . 24
N 1000

P4 = Probability that a staff will resign in the 4 th year

N (3) − N (4) 580 − 400

= = = 0 .18
N 1000

P5 = Probability that a staff will resign in the 5 th year

N (4) − N (5) 400 − 280
= = = 0 .12
N 1000

P6 = Probability that a staff will resign in the 6 th year

 311

N (5) − N (6) 280 − 190

= = = 0 . 09
N 1000

P7 = Probability that a staff will resign in the 7th year

N (6) − N (7) 190 − 130

= = = 0 . 06
N 1000

P8 = Probability that a staff will resign in the 8th year

N (7) − N (8) 130 − 70

= = = 0 . 06
N 1000

P9 = Probability that a staff will resign in the 9 th year

N (8) − N (9) 70 − 30
= = = 0 . 04
N 1000

P10 = Probability that a staff will resign in the 10 th year

N (9) − N (10) 30 − 0
= = = 0 . 03
N 1000

Total = 1. 00

Since the sum of all these probabilities is 1,

so P11 = P12 = P13 … = 6.

Example 22: A research team is planned to rise to a strength of 50 chemists and then to remain at
that level. The wastage of recruits depends on their length of service which is as follows:

Year 1 2 3 4 5 6 7 8 9 10

Total % who have left 5 36 56 63 68 73 79 87 97 100

upon the end of the year.

What is the recruitment per year necessary to maintain the required strength ? There are 8 senior
posts for which length of service is the main criterium. What is the average length of service after
which new entrant expect his promotion in one of these pasts.

Solution: We see from given table that none of the recruits is continues in the service for
more than 10 years.

Now, we construct the following table for 100 recruitment every year.
 312

Year No. of person who No. of persons in Prob. of a person to be in

( x ) (1) leave at the end of services at the end of service at the end of the
the year ( x ) (2) year x (3 ) = 100 − (2 ) year x (4 ) = (3 ) / 100 , Px

0 0 100 1.00

1 5 95 0.95

2 36 64 0.64

3 56 44 0.44

4 63 37 0.37

5 68 32 0.32

6 73 27 0.27

7 79 21 0.21

8 87 13 0.13

9 97 3 0.03

10 100 0 0.00

Total 664 436

We find from the table that there are 436 chemists in the organisation if 100 are
recruited every year.

Therefore, in order to maintain a strength of 50 chemists, one must recruit

100 × 50
= 12 chemists every year.
436

If Px is the probability of a person to be in service at the end of year x, then out of 12

recruits the total number of survivals at the end of year x will be equal to 12 Px .

Now, we construct a table showing the total number of chemists in service at the end of
each year.
 313

Year x No. of chemists in service at the end of xth year= 12 Px

0 12 × 1.00 = 12
1 12 × 0 .95 = 11
2 12 × 0 . 64 = 8
3 12 × 0 .44 = 5
4 12 × 0 .37 = 4
5 12 × 0 .32 = 4
6 12 × 0 .27 = 3
7 12 × 0 .21 = 2
8 12 × 0 .13 = 2
9 12 × 0 .03 = 0
10 12 × 0 .03 = 0

There are 8 senior posts for which the length of service is the main criterion.

Since there are 3 + 2 + 2 = 7 chemists in service whose lengths of services are from 6 years
to 8 years which is less than the total number of senior posts.

Hence, the proportions of the new entrants will start by the end of 5 th year.

P roblem S et
1. Explain how the theory of replacement is used in the following problems:
(i) Replacement of items whose maintenance cost varies with time.
(ii) Replacement of items that fail completely. [B.C.A. (Rohilkhand) 2002, 2004]

2. The following mortality rates have been observed for a certain type of light bulbs:

Week 1 2 3 4 5

Percent failing by week end 10 25 50 80 100

There are, 1000 bulbs in use and it costs ` 2 to replace an individual bulb which has
burnt out. If all bulbs were replaced simultaneously, it would cost 50 paise per bulb,
it is proposed to replace all the bulbs at fixed intervals, whether or not they have
burnt out, and to continue replacing burnt out bulbs as they fail. At what intervals
should all the bulbs be replaced ?

3. The probability Pn of failure just before age n is shown below. If individual

replacement cost ` 1.25 and group replacement cost ` 0.50 per item, find the
optimal group replacement policy.
 314

n 1 2 3 4 5 6 7 8 9 10 11

Pn 0.01 0.03 0.05 0.07 0.10 0.15 0.20 0.15 0.11 0.08 0.05

4. A computer contains 10000 resistors. When only resistor fails it is replaced. The
cost of replacing a resistor individually is ` 1 only. If all the resistor are replaced at
the same time, the per resistor would be reduced to 35 paise. The percent surviving
at the end of month t is given below :

Month 0 1 2 3 4 5 6

% surviving at the end of month 100 97 90 70 30 15 0

What is the optimum replacement plan? [B.B.A. (Meerut) 2010]

5. The following mortality rates have been found for a certain type of coal cutter motor.

Weeks 10 20 30 40 50

Total % failure up to end of 10 5 15 35 65 100

weeks period

If the motors are replaced over the week end the total cost is ` 20. If they fail during
the week the total cost is ` 100 per failure. Is it better to replace the motors failure
and if so when?
6. The probability Pn of failure just before age n are shown below. If individual
replacement costs ` 1.25 and group replacement costs ` 0.50 per item, find the
optimal group replacement policy (assume that there are 1000 bulbs in use)

n 1 2 3 4 5 6 7 8 9 10 11

Pn 0.01 0.03 0.05 0.07 0.10 0.15 0.20 0.15 0.11 0.08 0.05

7. Discuss briefly the various type of replacement problems.

[B.C.A. (Agra) 2004, 2012; B.C.A (Lucknow) 2011]
8. State some of the simple replacement policies.
9. What are three strategies of replacement of solving replacement problems?

A nswers
2. Every two weeks
3. After every 6 weeks
4. Replace every 3 months
5. After every 20 weeks
6. After every 5 weeks
mmm
 315

C HAPTER
8

Inventory Theory

8.1 Introduction
An inventory consists of a usable, but idle resources such as money, machines and men.
When the resource involved is a material, the inventory is also called ‘stock’. An
inventory problem is said to exist if either the resources are subjected to control or if
there is at least, one such cost that decreases as inventory increases. The objective is to
minimize total cost. However, in situations where inventory affects demand, the
objective may also be to maximize profit.

8.2 Necessity for Maintaining Inventory

[B.C.A. (Rohtak) 2007, 2012; B.C.A. (Bhopal) 2011]
1. It helps in smooth and efficient running of an enterprise. It decouples the
production from customers and venders and simplifies otherwise complex
organization for manufacture and reduces the co-ordination effort.
2. It provides service to the customer at a short notice. Timely deliveries can fetch
more goodwill and orders.
3. It helps in maintaining economy by absorbing some of the fluctuations when the
demand for an item fluctuates or is seasonal.
4. Process and moment inventories are quite necessary in big enterprises where
significant amounts of times are required to tranship items from one location to
another.
 316

5. It reduces product costs since there is an added advantage of batching and long,
uninterrupted production runs.
6. It acts as a buffer stock when raw materials are received late and shop rejections are
too many.

8.3 Inventory Costs

The three costs considered inventory control models are:
1. Inventory carrying cost or stock holding costs.
2. Procurement costs or set-up costs.
3. Storage costs.
4. Total inventory cost.

8.3.1 Inventory Carrying Costs or Stock Holding Costs

They arise an account of maintaining the stocks and the interest paid on the capital tied
up with stocks. They vary directly with the size of the inventory as well as the time the
item is held in stock. Some components of the stock holding cost are:
1. Cost of Money or Capital Tied up in Inventories: More borrowed from the
banks may cost interest of about 18%. It is generally taken somewhere around 15%
to 20% of the value of the inventories. [B.C.A. (Agra) 2008]

2. Cost of Storage Space: This consists of rent for space. Besides space expenses, this
will also include heating, lighting and other atmospheric control expenses. Typical
values may vary from 1 to 3%. [B.C.A. (Agra) 2008]

3. Depreciation and Deterioration Costs: They are important for fashion items or
items undergoing chemical changes during storage. Fragile items such as crockery
are liable to damage, breakage etc. 0.2% to 1% of the stock value may be lost due to
damage and deterioration.
4. Pil Ferage Cost: It depends upon the nature of the item. Valuables such as gun
metal bushes and expensive tools may be more tempting. While there is hardly any
possibility of heavy costing or forging being stolen. While the former must be kept
under lock and key, the latter may be simply dumped in the stockyard. It may be
consider as 1% of stock value.
5. Obsolescence Cost: It depends upon the nature of the item in stock. Changes in
design also lead to obsolescence. It may be taken as 5% of the stock value.
6. Handling Costs: Expenditure on stock holding is called handling costs. Such as
cost of labour, overhead cranes, gantries and other machinery used for this purpose.
[B.C.A. (Kashi) 2006, 2008; B.C.A. (Avadh) 2009]

7. Taxes and Insurance: Most organization have insurance cover and this may cost
1% to 2% of the invested capital.
 317

8.3.2 Ordering or Set-up Cost

[B.C.A. (Lucknow) 2010]
Ordering cost includes all cost that do not vary with the size of the order but are incurred
each time an order is placed for procuring items outside supplies. The cost per order
generally includes:
1. Requisition cost of handling of invoices, stationery, payments etc.
2. Cost of services which includes cost of mailing, telephone calls, transportation and
other follow up actions.
3. Materials handling cost in curred in receiving, sorting, inspecting and storing the
items included in the order.
4. Accounting and auditing, etc.

When an item is produced internally, ordering cost is referred as set-up cost which
includes both paper work costs and physical preparation costs.
Ordering or set-up cost is independent of the size of the order, if large number of orders
are placed, more money will required for procuring the items. Thus,
Ordering cost = (cost per order or per set-up) × (number of orders or set-ups places in the
planning period).

8.3.3 Shortage or Stock Out and Customer-service Cost

[B.C.A. (Rohilkhand) 2009]
The shortage of items occurs when items cannot be supplied on demand. Therefore,
shortage costs are usually interpreted in two ways:
1. The supply of items is awaited by the customers i.e., the items are back ordered.
2. Customers are not ready to wait.
This situation may lead to less of customer goodwill and therefore causes loss of sale.
Therefore,
Shortage cost = (cost of being short one unit of an item) × (average number of units cost)
The average number of units short in a planning period is obtain by average number of
units short
 minimum shortage + maximum shortage
=  × period of shortage
 2 

8.3.4 Total Inventory Cost

If unit of an item depends on the quantity purchase, then we should formulate an
inventory policy which takes into consideration the purchase cost of the items held in
stock also.
∴ Total inventory cost = purchase cost + ordering cost + carrying cost + shortage cost.
When price discounts are not offered, the purchase cost remains constant and is
independent of the quality purchased. Then,
Total variable inventory cost = ordering cost + carrying cost + shortage cost.
 318

8.4 Replenishment Lead Time

1. Order Cycle: The order cycle is the time period between two successive
replenishments which can be determined in one of the two ways.
(i) Continuous review

(ii) Periodic review

2. Lead Time or Delivery Lag: When an order is placed, it may require sometime
before delivery of the items ordered is reached. The time delay between placing an
order and receipt of delivery is called delivery lag or lead time. This time may be
deterministic or probabilistic.
3. Stock Replenishment: An inventory may operate with lead time, actual
replenishment of stock may occur instantaneously or gradual, the instantaneous
replenishment is possible when the stock is purchased from outside sources, while
gradual replenishment is possible due to finite production rate within the firm.

8.5 Inventory Control Problem

The inventory control problem consists of determination of two basic factors.
1. When to Order: This is related to the lead time of an item.
There should be sufficient stock for each item so that customers, order can be
reasonably met from this stock until replenishment. This stock level known as
reorder level. It is obtain by compromising the cost of maintaining these stocks and
the dis-service to the customer if his orders are not filled in time.
2. How must to Order: We know each order is related with its the ordering cost. To
maintain it low, the number of orders should be as few as possible. But large order
size would imply high inventory carrying cost. Thus over problem is determine how
must order is solved by compromising between the acquisition costs and inventory
carrying cost.

8.6 Concept of Economic Ordering Quantity (E.O.Q.)

The concept is that management is confronted with a set of opposing costs as the lot size
increases, the carrying charges will increase while the ordering cost will decrease or we
can say that as the lot size decreases the carrying cost will decrease but the ordering costs
will increase. Thus economic ordering quantity (E.O.Q.) is that size of order which
minimizes total annual cost of carrying inventory and cost of ordering under the assumed
conditions of certainly and that annual demands are known.
 319

The E.O.Q. concept is applicable under the following conditions:

1. The item is replenished in lots or batches, either by purchasing or by
manufacturing.
2. Consumption of items is uniform or continuous. E.O.Q. is that order quantity or
optimal order size which minimizes the total cost.
The model is describe under the following conditions:
(i) Planning period is one year.
(ii) Demand is deterministic.
(iii) Cost of purchases or of one unit is C.
(iv) Cost of ordering is C3 and or C0 . For manufacturing goods it is known as
set-up cost.
(v) The cost of holding or inventory carrying cost is C1 or Ch per unit per year
expressed.
(vi) Shortage cost or backorder cost is C2 or Cs per unit per year.
(vii) Leading time is L, expressed in unit time.
(viii) Cycle period is replenishment is t.
(ix) Order size is Q.

8.7 List of Symbols Used

The following symbols generally used in inventory theory.
C = purchase or manufacturing cost of an item (` per unit)
C0 = ordering or set-up cost per order (` per order)
Cs = shortage cost per unit per time (` per unit time)
Ch = C. r = cost of carrying one unit of an item in the inventory for a given length of time
(` per item per unit time)
D = annual demand (requirement) of an item (units per unit time)
Q = order quantity i.e., number of units ordered per order
R = demand rate
ROL = reorder level
LT = replenishment lead time
n = number of orders per time period
t = reorder cycle time
tp = production period
rp = production rate = K
Tc = total inventory cost (`)
 320

TVC = total variable inventory cost (`)

L = lead time
q = quantity already present in the beginning
z = order level or stock level
t = scheduling time period which is variable.

Note:
Sometime the cost Ch, Cs and C0 also denoted by C1, C2 and C3 .

8.8 Deterministic Models

8.8.1 Single Item Inventory Control Models without Shortages

Model (I): The economic lot size system with uniform demand.

Theorem:
Consider a period T during which R units of an item are required. The assumptions for
this model are as follows:
1. Demand is deterministic and is at uniform rates r units of quantity, per unit of
time.
2. The inventory is replenished as soon as the level of the inventory reaches to zero.
Thus the shortages are not allowed, in other word, we can say that shortage cost is
infinite so that it can not be considered in the cost of minimization problem.
3. The rate of replenishment of inventory is infinite i.e., the production or supply of
items to the inventory is instantaneous or say items are procured in one lot.
4. Lead time is zero.
5. The lot size is same for each cycle q i.e., 0 < q < ∞.
6. Set-up cost is C3 per cycle.
7. Carrying cost is C1 per unit of quantity per unit of time.
[B.C.A. (Lucknow) 2004, 2010; B.C.A. (Bhopal) 2009]
[B.C.A. (I.G.N.O.U.) 2007, 2012]

Proof: We have to find optimal value of q which is most economic i.e., gives the
minimum total cost.

Let after every time t, the quantity q is produced or ordered throughout the period T. Let
q0 and t0 be optimal value of q and t respectively which minimize the total cost per unit of
time. In this model, we consider only C1 and C3 costs.
The figure 8.1 shows the situation of inventory in the time period T.

This type of figure is known as time-inventory graph and is very much helpful in
understanding the inventory situation at various times.
 321

O t A t time
t

Fig. 8.1

1
The total inventory in one cycle i.e., for t units of time = area of the ∆ OAB = qt.
2
1
∵ The carrying cost for t units of time = q t C1.
2
The set-up cost for one cycle i.e., for t units of time = C3
1
∴ The total cost in t units of time = q C1 + C3 .
2
1 
∴ The total cost for one unit of time = C (q) =  q C1 + C3  / t
2 
1 C
q C1 + 3 .
=
2 t
q
But q = r t or t =
r
1 r
Then C (q) = q C1 + C3 ...(1)
2 q

This C (q) is called the cost equation.

d C (q)
This cost must be minimized, then =0
dt
dC (q) 1 C r
∴ = C1 − 32
dq 2 q
1 C r
⇒ 0= C1 − 32
2 q

3 C3 r
or q0 = ...(2)
C1
d2 C (q) 2 C3 r
Again 2
=
dq q3
 d2 C (q) 2 C3 r
At q = q0 , we have  2 
 = >0
 dq  q q 03
0
 322

⇒ q = q0 minimizes the cost C (q).

Equation (2) is known as Harri's or Wilson's economic lot size formula or simply
economic lot-size formula and q0 is also taking as E.O.Q.
q 2 C3
Again t0 = 0 = ...(3)
r C1 r

1 2 C3 r C1
Also minimum cost = C0 (q0 ) = C1 + r C3
2 C1 2 C3 r

1 1
= 2 C1C3 r + 2 C1C3 r = 2 C1C3 r ...(4)
2 2
Remark: 1. Total inventory in one cycle i.e., for t units of time
1
= q t.
2
1 1
The average inventory at any time = qt / t = q .
2 2
Therefore throughout the period T, average level of inventory is q / 2.
2. The optimum number of orders to be placed by the formula
total annual quantity requirement
N =r /q= .
economic ordering quantity

3. The number of days, supply per optimum is obtain by formula d = 365 / N .

S olved E xamples
Example 1: A particular item has a demand of 9,000 units per year. The cost of one procurement
is ` 100 and the holding cost per unit is ` 2.40 per year. The replacement is instantaneous and no
shortages are allowed. Determine
(i) The economic lot size
(ii) The number of orders per year
(iii) The time between orders
(iv) The total cost per year if the cost of one unit is ` 1.
[B.C.A. (I.G.N.O.U.) 2006; B.C.A. (Kanpur) 2007]

Solution: Here r = 9,000 units / year

C3 = ` 100 / procurement
C1 = ` 2 .40 / unit / year

2 C3 r 2 × 100 × 9,000
(i) q0 = = = 866 units /procurment
C1 2 .40
 323

(ii) The number of orders per year

1 C1 r
= N0 = =
t0 2 C3

2 .40 × 9,000
= = 108 = 10 .4 orders / year
2 × 100

1 1
(iii) The time between orders = t0 = = = 0 .0962 years between procurement.
N0 10 .4

(iv) The total cost per year

= 9, 000 + 2 C1C3 r

= 9, 000 + 2 × 2 .40 × 100 × 9,000

= 9, 000 + 2 , 080 = ` 11,080 / year.

Example 2: A precision engineering factory consumes 50,000 units of a component per year. The
ordering, receiving and handling costs are ` 3 per order while the trucking costs are ` 12 per order.
Further details are as follows:
Interest cost ` 0.06 per unit per year.
Deterioration and obsolescence cost ` 0.004 per unit per year.
Storage cost ` 1,000 per year for 5,000 units; calculate the economic order quantity.
[B.C.A. (Indore) 2012]
Solution: Here,
C3 = ` 3 + ` 12 = ` 15 / order
r = 50, 000
1000
C1 = ` 0 .06 + ` 0 .004 + ` per unit
50, 000

= ` 0 .084 / unit
2 C3 r 2 × 50, 000 × 15
∴ E.O.Q. = = = 4226 units.
C1 0 .084

Example 3: A stockist has to supply 400 units of a product every Monday to his customers. He gets
the product at ` 50 per unit from the manufacturer. The cost of ordering and transportation from the
manufacturer is ` 75 per order. The cost of carrying inventory is 7.5% per year of the cost of the
product. Find
(i) The economic lot size
(ii) The total optimal cost (including the capital cost). [B.C.A. (Avadh) 2003]

Solution: Here
r = 400 units / week
C3 = ` 75 / per order
C1 = 7 .5% / per year of the cost of the product
 324

 7 .5 
=` × 50 per unit per year
100 

 7 .5 50 
=` ×  per unit per week
100 52 

3 .75
=` per unit per week
52
2 C3 r
(i) The economic lot size q0 =
C1

2 × 75 × 400 × 52
= = 912 units /order
3 .75

(ii) The total capital cost = 400 × 50 + 2 C1 C3 r

2 × 3 .75
= 20, 000 + × 75 × 400
52
= 20, 000 + 65 .80 = ` 20, 065 .80 per week.

Example 4: A manufacturer has to supply 12,000 units of a product per year to his customer. The
demand is fixed and known and shortage cost is assumed is to be infinite. The inventory holding cost
is ` 0 .20 per unit per month and the set-up cost per run is ` 350. Determine
(i) The optimum run size q0
(ii) Optimum scheduling period t0
(iii) Minimum total variable yearly cost.

Solution: Here
12, 000
Supply rate r = = 1, 000 units / month
12

C1 = ` 0 .20 per unit per month

C3 = ` 350 per run

(i) 2 C3 r 2 × 350 × 1, 000

q0 = = = 1, 870 units /run
C1 0 .20

(ii) 2 C3 2 × 350
t0 = = = 1. 870 units /run
C1 r 0 .20 × 1, 000

= 8 .1 weeks between runs.

(iii) Minimum total variable yearly cost C0 = 2 C1 C3 r

= 2 × 0 .20 × 12 × 350 × (1, 000 × 12)

= ` 4, 490 per year

 325

Example 5: (i) Calculate the E.O.Q. in units and total variable cost for the following items,
assuming an ordering cost of ` 5 and a holding cost of 10%.

Item Annual Demand Unit Price (`)

A 800 units 0.02

B 400 units 1.00

C 392 units 8.00

D 13, 800 units 0.20

(ii) For the above problem, compute E.O.Q. in `, as well as in years of supply. Also calculate the
E.O.Q. frequency for each of the four items.

Solution: (i) For item A

2 C3 r 2 × 5 × 800 8,000
The E.O.Q. (q0 ) = = = = 2,000 units
C2 10 0 .002
0 .02 ×
100
Total variable cost (C0 ) = 2 C1 C3 r

10
= 2 × 5 × 800 × 0 .02 × =`4
100
For item B
2 × 5 × 400
q0 = = 200 units
10
1.00 ×
100
10
C0 = 2 × 5 × 400 × 1.00 × = ` 20
100
For item C
2 × 5 × 392
q0 = = 70 units
10
8 .00 ×
100
10
C0 = 2 × 5 × 392 × 8 .00 × = ` 56
100

For item D
2 × 5 × 13, 800
q0 = = 2, 627 units
10
0 .20 ×
100
10
C0 = 2 × 5 × 13, 800 × 0 .20 × = ` 52 .54
100
 326

(ii) E.O.Q. in `
For item A = 2, 000 × 0 .02 = 40
For item B = 200 × 1 = 200
For item C = 70 × 8 = 560
For item D = 2, 627 × 0 .20 = 525 .40

and E.O.Q. in the years of supply

2000
For item A = = 2 .5 years
800
200
For item B = = 0 .5 year
400
70
For item C = = 0 .18 year
392
2627
For item D = = 0 .19 year
13, 800
E.O.Q. frequency (number of orders per year)
For item A = 1 / 2 .5 = 0 .4
For item B = 1 / 0 .5 = 2
For item C = 1 / 0 .18 = 5 .6
and For item D = 1 / 0 .19 = 5 .25

8.9 Economic Lot Size with Different Rates of Demand in

Different Cycles
In that case all assumptions are same as in first model only change that the demand is
uniform with different rates in different cycles.

Let t1, t2 ,..., tn are the time of successive cycles and r1, r2 ,..., rn are the demand rates in
these cycles respectively, then
t1 + t2 + ... + tn = T
r2

rn
r1

t1 t2 tn
T
Fig. 8.2
 327

Now, we have to minimize the total cost for time period T.

The inventory carrying cost for the time period T

1 1 1
= C1 q t1 + C1 q t2 + ... + C1 q tn
2 2 2

1 1
= C1 q (t1 + t2 + ... + tn) = q C1 T
2 2
r
Since each time q units are produced, the number of cycles in the total period T is , r be
q
the demand of period T.
C r
Therefore the set-up cost for time T = 3
q

Now, the total cost for the time period T is given by

1 r
C (q) = C1 q T + C3
2 q

d C (q) 1 C r
⇒ = C1 T − 32 ...(1)
dt 2 q

d C(q) 2 C3 r
For minimum cost put = 0 we get q0 =
dt C1 T

This is optimal value of q.

2 C1 C3 r
The minimum cost = C0 (q0 ) = ...(2)
T

8.9.1 Economic Lot Size with Finite Rate of Replenishment

or
Production Lot Size Model (PLS)
In both models, we consider replenishment time is zero i.e., item order is produced in one
lot. This model is similar to first model except that the rate of replenishment of inventory
is finite say k unit of quantity per unit of time. We have to obtain an economic lot size q0
and this model having C1 and C3 costs.

Each cycle consists of two parts as shown in figure 8.3. In the first part of time i.e., t1 the
inventory is building up i.e., items are assembled at constant rate of (k − r ) units per unit
of time and in the second part t2 , there is only withdrawl of items from the inventory at a
constant rate of r units per unit of time.
 328

Inventory
B

r
r
k–

O A
t1 t2 time
T

Fig. 8.3

Let at the end of first part of time t1, the level of inventory is S. Then
S = t1 (k − r ) and S = t2 r
S S
or t1 = and t2 = ...(1)
k−r r

If q is the lot size, then q − t1 r = S

S k−r
or q− r = S or S = q ...(2)
k−r k

The carrying cost for one cycle is

1 1  1
 St1 + St2  C1 = (t1 + t2 ) S C1
2 2  2

1 (t1 + t2 ) S C1 1
Therefore, carrying cost per unit of time = = S C1
2 t1 + t2 2
1 k−r
= q C1
2 k
The set-up cost for one cycle = C3
C3 C
∴ Set-up cost for one unit of time = =
t1 + t2 S
+
S
k−r r
C r (k − r ) r (k − r ) k C r
= 3. = C3 . = 3
S k k (k − r ) q q

Hence, the total cost for one unit of time is C (q) which is given by
1  k − r C3 r
C (q) =   C1 q +
2 k  q
d C (q)
Then for C (q) is minimum ⇒ =0
dq
 329

1  r C r
⇒ = 1 −  C1 − 32 = 0
2  k q

2 C3 r 1
or q = q0 = . ...(3)
C1 1 − r 
 
 k

This is the required optimal lot size.

q
Also optimum time for one cycle = t0 = 0
r
2 C3
= ...(4)
r C1 1 − 
r
 k

Putting this value of q, we can obtain the minimum value of cost C0 (q0 ).

Example 6: A company has a demand of 12,000 units/year for an item and it can produce 2,000
such items per month.

The cost of one set-up is ` 400 and the holding cost/unit/month is ` 0.15. Find the optimum lot size
and the total cost per year, assuming the cost of 1 unit at ` 4. Also find the maximum inventory
manufacturing time and total time. [B.C.A. (Purvanchal) 2009]

Solution: Here
r =12, 000 units / year
k = 2, 000 × 12 = 24,000 units / year
C3 = ` 400 / set - up
C1 = ` 0 .15 × 12 = ` 1.80 / unit / year

2 C3 k r
The optimum lot size q0 =
C1 k − r

2 × 400 12, 000 × 24, 000

= × = 3, 264 units / set - up
1.8 12, 000

 k − r
The total cost per year = C0 = 12, 000 × 4 + 2 C1 C3 r  
 k 

12, 000
= 48, 000 + 2 × 1.8 × 400 × 12, 000 ×
24, 000

= 48, 000 + 2, 940 = ` 50, 940 / year

k−r
The maximum inventory I m = q0
0 k
24, 000 − 12, 000
= × 3, 264
24, 000
 330

=1, 632 units

Im 1, 632
The manufacturing time t1 = 0
= = 0 .136 years
k−r 12, 000
q 3, 264
The total time t0 = 0 = = 0 .272 years
r 12, 000

Example 7: An item is produced at the rate of 50 items per day. The demand occurs at the rate of
25 items per day. If the set-up cost is ` 100.00 per set-up and holding cost is ` 0.01 per unit of item
per day, find the economic lot size for one run, assuming that the shortages are not permitted.

Solution: Here, we have

k = 50 / day , r = 25 / day
C1 = ` 0.01 / day and C3 = ` 100 / run

 
 2 × 100 × 25  2 × 100 × 25 × 50
∴ q0 =  = = 1000 items
 25 
0 .01 × 1 −   0 .01 × 25
  50  

1000 
∴ t0 =   = 40 days
 25 

The minimum cost = ` 5 per day

Therefore, the total cost for one run = ` 200 .00.

Example 8: A contractor has to supply 10, 000 paper cones per day to a textile unit. He finds
that, when he starts a production run he can produce 25000 paper cones per day. The cost of holding
a paper cone is stock for one year is 2 paise and the set-up cost of a production run is ` 18. How
frequently should production run be made ?
[B.C.A. (Rohilkhand) 2004, 2008; B.C.A. (Kurukshetra) 2007, 2008]

Solution: Here, we have

r =10, 000 / day
k = 25, 000 / day, C3 = 18
 0 .02 
C1 = ` 0 .02 / year = `   / day
 365 
2 C3 r 1
∴ The optimal lot size q0 =
C1 1 − r 
 
 k

2 × 10000 × 18 × 365 25000

= ×
0 .02 25000 − 10000

=104447 paper cones

q 104447
Also t0 = 0 = = 4 days (approximately).
k 25000
 331

8.9.2 Limitations of E.O.Q. Formula

In that model the demand is neither exactly known nor uniform. If the fluctuations are
mild, the formula serves the purpose while in case of high fluctuations it is not valid. If an
inventory level reaches to zero, the instantaneous supply of items is not possible. The
determination of ordering cost is difficult and also it is not linearly related to the number
of orders. It may not be independent of the order size. Inventory carry charge is also
difficult to determine. We generally use only a simple assumption of about 20% while in
practice it may vary from 5% to 35%.

Example 9: A contractor has to supply 10,000 bearings per day to an automobile manufacturer.
He finds that, when he starts a production run, he can produce 25,000 bearings per day. The cost of
holding a bearing in stock for one year is 20 paise and set-up cost of a production run is ` 180.00.
How frequently should production run be made.

Solution: We have the optimum lot size

2 C3 r k 2 C3 rk
q0 = = ...(1)
C1 (k − r ) C1 k−r

2 C3 k 2 C3 k
and t0 = = ...(2)
r C1 (k − r ) r C1 k−r

Here, we have

C1 = ` 0 .20 per bearing per year

 0 .20 
=`  per bearing per day
 365 

= ` 0 .00055

C3 = ` 180 .00 per production run

r =10, 000 bearings per year

k = 25, 000 bearings per day

Put these values in (1) and (2), we find

2 × 180 × 10, 000 × 25, 000

q0 = = 1.09 × 1010 = 1, 05, 000 bearings
0 .00055 × (25, 000 − 10, 000)

2 × 180 × 25, 000

t0 = = 0 .3 day.
10, 000 × 0 .00055 × 15, 000
 332

8.10 IInd Model (The E.O.Q. Model with Shortage)

This model is the extension of model I allowing shortages. We discuss this model in three
form.

8.10.1 I st Form Single Item Inventory Control Models with

Shortages
1. C1 is holding cost per quantity unit per unit time.
2. C2 is the shortage cost per quantity unit per unit time.
3. r quantity per unit time is the demand rate.
4. tp is the scheduling time period which is constant.
5. q p is the fixed lot size (qp = r tp ).
6. z is the order level to which the inventory is raised in the beginning of each
scheduling period. Shortages, if any, have to be made-up. Here z is the variable.
7. Production rate is infinte.
8. Load time is zero.
Then obtain optimal ordering level and minimum average cost. [B.C.A. (Kanpur) 2010]

Proof: When 0 ≤ z ≤ qp , then there exist inventory carrying cost C1 and shortage cost C2 .

Inventory

Stock
qp (tp – z/r)

O Time
B Shortage D
(z/r)

(qp – z)

C
tp
tp

Fig. 8.4
 333

The dotted area (∆ DBC) represents the failure to meet the demand and shady area
(∆ AOB) show the inventory. Since qp is the lot size required to meet the demand for time
tp , but qp amount of stock is planned in order to meet the demand for time(z / r ).

Shortage of amount (qp − z) will arise for the entire remaining period (tp − z / r ).
C (∆ OAB)
The holding cost per unit time = 1
tp
2
C 1 z  1 z C1
= 1  ×z×  = (∵ qp = r tp ) ...(1)
tp  2 r  2 qp
C2 (∆ BDC)
and shortage cost per unit time =
tp

1 
C2  × BD × DC
2  C2  z
=  tp −  (qp − z)
tp 2 tp  r

C2 1 C2
= (r tp − z) (qp − z) = (qp − z)2 ...(2)
2 r tp 2 qp

Total average cost C (z)

2
1 z C1 1 C2
C ( z) = + (qp − z)2 ...(3)
2 qp 2 qp

This is called cost equation.

Now to find the optimum level z for this put

dC (z)
=0
dz
dC (z) 1 C1 1 C2
∴ = (2 z) + 2 (qp − z) (− 1)
dz 2 qp 2 qp
1 C1 1 C2
⇒ 0= (2 z) + 2 (qp − z) (−1)
2 qp 2 qp
C2 C2
⇒ z= qp = r tp ...(4)
C1 + C2 C1 + C2

d2 C (z) C C
Again = 1 + 2 >0
dz2 qp qp

i.e., C (z) is minimum, which is obtain by putting z from (4) in (3)

2 2
1 C1  C2 qp  1 C2  C2 qp 
Cmin =   +  qp − 
2 qp  C1 + C2  2 qp  C1 + C2 

1 C1 C2 C1 C2
or Cmin = . qp = r tp ...(5)
2 C1 + C2 2 (C1 + C2 )
 334

8.10.2 II nd Form the E.O.Q. with Constant Rate of Demand

Scheduling Time Variable
In a certain manufacturing situation:
1. R is demand per year.
2. The production is instantaneous.
3. q = r t is the order quantity per run.
4. t is the scheduling time period which is variable.
5. z is the order level to which the inventory is raised.
6. Lead time is zero.

Show that the optimal order quantity q per-run which minimizes the total cost is

2 r C3 (C1 + C2 )
q=
C1 C2

where C1 = holding cost per unit per year

C2 = shortage cost per unit per year
C3 = set-up cost per run. [B.C.A. (Agra) 2004, 2008]

Proof: This form is same as first form only change with difference that the scheduling
period t is not constant here, now to consider the average set-up cost (C3 / t) in the cost
equation.

Thus the cost equation in this form becomes

1  C z 
2
1 C2
C ( t, z ) =  1 + (r t − z)2 + C3  ...(1)
t  2 r 2 r 

∂C ∂C
This cost must be minimized so that = 0, =0
∂z ∂t
∂C 1 C z C 
∴ =  1 − 2 (rt − z) = 0
∂z t r r 
C2 rt
⇒ z= ...(2)
C1 + C2

∂C 1  C1 z2 C 
Also =− 2  + 2 (rt − z)2 + C3 
∂t t  2 r 2r 
1 C2 
+ 0 + 2(rt − z) r + 0  = 0
t  2r 
1  C z  C
2
C
⇒ − 2  1 + 2 (rt − z)2 + C3  + 2 (rt − z) = 0
t  2 r 2r  t
 335

⇒ − (C1 + C2 ) z2 + C2 r 2 t2 = 2 r C3 ...(3)

Put z from (2) in (3), we get

r 2 t2 C22
− + C2 r 2 t2 = 2 r C3
C1 + C2

 C2 
⇒ C2 r 2 t2 1 −  = 2 r C3
 C1 + C2 
2 C3 (C1 + C2 )
or t= ...(4)
r C1 C2
2
∂2 C ∂2 C  ∂2 C 
Also . −  >0
∂t2 ∂z2  ∂t ∂z 

∂2 C ∂2 C
and > 0, >0
∂ t2 ∂z2

Optimal order quantity q is given by

2 C3 (C1 + C2 )
q0 = r t0 = r
r C1 C2

2 r C3 (C1 + C2 )
or q0 = ...(5)
C1 C2

This is E.O.Q in this form.

For minimum cost put z from (2) in (1), we have

2
C1 C2 (rt)2 C  C2 rt  C
Cmin = . 2 + 2  rt −  + 3
2 r t (C1 + C2 )2 2 rt  C1 + C2  t

C1 C22 rt C2 (rt) C3
= 2
+ (C1 + C2 − C2 )2 +
2 (C1 + C2 ) 2 (C1 + C2 )2 t

C1 C2 r t C
Cmin = + 3 ...(6)
2 (C1 + C2 ) t

Put t from (4), we get

C1 C2 r 2 C3 (C1 + C2 ) r C1 C2
Cmin = . + C3 .
2 (C1 + C2 ) r C1 C2 2 C3 (C1 + C2 )

C1 C2 C3 r C1 C2 C3 r 2 C1 C2 C3 r
= + =
2 (C1 + C2 ) 2 (C1 + C2 ) (C1 + C2 )

C2
or Cmin = 2 C1 C3 r ...(7)
C1 + C2
 336

8.10.3 III rd Form the Production Lot Size Model with Shortages
To determine an economic lot-size formula for optimum production quantity q per cycle
of a single product so as to minimize the total average variable cost per unit time, where
1. r units per unit time is the uniform demand rate.
2. Lead time is zero.
3. Production rate is finite (k units per unit time), k > r .
4. Inventory carrying cost in ` C1 = IP per quantity per unit time.
5. Shortages are allowed and back logged.
6. Shortage cost in ` C2 per quantity per unit time.
7. Set-up cost in ` C3 per set-up.

Proof: This figure 8.5 shows that there is an inventory cycle. Stock short at zero and
increase for a period t1 and decrease for period t2 until they again reach zero at the point
where a back-log piles up for the time t3 at the end of time t3 production starts and
backlog is diminished for the time t4 when the backlog reaches zero.
Inventory

Q1 t3 t4
C D Time
O B E
t1 t2

Q2

F
Fig. 8.5

This cycle repeated after time period (t1 + t2 + t3 + t4 ).

Then,
Holding cost = C1 × area of triangle OAC
1
= C1 × Q1 (t1 + t2 )
2
Shortage cost = C2 × area of triangle EFC
1
= C2 × Q 2 (t3 + t4 )
2
Set-up cost = C3
 337

1
[C1 Q1 (t1 + t2 ) + C2 Q2 (t3 + t4 ) + C3 ]
Thus; average cost C= 2 ...(1)
t1 + t2 + t3 + t4

where C is the function of six variables (Q1, Q2 , t1, t2 , t3 , t4 ), but there are four
relationships.

From given we see that the inventory is zero at O and during the period t1 an account k t1
is produced but order are being filled up at a rate r, the net increase Q1 in inventory
during t1 is given by
Q1 = k t1 − r t1 = t1 (k − r ) ...(2)

After time t1, the production is stopped and the stock Q1 is used up during t1 and the
rate is r, then
Q1 = r t2 ...(3)
Q rt
From (2) and (3), we have t1 = 1 = 2 ...(4)
k−r k−r

During period t3 shortages accumulate at a rate r

Q2 = r t3 ...(5)

During period t4 , production rate is k and demand rate is r, so that the net rate of
reduction of shortage becomes k − r and thus, we have
Q2 = t4 (k − r ) ...(6)
Q rt
From (5) and (6), we have t4 = 2 = 3 ...(7)
k−r k−r

∴ q = r (t1 + t2 + t3 + t4 ) ...(8)

Put t1 and t4 from (4) and (7) in (8), we get

rt rt 
q = r  2 + (t2 + t3 ) + 3 
k−r k − r

(t + t ) k
or q= 2 3 ...(9)
k−r

Put t1, t4 , Q1, Q2 in (1), we get

1   r t2   rt  
C1 (rt2 )  + t2  + C2 r t3  t3 + 3  + C3 
2 k−r   k − r 
C= 
r t3 r t3
+ t2 + t3 +
k−r k−r

1  C1 t2 rk C2 t3 k r 
2 2
 +  + C3
2  k − r k − r 
C=
 r 
(t2 + t3 ) 1 + 
 k − r
 338

1 kr
{C1 t22 + C2 t32 } + C3
2 k−r
C=
 k 
(t2 + t3 )  
 k − r
1
[C1 t22 + C2 t32 ] k r + C3 (k − r )
= 2
k (t2 + t3 )

1
[C1 t22 + C2 t32 ] k r + C3 (k − r )
∴ C (t2 , t3 ) = 2 ...(10)
k (t2 + t3 )

To find best values of t2* and t3* of t2 and t3 for this put
∂C ∂C
=0 and = 0, we get
∂ t2 ∂ t3

2 C3 C2 1 − 
r
 k
t2* = ...(11)
r (C1 + C2 ) C1

2 C3 C1 1 − 
r
 k
t3* = ...(12)
r (C1 + C2 ) C2

Then put t2 = t2* , t3 = t3* in (9), we get

2 r C3 (C1 + C2 ) 1
q* =
C1 C2 1− r /k

 k
2 r C1 C3 1 − 
 r
= ...(13)
(C1 + C2 ) C2

Now, the minimum cost from (1), we get

2 r C1 C2 C3 1 − 
r
 k
C* = ...(14)
C1 + C2

when k → ∞ then (10) becomes

1 
C (t2 , t3 ) =  r (C1 t22 + C2 t32 ) + C3 
2  / (t
2 + t3 )

2 r C3 (C1 + C2 ) 2 r C1 C3
and q* = , Q* = ,
C1C2 (C1 + C2 ) C2

2 r C1 C2 C3
C* =
(C1 + C2 )
 339

Example 10: A sub-contractor undertakes a supply diesel engines to a truck manufacturer at the
rate of 25 per day. There is a clause in contract penalizing him ` 10 per engine per day late for
missing the scheduled delivery date. He finds that the cost of holding a completed engine in stock is
` 16 per month. This production process is such that each month (30 days) he starts a batch of
engines through the shops and all these engines are available for delivery any time after the end of
the month. What should his inventory level at the beginning of each month (i.e., immediately after
taking in) to stock the engines made in the previous month and then shipping engines to fill
unsatisfied demand from previous month ? [B.B.A. (Meerut) 2010]

Solution: Here, r = 25 engines per day, C1 = ` (16 / 30)

C2 = ` 10, tp = 30 days, then inventory level (z) is given by

C2 10
z= r tp = × 25 × 30 = 712 engines
C1 + C2  16 
10 +  
 30 

Example 11: The demand of an item is uniform at a rates of 25 units per month. The fixed cost is
` 15 each time a production run is made. The production cost is ` 1 per item and the inventory
carrying cost is ` 0.30 per item per month. If the shortage cost is `1.50 per item per month,
determine how often to make a production run and of what size it should be ?
[B.C.A. (Bundelkhand) 2007]
Solution: Here, we have
C1 = ` 0 .30 per item per month
C2 = ` 1.5 per item per month
C3 = ` 15 per production run
r = 25 units per month
b = `1 per item
Then optimum values of q and t are obtained by

2 r C3 (C1 + C2 ) 2 × 15 × 25 (0 .30 + 1.50)

q * or q0 = = = 54 items.
C1 C2 0 .30 × 1.50

q 54
and t * or t0 = 0 = months = 2 .16 months .
r 25

Example 12: The demand for an item is deterministic and constant over time and it is equal to
600 units per year. The unit cost of the item is ` 50 while the cost of placing an order is ` 5. The
inventory carrying cost is 20% of the cost of inventory per annum and the cost of holding is ` 1 per
month. Find the optimal ordering quantity when stock outs are permitted. If the stock outs are not
permitted, what would be the loss to the company.

Solution: We have, r = 600 units, C1 = 0 .20 × 50 = ` 10, C3 = ` 5 per order

and C2 = ` 1 per unit / month or `12 per year / unit

 340

When stock outs are permitted, the optimal ordering quantity is

2 r C3  (C1 + C2 ) 2 × 600 × 5 10 + 12 

q* =   =   = 33 units
C1  C2  10  12 

 C1   10 
Maximum number of back orders z * = q *   = 33   = 15 units
 1
C + C2 10 + 12 

Total expected yearly cost without shortages is given by

 C2 
C (q*) = 2 r C1C3  
 C1 + C2 

 12 
= 2 × 600 × 10 × 5   = 180 .91 = 181
10 + 12 

If stock outs or back-orderings are not permitted, the optimal order quantity is

2 r C3 2 × 600 × 5
q1* = = = 24 .5 units
C1 10

The total relevant cost is given by

C (q1*) = 2 C1C3 r = 2 × 10 × 5 × 600 = 60,000 = ` 245

Thus, the additional cost when back-ordering is not allowed

= ` (245 − 181) = ` 64.

Example 13: The demand for an item in a company is 18,000 units per year and the company
can produce the item at a rate of 3,000 per month is 15 paise. The shortage cost of one unit is ` 20
per year. Determine the optimum manufacturing quantity and the number of shortages. Also,
determine the manufacturing time and the time between set-ups. [B.C.A. (Kurukshetra) 2012]

Solution: Here, we have C1 = ` 0 .15, C2 = ` 20, C3 = ` 500

k = 3, 000 units /month, r = 18, 000 /12 units /month = 15, 00 units /month

∴ The optimal manufacturing quantity

2 C3 r (C1 + C2 ) k
q* = .
C1 C2 k−r

2 × 500 × 1500 × (0 .15 × 12 + 20) 3000

= . =1344 units
(12 × 0 .15) × 20 3000 − 1500

The number of shortages are given by

 0 .15 × 12   1500 
q * 1 −  = 
C1 r
S=  × 1344 1 − 
C1 + C2  k   0 .15 × 12 + 20   3000 
 341

1.8
= × 1344 × 0 .5 = 55.486 units
21.8
q* 1344
Manufacturing time = = = .037 year
k 3000 × 12
q* 1344
Time between set-ups = = = .074 year.
r 18, 000

P roblem S et
1. Find the E.O.Q. for the following data:

Annual usage =1, 000 pieces Expediting cost = ` 4 per order

Cost per piece = ` 250 Inventory holding cost =20% of average inventory

Ordering cost = ` 6 per order Material holding cost = `1 per piece

2. You have to supply your customers 100 units of a certain product every Monday
(and only then). You obtain the product from a local supplier at ` 60 per unit. The
costs of ordering and transportation from the supplier are `150 per order. The cost
of carrying inventory is estimated at 15% per year of the cost of the product carried.
(i) Find the lot size which will minimize the cost of the system.
(ii) Determine the optimal cost. [B.C.A. (I.G.N.O.U.) 2004, 2010]
3. An aircraft company uses rivets at an approximate customer rate of 2, 500 kg per
year. Each unit cost ` 30 per kg and the company personnel estimate that it costs
`130 to place an order and that the carrying cost of inventory is 10% per year. How
frequently should orders for rivets be placed ? Also determine the optimal size of
each order.
4. An aircraft uses rivets at an approximately constant rate of 5,000 kg per year. The
rivets cost ` 20 per kg and the company personnel estimate that it costs ` 200 to
place an order and the carrying cost of inventory is 10% per year.
(i) How frequently should orders for rivets be placed and what quantities should
be order for ?
(ii) If the actual costs are ` 500 to place an order and 15% for carrying cost, the
optimal policy would change. How must is the company losing per year
because of imperfect cost information ?
5. A contractor has to supply 10,000 bearings per day to an automobile manufacturer.
He finds that, when he starts production run, he can produce 25,000 bearings per
day. The cost of holding a bearing in stock for a year is ` 2 and the set-up cost of a
production run is ` 1,800. How frequently should production run be made ?
 342

6. A manufacturing company uses an E.O.Q. (Economic Order Quantity) approach in

planning its production of gears. The following information is available. Each gear
cost ` 250 per unit, annual demand is 60,000 gears, set-up cost are ` 4,000 per
set-up and the inventory carrying cost per month is established at 2% of the average
inventory value. When in production, these gears can be produced at the rate of
400 units per day and this company works only for 300 days in a year. Determine
the economic lot size, the number of production runs per year and the total
inventory costs. [B.C.A. (Delhi) 2012]

7. A dealer supplies you the following information with regard to a product dealt in by
him: Annual demand : 10,000 unit : ordering cost : ` 10 per order price ` 20 per
unit. Inventory carrying cost : 20% of the value of inventory per year. The dealer is
considering the possibility of allowing some back order to occur. He has estimated
that the annual cost of back ordering will be 25% of the value of inventory.
(i) What should be the optimum number of units of the product he should busy
in one let ?
(ii) What quantity of the product should be allowed to be back ordered, if any ?
(iii) What would be the maximum quantity of inventory at any time of the year ?
(iv) Would you recommend to allow back-ordering ? If so, what would be the
annual cost saving by adopting the policy of back-ordering ?

8. The annual demand for a product is 64,000 units or 1280 units per week. The
buying cost per order is ` 10 and the estimated cost of carrying one unit in the stock
for a year is 20%. The normal price of the product is ` 10 per unit. However, the
supplier offers a quantity discount of 2% on an order of at least 1000 units of a
time and a discount of 3% if the order is for at least 5,000 units.

9. An oil engine manufacturer purchases lubricants at the rate of ` 42 per piece from a
vendor, the requirements of these lubricants is 1,800 per year. What should be the
order quantity per order, if the cost per placement of an order is ` 16 and inventory
carrying charges per rupee per year is only 20 paise. [B.C.A. (Bhopal) 2008]

10. A company uses 24,000 units of a raw material which costs ` 12.5 per unit.
Planning each order costs ` 22.5 and the carrying cost is 5.4 percent per year of the
average inventory. Find the economic quantity and the total inventory cost
(including the cost of material).

11. A manufacturer has to supply his customers 24,000 units of his product per year.
The demand is fixed and known. The customer has no storage space and so the
manufacturer has to ship a day’s supply each day. If the manufacturer fails to
supply the penalty is ` 0 .20 per unit per month. The inventory holding cost
amounts to ` 0 . 10 per unit per month and the set-up cost is ` 350 per production
run. Find the optimum lot size for the manufacturer.
 343

12. A contractor undertakes to supply diesel engines to a truck manufacturer at rate of

25 per day. He finds that the cost of holding a completed engine in stock is `16 per
month, and there is a clause in the contract penalizing him `10 per engine per day
late for missing the scheduled delivery date. Production of engines is in batches,
and each time a new batch is started there are set-up costs of `10,000. How
frequently should batches be started and what should be the initial inventory level
at the time each batch is completed. [B.C.A. (Kashi) 2010]

13. The demand for an item is 18,000 units per year. The holding cost is `1.20 per unit
time and the cost of shortage is ` 5 .00, the production cost is ` 400. Assume that
replenishment rate is instantaneous, determine the optimal order quantity.

14. What are assumptions of economic lot size formula ? [B.B.A. (Meerut) 2011]

15. Write the note on economic lot size problem with uniform rate of demand.

16. Derive economic order quantity model for an inventory problem when shortages of
costs are not allowed.

17. Describe the single item production inventory model with no shortages and derive
the formula for optimum lot size for one run and the optimum time between two
runs.

18. Derive an economic lot size with different rates of demand in different cycles.
[B.C.A. (Rohilkhand) 2007]

19. Derive the optimal economic lot size formula

2 C3 r k
q= in the usual notations when the rate of replenishment is
C1 (k − r )
1 /2
 
finite. Also, derive the minimum cost formula Cmin = 2 C1 1 −  C3 r 
r
.
  k  

20. Discuss various inventory costs. Derive economic lot size formula when shortage
costs are allowed.
 344

A nswers
1. 20 [Hint, D = 1000
, , C3 = 6 + 4 = 10, 2. (i) 416 units; (ii) ` 6,072

C1 = C × I = 250 × 0 .20 = 50

2 × 10 × 1000
EOQ = = 20]
250 × 0 .20

3. 466 units, 2.16 months, n = 5 orders 4. (i) ` 2,000; (ii) ` 1,873

per year

5. 1,04,446 bearings, 10.44 days 6. Economic lot size = 4,000 gears,

Number of production = 15

Total inventory cost = ` 1,20,000

7. (i) 223.6 units, 300 units 9. q * = 83 lubricants

(ii) 133 units

(iii) 167 units

(iv) T.C. (223.6) = ` 894.43,

T.C. (300) = ` 666.67

10. q * = 4,000 units, cost = ` 30,270 11. 4,744 units per run

12. q * = 943 engines, t * = q */ 38 days 13. (i) q * = 3,600

(ii) 50 orders per year

(iii) C min = ` 6,000

mmm
 345

C HAPTER
9

Job Sequencing

9.1 Sequencing Problem

[B.C.A.(Agra) 2010]
The selection of an appropriate order for a series of jobs to be done on a finite number of a
service facilities is called sequencing. In a sequencing problem, we have to determine the
order (sequence) of performing the jobs in such a way that the total involved cost may be
minimum. A general sequencing problem may be defined as follows. Let there be n-jobs
(1, 2, 3... n) which have to be processed one at a time at each of m machines A, B, C.... . The
order of machines for each job, in which it should be go to the machines is given. The
time required by jobs on each of machines is also given. Then the problem is to find the
sequence out of (n !)m sequences, which optimizes (minimizes) the total time elapsed
from the start of the first job to the completion of the last job.

Mathematically,
Let Ai = time required for job i on machine A

Bi = time required for job i on machine B

T = total elapsed time for jobs 1, 2, 3 ... n i.e. time from start of the first job to
completion of the last job.

The problem is to determine a sequence (i1, i2 , … in) where (i1, i2 , ... in) is the permutation
of integers which will minimize T.
 346

Analytic methods have been developed for solving only four simple cases:

1. n jobs and two machines A and B: all jobs processed in the order AB.

2. n jobs and three machines A, B and C: all jobs processed in the order ABC.

3. n jobs and m machines A, B, C.... K: all jobs processed in the order ABC..... K.

4. two jobs and m machines: each job to be processed through the machines in a
prescribed order.

9.2 General Assumptions

Following are the basic assumptions under lying a sequencing problem:

1. No machine can process more than one job at a time.

2. Processing times are independent of processing the jobs.

3. The time involved in moving a job from one machine to another is negligibly small.

4. Each job, once started on machine is to be continued till its completion on it.

5. All machines are of different types.

6. All jobs are completely known and are ready for processing before the period under
consideration begins.

7. Only one machine of each type is available.

8. A job is processed as soon as possible, but only in the order specified.

9. The cost of in-process inventory for each jobs is same and negligibly small.

9.2.1 Terminology and Notations

1. Number of Machines: It means the service facilities through which a job must
pass before it is completed.

2. Processing Order: The order in which various machines are required for
completing the job.

3. Processing Time: It means the time required by each job on each machine.

4. Idle Time on a Machine: This is the time for which a machine remains idle during
the total elapsed time.
[B.B.A. (Garhwal) 2008]

5. Total Elapsed Time: This is the time between starting the first job and completing
the last job. It is denoted by T.
[B.B.A. (Meerut) 2007]
 347

9.3 Sequencing Decision Problem for n-jobs on Two

Machines
[B.C.A.(Rohilkhand) 2007]
Johnson's Method
Let there are n jobs to be processed on two machines, A and B in order AB. The processing
time for each job on each machine is known and is given by
Ai = Processing time for ith job on machine A

Bi = Processing time for ith job on machine B

The procedure for the solution of the above problem, was developed by Johnson and
Bellman. The method is based on minimizing an optimal sequence is as follows:

Step 1: Examine the Ai' s and Bi' s for i = 1, 2, 3 .... n and select the minimum of these. If
there are two or more minimum processing times, then select any one of them arbitrarily.

Step 2: (i) If the minimum processing time is for machine A, process (do) that job first
and place it at the beginning of the sequence.

(ii) If the minimum processing time is for machine B, process (do) that job first and place
it at the end of the sequence.

Step 3: Cross all the jobs already assigned and repeat step 1 and 2, placing the remaining
jobs next to first or next to last, until all the jobs have been assigned.

Step 4: Calculate the time at which each job in the sequence will be processed on
machine A. This time can be calculated as follows:

Time at which ith job in a sequence finishes on machine A = Time when the (i −1)th job in
a sequence finishes on machine A plus the time of processing the ith job on machine
Ai (i = 1, 2 ... n) and the time for start of first job on machine A is zero.

Step 5. Calculate the time at which each job in the sequence will start and finish on
machine B as follows:
(i) Time when first job in a sequence starts on machine B = time when the first job in a
sequence finishes on machine A.
(ii) Time the ith job in the sequence finishes at B = time when the ith job in a sequence
starts on machine B + the processing time of ith job on machine B. (i = 1, 2, 3 ... n)

(iii) Time at which the (i + 1)th job in a sequence finishes on machine B = maximum
(time when the (i + 1)th job in a sequence finishes on machine at the time when the
ith job in a sequence finishes on machine B. (i = 1, 2, 3 ... n)

Step 6: Calculate the total elapsed time to process all jobs through two machines i.e. time
when the nth job in a sequence finishes on machine B.
 348

Step 7: Compute the idle time for machines A and B as follows:

(i) Idle time for machine A = time when the nth job in a sequence finishes on machine
B minus the time when the nth job in a sequence finishes on machine B.
(ii) Idle time for machine B = time at which the first job in a sequence finishes on
N
machine A + ∑ (time when the ith job in a sequence starts on machine B − time
i =1

when the (i − 1)th job in sequence finishes on machine B.)

S olved E xamples
Example 1: A company has 3 jobs on hand. Each of these must be processed through two
departments, the sequential order for which is
Department A : Press shop
Department B : Finishing
The table below lists the number of days required by each job in each department:

Job I Job II Job III

Department A 8 6 5
Department B 8 3 4

Find the sequence in which the three jobs should be processed so as to take minimum time to finish all
the 3 jobs. [B.C.A. (Lucknow) 2010]

Solution:

Step 1: The smallest processing time in the two departments is 3.

Step 2: Since the smallest processing time corresponds to job II in department B. Job II
will be processed in the last as shown

II

Step 3: After job II which is put in the last. We are left with 2 jobs and their processing
times as given below:

Job I Job III

Department A 8 5

Department B 8 4
 349

Step 1 and 2: The minimum processing time in the reduced problem is 4 which
corresponds to job III in department B. Thus job III will be processed next to last and job
I will be processed first.

The optimal sequence, therefore is

I III II

Step 4 and 5: Now we can calculate the elapse time corresponding to the optimal
sequence.

Job Department A Department B Ideal time

of B
Time in Time out Time in Time out

I 0 8 8 16 8

III 8 13 16 20

II 13 19 20 23

Step 6: We see that job I starts at zero time in department A and is over by time 8 days,
when it passes to department B to be worked on till 16 days. Job III then start in the
department A at time 8 days as this department is free at that time. It is completed at
time 13 and has to wait for 3 days before it is passing in B, starting at time 16 days when
this department is free and completes its finishing by time 20 days. Job II starts in A at
time 13 and is completed at the time 19. Then it has to wait for one day before passing on
to B. It enters B at time 20 and works on till 23 days. Thus, the total minimum elapsed
time = 23 days

Step 7: Ideal time in the two departments is given below:

Department A : 4 days = (23 − 19)

Department B : 8 days = (8 − 0)

Example 2: There are 7 jobs, each of which has to go through the machines A and B in the order
AB. Processing times hours are given as

Job : 1 2 3 4 5 6 7

Machine A : 3 12 15 6 10 11 9

Machine B : 8 10 10 6 12 1 3

Determine a sequence of these jobs that will minimize the total elapsed time T.
[M.B.A. (Meerut) 2005, 2007, 2009]
 350

Solution: The smallest processing time in given example is 1 hour for job 6 on machine
B. Thus, we schedule job 6 last as shown below

The reduced set of processing times becomes

Job : 1 2 3 4 5 7

Machine A : 3 12 15 6 10 9

Machine B : 8 10 10 6 12 3

There are two equal minimum values, processing time of 3 hours for job 1 on machine A
and processing time of 3 hours for job seven on machine B. Job 1 is scheduled first and
job 7 next to job 6 as shown below.

1 7 6

The reduced set of processing time becomes

Job : 2 3 4 5

Machine A : 12 15 6 10

Machine B : 10 10 6 12

Again there are two equal minimum value, processing time of 6 hours for job 4 on
machine A as well as on machine B. We may choose arbitrarily to process job 4 next to job
1 or next to job 7 i.e.

1 4 7 6

The reduced set of processing times becomes

Job : 2 3 5

Machine A : 12 15 10

Machine B : 10 10 12

There are three equal minimal values, processing time of 10 hours for job 5 on machine A
and for job 2 and 3 on machine B. Then job 5 is scheduled next to job 4 and job 2 next to
job 7. The optimal sequence is shown below.

1 4 5 3 2 7 6
 351

For minimum elapsed time, we have

Job Machine A Machine B Ideal time

for Machine
Time in Time out Time in Time out
B

1 0 3 3 11 3

4 3 9 11 17 0

5 9 19 19 31 2

3 19 34 34 44 3

2 34 46 46 56 2

7 46 55 56 59 0

6 55 66 66 67 7

From above information we get minimum elapsed time T = 67 hours

Ideal time for machine A = 67 − 66 = 1 hour

Ideal time for machine B = 3 + 0 + 2 + 3 + 2 + 0 + 7 = 17 hours

Example 3: Six jobs go first over machine I and then over II. The order of the completion of jobs has
no significance. The following table gives the machine times in hours for six jobs on the two machines:

Job : 1 2 3 4 5 6

Time on machine I ( Ai ) : 5 9 4 7 8 6

Time on machine II ( Bi ) : 7 4 8 3 9 5

Find the sequence of the jobs that minimizes the total elapsed time to complete the jobs. Find the
minimum time using Gantt ‘s chart or by other method.
[B.C.A. (Rohilkhand) 2006, 2010]

Solution: The minimum time in the given table is 3 which is of B4 machine. Thus, we
schedule job 4 last as shown below

The reduced set of processing times becomes

Job : 1 2 3 5 6

Time on machine I ( Ai ) : 5 9 4 8 6

Time on machine II ( Bi ) : 7 4 8 9 5
 352

Now, minimum time in this table is 4 which is A3

Thus, we shall schedule job 3 first as shown below

3 4

The reduced set of processing times becomes

Job : 1 2 5 6

Time on machine I ( Ai ) : 5 9 8 6

Time on machine II ( Bi ) : 7 4 9 5

Now, minimum time in this table is 4 which is of B2 . Thus, we shall schedule job 2 just
before the last job 4 as shown below

3 2 4

The reduced set of processing times becomes

Job : 1 5 6

Time on machine I ( Ai ) : 5 8 6

Time on machine II ( Bi ) : 7 9 5

Now, the minimum time in this table is 5 which is of A1 and B6 . Thus we shall schedule
job 1 just after the first job 3 and job 6 just before the last job 2 as shown below

3 1 5 6 2 4

For minimum elapsed time, we have

Machine I Machine II Ideal time for

Job
Time in Time out Time in Time out machine II

3 0 4 4 12 4

1 4 9 12 19 0

5 9 17 19 28 0

6 17 23 28 33 0

2 23 32 33 37 0

4 32 39 39 42 2
 353

From the above information we get minimum elapsed time T = 42 hours.

Ideal time for machine I = (42 − 39) = 3 hours

Ideal time for machine II = 4 + 2 = 6

Example 4: In a factory, there are six jobs to perform, each of which should go through two
machines A and B, in the order AB. The processing timings (in hours) for the jobs are given here. You
are required to determine the sequence for performing the jobs that would minimize the total elapsed
time T. What is the value of T ?

Job : J1 J2 J3 J4 J5 J6

Machine A : 1 3 8 5 6 3

Machine B : 5 6 3 2 2 10

Solution:

Step 1: The smallest processing time in the given table is 1 on machine A. So perform J1
in the beginning. Now, delete this job from the given data.

Step 2: of all timings now, minimum is 2 which corresponds to J4 and J5 both on B.

Since the corresponding processing time of J4 on A is less than the corresponding
processing time of J5 on A, therefore, J4 will be processing in the last and J5 next to last.

Step 3: After deleting J4 and J5 also, the minimum time is 3 hours which corresponds to
J2 and J6 on machine A, and to J3 on machine B. Thus J2 will be placed in the IInd
sequence cell and J3 is placed in the sequence cell before J5 .

The resulting optimized sequence of jobs is

J1 J2 J6 J3 J5 J4

The total elapsed time is calculated below.

Machine A Machine B Ideal time for

Job
Time in Time out Time in Time out machine B

J1 0 1 1 6 1

J2 1 4 6 12 0

J6 4 7 12 22 0

J3 7 15 22 25 0

J5 15 21 25 27 0

J4 21 26 27 29 0
 354

From the above information, we get minimum elapsed time T = 29 hours.

Ideal time for machine A = (29 − 26) = 3 hours.

Ideal time for machine B = 1 hour.

Example 5: A book binder has one printing press, one binding machine, and the manuscripts of
number of different books. The time required to perform the printing and binding operations for each
book are shown below. Determine the order in which books should be processed, in order to minimize
the total time required to turn out all the books.

Book : 1 2 3 4 5 6

Printing time (hrs) : 30 120 50 20 90 110

Binding time (hrs) : 80 100 90 60 30 10

[M.B.A. (Meerut) 2008, M.C.A. (Meerut) 2009]

Solution: Here, the books will first go to the printing press and then on to the binding
machine. If
Ai (i = 1, 2, 3, 4, 5, 6) = the time in hours on printing press
Bi (i = 1, 2, 3, 4, 5, 6) = the binding time for book

Then using sequencing algorithm the following sequence is obtained :

4 1 3 2 5 6

The total elapsed time is calculated below.

Job Printing machine Binding machine Ideal time of

binding
Time in Time out Time in Time out
machine

4 0 20 20 80 20

1 20 50 80 160 0

3 50 100 160 250 0

2 100 220 250 350 0

5 220 310 350 380 0

6 310 420 420 430 40

From the above information, we get minimum elapsed time T = 430 hours
Ideal time for printing machine = 430 − 420 = 10 hours
Ideal time for binding machine = 20 + 40 = 60
 355

Example 6: Find the sequence that minimizes the total elapsed time required to complete the
following tasks on two machines:

Task A B C D E F G H I

Machine (I) 2 5 4 9 6 8 7 5 4

Machine (II) 6 8 7 4 3 9 3 8 11

[B.B.A. (Delhi) 2007, 2009; B.B.A. (Kanpur) 2008]

Solution: The minimum time in the table is 2 which is for job A on machine I.

∴ We must do the job A first. After assignment job A we are left with the remaining 8 jobs
with their processing times. Now the minimum of the processing time for the remaining
jobs is 3 which is for job E and G both on machine II. Therefore we must do the job E or G
(either of them) in the last. Proceeding in this way we get the following optimal
sequencings.

1. A C I B H F D G E

2. A I C H B F D G E

3. A I C B H F D G E

4. A C I H B F D G E

5. A C I B H F D E G

6. A I C H B F D E G

7. A C I H B F D E G

8. A I C B H F D E G

It may be seen that the total time elapsed for all sequencings are equal.
 356

The minimum elapsed time may be computed as follows:

Machine I Machine II Ideal time

Task for Machine
Time in Time out Time in Time out II

A 0 2 2 8 2

C 2 6 8 15 0

I 6 10 15 26 6

B 10 15 26 34 0

H 15 20 34 42 0

F 20 28 42 51 0

D 28 37 51 55 0

G 37 44 55 58 0

E 44 50 58 61 −

From the above information, we get minimum elapsed time = 61 hours

Ideal time for machine (I ) = 61 − 50 = 11 hours.
Ideal time for machine (II ) = 2 + 0 = 2 hours.

9.4 Sequence Decision Problem for n-jobs on Three

Machines
Modified Johnson's Method
Let there are n-jobs each of which is to be processed through three machines A, B and C.
This sequencing problem is completely described as follows:
1. Only three machines A, B and C are involved.
2. Each job is processed in the prescribed order ABC.
3. No passing of jobs is permitted.
4. The actual or expected processing times A1, A2 … An, B1, B2 … Bn and C1, C2 , … Cn
are known.
 357

The problem, again is to find the optimum sequence of jobs which minimizes the total
time T.

No general solution is available at present for such a case. The method of previous article
can be extended to the cases in which either one or both of the following conditions
holds.
1. The minimum time on machine A is ≥ maximum time on machine B.
and
2. The minimum time on machine C ≥ maximum time on machine B.

The method, described here is to replace the problem by an equivalent problem involving
n jobs and two machines. These two machines are denoted by G and H and
corresponding processing time are given by
Gi = Ai + Bi

 Hi = Bi + Ci

Now find the optimal sequence of jobs in the order GH of these two machines G and H
by the method of n-jobs and two machines. These resulting optimal sequence on two
machines G and H will also be the optimal sequence on three machines A, B and C.

Example 7: A machine operator has perform three operations: Turning, threading and knurling on
a number of different jobs. The time required to perform these operations (in minutes) for each job is
known. Determine the order in which the jobs should be processed in order to minimize the total time
required to turn out all the jobs.

Table

Jobs Time for Time for Time for

Turning (A) Threading (B) Knurling (C)
(minutes) (minutes) (minutes)

1 3 8 13

2 12 6 14

3 5 4 9

4 2 6 12

5 9 3 8

6 11 1 13

Solution: Here, min Ai = 2, Max Bi = 8 and mini (Ci) = 8 since min Ci = maxi Bi. Then
three machines can be converted in two machines G and H
 358

Processing times for 6 jobs and two machines

Job Gi = Ai + Bi Hi = Bi + Ci

1 11 21

2 18 20

3 9 13

4 8 18

5 12 11

6 12 14

Now, to form job sequence, we find that the smallest value is 8 under operation Gi in row
4. Thus we schedule job 4 first as shown below.

The reduced set of processing times becomes.

Job Gi Hi

1 11 21

2 18 20

3 9 13

5 12 11

6 12 14

The next smallest value is 9 under the column Gi for job 3 we can schedule job 3 as shown
below.

4 3

The reduced set of processing times becomes.

Job Gi Hi

1 11 21

2 18 20

5 12 11

6 12 14
 359

There are two equal minimal values, processing time of 11 minutes under column Gi for
job 1 and processing time of 11 minutes under column Hi for job 5. Then job 1 is placed
next to job 3 and job 5 is placed last as shown below.

4 3 1 5

The reduced set of processing times becomes

Job Gi Hi

2 18 20

6 12 14

The smallest value is 12 under column Gi, for job 6. Hence we schedule job 6 next to job 1
and the optimal sequence becomes.

4 3 1 6 2 5

For the total elapsed time, we have

Job Turning Threading Knurling Ideal Ideal

operation (A) operation (B) operation (C) time for time
B for C
Time Time Time Time Time Time
in out in out in out

4 0 2 2 8 8 20 2 8

3 2 7 8 12 20 29 0 0

1 7 10 12 20 29 42 0 0

6 10 21 21 22 42 55 1 0

2 21 33 33 39 55 69 11 0

5 33 42 42 45 69 77 3+ 0
(77 − 45)

From the given information, we get minimum elapsed time T = 77 minutes

Ideal time for turning operation

( A) = (77 − 42) = 35 minutes

Ideal time for threading operation

(B) = 2 + 0 + 0 + 1 + 11 + 3 + (77 − 45) = 49 minutes
 360

Ideal time for knurling operation

(C) = 8 + 0 + 0 + 0 + 0 + 0 = 8 minutes.

Example 8: We have five jobs, each of which must go through the machines A, B and C in the order
ABC, processing times are:

Job A B C

1 4 5 8

2 9 6 10

3 8 2 6

4 6 3 7

5 5 4 11

Determine a sequence for the five jobs that will minimize the elapsed time T.

Solution: Here mini Ai = 4, Max. Bi = 6, Mini Ci = 6

since Max. Bi ≤ Mini. Ci

∴ We may consider two fictitious machines G and H whose processing times are given in
the following table.

Jobs Gi = Ai + Bi Hi = Bi + Ci

1 9 13

2 15 16

3 10 8

4 9 10

5 9 15

Using sequencing algorithms we get the following optimal sequencings

1 4 5 2 3 1 5 4 2 3 4 5 1 2 3

5 4 1 2 3 5 1 4 2 3 4 1 5 2 3
 361

The minimum elapsed time can now be calculated corresponding to the first optimal
sequencing using the individual machining times given in the problem.

Jobs Machine (A) Machine (B) Machine (C) Ideal time

of machine
Time in Time out Time in Time out Time in Time out (C)

1 0 4 4 9 9 17 9

4 4 10 10 13 17 24 0

5 10 15 15 19 24 35 0

2 15 24 24 30 35 45 0

3 24 32 32 34 45 51 0

From the given information, we get the minimum elapsed time T = 51 hours

Ideal time for machine A = 51 − 32 = 19 hours

Ideal time for machine B = 4 + 1 + 2 + 5 + 2 + (51 − 34) = 31 hours

Ideal time for machine C = 9 + 0 + 0 + 0 + 0 = 9 hours.

Example 9: There are five jobs, each of which is to be processed through three machines A, B and C
in the order ABC. Processing times in hours are

Jobs A B C

1 3 4 7

2 8 5 9

3 7 1 5

4 5 2 6

5 4 3 10

Determine the optimum sequence for the five jobs and the minimum elapsed time.

Solution: Here mini Ai = 3, maxi Bi = 5 and min Ci = 5. Since min Ci = max Bi, then three
machines can be converted into two machines G and H.

Hence the equivalent problem become 5 jobs and two machines G and H.
 362

Machines times for five jobs and two machines G and H

G H
Jobs
(Gi = Ai + Bi ) ( Hi = Bi + Ci )
1 7 11
2 13 14
3 8 6
4 7 8
5 7 13

Examining the columns Gi and Hi, we find that the smallest value of machining time is 7
hrs, under the column Gi for job 1, 4 and 5. Since all values are in the same column Gi, we
schedule job 4 first since it has lower entry 8 in column Hi, job 1 next as it next height
entry 11 in column Hi and job 5 next to job 1, since it has still higher entry 13 in column
Hi then job sequence is shown below

4 1 5

The reduced set of machining time becomes

Jobs Gi Hi
2 13 14
3 8 6

The smallest value is 6 hrs in column Hi for job 3. Thus we put job 3 last and the
optimum sequence becomes

4 1 5 2 3

For total elapse time we have

Machine A Machine B Machine C Ideal time Ideal time

for for
Jobs Time Time Time Time Time Time
machine B machine C
in out in out in out

4 0 5 5 7 7 13 5 7

1 5 8 8 12 13 20 1 0

5 8 12 12 15 20 30 0 0

2 12 20 20 25 30 39 5 0

3 20 27 27 28 39 44 2 +(44 − 28) 0
 363

From the given information, we get the minimal elapsed time T = 44 hours

Ideal time for machine A = 44 − 27 = 17 hours.

Ideal time for machine B = 5 + 1 + 0 + 5 + 2 + (44 − 28)

= 13 + 16 = 29 hours

Ideal time for machine C = 7 + 0 + 0 + 0 + 0 = 7 hours

9.5 Sequence Decision Problem for n Jobs on m Machines

Let there are n jobs each of which is to be processed through m machines say
M1, M2 … Mm in the order M1 M2 … Mm . Now to find an optimal sequence as follows:

Step 1. Find minimum Mi1, min Mi m (i = 1, 2, 3 ... n) and maximum of each of

Mi2 , Mi3 , … Mi m − 1 for i = 1, 2, 3 … n

Step 2. Now to check either one or both conditions are satisfying

(i) min Mi1 (i = 1, 2 … n) ≥ max Mi j for j = 1, 2 … (m − 1)

(ii) min Mim (i = 1, 2 … n) ≥ max Mi j for j = 1, 2 … (m − 1)

Step 3. If the inequality of step 2 is not satisfying then method fail. Otherwise go to next
step.

Step 4. Convert the m machine problem into two machine problem G and H. Such that

MiG = Mi1 + Mi2 + … Mi (m −1)

and Mi H = Mi2 + Mi3 + … + Mim

Determine the optimal sequence of n jobs through 2 machines by using optimal sequence
algorithm.

Step 5. In addition to conditions given in step 4, if

Mi2 + Mi3 + … + Mi (m −1) = C

is a fixed positive constant for all i = 1, 2, 3 … n, then determine the optimal sequence for n
jobs and two machines M1 and Mm in order M1 Mm by using the optimal sequence
algorithm.
 364

Example 10: Solve the following sequencing problem when passing is not allowed.

Machines
Items
A B C D E

I 9 7 4 5 11

II 8 8 6 7 12

III 7 6 7 8 10

IV 10 5 5 4 8

Solution: Here, min Ai = 7, min Ei = 8 and max Bi = 8, maxi Ci = 7 and max Di = 8

Since min Ei ≥ {max Bi, max ci, max Di}, then given problem 4 jobs and 5 machines can
be converted into 4 jobs and two machines G and H.
where G = Ai + Bi + Ci + Di
H = Bi + Ci + Di + Ei
Hence,

Jobs : I II III IV

Machine (G) : 25 29 28 24

Machine (H) : 27 33 31 22

The optimal sequence of jobs is as follows by usual process.

I III II IV

Now, we may calculate the total elapsed time corresponding to optimal sequence, using
the individual processing time given in the original problem.

Jobs Machine A Ideal

Machine B Machine C Machine D Machine E
time
for E

Time Time Time Time Time Time Time Time Time Time
in out in out in out in out in out

I 0 9 9 16 16 20 20 25 25 36 25

III 9 16 16 22 22 29 29 37 37 47 1

II 16 24 24 32 32 38 38 45 47 59

IV 24 34 34 39 39 44 45 49 59 67
 365

From the above information, we get minimum elapsed time T = 67 hours

Ideal time for machine ( A) = (67 − 34) = 33 hours

Ideal time for machine (B) = 9 + 0 + 2 + 2 + (67 − 39)

= 9 + 2 + 2 + 28

= 41 hours

Ideal time for machine (C) = 16 + 2 + 3 + 1 + (67 − 44)

= 22 + 23

= 45 hours

Ideal time for machine D = 20 + 4 + 1 + (67 − 49)

= 25 + 18

= 43 hours

Ideal time for machine E = 25 + 1 + 0 + 0 = 26 hours.

Example 11: We have 4 jobs each of which go through the machine M j ( j = 1, 2, 3, 4, 5, 6) in order
M1 M2 … M6 processing time (in hours) is given below:

Jobs Machines

M1 M2 M3 M4 M5 M6

A 18 8 7 2 10 25

B 17 6 9 6 8 19

C 11 5 8 5 7 15

D 20 4 3 4 8 12

Determine a sequence of these four jobs that minimizes the total elapsed time T.

Solution: Here Min ( M1) = 11, Maxi ( M2 ) = 8, Max ( M3 ) = 9,

Max ( M4 ) = 6, Maxi ( M5 ) = 10, Min ( M6 ) = 12

Hence (i) Mini ( M1) ≥ max { M2 , M3 , M4 , M5 }

(ii) Min ( M6 ) ≥ Max { M2 , M3 , M4 , M5 }

Now, both conditions are satisfying then 6-machines can be converted into two
machines G and H
 366

Jobs : A B C D

Machine G : 45 46 36 39

Machine H : 52 48 40 31

Where Gi = M1 + M2 + M3 + M4 + M5

and Hi = M2 + M3 + M4 + M5 + M6

Using the optimal sequence algorithm, the following optimal sequence is easily obtained

C A B D

Total elapsed time is given in the following table

Machine Machine Machine Machine Machine Machine

M1 M2 M3 M4 M5 M6
Job

Time Time Time Time Time Time Time Time Time Time Time Time
in out in out in out in out in out in out

C 0 11 11 16 16 24 24 29 29 36 36 51

A 11 29 29 37 37 44 44 46 46 56 56 81

B 29 46 46 52 52 61 61 67 67 75 81 100

D 46 66 66 70 70 73 73 77 77 85 100 112

This table shows that the minimal total elapse time T = 112 hours.

9.6 Processing Two Jobs Through m Machines

When two jobs are to be processed on m machines, we generally use graphical method to
determine the total elapsed time as well as the optimal sequence. Let us consider the
following situation.
 367

1. There are m machines, denoted by A, B, C, … K

2. Only two jobs are to be performed : job 1 and job 2.

3. The technological ordering of each of the two jobs through m machines is known.

4. The actual or expected processing time A1, A2 … Ki, A2 , B2 … K2 are known.

5. Each machine can work only one job at a time and storage space for in-process
inventory is available.

Example 12: Using graphical method, calculate the minimum time needed to process job 1 and 2 on
five machines, A, B, C, D and E i.e. for each machine find the job which should be done first. Also
calculate the total time needed to complete both jobs.

Job 1 Job 2

Sequence of machines Time Sequence of machines Time

A 3 B 5

B 4 C 4

C 2 A 3

D 6 D 2

E 2 E 6

Solution: The graphical method is described with the help of the following steps:

Step 1: Draw two axes at right angles to each other, represent processing time on job 1
along horizontal axis and processing time on job 2 along vertical axis.

Step 2: Layout the machine time for the two jobs on corresponding axes in the given
technological order.

Step 3: Machine A requires 3 hour for job 1 and 3 hours for job 2. A rectangle LMNP is,
thus constructed for machine A. Similar rectangles are constructed for machines B, C, D
and E as shown.
 368

Idle time
Job 2 for job 2
Finish
20 P

E
E

14

Idle time
for job 1
D D
P N
12
11
A
A
9
L M

C C

5
Idle time
for job 1

3
B B

Start 0 3 7 9 15 17 Job 1

A B C D E

Fig. 9.1

Step 4: Draw the line by starting from origin (0) and moving through the various stages
of completion (points) till the point marked finish is reached. Then take path consisting
only of horizontal, vertical and 45° lines. A horizontal line represents work on job 1 while
job 2 remains idle, a vertical line represents work on job 2 while job 1 remains idle and a
45° line to the base represents simultaneous work on both jobs.

Step 5: Find the optimal path. An optimal path is one that minimizes idle time for job 1.
Likewise, an optimal path is one that minimizes idle time for job 2. Then the optimal
path is one which coincides with 45° line to maximum extent.

Further, both jobs can not be processed simultaneously on one machine. Graphically it
means that diagonal through the blocked out area is not allowed.

Step 6: Find the elapsed time i.e.

The elapsed time = processing time of job 1 + ideal time of job 1
= 17 + (2 + 3) = 22 hrs
or = Processing time of job 2 + ideal time of job 2
 369

= 20 + 2 = 22 hrs.
The optimal processing sequence for the two jobs on various machines, as evident from
the figure is
Job 1 before 2 on machine A
Job 2 before 1 on machine B
Job 2 before 1 on machine C
Job 2 before 1 on machine D
and Job 2 before 1 on machine E

Example 13: Use graphical method to minimize the time required to process the following jobs on
the machines i.e. for each machine specify the job which should done first. Also calculate the total
elapsed time to complete both jobs.
[M.B.A. (Meerut) 2007, 2009]

Job 1 Job 2

Sequence of machines Time Sequence of machines Time

A 6 B 10

B 8 C 8

C 4 A 6

D 12 D 4

E 4 E 12

Solution:

Step 1: Draw two axes at right angles to each other, represent processing time on job 1
along horizontal axis and processing time on job 2 along vertical axis.

Step 2: Layout the machine time for the two jobs on corresponding axes in the given
technological order.

Step 3: Machine A requires 6 hour for job 1 and 6 hours for job 2. A rectangle LMNP is,
thus constructed for machine A. Similar rectangles are constructed for machines B, C, D
and E as shown.

Step 4: Draw the line by starting from origin (O) and moving through the various stages
of completion (points) till the point marked finish is reached. Then take path consisting
only of horizontal, vertical and 45° lines. A horizontal line represents work on job 1 while
job 2 remains idle, a vertical line represents work on job 2 while job 1 remains idle and a
45° line to the base represents simultaneous work on both jobs.

Step 5: Find the optimal path. An optimal path is one that minimizes idle time for job 1.
Likewise, an optimal path is one that minimizes idle time for job 2. Then the optimal
path is one which coincides with 45° line to maximum extent.
 370

Further, both jobs can not be processed simultaneously on one machine. Graphically it
means that diagonal through the blocked out area is not allowed.

Step 6: Find the elasped time i.e.

The elapsed time = processing time of job 1 + ideal time of job 1
= 34 + 10 = 44 hrs.
The elapsed time = processing time of job 2 + ideal time of job 2
= 40 + 4 = 44 hrs.
The optimal processing sequence for the two jobs on various machines, as
Job 1 before job 2 on machine A
Job 2 before job 1 on machine B
Job 2 before job 1 on machine C
Job 2 before job 1 on machine D and
Job 2 before job 1 on machine E
Ideal time
Job 2 for Job 2
40 Finish

36
E
E
32

45°
28
D Ideal time D
P N for Job 1
24

A
20 A
L M
16
C
C
12

Ideal time
8 for Job 1

B B
4

45°
O 4 6 8 12 14 16 18 20 24 28 30 32 34 36 Job 1

A B C D E
Fig. 9.2
 371

P roblem S et
1. What is sequencing problem ?

2. Write short note on sequencing problem. [B.C.A. (Rohilkhand) 2008]

3. What is ‘no passing rule’ in a sequencing algorithm ?

4. Give the Johnson's method for determining an optimal sequence for processing.
[B.C.A. (Agra) 2010]

5. Explain how to process n jobs through m-machines.

6. State clearly the assumptions used in the study of sequencing problems.

7. Seven jobs each of which has to go through two machines M1 and M2 in order
M1 M2 take time on the machine as follows : Find in which order the jobs should be
performed to minimise the time.

Jobs : 1 2 3 4 5 6 7

M1 : 3 12 15 6 10 11 9

M2 : 8 10 10 6 12 1 3

8. Six jobs go over two machines I and II in that order. The order of the completion of
the jobs has no significance. From the data given below, find the sequence that
minimizes the total elapsed time, and also find the minimum time.

Jobs : 1 2 3 4 5 6

Machine I 4 8 3 6 7 5
Time in hours :
Machine II 6 3 7 2 8 4

9. The following table given the machine times (in hours) for 5 jobs to be processed on
two-different machines:

Jobs : 1 2 3 4 5

Machine A : 3 7 4 5 7

Machine B : 6 2 7 3 4

Passing time is not allowed. Find the optimal sequence in which jobs should be
processed. [B.C.A. (Kanpur) 2009]
 372

10. A readymade garment manufacturer has to process seven items through two stages
of production i.e., cutting and sewing. The time taken for each of these item at the
different stages are given below in appropriate units:

Item No. Processing Time Cutting Processing Time Sewing

1 5 2

2 7 6

3 3 7

4 4 5

5 6 9

6 7 5

7 12 8

Find an order in which these items are to be processed through these stages so as
to minimize the total processing time. [B.C.A. (Avadh) 2010]

11. We have 5 jobs, each of which must go through the two machine A and B in order
AB. Processing times are given in the table below:

Processing time in hours

Job : 1 2 3 4 5

Machine A : 5 1 9 3 10

Machine B : 2 6 7 8 4

Determine a sequence for the five jobs that will minimize the elapsed time T.
12. A book-binder has one printing press, one binding machine and manuscripts of a
number of different books. The time required in min. to perform the printing and
binding operation of each book are known. We wish to determine the order in
which books should be processed in order to minimize the total time required to
turn out all the books.

Book : 1 2 3 4 5 6

Printing time : 30 120 50 20 90 110

Binding time : 80 100 90 60 30 10

 373

13. Find the sequence, for the following eight jobs, that will minimize the total elapsed
time for the completion of all the jobs. Each job is processed in the same order CAB.
Entries give the time in hours on the machines.

Jobs 1 2 3 4 5 6 7 8

A 4 6 7 4 5 3 6 2
Times
on B 8 10 7 8 11 8 9 13
Machines
C 5 6 2 3 4 9 15 11

14. There are five jobs, each of which must go through the machines A, B and C in the
order ABC. Processing times are:

Job A B C

1 4 5 8

2 9 6 10

3 8 2 6

4 6 3 7

5 5 4 11

Determine a sequence for the five jobs that will minimize the elapsed time T.
15. We have 5 jobs, each of which must go through machines, A, B and C in the order
ABC. Processing times are given in the following table :

Jobs 1 2 3 4 5

Machines A ( Ai ) 8 10 6 7 11

Machines B ( Bi ) 5 6 2 3 4

Machines C (Ci ) 4 9 8 6 5

Determine a sequence for the five jobs that will minimize the elapsed time T.
16. Find the time that minimizes the total required to complete the following tasks.

Task A B C D E F G

Machines I 3 8 7 4 9 8 7

Machines II 4 3 2 5 1 4 3

Machines III 6 7 5 11 5 6 12

[B.B.A. (Rohilkhand) 2008]

 374

17. Find the sequence and idle time of each machine that minimized the total elapsed
time required to complete all the jobs on machines A, B, C in the order ABC.

Jobs
Machines
1 2 3 4 5

A 7 12 11 9 8

B 8 9 5 6 7

C 11 13 9 10 14

18. Solve the following sequencing problem giving an optimal solution when passing
time is not allowed.

Jobs
Machines
A B C D E

M1 10 12 8 15 16

M2 3 2 4 1 5

M3 5 6 4 7 3

M4 14 7 12 8 10

[B.B.A. (Delhi) 2005, 2007, 2009; B.B.A. (Bhopal) 2008, 2009; B.B.A. (Agra) 2009]

19. Solve the following sequencing problem of jobs on six machines M j ( j = 1, 2, 3, 4, 5, 6)

in the order M1, M2 , M3 , M4 , M5 , M6 . Processing times (in hours) are given below.

Machines
Jobs
M1 M2 M3 M4 M5 M6

A 18 8 7 2 10 25

B 17 6 9 6 8 19

C 11 5 8 5 7 15

D 20 4 3 4 8 12

[B.B.A. (Lucknow) 2008]

 375

20. Use the graphical method to minimize the time needed to process the following two
jobs on four machines A, B, C and D with technological orderings.

Job 1 : A B C D

Job 2 : D B A C

and the processing times:

A B C D

Job 1 : 2 4 5 1

Job 2 : 6 5 3 6

21. Use graphical method to minimize the time needed to process the following jobs on
two machines shown below i.e. for each machine find the job which should be done
first. Also calculate the total time needed to complete both jobs:

Job 1 Job 2

Sequence of Machines Time Sequence of Machines Time

A 2 C 4

B 3 A 5

C 4 D 3

D 6 E 2

E 2 B 6

[B.B.A. (Kanpur) 2007]

 376

A nswers
7. 1 − 4 − 5 − 3 − 2 − 7 − 6, 1−5 −3 −2 −4 −7 −6

1 − 4 − 5 − 2 − 3 − 7 −6 and 1 − 5 − 2 − 3 − 4 − 7 − 6

8. 3 − 1 − 5 − 6 − 2 − 4, Mini time = 35 hrs.

9. 1−3 −5 −4 −2

10. 3 − 4 − 5 − 7 − 2 − 6 −1

11. 2 − 4 − 3 − 5 − 1; Mini time T = 30 hrs.

12. 4 − 1 − 3 − 2 − 5 − 6 ; Mini time T = 430 minutes

13. 4 − 1 − 3 − 5 − 2 − 8 − 7 − 6; Mini time T = 81 hrs.

14. 4 − 1 − 5 − 2 − 3 ; 4 − 5 − 1 − 2 − 3 ;1 − 4 − 5 − 2 − 3 ;1 − 5 − 4 − 2 − 3 ,

5 − 1 − 4 − 2 − 3 ; 5 − 4 − 1 − 2 − 3 , Total elape time = 51 hrs.

15. 3 − 2 − 1 − 4 − 5 ; 3 − 2 − 4 − 1 − 5 ; 3 − 2 − 4 − 5 − 1,

3 − 2 − 5 − 4 − 1, 3 − 2 − 1 − 5 − 4 and 3 − 2 − 5 − 1 − 4

16. A − D − G − F − B − C − E, A − D − G − B − F − C − E

Minimum time T = 59 hours

17. 1 − 5 − 4 − 2 − 3 , Total elapsed time = 72 minutes

Ideal times are for A = 25 min, for B = 37m and for C = 15 m

18. C−A−E−D−B

19. C−A−B−D

20. Minimum time T = 21 hrs.

21. Job 1 before Job 2 on machine A

Job 1 before Job 2 on machine B
Job 1 before Job 2 on machine C
Job 2 before Job 1 on machine D
Job 1 before Job 2 on machine E
Total elapsed time 20 hrs.
 377

9.7 The Travelling Salesman Problem

[B.B.A. (Meerut) 2012]
There are number of cities a salesman wants to visit. The distance (or time or cost)
between every pair of cities is known. He starts from his home city, passes through each
city once and only once or exactly one time and return to his home city. The problem is to
find the routes shortest in distance (or time or cost).

If the distance or time or cost between every pair of cities is independent of the direction
of travel, the problem is said to be symmetrical if for one or more pair of cities (or time or
cost) varies with direction the problem is called asymmetrical. For example i.e. it take
more time to go up a hill between two stations then come down the hill between them;
similarly a flight may take more time against the wind direction compared to that in the
direction of wind. If the salesman is to visit only two cities there is of course no choice. If
the number of cities is three ( A, B and C) of which starting base A, there are two possible
routes A → B → C and A → C → B. In general, for n cities there are (n − 1) ! possible
routes. The problem is to find the best route without trying each one. Such types of
problems arise in the following areas of management:

1. Postal deliveries 2. Inspection 3. School bus routing 4. Television relays 5. Assembly

lines etc. The travelling salesman problem appear to be related to sequencing problem
but actually it is more similar to assignment problem with the difference that there is an
additional constraint mathematically, the problem may be stated as follows; given Cij as
the cost of going from. City i to city j and x ij = 1, for going directly from i to j and 0,
otherwise,
minimize ∑i ∑j cij xij
Subject to the additional constraint that x ij is to be so chosen that no city is visited twice
before the tour of all the cities is completed. In particular, going from i to i is not
permitted which means cij = ∞ it should be noted that there must be only one x ij = 1 for
each value of i and j. The travelling salesman problem can be written in the form of square
assignment problem.
To

1 2 3 … n

1 ∞ c12 c13 … c1n

2 c21 ∞ c23 … c2 n

From 3 c31 c32 ∞ … c3 n

⋮ ⋮ ⋮ ⋮ ⋮ ⋮

n cn1 cn2 cn3 ⋯ ∞

 379

or 1 → 2 → 4 → 1, 3 → 5 → 3

So this is not the solution to the travelling salesman problem, as it is not allowed to go
back to city 1 from city 4 without visiting cities 3 and 5.

Now we try to get next best solution which satisfy the extra restriction. The smallest
element in the table other than zero is 1. So we try to bring 1 into the solution. Since 1
occurs at two places, so we shall consider both the cases separately until the acceptable
solution is attained.

Now, if we put assignment in the cell (1, 3) having the element 1 instead of making 0
assignment in the cell (5, 3) and there will be no assignment in the first row and third
column then make assignment in the cell (5, 2) having element 0, instead of assignment
in (5, 3)

1 2 3 4 5

1 ∞ ×0 1 ×0 ×0
2 ×0 ∞ ×0 0 1

3 2 2 ∞ 6 0

4 0 ×0 2 1 3

5 2 0 ×0 6 ∞

Hence resulting feasible solution is 1 → 3 → 5 → 2 → 4 → 1 having the cost 1. If we put

assignment in the cell (2, 5) in place (2, 4) and change the assignment to (3, 4) from (3, 5)
in that case the feasible solution is

1 → 2 → 5 → 3 → 4 → 1, having cost 6 which is more than 1 from above. Hence the

optimal route is 1 → 3 → 5 → 2 → 4 → 1 and minimum cost 12 + 3 + 2 + 5 + 5 = 27

Example 15: Solve the following travelling salesman problem so as to minimize the cost per cycle:

To

A B C D E

A − 3 6 2 3

B 3 − 5 2 3

From C 6 5 − 6 4

D 2 2 6 − 6

E 3 3 4 6 −
 380

Solution: Reduce the cost matrix and make assignment in rows and columns having
single zeroes. Then draw the minimum number of lines to cover all the zeros by usual
assignment process, we get

A B C D E A B C D E

A M 1 3 0 1 P A M 0 2 ×0 ×0
B 1 M 2 ×0 1 P B ×0 M 1 0 ×0
C 2 1 M 12 0 C 2 1 M 3 0

D 0 ×0 3 M 4 D 0 ×0 3 M 4

E ×0 ×0 0 3 M E ×0 ×0 0 4 M

By subtracting the lowest element 1 from all the elements not covered by lines and
adding the some at the intersection of two lines.

The optimum assignment is A → B, B → D, C → E, D → A and E → C with minimum

cost = 3 + 2 + 4 + 2 + 4 = 15 since this assignment schedule does not provide us the
solution of travelling salesman problem. We try to find the next solution which satisfies
the extra condition also make assignment at (2, 3) instead of zero at (2, 4).

The resulting table is

A B C D E A B C D E

A M ×0 2 0 ×0 A M ×0 2 ×0 0

B ×0 M 1 ×0 ×0 B ×0 M ×0 0 ×0
C 2 ×0 M 3 0 C 2 1 M 3 ×0
D ×0 0 3 M 4 D 0 ×0 3 M 4

E 0 ×0 ×0 4 M E ×0 ×0 0 4 M

The optimum assignment schedule is

A→ D→ B→ C→ E→ A or A→ E→ C→ B→ D→ A

Total minimum cost per cycle in both the cases will be 16.

Example 16: Solve the travelling salesman problem given by the following data:
c12 = 20, c13 = 4, c14 = 10, c 23 = 5, c 34 = 6
c 25 = 10, c 35 = 6, c 45 = 20, c ij = c ji
 381

and there is no route between cities i and j if the value for cij is not given above.
[B.C.A. (Avadh) 2011]

Solution: The assignment of given problem can be written as

1 2 3 4 5

1 ∞ 20 4 10 ∞

2 20 ∞ 5 ∞ 10

3 4 5 ∞ 6 6

4 10 ∞ 6 ∞ 20

5 ∞ 10 6 20 ∞

If there is no route between cities i and j then we have taking cij = ∞ for i = j

Now we will solve the assignment by usual process. we get

1 2 3 4 5 1 2 3 4 5

1 ∞ 15 0 4 ∞ 1 ∞ 12 0 1 ∞

2 15 ∞ ×0 ∞ 3 2 12 ∞ ×0 ∞ 0

3 0 ×0 ∞ ×0 ×0 3 0 ×0 ∞ ×0 ×0
4 4 ∞ ×0 ∞ 12 4 1 ∞ ×0 ∞ 9

5 ∞ 3 ×0 12 ∞ 5 ∞ 0 ×0 9 ∞

1 2 3 4 5

1 ∞ 11 0 ×0 ∞

2 12 ∞ 1 ∞ 0

3 ×0 ×0 ∞ 0 ×0
4 0 ∞ ×0 ∞ 8

5 ∞ 0 1 9 ∞

Hence, the optimal solution of the assignment problem is:

1 → 3, 3 → 4, 4 → 1, 2 → 5, 5 → 2
 382

But this is not the solution to the travelling salesman problem, as it is not allowed to go
from city 4 to 1 without visiting the cities 2 and 5.

We try to find best solution which satisfying the additional condition. The smallest
element other then zero is one which is in two places. We consider both the cases
separately until the acceptable solution is attained.

Make assignment at 1 in cell (2, 3) instead of 0 in the cell (2, 5) making this assignment
we note that there are two assignments in column 3 while there is no assignment in
column 5. So we should shift the assignment at 0 in cell (1, 3) to cell (1, 5) which is not
possible, as this route is not allowed.

A B C D E

A ∞ 11 0 0 ∞

B 12 ∞ 1 ∞ 0

C 0 0 ∞ 0 0

D 0 ∞ 0 ∞ 8

E ∞ 0 1 9 ∞

Again we make assignment at 1 in cell (5, 3) instead 0 in cell (5, 2). Making this
assignment the assignment in cell (1, 3) is shifted to cell (1, 2). So we get the following
table.
This feasible solution is 1 → 2 → 5 → 3 → 4 → 1

In this case the cost is increased by (11 + 1) − 0 = 12. In comparision in previous cost 0 in
the last table

The next element in the table greater than 1 is 8 in cell (4, 5). So we shift the assignment
from 0 to cell (4, 1) to 8 of cell (4, 5) and then shift the assignment from cell (2, 5) to cell
(2, 1). So we get the following table.

∞ 11 0 0 ∞

12 ∞ 1 ∞ 0

0 0 ∞ 0 0

0 ∞ 0 ∞ 8

∞ 0 1 9 ∞

This feasible solution is 1 → 3 → 4 → 5 → 2 → 1

 383

In this case the cost is increase by (2 + 8) − 0 = 20

Again next greater element in this table is 9 in cell (5, 4). So we shift the assignment from
cell (3, 4) to (3, 2). This we get the following table and the feasible solution is
1 → 3 → 2 → 5 → 4 → 1:

1 2 3 4 5

1 ∞ 11 0 0 ∞

2 12 ∞ 1 ∞ 0

3 0 0 ∞ 0 0

4 0 ∞ 0 ∞ 0

5 ∞ 0 1 9 ∞

In this case the cost is increased by (9 + 0) − 0 = 9

Hence the optimal feasible solution of the problem i.e. the optimum route is

1→ 3 → 2 → 5 → 4 → 1

The minimum cost = 4 + 5 + 10 + 20 + 10 = 49.

P roblem S et
1. A salesman has to visit five cities A, B, C, D and E. The distance (in hundred miles)
between the five cities are as follows:

A B C D E

A − 7 6 8 4

B 7 − 8 5 6

C 6 8 − 9 7

D 8 5 9 − 8

E 4 6 7 8 −

If the salesman starts from city A and has to come back to city A, which route
should he select so that the total distance travelled is minimum.
[B.C.A. (Lucknow) 2010]
 384

2. Solve the travelling salesman problem given by the following data.

c12 = 16, c13 = 4, c14 = 12, c 23 = 6, c 34 = 5, c 25 = 8, c 35 = 6, c 45 = 20, c ij = c ji,
c ij = ∞ for all value of i and j not given in the data. Find the optimum sequence of
products in order to minimize the total set up cost.
3. Solve the travelling salesman problem given by the following data c12 = 4,
c13 = 7, c14 = 3, c23 = 6, c24 = 3, and c34 = 7 where cij = c ji.
4. Solve the following travelling salesman problem.

1 2 3 4 5 6

1 ∞ 20 23 27 29 34

2 21 ∞ 19 26 31 24

3 26 28 ∞ 15 36 26

4 25 16 25 ∞ 23 18

5 23 40 23 31 ∞ 10

6 27 18 12 35 16 ∞

A nswers
1. A→ C→ D→ B→ E→ A and A → E → B → D → C → A.
In both case total distance covered = 30 i.e. 3000 miles.

2. 1 → 2 → 5 → 3 → 4 → 1, Minimum total setup cost = ` 47

3. 1→ 3 → 2 → 4 → 1 , Minimum cost = ` 19

4. 1 → 5 → 6 → 3 → 4 → 2 → 1, Minimum cost = ` 103

mmm