Contents

Preface to the Sixth Edition v

Preface to the First Edition vi

Chapter 1 Operations Research: An Introduction 1–24

1.1 Operations Research – A Quantitative Approach to Decision-Making 2
1.2 The History of Operations Research 2
1.3 Definitions of Operations Research 4
1.4 Features of Operations Research Approach 5
1.5 Operations Research Approach to Problem Solving 6
Conceptual Questions A 7
1.6 Models and Modelling in Operations Research 7
 Classification Based on Structure 8  Classification Based on Function (or Purpose) 10
 Classification Based on Time Reference 10  Classification Based on Degree of Certainty 10
 Classification Based on Method of Solution or Quantification 11
1.7 Advantages of Model Building 11
1.8 Methods for Solving Operations Research Models 11
1.9 Methodology of Operations Research 12
1.10 Advantages of Operations Research Study 14
1.11 Opportunities and Shortcomings of the Operations Research Approach 14
1.12 Features of Operations Research Solution 15
1.13 Applications of Operations Research 15
1.14 Operations Research Models in Practice 16
1.15 Computer Software for Operations Research 17
Conceptual Questions B 18
Chapter Summary 19
Chapter Concepts Quiz 19
Case Study 20
Puzzles in Operations Research 22

Chapter 2 Linear Programming: Applications and Model Formulation 25–67

2.1 Introduction 26
2.2 Structure of Linear Programming Model 26
 General Structure of an LP Model 26  Assumptions of an LP Model 27
2.3 Advantages of Using Linear Programming 27
2.4 Limitations of Linear Programming 27
2.5 Application Areas of Linear Programming 28
2.6 General Mathematical Model of Linear Programming Problem 29
2.7 Guidelines on Linear Programming Model Formulation 30
2.8 Examples of LP Model Formulation 30
 Examples on Production 30  Examples on Marketing 41  Examples on Finance 43
 Examples on Agriculture 49  Example on Transportation 51  Examples on Personnel 53
Conceptual Questions 55
Self Practice Problems 56
Hints and Answers 61
viii Contents

Chapter Summary 64
Chapter Concepts Quiz 65
Case Study 66

Chapter 3 Linear Programming: The Graphical Method 68–99

3.1 Introduction 69
3.2 Important Definitions 69
3.3 Graphical Solution Methods of LP Problems 69
 Extreme Point Solution Method 70  Examples on Maximization LP Problem 70
 Examples on Minimization LP Problem 75  Examples on Mixed Constraints LP Problem 78
 Iso-Profit (Cost) Function Line Method 86  Comparison of Two Graphical Solution Methods 87
3.4 Special Cases in Linear programming 87
 Alternative (or Multiple) Optimal Solutions 87  Unbounded Solution 88
 Infeasible Solution 90  Redundancy 92
Conceptual Questions 92
Self Practice Problems 92
Hints and Answers 96
Chapter Summary 97
Chapter Concepts Quiz 97
Case Study 98

Chapter 4 Linear Programming: The Simplex Method 100–144

4.1 Introduction 101
4.2 Standard form of an LP Problem 101
4.3 Simplex Algorithm (Maximization Case) 103
4.4 Simplex Algorithm (Minimization Case) 112
 Two-Phase Method 114  Big-M Method 119
Self Practice Problems A 127
Hints and Answers 130
4.5 Some Complications and Their Resolution 131
 Unrestricted Variables 131  Tie for Entering Basic Variable (Key Column) 134
 Tie for Leaving Basic Variable (Key Row) – Degeneracy 134
4.6 Types of Linear Programming Solutions 135
 Alternative (Multiple) Optimal Solutions 136
 Unbounded Solution 137  Infeasible Solution 138
Conceptual Questions 139
Self Practice Problems B 139
Hints and Answers 141
Chapter Summary 142
Chapter Concepts Quiz 142
Case Study 143

Chapter 9 Transportation Problem 256–309

9.1 Introduction 257
9.2 Mathematical Model of Transportation Problem 257
 General Mathematical Model of Transportation Problem 258
9.3 The Transportation Algorithm 259
9.4 Methods for Finding Initial Solution 259
 North-West Corner Method (NWCM) 259  Least Cost Method (LCM) 260
 Vogel’s Approximation Method (VAM) 262
Conceptual Questions A 265
Self Practice Problems A 265
Hints and Answers 265
9.5 Test for Optimality 266
 Dual of Transportation Model 266  Economic Interpretation of ui’s and vj’s 267
 Steps of MODI Method (Transportation Algorithm) 268
xii Contents Contents xi

 Close-Loop in Transportation Table and its Properties 269

Conceptual Questions B 278
Self Practice Problems B 278
Hints and Answers 280
9.6 Variations in Transportation Problem 280
 Unbalanced Supply and Demand 280  Degeneracy and its Resolution 283
 Alternative Optimal Solutions 287  Prohibited Transportation Routes 290
9.7 Maximization Transportation Problem 294
9.8 Trans-Shipment Problem 296
Conceptual Questions C 298
Self Practice Problems C 298
Hints and Answers 302
Chapter Summary 304
Chapter Concepts Quiz 304
Case Study 305
Appendix: Theorems and Results 307

Chapter 10 Assignment Problem 310–338

10.1 Introduction 311
10.2 Mathematical Models of Assignment Problem 311
10.3 Solution Methods of Assignment Problem 312
 Hungarian Method for Solving Assignment Problem 312
Conceptual Questions A 318
Self Practice Problems A 318
Hints and Answers 320
10.4 Variations of the Assignment Problem 320
 Multiple Optimal Solutions 320  Maximization Case in Assignment Problem 320
 Unbalanced Assignment Problem 323  Restrictions on Assignments 323
Conceptual Questions B 327
Self Practice Problems B 327
Hints and Answers 329
10.5 A Typical Assignment Problem 330
10.6 Travelling Salesman Problem 331
Self Practice Problems C 334
Hints and Answers 335
Chapter Summary 335
Chapter Concepts Quiz 335
Case Study 337
Appendix: Important Results and Theorems 338

Chapter 13 Project Management: PERT and CPM 417–473

13.1 Introduction 418
13.2 Basic Differences Between PERT and CPM 418
 Significance of Using PERT/CPM 418
13.3 Phases of Project Management 419
13.4 PERT/CPM Network Components and Precedence Relationships 420
 Rules for AOA Network Construction 422  Errors and Dummies in Network 423
Conceptual Questions A 426
Self Practice Problems A 426
13.5 Critical Path Analysis 428
 Forward Pass Method (For Earliest Event Time) 428
 Backward Pass Method (For Latest Allowable Event Time) 429
 Float (Slack) of an Activity and Event 429  Critical Path 430
Conceptual Questions B 434
Self Practice Problems B 434
 Contents xiii

Hints and Answers 437

13.6 Project Scheduling with Uncertain Activity Times 437
 Estimation of Project Completion Time 438
Conceptual Questions C 441
Self Practice Problems C 441
Hints and Answers 445
13.7 Project Time-Cost Trade-Off 445
 Project Crashing 445  Time-Cost Trade-Off Procedure 445
Self Practice Problems D 454
Hints and Answers 457
13.8 Updating of the Project Progress 458
13.9 Resource Allocation 459
 Resource Levelling 459  Resource Smoothing 459
Self Practice Problems E 469
Chapter Summary 470
Chapter Concepts Quiz 471
Case Study 472

Chapter 20 Sequencing Problems 708–725

20.1 Introduction 709
20.2 Notations, Terminology and Assumptions 709
20.3 Processing n Jobs Through Two Machines 710
 Johnson Procedure 710
Conceptual Questions A 715
Self Practice Problems A 715
Hints and Answers 716
20.4 Processing n Jobs Through Three Machines 716
 The Procedure 716
Self Practice Problems B 718
Hints and Answers 719
20.5 Processing n Jobs Through m Machines 719
20.6 Processing Two Jobs Through m Machines 721
Conceptual Questions B 724
Self Practice Problems B 724
Hints and Answers 724
Chapter Summary 724
Chapter Concepts Quiz 725
 C h a p t e r 1

“The first rule of any technology used in a business is that automation applied to an efficient
operation will magnify the efficiency. The second is that automation applied to an inefficient
operation will magnify the inefficiency.”
– Bill Gates

This chapter presents a framework of a possible structural analysis of problems pertaining to an

organization in order to arrive at an optimal solution using operations research approach.

LEARNING OBJECTIVES

After reading this chapter you should be able to

 understand the need of using operations research – a quantitative approach for effective decision-
making.
 know the historical perspective of operations research approach.
 know various definitions of operations research, its characteristics and various phases of scientific
study.
 recognize, classify and use of various models for solving a problem under consideration.
 be familiar with several computer software available for solving an operations research model.

CHAPTER OUTLINE
1.1 Operations Research – A Quantitative 1.9 Methodology of Operations Research
Approach to Decision-Making 1.10 Advantages of Operations Research Study
1.2 The History of Operations Research 1.11 Opportunities and Shortcomings of the
1.3 Definitions of Operations Research Operations Research Approach
1.4 Features of Operations Research Approach 1.12 Features of Operations Research Solution
1.5 Operations Research Approach to Problem 1.13 Applications of Operations Research
Solving 1.14 Operations Research Models in Practice
 Conceptual Questions A 1.15 Computer Software for Operations Research
1.6 Models and Modelling in Operations  Conceptual Questions B
Research
 Chapter Summary
1.7 Advantages of Model Building
 Chapter Concepts Quiz
1.8 Methods for Solving Operations Research
 Case Study
Models
 Puzzles in Operations Research
2 Operations Research: Theory and Applications

1.1 OPERATIONS RESEARCH – A QUANTITATIVE PERSPECTIVE TO

DECISION-MAKING
Knowledge, innovations and technology are changing and hence decision-making in today’s social and
business environment has become a complex task due to little or no precedents. High cost of technology,
materials, labour, competitive pressures and so many economic, social, political factors and viewpoints, have
greatly increase the complexity of managerial decision-making. To effectively address the broader tactical
and strategic issues, also to provide leadership in the global business environment, decision-makers cannot
afford to make decisions based on their personal experiences, guesswork or intuition, because the
consequences of wrong decisions can prove to be serious and costly. For example, entering the wrong
Quantitative markets, producing the wrong products, providing inappropriate services, etc., may cause serious financial
analysis is the problems for organizations. Hence, an understanding of the use of quantitative methods to decision-making
scientific approach
to decision-making. is desirable to decision-makers.
Few decision-makers claim that the OR approach does not adequately meet the needs of business and
industry. Lack of implementation of findings is one of the major reasons. Among the reasons for
implementation failure is due to creative problem solving inabilities of the decision-maker. The implementation
process presumes that the definition, analysis, modeling, and solution phases of a project have been
performed as per prescribed guidelines.
Operations research approach helps in the comparison of all possible alternatives (courses of action or
acts) with respect to their potential outcomes and then sensitivity analysis of the solution to changesor
errors in numerical values. However, this approach (or technique) is an aid to the decision-makers’s
judgement not a substitute for it.
While attempting to solve a real-life problem, the decision-maker must examine the given problem from
both quantitative as well as qualitative perspective. For example, consider the problem of investments in three
alternatives: Stock Market, Real Estate and Bank Deposit. To arrive at any decision, the investor needsto
examine certain quantitative factors such as financial ratios from the balance sheets of companies whose
stocks are under consideration; real estate companies’ cash flows and rates of return for investment in
property; and how much investment will be worth in the future when deposited at a bank at a given interest
rate for a certain number of years? Also, certain qualitative factors, such as weather conditions, state and
central policies, new technology, political situation, etc.?
The evaluation of each alternative can be extremely difficult or time consuming for two reasons: First,
the amount and complexity of information that must be processed, and second the availability of large number
of alternative solutions. For these reasons, decision-makers increasingly turn to quantitative factors
and use computers to arrive at the optimal solution for problems.
Decision maker There is a need for structural analysis using operations research/quantitative techniques to arrive at
should consider a holistic solution to any managerial problem. This can be done by critically examining the levels of
both qualitative and
quantitative factors interaction between the application process of operations research, and various systems of an organization.
while solving a Figure 1.1 summarizes the conceptual framework of organizational/management structures and operations
problem. research application process.
This book introduces a set of operations research techniques that should help decision-makers in making
rational and effective decisions. It also gives a basic knowledge of the use of computer software needed for
computational purposes.

1.2 THE HISTORY OF OPERATIONS RESEARCH

It is generally agreed that operations research came into existence as a discipline during World War II when
Operations
research approach there was a critical need to manage scarce resources. However, a particular model and technique of OR can
considers be traced back as early as in World War I, when Thomas Edison (1914–15) made an effort to use a tactical
environmental game board for finding a solution to minimize shipping losses from enemy submarines, instead of risking
influences along with ships in actual war conditions. About the same time AK Erlang, a Danish engineer, carried out experiments
both organization
structures, and to study the fluctuations in demand for telephone facilities using automatic dialling equipment. Such
managerial experiments were later on used as the basis for the development of the waiting-line theory.
behaviour, for Since World War II involved strategic and tactical problems that were highly complicated, to expect
decision-making.
adequate solutions from individuals or specialists in a single discipline was unrealistic. Thus, groups of
 Operations Research: An Introduction 3

Fig. 1.1
Conceptual
Framework of an
Organization and
OR Application
Process

individuals who were collectively considered specialists in mathematics, economics, statistics and probability
theory, engineering, behavioural, and physical science, were formed as special units within the armed forces,
in order to deal with strategic and tactical problems of various military operations.
Such groups were first formed by the British Air Force and later the American armed forces formed similar
groups. One of the groups in Britain came to be known as Blackett’s Circus. This group, under the leadership
of Prof. P M S Blackett was attached to the Radar Operational Research unit and was assigned the problem
of analyzing the coordination of radar equipment at gun sites. Following the success of this group similar
mixed-team approach was also adopted in other allied nations.
After World War II, scientists who had been active in the military OR groups made efforts to apply
the operations research approach to civilian problems related to business, industry, research, etc. The
following three factors are behind the appreciation for the use of operations research approach:
(i) The economic and industrial boom resulted in mechanization, automation and decentralization of
operations and division of management functions. This industrialization resulted in complex
managerial problems, and therefore the application of operations research to managerial decision-
making became popular.
(ii) Continued research after war resulted in advancements in various operations research techniques.
The term
In 1947, G B Dantzig, developed the concept of linear programming, the solution of which is found ‘operations
by a method known as simplex method. Besides linear programming, many other techniques of OR, research’ was
such as statistical quality control, dynamic programming, queuing theory and inventory theory were coined as a result
well-developed before 1950’s. of conducting
research on military
(iii) The use of high speed computers made it possible to apply OR techniques for solving real-life operations during
decision problems. World War II
During the 1950s, there was substantial progress in the application of OR techniques for civilian
problems along with the professional development. Many colleges/schools of engineering, public
administration, business management, applied mathematics, computer science, etc. Today, however, service
organizations such as banks, hospitals, libraries, airlines, railways, etc., all recognize the usefulness of OR in
improving efficiency. In 1948, an OR club was formed in England which later changed its name to the
4 Operations Research: Theory and Applications

Operational Research Society of UK. Its journal, OR Quarterly first appeared in 1950. The Operations
Research Society of America (ORSA) was founded in 1952 and its journal, Operations Research was first
published in 1953. In the same year, The Institute of Management Sciences (TIMS) was founded as an
international society to identify, extend and unify scientific knowledge pertaining to management. Its journal,
Management Science, first appeared in 1954.
In India, during same period, Prof R S Verma set up an OR team at Defence Science Laboratory for
solving problems of store, purchase and planning. In 1953, Prof P C Mahalanobis established an OR team
in the Indian Statistical Institute, Kolkata for solving problems related to national planning and survey. The
A key person in the OR Society of India (ORSI) was founded in 1957 and it started publishing its journal OPSEARCH 1964
post-war onwards. In the same year, India along with Japan, became a member of the International Federation of
development of OR
was George B Operational Research Societies (IFORS) with its headquarters in London. The other members of IFORS were
Dantzig UK, USA, France and West Germany.
A year later, project scheduling techniques – Program Evaluation and Review Technique (PERT) and
Critical Path Method (CPM) were developed for scheduling and monitoring lengthy, complex and expensive
projects of that time.
The American Institute for Decision Sciences came into existence in 1967. It was formed to promote,
develop and apply quantitative approach to functional and behavioural problems of administration. It
started publishing a journal, Decision Science, in 1970.
Because of OR’S multi-disciplinary approach and its application in varied fields, it has a bright future,
provided people devoted to the study of OR help to meet the needs of society. Some of the problems in
In India, operations the area of hospital management, energy conservation, environmental pollution, etc., have been solved by
research came into OR specialists. This is an indication of the fact that OR can also contribute towards the improvement of
existence in 1949,
when an OR unit the social life and of areas of global need.
was established at
Regional Research 1.3 DEFINITIONS OF OPERATIONS RESEARCH
Laboratory,
Hyderabad, for
The wide scope of applications of operations research encouraged various organizations and individuals
planning and
organizing research. to define it as follows:
 Operations research is the application of the methods of science to complex problems in the
direction and management of large systems of men, machines, materials and money in industry,
business, government and defence. The distinctive approach is to develop a scientific model of
the system incorporating measurements of factors such as chance and risk, with which to predict
and compare the outcomes of alternative decisions, strategies or controls. The purpose is to help
management in determining its policy and actions scientifically.
– Operational Research Society, UK
 The application of the scientific method to the study of operations of large complex organizations
or activities. It provides top level administrators with a quantitative basis for decisions that will
increase the effectiveness of such organizations in carrying out their basic purposes.
– Committee on OR National Research Council, USA
The definition given by Operational Research Society of UK has been criticized because of the emphasis it
places on complex problems and large systems, leaving the reader with the impression that it is a highly
technical approach suitable only to large organizations.
Operations  Operations research is the systematic application of quantitative methods, techniques and tools
research is
to the analysis of problems involving the operation of systems.
concerned with
scientifically deciding – Daellenbach and George, 1978
how to best design  Operations research is essentially a collection of mathematical techniques and tools which in
and operate man- conjunction with a system’s approach, are applied to solve practical decision problems of an
machine systems
economic or engineering nature. – Daellenbach and George, 1978
that usually require
the allocation of These two definitions imply another view of OR – being the collection of models and methods that have been
scarce resources.
developed largely independent of one another.
 Operations research utilizes the planned approach (updated scientific method) and an
interdisciplinary team in order to represent complex functional relationships as mathematical
models for the purpose of providing a quantitative basis for decision-making and uncovering
new problems for quantitative analysis. – Thierauf and Klekamp, 1975
 Operations Research: An Introduction 5

 This new decision-making field has been characterized by the use of scientific knowledge through
interdisciplinary team effort for the purpose of determining the best utilization of limited
resources. – H A Taha, 1976
These two definitions refer to the interdisciplinary nature of OR. However, there is nothing that can stop
one person from considering several aspects of the problem under consideration.
 Operations research, in the most general sense, can be characterized as the application of scientific

methods, techniques and tools, to problems involving the operations of a system so asto provide
those in control of the operations with optimum solutions to the problems.
– Churchman, Ackoff and Arnoff, 1957
This definition refers operations research as a technique for selecting the best course of action out of the
several courses of action available, in order to reach the desirable solution of the problem.
 Operations research has been described as a method, an approach, a set of techniques, a team Operations
activity, a combination of many disciplines, an extension of particular disciplines (mathematics, research is the art
of winning wars
engineering, economics), a new discipline, a vocation, even a religion. It is perhaps some of all without actually
these things. – S L Cook, 1977 fighting them.
 Operations research may be described as a scientific approach to decision-making that involves the
operations of organizational system. – F S Hiller and G J Lieberman, 1980
 Operations research is a scientific method of providing executive departments with a quantitative
basis for decisions regarding the operations under their control.
– P M Morse and G E Kimball, 1951
 Operations research is applied decision theory. It uses any scientific, mathematical, or logical
means to attempt to cope with the problems that confront the executive, when he tries to achieve
a thorough-going rationality in dealing with his decision problems.
– D W Miller and M K Star, 1969 Operations
 Operations research is a scientific approach to problem-solving for executive management. research is the art
of finding bad
– H M Wagner answers to
As the discipline of operations research grew, numerous names such as Operations Analysis, Systems problems which
Analysis, Decision Analysis, Management Science, Quantitative Analysis, Decision Science were given otherwise have
to it. This is because of the fact that the types of problems encountered are always concerned with worse answers.
‘effective decision’.

1.4 FEATURES OF OPERATIONS RESEARCH APPROACH

OR utilizes a planned approach following a scientific method and an interdisciplinary team, in order to
represent complex functional relationship as mathematical models, for the purpose of providing a
quantitative basis for decision-making and uncovering new problems for quantitative analysis. This
definition implies additional features of OR approach. The broad features of OR approach in solving any
decision problem are summarized as follows:
Interdisciplinary approach For solving any managerial decision problem often an interdisciplinary
teamwork is essential. This is because while attempting to solve a complex management problem, one person
may not have the complete knowledge of all its aspects such as economic, social, political, psychological, Operations research
engineering, etc. Hence, a team of individuals specializing in various functional areas of management should uses:
(i) interdisciplinary,
be organized so that each aspect of the problem can be analysed to arrive at a solution acceptable to all
(ii) scientific,
sections of the organization. (iii) holistic, and
(iv) objective-
Scientific approach Operations research is the application of scientific methods, techniques and tools oriented
to problems involving the operations of systems so as to provide those in control of operations with optimum approaches to
solutions to the problems (Churchman et al.). The scientific method consists of observing and defining the decision making
problem; formulating and testing the hypothesis; and analysing the results of the test. Thedata so obtained
is then used to decide whether the hypothesis should be accepted or not. If the hypothesis is accepted, the
results should be implemented, otherwise not.
Holistic approach While arriving at a decision, an operations research team examines the relative
importance of all conflicting and multiple objectives. It also examines the validity of claims of various
departments of the organization from the perspective of its implications to the whole organization.
6 Operations Research: Theory and Applications

Objective-oriented approach An operations research approach seeks to obtain an optimal solution

to the problem under analysis. For this, a measure of desirability (or effectiveness) is defined, basedon the
objective(s) of the organization. A measure of desirability so defined is then used to compare alternative
courses of action with respect to their possible outcomes.
Illustration The OR approach attempts to find a solution acceptable to all sections of the organizations. One
such situation is described below.
A large organization that has a number of management specialists is faced with the basic problem of
maintaining stocks of finished goods. To the marketing manager, stocks of a large variety of products are
purely a means of supplying the company’s customers with what they want and when they want it. Clearly,
according to a marketing manager, a fully stocked warehouse is of prime importance to the company. But
the production manager argues for long production runs, preferably on a smaller product range, particularly
Operations research if a significant amount of time is lost when production is switched from one variety to another. The result
attempts to resolve would again be a tendency to increase the amount of stock carried but it is, of course, vital that the plant
the conflicts of
interest among should be kept running. On the other hand, the finance manager sees stocks in terms of capital that is
various sections of unproductively tied up and argues strongly for its reduction. Finally, there appears the personnel manager for
the organization and whom a steady level of production is advantageous for having better labour relations. Thus, all these people
seeks to find the
optimal solution that would claim to uphold the interests of the organization, but they do so only from their own specialized points
is in the interest of of view. They may come up with contradictory solutions and obviously, all of them cannot be right.
the organization as In view of such a problem that involves every section of an organization, the decision-maker, irrespective
a whole.
of his/her specialization, may require to seek assistance from OR professionals.
Remark A system is defined as an arrangement of components designed to achieve a particular objective
or objectives according to plan. The components may either be physical or conceptual or both, but they
all share a unique relationship with each other and with the overall objective of the system.

1.5 OPERATIONS RESEARCH APPROACH TO PROBLEM SOLVING

The most important feature of operations research is the use of the scientific method and the building of
decision models. The operations research approach to problem solving is based on three phases, namely
(i) Judgement Phase; (ii) Research Phase, and (iii) Action Phase.
Judgement phase This phase includes: (i) identification of the real-life problem, (ii) selection of an
appropriate objective and the values of various variables related to this objective, (iii) application of the
appropriate scale of measurement, i.e. deciding the measures of effectiveness (desirability), and (iv)
formulation of an appropriate model of the problem and the abstraction of the essential information, so that
a solution to the decision-maker’s goals can be obtained.
Research phase This phase is the largest and longest amongst all the phases. However, even though
the remaining two are not as long, they are also equally important as they provide the basis fora scientific
method. This phase utilizes: (i) observations and data collection for a better understanding ofthe problem,
(ii) formulation of hypothesis and model, (iii) observation and experimentation to test the hypothesis on the
basis of additional data, (iv) analysis of the available information and verification of the hypothesis using
Operations research pre-established measures of desirability, (v) prediction of various results from the hypothesis, and (iv)
approach to
decision making is generalization of the result and consideration of alternative methods.
based on three
phases: Action phase This phase consists of making recommendations for implementing the decision. This
(i) Judgement, decision is implemented by an individual who is in a position to implement results. This individual must
(ii) Research, and be aware of the environment in which the problem occurred, be aware of the objective, of assumptions behind
(iii) Action
the problem and the required omissions of the model.
 Operations Research: An Introduction 7

CONCEPTUAL QUESTIONS A
1. Briefly trace the history of operations research. How did operations 17. Comment on the following statements:
research develop after World War II? (a) OR is the art of winning war without actually fighting it.
2. Is operations research (i) a discipline (ii) a profession (iii) a set (b) OR is the art of finding bad answers where worse exist.
of techniques (iv) a philosophy or a new name for old things? (c) OR replaces management by personality.
Discuss. [Delhi Univ., MBA (HCA), 2006]
18. Explain critically the limitations of any three definitions of OR as
3. Discuss the following:
you understand them.
(a) OR as an interdisciplinary approach.
19. Discuss the significance and scope of operations research in
(b) Scientific method in OR.
modern management. [Delhi Univ., MBA (HCA), 2005]
(c) OR as more than a quantitative analysis of the problem.
20. Quantitative techniques complement the experience and
4. What are the essential characteristics of operations research? judgement of an executive in decision-making. They do not and
Mention different phases in an operations research study. Point out cannot replace it. Discuss. [Delhi Univ., MBA, 2004]
its limitations, if any. 21. Explain the difference between scientific management and
5. Define operations research as a decision-making science. operations research.
(a) Give the main characteristics of OR. 22. Decision-makers are quick to claim that quantitative analysis talks
(b) Discuss the scope of OR. to them in a jargon that does not sound like English. Listfour
6. (a) Outline the broad features of the judgement phase and the terms that might not be understood by a manager. Then explain, in
research phase of the scientific method in OR. Discuss in non-technical terms, what each term means.
detail any one of these phases. 23. It is said that operations research increases the creativecapabilities
(b) What are various phases of the OR problem? Explain them of a decision-maker. Do you agree with this view? Defend your
briefly. point of view with examples.
(c) State the phases of OR study and their importance in solving 24. (a) Why is the study of operations research important to the
problems. decision-maker?
7. What is the role of operations research in decision-making? (b) Operations research increases creative and judicious capa-
bilities of a decision-maker. Comment. [Delhi Univ., MBA, 2007]
[Punjab, BE (Mech. Engg.), 2000]
25. (a) ‘Operations research is the application of scientific methods,
8. Define operations research. Explain critically the limitations of techniques and tools to problems involving the operations of
various definitions as you understand them. a system so as to provide those in control of the system with
9. Give any three definitions of operations research and explain optimum solutions to the problem.’ Discuss.
them. (b) ‘Operations research is an aid for the executive in making
10. (a) Discuss various phases of solving an OR problem. his decisions by providing him with the needed quantitative
(b) State phases of an OR study and their importance in solving information, based on the scientific method analysis.’ Discuss
problems. this statement giving examples of OR methods that you
11. Discuss the role and scope of quantitative methods for scientific know.
decision-making in a business environment. 26. Comment on the following statements.
[Delhi Univ., MBA, 2003] (a) OR is a bunch of mathematical techniques.
(b) OR advocates a system’s approach and is concerned with
12. Discuss the advantage and limitations of operations research.
optimization. It provides a quantitative analysis for decision-
[AMIE, 2004 ] making.
13. Write a critical essay on the definition and scope of operations (c) OR has been defined semi-facetiously as the application of
research. [Meerut, MSc (Maths), 2001] big minds to small problems.
14. What post-World War II factors were important in the development 27. Discuss the points to justify the fact that the primary purpose of
of operations research? operations research is to resolve the conflicts resulting from the
15. Does the fact that OR takes the organizational point of viewinstead various subdivisions of the functional areas like production,
of the individual problem-centered point of view generate marketing, finance and personnel in an optimal, near optimal or
constraints on its increased usage? satisfying way.
28. How far can quantitative techniques be applied in management
16. What were the significant characteristics of OR applications during
decision-making? Discuss, in detail, with special reference to any
World War II? What caused the discipline of OR to takeon these
functional area of management pointing out their limitations, if any.
characteristics during that period?

1.6 MODELS AND MODELLING IN OPERATIONS RESEARCH

Models do not, and cannot, represent every aspect of a real-life problem/system because of its large and
changing characteristics. However, a model can be used to analyze, understand and describe certain aspects
(key features) of a system for the purpose of improving its performance as well as to examine changes (if
any) without disturbing the ongoing operations. For example, to study the flow of material in a manufacturing
firm, a scaled diagram (descriptive model) on paper showing the factory floor, position of equipment, tools,
and workers can be constructed. It would not be necessary to give details such as the colour of machines,
the heights of the workers, or the temperature of the building.
8 Operations Research: Theory and Applications

The key to model-building lies in abstracting only the relevant variables that affect the criteria of the
measures-of-performance of the given system and in expressing the relationship in a suitable form. However,
a model should be as simple as possible so as to give the desired result. On the other hand, oversimplifying
the problem can also lead to a poor decision. Model enrichment is done by changing value of variables, and
relaxing assumptions. The essential three qualities of any model are:
 Validity of the model – model should represent the critical aspects of the system/problem under
study,
 Usability of the model – a model can be used for the specific purposes, and
 Value of the model to the user.
Besides these three qualities, other consideration of interest are, (i) cost of the model and its sophistication,
(ii) time involved in formulating the model, etc.
An informal definition of model that applies to all of us is a tool for thinking and understanding features
A model is an of any problem/system before taking action. For example, a model tends to be formulated when (a) we think
approximation or
abstraction of reality about what someone will say in response to our act(s), (b) we try to decide how to spend our money, or
which considers (c) we attempt to predict the consequences of some activity (either ours, someone else’s or even a natural
only the essential event). In other words, we would not be able to derive or take any purposeful action if we do not form a model
variables (or of the activity first. OR approach uses this natural tendency to create models. This tendency forces to think
factors) and
parameters in the more rigorously and carefully about the models we intend to use.
system along with In general models are classified in eight categories as shown in Table 1.1. Such a classification provides a useful
their relationships. frame of reference for practioners/researchers.

1. Function 4. Degree of certainty 7. Degree of closure

 Descriptive  Certainty  Closed

 Predictive  Conflict  Open

 Normative  Risk 8. Degree of quantification

2. Structure  Uncertainty  Qualitative

 Iconic 5. Time reference  Mental

 Analog  Static  Verbal

Table 1.1  Symbolic  Dynamic  Quantitative

General 3. Dimensionality 6. Degree of generality  Statistical

Classification of  Two-dimensional  Specialized  Heuristic

Models  Multidimensional  General  Simulation

A summary classification of OR models based on different criteria is given in Fig. 1.2 and few of these
classifications are discussed below:

1.6.1 Classification Based on Structure

Physical models These models are used to represent the physical appearance of the real object under study,
either reduced in size or scaled up. Physical models are useful only in design problems because they are easy
to observe, build and describe. For example, in the aircraft industry, scale models of a proposed
new aircraft are built and tested in wind tunnels to record the stresses experienced by the air frame.
Physical model Physical models cannot be manipulated and are not very useful for prediction. Problems such as portfolio
represents the selection, media selection, production scheduling, etc., cannot be analysed with the help of these models.
physical appearance
of the real object
Physical models are classified into two categories.
under study, either (i) Iconic Models An iconic model is a scaled (small or big in size) version of the system. Such models
reduced in size or retain some of the physical characteristics of the system they represent.
scaled up.
Examples of iconic model are, blueprints of a home, maps, globes, photographs, drawings, air planes, trains, etc.
An iconic model is used to describe the characteristics of the system rather than explaining the system.
This means that such models are used to represent system’s characteristics that are not used in determining or
predicting effects that take place due to certain changes in the actual system. For example, (i) colour of an atom
does not play any vital role in the scientific study of its structure, (ii) type of engine in a car has no role to play
in the study of the problem of parking, etc.
(ii) Analogue Models An analogue model does not resemble physically the system they represent, but
retain a set of characteristics of the system. Such models are more general than iconic models and can
also be manipulated.
 Operations Research: An Introduction 9

For example, (i) oil dipstick in a car represents the amount of oil in the oil tank; (ii) organizational chart
represents the structure, authority, responsibilities and relationship, with boxes and arrows; (iii) maps in
different colours represent water, desert and other geographical features, (iv) Graphs of time series, stock-
market changes, frequency curves, etc., represent quantitative relationships between any two variables and
predict how a change in one variable effects the other, and so on.

Fig. 1.2
Classification of
Models

Symbolic models These models use algebraic symbols (letters, numbers) and functions to represent Mathematical
variables and their relationships for describing the properties of the system. Such relationships can also model uses
mathematical
be represented in a physical form. Symbolic models are precise and abstract and can be analysed by using
equations and
laws of mathematics. statements to
Symbolic models are classified into two categories. represent the
(i) Verbal Models These models describe properties of a system in written or spoken language. Written relationships within
sentences, books, etc., are examples of a verbal model. the model.
(ii) Mathematical Models These models use mathematical symbols, letters, numbers and mathematical
operators (+, –, ÷, ×) to represent relationships among variables of the system to describe its
properties or behaviour. The solution to such models is obtained by applying suitable
mathematical technique.
Few examples of mathematical model are (i) the relationship among velocity, distance and acceleration,
(ii) the relationship among cost-volume-profit, etc.
10 Operations Research: Theory and Applications

1.6.2 Classification Based on Function (or Purpose)

Descriptive models These models are used to investigate the outcomes or consequences of various
alternative courses of action (strategies, or actions). Since these models evaluate the consequence based
on a given condition (or alternative) rather than on all other conditions, there is no guarantee that an
alternative selected is optimal. These models are usually applied (i) in decision-making where optimization
models are not applicable, and (ii) when objective is to define the problem or to assess its seriousness rather
than to select the best alternative. These models are especially used for predicting the behaviour of a particular
system under various conditions.
Simulation is an example of a descriptive model for conducting experiments with the systems based
on given alternatives.
Predictive models These models represent a relationship between dependent and independent variables
and hence measure ‘cause and effect’ due to changes in independent variables.
These models do not have an objective function as a part of the model of evaluating decision alternatives
based on outcomes or pay off values. Also, through such models decision-maker does not attempt to choose
the best decision alternative, but can only have an idea about the possible alternatives available to him.
For example, the equation S = a + bA + cI relates dependent variable (S) with other independent variables
on the right hand side. This can be used to describe how the sale (S ) of a product changes witha change
in advertising expenditure (A) and disposable personal income (I ). Here, a, b and c are parameterswhose
values must be estimated. Thus, having estimated the values of a, b and c, the value of advertising expenditure
(A) can be adjusted for a given value of I, to study the impact of advertising on sales. Also, through such
models decision-maker does not attempt to choose the best decision alternative, but can only have an idea
about the possible alternative available to him.
Normative (or Optimization) models These models provide the ‘best’ or ‘optimal’ solution to problems
using an appropriate course of action (strategy) subject to certain limitations on the use of resources. For
example, in mathematical programming, models are formulated for optimizing the given objective function,
subject to restrictions on resources in the context of the problem under consideration and non-negativityof
variables. These models are also called prescriptive models because they prescribe what the decision- maker
ought to do.

1.6.3 Classification Based on Time Reference

Static models Static models represent a system at a particular point of time and do not take into account
changes over time. For example, an inventory model can be developed and solved to determine an economic
In a deterministic
model all values order quantity assuming that the demand and lead time would remain same throughout the planning period.
used are known
with certainty. Dynamic models Dynamic models take into account changes over time, i.e., time is considered as one of
the variables while deriving an optimal solution. Thus, a sequence of interrelated decisions over a period of
time are made to select the optimal course of action in order to achieve the given objective. Dynamic
programming is an example of a dynamic model.

1.6.4 Classification Based on Degree of Certainty

Deterministic models If all the parameters, constants and functional relationships are assumed to be
known with certainty when the decision is made, the model is said to be deterministic. Thus, the outcome
associated with a particular course of action is known, i.e. for a specific set of input values, there is only
one output value which is also the solution of the model. Linear programming models are example of
deterministic models.
Probabilistic (Stochastic) models If at least one parameter or decision variable is random (probabilistic or
stochastic) variable, then the model is said to be probabilistic. Since at least one decision variable is random,
In a Probabilistic
model all values the independent variable, which is the function of dependent variable(s), will also be random. This means
used are not known consequences (or payoff) due to certain changes in the independent variable(s) cannot be predicted with
with certainty; and certainty. However, it is possible to predict a pattern of values of both the variables by their probability
are often measured distribution.
as probability
values. Insurance against risk of fire, accidents, sickness, etc., are examples where the pattern of events is studied
in the form of a probability distribution.
 Operations Research: An Introduction 11

1.6.5 Classification Based on Method of Solution or Quantification

Heuristic models If certain sets of rules (may not be optimal) are applied in a consistent manner to facilitate
solution to a problem, then the model is said to be Heuristic.
Analytical models These models have a specific mathematical structure and thus can be solved by the
known analytical or mathematical techniques. Any optimization model (which requires maximization or
minimization of an objective function) is an analytical model.
Simulation models These models have a mathematical structure but cannot be solved by the known
mathematical techniques. A simulation model is essentially a computer-assisted experimentation on a
mathematical structure of a problem in order to describe and evaluate its behaviour under certain assumptions
over a period of time.
Simulation models are more flexible than mathematical models and can, therefore, be used to represent
a complex system that cannot be represented mathematically. These models do not provide general solution
like those of mathematical models.

1.7 ADVANTAGES OF MODEL BUILDING

Models, in general, are used as an aid for analysing complex problems. However, general advantages of
model building are as follows:
1. A model describes relationships between various variables (factors) present in a system more easily than
what is done by a verbal description. That is, models help the decision-maker to understand the system’s
structure or operation in a better way. For example, it is easier to represent a factory layout on paper than to
construct it. It is cheaper to try out modifications of such systems by rearrangement on paper.
2. The problem can be viewed in its entirety, with all the components being considered simultaneously.
3. Models serve as aids to transmit ideas among people in the organization. For example, a process chart
can help the management to communicate better work methods to workers.
4. A model allows to analyze and experiment on a complex system which would otherwise be impossible
on the actual system. For example, the experimental firing of INSAT satellite may cost millions of rupees
and require years of preparation.
5. Models considerably simplify the investigation and provide a powerful and flexible tool for predicting
the future state of the system (or process).

1.8 METHODS FOR SOLVING OPERATIONS RESEARCH MODELS

In general, the following three methods are used for solving OR models, where values of decision variables
are obtained that optimize the given objective function (a measure of effectiveness).

Analytical (or deductive) method In this method, classical optimization techniques such as calculus, In general,
operations research
finite difference and graphs are used for solving an OR model. The analytical methods are non- iterative models are solved
methods to obtain an optimal solution of a problem. For example, to calculate economic order quantity using either
(optimal order size), the analytical method requires that the first derivative of the mathematical expression (i) analytical,
(ii) numerical, or
TC = (D/Q) C0 + (Q/2) Ch (iii) Monte Carlo
where TC = total variable inventory cost; D = annual demand; method
C0 = ordering cost per order; Ch = carrying cost per time period;
Q = size of an order
be taken and equated to zero as the first step towards calculating, EOQ(Q) = 2 DC0 / Ch . This is because
of the concept of maxima and minima used for optimality.
Numerical (or iterative) method When analytical methods fail to obtain the solution of a particular
problem due to its complexity in terms of constraints or number of variables, a numerical (or iterative) method
is used to find the solution. In this method, instead of solving the problem directly, a general algorithm is
applied for obtaining a specific numerical solution.
The numerical method starts with a solution obtained by trial and error, and a set of rules for improving
it towards optimality. The solution so obtained is then replaced by the improved solution and the process
of getting an improved solution is repeated until such improvement is not possible or the cost of further
calculation cannot be justified.
12 Operations Research: Theory and Applications

Monte Carlo method This method is based upon the idea of experimenting on a mathematical modelby
inserting into the model specific values of decision variables for a selected period of time under different
conditions and then observing the effect on the criterion chosen. In this method, random samples of specified
random variables are drawn to know how the system is behaving for a selected period of time under different
conditions. The random samples form a probability distribution that represents the real-life system and from
this probability distribution, the value of the desired random variable can then be estimated.

1.9 METHODOLOGY OF OPERATIONS RESEARCH

Every OR specialist may have his/her own way of solving problems. However, the effective use of OR
techniques requires to follow a sequence of steps. Each of these steps is described in detail in Fig. 1.3:

A problem is the Step 1: Defining the problem Problem definition involves the process of identifying, understanding, and
statement that describing the problem being faced by an organization. It helps to define the objective(s) to be achieved, and
indicates an the alternative courses of action.
objective/goal that The procedure begins with gathering information (data) of the organizational structure, communication and
needs to be
achieved. control system, objectives and expectations. Such information will help in assessing the difficulty likely to be faced
in terms of costs, time, resources, probability of success of the study, etc.

Fig. 1.3
Methodology of
Operations
Research
The major components involved in problem formulation are as under:
Decision-maker’s Point of View The first component of the problem formulation is to know the pointof
view of the decision-maker who is not satisfied with the existing state of affairs. Such interaction withthe
decision-maker will help to understand whether he has already obtained the solution of the problemand
wants to retain it, or wants to improve it to next level of satisfaction. If the decision-maker has conflicting
multiple objectives, he may be advised to rank his objectives in order of preference; overlapping among
several objectives may be eliminated.

All view points Decision Environment It is desirable to know the details about resources (such as manpower, material,
should be money, etc.) that are required to attain objectives of the organization while keeping in mind the social
considered before norms in which the organization need to function. Knowledge of such factors will help in modifying the
defining the problem initial set of the decision-maker’s objectives.
to be solved in
order to arrive at an
Alternative Courses of Action If several courses of action are available for solution to a problem, then
optimal solution.
an exhaustive list of such courses of action should be prepared. Courses of action that are not feasible
with respect to objectives and resources may be ruled out.
 Operations Research: An Introduction 13

Measure of Effectiveness A measure of effectiveness (or performance) is required in order to evaluate the
merit of the several available courses of action. The effectiveness can be measured in different units such as
rupees (net profits), percentage (share of market desired) or time dimension (service or waiting time).

Step 2: Formulating a mathematical model After the problem is clearly defined and understood,
the next step is to collect required data and then formulate a mathematical model. Model formulation requires
to define relationships among decision variables. Certain basic components required in every decision
problem model are:

Controllable (Decision) Variables These are certain factors (or variables) associated with the problem
whose values are to be determined by solving the model. The possible values assigned to these variables
are called decision alternatives. For example, in queuing theory, the number of service facilities is the
decision variable.

Uncontrollable (Exogenous) Variables The values of these variables are not under the control of the Exogenous
variables are those
decision-maker and are also termed as state of nature. whose numerical
value depends upon
Objective Function It represents the criterion of evaluating alternative courses of action in terms of value of the external
decision variables so as to optimize (minimized or maximized) the desired performance. environment that
influences an
Policies and Constraints There are certain constraints (limitations) on the use of resources, and such organization.
constraints arise due to organizational policy, legal restraints or limited resources such as space, money,
manpower, material, etc. The constraints on the use of resources are expressed either in the form of equations
or inequalities.

Functional Relationships In a decision problem, the decision variables in the objective function and in the
constraints are connected by a specific functional relationship. A general decision problem model might take
the following form:
Optimize (Max. or Min.) Z = f (x)
subject to the constraints
gi (x) {, =, } bi ; i = 1, 2, . . ., m
and x0
where, x = a vector of decision variables (x1, x2, . . ., xn)
f (x) = criterion or objective function to be optimized
gi (x) = the ith constraint
bi = fixed amount of the ith resource
A model is referred to as a linear model if all functional relationships among decision variables
x1, x2, . . ., xn in f (x) and g (x) are linear. But if one or more of the relationships are non-linear, the model
is said to be a non-linear model.
Parameters These are constants in the functional relationships among decision variables. Parameters can Optimization
methods yield the
either be deterministic or probabilistic in nature. A deterministic parameter is one whose value is assumed best value for the
to occur with certainty. Otherwise, it is probabilistic. decision variables
both for
Step 3: Solving the mathematical model Once a mathematical model of the problem has been unconstrained and
formulated, the next step is to obtain numerical values of decision variables. Obtaining these values constrained
depends on the specific form or type of mathematical model. In general, the following two categories of problems.
methods are used for solving an OR model.
(i) Analytical Methods These methods are applied to solve both constrained and unconstrained
mathematical models. In constrained problems, value of decision variables satisfy all the constraints
simultaneously and provide an optimal value for the objective function.
However, in unconstrained problems, value of decision variables provide an acceptable value
for the objective function.
(ii) Heuristic Methods (also referred as rules of thumb which work). These methods are used when
obtaining optimal solution is either very time consuming or the model is too complex. A commonly
used heuristic is ‘stand in the shortest line’,
14 Operations Research: Theory and Applications

Step 4: Validating the solution After solving the mathematical model, it is important to review the
solution to see whether value of variables make sense and that the resulting decisions can be implemented.
Some of the reasons for validating the solution are:
(i) Mathematical model may not have considered all constraints present in study because few
constraints may either be omitted or incorrect.
(ii) Certain aspects of the problem may have been overlooked, omitted or simplified.
(iii) The data may have been incorrectly estimated or recorded.
Step 5: Implementing the solution Before implementing the solution, the decision-maker should ensure
to select alternatives that are capable of being implemented. Further, it is also important to ensure that any
solution implemented is continuously reviewed and updated in the light of a changing environment.In any
case, the decision-maker who is in the best position to implement solution (or result), must be aware of the
objective, assumption, omissions and limitations of the model.
Step 6: Modifying the model For a mathematical model to be useful, the degree to which it actually
represents the system or problem being modelled must be established. If during validation the solution cannot
be implemented, then identify the constraint(s) that were omitted during the mathematical model formulation
or find out whether some of the original constraints were incorrect. In all such cases, returnto the problem
formulation step and carefully make the appropriate modifications to represent the given problem more
accurately.
Heuristic Methods A model must be applicable for a reasonable time period and should be updated from time to time, taking
yield values of the into consideration the past, present and future aspects of the problem.
variables that satisfy
all the constraints, Step 7: Establishing controls over the solution The changes within the environment (society or
but do not business) influence the continuing validity of models and their solutions. Thus, a control procedure needs
necessarily provide
optimal solution.
to be developed for detecting significant changes in decision variables of the problem, without having to

build a model every time a significant change occurs.

1.10 ADVANTAGES OF OPERATIONS RESEARCH STUDY

In addition to learning about various OR models that are useful to arrive at an optimal decision, the knowledge of
operations research techniques can also help the decision-maker to attain several other advantages as listed
below.
1. Structured approach to problems: A substantial amount of time and effort can be saved in developing
and solving OR models if a logical and consistent approach is followed. This implies that the decision-
maker has to be careful while defining variables, availability of resources, and functional relationships
among variables in the objective function and constraints. This will also reduce the chance of conceptual
and computational errors. Any such error can also be detected easily and corrected at an early stage.
2. Critical approach to problem solving: The decision-maker will come to understand various components
of the problem and accordingly select a mathematical model for solving the given problem. He will
become aware of the explicit and implicit assumptions and limitations of such models. Problem solutions
are examined critically and the effect of any change and error in the problem data can be studied through
sensitivity analysis techniques.

1.11 OPPORTUNITIES AND SHORTCOMINGS OF THE OPERATIONS

RESEARCH APPROACH
The use of quantitative methods is appreciated to improve managerial decision-making. However, besides
certain opportunities OR approach has not been without its shortcomings. The main reasons for its failure are
due to unawareness on the part of decision-makers about their own role, as well as the avoidance of
behavioural/organizational issues while constructing a decision model. A few opportunities and shortcomings
of the OR approach are listed below.
Opportunities
The OR approach of problem solving facilitates.
 decision-maker to define explicitly his/her objectives, assumptions and constraints.
 decision-maker to identify variables that might influence decisions.
 Operations Research: An Introduction 15

 in identifying gaps in the data required to support solution to a problem.

 the use of computer software to solve a problem.
Shortcomings
 The solution to a problem is often derived by simplifying assumptions. Consequently, such solutions
are not optimal.
 Models may not represent the realistic situations in which decisions are made.
 Often the decision-maker is not fully aware of the limitations of the models.

1.12 FEATURES OF OPERATIONS RESEARCH SOLUTION

A solution that is expensive as compared to the potential savings from its implementation, should not be
considered. Even, if a solution is well within the budget but does not fulfill the objective is also not acceptable.
Few requirements of a good solution are:
1. Technically Appropriate: The solution should work technically, meet the constraints and operate in the
problem environment. A solution needs to
2. Reliable: The solution must be useful for a reasonable period of time under the conditions for which be: (i) technically
appropriate,
it was designed. (ii) reliable, and
3. Economically Viable: Its economic value should be more than what it costs to develop it and it should (iii) economically
be seen as a wise investment in hiring OR talent. viable

4. Behaviourally Appropriate: The solution should be behaviourally appropriate and must remain valid
for a reasonable period of time within the organization.

1.13 APPLICATIONS OF OPERATIONS RESEARCH

Some of the industrial/government/business problems that can be analysed by the OR approach have been
arranged by functional areas as follows:
Finance and Accounting
 Dividend policies, investment and portfolio  Break-even analysis, capital budgeting,
management, auditing, balance sheet and cash cost allocation and control, and financial
flow analysis planning
 Claim and complaint procedure, and public  Establishing costs for by-products and
accounting developing standard costs
Marketing
 Selection of product-mix, marketing and export  Sales effort allocation and assignment
planning  Launching a new product at the best possible
time
 Advertising, media planning, selection and  Predicting customer loyalty
effective packing alternatives
Purchasing, Procurement and Exploration
 Optimal buying and reordering with or  Replacement policies
without price quantity discount  Bidding policies
 Transportation planning  Vendor analysis
Production Management
Facilities planning
 Location and size of warehouse or new plant,  Logistics, layout and engineering design
distribution centres and retail outlets  Transportation, planning and scheduling
Manufacturing
 Aggregate production planning, assembly  Employment, training, layoffs and quality
line, blending, purchasing and inventory control
control  Allocating R&D budgets most effectively
16 Operations Research: Theory and Applications

Maintenance and project scheduling

 Maintenance policies and preventive  Maintenance crew size and scheduling
maintenance  Project scheduling and allocation of resources
Personnel Management
 Manpower planning, wage/salary  Negotiation in a bargaining situation
administration  Skills and wages balancing
 Designing organization structures more  Scheduling of training programmes to maximize
effectively skill development and retention
Techniques and General Management
 Decision support systems and MIS;  Making quality control more effective
forecasting  Project management and strategic planning
Government
 Economic planning, natural resources, social  Urban and housing problems
planning and energy  Military, police, pollution control, etc.

1.14 OPERATIONS RESEARCH MODELS IN PRACTICE

There is no unique set of problems that can be solved by using OR techniques. Several OR techniques
can be grouped into some basic categories as given below. In this book, a large number of OR techniques
have been discussed in detail.
 Allocation models Allocation models are used to allocate resources to activities so as to optimize measure
of effectiveness (objective function). Mathematical programming is the broad term for the OR techniques
used to solve allocation problems.
If the measure of effectiveness such as profit, cost, etc., is represented as a linear function of several
variables and limitations on resources (constraints) are expressed as a system of linear equalities or
inequalities, the allocation problem is classified as a linear programming problem. But, if the objective
function or all constraints cannot be expressed as a system of linear equalities or inequalities, the allocation
problem is classified as a non-linear programming problem.
When the value of decision variables in a problem is restricted to integer values or just zero-one values,
the problem is classified as an integer programming problem or a zero-one programming problem,
respectively.
A problem having multiple, conflicting and incommensurable objective functions (goals) subject to linear
constraints is called a goal programming problem. If values of decision variables in the linear programming
problem are not deterministic, then such a problem is called a stochastic programming problem.
If resources (such as workers, machines or salesmen) have to be assigned to perform a certain
number of activities (such as jobs or territories) on a one-to-one basis so as to minimize total time, cost
or distance involved in performing a given activity, such problems are classified as assignment problems.
But if the activities require more than one resource and conversely if the resources can be used for more
than one activity, the allocation problem is classified as a transportation problem.
 Inventory models These models deal with the problem of determination of how much to order at a pointin
time and when to place an order. The main objective is to minimize the sum of three conflicting inventory
costs: the cost of holding or carrying extra inventory, the cost of shortage or delay in the delivery of items
when it is needed and the cost of ordering or set-up. These are also useful in dealing with quantity
discounts and selective inventory control.
 Waiting line (or Queuing) models These models establish a trade-off between costs of providing service
and the waiting time of a customer in the queuing system. A queuing model describes: Arrival process,
queue structure and service process and solution for the measure of performance – average length of
waiting time, average time spent by the customer in the line, traffic intensity, etc., of the waiting system.
 Competitive (Game Theory) models These models are used to characterize the behaviour of two or more
competitors (called players) competing to achieve their conflicting goals. These models are classified
according to several factors such as number of competitors, sum of loss and gain, and thetype of
strategy which would yield the best or the worst outcomes.
 Operations Research: An Introduction 17

 Network models These models are applied to the management (planning, controlling and scheduling)
of large-scale projects. PERT/CPM techniques help in identifying delay and project critical path. These
techniques improve project coordination and enable the efficient use of resources. Network methods are
also used to determine time-cost trade-off, resource allocation and help in updating activity time.
 Sequencing models These models are used to determine the sequence (order) in which a number of tasks
can be performed by a number of service facilities such as hospital, plant, etc., in such a way that some
measure of performance (such as total time to process all the jobs on all the machines) is optimized.
 Replacement models These models are used to calculate optimal time to replace an equipment when
either its efficiency deteriorates with time or fails immediately and completely.
 Dynamic programming models These models are used where a problem requires optimization of
multistage (sequence of interrelated decisions) decision processes. The method starts by dividing a
given problem into stages or sub-problems and then solves those sub-problems sequentially until the
solution to the original problem is obtained.
 Markov-chain models These models are used for analyzing a system which changes over a period of
time among various possible outcomes or states. That is, these models describe transitions in terms of
transition probabilities of various states.
 Simulation models These models are used to evaluate alternative courses of action by experimenting with
a mathematical model of the problems with random variables. Thus, repetition of the process by using the
simulation model provides an indication of the merit of alternative course of action with respectto the
decision variables.
 Decision analysis models These models deal with the selection of an optimal course of action given the
possible payoffs and their associated probabilities of occurrence. These models are broadly appliedto
problems involving decision-making under risk and uncertainty.

1.15 COMPUTER SOFTWARE FOR OPERATIONS RESEARCH

Many real-life OR models require long and complex mathematical calculations. Thus, computer software
packages that are used to do these calculations rapidly and effectively have become a part of OR approach to
problem solving. Computer facilities such as spread sheets or statistical and mathematical software packages
that make such analysis readily available to a decision-maker.
Prentice-Hall distributes some of the leading software packages for quantitative approach to decision-
making. The features of a few such software packages are explained as follows.
 Quantitative Systems for Business Plus (QSB+), Version 3.0, by Yih-long Chang and Robert S Sullivan,
is a software package that contains problem-solving algorithms for operations research/management
science, as well as modules on basic statistics, non-linear programming and financial analysis.
QSB+ gives tips for getting started, presents tutorials on how to use it, details the special functions
and steps involved in using each program. Because of the friendliness of this microcomputer software, the
user will not encounter much difficulty.
 Quantitative Systems for Operations Management (QSOM), by Yih-long Chang, is an interactive user-
friendly system. It contains problem-solving algorithms for most of the important operations management
problems and associated information systems.
 Value STORM: MS Quantitative Modelling for Decision Support, by Hamilton Emmons, A D Flowers,
Chander Shekhar, M Khot and Kamlesh Mathur, is a special version of Personal STORM Version 3.0
developed for use in the operations research/management science.
 Excel 97 by Gene Weiss Kopf and distributed by BPB Publications, New Delhi, is an easy-to-use task-
oriented guide to Excel Spread sheet applications. It displays data in a wide variety of charts, creates,
maintains and sorts databases, while taking advantage of improved file-sharing capabilities.
 Linear Interactive Discrete Optimization (LINDO) is also a popular package for solving LP models on
personal computers. It was developed by Linus Schrage Lindo in his book: An Optimization Modeling
System, 4th ed. (Palo Alto, CA: Scientific Press, 1991)
18 Operations Research: Theory and Applications

CONCEPTUAL QUESTIONS B

1. Model building is the essence of the operations research approach. (c) The variables with which you propose to construct a model
Discuss. to assist you in making the decision.
2. What is meant by a mathematical model of a real situation?Discuss (d) The extent to which the variables and the relationships
the importance of models in the solution of OR problems. between them are quantifiable.
[AMIE, 2009] 18. What are the advantages and disadvantages of operations
3. What is the purpose of a mathematical model? How does a model research models? Why is it necessary to test models and how
achieve this purpose? In your answer consider the concept of ‘a would you go about testing a model?
model an abstraction of reality.’ 19. State any four areas for the application of OR techniques infinancial
4. Discuss in detail the three types of models with special emphasis management. How do they improve the performance of an
on important logical properties and the relationship the three types organization?
bear to each other and to modelled entities. 20. It is common in business to insure against the occurrence of
[Meerut, MSc (OR), 2000, AMIE, 2008 ] events which are subject to varying degrees of uncertainty, for
5. What is a model? Discuss various classification schemes of example, ill health of senior executives. At the same time, the
models. [CA, May, 2000 ] use of formal analytical models to assist in the process of
making decisions on business problems, which are generally
6. (a) Discuss, in brief, the role of OR models in decision-making.
subject to uncertainty, does not appear to be very widespread.
What areas of OR have made a significant impact on the
Describe the model building approach to the analysis of
decision-making process?
business problems under conditions of uncertainty. Discuss the
(b) Discuss the importance of operations research in the
apparent inconsistency in a company’s willingness to insure
decision-making process.
when formal analytical models of an operations research nature,
(c) Give some examples of various types of models. What is
a mathematical model? Develop two examples of which allow for uncertainty, are relatively employed.
mathematical models. 21. Recent developments in the field of computer technology have
enabled operations research to integrate its models into information
7. Explain how and why operations research methods have been
systems and thus make itself a part of decision-making procedures
valuable in aiding executive decisions.
of many organizations. Comment.
[Meerut, MSc (Stat.), 2001 ]
22. Explain various types of OR models and indicate their applications
8. State the different types of models used in OR. Explain briefly
the general methods for solving these OR models. to production, inventory and distribution systems.
[Indore, MSc (Maths), 2002; Punjab BSc 23. Explain the distinction between
(Mech. Engg.), 2003 ] (a) Static and dynamic models.
9. What reasons can you think of to explain the fact that many models (b) Analytical and simulation models.
are built and never implemented? Does the absence of (c) Descriptive and prescriptive models.
implementation mean that the entire activity of model development 24. Write briefly about the following:
was a waste. (a) Iconic models. (b) Analogue models.
(c) Mathematical models (or symbolic models).
10. ‘Quantifying the elements of a decision problem is the easy part; the
hard part is solving the model.’ Do you agree? Why or why not? 25. ‘Whether in a private, non-profit, or public organization, the most
important and the distinguishing function of a manager is problem
11. In constructing any OR model, it is essential to realize that the most
solving. The field of quantitative methods is dedicated to aiding
important purpose of the modelling process is ‘to help manager
better’. Keeping this purpose in mind, state any four OR models managers in their problem-solving efforts. This is accomplished
that can be of help chartered accountants in advising their through the use of mathematical models to analyse the problems.’
clients. [CA, May, 2001] Discuss. [Delhi Univ., MBA, 2004]
12. (a) Explain the scope and methodology of OR, the main phases 26. ‘Executives at all levels in business and industry come across
of OR and techniques used in solving an OR problem. the problems of making decisions at every stage in their day- to-
day activities. Operations research provides them with various
(b) Explain the steps involved in the solution of an operations
quantitative techniques for decision-making and enhances their
research problem. [Allahabad, MBA, 2002]
ability to make long-range plans and solve everyday problems
13. How can OR models be classified? What is the best classification of running a business and industry with greater efficiency,
in terms of learning and understanding the fundamentals of OR? competence and confidence.’ Elaborate the statement with
14. Explain the essential characteristics of the following types of examples.
models: [Delhi Univ., MBA, 2005; IGNOU, Diploma in Mgt, 2003]
(a) Allocation models (b) Competitive games 27. Comment on the following statements;
(c) Inventory (d) Waiting line. (a) Operations research advocates a system’s approach andis
15. (a) Describe briefly the applications of OR in managerial decision- concerned with optimization.
making. Give suitable examples from different areas. [Delhi Univ., MBA (HCA), 2004]
(b) Explain the concept, scope and tools of OR as applicable (b) Operations research replaces management by personality.
to business and industry. [Meerut, MSc (Stat.), 2003] [Delhi Univ, MBA (HCA), 2005]
16. Define an OR model and give four examples. State their 28. Suppose you are being interviewed by the manager of a
properties, advantages and limitations. commercial firm for a job in the research department which deals
17. Assume that you have to make a decision. Describe your with the applications of quantitative techniques. Explain the scope
decision with specific reference to: and purpose of quantitative techniques and its usefulness to the
(a) The alternative courses of action available. firm. Give some examples of the applications of quantitative
(b) The criteria you have set for the evaluation of the available techniques in industry. [Ajmer Univ., MBA, 2002]
alternatives. 29. Operations research is an ongoing process. Explain this statement
with examples.
 Operations Research: An Introduction 19

30. Distinguish between model results that recommend a decision best OR models are the ones that are most elaborate and
and model results that are descriptive. mathematically complex? Why?
31. ‘Mathematics . . . tends to lull the unsuspecting into believing 32. ‘The hard problems are those for which models do not exist’
that he who thinks elaborately thinks well’. Do you think that the Interpret this statement. Give some examples.

CHAPTER SUMMARY
Operations research is the scientific approach to decision making. This approach includes defining the problem, developing a model,
solving the model, testing the solution, analyzing the results, and implementing the results. In using operations research approach,
however, there can be potential problems, including conflicting viewpoints, the impact of quantitative analysis on various sections
of the organization, beginning assumptions, outdated solutions, understanding the model, obtaining relevantinput data, obtaining
optimal solution, testing the solution, and analyzing the results. In using the operations research approach, implementation is not
the final step because there can be a lack of commitment to the approach and resistance to change.

CHAPTER CONCEPTS QUIZ

True or False Multiple Choice
1. Operations research is that art of winning wars without actually 21. Operations research approach is
fighting them. (a) multi-disciplinary (b) scientific
2. Operations research approach does not adequately meet the (c) intuitive (d) all of the above
needs of business and industry. 22. Operations research practioners do not
3. Static model represents a system at some specified time without (a) predict future operations
doing any change over a period of time. (b) build more than one model
(c) collect relevant data
4. Operations research is a scientific approach to problem solving
(d) recommend decision and accept
for executives.
23. For analyzing a problem, decision-makers should study
5. Simulation is one of the most widely used quantitative techniques.
(a) its qualitative aspects (b) its quantitative aspects
6. The mathematical modeling of inventory control problem are (c) both (a) and (b) (d) neither (a) nor (b)
among the best operations research techniques.
24. Decision variables are
7. Simulation models are comparatively less flexible than any (a) controllable (b) uncontrollable
mathematical models. (c) parameters (d) none of the above
8. All decisions are made either on the basis of intuition or by using 25. A model is
more formal approaches. (a) an essence of reality (b) an approximation
9. The quantitative approach to decision modeling is a rational (c) an idealization (d) all of the above
approach. 26. Managerial decisions are based on
10. Operations research practioners can predict about the future (a) an evaluation of quantitative data
events. (b) the use of qualitative factors
(c) results generated by formal models
Fill in the Blanks (d) all of the above
27. The use of decision models
11. Dynamic programming involves the optimization of the .............. (a) is possible when the variable’s value is known
decision processes. (b) reduces the scope of judgement and intuition known with
12. Media planning is one of the .............. of operations research. certainty in decision-making
(c) require the use of computer software
13. There are many real world problems that don’t possess any
.............. solution. (d) none of the above
28. Every mathematical model
14. Operations research is a................ approach to problem solving
for executives. (a) must be deterministic
(b) requires computer aid for its solution
15. Operations research increases the .............. capabilities of a
(c) represents data in numerical form
decision-maker.
(d) all of the above
16. Model in which at least one decision variable is random is known
29. A physical model is example of
as ................. model.
(a) an iconic model (b) an analogue model
17. .............. is one of the situations which could cause decision- (c) a verbal model (d) a mathematical model
maker to move towards a quantitative approach.
30. An optimization model
18. The manager should rely on .............., .............. and .............. (a) provides the best decision
while using a quantitative approach for problem solving.
(b) provides decision within its limited context
19. The constraints may be in the form of .............. or ............... . (c) helps in evaluating various alternatives
20. A .............. is sometimes described as a rule of thumb that (d) all of the above
usually works. 31. The quantitative approach to decision analysis is a
(a) logical approach (b) rational approach
(c) scientific approach (d) all of the above
20 Operations Research: Theory and Applications

32. The process of modifying an OR model to observe the effect (c) predict future actions/operations
upon its outputs is called (d) build more than one model
(a) sensitivity analysis (b) cost/benefit analysis 34. OR approach is typically based on the use of
(c) model validation (d) input variation (a) Physical model (b) Mathematical model
33. Operations research practitioners do not (c) Iconic model (d) Descriptive model
(a) take responsibility for solution implementation 35. The qualitative approach to decision analysis relies on
(b) collect essential data (a) Experience (b) Judgement
(c) Intuition (d) All of the above

Answers to Quiz
1. T 2. T 3. T 4. T 5. T 6. T 7. F 8. F 9. T 10. F
11. Multistage 12. application 13. optimal 14. scientific 15. creativity
16. stochastic 17. problem is complex 18. experience, intuitions, judgement 19. equalities, inequalities
20. heuristic 21. (a) 22. (a) 23. (c) 24. (a) 25. (d) 26. (d) 27. (d) 28. (c) 29. (a)
30. (d) 31. (c) 32. (a) 33. (c) 34. (b) 35. (d)

CASE STUDY
Case 1.1: Daily Newspaper
A meeting of ‘Board of Governers’ of a daily newspaper was called by its Managing Director to consider business
proposals from the promotion department.
The promotion department submitted two alternatives proposals to increase daily circulation of the paper. The
proposals, indicating the increased circulation as well as the cost of obtaining this increase is shown in Table 1. The
cost is incurred in the first month, and the increase in circulation generally lasts about three months.
For each proposal, the advertising volume (in thousands of column cm) at the current rate of Rs 1,000 per column
cm, and at the increased rate of Rs 1,100 per column cm is also shown in Table 1. Even advertising also has seasonal
variations, above or below the normal level. In particular, next month should be at normal level, the succeeding month at
10 per cent above the normal, and the third month at 8 per cent below normal.

Proposal Daily Average Cost of Additional Advertising Volume

Circulation Circulation (in thousand column cm )
(Rs 1,000s) Current Rate Increased Rate
(Rs 1,000 per (Rs 1,100 per
column cm) column cm)

Table 1 A (Existing) 2,00,000 0 335 300

Promotion B 2,30,000 500 345 325
Alternatives C 2,40,000 900 350 335

Since various parts of the Sunday edition are printed during the entire week, the promotion department also proposed
to sub-contract the printing of various parts of the Sunday edition. It is believed that such an act will beable to
reduce (or eliminate) overtime in the press room. A local printer has consented to do the printing and supplythe
newsprint for Rs 1,04,000 per 10 lakh sheets printed, with the condition that he will be given three months’
contract with the same amount of work each month.

Type of Sheet Variable Cost (Rs)

(per column cm)
Table 2
Cost of Advertising sheet 180
Typesetting Editorial, news and other features 220

The main sources of revenue generation for the newspaper are: (i) daily sales, and (ii) sales of advertising space.
Operational costs incurred are broadly classified as costs related to: (i) distribution, (ii) editorial and news coverage,
(iii) typesetting, photo-engraving, layout and artwork, (iv) running of presses to produce the paper, and (v) newsprint.
It is understood that a larger circulation should generate more advertising revenue because the paper would be
read by more readers and therefore, advertisers would like to release more advertisements in the paper. Also, readers buy
the paper partly for the advertisements.
 Operations Research: An Introduction 21

Variable Cost Per Capacity Per

Lakh Sheets Lakh Sheets
Printed (Rs) Per Day Table 3
Press Room
Regular time 3,000 24 Costs and
Overtime 4,600 Unlimited Capacity
For planning purposes, a month is considered as one unit and average values for circulation, advertising and news
are used. The current average daily circulation of the paper is 2 lakh and average net profit is Re 0.50 on each paper,
per day.
The paper is currently using 4,000 column cm space for advertising per day and this generates revenue at the
rate of Rs 1,000 per column cm. The remaining, but fixed space of 800 column cm, is used for editorial features and news
makeup.
The variable costs of typesetting and press room are shown in Tables 2 and 3, respectively. A sheet is one copy of the
page of the newspaper. The regular capacity of the press room is 24 lakh sheets per day.

Department Fixed Cost per Month

(Rs 100s)
Typesetting 2,500
Editorial, news and features 3,200
Press room 9,000
Table 4
Distribution 26,000
Fixed Costs
The variable distribution costs depend on the circulation and size. The average distribution cost is Rs 1,600 per
tonne of newspaper. The fixed costs for other activities are shown in Table 4. The variable cost of newsprint is
Rs 4,800 per tonne. The press room can effectively print 50,000 sheets from one tonne of newsprint. The quota of
the newspaper is 48 tonnes of newsprint per month. However, additional newsprint can be obtained at an extra cost
of Rs 6,400 per tonne.
Questions for Discussion
1. Identify decision variables and constraints to build a model. Write down the relationships in the form of equations/
inequalities.
2. Consider the following initial values for the decision variables.
Initial circulation : 2 lakh/day
Price of advertising : Rs 1,000/cm
Initial space used for advertising : 4,000 column cm
Sheets printed outside : Nil
Examine (a) the financial viability of various proposals.
(b) the possibility of increasing circulation by one per cent other conditions remaining same, as well
as the effect of the increase on total profit per year.

Case 1.2: Supply Chain Management

A global consumer-electronics firm wants to automate its existing fulfillment process using ERP, APS, and the Internet.
Currently, its regional offices around the globe group, orders from customers, typically consumer electronics chains,
in their regions once a week for 26 weeks of a rolling horizon. They then send the grouped orders to the headquarters.
The headquarters assign current and planned inventory to these orders and then spends them directly from warehouses at
the plant locations to customer warehouses around the globe. In the future, customers may order from headquarters
directly, and external suppliers may fill orders. For the first four weeks of the planning horizon, the orders are firm, while
later orders may change in time and therefore serve as forecasts. The headquarters, regional offices (or customers), and
plants (or suppliers) can use the Internet, ERP, and APS systems (as shown in Fig. 1.4) as follows, while maintaining the
existing fulfillment process, with all steps except (f) and (g) to be run once a week in the following sequence:

(a) Regional offices (or customers) send orders through the Internet to a database at headquarters and overwrite the
orders for corresponding weeks placed a week earlier. Unfilled orders from past weeks in the ERP system also
go to this database.
(b) Plants (or suppliers) use the Internet to enter the planned replenishments by date and ship-from location in the
same database. The ERP system maintains records of current inventory at each plant (supplier) location as stocks
are added or taken out.
22 Operations Research: Theory and Applications

Fig. 1.4
Illustration of
using Web in
Conjunction with
its ERP and APS
Systems to
Improve Supply
Chain
Management

(c) The APS system takes all the orders and the planned replenishments for the entire time horizon of 26 weeks from the
database and takes the current inventory from the ERP system. It runs either a heuristic or a linear-programming-based
model to allocate current and planned inventory among these orders, balancing on the way delays and customers
according to their importance. With either model, the system gives each order an allocated quantity, target shipping date,
and a target shipper location. The firm considers planned replenishments fixed for the first 12 weeks and flexible for
weeks 13 through 26 weeks, so the APS system take the first 12 weeks’ replenishment from suppliers as a constraint and
the next 14 weeks’ replenishment as a decision variable. (The heuristic for a single shipping location works as follows:
taking the time periods in sequence, for each time period, it sorts unfilled orders by due date and customer importance
and then allocates the maximum possible inventory to orders starting from the top of the list. The linear-programming-
based solution is a multi-time-period network model with current and planned inventory as source nodes, orders as
demand nodes, and delay penalties per unit, time-based on customer importance.)
(d) Through Internet-enabled collaboration, the APS system makes the forecast for weeks 13 through to 26 available
to the suppliers.
(e) The ERP system imports orders for the current week from the database and updates unfilled orders from previous
weeks. All of these orders now have a shipping quantity and date as determined by the APS module in step (c).
(f) The ERP system on a daily basis executes orders based on the shipping date and the inventory at the shipping
location, issuing transportation orders and updating order statuses accordingly.
(g) The regional offices (or customers) can use the Internet any time after the APS run to view the shipping quantities
and shipping dates for their planned orders. The ERP system provides information on later orders.

Questions for Discussion

1. If this firm were to join an electronic marketplace, it might have to radically change and use OR to make
improvements even before joining such a marketplace. Suggest the kind of improvement it should adopt in order
to make the process more responsive to changes in demand.
2. The firm could make the supply chain even more responsive if it allowed changes to the suppliers’ replenishment
schedule within the first 12 weeks. Then, it could tie the suppliers’ production scheduling system(s) to its APS system.
This would let a customer enter a tentative order through the Web to get an immediate response as to whether or not
the firm could fill the order by the specified date. Suggest the method which could provide this functionality in the
present context.

PUZZLES IN OPERATIONS RESEARCH*

Students are advised to read the chapter on ‘Integer Programming’ and ‘Assignment Problems’ before attempting to
solve the following puzzles.
Less and Less # 1 The puzzle given below can initially be formulated as the integer programming problem.
47
Maximize Z =  ( c1 j  c2 j  c3 j ) x j
j 1

* Based on Glenn Weber, Puzzle Contests in MS/OR Education, Interface 20, March-April 1990, pp. 72–76.
subject to the constraints
47 47 47
 xj = 8;  ( c1 j  c2 j ) xj  – 1;  ( c1 j  c3 j ) xj  – 1
j 1 j 1 j 1

and xj = 0, 1 for j = 1, 2, . . ., 47.

where cij is the numerical value of letter i in word j and xj = 1 if word j is chosen.

Less and Less # 1

Place a different word from the word list on each of the 8 lines below. All letters are assigned values. Use the letter values
to score the columns of the words. The total of column 1 (extreme left column) must be less than the totalof column
2 . . . and the total of column 1 must also be less than the total of column 3. Add the column totals to determine your
score. High score wins. Ties, if any, will be broken by the highest total for column 1. If there arestill ties, column
2 will be judged for the highest total.

Letter Values
A- 1 H - 8 O - 15 V - 22 1. —— —— ——
B- 2 I - 9 P - 16 W - 23 2. —— —— ——
C- 3 J - 10 Q - 17 X - 24 3. —— —— ——
D- 4 K - 11 R - 18 Y - 25 4. —— —— ——
E- 5 L - 12 S - 19 Z - 26 5. —— —— ——
F - 6 M - 13 T - 20 6. —— —— ——
G- 7 N - 14 U - 21 7. —— —— ——
8. —— —— ——
Total =
Word List
ADV AFT BET BKS CCW CIR DER DIP EAT EGO FAR
FIN GHQ GOO HAT HOI HUG ION IVE JCS JOE KEN
KKK LIP LYE MOL MTG NES NTH OIL OSF PIP PRF
QMG QUE ROE RUG STG SIP TUE TVA UTE VIP WHO
XIN YES ZIP

Hint The solution of this IP problem has an objective function value of Z = 330. The alternative optimal solutions
must be considered in order to handle ties, as stipulated in the rules. The easiest way to handle this is to include
in the constraint set the objective function.
47 47
 ( c1 j  c2 j  c3 j ) x j = 330 and to use Max  c1 j xj
j 1 j 1

as the new objective function. The solution to this revised formulation includes the words BKS, GOD, ION, LYE, NTH,
PIP, VIP, ZIP with column totals of 108, 113, and 109 and the overall total of 330. There was no need to consider
maximizing the second column to handle additional ties since, if 113 could be increased, 109 would have to decrease and
this world violate the condition that the total of column 1 must be less than the total of column 3.

Assign the Values # 2 This puzzle is a variation of assignment problem in which an even letter is assigned a number,
and to guarantee that they are different, every number is assigned a letter. Assuming that the group totals can be made

RS| GF JIUV |
equal, the objective is to solve for ties by maximizing the total of group 1.
9 9
Maximize Z =  ci1  [ j] xij

j 1 |T H j 1 K|W
subject to the constraints

R(|c F IU
S G J V| = 0;
9 9 9
 i1  ci 2 )  [ j ] xij  xij = 1, i = 1, 2, . . ., 9

j 1 |T H j 1 K |W j 1
9

 xij = 1, j = 1, 2, . . ., 9; and xij = 0, 1 for all i and j

j 1
 Operations Research: An Introduction 23

where cik is the number of times letters i appears in group k and xij = 1 if letter i is assigned value j.
24 Operations Research: Theory and Applications

Assign the Values # 2

Assign values 1 to 9 to the letters in the letter-value chart below. Each letter must have a different value. Determine totals
for both groups by using the letter values you assigned. You must now find the difference between the totalsof
Groups 1 and 2 to determine your score. Low score wins. A score of 0 (equal totals) would be perfect. Ties, ifany,
will be broken by judging the total of Group 1 for the highest score.

Group 1 Group 2
1. AREA - 1. ERST -
2. FORT - 2. FOOT -
3. HOPE - 3. HEAT -
4. SPAR - 4. PAST -
5. THAT - 5. PROG -
6. TREE - 6. STOP -
Letter Values
A- P- E- Total — Group 1 = ––––––––
R- F- S - Total — Group 2 = ––––––––
H- T- O- Score = ––––––––

Hint The solution of this equal-group-total-assumption formulation has the letter values A – 6, E – 7, F – 2,H
– 1, O – 8, P – 4, R – 5, S – 3, T – 9 with equal-group totals of 139. If this formulation had no feasible solution,then
it would be necessary to solve a sequence of problems in which the right-hand side of the first constraint is varied with
increasing positive and negative integer values until a feasible solution is found.
 C h a p t e r 2
“Whenever there is a hard job to be done I assign it to a lazy man; he is sure to find an easy
way of doing it.”
– Walter Chrysler

Linear programming (LP) is a widely used mathematical modelling technique developed to help decision
makers in planning and decision-making regarding optimal use of scarce resources. This chapter is devoted
to illustrate the applications of LP programming in different functional areas of management and how LP
models are formulated.

LEARNING OBJECTIVES

After reading this chapter you should be able to

 identify situations in which linear programming technique can be applied.
 understand fundamental concepts and general mathematical structure of a linear programming model.
 express objective function and resource constraints in LP model in terms of decision variables and
parameters.
 appreciate the limitations and assumptions of linear programming technique with a view to interpret the
solution.

CHAPTER OUTLINE
2.1 Introduction 2.8 Examples of LP Model Formulation
2.2 Structure of Linear Programming Model  Conceptual Questions
2.3 Advantages of Using Linear Programming  Self Practice Problems
2.4 Limitations of Linear Programming  Hints and Answers
2.5 Application Areas of Linear Programming  Chapter Summary
2.6 General Mathematical Model of Linear  Chapter Concepts Quiz
Programming Problem  Case Study
2.7 Guidelines on Linear Programming Model
Formulation
26 Operations Research: Theory and Applications

2.1 INTRODUCTION
The application of specific operations research techniques to determine the choice among several courses
of action, so as to get an optimal value of the measures of effectiveness (objective or goal), requires to
formulate (or construct) a mathematical model. Such a model helps to represent the essence of a system
that is required for decision-analysis. The term formulation refers to the process of converting the verbal
description and numerical data into mathematical expressions, which represents the relationship among
relevant decision variables (or factors), objective and restrictions (constraints) on the use of scarce resources
(such as labour, material, machine, time, warehouse space, capital, energy, etc.) to several competing activities
(such as products, services, jobs, new equipment, projects, etc.) on the basis of a given criterion of optimality.
The term scarce resources refers to resources that are not available in infinite quantity during the planning
period. The criterion of optimality is generally either performance, return on investment, profit, cost, utility,
Linear
time, distance and the like.
Programming is a
mathematical In 1947, during World War II, George B Dantzing while working with the US Air Force, developed LP
technique useful for model, primarily for solving military logistics problems. But now, it is extensively being used in all functional
allocation of ‘scarce’
areas of management, airlines, agriculture, military operations, education, energy planning, pollution control,
or ‘limited’
resources, to transportation planning and scheduling, research and development, health care systems, etc. Though these
several competing applications are diverse, all LP models have certain common properties and assumptions – that are essential
activities on the for decision-makers to understand before their use.
basis of a given
criterion of Before discussing the basic concepts and applications of linear programming, it is important to
optimality. understand the meaning of the words – linear and programming. The word linear refers to linear relationship
among variables in a model. That is, a given change in one variable causes a proportional change in another
variable. For example, doubling the investment on a certain project will also double the rate of return. The
word programming refers to the mathematical modelling and solving of a problem that involves the use
of limited resources, by choosing a particular course of action (or strategy) among the given courses of action
(or strategies) in order to achieve the desired objective.
The usefulness of this technique is enhanced by the availability of several user-friendly computer
software such as STORM, TORA, QSB+, LINDO, etc. However, there is no computer software for building
an LP model. Model building is an art that improves with practice. A variety of examples are given in this
chapter to illustrate the formulation of an LP model.

2.2 STRUCTURE OF LINEAR PROGRAMMING MODEL

2.2.1 General Structure of an LP Model

The general structure of an LP model consists of following three basic components (or parts).
Decision variables (activities) The evaluation of various courses of action (alternatives) and select the
best to arrive at the optimal value of objective function, is guided by the nature of objective function and
availability of resources. For this, certain activities (also called decision variables) usually denoted by
x1, x2, . . ., xn are conducted. The value of these variables (activities) represents the extent to which each
of these is performed. For example, in a product-mix manufacturing problem, an LP model may be used to
Four components of determine units of each of the products to be manufactured by using limited resources such as personnel,
any LP model are: machinery, money, material, etc.
(i) decision
variables, (ii) The value of certain variables may or may not be under the decision-maker’s control. If values are under
objective function, the control of the decision-maker, then such variables are said to be controllable, otherwise they are said
(iii) constraints, and to be uncontrollable. These decision variables, usually interrelated in terms of consumption of resources,
(iv) non-negativity require simultaneous solutions. In an LP model all decision variables are continuous, controllable and non-
negative. That is, x1  0, x2  0, . . ., xn  0.

The objective function The objective function of each LP problem is expressed in terms of decision
variables to optimize the criterion of optimality (also called measure-of-performance) such as profit, cost,
revenue, distance etc. In its general form, it is represented as:
Optimize (Maximize or Minimize) Z = c1x1 + c2 x2 + . . . + cn xn,
where Z is the measure-of-performance variable, which is a function of x1, x2, . . ., xn. Quantities c1, c2, . . .,
cn are parameters that represent the contribution of a unit of the respective variable x1, x2, . . ., xn to the
 Linear Programming: Applications and Model Formulation 27

measure-of-performance Z. The optimal value of the given objective function is obtained by the graphical
method or simplex method.
The constraints There are always certain limitations (or constraints) on the use of resources, such as:
labour, machine, raw material, space, money, etc., that limit the degree to which an objective can be achieved.
Such constraints must be expressed as linear equalities or inequalities in terms of decision variables. The
solution of an LP model must satisfy these constraints.

2.2.2 Assumptions of an LP Model

In all mathematical models, assumptions are made for reducing the complex real-world problems into a
simplified form that can be more readily analyzed. The following are the major assumptions of an LP model:
Assumptions of an
LP model are:
1. Certainty: In LP models, it is assumed that all its parameters such as: availability of resources, profit (i) certainty,
(or cost) contribution per unit of decision variable and consumption of resources per unit of decision (ii) additivity,
variable must be known and constant. (iii) proportionality, &
(iv) divisibility
2. Additivity: The value of the objective function and the total amount of each resource used (or supplied),
must be equal to the sum of the respective individual contribution (profit or cost) of the decision
variables. For example, the total profit earned from the sale of two products A and B mustbe equal
to the sum of the profits earned separately from A and B. Similarly, the amount of a resource consumed
for producing A and B must be equal to the total sum of resources used for A and B individually.

3. Linearity (or proportionality): The amount of each resource used (or supplied) and its contributionto
the profit (or cost) in objective function must be proportional to the value of each decision variable. For
example, if production of one unit of a product uses 5 hours of a particular resource, then making3
units of that product uses 3×5 = 15 hours of that resource.

4. Divisibility (or continuity): The solution values of decision variables are allowed to assume
continuous values. For instance, it is possible to collect 6.254 thousand litres of milk by a milk dairy and
such variables are divisible. But, it is not desirable to produce 2.5 machines and such variables are
not divisible and therefore must be assigned integer values. Hence, if any of the variable can
assume only integer values or are limited to discrete number of values, LP model is no longer applicable.

2.3 ADVANTAGES OF USING LINEAR PROGRAMMING

Following are certain advantages of using linear programming technique:

1. Linear programming technique helps decision-makers to use their productive resources effectively.
2. Linear programming technique improves the quality of decisions. The decision-making approach of the
user of this technique becomes more objective and less subjective.
3. Linear programming technique helps to arrive at optimal solution of a decision problem by taking into
account constraints on the use of resources. For example, saying that so many units of any product
may be produced does not mean that all units can be sold.
4. Linear programming approach for solving decision problem highlight bottlenecks in the production
processes. For example, when a bottleneck occurs, machine cannot produce sufficient number of units
of a product to meet demand. Also, machines may remain idle.

2.4 LIMITATIONS OF LINEAR PROGRAMMING

In spite of having many advantages and wide areas of applications, there are some limitations associated
with this technique. These are as follows:
1. Linear programming assumes linear relationships among decision variables. However, in real-life problems,
decision variables, neither in the objective function nor in the constraints are linearly related.
28 Operations Research: Theory and Applications

2. While solving an LP model there is no guarantee that decision variables will get integer value. For
example, how many men/machines would be required to perform a particular job, a non-integer valued
solution will be meaningless. Rounding off the solution to the nearest integer will not yield an optimal
solution.
3. The linear programming model does not take into consideration the effect of time and uncertainty.
4. Parameters in the model are assumed to be constant but in real-life situations, they are frequently neither
known nor constant.
5. Linear programming deals with only single objective, whereas in real-life situations a decision problem
may have conflicting and multiple objectives.

2.5 APPLICATION AREAS OF LINEAR PROGRAMMING

Linear programming is the most widely used technique of decision-making in business and industry and
in various other fields. In this section, broad application areas of linear programming are discussed:

Applications in Agriculture
These applications fall into categories of farm economics and farm management. The former deals with inter-
regional competition, optimum allocation of crop production, efficient production patterns under regional
land resources and national demand constraints, while the latter is concerned with the problems of the
individual farm such as allocation of limited resources such as acreage, labour, water supply, working capital,
etc., so as to maximize the net revenue.

Applications in Military
Military applications include (i) selection of an air weapon system against the enemy, (ii) ensuring minimum
use of aviation gasoline (iii) updating supply-chain to maximize the total tonnage of bombs dropped on a
set of targets and takes care of the problem of community defence against disaster at the lowest possible cost.
Production Management
Product Mix To determine the quantity of several different products to be produced, knowing their per
unit profit (cost) contribution and amount of limited production resources used. The objective isto
maximize the total profit subject to all constraints.
 Production Planning This deals with the determination of minimum cost production plan over the
planning period, of an item with a fluctuating demand, while considering the initial number of units in
inventory, production capacity, constraints on production, manpower and all relevant cost factors. The
objective is to minimize total operation costs.
 Assembly-line Balancing This problem is likely to arise when an item can be made by assembling
different components. The process of assembling requires some specified sequence(s). The objective
is to minimize the total elapse time.
 Blending Problems These problems arise when a product can be made from a variety of available
raw materials, each of which has a particular composition and price. The objective here is to determine
the minimum cost blend, subject to availability of the raw materials, and to minimum and maximum
constraints on certain product constituents.
 Trim Loss When an item is made to a standard size (e.g. glass, paper sheet), the problem of determining
which combination of requirements should be produced from standard materials in order to minimize
the trim loss, arises.
Financial Management
 Portfolio Selection This deals with the selection of specific investment activity among several other
activities. The objective here is to find the allocation which maximizes the total expected return or
minimizes risk under certain limitations.
 Profit Planning This deals with the maximization of the profit margin from investment in plant facilities
and equipment, cash in hand and inventory.
 Linear Programming: Applications and Model Formulation 29

Marketing Management
 Media Selection The linear programming technique helps in determining the advertising media mix so
as to maximize the effective exposure, subject to limitation of budget, specified exposure rates to different
market segments, specified minimum and maximum number of advertisements in various media.
 Travelling Salesman Problem The salesman’s problem is to find the shortest route from a given city
to each of the specified cities and then returning to the original point of departure, provided no city would
be visited twice during the tour. Such type of problems can be solved with the help of the modified
assignment technique.
 Physical Distribution Linear programming determines the most economic and efficient manner of
locating manufacturing plants and distribution centres for physical distribution.
Personnel Management
 Staffing Problem Linear programming is used to allocate optimum manpower to a particular job so
as to minimize the total overtime cost or total manpower.
 Determination of Equitable Salaries Linear programming technique has been used in determining
equitable salaries and sales incentives.
 Job Evaluation and Selection Selection of suitable person for a specified job and evaluation of job
in organizations has been done with the help of the linear programming technique.
Other applications of linear programming lie in the area of administration, education, fleet utilization,
awarding contracts, hospital administration, capital budgeting, etc.

2.6 GENERAL MATHEMATICAL MODEL OF

LINEAR PROGRAMMING PROBLEM
The general linear programming problem (or model) with n decision variables and m constraints can be
stated in the following form:
Optimize (Max. or Min.) Z = c1x1 + c2 x2 + . . . + cn xn
subject to the linear constraints,
a11 x1  a12 x 2  . . .  a1 n xn ( , , ) b1
a21 x1  a22 x2  .. .  a2 n xn (, , ) b2

⁝ ⁝ ⁝ ⁝
am 1 x1  am 2 x 2  .. .  amn xn ( , , ) bm

and x1 , x2 , . . ., xn ≥ 0
The above formulation can also be expressed in a compact form as follows.
n
Optimize (Max. or Min.) Z = Â cj xj (Objective function) (1)
j =1

subject to the linear constraints

n
 aij xj ( £, =, ≥) bi ; i = 1, 2, ..., m (Constraints) (2)
j =1

and xj  0 ; j  1, 2, .. ., n (Non-negativity conditions) (3)

where, the cj s are coefficients representing the per unit profit (or cost) of decision variable xj to the value
of objective function. The aij’s are referred as technological coefficients (or input-output coefficients).
These represent the amount of resource, say i consumed per unit of variable (activity) xj. These coefficients
can be positive, negative or zero. The bi represents the total availability of the ith resource. The term resource
is used in a very general sense to include any numerical value associated with the right-hand side
of a constraint. It is assumed that bi  0 for all i. However, if any bi < 0, then both sides of constraint
i is multiplied by –1 to make bi > 0 and reverse the inequality of the constraint.
In the general LP problem, the expression (, =, ) means that in any specific problem each constraint
may take only one of the three possible forms:
(i) less than or equal to ()
(ii) equal to (=)
(iii) greater than or equal to ()
30 Operations Research: Theory and Applications

2.7 GUIDELINES ON LINEAR PROGRAMMING MODEL FORMULATION

The effective use and application requires, as a first step, the mathematical formulation of an LP model. Steps
of LP model formulation are summarized as follows:
Step 1: Identify the decision variables
(a) Express each constraint in words. For this you should first see whether the constraint is of the form
 (at least as large as), of the form  (no larger than) or of the form = (exactly equal to).
LP model
(b) Express verbally the objective function.
formulation requires: (c) Verbally identify the decision variables with the help of Step (a) and (b). For this you need to ask yourself
(i) identification of the question – What decisions must be made in order to optimize the objective function? Having followed
decision variables
Step 1(a) to (c) decide the symbolic notation for the decision variables and specify
and input data,
(ii) formulation of their units of measurement. Such specification of units of measurement would help in interpreting the final
constraints, and solution of the LP problem.
(iii) objective
function Step 2: Identify the problem data
To formulate an LP model, identify the problem data in terms of constants, and parameters associated with
decision variables. It may be noted that the decision-maker can control values of the variables but cannot
control values in the data set.
Step 3: Formulate the constraints
Convert the verbal expression of the constraints in terms of resource requirement and availability of each
resource. Then express each of them as linear equality or inequality, in terms of the decision variables defined
in Step 1.
Values of these decision variables in the optimal LP problem solution must satisfy these constraints
in order to constitute an acceptable (feasible) solution. Wrong formulation can either lead to a solution that
is not feasible or to the exclusion of a solution that is actually feasible and possibly optimal.
Step 4: Formulate the objective function
Identify whether the objective function is to be maximized or minimized. Then express it in the form of linear
mathematical expression in terms of decision variables along with profit (cost) contributions associated with
them.
After gaining enough experience in model building, readers may skip verbal description. The following
are certain examples of LP model formulation that may be used to strengthen the ability to translate a real-
life problem into a mathematical model.

2.8 EXAMPLES OF LP MODEL FORMULATION

In this section a number of illustrations have been presented on LP model formulation with the hope that
readers may gain enough experience in model building.

2.8.1 Examples on production

Example 2.1 A manufacturing company is engaged in producing three types of products: A, B and C. The
production department produces, each day, components sufficient to make 50 units of A, 25 units of B and
30 units of C. The management is confronted with the problem of optimizing the daily production of the
products in the assembly department, where only 100 man-hours are available daily for assembling the
products. The following additional information is available:

Type Profit Contribution per Assembly Time

of Product Unit of Product (Rs) per Product (hrs)

A 12 0.8
B 20 1.7
C 45 2.5

The company has a daily order commitment for 20 units of products A and a total of 15 units of
products B and C. Formulate this problem as an LP model so as to maximize the total profit.
 Linear Programming: Applications and Model Formulation 31

LP model formulation The data of the problem is summarized as follows:

Resources/Constraints Product Type Total

A B C
Production capacity (units) 50 25 30
Man-hours per unit 0.8 1.7 2.5 100
Order commitment (units) 20 15 (both for B and C)
Profit contribution (Rs/unit) 12 20 45

Decision variables Let x1, x2 and x3 = number of units of products A, B and C to be produced, respectively.
The LP model
Maximize (total profit) Z = 12x1 + 20x2 + 45x3
subject to the constraints
(i) Labour and materials
(a) 0.8x1 + 1.7x2 + 2.5x3  100, (b) x 1  50, (c) x2  25, (d) x3  30
(ii) Order commitment
(a) x1  20; (b) x2 + x3  15
and x1, x2, x3  0.

Example 2.2 A company has two plants, each of which produces and supplies two products: A and B.
The plants can each work up to 16 hours a day. In plant 1, it takes three hours to prepare and pack 1,000
gallons of A and one hour to prepare and pack one quintal of B. In plant 2, it takes two hours to prepare
and pack 1,000 gallons of A and 1.5 hours to prepare and pack a quintal of B. In plant 1, it costs Rs 15,000
to prepare and pack 1,000 gallons of A and Rs 28,000 to prepare and pack a quintal of B, whereas in plant
2 these costs are Rs 18,000 and Rs 26,000, respectively. The company is obliged to produce daily at least
10 thousand gallons of A and 8 quintals of B.
Formulate this problem as an LP model to find out as to how the company should organize its production
so that the required amounts of the two products be obtained at the minimum cost.
LP model formulation The data of the problem is summarized as follows:

Resources/Constraints Product Total

A B Availability (hrs)

Preparation time (hrs) Plant 1: 3 hrs/thousand gallons 1 hr/quintal 16

Plant 2: 2 hrs/thousand gallons 1.5 hr/quintal 16
Minimum daily production 10 thousand gallons 8 quintals
Cost of production (Rs) Plant 1: 15,000/thousand gallons 28,000/quintals
Plant 2: 18,000/thousand gallons 26,000/quintals

Decision variables Let

x1, x2 = quantity of product A (in ’000 gallons) to be produced in plant 1 and 2, respectively.
x3, x4 = quantity of product B (in quintals) to be produced in plant 1 and 2, respectively.

The LP model
Minimize (total cost) Z = 15,000x1 + 18,000x2 + 28,000x3 + 26,000x4
subject to the constraints
(i) Preparation time
(a) 3x1 + 2x2  16, (b) x3 + 1.5x4  16
(ii) Minimum daily production requirement
(a) x1 + x2  10, (b) x3 + x4  8
and x1, x2, x3, x4  0.
32 Operations Research: Theory and Applications

Example 2.3 An electronic company is engaged in the production of two components C1 and C2 that are used
in radio sets. Each unit of C1 costs the company Rs 5 in wages and Rs 5 in material, while each ofC2
costs the company Rs 25 in wages and Rs 15 in material. The company sells both products on one- period
credit terms, but the company’s labour and material expenses must be paid in cash. The selling price of C1 is
Rs 30 per unit and of C2 it is Rs 70 per unit. Because of the company’s strong monopoly in these components,
it is assumed that the company can sell, at the prevailing prices, as many units as it produces. The company’s
production capacity is, however, limited by two considerations. First, at the beginning of period 1, the
company has an initial balance of Rs 4,000 (cash plus bank credit plus collections from past credit sales).
Second, the company has, in each period, 2,000 hours of machine time and 1,400 hours of assembly time.
The production of each C1 requires 3 hours of machine time and 2 hours of assembly time, whereas the
production of each C2 requires 2 hours of machine time and 3 hours of assembly time. Formulatethis problem
as an LP model so as to maximize the total profit to the company.
LP model formulation The data of the problem is summarized as follows:

Resources/Constraints Components Total Availability

C1 C2

Budget (Rs) 10/unit 40/unit Rs 4,000

Machine time 3 hrs/unit 2 hrs/unit 2,000 hours
Assembly time 2 hrs/unit 3 hrs/unit 1,400 hours
Selling price Rs 30 Rs 70
Cost (wages + material) price Rs 10 Rs 40

Decision variables Let x1 and x2 = number of units of components C1 and C2 to be produced, respectively.
The LP model
Maximize (total profit) Z = Selling price – Cost price
= (30 – 10) x1 + (70 – 40) x2 = 20x1 + 30x2
subject to the constraints
(i) The total budget available
10x1 + 40x2  4,000
(ii) Production time
(a) 3x1 + 2x2  2,000; (b) 2x1 + 3x2  1,400
and x1, x2  0.
Example 2.4 A company has two grades of inspectors 1 and 2, the members of which are to be assigned for
a quality control inspection. It is required that at least 2,000 pieces be inspected per 8-hour day. Grade1
inspectors can check pieces at the rate of 40 per hour, with an accuracy of 97 per cent. Grade 2 inspectors
check at the rate of 30 pieces per hour with an accuracy of 95 per cent.
The wage rate of a Grade 1 inspector is Rs 5 per hour while that of a Grade 2 inspector is Rs 4 per
hour. An error made by an inspector costs Rs 3 to the company. There are only nine Grade 1 inspectors
and eleven Grade 2 inspectors available to the company. The company wishes to assign work to the available
inspectors so as to minimize the total cost of the inspection. Formulate this problem as an LP modelso as to
minimize the daily inspection cost. [Delhi Univ., MBA, 2004, 2006]
LP model formulation The data of the problem is summarized as follows:
Inspector

Grade 1 Grade 2
Number of inspectors 9 11
Rate of checking 40 pieces/hr 30 pieces/hr
Inaccuracy in checking 1 – 0.97 = 0.03 1 – 0.95 = 0.05
Cost of inaccuracy in checking Rs 3/piece Rs 3/piece
Wage rate/hour Rs 5 Rs 4
Duration of inspection = 8 hrs per day
Total pieces which must be inspected = 2,000
 Linear Programming: Applications and Model Formulation 33

Decision variables Let x1 and x2 = number of Grade 1 and 2 inspectors to be assigned for inspection,
respectively.
The LP model
Hourly cost of each inspector of Grade 1 and 2 can be computed as follows:
Inspector Grade 1 : Rs (5 + 3 × 40 × 0.03) = Rs 8.60
Inspector Grade 2 : Rs (4 + 3 × 30 × 0.05) = Rs 8.50
Based on the given data, the LP model can be formulated as follows:
Minimize (daily inspection cost) Z = 8 (8.60x1 + 8.50x2) = 68.80x1 + 68.00x2
subject to the constraints
(i) Total number of pieces that must be inspected in an 8-hour day
8 × 40x1 + 8 × 30x2  2000
(ii) Number of inspectors of Grade 1 and 2 available
(a) x1  9, (b) x2  11
and x1, x2  0.
Example 2.5 An electronic company produces three types of parts for automatic washing machines. It
purchases casting of the parts from a local foundry and then finishes the part on drilling, shaping and polishing
machines.
The selling prices of parts A, B and C are Rs 8, Rs 10 and Rs 14 respectively. All parts made can be
sold. Castings for parts A, B and C, respectively cost Rs 5, Rs 6 and Rs 10.
The shop possesses only one of each type of casting machine. Costs per hour to run each of the three
machines are Rs 20 for drilling, Rs 30 for shaping and Rs 30 for polishing. The capacities (parts per hour)
for each part on each machine are shown in the table:
Machine Capacity per Hour
Part A Part B Part C
Drilling 25 40 25
Shaping 25 20 20
Polishing 40 30 40

The management of the shop wants to know how many parts of each type it should produce per hour
in order to maximize profit for an hour’s run. Formulate this problem as an LP model so as to maximize total
profit to the company. [Delhi Univ., MBA, 2001, 2004, 2007]
LP model formulation Let x1, x2 and x3 = numbers of type A, B and C parts to be produced per hour,
respectively.
Since 25 type A parts per hour can be run on the drilling machine at a cost of Rs 20, then
Rs 20/25 = Re 0.80 is the drilling cost per type A part. Similar reasoning for shaping and polishing gives

 20 30 30 
Profit per type A part = (8 – 5) –      = 0.25
 25 25 40 

 20 30 30 
Profit per type B part = (10 – 6) –     = 1
 40 20 30 
 20 30 30 
Profit per type C part = (14 – 10) –   
  = 0.95
 25 20 40 

On the drilling machine, one type A part consumes 1/25th of the available hour, a type B part consumes
1/40th, and a type C part consumes 1/25th of an hour. Thus, the drilling machine constraint is
x1 x2 x3
  1
25 40 25
Similarly, other constraints can be established.
The LP model
Maximize (total profit) Z = 0.25 x1 + 1.00 x2 + 0.95 x3
subject to the constraints
x1 x2 x3 x1 x2 x3
(i) Drilling machine:    1, (ii) Shaping machine:    1,
25 40 25 25 20 20
34 Operations Research: Theory and Applications

x1 x2 x3
(iii) Polishing machine:    1,
40 30 40

and x1 , x2 , x3  0.
Example 2.6 A pharmaceutical company produces two pharmaceutical products: A and B. Production of
both these products requires the same process – I and II. The production of B also results in a by-product
C at no extra cost. The product A can be sold at a profit of Rs 3 per unit and B at a profit of Rs 8 per
unit. Some quantity of this by-product can be sold at a unit profit of Rs 2, the remainder has to be destroyed
and the destruction cost is Re 1 per unit. Forecasts show that only up to 5 units of C can be sold. The company
gets 3 units of C for each unit of B produced. The manufacturing times are 3 hours per unit forA on
process I and II, respectively, and 4 hours and 5 hours per unit for B on process I and II, respectively. Because
the product C is a by product of B, no time is used in producing C. The available times are 18and 21
hours of process I and II, respectively. Formulate this problem as an LP model to determine the quantity of
A and B which should be produced, keeping C in mind, to make the highest total profit to the company.
[Delhi Univ., MBA (HCA), 2001, 2008]
LP model formulation The data of the problem is summarized as follows:
Constraints/Resources Time (hrs) Required by Availability

A B C
Process I 3 4 – 18 hrs
Process II 3 5 – 21 hrs
By-product ratio from B – 1 3 5 units (max. units that
Profit per unit (Rs) 3 8 2 can be sold)

Decision variables Let

x1, x2 = units of product A and B to be produced, respectively
x3, x4 = units of product C to be produced and destroyed, respectively.
The LP model
Maximize (total profit) Z = 3x1 + 8x2 + 2x3 – x4
subject to the constraints
(i) Manufacturing constraints for product A and B
(a) 3x1 + 4x2  18, (b) 3x1 + 5x2  21
(ii) Manufacturing constraints for by-product C
(a) x3  5, (b) – 3x2 + x3 + x4 = 0
and x1, x2, x3, x4  0.
Example 2.7 A tape recorder company manufactures models A, B and C, which have profit contributions per
unit of Rs 15, Rs 40 and Rs 60, respectively. The weekly minimum production requirements are 25 unitsfor
model A, 130 units for model B and 55 units for model C. Each type of recorder requires a certain amountof
time for the manufacturing of the component parts for assembling and for packing. Specifically, a dozen units
of model A require 4 hours for manufacturing, 3 hours for assembling and 1 hour for packaging. The
corresponding figures for a dozen units of model B are 2.5, 4 and 2 and for a dozen units of model C are
6, 9 and 4. During the forthcoming week, the company has available 130 hours of manufacturing, 170 hours
of assembling and 52 hours of packaging time. Formulate this problem as an LP model so as to maximize the
total profit to the company.
LP model formulation The data of the problem is summarized as follows:

Resources/Constraints Models Total Availability

A B C (hrs)

Production requirement (units) 25 130 55

Manufacturing time (per dozen) 4 2.5 6 130
Assembling time (per dozen) 3 4 9 170
Packaging time (per dozen) 1 2 4 52
Contribution per unit (Rs) 15 40 60
 Linear Programming: Applications and Model Formulation 35

Decision variables Let x1, x2 and x3 = units of model A, B and C to be produced per week,
respectively.
The LP model
Maximize (total profit) = 15x1 + 40x2 + 60x3
subject to the constraints
(i) Minimum production requirement:
(a) x1  25, (b) x2  130, (c) x3  55
4x1 2.5x2 6x3  130
(ii) Manufacturing time :  
12 12 12
9x3
(iii) Assembling time : 3x1

4x2   170

12 12 12
x1 2x2 4x3  52
(iv) Packaging time :  
12 12 12
and x1, x2, x3  0.

Example 2.8 Consider the following problem faced by a production planner of a soft drink plant. He has two
bottling machines A and B. A is designed for 8-ounce bottles and B for 16-ounce bottles. However, each can
also be used for both types of bottles with some loss of efficiency. The manufacturing data isas follows:
Machine 8-ounce Bottles 16-ounce Bottles
A 100/minute 40/minute
B 60/minute 75/minute
The machines can be run for 8 hours per day, 5 days per week. The profit on an 8-ounce bottle is
Rs 1.5 and on a 16-ounce bottle is Rs 2.5. Weekly production of the drink cannot exceed 3,00,000 bottles
and the market can absorb 25,000, 8-ounce bottles and 7,000, 16-ounce bottles per week. The planner wishes
to maximize his profit, subject of course, to all the production and marketing restrictions. Formulate this
problem as an LP model to maximize total profit.
LP model formulation The data of the problem is summarized as follows:

Constraints Production Availability

8-ounce Bottles 16-ounce Bottles
Machine A time 100/minute 40/minute 8 × 5 × 60 = 2,400 minutes
Machine B time 60/minute 75/minute 8 × 5 × 60 = 2,400 minutes
Production 1 1 3,00,000 units/week
Marketing 1 – 25,000 units/week
– 1 7,000 units/week
Profit/unit (Rs) 1.5 2.5

Decision variables Let x1 and x2 = units of 8-ounce and 16-ounce bottles to be produced weekly,
respectively
The LP model
Maximize (total profit) Z = 1.5x1 + 2.5x2
subject to the constraints
(i) Machine time : (a) x1  x2  2, 400 and (b) x1  x2  2, 400
100 40 60 75
(ii) Production : x1 + x2  3,00,000

(iii) Marketing : (a) x1  25,000, (b) x2  7,000

and x1, x2  0.
Example 2.9 A company engaged in producing tinned food has 300 trained employees on its rolls, each of
whom can produce one can of food in a week. Due to the developing taste of public for this kind of food, the
company plans to add to the existing labour force, by employing 150 people, in a phased manner, over the
next five weeks. The newcomers would have to undergo a two-week training programme before being
36 Operations Research: Theory and Applications

put to work. The training is to be given by employees from among the existing ones and it is a known fact
that one employee can train three trainees. Assume that there would be no production from the trainers and
the trainees during training period, as the training is off-the-job. However, the trainees would be remunerated
at the rate of Rs 300 per week, the same rate would apply as for the trainers.
The company has booked the following orders to supply during the next five weeks:
Week : 1 2 3 4 5
No. of cans : 280 298 305 360 400
Assume that the production in any week would not be more than the number of cans ordered for, so that every
delivery of the food would be ‘fresh’.
Formulate this problem as an LP model to develop a training schedule that minimizes the labour cost
over the five-week period. [Delhi Univ., MBA, 2003, 2005]

LP model formulation The data of the problem is summarized as given below:

(i) Cans supplied Week : 1 2 3 4 5
Number : 280 298 305 360 400
(ii) Each trainee has to undergo a two-week training.
(iii) One employee is required to train three trainees.
(iv) Every trained worker produces one can/week but there would be no production from trainers and
trainees during training.
(v) Number of employees to be employed = 150
(vi) The production in any week is not to exceed the cans required.
(vii) Number of weeks for which newcomers would be employed: 5, 4, 3, 2, 1.
Observations based on given data are as follows:
(a) Workers employed at the beginning of the first week would get salary for all the five weeks; those
employed at the beginning of the second week would get salary for four weeks and so on.
(b) The value of the objective function would be obtained by multiplying it by 300 because each
person would get a salary of Rs 300 per week.
(c) Inequalities have been used in the constraints because some workers might remain idle in some
week(s).
Decision variables Let x1, x2, x3, x4 and x5 = number of trainees appointed in the beginning of
week 1, 2, 3, 4 and 5, respectively.
The LP model
Minimize (total labour force) Z = 5x1 + 4x2 + 3x3 + 2x4 + x5
subject to the constraints
(i) Capacity
x x x
(a) 300  1  280 , (b) 300  1  2  298

3 3 3
x2 x3
(c) 300  x    305, (d) 300  x  x

x3

x4  360

1 1 2
3 3 3 3
x4 x5
(e) 300  x  x  x    400
1 2 3
3 3
(ii) New recruitment
x1 + x2 + x3 + x4 + x5 = 150
and x1, x2, x3, x4, x5  0.
Example 2.10 XYZ company assembles and markets two types of mobiles – A and B. Presently 200 mobiles
of each type are manufactured per week. You are advised to formulate the production schedule which will
maximize the profits in the light of the following information:

Type Total Component Cost Man-Hours of Average Man-Minutes of Selling Price Per
Per Mobile (Rs) Assembly Time Per Inspection and Correction Mobile (Rs)
Mobile
A 2000 12 10 6000
B 1600 6 35 4800
 Linear Programming: Applications and Model Formulation 37

The company employs 100 assemblers who are paid Rs 50 per hour actually worked and who will work
up to a maximum of 48 hours per week. The inspectors, who are presently four, have agreed to a plan,
whereby they average 40 hours of work per week each. However, the four inspectors have certain other
administrative duties which have been found to take up an average of 8 hours per week between them.
The inspectors are each paid a fixed wage of Rs 12000 per week.
Each mobile of either type requires one camera of same type. However, the company can obtain a
maximum supply of 600 cameras per week. Their cost has been included in the component's cost, given
for each mobile in the table above. The other cost incurred by the company are fixed overheads of Rs 20,000
per week.
LP model formulation Computation of contribution from radio types A and B is as follows:
A B
Component cost : 2000 1600
Labour cost in assembly (Rs 50 per hour) : 600 300
Labour cost for inspection

1200 1
  Re 0.50 per minute 5.00 17.5
40 60

Total variable cost 2605.00 1917.50

Selling price 6000 4800
Contribution (selling price – cost price) 3395 2882.50
Decision variables Let x1 and x2 = number of units of radio types A and B to be produced, respectively.
The LP model
Maximize (total contribution) Z = 3395x1 + 2882.5x2 – 20,000
subject to the constraints
 25  
(i) 12x + 6x  48 × 100, (ii) 10x + 35x  4 × 40 

 × 60 = 7,600


1 2 1 2  3 

(iii) x1 + x2  600
and x1, x2  0.
Example 2.11 A plastic products manufacturer has 1,200 boxes of transparent wrap in stock at one factory
and another 1,200 boxes at its second factory. The manufacturer has orders for this product from three
different retailers, in quantities of 1,000, 700 and 500 boxes, respectively. The unit shipping costs (in rupees
per box) from the factories to the retailers are as follows:

Retailer I Retailer II Retailer III

Factory A 14 11 13
Factory B 13 13 12

Determine a minimum cost shipping schedule for satisfying all demands from current inventory.
Formulate this problem as an LP model.
LP model formulation Given that the total number of boxes available at factory A and B = total numberof
boxes required by retailers 1, 2 and 3.
Decision variables Let x1, x2 and x3 = number of boxes to be sent from factory A to retailer 1; factory B
to retailer 2 and factory C to retailer 3, respectively.

Number of Boxes to be Sent

Retailer 1 Retailer 2 Retailer 3
Factory A x1 x2 1,200 – (x1 + x2)
Factory B 1,000 – x1 700 – x2 500 – [(1,000 – x1) + (700 – x2)]

The LP model
Minimize (total distance) Z = 14x1 + 13x2 + 11 (1,200 – x1 – x2) + 13 (1,000 – x1) +
13 (700 – x2) + 12 (x1 – x2 – 700) = 2x1 + x2 + 26,900
38 Operations Research: Theory and Applications

subject to the constraints

(i) x1 + x2  1,200, (ii) x1  1,000, (iii) x2  700;
and x1, x2  0.
Example 2.12 A company produces two types of sauces: A and B. Both these sauces are made by blending
two ingredients – X and Y. A certain level of flexibility is permitted in the formulae of these products. Indeed,
the restrictions are that (i) B must contain no more than 75 per cent of X, and
(ii) A must contain no less than 25 per cent of X, and no less than 50 per cent of Y. Up to 400 kg of X
and 300 kg of Y could be purchased. The company can sell as much of these sauces as it produces at
a price of Rs 18 for A and Rs 17 for B. The X and Y cost Rs 1.60 and 2.05 per kg, respectively.
The company wishes to maximize its net revenue from the sale of these sauces. Formulate this problem
as an LP model.
LP model formulation Let x1, x2 = kg of sauces A and B to be produced, respectively.
y1, y2 = kg of ingredient X used to make sauces A and B, respectively.
y3, y4 = kg of ingredient Y used to make sauces A and B, respectively.
The LP model
Maximize Z = 18x1 + 17x2 – 1.60 (y1 + y2) – 2.05 (y3 + y4)
subject to the constraints
(i) y1 + y3 – x1 = 0, (ii) y2 + y4 – x2 = 0
(iii) y1 + y2  400  (Purchase) (iv) y1 – 0.25 x1  0 
(Sauce A)
 
y3 + y4  300  y2 – 0.50 x2  0 

(v) y2 – 0.75x2  0 (Sauce B)

and x1, x2, y1, y2, y3, y4  0.
Example 2.13 A complete unit of a certain product consists of four units of component A and three units of
component B. The two components (A and B) are manufactured from two different raw materials of which100
units and 200 units, respectively, are available. Three departments are engaged in the production processwith
each department using a different method for manufacturing the components per production run and the
resulting units of each component are given below:

Department Input of Raw Materials Output of Components

per Run (units) per Run (units)
I II A B
1 7 5 6 4
2 4 8 5 8
3 2 7 7 3

Formulate this problem as an LP model to determine the number of production runs for each department
which will maximize the total number of complete units of the final product.
LP model formulation Let x1, x2 and x3 = number of production runs for departments 1, 2 and 3,
respectively.
Since each unit of the final product requires 4 units of component A and 3 units of component B,

R UV
therefore maximum number of units of the final product cannot exceed the smaller value of

TRS
Total number of units of A produced Total number of units of B produced

6 x1  5 x 2  7 x 3
4
;
4 x1  8 x2  3 x3 U 3 W

or ST 4
and
3
VW
Also if y is the number of component units of final product, then we obviously have
6 x1  5 x2  7 x3 4 x1  8 x2  3 x3
 y and  y
4 3
The LP model

R6x 1  5 x2  7 x3 4 x1  8 x 2  3 x 3 U
Maximize Z = Min ST 4
;
3
VW
 Linear Programming: Applications and Model Formulation 39

subject to the constraints

(i) Raw material
(a) 7x1 + 4x2 + 2x3  100 (Material I ), (b) 5x1 + 8x2 + 7x3  200 (Material II)
(ii) Number of component units of final product
(a) 6x1 + 5x2 + 7x3 – 4y  0, (b) 4x1 + 8x2 + 3x3 – 4y  0
and x1, x2, x3  0.
Example 2.14 ABC company manufactures three grades of paint: Venus, Diana and Aurora. The plant
operates on a three-shift basis and the following data is available from the production records:

Requirement Grade Availability

of Resource Venus Diana Aurora (capacity/month)
Special additive (kg/litre) 0.30 0.15 0.75 600 tonnes
Milling (kilolitres per machine shift) 2.00 3.00 5.00 100 machine shifts
Packing (kilolitres per shift) 12.00 12.00 12.00 80 shifts

There are no limitations on the other resources. The particulars of sales forecasts and the estimated
contribution to overheads and profits are given below:
Venus Diana Aurora
Maximum possible sales
per month (kilolitres) 100 400 600
Contribution (Rs/kilolitre) 4,000 3,500 2,000
Due to the commitments already made, a minimum of 200 kilolitres per month, of Aurora, must be
supplied the next year.
Just when the company was able to finalize the monthly production programme for the next 12 months,
it received an offer from a nearby competitor for hiring 40 machine shifts per month of milling capacity for
grinding Diana paint that could be spared for at least a year. However, due to additional handling at the
competitor’s facility, the contribution from Diana would be reduced by Re 1 per litre.
Formulate this problem as an LP model for determining the monthly production programme to maximize
contribution. [Delhi Univ., MBA, 2006]
LP model formulation Let
x1 = quantity of Venus (kilolitres) produced in the company
x2 = quantity of Diana (kilolitres) produced in the company
x3 = quantity of Diana (kilolitres) produced by hired facilities
x4 = quantity of Aurora (kilolitres) produced in the company
The LP model
Maximize (total profit) Z = 4,000x1 + 3,500x2 + (3,500 – 1,000)x3 + 2,000x4
subject to the constraints
(i) Special additive : 0.30x1 + 0.15x2 + 0.15x3 + 0.75x4  600
x x x
1  2  4  100
(ii) Own milling facility :
2 3 5
x3
(iii) Hired milling facility :  40
3
x1 x2  x3 x4
(iv) Packing :   80
12 12 12
(v) Marketing:
(i) x1  100 (Venus); (ii) x2 + x3  400 (Diana); (iii) 200  x4  600 (Aurora)
and x1, x2, x3, x4  0.
Example 2.15 Four products have to be processed through a particular plant, the quantities required for the
next production period are:
Product 1 : 2,000 units Product 2 : 3,000 units
Product 3 : 3,000 units Product 4 : 6,000 units
There are three production lines on which the products could be processed. The rates of production
in units per day and the total available capacity in days are given in the following table. The corresponding
cost of using the lines is Rs 600, Rs 500 and Rs 400 per day, respectively.
40 Operations Research: Theory and Applications

Production Line Product Maximum

(days) 1 2 3 4 Line
1 150 100 500 400 20
2 200 100 760 400 20
3 160 80 890 600 18

Total 2,000 3,000 3,000 6,000

Formulate this problem as an LP model to minimize the cost of operation.
LP model formulation Let xij = number of units of product i (i = 1, 2, 3, 4) produced on production
line j ( j = 1, 2, 3)
The LP model
4 4 4
Minimize (total cost) Z = 600  xi 1  500  xi 2  400  xi 3
i 1 i 1 i 1
subject to the constraints
3 3

(i) Production: (a)  xi1  2, 000, (b)  xi 2  3,000

i1 i 
1
3 3
(c)  xi3  3, 000, (d)  xi 4  6,000
i1 i 
1
(ii) Line capacity

x11 x12 x13

x14  20 , x21 x22 x23 x24
(a)    (b)     20
150 100 500 400 200 100 760 400
x x x x
(c) 31  32  33  34  18
160 80 890 600
and xij  0 for all i and j.

Example 2.16 XYZ company produces a specific automobile spare part. A contract that the company has
signed with a large truck manufacturer calls for the following 4-month shipping schedule.

Month Number of Parts

to be Shipped
January 3,000
February 4,000
March 5,000
April 5,000

The company can manufacture 3,000 parts per month on a regular time basis and 2,000 parts per month
on an overtime basis. Its production cost is Rs 15,000 for a part produced during regular time and 25,000
for a part produced during overtime. Its monthly inventory holding cost is Rs 500. Formulate this problem as
an LP model to minimize the overall cost.
LP model formulation Let xijk = number of units of automobile spare part manufactured in month
i (i = 1, 2, 3, 4) using shift j ( j = 1, 2) and shipped in month
k (k = 1, 2, 3, 4)
The LP model
Minimize (total cost) Z = Regular time production cost + Overtime production cost
+ One-month inventory cost + Two-month inventory cost
+ Three-month inventory cost
= 15,000(x111 + x112 + x113 + x114 + x212 + x213 + x214 + x313 + x314 + x414)
+ 25,000(x121 + x122 + x123 + x124 + x222 + x223 + x224 + x323 + x324 + x424)
+ 500(x112 + x122 + x213 + x223 + x314 + x324) + 1,000(x113 + x123 + x214
+ x224) + 1,500(x114 + x124)
 Linear Programming: Applications and Model Formulation 41

subject to the constraints

(i) Monthly regular time production
(a) x111 + x112 + x113 + x114  3,000, (b) x212 + x213 + x214  3,000
(c) x313 + x314  3,000, (d) x414  3,000
(ii) Monthly overtime production constraints
(a) x121 + x122 + x123 + x124  2,000, (b) x222 + x223 + x224  2,000
(c) x323 + x324  2,000, (d) x424  2,000
(iii) Monthly demand constraints
(a) x111 + x121 = 3,000, (b) x112 + x122 + x212 + x222 = 4,000
(c) x113 + x123 + x213 + x223 + x313 + x323 = 5,000
(d) x114 + x124 + x214 + x224 + x314 + x324 + x414 + x424 = 5,000
and xijk  0 for all i, j and k.

2.8.2 Examples on Marketing

Example 2.17 An advertising company wishes to plan an advertising campaign for three different media:
television, radio and a magazine. The purpose of the advertising is to reach as many potential customers
as possible. The following are the results of a market study:
Television
Prime Day Prime Time Radio Magazine
(Rs) (Rs) (Rs) (Rs)
Cost of an advertising unit 40,000 75,000 30,000 15,000
Number of potential customers reached per unit 4,00,000 9,00,000 5,00,000 2,00,000
Number of women customers reached per unit 3,00,000 4,00,000 2,00,000 1,00,000

The company does not want to spend more than Rs 8,00,000 on advertising. It is further required that
(i) at least 2 million exposures take place amongst women,
(ii) the cost of advertising on television be limited to Rs 5,00,000,
(iii) at least 3 advertising units be bought on prime day and two units during prime time; and
(iv) the number of advertising units on the radio and the magazine should each be between 5 and 10.
Formulate this problem as an LP model to maximize potential customer reach.
LP model formulation Let x1, x2, x3 and x4 = number of advertising units bought in prime day and
time on television, radio and magazine, respectively.
The LP model
Maximize (total potential customer reach) Z = 4,00,000x1 + 9,00,000x2 + 5,00,000x3 + 2,00,000x4
subject to the constraints
(i) Advertising budget: 40,000x1 + 75,000x2 + 30,000x3 + 15,000x4  8,00,000
(ii) Number of women customers reached by the advertising campaign
3,00,000x1 + 4,00,000x2 + 2,00,000x3 + 1,00,000x4  20,00,000
(iii) Television advertising : (a) 40,000x1 + 75,000x2  5,00,000; (b) x1  3; (c) x2  2
(iv) Radio and magazine advertising : (a) 5  x3  10; (b) 5  x4  10
and x1, x2, x3, x4  0.
Example 2.18 A businessman is opening a new restaurant and has budgeted Rs 8,00,000 for advertisement,
for the coming month. He is considering four types of advertising:
(i) 30 second television commercials
(ii) 30 second radio commercials
(iii) Half-page advertisement in a newspaper
(iv) Full-page advertisement in a weekly magazine which will appear four times during the coming month.
The owner wishes to reach families (a) with income over Rs 50,000 and (b) with income under Rs 50,000.
The amount of exposure of each media to families of type (a) and (b) and the cost of each media is shown
below:
42 Operations Research: Theory and Applications

Media Cost of Advertisement Exposure to Families with Exposure to Families with

(Rs) Annual Income Over Annual Income Under
Rs 50,000 (a) Rs 50,000 (b)
Television 40,000 2,00,000 3,00,000
Radio 20,000 5,00,000 7,00,000
Newspaper 15,000 3,00,000 1,50,000
Magazine 5,000 1,00,000 1,00.000

To have a balanced campaign, the owner has determined the following four restrictions:
(i) there should be no more than four television advertisements
(ii) there should be no more than four advertisements in the magazine
(iii) there should not be more than 60 per cent of all advertisements in newspaper and magazine put together
(iv) there must be at least 45,00,000 exposures to families with annual income of over Rs 50,000.
Formulate this problem as an LP model to determine the number of each type of advertisement to be
given so as to maximize the total number of exposures.
LP model formulation Let x1, x2, x3 and x4 = number of television, radio, newspaper, magazine
advertisements to be pursued, respectively.
The LP model
Maximize (total number of exposures of both groups) Z
= (2,00,000 + 3,00,000) x1 + (5,00,000 + 7,00,000) x2 + (3,00,000 + 1,50,000) x3
+ (1,00,000 + 1,00,000) x4
= 5,00,000 x1 + 12,00,000 x2 + 4,50,000 x3 + 2,00,000 x4
subject to the constraints
(i) Available budget : 40,000x1 + 20,000x2 + l5,000x3  5,000x4  8,00,000
(ii) Maximum television advertisement : x1  4
(iii) Maximum magazine advertisement
x4  4 (because magazine will appear only four times in the next month)
(iv) Maximum newspaper and magazine advertisement
x3  x4
 0.60 or – 0.6x – 0.6x + 0.4x + 0.4x  0
x x x x
1 2 3 4 1 2 3 4

(v) Exposure to families with income over Rs 50,000

2,00,000x1 + 5,00,000x2 + 3,00,000x3 + 1,00,000x4  45,00,000
and x1, x2, x3, x4  0.
Example 2.19 An advertising agency is preparing an advertising campaign for a group of agencies. These
agencies have decided that different characteristics of their target customers should be given different importance
(weightage). The following table gives the characteristics with their corresponding importance (weightage).

Characteristics Weightage (%)

Age 25– 40 years 20
Annual income Above Rs 60,000 30
Female Married 50

The agency has carefully analyzed three media and has compiled the following data:

Data Item Media

Women’s Magazine (%) Radio (%) Television (%)
Reader characteristics
(i) Age: 25–40 years 80 70 60
(ii) Annual income: Above Rs 60,000 60 50 45
(iii) Females/Married 40 35 25
Cost per advertisement (Rs) 9,500 25,000 1,00,000
Minimum number of advertisement allowed 10 5 5
Maximum number of advertisement allowed 20 10 10
Audience size (1000s) 750 1,000 1,500
 Linear Programming: Applications and Model Formulation 43

The budget for launching the advertising campaign is Rs 5,00,000. Formulate this problem as an LP
model for the agency to maximize the total expected effective exposure.
LP model formulation Let x1, x2 and x3 = number of advertisements made using advertising media:
women’s magazines, radio and television, respectively.
The effectiveness coefficient corresponding to each of the advertising media is calculated as follows:

Media Effectiveness Coefficient

0.80 (0.20) + 0.60 (0.30) + 0.40 (0.50) = 0.54

Radio 0.70 (0.20) + 0.50 (0.30) + 0.35 (0.50) = 0.46
Television 0.60 (0.20) + 0.45 (0.30) + 0.25 (0.50) = 0.38
The coefficient of the objective function, i.e. effective exposure for all the three media employed, can
be computed as follows:
Effective exposure = Effectiveness coefficient × Audience size
where effectiveness coefficient is a weighted average of audience characteristics. Thus, the effective exposure
of each media is as follows:
Women’s magazine = 0.54 × 7,50,000 = 4,05,000
Radio = 0.46 × 10,00,000 = 4,60,000
Television = 0.38 × 15,00,000 = 5,70,000
The LP model
Maximize (effective exposure) Z = 4,05,000x1 + 4,60,000x2 + 5,70,000x3
subject to the constraints
(i) Budget: 9,500x1 + 25,000x2 + 1,00,000x3  5,00,000
(ii) Minimum number of advertisements allowed
(a) x1  10; (b) x2  5; and (c) x3  5
(iii) Maximum number of advertisements allowed constraints
(a) x1  20; (b) x2  10; and (c) x3  10
and x1, x2, x3  0.

2.8.3 Examples on Finance

Example 2.20 An engineering company planned to diversify its operations during the year 2005-06.
The company allocated capital expenditure budget equal to Rs 5.15 crore in the year 2005 andRs 6.50
crore in the year 2006. The company had to take five investment projects under consideration. The estimated
net returns at that present value and the expected cash expenditures on each project in thosetwo years
are as follows.
Assume that the return from a particular project would be in direct proportion to the investment in it,
so that, for example, if in a project, say A, 20% (of 120 in 2005 and of 320 in 2006) was invested, then the
resulting net return in it would be 20% (of 240). This assumption also implies that individuality of the project
should be ignored. Formulate this capital budgeting problem as an LP model to maximize the net return.

Project Estimated Net Returns Cash Expenditure

(in ’000 Rs) (in ’000 Rs)
Year 2005 Year 2006
A 240 120 320
B 390 550 594
C 80 118 202
D 150 250 340
E 182 324 474

LP model formulation Let x1, x2, x3, x4 and x5 = proportion of investment in projects A, B, C, D and
E, respectively.
The LP model
Maximize (net return) = 240x1 + 390x2 + 80x3 + 150x4 + 182x5
44 Operations Research: Theory and Applications

subject to the constraints

(i) Capital expenditure budget
(a) 120x1 + 550x2 + 118x3 + 250x4 + 324x5  515 [For year 2005]
(b) 320x1 + 594x2 + 202x3 + 340x4 + 474x5  650 [For year 2006]
(ii) 0-1 integer requirement constraints
(a) x1  1, (b) x2  1, (c) x3  1, (d) x4  1, (e) x5  1
and x1, x2, x3, x4, x5  0
Example 2.21 XYZ is an investment company. To aid in its investment decision, the company has developedthe
investment alternatives for a 10-year period, as given in the following table. The return on investmentis
expressed as an annual rate of return on the invested capital. The risk coefficient and growth potentialare
subjective estimates made by the portfolio manager of the company. The terms of investment is the average
length of time period required to realize the return on investment as indicated.

Investment Length of Annual Rate of Risk Coefficient Growth Potential

Alternative Investment Return (Year) Return (%)
A 4 3 1 0
B 7 12 5 18
C 8 9 4 10
D 6 20 8 32
E 10 15 6 20
F 3 6 3 7
Cash 0 0 0 0

The objective of the company is to maximize the return on its investments. The guidelines for selecting
the portfolio are:
(i) The average length of the investment for the portfolio should not exceed 7 years.
(ii) The average risk for the portfolio should not exceed 5.
(iii) The average growth potential for the portfolio should be at least 10%.
(iv) At least 10% of all available funds must be retained in the form of cash, at all times.
Formulate this problem as an LP model to maximize total return.
LP model formulation Let xj = proportion of funds to be invested in the jth investment alternative
( j = 1, 2, . . ., 7)
The LP model
Maximize (total return) Z = 0.03x1 + 0.12x2 + 0.09x3 + 0.20x4 + 0.15x5 + 0.06x6 + 0.00x7
subject to the constraints
(i) Length of investment : 4x1 + 7x2 + 8x3 + 6x4 + 10x5 + 3x6 + 0x7  7
(ii) Risk level : x1 + 5x2 + 4x3 + 8x4 + 6x5 + 3x6 + 0x7  5
(iii) Growth potential : 0x1 + 0.18x2 + 0.10x3 + 0.32x4 + 0.20x5 + 0.07x6 + 0x7  0.10
(iv) Cash requirement : x7  0.10
(v) Proportion of funds : x1 + x2 + x3 + x4 + x5 + x6 + x7 = 1
and x1, x2, x3, x4, x5, x6, x7  0.
Example 2.22 An investor has three investment opportunities available to him at the beginning of each years,
for the next 5 years. He has a total of Rs 5,00,000 available for investment at the beginning of thefirst
year. A summary of the financial characteristics of the three investment alternatives is presented in the
following table:

Investment Allowable Size of Return (%) Timing of Return Immedia te

Alternative Initial Investment Reinvestment Possible?
(Rs)
1 1,00,000 19 1 year later yes
2 unlimited 16 2 years later yes
3 50,000 20 3 years later yes
 Linear Programming: Applications and Model Formulation 45

The investor wishes to determine the investment plan that will maximize the amount of money which
can be accumulated by the beginning of the 6th year in the future. Formulate this problem as an LP model
to maximize total return. [Delhi Univ., MBA, 2002, 2008]
LP model formulation Let
xij = amount to be invested in investment alternative, i (i = 1, 2, 3) at the beginning of the year
j ( j =1, 2, . . ., 5)
yj = amount not invested in any of the investment alternatives in period j
The LP model
Minimize (total return) Z = 1.19x15 + 1.16x24 + 1.20x33 + y5
subject to the constraints
(i) Yearly cash flow
(a) x11  x21  x31  y1  5,00,000 (year 1)
(b)  y1  1.19x11  x12  x22  x32  y2  0 (year 2)
(c) – y2 – 1.16x21 – 1.19x12 + x23 + x23 + x33 + y3 = 0
(d) – y3 – 1.20x31 – 1.16x22 – 1.19x13 + x14 + x24 + x34 + y4 = 0 (year 4)
(e) – y4 – 1.20x32 – 1.16x23 – 1.19x14 + x15 + x25 + x35 + y5 = 0 (year 5)
(ii) Size of investment
x11  1,00,000 , x12  1,00,000 , x13  1,00,000 , x14  1,00,000 , x15  1,00,000
x31  50,000 , x32  50,000 , x33  50,000 , x34  50,000 , x35  50,000
and xij , y j  0 for all i and j.
Remark To formulate the first set of constraints of yearly cash flow, the following situation is adopted:
Investment alternatives Investment alternatives

x12  x22  x32  y2 y1 1.19x11

or – y1 – 1.19x11 + x12 + x22 + x32 + y2 = 0.

Example 2.23 A leading CA is attempting to determine the ‘best’ investment portfolio and is considering six
alternative investment proposals. The following table indicates point estimates for the price per share, the
annual growth rate in the price per share, the annual dividend per share and a measure of the risk associated
with each investment.
Portfolio Data
Shares Under Consideration A B C D E F
Current price per share (Rs) 80.00 100.00 160.00 120.00 150.00 200.00
Projected annual growth rate 0.08 0.07 0.10 0.12 0.09 0.15
Projected annual dividend per share (Rs) 4.00 4.50 7.50 5.50 5.75 0.00
Projected risk return 0.05 0.03 0.10 0.20 0.06 0.08

The total amount available for investment is Rs 25 lakh and the following conditions are required to
be satisfied:
(i) The maximum rupee amount to be invested in alternative F is Rs 2,50,000.
(ii) No more than Rs 5,00,000 should be invested in alternatives A and B combined.
(iii) Total weighted risk should not be greater than 0.10, where
(Amount invested in alternative j) (Risk of alternative j)
Total weighted risk =
Total amount invested in all the alternatives
(iv) For the sake of diversity, at least 100 shares of each stock should be purchased.
(v) At least 10 per cent of the total investment should be in alternatives A and B combined.
(vi) Dividends for the year should be at least 10,000.
Rupee return per share of stock is defined as the price per share one year hence, less current price
per share plus dividend per share. If the objective is to maximize total rupee return, formulate this problem
as an LP model for determining the optimal number of shares to be purchased in each of the shares under
consideration. You may assume that the time horizon for the investment is one year.
46 Operations Research: Theory and Applications

LP model formulation Let x1, x2, x3, x4, x5 and x6 = number of shares to be purchased in each of
the six investment proposals A, B, C, D, E and
F, respectively.
Rupee return per share = Price per share one year hence – Current price per share + Dividend
per share
= Current price per share × Projected annual growth rate (i.e. Projected
growth each year + Dividend per share).
Thus, we compute the following data:

Investment Alternatives : A B C D E F
No. of shares purchased : x1 x2 x3 x4 x5 x6
Projected growth for each share (Rs) : 6.40 7.00 16.00 14.40 13.50 30.00
Projected annual dividend per share (Rs) : 4.00 4.50 7.50 5.50 5.75 0.00
Return per share (Rs) : 10.40 11.50 23.50 19.90 19.25 30.00

The LP model
Maximize (total return) R = 10.40x1 + 11.50x2 + 23.50x3 + 19.90x4 + 19.25x5 + 30.00x6
subject to the constraints
(i) 80x1 + l00x2 + 160x3 + 120x4 + 150x5 + 200x6  25,00,000 (total fund available)
(ii) 200x6  2,50,000 [from condition (i)]
(iii) 80x1 + l00x2  5,00,000 [from condition (ii)]
80 x1 ( 0.05)  100 x2 ( 0.03)  160 x3 ( 0.10)  120 x4 ( 0.02)  150 x5 (0.06)  200 x6 (0.08)
(iv) 1
80 x1  100 x2  160 x3  120 x4  150 x5  200 x6
4x1 + 3x2 + 16x3 + 24x4 + 9x5 + 16x6  8x1 + l0x2 + 16x3 + 12x4 + 15x5 + 20x6
– 4x1 – 7x1 + 0x3 + 12x4 – 6x5 – 4x6  0
(v) x1  100, x2  100, x3  100, x4  100, x5  100, x6  100 [from condition (iv)]
(vi) 80x1 + l00x2  0.10 (80x1 + l00x2 + 160x3 + 120x4 + 150x5 + 200x6) [from condition (v)]
80x1 + l00x2  8x1 + l0x2 + 16x3 + 12x4 + 15x5 + 20x6
72x1 + 90x2 – 16x3 – 12x4 – 15x5 – 20x6  0
(vii) 4x1 + 4.5x2 + 7.5x3 + 5.5x4 + 5.75x5  10,000 [from condition (vi)]
and xj  0; j = 1, 2, 3, 4, 5 and 6.
Example 2.24 A company must produce two products over a period of three months. The company can
pay for materials and labour from two sources: company funds and borrowed funds.
The firm has to take three decisions:
(a) How many units of product 1 should it produce?
(b) How many units of product 2 should it produce?
(c) How much money should it borrow to support the production of the products?
The firm must take these decisions in order to maximize the profit contribution, subject to the conditions
stated below:
(i) Since the company’s products enjoy a seller’s market, the company can sell as many units as it
can produce. The company would therefore like to produce as many units as possible, subject
to its production capacity and financial constraints. The capacity constraints, together with cost
and price data, are shown in the following table:
Capacity, Price and Cost Data
Product Selling Price Cost of Production Required Hours per Unit in
(Rs per Unit) (Rs per Unit) Department

A B C
1 14 10 0.5 0.3 0.2
2 11 8 0.3 0.4 0.1
Available hours per production period of three months : 500.00 400.00 200.00
(ii) The available company funds during the production period will be Rs 3 lakh.
 Linear Programming: Applications and Model Formulation 47

(iii) A bank will give loans up to Rs 2 lakh per production period at an interest rate of 20 per cent
per annum provided that company’s acid (quick) test ratio is at 1 to 1 while the loan is outstanding.
Take a simplified acid-test ratio given by
Surplus cash on hand after production + Accounts receivable
Bank borrowings + Interest occurred thereon
(iv) Also make sure that the needed funds are made available for meeting production costs. Formulate
this problem as an LP model.
LP model formulation Let x1, x2 = number of units of products 1 and 2 produced, respectively.
x3 = amount of money borrowed.
The LP model
Profit contribution per unit of each product = (Selling price – Variable cost of production)
Maximize Z = Total profit by producing two products – Cost of borrowed money
= (14 – 10)x1 + (11 – 8)x2 – 0.05x3 = 4x1 + 3x2 – 0.05x3
(since the interest rate is 20 per cent per annum, it will be 5 per cent for a period
of three months)
subject to the constraints
(i) The production capacity constraints for each department
(a) 0.5x1 + 0.3x2  500, (b) 0.3x1 + 0.4x2  400, (c) 0.2x1 + 0.lx2  200
(ii) The funds available for production are the sum of Rs 3,00,000 in cash that the firm has and
borrowed funds maximum up to Rs 2,00,000. Consequently, production is limited to the extent
that the funds are available to pay for production costs. Thus, we write the constraint as:
Funds required for production  Funds available
10x1 + 8x2  3,00,000 + x3
10x1 + 8x2 – x3  3,00,000
(iii) Borrowed funds constraint [from condition (iii) of the problem]
x1  2,00,000
(iv) Acid-test condition constraint
Surplus cash on hand after production + Accounts receivable
1
Bank borrowings + Interest accrued thereon
(3, 00, 000  x3 10x1  8x2 ) 14x1 11x2
1
x3  0.2x3
3, 00, 000  x3  4x1  3x2  x3  0.2x3
or  4 x1  3x2  0.2 x3  3,00,000
and x1, x2, x3  0.
Example 2.25 The most recent audited summarized balance sheet of Shop Financial Service is given below:
The company intends to enhance its investment in the lease portfolio by another Rs 1,000 lakh. For
this purpose, it would like to raise a mix of debt and equity in such a way that the overall cost of raising
additional funds is minimized. The following constraints apply to the way the funds can be mobilized:
(i) Total debt divided by net owned funds, cannot exceed 10.
(ii) Amount borrowed from financial institutions cannot exceed 25 per cent of the net worth.
(iii) Maximum amount of bank borrowings cannot exceed three times the net owned funds.
Balance Sheet as on 31 March 2008
Liabilities (Rs lakh) Assets (Rs lakh)
Equity Share Capital 65 Fixed Assets:
Reserves & Surplus 110 Assets on Lease
(Original Cost: Rs 550 lakhs) 375
Term Loan from IFCI 80 Other Fixed Assets 50
Public Deposits 150 Investments (on wholly owned subsidiaries) 20
Bank Borrowings 147 Current Assets:
Other Current Liabilities 50 Stock on Hire 80
602 Receivables 30
Other Current Assets 35
Miscellaneous Expenditure (not written off) 12
602
48 Operations Research: Theory and Applications

(iv) The company would like to keep the total public deposit limited to 40 per cent of the total debt.
The post-tax costs of the different sources of finance are as follows:
Equity Term Loans Public Deposits Bank Borrowings
2.5% 8.5% 7% 10%
Formulate this problem as an LP model to minimize cost of funds raised.
Note: (a) Total Debt = Term loans from Financial Institutions + Public deposits + Bank borrowings
(b) Net worth = Equity share capital + Reserves and surplus
(c) Net owned funds = Net worth – Miscellaneous expenditures
LP model formulation Let x1, x2, x3 and x4 = quantity of additional funds (in lakh) raised on account
of additional equity, term loans, public deposits, bank
borrowings, respectively.
The LP model
Minimize (cost of additional funds raised) Z = 0.025x1 + 0.085x2 + 0.07x3 + 0.1x4
subject to the constraints
Total Debt Existing debt+Additional total debt
(i) Net owned funds  10 or  10
(Equity share capital + Reserve & surplus
+ Additional equity  Misc. exp.)

80  150  147  x2  x3  x4 x2  x3  x4  377

 10 or  10
(65  110  x1)  12 x1  163

x2 + x3 + x4 + 377  10 x1 + 1,630 or –10x1 + x2 + x3 + x4  1,253.

Amount borrowed (from financial institutions)  25% of net worth
(ii)
or (Existing long-term loan from financial institutions + Additional loan)
 25% (Existing equity capital + Reserve & surplus + Addl. equity capital)
80 + x2  0.25 (175 + x1)
320 + 4x1  175 + x1
– x1 + 4x2  – 145 or x1 – 4x2  145.
Maximum bank borrowings  3 (Net owned funds)
(iii)
or (Existing bank borrowings + Addl. bank borrowings  3 (Existing equity capital + Reserves
& surplus + Addl. equity capital – Misc. exp.)
(147 + x4)  3 (65 + 110 + x1 – 12)
x4 – 3x1  525 – 36 – 147
–3x1 + x4  342.
Total public deposit  40% of total debt.
(iv)
or (Existing public deposits + Addl. public deposits)  0.40 (Existing total debt + Addl. total
debt)
or 150 + x3  0.40 (80 + 150 + 147 + x2 + x3 + x4) or 150 + x3  0.40 (x2 + x3 + x4 + 377)
1,500 + l0x3  4x2 + 4x3 + 4x4 + 1,508 or – 4x2 + 6x3 – 4x4  8.
Addl. equity capital + Addl. term loan + Addl. public deposits + Addl. bank borrowings = 1,000
(v)
(since the company wants to enhance the investment by Rs 1,000 lakh)
or x1 + x2 + x3 + x4 = 1,000
and x1, x2, x3, x4  0.
Example 2.26 Renco-Foundries is in the process of drawing up a Capital Budget for the next three years. It
has funds to the tune of Rs 1,00,000 that can be allocated among projects A, B, C, D and E. The net cash
flows associated with an investment of Re 1 in each project are provided in the following table.
Cash Flow at Time
Investment in 0 1 2 3
A – Re 1 + Re 0.5 + Re 1 Re 0
B Re 0 – Re 1 + Re 0.5 + Re 1
C – Re 1 + Rs 1.2 Re 0 Re 0
D – Re 1 Re 0 Re 0 Rs 1.9
E Re 0 Re 0 – Re 1 Rs 1.5

Note: Time 0 = present, Time 1 = 1 year from now. Time 2 = 2 years from now. Time 3 = 3 years from now.
 Linear Programming: Applications and Model Formulation 49

For example, Re 1 invested in investment B requires a Re 1 cash outflow at time 1 and returns Re 0.50
at time 2 and Re 1 at time 3.
To ensure that the firm remains reasonably diversified, the firm will not commit an investment exceeding
Rs 75,000 for any project. The firm cannot borrow funds and therefore, the cash available for investment
at any time is limited to the cash in hand. The firm will earn interest at 8 per cent per annum by parking
the un-invested funds in money market investments. Assume that the returns from investments can be
immediately re-invested. For example, the positive cash flow received from project C at time 1 can immediately
be re-invested in project B. Formulate this problem as an LP model so as to maximize cash on hand at
time 3. [CA, 2000; Delhi Univ., MBA, 2007]
LP model formulation Let x1, x2, x3, x4 and x5 = Amount of rupees invested in investments A, B,
C, D and E, respectively.
si = Money invested in money market instruments at
time i (for i = 0, 1, 2).
Firm earns interest at 8 per cent per annum by parking the un-invested funds in money market
instruments, hence Rs s0, Rs s1 and Rs s2 which are invested in these instruments at times 0, 1 and 2 will
become 1.08s0, 1.08s1 and 1.08s2 at times 1, 2 and 3, respectively.
Note: Cash available for investment in time t = cash on hand at time t.
From the given data, it can be computed that at time 3:
Cash on hand = x1 × 0 + x2 × 1 + x3 × 0 + 1.9x4 + 1.5x5 + 1.08s2
= Rs (x2 + 1.9x4 + 1.5x5 + 1.08s2)
The LP model
Maximize (Cash on hand at time 3) Z = x2 + 1.9x4 + 1.5x5 + 1.08s2
subject to the constraints
At time 0: Total fund of Rs 1,00,000 is available for investing on projects A, C and D. That is
x1 + x2 + x3 + s0 = 1,00,000
At time 1: Rs 0.5x1, Rs 1.2x3, and Rs 1.08s0 will be available as a result of investment made at time 0. Since
Rs x2 and s1 are invested in project B and money market instruments, respectively at time 1,
therefore we write
0.5x1 + 1.2x3 + 1.08s0 = x2 + s1
At time 2: Rs x1; Rs 0.5x2 and Rs 1.08 s1 will be available for investment. As Rs x5 and Rs s2 are invested
at time 2. Thus
x1 + 0.5x2 + 1.08s1 = x5 + s2
Also, since the company will not commit an investment exceeding Rs 75,000 in any project,
therefore the constraint becomes: xi  75,000 for i = 1, 2, 3, 4, 5.
and x1, x2, x3, x4, x5, s0, s1, s2  0.
2.8.4 Examples on Agriculture
Example 2.27 A cooperative farm owns 100 acres of land and has Rs 25,000 in funds available for
investment. The farm members can produce a total of 3,500 man-hours worth of labour during September – May
and 4,000 man-hours during June – August. If any of these man-hours are not needed, some members ofthe
firm would use them to work on a neighbouring farm for Rs 2 per hour during September – May and Rs 3 per hour
during June – August. Cash income can be obtained from the three main crops and two types of livestock: dairy
cows and laying hens. No investment funds are needed for the crops. However, each cow will require an
investment outlay of Rs 3,200 and each hen will require Rs 15.
In additon each cow will also require 15 acres of land, 100 man-hours during the summer. Each cow will
produce a net annual cash income of Rs 3,500 for the farm. The corresponding figures for each henare:
no acreage, 0.6 man-hours during September – May; 0.4 man-hours during June – August, and an annual net
cash income of Rs 200. The chicken house can accommodate a maximum of 4,000 hens and the size of the
cattle-shed limits the members to a maximum of 32 cows.
Estimated man-hours and income per acre planted in each of the three crops are:
Paddy Bajra Jowar
Man-hours
September-May 40 20 25
June-August 50 35 40
Net annual cash income (Rs) 1,200 800 850
50 Operations Research: Theory and Applications

The cooperative farm wishes to determine how much acreage should be planted in each of the crops
and how many cows and hens should be kept in order to maximize its net cash income. Formulate this
problem as an LP model to maximize net annual cash income.
LP model formulation The data of the problem is summarized as follows:
Constraints Cows Hens Crop Extra Hours Total
Paddy Bajra Jowar Sept–May June–Aug Availability

Man-hours
Sept–May 100 0.6 40 20 25 1 – 3,500
June–Aug 50 0.4 50 35 40 – 1 4,000
Land 1.5 – 1 1 1 – – 100
Cow 1 – – – – – – 32
Hens – 1 – – – – – 4,000
Net annual cash
income (Rs) 3,500 200 1,200 800 850 2 3

Decision variables Let

x1 and x2 = number of dairy cows and laying hens, respectively.
x3, x4 and x5 = average of paddy crop, bajra crop and jowar crop, respectively.
x6 = extra man-hours utilized in Sept–May.
x7 = extra man-hours utilized in June–Aug.
The LP model
Maximize (net cash income) Z = 3,500x1 + 200x2 + 1,200x3 + 800x4 + 850x5 + 2x6 + 3x7
subject to the constraints
(i) Man-hours: 100x1 + 0.6x2 + 40x3 + 20x4 + 25x5 + x6 = 3,500 (Sept-May duration)
50x1 + 0.4x2 + 50x3 + 35x4 + 40x5 + x7 = 4,000 (June-Aug duration)
(ii) Land availability: 1.5x1 + x3 + x4 + x5  100
(iii) Livestock: (a) x1  32 (dairy cows), (b) x2  4,000 (laying hens)
and x1, x2, x3, x4, x5, x6, x7  0.
Example 2.28 A certain farming organization operates three farms of comparable productivity. The output
of each farm is limited both by the usable acreage and by the amount of water available for irrigation. The
data for the upcoming season is as shown below:

Farm Usable Acreage Water Available

(in cubic feet)
1 400 1,500
2 600 2,000
3 300 900

The organization is considering planting crops which differ primarily in their expected profit per acre
and in their consumption of water. Furthermore, the total acreage that can be devoted to each of the crops
is limited by the amount of appropriate harvesting equipment available.

Crop Maximum Water Consumption Expected Profit

Acreage (in cubic feet) per Acre (Rs)
A 700 5 4,000
B 800 4 3,000
C 300 3 1,000

In order to maintain a uniform workload among the three farms, it is the policy of the organization that
the percentage of the usable acreage planted be the same for each farm. However, any combination of the
crops may be grown at any of the farms. The organization wishes to know how much of each crop should
be planted at the respective farms in order to maximize expected profit.
Formulate this problem as an LP model in order to maximize the total expected profit.
 Linear Programming: Applications and Model Formulation 51

LP model formulation The data of the problem is summarized below:

Crop Farm Crop Expected

Requirement Profit per Acre
1 2 3
(in acres) (Rs)

A x11 x12 x13 700 4,000

B x21 x22 x23 800 3,000
C x31 x32 x 300 1,000
Usable acreage 400 600 300
Water available per acre 1,500 2,000 900

Decision variables Let xij = number of acres to be allocated to crop i (i = 1, 2, 3) to farm j ( j = 1, 2)

The LP model
Maximize (net profit) Z = 4,000(x11 + x12 + x13) + 3,000(x21 + x22 + x23) + l,000(x31 + x32 + x)
subject to the constraints
(i) Crop requirement
(a) x11 + x12 + x13  700, (b) x21 + x22 + x23  800, (c) x31 + x32 + x  300
(ii) Available acreage
(a) x11 + x21 + x31  400, (b) x12 + x22 + x32  600, (c) x13 + x23 + x  300
(iii) Water available (in acre feet)
(a) 5x11 + 4x21 + 3x31  1,500, (b) 5x12 + 4x22 + 3x32  2,000, (c) 5x13 + 4x23 + 3x  900
(iv) Social equality
x11  x21  x31 x12  x22  x32 x12  x22  x32 x13  x23  x33
(a)  , (b)  ,
400 600 600 300
x13  x23  x33 x11  x21  x31
(c) 
300 400
and xij  0 for all i and j.

2.8.5 Examples on Transportation

Example 2.29 ABC manufacturing company wishes to develop its monthly production schedule for the next
three months. Depending upon the sales commitments, the company can either keep the production constant,
allowing fluctuation in inventory; or its inventories can be maintained at a constant level, with fluctuating
production. Fluctuating production makes overtime work necessary, the cost of which is estimated to be double
the normal production cost of
Rs 12 per unit. Fluctuating inventories Month Production Capacity (units) Sales
Regular Overtime (units)
result in an inventory carrying cost of Rs
2 per unit/month. If the company fails to 1 50 30 60
fulfil its sales commitment, it incurs a 2 50 0 120
shortage cost of Rs 4 per unit/month. The 3 60 50 40
production capacities for the next three
months are in the table:
Formulate this problem as an LP model to minimize the total production cost.
[Delhi Univ., MBA, 2008]
LP model formulation The data of the problem is summarized as follows:

Month Production Capacity Sales

Regular Overtime
1 50 30 60
2 50 0 120
3 60 50 40
52 Operations Research: Theory and Applications

Normal production cost : Rs 12 per unit Overtime cost : Rs 24 per unit

Carrying cost : Rs 2 per unit per month Shortage cost : Rs 4 per unit per month Assume
five sources of supply: three regular and two overtime (because the second months overtime
production is zero) production capacities. The demand for the three months will be the sales during these months.
All supplies against the order have to be made and can be made in the subsequent month if it is not possible
to make them during the month of order, with additional cost equivalent to shortage cost, i.e. in month 2. The
cumulative production of months 1 and 2 in regular and overtime is 130 units while the orders are for 180 units. This
balance can be supplied during month 3 at an additional production cost of Rs 4.
The given information can now be presented in matrix form as follows:
M1 M2 M3 Production (supply)
M1 12 14 16 50
M2 16 12 14 50
M3 20 16 12 60

M1(OT) 24 26 28 30
M2(OT) 32 28 24 50
Sales (demand) 60 120 40
Decision variables Let xij = amount of commodity sent from source of supply i (i = 1, 2, . . ., 5) to
destination j ( j = 1, 2, 3)
The LP model
Minimize (total cost) Z = 12x11 + l4x12 + 16x13 + 16x21 + l2x22 + 14x23 + 20x31 + 16x32 + 12x + 24x41
+ 26x42 + 28x43 + 32x51 + 28x52 + 24x53
subject to the constraints
(i) Production (supply) constraints
(a) x11 + x12 + x13 = 50, (b) x21 + x22 + x23 = 50, (c) x31 + x32 + x = 60,
(d) x41 + x42 + x43 = 30, (e) x51 + x52 + x53 = 30
(ii) Sales (demand) constraints
(a) x11 + x21 + x31 + x41 + x51 = 60, (b) x12 + x22 + x32 + x42 + x52 = 120,
(c) x13 + x23 + x + x43 + x53 = 40
and xij  0 for all i and j.
Example 2.30 A trucking firm has received an order to move 3,000 tonnes of industrial material to a
destination 1,000 km away. The firm has available, at the moment, a fleet of 150 class-A 15-tonne trailer
trucks and another fleet of 100 class-B 10-tonne trailer trucks. The operating costs of these trucks are
Rs 3 and Rs 4 per tonne per km, respectively. Based on past experience, the firm has a policy of retaining
at least one class-A truck with every two class-B trucks in reserve. It is desired to know how many of
these two classes of vehicles should be despatched to move the material at minimal operating costs. Formulate
this problem as an LP model.
LP model formulation Let x1 and x2 = number of class A and B trucks to be despatched, respectively.
The LP model
Minimize (total operating cost) Z = 3x1 + 4x2
subject to the constraints
15x1 + 10x2  3,000
UV
x 1  149 (due to the policy of retaining at least one class-
x  98 W
2
A truck with every two class-B truck in reserve)
and x , x  0.
1 2

Example 2.31 A ship has three cargo loads – forward, after and centre. Their capacity limits are:

Weight (kg) Volume (cu cm)

Forward 2,000 1,00,000
Centre 3,000 1,35,000
After 1,500 30,000
 Linear Programming: Applications and Model Formulation 53

The following cargos are offered to be carried in the ship. The ship owner may accept all or any part
of each commodity:

Commodity Weight (kg) Volume (cu cm) Profit (in Rs) per kg
A 6,000 60 60
B 4,000 50 80
C 2,000 25 50

In order to preserve the trim of the ship, the weight in each cargo must be proportional to the capacity
in kg. The cargo is to be distributed in a way so as to maximize profit. Formulate this problem as an LP
model.
LP model formulation xiA, xiB and xiC = weight (in kg) of commodities A, B and C to be
accommodated in the direction i(i = 1, 2, 3 – forward, centre
and after), respectively.
The LP model
Maximize (total profit) Z  60 (x1A  x2 A  x3 A )  80 (x1B  x2B  x3B )  50 (x1C  x2C  x3C )
subject to the constraints
x1B + x2B + x3B  4,000; x1B + x2B + x3B  4,000; x1B
+ x2B + x3B  4,000; x1A + x1B + x1C  2,000 x1A +
x2B + x3C  4,000; x3A + x3B + x3C  1,500
60x1A + 50x1B + 25x1C  1,00,000
60x2A + 50x2B + 25x2C  1,35,000
60x3A + 50x3B + 25x3C  30,000
and xiA , xiB , xiC  0, for all i.

2.8.6 Examples on Personnel

Example 2.32 Evening shift resident doctors in a government hospital work five consecutive days and have
two consecutive days off. Their five days of work can start on any day of the week and their schedule rotates
indefinitely. The hospital requires the following minimum number of doctors to work on the given days:
Sun Mon Tues Wed Thus Fri Sat
35 55 60 50 60 50 45
No more than 40 doctors can start their five working days on the same day. Formulate this problem
as an LP model to minimize the number of doctors employed by the hospital.
[Delhi Univ., MBA (HCA), 2006]
LP model formulation Let xj = number of doctors who start their duty on day j ( j = 1, 2, . . ., 7) of
the week.
The LP model
Minimize (total number of doctors) Z = x1 + x2 + x3 + x4 + x5 + x6 + x7
subject to the constraints
(i) x1 + x4 + x5 + x6 + x7  35, (ii) x2 + x5 + x6 + x7 + x1  55 (iii)
x3 + x6 + x7 + x1 + x2  60, (iv) x4 + x7 + x1 + x2 + x3  50 (v) x5 +
x1 + x2 + x3 + x4  60, (vi) x6 + x2 + x3 + x4 + x5  50
(vii) x7 + x3 + x4 + x5 + x6  45, (viii) xj  40
and x j  0 for all j.
Example 2.33 A machine tool company conducts on-the-job training programme for machinists. Trained
machinists are used as teachers for the programme, in the ratio of one for every ten trainees. The training
programme lasts for one month. From past experience it has been found that out of the ten trainees hired,
only seven complete the programme successfully and the rest are released.
Trained machinists are also needed for machining. The company’s requirement for machining for the
next three months is as follows: January 100, February 150 and March 200. In addition, the company requires
250 machinists by April. There are 130 trained machinists available at the beginning of the year. Pays per
month are:
54 Operations Research: Theory and Applications

Each trainee : Rs 4,400

Each trained machinist
(machining and teaching) : Rs 4,900
Each trained machinist idle : Rs 4,700
Formulate this problem as an LP model to minimize the cost of hiring and training schedule and the
company’s requirements.
LP model formulation Let
x1, x2 = trained machinist teaching and idle in January, respectively
x3, x4 = trained machinist teaching and idle in February, respectively
x5, x6 = trained machinist teaching and idle in March, respectively
The LP model
Minimize (total cost) Z = Cost of training programme (teachers and trainees) + Cost of idle machinists
+ Cost of machinists doing machine work (constant)
= 4,400 (10x1 + 10x3+ 10x5) + 4,900 (x1 + x3 + x5) + 4,700 (x2 + x4 + x6)
subject to the constraints
(i) Total trained machinists available at the beginning of January
= Number of machinists doing machining + Teaching + Idle
130 = 100 + x1 + x2 or x1 + x2 = 30
(ii) Total trained machinists available at the beginning of February
= Number of machinists in January + Joining after training programme
130 + 7x1 = 150 + x3 + x4 or 7x1 – x3 – x4 = 20
In January there are 10x1 trainees in the programme and out of those only 7x1 will become trained
machinists.
(iii) Total trained machinists available at the beginning of March
= Number of machinists in January + Joining after training
programme in January and February
130 + 7x1 + 7x3 = 200 + x5 + x6
7x1 + 7x3 – x5 – x6 = 70
(iv) Company requires 250 trained machinists by April
130 + 7x1 + 7x3 + 7x5 = 250
7x1 + 7x3 + 7x5 = 120
and x1, x2, x3, x4, x5, x6  0.
Example 2.34 The super bazzar in a city daily needs anything between 22 to 30 workers in the bazzar
depending on the time of day. The rush hours are between noon and 2 pm. The table indicates the number
of workers needed at various hours when the bazzar is open.
The super bazzar now employs 24 full-time workers, but also needs a few part-time workers. A part-
time worker must put in exactly 4 hours per day, but can start any time between 9 am and 1 pm. Full- time
workers work from 9 am to 5 pm but are allowed an hour for lunch (half of the full-timers eat at 12 noon, the
other half at 1 am). Full-timers thus provide 35 hours per week of productive labour time.
The management of the super bazzar limits part-time
hours to a maximum of 50 per cent of the day’s total Time Period Number of Workers
requirement. Needed
Part-timers earn Rs 28 per day on the average, while 9 AM – 11 A M 22
full-timers earn Rs 90 per day in salary and benefits on the 11 AM – 1 P M 30
average. The management wants to set a schedule that would 1 PM – 3 P M 25
minimize total manpower costs. 3 PM – 5 P M 23
Formulate this problem as an LP model to minimize total
daily manpower cost.
LP model formulation Let
y = full-time workers
xj = part-time workers starting at 9 am, 11 am and 1 pm, respectively ( j = 1, 2, 3)
 Linear Programming: Applications and Model Formulation 55

The LP model
Minimize (total daily manpower cost) Z = 90y + 28(x1 + x2 + x3)
subject to the constraints
1
y  x  x  30
(i) y  x1  22 [9 am – 11 am], (ii) 1 2 [11 am – 1 pm],
2

1
(iii) y  x  x  25
2 3 [1 pm – 3 pm], (iv) y  x3  23 [3 pm – 5 pm],
2

(v) y  24 [Full-timers available], (iv) 4 ( x1  x2  x3 )  0.50 (22  30  25  23)

[Part-timers’ hours cannot exceed 50% of total hours required each day which is the sum of the
workers needed each hour]
and y, xj  0 for all j.
Example 2.35 The security and traffic force, on the eve of Republic
Time Number of Officers
Day, must satisfy the staffing requirements as shown in the table.
Required
Officers work 8-hour shifts starting at each of the 4-hour intervals
as shown below. How many officers should report for duty at the 0:01 – 4:00 5
beginning of each time period in order to minimize the total number 4:01 – 8:00 7
of officers needed to satisfy the requirements? 8:01 – 12:00 15
Formulate this problem as an LP model so as to determine the 12:01 – 16:00 7
16:01 – 20:00 12
minimum number of officers required on duty at beginning of each
20:01 – 24:00 9
time period.
LP model formulation xi = number of officers who start in shift i (i = 1, 2, 3, . . ., 6)
The LP model
Minimize (number of officers required on duty) Z = x1  x2  x3  x4  x5  x6
subject to the constraints
(i) x1  x2  7 , (ii) x2  x3  15 , (iii) x3  x4  7 ,
(iv) x4  x5  12 , (v) x5  x6  9 , (vi) x6  x1  5

and x j  0, for all j.

CONCEPTUAL QUESTIONS

1. (a) What is linear programming? What are its major assumptions (b) Discuss and describe the role of linear programming in
and limitations? [Delhi Univ., MBA, Nov. 2005] managerial decision-making, bringing out limitations, if any.
(b) Two of the major limitations of linear programming are: [Delhi Univ., MBA, 2003]
assumption of ‘additivity’ and ‘single objective’. Elaborate by 7. Regardless of the way one defines linear programming, certain
giving appropriate examples. [Delhi Univ., MBA, Nov. 2009 ] basic requirements are necessary before this technique can be
2. Linear programming has no real-life applications’. Do you agree employed to business problems. What are these basic
with this statement? Discuss. [Delhi Univ., MBA, 2004 ] requirements in formulation? Explain briefly.
8. Discuss in brief linear programming as a technique for resource
3. In relation to the LP problem, explain the implications of the
following assumptions of the model: utilization. [Delhi Univ., MBA (HCA), 2004]
(i) Linearity of the objective function and constraints, 9. What are the four major types of allocation problems that can
(ii) Continuous variables, be solved using the linear programming technique? Briefly explain
(iii) Certainty. each with an example.
4. What is meant by a feasible solution of an LP problem? 10. Give the mathematical and economic structure of linear
programming problems. What requirements should be met in order
5. ‘Linear programming is one of the most frequently and successfully
to apply linear programming?
applied operations research technique to managerial decisions.’
Elucidate this statement with some examples. 11. Discuss and describe the role of linear programming in managerial
[Delhi Univ., MBA, 2008] decision-making bringing out limitations, if any.
6. (a) What are the advantages and limitations of LP models? [Delhi Univ., MBA, 2004, 2009]
56 Operations Research: Theory and Applications

SELF PRACTICE PROBLEMS

Problems on Production 5. Consider a small plant which makes two types of automobile
parts, say A and B. It buys castings that are machined, bored
1. A company sells two different products A and B, making a profit and polished. The capacity of machining is 25 per hour for A
of Rs 40 and Rs 30 per unit, respectively. They are both produced and 24 per hour for B, capacity of boring is 28 per hour forA
with the help of a common production process and are sold in and 35 per hour for B, and the capacity of polishing is 35per
two different markets. The production process has a total capacity hour for A and 25 per hour for B. Castings for part A costRs 2
of 30,000 man-hours. It takes three hours to produce a unit of and sell for Rs 5 each and those for part B cost Rs 3and sell
A and one hour to produce a unit of B. The market has been for Rs 6 each. The three machines have running costsof Rs 20,
surveyed and company officials feel that the maximum number Rs 14 and Rs 17.50 per hour. Assuming that anycombination of
of units of A that can be sold is 8,000 units and that of B is parts A and B can be sold, formulate this problem as an LP model
12,000 units. Subject to these limitations, products can be sold to determine the product mix which would maximizes profit.
in any combination. Formulate this problem as an LP model to 6. On October 1, a company received a contract to supply 6,000 units
maximize profit. of a specialized product. The terms of contract require that 1,000
2. The manager of an oil refinery must decide on the optimal mix units of the product be shipped in October; 3,000 units in
of two possible blending processes of which the input and output November and 2,000 units in December. The company can
per production run are given as follows: manufacture 1,500 units per month on regular time and 750
units per month in overtime. The manufacturing cost per item
Process Input (units) Output produced during regular time is Rs 3 and the cost per item
(units) Crude A Crude B Gasoline X Gasoline Y produced during overtime is Rs 5. The monthly storage cost
is Re 1. Formulate this problem as an LP model so as to minimize
1 5 3 5 8 total costs. [Delhi Univ., MBA, 1999]
2 4 5 4 4
7. A wine maker has a stock of three different wines with the following
The maximum available amount of crude A and B are characteristics:
200 units and 150 units, respectively. Market requirements show
Wine Proofs Acids% Specific Gravity Stock (gallons)
that at least 100 units of gasoline X and 80 units ofgasoline Y
must be produced. The profit per production run from process A 27 0.32 1.70 20
1 and process 2 are Rs 300 and Rs 400, respectively. Formulate B 33 0.20 1.08 34
this problem as an LP model to maximize profit.
C 32 0.30 1.04 22
3. A firm places an order for a particular product at the beginning
of each month and that product is received at the end of the month. A good dry table wine should be between 30 and 31 degree
The firm sells during the month from the stocks and itcan sell proof, it should contain at least 0.25% acid and should havea
any quantity. specific gravity of at least 1.06. The wine maker wishes to
The prices at which the firm buys and sells vary every month. blend the three types of wine to produce as large a quantityas
The following table shows the projected buying and selling prices possible of a satisfactory dry table wine. However, his stockof
for the next four months: wine A must be completely used in the blend because further
storage would cause it to deteriorate. What quantities of wines
Month Selling Price (Rs) Purchase Price (Rs) B and C should be used in the blend. Formulate this problem
(During the Month) (Beginning of the Month) as an LP model.
April – 75 8. ABC foods company is developing a low-calorie high-protein diet
May 90 75 supplement called Hi-Pro. The specifications of Hi-Pro have been
June 60 60 established by a panel of medical experts. These specifications
July 75 – along with the calorie, protein and vitamin content of three basic
foods, are given in the following table:
As on April 1 the firm has no stocks on hand, and does not
wish to have any stocks at the end of July. The firm has a Nutritional Units of Nutritional Elements Basic Foods
warehouse of limited size, which can hold a maximum of 150 Elements (Per 100 gm Serving of Hi-Pro
units of the product. Basic Foods) Specifications
Formulate this problem as an LP model to determine the number 1 2 3
of units to buy and sell each month so as to maximizethe profits
from its operations. [Delhi Univ., MBA, 2007] Calories 350 250 200 300
4. A manufacturer produces three models (I, II and III) of a certain Proteins 250 300 150 200
product. He uses two types of raw materials (A and B) of which Vitamin A 100 150 75 100
4,000 and 6,000 units, respectively, are available. The raw Vitamin C 75 125 150 100
material requirements per unit of the three models are as follows: Cost per
The labour time of each unit of model I is twice that of model serving (Rs) 1.50 2.00 1.20
II and three times that of model III. The entire labour force of What quantities of foods 1, 2, and 3 should be used? Formulate
the factory can produce equivalent of 2,500 units of model I. this problem as an LP model to minimize cost of serving.
A market survey indicates that the minimum demand of the three [Delhi Univ., MBA (HCA), 2009]
models is: 500, 500 and 375 units, respectively. However, the
ratios of the number of units produced must be equal to 9. Omega leather goods company manufactures two types of leather
soccer balls X and Y. Each type of ball requires workby two
3 : 2 : 5. Assume that the profit per unit of models I, II and
types of employees – semi-skilled and skilled. Basically,the semi-
III are Rs 60, 40 and 100, respectively. Formulate this problem
skilled employees use machines, while the skilled employees stitch
as an LP model to determine the number of units of each product
the balls. The available time (per week) for manufacturing each
which will maximize profit.
type of employee and the time requirementfor each type of ball
are given below:
 Linear Programming: Applications and Model Formulation 57

B. One unit of food F2 contains six units of Vitamin A and three units
Type of Employee Manufacturing Time Time Available of Vitamin B, One unit of food F1 and F2 cost Rs 4 and5,
Requirement (hr) (hr/week) respectively. The minimum daily requirement (for a person)of
Vitamins A and B is 80 and 100 units, respectively. Assuming that
Ball X Ball Y
anything in excess of the daily minimum requirement of Aand
B is not harmful. Formulate this problem as an LP modelto find
Semi-skilled 2 3 80
out the optimum mixture of food F1 and F2 at the minimum cost
Skilled 4 6 150 which meets the daily minimum requirement of VitaminsA and
The cost of an hour of semi-skilled labour is Rs 5.50 and that B. [Delhi Univ., MBA (HCA), 2009]
of an hour of skilled labour is Rs 8.50. To meet the weeklydemand 13. The Omega Data Processing Company performs three types of
requirements, at least 15 balls of type X and at least10 balls of activities: payrolls, account receivables, and inventories. The profit
type Y must be manufactured. Formulate this problem as an LP and time requirements for keypunch, computation and office
model so as to minimize cost of production. printing for a ‘standard job’ are shown in the following table:
10. A pharmaceutical company has developed a new pill to be taken
by smokers that will nauseate them if they smoke. Thisnew pill Job Profit/Standard Time Requirement (Min.)
is a combination of four ingredients that are costly andin limited Job (Rs) Keypunch Computation Print
supply. The available supply and costs are as follows:
Payroll 275 1,200 20 100
Ingredient Supply Cost A/c Receivable 125 1,400 15 60
Availability (kg) (Rs/kg) Inventory 225 800 35 80
1 22 28 Omega guarantees overnight completion of all standard jobs. Any
2 18 25 job scheduled during the day can be completed during theday
3 20 52 or night. Any job scheduled during the night, however, must be
4 24 26 completed during the night. The capacity for both dayand night
are shown in the following table:
Blending requirements for this new pill are as follows:
Capacity (Min.) Keypunch Computation Print
(i) Ingredient 1 must be at least 45 per cent of the total quantity,
but cannot exceed 60 per cent of the total. Day 4,200 150 400
(ii) Ingredients 2 and 3 must each comprise at least 10 per Night 9,200 250 650
cent of the mixture, but their combined percentage cannot
exceed 25 per cent of the total quantity. Formulate this problem as an LP model to determine the ‘mix’
(iii) Ingredient 4 must not be more than 50 per cent of the total of standard jobs that should be accepted during the day and
quantity. Additionally, at least 25 kg of the pill must be night. [Delhi Univ., MBA, 2008]
produced. 14. PQR coffee company mixes South Indian, Assamese and imported
Formulate this problem as an LP model to determine optimum coffee for making two brands of coffee: plains X and plains XX. The
blending of ingredients. characteristics used in blending the coffee include strength, acidity
[Delhi Univ., MBA (HCA), 2008] and caffeine. The test results of the available supplies of South
Indian, Assamese and imported coffee are shown in the following
11. A paint manufacturing company manufactures paints at two of table:
its plants. Firm orders have been received from three large
contractors. The firm has determined that the following shipping Price per kg Strength Acidity Per cent Supply
cost data is appropriate for these contractors with respect toits (Rs) Index Index Caffeine Available
two plants: (kg)

Contractor Order Size Shipping Cost/Gallon (Rs)

South Indian, 30 6 4.0 2.0 40,000
(gallon) From Plant 1 From Plant 2 Assamese, 40 8 3.0 2.5 20,000
Imported, 35 5 3.5 1.5 15,000
A 750 1.80 2.00
B 1,500 2.60 2.20 The requirements for plains X and plains XX coffees are given
C 1,500 2.10 2.25 in the following table:

Each gallon of paint must be blended and tinted. The Plains Price Minimum Maximum Maximum Quantity
company’s costs with respect to these two operations at both Coffee per kg Strength Acidity Per Demanded
of the plants are as follows: (Rs) Caffeine (kg) Cent
X 45 6.5 3.8 2.2 35,000
Plant/Operation Hours Cost/hour Hours XX 55 6.0 3.5 2.0 25,000
Required (Rs) Available
per Gallon Assume that 35,000 kg of plains X and 25,000 kg of plains XX,
are to be sold. Formulate this problem as an LP model to maximize
Blending 0.10 3.80 300 sales.
Tinting Plant 1 0.25 3.20 360
15. A manufacturer of metal office equipments makes desks, chairs,
Blending 0.15 4.00 600
cabinets and book cases. The work is carried out in the three major
Tinting Plant 2 0.20 3.10 720 manufacturing departments: Metal stamping, Assembly and
Formulate this problem as an LP model. Finishing. Exhibits, A, B the and C give the requisite dataof the
problem.
12. Vitamins A and B are found in foods F1 and F2. One unit of food
F1 contains three units of Vitamin A and four units of Vitamin
58 Operations Research: Theory and Applications

Exhibit A
Model Foundry Machine-shop Contribution per
Department Time Required per Unit Available Model (Rs)
of Product (hrs) Products Time per X 100 200 120
Desk Chair Cabinet Bookcase Week (hrs) Y 240 150 90
Stamping 4 2 3 3 800 In addition, the material required for model X was in short supply
Assembly 10 6 8 7 1,200 and sufficient only for 140 units per week, guaranteed for the
Finishing 10 8 8 8 800 year. Formulate this problem as an LP model to determine the
Exhibit B optimal combination of output.
19. A company manufacturing television and radio sets has four major
Department Cost (Rs) of Operation per Unit of Product departments: chassis, cabinet, assembly and final testing. The
monthly capacities of these are as follows:
Desk Chair Cabinet Bookcase
Television Radio
Stamping 15 8 12 12
Assembly 30 18 24 21 Chassis 1,500 or 4,500
Finishing 35 28 25 21 Cabinet 1,000 or 8,000
Assembly 2,000 or 4,000
Exhibit C: Selling price (Rs) per unit of product Testing 3,000 or 9,000
Desk : 175 Chair : 95
Cabinet : 145 Bookcase : 130 The contribution of a television set is Rs 500 and that of a radio
set Rs 250. Assume that the company can sell any quantity of
In order to maximize weekly profits, what should be all production
either product. Formulate this problem as an LP model todetermine
programme? Assuming that the items produced can be sold,
the optimal combination of television and radio sets.
which department needs to be expanded for increasing profits?
Formulate this problem as an LP model. 20. A company wants to plan production for the ensuing year so
as to minimize the combined cost of production and inventory
16. The PQR stone company sells stone procured from any of three
storage. In each quarter of the year, demand is anticipated to
adjacent quarries. The stone sold by the companyconforms to the
be 65, 80, 135 and 75 respectively. The product can be
following specification:
manufactured during regular time at a cost of Rs 16 per unit
Material X equal to 30%
produced, or during overtime at a cost of Rs 20 per unit. The
Material Y equal to or less than 40%
table given below gives data pertinent to production capacities. The
Material Z between 30% and 40%
cost of carrying one unit in inventory per quarter is Rs 2.The
Stone from quarry A costs Rs 10 per tonne and has the following
inventory level at the beginning of the first quarter is zero.
properties:
Material X : 20%, Material Y : 60%, Material Z : 20% Stone
Quarter Capacities (units) Quarterly
from quarry B costs Rs 12 per tonne and has the following
Regular Time Overtime Demand
properties:
Material X : 40%, Material Y : 30%, Material Z : 30% Stone 1 80 10 65
from quarry C costs Rs 15 per tonne and has the following 2 90 10 80
properties: 3 95 20 135
Material X : 10%, Material Y : 40%, Material Z : 50%
4 70 10 75
From what quarries should the PQR stone company procure rocks
in order to minimize cost per tonne of rock? Formulatethis Formulate this problem as an LP model so as to minimize
problem as an LP model. [Delhi Univ., MBA, 2006] the production plus storage costs for the entire year.
17. A manufacturer produces three models (I, II and III) of a certain [Delhi Univ., MBA, 2005]
product. He uses two types of raw material: A and B of which 4,000 Problems on Marketing
and 6,000 units are available, respectively. The raw material
requirements per unit of the three models are given below: 21. Suppose a media specialist has to decide how to allocate
advertising in three media vehicles. Let xi be the number of

Raw Material Requirements per Unit of Given Model messages carried in the media, i = 1, 2, 3. The unit costs of
a message in the three media are Rs 1,000, Rs 750 andRs
I II III 500. The total budget available for the campaign is Rs 2,00,000
A 2 3 5 period of a year. The first media is a monthly magazine and it is
B 4 2 7 desired to advertise not more than one insertion in one issue. At
least six messages should appear in the second media. The
The labour time for each unit of model I is twice that of model II and number of messages in the third media should strictly lie between
three times that of model III. The entire labour force of the factory 4 and 8. The expected effective audiencefor unit message in
can produce the equivalent of 2,500 units of model I.A market the media vehicles is shown below:
survey indicates that the minimum demand for the three models
is 500, 500 and 375 units, respectively. However,the ratios of Vehicle Expected Effective Audience
the number of units produced must be equal to
1 80,000
3 : 2 : 5. Assume that the profit per unit of models I, II and
2 60,000
III is Rs 60, Rs 40 and Rs 100, respectively. Formulate thisproblem
as an LP model to determine the number of units ofeach 3 45,000
product that will the maximize amount of profit. Formulate this problem as an LP model to determine the
18. A company manufactures two models of garden rollers: X and optimum allocation that would maximize total effective audience.
Y. When preparing the 2008 budget, it was found that the 22. An advertising company is planning an advertisement campaign for
limitations on capacity were represented by the following weekly a new product recently introduced in the market. It is decidedto
production maxima: insert advertisements in three leading magazines. The
 Linear Programming: Applications and Model Formulation 59

company has made a careful analysis of three available media, and The relative importance of the three characteristics is:
has compiled the following set of relevant data: Homeowner, 0.4; Income, 0.2; Gardener, 0.4. The advertising
budget is Rs 2,00,000. Formulate this problem as an LP model
Data Item Magazines to find the most effective number of exposures in each magazine.
1 2 3 Problems on Finance
Reader characteristics 25. A gambler plays a game that requires dividing bet money among
(a) Age: 25–35 yrs 70% 80% 40% four different choices. The game has three outcomes. The following
(b) Education level: Graduation table gives the corresponding gain (or loss) per rupee deposited in
and above 80% 60% 50% each of the four choices for the three outcomes:
(c) Income: Rs 5,000 and above 60% 70% 40%
Outcome Gain (or loss) Deposited in
Minimum number of advertisements 20 10 5
per Rupee Given Choice
Maximum number of advertisements 50 40 30
Cost per advertisement (Rs) 1,000 700 500 1 2 3 4
Readership 3,20,000 5,00,000 2,00,000
1 – 3 4 – 7 15
Additionally, the company has specified that the relative 2 5 – 3 9 4
importance of the reader characteristics should be weighted as 3 3 – 9 10 – 8
follows:
Assume that the gambler has a total of Rs 500 with which he
Reader Characteristics Weightage may play only once. The exact outcome of the game is not known
in advance and in the face of this uncertainty the gambler decides
Age: 25–35 yrs 0.4 to make the allocation that would maximize the minimum return.
Graduate and above 0.4 Formulate this problem as an LP model.
Income  Rs 5,000 0.2 26. An investor has money-making activities A1, A2, A3 and A4. He has
At this point in time the company has Rs 5,00,000 to spend. only one lakh rupees to invest. In order to avoid excessive
Formulate this problem as an LP model to maximize the effective investment, no more than 50 per cent of the total investment
exposure level. can be placed in activity A2 and/or activity A3. Activity A1 is very
conservative, while activity A4 is speculative. To avoid excessive
23. The owner of Metro Sports wishes to determine the number of speculation, at least Re 1 must be invested in activityA1 for every
advertisements to be placed in the selected three monthly Rs 3 invested in activity A4. The data on the return on investment
magazines A, B and C. His objective is to advertise in sucha is as follows:
way that total exposure to principal buyers of expensive sports
goods is maximized. The percentage of readers for each Activity Anticipated Return on Investment (%)
magazine is known. Exposure to any particular magazineis the
number of advertisements placed multiplied by the number of A1 10
principal buyers. The following data may be used: A2 12
A3 14
Magazines A4 16
A B C
The investor wishes to know how much to invest in each in order
Readers 1 lakh 0.6 lakh 0.4 lakh activity to maximize the total return on the investment. Formulate
Principal buyers 10 % 15 % 7% this problem as an LP model.
Cost per advertisement (Rs) 5,000 4,500 4,250 27. The board of directors of a company has given approval for the
construction of a new plant. The plant will require an investment of
The budget amount is at the most Rs 2,00,000 for the
Rs 50 lakh. The required funds will come from the sale of a
advertisements. The owner has already decided that magazine
proposed bond issue and by taking loans from two financial
A should have no more than six advertisements and that B and
corporations. For the company, it will not be possible to sell more
C each should have at least two advertisements. Formulate this
than Rs 20 lakh worth of bonds at the proposed rate of 12%.
problem as an LP model to determine the number of
Financial corporation A will give loan upto Rs 30 lakh at an interest
advertisements that should be placed in each magazine.
rate of 16% but insists that the amount of bond debt plus the
24. The XYZ company is preparing a proposal for an advertising amount owned to financial corporation B be no more than twice the
campaign for a client who is a publisher of law books. An optimal amount owed to financial corporation A. Financial corporation B
allocation of advertising funds to maximize the total number of will loan the same amount as that loaned by financial
exposures has to be made for the client. The relevant corporation A but itwould do so at an interest rate of 18%.
characteristics of the three alternative publications are shownin Formulate this problem as an LP model to determine the amount of
the following table: funds to be obtained from each source in a manner that minimizes
the total annual interest charges.
Home Home and Care
Beautiful Garden (Rs) 28. An investor wishes to diversify his portfolio and make due
(Rs) (Rs) allowance for long-term potentialities, but at the same time wishes
to maximize his current dividend income. He has considered
Cost/advertisement 600 800 450 various securities in which he might invest, and has classified
Max. number of ads 12 24 12 them into four types:
Min. number of ads 3 6 2 Type A : Relatively high element of risk, with commensurately high
Characteristics dividend and considerable growth potential.
Homeowner 80% 70% 20% Type B : Speculative stock with considerable risk, high dividends,
Income: Rs 10,000 or more 70% 80% 60% but less growth potential than type A.
Occupation: gardener 15% 20% 40%
Audience size 6,00,000 8,00,000 3,00,000
60 Operations Research: Theory and Applications

Type C : Stock with little risk, considerable growth potential,but 32. A transport company is considering the purchase of new vehicles
relatively low dividend income at present. for transportation between Delhi airport and hotels inthe city.
Type D : Stock with little risk, not much growth potential, and fairly There are three vehicles under consideration – station wagons,
high dividends. mini buses and large buses. The purchase price would be Rs
Because of the element of risk, the investor wishes to restrict 2,45,000 for each station wagon, Rs 3,50,000 for a minibus and
purchases of types A and B to not more than 30% of his investment. Rs 5,00,000 for a large bus. The board of directors has
To enhance prospects for long-term growth of his invest- ments, authorized a maximum amount of Rs 50,00,000 for these
he wishes to have at least 40% of his total outlay intypes A and purchases. Because of the heavy air travel involved, the new
C. Within these restrictions, he wishes to maximize his current vehicles would be utilized at maximum capacity, regardless of
dividend income. Total investment is Rs 1,00,000. Dividend returns the type of vehicles purchased. The expected net annual profit
on the four types of investments are A: 6%,B: 7%, C: 3%, D: would be Rs 15,000 for the station wagon, Rs 35,000 for the
5%. Formulate this problem as an LP modelto suggest the total mini bus, and Rs 45,000 for the large bus. The company has
investment to be allocated. hired 30 new drivers for the new vehicles. They are qualified drivers
29. The Agro Promotion Bank is trying to select an investment portfolio for all the three types of vehicles. The maintenance department has
for a cotton farmer. The bank has chosen a set of five investment the capacity to handle an additional 80 station wagons. A mini bus
alternatives, with subjective estimates of rates of return and risk is equivalent to 5/3 station wagons and each large bus is
as follows: equivalent to two station wagons in terms of their use of the
maintenance department. Formulate this problem as an LP model
Investment Annual Rate of Return Risk to determine optimal number of each type of vehicle to be
purchased in order to maximize profit.
Tax-free municipal bonds 6.0 1.3
Corporate bonds 8.0 1.5 [Delhi Univ., MBA, Oct. 2000]
High grade common stock 5.0 1.9 33. The managers of several cattle feed lots are interested in
Mutual fund 7.0 1.7 determining how many of each of several types of livestock feeds
Real estate 15.0 2.7 should be purchased in order to satisfy the nutritional requirements
for their livestock. They wish to purchase such food in a manner
The bank officer incharge of the portfolio would like to maximize the that minimizes the cost of feeding their livestock. Relevant costs
average annual rate of return on the portfolio. However, the wealthy and nutritional data are as below:
investor has specified that the average risk of the portfolio should
not exceed 2.0. The investor and does not want more than 20% Required Units of Nutritional Element Minimum
of the investment to be put into real estate. Formulate this problem Nutrient Nutrient
as an LP model. Alfa Corn Soyabean Sorghum Requirements
30. Raj, a retired government officer, has recently received his
retirement benefits, viz., provident fund, gratuity, etc. He is A 40 50 30 60 500
contemplating how much money he should invest in various B 30 60 35 40 750
alternatives open to him so as to maximize return on his C 25 30 25 50 600
investment. The investment alternatives are: government Cost per
securities, fixed deposits of a public limited company, equity unit (Rs) 1.00 1.25 0.95 1.35
shares, time deposits in a bank, and house construction. He has
made a subjective estimate of the risk involved on a five-point Formulate this problem as an LP model.
scale. The data on the return on investment, the numberof years
34. Old hens can be bought at Rs 100 each and young ones at
for which the funds will be blocked to earn this return on
Rs 250 each. The old hens lay 3 eggs per week and the young
investment and the subjective risk involved are as follows:
ones 5 eggs per week, each egg being worth 50 paise. A hen costs
Rs 20 per week to be fed. There are only Rs 8,000 available to
Return (%) Number Risk
be spent on purchasing the hens and at the most
of Years
20 hens can be accommodated in the space. Formulate this
Government securities 6 15 1 problem as an LP model to determine each kind of hen that
Company deposits 13 3 3 should be bought in order to yield the maximum profit per week.
Time deposits 10 5 2 35. A pension fund manager is considering investing in two shares
Equity share 20 6 5 A and B. It is estimated that:
House construction 25 10 1
(i) Share A will earn a dividend of 12 per cent per annum
He is wondering as to what percentage of funds he should invest and share B, 4 per cent per annum.
in each alternative so as to maximize the return on investment. He (ii) Growth in the market value in one year of share A will
has decided that the risk should not be morethan 4, and funds be 10 paise per Re l invested and in B, 40 paise perRe
should not be locked up for more than 15years. He would 1 invested.
necessarily invest at least 25% in house construction. Formulate He requires to invest the maximum total sum which will give:
this problem as an LP model. (i) dividend income of at least Rs 600 per annum; and
31. A dealer of used scooters wishes to stock up his lot to maximize his (ii) growth in one year of at least Rs 1,000 on the initial
profit. He can select scooters A, B and C which are valued on investment.
wholesale at Rs 5,000, Rs 7,000 and Rs 8,500 respectively. These Formulate this problem as an LP model to compute the minimum
can be sold at Rs 6,000, Rs 8,500 and Rs 10,500, respectively. For sum in order to be invested to meet the manager’s objective.
each type of scooter, the probabilities of sale are: 36. A scrap metal dealer has received an order from a customer
Type of scooter : A B C for at least 2,000 kg of scrap metal. The customer requires that
Prob. of sale in 90 days : 0.7 0.8 0.6 at least 1,000 kg of the shipment of the metal be high quality
copper that can be melted down and further used to produce
For every two scooters of B-type he should buy one scooterof copper tubings. Furthermore, the customer will not accept delivery
type A or type C. If he has Rs 1,00,000 to invest, what should of the order if it contains more than 175 kg ofmetal that he
he buy in order to maximize his expected gain. Formulate this deems unfit for commercial use, i.e. metal that contains an
problem as an LP model. excessive amount of impurity and cannot be melted down and
defined profitably.
 The dealer can purchase scrap metal from two different year. The company wants to minimize the total cost of using
suppliers in unlimited quantities with the following percentage(in 50 machines over a two-year period. The objective is to minimize
terms of weight) of high quality copper and unfit scrap: the costs of purchasing machines or leasing machines during the
years 1 and 2, and to minimize the interest paymentson funds
Supplier A Supplier B borrowed to obtain the machines. Formulate this problem as a
Copper 25% 75% linear programming problem.
Unfit scrap 5% 10% [Delhi Univ., MBA, 2009]
38. A trucking company with Rs 40,00,000 to spend on new equipment
The cost per kg of metal purchased from supplier A and supplier is contemplating three types of vehicles. Vehicle Ahas a 10
B is Re 1 and Rs 4, respectively. Formulate this problem as tonne payload and is expected to average 35 km per hour. It costs
an LP model so as to determine the optimal quantities of metal that Rs 80,000. Vehicle B has a 20-tonne payload and is expected
the dealer should purchase from each of the two suppliersin to average 30 km per hour. It costs Rs 1,30,000. Vehicle C is a
order to minimize total the purchase cost. modified form of vehicle B; it carries sleeping quarters for one
[Delhi Univ., MBA, 2008] driver and then reduces its capacityto 18 tonnes and raises the
37. A company needs 50 new machines. The machines have an cost to Rs 1,50,000. Vehicle A requires a crew of one average man,
economic life of two years and can be purchased for Rs 4,500 and if driven on three shifts per day, could be run for an average
or be leased for Rs 2,800 per year. The purchased machines, of 18 hours perday. Vehicles B and C require a crew of two
at the end of two years, have no salvage value. Company has men each, whileB would be driven 18 hours per day with three
Rs 1,00,000 in uncommitted funds that can be used for the shifts, C however would average 21 hours per day. The company
purchase or the lease of machines at the beginning of year 1. has 150 drivers available each day and would find it very difficult
The company can obtain a loan of upto Rs 2,00,000 at 18 per to obtain further crews. Maintenance facilities are such that the
cent interest per year. According to the terms of loans, the company total number of vehicles must not exceed 30. How many
has to repay the amount borrowed plus the interestat the end vehiclesof each type should be purchased if the company wishes
of each year. Each machine can earn Rs 3,000 per year. The to maximize its capacity in tonne-kms per day? Formulate this
earnings from the first year can be used to lease costs and the problem as an LP model. [Delhi Univ., MBA, 2009]
repayment of debt at the start of the second

HINTS AND ANSWERS

subject to 2 x1  3 x2  5 x3  4,000 U
V
1. Let x1, x2 = number of units of products A and B to be produced,

respectively. 4x  2x  7x  6,000

Max Z = 40x1 + 30x2

1 2 3 W
(Raw material requirement)

subject to 3x1 + x2  3,000 (Man-hours) 1 1

x1  8,000 ; x2  1,200 (Marketing) x  x  x  2 ,500 (Production limitation)
and x , x  0. 1
2 2
3 3

x  500; x  500; x  375 (Market demand)

1 2

2. Let x1, x2 = number of productive runs of process 1 and 2, 1 2 3

1 1 1 1
respectively. x  x ; x  x (Ratios of production)
1 2 2
Max Z = 300x + 400x 3 2 2 5 3
1 2

U
5. Let x and x = parts of A and B per hour manufactured,
5 x1  4 x 2  200 (Max amount of crude A
WV
subject to 1 2
respectively.
3 x1  5 x2  150 and B)  20
14 17.50 

Max Z = (5x1  6x2 )      2.00  x1


5 x  4 x  100 UV (Market requirement of  25 28 35 

8 x  4 x  80 W gasoline X and Y)
1 2
 20 14 17.50 
1 2
    3.00 x
  2
and x1, x2  0.  24 35 25 

3. Let xi = number of units purchased per month = 1.20x1 + 1.40x2

(i = 1, 2, 3 – April, May, June) x1 x2 x x

subject to (i) x1  x2  1; (ii)   1; (iii) 1  2  1

yi = number of units sold per month x1  x2  x3  y1  y2  y 3

(i = 1, 2, 3 – May, June, July)
and xi , yi  0 for all i.
Max Z = (90y1 + 60y2 + 75y3) – (75x1 + 75x2 + 60x3)
subject to y1  x1  150 4. Let x1, x2 and x3 = number of units of models I, II and III,

y2  x1 + x2 – y1  150
y3  x1 + x2 + x3 – y1 – y2  150
 Linear Programming: Applications and Model Formulation 61

25 24 28 35 35 25 j ( j = 1, 2 – regular, overtime) and shipped in month

(Manufacturing capacity)
k (k = 1, 2, 3 – Oct., Nov., Dec.)
and x1 , x2  0.
Min Z = 3x111 + 5x121 + 4x112 + 6x122 + 5x113 + 7x123
6. Let xijk = number of units manufactured in month + 3x212 + 5x222 + 4x213 + 6x223 + 3x313 + 5x323
i (1, 2, 3 – Oct., Nov., Dec.) during shift
respectively to be manufactured U
|
subject to x111  x112  x113  1,500
1,500 V (Regular time)
Max Z = 60x1 + 40x2 + 100x3
|
x212  x 213 

1,500 W
x313

  U x1 + x2 + x3 + x4  25
|
x121  x122  x123 750 (Total requirement)
750 V
and x1, x2, x3, x4  0.
|
x222  x 223  (Overtime)
11. Let xij = gallons of paint manufactured at plant i (i = 1, 2) and
750 W
x323
 shipped to contractor j ( j = 1, 2, 3) Min (production

x111 + x121 = 1,000 cost + shipping cost)

x121 + x122 + x212 + x222 = 3,000 Z = (x11 + x12 + x13) {(0.10) (3.80) + (0.25) (3.20)}
x113 + x123 + x213 + x223 + x313 + x323 = 2,000 + (x21 + x22 + x23) {(0.15) (4.00) + (0.20) (3.10)}
and xijk  0 for all i, j, k. + 1.80x11 + 2.00x21 + 2.60x12 + 2.20x22 + 2.10x13

+ 2.25x23
7. Let x1, x2 = number of gallons of wine B and C in the blend,
respectively. = 2.98x11 + 3.78x12 + 3.28x13 + 3.22x21 + 3.42x22

Max Z = 20 + x1 + x2
( 20  27)  33x1  32 x2
+ 3.47x23
subject to x  x  750 U

V|
subject to 30  20  x  x  31 11 21
1 2 x12  x 22  1,500 (Order size)
(Resultant degrees proof of blend) x13  x 23  1,500 W|
( 20  0.32)  0.2 x1  0.3x2  0.25 (Acidity) ( x11  x12  x13 )( 0.10)  300 Plant 1 
 (Blending)
20  x1  x2 ( x  x  x )( 0.15)  600 Plant 2




11  x12  x13 )( 0.25)  360 Plant 1 
( x21 22 23
( 20  1.07)  1.08 x1  1.04 x2  1.06 (Specific gravity) (Tinting)

(x x  x )( 0.20)  720 Plant 2 



20  x1  x 2 21 22 23
and xij  0 for all i and j.
x1  34 (Quality)
12. Let x1 and x2 = number of vitamin units purchased of food F1
and x 1 , x2  0.
and F2, respectively.

8. Let x1, x2 and x3 = quantity of foods 1, 2 and 3 to be used,

Min Z = 4x1 + 5x2
respectively.
subject to (i) 3x1 + 6x2  80; (ii) 4x1 + 3x2  100
Min Z = 1.50x1 + 2.00x1 + 1.20x3
and x1, x2  0.
subject to x1  y1  12,
13. Let xij = number of jobs accepted during day and night
350x1  250x2  200x3  300

Max Z = 275 (x11 + x12) + 125 (x21 + x22) + 225 (x31 + x32)
250x1  300x2  150x3  200 subject to 1,200 (x + x ) + 1,400 (x + x )
100x1  150x2  75x3  100 11 12 21 22

75x  125x  150x  100 + 800 (x31 + x32)  13,400

1 2 3
and x1, x2 , x3  0. 100 (x11 + x12) + 60 (x21 + x22)

+ 80 (x31 + x32)  1,050

9. Let x1 and x2 = number of soccer balls of types X and Y,
1,200x12 + 1,400x22 + 800x32  9,200
respectively.
100x12 + 60x22 + 80x32  650
Min Z = (2 hrs) (Rs 5.50/hr) x1 + (4 hrs) (Rs 8.50/hr) x1
+ (3 hrs) (Rs 5.50/hr) x2 and xij  0 for all i and j.
+ (6 hrs) (Rs 8.50/hr) x2 = 45x1 + 67.50x2 14. For plain coffee X
subject to 2x1 + 3x2  80 (Semi-skilled hours) x11, x12 and x13 = quantity (in kg) of the three coffees,
4x1 + 6x2  150 (Skilled hours) respectively.
x1  15 (Ball X) For plain coffee XX
x2  10 (Ball Y) x21, x22 and x23 = quantity (in kg) of the three coffees,
and x1, x2  0. respectively.

10. L e
62 Operations Research: Theory and Applications

t xj = number of kg of ingredient j ( j = 1, 2, 3, 4) used in Max Z = 45 (30x11 + 40x12 + 35x 13)

the mixture

Min Z = 28x + 25x + 52x + 26x + 55 (30x21 + 40x22 + 35x23)

1 2
x2  18 U3 4 subject 6x11  8x12  5x13  6.5 

V
subject to x1  22;
x  20; x  24 
(Supplies) 4x11  3x12  3.5x13  3.8 (Plain coffee X)
3 4 W 2x  2.5x  1.5x  2.2


0.55x  0.45x  0.45x  0.45x  0 U 11 12 13 

1 2 3
| 4 x11  x12  x13  35,000 

0.40 x  0.60 x  0.60 x  0.60 x  0

1 2 3
| 4 6x  8x  5x  6.0 U
 0.10 x  0.90 x  0.90 x  0.10 x  0 V
1 2 3
|
4 21 22 23
||
 0.25 x  0.75x  0.75 x  0.25x  0 4 x21  3x22  3.5x23  3.5
V
(Plain coffee XX)
1 2 3 | 4 2x  2.5x  1.5 x  2.0

 0.50 x  0.50 x  0.50 x  0.50 x  0 |W

|
1 2 3 4 21 22 23

(Mixing requirements) x 21  x 22  x 23  25,000 W

 x11 + x21  40,000; x12 + x22  20,000 ; subject to (1/1,500) x1 + (1/4,500) x2  1,
x13 + x23  15,000 (1/1,000) x1 + (1/8,000) x2  1,
and xij  0 for all i and j. (1/2,000) x1 + (1/4,000) x2  1,
15. Let x , x , x and x = quantities of four products to be (1/3,000) x1 + (1/9,000) x2  1,

1 2 3 4 and x1, x2  0.
manufactured, respectively. 21. Let x1, x2 and x3 = number of messages carried in media 1, 2

Max Z = [175 – (15 + 30 + 35)] x1 + [95 – (8 + 18 + 28)] x2 and 3, respectively.

+ [145 – (12 + 24 + 25)] x3 + [130 – (12 + 21 + 21)] x4 Max Z = 80,000x1 + 60,000x2 + 45,000x3
= 95x1 + 41x2 + 84x3 + 76x4 subject to 1,000 x1  750 x2  500 x3  2,00,000

subject to 4x1  2x2  3x3  3x4  800, x1  12; x2  6, 4  x3  8,

10 x1  6 x2  8 x3  7 x4  1,200 and x1 , x2  0.
10x1  8x2  8x3  8x4  800 22. Let x1, x2 and x3 = number of advertisements in magazine 1, 2
and x1 , x2 , x3 , x4  0 and 3, repectively.
16. The data of the problem can be summarized as given below: Max Z = 3,20,000 (0.7 × 0.4 + 0.8 × 0.4 + 0.6 × 0.2) x1
+ 5,00,000 (0.8 × 0.4 + 0.6 × 0.4 + 0.7 × 0.2) x2
Quarry A Quarry B Quarry C Specifications
+ 2,00,000 (0.4 × 0.4 + 0.5 × 0.4 + 0.4 × 0.2) x3
Material X 20% 40% 10% = 30% = 2,30,400x1 + 35,000x2 + 88,000x3
Material Y 60% 30% 40%  40%
subject to 1,000x1  700x2  500x3  5,00,000
Material Z 20% 30% 50% between 30%
and 40% 20  x1  50; 10  x1  40; 5  x3  30
Cost/tonne (Rs) 10 12 15 and x j  0 for all j.
Let x1, x2, and x3 = number of tonnes procured from 23. Let x1, x2 and x3 = number of insertions in magazines A, B and
quarry A, B and C, respectively. C, respectively.
Min (total cost) Z = 10x1 + 12x2 + 15x3 Max (total exposure) Z = (10% of 1,00,000)x1
subject to 2x1 + 4x2 + x3 = 3 (Material X) + (15% of 60,000)x2 + (7% of 40,000)x3
6x1 + 3x2 + 4x3  4 (Material Y) subject to 5,000 x1  4,500 x2  4,250 x3  1,00,000

2x1 + 3x2 + 5x3  4 (Material Z) x1  6; x2  2; x3  2

2x1 + 3x2 + 5x3  3 and x1 , x2 , x3  0.

17. Let x1, x2 and x3 = number of units of types I, II and III model, 24. Let xj = number of advertisements in media j ( j = 1, 2, 3).
respectively. Media Effectiveness coefficient
Max (total profit) Z = 60x1 + 40x2 + 1,000x3 1 0.80 (0.4) + 0.70 (0.2) + 0.15 (0.4) = 0.52
subject to 2x1 + 3x2 + 5x3  4,000 (Raw material A) 2 0.70 (0.4) + 0.80 (0.2) + 0.20 (0.4) = 0.52
3 0.20 (0.4) + 0.60 (0.2) + 0.40 (0.4) = 0.36
4x1 + 2x2 + 7x3  6,000 (Raw material B)
Max Z = 0.52 (6,00,000)x1 + 0.52 (8,00,000)x2
x2 x3
x   + 0.36 (3,00,000)x
3  2,500 (Labour force)
1 3
2 subject to 600 x 1  800 x2  450 x3  2,00,000

x1 x2 x2 x x1  12; x2  24; x3  12
 ;  3 (Number of units produced)
3 2 2 5 x1  3; x2  6; x3  2

and x1, x2, x3  0.

x1  500; x2  500; x3  375 (Market demand)
25. y = minimum expected gain per rupee deposited in the given
and x1, x2, x3  0.
choice j ( j = 1, 2, 3, 4) by the gambler
18. Let x1, and x2 = number of units of models X and Y, respectively. xj = amount of bet money used among four different choices,
Max Z = 120x1 + 90x2 respectively ( j = 1, 2, 3, 4)
x1 x x1 x2 Max Z = y
 2 
subject to 100 240
1;
subject to  4 x  7 x  15x y
  1 ; x1  140 3 x
x1, x2  0.
and
 Linear Programming: Applications and Model Formulation 63

200 150 1 2
3
4
5 x1  3 x2
 9 x3 
4 x4  y
19. Let x1 and x2 = number of television and radio sets to be 3 x1  9 x2  10 x3  8 x4  y
manufactured, respectively. x1  x2  x3  x4  500
Max Z = 500x1 + 250x2 and xj may be negative, zero or positive.
64 Operations Research: Theory and Applications

26. Let x1, x2, x3 and x4 = percentage of the total fund that should subject to 5,000x1 + 2 × 7,000x2  1,00,000
be invested in activities A1, A2, A3 and 2 × 7,000x2 + 8,500x3  1,00,000
A4, respectively. and x1, x2, x3  0.
Max Z = 0.10x1 + 0.12x2 + 0.14x3 + 0.16x4 33. Let x1, x2, x3 = number of station wagons, minibuses and larges
subject to x1  x2  x3  x4  1,00,000 buses, respectively to be purchased.
(total money invested) Max Z (total profit) = 15,000x1 + 35,000x2 + 45,000x3
x2  x3  0.50 ( x1  x2  x3  x4 ) subject to x1 + x2 + x3  30 (Availability of drivers)
1

x1  x4 2,45,000 x1  3,50,000 x2  5,00,000 x3  50,00,000

3 (Budget)

and x j  0 for all j.

3 1
x1  x  x  80 (Maintenance capacity)
27. Let x1 = bond debt to be obtained. 5 2 2 3
x2 and x3 = loan to be obtained from financial corporations A and x1, x2, x3  0.
and B, respectively. Note: 1 S.W. = (3/5) M.B., because (5/3) S.W. = 1 M.B.;
Min Z = 0.12x1 + 0.16x2 + 0.18x3 and 1 S.W. = (3/5) M.B., because (5/3) S.W. = 1 M.B.
subject to x1  x2  x3  50 34. Let xj = number of units of food type j ( j = 1, 2, 3, 4) used.
x1  20; x2  30 Min (total cost) Z = 1.00x1 + 1.25x2 + 0.95x3 + 1.35x4
x1  x3  2 x2 ; x1  x2 subject to 40x1 + 50x2 + 30x3 + 60x4  500
30x1 + 60x2 + 35x3 + 40x4  750
and x1 , x2  0.
25x1 + 30x2 + 25x3 + 50x4  600
28. Let x1, x2, x3 and x4 = amount of money to be invested in A, and x1, x2, x3, x4  0.
B, C and D securities, respectively. 35. Let x1 and x2 = number of old hens and young hens bought,
Max (current dividend return)
respectively.
Z = 0.06 x1  0.07 x2  0.03x3  0.05x4 Max Z = 0.5 (3x + 5x ) – (x + x ) = 0.5x – 1.5x

1 2 1 2 2 1
subject to x1  x2  x3  x4  1,00,000 subject to 100x1 + 250x1  8,000; x1 + x2  20

x1  x2  30, 000 ; x1  x3  40,000 and x1, x2  0.

and x1 , x2 , x3 , x4  0. 36. Let x1 and x2 = number of units of share A and B respectively.

to Min (total investment) Z = x1 + x2

31. Let x1, x2, x3, x4 and x5 = percentage of the total fund that
should be invested in all given five subject to 0.12x1 + 0.04x2  600 (Dividend constraint)
schemes, respectively. 0.10x1 + 0.40x2  1,000 (Investment constraint)
Max Z  6 x1  13x2  10 x3  20 x4  25x5 and x1, x2  0.
subject to 15x1  3x2  5x3  6 x4  10 x5  15 37. Let x1 and x2 = volume of purchase from supplier A and B,
respectively.
x1  3x2  2x3  5x4  x5  4
x5  0.25 Min Z = x1 + 4x2

x1 + x2 + x3 + x4 + x5 = 1 1 3
subject to (i) x1 + x2  2,000; (ii) x1  x2  1,000
and x j  0 for all j. 4 4

1 1
(iii) x  x  175
32. Let x1, x2, x3 = number of scooters of types A, B and C, 20 1 10 2

respectively. and x1, x2  0.

Max Z = 0.7 (6,000 – 5,000)x1 + 0.8 (8,500 – 7,000)x2
+ 0.6 (10,500 – 8,500)x3

CHAPTER SUMMARY
This chapter presents basic assumptions, limitations, components of any linear programming model and broad application areas
of linear programming. The guidelines of mathematical modelling of any decision problem were explained followed by a large
number of model building solved exercises in all functional areas of management and allied areas. These exercises are illustrative
for students to deal with more complex and real-life problems.
 Linear Programming: Applications and Model Formulation 65

CHAPTER CONCEPTS QUIZ

True or False
1. In a Linear Programming model, all parameter are assumed to (c) use of the available resource
be known as constant. (d) all of the above
2. In LP model, any variable can assume to take only integer values 24. The distinguishing feature of an LP model is
or restricted to take discrete number of values. (a) relationship among all variables is linear
3. Total contribution is used in place of profit in the objective function (b) it has single objective function and constraints
of maximization. problem because whole profit is not linearly (c) value of decision variables is non-negative
related to sales volume. (d) all of the above
4. An equation is more restrictive than an inequality. 25. Constraints in an LP model represents
5. All the variables in the solution of a linear programming problem (a) limitations
are either positive or negative because of the existence of (b) requirements
structural constraints. (c) balancing limitations and requirements
(d) all of the above
6. Linear programming is a technique for finding the best uses of
an organizations manpower, money and machinery. 26. Non-negativity condition is an important component of LP model
because
7. Production planning is one of the application areas of the linear (a) variables value should remain under the control of the
programming.
decision-maker
8. The effect of time and uncertainty are taken into consideration (b) value of variables make sense and correspond to real-
by linear programming model. world problems
9. Linear Programming determines the economic and efficient way (c) variables are interrelated in terms of limited resources
of locating manufacturing plants for physical distribution. (d) none of the above
10. All variables in the linear programming problem must take one 27. Before formulating a formal LP model, it is better to
negative values. (a) express each constraint in words
(b) express the objective function in words
Fill in the Blanks (c) verbally identify decision variables
11. Linear programming is a technique which attempts to determine (d) all of the above
how best to allocate in order achieve some . 28. Each constraint in an LP model is expressed as an
12. A linear programming technique improves the quality of . (a) inequality with ³ sign
(b) inequality with £ sign
13. In a linear programming, all relationships among decision variables
(c) equation with = sign
are .
(d) none of the above
14. If two variables always take on values which are in the same
29. Maximization of objective function in an LP model means
proportion, the variables are related.
(a) value occurs at allowable set of decisions
15. appearing in the models are assumed to be constant (b) highest value is chosen among allowable decisions
but in real life situations. (c) neither of above

16. Every linear programming problem includes which (d) both (a) and (b)

relates variable in the problem to the goal of the firm and 30. Which of the following is not a characteristic of the LP model
which represent the limit on resource available to the (a) alternative courses of action
firm. (b) an objective function of maximization type
17. Most of the constraints in the linear programming problem are (c) limited amount of resources
expressed as . (d) non-negativity condition on the value of decision variables
18. Linear programming is used to allocate to activities 31. The best use of linear programming technique is to find an optimal
so as to optimize the value of objective function. use of
19. If the value of the variables are under the control of decision (a) money (b) manpower
makers then variables are said to be otherwise (c) machine (d) all of the above
. 32. Which of the following is not the characteristic of linear
20. of the decision variables is one of the assumption of programming
the linear programming model. (a) resources must be limited
(b) only one objective function
Multiple Choice (c) parameters value remains constant during the planning period
(d) the problem must be of minimization type
21. The mathematical model of an LP problem is important because
(a) it helps in converting the verbal description and numerical 33. Non-negativity condition in an LP model implies
data into mathematical expression (a) a positive coefficient of variables in objective function
(b) decision-makers prefer to work with formal models (b) a positive coefficient of variables in any constraint
(c) it captures the relevant relationship among decision factors (c) non-negative value of resources
(d) it enables the use of algebraic technique (d) none of the above
22. Linear programming is a 34. Which of the following is an assumption of an LP model
(a) constrained optimization technique (a) divisibility (b) proportionality
(b) technique for economic allocation of limited resources (c) additivity (d) all of the above
(c) mathematical technique 35. Which of the following is a limitation associated with an LP Model
(d) all of the above (a) the relationship among decision variables in linear
(b) no guarantee to get integer valued solutions
23. A constraint in an LP model restricts (c) no consideration of effect of time and uncertainty on LP model
(a) value of objective function (d) all of the above
(b) value of a decision variable
66 Operations Research: Theory and Applications

Answers to Quiz
1. T 2. F 3. T 4. T 5. F 6. T 7. T
8. F 9. T 10. T 11. resources, objective 12. decisions 13. linear
14. linearly 15. parameters, unknown 16. objective function, constraints 17. inequalities 18. scarce resources
19. controllable and uncontrollable 20. certainty 21. (a) 22. (d) 23. (d) 24. (a)
25. (d) 26. (b) 27. (d) 28. (d) 29. (a) 30. (b) 31. (d)
32. (d) 33. (d) 34. (d) 35. (d)

CASE STUDY

Case 2.1: Welltype Manufacturing

Welltype manufacturing company produces three types of typewriters. All the three models are required to be
machined first and then assembled. The time required for the various models are as follows:
Types Manual Electronic Deluxe
Typewriters Typewriters Electronic
Typewriters
Machine time (in hours) 15 12 14
Assembly time (in hours) 4 3 5

The total available machine time and assembly time are 3,000 hours and 1,200 hours, respectively. The data regarding
the selling price and variable costs for the three types are:

Types Manual Electronic Deluxe

Typewriters Typewriters Electronic
Typewriters
Selling price (Rs) 4,100 7,500 14,600
Labour, material and
other variable costs
(Rs) 2,500 4,500 9,000
The company sells all the three types on credit, but can only collect the amounts on the first of the following month.
The labour, material and other variable expenses will have to be paid in cash. This company has taken a loan of Rs 40,000
from a co-operative bank and will have to repay it to the bank on 1 April 2008. The TNC Bank from whom this company
has borrowed Rs 60,000 has expressed its approval to renew the loan.
Balance Sheet of the Company as on 31 March 2008

Liabilities Rs Assets Rs
Equity Share Capital 1,50,000 Land 90,000
Capital Reserve 15,000 Building 70,000
General Reserve 1,10,000 Plant &
Profit & Loss A/c 25,000 Machinery 1,00,000
Long-term Loan 1,00,000 Furniture &
Loan from TNC Fixtures 15,000
Bank 60,000 Vehicles 30,000
Loan from Inventory 5,000
Co-operative Bank 40,000 Receivables 50,000
Cash 1,40,000

Total 5,00,000 Total 5,00,000

The company will have to pay a sum of Rs 10,000 towards salary of top management executives and other
fixed overheads for the month. Interest on long-term loans is to be paid every month at 24% per annum. Interest
on loans from TNC and cooperative banks may be taken to be 1,200 for the month. Also this company has promised
to deliver 2 manual typewriters and 8 deluxe electronic typewriters to one of its valued customers next month. Keep
 Linear Programming: Applications and Model Formulation 67

in mind the fact that the level of operations in this company is subject to the availability of cash next month. This company
will also be able to sell all types of typewriters in the market. The senior manager of this company desires to know
as to how many units of each typewriter must be manufactured in the factory next month so as maximize the profits
of the company. Advise the management of the company for manufacturing strategy with an aim to maximize
profit.

Case 2.2: Shreya Advertizers

Shreya Advertizers – a marketing company dealing with laminated sheets ‘Gloss’ in the western zone covering
Maharashtra, Gujarat and Madhya Pradesh, is considering to launch an advertisement campaign within a budget of
Rs 2.5 lakh. On the basis of advertisement testing of the previous year, the company’s research department has
found that magazines and films are the ideal media for advertising laminated sheets. The company is not in a position
to use the audiovisual medium due to limitation of funds. The magazines enjoying good recall in last year’s campaign
are Stardust, Filmfare, Reader’s Digest and Madhuri. This is attributed to the effective visual impact made by the
good reproduction of the advertisements both in colour, and black and white.
The characteristics of target audience for ‘Gloss’ and weightage for each characteristic are as follows:

Characteristics Weightage (%)

Age 15–34 yrs 20
Monthly income Over Rs 5,000 70
Education Above S.S.C. 10

The audience characteristics for the four magazines selected are given below:

Characteristics Stardust Filmfare Reader’s Madhuri

(%) (%) Digest (%) (%)
Age: 15–34yrs 75 45 56 80
Monthly income: Over Rs 5,000 52 43 47 25
Education: Above S.S.C. 83 53 72 34

The efficacy index for a black and white advertisement may be taken as 0.15 and that for a colour advertisement
as 0.20. The cost per insertion of a black and white, and a colour advertisement and the readership for the four magazines
are as follows:

Magazines Cost (Rs) per Insertion Readership

(in ’000 Rs)
Black and White Colour
Stardust (Monthly) 4,500 8,400 189
Filmfare (Fortnightly) 4,200 8,400 256
Reader’s Digest (Monthly) 6,400 9,600 136
Madhuri (Fortnightly) 3,300 6,600 205

It has also been found that for creating an impact at least 03 insertions are necessary in Stardust and Reader’s Digest,
while a minimum of 04 insertions will be required in the case of Filmfare.
Suggest an advertising strategy for the company to maximize the expected effective exposure.
 C h a p t e r 3
“People realize that technology certainly is a tool. This tool can be used to enhance operations,
improve efficiencies and really add value to academic research and teaching exercises.”
– Peter Murray

This chapter presents graphical solution method for solving any LP problem with only two decision
variables. This method provides a conceptual basis for solving large and complex LP problems.

LEARNING OBJECTIVES

After studying this chapter, you should be able to

 solve an LP problem by the graphical method.
 understand various important terms such as extreme points, infeasibility, redundancy and multiple
solutions and demonstrate them with the help of the graphical method.
 interpret the solution of an LP model.

CHAPTER OUTLINE
3.1 Introduction  Self Practice Problems
3.2 Important Definitions  Hints and Answers
3.3 Graphical Solution Methods of LP Problem  Chapter Summary
3.4 Special Cases in Linear Programming  Chapter Concepts Quiz
 Conceptual Questions  Case Study
 Linear Programming: The Graphical Method 69

3.1 INTRODUCTION
An optimal as well as a feasible solution to an LP problem is obtained by choosing one set of values from
several possible values of decision variables x1, x2, . . ., xn, that satisfies the given constraints simultaneously
and also provides an optimal (maximum or minimum) value of the given objective function.
For LP problems that have only two variables, it is possible that the entire set of feasible solutions
can be displayed graphically by plotting linear constraints on a graph paper in order to locate the best
(optimal) solution. The technique used to identify the optimal solution is called the graphical solution method
(approach or technique) for an LP problem with two variables.
Since most real-world problems have more than two decision variables, such problems cannot be solved
graphically. However, graphical approach provides understanding of solving an LP problem algebraically,
involving more than two variables. Optimal solution
to an LP problem is
In this chapter, we shall discuss the following two graphical solution methods (or approaches):
obtained by
(i) Extreme point solution method (i) Extreme (corner)
point, and
(ii) Iso-profit (cost) function line method to
(ii) Iso-profit (cost)
find the optimal solution to an LP problem. function line
method
3.2 IMPORTANT DEFINITIONS
Solution The set of values of decision variables xj ( j = 1, 2, . . ., n) that satisfy the constraints of an
LP problem is said to constitute the solution to that LP problem.
Feasible solution The set of values of decision variables xj ( j = 1, 2, . . ., n) that satisfy all the constraints
and non-negativity conditions of an LP problem simultaneously is said to constitute the feasible solution to
that LP problem.
Infeasible solution The set of values of decision variables xj ( j = 1, 2, . . ., n) that do not satisfy all the
constraints and non-negativity conditions of an LP problem simultaneously is said to constitute the infeasible
solution to that LP problem.
Basic solution For a set of m simultaneous equations in n variables (n > m) in an LP problem, a solution
obtained by setting (n – m) variables equal to zero and solving for remaining m equations in m variables
is called a basic solution of that LP problem.
The (n – m) variables whose value did not appear in basic solution are called non-basic variables and
the remaining m variables are called basic variables.
Basic feasible solution A feasible solution to an LP problem which is also the basic solution is called the
basic feasible solution. That is, all basic variables assume non-negative values. Basic feasible solutionis of
two types:
(a) Degenerate A basic feasible solution is called degenerate if the value of at least one basic
variable is zero.
(b) Non-degenerate A basic feasible solution is called non-degenerate if value of all m basic
variables is non-zero and positive.
Optimum basic feasible solution A basic feasible solution that optimizes (maximizes or minimizes) the
objective function value of the given LP problem is called an optimum basic feasible solution.
Unbounded solution A solution that can increase or decrease infinitely the value of the objective function
of the LP problem is called an unbounded solution.

3.3 GRAPHICAL SOLUTION METHODS OF LP PROBLEM

While obtaining the optimal solution to the LP problem by the graphical method, the statement of the
following theorems of linear programming is used
 The collection of all feasible solutions to an LP problem constitutes a convex set whose extreme
points correspond to the basic feasible solutions.
70 Operations Research: Theory and Applications

 There are a finite number of basic feasible solutions within the feasible solution space.
 If the convex set of the feasible solutions of the system of simultaneous equations: Ax = b,
x  0, is a convex polyhedron, then at least one of the extreme points gives an optimal solution.
 If the optimal solution occurs at more than one extreme point, the value of the objective function
will be the same for all convex combinations of these extreme points.

Remarks 1. A convex set is a polygon and by ‘convex’ we mean that if any two points of a polygon
are selected arbitrarily, a straight line segment joining these two points lies completely within the
polygon.
2. Each corner (extreme or vertex) point of the feasible region (space or area) falls at the intersection of
two constraint equalities.
3. The extreme points of the convex set provide the basic feasible solution to the LP problem.

3.3.1 Extreme Point Solution Method

Extreme point In this method, the coordinates of all corner (or extreme) points of the feasible region (space or area) are
refers to the corner
of the feasible
determined and then value of the objective function at each of these points is computed and compared.
region (or space), The coordinates of an extreme point where the optimal (maximum or minimum) value of the objective
i.e. this point lies at function is found represent solution of the given LP problem. The steps of the method are summarized as
the intersection of follows:
two constraint
equations.
Step 1 : Develop an LP model State the given problem in the mathematical LP model as illustrated in
the previous chapter.
Step 2 : Plot constraints on graph paper and decide the feasible region
(a) Replace the inequality sign in each constraint by an equality sign.
(b) Draw these straight lines on the graph paper and decide each time the area of feasible solutions according
to the inequality sign of the constraint. Shade the common portion of the graph that satisfies all the
constraints simultaneously drawn so far.
(c) The final shaded area is called the feasible region (or solution space) of the given LP problem. Any point
inside this region is called feasible solution and this provides values of x1 and x2 that satisfy all the
constraints.
Step 3 : Examine extreme points of the feasible solution space
to find an optimal solution
(a) Determine the coordinates of each extreme point of the feasible solution space.
(b) Compute and compare the value of the objective function at each extreme point.
(c) Identify the extreme point that gives optimal (max. or min.) value of the objective function.
Extreme point
method is one of
the methods of 3.3.2 Examples on Maximization LP Problem
finding the optimal
solution to an LP Example 3.1 Use the graphical method to solve the following LP problem.
problem by Maximize Z = 15x1 + 10x2
examining the profit
(or cost) level at subject to the constraints
each corner point of (i) 4x1 + 6x2  360, (ii) 3x1 + 0x2  180, (iii) 0x1 + 5x2  200
feasible region.
and x1, x2  0.

Solution 1. The given LP problem is already in mathematical form.

2. Treat x1 as the horizontal axis and x2 as the vertical axis. Plot each constraint on the graph by treating
it as a linear equation and it is then that the appropriate inequality conditions will be used to mark
the area of feasible solutions.
Consider the first constraint 4x1 + 6x2  360. Treat this as the equation 4x1 + 6x2 = 360. For this
find any two points that satisfy the equation and then draw a straight line through them. The two
points are generally the points at which the line intersects the x1 and x2 axes. For example, when
x1 = 0 we get 6x2 = 360 or x2 = 60. Similarly when x2 = 0, 4x1 = 360, x1 = 90.
 Linear Programming: The Graphical Method 71

These two points are then connected by a straight line as shown in Fig. 3.1(a). But the question is: Where are
these points satisfying 4x1 + 6x2  360. Any point above the constraint line violates the inequality condition. But
any point below the line does not violate the constraint. Thus, the inequality and non-negativity condition can only
be satisfied by the shaded area (feasible region) as shown in Fig. 3.1(a).

Fig. 3.1
Graphical Solution
of LP Problem

(a) (b)
Similarly, the constraints 3x1  180 and 5x2  200 are also plotted on the graph and are indicated by
the shaded area as shown in Fig. 3.1(b).
Since all constraints have been graphed, the area which is bounded by all the constraints linesincluding
all the boundary points is called the feasible region (or solution space). The feasible region is shown in Fig.
3.1(b) by the shaded area OABCD.
3. (i) Since the optimal value of the objective function occurs at one of the extreme points of the feasible
region, it is necessary to determine their coordinates. The coordinates of extreme points of the feasible
region are: O = (0, 0), A = (60, 0), B = (60, 20), C = (30, 40), D = (0, 40).
(ii) Evaluate objective function value at each extreme point of the feasible region as shown in the Table 3.1:
Extreme Point Coordinates Objective Function Value
(x1, x2) Z = 15x1 + 10x2
O (0, 0) 15(0) + 10(0) = 0
A (60, 0) 15(60) + 10(0) = 900
B (60, 20) 15(60) + 10(20) = 1,100 Table 3.1
C (30, 40) 15(30) + 10(40) = 850 Set of Feasible
D (0, 40) 15(0) + 10(40) = 400 Solutions

(iii) Since objective function Z is to be maximized, from Table 3.1 we conclude that maximum value of Z
= 1,100 is achieved at the point extreme B (60, 20). Hence the optimal solution to the given LP problem
is: x1 = 60, x2 = 20 and Max Z = 1,100.
Remark To determine which side of a constraint equation is in the feasible region, examine whether the Feasible region is
origin (0, 0) satisfies the constraints. If it does, then all points on and below the constraint equation towards the overlapping area
the origin are feasible points. If it does not, then all points on and above the constraint equation away of constraints that
satisfies all of the
from the origin are feasible points. constraints on
resources.
Example 3.2 Use the graphical method to solve the following LP problem.
Maximize Z = 2 x1 + x2
subject to the constraints
(i) x1 + 2x2  10, (ii) x1 + x2  6,
(iii) x1 – x2  2, (iv) x1 – 2x2  1
and x1, x2  0.
Solution Plot on a graph each constraint by first treating it as a linear equation. Then use inequality condition
of each constraint to mark the feasible region by shaded area as shown in Fig. 3.2. It may be noted
72 Operations Research: Theory and Applications

that we have not considered the area below the lines x1 – x2 = 2 and x1 – 2x2 = 1 for the negative values
of x2. This is because of the non-negativity condition, x2  0.

Fig. 3.2
Graphical Solution
of LP Problem

The coordinates of extreme points of the feasible region are: O = (0, 0), A = (1, 0), B = (3, 1), C = (4, 2),
D = (2, 4), and E = (0, 5). The value of objective function at each of these extreme points is shown in Table 3.2.

Extreme Point Coordinates Objective Function Value

(x1, x2) Z = 2x1 + x2

O (0, 0) 2(0) + 1(0) = 0

A (1, 0) 2(1) + 1(0) = 2
B (3, 1) 2(3) + 1(1) = 7
Table 3.2 C (4, 2) 2(4) + 1(2) = 10
Set of Feasible D (2, 4) 2(2) + 1(4) = 8
Solutions E (0, 5) 2(0) + 1(5) = 5

The maximum value of the objective function Z = 10 occurs at the extreme point (4, 2). Hence, the
optimal solution to the given LP problem is: x1 = 4, x2 = 2 and Max Z = 10.
Example 3.3 Solve the following LP problem graphically:
Maximize Z = – x1 + 2x2
subject to the constraints
(i) x1 – x2  –1; (ii) – 0.5x1 + x2  2
and x1, x2  0. [Punjab Univ., BE (CS and E), 2006]
Solution Since resource value (RHS) of the first constraint is negative, multiplying both sides of this
constraint by –1, the constraint becomes: – x1 + x2  1. Plot on a graph each constraint by first treating them
as a linear equation and mark the feasible region as shown in Fig. 3.3.

Fig. 3.3
Graphical Solution
of LP Problem

The value of the objective function at each of the extreme points A(0, 1), B(0, 2) and C(2, 3) is shown in
Table 3.3.
 Linear Programming: The Graphical Method 73

Extreme Coordinates Objective Function Value

Point (x1, x2) Z = – x1 + 2x2

A (0, 1) 0+2×1 =2 Table 3.3

B (0, 2) 0 + 2 × 2 = 4Multiple optimal Set of Feasible

C (2, 3) – 1 × 2 + 2 × 3 = 4 so Solutions
 lution
The maximum value of objective function Z = 4 occurs at extreme points B and C. This implies that every
point between B and C on the line BC also gives the same value of Z. Hence, problem has multiple optimal
solultions: x1 = 0, x2 = 2 and x1 = 2, x2 = 3 and Max Z = 4.
Example 3.4 The ABC Company has been a producer of picture tubes for television sets and certain printed
circuits for radios. The company has just expanded into full scale production and marketing of AM and AM-FM
radios. It has built a new plant that can operate 48 hours per week. Production of an AM radio in the new plant will
require 2 hours and production of an AM-FM radio will require 3 hours. Each AM radio will contribute
Rs 40 to profits while an AM-FM radio will contribute Rs 80 to profits. The marketing department, after extensive
research has determined that a maximum of 15AM radios and 10 AM-FM radios can be sold each week.
(a) Formulate a linear programming model to determine the optimum production mix of AM and FM radios that
will maximize profits.
(b) Solve this problem using the graphical method. [Delhi Univ., MBA, 2002, 2008]
Solution Let us define the following decision variables
x1 and x2 = number of units of AM radio and AM-FM radio to be produced, respectively.
Then LP model of the given problem is:
Maximize (total profit) Z = 40x1 + 80x2
subject to the constraints
(i) Plant : 2x1 + 3x2  48, (ii) AM radio : x1  15, (iii) AM-FM radio : x2  10
and x1, x2  0.
Plot on a graph each constraint by first treating it as a linear equation. Then use inequality condition of each
constraint to mark the feasible region. The feasible solution space (or region) is shown in Fig. 3.4 by shaded
area.

Fig. 3.4
Graphical Solution
of LP Problems

The coordinates of extreme points of the feasible region are: O (0, 0), A (0, 10), B (9, 10), C (15, 6) and
D (15, 0). The value of objective function at each of corner (or extreme) points is shown in Table 3.4.
74 Operations Research: Theory and Applications

Extreme Point Coordinates Objective Function Value

(x1, x2) Z = 40x1 + 80x2
O (0, 0) 40(0) + 80(0) =0
A (0, 10) 40(0) + 80(10) = 800
Table 3.4 B (9, 10) 40(9) + 80(10) = 1,160
Set of Feasible C (15, 6) 40(15) + 80(6) = 1,080
Solutions D (15, 0) 40(15) + 80(0) = 600

Since the maximum value of the objective function Z = 1,160 occurs at the extreme point (9, 10), the optimum
solution to the given LP problem is: x1 = 9, x2 = 10 and Max. Z = Rs 1,160.
Example 3.5 Anita Electric Company produces two products P1 and P2. Products are produced and sold ona
weekly basis. The weekly production cannot exceed 25 for product P1 and 35 for product P2 because of limited
available facilities. The company employs total of 60 workers. Product P1 requires 2 man-weeks of labour, while
P2 requires one man-week of labour. Profit margin on P1 is Rs. 60 and on P2 is Rs. 40. Formulate this problem as
an LP problem and solve that using graphical method.
Solution Let us define the following decision variables:
x1 and x2 = number of units of product P1 and P2, to be produced, respectively.
Then LP model of the given problem is:
Maximize Z = 60x1 + 40x2
subject to the constraints
(i) Weekly productions for P1 : x1  25, (ii) Weekly production for P2 : x2  35,
(iii) Workers: 2x1 + x2 = 60
and x1, x2  0.

Fig. 3.5
Graphical Solution
of LP Problem

Plot on a graph each constraint by first treating it as linear equation. Then use inequality condition of each
constraint to mark the feasible region shown in Fig. 3.6.
The coordinates of extreme points of the feasible solution space (or region) are: A(0, 35), B(12.5, 35), C(25, 10)
and D(25, 0). The value of the objective function at each of these extreme points is shown in Table 3.5.

Extreme Point Coordinates Objective Function Value

(x1, x2) Z = 60x1 + 40x2
A (0, 35) 60(0) + 40(35) = 1,400
Table 3.5 B (12.5, 35) 60(12.5) + 40(35) = 2,150
Set of Feasible C (25, 10) 60(25) + 40(10) = 1,900
Solutions D (25, 0) 60(25) + 40(0) = 1,500
 Linear Programming: The Graphical Method 75

The optimal (maximum) value, Z = 2,150 is obtained at the point B (12.5, 35). Hence, x1 = 12.5 units of product
P1 and x1 = 35 units of product P2 should be produced in order to have the maximum profit, Z = Rs. 2,150.
3.3.3 Examples on Minimization LP Problem
Example 3.6 Use the graphical method to solve the following LP problem.
Minimize Z = 3x1 + 2x2
subject to the constraints
(i) 5x1 + x2  10, (ii) x1 + x2  6, (iii) x1 + 4x2  12
and x1, x2  0.
Solution Plot on a graph each constraint by first treating it as a linear equation. Then use inequality condition
of each constraint to mark the feasible region by shaded area as shown in Fig. 3.6. This regionis bounded
from below by extreme points A, B, C and D.

Fig. 3.6
Graphical Solution
of LP Problem

The coordinates of the extreme points of the feasible region (bounded from below) are: A = (12, 0), B = (4, 2),
C = (1, 5) and D = (0, 10). The value of objective function at each of these extreme points is shown in Table 3.6.
Extreme Point Coordinates Objective Function Value
(x1, x2) Z = 3x1 + 2x2
A (12, 0) 3(12) + 2(0) = 36
B (4, 2) 3(4) + 2(2) = 16 Table 3.6
C (1, 5) 3(1) + 2(5) = 13 Set of Feasible
D (0, 10) 3(0) + 2(10) = 20 Solutions

The minimum (optimal) value of the objective function Z = 13 occurs at the extreme point C (1, 5). Hence,
the optimal solution to the given LP problem is: x1 = 1, x2 = 5, and Min Z = 13.
Example 3.7 Use the graphical method to solve the following LP problem.
Minimize Z = – x1 + 2x2
subject to the constraints
(i) – x1 + 3x2  10, (ii) x1 + x2  6, (iii) x1 – x2  2
and x1, x2  0.

Fig. 3.7
Graphical Solution
of LP Problem
76 Operations Research: Theory and Applications

Solution Plot on a graph each constraint by first treating it as a linear equation. Then use inequality condition
of each constraint to mark the feasible region by shaded area as shown in Fig. 3.7. It may be noted that the
area below the lines – x1 + 3x2 = 10 and x1 – x2 = 2 is not desirable, due to the reason that values of x1 and
x2 are desired to be non-negative, i.e. x1  0, x2  0.
The coordinates of the extreme points of the feasible region are: O = (0, 0), A = (2, 0), B = (4, 2), C = (2, 4),
and D = (0, 10/3). The value of the objective function at each of these extreme points is shown in Table 3.7.

Extreme Point Coordinates Objective Function Value

(x1, x2) Z = – x1 + 2x2
O (0, 0) – 1(0) + 2(0) = 0
A (2, 0) – 1(2) + 2(0) = –2
B (4, 2) – 1(4) + 2(2) = 0
Table 3.7 Set of C (2, 4) – 1(2) + 2(4) = 6
Feasible Solutions D (0, 10/3) – 1(0) + 2(10/3) = 20/3

The minimum (optimal) value of the objective function Z = – 2 occurs at the extreme point A (2, 0).
Hence, the optimal solution to the given LP problem is: x1 = 2, x2 = 0 and Min Z = – 2.
Example 3.8 G.J. Breweries Ltd have two bottling plants, one located at ‘G’ and the other at ‘J’. Each plant
produces three drinks, whisky, beer and brandy named A, B and C respectively. The number of the bottles
produced per day are shown in the table.
Drink Plant at
G J
Whisky 1,500 1,500
Beer 3,000 1,000
Brandy 2,000 5,000

A market survey indicates that during the month of July, there will be a demand of 20,000 bottles of whisky,
40,000 bottles of beer and 44,000 bottles of brandy. The operating cost per day for plants at G and J are 600 and
400 monetary units. For how many days each plant be run in July so as to minimize the production cost, while
still meeting the market demand? Formulate this problem as an LP problem and solve that using graphical
method. [Pune Univ., MBA, 2009]
Solution Let us define the following decision variables:
x1 and x2 = number of days of work at plant G and J, respectively.
Then LP model of the given problem is:
Minimize Z = 600x1 + 400x2
subject to the constraints
(i) Whisky : 1500x1 + 1500x2 = 20,000 (ii) Beer : 3000x1 + 1000x2 = 40,000,
(iii) Brandy : 2000x1 + 5000x2 = 44,000
and x1, x2  0.
Plot on a graph each constraint by first treating it as linear equation. The feasible solution space (region) is shown
in Fig. 3.8 by shaded area. This region is bounded from below by the extreme points A, B, and C. The coordinates of
extreme points of the feasible solution space marked by shaded area are: A(22, 0), B(12, 4) and C(0, 40).

Fig. 3.8
Graphical Solution
of LP Problem
 Linear Programming: The Graphical Method 77

The value of objective function at each of the extreme points is shown shown is Table 3.8.

Extreme Point Coordinates Objective Function Value

(x1, x2) Z = 600x1 + 400x2
A (22, 0) 600(22) + 400(0) = 13,200 Table 3.8
B (12, 4) 600(12) + 400(4) = 8,800 Set of Feasible
C (0, 40) 600(0) + 400(40) = 16,000 Solutions

The minimum (optimal) value of objective function occurs at point B(12, 4). Hence, the Plant G should run for
x1 = 12 days and plant J for x2 = 4 days to have a minimum production cost of Rs. 8,800.
Example 3.9 A diet for a sick person must contain at least 4,000 units of vitamins, 50 units of minerals and 1,400
calories. Two foods A and B are available at a cost of Rs. 4 and Rs. 3 per unit, respectively. If one of A contains 200
units of vitamins, 1 unit of mineral and 40 calories and one unit of food B contains 100 units of vitamins, 2 units of
minerals and 40 calories. Formulate this problem as an LP model and solve it by graphical method to find
combination of foods to be used to have least cost?
Solution The data of the given problem can be summarized as shown below.

Food Units Content of Cost per

Vitamins Mineral Calories Unit (Rs.)

A 200 1 40 4
B 100 2 40 3
Minimum requirement 4,000 50 1,400

Let us define the following decision variables:

x1 and x2 = number of units of food A and B to be used, respectively.
Then LP model of the given problem is:
Minimize (total cost) Z = 4x1 + 3x2
subject to the constraints
(i) Vitamins : 200x1 + 100x2  4,000, (ii) Mineral : x1 + 2x2  50,
(iii) Calories : 40x1 + 40x2  1,400
and x1, x2  0

Fig. 3.9
Graphical Solution
of LP Problem

Plot on a graph each constraint by treating it as a linear equation. Then use inequality sign of each constraint
to mark the feasible solution space (region) by shaded area as shown in Fig. 3.9. The coordinates of the extreme
points of the feasible solution space are: A(0, 40), B(5, 30), C(20, 15) and D(50, 0). The value of objective
function at each of these extreme points is shown in Table 3.9.
78 Operations Research: Theory and Applications

Extreme Point Coordinates Objective Function Value

(x1, x2) Z = 4x1 + 3x2

A (0, 40) 4(0) + 3(40) = 120

Table 3.9 B (5, 30) 4(5) + 3(30) = 110
Set of Feasible C (20, 15) 4(20) + 3(15) = 125
Solutions D (50, 0) 4(50) + 3(0) = 200

The minimum (optimal) value of the objective function Z = 110 occurs at the extreme point B(5, 30). Hence, to
have least cost of Rs. 110, the diet should contain x1 = 5 units of food A and x2 = 30 units of food B.

3.3.4 Examples on Mixed Constraints LP Problems

Example 3.10 A firm plans to purchase at least 200 quintals of scrap containing high quality metal X and
low quality metal Y. It decides that the scrap to be purchased must contain at least 100 quintals ofmetal
X and not more than 35 quintals of metal Y. The firm can purchase the scrap from two suppliers (A and B)
in unlimited quantities. The percentage of X and Y metals in terms of weight in the scrap suppliedby A
and B is given below.
Metals Supplier A Supplier B
X 25% 75%
Y 10% 20%

The price of A’s scrap is Rs 200 per quintal and that of B is Rs 400 per quintal. The firm wants to
determine the quantities that it should buy from the two suppliers so that the total cost is minimized.
[Delhi Univ., MBA, 1998, 2001]
Solution Let us define the following decision variables:
x1 and x2 = quantity (in quintals) of scrap to be purchased from suppliers A and B, respectively.
Then LP model of the given problem is:
Minimize Z = 200x1 + 400x2
subject to the constraints
(i) Maximum purchase : x1 + x2  200
x1 3x2
(ii) Scrap containing :   100 or x1 + 3x2  400 (Metal X)
4 4

x1 x2
  35 or x1 + 2x2  350 (Metal Y)
10 5
and x1, x2  0.

Fig. 3.10
Graphical Solution
of LP Problem
 Linear Programming: The Graphical Method 79

Plot on a graph each constraint by treating it as a linear equation. Then use inequality sign of each
constraint to mark the feasible solution space (or region) by shaded area as shown in Fig. 3.10. The feasible
region is bounded by the corners, A, B and C.
The coordinates of the extreme points of the feasible region are: A = (100, 100), B = (250, 50), and
C = (50, 150). The value of objective function at each of these extreme points is shown in Table
3.10.

Extreme Coordinates Objective Function Value

Point (x1, x2) Z = 200x1 + 400x2
A (100, 100) 200(100) + 400(100) = 60,000 Table 3.10
B (250, 50) 200(250) + 400(50) = 70,000 Set of Feasible
C (50, 150) 200(50) + 400(150) = 70,000 Solutions

Since minimum (optimal) value of objective function, Z = Rs 60,000 occurs at the extreme point
A (100, 100), firm should buy x1 = x2 = 100 quintals of scrap each from suppliers A and B.

Example 3.11 Use the graphical method to solve the following LP problem.
Maximize Z = 7x1 + 3x2
subject to the constraints
(i) x1 + 2x2  3 (ii) x1 + x2  4
(iii) 0  x1  5/2 (iv) 0  x2  3/2
and x1, x2  0.
Solution Plot on a graph each constraint by first treating it as a linear equation. Then use the inequality
condition of each constraint to mark the feasible region by shaded area as shown in Fig. 3.11.

Fig. 3.11
Graphical Solution
of LP Problem
The coordinates of the extreme points of the feasible region are: A = (5/2, 1/4), B = (5/2, 3/2), and
C = (0, 3/2). The value of the objective function at each of these extreme points is shown in Table 3.11.
Extreme Point Coordinates Objective Function Value
(x1, x2) Z = 7x1 + 3x2

A (5/2, 1/4) 7(5/2) + 3(1/4) = 73/4

B (5/2, 3/2) 7(5/2) + 3(3/2) = 22 Table 3.11
C (0, 3/2) 7(0) + 3(3/2) = 9/2 Set of Feasible
Solutions
The maximum (optimal) value of the objective function, Z = 22 occurs at the extreme point B (5/2, 3/2).
Hence, the optimal solution to the given LP problem is: x1 = 5/2, x2 = 3/2 and Max Z = 22.

Example 3.12 Use the graphical method to solve the following LP problem.
Minimize Z = 20x1 + 10x2
subject to the constraints
(i) x1 + 2x2  40, (ii) 3x1 + x2  30, (iii) 4x1 + 3x2  60
and x1, x2  0.
80 Operations Research: Theory and Applications

Solution Plot on a graph each constraint by first treating it as a linear equation. Then use the inequality
condition of each constraint to mark the feasible region by shaded area as shown in Fig. 3.12.
The coordinates of the extreme points of the feasible region are: A = (15, 0), B = (40, 0), C = (4, 18)
and D = (6, 12). The value of the objective function at each of these extreme points is shown Table 3.12.

Extreme Point Coordinates Objective Function Value

(x1, x2) Z = 20x1 + 10x2
A (15,0) 20(15) + 10(0) = 300
Table 3.12 B (40,0) 20(40) + 10(0) = 800
Set of Feasible C (4,18) 20(4) + 10(18) = 260
Solutions D (6,12) 20(6) + 10(12) = 240

Fig. 3.12
Graphical Solution
of LP Problem

The minimum (optimal) value of the objective function, Z = 240 occurs at the extreme point D (6, 12).
Hence, the optimal solution to the given LP problem is: x1 = 6, x2 = 12 and Min Z = 240.
Example 3.13 Use the graphical method to solve the following LP problem.
Maximize Z = 2x1 + 3x2
subject to the constraints
(i) x1 + x2  30 (ii) x2  3 (iii) 0  x2  12
(iv) 0  x1  20 (v) x1 – x2  0
and x1, x2  0.
Solution Plot on a graph each constraint by first treating it as a linear equation. Then use the inequality
condition of each constraint to mark the feasible region by shaded area as shown in Fig. 3.13.

Fig. 3.13
Graphical Solution
of LP Problem
 Linear Programming: The Graphical Method 81

The coordinates of the extreme points of the feasible region are : A = (3, 3), B = (20, 3), C = (20, 10), D = (18, 12)
and E = (12, 12). The value of the objective function at each of these extreme points is shown in Table 3.13.

Extreme Point Coordinates Objective Function Value

(x1, x2) Z = 2x1 + 3x2
A (3, 3) 2(3) + 3(3) = 15
B (20, 3) 2(20) + 3(3) = 49
C (20, 10) 2(20) + 3(10) = 70 Table 3.13
D (18, 12) 2(18) + 3(12) = 72 Set of Feasible
E (12, 12) 2(12) + 3(12) = 60 Solutions

The maximum value of the objective function, Z = 72 occurs at the extreme point D (18, 12). Hence,
the optimal solution to the given LP problem is: x1 = 18, x2 = 12 and Max Z = 72.

Example 3.14 A firm makes two products X and Y, and has a total production capacity of 9 tonnes per day.
Both X and Y require the same production capacity. The firm has a permanent contract to supply at least 2
tonnes of X and at least 3 tonnes of Y per day to another company. Each tonne of X requires 20 machine
hours of production time and each tonne of Y requires 50 machine hours of production time. The daily
maximum possible number of machine hours is 360. All of the firm’s output can be sold. The profit made is
Rs 80 per tonne of X and Rs 120 per tonne of Y. Formulate this problem as an LP model and solve it by
using graphical method to determine the production schedule that yields the maximum profit.
[Delhi Univ., MBA, 1999]

Solution Let us define the following decision variables:

x1 and x2 = number of units (in tonnes) of products X and Y to be manufactured, respectively.
Then the LP model of the given problem can be written as
Maximize (total profit) Z = 80x1 + 120x2
subject to the constraints
(i) x1 + x2  9 (Production capacity)
(ii) x1  2; x2  3 (Supply)
(iii) 20x1 + 50x2  360 (Machine hours)
and x1, x2  0
For solving this LP problem graphically, plot on a graph each constraint by first treating it as a linear
equation. Then use the inequality sign of each constraint to mark the feasible region as shown in Fig. 3.14.

Fig. 3.14
Graphical Solution
of LP Problem

The coordinates of the extreme points of the feasible region are: A = (2, 3), B = (6, 3), C = (3, 6), and
D = (2, 6.4). The value of the objective function at each of these extreme points is shown in Table 3.14.
82 Operations Research: Theory and Applications

Extreme Point Coordinates Objective Function Value

(x1, x2) Z = 80x1 + 120x2
A (2, 3) 80(2) + 120(3) = 520
Table 3.14 B (6, 3) 80(6) + 120(3) = 840
Set of Feasible C (3, 6) 80(3) + 120(6) = 960
Solutions D (2, 6.4) 80(2) + 120(6.4) = 928

The maximum (optimal) value of the objective function, Z = 960 occurs at the extreme point C (3, 6).
Hence the company should produce, x1 = 3 tonnes of product X and x2 = 6 tonnes of product Y in order
to yield a maximum profit of Rs 960.
Example 3.15 The standard weight of a special purpose brick is 5 kg and it contains two basic ingredients
B1 and B2. B1 costs Rs 5 per kg and B2 costs Rs 8 per kg. Strength considerations dictate that the brick should
contain not more than 4 kg of B1 and a minimum of 2 kg of B2. Since the demand for the product is likely to
be related to the price of the brick, graphically find out the minimum cost of the brick satisfying the above
conditions.
Solution Let us define the following decision variables.
x1 and x2 = amount (in kg) of ingredient B1 and B2 contained in the brick, respectively.
Then the LP model of the given problem can be written as:
Minimize (total cost) Z = 5x1 + 8x2
subject to the constraints
(i) x1  4, (ii) x2  2, (iii) x1 + x2 = 5 (Brick strength)
and x1, x2  0
For solving this LP problem graphically, plotting on a graph each constraint by first treating it as a linear
equation. Then use the inequality condition of each constraint to mark the feasible region as shownin Fig.
3.15.

Fig. 3.15
Graphical Solution
of LP Problem

It may be noted that in Fig. 3.15 there is no unique (or single) feasible region. The value of the objective
function at extreme points A = (3, 2) and B = (4, 2) is given in Table 3.15.

Extreme Point Coordinates Objective Function Value

(x1, x2) Z = 5x1 + 8x2
Table 3.15
Set of Feasible A (3, 2) 5(3) + 8(2) = 31
Solutions B (4, 2) 5(4) + 8(2) = 36

The minimum value of objective function Z = 31 occurs at the extreme point A (3, 2). Hence, the optimal
product mix is: 3 kg of ingredient B1 and 2 kg of ingredient B2 of a special purpose brick to achieve the
minimum cost of Rs 31.
 Linear Programming: The Graphical Method 83

Example 3.16 The manager of an oil refinery must decide on the optimal mix of two possible blending
processes of which the inputs and outputs per production run are as follows:
The maximum amounts
available of crudes A and B are Process Input (units) Output (units)
200 units and 150 units, respec- (units) Grade A Grade B Gasoline X Gasoline Y
tively. Market requirements show
1 5 3 5 8
that at least 100 units of gasoline X
and 80 units of gasoline Y mustbe 2 4 5 4 4
produced. The profits per
production run for process 1 and process 2 are Rs 300 and Rs 400, respectively. Formulate this problem
as an LP model and solve it using graphical method to determine the production run for process 1 and 2.
[Gujarat Univ., MBA, 1999]
Solution Let us define the following decision variables
x1 and x2 = number of production run for process 1 and process 2, respectively.
The LP model of the given problem can be written as:
Maximize (total profit) Z = 300x1 + 400x2
subject to the constraints
(i) 5x1 + 4x2  200
3x1 + 5x2  150  (Input)
(ii) 5x1 + 4x2  100
8x1 + 4x2  80  (Output)
and x1 , x2  0

For solving this LP problem graphically, plotting on a graph each constraint by treating it as a linear
equation. Then use the inequality condition of each constraint to mark the feasible region as shown in Fig. 3.16.

Fig. 3.16
Graphical Solution
of LP Problem

The coordinates of extreme points of the feasible region are: A = (20, 0), B = (40, 0), C = (400/13,
150/13), D = (0, 30) and E = (0, 25). The value of the objective function at each of these extreme points
is give in Table 3.16.

Extreme Coordinates Objective Function Value

Point (x1, x2) Z = 300x1 + 400x2
A (20, 0) 300(20) + 400(0) = 6,000
B (40, 0) 300(40) + 400(0) = 12,000
C (400/13, 150/13) 300(400/13) + 400(150/13) = 1,80,000/13
Table 3.16
D (0, 30) 300(0) + 400(30) = 12,000
Set of Feasible
E (0, 25) 300(0) + 400(25) = 10,000 Solutions
84 Operations Research: Theory and Applications

The maximum (optimal) value of the objective function occurs at the extreme point (400/13, 150/13). Hence,
the manager of the oil refinery should produce, x1 = 400/13 units under process 1 and x2 = 150/13 units under
process 2 in order to achieve the maximum profit of Rs 1,80,000/13.
Example 3.17 A manufacturer produces two different models – X and Y of the same product. Model X
makes a contribution of Rs 50 per unit and model Y, Rs 30 per unit, towards total profit. Raw materials r1 and r2
are required for production. At least 18 kg of r1 and 12 kg of r2 must be used daily. Also at most 34 hours of
labour are to be utilized. A quantity of 2 kg of r1 is needed for model X and 1 kg of r1 for model Y. For each of X
and Y, 1 kg of r2 is required. It takes 3 hours to manufacture model X and 2 hours to manufacture model Y. How
many units of each model should be produced in order to maximize the profit?
Solution Let us define the following decision variables:
x1 and x2 = number of units of model X and Y to be produced, respectively.
Then the LP model of the given problem can be written as:
Maximize (total profit) Z = 50x1 + 30x2
subject to the constraints
(i) 2x1 + x2  18
x1 + x2  12  (Raw material)
(ii) 3x1 + 2x2  34 (Labour hours)
and x1, x2  0
For solving this LP problem graphically, plotting on a graph each constraint by treating it as a linear
equation. Then use the inequality sign of each constraint to mark the feasible region (shaded area) as shown in
Fig. 3.17.

Fig. 3.17
Graphical Solution
of LP Problem

The coordinates of extreme points of the feasible region are: A = (6, 6), B = (2, 14), and C = (10, 2). The value
of the objective function at each of these points is given in Table 3.17.

Extreme Point Coordinates Objective Function Value

(x1, x2) Z = 50x1 + 30x2

Table 3.17 A (6, 6) 50(6) + 30(6) = 480

Set of Feasible B (2, 14) 50(2) + 30(14) = 520
Solutions C (10, 2) 50(10) + 30(2) = 560

Since the maximum (optimal) value of Z = 560 occurs at the point C (10, 2), the manufacturer should produce
x1 = 10 units of model X and x2 = 2 units of Y in order to yield the maximum profit of Rs 560.
 Linear Programming: The Graphical Method 85

Example 3.18 An advertising agency wishes to reach two types of audiences – customers with annual
income greater than one lakh rupees (target audience A) and customers with annual income of less than
one lakh rupees (target audience B). The total advertising budget is Rs 2,00,000. One programme of TV
advertising costs Rs 50,000; one programme of radio advertising costs Rs 20,000. For contract reasons, at
least three programmes ought to be on TV and the number of radio programmes must be limited to 5. Surveys
indicate that a single TV programme reaches 4,50,000 prospective customers in target audience A and 50,000
in target audience B. One radio programme reaches 20,000 prospective customers in target audience A and
80,000 in target audience B. Determine the media mix so as to maximize the total reach.
[Delhi Univ., MBA, 2008]

Solution Let us define the following decision variables:

x1 and x2 = number of programmes to be released on TV and radio, respectively.
Then the LP model of the given problem can be expressed as:
Maximize Z = (4,50,000 + 50,000) x1 + (20,000 + 80,000) x2
subject to the constraints
(i) 50,000x1 + 20,000x2  2,00,000 or 5x1 + 2x2  20 (Budget available)
(ii) x1  3; x2  5 (Programmes)
and x1, x2  0
For solving this LP problem by graphical method, plot each constraint on a graph first treating it as
a liner equation. Then use inequality sign of each constraint to mark feasible solution region (shaded area) as
shown in Fig. 3.18.

Fig. 3.18
Graphical Solution
of LP Problem

The coordinates of the extreme points of feasible region are: A = (3, 0), B = (4, 0), and C = (3, 5/2). The
value of the objective function at each of these three corner points is given in Table 3.18.

Extreme Point Coordinates Objective Function Value

(x1, x2) Z = 5,00,000x1 + 1,00,000x2

A (3, 0) 5,00,000(3) + 1,00,000(0) = 15,00,000

Table 3.18
B (4, 0) 5,00,000(4) + 1,00,000(0) = 20,00,000 Set of Feasible
C (3, 5/2) 5,00,000(3) + 1,00,000(5/2) = 17,50,000 Solutions

Since the maximum (optimal) value of Z = 20,00,000 occurs at the point B (4, 0), the agency must release,
x2 = 4 programmes on TV and x2 = 0 (no programme on radio) in order to achieve the maximum reach of
Z = 20,00,000 audiences.
86 Operations Research: Theory and Applications

3.3.5 Iso-profit (Cost) Function Line Method

According to this method, the optimal solution is found by using the slope of the objective function line
(or equation). An iso-profit (or cost) line is a collection of points that give solution with the same value
of objective function. By assigning various values to Z, we get different profit (cost) lines. Graphically many
such lines can be plotted parallel to each other. The steps of iso-profit (cost) function method are as follows:
Step 1: Identify the feasible region and extreme points of the feasible region.
Step 2: Draw an iso-profit (iso-cost) line for an arbitrary but small value of the objective function without
violating any of the constraints of the given LP problem. However, it is simple to pick a value that gives
an integer value to x1 when we set x2 = 0 and vice-versa. A good choice is to use a number that is divided
by the coefficients of both variables.
Step 3: Move iso-profit (iso-cost) lines parallel in the direction of increasing (decreasing) objective function
Iso-profit (or cost)
function line is a values. The farthest iso-profit line may intersect only at one corner point of feasible region providing a single
straight line that optimal solution. Also, this line may coincide with one of the boundary lines of the feasible area. Then at
represents all non- least two optimal solutions must lie on two adjoining corners and others will lie onthe boundary connecting
negative
combinations of x1 them. However, if the iso-profit line goes on without limit from the constraints, then an unbounded solution
and x2 variable would exist. This usually indicates that an error has been made in formulating the LP model.
values for a
particular profit (or Step 4: An extreme (corner) point touched by an iso-profit (or cost) line is considered as the optimal solution
cost) level. point. The coordinates of this extreme point give the value of the objective function.

Example 3.19 Consider the LP problem

Maximize Z = 15x1 + 10x2
subject to the constraints
(i) 4x1 + 6x2  360, (ii) 3x1  180, (iii) 5x2  200
and x1, x2  0.
Solution: Plot each constraint on a graph first treating it as a linear equation. Then use inequality sign of each
constraint to mark feasible solution region (shaded area) as shown in Fig. 3.19.
A family of lines (shown by dotted lines) that represents various values of objective function is shown
in Fig. 3.19. Such lines are referred as iso-profit lines.

Fig. 3.19
Optimal Solution
(Iso-profit Function
Approach)

In Fig. 3.19, a value of Z = 300 is arbitrarily selected. The iso-profit (objective) function equation then
becomes: 15x1 + 10x2 = 300. This equation is also plotted in the same way as other equality constraints plotted
before. This line is then moved upward until it first intersects a corner (or corners) in the feasible region
(corner B). The coordinates of corner point B can be read from the graph or can be computed asthe
intersection of the two linear equations.
 Linear Programming: The Graphical Method 87

The coordinates x1 = 60 and x2 = 0 of corner point B satisfy the given constraints and the total profit
obtained is Z = 1,100.

3.3.3 Comparison of Two Graphical Solution Methods

After having plotted the constraints of the given LP problem to locate the feasible solution region (area) one
of the two graphical solution methods may be used to get optimal value of the given LP problem.

Extreme Point Method Iso-Profit (or Cost) Method

(i) Identify coordinates of each of the extreme (or (i) Determine the slope (x1, x2) of the objective func-
corner) points of the feasible region by either tion and then join intercepts to identify the profit
drawing perpendiculars on the x-axis and the y- (or cost) line.
axis or by solving two intersecting equations.
(ii) Compute the profit (or cost) at each extreme point (ii) In case of maximization, maintain the same slope
by substituting that point’s coordinates into the through a series of parallel lines, and move the line
objective function. upward towards the right until it touches the feasible
region at only one point. But in case of minimization,
move downward towards left until it touches only
one point in the feasible region.
(iii) Identify the optimal solution at that extreme point (iii) Compute the coordinates of the point touched by the
with highest profit in a maximization prob- lem or iso-profit (or cost) line on the feasible region.
lowest cost in a minimization problem. (iv) Compute the profit or cost.

3.4 SPECIAL CASES IN LINEAR PROGRAMMING

3.4.1 Alternative (or Multiple) Optimal Solutions

We have seen that the optimal solution of any linear programming problem occurs at an extreme point of
the feasible region and that the solution is unique, i.e. no other solution yields the same value of the objective Alternative optimal
function. However, in certain cases, a given LP problem may have more than one solution yieldingthe same solutions are
arrived only when
optimal objective function value. Each of such optimal solutions is termed as alternative optimal solution. slope of the
There are two conditions that should be satisfied for an alternative optimal solution to exist: objective function is
(i) The slope of the objective function should be the same as that of the constraint forming the boundary same as the slope
of a constraint in
of the feasible solutions region, and the LP problem.
(ii) The constraint should form a boundary on the feasible region in the direction of optimal movement
of the objective function. In other words, the constraint should be an active constraint.
Remark The constraint is said to be active (binding or tight), if at the point of optimality, the left-hand side
of a constraint equals the right-hand side. In other words, an equality constraint is always active, and
inequality constraint may or may not be active.
Geometrically, an active constraint is one that passes through one of the extreme points of the feasible
solution space.
Example 3.20 Use the graphical method to solve the following LP problem.
Maximize Z = 10x1 + 6x2
subject to the constraints
(i) 5x1 + 3x2  30, (ii) x1 + 2x2  18
and x1, x2  0.
Solution The constraints are plotted on a graph by first treating these as equations and then their inequality
signs are used to identify feasible region (shaded area) as shown in Fig. 3.20. The extreme pointsof the region
are O, A, B and C.
88 Operations Research: Theory and Applications

Fig. 3.20
Graphical Solution
— Multiple Optima

Since objective function (iso-profit line) is parallel to the line BC (first constraint : 5x1 + 3x2 = 30), which
also falls on the boundary of the feasible region. Thus, as the iso-profit line moves away from the origin,
it coincides with the line BC of the constraint equation line that falls on the boundary of the feasible region.
This implies that an optimal solution of LP problem can be obtained at any point between B and C including
extreme points B and C on the same line. Therefore, several combinations of values of x1 and x2 give the
same value of objective function.
The value of variables x1 and x2 obtained at extreme points B and C should only be considered to establish
that the solution to an LP problem will always lie at an extreme point of the feasible region.
The value of objective function at each of the extreme points is shown in Table 3.19.

Extreme Point Coordinates Objective Function Value

(x1, x2) Z = 10x1 + 6x2
O (0, 0) 10(0) + 6(0) = 0
Table 3.19 A (0, 9) 10(0) + 6(9) = 54
Set of Feasible B (6/7, 60/7) 10(6/7) + 6(60/7) = 60
Solutions C (6, 0) 10(6) + 6(0) = 60

Since value (maximum) of objective function, Z = 60 at two different extreme points B and C is same,
therefore two alternative solutions: x1 = 6/7, x2 = 60/7 and x1= 6, x2 = 0 exist.
Unbounded Remark If slope of a constraint is parallel to the slope of the objective function, but it does not fallon
solution exists the boundary of the feasible region, the multiple solutions will not exist. This type of a constraint is called
when the value of a redundant constraint (a constraint whose removal does not change the feasible region.)
decision variables
can be made
infinitely large 3.4.2 Unbounded Solution
without violating any
of the LP problem Sometimes an LP problem may have an infinite solution. Such a solution is referred as an unbounded solution.
constraints. It happens when value of certain decision variables and the value of the objective function (maximization
case) are permitted to increase infinitely, without violating the feasibility condition.
It may be noted that there is a difference between unbounded feasible region and unbounded solution
to a LP problem. It is possible that for a particular LP problem the feasible region may be unbounded but LP
problem solution may not be unbounded, i.e. an unbounded feasible region may yield some definite value
of the objective function. In general, an unbounded LP problem solution exists dueto improper formulation
of the real-life problem.
Example 3.21 Use the graphical method to solve the following LP problem:
Maximize Z = 3x1 + 4x2
subject to the constraints
(i) x1 – x2  –1 (ii) – x1 + x2  0
and x1, x2  0.
 Linear Programming: The Graphical Method 89

Solution Plot on a graph each constraint by first treating it as a linear equation. Then use the inequality
condition of each constraint to mark the feasible region (shaded area) as shown in Fig. 3.21.

Fig. 3.21
Unbounded
Solution

It may be noted from Fig. 3.21 that there exist an infinite number of points in the convex region for
which the value of the objective function increases as we move from the extreme point (origin), to the right.
That is, the value of variables x1 and x2 can be made arbitrarily large and accordingly the value of objective
function Z will also increase. Thus, the LP problem has an unbounded solution.
Example 3.22 Use graphical method to solve the following LP problem:
Maximize Z = 3x1 + 2x2
subject to the constraints
(i) x1 – x2  1 (ii) x1 + x2  3
and x1, x2  0.
Solution Plot on a graph each constraint by first treating it as a linear equation. Then use the inequality sign
of each constraint to mark the feasible region (shaded area) as shown in Fig. 3.22.
It may be noted that the shaded region (solution space) is unbounded from above. The two corners of the
region are, A = (0, 3) and B = (2, 1). The value of the objective function at these corners is: Z(A) = 6 andZ(B)
= 8.

Fig. 3.22
Graphical Solution
of LP Problem

Since the given LP problem is of maximization, there exist a number of points in the shaded region
for which the value of the objective function is more than 8. For example, the point (2, 2) lies in the
region and the objective function value at this point is 10 which is more than 8. Thus, as value of
variables x1 and x2 increases arbitrarily large, the value of Z also starts increasing. Hence, the LP problem
has an unbounded solution.
Example 3.23 Use graphical method to solve the following LP problem:
Maximize Z = 5x1 + 4x2
subject to the constraints
(i) x1 – 2x2  1, (ii) x1 + 2x2  3
and x1, x2  0 [PT Univ., BE, 2002]
Solution Constraints are plotted on a graph as usual as shown in Fig. 3.23. The solution space (shaded area)
is bounded from below and unbounded from above.
90 Operations Research: Theory and Applications

The two extreme points of the solution space are, A(0, 3/2) and B(2, 1/2). The value of objective function at
these points is Z(A) = 6 and Z(B) = 12. Since the given LP problem is of maximization, there exists a number of
points in the solution space where the value of objective function is much more than 12. Hence, the unique
value of Z cannot be found. The problem, therefore, has an unbounded solution.

Fig. 3.23
Unbounded
Solution

Example 3.24 Use graphical method to solve the following LP problem.

Maximize Z = – 4x1 + 3x2
subject to the constraints
(i) x1 – x2  0, (ii) x1  4
and x1, x2  0. [Punjab Univ., B Com, 2009]
Solution The solution space satisfying the constraints is shown shaded in Fig 3.24. The line x1 – x2 = 0 is
drawn by joining origin point (0, 0) and has a slope of 45°. Also as x1 – x2  0, i.e. x1  x2, the solution space due
to this line is in the upward direction.

An infeasible
solution lies
outside the feasible
region, it violates
one or more of the
given constraints.

Fig. 3.24
Unbounded
Solution

Since objective function is of maximization, therefore, the value of Z can be made arbitrarily large. Hence,
this LP problem has an unbounded solution. Value of variable x1 is limited to 4, while value of variable x2 can be
increased indefinitely.

3.4.3 Infeasible Solution

An infeasible solution to an LP problem arises when there is no solution that satisfies all the constraints
simultaneously. This happens when there is no unique (single) feasible region. This situation arises when
a LP model that has conflicting constraints. Any point lying outside the feasible region violates one or more
of the given constraints.
 Linear Programming: The Graphical Method 91

Example 3.25 Use the graphical method to solve the following LP problem:
Maximize Z = 6x1 – 4x2
subject to the constraints
(i) 2x1 + 4x2  4 (ii) 4x1 + 8x2  16
and x1, x2  0

Fig. 3.25
An Infeasible
Solution
Solution The constraints are plotted on graph as usual as shown in Fig. 3.25. Since there is no unique
feasible solution space, therefore a unique set of values of variables x1 and x2 that satisfy all the constraints
cannot be determined. Hence, there is no feasible solution to this LP problem because of the conflicting
constraints.
Example 3.26 Use the graphical method to solve the following LP problem:

x2
Maximize Z = x1 +
2
subject to the constraints
(i) 3x1 + 2x2  12 (ii) 5x1  10
(iii) x1 + x2  8 (iv) – x1 + x2  4
and x1, x2  0
Solution The constraints are plotted on graph as usual and feasible regions are shaded as shown in
Fig. 3.26. The three shaded areas indicate non-overlapping regions. All of these can be considered feasible
solution space because they all satisfy some subsets of the constraints.

Fig. 3.26
An Infeasible
Solution
However, there is no unique point (x1, x2) in these shaded regions that can satisfy all the constraints
simultaneously. Thus, the LP problem has an infeasible solution.
Example 3.27 Use the graphical method to solve the following LP problem:
Maximize Z = 3x + 2y
subject to the constraints
(i) – 2x + 3y  9, (ii) 3x – 2y  –20

and x, y  0. [Punjab Univ., BE (Elect.) 2006]

92 Operations Research: Theory and Applications

Solution There are two solution spaces (shaded areas) shown in the Fig. 3.27. One of these solution spaceis
satisfying the constraint – 2x + 3y  9 while the other is satisfying the constraint 3x – 2y  –20. These two
shaded regions in the first quadrant do not overlap and hence there is no point (x, y) common to both the
shaded regions. This implies that the feasible solution to the problem does not exist. Consequently, this LP
problem has one infeasible solution.

Redundancy is a
situation in which
one or more
constraints do not
affect the feasible
solution region.

Fig. 3.27
No Feasible
Solution

3.4.4 Redundancy
A redundant constraint is one that does not affect the feasible solution region (or space) and thus redundancy
of any constraint does not cause any difficulty in solving an LP problem graphically. As shown in Fig. 3.16,
constraint 8x1 + 4x2  80 (Ex. 3.16) is redundant. In other words, a constraint is said to be redundant when it
may be more binding (restrictive) than another.

CONCEPTUAL QUESTIONS

1. Explain graphical method of solving an LP problem. 7. You have just formulated a maximization LP problem and are
2. Give a brief description of an LP problem with illustrations. How preparing to solve it graphically. What criteria should you
can it be solved graphically? consider in deciding whether it would be easier to solve the
problem by the extreme point enumeration method or the iso-
3. What is meant by the term ‘feasible region’? Why must this be a
profit line method.
well-defined boundary for the maximization problem?
8. Under what condition is it possible for an LP problem to have
4. Define iso-profit and iso-cost lines. How do these help us to
obtain a solution to an LP problem? more than one optimal solution? What do these alternative
optimal solutions represent?
5. Explain the procedure of generating extreme point solutions to
9. What is an infeasible solution, and how does it occur? How is this
an LP problem, pointing out the assumptions made, if any.
condition recognized in the graphical method?
6. It has been said that each LP problem, that has a feasible
region, has an infinite number of solutions. Explain. 10. What is an unbounded solution, and how is this condition
recognized in the graphical method?

SELF PRACTICE PROBLEMS

1. Comment on the solution of the following LP problems.

(i) Max Z = 4x1 + 4x2 (ii) Max Z = 5x1 + 3x2 (iii) Max Z = 4x1 + 2x2
subject to x1 + 2x2  10 subject to 4x1 + 2x2  8 subject to – x1 + 2x2  6
6x1 + 6x2  36 x1  3 – x1 + x2  2
x1  6 x2  7 and x1, x2  0
and x1, x2  0 and x1, x2  0
 Linear Programming: The Graphical Method 93

(iv) Min Z = x1 – 2x2 (v) Max Z = 6x1 + 12x2 (vi) Max Z = 3x1 + 5x2
subject to – 2x1 + x2  8 subject to x1 + 2x2  10 subject to x1 + x2  100
– x1 + 2x2  –24 2x1 – 5x2  20 5x1 + 10x2  400
and x1, x2  0 x1 + x2  15 6x1 + 8x2  440
and x 1, x 2  0 and x1, x2  0

(vii) Max Z = 1.75 x1 + 1.5 x2 (viii) Max Z = 3x1 + 2x2 (ix) Max Z = 5x1 + 3x2
subject to – 2x1 + 3x2  9 3x1 + 5x2  15
subject to 8x1 + 5x2  320 subject to
4x1 + 5x2  20 x1 – 5x2  –20 5x1 + 2x2  10
x1  15; x2  10 and x1, x2  0 and x1, x2  0
and x1, x2  0 [Kerala, BSc (Engg.), 1995]

(x) Max Z = x1 + x2
subject to x1 + x2  1
– 3x1 + x2  3
and x1, x2  0

2. Solve the following LP problems graphically and state what your solution indicates.
(i) Min Z = 4x1 – 2x2 (ii) Min Z = 3x1 + 5x2 (iii) Min Z = 20x1 + 10x2
subject to x1 + x2  14 subject to – 3x1 + 4x2  12 subject to x1 + 2x2  40
3x1 + 2x2  36 2x1 – x2  – 2 3x1 + x2  30
2x1 + x2  24 2x1 + 3x2  12 4x1 + 3x2  60
and x1, x2  0 x1  4; x2  2 and x 1, x 2  0
and
x 1, x 2  0
(iv) Max Z = x1 + x2 are available per day. For belt B only 700 buckles are available per
subject to x1 – x2  0 day. Solve this problem to determine how many units of the two
types of belts the company should manufacture in order to have
3x1 – x2  – 3
the maximum overall profit?
and x1, x2  0 6. Consider a small plant which makes two types of automobile
3. A manufacturing firm produces two machine parts P1 and P2 parts, say A and B. It buys castings that are machined, bored
of a machine. For this it makes use of milling and grinding and polished. The capacity of machining is 25 per hour for A and
machines. The different machining times required for each part, 40 per hour for B, the capacity of boring is 28 per hourfor A
the machining times available on different machines and the and 35 per hour for B, and the capacity of polishing is 35 per
hour for A and 25 per hour for B. Casting for part Acosts Rs
profit on each machine part are as given below:
2 each and for part B it costs Rs 3 each. The company sells
these for Rs 5 and Rs 6, respectively. Thethree machines have
Manufacturing Maximum Time running costs of Rs 20, Rs 14 and Rs
Machine Time Required (min) Available 17.50 per hour. Assuming that any combination of parts A and
P1 P2 per Week (min) B can be sold, what product mix would maximize profit? Solve
this problem using the graphical method.
Lathe 10 1.5 2,500
7. A company machines and drills two castings X and Y. The
Milling machine 4 10 2,000
time required to machine and drill one casting including machine
Grinding machine 1 1.5 450 set-up time is as follows:
Profit per unit (Rs) 50 100

Determine the number of pieces of P1 and P2 to be manufactured Casting Machine Hours Drilling Hours
per week in order to maximize profit.
X 4 2
4. A furniture manufacturer makes two products: chairs and tables.
Y 2 5
These products are processed using two machines – A and
B. A chair requires 2 hours on machine A and 6 hours on
There are two lathes and three drilling machines. The working
machine B. A table requires 5 hours on machine A and no time week is of 40 hours; there is no lost time and overtime. Variables
on machine B. There are 16 hours per day available on machine costs for both castings are Rs 120 per unit while total fixed
A and 30 hours on machine B. The profit gained by the costs amount to Rs 1,000 per week. The selling price of
manufacturer from a chair is Rs 2 and from a table is Rs casting X is Rs 300 per unit and that of Y is Rs 360per unit.
10. Solve this problem to find the daily production of each of There are no limitations on the number of X and Y casting that
the two products. can be sold. The company wishes to maximizeits profit. You
5. A company produces two types of leather belts, A and B. Belt A is are required to (i) formulate a linear programmingmodel for the
of a superior quality and B is of an inferior quality. The profit from problem and (ii) solve this problem using the graphical method.
the two are 40 and 30 paise per belt, respectively. Each belt of [Delhi Univ., MBA, 2007]
type A requires twice as much time as required by a belt of type B. 8. A company possesses two manufacturing plants, each of which
If all the belts were of type B, the company could produce 1,000 can produce three products: X, Y and Z from a common raw
belts per day. But the supply of leather is sufficient only for 800 material. However, the proportions in which these products can
belts per day. Belt A requires a fancy buckle and only 400 of them be produced are different in each plant
94 Operations Research: Theory and Applications

and so are the plant’s operating costs per hour. The data on A and of Rs 120 for each unit of product B. With this information,
production per hour and costs together with current orders in formulate this problem as a linear programming model and
hand for each product, is as follows: determine the optimal level of output for the two products using
the graphic method.
Plant Products Operating Cost 13. A publisher of textbooks is in the process of bringing a new book
X Y Z per Hour (Rs) in the market. The book may be bound by either cloth or hard
paper. Each cloth-bound book sold contributes Rs 24 towards
A 2 4 3 9 the profit and each paperbound book contributes Rs 23. It
B 4 3 2 10 takes 10 minutes to bind a cloth cover, and 9 minutes to bind
Orders on hand 50 24 60 a paperback. The total available time for binding is 80 hours.
After considering a number of marketsurveys, it is predicted that
You are required to use the graphical method to find out the the cloth-cover sales will be anything more than 10,000 copies,
number of production hours needed to fulfil the orders on hand but the paperback sales will not be more than 6,000 copies.
at minimum cost. Formulate this problem as a LP problem and solve it
9. A company manufactures two products: A and B. Product A graphically.
yields a contribution of Rs 30 per unit and product B, Rs 40 14. PQR Feed Company markets two feed mixes for cattle. The feed
per unit, towards the fixed costs. It is estimated that the sales mix, Fertilex, requires at least twice as much wheat as barley.
of product A for the coming month will not exceed 20 units. Sales The second mix, Multiplex, requires at least twice as much barley
of product B have not yet been estimated but the company does as wheat. Wheat costs Rs 1.50 per kg, and only 1,000 kg are
have a contract to supply at least 10 units to a regular available this month. Barley costs Rs 1.25 per kg and 1,200 kg
customer. is available: Fertilex sells for Rs 1.80 per kg up to 99 kg and each
There are 100 machine hours available for the coming month. It additional kg over 99 sells for Rs 1.65. Multiplex sells at Rs 1.70
takes 4 hours to produce both – product A and product B. There per kg up to 99 kg and each additional kg over 99 sells for Rs
are 180 labour hours available and products A and B require 4 1.55 per kg. Bharat Farms is sure to buy any andall amounts of
hours and 6 hours of labour, respectively. Materials available are both mixes that PQR Feed Company will mix. Formulate this
restricted to 40 units for the two products. One unit of each problem as a LP problem to determine the product mix so as to
product requires one unit of material. The company wishes to maximize the profits.
maximize its profits. Use the graphical method, to find the optimum
15. On October 1st, a company received a contract to supply 6,000
product mix.
units of a specialized product. The terms of contract requires that
10. A company manufactures two kinds of machines, eachrequiring 1,000 units be shipped in the north of October; 3,000 units in
a different manufacturing technique. The deluxe machine November and 2,000 units in December. The company can
requires 18 hours of labour, 8 hours of testing and yields a profit manufacture 1,500 units per month on regular time and 750 units
of Rs 400. The standard machine requires per month on overtime. The manufacturing cost per item
3 hours of labour, 4 hours of testing and yields a profit of produced during regular time is Rs 3 and thecost per item
Rs 200. There are 800 hours of labour and 600 hours of testing produced during overtime is Rs 5. The monthly storage cost is
available each month. A marketing forecast has shown that the Re 1. Formulate this problem as a linear programming problem
monthly demand for the standard machine is to bemore than so as to minimize the total cost.
150. The management wants to know the number of each
16. A small-scale manufacturer has production facilities for
model to be produced monthly that would maximize total profit.
producing two different products. Each of the products requires
Formulate and solve this as a linear programming problem.
three different operations: grinding, assembly and testing.
[Delhi Univ., MBA, 2003, 2008]
Product I requires 15, 20 and 10 minutes to grind, assemble
11. A firm makes two types of furniture: chairs and tables. The
and test, respectively, on the other hand product II requires
contribution to profit by each product as calculated by the
7.5, 40 and 45 minutes for grinding, assembling and testing,
accounting department is – Rs 20 per chair and Rs 30 per table. respectively. The production run calls for at least 7.5 hours
Both products are to be processed on three machines M1, M2
of grinding time, at least 20 hours of assembly and at least
and M3. The time required in hours by each product and total
15 hours of testing time. If manufacturing product I costs Rs
time available in hours per week on each machine are as follows:
60 and product II costs Rs 90, determine the number ofunits
of each product the firm should produce in order to minimize the
Machine Chair Table Available Time (hrs) cost of operation.
17. A firm is engaged in breeding pigs. The pigs are fed on various
M1 3 3 36
products grown in the farm. Because of the need to ensure
M2 5 2 50
certain nutrient constituents, it is necessary to additionaly buy
M3 2 6 60
one or two products (call them A and B).
How should the manufacturer schedule his production in order The content of the various products (per unit) in the nutrient
to maximize profit? constituents (e.g. vitamins, proteins, etc.) is given in the following
12. The production of a certain manufacturing firm involves a table:
machining process that acquires raw materials and then
converts them into (unassembled) parts. These parts are then Nutrient Nutrient Content in Minimum Amount
sent to one of the two divisions for being assembly intothe Product of Nutrient
final product. Division 1 is used for product A, and Division 2 for
A B
product B. Product A requires 40 units of raw materialand 10
hours of machine processing time. Product B requires 80 units M1 36 6 108
of raw material and 4 hours of machine processing time. During
M2 3 12 36
the period, 800 units of raw material and 80 hours of machine
processing time are available. The capabilities of the two M3 20 10 100
assembly divisions during the period are 6 and 9 units,
respectively. The profit contribution per unit to profit and The last column of the above table gives the minimum quantity
overhead (fixed costs) is of Rs 200 for each unit of product of nutrient constituents M1, M2 and M3 that must be given to
 Linear Programming: The Graphical Method 95

the pigs. If products A and B cost Rs 20 and Rs 40 per unit, 22. A manufacturer produces electric hand saws and electric drills,
respectively, how much of each of these two products should for which the demand exceeds his capacity. The production cost
be bought so that the total cost is minimized? Solve this LP of a saw is Rs 6 and the production cost ofa drill Rs 4. The
problem graphically. shipping cost is 20 paise for a saw and 30 paise for a drill. A
18. A company manufacturing television sets and radios has four saw sells for Rs 9 and a drill sells for Rs
5.50. The budget allows a maximum of Rs 2,400 for production
major departments: chassis, cabinet, assembly and final testing.
costs and Rs 120 for shipping costs. Formulate this problem
The monthly capacities of these are as follows:
as an LP model and solve it to determine the number of saws
and drills that should be produced in order to maximize the
Department Television Radio Capacity excess of sales over production and shipping costs.
Capacity 23. A company produces two types of pens, say A and B. Pen A
is of a superior quality and pen B is of an inferior quality. Profits
Chassis 1,500 or 4,500
on pen A and B are Rs 5 and Rs 3 per pen, respectively. The
Cabinet 1,000 or 8,000 raw material required for producing single pen A is twice as that
Assembly 2,000 or 4,000 of pen B. The supply of raw material is sufficient only for 1,000
Testing 3,000 or 9,000 pens of B per day. Pen A requires a special clip and only 400
such clips are available per day. For pen B, only 700 clips are
The contribution of television is Rs 150 each and the available per day. Solve this LP problem graphically to find the
contribution of radio is Rs 250 each. Assume that the company product mix so that the company can make the maximum profit.
can sell any quantity of either product. Formulate this problem [Delhi Univ., MBA, Nov. 1998]
as an LP problem and solve it to determine the optimal 24. Two products A and B are to be manufactured. One single
combination of output. unit of product A requires 2.4 minutes of punch press time
and 5 minutes of assembly time. The profit for product A is
19. The commander of a small tank has been ordered to win and Re 0.60 per unit. One single unit of product B requires 3 minutes
occupy a valley located in a river delta area. He has 4 heavy of punch press time and 2.5 minutes of welding time. The profit
tanks and 10 light tanks. Each heavy tank requires 4 men to for product B is Re 0.70 per unit. The capacity of the punch
operate, whereas the light tank requires 2 men. The total number press department available for these products is 1,200
of men available is 29. The fire power of the heavytank is minutes/week. The welding department has an idle capacity of
three times that of the light tank. Yet the commander feels that 600 minutes/week and the assembly department has the
he should use more light tanks than heavy onessince the capacity of 1,500 minutes/week. Formulate this problem as an
light ones are more effective against guerillas. LP model to determine the quantities of products A and B that
(i) Formulate this problem as an LP model so as to would yield the maximum.
determine the number of tanks in each type to be sent 25. A rubber company is engaged in producing three different kinds
into the combat, keeping in mind that the commander of tyres A, B and C. These three different tyres areproduced at
wishes to maximize the total fire power. two different plants of the company, which have different
(ii) If you could persuade the commander that the rule‘more production capacities. In a normal 8-hour working day, Plant 1
light tanks than heavy tanks’ may not be applicable due produces 50, 100 and 100 tyres of types A, B and C,
to the fact that guerilla warfare is absent in the area, respectively. Plant 2, produces 60, 60 and 200 tyresof types
how many tanks of each type could be used? A, B and C, respectively. The monthly demand for types A, B and
20. A timber company cuts raw timber – oak and pine logs – into C is 2,500, 3,000 and 7,000 units, respectively. The daily cost
wooden boards. Two steps are required to produce boards of operation of Plants 1 and 2 is Rs 2,500 and Rs 3,500,
from logs. The first step involves removing the bark from the respectively. Formulate this problem as an LP model and solve
logs. Two hours are required to remove the bark from 1,000 it to determine how the company can minimize the number of
feet of oak logs and three hours per 1,000 feet of pine logs. After days on which it operates, per month, at the two plants, so that
the logs have been debarked, they must be cut into boards. It the total cost is also minimized, while the demand is also met.
takes 2.4 hours per 1,000 feet of oak logs to be cut into boards 26. Kishore Joshi mixes pet food in his basement on a small scale.
and 1.2 hours per 1,000 feet of pine logs. The bark removing He advertises two types of pet food: Diet-Sup and Gro-More.
machines can operate for up to 60 hours per week, while for Contribution from Diet-Sup is Rs 1.50 a bag and from Gro-
cutting machines this number is limited to 48 hoursper week. More Rs 1.10 a bag. Both are mixed from two basic ingredients
The company can buy a maximum of 18,000 feetof raw oak – a protein source and a carbohydrate source. Diet-Sup and
logs and 12,000 feet of raw pine logs each week. The profit per Gro-More require ingredients in these amounts:
1,000 feet of processed logs is Rs 1,800 andRs 1,200 for oak
and pine logs, respectively. Formulate this problem as an LP Protein Carbohydrate
model and solve it to determine how many feet of each type of
log should be processed each week inorder to maximize the Diet-Sup (7 kg bag) 4 kg 3 kg
profit. Gro-More (3 kg bag) 2 kg 1 kg
21. Upon completing the construction of his house, Mr Sharma
discovers that 100 square feet of plywood scrap and 80 square Kishore has the whole weekend ahead of him, but will not
feet of white pine scrap are in unusable form, which can be be able to procure more ingredients over the weekend. He
used for the construction of tables and bookcases. It takes 16 checks his bins and finds he has 700 kg of protein source
square feet of plywood and 16 square feet of white pine to and 500 kg of carbohydrate in the house. How many bags
construct a book case. By selling the finished products to a local of each food should he mix in order to maximize his profits?
furniture store, Mr Sharma can realize a profit of Rs25 on each Formulate this problem as an LP model and solve it to determine
table and Rs 20 on each bookcase. How can he most profitably the optimum food-mix that would maximize the profit.
use the leftover wood? Use the graphical method to solve this
LP problem.
96 Operations Research: Theory and Applications

HINTS AND ANSWERS

1. (i) Alternative solutions (ii) Infeasible solution 8. Let x1 and x2 = number of production hours for plants A and
(iii) Unbounded solution B needed to complete the orders, respectively.
(iv) Unbounded solution (v) Redundant Min Z = 9x1 + 10x2
constraints are second and third (vi) Infeasible solution subject to (i) 2x1 + 4x2  50 ; (ii) 4x1 + 3x2  24
(vii) No feasible solution (viii) Unbounded (iii) 3x1 + 2x2  60
solution and x1, x2  0
(ix) x1 = 20/19, x2 = 45/19 and Max Z = 535/19 Ans. x1 = 35/2 and x2 = 15/4 and Min Z = Rs 195.
(x) No feasible solution
9. Let x1 and x2 = number of units of products A and B,
2. (i) x1 = 8, x2 = 6 and Min Z = 20 respectively.
(ii) x1 = 3, x2 = 2 and Min Z = 19 Max Z = 30x1 + 40x2
(iii) x1 = 6, x2 = 12 and Min Z = 240
subject to (i) x1  20; x2  10, (ii) 4x1 + 4x2  100,
3. Let x1 and x2 = number of pieces of P1 and P2 to be (iii) 4x1 + 6x2  180, (iv) x1 + x2  40
manufactured, respectively. and x1, x2  0
Max Z = 50x1 + 100x2
11. Let x1 and x2 = number of chairs and tables to be made,
subject to (i) 10x1 + 15x2  2,500 respectively.
(ii) 4x1 + 10x2  2,000 Max Z = 20x1 + 30x2
(iii) x1 + 15x2  450 subject to (i) 3x1 + 3x2  36 ; (ii) 5x1 + 2x2  50
and x1, x2  0 (iii) 2x1 + 6x2  60
Ans. x1 = 187.5, x2 = 125 and Max Z = Rs 21,875. and x1, x2  0
4. Let x1 and x2 = number of chairs and tables produced, respectively. Ans. x1 = 3, x2 = 9 and Max Z = Rs 330.
Max Z = 2x1 + 10x2 12. Let x1 and x2 = number of units to be produced of products A
subject to (i) 2x1 + 5x2 16 ; (ii) 6x1  30 and B, respectively.
and x1, x2  0 Max Z = 200x1 + 120x2
Ans. x1 = 0, x2 = 3.2 and Max Z = Rs 32. subject to (i) 40x1 + 80x2  800, (ii) 10x1 + 04x2  80,
5. Let x1 and x2 = number of belts of types A and B produced, (iii) x1  6; (iv) x2  9
respectively. and x1, x2  0
Max Z = 0.40x1 + 0.30x2 Ans. x1 = 5, x2 = 7.5 and Max Z = Rs 1,900.
subject to (i) x1 + x2  800 ; (ii) 2x1 + x2  1,000 17. Let x1 and x2 = number of units of products A and B to be
(ii) x1  400 ; (iv) x2  700 bought, respectively
and x1, x2 0 Min (total cost) Z = 20x1 + 40x2
Ans. x1 = 200, x2 = 600 and Max = Rs 260.
subject to (i) 36x1 + 6x2  108 ; (ii) 3x1 + 12x2  36
6. Let x1 and x2 = number of parts manufactured of automobile A
(iii) 20x1 + l0x2  100
and B, respectively.
Max Z = {5  2  20 / 25  14 / 28  17.5 / 35}x1 and x1, x2  0

Ans. x1 = 4, x2 = 2 and Min Z = Rs 160.

 {6  3  20 / 40 14 / 35 17.5/ 25}x2 18. Let x1 and x2 = number of units of television and radio tubes
= 4.80x1 + 4.60x2 produced, respectively.
subject to Max (total profit) Z = 150x1 + 250x2

x1 x x x x1 x2 subject to (i) x1/1,500 + x2/4,500  1

(i) + 2 ≤ 1, (ii) 1 + 2 ≤ 1, (iii) + ≤1
25 40 28 35 35 25 (ii) x1/1,000 + x2/8,000  1
and x1, x2  0 (iii) x /2,000 + x /2,000  1
2
1
7. Let x1 and x2 = number of units of casting X and Y to be
manufactured, respectively. (iv) x1/3,000 + x2/9,000  1
Max Z = (300x + 360x ) – (120x + 120x ) – 1,000 and x1, x2  0

1 2 1 2
20. Let x1 and x2 = number of feet of raw oak logs and raw pine
= 180x1 + 240x2 – 1,000 logs to be processed each week, respectively

subject to (i) 4x1 + 2x2  2 × 40

Max Z = 1,800x1 + 1,200x2
(ii) 2x1 + 5x2  3 × 40
subject to (i) 2x1 + 3x2  60, (ii) 2.4x1 + 1.2x2  48,
and x1, x2  0
(iii) x1  18, (iv) x2  12
Ans. x1 = 10, x2 = 20 and Max Z = 5,600.
and x1, x2  0
 Linear Programming: The Graphical Method 97

CHAPTER SUMMARY
The graphical solution approaches (or methods) provide a conceptual basis for solving large and complex LP problems. A graphical
method is used to reach an optimal solution to an LP problem that has a number of constraints binding the objective function.
Both extreme point methods and the iso-profit (or cost) function line method are used for graphically solving any LP problem
that has only two decision variables.

CHAPTER CONCEPTS QUIZ

True or False Multiple Choice

1. The graphical method of solving linear programming problem 21. The graphical method of LP problem uses
is useful because of its applicability to many real life situations. (a) objective function equation
2. The problem of infeasibility in linear programming can only be (b) constraint equations
solved by making additional resources available which in turn (c) linear equations
changes the constraints of the problem. (d) all of the above
3. A linear programming problem is unbounded because 22. A feasible solution to an LP problem
constraints are incorrectly formulated. (a) must satisfy all of the problem’s constraints
4. If a LP problem has more than one solution yielding the same simultaneously
objective functional value, then such optimal solutions are (b) need not satisfy all of the constraints, only
known as alternative optimal solution. some of them
5. Since the constraints to a linear programming are always linear, (c) must be a corner point of the feasible region
we can graph them by locating only two different points on a (d) must optimize the value of the objective function
line. 23. Which of the following statements is true with respect to the
6. An optimal solution does not necessarily use up all the limited optimal solution of an LP problem
resources available. (a) every LP problem has an optimal solution
(b) optimal solution of an LP problem always occurs at an
7. The intersection of any two constraints is an extreme point which
extreme point
is a corner of the feasible region.
(c) at optimal solution all resources are completely used
8. When there are more than one optimal solution to the problem (d) if an optimal solution exists, there will always be at least
then the decision-maker will be unable to judge the best optimal one at a corner
solution among them.
24. An iso-profit line represents
9. If all the constraints are  inequalities in a linear programming (a) an infinite number of solutions all of which yield the same
problem whose objective function is to be maximized, then the profit
solution of the problem is unbounded. (b) an infinite number of solution all of which yield the same
10. The problem caused by redundant constraints is that two cost
isoprofit lines may not be parallel to each other. (c) an infinite number of optimal solutions
(d) a boundary of the feasible region
Fill in the Blanks
25. If an iso-profit line yielding the optimal solution coincides with
11. Constraints in a linear programming requiring all variables to a constraint line, then
be zero or positive are known as .......................... constraints (a) the solution is unbounded
and rest from limited resources are referred as ........................ (b) the solution is infeasible
constraints. (c) the constraint which coincides is redundant
12. Instead of maximizing profit in linear programming problem, (d) none of the above
we ensure linearity of the objective function by maximizing 26. While plotting constraints on a graph paper, terminal points on
........................ both the axes are connected by a straight line because
13. The ........................ points of the convex set give the basic (a) the resources are limited in supply
feasible solution to the linear programming. (b) the objective function is a linear function
14. A basic feasible solution is said to be ........................ if the (c) the constraints are linear equations or inequalities
values of all ......................... variables are nonzero and positive. (d) all of the above
15. If an optimal solution to a linear programming problem exists, 27. A constraint in an LP model becomes redundant because
it will lie at.......................... of the feasible solution. (a) two iso-profit line may be parallel to each other
16. A constraint in linear programming must be expressed as a (b) the solution is unbounded
linear ........................ or a linear ........................ . (c) this constraint is not satisfied by the solution values
17. When more than one solution best meets the objective of the (d) none of the above
linear programming problem then it is said to have 28. If two constraints do not intersect in the positive quadrant of
........................ solution. the graph, then
18 ............................. occurs when no value of the variable is able (a) the problem is infeasible
to satisfy all the constraints in linear programming problem (b) the solution is unbounded
simultaneously. (c) one of the constraints is redundant
19. An existence of objective of the problem for any firm is one (d) none of the above
of the major .........................of the linear programming problem. 29. Constraints in LP problem are called active if they
20. Any two isoprofit or isocost lines for a given linear programming (a) represent optimal solution
problem are ........................ to each other. (b) at optimality do not consume all the available resources
98 Operations Research: Theory and Applications

(c) both of (a) and (b) (d) all of the above

(d) none of the above 33. While solving a LP problem, infeasibility may be removed by
30. The solution space (region) of an LP problem is unbounded (a) adding another constraint
due to (b) adding another variable
(a) an incorrect formulation of the LP model (c) removing a constraint
(b) objective function is unbounded (d) removing a variable
(c) neither (a) nor (b) 34. If a non-redundant constraint is removed from an LP problem,
(d) both (a) and (b) then
31. While solving a LP model graphically, the area bounded by the (a) feasible region will become larger
constraints is called (b) feasible region will become smaller
(a) feasible region (b) infeasible region (c) solution will become infeasible
(c) unbounded solution (d) none of the above (d) none of the above
32. Alternative solutions exist of an LP model when 35. If one of the constraint of an equation in an LP problem has
(a) one of the constraints is redundant an unbounded solution, then
(b) objective function equation is parallel to one of the (a) solution to such LP problem must be degenerate
constraints (b) feasible region should have a line segment
(c) two constraints are parallel (c) alternative solutions exist
(d) none of the above

Answers to Quiz
1. T 2. F 3. F 4. T 5. T 6. T 7. T 8. F 9. T 10. T
11. non-negative, structural 12. Contribution 13. extreme 14. non-degenerative, basic,
15. an extreme point 16. equation, inequality 17. alternative 18. infeasibility
19. requirement 20. parallel.
21. (d) 22. (a) 23. (d) 24. (a) 25. (d) 26. (c) 27. (d) 28. (a) 29. (a) 30. (c)
31. (a) 32. (b) 33. (c) 34. (a) 35. (b)

CASE STUDY

Case 3.1: Raman Traveller

Mr. Raman, a local travel agent is planning a charter trip to a famous sea resort. The eight-day seven-night package
includes the round-trip fare, surface transportation, boarding and lodging and selected tour options. The charter trip
is restricted to 200 persons and past experience indicates that there will not be any problem in arranging 200 passangers.
The problem faced by the travel agent is to determine the number of Deluxe Standard and Economy tour packages
to offer for this charter. In all three of these plans, each differ in terms of their seating and service for the flight, quality
of accommodation, meal plans and tour options. The following table summarizes the estimated prices of the three
packages and the corresponding expenses of the travel agent. The travel agent has hired an aircraft for a flat fee of
Rs 2,00,000 for the entire trip. The per person price and costs for the three packages are as follows:

Tour Plan Price (Rs) Hotel Meals and

Costs (Rs) Other Expenses (Rs)
Deluxe 10,000 3,000 4,750
Standard 7,000 2,200 2,500
Economy 6,500 1,900 2,200

In planning the trip, the following considerations have come up that need to be taken into account:
(i) At least 10 per cent of the packages must be of the deluxe type.
(ii) At least 35 per cent but not more than 70 per cent must be of the standard type.
(iii) At least 30 per cent must be of the economy type.
(iv) The maximum number of deluxe packages available in any aircraft is restricted to 60.
(v) The hotel desires that at least 120 of the tourists should be on the deluxe standard packages together.
Mr. Raman wishes to determine the number of packages to offer in each type of trip so as to maximize the total
profit.
 Linear Programming: The Graphical Method 99

Case 3.2: Jain Clothing Stores

Mr. Vipin Jain, owner of Clothing Stores, is planning annual sale of shirts and pants. Mr Jain is planning to use two
different forms of advertising, viz., radio and newspaper ads, in order to promote the sale. Based on past experience,
Mr Jain feels confident that each newspaper advertisement will reach 40,000 shirt customers and 80,000 pant customers.
Each radio advertisement, he believes, will reach 30,000 shirt customers and 20,000 pant customers. The cost of each
newspaper advertisement is Rs 30,000 and the cost of each radio spot is Rs 45,000. The advertisingagency will
prepare the advertisements and it will require 5 man-hours of preparation for each newspaper advertisement and 15 man-
hours of preparation for each radio spot. Mr Jain’s sales manager says that a minimum of 75 man-hours should be spent
on preparation of advertising in order to fully utilize the services of advertising agency. Mr Jain feels that in order to have
a successful sale, the advertising must reach at least 3,60,000 shirt customers and at least 4,00,000 pant customers.
You as newly appointed management trainee are expected to suggest the media planning for advertisement at a
minimum cost and still attain the objective of Mr. Jain.
 C h a p t e r 4
“Checking the results of a decision against its expectations shows executives what their
strengths are, where they need to improve, and where they lack knowledge or information.”
– Peter Drucker

The purpose of this chapter is to help you gain an understanding of how the simplex method works.
Understanding the underlying principles help to interpret and analyze solution of any LP problem.

LEARNING OBJECTIVES

After studying this chapter, you should be able to

 understand the meaning of the word ‘simplex’ and logic of using simplex method.
 convert an LP problem into its standard form by adding slack, surplus and/or artificial variables.
 set-up simplex tables and solve LP problems using the simplex algorithm.
 interpret the optimal solution of LP problems.
 recognize special cases such as degeneracy, multiple optimal solution, unbounded and infeasible
solutions.

CHAPTER OUTLINE
4.1 Introduction 4.6 Types of Linear Programming Solutions
4.2 Standard Form of an LP Problem  Conceptual Questions
4.3 Simplex Algorithm (Maximization Case)  Self Practice Problems B
4.4 Simplex Algorithm (Minimization Case)  Hints and Answers
 Self Practice Problems A  Chapter Summary
 Hints and Answers  Chapter Concepts Quiz
4.5 Some Complications and their Resolution  Case Study
 Linear Programming: The Simplex Method 101

4.1 INTRODUCTION
Most real-life problems when formulated as an LP model have more than two variables and therefore need
a more efficient method to suggest an optimal solution for such problems. In this chapter, we shall discuss
a procedure called the simplex method for solving an LP model of such problems. This method was developed
by G B Dantzig in 1947.
The simplex, an important term in mathematics, represents an object in an n-dimensional space,
connecting n + 1 points. In one dimension, the simplex represents a line segment connecting two points;
in two dimensions, it represents a triangle formed by joining three points; in three dimensions, it represents
a four-sided pyramid.
The concept of simplex method is similar to the graphical method. In the graphical method, extreme
points of the feasible solution space are examined in order to search for the optimal solution that lies at
one of these points. For LP problems with several variables, we may not be able to graph the feasible region,
but the optimal solution will still lie at an extreme point of the many-sided, multidimensional figure (called
an n-dimensional polyhedron) that represents the feasible solution space. The simplex method examines these
extreme points in a systematic manner, repeating the same set of steps of the algorithm until an optimalsolution
is found. It is for this reason that it is also called the iterative method.
Since the number of extreme points of the feasible solution space are finite, the method assures an
improvement in the value of objective function as we move from one iteration (extreme point) to another and
achieve the optimal solution in a finite number of steps. The method also indicates when an unbounded
solution is reached.

4.2 STANDARD FORM OF AN LP PROBLEM

The use of the simplex method to solve an LP problem requires that the problem be converted into its Simplex method
standard form. The standard form of the LP problem should have the following characteristics: examines corner
points of the
(i) All the constraints should be expressed as equations by adding slack or surplus and/or artificial feasible region,
variables. using matrix row
(ii) The right-hand side of each constraint should be made non-negative (if not). This is done by operations, until an
multiplying both sides of the resulting constraint by – 1. optimal solution is
found.
(iii) The objective function should be of the maximization type.
The standard form of the LP problem is expressed as:
Optimize (Max or Min) Z = c1 x1 + c2 x2 + . . . + cn xn + 0s1 + 0s2 + . . . + 0sm
subject to the linear constraints
a11 x1 + a12 x2 + . . . + a1n xn + s1 = b1
a21 x1 + a22 x2 + . . . + a2n xn + s2 = b2
. . .
. . .
. . .
am1 x1 + am2 x2 + . . . + amn xn + sm = bm
and x1, x2, . . . , xn, s1, s2, . . ., sm ≥ 0

This standard form of the LP problem can also be expressed in the compact form as follows:

n m
Optimize (Max or Min) Z =  cj xj   0si (Objective function)

j 1 i 1

subject to the linear constraints

n
 aij xj  si = bi ; i = 1, 2, . . ., m (Constraints)
j 1
and xj, si  0, for all i and j (Non-negativity conditions)
102 Operations Research: Theory and Applications

In matrix notations the standard form is expressed as:

Optimize (Max or Min) Z = cx + 0s
subject to the linear constraints
Ax + s = b, and x, s ≥ 0
where, c = (c1, c2, . . ., cn ) is the row vector, x = (x1, x2, . . ., xn )T, b = (b1, b2, . . ., bm )T and s = (s1, s2, . . ., sm ) are
column vectors, and A is the m × n matrix of coefficients, aij of variables x1, x2, . . ., xn in the constraints.
Remarks Any LP problem (maximization or minimization) may have
1. (a) no feasible solution, i.e. value of decision variables, xj ( j = 1, 2, . . ., n) may not satisfy every constraint.
(b) a unique optimum feasible solution.
(c) more than one optimum feasible solution, i.e. alternative optimum feasible solutions.
(d) a feasible solution for which the objective function is unbounded (value of the objective function can
be made as large as possible in a maximization problem or as small as possible in a minimization
problem).
2. Any minimization LP problem can be converted into an equivalent maximization problem by changing the
sign of cj’s in the objective function. That is,
n n
Minimize  cj xj = Maximize  (  cj ) xj
j 1 j 1

3. Any constraint expressed by equality (=) sign may be replaced by two weak inequalities. For example,
a11 x1 + a12 x2 + . . . + a1n xn = b1 is equivalent to following two simultaneous constraints,
Slack variable
represents an a11 x1 + a12 x2 + . . . + a1n xn  b1 and a11 x1 + a12 x2 + . . . + a1n xn  b1
unused quantity of
resource; it is 4. Three types of additional variables, namely (i) slack variables (s) (ii) surplus variables (– s), and
added to less-than (iii) artificial variables (A) are added in the given LP problem to convert it into the standard form for the
or equal-to type following reasons:
constraints in order
to get an equality (a) These variables allow us to convert inequalities into equalities, thereby converting the given LP
constraint. problem into its standard form. Such form will help in getting solution of the LP problem.
(b) These variables help decision-makers to draw economic interpretation from the final solution.
(c) These variables are used to get an initial feasible solution represented by the columns of the identity
matrix.
The summary of the additional variables to be added in the given LP problem in order to convert it
into a standard form is given in Table 4.1.

Types of Constraint Extra Variable Coefficient of Additional Presence of Additional

Needed Variables in the Variables in the
Objective Function Initial Solution

Max Z Min Z
 Less than or A slack variable 0 0 Yes
equal to () is added
 Greater than or A surplus variable 0 0 No
Table 4.1 equal to () is subtracted, and
Summary of an artificial variable –M +M Yes
Additional is added
Variables Added  Equal to (=) Only an artificial –M +M Yes
in an LP Problem variable is added

Remark A slack variable represents an unused resource, either in the form of time on a machine, labour
hours, money, warehouse space, etc. Since these variables don’t yield any profit, therefore such variables are
added to the objective function with zero coefficients.
A surplus variable represents the amount by which solution values exceed a resource. These variables
are also called negative slack variables. Surplus variables, like slack variables carry a zero coefficient in
the objective function.
 Linear Programming: The Simplex Method 103

Definitions
Basic solution Given a system of m simultaneous linear equations with n (> m) variables: Ax = B, where Ais
an m × n matrix and rank (A) = m. Let B be any m × m non-singular submatrix of A obtained by reordering m
linearly independent columns of A. Then, a solution obtained by setting n – m variables not associated with the
columns of B, equal to zero, and solving the resulting system is called a basic solution to the given system of
equations.
The m variables (may be all non zero) are called basic variables. The m × m non-singular sub-matrix B is
called a basis matrix and the columns of B as basis vectors.
If B is the basis matrix, then the basic solution to the system of equations will be xB = B–1b.
Basic feasible solution A basic solution to the system of simultaneous equations, Ax = b is called basic
feasible if xB  0.
Degenerate solution A basic solution to the system of simultaneous equations, Ax = b is called degenerate
if one or more of the basic variables assume zero value.
Cost vector Let xB be a basic feasible solution to the LP problem.
Maximize Z = cx
subject to the constraints
Ax = b, and x  0.
Then the vector cB = (cB1, cB2, . . ., cBm), is called cost vector that represents the coefficient of basic variable,
xB in the basic feasible solution.

4.3 SIMPLEX ALGORITHM (MAXIMIZATION CASE)

The steps of the simplex algorithm for obtaining an optimal solution (if it exists) to a linear programming
problem are as follows:
Step 1: Formulation of the mathematical model
(i) Formulate the LP model of the given problem. Surplus variable
(ii) If the objective function is of minimization, then convert it into equivalent maximization, by using represents the
the following relationship amount of resource
usage above the
Minimize Z = – Maximize Z *, where Z * = – Z. minimum required
and is added to
(iii) Check whether all the bi (i = 1, 2, . . . , m) values are positive. If any one of them is negative, then greater-than or
multiply the corresponding constraint by – 1 in order to make bi > 0. In doing so, remember to equal-to constraints
change a  type constraint to a  type constraint, and vice versa. in order to get
(iv) Express the given LP problem in the standard form by adding additional variables in constraints equality constraint.
(as per requirement) and assign a zero-cost coefficient to these variables in the objective function.
(v) Replace each unrestricted variable (if any) with the difference of the two non-negative variables.
Step 2: Set-up the initial solution Write down the coefficients of all the variables in the LP problem
in a tabular form, as shown in Table 4.2, in order to get an initial basic feasible solution [xB = B–1 b].

cj  c1 c2 ... cn 0 0 ... 0
Basic Variables Basic Basic Variables Variables
Coefficient Variables Value x1 x2 ... xn s1 s2 . . . sm
(cB) B b (= xB)
cB1 s1 xB1 = b1 a11 a12 ... a1n 1 0 ... 0
cB2 s2 xB2 = b2 a21 a22 ... a2n 0 1 ... 0
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
cBm sm xBm = bm am1 am2 ... amn 0 0 ... 1
Table 4.2
Z =  cBi xBi zj =  cBi xj 0 0 ... 0 0 0 ... 0 Initial Simplex
cj – zj c1 – z1 c2 – z2 ... cn – zn 0 0 ... 0 Table
104 Operations Research: Theory and Applications

After having set up the initial simplex table, locate the identity matrix (contains all zeros except
positive elements 1’s on the diagonal). The identity matrix so obtained is also called a basis matrix
[because basic feasible solution is represented by B = I ].
The columns of identity matrix represent the coefficients of slack variables that have been added to
the constraints. Each column of the identity matrix also represents a basic variable.
Assign values of the constants (bi’s) to the column variables in the identity matrix [because
Degeneracy arises xB = B –1 b = I b = b].
when there is a tie The variables corresponding to the columns of the identity matrix are called basic variables and the
in the minimum ratio remaining ones are called non-basic variables. In general, if an LP model has n variables and m (< n)
value that
constraints, then m variables would be basic and n – m variables non-basic. In certain cases one or more
determines the
variable to enter than one basic variables may also have zero values. It one or more basic variables have zero value, then
into the next this situation is called degeneracy and will be discussed later.
solution. The first row in Table 4.2 indicates the coefficients cj of variables in the objective function. These values
represent the cost (or profit) per unit associated with a variable in the objective function and these
are used to determine the variable to be entered into the basis matrix B.
The column ‘cB’ lists the coefficients of the current basic variables in the objective function. These values
are used to calculate the value of Z when one unit of any variable is brought into the solution.
Column ‘xB’ represents the values of the basic variables in the current basic solution.
Basis is the set of Numbers, aij in the columns under each variable are also called substitution rates (or exchange
variables present in coefficients) because these represent the rate at which resource i (i = 1, 2, . . ., m) is consumed by each unit
the solution, have
positive non-zero of an activity j ( j = 1, 2, . . ., n).
value; these The values zj represent the amount by which the value of objective function Z would be decreased
variables are also (or increased ) if one unit of the given variable is added to the new solution. Each of the values in the
called basic
variables.
cj – zj row represents the net amount of increase (or decrease) in the objective function that would occur
when one unit of the variable represented by the column head is introduced into the solution. That is:
cj – zj (net effect) = cj (incoming unit profit/cost) – zj (outgoing total profit/cost) where
zj = Coefficient of basic variables column × Exchange coefficient column j

Step 3: Test for optimality Calculate the cj – zj value for all non-basic variables. To obtain the value
of zj multiply each element under ‘Variables’ column (columns, aj of the coefficient matrix) with the
corresponding elements in cB-column. Examine values of cj – zj. The following three cases may arise:
(i) If all cj – zj  0, then the basic feasible solution is optimal.
(ii) If at least one column of the coefficients matrix (i.e. ak ) for which ck – zk > 0 and all other elements
are negative (i.e. aik < 0), then there exists an unbounded solution to the given problem.
(iii) If at least one cj – zj > 0, and each of these columns have at least one positive element (i.e. aij )
for some row, then this indicates that an improvement in the value of objective function Z is
possible.

Non-basic Step 4: Select the variable to enter the basis If Case (iii) of Step 3 holds, then select a variable that
variables are those has the largest cj – zj value to enter into the new solution. That is,
which are not in the
‘basis’ and have ck – zk = Max {(cj – zj); cj – zj > 0}
zero-value.
The column to be entered is called the key or pivot column. Obviously, such a variable indicates the
largest per unit improvement in the current solution.
Step 5: Test for feasibility (variable to leave the basis) After identifying the variable to become the
basic variable, the variable to be removed from the existing set of basic variables is determined. For this,
each number in xB-column (i.e. bi values) is divided by the corresponding (but positive) number in
the key column and a row is selected for which this ratio is non-negative and minimum. This ratio is called
the replacement (exchange) ratio. That is,
R| x U
xBr
a = Min S| a
Bi
; a rj  0 V||

rj T rj W
 Linear Programming: The Simplex Method 105

This ratio restricts the number of units of the incoming variable that can be obtained from the exchange. Infeasibility is the
It may be noted that the division by a negative or by a zero element in key column is not permitted. situation in which no
The row selected in this manner is called the key or pivot row and it represents the variable which solution satisfies all
constraints of an LP
will leave the solution. problem.
The element that lies at the intersection of the key row and key column of the simplex table is called
key or pivot element.

Step 6: Finding the new solution

(i) If the key element is 1, then the row remains the same in the new simplex table.
(ii) If the key element is other than 1, then divide each element in the key row (including the elements
in xB-column ) by the key element, to find the new values for that row.
(iii) The new values of the elements in the remaining rows of the new simplex table can be obtained
by performing elementary row operations on all rows so that all elements except the key element
in the key column are zero. In other words, for each row other than the key row, we use the formula:

Number in F Number inI LMF Number above or belowI F Corresponding number inI OP
H replaced in Step 6 (ii)
= G old row JK  MGH
JK × G the new row, that is row JK P
H Q
key element
new row
N
The new entries in cB and xB columns are updated in the new simplex table of the current solution.
Step 7: Repeat the procedure Go to Step 3 and repeat the procedure until all entries in the cj – zj row values
cj – zj row are either negative or zero. represent net profit
or loss resulting
Remark The flow chart of the simplex algorithm for both the maximization and the minimization LP from introducing
problem is shown in Fig. 4.1. one unit of any
variable into the
Example 4.1 Use the simplex method to solve the following LP problem. ‘basis’ or solution
mix.
Maximize Z = 3x1 + 5x2 + 4x3
subject to the constraints
(i) 2x1 + 3x2  8, (ii) 2x2 + 5x3  10, (iii) 3x1 + 2x2 + 4x3  15
and x1, x2, x3  0

[Rewa MSc (Maths), 1995; Meerut MSc (Stat.), 1996; MSc (Maths), 1998;
Kurukshetra, MSc (Stat.), 1998]
Solution Step 1: Introducing non-negative slack variables s1, s2 and s3 to convert the given LP problem
into its standard form:
Maximize Z = 3x1 + 5x2 + 4x3 + 0s1 + 0s2 + 0s3
subject to the constraints
(i) 2x1 + 3x2 + s1 = 8, (ii) 2x2 + 5x3 + s2 = 10, (iii) 3x1 + 2x2 + 4x3 + s3 = 15
and x1, x2, x3, s1, s2, s3  0
Step 2: Since all bi (RHS values) > 0, (i = 1, 2, 3), therefore choose the initial basic feasible solution as:
x1 = x2 = x3 = 0; s1 = 8, s2 = 10, s3 = 15 and Max Z = 0
This solution can also be read from the initial simplex Table 4.3 by equating row-wise values in the basis
(B) column and solution values (xB ) column.
Step 3: To see whether the current solution given in Table 4.3 is optimal or not, calculate
cj – zj = cj – cB B –1 aj = cj – cB yj
for non-basic variables x1, x2 and x3 as follows.
zj = (Basic variable coefficients, cB ) × ( jth column of data matrix)
That is, z1 = 0 (2) + 0 (0) + 0 (3) = 0 for x1-column
z2 = 0 (3) + 0 (2) + 0 (2) = 0 for x2-column
z3 = 0 (0) + 0 (5) + 0 (4) = 0 for x3-column
106 Operations Research: Theory and Applications

These zj values are now subtracted from cj values in order to calculate the net profit likely to be gained
by introducing (performing) a particular variable (activity) x1, x2 or x3 into the new solution mix. This is done
by introducing one unit of each variable x1, x2 and x3 into the new solution mix.
c1 – z1 = 3 – 0 = 3, c2 – z2 = 5 – 0 = 5, c3 – z3 = 4 – 0 = 4
The zj and cj – zj rows are added into the Table 4.3.
The values of basic variables, s1, s2 and s3 are given in the solution values (xB ) column of Table 4.3. The
remaining variables that are non-basic at the current solution have zero value. The value of objective
function at the current solution is given by
Z = (Basic variables coefficient, cB ) × (Basic variables value, xB )
= 0 (8) + 0 (10) + 0 (15) = 0

cj  3 5 4 0 0 0
Basic Variables Basic Basic Variables x1 x2 x3 s1 s2 s3 Min Ratio
Coefficient Variables Value xB/x2
cB B b (= xB )
0 s1 8 2 3 0 1 0 0 8/3 
0 s2 10 0 2 5 0 1 0 10/2
0 s3 15 3 2 4 0 0 1 15/2

Z = 0 zj 0 0 0 0 0 0
Table 4.3 cj – zj 3 5 4 0 0 0
Initial Solution 

Since all cj – zj  0 ( j = 1, 2, 3), the current solution is not optimal. Variable x2 is chosen to enter into the
basis because c2 – z2 = 5 is the largest positive number in the x2-column, where all elements are positive.This
means that for every unit of variable x2, the objective function will increase in value by 5. The x2- column is
the key column.

Step 4: The variable that is to leave the basis is determined by dividing the values in the xB-column by
the corresponding elements in the key column as shown in Table 4.3. Since the ratio, 8/3
is minimum in row 1, the basic variable s1 is chosen to leave the solution (basis).
Step 5 (Iteration 1): Since the key element enclosed in the circle in Table 4.3 is not 1, divide all elements
of the key row by 3 in order to obtain new values of the elements in this row. The new values of the elements
in the remaining rows for the new Table 4.4 can be obtained by performing the following elementary row
operations on all rows so that all elements except the key element 1 in the key column are zero.

R1 (new)  R1 (old)  3 (key element)

 (8/3, 2/3, 3/3, 0/3, 1/3, 0/3, 0/3) = (8/3, 2/3, 1, 0, 1/3, 0, 0)

R2 (new)  R2 (old) – 2R1 (new) R3 (new)  R3 (old) – 2R1 (new)

10 – 2 × 8/3 = 14/3 15 – 2 × 8/3 = 29/3
0 – 2 × 2/3 = – 4/3 3 – 2 × 2/3 = 5/3
2 – 2 × 1= 0 2 – 2 × 1= 0
5 – 2 × 0= 5 4 – 2 × 0= 4
0 – 2 × 1/3 = – 2/3 0 – 2 × 1/3 = – 2/3
1 – 2 × 0= 1 0 – 2 × 0= 0
0 – 2 × 0= 0 1 – 2 × 0= 1
 Linear Programming: The Simplex Method 107

cj  3 5 4 0 0 0

Basics Variables Basic Basic Variables x1 x2 x3 s1 s2 s3 Min Ratio

Coefficient Variables Value xB/x3
cB B b (= xB)

5 x2 8/3 2/3 1 0 1/3 0 0 –

0 s2 14/3 – 4/3 0 5 – 2/3 1 0 (14/3)/5 
0 s3 29/3 5/3 0 4 – 2/3 0 1 (29/3)/4

Z = 40/3 zj 10/3 5 0 5/3 0 0

cj – zj – 1/3 0 4 – 5/3 0 0 Table 4.4
 Improved Solution

The improved basic feasible solution can be read from Table 4.4 as: x2 = 8/3, s2 = 14/3, s3 = 29/3 and
x1 = x3 = s1 = 0. The improved value of the objective function becomes:
Z = (Basic variable coefficients, cB ) × (Basic variable values, xB )
= 5 (8/3) + 0 (14/3) + 0 (29/3) = 40/3
Once again, calculate values of cj – zj to check whether the solution shown in Table 4.4 is optimal or
not. Since c3 – z3 > 0, the current solution is not optimal.
Iteration 2: Repeat Steps 3 to 5. Table 4.5 is obtained by performing following row operations to enter
variable x3 into the basis and to drive out s2 from the basis.
R2 (new) = R2 (old)  5 (key element) = (14/15, – 4/15, 0, 1, – 2/15, 1/5, 0)

R3 (new)  R3 (old) – 4R2 (new)

29/3 – 4 × 14/15 = 89/15

5/3 – 4 × – 4/15 = 41/15
0 – 4× 0= 0
4 – 4× 1= 0
– 2/3 – 4 × – 2/15 = – 2/15
0 – 4× 1/5 = – 4/5
1 – 4× 0= 0

Table 4.5 is completed by calculating the new zj and cj – zj values and the new value of objective
function:
z1 = 5 (2/3) + 4 (– 4/15) + 0 (41/15) = 34/15 c1 – z1 = 3 – 34/15 = 11/15 for x1 -column
z4 = 5 (1/3) + 4 (– 2/15) + 0 (2/15) = 17/15 c4 – z4 = 0 – 17/15 = – 17/15 for s1-column
z5 = 5 (0) + 4 (1/5) + 0 (– 4/5) = 4/5 c5 – z5 = 0 – 4/5 = – 4/5 for s2-column
The new objective function value is:
Z = (Basic variables coefficient, cB ) × (Basic variables value, xB )
= 5 (8/3) + 4 (14/15) + 0 (89/15) = 256/15
The improved basic feasible solution is shown in Table 4.5.

cj  3 5 4 0 0 0

Basic Variables Basic Basic Variables x1 x2 x3 s1 s2 s3 Min Ratio

Coefficient Variables Value xB /x1
cB B b (= xB)

5 x2 8/3 2/3 1 0 1/3 0 0 (8/3)/(2/3) = 4

4 x3 14/15 – 4/15 0 1 – 2/15 1/5 0 –
0 s3 89/15 41/15 0 0 2/15 – 4/5 1 (89/15)/(41/15) = 2.17 

Z = 256/15 zj 34/15 5 4 17/15 4/5 0

cj – zj 11/15 0 0 – 17/15 – 4/5 0
Table 4.5
 Improved Solution
108 Operations Research: Theory and Applications

Iteration 3: In Table 4.5, since c1 – z1 is still a positive value, the current solution is not optimal. Thus,
the variable x1 enters the basis and s3 leaves the basis. To get another improved solution as shown in Table
4.5 perform the following row operations in the same manner as discussed earlier.
R3 (new)  R3 (old) × 15/41 (key element)
 (89/15 × 15/41, 41/15 × 15/41, 0 × 15/41, 0 × 15/41,
– 2/15 × 15/41, – 4/5 × 15/41, 1 × 15/41)
 (89/41, 1, 0, 0, – 2/41, – 12/41, 15/41)

R1 (new)  R1 (old) – (2/3) R3 (new) R2 (new)  R2 (old)+ (4/15) R3 (new)

8/3 – 2/3 × 89/3 = 50/41 14/15 + 4/15 × 89/41 = 62/41

2/3 – 2/3 × 1 = 0 – 4/15 + 4/15 × 1 = 0
1 – 2/3 × 0 = 1 0 + 4/15 × 0 = 0
0 – 2/3 × 0 = 0 1 + 4/15 × 0 = 1
1/3 – 2/3 × – 2/41 = 15/41 – 2/15 + 4/15 × – 2/41 = – 6/41
0 – 2/3 × – 2/41 = 8/41 1/5 + 4/15 × – 12/41 = 5/41
0 – 2/3 × 15/41 = – 10/41 0 + 4/15 × 15/41 = 4/41

cj  3 5 4 0 0 0
Basic Variables Basic Basic Variables x1 x2 x3 s1 s2 s3
Coefficient Variables Value
cB B b (= xB)

5 x2 50/41 0 1 0 15/41 8/41 – 10/41

4 x3 62/41 0 0 1 – 6/41 5/41 4/41
3 x1 89/41 1 0 0 – 2/41 – 12/41 15/41

Table 4.6 Z = 765/41 zj 3 5 4 45/41 24/41 11/41

Optimal Solution cj – zj 0 0 0 – 45/41 – 24/41 – 11/41

In Table 4.6, all cj – zj < 0 for non-basic variables. Therefore, the optimal solution is reached with,
x1 = 89/41, x2 = 50/41, x3 = 62/41 and the optimal value of Z = 765/41.
Example 4.2 A company makes two kinds of leather belts, belt A and belt B. Belt A is a high quality belt
and belt B is of lower quality. The respective profits are Rs 4 and Rs 3 per belt. The production ofeach
of type A requires twice as much time as a belt of type B, and if all belts were of type B, the company could
make 1,000 belts per day. The supply of leather is sufficient for only 800 belts per day (both A andB
combined). Belt A requires a fancy buckle and only 400 of these are available per day. There are only700
buckles a day available for belt B.
What should be the daily production of each type of belt? Formulate this problem as an LP model and
solve it using the simplex method.
Solution Let x1 and x2 be the number of belts of type A and B, respectively, manufactured each day. Then
the LP model would be as follows:
Maximize (total profit) Z = 4x1 + 3x2
subject to the constraints
(i) 2x1 + x2  1,000 (Time availability), (ii) x1 + x2  800 (Supply of leather)
(iii) x  400 
1  (Buckles availability)
x2  700 
and x1, x2  0

Standard form Introducing slack variables s1, s2, s3 and s4 to convert given LP model into its standard form
as follows.
 Linear Programming: The Simplex Method 109

Maximize Z = 4x1 + 3x2 + 0s1 + 0s2 + 0s3 + 0s4

subject to the constraints
(i) 2x1 + x2 + s1 = 1,000, (ii) x1 + x2 + s2 = 800
(iii) x1 + s3 = 400, (iv) x2 + s4 = 700
and x1, x2, s1, s2, s3, s4  0

Solution by simplex method An initial feasible solution is obtained by setting x1 = x2 = 0. Thus, the initial
solution is: s1 = 1,000, s2 = 800, s3 = 400, s4 = 700 and Max Z = 0. This solution can also be read from the
initial simplex Table 4.7.
cj  4 3 0 0 0 0
Basic Variables Basic Basic Variables x1 x2 s1 s2 s3 s4 Min Ratio
Coefficient Variables Value xB /x1
cB B b (= xB )
0 s1 1,000 2 1 1 0 0 0 1,000/2 = 500
0 s2 800 1 1 0 1 0 0 800/1 = 800
0 s3 400 1 0 0 0 1 0 400/1 = 400 
0 s4 700 0 1 0 0 0 1 not defined

Z = 0 zj 0 0 0 0 0 0
cj – zj 4 3 0 0 0 0 Table 4.7
 Initial Solution

In Table 4.7, since c1 – z1 = 4 is the largest positive number, we apply the following row operations
in order to get an improved basic feasible solution by entering variable x1 into the basis and removing variable
s3 from the basis.
R3 (new)  R3 (old)  1 (key element) R1 (new)  R1 (old) – 2R3 (new)
R2 (new)  R2 (old) – R3 (new)
The new solution is shown in Table 4.8.

cj  4 3 0 0 0 0
Basic Variables Basis Basic Variables x1 x2 s1 s2 s3 s4 Min Ratio
Coefficient Variables Value xB/x2
cB B b (= xB)
0 s1 200 0 1 1 0 – 2 0 200/1 = 200 
0 s2 400 0 1 0 1 – 1 0 400/1 = 400
4 x1 400 1 0 0 0 1 0 –
0 s4 700 0 1 0 0 0 1 700/1 = 700
Z = 1,600 zj 4 0 0 0 4 0 Table 4.8
cj – zj 0 3 0 0 – 4 0 An Improved
 Solution

The solution shown in Table 4.8 is not optimal because c2 – z2 > 0 in x2 - column. Thus, again applying
the following row operations to get a new solution by entering variable x2 into the basis and removing variable
s1 from the basis, we get
R1 (new)  R1 (old)  1 (key element) R2 (new)  R2 (old) – R1 (new)
R4 (new)  R4 (old) – R1 (new)
The improved solution is shown in Table 4.9.
110 Operations Research: Theory and Applications

cj  4 3 0 0 0 0
Basic Variables Basic Basic Variables x1 x2 s1 s2 s3 s4 Min Ratio
Coefficient Variables Value xB /s3
cB B b (= xB)
3 x2 200 0 1 1 0 –2 0 —
0 s2 200 0 0 –1 1 1 0 200/1 = 200 
4 x1 400 1 0 0 0 1 0 400/1 = 400
0 s4 500 0 0 –1 0 2 1 500/2 = 250
Z = 2,200 zj 4 3 3 0 –2 0
Table 4.9 cj – zj 0 0 –3 0 2 0
Improved Solution 

The solution shown in Table 4.9 is not optimal because c5 – z5 > 0 in s3-column. Thus, again applying
the following row operations to get a new solution by entering variable s3, into the basis and removing
variable s2 from the basis, we get
R2 (new)  R2 (old)  1 (key element) R1 (new)  R1 (old) + 2R2 (new)
R3 (new)  R3 (old) – R2 (new); R4 (new)  R4 (old) – 2R2 (new)
The new improved solution is shown in Table 4.10.

cj  4 3 0 0 0 0
Basic Variables Basic Basic Variables x1 x2 s1 s2 s3 s4
Coefficient Variables Value
cB B b (= xB)
3 x2 600 0 1 –1 2 0 0
0 s3 200 0 0 –1 1 1 0
4 x1 200 1 0 1 – 1 0 0
0 s4 100 0 0 1 – 2 0 1
Table 4.10 Z = 2,600 zj 4 3 1 2 0 0
Optimal Solution cj – zj 0 0 – 1 – 2 0 0

Since all cj – zj < 0 correspond to non-basic variables columns, the current basic feasible solution is also
the optimal solution. Thus, the company must manufacture, x1 = 200 belts of type A and x2 = 600 belts of
type B in order to obtain the maximum profit of Rs 2,600.
Example 4.3 A pharmaceutical company has 100 kg of A, 180 kg of B and 120 kg of C ingredients available
per month. The company can use these materials to make three basic pharmaceutical products namely 5-10-
5, 5-5-10 and 20-5-10, where the numbers in each case represent the percentage of weight of A, B and C,
respectively, in each of the products. The cost of these raw materials is as follows:

Ingredient Cost per kg (Rs)

A 80
B 20
C 50
Inert ingredients 20

The selling prices of these products are Rs 40.5, Rs 43 and 45 per kg , respectively. There is a capacity
restriction of the company for product 5-10-5, because of which the company cannot produce more than 30 kg
per month. Determine how much of each of the products the company should produce in order to maximize its
monthly profit. [Delhi Univ., MBA, 2004, AMIE, 2005]
Solution Let the P1, P2 and P3 be the three products to be manufactured. The data of the problem can then
be summarized as follows:
 Linear Programming: The Simplex Method 111

Product Product Ingredients

Inert
A B C
P1 5% 10% 5% 80%
P2 5% 5% 10% 80%
P3 20% 5% 10% 65%
Cost per kg (Rs) 80 20 50 20

Cost of P1 = 5% × 80 + 10% × 20 + 5% × 50 + 80% × 20 = 4 + 2 + 2.50 + 16 = Rs 24.50 per kg

Cost of P2 = 5% × 80 + 5% × 20 + 10% × 50 + 80% × 20 = 4 + 1 + 5 + 16 = Rs 26 per kg
Cost of P3 = 20% × 80 + 5% × 20 + 10% × 50 + 65% × 20 = 16 + 1 + 5 + 13 = Rs 35 per kg
Let x1, x2 and x3 be the quantity (in kg) of P1, P2 and P3, respectively to be manufactured. The LP
problem can then be formulated as
Maximize (net profit) Z = (Selling price – Cost price) × (Quantity of product)
= (40.50 – 24.50) x1 + (43 – 26) x2 + (45 – 35) x3 = 16x1 + 17x2 + 10x3
subject to the constraints
1
x2  x3  100 or x1 + x2 + 4x3  2,000
1 1
x1 
20 20 5

1 1 1
x1  x2  x  180 or 2x1 + x2 + x3  3,600
10 20 20 3

1 1 1
x  x  x  120 or x1 + 2x2 + 2x3  2,400
20 1 10 2 10 3

x1  30
and x1, x2, x3  0.
Standard form Introducing slack variables s1, s2 and s3 to convert the given LP model into its standard
form as follows:
Maximize Z = 16x1 + 17x2 + 10x3 + 0s1 + 0s2 + 0s3 + 0s4
subject to the constraints
(i) x1 + x2 + 4x3 + s1 = 2,000, (ii) 2x1 + x2 + x3 + s2 = 3,600
(iii) x1 + 2x2 + 2x3 + s3 = 2,400, (iv) x1 + s4 = 30
and x1, x2, x3, s1, s2, s3, s4  0
Solution by simplex method An initial basic feasible solution is obtained by setting x1 = x2 = x3 = 0. Thus,
the initial solution shown in Table 4.11 is: s1 = 2,000, s2 = 3,600, s3 = 2,400, s4 = 30 and Max Z = 0.

cj  16 17 10 0 0 0 0
Basic Variables Basis Basic Variables x1 x2 x3 s1 s2 s3 s4 Min Ratio
Coefficient Variables Value xB /x2
cB B b (= xB)
0 s1 2,000 1 1 4 1 0 0 0 2,000/1 = 2,000
0 s2 3,600 2 1 1 0 1 0 0 3,600/1 = 3,600
0 s3 2,400 1 2 2 0 0 1 0 2,400/2 = 1,200 
0 s4 30 1 0 0 0 0 0 1 —
Z = 0 zj 0 0 0 0 0 0 0
cj – zj 16 17 10 0 0 0 0 Table 4.11
 Initial Solution

Since c2 – z2 = 17 in x2-column is the largest positive value, we apply the following row operations
in order to get a new improved solution by entering variable x2 into the basis and removing variable s3
from the basis.
R3 (new)  R3 (old)  2 (key element) R1 (new)  R1 (old) – R3 (new)
R2 (new)  R2 (old) – R3 (new)
112 Operations Research: Theory and Applications

The new solution is shown in Table 4.12.

cj  16 17 10 0 0 0 0
Basic Variables Basic Basic Variables x1 x2 x3 s1 s2 s3 s4 Min Ratio
Coefficient Variables Value xB /x1
cB B b (= xB )
0 s1 800 1/2 0 3 1 0 – 1/2 0 800/(1/2) = 1,600
0 s2 2,400 3/2 0 0 0 1 – 1/2 0 2,400/(3/2) = 1,600
17 x2 1,200 1/2 1 1 0 0 1/2 0 1,200/(1/2) = 2,400
0 s4 30 1 0 0 0 0 0 1 30/1 = 30 
Z = 20,400 zj 17/2 17 17 0 0 17/2 0
Table 4.12 cj – zj 15/2 0 – 7 0 0 – 17/2 0
Improve Solution


The solution shown in Table 4.12 is not optimal because c1 – z1 > 0 in x1- column. Thus, applying the
following row operations to get a new improved solution by entering variable x1 into the basis and removing
the variable s4 from the basis, we get
R4 (new)  R4 (old)  1 (key element); R1 (new)  R1 (old) – (1/2) R4 (new)
R2 (new)  R2 (old) – (3/2) R4 (new); R3 (new)  R3 (old) – (1/2) R4 (new) The
new solution is shown in Table 4.13.
cj  16 17 10 0 0 0 0

Basic Variables Basic Basic Variables x1 x2 x3 s1 s2 s3 s4

Coefficient Variables Value
cB B b (= xB )
0 s1 785 0 0 3 1 0 – 1/2 – 1/2
0 s2 2,355 0 0 0 0 1 – 1/2 – 3/2
17 x2 1,185 0 1 1 0 0 1/2 – 1/2
16 x1 30 1 0 0 0 0 0 1

Table 4.13 Z = 20,625 zj 16 17 17 0 0 17/2 15/2

Optimal Solution cj – zj 0 0 –7 0 0 – 17/2 – 15/2

Since all cj – zj < 0 corresponding to non-basic variables columns, the current solution is an optimal
solution. Thus, the company must manufacture, x1 = 30 kg of P1, x2 = 1,185 kg of P2 and x3 = 0 kg of
P3 in order to obtain the maximum net profit of Rs 20,625.

4.4 SIMPLEX ALGORITHM (MINIMIZATION CASE)

In certain cases, it is difficult to obtain an initial basic feasible solution of the given LP problem. Such cases
arise
(i) when the constraints are of the  type,
n
 aij x j  bi , xj  0
j 1

and value of few right-hand side constants is negative [i.e. bi < 0]. After adding the non-negative slack
variable si (i = 1, 2, . . ., m), the initial solution so obtained will be si = – bi for a particular resource, i.
This solution is not feasible because it does not satisfy non-negativity conditions ofslack variables
(i.e. si  0).
(ii) when the constraints are of the  type,
n
 aij x j  bi , xj  0
j 1
 Linear Programming: The Simplex Method 113

After adding surplus (negative slack) variable si, the initial solution so obtained will be – si = bi or
si = – bi
n
 aij x j  si = bi , xj  0, si  0
j 1

This solution is not feasible because it does not satisfy non-negativity conditions of surplus variables (i.e. si
 0). In such a case, artificial variables, Ai (i = 1, 2, . . ., m) are added to get an initial basic feasible solution. The
resulting system of equations then becomes:
n
 aij x j  si  Ai = bi
j 1

xj, si, Ai  0, i = 1, 2, . . ., m
These are m simultaneous equations with (n + m + m) variables (n decision variables, m artificial variables
and m surplus variables). An initial basic feasible solution of LP problem with such constraints can be
obtained by equating (n + 2m – m) = (n + m) variables equal to zero. Thus the new solution to the given
LP problem is: Ai = bi (i = 1, 2, . . . , m), which is not the solution to the original system of equations because
the two systems of equations are not equivalent. Thus, to get back to the original problem, artificial
variables must be removed from the optimal solution. There are two methods for removing artificial variables
from the solution.
 Two-Phase Method
 Big-M Method or Method of Penalties
The simplex method, both for the minimization and the maximization LP problem, may be summarized through
a flow chart shown in Fig. 4.1.

Fig. 4.1
Flow Chart of
Simplex Algorithm
114 Operations Research: Theory and Applications

Remark Artificial variables have no meaning in a physical sense and are only used as a tool forgenerating
an initial solution to an LP problem. Before the optimal solution is reached, all artificial variablesmust be
dropped out from the solution mix. This is done by assigning appropriate coefficients to these variables in the
objective function. These variables are added to those constraints with equality (=) and greater than or equal
to () sign.

4.4.1 Two-Phase Method

In the first phase of this method, the sum of the artificial variables is minimized subject to the given constraints
in order to get a basic feasible solution of the LP problem. The second phase minimizes the original objective
function starting with the basic feasible solution obtained at the end of the first phase. Since the solution of
the LP problem is completed in two phases, this method is called the two-phase method.

Advantages of the method

1. No assumptions on the original system of constraints are made, i.e. the system may be redundant,
inconsistent or not solvable in non-negative numbers.
2. It is easy to obtain an initial basic feasible solution for Phase I.
3. The basic feasible solution (if it exists) obtained at the end of phase I is used as initial solution for
Phase II.

Steps of the Algorithm: Phase I

An artificial
variable is added to Step 1 (a): If all the constraints in the given LP problem are ‘less than or equal to’ () type, then
the constraints to
get an initial solution Phase II can be directly used to solve the problem. Otherwise, the necessary number of surplus and artificial
to an LP problem. variables are added to convert constraints into equality constraints.
(b) If the given LP problem is of minimization, then convert it to the maximization type by the usual method.
Step 2: Assign zero coefficient to each of the decision variables (xj) and to the surplus variables; and assign
– 1 coefficient to each of the artificial variables. This yields the following auxiliary LP problem.
m
Maximize Z* =  ( 1) Ai
i 1

subject to the constraints

n
 aij x j  Ai  bi , i = 1, 2, . . ., m
i 1

and xj, Aj  0

Step 3: Apply the simplex algorithm to solve this auxiliary LP problem. The following three cases may arise
at optimality.
(a) Max Z * = 0 and at least one artificial variable is present in the basis with positive value. This means
that no feasible solution exists for the original LP problem.
(b) Max Z * = 0 and no artificial variable is present in the basis. This means that only decision variables
(xj’s) are present in the basis and hence proceed to Phase II to obtain an optimal basic feasible
solution on the original LP problem.
(c) Max Z * = 0 and at least one artificial variable is present in the basis at zero value. This means
that a feasible solution to the auxiliary LP problem is also a feasible solution to the original LP
problem. In order to arrive at the basic feasible solution, proceed directly to Phase II or else eliminate
the artificial basic variable and then proceed to Phase II.
Remark Once an artificial variable has left the basis, it has served its purpose and can, therefore, be removed
from the simplex table. An artificial variable is never considered for re-entry into the basis.
Phase II: Assign actual coefficients to the variables in the objective function and zero coefficient to the
artificial variables which appear at zero value in the basis at the end of Phase I. The last simplex table of
 Linear Programming: The Simplex Method 115

Phase I can be used as the initial simplex table for Phase II. Then apply the usual simplex algorithm to the
modified simplex table in order to get the optimal solution to the original problem. Artificial variables that
do not appear in the basis may be removed.
Example 4.4 Use two-phase simplex method to solve the following LP problem:
Minimize Z = x1 + x2
subject to the constraints
(i) 2x1 + x2  4, (ii) x1 + 7x2  7
and x1, x2  0
Solution Converting the given LP problem objective function into the maximization form and then adding
surplus variables s1 and s2 and artificial variables A1 and A2 in the constraints, the problem becomes:
Maximize Z * = – x1 – x2
subject to the constraints
(i) 2x1 + x2 – s1 + A1 = 4, (ii) x1 + 7x2 – s2 + A2 = 7
and x1, x2, s1, s2, A1, A2  0
where Z * = – Z
Phase I: This phase starts by considering the following auxiliary LP problem:
Maximize Z * = – A1 – A2
subject to the constraints
(i) 2x1 + x2 – s1 + A1 = 4, (ii) x1 + 7x2 – s2 + A2 = 7
and x1, x2, s1, s2, A1, A2  0
The initial solution is presented in Table 4.14.

cj  0 0 0 0 – 1 1
Basic Variables Basic Basic Variables x1 x2 s1 s2 A1 A2
Coefficient Variables Value
cB B b(= xB )

–1 A1 4 2 1 –1 0 1 0
–1 A2 7 1 7 0 –1 0 1 

Z* = – 11 zj –3 –8 1 1 –1 –1
cj – zj 3 8 –1 –1 0 0 Table 4.14
 Initial Solution

Artificial variables A1 and A2 are now removed, one after the other, maintaining the feasibility of the solution.
Iteration 1: Applying the following row operations to get an improved solution by entering variable x2
in the basis and first removing variable A2 from the basis. The improved solution is shown in Table 4.15.
Note that the variable x1 cannot be entered into the basis as this would lead to an infeasible solution.
R2 (new)  R2 (old)  7 (key element); R1 (new)  R1 (old) – R2 (new)

cj  0 0 0 0 – 1 1
Basic Variables Basic Basic Variables x1 x2 s1 s2 A1 A2*
Coefficient Variables Value
cB B b ( = xB )
–1 A1 3 13/7 0 – 1 1/7 1 – 1/7 
0 x2 1 1/7 1 0 – 1/7 0 1/7
Z* = – 3 zj – 13/7 0 1 – 1/7 – 1 1/7 Table 4.15
cj – zj 13/7 0 – 1 1/7 0 – 8/7 Improved
Solution


* This column can permanently be removed at this stage.

116 Operations Research: Theory and Applications

Iteration 2: To remove A1 from the solution shown in Table 4.15, we enter variable s2 in the basis by applying
the following row operations. The new solution is shown in Table 4.16. It may be noted that if instead of s2,
variable x1 is chosen to be entered into the basis, it will lead to an infeasible solution
R1 (new)  R1 (old) × 7; R2 (new)  R2 (old) + (1/7) R1 (new)
cj  0 0 0 0 – 1 1
Basic Variables Basic Basic Variables x1 x2 s1 s2 A1* A2*
Coefficient Variables Value
cB B b( = xB )

0 s2 21 13 0 –7 1 7 –1
0 x2 4 2 1 –1 0 1 0
Table 4.16 Z* = 0 zj 0 0 0 0 0 0
Improved Solution cj – zj 0 0 0 0 –1 –1

* Remove columns A1 and A2 from Table 4.16.

Since all cj – zj  0 correspond to non-basic variables, the optimal solution: x1 = 0, x2 = 4,s1
= 0, s2 = 21, A1 = 0, A2 = 0 with Z * = 0 is arrived at. However, this solution may or may not be the basic
feasible solution to the original LP problem. Thus, we have to move to Phase II to get an optimal solution
to our original LP problem.
Phase II: The modified simplex table obtained from Table 4.16 is represented in Table 4.17.
cj  – 1 – 1 0 0
Basic Variables Basic Basic Variables x1 x2 s1 s2 Min Ratio
Coefficient Variables Value xB /x1
cB B b( = xB )

0 s2 21 13 0 –7 1 21/13 
–1 x2 4 2 1 –1 0 4/2

Z* = – 4 zj – 2 –1 1 0
cj – zj 1 0 –1 0
Table 4.17


Iteration 1: Introducing variable x1 into the basis and removing variable s2 from the basis by applying the
following row operations:
R1 (new)  R1 (old)  13 (key element); R2 (new)  R2 (old) – 2R1 (new)
The improved basic feasible solution so obtained is given in Table 4.18. Since in Table 4.18, cj – zj  0
for all non-basic variables, the current solution is optimal. Thus, the optimal basic feasible solution to the given
LP problem is: x1 = 21/13, x2 = 10/13 and Max Z* = – 31/13 or Min Z = 31/13.
cj  –1 –1 0 0
Basic Variables Basic Basic Variables x1 x2 s1 s2
Coefficient Variables Value
cB B b( = xB )

–1 x1 21/13 1 0 – 7/13 1/13

–1 x2 10/13 0 1 1/13 – 2/13

Table 4.18 Z* = – 31/13 zj –1 –1 6/13 1/13

Optimal Solution cj – zj 0 0 – 6/13 – 1/13

Example 4.5 Solve the following LP problem by using the two-phase simplex method.
Minimize Z = x1 – 2x2 – 3x3
subject to the constraints
(i) – 2x1 + x2 + 3x3 = 2, (ii) 2x1 + 3x2 + 4x3 = 1
and x1, x2, x3  0. [Meerut, MSc (Math), 2003]
 Linear Programming: The Simplex Method 117

Solution After converting the objective function into the maximization form and by adding artificial
variables A1 and A2 in the constraints, the given LP problem becomes:
Maximize Z* = – x + 2x + 3x
1 2 3
subject to the constraints
(i) – 2x1 + x2 + 3x3 + A1 = 2, (ii) 2x1 + 3x2 + 4x3 + A2 = 1
and x1, x2, x3, A1, A2  0
where Z* = – Z
Phase I: This phase starts by considering the following auxiliary LP problem:
Maximize Z* = – A1 – A2
subject to the constraints
(i) – 2x1 + x2 + 3x3 + A1 = 2 (ii) 2x1 + 3x2 + 4x3 + A2 = 1
and x1, x2, x3, A1, A2  0
The initial solution is presented in Table 4.19.
cj  0 0 0 – 1 1
Basic Variables Basic Basic Variables x1 x2 x3 A1 A2
Coefficient Variables Value
cB B b ( = xB)
–1 A1 2 –2 1 3 1 0
–1 A2 1 2 3 4 0 1 
Z* = – 3 zj 0 –4 –7 –1 –1
cj – zj 0 4 7 0 0 Table 4.19
Initial Solution


To first remove the artificial variable A2 from the solution shown in Table 4.19, introduce variable
x3 into the basis by applying the following row operations:
R1 (new)  R2 (old)  4 (key element) ; R1 (new)  R1 (old) – 3R2 (new)
The improved solution so obtained is given in Table 4.20. Since in Table 4.20, cj – zj  0 corresponds
to non-basic variables, the optimal solution is: x1 = 0, x2 = 0, x3 = 1/4, A1 = 5/4 and A2 = 0 withMax
Z * = – 5/4. But at the same time, the value of Z * < 0 and the artificial variable A1 appears in the basis with
positive value 5/4. Hence, feasible solution to the given original LP problem does not exist.

cj  0 0 0 1
Basic Variables Basic Basic Variables x1 x2 x3 A1
Coefficient Variables Value
cB B b( = xB )
–1 A1 5/4 –7/2 –5/4 0 1
0 x3 1/4 1/2 3/4 1 0
Table 4.20
Z * = – 5/4 zj 7/2 5/4 0 –1 Optimal but not
cj – zj – 7/2 – 5/4 0 0 Feasible Solution

Example 4.6 Use two-phase simplex method to solve following LP problem.

Maximize Z = 3x1 + 2x2 + 2x3
subject to the constraints
(i) 5x1 + 7x2 + 4x3  7, (ii) – 4x1 + 7x2 + 5x3  – 2, (iii) 3x1 + 4x2 – 6x3  29/7
and x1, x2, x3  0. [Punjab Univ., BE (E & C) 2006 ; BE (IT) 2006]
Solution Since RHS of constraint 2 is negative, multiplying it by – 1 on both sides and express it as:
4x1 – 7x2 – 5x3  2.
Phase I: Introducing slack, surplus and artificial variables in the constraints, the standard form of LP
problem becomes:
118 Operations Research: Theory and Applications

Minimize Z* = A1
subject to the constraints
(i) 5x1 + 7x2 + 4x3 + s1 = 7, (ii) 4x1 – 7x2 – 5x3 + s2 = 2, (iii) 3x1 + 4x2 – 6x3 – s3 + A1 = 29/7
and x1, x2, x3, s1, s2, s3, A1  0.
Setting variables x1 = x2 = x3 = s3 = 0, the basic feasible solution shown in Table 4.21 to the auxiliary LP
problem is: s1 = 7, s2 = 2, A1 = 29/7
cj Æ 0 0 0 0 0 0 1
Basic Variables Basic Basic Variables x1 x2 x3 s1 s2 s3 A1 Min Ratio
Coefficient Variables Value xB /x2
cB B b (= xB)
0 s1 7 5 7 4 1 0 0 0 1 
0 s2 2 4 –7 –5 0 1 0 0 —
1 A1 29/7 3 4 –6 0 0 –1 1 29/28

Z = 29/7 zj 3 4 –6 0 0 –1 1
Table 4.21
Initial Solution cj – zj –3 –4 6 0 0 1 0


Since c2 – z2 = – 4 is the largest negative value under x2-column in Table 4.21, replacing basic variable s1
with the non-basic variable x2 into the basis. For this, apply following row operations:
R1(new)  R1(old) ÷ 7(key element); R2(new)  R2(old) + 7R1(new);
R3(new)  R3(old) – 4R1(new)
to get an improved solution as shown in Table 4.22.
cj Æ 0 0 0 0 0 0 1
Basic Variables Basic Basic Variables x1 x2 x3 s1 s2 s3 A1 Min Ratio
Coefficient Variables Value xB /x1
cB B b (= xB)
0 x2 1 5/7 1 4/7 1/7 0 0 0 7/5
1 
0 s2 9 9 0 –1 1 1 0 0 Tie
1 A1 1/7 1/7 0 – 58/7 – 4/7 0 –1 1 1 

Table 4.22 Z = 1/7 zj 1/7 0 – 58/7 – 4/7 0 –1 1
An Improved cj – zj –1/7 0 58/7 4/7 0 1 0
Solution


Solution shown in Table 4.22 is not optimal and there is a tie between the s2 and A1-rows for the key row.
The A1-row is selected as the key row because preference is given to the artificial variable. Thus (1/7) is the key
element. Replace basic variable A1 in the basis with the non-basic variable x1. Applying following row opera-
tions to get the new solution as shown in Table 4.23.
R3(new)  R3(old) ÷ (1/7) key element; R1(new)  R1(old) – (5/7) R3 (new);
R2(new)  R2(old) – 9 R3(new)
cj  0 0 0 0 0 0
Basic Variables Variables Basic Variables x1 x2 x3 s1 s2 s3
Coefficient in Basis Value
cB B b (= xB)
0 x2 2/7 0 1 42 3 0 5
0 s2 0 0 0 521 37 1 63
0 x1 1 1 0 – 58 –4 0 –7
Table 4.23
Basic Feasible Z=0 zj 0 0 0 0 0 0
Solution
cj – zj 0 0 0 0 0 0
 Linear Programming: The Simplex Method 119

In Table 4.23, all cj – zj = 0 under non-basic variable columns. Thus, the current solution is optimal. Also
Min Z* = 0 and no artificial variable appears in the basis, and hence this solution is the basic feasible solution
to the original LP problem.
Phase II: Using the solution shown in Table 4.23 as the initial solution for Phase II and carrying out
computations to get optimal solution as shown in Table 4.24.
cj  3 2 2 0 0 0
Basic Variables Basic Basic Variables x1 x2 x3 s1 s2 s3 Min Ratio
Coefficient Variables Value xB /x3
cB B b (= xB)
2 x2 2/7 0 1 42 3 0 5 1/147
0 s2 0 0 0 521 37 1 63 0 
3 x1 1 1 0 – 58 –6 0 –7 —
Z = 25/7 zj 3 2 – 90 –6 0 – 11 Table 4.24
cj – zj 0 0 92 6 0 11 Initial Solution


In Table 4.24, c3 – z3 = 92 is the largest positive value corresponding non-basic variable x3, replacing basic
variable s2 with non-basic variable x3 into the basis. For this apply necessary row operations as usual. The new
solution is shown in Table 4.25.

cj  3 2 2 0 0 0
Basic Variables Basic Basic Variables x1 x2 x3 s1 s2 s3
Coefficient Variables Value
cB B b (= xB)
2 x2 2/7 0 1 0 9/521 – 42/521 – 41/521
2 x3 0 0 0 1 37/521 1/521 63/521
3 x1 1 1 0 0 62/521 58/521 7/521
Z = 25/7 zj 3 2 2 278/521 92/521 65/521 Table 4.25
cj – zj 0 0 0 –278/521 – 92/521 – 65/521 Optimal Solution

Since all cj – zj  0 in Table 4.25, the current solution is the optimal basic feasible solution: x1 = 1, x2 = 2/7,
x3 = 0 and Max Z = 25/7.

4.4.2 Big-M Method

Big-M method is another method of removing artificial variables from the basis. In this method, large
undesirable (unacceptable penalty) coefficients to artificial variables are assigned from the point of view
of the objective function. If the objective function Z is to be minimized, then a very large positive price (called
penalty) is assigned to each artificial variable. Similarly, if Z is to be maximized, then a very large negative
price (also called penalty) is assigned to each of these variables. The penalty is supposed to be designated by
– M, for a maximization problem, and + M, for a minimization problem, where M > 0. The Big-M method
for solving an LP problem can be summarized in the following steps:

Step 1: Express the LP problem in the standard form by adding slack variables, surplus variables and/or
artificial variables. Assign a zero coefficient to both slack and surplus variables. Then assign a very large
coefficient + M (minimization case) and – M (maximization case) to artificial variable in the objective function.
Step 2: The initial basic feasible solution is obtained by assigning zero value to decision variables, x1,
x2, ..., etc.
Step 3: Calculate the values of cj – zj in last row of the simplex table and examine these values.
(i) If all cj – zj  0, then the current basic feasible solution is optimal.
(ii) If for a column, k, ck – zk is most negative and all entries in this column are negative, then the
problem has an unbounded optimal solution.
120 Operations Research: Theory and Applications

(iii) If one or more cj – zj < 0 (minimization case), then select the variable to enter into the basis (solution
mix) with the largest negative cj – zj value (largest per unit reduction in the objective function value).
This value also represents the opportunity cost of not having one unit of the variable in the solution.
That is,
ck – zk = Min {cj – zj : cj – zj < 0}
Step 4: Determine the key row and key element in the same manner as discussed in the simplex algorithm
for the maximization case.
Step 5: Continue with the procedure to update solution at each iteration till optimal solution is obtained.
Remarks At any iteration of the simplex algorithm one of the following cases may arise:
1. If at least one artificial variable is a basic variable (i.e., variable that is present in the basis) with zero
value and the coefficient it M in each cj – zj ( j = 1, 2, . . ., n) values is non-negative, then the given
LP problem has no solution. That is, the current basic feasible solution is degenerate.
2. If at least one artificial variable is present in the basis with a positive value and the coefficients M in each
cj – zj ( j = 1, 2, . . ., n) values is non-negative, then the given LP problem has no optimum basic feasible
solution. In this case, the given LP problem has a pseudo optimum basic feasible solution.
Example 4.7 Use penalty (Big-M) method to solve the following LP problem.
Minimize Z = 5x1 + 3x2
subject to the constraints
(i) 2x1 + 4x2  12, (ii) 2x1 + 2x2 = 10, (iii) 5x1 + 2x2  10
and x1, x2  0.
Solution Adding slack variable, s1; surplus variable, s2 and artificial variables, A1 and A2 in the constraints
of the given LP problem, the standard form of the LP problem becomes.
Minimize Z = 5x1 + 3x2 + 0s1 + 0s2 + MA1 + MA2
subject to the constraints
(i) 2x1 + 4x2 + s1 = 12, (ii) 2x1 + 2x2 + A1 = 10 (iii) 5x1 + 2x2 – s2 + A2 = 10
and x1, x2, s1, s2, A1, A2  0
An initial basic feasible solution: s1 = 12, A1 = 10, A2 = 10 and Min Z = 10M + 10M = 20M is obtained
by putting x1 = x2 = s2 = 0. It may be noted that the columns that correspond to the current basic variables and
form the basis (identity matrix) are s1 (slack variable), A1 and A2 (both artificial variables). The initial basic
feasible solution is given in Table 4.26.
Since the value c1 – z1= 5 – 7M is the smallest value, therefore variable x1 is chosen to enter into the basis
(solution mix). To decide a current basic variable to leave the basis, calculate minimum ratio as shown in
Table 4.26.
cj  5 3 0 0 M M
Basic Variables Basic Basic Variables x1 x2 s1 s2 A1 A2 Min Ratio
Coefficient Variables Value xB/x1
cB B b (= xB )
0 s1 12 2 4 1 0 0 0 12/2 = 6
M A1 10 2 2 0 0 1 0 10/2 = 5
M A2 10 5 2 0 –1 0 1 10/5  2 
Z = 20M zj 7M 4M 0 – M M M
Table 4.26 cj – zj 5 – 7M 3 – 4M 0 M 0 0
Initial Solution 

Iteration 1: Introduce variable x1 into the basis and remove A2 from the basis by applying the following
row operations.
R3 (new)  R3 (old)  5 (key element); R2 (new)  R2 (old) – 2R3 (new).
R1 (new)  R1 (old) – 2R3 (new).
The improved basic feasible solution is shown in Table 4.27.
 Linear Programming: The Simplex Method 121

cj  5 3 0 0 M
Basic Variables Basic Basic Variables x1 x2 s1 s2 A1 Min Ratio
Coefficient Variables Value xB/x2
cB B b (= xB )

0 s1 8 0 16/5 1 2/5 0 8/(16/5) = 5/2 

M A1 6 0 6/5 0 2/5 1 6/(6/5) = 5
5 x1 2 1 2/5 0 – 1/5 0 2/(2/5) = 5
Z = 10 + 6M zj 5 (6M/5) + 2 0 (2M/5) – 1 M
cj – zj 0 (– 6M/5) + 1 0 (– 2M/5) + 1 0 Table 4.27
 Improved Solution

Iteration 2: Since the value of c2 – z2 in Table 4.27 is the largest negative value, variable x2 is chosen
to replace basic variable s1 in the basis. Thus, to get an improved basic feasible solution, apply, the following
row operations:
R1 (new)  R1 (old) × 5/16 (key element); R2 (new)  R2 (old) – (6/5) R1 (new).
R3 (new)  R3 (old) – (2/5) R1 (new).
The new solution is shown in Table 4.28.

cj  5 3 0 0 M
Basic Variables Basic Basic Variables Min Ratio
Coefficient Variables Value x1 x2 s1 s2 A1 xB /s2
cB B b (= xB)
3 x2 5/2 0 1 5/16 1/8 0 (5/2)/(1/8) = 40
M A1 3 0 0 – 3/8 1/4 1 3/(1/4) = 12 
5 x1 1 1 0 – 1/8 – 1/4 0
Z = 25/2 + 3M zj 5 3– 3M/8 + 5/16 M/4 – 7/8 M
cj – zj 0 03M/8 – 5/16 – M/4 + 7/8 0 Table 4.28
Improved Solution


Iteration 3: Since c4 – z4 < 0 (negative) in s2-column, the current solution is not optimal. Thus, non-basic
variable s2 is chosen to replace artificial variable A1 in the basis. To get an improved basic feasible solution,
apply the following row operations:
R2 (new)  R2 (old) × 4 (key element); R1 (new)  R1 (old) – (1/8) R2 (new)
R3 (new)  R3 (old) + (1/4) R2 (new).
The improved basic feasible solution is shown in Table 4.29.

cj  5 3 0 0
Basic Variables Basic Basic Variables x1 x2 s1 s2
Coefficient Variables Value
cB B b(= xB )
3 x2 1 0 1 1/2 0
0 s2 12 0 0 – 3/2 1
5 x1 4 1 0 – 1/2 0
Z = 23 zj 5 3 – 1 0 Table 4.29
cj – zj 0 0 1 0 Optimal Solution

In Table 4.24, all cj – zj  0. Thus an optimal solution is arrived at with the value of variables as:
x1 = 4, x2 = 1, s1 = 0, s2 = 12 and Min Z = 23.
122 Operations Research: Theory and Applications

Example 4.8 Use penalty (Big-M) method to solve the following LP problem.
Maximize Z = x1 + 2x2 + 3x3 – x4
subject to the constraints
(i) x1 + 2x2 + 3x3 = 15, (ii) 2x1 + x2 + 5x3 = 20, (iii) x1 + 2x2 + x3 + x4 = 10
and x1, x2, x3, x4  0 [Calicut, BTech. (Engg), 2000; Bangalore BE (Mech.), 2000; AMIE, 2009]

Solution Since all constraints of the given LP problem are equations, therefore only artificial variables A1
and A2 are added in the constraints to convert given LP problem to its standard form. The standard form
of the problem is stated as follows:
Maximize Z = x1 + 2x2 + 3x3 – x4 – MA1 – MA2
subject to the constraints
(i) x1 + 2x2 + 3x3 + A1 = 15, (ii) 2x1 + x2 + 5x3 + A2 = 20
(iii) x1 + 2x2 + x3 + x4 = 10
and x1, x2, x3, x4, A1, A2  0
An initial basic feasible solution is given in Table 4.30.
cj  1 2 3 –1 – M – M
Basic Variables Basic Basic Variables Min Ratio
Coefficient Variables Value x1 x2 x3 x4 A1 A2 xB/x3
cB B b (= xB )

– M A1 15 1 2 3 0 1 0 15/3 = 5
– M A2 20 2 1 5 0 0 1 20/5 = 4 
– 1 x4 10 1 2 1 1 0 0 10/1 = 10
Z = – 35M – 10 zj – 3M – 1 – 3M – 2 – 8M – 1 –1 – M – M
Table 4.30 cj – zj 3M + 2 3M + 4 8M + 4 0 0 0
Initial Solution 

Since value of c3 – z3 in Table 4.30 is the largest positive value, the non-basic variable x3 is chosen to replace
the artificial variable A2 in the basis. Thus, to get an improved solution, apply the following row operations.
R2 (new)  R1 (old)  5 (key element) ; R1 (new)  R1 (old) – 3 R2 (new).
R3 (new)  R3 (old) – R2 (new)
The improved basic feasible solution is shown in Table 4.31.
cj  1 2 3 –1 –M

Basic Variables Basic Basic Variables Min Ratio

Coefficient Variables Value x1 x2 x3 x4 A1 xB/x2
cB B b (= xB )

3 15
– M A1 3 – 1/5 7/5 0 0 1 = 
7/5 7
4
3 x3 4 2/5 1/5 1 0 0 = 20
1/5
6 30
– 1 x4 6 3/5 9/5 0 1 0 =
9/5 9

Z = – 3M + 6 zj M/5 + 3/5 – 7M/5 – 6/5 3 – 1 – M

Table 4.31 cj – zj – M/5 – 2/5 7M/5 + 16/5 0 0 0
Improved Solution

Since value of c2 – z2 > 0 (positive) in Table 4.31, the non-basic variable x2 is chosen to replace the
artificial variable A1 in the basis. Thus, to get an improved basic feasible solution, apply the following row
operations:
 Linear Programming: The Simplex Method 123

R1 (new)  R1 (old) × (5/7) (key element); R2 (new)  R2 (old) – (1/5) R1 (new).

R3 (new)  R3 (old) – (9/5) R1 (new).
The improved basic feasible solution is shown in Table 4.32.

cj 1 2 3 1
Basic Variables Basic Basic Variables x1 x2 x3 x4 Min Ratio
Coefficient Variables Value xB /x1
cB B b(= xB )

2 x2 15/7 – 1/7 1 0 0 –
3 x3 25/7 3/7 0 1 0 25/7 × 7/3 = 25/3
–1 x4 15/7 6/7 0 0 1 15/7 × 7/6 = 15/6 
Z = 90/7 zj 1/7 2 3 –1
cj – zj 6/7 0 0 0 Table 4.32
Improved Solution


Once again, the solution shown in Table 4.32 is not optimal as c1 – z1 > 0 in x1-column. Thus, applying
the following row operations for replacing non-basic variable x1 in the basis with basic variable x4:

R3 (new)  R3 (old) × (7/6) (key element) ; R1 (new)  R1 (old) + (1/7) R3 (new).

R2 (new)  R2 (old) – (3/7) R3 (new).

The improved basic feasible solution is shown in Table 4.33.

cj  1 2 3 1
Basic Variables Basic Basic Variables x1 x2 x3 x4
Coefficient Variables Value
cB B b (= xB )
2 x2 15/6 0 1 0 1/6
3 x3 15/6 0 0 1 – 3/6
1 x1 15/6 1 0 0 7/6
Z = 15 zj 1 2 3 0 Table 4.33
cj – zj 0 0 0 – 1 Optimal Solution

In Table 4.28, since all cj – zj  0, therefore, an optimal solution is arrived at with value of variables
as: x1 = 15/6, x2 = 15/6, x3 = 15/6 and Max Z = 15.
Example 4.9 ABC Printing Company is facing a tight financial squeeze and is attempting to cut costs
wherever possible. At present it has only one printing contract, and luckily, the book is selling well in both the
hardcover and the paperback editions. It has just received a request to print more copies of this book in either
the hardcover or the paperback form. The printing cost for the hardcover books is Rs 600 per 100 books while
that for paperback is only Rs 500 per 100. Although the company is attempting to economize, it does not wish
to lay off any employee. Therefore, it feels obliged to run its two printing presses – I and II, at least 80 and 60
hours per week, respectively. Press I can produce 100 hardcover books in 2 hours or 100 paperback books in 1
hour. Press II can produce 100 hardcover books in 1 hour or 100 paperbacks books in 2 hours. Determine how
many books of each type should be printed in order to minimize costs.
Solution Let x1 and x2 be the number of batches containing 100 hard cover and paperback books, tobe
printed respectively. The LP problem can then be formulated as follows.
Minimize Z = 600x1 + 500x2
subject to the constraints

(i) 2x1 + x2  80, (ii) x1 + 2x2  60 (Printing press hours)

and x1, x2  0
Standard form Adding surplus variables s1, s2 and artificial variables A1, A2 in the the constraints, the
standard form of the LP problem becomes
124 Operations Research: Theory and Applications

Minimize Z = 600x1 + 500x2 + 0s1 + 0s2 + MA1 + MA2

subject to the constraints
(i) 2x1 + x2 – s1 + A1 = 80, (ii) x1 + 2x2 – s2 + A2 = 60
and x1, x2, s1, s2, A1, A2  0
Solution by simplex method The initial basic feasible solution: A1 = 80, A2 = 60 and Min Z = 80M + 60M = 140M
obtained by putting x1 = x2 = s1 = s2 = 0 is shown in Table 4.34.
cj  600 500 0 0 M M
Basic Variables Basic Basic Variables x1 x2 s1 s2 A1 A2 Min Ratio
Coefficient Variables Value xB /x2
cB B b(= xB )
M A1 80 2 1 –1 0 1 0 80/1
M A2 60 1 2 0 – 1 0 1 60/2 
Z = 140M zj 3M 3M – M – M M M
Table 4.34 cj – zj 600 – 3M 500 – 3M M M 0 0
Initial Solution

Since c2 – z2 value in x2-column of Table 4.34 is the largest negative, therefore non-basic variable x2
is chosen to replace basic variable A2 in the basis. For this, apply following row operations.
R2 (new) = R2 (old)  2 (key element); R1 (new)  R1 (old) – R2 (new)
to get an improved basic feasible solution as shown in Table 4.35.
cj  600 500 0 0 M
Basic Variables Basic Basic Variables x1 x2 s1 s2 A1 Min Ratio
Coefficient Variables Value xB /x1
cB B b (= xB )
M A1 50 3/2 0 – 1 1/2 1 100/3 
500 x2 30 1/2 1 0 – 1/2 0 60
Z = 15,000 + 50M zj 3M/2 + 250 500 –M M/2 – 250 M
Table 4.35
cj – zj 350 – 3M/2 0 M 250 – M/2 0
Improved Solution


Again, since value c1 – z1 of in x1-column of Table 4.35 is the largest negative, non-basic variable x1
is chosen to replace basic variable A1 in the basis. For this, apply following row operations:
R1 (new)  R1 (old) × (2/3) (key element) ; R2 (new)  R2 (old) – (1/2) R1 (new)
to get an improved basic feasible solution as shown in Table 4.36.
cj  600 500 0 0
Basic Variables Basic Basic Variables x1 x2 s1 s2
Coefficient Variables Value
cB B b (= xB )
600 x1 100/3 1 0 – 2/3 1/3
500 x2 40/3 0 1 1/3 – 2/3

Table 4.36 Z = 80,000/3 zj 600 500 – 700/3 – 400/3

Optimal Solution cj – zj 0 0 700/3 400/3

In Table 4.36, all cj – zj  0 and no artificial variable is present in the basis (solution mix). Hence, an
optimum solution is arrived at with x1 = 100/3 batches of hardcover books, x2 = 40/3 batches of paperback
books, at a total minimum cost, Z = Rs. 80,000/3.
Example 4.10 An advertising agency wishes to reach two types of audiences: Customers with annual
income greater than Rs 15,000 (target audience A) and customers with annual income less than Rs 15,000
(target audience B). The total advertising budget is Rs 2,00,000. One programme of TV advertising costs
Rs 50,000; one programme on radio advertising costs Rs 20,000. For contract reasons, at least three
programmes ought to be on TV, and the number of radio programmes must be limited to five. Surveys
indicate that a single TV programme reaches 4,50,000 customers in target audience A and 50,000 in target
 Linear Programming: The Simplex Method 125

audience B. One radio programme reaches 20,000 in target audience A and 80,000 in target audience B.
Determine the media mix to maximize the total reach. [Delhi Univ., MBA, 2000]
Solution Let x1 and x2 be the number of insertions in TV and radio, respectively. The LP problem can
then be formulated as follows:
Maximize (total reach) Z = (4,50,000 + 50,000) x1 + (20,000 + 80,000) x2
= 5,00,000 x1 + 1,00,000 x2 = 5x1 + x2
subject to the constraints
(i) 50,000 x1 + 20,000 x2  2,00,000 or 5x1 + 2x2  20 (Advt. budget)
(ii) x1  3 (Advt. on TV) (iii) x2  5 (Advt. on Radio)
and x1, x2  0.
Standard form Adding slack/surplus and/or artificial variables in the constraints of LP problem. Then
standard form of given LP problem becomes
Maximize Z = 5x1 + x2 + 0s1 + 0s2 + 0s3 – MA1
subject to the constraints
(i) 5x1 + 2x2 + s1 = 20, (ii) x1 – s2 + A1 = 3, (iii) x2 + s3 = 5
and x1, x2, s1, s2, s3, A1  0
Solution by simplex method An initial basic feasible solution: s1 = 20, A1 = 3, s3 = 5 and Max Z = – 3M
obtained by putting x1 = x2 = s2 = 0 is shown in simplex Table 4.37.
cj  5 1 0 0 0 –M
Basic Variables Basic Basic Variables x1 x2 s1 s2 s3 A1 Min Ratio
Coefficient Variables Value xB/x1
cB B b(= xB )
0 s1 20 5 2 1 0 0 0 20/5 = 4
–M A2 3 1 0 0 –1 0 1 3/1 = 3 
0 s3 5 0 1 0 0 1 0 —
Z = – 3M zj –M 0 0 M 0 –M
Table 4.37
cj – zj M+5 1 0 –M 0 0
Initial Solution


The c1 – z1 value in x1-column of Table 4.37 is the largest positive value. The non-basic variable x1
is chosen to replace basic variable A1 in the basis. For this apply following row operations
R2 (new)  R2 (old)  1 (key element); R1 (new)  R1 (old) – 5R2 (new)
to get the improved basic feasible solution as shown in Table 4.38.

cj  5 1 0 0 0
Basic Variables Basic Basic Variables x1 x2 s1 s2 s3 Min Ratio
Coefficient Variable Value xB /s2
cB B b ( = xB )
0 s1 5 0 2 1 5 0 5/5 = 1 
5 x1 3 1 0 0 – 1 0 —
0 s3 5 0 1 0 0 1 —
Z = 15 zj 5 0 0 – 5 0
cj – zj 0 1 0 5 0 Table 4.38
Improved Solution

Again in Table 4.38, c4 – z4 value in s2-column is the largest positive. Thus non-basic variable s2 is
chosen to replace basic variable s1 into the basis. For this apply the following row operations:
R1 (new)  R1 (old)  5 (key element); R2 (new)  R2 (old) + R1 (new)
to get the improved basic feasible solution as shown in Table 4.39.
126 Operations Research: Theory and Applications

cj  5 1 0 0 0
Basic Variables Basic Basic Variables x1 x2 s1 s2 s3
Coefficient Variables Value
cB B b (= xB )
0 s2 1 0 2/5 1/5 1 0
5 x1 4 1 2/5 1/5 0 0
0 s3 5 0 1 0 0 1
Table 4.39 Z = 20 zj 5 2 1 0 0
Optimal Solution – 1 – 1
cj – zj 0 0 0
Since all cj – zj  0 in Table 4.39, the total reach of target audience cannot be increased further. Hence,
the optimal solution is: x1 = 4 insertions in TV and x2= 0 in radio with Max (total audience) Z = 20,00,000.
Example 4.11 An Air Force is experimenting with three types of bombs P, Q and R in which three kinds of
explosives, viz., A, B and C will be used. Taking the various factors into account, it has been decided to use the
maximum 600 kg of explosive A, at least 480 kg of explosive B and exactly 540 kg of explosive C. Bomb P requires
3, 2, 2 kg, bomb Q requires 1, 4, 3 kg and bomb R requires 4, 2, 3 kg of explosives A, B and C respectively. Bomb
P is estimated to give the equivalent of a 2 ton explosion, bomb Q, a 3 ton explosion and bomb R, a 4 ton
explosion respectively. Under what production schedule can the Air Force make the biggest bang?
Solution Let x1, x2 and x3 be the number of bombs of type P, Q and R to be experimented, respectively.
Then the LP problem can be formulated as:
Maximize Z = 2x1 + 3x2 + 4x3
subject to the constraints
(i) Explosive A : 3x1 + x2 + 4x3  600, (ii) Explosive B : 2x1 + 4x2 + 2x3  480,
(iii) Explosive C : 2x1 + 3x2 + 3x3 = 540,
and x1, x2, x3  0.
Standard form Adding slack, surplus and artificial variables in the constraints of LP problem, the standard
form of LP problem becomes:
Maximize Z = 2x1 + 3x2 + 4x3, + 0s1 + 0s2 – MA1 – MA2
subject to the constraints
(i) 3x1 + x2 + 4x3 + s1 = 600, (ii) 2x1 + 4x2 + 2x3 – s2 + A1 = 480, (iii) 2x1 + 3x2 + 3x3 + A2 = 540,
and x1, x2, x3 s1, s2, A1, A2  0.
Solution by simplex method The initial basic feasible solution: s1 = 600, A1 = 480, A2 = 540,
Max Z = –1,020 M obtained by putting basic variables x1 = x2 = s2 = 0 is shown in Table 4.40.
cj Æ 2 3 4 0 0 –M –M
Basic Variables Basic Basic Variables x1 x2 x3 s1 s2 A1 A2 Min Ratio
Coefficient Variables Value xB /x2
cB B b (= xB)
0 s1 600 3 1 4 1 0 0 0 600/1
–M A1 480 2 4 2 0 –1 1 0 480/4 
–M A2 540 2 3 3 0 0 0 1 540/3
Z = 1,020 M zj – 4M – 7M – 5M 0 M –M –M
Table 4.40 cj – zj 2 + 4M 3 + 7M 4 + 5M 0 –M 0 0
Initial Solution 

Since in Table 4.40, c2 – z2 value in x2–column is largest positive, non-basic variable x2 is chosen to replace
basic variable A1 in the basis. For this, apply the following row operations:
R2(new)  R2(old) × (1/4) (key element); R1(new)  R1(old) – R2(new);
R3(new)  R3(old) – 3R2(new)
to get the improved basic feasible solution as shown in Table 4.41.
 Linear Programming: The Simplex Method 127

cj Æ 2 3 4 0 0 –M
Basic Variables Basic Basic Variables x1 x2 x3 s1 s2 A1 Min Ratio
Coefficient Variables Value xB /x3
cB B b (= xB)
0 s1 480 5/2 0 7/2 1 1/4 0 960/7
3 x2 120 1/2 1 1/2 0 –1/4 0 240
–M A1 180 1/3 0 3/2 0 3/4 1 120 
Z = 360 – 180M zj 3/2 – M/2 3 3/2 – 3M/2 0 –3/4 – 3M/4 –M
cj – zj 1/2 + M/2 0 5/2 + 3M/2 0 3/4 + 3M/4 0 Table 4.41
 Improved Solution

In Table 4.41, c3 – z3 value in x3-column is largest positive, non-basic variable x3 is chosen to replace basic
variable A2 into the basis. For this, apply following row operations:
R3(new)  R3(old) × (2/3) (key element); R1(new)  R1(old) – (7/2) R3(new);
R2(new)  R2(old) – (1/2) R3(new).
to get the improved basic feasible solution as shown in Table 4.42.
cj Æ 2 3 4 0 0

Basic Variables Basic Basic Variables x1 x2 x3 s1 s2

Coefficient Variables Value
cB B b (= xB )

0 s1 60 4/3 0 0 1 –3/2
3 x2 60 1/3 1 0 0 –1/2
4 x3 120 1/3 0 1 0 1/2
Z = 660 zj 7/3 3 4 0 1/2 Table 4.42
cj – zj –1/3 0 0 0 –1/2 Optimal Solution

In Table 4.42, all cj – zj  0 and artificial variables A1 and A2 have been removed from the basis (solution
mix). Thus, an optimal solution is arrived at with x1 = 0 bombs of type P, x2 = 60 bombs of type Q, x3 = 120 bombs
of type R, at largest benefit of Z = 660.

SELF PRACTICE PROBLEMS A

1. A television company has three major departments for manufac- M3 and Belt C requires 5 hours on machine M2 and 4 hours on
turing two of its models – A and B. The monthly capacities of machine M3. There are 8 hours of time per day available on
the departments are given as follows: machine M1, 10 hours of time per day available on machine M2 and
15 hours of time per day available on machine M3. The profit gained
Per Unit Time from belt A is Rs 3.00 per unit, from Belt B is Rs 5.00 per unit,
Requirement (hours) Hours Available from belt C is Rs 4.00 per unit. What should be the daily production
Model A Model B this Month of each type of belt so that the products yield the maximum profit?
3. A company produces three products A, B and C. These products
Department I 4.0 2.0 1,600 require three ores O1, O2 and O3. The maximum quantities of the
Department II 2.5 1.0 1,200 ores O1, O2 and O3 available are 22 tonnes,
Department III 4.5 1.5 1,600 14 tonnes and 14 tonnes, respectively. For one tonne of each
of these products, the ore requirements are:
The marginal profit per unit from model A is Rs 400 and from model
B is Rs 100. Assuming that the company can sell any quantity of A B C
either product due to favourable market conditions, determine the
O1 3 – 3
optimum output for both the models, the highest possible profit for
this month and the slack time in the threedepartments. O2 1 2 3
O3 3 2 3
2. A manufacturer of leather belts makes three types of belts A,
Profit per tonne 1 4 5
B and C which are processed on three machines M1, M2 and
(Rs in thousand)
M3. Belt A requires 2 hours on machine M1 and 3 hours on
machine M2 and 2 hours on machine M3. Belt B requires 3 hours The company makes a profit of Rs 1,000, 4,000 and 5,000 on each
on machine M1, 2 hours on machine M2 and 2 hours on machine tonne of the products A, B and C, respectively. How many
128 Operations Research: Theory and Applications

tonnes of each product should the company produce in order 7. Mr Jain, the marketing manager of ABC Typewriter Company is
to maximize its profits. trying to decide how he should allocate his salesmen to the
4. A manufacturing firm has discontinued the production of a certain company’s three primary markets. Market I is in the urban area and
unprofitable product line. This has created considerable excess the salesman can sell, on the average, 40 typewriters a week.
production capacity. Management is considering to devote this Salesmen in the other two markets, II and III can sell, on the
excess capacity to one or more of three products; call them average, 36 and 25 typewriters per week, respectively. For the
product 1, 2 and 3. The available capacity on the machines that coming week, three of the salesmen will be on vacation, leaving
might limit output is summarized in the following table: only 12 men available for duty. Also because of lack of company
care, a maximum of 5 salesmen can be allocated to market area I.
Machine Type Available Time The selling expenses per week for salesmen in each area are Rs
(in Machine-hours per Week) 800 per week for area I, Rs 700 per weekfor area II, and Rs
500 per week for area III. The budget for the next week is Rs
Milling Machine 250 7,500. The profit margin per typewriter isRs 150. Determine how
Lathe 150 many salesmen should be assigned to each area in order to
maximize profits?
Grinder 50
8. Three products – A, B and C – are produced in three machine
The number of machine-hours required for each unit of the centres X, Y and Z. All three products require a part of their
respective product is as follows: manufacturing operation at each of the machine centres. Thetime
required for each operation on various products is indicated in the
Machine Type Productivity following table. Only 100, 77 and 80 hours are available at
(in Machine-hours per Unit) machine centres X, Y and Z, respectively. The profit per unit from
A, B and C is Rs 12, Rs 3 and Re 1, respectively.
Product Product Product
1 2 3 Products Machine Centres Profit per
Milling Machine 8 2 3 Unit (Rs)
X Y Z
Lathe 4 3 0
A 10 7 2 12
Grinder 2 – 1
B 2 3 4 3
The profit per unit would be Rs 20, Rs 6 and Rs 8, respectively C 1 2 1 1
for product 1, 2 and 3. Find how much of each product the firm Available hours 100 77 80
should produce in order to maximize its profit.
(a) Determine suitable product mix so as to maximize the profit.
[Delhi Univ., MBA 2006]
Comment on the queries (b) and (c) from the solution table
5. A farmer has 1,000 acres of land on which he can grow corn, wheat obtained.
or soyabean. Each acre of corn costs Rs 100 for preparation, (b) Satisfy that full available hours of X and Y have been utilized
requires 7 men-days of work and yields a profit of Rs 30. An and there is surplus hours of Z. Find out the surplus hours of
acre of wheat costs Rs 120 to prepare, requires 10 men-days of Z.
work and yields a profit of Rs 40. An acre ofsoyabean costs Rs 70 (c) Your aim is to utilize surplus capacity of Z. Can you say
to prepare, requires 8 men-days of work and yields a profit of Rs from the table that the introduction of more units of Y is
20. If the farmer has Rs 1,00,000 for preparation and can count on required ?
8,000 men-days of work, deter- mine how many acres should be 9. A certain manufacturer of screw fastenings found that there is
allocated to each crop in orderto maximize profits? [Delhi Univ., a market for packages of mixed screw sizes. His market research
MBA, 2004] data indicated that two mixtures of three screw types (1, 2 and
6. The annual handmade furniture show and sale is supposed to 3), properly priced, could be most acceptable to the public. The
take place next month and the school of vocational studies is also relevant data is:
planning to make furniture for this sale. There are three wood-
working classes – I year, II year and III year, at the school and they Mixture Specifications Selling Price (Rs/kg)
have decided to make styles of chairs – A, B and C. Each chair
A  50% type 1 5
must receive work in each class. The time in hours required for
each chair in each class is:  30% type 2
and quantity of type 3
Chair I Year II Year III Year B  35% type 1 4
 45% type 2
A 2 4 3
and quantity of type 3
B 3 3 2
C 2 1 4 For these screws, the plant capacity and manufacturing cost
are as follows:
During the next month there will be 120 hours available to the
I year class, 160 hours to the II year class, and 100 hours to
Screw Type Plant Capacity Manufacturing Cost
the III year class for producing the chairs. The teacher of the wood-
(kg/day × 100) (Rs/kg)
working classes feels that a maximum of 40 chairs can be sold
at the show. The teacher has determined that the profit from each 1 10 4.50
type of chair will be: A, Rs 40; B, Rs 35 and C, Rs. 2 10 3.50
30. How many chairs of each type should be made in order to 3 6 2.70
maximize profits at the show and sale?
What production shall this manufacturer schedule for greatest
profit, assuming that he can sell all that he manufactures?
 Linear Programming: The Simplex Method 129

10. A blender of whisky imports three grades A, B and C. He mixes 15. A marketing manager wishes to allocate his annual advertising
them according to the recipes that specify the maximum or budget of Rs 2,00,000 to two media vehicles – A and B. The
minimum percentages of grades A and C in each blend. These unit cost of a message in media A is Rs 1,000 and that of B
are shown in the table below. is Rs 1,500. Media A is a monthly magazine and not more than one
insertion is desired in one issue, whereas at least five messages
Blend Specification Price per Unit should appear in media B. The expected audience for unit
(Rs) messages in media A is 40,000 and that of media B is 55,000.
Develop an LP model and solve it for maximizing the total
Blue Dot Not less than 60% of A 6.80
Not more than 20% of C effective audience.
16. A transistor radio company manufactures four models A, B, C
Highland Not more than 60% of C 5.70
and D. Models A, B and C, have profit contributions of Rs 8, Rs
Fling Not less than 15% of A 15 and Rs 25 respectively and has model D a loss of Re 1.
Old Frenzy Not more than 50% of C 4.50 Each type of radio requires a certain amount of time for the
manufacturing of components, for assembling and for packing.
Following are the supplies of the three whiskies along with their
A dozen units of model A require one hour for manufacturing,
cost.
two hours for assembling and one hour for packing. The
Whisky Maximum Quantity Cost per corresponding figures for a dozen units of model B are 2, 1 and
Available per Day Unit (Rs) 2, and for a dozen units of C are 3, 5 and 1. A dozen units of model
D however, only require 1 hour of packing. During the forthcoming
A 2,000 7.00 week, the company will be able to make available
B 2,500 5.00 15 hours of manufacturing, 20 hours of assembling and 10 hours
C 1,200 4.00 of packing time. Determine the optimal production sched- ule for
the company.
Show how to obtain the first matrix in a simplex computation of
a production policy that will maximize profits. 17. A transport company is considering the purchase of new
11. An animal feed company must produce on a daily basis 200 kg vehicles for providing transportation between the Delhi Airportand
of a mixture that consists ingredients x1 and x2 ingredient. x1 hotels in the city. There are three vehicles under consid- eration:
costs Rs 3 per kg and x2 costs Rs 8 per kg. Not more than Station wagons, minibuses and large buses. The pur- chase price
80 kg of x1 can be used and at least 60 kg of x2 must be used. Find would be Rs 1,45,000 for each station wagon, Rs 2,50,000 for each
out how much of each ingredient should be used if the company minibus and Rs 4,00,000 for each large bus. The board of directors
wants to minimize costs. has authorized a maximum amount of Rs 50,00,000 for these
purchases. Because of the heavy air travel, the new vehicles would
12. A diet is to contain at least 20 ounces of protein and 15 ounces
be utilized at maximum capacity, regardless of the type of vehicles
of carbohydrate. There are three foods A, B and C available in the
purchased. The expected net annual profit would be Rs 15,000
market, costing Rs 2, Re 1 and Rs 3 per unit, respectively. Each
for the station wagon,Rs 35,000 for the minibus and Rs 45,000
unit of A contains 2 ounces of protein and 4 ounces of carbohy-
for the large bus. The company has hired 30 new drivers for the
drate; each unit of B contains 3 ounces of protein and 2 ounces
new vehicles. They are qualified drivers for all three types of
of carbohydrate; and each unit of C contains 4 ounces of protein
vehicles. The mainte- nance department has the capacity to handle
and 2 ounces of carbohydrate. How many units of each food should
an additional 80 station wagons. A minibus is equivalent to 1.67
the diet contain so that the cost per unit diet is minimum?
station wagons and each large bus is equivalent to 2 station
13. A person requires 10, 12 and 12 units of chemicals A, B and C, wagons in terms of their use of the maintenance department.
respectively for his garden. A typical liquid product contains 5, 2 Determine the number of each type of vehicle that should be
and 1 unit of A, B and C, respectively per jar. On the other hand purchased in order to maximize profit.
a typical dry product contains 1, 2 and 4 units of A, B and C per
18. Omega Data Processing Company performs three types of activi-
unit. If the liquid product sells for Rs 3 per jar and the dry product
ties: Payroll, accounts receivables, and inventories. The profit and
for Rs 2 per carton, how many of each should be purchased in
time requirement for keypunch, computation and office printing, for
order to minimize the cost and meet the requirement?
a standard job, are shown in the following table:
14. A scrap metal dealer has received an order from a customerfor
a minimum of 2,000 kg of scrap metal. The customer requires Job Profit/ Time Requirement (Min)
that at least 1,000 kg of the shipment of metal be of high quality Standard
copper that can be melted down and used to produce copper
Job (Rs) Keypunch Computation Printing
tubing. Furthermore, the customer will not accept delivery of the
order if it contains more than 175 kg metal that he deems unfit for Payroll 275 1,200 20 100
commercial use, i.e. metal that contains an excessive amount of
A/c Receivable 125 1,400 15 60
impurities and cannot be melted down and refined profitably.
Inventory 225 800 35 80
The dealer can purchase scrap metal from two different suppli-
ers in unlimited quantities with following percentages (by weight) of Omega guarantees overnight completion of the job. Any job
high quality copper and unfit scrap. scheduled during the day can be completed during the day or night.
Any job scheduled during the night, however, must becompleted
Supplier A Supplier B
during the night. The capacities for both day and night are shown
Copper 25% 75% in the following table:
Unfit scrap 5% 10%
Capacity (Min) Keypunch Computation Print
The cost per kg of metal purchased from supplier A and B are Re 1
and Rs 4, respectively. Determine the optimal quantities of metal to Day 4,200 150 400
be purchased for the dealer from each of the two suppliers. Night 9,200 250 650
130 Operations Research: Theory and Applications

Determine the mixture of standard jobs that should be accepted The sales revenue per unit of waste cans, filing cabinets,
during the day and night. correspondence boxes and lunch boxes are Rs 20, Rs 400, Rs
19. A furniture company can produce four types of chairs. Each chair 90 and Rs 20, respectively. There are 225 units of sheetmetal
is first made in the carpentry shop and then furnished, waxed and A available in the company’s inventory, 300 of sheet metal B, and
polished in the finishing shop. The man-hours required in each are: a total of 190 units of manual labour. What is the company’s optimal
sales revenue?
Chair Type [Delhi Univ., MBA, 2003, 2005]
21. A company mined diamonds in three locations in the country. The
1 2 3 4 three mines differed in terms of their capacities, number, weight
Carpentry shop 4 9 7 10 of stones mined, and costs. These are shown in the table below:
Due to marketing considerations, a monthly production of exactly
Finishing shop 1 1 3 40
1,48,000 stones was required. A similar requirement called for at
Contribution per chair (Rs) 12 20 18 40 least 1,30,000 carats (The average stone size was at least 130/148
= 0.88 carats). The capacity of each mine is measured in cubic
The total number of man-hours available per month in carpentry meter. The mining costs are not included from the treatment costs
and finishing shops are 6,000 and 4,000, respectively. and assume to be same at each mine. The problem for the company
Assuming an abundant supply of raw material and an abundant was to meet the marketing requirements at the least cost.
demand for finished products, determine the number of eachtype
of chairs that should be produced for profit maximization. Capacity Treatment Crude Stone Count
20. A metal products company produces waste cans, filing cabinets, (M3 of Costs (Carats (Number of
file boxes for correspondence, and lunch boxes. Its inputs are Mine earth (Rs. per M3) per M3) stone per
sheet metal of two different thickness, called A and B, and manual processed) M3)
labour. The input-output relationship for the company are shown in
the table given below: Plant 1 83,000 0.60 0.360 0.58
Plant 2 3,10,000 0.36 0.220 0.26
Waste Filing Correspondence Lunch Plant 3 1,90,000 0.50 0.263 0.21
Cans Cabinets Boxes Boxes
Formulate a linear programming model to determine how much
Sheet metal A 6 0 2 3 should be mined at each location.
Sheet metal B 0 10 0 0
Manual labour 4 8 2 3

HINTS AND ANSWERS

1. Let x1 and x2 = units of models A and B to be manufactured, 4. Let x1, x2 and x3 = number of units of products 1, 2 and 3
respectively. to be produced per week, respectively.
Max Z = 400x1 + 400x2 Max Z = 20x1 + 6x2 + 8x3
subject to 4x1 + 2x2  1,600 subject to 8x1 + 2x2 + 3x3  250
5x1/2 + x2  1,200 4x1 + 3x2  150
9x1/2 + 3x2/2  1,600 2x1 + x3  50
and x1, x2, x3  0
and x1, x2  0 Ans. x1 = 0, x2 = 50, x3 = 50 and Max Z = 700.

Ans. x1 = 355.5, x2 = 0 and Max Z = 1,42,222.2.

5. Let x1, x2 and x3 = acreage of corn, wheat and soyabean,
2. Let x1, x2 and x3 = units of types A, B and C belt to be respectively.
manufactured, respectively.
Max Z = 30x1 + 40x2 + 20x3
Max Z = 3x1 + 5x2 + 4x3
subject to 10x1 + 12x2 + 7x3  10,000
subject to 2x + 3x  8
1 2 + 10x + 8x  8,000
7x 1
2x2 + 5x3  10 2 3
x1 + x2 + x3  1,000
3x1 + 2x2 + 4x3  15
and x1, x2, x3  0 and x1, x2, x3  0
Ans. x1 = 89/41, x2 = 50/41 x3 = 64/41 and Max Z = 775/41. Ans. x1 = 250, x2 = 625, x3 = 0 and Max Z = Rs 32,500.
3. Let x , x and x = quantity of products A, B and C to be 6. Let x1, x2 and x3 = number of units of chair of styles A, B
1 2 3
produced, respectively. and C, respectively.

Max Z = 40x1 + 35x2 + 30x3

Max Z = x1 + 4x2 + 5x3 subject to 2x1 + 3x2 + 2x3  120

subject to 3x1 + 3x2  22 and x1, x2, x3  0

x1 + 2x2 + 3x3  14 Ans. x1 = 0, x2 = 7, x3 = 0 and Max Z = Rs 28,000.
3x1 + 2x2  14
 4x1 + 3x2 + x3  160
3x1 + 2x2 + 4x3  100
x1 + x2 + x3  40
and x1, x2, x3  0
Ans. x1 = 20, x2 = 20, x3 = 0 and Max Z = Rs 1,500.
 Linear Programming: The Simplex Method 131

7. Let x1, x2 and x3 = salesman assigned to area, 1, 2 and 3, subject to (i) x1 + 2x2 + 3x3 = 15; (ii) 2x1 + x2 + 5x3 = 20
respectively. (iii) x1 + 2x2 + x3 + x4 = 10
Max. Z = 40 × 150x1 + 36 × 150x2 + 25 × 150x3 and x1, x2, x3, x4  0
– (800x1 + 700x2 + 500x3) Ans. x1 = 5/2, x2 = 5/2, x3 = 5/2, x4 = 0 and Max Z = 120.
subject to (i) x1 + x2 + x3  12; (ii) x1  5; 17. Let x , x and x = number of station wagons, minibuses and

1 2 3
(iii) 800x1 + 700x2 + 500x3  7,500 large buses to be purchased, respectively.

and x1, x2, x3  0 Max Z = 15,000x1 + 35,000x2 + 45,000x3

11. Let x and y = number of kg of ingredients x1, x2, respectively.
subject to x1 + x2 + x3  30
Min (total cost) Z = 3x + 8y
1,45,000x1 + 2,50,000x2 + 4,00,000x3  50,00,000
subject to (i) x + y = 200; (ii) x  80; (iii) y  60 x1 + 1.67x2 + 0.5x3  80
and x, y  0 and x1, x2, x3  0
Ans. x = 80, y = 120 and Min Z = Rs 1,200.
12. Let x1, x2 and x3 = number of units of food A, B and C, 18. Let xij represents ith job and jth activity

Max Z = 275 (x11 + x12 ) + 125(x21 + x22 ) + 225 (x31 + x32 )

respectively which a diet must contain.
subject to
Min (total cost ) Z = 2x1 + x2 + x3 1,200 (x + x ) + 1,400 (x + x )

subject to 2x1 + 2x2 + 4x3  20 11 12 21 22

4x + 2x + 2x  15 + 800 (x31 + x32 )  13,400

1 2 3 20 (x + x ) + 15 (x + x ) + 35 (x + x )  400
and x1, x2, x3  0 11 12 21 22 31 32

13. Let x 1 and x2 = number of units of liquid and dry product 100 (x11 + x12 ) + 60 (x21 + x22 ) + 80 (x31 + x32 )  1,050
produced, respectively. 1,200x12 + 1,400x22 + 800x32  9,200

Min (total cost) Z = 3x1 + 2x2 20x12 + 15x22 + 35x32  250

subject to (i) 5x1 + x2  10; (ii) 2x1 + 2x2  12; 100x12 + 60x22 + 80x32  650
(iii) x1 + 4x2  12 and xij  0 for all i, j
and x1, x2  0 19. Let x1, x2, x3 and x4 = chair types 1, 2, 3 and 4 to be produced,
Ans. x1 = 1, x2 = 5 and Min Z = 13. respectively.
14. Let x1 and x2 = number of scrap (in kg) purchased from Max Z = 12x1 + 20x2 + 18x3 + 40x4
suppliers A and B, respectively.
subject to 4x1 + 9x2 + 7x3 + 10x4  6,000
Min (total cost) Z = x1 + 4x2
x1 + x2 + 3x3 + 40x4  4,000
subject to 2.25x1 + 0.75x2  1,000
0.05x1 + 0.10x2  175 and x1, x2, x3, x4  0
x1 + x2  2,000 Ans. x1 = 4,000/3, x2 = x3 = 0, x4 = 200/3 and Max Z =
and x1, x2  0 Rs 56,000/3.
Ans. x1 = 2,500, x2 = 500 and Min Z = 4,500. 21. Let x1, x2 and x3 = cubic metric of earth processed at plants
1, 2 and 3, respectively.
15. Let x1 and x2 = number of insertions messages for media A and
B, respectively. Min Z = 0.60x1 + 0.36x2 + 0.50x3
Min (total effective audience) Z = 40,000x1 + 55,000x2 subject to

subject to (i) 1,000x1 + 1500x2  2,00,000; 0.58x 1 + 0.26x2 + 0.21x3= 1,48,000 (Stone count requirement)
(ii) x  12; (iii) x  5

1 2
0.36x1 + 0.22x2 + 0.263x3 1,30.000 (Carat requirement)
and x1, x2  0
x1  83,000; x2  3,10,000;
Ans. x1 = 12, x2 = 16/3 and Max Z = 77,3333.33.
x3  1,90,000 (Capacity requirement)
16. Let x1, x2, x3 and x4 = unit of models A, B, C and D to be
produce, respectively. Ans. x1 = 61,700; x2 = 3,10,000; x3 = 1,50,500
Max (total income) Z = 8x1 + 15x2 + 25x3 – x4 and Min Z = Rs 2,23,880.
4.5 SOME COMPLICATIONS AND THEIR RESOLUTION
In this section, some of the complications that may arise in applying the simplex method for solving both
maximization and minimization LP problems and their resolution are discussed.

4.5.1 Unrestricted Variables

In actual practice, decision variables, xj ( j = 1, 2, . . ., n) should have non-negative values. However, in many
situations, one or more of these variables may have either positive, negative or zero value. Variables that can
 assume positive, negative or zero value are called unrestricted variables. Since use of the simplex method
requires that all the decision variables must have non-negative value at each iteration, therefore in order to
convert an LP problem involving unrestricted variables into an equivalent LP problem having only non-
negative variables, each of unrestricted variable is expressed as the difference of two non-negative variables.
Let variable xr be unrestricted in sign. We define two new variables say xr and xr such that

xr = xr – xr ; xr , xr  0

If xr  xr , then xr  0, but and if xr  xr , then xr  0. Also if xr = xr , then xr = 0. Hence depending on the
value of xr and xr , the variable xr can have either positive or negative sign. For example, the following LP
problem
n
Maximize Z =  c j x j  cr xr
jr

An unrestricted subject to the constraints

variable in an LP

model can have n a x  a x = b , i = 1, 2, . . ., m

either positive,  ij j ir r i
negative or zero jr

value. and xj  0, and xr unrestricted in sign; j = 1, 2, . . ., n, j  r;

can be converted into its equivalent standard form as follows:
n
Maximize Z =  cj xj  c r ( x r  x r )
jr

subject to the constraints

n
 aij xj  air ( xr  xr ) = bi , i = 1, 2, . . ., m
jr

and xj , xr , xr  0; j = 1, 2, . . ., n, j  r.

Variables xr and xr simultaneously cannot appear in the basis (since the column vectors correspond-

ing to these variables are linearly dependent). Thus any of the following three cases may arise at the optimal
solution:
(i) xr = 0  xr = – xr (ii) xr = 0  xr = xr (iii) xr = xr = 0  xr = 0

This indicates that the value of xr and xr uniquely determines the values of the variable xr.

Example 4.12 Use the simplex method to solve the following LP problem.
Maximize Z = 3x1 + 2x2 + x3
subject to the constraints
(i) 2x1 + 5x2 + x3 = 12, (ii) 3x1 + 4x2 = 11
and x2, x3  0, and x1 unrestricted [Bombay, BSc (Maths), 2001]

Solution Introducing an artificial variable A1 in the second constraint of the given LP problem in order
to obtain the basis matrix as shown in Table 4.43. Since x1 is unrestricted in sign, introducing the non-
negative variables x1 and x 2 so that x1 = x1 – x1 , where x1 , x1  0. The standard form of the LP
problem now becomes:
132 Operations Research: Theory and Applications

Maximize Z = 3 ( x1 – x1 ) + 2x2 + x3 – MA1

subject to the constraints
(i) 2( x1  x1 ) + 5x2 + x3 = 12, (ii) 3( x1  x1 ) + 4x2 + A1 = 11
and x1 , x1 , x2, x3 , A1  0
The initial solution is shown in Table 4.43.
 Linear Programming: The Simplex Method 133

cj  3 – 3 2 1 – M

Basic Variables Basic Basic Variables x1 x1 x2 x3 A1 Min Ratio

Coefficient Variables Value xB/x2
B b(= xB)
1 x3 12 2 –2 5 1 0 12/5 
– M A1 11 3 –3 4 0 1 11/4
Z = – 11M + 12 zj – 3M + 2 3M – 2 – 4M + 5 1 – M
Table 4.43
cj – zj 3M + 1 – 3M – 1 4M – 3 0 0
Initial Solution


In Table 4.43, c3 – z3 is the largest positive value, the non-basic variable x2 is chosen to enter into
the basis to replace basic variable x3 in the basis. For this, apply the following row operations:
R1 (new)  R1 (old)5 (key element) R2 (new)  R2 (old) – 4R1 (new).
to get an improved solution shown in Table 4.44.

cj  3 –3 2 1 –M
Basic Variables Basic Basic Variables x1 x1 x2 x3 A1 Min Ratio
Coefficient Variables Value xB / x1
B b(= xB)
12 5
2 x2 12/5 2/5 –2/5 1 1/5 0  =6
5 2
7 5
– M A1 7/5 7/5 –7/5 0 – 4/5 1  = 1 
5 7
Z = – 7M/5 + 24/5 zj – 7M/5 + 4/5 7M/5 – 4/5 2 4M/5 + 2/5 – M
cj – zj 7M/5 + 11/5 – 7M/5 – 11/5 0 – 4M/5 + 3/5 0 Table 4.44
 Improved Solution

In Table 4.44, as c1 – z1 in x'1-column is positive, the non-basic variable x'1 is chosen to enter into
the basis to replace basic variable A1 in the basis. For this, apply the following row operations:
R2 (new)  R2 (old) × (5/7) (key element); R1 (new)  R1 (old) – (2/5) R2 (new).
to get an improved solution shown in Table 4.45.

cj  3 –3 2 1

Basic Variables Basic Basic Variables x1 x1 x2 x3 Min Ratio

Coefficient Variables Value xB /x3
B b(= xB )
2 x2 2 0 0 1 3/7 2/(3/7) = 14/3 
3 x1 1 1 –1 0 – 4/7 —

Z = 7 zj 3 – 3 2 – 6/7
cj – zj 0 0 0 13/7 Table 4.45
Improved Solution
 
Further, in order to improve the solution given in Table 4.45, the non-basic variable x3 is chosen to enter
into the basis and remove basic variable x2 from the basis. For this, apply the following rowoperations:

R1 (new)  R1 (old ) × (7/3) (key element); R2 (new)  R2 (old) + (4/7) R1 (new)

to get an improved solution given in Table 4.46.

134 Operations Research: Theory and Applications

cj  3 – 3 2 1

Basic Variables Basic Basic Variables x1 x1 x2 x3

Coefficient Variables Value
. B b(= xB )
1 x3 14/3 0 0 7/3 1
3 x1 11/3 1 –1 4/3 0
Table 4.46 Z = 47/3 zj 3 –3 19/3 1
Optimal Solution cj – zj 0 0 – 13/3 0

Since in Table 4.46 all cj – zj  0, an optimal solution: x1 = 11/3 or x1 = x1 – x1 = 11/3 – 0 = 11/3;
x3 = 14/3 is arrived with Max Z = 47/3.

4.5.2 Tie for Entering Basic Variable (Key Column)

While solving an LP problem using simplex method two or more columns of simplex table may have same
cj – zj value (positive or negative depending upon the type of LP problem). In order to break this tie, the
selection for key column (entering variable) can be made arbitrary. However, the number of iterations
required to arrive at the optimal solution can be minimized by adopting the following rules:
(i) If there is a tie between two decision variables, then the selection can be made arbitrarily.
(ii) If there is a tie between a decision variable and a slack (or surplus) variable, then select the
decision variable to enter into basis.
(iii) If there is a tie between two slack (or surplus) variables, then the selection can be made arbitrarily.

4.5.3 Tie for Leaving Basic Variable (Key Row) – Degeneracy

While solving an LP problem a situation may arise where either the minimum ratio to identify the basic
variable to leave the basis is not unique or value of one or more basic variables in the xB becomes zero.
This causes the problem of degeneracy.
Usually, the problem of degeneracy arises due to redundant constraint, i.e. one or more of the constraints
of LP problem are not part of feasible solution space. For example, constraints such as x1  5, x2  5 and x1
+ x2  5 in the LP problem make constraint x1 + x2  5 unnecessary (redundant).
Degeneracy may occur at any iteration of the simplex method. In order to break the tie in the minimum
ratios, the selection can be made arbitrarily. However, the number of iterations required to arrive at the
optimal solution can be minimized by adopting the following rules.
(i) Divide the coefficients of slack variables in the simplex table where degeneracy is seen by the
corresponding positive numbers of the key column in the row, starting from left to right.
(ii) Compare the ratios in step (i) from left to right columnwise, select the row that contains the smallest
ratio.
Remark When there is a tie between a slack and artificial variable to leave the basis, preference should be
given to the artificial variable for leaving the basis.

Example 4.13 Solve the following LP problem

Maximize Z = 3x1 + 9x2
subject to the constraints
(i) x1 + 4x2  8, (ii) x1 + 2x2  4
and x1, x2  0
Solution Adding slack variables s1 and s2 to the constraints, the problem can be expressed as
Maximize Z = 3x1 + 9x2 + 0s2 + 0s2
subject to the constraints
(i) x1 + 4x2 + s1 = 8, (ii) x1 + 2x2 + s2 = 4
and x1, x2, s1, s2  0
 Linear Programming: The Simplex Method 135

The initial basic feasible solution is given in Table 4.47. Since c2 – z2 = 9 is the largest positive value,
therefore, variable x2 is selected to be entered into the basis. However, both variables s1 and s2 are eligible to
leave the basis as the minimum ratio is same, i.e. 2, so there is a tie among the ratio in rows s1 and s2. This
causes the problem of degeneracy. To obtain the key row for resolving degeneracy, apply the following
procedure:
(i) Write the coefficients of the slack variables as shown in Table 4.47.
cj  3 9 0 0
Basic Variables Basic Basic Variables x1 x2 s1 s2 Min Ratio
Coefficient Variables Value xB/x2
B b(= xB )

0 s1 8 1 4 1 0 8/4 = 2 

0 s2 4 1 2 0 1 4/2 = 2 

Z = 0 zj 0 0 0 0
cj – zj 3 9 0 0 Table 4.47
 Initial Solution

x2-column Column
Row (Key Column) s1 s2

s1 4 1 0
s2 2 0 1

(ii) Dividing the coefficients of slack variables s1 and s2 by the corresponding elements in the key
column as shown below in the table.

x2-column Column
Row (Key Column) s1 s2

s1 4 1/4 = 1/4 0/4 = 0

s2 2 0/2 = 0 1/2 = 1/2

(iii) Comparing the ratios of Step (ii) from left to right columnwise, the minimum ratio (i.e., 0/2 = 0)
occurs in the s2-row. Thus, variable s2 is selected to leave the basis. The new solution is shown
in Table 4.48.

cj  3 9 0 0
Basic Variables Basic Basic Variables x1 x2 s1 s2
Coefficient Variables Value
B b(= xB)
0 s1 0 –1 0 1 –2
9 x2 2 1/2 1 0 1/2
Z = 18 zj 9/2 9 0 9/2 Table 4.48
cj – zj – 3/2 0 0 – 9/2 Optimal Solution

In Table 4.48, all cj – zj  0. Hence, an optimal solution is arrived at. The optimal basic feasible solution
is: x1 = 0, x2 = 2 and Max Z = 18.

4.6 TYPES OF LINEAR PROGRAMMING SOLUTIONS

While solving any LP problem using simplex method, at the stage of optimal solution, the following three
types of solutions may exist:
136 Operations Research: Theory and Applications

4.6.1 Alternative (Multiple) Optimal Solutions

The cj – zj values in the simplex table indicates the contribution in the objective function value by each unit of
a variable chosen to enter into the basis. Also, an optimal solution to a maximization LP problem is reached
when all cj – zj  0. But, if cj – zj = 0 for a non-basic variable column in the optimal simplex table and such non-
basic variable is chosen to enter into the basis, then another optimal solution so obtained will show no
improvement in the value of objective function.

Example 4.14 Solve the following LP problem.

Alternative optimal Maximize Z = 6x1 + 4x2
solutions arise
when cj – zj = 0 for
subject to the constraints
non-basic variable (i) 2x1 + 3x2  30, (ii) 3x1 + 2x2  24, (iii) x1 + x2  3
columns in the and x1, x2  0.
simplex table.
Solution Adding slack variables s1, s2, surplus variable s3 and artificial variable A1 in the constraint set,
the LP problem becomes
Maximize Z = 6x1+ 4x2 + 0s1 + 0s2 + 0s3 – MA1
subject to the constraints
(i) 2x1 + 3x2 + s1 = 30, (ii) 3x1 + 2x2 + s2 = 24 (iii) x1 + x2 – s3 + A1 = 3
and x1, x2, s1, s2, s3, A1  0
The optimal solution: x1 = 8, x2 = 0 and Max Z = 48 for this LP problem is shown in Table 4.49.
cj  6 4 0 0 0

Basic Variables Basic Basic Variables x1 x2 s1 s2 s3 Min Ratio

Coefficient Variables Value xB /x2
B b(= xB )

0 s1 14 0 5/3 1 – 2/3 0 14/(15/3) = 42/5 

0 s3 5 0 – 1/3 0 1/3 1 —
6 x1 8 1 2/3 0 1/3 0 8/(2/3) = 12
Z = 48 zj 6 4 0 2 0
Table 4.49 cj – zj 0 0 0 – 2 0
Optimal Solution 

In Table 4.49, c2 – z2 = 0 corresponds to a non-basic variable, x2. Thus, an alternative optimal solution
can also be obtained by entering variable x2 into the basis and removing basic variable, s1 from the basis. The
new solution is shown in Table 4.50.
cj  6 4 0 0 0
Basic Variables Basic Basic Variables x1 x2 s1 s2 s3
Coefficient Variables Value
B b (= xB )

4 x2 42/5 0 1 3/5 – 2/5 0

0 s3 39/5 0 0 1/5 1/5 1
6 x1 12/5 1 0 – 2/5 3/5 0
Table 4.50
Alternative Z = 48 zj 6 4 0 2 0
Solution cj – zj 0 0 0 –2 0

The optimal solution shown in Table 4.50 is: x1 = 12/5, x2 = 42/5 and Max Z = 48. Since this optimal
solution shows no change in the value of objective function, it is an alternative solution.
Once again, c3 – z3 = 0 corresponds to non-basic variable, s1. This again indicates that an alternative
optimal solution exists. The infinite number of solutions that can be obtained for this LP problem are as
follows:
 Linear Programming: The Simplex Method 137

Variables Solution Values General Solution

1 2

x1 8 12/5 x1 = 8  + (12/5) (1 – )
x2 0 42/5 x2 = 0  + (42/5) (1 – )
s1 14 0 s1 = l4  + (0) (1 – )
s3 5 39/5 s3 = 5  + (39/5) (1 – )

For each arbitrary value of , the value of objective function will remain same.

4.6.2 Unbounded Solution

In a maximization LP problem, if cj – zj > 0 (cj – zj < 0 for a minimization case) corresponds to a non-basic
variable column in simplex table, and all aij values in this column are negative, then minimum ratio to decide
basic variable to leave the basis can not be calculated. It is because negative value in denominator would
indicate the entry of a non-basic variable in the basis with a negative value (an infeasible solution). Also,
a zero value in the denominator would result in a ratio having an infinite value and would indicate that the
value of non-basic variable could be increased infinitely with any of the current basic variables being removed
from the basis.
Example 4.15 Solve the following LP problem.
Maximize Z = 3x1 + 5x2
subject to the constraints
(i) x1 – 2x2  6, (ii) x1  10, (iii) x2  1
and x1, x2  0

Solution Adding slack variables s1, s2, surplus variable s3 and artificial variable A1 in the constraint set. Unbounded
Then the standard form of LP problem becomes solution occurs
when value of
Maximize Z = 3x1 + 5x2 + 0s1 + 0s2 + 0s3 – MA1 decision variables in
the solution of LP
subject to the constraints problem becomes
(i) x1 – 2x2 + s1 = 6, (ii) x1 + s2 = 10, (iii) x2 – s3 + A1 = 1 infinitely large
without violating any
and x1, x2, s1, s2, s3, A1  0 given constraints.
The initial solution to this LP problem is shown in Table 4.51.

cj  3 5 0 0 0 – M

Basic Variables Basic Basic Variables x1 x2 s1 s2 s3 A1 Min Ratio

Coefficient Variables Value xB /x2
B b (= xB)

0 s1 6 1 – 2 1 0 0 0 –
0 s2 10 1 0 0 1 0 0 –
– M A1 1 0 1 0 0 – 1 1 1 

Z = – M zj 0 – M 0 0 M – M
cj – zj 3 5 + M 0 0 – M 0 Table 4.51
 Initial Solution

Iteration 1: Since c2 – z2  0, non-basic variable x2 is chosen to enter into the basis in place of basic variable
A1. The new solution is shown in Table 4.52.
138 Operations Research: Theory and Applications

cj  3 5 0 0 0 –M

Basic Variables Basic Basic Variables x1 x2 s1 s2 s3 A1

Coefficient Variables Value
B b (= xB)

0 s1 8 1 0 1 0 – 2 2
0 s2 10 1 0 0 1 0 0
5 x2 1 0 1 0 0 – 1 1

Z = 5 zj 0 5 0 0 – 5 5
Table 4.52 cj – zj 3 0 0 0 5 – M – 5

In Table 4.52, c5 – z5 = 5 is largest positive, so variable s3 should enter into the basis. But, coefficients in
the ‘s3’ column are all negative or zero. This indicates that s3 cannot be entered into the basis. However, the
value of s3 can be increased infinitely without removing any one of the basic variables. Further, since s3 is
associated with x2 in the third constraint, x2 will also be increased infinitely because it can be expressed as x2
= 1 + s3 – A1. Hence, the solution to the given problem is unbounded.

4.6.3 Infeasible Solution

If LP problem solution does not satisfy all of the constraints, then such a solution is called infeasible solution.
Also, infeasible solution occurs when all cj – zj values satisfy optimal solution condition but at least one of
artificial variables appears in the basis with a positive value. This situation may occur whenan LP model
is either improperly formulated or more than two of the constraints are incompatible.
Example 4.16 Solve the following LP problem
Maximize Z = 6x1 + 4x2
subject to the constraints
(i) x1 + x2  5, (ii) x2  8
and x1, x2  0.
Solution Adding slack, surplus and artificial variables, the standard form of LP problem becomes
Maximize Z = 6x1+ 4x2 + 0s1 + 0s2 – MA1
subject to the constraints
(i) x1 + x2 + s1 = 5, (ii) x2 – s2 + A1 = 8
and x1, x2, s1, s2, A1  0
The initial solution to this LP problem is shown in Table 4.53.
cj  6 4 0 0 – M

Basic Variables Basic Basic Variables x1 x2 s1 s2 A1 Min Ratio

Coefficient Variables Value xB/x2
B b (= xB)

6 s1 5 1 1 1 0 0 5/1 
– M A1 8 0 1 0 –1 1 8/1

Z = 30 – 8M zj 6 6 – M 0 M – M
Table 4.53 cj – zj 0 – 2 + M 0 – M 0
Initial Solution 
Iteration 1: Since c2 – z2 = M – 2 ( 0), non-basic variable x2 is chosen to enter into the basis to replace basic
variable x1. The new solution is shown in Table 4.58.
 Linear Programming: The Simplex Method 139

cj  6 4 0 0 – M

Basic Variables Basic Basic Variables x1 x2 s1 s2 A1

Coefficients Variables Values
B b (= xB )

4 x2 5 1 1 1 0 0
–M A1 3 –1 0 –1 –1 1 Table 4.54
Optimal but
Z = 20 – 3M zj 4 + M 4 4 + M M –M Infeasible
cj – zj 2 – M 0 – 4 – M –M 0 Solution

In table 5.44, since all cj – zj  0, the current solution is optimal. But this solution is not feasible for
the given LP problem because values of decision variables are: x1 = 0 and x2 = 5 violates second constraint,
x2  8. The presence of artificial variable A1 = 3 in the solution also indicates that the optimal solution
violates the second constraint (x2  8) by 3 units.

CONCEPTUAL QUESTIONS
1. Define slack and surplus variables in a linear programming 7. Explain what is meant by the terms degeneracy and cycling in
problem. linear programming? How can these problems be resolved?
2. Explain the various steps of the simplex method involved in the 8. Explain the term artificial variables and its use in linear programming.
computation of an optimum solution to a linear programming problem. 9. What are artificial variables? Why do we need them? Describe the
3. (a) Give outlines of the simplex method in linear programming. two-phase method of solving an LP problem with artificial variables.
(b) What is simplex? Describe the simplex method of solving 10. What is the significance of cj – zj numbers in the simplex table?
linear programming problem. Interpret their economic significance in terms of marginal worth.
(c) What do you understand by the term two-phase method of
11. What is meant by the term opportunity cost? How is this concept
solving linear programming problem? used in designing a test for optimality?
(d) Outline the simplex method in linear programming. Why is
it called so? 12. How do the graphical and simplex methods of solving LP problems
(e) Explain the purpose and procedure of the simplex method. differ from each other? In what ways are they same?
4. What do you mean by an optimal basic feasible solution to a 13. How do maximization and minimization problems differ when
linear programming problem? applying the simplex method?
5. Given a general linear programming problem, explain how you 14. What is the reason behind the use of the minimum ratio test in
would test whether a basic feasible solution is an optimal solution selecting the key row? What might happen without it?
or not. How would you proceed to change the basic feasible 15. What conditions must exist in a simplex table to establish the
solution in case it is not optimal? existence of an alternative solution? No feasible solution? Un-
6. Explain the meaning of basic feasible solution and degenerate bounded solution? Degeneracy?
solution in a linear programming problem.

SELF PRACTICE PROBLEMS B

1. Solve the following LP problems using the simplex method.
(i) Max Z = 3x1 + 2x2 (ii) Max Z = 5x1 + 3x2 (iii) Max Z = 3x1 + 2x2 + 5x3
subject to x1 + x2  4 subject to x1 + x2  2 subject to x1 + 2x2 + x3  430
x1 – x2  2 5x1 + 2x2  10 3x1 + 2x3  460
and x 1, x 2  0 3x1 + 8x2  12 x1 + 4x3  420
and x 1, x 2  0 and x1, x2, x3  0
[Shivaji, MSc (Maths), 1995]
(iv) Max Z = x1 – 3x2 + 2x3 (v) Max Z = 4 x1 + 5x2 + 9x3 + 11x4 (vi) Max Z = x1 + x2 + x3
subject to 3x1 – x2 + 3x3  7 subject to subject to 4x1 + 5x2 + 3x3  15
–2x1 + 4x2  12 x1 + x2 + x3 + x4  15 10x1 + 7x2 + x3  12
– 4x1 + 3x2 + 8x3  10 7x1 + 5x2 + 3x3 + 2x4  120 and x1, x2, x3  0
and x1, x2, x3  0 3x1 + 5x2 + 10x3 + 15x4  100
[Meerut, MSc (Maths), 1997; and x1, x2, x3, x4  0
Bhathiar MSc (Maths), 1996]
140 Operations Research: Theory and Applications

(vii) Max Z = x1 + x2 + x3 (viii) Max Z = 2x1 + 4x2 + 3x + x4 (i x ) Max Z = 107x1 + x2 + 2x3

subject to 3x1 + 2x2 + x3  3 subject to x1 + 3x2 + x4  4 subject to 14x1 + x2 – 6x3 + 3x4  7
2x1 + x2 + 2x3  2 2x1 + x2  3 16x1 + 0.5x2 + 6x3  5
and x1, x2, x3  0 x2 + 4x3 + x4  3 3x1 – x2 – x3  10
and x1, x2, x3, x4  0 and x1, x2, x3, x4  0
[Sambalpur, MSc (Maths), 2003 ]
(x) Max Z = 4x1 + 10x2 (xi) Max Z = x1 – x2 + 3x3 (xii) Max Z = 4x1 + x2 + 3x3 + 5x4
subject to 2x1 + x2  50 subject to x1 + x2 + x3  10 subject to
2x1 + 5x2  100 2x1 – x3  2 4x1 – 6x2 – 5x3 + 4x4  – 20
2x1 + 3x2  90 2x1 – 2x2 + 3x3  0 3x1 – 2x2 + 4x3 + x4  10
and x 1, x 2  0 and x1, x2  0 8x1 – 3x2 + 3x3 + 2x4  20
[Karnataka, MSc (Maths), 2004 ] and x1, x2, x3, x4  0
has an unbound solution.
2. Use the two-phase method to solve the following LP problems:
(i) Max Z = 3x1 – x2 (ii) Min Z = 3x1 – x2 – x3 (iii) Min Z = 7.5x1 – 3x2
subject to 2x1 + x2  2 subject to x1 – 2x2 + x3  11 subject to 3x1 – x2 – x3  3
x1 + 3x2  2 – 4x1 + x2 + 2x3 3 x1 – x2 + x3  2
x2  4 – 2x1 + x3 = 1 and x1, x2, x3  0
and x 1, x 2  0 and x 1, x 2, x 3  0
(iv) Min Z = 3x1 – x2 (v) Min Z = 5x1 + 8x2 (vi) Min Z = 3x1 + 2x2
subject to 2x1 + x2  2 subject to 3x1 + 2x2  3 subject to 2x1 + x2  2
x1 + 3x2  2 x1 + 4x2  4 3x1 + 4x2  12
x2  4 x1 + x2  5 and x1, x2  0
and x 1, x 2  0 and x 1, x 2  0
3. Use penalty (Big-M) method to solve the following LP problems:
(i) Min Z = 3x1 – x2 (ii) Min Z = 2x1 + x2 (iii) Max Z = 3x1 + 2x2 + 3x3
subject to 2x1 + x2  2 subject to 3x1 + x2 = 3 subject to 2x1 + x2 + x3  2
x1 + 3x2  3 4x1 + 3x2  6 3x1 + 4x2 + 2x3  8
x2  4 x1 + 2x2  4 and x1, x2, x3  0
x 1, x 2  0 x 1, x 2  0
(i v) Max Z = x1 + x2 + x4 (v) Min Z = x1 – 3x2 + 2x3 (vi) Min Z = 5x1 + 2x2 + 10x3
subject to x1 + x2 + x3 + x4 = 4 subject to 3x1 – x2 + 2x3  7 subject to x1 – x3  10
x1 + 2x2 + x3  x = 4 –2x1 + 4x2 +  12 x2 + x3  10
x1 + 2x2 + x3 = 4 – 4x1 + 3x2 + 8x3  10 and x1, x2, x3  0
and x1, x2, x3  0
[Meerut MSc (Maths), 2005]
4. Solve the following LP problems and remove the complication (if any).
(i) Max Z = 2x1 + 3x2 + 10x3 (ii) Max Z = 5x1 – 2x2 + 3x3 (iii) Max Z = 5x1 + 3x2
subject to x1 + 2x3 = 2 subject to 2x1 + 2x2 – x3  2 subect to x1 + x2  2
x2 + x3 = 1 3x1 – 4x2  3 5x1 + 2x2  10
and x1, x2, x3  0 x2 + 3x3  5 3x1 + 8x2  12
and x 1, x 2, x 3  0 and x1, x2  0
(iv) Max Z = 22x1 + 30x2 + 25x3 (v) Max Z = 8x2
subject to 2x1 + 2x2 +  100 subject to x1 – x2  0
2x1 + x2 + x3  100 2x1 + 3x2  –6
x1 + 2x2 + 2x3  100 and x1, x2 unrestricted.
and x1, x2, x3  0
5. Solve the following LP problems to show that these have alternative optimal solutions.
(i) Max Z = 6x1 + 3x2 (ii) Min Z = 2x1 + 8x2 (iii) Max Z = x1 + 2x2 + 3x3 – x4
subject to 2x1 + x2  8 subject to 5x1 + x2  10 subject to x1 + 2x2 + 3x3 = 15
3x1 + 3x2  18 2x1 + 2x2  14 2x1 + x2 + 5x3  20
x2  3 x1 + 4x2  12 x1 + x2 + x3 + x4  10
and x 1, x 2  0 and x 1, x 2  0 and x1, x2, x3, x4  0
 Linear Programming: The Simplex Method 141

6. Solve the following LP problems to show that these have an unbounded solution.
(i) Max Z = – 2x1 + 3x2 (ii) Max Z = 3x1 + 6x2 (iii) Max Z = 107x1 + x2 + 2x3
subject to x1  5 subject to 3x1 + 4x2  12 subject to 14x1 + x2 – 6x3 + 3x4 = 7
2x1 – 3x2  6 – 2x1 + x2  4 16x1 + 0.5x2 – 6x3  5
and x 1, x 2  0 and x 1, x 2  0 3x1 – x2 – x3  0
and x1, x2  0
(iv) Max Z = 6x1 – 2x2
subject to 2x1 – x2  2
x1  4
and x1, x2, x3, x4  0
7. Solve the following LP problems to show that these have no feasible solution.
(i) Max Z = 2x1 + 3x2 (ii) Max Z = 4x1 + x2 + 4x3 + 5x4 (iii) Max Z = x1 + 3x2
subject to x1 – x2  4 subject to 4x1 + 6x2 – 5x3 + 4x4  – 20 subject to x1 – x2  1
x 1 + x2  6 3x1 – 2x2 + 4x3 + x4  10 3x1 – x2  – 3
x1  2 3x1 – 2x2 + 4x3 + x4  10 and x 1, x 2  0
and x 1, x 2  0 8x1 – 3x2 – 3x3 + 2x4  20
and x1, x2, x3, x4  0
(iv) Max Z = 3x1 + 2x2
subject to 2x1 + x2  2
3x1 + 4x2  12
and x1, x2  0
[Meerut, MSc (Maths), 2006]

HINTS AND ANSWERS

1. (i) x1 = 3, x2 = 1 and Max Z = 11 (ii) x1 = 3/5, x2 = 6/5 and Min Z = 12/5

(ii) x1 = 2, x2 = 0 and Max Z = 10 (iii) All cj – zj  0 artificial variable A1 = 0 appears in the basis
(iii) x1 = 0, x2 = 100, x3 = 230 and Max Z = 1,350 with zero value. Thus an optimal solution to the given LP
(iv) Max Z * = – x1 + 3x2 – 2x3 where Z * = – Z problem exists.
x1 = 4, x2 = 5, x3 = 0 and Z * = 11. (iv) Introduce artificial variable only in the third constraint.
(v) x1 = 50/7, x2 = 0, x3 = 55/7 and Max Z = 695/7 x1 = 0, x2 = 0, x3 = 0, x5 = 0 and Max Z = 4.
(vi) x1 = 0, x2 = 0, x3 = 5 and Max Z = 5 (v) x1 = 4, x2 = 5 and Min Z = – 11
(vii) x1 = 0, x2 = 0, x3 = 1 and Max Z = 3 (vi) x1 = 0, x2 = 10, x3 = 0 and Min Z = 20
(viii) x1 = 1, x2 = 1, x3 = 1/2 and Max Z = 13/2 4. (i) x1 = 0, x2 = 1, x3 = 0 and Max Z = 3
(ix) Divide the first equation by 3 (coefficient of x4 ) (ii) Degeneracy occurs at the initial stage. One of the variable
(x) x1 = 0, x2 = 0 and Max Z = 200 eligible to leave the basis is artificial variable, therefore,
there is no need of resolving degeneracy. Remove the
(xi) x1 = 0, x2 = 6, x3 = 4 and Max Z = 6
artificial variable from the basis.
2. (i) x1 = 1, x2 = 0, and Max Z = 6
x1 = 23/3, x2 = 5, x3 = 0 and Max Z = 85/3
(ii) All cj – zj  0 but Z = – 5/4 (< 0) and artificial variable (iii) x1 = 2, x2 = 0 and Max Z = 10
A1 = 5/4 appears in the basis with positive value. Thus
(iv) x1 = 100/3, x2 = 50/3, x3 = 50/3 and Max Z = 1,650
the given LP problem has no feasible solution.
(iii) x = 5/4, x = 0, x = 0 and Min Z = 75/8 (v) x1 = 6/5 or x1 = – 6/5 or x 2 = 6/5 or x2 = – 6/5 and

1 2 3 Max Z = – 48/5.
(iv) x1 = 2, x2 = 0 and Max Z = 6 5. (i) (a) x1 = 4, x2 = 0 and Max Z = 24 (b)
(v) x1 = 0, x2 = 5 and Max Z = 40 x1 = 5/2, x2 = 3, and Max Z = 24
(vi) All cj – zj  0, but Z = – 4 (< 0) and artificial variable (ii) (a) x1 = 32/6, x2 = 10/6 and Min Z = 24
A1 = 4 appears in the basis with a positive value. Thus (b) x1 = 12, x2 = 0 and Min Z = 24
the given LP problem has no feasible solution. 6. (i) At the current solution: x1 = 5, x2 = 9 and Max Z = 15,it
3. (i) x1 = 3, x2 = 0 and Max Z = 9 may be observed that c2 – z2 = 3/2 but all elementsin
the second column are negative. Solution is unbounded.
142 Operations Research: Theory and Applications

(ii) At the current solution: x2 = 4, s1 = 4 and Max Z = 24, (iv) Optimal solution: x1 = 4, x2 = 6 and Max Z = 12. Since
it may be observed that c2 – z2 = 15 but all elements in the in the initial simplex table all the elements are negative
second column are negative. Solution is unbounded. in the second column, the feasible solution is unbounded
(iii) At second best solution, c3 – z3 = 113/3 but all elements but the optimal solution is bounded.
in the third column are negative. Solution is unbounded; 7. (i) x1 = 2, x2 = 0, A1 = 2 and M Max Z = 9 – 4M; Infeasible
x1 = 0, x4 = 7/3, s1 = 5 and Max Z = 0. solution.
(iv) x1 = 0, x2 = 2, A1 = 2 and Max Z = 4 – 4M; Infeasible
solution.

CHAPTER SUMMARY
The simplex method is an interactive procedure for reaching the optimal solution to any LP problem. It consists of a series of
rules that, in effect, algebraically examine corner (extreme) points of the solution space in a systematic way. Each step moves
towards the optimal solution by increasing profit or decreasing cost, while maintaining feasibility. The simplex method consists
of five steps: (i) identifying the pivot column, (ii) identifying the pivot row and key element, (iii) replacing the pivot row, (iv)
computing new values for each remaining row; and (v) computing the zj and cj – zj values and examining for optimality. Each step
of this iterative procedure is displayed and explained in a simplex table both for maximization and minimization LP problems.

CHAPTER CONCEPTS QUIZ

True or False

1. The major difference between slack and artificial variables is that 15. The value for the replacing row must be ————— before
an artificial can never be zero. computing the values for the ————— rows.
2. If an optimal solution is degenerate, then there are alternate 16. Entries in the cj – zj rows are known as ————— costs.
optimal solutions of the LP problem. 17. Optimality is indicated for a maximization problem when all
3. An infeasible solution is characterized as one where one elements in the cj – zj rows are —————, while for a
constraint is violated. minimization problem all elements must be —————.
4. If the objective function coefficient in the cj row above an artificial 18. In Big-M method, ————— basic feasible solution is obtained
variable is – M, then the problem is a minimization problem. by assigning ————— value to the original value.
19. In the two-phase method, an ————— variable is never
5. In the simplex method, the initial solution contains only slack
considered for re-entry into the basis.
variable in the product mix.
20. ————— occurs when there is no finite solution in the LP
6. If there is a tie between decision variable and a slack (or surplus)
problem.
variable, then select the decision variable to enter into the first
basis.
Multiple Choice
7. An optimal solution to the maximized LP problem is reached if
all cj – zj  0. 21. The role of artificial variables in the simplex method is
8. Variables which can assume negative, positive or zero value are (a) to aid in finding an initial solution
called unrestricted variables. (b) to find optimal dual prices in the final simplex table
9. All the rules and procedures of the simplex method are identical (c) to start phases of simplex method
whether solving a maximization or minimization LP problem. (d) all of the above
10. Artificial variables are added to a linear programming problem to 22. For a maximization problem, the objective function coefficient for
aid in the finding an optimal solution. an artificial variable is
(a) + M (b) – M
(c) Zero (d) none of the above
Fill in the Blanks 23. If a negative value appears in the solution values (xB )
column of the simplex table, then
11. In the simplex method, the ————— column contains the
(a) the solution is optimal
variables which are currently in the solution; the values of these (b) the solution is infeasible
variables can be read from the ————— column.
(c) the solution is unbounded
12. A ————— variable represents amounts by which solution (d) all of the above
values exceed a resource. 24. At every iteration of the simplex method, for minimization problem,
13. The simplex method examines the ————— points in a a variable in the current basis is replaced with another variable that
systematic manner, repeating the same set of steps of the has
algorithm until an ————— solution is reached.
(a) a positive cj – zj value
14. ————— occurs when there is no solution that satisfies all of (b) a negative cj – zj value
the constraints in the linear programming problem. (c) cj – zj = 0
(d) none of the above
 Linear Programming: The Simplex Method 143

25. In the optimal simplex table, cj – zj = 0 value indicates (d) none of the above
(a) unbounded solution (b) cycling 31. If any value in xB – Column of final simplex table is negative, then
(c) alternative solution (d) infeasible solution the solution is
26. For maximization LP model, the simplex method is terminated (a) unbounded (b) infeasible
when all values (c) optimal (d) none of the above
(a) cj – zj  0 (b) cj – zj  0 32. If all aij values in the incoming variable column of the simplex
(c) cj – zj = 0 (d) zj  0 table are negative, then
27. A variable which does not appear in the basic variable (B) (a) solution is unbounded (b) there are multiple solutions
column of simplex table is (c) there exist no solution (c) the solution is degenerate
(a) never equal to zero (b) always equal to zero 33. If an artificial variable is present in the ‘basic variable’ column
(c) called a basic variable (d) none of the above of optimal simplex table, then the solution is
(a) infeasible (b) unbounded
28. If for a given solution a slack variable is equal to zero then
(c) degenerate (d) none of the above
(a) the solution is optimal (b) the solution is infeasible
(c) the entire amount of resource with the constraint in which 34. The per unit improvement in the solution of the minimization
the slack variable appears has been consumed. LP problem is indicated by the negative value of
(d) all of the above (a) cj – zj in the decision-variable column
29. If an optimal solution is degenerate, then (b) cj – zj in the slack variable column
(a) there are alternative optimal solutions (c) cj – zj in the surplus variable column
(b) the solution is infeasible (d) none of the above
(c) the solution is of no use to the decision-maker
35. To convert  inequality constraints into equality constraints, we
(d) none of the above
must
30. To formulate a problem for solution by the simplex method, we (a) add a surplus variable
must add artificial variable to (b) subtract an artificial variable
(a) only equality constraints (c) substract a surplus variable and an artificial variable
(b) only ‘greater than’ constraints (d) add a surplus variable and substract an artificial variable
(c) both (a) and (b)

Answers to Quiz
1. F 2. F 3. T 4. F 5. F 6. T 7. T 8. T 9. F 10. T
11. product mix; quantity 12. Suplus 13. Extreme; optimal 14. Infeasibility 15. Computed; remaining
16. reduced 17. non positive, non negative 18. Initial; zero 19. Artificial 20. Unboundedness
21. (d) 22. (b) 23. (b) 24. (b) 25. (c) 26. (a) 27. (b) 28. (c) 29. (d) 30. (c)
31. (b) 32. (a) 33. (a) 34. (d) 35. (c)

CASE STUDY
Case 4.1: New Bharat Bank
New Bharat Bank is planning its operations for the next year. The bank makes five types of loans. These are listed
below, together with the annual return (in per cent);

Type of Loan Annual Return (per cent)

Education 15
Furniture 12
Automobile 9
Home Construction 10
Home repair 7

Legal requirements and bank policy place the following limits on the amounts of the various types of loans:
Education loans cannot exceed 10 per cent of the total amount of loans. The amount of education and furniture loans
together cannot exceed 20 per cent of the total amount of loans. Home construction loan must be at least 40 per cent
of the total amount of home loans and at least 20 per cent of the total amount of loans. Home construction loan must
not exceed 25 per cent of the total amount of loans.
The bank wishes to maximize its revenue from loan interest, subject to the above restrictions. The bank can lend
a maximum of Rs 75 crore. You are expected to suggest to the management of the bank for maximizing revenue keeping
in view operational restrictions.
144 Operations Research: Theory and Applications

Case 4.2: Accurate Transformers

Accurate Transformers produces electric transformers. The company has orders for transformer for the next six months.
The cost of manufacturing a transformer is expected to somewhat vary over the next few months due to expected changes
in materials costs and in labor rates. The company can produce up to 50 units per month in regular timeand up to
an additional 20 units per month during overtime. The costs for both regular and overtime production are shown in the
table below:

Month
Jan. Feb. Mar. April May June

Orders (units) 58 36 34 69 72 43
Cost per unit at 180 170 170 185 190 190
regular time (in '000 Rs)
Cost per unit at 200 190 190 210 220 220
overtime (in '000 Rs)

The cost of carrying an unsold transformer in stock is Rs 25,000 per month. As on January 1 the company has
15 transformers in stock on and it wishes to have no less than 5 in stock on June 30. Suggest to the management
of the company to determine the optimal production schedule.

Case 4.3: Nationwide Air lines*

Nationwide Airlines, faced with a sharply escalating cost of jet fuel, is interested in optimizing its purchase of jet
fuel at its various locations around the country. Typically, there is some choice concerning the amount of fuel that
can be placed on board any aircraft for any flight segment, as long as minimum and maximum limits are not violated.
The flight schedule is considered as a chain of flight segments, or legs, that each aircraft follows. The schedule ultimately
returns the aircraft to its starting point, resulting in a ‘rotation’. Consider the following rotation: Delhi
– Hyderabad – Cochin – Chennai – Delhi
Fuel Requirements and Limits (1,000 gallons unless specified otherwise)

City Flight Sequence Minimum Fuel Maximum Fuel Regular Fuel Additional Fuel
Required Allowed Consumption if Burned per
Minimum Fuel Gallon of
Boarded Tankered Fuel
(i.e. fuel above
minimum Price per
– in gallons) Gallons (Rs)

1. Delhi to Hyderabad 23 33 12.1 0.040 8,200

2. Hyderabad to Cochin 8 19 2.0 0.005 7,500
3. Cochin to Chennai 19 33 9.5 0.025 7,700
4. Chennai to Delhi 25 33 13.0 0.045 8,900

The fuel for any one of these flight segments may be bought at its departure city, or it may be purchased at a previous
city in the sequence and ‘tankered’ for the flight. Of course, it takes fuel to carry fuel, and thus an economic trade-off
between purchasing fuel at the lowest-cost location and tankering it all around the country must be made.
In the table above the column ‘Regular Fuel Consumption’ takes into account fuel consumption if the minimum
amount of fuel is on board, and the column ‘Additional Fuel Burned’ indicates the additional fuel burned in each flight
segment per gallon of ‘tankered’ fuel carried; tankered fuel refers to fuel above the minimum amount.
The fuel originally carried into Delhi should equal fuel carried into Delhi on the next rotation, in order for the system
to be in equilibrium.
If l1= Leftover fuel inventory coming into city i (1,000 gallons) and xi = Amount of fuel purchased at city i (1,000
gallons) then, (li + xi) is the amount of fuel on board the aircraft when it departs city i.
Suggest the optimal fuel quantity to be purchased by the Airlines.

* This case is based on ‘D. Wayne Darnell and Carln Loflin, National Airlines Fuel Management and Allocation
Model, Interfaces’ Feburary, 1977.
 C h a p t e r 9
“We want these assets to be productive. We buy them. We own them. To say we care only
about the short term is wrong. What I care about is seeing these assets in the best hands.”
– Carl Icahn

The structure of transportation problem involves a large number of shipping routes from several supply
centres to several demand centres. The objective is to determine the number of units of an item (commodity
or product) that should be shipped from an origin to a destination in order to satisfy the required quantity of
goods or services at each destination centre.

LEARNING OBJECTIVES

After studying this chapter, you should be able to

 recognize and formulate a transportation problem involving a large number of shipping routes.
 drive initial feasible solution using several methods.
 drive optimal solution by using Modified Distribution Method.
 handle the problem of degenerate and unbalanced transportation problem.
 examine multiple optimal solutions, and prohibited routes in the transportation problem.
 construct the initial transportation table for a trans-shipment problem.
 solve a profit maximization transportation problem using suitable changes in the transportation algorithm.

CHAPTER OUTLINE
9.1 Introduction 9.6 Variations in Transportation Problem
9.2 Mathematical Model of Transportation Problem 9.7 Maximization Transportation Problem
9.3 The Transportation Algorithm 9.8 Trans-shipment Problem
9.4 Methods of Finding Initial Solution  Conceptual Questions C
 Conceptual Questions A  Self Practice Problems C
 Self Practice Problems A  Hints and Answers
 Hints and Answers  Chapter Summary
9.5 Test for Optimality  Chapter Concepts Quiz
 Conceptual Questions B  Case Study
 Self Practice Problems B  Appendix: Theorem and Results
 Hints and Answers



Transportation Problem 257

9.1 INTRODUCTION
One important application of linear programming is in the area of physical distribution (transportation) of
goods and services from several supply centres to several demand centres. A transportation problem when
expressed in terms of an LP model can also be solved by the simplex method. However a transportation
problem involves a large number of variables and constraints, solving it using simplex methods takes a
long time. Two transportation algorithms, namely Stepping Stone Method and the MODI (modified
distribution) Method have been developed for solving a transportation problem.
The structure of transportation problem involves a large number of shipping routes from several supply
centres to several demand centres. Thus, objective is to determine shipping routes between supply centres
and demand centres in order to satisfy the required quantity of goods or services at each destination centre,
with available quantity of goods or services at each supply centre at the minimum transportation cost and/
or time.
The transportation algorithms help to minimize the total cost of transporting a homogeneous commodity The study of
(product) from supply centres to demand centres. However, it can also be applied to the maximization of total transportation
value or utility. problem helps to
identify optimal
There are various types of transportation models and the simplest of them was first presented by transportation routes
F L Hitchcock (1941). It was further developed by T C Koopmans (1949) and G B Dantzig (1951). Several along with units of
extensions of transportation models and methods have been subsequently developed. commodity to be
shipped in order to
minimize total
9.2 MATHEMATICAL MODEL OF TRANSPORTATION PROBLEM transportation cost.

Let us consider Example 9.1 to illustrate the mathematical model formulation of transportation problem of
transporting a single commodity from three sources of supply to four demand destinations. The sources
of supply are production facilities, warehouses, or supply centres, each having certain amount of commodity
to supply. The destinations are consumption facilities, warehouses or demand centres each having certain
amount of requirement (or demand) of the commodity.

Example 9.1 A company has three production facilities S1, S2 and S3 with production capacity of 7,
9 and 18 units (in 100s) per week of a product, respectively. These units are to be shipped to four warehouses
D1, D2, D3 and D4 with requirement of 5, 6, 7 and 14 units (in 100s) per week, respectively. The transportation
costs (in rupees) per unit between factories to warehouses are given in the table below:

D1 D2 D3 D4 Supply
(Availability)

S1 19 30 50 10 7

S2 70 30 40 60 9

S3 40 8 70 20 18

Demand 5 8 7 14 34
(Requirement)

Formulate this transportation problem as an LP model to minimize the total transportation cost.

Model formulation Let xij = number of units of the product to be transported from a production facility
i (i = 1, 2, 3) to a warehouse j ( j = 1, 2, 3, 4)
The transportation problem is stated as an LP model as follows:
Minimize (total transportation cost) Z = 19x11 + 30x12 + 50x13 + 10x14 + 70x21 + 30x22 + 40x23
+ 60x24 + 40x31 + 8x32 + 70x33 + 20x34

subject to the constraints

x11 + x12 + x13 + x14 = 7 
x21 + x + x + x = 9  (Supply)
22 23 24 

x31 + x32 + x33 + x34 = 18 





 258 Operations Research: Theory and Applications

x11 + x21 + x31 = 5 

x12 + x22 + x32 = 8 
x13 + x + x = 7 
(Demand)
23 33

x14 + x24 + x34 = 14 

and xij  0 for i = 1, 2, 3 and j = 1, 2, 3, and 4.
In the above LP model, there are m × n = 3 × 4 = 12 decision variables, xij and m + n = 7 constraints,
where m are the number of rows and n are the number of columns in a general transportation table.

9.2.1 General Mathematical Model of Transportation Problem

Let there be m sources of supply, S1, S2, . . ., Sm having ai (i = 1, 2, . . ., m) units of supply (or capacity),
respectively to be transported to n destinations, D1, D2, . . ., Dn with bj ( j = 1, 2, . . ., n)units of demand
(or requirement), respectively. Let cij be the cost of shipping one unit of the commodity from source i to
destination j. If xij represents number of units shipped from source i to destination j, the problem is to
determine the transportation schedule so as to minimize the total transportation cost while satisfying the
supply and demand conditions. Mathematically, the transportation problem, in general, may be stated
as follows:
m n
Minimize (total cost) Z =   cij xij (1)
i 1 j 1

subject to the constraints

n
 xij  ai , i = 1, 2, . . ., m (supply constraints) (2)
j 1
m

 xij  bj , j = 1, 2, . . ., n (demand constraints) (3)

i 1

and xij  0 for all i and j. (4)

For easy presentation and solution, a transportation problem data is generally presented as shown
in Table 9.1.
Transportation
table is a Existence of feasible solution A necessary and sufficient condition for a feasible solution to the
convenient way to transportation problems is:
summarize data.
Total supply = Total demand
m n
 ai =  bj (also called rim conditions)
i 1 j 1
For proof, see Appendix at the end of this chapter.

To D1 D2 .. . Dn Supply
From ai
c11 c12 .. . c1n
S1 x11 x12 x1n a1

c21 c22 .. . c2n

S2 x21 x22 x2n a2

. . . . .
cm1 cm2 .. . cmn
Sm xm1 xm2 xmn am
Table 9.1
General m n
Transportation Demand b1 b2 .. . bn  ai =  bj
Table. bj i 1 j 1
 Transportation Problem 259

In this problem, there are (m + n) constraints, one for each source of supply, and distinction and
m × n variables. Since all (m + n) constraints are equations, therefore, one of these equations is extra
(redundant). The extra constraint (equation) can be derived from the other constraints (equations), without
affecting the feasible solution. It follows that any feasible solution for a transportation problem must have
exactly (m + n – 1) non-negative basic variables (or allocations) xij satisfying the rim conditions.
Remarks 1. When the total supply is equal to the total demand, the problem is called a balanced
transportation problem, otherwise it is called an unbalanced transportation problem. The unbalanced
transportation problem can be made balanced by adding a dummy supply centre (row) or a dummy
demand centre (column) as the need arises.
2. When the number of positive allocations (values of decision variables) at any stage of the feasible When total demand
solution is less than the required number (rows + columns – 1), i.e. number of independent constraint equals total supply,
equations, the solution is said to be degenerate, otherwise non-degenerate. For proof, see Appendix the transportation
problem is said to
at the end of this chapter.
be balanced.
3. Cells in the transportation table having positive allocation, i.e., xij > 0 are called occupied cells,
otherwise are known as non-occupied (or empty) cells.

9.3 THE TRANSPORTATION ALGORITHM

The algorithm for solving a transportation problem may be summarized into the following steps:

Step 1: Formulate the problem and arrange the data in the matrix form The formulationof
the transportation problem is similar to the LP problem formulation. In transportation problem, the objective
function is the total transportation cost and the constraints are the amount of supply and demand available at
each source and destination, respectively.
Step 2: Obtain an initial basic feasible solution In this chapter, following three different methods
are discussed to obtain an initial solution:
 North-West Corner Method,
 Least Cost Method, and
 Vogel’s Approximation (or Penalty) Method.
The initial solution obtained by any of the three methods must satisfy the following conditions:
(i) The solution must be feasible, i.e. it must satisfy all the supply and demand constraints (also called
rim conditions).
(ii) The number of positive allocations must be equal to m + n – 1, where m is the number of rows
and n is the number of columns.
Any solution that satisfies the above conditions is called non-degenerate basic feasible solution,
otherwise, degenerate solution.
Step 3: Test the initial solution for optimality In this chapter, the Modified Distribution (MODI)
method is discussed to test the optimality of the solution obtained in Step 2. If the current solution is optimal,
then stop. Otherwise, determine a new improved solution.
Step 4: Updating the solution Repeat Step 3 until an optimal solution is reached.

9.4 METHODS OF FINDING INITIAL SOLUTION

There are several methods available to obtain an initial basic feasible solution. In this chapter, we shall
discuss only following three methods:

9.4.1 North-West Corner Method (NWCM)

This method does not take into account the cost of transportation on any route of transportation. The
method can be summarized as follows:
Step 1: Start with the cell at the upper left (north-west) corner of the transportation table (or matrix) and
allocate commodity equal to the minimum of the rim values for the first row and first column, i.e. min (a1, b1).
260 Operations Research: Theory and Applications

Step 2: (a) If allocation made in Step 1 is equal to the supply available at first source (a1, in first row), then
move vertically down to the cell (2, 1), i.e., second row and first column. Apply Step 1 again, for next
allocation.
(a) If allocation made in Step 1 is equal to the demand of the first destination (b1 in first column),
then move horizontally to the cell (1, 2), i.e., first row and second column. Apply Step 1 again for next
allocation.
(b) If a1 = b1, allocate x11 = a1 or b1 and move diagonally to the cell (2, 2).
Step 3: Continue the procedure step by step till an allocation is made in the south-east corner cell of
the transportation table.
Remark If during the process of making allocation at a particular cell, the supply equals demand, then the
next allocation of magnitude zero can be made in a cell either in the next row or column. This condition is
known as degeneracy.
Example 9.2 Use North-West Corner Method (NWCM) to find an initial basic feasible solution to the
transportation problem using data of Example 9.1
Solution The cell (S1, D1) is the north-west corner cell in the given transportation table. The rim values for
row S1 and column D1 are compared. The smaller of the two, i.e. 5, is assigned as the first allocation; otherwise
it will violate the feasibility condition. This means that 5 units of a commodity are to be transported from
source S1 to destination D1. However, this allocation leaves a supply of 7 – 5 = 2 unitsof commodity at
S 1.
Move horizontally and allocate as much as possible to cell (S1, D2). The rim value for row S1 is 2 and for
column D2 is 8. The smaller of the two, i.e. 2, is placed in the cell. Proceeding to row S2, since the demandof
D1 is fullfilled. The unfulfilled demand of D2 is now 8 – 2 = 6 units. This can be fulfilled by S2 with capacity
of 9 units. So 6 units are allocated to cell (S2, D2). The demand of D2 is now satisfied and a balanceof 9 – 6
= 3 units remains with S2.

D1 D2 D3 D4 Supply

19 30 50 10
S1 5 2 7

70 30 40 60
S2 6 3 9

40 8 70 20
S3 4 14 18
Table 9.2
Initial Solution
using NWCM Demand 5 8 7 14 34

Continue to move horizontally and vertically in the same manner to make desired allocations. Once the
procedure is over, count the number of positive allocations. These allocations (occupied cells) should be
equal to m + n – l = 3 + 4 – l = 6. If yes, then solution is non-degenerate feasible solution. Otherwise
degenerate solution.
The total transportation cost of the initial solution is obtained by multiplying the quantity xij in the
occupied cells with the corresponding unit cost cij and adding all the values together. Thus, the total
transportation cost of this solution is
Total cost = 5 × 19 + 2 × 30 + 6 × 30 + 3 × 40 + 4 × 70 + 14 × 20 = Rs 1,015

9.4.2 Least Cost Method (LCM)

Since the main objective is to minimize the total transportation cost, transport as much as possible
through those routes (cells) where the unit transportation cost is lowest. This method takes into account the
minimum unit cost of transportation for obtaining the initial solution and can be summarizedas follows:
 Transportation Problem 261

Step 1: Select the cell with the lowest unit cost in the entire transportation table and allocate as much
as possible to this cell. Then eliminate (line out) that row or column in which either the supply or demand
is fulfilled. If a row and a column are both satisfied simultaneously, then crossed off either a row or a column.
In case the smallest unit cost cell is not unique, then select the cell where the maximum allocation can
be made.
Step 2: After adjusting the supply and demand for all uncrossed rows and columns repeat the procedure to
select a cell with the next lowest unit cost among the remaining rows and columns of the transportation table
and allocate as much as possible to this cell. Then crossed off that row and column in which either supply or
demand is exhausted.
Step 3: Repeat the procedure until the available supply at various sources and demand at various destinations
is satisfied. The solution so obtained need not be non-degenerate.
Example 9.3 Use Least Cost Method (LCM) to find initial basic feasible solution to the transportation
problem using data of Example 9.1.
Solution The cell with lowest unit cost (i.e., 8) is (S3, D2). The maximum units which can be allocated
to this cell is 8. This meets the complete demand of D2 and leave l0 units with S3, as shown in Table 9.3.
In the reduced table without column D2, the next smallest unit transportation cost, is 10 in cell
(S1, D4). The maximum which can be allocated to this cell is 7. This exhausts the capacity of S1 and
leaves 7 units with D4 as unsatisfied demand. This is shown in Table 9.3.

D1 D2 D3 D4 Supply

19 30 50 10
S1 7
7
70 30 40 60
S2 9

40 8 70 20
S3 18
8
Table 9.3
Demand 5 8 7 14 34

In Table 9.3, the next smallest cost is 20 in cell (S3, D4). The maximum units that can be allocated to this
cell is 7 units. This satisfies the entire demand of D4 and leaves 3 units with S3, as the remaining supply,shown
in Table 9.4.
In Table 9.4, the next smallest unit cost cell is not unique. That is, there are two cells – (S2, D3) and (S3,
D1) – that have the same unit transportation cost of 40. Allocate 7 units in cell (S2, D3) first because it can
accommodate more units as compared to cell (S3, D1). Then allocate 3 units (only supply left with S3) to cell
(S3, D1). The remaining demand of 2 units of D1 is fulfilled from S2. Since supply and demand at each supply
centre and demand centre is exhausted, the initial solution is arrived at, and is shown in Table 9.4.

D1 D2 D3 D4 Supply

19 30 50 10
S1 7
7
70 30 40 60
S2 9
2 7
40 8 70 20
S3 18
3 8 7

Demand 5 8 7 14 34 Table 9.4

The total transportation cost of the initial solution by LCM is calculated as given below:
Total cost = 7 × 10 + 2 × 70 + 7 × 40 + 3 × 40 + 8 × 8 + 7 × 20 = Rs 814
The total transportation cost obtained by LCM is less than the cost obtained by NWCM.
262 Operations Research: Theory and Applications

9.4.3 Vogel’s Approximation Method (VAM)

Vogel’s approximation (penalty or regret) is preferred over NWCR and LCM methods. In this method, an
allocation is made on the basis of the opportunity (or penalty or extra) cost that would have been incurred
if the allocation in certain cells with minimum unit transportation cost were missed. Hence, allocations are
made in such a way that the penalty cost is minimized. An initial solution obtained by using this method
is nearer to an optimal solution or is the optimal solution itself. The steps of VAM are as follows:
Step 1: Calculate the penalties for each row (column) by taking the difference between the smallest and next
smallest unit transportation cost in the same row (column). This difference indicates the penalty or extra cost
that has to be paid if decision-maker fails to allocate to the cell with the minimum unittransportation cost.
Step 2: Select the row or column with the largest penalty and allocate as much as possible in the cell
that has the least cost in the selected row or column and satisfies the rim conditions. If there is a tie in
the values of penalties, it can be broken by selecting the cell where the maximum allocation can be made.
Step 3: Adjust the supply and demand and cross out the satisfied row or column. If a row and a column are
satisfied simultaneously, only one of them is crossed out and the remaining row (column) is assigneda
zero supply (demand). Any row or column with zero supply or demand should not be used in computing
future penalties.
Step 4: Repeat Steps 1 to 3 until the available supply at various sources and demand at various destinations
is satisfied.
Example 9.4 Use Vagel’s Approximation Method (VAM) to find the initial basic feasible solution to
the transportation problem using the data of Example 9.1.
Solution The differences (penalty costs) for each row and column have been calculated as shown in Table
9.5. In the first round, the maximum penalty, 22 occurs in column D2. Thus the cell (S3, D2) having the least
transportation cost is chosen for allocation. The maximum possible allocation in this cell is 8 units and it
satisfies demand in column D2. Adjust the supply of S3 from 18 to 10 (18 – 8 = 10).

D1 D2 D3 D4 Supply Row differences

S1 19 30 50 10 7 9 9 40 40
5 2

S2 70 30 40 60 9 10 20 20 20
7 2
S3 40 8 70 20 18 12 20 50 –
8 10

Demand 5 8 7 14 34

Column 21 22 10 10
Table 9.5
Initial Solution differences 21 – 10 10
Using VAM – – 10 10
– – 10 50
The new row and column penalties are calculated except column D2 because D2’s demand has been
satisfied. In the second round, the largest penalty, 21 appears at column D1. Thus the cell (S1, D1) having
the least transportation cost is chosen for allocating 5 units as shown in Table 9.5. After adjusting the
supply and demand in the table, we move to the third round of penalty calculations.
In the third round, the maximum penalty 50 appears at row S3. The maximum possible allocation of 10
units is made in cell (S3, D4) that has the least transportation cost of 20 as shown in Table 9.5.
The process is continued with new allocations till a complete solution is obtained. The initial solution
using VAM is shown in Table 9.5. The total transportation cost associated with this method is:
Total cost = 5 × 19 + 2 × 10 + 7 × 40 + 2 × 60 + 8 × 8 + 10 × 20 = Rs 779
 Transportation Problem 263

Example 9.5 A dairy firm has three plants located in a state. The daily milk production at each plant
is as follows:
Plant 1 : 6 million litres, Plant 2 : 1 million litres, and Plant 3 : 10 million litres
Each day, the firm must fulfil the needs of its four distribution centres. The minimum requirement of each
centre is as follows:
Distribution centre 1 : 7 million litres, Distribution centre 2 : 5 million litres,
Distribution centre 3 : 3 million litres, and Distribution centre 4 : 2 million litres
Cost (in hundreds of rupees) of shipping one million litre from each plant to each distribution centre is given
in the following table:

Distribution Centre
D1 D2 D3 D4
P1 2 3 11 7
Plant P2 1 0 6 1
P3 5 8 15 9

Find the initial basic feasible solution for given problem by using following methods:
(a) North-west corner rule
(b) Least cost method
(c) Vogel’s approximation method

Solution (a) North-West Corner Rule

Distribution Centre

D1 D2 D3 D4 Supply

2 3 11 7
P1 6 = a1
6

1 0 6 1
Plant P2 1 = a2
1

5 8 15 9
P3 10 = a3 Table 9.6
5 3 2
Initial Solution by
NWCR
Demand 7 = b1 5 = b2 3 = b3 2 = b4

(i) Comparing a1 and b1, since a1 < b1; allocate x11 = 6. This exhausts the supply at P1 and leaves
1 unit as unsatisfied demand at D1.
(ii) Move to cell (P2, D1). Compare a2 and b1 (i.e. 1 and 1). Since a2 = b1, allocate x21 = 1.
(iii) Move to cell (P3, D2). Since supply at P3, is equal to the demand at D2, D3 and D4, therefore,
allocate x32 = 5, x33 = 3 and x34 = 2.
It may be noted that the number of allocated cells (also called basic cells) are 5 which is one less than
the required number m + n – 1 (3 + 4 – 1 = 6). Thus, this solution is the degenerate solution. The transportation
cost associated with this solution is:
Total cost = Rs (2 × 6 + l × l + 8 × 5 + 15 × 3 + 9 × 2) × 100 = Rs 11,600
264 Operations Research: Theory and Applications

(b) Least Cost Method

Distribution Centre

D1 D2 D3 D4 Supply

2 3 11 7
P1 6
6

1 0 6 1
Plant P2 1
1
Table 9.7
Initial Solution by 5 8 15 9
P3 10
LCM 1 4 3 2

Demand 7 5 3 2

(i) The lowest unit cost in Table 9.7 is 0 in cell (P2, D2), therefore the maximum possible allocation that
can be made is 1 unit. Since this allocation exhausts the supply at plant P2, therefore row 2 is crossed
off.
(ii) The next lowest unit cost is 2 in cell (P1, D1). The maximum possible allocation that can be made
is 6 units. This exhausts the supply at plant P1, therefore, row P1 is crossed off.
(iii) Since the total supply at plant P3 is now equal to the unsatisfied demand at all the four distribution
centres, therefore, the maximum possible allocations satisfying the supply and demand conditions,
are made in cells (P3, D1), (P3, D2), (P3, D3) and (P3, D4).
The number of allocated cells in this case are six, which is equal to the required number
m + n – 1 (3 + 4 – 1 = 6). Thus, this solution is non-degenerate. The transportation cost associated with
this solution is
Total cost = Rs (2 × 6 + 5 × l + 8 × 4 + 15 × 3 + 9 × 2) × 100 = Rs 11,200
(c) Vogel’s Approximation Method: First calculating penalties as per rules and then allocations are made
in accordance of penalties as shown in Table 9.8.

D1 D2 D3 D4 Supply Row penalty

2 3 11 7 6 1 1 5
P1
1 5

1 0 6 1 1 0 – –
Plant P2
1

5 8 15 9 10 3 3 4
P3
6 3 1

Demand 7 5 3 2
Column penalty 1 3 5 6
Table 9.8 3 5 4 2
Initial Solution by 3 – 4 2
VAM Distribution Centre

The number of allocated cells in Table 9.8 are six, which is equal to the required numberm
+ n – 1 (3 + 4 – 1 = 6), therefore, this solution is non-degenerate. The transportation cost associated with
this solution is:
Total cost = Rs (2 × 1 + 3 × 5 + 1 × 1 + 5 × 6 + 15 × 3 + 9 × 1) × 100 = Rs 10,200
Remark: Total transportation cost found by VAM is lower than the costs of transportation determined
by the NWCR and LCM methods. Therefore, it is of advantage to use this method in order to reduce
computational time required to obtain optimum solution.
 Transportation Problem 265

CONCEPTUAL QUESTIONS A
1. Show that all the bases for a transportation problem are 8. What is meant by the triangular form of a system of linear
triangular. equations? When does a system of linear equations have a
2. With reference to a transportation problem define the following triangular basis? (See Appendix for proof.)
terms: 9. What is meant by non-degenerate basic feasible solution of a
(i) Feasible solution (ii) Basic feasible solution transportation problem?
(iii) Optimal solution 10. Explain in brief three, methods of initial feasible solution for
(iv) Non-degenerate basic feasible solution transportation problem.
3. Given a mathematical formulation of the transportation problem 11. Explain the various steps involved in solving transportation problem
and the simplex methods, what are the differences in the nature using (i) Least cost method, and (ii) Vogel’s approximation method.
of problems that can be solved by using these methods?
12. Explain the (i) North-West Corner method, (ii) Least-Cost method,
4. Prove that there are only m + n – 1 independent equations in and (iii) Vogel’s Approximation method, for obtaining an initial basic
a transportation problem, m and n being the number of origins and feasible solution of a transportation problem.
destination, and only one equation can be dropped as being
redundant. (For proof see Appendix). 13. State the transportation problem. Describe clearly the steps
involved in solving it.
5. Describe the transportation problem with its general mathematical
formulation. 14. Is the transportation model an example of decision-making under
certainty or under uncertainty? Why?
6. Show that a transportation problem is a special type of LP problem.
In what areas of management can the transportation model be 15. Why does Vogel’s approximation method provide a good initial
effectively used? Discuss. feasible solution? Can the North-West Corner method ever be able
to provide an initial solution with a cost as low as this?
7. What are the characteristics of transportation problem of linear
programming?

SELF PRACTICE PROBLEMS A

1. Determine an initial basic feasible solution to the following 3. Determine an initial basic feasible solution to the following
transportation problem by using (a) NWCR, (b) LCM and (c) VAM. transportation problem by using (a) NWCM, (b) LCM, and (c)
Destination VAM.
Destination
D1 D2 D3 D4 Supply
D1 D2 D3 D4 Supply
S1 21 16 15 3 11
A 11 13 17 14 250
Source S2 17 18 14 23 13
Source B 16 18 14 10 300
S3 32 27 18 41 19
C 21 24 13 10 400
Demand 6 6 8 23
Demand 200 225 275 250
2. Determine an initial basic feasible solution to the following
transportation problem by using (a) the least cost method, and 4. Determine an initial basic feasible solution to the following
(b) Vogel’s approximation method. transportation problem by using the North-West corner rule, where
Oi and Dj represent ith origin and jth destination, respectively.
Destination
Destination

D1 D2 D3 D4 Supply D1 D2 D3 D4 Supply
S1 1 2 1 4 30 O1 6 4 1 5 14
Source S2 3 3 2 1 30 Source O2 8 9 2 7 16
S3 4 2 5 9 40 4 3 6 2 5
O3
Demand 20 40 30 10 Demand 6 10 15 4

HINTS AND ANSWERS

1. x14 = 11, x21 = 6, x22 = 3, x24 = 4, x32 = 3, x33 = 4, x34 = 12; Total cost = 686.
2. (a) and (b): x11 = 20, x13 = 10, x22 = 20, x33 = 20, x24 = 10, x32 = 20; Total cost = 180.
3. (a) x11 = 200, x12 = 50, x22 = 175, x23 = 125, x33 = 150, x34 = 250; Total cost = 12,200.
(b) x11 = 200, x12 = 50, x22 = 175, x23 = 125, x33 = 150, x34 = 250; Total cost = 12,200.
(c) x11 = 200, x12 = 50, x22 = 175, x24 = 125, x33 = 275, x34 = 125; Total cost = 12,075.
4. x11 = 6; x12 = 8; x22 = 2; x23 = 14; x33 = 1; x34 = 4 ; Total cost = Rs 128.
266 Operations Research: Theory and Applications

9.5 TEST FOR OPTIMALITY

Once an initial solution is obtained, the next step is to check its optimality in terms of feasibility of the solution
and total minimum transportation cost.
The test of optimality begins by calculating an opportunity cost associated with each unoccupied cell
(represents unused route) in the transportation table. An unoccupied cell with the largest negative opportunity
cost is selected to include in the new set of transportation routes (allocations). This value indicates the per
unit cost reduction that can be achieved by making appropriate allocation in the
unoccupied cell. This cell is also known as an incoming cell (or variable). The outgoing cell (or variable)
The negative from the current solution is the occupied cell (basic variable) where allocation will become zero as allocation
opportunity cost is made in the unoccupied cell with the largest negative opportunity cost. Such an exchange reduces the
indicates the per
unit cost reduction
total transportation cost. The process is continued until there is no negative opportunity cost. That is, the
that can be current solution is an optimal solution.
achieved by raising The Modified-distribution (MODI) method (also called u-v method or method of multipliers) is used
the shipment to calculate opportunity cost associated with each unoccupied cell and then improving the current solution
allocation in the
unoccupied cell from
leading to an optimal solution. The steps of MODI method based on the concept of duality are summarized
its present level of in section 9.5.3.
zero.
9.5.1 Dual of Transportation Model
For a given basic feasible solution if we associate numbers (also called dual variables or multipliers) ui
and vj with row i (i = 1, 2, . . ., m) and column j ( j = 1, 2, . . ., n) of the transportation table, respectively, then
ui and vj must satisfy the equation
ui + vj = cij , for each occupied cell (i, j)
These equations yield m + n – 1 equations in m + n unknown dual variables. The values of these
variables can be determined by arbitrarily assigning a zero value to any one of these variables. The value
of the remaining m + n – 2 variables can then be obtained algebraically by using the above equation for
the occupied cells. The opportunity cost of each unoccupied cell (called non-basic variable or unused route)
is calculated by using following equation that involves ui and vj values.:
drs = crs – (ur + vs ) , for each unoccupied cell (r, s)
This equation also indicates the per unit reduction in the total transportation cost for the route (r, s). To
prove these two results, consider the general transportation model:
m n
Minimize Z =   cij xij
i 1 j 1

subject to the constraints

n
 xij = ai , i = 1, 2, . . ., m (Supply)
j 1
Modi method helps m
in comparing the  xij = bj , j = 1, 2, . . ., n (Demand)
relative advantage i 1

of alternative
allocations for all the
and xij  0 for all i and j
unoccupied cells Since all of the constraints are equalities, write each equality constraint equivalent to two inequalities
simultaneously.
as follows:

n 
 xij  a , i = 1, 2, . . ., m 
j 1 i
 (Supply constraints)
n – (x )  – a 

ij i 
j 1 
m
 xij  b , j = 1, 2, . . ., m 
j
i 1 

m – (x )  – b  (Demand constraints)
 ij j 
i 1 

Let u and u be the dual variables, one for each supply constraint i. Similarly v  , v  be the dual
 i i j j
variables one for each demand constraint j. Then, the dual of the transportation model can be written as:
 m n
Z*    

Maximize   ( u i  ui ) ai   ( vj  vj ) bj
i 1 j 1

subject to the constraints

(u   u  )  (v   v  )  c
i i j j ij
and u i , u i , vj  , vj  0 , for all i and j.
The variables u and u that appear in the objective function, may take positive, negative or zero
i i
values. Thus, either of these will appear in the optimal basic feasible solution because one is the negative

of the other. The same argument may be given for vj and vj . Thus, let

ui  ui  ui , i = 1, 2, . . ., m

vj  vj   vj  , j = 1, 2, . . ., n`

The values of ui and vj will then be unrestricted in sign. Hence, the dual of the transportation model can
now be written as
m n
Maximize Z*   ui ai   v j bj
i 1 j 1

subject to the constraints

ui + vj  cij
and ui, vj unrestricted in sign for all i and j.
The relationship (cij – ui – vj) xij = 0 is known as complementary slackness for a transportation problem
and indicates that
(a) if xij > 0 and solution is feasible, then cij – ui – vj = 0 or cij = ui + vj, for each occupied cell,
(b) if xij = 0 and cij > ui + vj, then it is not desirable to have xij > 0 in the solution mix because it would
cost more to transport on a route (i, j),
(c) if cij  ui + vj for some xij = 0, then xij can be brought into the solution mix.

9.5.2 Economic Interpretation of ui’s and vj’s

The ui values measures the comparative advantage of additional unit of supply or shadow price (or value) The dual variables
of available supply at centre i. This may also be termed as location rent. Similarly, the vj values measures the ui s and vj s
comparative advantage of an additional unit of commodity demanded at demand centre j. This may alsobe represent the
shadow price (value
termed as market price.
of the commodity)
Illustration The concept of duality in transportation problem is applied on Example 9.1 in the following manner: for the supply
centres and demand
Reproducing transportation data of Example 9.1 for ready reference in Table 9.9. In Table 9.9, there are centres,
m = 3 rows and n = 4 columns. Let u1, u2 and u3 be dual variables corresponding to each of the supply respectively.
constraint in that order. Similarly, v1, v2, v3 and v4 be dual variables corresponding to each of demand
constraint in that order. The dual problem then becomes
D1 D2 D3 D4 Supply ui

S1 19 30 50 10 7 u1

S2 70 30 40 60 9 u2

S3 40 8 70 20 18 u3

Demand 5 8 7 14 34
vj v1 v2 v3 v4 Table 9.9

Maximize Z = (7u1 + 9u2 + 18u3) + (5v1 + 8v2 + 7v3 + 14v4)

subject to the constraints

(i) u1 + v1  19, (ii) u1 + v2  30, (iii) u1 + v3  50, (iv) u1 + v4  10,

 Transportation Problem 267

(v) u2 + v1  70, (vi) u2 + v2  30, (vii) u2 + v3  40, (viii) u2 + v4  60,

(ix) u3 + v1  40, (x) u3 + v2  8, (xi) u3 + v3  70, (xii) u3 + v4  20,
and ui, vj unrestricted in sign for all i and j.
268 Operations Research: Theory and Applications

Interpretation Consider the dual constraint u1 + v1  19 or v1  19 – u1. This represents the delivered market
value of the commodity at destination D1 which should be less than or equal to the unit cost of transportation
from S1 to D1 minus the per unit value of commodity at D1. A similar interpretation can also be given for
other constraints.
The optimal value of dual variables can be obtained either by simplex method or by reading values of
these variables from the optimal solution of transportation problem. It may be noted that the total
transportation cost at optimal solution would be the same as obtained by putting values of ui’s and vj’s from
optimal solution of transportation problem in the dual objective function:
3 4
Maximize Z   ai ui   bj v j
i 1 j 1

9.5.3 Steps of MODI Method (Transportation Algorithm)

The steps to evaluate unoccupied cells are as follows:
Step 1: For an initial basic feasible solution with m + n – 1 occupied cells, calculate ui and vj for rows and
columns. The initial solution can be obtained by any of the three methods discussed earlier.
To start with, any one of uis or vjs is assigned the value zero. It is better to assign zero to a particular
ui or vj where there are maximum number of allocations in a row or column respectively, as this will reduce
the considerably arithmetic work. The value of uis and vjs for other rows and columns is calculated by using
the relationship.
cij = ui + vj , for all occupied cells (i, j).
Step 2: For unoccupied cells, calculate the opportunity cost by using the relationship
dij = cij – (ui + vj) , for all i and j.
Step 3: Examine sign of each dij
(i) If dij > 0, then the current basic feasible solution is optimal.
(ii) If dij = 0, then the current basic feasible solution will remain unaffected but an alternative solution
Changing the exists.
shipping route
involves adding to (iii) If one or more dij < 0, then an improved solution can be obtained by entering an unoccupied cell
cells on the closed (i, j) into the solution mix (basis). An unoccupied cell having the largest negative value of dij is
path with plus signs chosen for entering into the solution mix (new transportation schedule).
and subtracting from
cells with negative Step 4: Construct a closed-path (or loop) for the unoccupied cell with largest negative value of dij. Start the
signs. closed path with the selected unoccupied cell and mark a plus sign (+) in this cell. Trace a path alongthe
rows (or columns) to an occupied cell, mark the corner with a minus sign (–) and continue down the column
(or row) to an occupied cell. Then mark the corner with plus sign (+) and minus sign (–) alternatively.Close the
path back to the selected unoccupied cell.
Step 5: Select the smallest quantity amongst the cells marked with minus sign on the corners of closed loop.
Allocate this value to the selected unoccupied cell, add it to occupied cells marked with plus signs,and
subtract it from the occupied cells marked with minus signs.
Step 6: Obtain a new improved solution by allocating units to the unoccupied cell according to Step5
and calculate the new total transportation cost.
Step 7: Test optimality of the revised solution. The procedure terminates when all dij  0 for unoccupied
cells.
Remarks 1. The closed-loop (path) starts and ends at the selected unoccupied cell. It consists of successive
horizontal and vertical (connected) lines whose end points must be occupied cells, except anend point
associated with entering unoccupied cell. This means that every corner element of the loop must be an
occupied cell.
It is immaterial whether the loop is traced in a clockwise or anti-clockwise direction and whether it starts
up, down, right or left (but never diagonally). However, for a given solution only one loop can beconstructed
for each unoccupied cell.
2. There can only be one plus (+) sign and only one minus (–) sign in any given row or column.
3. The closed path indicates changes involved in reallocating the shipments.
 Transportation Problem 269

Fig. 9.1
Flow Chart of
MODI Method

The steps of MODI method for solving a transportation problem are summarized in the flow chart shown
in Fig. 9.1.

9.5.4 Close-Loop in Transportation Table and its Properties

Any basic feasible solution must contain m + n – 1 non-zero allocations provided. An ordered set of at
(i) any two adjacent cells of the ordered set lie either in the same row or in the same column, and least four cells in a
(ii) no three or more adjacent cells in the ordered set lie in the same row or column. The first cell of the set must transportation table
forms a loop.
follow the last in the set, i.e. each cell (except the last) must appear only once in the ordered set.
Consider the following two cases represented in Tables 9.10(a) and 9.10(b). In Table 9.10(a), if we join
the positive allocations by horizontal and vertical lines, then a closed loop is obtained. The ordered set
of cells forming a loop is:
L = {(a, 2), (a, 4), (e, 4), (e, 1), (b, 1), (b, 2), (a, 2)}
The loop in Table 9.10(b) is not allowed because it does not satisfy the conditions in the definition
of a loop. That is, the cell (b, 2) appears twice.

Table 9.10

(a) (b)
270 Operations Research: Theory and Applications

Remarks 1. Every loop has an even number of cells and has at least four cells.
2. The allocations are said to be in independent position if it is not possible to increase or decrease any
individual allocation without changing the positions of these allocations, or if a closed loop cannot
be formed through these allocations without violating the rim conditions.
3. Each row and column in the transportation table should have only one plus and minus sign. All cells that
have a plus or a minus sign, except the starting unoccupied cell, must be occupied cells.
4. Closed loops may or may not be in the shape of a square.
Example 9.6 Apply MODI method to obtain optimal solution of transportation problem using the data
of Example 9.1.

D1 D2 D3 D4 Supply

S1 19 30 50 10 7

S2 70 30 40 60 9

S3 40 8 70 20 18

Demand 5 8 7 14 34

Solution Applying Vogel’s approximation method to obtain an initial basic feasible solution. This
solution is shown in Table 9.11 [for ready reference see Table 9.5].
1. In Table 9.11, since number of occupied cells are m + n – 1 = 3 + 4 – 1 = 6 (as required), therefore
this initial solution is non-degenerate. Thus, an optimal solution can be obtained. The total transportation
cost associated with this solution is Rs 779.
2. In order to calculate the values of uis (i = 1, 2, 3) and vjs ( j = 1, 2, 3, 4) for each occupied cell, assigning
arbitrarily, v4 = 0 in order to simplify calculations. Given v4 = 0, u1, u2 and u3 can be immediately
computed by using the relation cij = ui + vj for occupied cells, as shown in Table 9.11.

Supply

19 30 50 10 u1 = 10
5 + 32 + 60 2 7

70 30 40 60 u2 = 60
+1 (+) 7 2 (–) 9
– 18
40 8 70 20 u3 = 20
(–) 8 10 (+) 18
+ 11 + 70

Table 9.11 Demand 5 8 7 14 34

Initial Solution,
VAM v1 = 9 v2 = – 12 v3 = – 20 v4 = 0

c34 = u3 + v4 or 20 = u3 + 0 or u3 = 20 c24
= u2 + v4 or 60 = u2 + 0 or u2 = 60 c14 =
u1 + v4 or 10 = u1 + 0 or u1 = 10
Given u1, u2, and u3, value of v1, v2 and v3 can also be calculated as shown below:
c11 = u1 + v1 or 19 = 10 + v1 or v1 = 9 c23
= u2 + u3 or 40 = 60 + v3 or v3 = – 20 c32 =
u3 + v2 or 8 = 20 + v2 or v2 = – 12
 Transportation Problem 271

3. The opportunity cost for each of the occupied cell is determined by using the relation
dij = cij – (ui + vj) and is shown below.
d12 = c12 – (u1 + v2) = 30 – (10 – 12) = 32 d13
= c13 – (u1 + v3) = 50 – (10 – 20) = 60 d21 =
c2l – (u2 + v1) = 70 – (60 + 9) = 1 d22 = c22
– (u2 + v2) = 30 – (60 – 12) = – 18 d31 = c31 –
(u3 + v1) = 40 – (20 + 9) = 11 d33 = c33 – (u3
+ v3) = 70 – (20 – 20) = 70
4. According to the optimality criterion for cost minimizing transportation problem, the current solution
is not optimal, since the opportunity costs of the unoccupied cells are not all zero or positive. The
value of d22 = – 18 in cell (S2, D2) is indicating that the total transportation cost can be reduced in
the multiple of 18 by shifting an allocation to this cell.
5. A closed-loop (path) is traced along row S2 to an occupied cell (S3, D2). A plus sign is placed in cell (S2,
D2) and minus sign in cell (S3, D2). Now take a right-angle turn and locate an occupied cell in column
D4. An occupied cell (S3, D4) exists at row S3, and a plus sign is placed in this cell.
Continue this process and complete the closed path. The occupied cell (S2, D3) must be bypassed
otherwise they will violate the rules of constructing closed path.
6. In order to maintain feasibility, examine the occupied cells with minus sign at the corners of closed loop,
and select the one that has the smallest allocation. This determines the maximum number of units that
can be shifted along the closed path. The minus signs are in cells (S3, D2) and (S2, D4). The cell
(S2, D4) is selected because it has the smaller allocation, i.e. 2. The value of this allocation is then added
to cell (S2, D2) and (S3, D4), which carry plus signs. The same value is subtracted from cells (S2, D4) and
(S3, D2) because they carry minus signs.
7. The revised solution is shown in Table 9.12. The total transportation cost associated with this solution
is:
Total cost = 5 × 19 + 2 × 10 + 2 × 30 + 7 × 40 + 6 × 8 + 12 × 20 = Rs 743

D1 D2 D3 D4 Supply ui

S1 19 30 50 10 u1 = 0
5 +32 +42 2 7

S2 70 30 40 60 u2 = 32
+ 19 2 7 +14 9

S3 40 8 70 20 u3 = 10
+ 11 6 + 52 12 18

Demand 5 8 7 14 34
Table 9.12
Optimal Solution
vj v1 = 19 v2 = – 2 v3 = 8 v4 = 10

8. Test the optimality of the revised solution once again in the same way as discussed in earlier steps.
The values of uis, vjs and dijs are shown in Table 9.12. Since each of dijs is positive, therefore, the current
basic feasible solution is optimal with a mi]nimum total transportation cost of Rs 743.

Example 9.7 A company has factories at F1, F2, and F3 that supply products to warehouses at W1, W2 and
W3. The weekly capacities of the factories are 200, 160 and 90 units, respectively. The weekly warehouse
requirements are 180, 120 and 150 units, respectively. The unit shipping costs (in rupees) are as follows:
272 Operations Research: Theory and Applications

Warehouse
W1 W2 W3 Supply

F1 16 20 12 200
Factory F2 14 8 18 160
F3 26 24 16 90
Demand 180 120 150 450

Determine the optimal distribution for this company in order to minimize its total shipping cost.
Solution Initial basic feasible solution obtained by North-West Corner Rule is given in Table 9.13.
Since, as required, this initial solution has m + n – l = 3 + 3 – l = 5 allocations, it is a non-degenerate solution.
The optimality test can, therefore, be performed. The total transportation cost associated with this solution is:

W1 W2 W3 Supply

F1 16 20 12
180 20 200

F2 14 8 18
100 60 160

F3 26 24 16
90 90
Table 9.13
Initial Solution
Demand 180 120 150 450

Total cost = 16 × 180 + 20 × 20 + 8 × 100 + 18 × 60 + 16 × 90 = Rs 6,800

Determine the values of uis and vjs as usual, by arbitrarily assigning u1 = 0. Given u1 = 0, the values
of others variables obtained by using the equation cij = ui + vj for occupied cells, are shown in Table 9.14.

W1 W2 W3 Supply ui

At each step a non- F1 16 20 12 u1 = 0

occupied cell with 180 (–) 20 (+) 200
largest negative – 18
opportunity cost is
selected to get F2 14 8 18 u2 = – 12
maximum reduction (+) 100 60 (–) 160
in total
transportation cost.
+10

F3 26 24 16 u3 = – 14
90 90
+24 +18

Demand 180 120 150 450

Table 9.14 vj v1 = 16 v2 = 20 v3 = 30

c11 = u1 + v1 or 16 = 0 + v1 or v1 = 16
c12 = u1 + v2 or 20 = 0 + v2 or v2 = 20
c22 = u2 + v2 or 8 = u2 + 20 or u2 = – 12
c23 = u2 + v3 or 18 = –12 + v3 or v3 = 30
c33 = u3 + v3 or 16 = u3 + 30 or u3 = – 14
 Transportation Problem 273

The opportunity cost for each of the unoccupied cells is determined by using the equation,
dij = cij – (ui + vj) as follows:
d13 = c13 – (u1 + v3) = 12 – (0 + 30) = – 18
d21 = c21 – (u2 + v1) = 14 – (–12 + 16) = 10
d31 = c31 – (u3 + v1) = 26 – (–14 + 16) = 24
d32 = c32 – (u3 + v2) = 24 – (–14 + 20) = 18

The value of d13 = – 18 in the cell (F1, W3) indicates that the total transportation cost can be reduced
in a multiple of 18 by introducing this cell in the new transportation schedule. To see how many units of the
commodity could be allocated to this cell (route), form a closed path as shown in Table 9.14.
The largest number of units of the commodity that should be allocated to the cell (F1, W3) is 20 units
because it does not violate the supply and demand restrictions (minimum allocation among the occupied cells
bearing negative sign at the corners of the loop). The new transportation schedule (solution) so obtained is
shown in Table 9.15.

W1 W2 W3 Supply

F1 16 20 12
180 20 200

F2 14 8 18
120 40 160

F3 26 24 16
90 90

Demand 180 120 150 450

Table 9.15
Revised Schedule

The total transportation cost associated with this solution is

Total cost = 16 × 180 + 12 × 20 + 8 × 120 + 18 × 40 + 16 × 90 = Rs 6,240
To test the optimality of the new solution shown in Table 9.15, again calculate the opportunity cost
of each unoccupied cell in the same manner as discussed earlier. The calculations for uis, vjs and dijs are
shown in Table 9.16.
c13 = u1 + v3 or 12 = u1 + 0 or u1 = 12
c23 = u2 + v3 or 18 = u2 + 0 or u2 = 18
c33 = u3 + v3 or 16 = u3 + 0 or u3 = 16
c11 = u1 + v1 or 16 = 12 + v1 or v1 = 4
c22 = u2 + v2 or 8 = 18 + v2 or v2 = – 10

d12 = c12 – (u1 + v2) or 20 – (12 – 10) = 18

d21 = c21 – (u2 + v1) or 14 – (18 + 4) = – 8
d31 = c31 – (u3 + v1) or 26 – (16 + 4) = 6
d32 = c32 – (u3 + v2) or 24 – (16 – 10) = 18

The value of d21 = – 8 in the cell (F2, W1) indicates that the total cost of transportation can further
be reduced in a multiple of 8 by introducing this cell in the new transportation schedule. The new solution
is obtained by introducing 40 units of the commodity in the cell (F2, W1), as indicated in Table 9.16. The new
solution is shown in Table 9.17.
274 Operations Research: Theory and Applications

W1 W2 W3 Supply ui

F1 16 20 12 u1 = 12
(–) 180 20 (+) 200
18

F2 14 8 18 u2 = 18
(+) 120 40 (–) 160
–8

F3 26 24 16 u3 = 16
90 90
6

Demand 180 120 150

Table 9.16 vj v1 = 4 v2 = – 10 v3 = 0

W1 W2 W3 Supply

F1 16 20 12
140 60 200

F2 14 8 18
40 120 160

F3 26 24 16
90 90
Table 9.17
Revised Schedule Demand 180 120 150

The total transportation cost associated with this solution is

Total cost = 16 × 140 + 12 × 60 + 14 × 40 + 8 × 120 + 16 × 90 = Rs 5,920
To test the optimality of the new solution shown in Table 9.17, calculate again the opportunity cost
of each unoccupied cell in the same manner as discussed earlier. The calculations are shown in Table 9.18.

W1 W2 W3 Supply ui

F1 16 20 12 u1 = 16
140 +10 60 200

F2 14 8 18 u2 = 14
40 120 +8 160

F3 26 24 16 u3 = 20
+6 +10 90 90

Demand 180 120 150

Table 9.18
vj v1 = 0 v2 = – 6 v3 = – 4

d12 = c12 – (u1 + v2) or 20 – (16 – 6) = 10 d23

= c23 – (u2 + v3) or 18 – (14 – 4) = 8 d31 =
c31 – (u3 + v1) or 26 – (20 + 0) = 6 d32 = c32
– (u3 + v2) or 24 – (20 – 6) = 10
 Transportation Problem 275

Since none of the unoccupied cells in Table 9.18 has a negative opportunity cost value, therefore, the
total transportation cost cannot be reduced further. Thus, the solution shown in Table 9.18 is the optimal
solution, giving the optimal transportation schedule with a total cost of Rs 5,920.

Example 9.8 The following table provides all the necessary information on the availability of supplyto
each warehouse, the requirement of each market, and the unit transportation cost (in Rs) from each warehouse
to each market.
Market

P Q R S Supply

Warehouse A 6 3 5 4 22
B 5 9 2 7 15
C 5 7 8 6 8
Demand 7 12 17 9 45

The shipping clerk of the shipping agency has worked out the following schedule, based on his own
experience: 12 units from A to Q, 1 unit from A to R, 9 units from A to S, 15 units from B to R, 7 units from
C to P and 1 unit from C to R.
(a) Check and see if the clerk has the optimal schedule.
(b) Find the optimal schedule and minimum total transport cost.
(c) If the clerk is approached by a carrier of route C to Q, who offers to reduce his rate in the hope of getting
some business, by how much should the rate be reduced before the clerk would offer him the business.
Solution (a) The shipping schedule determined by the clerk based on his experience is shown inTable
9.19. The total transportation cost associated with this solution is
Total cost = 3 × 12 + 5 × 1 + 4 × 9 + 2 × 15 + 5 × 7 + 8 × 1 = Rs 150
Since the number of occupied cells (i.e. 6) is equal to the required number of occupied cells (i.e.m
+ n – 1) in a feasible solution, therefore the solution is non-generate feasible solution. Now, to test the
optimality of the solution given in Table 9.19, evaluate each unoccupied cell in terms of the opportunity
cost associated with it. This is done in the usual manner and is shown in Table 9.20.

P Q R S Supply

A 6 3 5 4
12 1 9 22
B 5 9 2 7
15 15
C 5 7 8 6
7 1 8

Demand 7 12 17 9 45 Table 9.19

Initial Solution

In Table 9.20, cell (C, S) has a negative opportunity cost (i.e. – 1). Thus, this solution is not the optimal
solution and, therefore, the schedule prepared by the shipping clerk is not optimal.
(b) By forming a closed-loop to introduce the cell (C, S) into the new transportation schedule as shown
in Table 9.20, we get a new solution that is shown in Table 9.21.
276 Operations Research: Theory and Applications

P Q R S Supply ui

A 6 3 5 4 u1 = 0
+4 12 (+) 1 9 (–) 22

B 5 9 2 7 u2 = – 3
15 15
+9 +6
C 5 7 8 6 u3 = 3
7 (–) 1 (+) 8
+1 –1

Demand 7 12 17 9 45

Table 9.20 vj v1 = 2 v2 = 3 v3 = 5 v4 = 4

While testing the optimality of the improved solution shown in Table 9.21, we found that the opportunity
costs in all the unoccupied cells are positive. Thus the current solution is optimal and the optimal schedule is
to transport 12 units from A to Q; 2 units from A to R; 8 units from A to S; 15 units from B to R; 7 units from
C to P and 1 unit from C to S. The total minimum transportation cost associated with this solution is
Total cost = 3 × 12 + 5 × 2 + 4 × 8 + 2 × 15 + 5 × 7 + 6 × 1 = Rs 149

P Q R S Supply ui

A 6 3 5 4 u1 = 0
+3 12 2 8 22

B 5 9 2 7 u2 = – 3
+5 +9 15 +6 15

C 5 7 8 6 u3 = 2
7 +2 +1 1 8

Demand 7 12 17 9 45

Table 9.21
vj v1 = 3 v2 = 3 v3 = 5 v4 = 4
Optimal Solution

(c) The total transportation cost will increase by Rs 2 (opportunity cost) if one unit of commodity is
transported from C to Q. This means that the rate of the carrier on the route C to Q should be reduced by
Rs 2, i.e. from Rs 7 to Rs 5 so as to get some business of one unit of commodity only.
In case all the 8 units available at C are shipped through the route (C, Q), then the solution presented
in Table 9.21 may be read as shown in Table 9.22.

P Q R S Supply

A 6 3 5 4
7 4 2 9 22

B 5 9 2 7
15 15

C 5 7 8 6
8 8

Table 9.22 Demand 7 12 17 9 45

 Transportation Problem 277

The total cost of transportation associated with this solution is

Total cost = 6 × 7 + 3 × 4 + 5 × 2 + 4 × 9 + 2 × 15 + 7 × 8 = Rs 186.
Thus, the additional cost of Rs 37 (= 186 – 149) should be reduced from the transportation cost of 8 units
from C to Q. Hence transportation cost per unit from C to Q should be at the most 7 – (37/8) = Rs 2.38.

Example 9.10 ABC Limited has three production shops that supply a product to five warehouses. The cost
of production varies from shop to shop and cost of transportation from one shop to a warehouse also varies.
Each shop has a specific production capacity and each warehouse has certain amount of requirement. The
costs of transportation are given below:
Warehouse
I II III IV V Supply

A 6 4 4 7 5 100
Shop B 5 6 7 4 8 125

C 3 4 6 3 4 175

Demand 60 80 85 105 70 400

The cost of manufacturing the product at different production shops is

Shop Variable Cost Fixed Cost

A 14 7,000
B 16 4,000
C 15 5,000

Find the optimum quantity to be supplied from each shop to different warehouses at the minimum total
cost. [Delhi Univ., MBA, 2007]
Solution In this case, the fixed cost data is of no use. The transportation cost matrix will include the given
transportation cost plus the variable cost, as shown in Table 9.23.

I II III IV V Supply

A 6 + 14 = 20 4 + 14 = 18 4 + 14 = 18 7 + 14 = 21 5 + 14 = 19 100

B 5 + 16 = 21 6 + 16 = 22 7 + 16 = 23 4 + 16 = 20 8 + 16 = 24 125

C 3 + 15 = 18 4 + 15 = 19 6 + 15 = 21 3 + 15 = 18 4 + 15 = 19 175
Table 9.23
Demand 60 80 85 105 70 400
The optimal solution obtained by applying MODI method is shown in Table 9.24.

I II III IV V Supply ui

A 20 18 18 21 19 u1 = 18
15 85 100
+3 +5 +1

B 21 22 23 20 24 u2 = 22
20 105 125
0 +1 +2

C 18 19 21 18 19 u3 = 19
60 45 70 175
+2 +1

Demand 60 80 85 105 70
Table 9.24
vj v1 = – 1 v2 = 0 v3 = 0 v4 = – 2 v5 = 0 Optimal Solution
278 Operations Research: Theory and Applications

The transportation cost associated with the solution is

Total cost = 18 × 15 + 18 × 85 + 22 × 20 + 20 × 105 + 18 × 60 + 19 × 45 + 19 × 70 = Rs 7,605

CONCEPTUAL QUESTIONS B
1. Describe the computational procedure of the optimality test in m n
a transportation problem.  Si   D j
i 1 j 1

2. Indicate how you will test for optimality of initial feasible solution
Symbolically state the transportation problem. Establish that the
of a transportation problem.
optimal solution is not altered when the cij’s are replaced byc i*j
3. Let Si and Dj (i = 1, 2, . . ., m; j = 1, 2, . . ., n) be the supply
s, where c i*j = cij + ui + vj, ui (i = 1, 2, . . ., m) and vj ( j
and demand available, respectively, for a commodity at m godowns
and n markets. Let cij be the cost of transporting oneunit of the = 1, 2, . . ., n) are arbitrary real numbers.
commodity from godown i to market j. Assuming that

SELF PRACTICE PROBLEMS B

1. Consider four bases of operation Bi and three targets Tj. Thetons product to the five customers. The net unit cost of transporting
of bombs per aircraft from any base that can be deliveredto any from the three plants to the five customers is given below:
target are given in the following table:
Customers
Target (Tj )
A B C D E
T1 T2 T3
I 4 1 3 4 4
B1 8 6 5
Base (B ) B Plants II 2 3 2 2 3
6 6 6
i 2 10 8 4 III 3 5 2 4 4
B3 8 6 4
B4 4. The following table gives the cost of transporting material from
supply points A, B, C and D to demand points E, F, G, H and I.
The daily sortie capability of each of the four bases is 150
sorties per day. The daily requirement of sorties spread overeach To
individual target is 200. Find the allocation of sorties from each
E F G H I
base to each target which maximizes the total tonnageover all
three targets. Explain each step in the process. A 8 10 12 17 15
2. A company has four warehouses, a, b, c and d. It is required From B 15 13 18 11 9
to deliver a product from these warehouses to three customers
A, B and C. The warehouses have the following amounts instock: C 14 20 6 10 13

Warehouse : a b c d D 13 19 7 5 12
No. of units : 15 16 12 13
The present allocation is as follows:
and the customers’ requirements are A to E 90; A to F 10; B to F 150; C to F 10; C to G 50; C to
Customer : A B C I 120; D to H 210; D to I 70.
No. of units : 18 20 18 (a) Check if this allocation is optimum. If not, find an optimum
The table below shows the costs of transporting one unit from schedule.
warehouse to the customer. (b) If in the above problem, the transportation cost from A to
G is reduced to 10, what will be the new optimum schedule?
Warehouse
5. A whole selling company has three warehouses from which the
a b c d supplies are drawn for four retail customers. The company
A 8 9 6 3 deals in a single product, the supplies of which at each
warehouse are:
Customer B 6 11 5 10
Warehouse Supply Customer Demand
C 3 8 7 9 Number (units) Number (units)

Find the optimal transportation routes. 1 20 1 15

3. A firm manufacturing a single product has three plants I, II and 2 28 2 19
II. They have produced 60, 35 and 40 units, respectively during 3 17 3 13
this month. The firm had made a commitment to sell 22 units to 4 18
customer A, 45 units to customer B, 20 units to customer C, 18
Total supply at the warehouses is equal to total demand from
units to customer D and 30 units to customer E. Find the minimum
the customers. The table below gives the transportation costs,
possible transportation cost of shifting the manufactured
per unit, shipped from each warehouse to each customer.
 Transportation Problem 279

Customer Calculate a transportation plan that will minimize the distribution

C1 C2 C3 C4 cost for the steel company. State the cost of this distributionplan.
9. A company has three factories at Amethi, Baghpat and Gwalior that
W1 3 6 8 5 have a production capacity of 5,000, 6,000, and 2,500 tonnes,
respectively. Four distribution centres at Allahabad, Bombay,
W2 6 1 2 5 Kolkata and Delhi, require 6,000 tonnes, 4,000 tonnes, 2,000
Warehouse tonnes and 1,500 tonnes, respectively, of the product. The
W3 7 8 3 9
transportation costs per tonne from different factories to
Determine what supplies should be dispatched from each of the different centres are given below:
warehouses to each customer so as to minimize the overall
transportation cost. Factories Distribution Centres
6. A manufacturer has distribution centres at Agra, Allahabad and
Allahabad Bombay Kolkata Delhi
Kolkata. These centres have availability of 40, 20 and 40 units of
his product, respectively. His retail outlets at A, B, C, D and E Amethi 3 2 7 6
require 25, 10, 20, 30 and 15 units of the products, respectively.
The transport cost (in rupees) per unit between each centre outlet Baghpat 7 5 2 3
is given below: Gwalior 2 5 4 5

Distribution Retail Outlets Suggest an optimum transportation schedule and find the minimum
Centre cost of transportation.
A B C D E
10. A company has three plants and four warehouses. The supply
Agra 55 30 40 50 40 and demand in units and the corresponding transportation costs
are given. The table below has been taken from the solution
Allahabad 35 30 100 45 60 procedure of a transportation problem:
Kolkata 40 60 95 35 30 Warehouses

Determine the optimal distribution so as to minimize the cost of I II III IV Supply

transportation. [Delhi Univ., MBA, 2002]
7. A manufacturer has distribution centres located at Agra, Allahabad A 5 10 4 10 5 10
and Kolkata. These centres have available 40, 20 and 40 units
of his product. His retail outlets at A, B, C, D and E requires 25, 10,
Plants B 6 20 8 7 2 5 25
20, 30 and 15 units of the product, respectively. The shipping
cost per unit (in rupees) between each centre and outlet is given
C4 5 2 10 5 5 7 20
in the following table.

Retail Outlets Demand 25 10 15 5 55

Distribution

Centre
A B C D E
Answer the following questions, giving brief reasons for the
Agra 55 30 40 50 40 same:
Allahabad 35 30 100 45 60 (a) Is this solution feasible?
(b) Is this solution degenerate?
Kolkata 40 60 95 35 30 (c) Is this solution optimum?
(d) Does this problem have more than one optimum solution?
Determine the optimal shipping cost.
If so, show all of them.
[Delhi Univ. MBA, 2003]
(e) If the cost for the route B-III is reduced from Rs 7 to Rs
8. A steel company is concerned with the problem of distributing
6 per unit, what will be the optimum solution?
imported ore from three ports to four steel mills. The supplies
11. A baking firm can produce a special bread in either of its two
of ore arriving at the ports are:
plants, the details of which are as follows:
Port Tonnes per week
a 20,000 Plant Production Capacity Production Cost
b 38,000 Loaves Rs/Loaf
c 16,000
The demand at the steel mills is as follows: A 2,500 2.30
Steel mills : A B C D B 2,100 2.50

Tonnes per week : 10,000 18,000 22,000 24,000

Four restaurant chains are willing to purchase this bread; their
The transportation cost is Re 0.05 per tonne per km. The demand and the prices they are willing to pay are as follows:
distance between the ports and the steel mills is as given below:

A B C D Chain Maximum Demand Price Offered

Loaves Rs/Loaf
a 50 60 100 50
1 1,800 3.90
b 80 40 70 50 2 2,300 3.70
3 550 4.00
c 90 70 30 50
4 1,750 3.60
280 Operations Research: Theory and Applications

The cost (in paise) of shipping a loaf from a plant to a restaurant Determine a delivery schedule for the baking firm that will
chain is: maximize its profit from this bread.
Write the dual of this transportation problem and use it for
Chain 1 Chain 2 Chain 3 Chain 4 checking the optimal solution to the given problem.
[Delhi Univ., MBA, 2005]
Plant A 6 8 11 9
Plant B 12 6 8 5

HINTS AND ANSWERS

1. The initial solution obtained by VAM is also the optimal 8. x11 = 10,000, x14 = 10,000, x22 = 18,000, x23 = 6,000,
solution: x11 = 50, x12 = 100, x21 = 150, x33 = 150, x24 = 14,000, x33 = 16,000, Total cost = Rs 1,66,000.
x42 = 100, x43 = 50. Maximum total tonnage = 3,300. 9. x11 = 3,500, x12 = 1,500, x22 = 2,500, x23 = 2,000, x24 = 1,500,
2. (A, b) = 5, (A, d ) = 13, (B, b) = 8, (B, c) = 12, (C, a ) = 15 x31 = 2,500, Total cost = Rs 39,500.
and (C, b) = 3, Total cost = Rs 301. 10. (a) The solution is feasible because it satisfies supply and demand
3. (I, B) = 45, (I, F) = 15, (II, A) = 17, (II, D) = 18, constraints.
(III, A) = 5, (III, C) = 20 and (III, E) = 15, (b) The solution is non-degenerate because the number of
Total cost = Rs 290. occupied cells are equal to the required number of(m
4. (a) (A, F) = 100, (B, F) = 70, (B, I ) = 80, (C, E) = 90, + n – 1) of occupied cells in the solution.
(C, G) = 50, (C, I ) = 40, (D, H) = 210, (D, I ) = 70, (c) Solution is optimal.
Total cost = Rs 6,600. (d) The problem has alternative optimal solution because
(b) When transportation cost from A to G is reduced to 10, the opportunity cost for cell (B, III) is zero;
optimal schedule given in (a) remains unchanged. x11 = 10, x21 = 15, x23 = 5, x24 = 5, x31 = 10 and x32 = 10,
Total cost = Rs 235.
5. x11 = 15, x14 = 5, x22 = 19, x24 = 9, x33 = 13, x34 = 4, Total (e) If cell (B, III) has a unit cost of 6, the opportunity cost
cost = Rs 209.
in this cell will be – 1 and, therefore, the given solution will
6. x11 = 5, x12 = 10, x13 = 20, x14 = 5, x21 = 20, x34 = 25, not be optimal. The new solution obtained will be:
x35 = 15, Total cost = Rs 3,650. x13 = 10, x21 = 15, x23 = 5, x24 = 5, x31 = 10, x32 = 10,
7. x12 = 2, x13 = 4, x15 = 2, x21 = 4, x31 = 1, x34 = 6, x35 = 1, Total cost = Rs 230.
Total cost = Rs 720.

9.6 VARIATIONS IN TRANSPORTATION PROBLEM

Some of the variations that often arise while solving a transportation problem are as follows.

9.6.1 Unbalanced Supply and Demand

For a feasible solution to exist, it is necessary that the total supply must equal the total demand. That is,
Total supply = Total demand
m n
 ai =  bj
A dummy source i 1 j 1
or destination is

added to balance But a situation may arise when the total available supply is not equal to the total demand [For proof see
transportation appendix]. The following two cases may arise:
problem where
demand is not equal (a) If the total supply exceeds the total demand, then an additional column (called a dummy demand centre)
to supply. can be added to the transportation table in order to absorb the excess supply. The unit transportation cost
for the cells in this column is set equal to zero because these represent product items that are neither
made nor sent.
(b) If the total demand exceeds the total supply, a dummy row (called a dummy supply centre) can be added
to the transportation table to account for the excess demand quality. The unit transportation cost in
such a case also, for the cells in the dummy row is set equal to zero.
Example 9.10 A company has received a contract to supply gravel to three new construction projects
located in towns A, B and C. The construction engineers have estimated that the required amounts of gravel
which will be needed at these construction projects are:
 Transportation Problem 281

Project Location Weekly Requirement

(Truckloads)
A 72
B 102
C 41

The company has 3 gravel pits located in towns X, Y and Z. The gravel required by the construction
projects can be supplied by three pits. The amount of gravel that can be supplied by each pit is as follows:
Plant : X Y Z
Amount available (truckloads) : 76 82 77
The company has computed the delivery cost from each pit to each project site. These costs (in Rs)
are shown in the following table:
Project Location
A B C
X 4 8 8
Pit Y 16 24 16
Z 8 16 24
Schedule the shipment from each pit to each project in such a manner that it minimizes the total
transportation cost within the constraints imposed by pit capacities and project requirements. Also find the
minimum cost.
Solution The total plant availability of 235 truckloads exceeds the total requirement of 215 truckloadsby
20 truckloads. The excess truckload capacity of 20 is handled by adding a dummy project location (column),
Dexcess with a requirement of 20 truck loads, Assigning unit transportation costs to the dummy
project location, modified transportation table is shown in Table 9.25.
A B C Dexcess Supply

W 4 8 8 0
35 41 76

X 16 24 16 0
62 20 82

Y 8 16 24 0
72 5 77
Table 9.25
Initial Solution
Demand 72 102 41 20 235

The initial solution is obtained by using Vogel’s approximation method as shown in Table 9.25. It may be
noted that 20 units are allocated from pit X to dummy project location D. This means pit X is short by 20 units.
To test the optimality of the initial solution shown in Table 9.25, calculate uis and vjs corresponding
to rows and columns respectively. These values are shown in Table 9.26.

A B C Supply ui

W 4 8 8 0 u1 = 8
+4 (+) 35 41 (–) +16 76

X 16 24 16 0 u2 = 24
(–) 62 (+) 20 82
–8
Y 8 16 24 0 u3 = 16
72 5 77
+8 +8

Demand 72 102 41 20 235

vj v1 = – 8 v2 = 0 v3 = 0 v4 = – 24 Table 9.26
282 Operations Research: Theory and Applications

In Table 9.26, opportunity cost shown in the cell (X, C) is negative, the current solution is not optimal.
Thus, the unoccupied cell (X, C), where d23 = – 8 must enter into the basis and cell (W, C) must leave the
basis, as shown by the closed path in Table 9.26. The new solution is shown in Table 9.27.

A B C Dexcess Supply ui

W 4 8 8 0 u1 = – 16
+4 76 +8 +16 76

X 16 24 16 0 u2 = 0
0 21 41 20 82

Y 8 16 24 0 u3 = – 8
72 5 +16 +8 77

Table 9.27 Demand 72 102 41 20 235

Optimal Solution
vj v1 = 16 v2 = 24 v3 = 16 v4 = 0
Since all opportunity costs, dij are non-negative in Table 9.27, the current solution is optimal. The total
minimum transportation cost associated with this solution is:
Total cost = 8 × 76 + 24 × 21 + 16 × 41 + 0 × 20 + 8 × 72 + 16 × 5 = Rs 2,424.
Example 9.11 A product is manufactured at four factories A, B, C and D. Their unit production costs are
Rs 2, Rs 3, Re 1 and Rs 5, respectively. Their production capacities are 50, 70, 30 and 50 units, respectively.
These factories supply the product to four stores, demands of which are 25, 35, 105 and 20 units respectively.
Unit transportation cost in rupees from each factory to each store is given in the table below.
Stores
I II III IV
A
B 2 4 6 11
Factories
C 10 8 7 5
D 13 3 9 12
4 6 8 3

Determine the extent of deliveries from each of the factories to each of the stores, so that the total
production and transportation cost is the minimum.
Solution The new transportation costs that include both the production and the transportation costs
is given in Table 9.28.

I II III IV Supply

A 2 + 2 = 4 4 + 2 = 6 6 + 2 = 8 11 + 2 = 13 50

B 10 + 3 = 13 8 + 3 = 11 7 + 3 = 10 5 + 3 = 8 70

C 13 + 1 = 14 3 + 1 = 4 9 + 1 = 10 12 + 1 = 13 30

D 4 + 5 = 9 6 + 5 = 11 8 + 5 = 13 3 + 5 = 8 50

Demand 25 35 105 20 200

Table 9.28
185

Since the total supply of 200 units exceeds the total demand of 185 units by 15 units, a dummy destination
(store) is added (or created) to absorb the excess capacity. The associated cost coefficients in dummy store
are taken as zero. This may be due to the reason that the surplus quantity remains lying inthe respective
factories and is not shipped at all. The modified table is shown in Table 9.29.
 Transportation Problem 283

I II III IV Dummy Supply

A 4 6 8 13 0 50
25 5 20

B 13 11 10 8 0 70
70

C 14 4 10 13 0 30
30

D 9 11 13 8 0 50
15 20 15
Table 9.29
Initial Solution
Demand 25 35 105 20 15 200

Using the VAM method the initial solution is shown in Table 9.29. It can be seen that 15 units are
allocated to dummy store from factory D. This means that the company may cut down the production by
15 units at the factory that is proving to be uneconomical. Now to test the optimality of the solution shown
in Table 9.29, evaluate each unoccupied cell in terms of opportunity cost associated with it as shown in Table
9.30.

I II III IV Dummy Supply ui

A 4 6 8 13 0 50 u1 = –5
25 5 20 +10 +5

B 13 11 10 8 0 70 u2 = –3
+7 +3 70 +3 +3

C 14 4 10 13 0 30 u3 = –7
30
+12 +4 +12 +7
D 9 11 13 8 0 50 u4 = 0
0 0 15 20 15

Demand 25 35 105 20 15 200

Table 9.30
vj v1 = 9 v2 = 11 v3 = 13 v4 = 8 v5 = 0

Since the opportunity cost in all the unoccupied cells is positive, the initial solution is an optimal solution.
The total cost of transportation associated with this solution is
Total cost = 4 × 25 + 6 × 5 + 8 × 20 + 10 × 70 + 4 × 30 + 13 × l5 + 8 × 20 + 0 × l5 = Rs l,465.

9.6.2 Degeneracy and its Resolution

A basic feasible solution for the general transportation problem must have exactly m + n – 1 (number of rows
+ number of columns – 1) positive allocations in the transportation table. If the number of occupied cells is
less than the required number, m + n – 1, then such a solution is called degenerate solution. Insuch
cases, the current solution cannot be improved further because it is not possible to draw a closed Degeneracy arises
path for every occupied cell. Also, the values of dual variables ui and vj that are used to test the optimality when the number of
occupied cells are
cannot be computed. Thus, degeneracy needs to be removed in order to improve the given solution. The less than the
degeneracy in the transportation problems may occur at two stages: number of rows +
(a) At initial basic feasible solution the number of occupied cells may be less than m + n – 1 allocations. columns – 1.
(b) At any stage while moving towards optimal solution two or more occupied cells may become
simultaneously unoccupied.

Case 1: Degeneracy at the initial solution To resolve degeneracy at the initial solution, allocate
a very small quantity close to zero to one or more (if needed) unoccupied cells so as to get m + n – 1 number
of occupied cells. This amount is denoted by a Greek letter  (epsilon) or  (delta). This quantity
284 Operations Research: Theory and Applications

would neither affect the total cost nor the supply and demand values. In a minimization transportation
problem it is better to allocate  to unoccupied cells that have lowest transportation costs, whereas in
maximization problems it should be allocated to a cell that has a high payoff value. In some cases,  must
be added in one of those unoccupied cells that may help in the determination of ui and vj values.
The quantity  is considered to be so small that if it is transferred to an occupied cell it does not
change the quantity of allocation. That is,
xij +  = xij –  = xij
 –  = 0;  +  = 
0 +  = ; k ×  = 
The  value also does not affect the total transportation cost. Hence, the quantity  is used to evaluate
unoccupied cells and to reduce the number of improvement cycles necessary to reach an optimal solution.
Once the purpose is over,  can be removed from the transportation table.

Example 9.12 A manufacturer wants to ship 22 loads of his product as shown below. The matrix gives
the kilometres from sources of supply to the destinations.

Destination

D1 D2 D3 D4 D5 Supply

S1 5 8 6 6 3 8
Source S2 4 7 7 6 5 5
S3 8 4 6 6 4 9
Demand 4 4 5 4 8 22
25

The shipping cost is Rs 10 per load per km. What shipping schedule should be used in order to minimize
the total transportation cost? [Delhi Univ., MBA, 2001]

Solution Since the total destination requirement of 25 units exceeds the total resource capacity of 22 by
3 units, the problem is unbalanced. The excess requirement is handled by adding a dummy plant, Sexcess with
a capacity equal to 3 units. Assign zero unit transportation cost to the dummy plant, the modified
transportation table is shown in Table 9.31.

D1 D2 D3 D4 D5 Supply

S1 5 8 6 6 3
5 3 8

S2 4 7 7 6 5
4 1 5

S3 8 4 6 6 4
4 5 9

Sexcess 0 0 0 0 0
Table 9.31 3 3
Initial Solution,
but Degenerate Demand 4 4 5 4 8 25

The initial solution is obtained by using Vogel’s approximation method as shown in Table 9.31. Since
solution does not have required number of m + n – 1 = 4 + 5 –1 = 8 occupied cells, therefore, the initial
solution is degenerate. In order to remove degeneracy, assign  to unoccupied cell (S2, D5), which has the
minimum cost amongst the unoccupied cells, as shown in Table 9.32.
 Transportation Problem 285

D1 D2 D3 D4 D5 Supply ui

S1 5 8 6 6 3 u1 = 0
(–) 5 3 8
+3 +5 +2 (+)

S2 4 7 7 6 5 u2 = 2
4 +2 1 (+)  (–)
5
–1
S3 8 4 6 6 4 u3 = 1
4 5 9
+5 –1 +1
0 0 0 0 0 u4 = – 4
3
(+) 3
+2 +1 –2 (–) +1 1

Demand 4 4 5 4 8 25

vj v1 = 2 v2 = 3 v3 = 6 v4 = 4 v5 = 3 Table 9.32

Determine ui and vj for occupied cells as shown in Table 9.32. Since opportunity cost in the cell (Sexcess,
D3) is largest negative, it is entered into the new solution mix and the cell (S2, D5) leaves the current solution
mix. The new solution is shown in Table 9.33.

D1 D2 D3 D4 D5 Supply ui

S1 5 8 6 6 3 u1 = 0
5 3 8
+1 +5 (–) 0 (+)
S2 4 7 7 6 5 u2 = 0
4 1 5
+4 +1 +2
S3 8 4 6 6 4 u3 = 1
4 (+) 5 9
+3 –1 –1 (–)

0 0 0 0 0 u4 = – 6
(+)  3 (–) 3
+2 +3 +3

Demand 4 4 5 4 8 25

vj v1 = 4 v2 = 3 v3 = 6 v4 = 6 v5 = 3 Table 9.33

Repeat the procedure of testing optimality of the solution given in Table 9.33. The optimal solution
is shown in Table 9.34.
D1 D2 D3 D4 D5 Supply

S1 5 8 6 6 3
8 8

S2 4 7 7 6 5
4 1 5

S3 8 4 6 6 4
4 2 3 9

Sexcess 0 0 0 0 0
3 3
Table 9.34
Demand 4 4 5 4 8 25 Optimal Solution
286 Operations Research: Theory and Applications

The minimum total transportation cost associated with this solution is

Total cost = (3 × 8 + 4 × 4 + 6 × 1 + 4 × 4 + 6 × 2 + 6 × 3) × 10 = Rs 920

Case 2: Degeneracy at subsequent iterations To resolve degeneracy, which occurs during optimality
test, the quantity may be allocated to one or more cells that have recently been unoccupied, to have m + n –
1 number of occupied cells in the new solution.
Example 9.13 Goods have to be transported from sources S1, S2 and S3 to destinations D1, D2 and D3. The
transportation cost per unit, capacities of the sources, and the requirements of the destinations aregiven
in the following table.
D1 D2 D3 Supply
S1 8 5 6 120
S2 15 10 12 80
S3 3 9 10 80
Demand 150 80 50

Determine a transportation schedule so that cost is minimized.

Solution Using North-West Corner Method, the non-degenerate initial basic feasible solution so
obtained is given in Table 9.35.

D1 D2 D3 Supply

S1 8 5 6
120
120

S2 15 10 12
80
30 50

S3 3 9 10
80
30 50
Table 9.35
Initial Solution Demand 150 80 50 280

To test the optimality of the solution given in Table 9.35, calculate ui, vj and dij as usual as shown
in Table 9.36.

D1 D2 D3 Supply ui

S1 8 5 6 120 u1 = –7
120 +2 +2
S2 15 10 12 80 u1 = 0
(–) 30 50 (+)
+1
S3 3 9 10 80 u1 = –1
(+) –11 30 (–) 50

Demand 150 80 50 280

Table 9.36 vj v1 = 15 v2 = 10 v3 = 11

Since the unoccupied cell (S3, D1) has the largest negative opportunity cost of – 11, therefore, cell
(S3, D1) is entered into the new solution mix. The closed path for (S3, D1) is shown in Table 9.36. The
maximum allocation to (S3, D1) is 30. However, when this amount is allocated to (S3, D1) both cells (S2,
D1) and (S3, D2) become unoccupied because these two have same allocations. Thus, the number of positive
allocations become less than the required number, m + n – 1 = 3 + 3 – 1 = 5. Hence, this is a degenerate
solution as shown in Table 9.37.
 Transportation Problem 287

D1 D2 D3 Supply

S1 8 5 6
120
120
S2 15 10 12
80
80
S3 3 9 10
80
30  50 Table 9.37
Degenerate
Demand 150 80 50 280 Solution

To remove degeneracy a quantity  is assigned to one of the cells that has become unoccupied so
that there are m + n – 1 occupied cells. Assign  to (S3, D2) and proceed with the usual solution procedure.
The optimal solution is given in Table 9.38.

D1 D2 D3 Supply

S1 8 5 6
120
70  50
S2 15 10 12
80
80
S3 3 9 10
80
80
Table 9.38
Demand 150 80 50 280 Optimal Solution

The total minimum transportation cost associated with this solution is:
Total cost = 8 × 70 + 6 × 50 + 10 × 80 + 3 × 80 = Rs 1,900

9.6.3 Alternative Optimal Solutions

The existence of alternative optimal solutions can be determined by an inspection of the opportunity costs,
dij for the unoccupied cells. If dij = 0, for an unoccupied cell in an optimal solution, then an alternative optimal
solution exists and can be obtained by bringing such an unoccupied cell in the solution mix withoutincreasing
the total transportation cost.
Illustration Consider the optimal solution of Example 9.10 given in Table 9.27. For ready reference Table
9.27 is reproduced as Table 9.39.

A B C Dexcess Supply ui

W 4 8 8 0
76 u1 = – 16
+4 76 +8 + 16

X 16 24 16 0
82 u2 = 0
0 21 41 20

Y 8 16 24 0
77 u3 = – 8
72 5 + 16 +8

Demand 72 102 41 20 235

Table 9.39
vj v1 = 16 v2 = 24 v3 = 16 v4 = 0
288 Operations Research: Theory and Applications

The opportunity costs in all unoccupied cells are positive except for the cell (X, A) where d21 = 0. This
means, if (X, A) is entered into the new solution mix, no change in the transportation cost would occur.
To determine this alternative solution, form a closed path for cell (X, A) as shown in Table 9.40.

A B C Supply ui

W 4 8 8 0 76 u1 = – 16
+4 76 +8 +16

X 16 24 16 0 82 u2 = 0
(+) 21 (–) 41 20
0
Y 8 16 24 0 77 u3 = – 8
(–) 72 5 (+) + 16 +8

Demand 72 102 41 20 235

Table 9.40
Optimal Solution vj v1 = 16 v2 = 24 v3 = 16 v4 = 0

The maximum quantity that can be allocated to cell (X, A) is 21. After this change, the new solution
is shown in Table 9.41.
Since all dij values are positive or zero, the solution given in Table 9.41 is optimal with a minimum total
transportation cost of Rs 2,424, which is same as in the previous solution.

A B C Dexcess Supply ui

W 4 8 8 0 76 u1 = – 16
+4 76 +8 + 16

X 16 24 16 0 82 u2 = –0
21 0 41 20

Y 8 16 24 0 77 u3 = – 8
51 26 + 16 +8
Table 9.41
Alternative Demand 72 102 41 20 235
Optimal Solution
vj v1 = 16 v2 = 24 v3 = 16 v4 = 0

Example 9.14 XYZ tobacco company purchases tobacco and stores in warehouses located in the
following four cities:
Warehouse Location (City) : A B C D
Capacity (Tonnes) : 90 50 80 60
The warehouses supply tobacco to cigarette companies in three cities that have the following demand:

Cigarette Company Demand (Tonnes)

Bharat 120
Janata 100
Red Lamp 110
 Transportation Problem 289

The following railroad shipping costs per tonne (in hundred rupees) have been determined:

Warehouse Location Bharat Janata Red Lamp

A 7 10 5
B 12 9 4
C 7 3 11
D 9 5 7

Because of railroad construction, shipments are temporarily prohibited from warehouse at city A to Bharat
Cigarette company.
(a) Find the optimum distribution for XYZ tobacco company.
(b) Are there multiple optimum solutions? If yes, identify them.
(c) Write the dual of the given transportation problem and use it for checking the optimum solution.
[Delhi Univ., MBA, 2003, 2005, 2008]
Solution Since the total demand of 330 units exceeds the total capacity of 280 units by 50 units of
the product, a dummy company is created to handle the excess demand. The associated cost coefficients
for the dummy warehouse location are taken as zero. Further, the cost element (i.e. 7) on the route city A–
Bharat company is replaced by M, since the route is prohibited. The modified table is shown in Table
9.42.

Bharat Janata Red Lamp Supply

A M 10 5 90

B 12 9 4 50

C 7 3 11 80

D 9 5 7 60

Dummy 0 0 0 50

Demand 120 100 110 330 Table 9.42

Using the VAM method, the initial solution is shown in Table 9.43. To test the optimality of this solution
evaluate each unoccupied cell in terms of opportunity cost associated with it as shown in Table 9.43.

Bharat Janata Red Lamp Supply ui

A M 10 5
M – 13 +1 90 90 u1 = 13

B 12 9 4
30 20 50 u2 = 12

C 7 3 11 11
(+) 80 (–) 80 u3 = 7
0 +12
D 9 5 7
(–) 40 20 (+) +6 60 u4 = 9
Dummy 0 0 0
50 +4 +8 50 u5 = 0
Demand 120 100 110 330
Table 9.43
vj v1 = 0 v2 = – 4 v3 = – 8 Optimal Solution
290 Operations Research: Theory and Applications

Since the opportunity cost in all the unoccupied cells is positive, the initial solution shown in
Table 9.43 is also an optimal solution. The total transport cost associated with this solution is
Total cost = 5 × 90 + 12 × 30 + 4 × 20 + 3 × 80 + 9 × 40 + 5 × 20 = Rs 1,59,000
(b) Since opportunity cost in cell (C, Bharat), d31 = 0, there exists an alternative optimal solution:
x13 = 90, x21 = 30, x23 = 20, x31 = 40, x32 = 40, x42 = 60 and x51 = 50
and total cost = Rs 1,59,000
(c) The dual of the given problem is
Maximize Z = (90 u1 + 50 u2 + 80 u3 + 60 u4 + 50 u5) + (120 v1 + 100 v2 + 110 v3)
subject to the constraints
u1 + v1  M u2 + v1  12 u3 + v1  7 u 4 + v1  9
u1 + v2  10 u2 + v2  9 u3 + v2  3 u 4 + v2  5
u1 + v3  5 u2 + v3  4 u3 + v3  11 u4 + v3  7
and ui, vj unrestricted in sign, for all i and j.
Substituting the values of uis and vjs from the optimal solution of transportation problem shown in
Table 9.43, we get
Maximize Z = 90 × 13 + 50 × 12 + 80 × 7 + 60 × 9 + 50 × 0 + 120 × 0 + 100 × – 4
+ 110 × – 8 = Rs 1,59,000
which is the same value as obtained earlier.

9.6.4 Prohibited Transportation Routes

If situations like road hazards (snow, flood, etc.), traffic regulations, etc., arise, then it may not be possible
to transport goods from certain sources to certain destinations. Such situations can be handled by assigning
a very large cost, say M (or  ) to such a route(s) (or cell).
Example 9.15 Consider the problem of scheduling the weekly production of certain items for the next four
weeks. The production cost of the item is Rs 10 for the first two weeks and Rs 15 for the last two weeks. The
weekly demands are 300, 700, 900 and 800, which must be met. The plant can produce a maximumof 700 units
per week. In addition, the company can use overtime during the second and third week. This increases the
weekly production by an additional 200 units, but the production cost also increases byRs 5. Excess
production can be stored at a unit cost of Rs 3 per week. How should the production be scheduled so as to
minimize the total cost?
Solution The given information is presented as a transportation problem in Table 9.44. The cost elementsin
each cell are determined by adding the production cost, the overtime cost of Rs 5, and the storage costof
Rs 3. Thus, in the first row, the cost of Rs 3 is added during second week onward. Since the outputof
any period cannot be used in a period preceding it, the cost element is written in the appropriate cells.A
dummy column has been added because the supply exceeds demand.

Week (Origin) Production Cost per Week (Destination) Supply

I II III IV Dummy
Rl 10 13 16 19 0 700
R2 – 10 13 16 0 700
O2 – 15 18 21 0 200
R3 – – 15 18 0 700
O3 – – 20 23 0 200
R4 – – – 15 0 700

Table 9.44 Demand 300 700 900 800 500 3,200

Apply Vogel’s method to get initial solution of the given transportation problem and then apply
MODI method to get optimal transportation schedule. Degeneracy occurs at the initial basic feasible
solution stage. Degeneracy may be removed by adding  in the cell (R2, Dummy). The optimal solution
is shown in Table 9.45.
 Transportation Problem 291

I II III IV Dummy Supply

R1 10 13 16 19 0 700
300 200 100 100
R2 – 10 13 16 0 700
700 
O2 – 15 18 21 0 200
200
R3 – – 15 18 0 700
700
O3 – – 20 23 0 200
200
R4 – – – 15 0 700
700
Table 9.45
Demand 300 700 900 800 500 3,200 Optimal Solution

The production schedule is given in Table 9.46.

Production Units For Use in Week

in Week I II III IV

I 700 300 – 200 100

RS R2 700 – 700 – –
II
OT2 Nil
RS R3 700 – – 700 –
III
OT3 Nil
Table 9.46
IV 700 – – – 700
Production
Demand 300 700 900 800 Schedule

The total minimum cost for the optimal production schedule given in Table 9.47 is
Total cost = 10 × 300 + 16 × 200 + 19 × 100 + 10 × 700 + 15 × 700 + 15 × 700 = Rs 36,100

Example 9.16 ABC company wishes to develop a monthly production schedule for the next three months.
Depending upon the sales commitments, the company can either keep the production constant, allowing
fluctuations in inventory, or inventories can be maintained at a constant level, with fluctuating production.
Fluctuating production necessitates working overtime, the cost of which is estimated to be double the normal
production cost of Rs 12 per unit. Fluctuating inventories result in inventory carryingcost of Rs 2 per
unit. If the company fails to fulfill its sales commitment, it incurs a shortage cost of Rs4 per unit per
month. The production capacities for the next three months are shown in the following table:

Production Capacity
Month Regular Overtime Sales

M1 50 30 60
M2 50 0 120
M3 60 50 40

Determine the optimal production schedule. [Delhi Univ, MBA, 2002, 2005, AMIE 2009]
Solution The given information is presented as a transportation problem in Table 9.47. The cost elements
in each cell are determined as follows:
292 Operations Research: Theory and Applications

(i) If items are produced in a month for sales during the same month, there will be no inventory carrying
cost. Thus, the total cost will either be the normal production cost or the overtime production cost.
Thus the cost elements for cells (R1, 1), (R2, 2) and (R3, 3) are Rs 12 each and for cells (O1, 1),
(O2, 3) are Rs 24 each.
(ii) If items are produced in a particular month for sales during the subsequent month, in addition to the
production costs (normal or overtime) inventory carrying cost at the rate of Rs 2 per month will be
incurred.

Cell Production Inventory Total Cost

Cost (Rs) Carrying Cost (Rs) (Rs)

(R1, 2) 12 2 14
(R1, 3) 12 4 16
(O1, 2) 24 2 26
(O1, 3) 24 4 28
(R2, 3) 12 2 14

(iii) If the company fails to fulfil its sales commitment, it incurs a shortage cost of Rs 4 per unit per month,
in addition to the production costs (normal or overtime), carrying (storage) cost at the rateof Rs 2
per month will be incurred.

Cell Production Cost Shortage Cost Total Cost

(Rs) (Rs) (Rs)

(R2, 1) 12 4 16
(R3, 2) 12 4 16
(R3, 1) 12 8 20
(O3, 2) 24 4 28
(O3, 1) 24 8 32

The solution is left as an exercise for the reader. The initial basic feasible solution obtained by Vogel’s
method (shown in Table 9.47) is also the optimal solution.

M1 M2 M3 Dummy Product
Supply

R1 12 14 16 0
50 50

O1 24 26 28 0
10 20 30

R2 16 12 14 0
50 50

R3 20 16 12 0
20 40 60

O3 32 28 24 0
30 20 50

Table 9.47 Sales Demand 60 120 40 20 240

 Transportation Problem 293

The production schedule is given in Table 9.48.

Production Units For Use in Month

in Month M1 M2 M3

M1 RS R 50 50 – –

TOR
1
30 10 20 –
1
M2 2 50 – 50 –

RS R 60 – 20 40

TO
M3 3
30 – 30 – Table 9.48
3 Production
Demand 60 120 40 Schedule

The total minimum cost for the optimal production schedule given in Table 9.49 is
Total cost = 12 × 50 + 24 × 10 + 16 × 20 + 12 × 50 + 16 × 20 + 12 × 40 + 28 × 30
= Rs 3,400.
Example 9.17 The following is the information that concerns the operations of the XYZ manufacturing
company. The production cost of the company is estimated to be Rs 5 per unit.

Month 1 Month 2

Units on order 800 1,400

Production Capacity
Regular time 920 920
Overtime 250 250
Excess cost/unit (overtime) 1.25 1.25
Storage cost/unit 0.50 0.50
Formulate and solve the above problem as transportation problem. [Delhi Univ., MBA, 2004; AMIE 2006]
Solution The storage cost of Re 0.50 per unit per month is charged only if production during the month 1 is
used for supplies during month 2.
Costs for supplies against first month’s order from the previous month have been assumed at infinity as
this is treated not only as prohibitive but also as undesirable. This is because the order quantity of first month
is even less than a regular time production capacity for the same month.
This problem is unbalanced as net supply is of 2,340 units while the demand is only of 2,200 units. A
dummy demand centre of 140 units with supply cost zero is added to solve the problem. The data are summa-
rized in Table 9.49 along with initial solution.
The initial basic solution obtained by Vogel’s method (shown in Table 9.49) is updated in order to obtain
optimal solution, which is shown in Table 9.50.
The least cost production schedule to meet the sale demand is shown below:
M1 M2 Dummy Supply M1 M2 Dummy Supply
5.0 5.5 0 5.0 5.5 0
Table 9.49
M1 800 120 920 M1 800 120 920 Initial Solution

6.25 6.75 0 6.25 6.75 0

M1(OT) 250 250 M1(OT) 110 140 250

 5.0 0  5.0 0

M2 920 920 M2 920 920

 6.25 0  6.25 0

M2(OT) 110 140 250 M2(OT) 250 250

Table 9.50
Demand 800 1,400 140 Demand 800 1,400 140 Optimal Solution
294 Operations Research: Theory and Applications

Production Supply
M1 M2 M1 M2

Regular time 920 110 800 230

Overtime 920 250 – 1,170
800 1,400

9.7 MAXIMIZATION TRANSPORTATION PROBLEM

In general, the transportation model is used for cost minimization problems. However, it may also be used
to solve problems in which the objective is to maximize total profit. That is, instead of unit cost cij, the
unit profit or payoff pij associated with each route, (i, j) is given. The objective function in terms of total profit
(or payoff) is then stated as follows:
m n
Maximize Z =   pij xij
i 1 j 1
The procedure for solving such problems is same as that for the minimization problem. However, a few
adjustments in Vogel’s approximation method (VAM) for finding initial solution and in the MODI optimality
test are required.
For finding the initial solution by VAM, the penalties are computed as difference between the largest
and next largest payoff in each row or column. In this case, row and column differences represent payoffs.
Allocations are made in those cells where the payoff is largest, corresponding to the highest row or column
difference.
Since it is a maximization problem, the criterion of optimality is the converse of the rule for minimi-
zation. The rule is: A solution is optimal if all opportunity costs d ij for the unoccupied cells are zero or
negative.
Example 9.18 A company has four manufacturing plants and five warehouses. Each plant manufactures the
same product, which is sold at different prices in each warehouse area. The cost of manufacturing and cost
of raw materials are different in each plant due to various factors. The capacities of the plants are also
different. The relevant data is given in the following table:

Plant
Item 1 2 3 4
Manufacturing cost (Rs) per unit 12 10 8 8
Raw material cost (Rs) per unit 8 7 7 5
Capacity per unit time 100 200 120 80

The company has five warehouses. The sale prices, transportation costs and demands are given in the
following table:

Warehouse Transportation Cost (Rs) per Unit Sale Price Demand

1 2 3 4 per Unit (Rs)

A 4 7 4 3 30 80
B 8 9 7 8 32 120
C 2 7 6 10 28 150
D 10 7 5 8 34 70
E 2 5 8 9 30 90

(a) Formulate this problem as a transportation problem in order to maximize profit.

(b) Find the solution using VAM method.
(c) Test for optimality and find the optimal solution.

Solution Based on the given data, the profit matrix can be derived by using following equation.
Profit = Sales price – Production cost – Raw material cost – Transportation cost
The matrix, so obtained, is shown in Table 9.51.
 Transportation Problem 295

Converting data on profit into cost by subtracting all the profit values in the table from the highest profit
value. Since highest profit value is 15, therefore, subtracting all cell values including itself from it. The new
values so obtained are shown in Table 9.52. The problem now becomes a usual cost minimizing
transportation problem.

1 2 3 4 Dummy Demand

A 6 6 11 15 0 80
B 4 6 10 12 0 120
C 6 4 7 6 0 150
D 4 10 14 14 0 70
E 8 8 7 9 0 90
Table 9.51
Supply 100 200 120 80 10 510 Profit Matrix

Apply Vogel’s method to find the initial basic feasible solution, as shown in Table 9.52.

1 2 3 4 Dummy Demand ui

A 9 9 4 0 15 80 u1 = 4
+7 +5 +4 80 + 12
B 11 9 5 3 15 120 u2 = 9
(+)
+4 70 (–) 50 –2 +7
C 9 11 8 11 15 150 u3 = 11
100 40 +1 +4 10

D 11 5 1 1 15 70 u4 = 5
+8 0 (+) 70  (–) +11
E 7 7 8 6 15 90 u5 = 7
+2 90 +5 +3 +9

Supply 100 200 120 80 10 510 Table 9.52

Initial Feasible
vj v1 = – 2 v2 = 0 v3 = – 4 v4 = – 4 v5 = – 1 Solution

Since at initial solution stage, the number of occupied cells are only 8, which is one less than the required
number, m + n – 1 = 9, therefore, the solution is degenerate. However, after making an allocation of  to the
cell (D, 4), the initial solution has now become eligible for optimality test.
Apply MODI method to evaluate each unoccupied cell in terms of opportunity cost associated with
it in the usual manner. This is shown in Table 9.52.
The cell (B, 4) has a negative opportunity cost (i.e. – 2) as shown in Table 9.52. Introduce it into the
new solution by constructing a loop shown in Table 9.52. The new solution is given in Table 9.53, where
 has shifted from cell (D, 4) to cell (B, 4).
296 Operations Research: Theory and Applications

1 2 3 4 Dummy Demand ui

A 9 9 4 0 15 80 u1 = – 5
80
+5 +3 +2 +5
B 11 9 5 3 15 120 u2 = – 2
+4
70 50  +2
C 9 11 8 11 15 150 u3 = 0
100 40 10
+1 +6
D 11 5 1 1 15 70 u4 = – 6
0 70
+8 +2 +6
E 7 7 8 6 15 90 u5 = – 4
90
+2 +5 +5 +4

Table 9.53 Supply 100 200 120 80 10 510

Optimal Solution
vj v1 = 9 v2 = 11 v3 = 7 v4 = 5 v5 = 15

Since there is no negative opportunity cost in any of the unoccupied cells in Table 9.53, therefore, this
solution is the optimal solution. However, the zero opportunity cost in cell (D, 2) indicates the existence
of an alternative solution. The total maximization profit associated with the solution is
Total profit = 9 × 70 + 5 × 50 + 9 × 100 + 11 × 40 + 15 × 10 + 1 × 70 + 7 × 90 = Rs 4,580.

9.8 TRANS-SHIPMENT PROBLEM

In a transportation problem, the shipment of a commodity takes place among sources and destinations. But
instead of direct shipments to destinations, the commodity can be transported to a particular destination
through one or more intermediate or trans-shipment points. Each of these points, in turn supply to other
points. Thus, when the shipments are done from destination to destination and from source to source, such
type of transportation problems are referred as trans-shipment problems. A trans-shipment problem involving
four sources and three destinations is shown diagrammatically in Figs. 9.2(a) and (b).

Fig. 9.2

Since the flow of commodity can be in both directions, arrows are not shown in Fig. 9.2(b). The solution
to this problem can be obtained by using the transportation model. The solution procedure is as follows:
If there are m sources and n destinations, a transportation table shall be of size (m + n) × (m + n) instead
 Transportation Problem 297

of m × n. If the total number of units transported from all sources to all destinations is N, then the given
supply at each source and demand at each destination are added to N. The demand at source and the
supply at each destination are set to be equal to N. The problem can then be solved by the usual MODI
method for transportation problems. In the final solution, ignore the units transported from a point to itself,
i.e. diagonal cells, because they do not have any physical meaning (no transportation).
Example 9.19 Consider a firm having two factories. The firm is to ship its products from the factoriesto
three retail stores. The number of units available at factories X and Y are 200 and 300, respectively, while
those demanded at retail stores A, B and C are 100, 150 and 250, respectively. Rather than shipping the
products directly from factories to retail stores, it is asked to investigate the possibility of trans-shipment.
The transportation cost (in rupees) per unit is given in the table:

Factory Retail Store

X Y A B C
Factory
RS X 0 8 7 8 9
TY 6 0 5 4 3

RA 7 2 0 5 1
Retail Store
STCB 1
8
5
9
1
7
0
8
4
0

Find the optimal shipping schedule.

Solution The number of units available at X and Y are 200 and 300, respectively and the demand at A,B
and C is 100, 150 and 250, respectively. The maximum amount which can be transported through a factoryor
retail store is the total supply and demand, i.e. N = 500 units. If all these 500 units are not transported through
a factory or retail store, then the remaining units will play the role of dummy. The trans-shipment table is
shown in Table 9.54 where 500 units have been added to the supply and demand at factory andto a retail
store.

X Y A B C Supply
X 0 8 7 8 9 200 + 500
Y 6 0 5 4 3 300 + 500
A 7 2 0 5 1 500
B 1 5 1 0 4 500
C 8 9 7 8 0 500
Table 9.54
Demand 500 500 100 + 500 150 + 500 250 + 500

The initial solution to trans-shipment problem given in Table 9.55 can be obtained by putting 500 units
to each route on the diagonal and making allocation in the matrix, by using Vogel’s approximation method.

A B C Supply

X 7 8 9 700

Y 5 4 3 800

Demand 600 650 750

Applying the MODI method to test the optimality of the solution given in Table 9.55. Since the
opportunity cost corresponding to each unoccupied cell is positive, the solution given in Table 9.55 is also
optimal. In order to interpret the optimal solution, allocations in the diagonal cells are ignored as these values
show the extra dummies that have been added in order to allow as much as flow possible.
298 Operations Research: Theory and Applications

X Y A B C Supply ui

X 0 8 7 8 9
500 100 100 700 u1 = 4
+4 +2
Y 6 0 5 4 3
500 50 250 800 u2 = 0
+10 +2
A 7 2 0 5 1
500 500 u3 = – 3
+14 +5 +4 +1
B 1 5 1 0 4
500 500 u4 = – 4
+9 +9 +2 +5
C 8 9 7 8 0
500 500 u5 = – 3
+15 +12 +7 +7

Table 9.55 Demand 500 500 600 650 750 3,000

Optimal Solution vj v1 = – 4 v2 = 0 v3 = 3 v4 = 4 v5 = 3

CONCEPTUAL QUESTIONS C

1. What is meant by unbalanced transportation problem? Explain 5. State a transportation problem in general terms and explain the
the method for solving such a problem. problem of degeneracy. How does one overcome it?
2. Explain how a profit maximization transportation problem can be 6. Explain a trans-shipment problem.
converted to an equivalent cost minimization transportation 7. What are the main characteristics of a transshipment problem?
problem.
8. Explain how a trans-shipment problem can be solved as a
3. What is degeneracy in transportation problems? How is a transportation problem.
transportation problem solved when the demand and supply are
9. What is a trans-shipment problem? Explain how it can be
not equal?
formulated and solved as a transportation problem.
4. (a) Explain how to resolve degeneracy in a transportation problem.
[Delhi Univ., MBA, 2006]
(b) How does the problem of degeneracy arise in a transportation
10. Explain the method for solving trans-shipment problem.
problem? Explain how one can overcome it.

SELF PRACTICE PROBLEMS C

1. A steel company has three open hearth furnaces and five rolling To
mills. The transportation costs (rupees per quintal) for shipping
steel from furnaces to rolling mills are given in the following table: I II III Supply
A 5 1 7 10
M1 M2 M3 M4 M5 Supply
From B 6 4 6 80

4 2 3 2 6 3 2 5
F1 8
C 15
F2 5 4 5 2 1 12
F3 14 Demand 75 20 50
6 5 4 7 7
4 4 6 8 8
Demand Since there is not enough supply, some of the demands at these
destinations may not be satisfied. Suppose there are penaltycosts
for every unsatisfied demand unit which are given by 5, 3 and
What is the optimal shipping schedule? 2 for destinations I, II and III, respectively. Find the optimal solution.
2. Consider the following unbalanced transportation problem.
 Transportation Problem 299

3. A company produces a small component for an industrial product 6. A departmental store wishes to purchase the following quantities
and distributes it to five wholesalers at a fixed delivered price of Rs of sarees:
2.50 per unit. The sales forecasts indicate that the monthly
Types of sarees : A B C D E
deliveries will be 3,000, 3,000, 10,000, 5,000 and 4,000 units to
Quantity : 150 100 75 250 200
wholesalers I, II, III, IV and V, respectively. The company’s monthly
production capacities are 5,000, 10,000 and 12,500 at plants 1, 2 Tenders are submitted by four different manufacturers who
and 3, respectively. Respective direct costs of production of undertake to supply not more than the quantities mentioned below
each unit are Re 1.00, Re 0.90, and Re (all types of sarees combined);
0.80 at plants W, X and Y. Transportation costs of shipping a
unit from a plant to a wholesaler are as follows. Manufacturer : W X Y Z
Total quantity : 300 250 150 200
Wholesaler
Sarees
I II III IV V
A B C D E
W 0.05 0.07 0.10 0.25 0.15
W 275 350 425 225 150
Plant X 0.08 0.06 0.09 0.12 0.14
X 300 325 450 175 100
Y 0.10 0.09 0.08 0.10 0.15 Manufacture
Y 250 350 475 200 125

Find how many components each plant should supply to each Z 325 275 400 250 175
wholesaler in order to maximize its profit.
4. ABC Tool Company has a sales force of 25 men, who operate from How should the orders be placed?
three regional offices. The company produces four basic product 7. A company has four factories F1, F2, F3 and F4 that manufacture the
lines of hand tools. Mr Jain, the sales manager, feels that 6 same product. Production and raw material costs differ from factory
salesmen are needed to distribute product line I, 10 to distribute to factory and are given in the following table in the first two rows.
product line II, 4 for product line III and 5 salesmen for product The transportation costs from the factories to the sales depots,
line IV. The cost (in Rs) per day of assigning salesmen from each S1, S2 and S3 are also given. The last two columns in the table give
of the offices for selling each of the product lines are as follows: the sales price and the total requirement at each depot. The
Product Lines production capacity of each factory is given in the last row.

I II III IV F1 F2 F3 F4 Sales Requirement

Price
A 20 21 16 18 per Unit
Production 15 18 14 13
Regional Office B 17 28 14 16
cost/unit
Raw material 10 9 12 9
C 29 23 19 20
cost/unit
At the present time, 10 salesmen are allocated to office A, 9 to S1  3 9 5 4 34 80
office B and 7 salesmen to office C. How many salesmen should Transportation S2  1 7 4 5 32 120
be assigned from each office to sell each product line in order cost/unit S3  5 8 3 6 31 150

to minimize costs? Identify alternate optimum solutions, if any. Supply 50 150 50 100
[Delhi Univ., MBA, 2001]
5. The Purchase Manager, Mr Shah, of the State Road Transport Determine the most profitable production and distribution
Corporation must decide on the amount of fuel that should be schedule and the corresponding profit. The deficit production
bought from three possible vendors. The corporation refuels its should be taken to yield zero profit. Write the dual of this
buses regularly at the four depots within the area of its operations. transportation problem and use it for checking the optimal solution.
The oil companies have said that they can furnish up to the [Delhi Univ., MBA, 2000]
following amounts of fuel during the coming month: 2,75,000
litres by oil company 1; 5,50,000 litres by oil company 2; and 8. Link Manufacturing Company has several plants, three ofwhich
6,60,000 litres by oil company 3. The required amount of the fuel manufacture two principal products – a standard card table and
is 1,10,000 litres by depot 1; 2,20,000 litres at depot 2; 3,30,000 deluxe card table. A new deluxe card table will beintroduced,
litres at depot 3; and 4,40,000 litres at depot 4. which must be considered in terms of selling priceand costs.
When the transportation costs are added to the bid price per The selling prices are: Standard, Rs 14.95, deluxeRs 18.85 and
litre supplied, the combined cost per litre for fuel from each vendor, New Deluxe, Rs 21.95.
servicing a specific depot, is as under:
Variable Costs (in Rs)
Company 1 Company 2 Company 3
Model Qty. Plant A Plant B Plant C Plant Capacity

Depot 1 25.00 24.75 24.25 Standard 450 8.00 7.95 8.10 A 800
Deluxe 1,050 8.50 8.60 8.45 B 600
Depot 2 25.00 25.50 26.75 New Deluxe 600 9.25 9.20 9.30 C 700

Depot 3 24.50 26.00 25.00 Solve this problem by the transportation technique for the
maximum contribution.
Depot 4 25.50 26.00 24.50
9. City Super Market keeps five different patterns of a particular
size of readymade garment for sale. There are four manufacturers
Determine the optimal schedule. [Delhi Univ., MBA, 2002]
300 Operations Research: Theory and Applications

available to the manager of the market to whom be can order

Investment Net Return Data Amount
the required quantity. The demand for the coming season and
Made at the (in paise) of Available
the maximum quantity that can be produced by a manufacturer
Beginning Selected Investment (lacs)
is as follows:
of Year
P Q R S
Pattern 1 2 3 4 5 1 95 80 70 60 70
U. Cut P. Form Maxi Mini Midi 2 75 65 60 50 40
Quantity required 200 150 200 150 300 3 70 45 50 40 90
4 60 40 40 30 30
Manufacturer A B C D
Max.
Capacity 200 300 250 250 Investment
(Rs lakh) 40 50 60 60 –
The following quotations have been submitted by the
manufacturers for different patterns: The following additional information is also provided:
(a) P, Q, R and S represent the selected investments.
1 2 3 4 5
(b) The company has decided to have a four years investment
U. Cut P. Form Maxi Mini Midi
plan.
A Rs 100 Rs 130 Rs 160 Rs 70 Rs 140 (c) The policy of the company is that the amount invested in
any year will remain so until the end of fourth year.
B 100 120 150 75 130 (d) The values (paise) in the table represent net return on
C 115 140 155 80 120 investment of one rupee till the end of the planning horizon
(for example, a rupee in investment P at the beginning of year
D 125 145 140 60 125 will grow to Rs 1.95 by the end of the fourth year, yielding a
return of 95 paise).
The super market is selling different patterns at the following
prices: Using the above, determine the optimum investment strategy.
12. A manufacturer must produce a certain product in sufficient
Pattern : 1 2 3 4 5 quantity in order to meet contracted sales of the next four months.
Selling Price (Rs) : 120 150 170 80 140 The production facilities available for this product are limited, and
How should the super market manager place the order? Write vary in different months. The unit cost of production also changes
the dual of this transportation problem and give its economic accordingly to the facilities and personnel available. The product
interpretation. may be produced in one month and then held for sale in a later
month, at an estimated storage cost of Re 1 perunit per month.
10. A company wishes to determine an investment strategy for each No storage cost is incurred for goods that aresold in the same
of the next four years. Five investment types have been selected. month in which they are produced. Presently there is no inventory
The investment capital has been allocated for each of the coming of this product and none is desired at theend of four months.
four years, and maximum investment levels have been established Given the following table, show how much to produce in each of
for each investment type. There is an assumption that amounts four months in order to minimize total cost.
invested in any year will remain invested until the end of the planning
horizon of four years. The following table summarizes the data of
Month Contracted Maximum Unit Cost of Unit Storage
this problem. The values in the body of the table represent the net
Sales Production Production Cost per Month
return on investment of one rupee up to the end of the planning
(in units) (in units) (Rs) (Rs)
horizon. For example, a rupee invested in investment type B at the
beginning of year will grow to Rs 1.90 by the end of the fourth year, 1 20 40 14 1
yielding a net return of Re 0.90. 2 30 50 16 1
3 50 30 15 1
Investment Investment Type Rupees 4 40 50 17 1
Made at the Available
Beginning A B C D E (in ’000s) Formulate the problem as a transportation problem and solve it.
of Year 13. A company has factories at A, B and C, which supply its products
1 0.80 0.90 0.60 0.75 1.00 500 to warehouses at D, E, F and G. The factorycapacities are 230,
2 0.55 0.65 0.40 0.60 0.50 600 280 and 180, respectively for regular production. If overtime
3 0.30 0.25 0.30 0.50 0.20 750 production is utilized, the capacities can be increased to 300, 360
4 0.15 0.12 0.25 0.35 0.10 800 and 190, respectively. Increment unit overtime costs are Rs 5, Rs
4 and Rs 6, respectively. The current warehouse requirements
Maximum are 165, 175, 205 and 165, respectively. Unit shipping costs in
rupees 750 600 500 800 1,000 rupees between the factories and the warehouses are:
investment To
(in ’000s)
The objective in this problem is to determine the amount to be D E F G
invested at the beginning of each year in an investment type so
as to maximize the net rupee return for the four year period. A 6 7 8 10
Solve the above transportation problem and get an optimal
solution. Also calculate the net return on investment for the From B 4 10 7 6
planning horizon of four-year period.
C 3 22 2 11
11. XYZ Company has provided the following data and seek your
advice on the optimum investment strategy:
Determine the optimum distribution for the company to minimize
costs.
 Transportation Problem 301

14. The products of three plants X, Y and Z are to be transported company standards and evaluated. The applicants have been
to four warehouses I, II, III and IV. The cost of transportation of each grouped into four categories by their abilities; the grouping and
unit from plants to the warehouses along with the normal capacities values of each category to the company are shown in the table
of plants and warehouses are indicated below: below:

Warehouse Applicant Category Value (Rs per month) Number of

I II III IV Available Category A B C D E F Applicants

X 25 17 25 14 300 I 1,000 1,100 1,500 1,400 1,400 1,450 54

II 1,200 1,250 1,200 1,350 1,400 1,400 54
Plant Y 15 10 18 24 500 III 1,000 1,100 1,200 1,400 1,500 1,600 45
IV 1,500 1,500 1,600 1,400 1,400 1,500 74
Z 16 20 8 13 600

Required 300 300 500 500 How many applicants of each category should the personnel
manager hire and for which jobs?
(a) Solve the problem for minimum cost of transportation. Are 17. Debonair Private Ltd. is in the business of manufacturing and

there any alternative solutions? If any, explain the selling office shirts for men. It has four factories located in different
methodology. parts of the country and the monthly capacities of the factories in
(b) Overtime can be used in each plant to raise the capacity thousand are as given below. The shirts are made in a few
by 50 per cent of the normal but the corresponding cost standard designs and colours, and each factory canmake all types
of trans-shipment will also increase by 10, 15 and 20 tothe of shirts in any size subject to the overall capacity of the factory.
unit costs of production at each plant. Factories : I II III IV
15. A firm manufacturing industrial chemicals has got 3 plants P1, P2 Monthly capacity : 3 4.5 2.5 5
and P3 each having capacities to produce 300 kg, 200 kg, and From the factories, the shirts are transported to five warehouses
500 kg, respectively of a particular chemical per day. The located in five different regions in India. The warehouses in turn
production costs per kg in plants P1, P2 and P3, respectively are Re supply to the distributors and the retailers. The monthly demand
0.70, Re 0.60 and Re 0.66. Four bulk consumers have placed of shirts (in thousand) from the warehouses is as follows:
orders for the products on the following basis:
Warehouses : A B C D E
Consumer kg Required Price Offered
per Day Rs/kg Monthly capacity : 3 5 1.5 2 2.5

I 400 1.00 The cost of transporting a shirt from a factory to a warehouse

depends on the distance between them and the cost oftransporting
II 250 1.00
a shirt from each factory to each warehouse is given in the table
III 350 1.02 below:
IV 150 1.03
Warehouse
Shipping costs (paise per kg) from plants to consumers are A B C D E
given in the table below:

Consumer I 6 3 4 2 5
C1 C2 C3 C4
II 11 7 5 10 9
P1 3 5 4 6 Factory III 10 7 1 2 8
Plant P2 8 11 9 12 IV 12 10 5 3 5
P3 4 6 2 8
How many shirts are to be produced, in which factory, and how

Work out an optimal schedule for the above situation. Under what are these to be dispatched to the warehouse so that the total
conditions would you change schedule? cost involved in transportation is minimized.
(a) Use the North-West Corner Method to get an initial feasible
16. The personnel manager of a manufacturing company is in the
solution.
process of filling 175 jobs in six different entry level skills due
(b) Check if the solution obtained in (a) above is an optimal
to the establishment of a third shift by the company. Union wage
allocation and if not, then find the optimal solution.
scale and requirement for the skills are shown in the following table:
Pay Scales and Skills Requirements 18. A manufacturer of jeans is interested in developing an advertising
campaign that will reach four different age groups. Advertising
Entry Level Skills A B C D E F campaigns can be conducted through TV, radio and magazines.
The following table gives the estimated cost in paise per exposure
Wage scale 1,000 1,100 1,200 1,300 1,400 1,500 for each age group according to the medium employed. In addition,
Rs/month maximum exposure levels possible in each of the media, namely
TV, radio and magazines are 40, 30 and 20 millions, respectively.
No. required 25 29 31 40 33 17
Also the minimum desired exposure within each age group, namely
230 applicants for the jobs have been tested and their aptitudes and 13–18, 19–25, 26–35, 36 and older are 30, 25, 15 and 10 millions.
skills for the jobs in question have been matched against The objective is to minimize the cost of attaining the minimum
exposure level in each age group.
302 Operations Research: Theory and Applications

Media Age Groups To

13–18 l9–25 26–35 36 and older A B C Supply
TV 12 7 10 10 X 7 3 4 2
Radio 10 9 12 10 From Y 2 1 3 3
Magazine 14 12 9 12
Z 3 4 6 5

(a) Formulate the above as a transportation problem, and find Demand 4 1 5

the optimal solution.
(b) Solve this problem if the policy is to provide at least 4million 22. A manufacturer wants to ship 8 loads of his product as shown
exposures through TV in the 13–18 age group in the table. The matrix gives the mileage from origin to destination.
and at least 8 million exposures through TV in the age group Shipping costs are Rs 10 per load per mile. What shipping
19–25. schedule should be used?

19. Two drug companies have inventories of 1.1 and 0.9 million doses D1 D2 D3 Supply
of a particular flu vaccine, and an epidemic of the flu seems
imminent in three states. Since the flu could be fatal to senior
O1 50 30 220 1
citizens, it is imperative that they be vaccinated first; others will be
vaccinated on a first-come-first-served basis while the vaccine O2 90 45 170 3
supply lasts. The amounts of vaccine (in millions of doses) each
state estimates it could administer are as follows: O3 250 200 50 4
State 1 State 2 State 3 Demand 4 2 2
Elders 0.325 0.260 0.195 Trans-shipment
Others 0.750 0.800 0.650 23. Consider the following trans-shipment problem with two sources
The shipping costs (paise per dose) between drug companies and three destinations, the cost for shipments (in rupees) is given
and states are as follows: below.

State 1 State 2 State 3 Source Destination Supply

Company 1 30 30 60 S1 S2 D1 D2 D3

Company 2 10 40 70  S1 0 80 10 20 30 100 + 300

Source  S2 10 0 20 50 40 200 + 300
Determine a minimum-cost shipping schedule which will provide
each state with at least enough vaccine to care for its senior  D1 20 30 0 4 10 300
citizens. Write the dual of this transportation problem and use
Destination  D2 40 20 10 0 20 300
it for checking the optimal solution.
 D3 60 70 80 20 0 300
Demand 300 300 100 + 300 100 + 300 100 + 300
[Delhi Univ., MBA (HCA), 2000, 2005]
20. A leading firm has three auditors. Each auditor can work up to Determine the optimal shipping schedule.
160 hours during the next month, during which time three projects
must be completed. Project 1 will take 130 hours, project 2 will 24. A firm having two sources, S1 and S2 wishes to ship its product
take 140 hours, and project 3 will take 160 hours. The amount per to two destinations, D1 and D2. The number of units available at
hour that can be billed for assigning each auditor to each project S1 and S2 are 10 and 30 and the product demanded at D1 and D2
is given in the table. are 25 and 15 units, respectively. The firm instead of shipping from
Project sources to destinations, decides to investigate the possibility of
transshipment. The unit transportation cost (in Rs) is given in
Auditor 1(Rs) 2(Rs) 3(Rs) the following table:
1 1,200 1,500 1,900
Source Destination Supply
2 1,400 1,300 1,200
3 1,600 1,400 1,500 S1 S2 D1 D2
S
Source 1 0 3 4 5 10 + 40
Formulate this as a transportation problem and find the optimal 
solution. Also find out the maximum total billing during the next  S2 3 0 3 5 30 + 40
month. D
Destination 1 4 3 0 2 40
Demand D2 5 5 2 0 40
Degeneracy 40 40 25 + 40 15 + 40
21. Obtain an optimum basic feasible solution to the following 
degenerate transportation problem. Determine the optimal shipping schedule.

HINTS AND ANSWERS

1. Total requirement (30) < Total capacity (34), add a dummy mill Ans. x12 = 4, x14 = 4, x24 = 2, x25 = 8, x31 = 4, x33 = 6 and
with requirement (34 – 30) = 4. Degeneracy occur at the initial x36 = 4. Total cost = Rs 80.
solution (VAM).
 Transportation Problem 303

2. Demand (145) > Supply (105), add dummy source with supply x22 = 165, x31 = 220, x33 = 110, x43 = 440, x52 = 385.
(145 – 105) = 40 and transportation costs 5, 3 and 2 for destination Total cost = Rs 51,700.
1, 2 and 3, respectively. 6. Profit maximization problem. First add a dummy column with
Ans. x12 = 0, x21 = 60, x22 = 10, x23 = 10, x31 = 5 and x43 = 40. demand of 125 sarees and then subtract all elements of the profit
Transportation cost = Rs 515. Penalty for transportation of 40 matrix from the highest element 475.
units to destination 3 at the cost of Rs 2 per unit = Rs 80. Thus,
total cost = 515 + 80 = Rs 595. Manufacturer Variety Quantity
3. Maximization as well as unbalanced problem. Total capacity
W B, D, E 25, 50, 200
(27,500) > demand (25,000). Add a dummy wholesaler with
demand 2,500 units. Since direct costs of production of each unit X A 150
are Re 1, Re 0.90 and Re 0.80 at plants 1, 2 and 3, add this cost Y B, C 75, 75
to each figure row wise to get exact data of the problem. Z D 200
Ans. x11 = 2,500, x21 = 500, x22 = 3,000, x23 = 2,500, x25 = 7. Profit maximization problem. First add dummy row with supply
4,000, x33 = 7,500 and x34 = 5,000. Total cost Rs 23,730. of 40 units and then subtract all elements of the profit matrix from
Since in total 25,000 units are supplied to the wholesalers highest element 8.
at a fixed price of Rs 2.50 per unit, therefore, the total cost is
Rs 62,500. Profit = Sales price – (Production cost + Raw material cost
The net maximum profit to the manufacturer = Rs 62,500 – + Transportation cost)
Rs 23,730 = Rs 38,770
Factory Sales Depot Unit Profit
4. Supply (26) > Demand (25), add dummy column with demand 1.
Ans. x12 = 4, x13 = 1, x14 = 5; x21 = 6, x23 = 3, x32 = 2, x35 F1 S2 6
= 1 Total cost = Rs 472 F2 S 2, S 3 – 2, – 4
Alternative solution exists because opportunity cost is zero in F3 S3 2
cell (B, 4) and (C, 4). F4 S 1, S 2 8, 5
5. Supply (14,85,000) > Demand (11,00,000), add dummy column
with demand of 3,85,000 litres of oil; x13 = 110, x21 = 55, Total profit = Rs 2,000
12. Transportation cost table and optimal solution

From/To 1 2 3 4 Dummy Supply

1 14 15 16 17 0
20 20 40

2  16 17 18 0
10 20 20 50

3   15 16 0
30 30

4    17 0
40 10 50

Demand 20 30 50 40 30 170

13. Transportation cost table and optimal solution

D E F G Dummy Supply
A 6 7 8 10 0
40 175 15 230

B 4 10 7 6 0
115 165 280
C 3 22 2 11 0
180 180
A1 11 12 13 15 0
70 70

B1 8 14 11 10 0
10 70 80

C1 9 28 8 17 0
10 10

Demand 165 175 205 165 140 850

304 Operations Research: Theory and Applications

22. Using NWCM, initial solution is degenerate. 23. Ans. x12 = 10, x13 = 20, x15 = 10, x21 = 20, x31 = 5,
Ans. x13 = 2, x22 = l, x23 = 2, x31 = 4 and x33 = 1, x34 = 30, x35 = 5, Total cost = Rs 3,600.
Total cost = Rs 33.

CHAPTER SUMMARY
The transportation problem (a special case of LP problem) is concerned with distributing a commodity from given sources to their
respective destinations. Each source has a fixed supply of the commodity and each destination has a fixed demand. A basic
assumption is that the cost of distribution from each source to each destination is directly proportional to the amount distributed.
Formulating a transportation problem requires constructing a special purpose table that gives the unit costs of distribution, the
supplies and the demands.
The objective of this chapter is to enable readers to recognize a problem that can be formulated and analyzed as atransportation
problem or as a variant of one of these problem types.

CHAPTER CONCEPTS QUIZ

True or False
1. The test for degeneracy of a solution is to check if there are unused 14. An improvement index in the transportation method is analogous
cells with an improvement index equal to zero. If so, degeneracy to a value in the quantity column in the method.
exists. 15. The serves the same purpose for the transportation
2. The advantage of the most method over the stepping stone method method as all slack variables in the simplex method.
of computing improvement indices for unused cells lies in the 16. An unused cell in the transportation table is analogous to a variable
greater computational efficiency. not in the solution column in the simplex method.
3. If you have to solve a transportation problem with m rows and 17. The end result of an iteration in the transportation method is the
n columns using the simplex method, you would have to formulate same as the end result in the simplex method in that value of
the problem with m variables and n constraints. the solution is .
4. A solution to the cost minimization transportation problem is 18. In Vogel’s approximation method, each allocation is made on the
optimal when all unused cells have an improvement index which basis of the cost that would have been incurred if
is non-negative. allocation in certain cells with unit transportation
5. A dummy row or column is introduced in the transportation method cost were missed.
is order to handle an unbalanced problem. The dummy serve, the 19. method is based on the concept of duality.
same purpose as a slack variable in the simplex method. 20. If the total supply total demand, then an additional
6. Each iteration of the transportation method involves the elimination column known as added to the transportation tableto
of one occupied cell and the introduction of one unoccupied cell absorb the same.
which is similar to a pivot in the simplex method.
7. In the transportation problem, extra constraint equation can not Multiple Choice
be derived from any other constraint equations as it affects the
21. The initial solution of a transportation problem can be obtained
feasible solution of the problem.
by applying any known method. However, the only condition is
8. In the north-west corner method, the cost of transportation on that
any route of transportation is taken into account. (a) the solution be optimal
9. The unbalanced transportation problem can be balanced by adding (b) the rim conditions are satisfied
a dummy supply row or a demand column as per the need. (c) the solution not be degenerate
10. In the transportation problem, the rim requirement for a row is (d) all of the above
the capacity of a supplier, white a rim requirement for a column 22. The dummy source or destination in a transportation problem is
is the demand of a user. added to
(a) satisfy rim conditions
Fill in the Blanks (b) prevent solution from becoming degenerate
11. The northwest corner rule provides a for obtaining (c) ensure that total cost does not exceed a limit
an solution to the transportation problem. (d) none of the above
12. The degeneracy may occurs when there are two or more cells 23. The occurrence of degeneracy while solving a transportation
with the same smallest value in a closed problem means that
for an incoming cell. (a) total supply equals total demand
13. We have alternative optimal solutions to a minimization (b) the solution so obtained is not feasible
transportation problem whenever we find a solution where the (c) the few allocations become negative
improvement indices are all with at least (d) none of the above
to zero. 24. An alternative optimal solution to a minimization transportation
problem exists whenever opportunity cost corresponding to unused
route of transportation is:
 Transportation Problem 305

(a) positive and greater than zero 30. During an iteration while moving from one solution to the next,
(b) positive with at least one equal to zero degeneracy may occur when
(c) negative with at least one equal to zero (a) the closed path indicates a diagonal move
(d) none of the above (b) two or more occupied cells are on the closed path but neither
25. One disadvantage of using North-West Corner Rule to find initial of them represents a corner of the path.
solution to the transportation problem is that (c) two or more occupied cells on the closed path with minus sign
(a) it is complicated to use are tied for lowest circled value
(b) it does not take into account cost of trans-portation (d) either of the above
(c) it leads to a degenerate initial solution 31. The large negative opportunity cost value in an unused cell in a
(d) all of the above transportation table is chosen to improve the current solution be-
26. The solution to a transportation problem with m-rows (supplies) and cause
n-columns (destination) is feasible if number of positive allocations (a) it represents per unit cost reduction
are (b) it represents per unit cost improvement
(a) m + n (c) it ensure no rim requirement violation
(b) m × n
(d) none of the above
(c) m + n – 1
32. The smallest quantity is chosen at the corners of the closed path
(d) m + n + 1 with negative sign to be assigned at unused cell because
27. The calculation of opportunity cost in the MODI method is (a) it improve the total cost
analogous to a
(b) it does not disturb rim conditions
(a) cj – zj value for non-basic variable columns in the simplex
(c) it ensure feasible solution
method
(d) all of the above
(b) value of a variable in xB-column of the simplex method
33. When total supply is equal to total demand in a transportation
(c) variable in the B-column in the simplex method
problem, the problem is said to be
(d) none of the above
(a) balanced
28. An unoccupied cell in the transportation method is analogous to
(b) unbalanced
a
(c) degenerate
(a) cj – zj value in the simplex table (d) none of the above
(b) variable in the B-column in the simplex table
34. Which of the following methods is used to verify the optimality of
(c) variable not in the B-column in the simplex table the current solution of the transportation problem,
(d) value in the xB-column in the simplex table (a) Least cost method
29. If an opportunity cost value is used for an unused cell to test (b) Vagel’s approximation method
optimality, it should be (c) Modified distribution method
(a) equal to zero
(d) all of the above
(b) most negative number 35. The degeneracy in the transportation problem indicates that
(c) most positive number (a) dummy allocation (s) needs to be added
(d) any value (b) the problem has no feasible solution
(c) the multiple optimal solution exist
(d) (a) and (b) but not (c)

Answers to Quiz
1. F 2. T 3. F 4. T 5. T 6. T 7. F 8. F 9. T 10. T
11. mechanism; initial 12. negative; path 13. non-negative; equals 14. simplex
15. north west corner rule 16. product mix 17. improved 18. opportunity; minimum
19. modified, distribution 20. exceed; dummy demand centre
21. (b) 22. (a) 23. (b) 24. (b) 25. (b) 26. (c) 27. (a) 28. (c) 29. (b) 30. (c)
31. (a) 32. (c) 33. (a) 34. (c) 35. (d)

CASE STUDY
Case 9.1: Asian Games*
For maintaining law and order during the Asian Games, 1982 police force has been requisitioned from the various central
police organizations like CRPF, BSF, ITBP and other states such as Haryana, Himachal Pradesh, Uttar Pradesh, Orissa and
Gujarat. Besides this there are 10 batalions of Delhi Police available for deployment. The outside and the local force have
been stationed at various places in the Union Territory of Delhi. The figures are in the number of sections (10 Men)
available for deployment.
306 Operations Research: Theory and Applications

Place Abbreviation Force (Section)

New Police Lines NPL 400

Jharoda Kalan JK 160
Pitam Pura PP 160
Model Town MT 90
Police Training School PTS 90
Old Police Lines OPL 120
New Kotwali NK 20
Kamala Market KMT 120
Parliament Street PST 240
Rajpura Lines RPL 120
Shakarpur SP 120
Moti Nagar MN 120
Mehram Nagar MRN 240
Total 2,000

The force is to be deployed at various stadia on the day of events. The requirement of the force has been calculated
keeping in view the factors like capacity of the stadium, traffic problems, security of VIPs and participants etc. The
requirement of force is also mentioned in number of sections.

Stadium Abbreviation Force (Section)

Chhatrasal CS 60
Delhi University DU 40
Indraprastha IP 400
Yamuna Velodrome YV 40
Talkatora Indoor TKI 80
Talkatora Swimming TKS 80
National NS 240
Hall of sports HS 60
Harbaksh HB 60
Shivaji SV 60
Jawaharlal Nehru JLN 500
Hauz Khas HK 20
Games Village GV 60
Ambedkar AK 200
Tughlakabad Range TR 40
Golf Club GC 20
Karnail Singh KS 40
Total 2,000

The cost of transportation of one section, i.e. 10 Men from every location to each stadium has been calculated as
shown in the matrix below.
It is assumed that the availability and requirement of force is equal, i.e. 2,000 sections. It will not be possible on any
day during the games as all stadia are not having games on all the days. Hence, there will always be a surplus force available.
Consequently dummy stadium will have to be planned everyday so as to absorb the unutilized force. The cost (in Rs 100s
for 10 persons – one section) matrix for the movement of the police force is given below:

* Based on the class assignment prepared by the MBA(FMS) student Mr M.S Upadhya, IPS.
 CS DU IP YV TKI TKS NS HS HB JLN SV HK GV AK TR GC KS Force
Availability

NPL 2 1 13 12 15 15 16 14 25 20 14 22 24 11 35 16 10 400
JK 28 31 36 37 130 30 35 36 16 42 30 35 36 35 50 35 36 160
PP 15 14 26 25 28 28 28 17 38 33 17 35 37 24 48 29 23 160
MT 4 3 15 14 17 17 18 16 27 22 16 24 26 13 37 18 12 90
PTS 26 24 15 16 13 13 12 15 15 12 15 33 3 16 10 17 16 90
OPL 7 2 8 7 10 10 11 13 20 15 9 17 13 6 30 11 6 120
NK 11 6 4 3 6 6 7 9 16 11 5 13 15 2 26 7 4 20
KMT 13 8 3 4 4 4 5 5 21 16 2 25 25 1 24 10 2 120
PST 15 12 5 6 2 2 4 5 16 10 12 12 12 4 22 4 3 240
RPL 4 1 11 10 13 13 14 14 23 18 12 20 22 9 33 14 8 120
SP 19 16 3 4 7 7 6 5 21 12 8 18 18 5 30 8 8 120
MN 10 12 16 17 13 13 15 16 20 20 15 25 25 14 35 18 12 120
MRN 30 28 18 19 15 15 15 16 6 22 15 12 12 20 20 20 20 240
Force
Requirement 60 40 400 40 80 80 180 240 60 60 500 60 20 200 40 20 40 2,000

Suggest an optimal transportation schedule of the police force so as to minimize the total transportation cost.

APPENDIX: THEOREMS AND RESULTS

Theorem 9.1 (Existence of Feasible Solution) A necessary and sufficient condition for the existence of a feasible
solution to the transportation problem is
m n
 ai =  b j (Rim condition)
i 1 j 1

That is, the total capacity (or supply) must equal total requirement (or demand).
Proof (a) Necessary Condition Let there exist a feasible solution to the transportation problem. Then, we have
m n m
  xij =  ai (1)

i 1 j 1 i 1
n m n
and   xij =  b j (2)

j 1 i 1 j 1
m n

Since the left-hand side of (1) and (2) are the same, therefore  ai =  b j
i 1 j 1
(b) Sufficient Condition Suppose

m n
 ai =  b j = k (say)
i 1 j 1

If there exists a real number i  0 such xij = i bj for all i and j, then value of i is given by
n n n 1 n ai
 xij =   ib j = i  b j = ki or i =  xij =
j 1 j 1 j 1 k j 1 k

ai b j
Thus xij = i bj = for all i and j.
k

Since ai > 0 and bj > 0 for all i and j, therefore ai bj /k  0 and hence a feasible solution exists, i.e. xij  0.
Theorem 9.2 (Basic Feasible Solution) The number of basic variables (positive allocations) in any basic feasible
solution are m + n – 1 (the number of independent constraint equations) satisfying all the rim conditions.
Proof In the mathematical model of a transportation problem it can be seen that there are m rows (capacity or supply
constraint equations) and n columns (requirement or demand constraint equations). Thus there are in total m + n constraint
equations. But because of Theorem 9.1 (total capacity be equal to the total requirement) out of m + n constraint equations
one of the equations is redundant and can also be eliminated. Thus there are m + n – 1 linearly independent equations. It
 Transportation Problem 307

can be verified by adding all the m rows equations and subtracting from the sum the firstn – 1 column equations,
thereby getting the last column equation. That is,
m n n 1 m m n 1
  xij –   xij =  ai –  b j

i 1 j 1 j 1 i 1 i 1 j 1
 ML  PO = ML  P
O
m n n m m m n
  xij –  xij   xin  ai – b j  bn

i 1 j 1
m
MN j  1i  1 i1
m
PQ i 1
m
MN j 1 PQ
 xin = bn ; since  ai =  b j
i 1 i 1 j 1

Triangular Basis We know that the number of basic variables are equal to the number of constraints in linear
programming. In the same way when capacity and requirement constraint equations are expressed in terms of basic
variables and all non-basic variables are given zero value, the matrix of coefficients of variables in the system of equations
is triangular, that is, there is an equation in which a single basic variable occurs; in the second equation one more basic
variable is added but the total number of variables does not exceed two; similarly, in the third equation one more basic
variable occurs, but total does not exceed three, and so on.
Theorem 9.3 The transportation problem has a triangular basis.
Proof To prove this theorem, consider capacity and requirement constraint equations represented in the tabular form
(Table 9.1).
In the system of capacity and requirement constraint equations, every equation has a basic variable otherwise
the equation cannot be satisfied for ai  0 or bj  0. Suppose, every row and column equation has at least two basic
variables. Since there are m rows and n columns, therefore, the total number of basic variables in row equations and
column equations will be at least 2m and 2n, respectively. If N be the total number of basic variables, then obviously
N  2m, N  2n. Now three cases may arise:
Case I : If m > n, then m + m > m + n, or 2m > m + n. Thus N  2m  m + n.
Case II : If m < n, then m + n < n + n or m + n < 2n. Thus N  2n > m + n.
Case III : If m = n, then m + m = m + n or 2m = m + n. Thus N  2m = m + n.
In each of these cases, we observed that N > m + n. But the number of basic variables in the transportation problem
are N = m + n – 1. This is a contradiction. Thus our assumption that every equation has at least two basic variables is wrong.
Therefore, there is at least one equation, either row or column, having only one basic variable.
Let the rth equation have only one basic variable, and let xrt be the only basic variable in row r and column t. Then
xrt = ar. Eliminate rth row from the system of equations and substituting xrt = ar in t-th column equation and
replace bt by bt = bt – ar.
After eliminating the r-th row, the system has m – 1 row equations and n column equations of which
m + n – 2 are linearly independent. This implies that the number of basic variables are m + n – 2. Repeating the argument
given earlier and conclude that in the reduced system of equation, there is an equation which has only one basic variable.
But if this equation happens to be the t-th column equation in the original system, then it will have two basic variables.
This indicates that in our original system of equations, there is an equation which has at least two basic variables.
Continue, repeating the argument to prove that the system has an equation which has at least three basic variables
and so on.

Unbalanced Transportation Problem For a feasible solution to exist it is necessary that the total
supply must equal total demand. That is,
m n
 ai =  b j
i 1 j 1

But a situation may arise when the total available supply is not equal to the total requirement.
Case 1: When the supply exceeds demand, the constraints of the transportation problem will appear as,
n
 xij  ai ; i = 1, 2, . . ., m
j 1
m
 xij = bj ; j = 1, 2, . . ., n
i 1

xij  0 for all i, j

Adding slack variables si, n + 1, (i = 1, 2, . . ., m) in the first m constraints, we get

m
 xij + si, n + 1 = ai
j 1
308 Operations Research: Theory and Applications
 m
x s = a
m  n


   ij i, n  1  i
i  1
 j 1  i 1
 m m n
 si, n  1 =  ai   b j = excess supply available.

i 1 i 1 j 1

If bn + 1 denotes the excess supply available, then the modified transportation model can be stated as
m n

Minimize Z =    cij xij  ci, n  1 si, n  1 

i 1 j 1
n
subject to  xij  si, n  1 = ai ; i = 1, 2, . . ., m
j 1
n

 xij = bj ; j = 1, 2, . . ., n + 1
j 1

and xij  0 for all i, j

where ci, n + 1 = 0 (i = 1, 2, . . ., m) and
m n m n
 ai =  b j  bn  1 or bn + 1 =  ai –  b j
i 1 j 1 i 1 j 1
m n

It follows that, if  ai >  b j , then a dummy column (demand centre) can be added to the transportation
i 1 j 1

table to absorb the excess supply. The unit transportation cost for the cells in this column is set equal to zero because
these represent product items that are not being made and not being sent.
Case 2: When demand exceeds supply, the constraints of the transportation table will appear as
n
 xij = ai ; i = 1, 2, . . ., m
j 1
m
 xij  bj ; j = 1, 2, . . ., n
i 1

and xij  0 for all i, j

Adding slack variables sm + 1, j ( j = 1, 2, . . ., n) in the last n constraints, we get
n
 xij = ai ; i = 1, 2, . . ., m
j 1
m

 xij  sm  1, j = bj ; j = 1, 2, . . ., n
i 1

n n m
 sm  1, j =  b j   ai = excess demand.
j 1 j 1 i 1

If am + 1 denotes the excess demand, then the modified transportation model can be stated as:

e j
m m
Minimize Z =   c x  c
ij ij m  1, j sm  1, j
i 1 j 1
n

subject to  xij = ai ; i = 1, 2, . . ., m + 1
j 1
m
 xij + sm + 1, j = bj ; j = 1, 2, . . ., n
i 1
 Transportation Problem 309

and xij  0 for all i, j

where cm + 1, j = 0, for all j and
m n n m
 ai + a m + 1 =  b j or am + 1 =  bj
i 1 j 1 j 1
n m – 
a
i
i

1
It follows that if  bj >  ai , then a dummy row (supply centre) can be added to the transportation table
j 1 i1

to account for excess demand quantity. The unit transportation cost here also for the cells in the dummy row is set
equal to zero.
 10
“We don’t have as many managers as we should, but we would rather have too few than too
many.”
– Larry Page

An assignment problem is a particular case of a transportation problem where the given resources are
allocated to an equal number of activities with an aim of either minimizing total cost, distance, time or
maximizing profit. Travelling salesman is a specific application of assignment model.

LEARNING OBJECTIVES

After studying this chapter, you should be able to

 understand the features of assignment problems and transportation problems.
 formulate an assignment problem as a square matrix.
 apply the Hungarian method to solve an assignment problem.
 make appropriate changes in the Hungarian method to solve an unbalanced assignment problem,
profit maximization assignment problem, etc.
 solve a travelling salesman problem.

CHAPTER OUTLINE
10.1 Introduction 10.5 A Typical Assignment Problem
10.2 Mathematical Model of Assignment Problem 10.6 Travelling Salesman Problem
10.3 Solution Methods of Assignment Problem  Self Practice Problems C
 Conceptual Questions A  Hints and Answers
 Self Practice Problems A  Chapter Summary
 Hints and Answers  Chapter Concepts Quiz
10.4 Variations of the Assignment Problem  Case Study
 Conceptual Questions B  Appendix: Important Results and Theorems
 Self Practice Problems B
 Hints and Answers
 Assignment Problem 311

10.1 INTRODUCTION
An assignment problem is a particular case of a transportation problem where the resources (say facilities)
are assignees and the destinations are activities (say jobs). Given n resources (or facilities) and n activities
(or jobs), with effectiveness (in terms of cost, profit, time, etc.) of each resource for each activity. Then
problem becomes to assign (or allocate) each resource to only one activity (job) and vice-versa so that the
given measure of effectiveness is optimized.
The problem of assignment arises because the resources that are available such as men, machines, etc.,
have varying degree of efficiency for performing different activities. Therefore, the cost, profit or time of
performing different activities is also different. Thus, the problem becomes: How should the assignments
be made in order to optimize the given objective.
Some of the problems where the assignment technique may be useful are assignment of (i) workers
to machines, (ii) salesmen to different sales areas, (iii) clerks to various checkout counters, (iv) classes to Assignment table
rooms, (v) vehicles to routes, (vi) contracts to bidders, etc. is a convenient way
to summarize
available data.
10.2 MATHEMATICAL MODEL OF ASSIGNMENT PROBLEM
The general data matrix for assignment problem is shown in Table 10.1. It may be noted that this data matrix
is the same as the transportation cost matrix except that the supply (or availability) of each of the resources
and the demand at each of the destinations is taken to be one. It is due to this fact that assignments are
made on a one-to-one basis.

Resources Activities ( jobs) Supply

(workers) J1 J2 . . . Jn
W1 c11 c12 . . . c1n 1
W2 c21 c22 . . . c2n 1
. . . . .
. . . . .
. . . . .
Table 10.1
Wn cn1 cn2 . . . cnn 1 Assignment Data
Demand 1 1 . . . 1 n Matrix

Suppose, xij represents the assignment of resource (facility) i to activity (job) j such that

1 if resource i is assigned to activity j

xij  
0 otherwise
Then mathematical model of the assignment problem can be stated as:
n n
Minimize Z    cij xij
i 1 j 1

subject to the constraints

n
 xij = 1, for all i (resource availability)
j 1

n
 xij = 1, for all j (activity requirement)
i 1

and xij = 0 or 1, for all i and j

where cij represents the cost of assignment of resource i to activity j.
This mathematical model of assignment problem is a particular case of the transportation problem for
two reasons: (i) the cost matrix is a square matrix, and (ii) the optimal solution table (matrix) for the problem
would have only one assignment in a given row or a column.
312 Operations Research: Theory and Applications

Remark If a constant is added to or subtracted from every element of any row or column of the given
cost matrix, then it would not change optimal assignments and value of objective function (See Appendix
10.A for proof.)

10.3 SOLUTION METHODS OF ASSIGNMENT PROBLEM

An assignment problem can be solved by any of the following methods:
 Enumeration method  Simplex method
 Transportation method  Hungarian method
1. Enumeration Method In this method, a list of all possible assignments among the given resources
(men, machines, etc.) and activities (jobs, sales areas, etc.) is prepared. Then an assignment that involves
the minimum cost (or maximum profit), time or distance is selected. If two or more assignmentshave the
same minimum cost (or maximum profit), time or distance, the problem has multiple optimal solutions.
In general, if an assignment problem involves n workers/jobs, then in total there are n! possible
assignments. For example, for an n = 5 workers/jobs problem, we have to evaluate a total of 5! or 120
assignments. However, when n is large, the method is unsuitable for manual calculations. Hence, this
method is suitable only when the value of n is small.
2. Simplex Method Since each assignment problem can be formulated as a 0 or 1 integer linear
programming problem, such a problem can also be solved by the simplex method. The general
mathematical model of the assignment problem involves n × n decision variables and n + n or 2n
equalities. Thus, for any assignment problem that involves 5 workers/jobs, there will be 25 decision
variables and 10 equalities. Solving such an assignment problem manually is difficult.
3. Transportation Method Since an assignment problem is a special case of the transportation problem,it
can also be solved by using MODI method. However, every basic feasible solution of a general
assignment problem that has a square matrix of order n should have m + n – 1 = n + n – l = 2n – 1
Hungarian method assignments. But due to the special structure of this problem, none of the solutions can have more than
to solve an n assignments. Consequently, solutions will be degenerate. To remove degeneracy, (n – 1) number of
assignment problem dummy allocations (deltas or epsilons) will be required in order to proceed with MODI method. Thus, the
is more efficient problem of degeneracy at each solution makes the procedure computationally inefficient for solving an
than using simplex
method. assignment problem.
4. Hungarian Method The Hungarian method (developed by Hungarian mathematician D. Konig) is an
efficient method of finding the optimal solution of an assignment problem without making a direct
comparison of every solution. The method works on the principle of reducing the given cost matrix
to a matrix of opportunity costs. Opportunity costs show the relative penalties associated with assigning
a resource to an activity. Hungarian method reduces the cost matrix to the extent of havingat least one
zero in each row and column so as to make optimal assignments.

10.3.1 Hungarian Method for Solving Assignment Problem

The Hungarian method (minimization case) can be summarized in the following steps:

Step 1: Develop the cost matrix from the given problem If the number of rows are not equalto
the number of columns, then add required number of dummy rows or columns. The cost element in dummy
rows/columns are always zero.

Step 2: Find the opportunity cost matrix

(a) Identify the smallest element in each row of cost matrix and then subtract it from each element of that
row, and
(b) In the reduced matrix obtained from 2(a), identify the smallest element in each column and then subtract
it from each element of that column. Each row and column now have at least one zero element.

Step 3: Make assignments in the opportunity cost matrix The procedure of making assignmentsis
as follows:
 Assignment Problem 313

(a) First round for making assignments

 Identify rows successively from top to bottom until a row with exactly one zero element is found.
Make an assignment to this single zero by making a square () around it. Then cross off (×) all
other zeros in the corresponding column.
 Identify columns successively from left to right hand with exactly one zero element that has not
been assigned. Make assignment to this single zero by making a square () around it and then cross
off (×) all other zero elements in the corresponding row.
(b) Second round for making assignments
 If a row and/or column has two or more unmarked zeros and one cannot be chosen by inspection,
then choose zero element arbitrarily for assignment.
 Repeat steps (a) and (b) successively until one of the following situations arise.

Step 4: Optimality criterion

(a) If all zero elements in the cost matrix are either marked with square () or are crossed off (×) and there is
exactly one assignment in each row and column, then it is an optimal solution. The total cost associated
with this solution is obtained by adding the original cost elements in the occupied cells.
(b) If a zero element in a row or column was chosen arbitrarily for assignment in Step 4(a), there exists
an alternative optimal solution.
(c) If there is no assignment in a row (or column), then this implies that the total number of assignments are
less than the number of rows/columns in the square matrix. In such a situation proceed to Step 5.

Fig. 10.1
Flow Chart of
Hungarian
Method
314 Operations Research: Theory and Applications

Step 5: Revise the opportunity cost matrix Draw a set of horizontal and vertical lines to cover
all the zeros in the revised cost matrix obtained from Step 3, by using the following procedure:
(a) For each row in which no assignment was made, mark a tick ()
(b) Examine the marked rows. If any zero element is present in these rows, mark a tick () to the respective
columns containing zeros.
(c) Examine marked columns. If any assigned zero element is present in these columns, tick () the
respective rows containing assigned zeros.
(d) Repeat this process until no more rows or columns can be marked.
(e) Draw a straight line through each marked column and each unmarked row.
If the number of lines drawn (or total assignments) is equal to the number of rows (or columns), the
current solution is the optimal solution, otherwise go to Step 6.
Zeros in the
assignment table
indicate opportunity
Step 6: Develop the new revised opportunity cost matrix
cost and show the
(a) Among the elements in the matrix not covered by any line, choose the smallest element. Call this
penalty of not
making the least value k.
cost (or best) (b) Subtract k from every element in the matrix that is not covered by a line.
assignment. (c) Add k to every element in the matrix covered by the two lines, i.e. intersection of two lines.
(d) Elements in the matrix covered by one line remain unchanged.

Step 7: Repeat steps Repeat Steps 3 to 6 until an optimal solution is obtained.

The flow chart of steps in the Hungarian method for solving an assignment problem is shown in
Fig. 10.1.
Example 10.1 A computer centre has three expert programmers. The centre wants three application
programmes to be developed. The head of the computer centre, after carefully studying the programmes
to be developed, estimates the computer time in minutes required by the experts for the application
programmes as follows:
Programmers
A B C
1 120 100 80
Programmes 2 80 90 110
3 110 140 120
Assign the programmers to the programmes in such a way that the total computer time is minimum.
Solution: Steps 1 and 2 The minimum time element in rows 1, 2 and 3 is 80, 80 and 110, respectively.
Subtract these elements from all elements in their respective row. The reduced time matrix is shown in Table
10.2(a).
(a) A B C (b) A B C
1 40 20 0 1 40 10 0
2 0 10 30 2 0 0 30
Table 10.2 3 0 30 10 3 0 20 10

In reduced Table 10.2(a) the minimum time element in columns A, B and C is 0, 10 and 0, respectively.
Subtract these elements from all elements in their respective column in order to get the reduced time matrix.
This is shown in Table 10.2(b).
Step 3 (a) Examine all the rows starting from the first, one-by-one, until a row containing single zero element
is found. In Table 10.2(b) rows 1 and 3 have only one zero in the cells (1, C) and (3, A), respectively.Make an
assignment in these cells and cross off all zero elements in the assigned column as shown in Table 10.3(a).

(a) A B C (b) A B C
1 40 20 0 1 40 10 0

2 0 0 30 2 0 0 30
Table 10.3 3 0 30 10 3 0 20 10
 Assignment Problem 315

(b) Now examine each column starting from A in Table 10.3(a). There is one zero in column B in the cell
(2, B). Make an assignment in this cell as shown in Table 10.3(b).
(c) Since the number of assignments (= 3) equals the number of rows (= 3), the optimal solution is
obtained.
The pattern of assignments among programmers and programmes with their respective time (in minutes)
is given below:

Programmer Programme Time (in minutes)

1 C 80
2 B 90
3 A 110
Total 280

Example 10.2 A department of a company has five employees with five jobs to be performed. The time (in
hours) that each man takes to perform each job is given in the effectiveness matrix.

Employees
I II III IV V

A 10 5 13 15 16
B 3 9 18 13 6
Jobs C 10 7 2 2 2
D 7 11 9 7 12
E 7 9 10 4 12

How should the jobs be allocated, one per employee, so as to minimize the total man-hours?
Solution Applying Step 2 of Hungarian algorithm, the reduced opportunity time matrix is shown in Table
10.4(a).

(a) I II III IV V (b) I II III IV V

A 5 0 8 10 11 A 5 0 8 10 11
B 0 6 15 10 3 B 0 6 15 10 3
C 8 5 0 0 0 C 8 5 0 0 0
D 0 4 2 0 5 D 0 4 2 0 5
E 3 5 6 0 8 E 3 5 6 0 8 Table 10.4

Steps 3 and 4: (a) Examine all the rows starting from A, one-by-one, until a row containing only single zero
element is found. Rows A, B and E have only one zero element in the cells (A, II ), (B, I ) and (E, IV ). Make an
assignment in these cells, and cross off all zeros in the assigned columns as shown in Table 10.4(b).
(b) Now examine each column starting from column I. There is one zero in column III, cell (C, III ).
Assignment is made in this cell. Thus cell (C, V ) is crossed off. All zeros in the table are now either
assigned or crossed off as shown in Table 10.4(b). The solution is not optimal because only four
assignments are made.

(a) I II III IV V (b) I II III IV V

A 5 0 8 10 11 A 7 0 8 12 11
B 0 6 15 10 3  B 0 4 13 10 1
C 8 5 0 0 0 C 10 5 0 2 0
D 0 4 2 0 5  D 0 2 0 0 3
E 3 5 6 0 8  E 3 3 4 0 6
Table 10.5
 
316 Operations Research: Theory and Applications

Step 5: Cover the zeros with minimum number of lines (= 4) as explained below:
(a) Mark () row D where there is no assignment.
(b) Mark () columns I and IV since row D has zero element in these columns.
(c) Mark () rows B and E since columns I and IV have an assignment in rows B and E, respectively.
(d) Since no other rows or columns can be marked, draw straight lines through the unmarked rows A and
C and the marked columns I and IV, as shown in Table 10.5(a).

Step 6: Develop the revised matrix by selecting the smallest element among all uncovered elements by the
lines in Table 10.5(a); viz., 2. Subtract k = 2 from uncovered elements including itself and add it to elements
5, 10, 8 and 0 in cells (A, I), (A, IV), (C, I) and (C, IV), respectively, which lie at the intersectionof two
lines. The revised matrix, so obtained is shown in Table 10.5(b).
Step 7: Repeat Steps 3 to 6 to find a new solution. The new assignments are shown in Table 10.6.

I II III IV V

A 7 0 8 12 11
B 0 4 13 10 1
C 10 5 0 2 0
D 0 2 0 0 3

Table 10.6 E 3 3 4 0 6

Since the number of assignments (= 5) equals the number of rows (or columns), the solution is optimal.
The pattern of assignments among jobs and employees with their respective time (in hours) is given below:

Job Employee Time (in hours)

A II 5
B I 3
C V 2
D III 9
E IV 4
Total 23

Example 10.3 A solicitors’ firm employs typists on hourly piece-rate basis for their daily work. There are
five typists and their charges and speed are different. According to an earlier understanding only onejob
was given to one typist and the typist was paid for a full hour, even if he worked for a fraction of an hour.
Find the least cost allocation for the following data:

Typist Rate per hour No. of Pages Job No. of Pages

(Rs) Typed/Hour
A 5 12 P 199
B 6 14 Q 175
C 3 8 R 145
D 4 10 S 298
E 4 11 T 178

[Delhi Univ., MBA, 2000, 2002]

Solution Develop a cost matrix based on the data of the problem as shown in Table 10.7(a), where elements
represent the cost to be incurred due to assignment of jobs to various typists on a one-to-one basis.

(a) P Q R S T (b) P Q R S T
A 85 75 65 125 75 A 2 2 0 4 0
B 90 78 66 132 78 B 6 4 0 10 2
C 75 66 57 114 69 C 0 1 0 1 2
D 80 72 60 120 72 D 2 4 0 4 2
Table 10.7
E 76 64 56 112 68 E 2 0 0 0 2
 Assignment Problem 317

Applying Step 2 of Hungarian method, the reduced opportunity cost matrix is shown in Table 10.7(b).
To make assignments in Table 10.7(b), examine all the rows starting from A until a row containing only
one zero element is found. Rows B and D have only one zero element in the cells (B, R) and (D, R),
respectively. First make assignment in cells (B, R), and cross off all zeros in the assigned columns as shown
in Table 10.7(b).
Now examine each column starting from column P. There is one zero in columns P, Q, S and T in the
cells (C, P), (E, Q ), (E, S ) and (A, T ). Assignment is made in these cells. All zeros in the matrix are either
assigned or crossed off, as shown in Table 10.8(a).

(a) P Q R S T (b) P Q R S T

A 2 2 0 4 0 A 2 2 2 4 0
B 6 4 0 10 2  B 4 2 0 8 0
C 0 1 0 1 2 C 0 1 2 1 2
D 2 4 0 4 2  D 0 2 0 2 0
E 2 0 0 0 2 E 2 0 2 0 2
Table 10.8


The solution shown in Table 10.8(a) is not optimal since only four assignments are made. Thus, in order
to get the next best solution, apply following steps.
(a) Mark () row D since it has no assignment.
(b) Mark () column R since row D has zero in this column.
(c) Mark () row B since column R has an assignment in row B.
(d) Since no other rows or columns can be marked, therefore draw straight lines through the unmarked
rows A, C and E and marked column R as shown in Table 10.8(a).
Develop the revised cost matrix by selecting the smallest element among all uncovered elements by
the lines (i.e., k = 2) in Table 10.8(a). Subtract this element (k = 2) from all uncovered elements including
itself and add it to elements in the cells (A, R), (C, R) and (E, R), respectively which lie at the intersection
of two lines. Another revised cost matrix so obtained is shown in Table 10.8(b).
Again repeat the procedure to find a new solution. The new assignments are shown in Table 10.9(a).

(a) P Q R S T (b) P Q R S T

A 2 2 2 4 0  A 2 1 2 3 0
B 4 2 0 8 0  B 4 1 0 7 0
C 0 1 2 1 2  C 0 0 2 0 2
D 0 2 0 2 0  D 0 1 0 1 0
E 2 0 2 0 2 E 3 0 3 0 3
Table 10.9
  

The solution shown in Table 10.9(a) is also not optimal since only four assignments are made. Thus, to
get the next best solution, follow Steps 6(a) to (d) of the algorithm to draw a minimum number of horizontal
and vertical lines to cover all zero elements in Table 10.9(a). The new opportunity cost matrix obtained from
Table 10.9(a) by subtracting the smallest element (= l) among all uncovered elements, including itself, by the
lines and adding it to elements at the intersection of two lines is shown in Table 10.9(b).
The new solution obtained by repeating the procedure as explained earlier is shown in Table 10.9(b).
Since both columns Q and S have two zeros, the arbitrary selection of a cell in any of these columns will give
us an alternative solution with same total cost of assignment.
The pattern of assignments among typists and jobs, along with cost is as follows:
318 Operations Research: Theory and Applications

Typist Job Cost (in Rs)

A T 75
B R 66
C S 114
D P 80
E Q 64
Total 399

CONCEPTUAL QUESTIONS A
1. What is an assignment problem? Give two applications. 6. State and discuss the methods of solving an assignment problem.
2. Give the mathematical formulation of an assignment problem. How is the Hungarian method better than other methodsfor solving
How does it differ from a transportation problem? an assignment problem?
3. Explain the conceptual justification that an assignment problem 7. (a) Give an algorithm to solve an assignment problem.
can be viewed as a linear programming problem. (b) Show that an assignment problem is a special case of a
4. Explain the difference between a transportation problem and an transportation problem.
assignment problem. 8. Explain how an assignment problem can be solved by using the
5. Specify the dual of an assignment problem. What are the transportation approach?
techniques used for solving an assignment problem?

SELF PRACTICE PROBLEMS A

1. Five men are available to do five different jobs. From past records, To
the time (in hours) that each man takes to do each job is known
and is given in the following table: 7 8 9 10 11 12

Jobs 1 31 62 29 42 15 41
I II III IV V 2 12 19 39 55 71 40
3 17 29 50 41 22 22
A 2 9 2 7 1 From 4 35 40 38 42 27 33
B 6 8 7 6 1 5 19 30 29 16 20 23
Men C 4 6 5 3 1
6 72 30 30 50 41 20
D 4 2 7 3 1
E 5 3 9 5 1 How should the trucks be displayed so as to minimize the total
distance travelled?
Find out how men should be assigned the jobs in way that will
minimize the total time taken. 4. An air freight company picks up and delivers freight where
customers want. The company has two types of aircraft, X and
2. A pharmaceutical company producing a single product sold it Y, with equal loading capacities but different operations costs.
through five agencies situated in different cities. All of a sudden, These are shown in the following Table.
there rouse a demand for the product in another five cities that
didn’t any agency of the company. The company is now facing
Type of Aircraft Operating Costs (Rs)
the problem of deciding on how to assign the existing agencies
in order to despatch the product to needy cities in such a way that Empty Loaded
the travelling distance is minimized. The distance between the
X 1.00 2.00
surplus and deficit cities (in km) is given in the following table.
Y 1.50 3.00
Deficit cities
a b c d e The present four locations of the aircraft that the company has
are; J – X; K – Y; L – Y, and M – X. Four customers of the company
A 160 130 115 190 200 located at A, B, C and D want to transport nearly the same size of
B 135 120 130 160 175 load to their final destinations. The final destinations are 600, 300,
Surplus Cities C 140 110 125 170 185 1,000 and 500 km from the loading points A, B, C and D,
D 50 50 80 80 110 respectively.
E 55 35 80 80 105 The distances (in km) between the aircraft and the loading
points are as follows:
Determine the optimum assignment schedule. Loading Point
3. A national truck rental service has a surplus of one truck in each
of the cities, 1, 2, 3, 4, 5 and 6; and a deficit of one truck in A B C D
each of the cities 7, 8, 9, 10, 11 and 12. The distances (in km) J 200 200 400 100
between the cities with a surplus and cities with deficit are Aircraft K 300 100 300 300
displayed in the table below:
Location L 400 100 100 500
M 200 200 400 200
 Assignment Problem 319

Determine the allocations which minimize the total cost of 9. A shipbuilding company has been awarded a big contract for the
transportation. construction of five cargo vessels. The contract stipulates that
5. A departmental head has four subordinates and four tasks to be the company must subcontract a portion of the total work to at least
performed. The subordinates differ in efficiency and the tasks differ five small ancillary companies. The company has invited bids
in their intrinsic difficulty. His estimates of the times that each from the small ancillary companies (A1, A2, A3, A4, and A5) to take
man would take to perform each task is given in the matrix below: care of the subcontract work in five fields – materials testing,
Tasks fabrication, assembly, scrap removal and painting. The bids
I II III IV received from the ancillary companies are given in the table.
Subcontract Bids (Rs)
A 8 26 17 11
Ancillary Materials Fabrication Assembly Scrap Painting
B 13 28 4 26 Companies Testing Removal
Subordinates C 38 19 18 15
A1 2,50,000 3,00,000 3,80,000 5,00,000 1,50,000
D 19 26 24 10
A2 2,80,000 2,60,000 3,50,000 5,00,000 2,00,000
A3 3,00,000 3,50,000 4,00,000 5,50,000 1,80,000
How should the tasks be allocated to subordinates so as to
A4 1,50,000 2,50,000 3,00,000 4,80,000 1,20,000
minimize the total man-hours? A5 3,00,000 2,70,000 3,20,000 4,80,000 1,60,000
6. An automobile dealer wishes to put four repairmen to four different
jobs. The repairmen have somewhat different kinds of skills and Which bids should the company accept in order to complete the
they exhibit different levels of efficiency from one job to another. contract at minimum cost? What is the total cost of the
The dealer has estimated the number of man-hours that would be subcontracts?
required for each job-man combination. This is given in matrix 10. In a textile sales emporium, four salesmen A, B, C and D are
form in the following table: available to four counters W, X, Y and Z. Each salesman can

Jobs handle any counter. The service (in hours) of each counter
when manned by each salesman is given below:
A B C D
Salesmen
1 5 3 2 8
A B C D
2 7 9 2 6
Men W 41 72 39 52
3 6 4 5 7
Counters X 22 29 49 65
4 5 7 7 8
Y 27 39 60 51
Z 45 50 48 52
Find the optimal assignment that will result in minimum man- hours
needed.
How should the salesmen be allocated to appropriate counters
7. A lead draftsman has five drafting tasks to accomplish and five idle so that the service time is minimized? Each salesman must handle
draftsmen. Each draftsman is estimated to require the following only one counter.
number of hours for each task.
11. A hospital wants to purchase three different types of medical
Tasks equipments and five manufacturers have come forward to supply
A B C D E one or all the three machines. However, the hospital’s policy is not
to accept more than one machine from any one of the
1 60 50 100 85 95 manufacturers. The data relating to the price (in thousand of
2 65 45 100 75 90 rupees) quoted by the different manufacturers is given below:
Draftsmen 3 70 60 110 97 85
Machines
4 70 55 105 90 93
1 2 3
5 60 40 120 85 97
A 30 31 27
If each draftsman costs the company Rs 15.80 per hour, including B 28 29 26
overhead, find the assignment of draftsmen to tasks that will Manufacturers C 29 30 28
result in the minimum total cost. What would be the total cost? D 28 31 27
8. A construction company has requested bids for subcontracts on five E 31 29 26
different projects. Five companies have responded. Theirbids are
Determine how best the hospital can purchase the three
represented below.
machines. [Delhi Univ., MBA (HCA), 2008]

12. The secretary of a school is taking bids on the city’s four school
Bid Amounts (’000s Rs) bus routes. Four companies have made the bids (in Rs), asdetailed
I II III IV V in the following table:

1 41 72 39 52 25 Route 1 Route 2 Route 3 Route 4

2 22 29 49 65 81 1 4,000 5,000 – –
Bidders 3 27 39 60 51 40 Bus 2 – 4,000 – 4,000
4 45 50 48 52 37 3 3,000 – 2,000 –
5 29 40 45 26 30 4 – – 4,000 5,000

Determine the minimum cost assignment of subcontracts to Suppose each bidder can be assigned only one route. Use the
bidders, assuming that each bidder can receive only one contract. assignment model to minimize the school’s cost of running the
four bus routes.
320 Operations Research: Theory and Applications

13. A large oil company operating a number of drilling platforms in Based on the marks awarded, what role should each of the trainees
the North Sea is forming a high speed rescue unit in order to be given in the rescue unit?
cope with emergency situations that may occur. The rescue unit 14. The personnel manager of ABC Company wants to assign Mr X,
comprises 6 personnel who, for reasons of flexibility, undergo Mr Y and Mr Z to regional offices. But the firm also has an opening
the same comprehensive training programme. The six personnel in its Chennai office and would send one of the three to that branch
are assessed as to their suitability for various specialist tasks if it were more economical than a move to Delhi, Mumbai or
and the marks they received in the training programme are given Kolkata. It will cost Rs 2,000 to relocate Mr X to Chennai, Rs 1,600
in the following table: to reallocate Mr Y there, and Rs 3,000 to move Mr Z. What is the
optimal assignment of personnel to offices?
Trainee Number
Specialist Task I II III IV V VI Office
Delhi Mumbai Kolkata
Unit Leader 21 5 21 15 15 28
Helicopter Pilot 30 11 16 8 16 4 Mr X 1,600 2,200 2,400
First Aid 28 2 11 16 25 25 Personnel Mr Y 1,000 3,200 2,600
Drilling Technology 19 16 17 15 19 8 Mr Z 1,000 2,000 4,600
Firefighting 26 21 22 28 29 24
Communications 3 21 21 11 26 26

HINTS AND ANSWERS

1. A – III, B – V, C – I, D – IV, E – II; 6. 1 – B, 2 – C, 3 – D, 4 – A ; 1 – C, 2 – D, 3 – B, 4 – A;

Optimal value = 13 hours. Total man-hours = 17 hours.
2. A – e, B – c, C – b, D – a, E – d; 7. 1 – A, 2 – D, 3 – E, 4 – C, 5 – B;
Minimum distance = 570 km. Minimum cost = Rs 365 × 15.8
4. 1 – 11, 2 – 8, 3 – 7, 4 – 9, 5 – 10, 6 – 12; 8. 1 – V, 2 – II, 3 – I, 4 – III, 5 – IV;
Minimum distance = 125 km. Minimum cost = Rs 155.
5. A – I, B – III, C – II, D – IV; 9. A1 – scrap, A2 – Fabrication, A3 – Painting, A4 – Testing,
Total man-hours = 41 hours. A5 – Assembly; Minimum cost = Rs 14,10,000
10. W – C, X – B, Y – A, Z – D; Optimal value = 147 hours.

10.4 VARIATIONS OF THE ASSIGNMENT PROBLEM

10.4.1 Multiple Optimal Solutions
While making an assignment in the reduced assignment matrix, it is possible to have two or more ways
to strike off a certain number of zeros. Such a situation indicates that there are multiple optimal solutions
with the same optimal value of objective function.

10.4.2 Maximization Case in Assignment Problem

If instead of cost matrix, a profit (or revenue) matrix is given, then assignments are mode in such a way
that total profit is maximized. The profit maximization assignment problems are solved by converting them
into a cost minimization problem in either of the following two ways:
(i) Put a negative sign before each of the elements in the profit matrix in order to convert the profit
values into cost values.
(ii) Locate the largest element in the profit matrix and then subtract all the elements of the matrix from
the largest element including itself.
The transformed assignment problem can be solved by using usual Hungarian method.
Example 10.4 A company operates in four territories, and four salesmen available for an assignment. The
territories are not equally rich in their sales potential. It is estimated that a typical salesman operatingin
each territory would bring in the following annual sales:
Territory : I II III IV
Annual sales (Rs) : 1,26,000 1,05,000 84,000 63,000
 Assignment Problem 321

The four salesmen also differ in their ability. It is estimated that, working under the same conditions, their
yearly sales would be proportionately as follows:
Salesmen : A B C D
Proportion : 7 5 5 4
If the criterion is maximum expected total sales, the intuitive answer is to assign the best salesman to the
richest territory, the next best salesman to the second richest, and so on; verify this answer by theassignment
technique. [Delhi Univ., MBA, 2004]
Solution Construct the Effectiveness Matrix: To avoid the fractional values of annual sales of each salesman
in each territory, for convenience, consider their yearly sales as 21 (i.e. the sum of sales proportions), taking
Rs 1,000 as one unit. Now divide the individual sales in each territory by 21 in orderto obtain the required
annual sales by each salesman. The maximum sales matrix so obtained is given in Table 10.10.

Territory
I II III IV Sales Proportion
A 42 35 28 21 7
B 30 25 20 15 5 Table 10.10
Salesman
C 30 25 20 15 5 Effectiveness
D 24 20 16 12 4 Matrix

Sales (in ’000 Rs) 6 5 4 3

Converting Maximization Problem into Minimization Problem: The given maximization assignment problem
(Table 10.10) can be converted into a minimization assignment problem by subtracting from the highest element
(i.e. 42) all the elements of the given table. The new data so obtained is given in Table 10.11(a).

(a) I II III IV (b) I II III IV

A 0 7 14 21 A 0 3 6 9
B 12 17 22 27 B 0 1 2 3 Table 10.11
C 12 17 22 27 C 0 1 2 3 Equivalent Cost
D 18 22 26 30 D 0 0 0 0 Matrix

Apply Hungarian Method to get Optimal Solution: Apply Step 2 of the Hungarian method in order to
get the reduced matrix that has at least one zero, either in a row or column, as shown in Table 10.11(b).
The assignment is made in row A. All zeros in the assigned column I are crossed as shown in Table
10.12. Column II has only one zero, in cell (D, II). Assignment is made in this column, and other zeros are
crossed in row D. Now all zeros are either assigned or crossed off as shown in Table 10.12.

I II III IV

A 0 3 6 9 
B 0 1 2 3 
C 0 1 2 3 
D 0 0 0 0
 Table 10.12

The solution shown in Table 10.12 is not optimal since only two assignments are made. Draw lines
through marked columns and unmarked rows to cover zeros as shown in Table 10.12.
Develop the revised cost matrix by selecting the minimum element (= 1) among all uncovered elements
by the lines. Subtract 1 from each uncovered element, including itself, and add it to the elements at the
intersection of two lines. A revised cost table so obtained is shown in Table 10.13.
322 Operations Research: Theory and Applications

I II III IV
A 0 2 5 8 
B 0 0 1 2 
C 0 0 1 2 

D 1 0 0 0
Table 10.13
 

Repeat Steps 1 to 3 to mark the assignments in Table 10.13. Two alternative optimal assignments are
shown in Tables 10.14(a) and (b).

(a) I II III IV (b) I II III IV

A 0 2 4 7 A 0 2 4 7
B 0 0 0 1 B 0 0 0 1
Table 10.14
Alternative C 0 0 0 1 C 0 0 0 1
Optimal Solutions D 2 1 0 0 D 2 1 0 0

The pattern of two alternative optimal assignments among territories and salesmen with their respective
sales volume (in Rs 1,000) is given in the table.
Assignment Set I Assignment Set II

Salesman Territory Sales (Rs) Salesman Territory Sales (Rs)

A I 42 A I 42
B III 20 B II 25
C II 25 C III 20
D IV 12 D IV 12
Total 99 Total 99

Example 10.5 A marketing manager has five salesmen and five sales districts. Considering the capabilities
of the salesmen and the nature of districts, the marketing manager estimates that the sales per month (in
hundred rupees) for each salesman in each district would be as follows:
Districts
A B C D E
1 32 38 40 28 40
2 40 24 28 21 36
Salesmen 3 41 27 33 30 37
4 22 38 41 36 36
5 29 33 40 35 39
Find the assignment of salesmen to districts that will result in maximum sales.
Solution The given maximization problem can be converted into a minimization problem by subtracting
from the largest element (i.e. 41) all the elements of the given table. The new cost data so obtained is given
in Table 10.15.
(a) A B C D E (b) A B C D E
1 9 3 1 13 1 1 8 0 0 7 0
2 1 17 13 20 5 2 0 14 12 14 4
Table 10.15 3 0 14 8 11 4 3 0 12 8 6 4
Equivalent Cost 4 19 3 0 5 5 4 19 1 0 0 5
Data 5 12 8 1 6 2 5 11 5 0 0 1

Apply Step 2 of the Hungarian method to get the opportunity cost table as shown in Table 10.15(b).
 Assignment Problem 323

Make assignments in Table 10.15(b) by applying Hungarian method as shown in Table 10.16(a).

(a) A B C D E (b) A B C D E

1 8 0 0 7 0 1 12 0 0 7 0
2 0 14 12 14 4  2 0 10 8 10 0
3 0 12 8 6 4  3 0 8 4 2 0
4 19 1 0 0 5 4 23 1 0 0 5
5 11 5 0 0 1 5 15 5 0 0 1
Table 10.16

The solution shown in Table 10.16(a) is not optimal since only four assignments are made. Cover the zeros
with the minimum number of lines (= 4) as shown in Table 10.16(a).
Develop the revised cost matrix by selecting the minimum element (= 4) among all uncovered elements
by the lines. Subtract 4 from all uncovered elements, including itself, and add it to the element at the
intersection of the lines. A revised cost table, so obtained, is shown in Table 10.16(b).
Repeat the above procedure again to make the assignments in the reduced Table 10.16(b). The two
alternative assignments are shown in Tables 10.17(a) and (b). Two more alternative solutions exist due to
presence of zero element in cells (4, C ), (4, D) and cells (5, C ), (5, D).

(a) A B C D E (b) A B C D E

1 12 0 0 7 0 1 12 0 0 7 0
2 0 12 10 12 0 2 0 12 10 12 0
3 0 10 4 2 0 3 0 10 4 2 0
4 23 1 0 0 5 4 23 1 0 0 5
Table 10.17
5 15 5 0 0 1 5 15 5 0 0 1 Alternative
Optimal Solutions
Two alternative optimal assignments are as follows:
Assignment Set I Assignment Set II
Salesman District Sales Salesman District Sales
(in ’000 Rs) (in ’000 Rs)

1 B 38 1 B 38
2 A 40 2 E 36
3 E 37 3 A 41
4 C 41 4 C 41
5 D 35 5 D 35
Total 191 Total 191

10.4.3 Unbalanced Assignment Problem

The Hungarian method for solving an assignment problem requires that the number of columns and rows
in the assignment matrix should be equal. However, when the given cost matrix is not a square matrix, the
assignment problem is called an unbalanced problem. In such cases before applying Hungarian method,
dummy row(s) or column(s) are added in the matrix (with zeros as the cost elements) in order to make it
a square matrix.

10.4.4 Restrictions on Assignments

Sometimes it may so happen that a particular resource (say a man or machine) cannot be assigned to a
particular activity (say territory or job). In such cases, the cost of performing that particular activity by a
324 Operations Research: Theory and Applications

particular resource is considered to be very large (written as M or ) so as to prohibit the entry of this
pair of resource-activity into the final solution.
Example 10.6 In the modification of a plant layout of a factory four new machines M1, M2, M3, and
M4 are to be installed in a machine shop. There are five vacant places A, B, C, D and E available. Because of
limited space, machine M2 cannot be placed at C and M3 cannot be placed at A. The cost of locatinga
machine at a place (in hundred rupees) is as follows.
Location
A B C D E
M1 9 11 15 10 11
M2 12 9 – 10 9
Machine
M3 – 11 14 11 7
M4 14 8 12 7 8

Find the optimal assignment schedule. [Delhi Univ., MBA, 2004, 2006]
Solution Since cost matrix is not balanced, add one dummy row (machine) with a zero cost elements in that
row. Also assign a high cost, denoted by M, to the pair (M2, C) and (M3, A). The cost matrix so obtainedis
given in Table 10.18(a).
Apply Hungarian method for solving this problem. An optimal assignment is shown in Table 10.18(b).

(a) (b) A B C D E A B C D E
M1 9 11 15 10 11 M1 0 2 6 1 2
M2 12 9 M 10 9 M2 3 0 M 1 0
M3 M 11 14 11 7 M3 M 4 7 4 0
M4 14 8 12 7 8 M4 7 1 5 0 1
Table 10.18 M5 0 0 0 0 0 M5 0 0 0 0 0

The total minimum cost (Rs) and optimal assignments made are as follows:

Machine Location Cost (in ’000 Rs)

M1 A 9
M2 B 9
M3 E 7
M4 D 7
M5 (dummy) C 0
Total 32

Example 10.7 An airline company has drawn up a new flight schedule that involves five flights. To assist
in allocating five pilots to the flights, it has asked them to state their preference scores by givingeach
flight a number out of 10. The higher the number, the greater is the preference. A few of these flights are
unsuitable to some pilots, owing to domestic reasons. These have been marked with ‘×’.

Flight Number
1 2 3 4 5
A 8 2 × 5 4
B 10 9 2 8 4
Pilot C 5 4 9 6 ×
D 3 6 2 8 7
E 5 6 10 4 3

What should be the allocation of the pilots to flights in order to meet as many preferences as possible?
[AIMA (Dip. in Mgt.), 2005]
 Assignment Problem 325

Solution Since the problem is to maximize the total preference score, in order to apply the Hungarian
method to solve this assignment problem, the equivalent cost matrix is required. This is obtained by
subtracting all the elements of the given matrix from the largest element (= 10) including itself as shown
in Table 10.19.
1 2 3 4 5
A 2 8 M 5 6
B 0 1 8 2 6
C 5 6 1 4 M Table 10.19
D 7 4 8 2 3 Equivalent Cost
E 5 4 0 6 7 Matrix

Perform the Hungarian method on Table 10.19 to make assignments as shown in Table 10.20.

1 2 3 4 5

A 0 5 M 3 3

B 0 0 8 2 5

C 4 4 0 3 M 

D 5 1 6 0 0
E 5 3 0 6 6 
Table 10.20
 Opportunity Cost
Table

The solution shown in Table 10.20 is not the optimal solution because there is no assignment in row
E. Draw minimum number of lines to cover all zeros in the table and then subtract the smallest element
(= 3) from all uncovered elements including itself and add it to the element at the intersection of two lines.
The new table so obtained is shown in Table 10.21.

1 2 3 4 5

A 0 5 M 3 3

B 0 0 11 2 5

C 1 1 0 0 M

D 5 1 9 0 0

E 2 0 0 3 3 Table 10.21

Repeat the Hungarian method to make assignments in Table 10.21. Since the number of assignments
in Table 10.21 is equal to the number of rows or columns, this solution is the optimal solution. The optimal
assignment is as follows:
Pilot Flight Number Preference Score

A 1 8
B 2 9
C 4 6
D 5 7
E 3 10
Total 40

Example 10.8 A city corporation has decided to carry out road repairs on four main arteries of the city. The
government has agreed to make a special grant of Rs 50 lakh towards the cost with a condition thatthe
repairs be done at the lowest cost and quickest time. If the conditions warrant, a supplementary token grant
will also be considered favourably. The corporation has floated tenders and five contractors have sent in their
bids. In order to expedite work, one road will be awarded to only one contractor.
326 Operations Research: Theory and Applications

Cost of Repairs (Rs in lakh)

R1 R2 R3 R4
C1 9 14 19 15
C2 7 17 20 19
Contractors/Road C3 9 18 21 18
C4 10 12 18 19
C5 10 15 21 16

(a) Find the best way of assigning the repair work to the contractors and the costs.
(b) If it is necessary to seek supplementary grants, what should be the amount sought?
(c) Which of the five contractors will be unsuccessful in his bid? [AMIE 2005]
Solution (a) Since cost matrix is not balanced, therefore add one dummy column (road, R5) with a zero
cost elements. The revised cost matrix is given in Table 10.22.

R1 R2 R3 R4 R5

C1 9 14 19 15 0
C2 7 17 20 19 0
C3 9 18 21 18 0

Table 10.22 C4 10 12 18 19 0
Cost Matrix C5 10 15 21 16 0

Apply the Hungarian method to solve this problem. This is left as an exercise for the reader. An optimal
assignment is shown in Table 10.23.

R1 R2 R3 R4 R5

C1 1 1 0 0 1

C2 0 5 2 5 2

C3 0 4 1 2 0
C4 3 0 0 5 2
Table 10.23
Optimal Solution C5 1 1 1 0 0

The total minimum cost (in rupees) and optimal assignment made are as follows:

Road Contractor Cost (Rs in lakh)

R1 C2 7
R2 C4 12
R3 C1 19
R4 C5 16
R5 C3 0
Total 54

(b) Since the total cost exceeds 50 lakh, the excess amount of Rs 4 lakh (= 54 – 50) is to be sought
as supplementary grant.
(c) Contractor C3 who has been assigned to dummy row, R5 (roads) loses out in the bid.
 Assignment Problem 327

CONCEPTUAL QUESTIONS B
1. Can there be multiple optimal solutions to an assignment problem? 3. Explain how can one modify an effectiveness matrix in an
How would you identify the existence of multiple solutions, if any? assignment problem, if a particular assignment is prohibited.
2. How would you deal with the assignment problems, where (a) 4. What is an unbalanced assignment problem? How is the Hungarian
the objective function is to be maximized? (b) some assignments method applied for obtaining a solution if the matrix is
are prohibited? rectangular?

SELF PRACTICE PROBLEMS B

1. A project work consists of four major jobs for which an equal available to replace one of the existing ones and the associated
number of contractors have submitted tenders. The tender amount of that machine costs are also given below.
quoted (in lakh of rupees) is given in the matrix.
Machines
Job M1 M2 M3 M4 M5 M6
a b c d
W1 12 3 6 – 5 9
1 10 24 30 15
W2 4 11 – 5 – 8
Contractor 2 16 22 28 12
3 12 20 32 10 Workers W3 8 2 10 9 7 5
4 9 26 34 16 W4 – 7 8 6 12 10
W5 5 8 9 4 6 1
Find the assignment which minimizes the total cost of the project
when each contractor has to be assigned at least one job. (a) Determine whether the new machine can be accepted.
2. Alpha Corporation has four plants, each of which can manufacture (b) Also determine the optimal assignment and the associated
any one of four products A, B, C or D. Production costs differ from saving in cost.
one plant to another and so do the sales revenue. The revenue and 5. A fast-food chain wants to build four stores. In the past, thechain
the cost data are given below. Determine which product should has used six different construction companies, and having been
each plant produce in order to maximize profit. satisfied with each, has invited each to bid on each job. The
Sales Revenue (in ’000 Rs) final bids (in lakh of rupees) are shown in the following table:
Plant
Construction Companies
1 2 3 4
1 2 3 4 5 6
A 50 68 49 62
Product B 60 70 51 74 Store 1 85.3 88.0 87.5 82.4 89.1 86.7
C 52 62 49 68 Store 2 78.9 77.4 77.4 76.5 79.3 78.3
D 55 64 48 66 Store 3 82.0 81.3 82.4 80.6 83.5 81.7
Store 4 84.3 84.6 86.2 83.3 84.4 85.5
Production Cost (in ’000 Rs)
Plant
Since the fast-food chain wants to have each of the new stores
1 2 3 4
ready as quickly as possible, it will award at the most one job to
A 49 60 45 61 a construction company. What assignment would result in
Product B 55 63 45 49 minimum total cost to the fast-food chain?
C 55 67 53 70 [Delhi Univ., MBA, 2001, 2003]
D 58 65 54 68 6. A methods engineer wants to assign four new methods to three
work centres. The assignment of the new methods will increase
3. A company has four machines that are to be used for three jobs. production. The methods are given below.
Each job can be assigned to one and only one machine. The
cost of each job on each machine is given in the following table. Increase in Production (unit)
Machines Work Centres
W X Y Z
A B C
A 18 24 28 32
Jobs B 8 13 17 18 1 10 7 8
C 10 15 19 22 2 8 9 7
Method 3 7 12 6
What are the job-assignment pairs that shall minimize the cost? 4 10 10 8
[Gauhati, MCA, 2001]
If only one method can be assigned to a work centre, determine the
4. Five workers are available to work with the machines and the optimum assignment.
respective costs (in rupees) associated with each worker- machine 7. Consider a problem of assigning four clerks to four tasks. The
assignment are given below. A sixth machine is time (hours) required to complete the task is given below:
328 Operations Research: Theory and Applications

Tasks Territory
A B C D 1 2 3 4 5
1 4 7 5 6 A 75 80 85 70 90
2 – 8 7 4 Salesmen B 91 71 82 75 85
Clerks C 78 90 85 80 80
3 3 – 5 3
4 6 6 4 2 D 65 75 88 85 90

Clerk 2 cannot be assigned task A and clerk 3 cannot be assigned Suggest optimal assignment of the salesmen. If for certain reasons,
task B. Find all the optimum assignment schedules. salesman D cannot be assigned to territory 3, will the optimal
assignment be different? If so, what would be the new assignment
8. The marketing director of a multi-unit company is faced with a schedule? [Delhi Univ., MCom, 2000]
problem of assigning 5 senior managers to six zones. From past 12. The personnel manager of a medium-sized company hasdecided
experience he knows that the efficiency percentage judged by to recruit two employees D and E in a particular section of the
sales, operating costs, etc., depends on the manager-zone organization. The section has five fairly defined tasks 1, 2, 3, 4
combination. The efficiency of different managers is given below: and 5; and three employees A, B and C are already employed in
Zones the section. Considering the specialized nature of task 3 and the
special qualifications of the recruit D for task 3, the manager has
I II III IV V VI decided to assign task 3 to employee D and then assign the
A 73 91 87 82 78 80 remaining tasks to remaining employees so as to maximize the
B 81 85 69 76 74 85 total effectiveness. The index of effectiveness of each employee
Manager C 75 72 83 84 78 91 of different tasks is as under.
D 93 96 86 91 83 82
Tasks
E 90 91 79 89 69 76
1 2 3 4 5
Find out which zone should be managed by a junior manager A 25 55 60 45 30
due to the non-availability of a senior manager. B 45 65 55 35 40
9. A head of department in a college has the problem of assigning Employees C 10 35 45 55 65
courses to teachers with a view to maximize educational quality D 40 30 70 40 60
in his department. He has available to him one professor, two E 55 45 40 55 10
associate professors, and one teaching assistant (TA). Four
courses must be offered. After appropriate evaluation, hehas Assign the tasks for maximizing total effectiveness. Critically
arrived at the following relative ratings (100 = best rating) regarding examine whether the decision of the manager to assign task 3
the ability of each instructor to teach each of the four courses. to employee D was correct. [Delhi Univ., MBA, 2003]
Course 1 Course 2 Course 3 Course 4
13. The casualty medical officer of a hospital has received four
Prof. 1 60 40 60 70 requests for Ambulance van facility. Currently, six vans are
Prof. 2 20 60 50 70 available for assignment and their estimated response time (in
Prof. 3 20 30 40 60 minutes) are shown in the table below:
TA 30 10 20 40 Van
How should he assign his staff to the courses in order to realize his 1 2 3 4 5 6
objective? [Delhi Univ., MBA (HCA), 1999] 1 16 15 13 14 15 18
10. At the end of a cycle of schedules, a transport company has 2 18 16 12 13 17 16
a surplus of one truck in each of the cities 1, 2, 3, 4, 5 and a Incident 3 14 14 17 16 15 15
deficit of one truck in each of the cities A, B, C, D, E and F. 4 13 17 19 18 14 17
The distance (in kilometres) between the cities with a surplus,
and cities with a deficit, is given below: Determine which van should respond, and what will be the average
To City response time. [Delhi Univ., MBA (HCA), 1990, 96, 98]
14. To stimulate interest and provide an atmosphere for intellectual
A B C D E F
discussion a finance faculty in a management school decides to
1 80 140 80 100 56 98 hold special seminars on four contemporary topics: Leasing,
2 48 64 94 126 170 100 portfolio management, private mutual funds, swaps and options.
From City 3 56 80 120 100 70 64 Such seminars are to be held be held once par week in the
4 99 100 1,100 104 80 90 afternoons. However, scheduling these seminars (one for each
5 64 80 90 60 60 70 topic, and not more than one seminar per afternoon) has to be done
carefully so that the number of students unable to attend is kept
How should the trucks be despatched so as to minimize the total to a minimum. A careful study indicates that the number of
distance travelled? Which city will not receive a truck? students who cannot attend a particular seminar on a specific day
[Madras, MBA, 2000] is as follows:
11. A company is considering expanding into five new sales territories.
The company has recruited four new salesmen. Based on the
salesmen’s experience and personality traits, the sales manager
has assigned ratings to each of the salesmen for each of the
sales territories. The ratings are as follows:
 Assignment Problem 329

 Chandru swims all styles: back 1 : 10, butterfly 1 : 12, free

Leasing Portfolio Private Swaps style 1 : 05 and breast stroke 1 : 20.
Management Mutual and
 Dorai swims only the butterfly 1 : 11, while Easwar swims
Funds Options
the back stroke 1 : 20, breast stroke 1 : 16, free style 1
Monday 50 40 60 20 : 06 and the butterfly 1 : 10.
Tuesday 40 30 40 30 Which swimmer should be assigned which swimming style?
Wednesday 60 20 30 20 Who will not be in the relay?
Thursday 30 30 20 30 16. (a) At the end of a cycle of schedules, a trucking company has
Friday 10 20 10 30 a surplus of one vehicle in each of the cities
– 1, 2, 3, 4 and 5 and a deficit of one vehicle in each of the cities A,
Find an optimal schedule of the seminars. Also find out the total B, C, D, E and F. The cost (in rupees) of transportation and
number of students who will be missing at least one seminar. handling between the cities with a surplus and the cities with
15. Five swimmers are eligible to compete in a relay team that is deficits are shown in the following table:
to consist of four swimmers swimming four different swimming To City
styles. The styles are – back stroke, breast stroke, free style
A B C D E F
and butterfly. The time taken by the five swimmers – Anand,
Bhasker, Chandru, Dorai and Easwar, to cover a distance of 1 134 116 167 233 164 97
100 metres in various swimming styles is given below, in minutes, 2 114 195 260 166 178 130
seconds. From City 3 129 117 48 94 66 101
 Anand swims the back stroke in 1 : 09, the breast stroke 4 71 156 92 143 114 136
in 1 : 15 and has never competed in the free style or butterfly. 5 97 134 125 83 142 118
 Bhasker is a free style specialist averaging 1 : 01 for the Find the assignment of surplus vehicles to deficit cities that will
100 metres but can also swim the breast stroke in 1 : 16 result in a minimum total cost? Which city will not receive avehicle?
and butterfly in 1 : 20. [Delhi Univ., MBA, 2002]

HINTS AND ANSWERS

1. (i) 1 – b, 2 – c, 3 – d, 4 – a, 7. (i) 1 – B, 2 – D, 3 – A, 4 – C
(ii) 1 – c, 2 – b, 3 – d, 4 – a; Minimum cost = Rs 71,00,000. (ii) 1 – C, 2 – D, 3 – A, 4 – B. Minimum hours = 18.
(iii) 1 – c, 2 – d, 3 – b, 4 – a, 8. A – III, B – II, C – III, D – I, E – IV, dummy – V.
(iv) 1 – b, 2 – c, 3 – d, 4 – a 9. (i) Prof. 1 – Course 3, Prof. 2 – Course 2, Prof. 3 – Course 4,
2. Construct the profit matrix using the relationship: Profit = T.A. – Course 1
Revenue – Cost (ii) Prof. 1 – Course 1, Prof. 2 – Course 2, Prof. 3 – Course 4,
A – 2, B – 4, C – 1, D – 3 and Maximum profit = Rs 42,000 T.A. – Course 3. Maximum educational quality = 210.
3. (i) A – W, B – X, C – Y, 10. 1 – E, 2 – B, 3 – A, 4 – F, 5 – D; Cost = Rs 326.
(ii) A – W, B – Y, C – X. Minimum cost = Rs 50. 11. A – 5, B – 1, C – 2, D – 3; 359.
4. (a) W1 – M5, W2 – M6, W3 – M2, W4 – M4, W5 – M1 and W6 12. A – 4, B – 2, C – 5, D – 3, E – 1; 300.
(dummy) – M3; Minimum cost = Rs 21 15. The assignment matrix with time expressed in seconds and
(b) W1 – M5, W2 – M1, W3 – M2, W4 – M3 and W5 – M4, adding a dummy style to balance it is given by
Minimum cost = Rs 23.
Back Stroke Breast Stroke Free Style Butterfly
The sixth machine should be accepted because saving in
cost is Rs (23 – 21) = Rs 2 Anand 69 75 – –
5. S1 – 4, S2 – 3, S3 – 2, S4 – 1; Cost = Rs 325.4 Bhasker – 76 61 80
Chandru 70 80 65 72
6. (i) 1 – A, 2 – dummy, 3 – B, 4 – C
Dorai – – – 71
(ii) 1 – C, 2 – dummy, 3 – B, 4 – A. Total production = 30 Easwar 80 76 66 70
units
330 Operations Research: Theory and Applications

10.5 A TYPICAL ASSIGNMENT PROBLEM

Example 10.9 A small airline company that owns five planes operates on all the seven days of a week.
Flights between three cities A, B and C, according to the schedule, is given below. The layover cost per
stop is roughly proportional to the square of the layover time:
Flight No. From Departure Time To Arrival Time
(in hrs) (in hrs)

1 A 09.00 B 12.00
2 A 10.00 B 13.00
3 A 15.00 B 18.00
4 A 20.00 C Midnight
5 A 22.00 C 02.00
6 B 04.00 A 07.00
7 B 11.00 A 14.00
8 B 15.00 A 18.00
9 C 07.00 A 11.00
10 C 15.00 A 19.00

Find out how the planes should be assigned to the flights so as to minimize the total layover cost. If you have
made any particular assumptions, state them clearly.
Solution We have made the following assumptions:
(i) A plane cannot make more than two trips (to and fro).
(ii) A plane flying from a particular city must be back within 24 hours for the next scheduled trip from
the city.
From the given data, it is clear that there is no route between city B and C. Therefore, the present problem
can be divided into two subproblems: (i) routes between A and C, and (ii) routes betweenA and B.
Let us now construct the cost matrix for the routes connecting cities A and C. This is given in Table 10.24.

Flight Number 9 10

4 130 226
5 146 178
Table 10.24
The elements in the Table 10.24 are interpreted as follows: For route 4 – 9 a plane taking flight number 4
from A to C and number 9 from C to A, would have a layover time of 9 hours (11.00 to 20.00) at city A
and 7 hours (midnight – 7.00) at city C. Thus, the layover cost for the route 4 – 9 would be (9)2 + (7)2
= 130 units. Similarly, the other route costs elements are also obtained.
The optimal assignment can now be obtained by applying the Hungarian method of assignment. The
optimal solution is given in Table 10.25.

Flight Number 9 10

4 0 64
Table 10.25 5 0 0

Optimal assignment : 4 – 9, 5 – 10
Total cost : 308 units.
Similarly, cost matrix for the routes connecting cities A and B can be constructed as shown in Table 10.26.

Flight Number 6 7 8

1 260 M 234
2 234 M 260
Table 10.26 3 164 290 M
 Assignment Problem 331

The elements in Table 10.26 are interpreted as follows: For route 1–6, a plane taking flight number 1 from A
to B and number 6 from B to A would have a total layover time of 18 hours (16 + 2) with an associated
layover cost of (16)2 + (2)2 = 260 units. Since a plane taking flight number 1 cannot return for flight number
7 because of assumption (ii), the layover cost associated with this flight is considered very high, say M.
Similarly, the other route cost elements are also obtained.
The optimal assignment can now be obtained by applying the Hungarian method of assignment. The
optimal solution is given in Table 10.27.

Flight Number 6 7 8

1 26 M 0

2 0 M 26

3 0 0 M Table 10.27

Optimal assignment : 1 – 8, 2 – 6, 3 – 7
Total cost : 758 units.
Hence, from the two solutions obtained above, the complete flight schedule of planes is arrived at and is
given in Table 10.28.
Plane Flight Departure Arrival
Number Number City Time (hrs) City Time (hrs)
1 1 A 09.00 B 12.00
8 B 15.00 A 18.00
-- -----------------------------------------------------
----- - ----
2 A 10.00 B 13.00
- - - - - - - 2 - - - - - - - - - 6- - - - - - - - - -B- - - - - - - - -04-.0-0- - - - - - - - -A- - - - - - - - 0-7-.0-0- - -
----- - ----
3 A 15.00 B 18.00
----- 3
- - - - - - - - - - - 7- - - - - - - - - -B- - - - - - - - -11-.0-0- - - - - - - - -A- - - - - - - - 1- 4-.0-0- - -
----- - ----
4 4 A 20.00 C Midnight
----- 9 C 07.00 A 11.00
-- ----------------------------------------------------- Table 10.28
----- - ---- Optimal Flight
5 A 22.00 C 0.200
----- 5 Schedule
10 C 15.00 A 19.00
The total minimum layover cost is 308 + 758 = 1,066 units.

10.6 TRAVELLING SALESMAN PROBLEM

The travelling salesman problem may be solved as an assignment problem, with two additional conditions
on the choice of assignment. That is, how should a travelling salesman travel starting from his home city (the
city from where he started), visiting each city only once and returning to his home city, so that thetotal
distance (cost or time) covered is minimum. For example, given n cities and distances dij (cost cij or time tij)
from city i to city j, the salesman starts from city 1, then any permutation of 2, 3, . . ., n represents the number
of possible ways of his tour. Thus, there are (n – 1)! possible ways of his tour. Now the problemis to select
an optimal route that is able to achieve the objective of the salesman.
To formulate and solve this problem, let us define:
RS1, if salesman travels from city i to city j
xij 
T0, otherwise

A travelling
Since each city can be visited only once, we have salesman starts
n 1 from his home city

 xij  1, j  1, 2, . . ., n ; i  j and returns back

i 1
visiting each city, so

Again, since the salesman has to leave each city except city n, we have that the total
n distance (cost or
 xij  1, i  1, 2, .. ., n  1; i  j time) covered is the
j 1 minimum.
332 Operations Research: Theory and Applications

The objective function is then 

n 1 n
Minimize Z    dij xij

i 1 j 1

Since dji = dij is not required, therefore dij =  for i = j. However, all dijs must be non-negative, i.e.
dij  0 and dij + djk  djk for all i, j, k.
Example 10.10 A travelling salesman has to visit five cities. He wishes to start from a particular city, visit
each city once and then return to his starting point. The travelling cost (in ’000 Rs) of each city froma
particular city is given below:

To City
A B C D E
A  2 5 7 1
B 6  3 8 2
From City C 8 7  4 7
D 12 4 6  5
E 1 3 2 8 

What should be the sequence of visit of the salesman so that the cost is minimum? [Delhi Univ., MBA, 2004]
Solution Solving the given travelling salesman problem as an assignment problem, by Hungarian methodof
assignment, an optimal solution is shown in Table 10.29. However, this solution is not the solution tothe
travelling salesman problem as it gives the sequence A – E – A. This violates the condition that
salesmen can visit each city only once.
To City
A B C D E

A  1 3 6 0

B 4  0 6 0

From City C 4 3  0 3

D 8 0 1  l
Table 10.29
Optimal Solution E 0 2 0 7 

The ‘next best’ solution to the problem that also satisfies the extra condition of unbroken sequence
of visit to all cities, can be obtained by bringing the next (non-zero) minimum element, i.e. 1, into the solution.
In Table 10.29, the cost element 1 occurs at three different places. Therefore, consider all three different
cases one by one until the acceptable solution is reached.
Case 1: Make the unit assignment in the cell (A, B) instead of zero assignment in the cell (A, E) and
delete row A and column B so as to eliminate the possibility of any other assignment in row A and column
B. Now make the assignments in the usual manner. The resulting assignments are shown in Table 10.30.

To City
A B C D E

A  1 3 6 0
B 4  0 6 0
From City CD 4 3  0 3
E 8 0 1  1
Table 10.30 0 2 0 7 
 Assignment Problem 333

The solution given in Table 10.30 gives the sequence: AB, BC, CD, DE, EA. The cost
corresponding to this feasible solution is Rs 15,000.
Case 2: If we make the assignment in the cell (D, C) instead of (D, E), then no feasible solution is obtained
in terms of zeros or in terms of element 1 which may give cost less than Rs 15,000.
Hence, the best solution is: A – B – C – D – E – A, and the total cost associated with this solution is
Rs 15,000.
Example 10.12 ABC Ice Cream Company has a distribution depot in Greater Kailash Part I for distributing
ice-cream in South Delhi. There are four vendors located in different parts of South Delhi (call them A, B, C
and D) who have to be supplied ice-cream everyday. The following matrix displays the distances (in
kilometres) between the depot and the four vendors:
To
Depot Vendor A Vendor B Vendor C Vendor D
Depot – 3.5 3 4 2
Vendor A 3.5 – 4 2.5 3
From Vendor B 3 4 – 4.5 3.5
Vendor C 4 2.5 4.5 – 4
Vendor D 2 3 3.5 4 –

What route should the company van follow so that the total distance travelled is minimized?
[Delhi Univ., MBA, 2003, 2004, 2006]
Solution Solving the given travelling salesman problem as an assignment problem, by using the Hungarian
method of assignment, an optimal solution is shown in Table 10.31.

To
Depot Vendor A Vendor B Vendor C Vendor D

Depot – 1.5 0 2 0

Vendor A 1.5 – 0.5 0 0.5

From Vendor B 0 0.5 – 1 0

Vendor C 2 0 1 – 1.5
Table 10.31
Vendor D 0 0.5 0 1.5 – Optimal Solution

The solution shown in Table 10.31 is an optimal solution to an assignment problem, but it is not the optional
solution to the travelling salesman problem. This is because it gives the sequence: Depot  Vendor D 
Vendor B  Depot. This violates the condition of the travelling salesman problem.
The ‘next best’ solution to the problem that also satisfies travelling salesman condition of unbroken
sequence, can be obtained by bringing next (non-zero) minimum element, i.e. 0.5, into the solution. In Table
10.31, the element 0.5 occurs at four different places. Therefore, consider all four different cases, one by one,
until an acceptable solution is reached.
All four cases of possible solution to travelling salesman problem tried with element 0.5 as well as zero
element do not provide a desired solution. Thus, we look for the ‘next best’ solution by bringing the next
(non-zero) element 1 along with 0.5 and zero elements into the solution.
Make assignment in cell (C, B) and delete row 4 and column 3. Then make assignments in the usual
manner using elements 1, 0.5 and 0 in the cells as shown in Table 10.32.
334 Operations Research: Theory and Applications

To

Depot Vendor A Vendor B Vendor C Vendor D

Depot – 1.5 0 2 0

Vendor A 1.5 – 0.5 0 0.5

From Vendor B 0 0.5 – 1 0

Vendor C 2 0 1 — 1.5
Table 10.32 Vendor D 0 0.5 0 1.5 –

The set of assignments given in Table 10.38 is a feasible solution to the travelling salesman problem.
The route for the salesman is: Vendor C  Vendor B  Depot  Vendor D  Vendor A  Vendor C. The
total distance (in km) to be covered in this sequence is 15 km.
The reader may try other cases. This is left as an exercise.

SELF PRACTICE PROBLEMS C

1. An airline that operates seven days a week has the timetable which If the salesman starts from city A and has to come back to city
is given below. The crew must have a minimum layovertime of A, which route should he select so that the total distance travelled
five hours between flights. is minimum?
4. A salesman must travel from city to city to maintain his accounts.
Flight No. Delhi-Jaipur Flight No. Jaipur-Delhi This week he has to leave his home base and visit other cities
Dep. Arrival Dep. Arrival and the return home. The table shows the distances (in km)
between the various cities. His home city is city A.
101 7.00 am 8.00 am 201 8.00 am 9.15 am
102 8.00 am 9.00 am 202 8.30 am 9.45 am To City
103 1.30 pm 2.30 pm 203 12.00 noon 1.15 pm A B C D E
104 6.30 pm 7.30 pm 204 5.30 pm 6.45 pm
A – 375 600 150 190
Obtain the pairing of flights that minimizes layover time away B 375 – 300 350 175
from home. For any given pairing, the crew will be based at the city From City C 600 300 – 350 500
that results in the smaller layover. For each pair also mention the
D 160 350 350 – 300
town where the crew should be based.
E 190 175 500 300 –
2. A machine operator processes four types of items on his machine
and he must choose a sequence for them. The set- Use the assignment method to determine the tour that will

up cost per change depends on the items currently on machine minimize the total distance of visiting all cities and then returning
and the set-up to be made according to the following table: home. [Delhi Univ., MBA, 2002]
To 5. A salesman travels from one place to another; he cannot, however,
A B C D travel from one place and back. The distances (in km) between
pairs of cities are given below:
A – 4 7 3
B 4 – 6 3 To City
From Item C 7 6 – 7 P Q R S
D 3 3 7 –
P – 15 25 20
If he processes each of the item once and only once each From City Q 22 – 45 55
week, then how should he sequence the item on his machine? R 40 30 – 25
Use the method for the problem of travelling salesman.
[Delhi Univ., MBA, 2001] S 20 26 38 –

3. A salesman has to visit five cities A, B, C, D and E. The distances The problem is to chalk out a route which enables him to visit
(in hundred kilometres) between the five cities are as follows: each of the cities only once, so that the total distance covered
by him is minimum. [Delhi Univ., MBA, 2003]
To City
6. Products 1, 2, 3, 4 and 5 are to be processed on a machine.
A B C D E
The set-up costs in rupees per change depend upon the product
A – 1 6 8 4 presently on the machine and the set-up to be made. These are
B 7 – 8 5 6 given by the following data:
From City C 6 8 – 9 7 C12 = 16, C13 = 4, C14 = 12, C23 = 6, C24 = 5, C25 = 8,
D 8 5 9 – 8 C35 = 6, C45 = 20; Cij = Cji, Cij =  for i = j
E 4 6 7 8 – for all values of i and j not given in the data. Find the optimum
sequence of products in order to minimize the total set-up cost.
 Assignment Problem 335

7. A salesman has to visit five cities A, B, C, D and E. The To City

distances (in hundred km) between the five cities are as follows:
C1 C2 C3 C4 C5
To City
C1 – 10 13 11 –
A B C D E C2 10 – 12 10 12
A – 17 16 18 14 From City C3 14 13 – 13 11
B 17 – 18 15 16 C4 11 10 14 – 10
From City C 16 18 – 19 17 C5 12 11 12 10 –
D 18 15 19 – 18 How should the salesman plan his trip so that he covers each
E 14 16 17 18 – of these cities no more than once, and completes his trip in
minimum possible time required for travelling?
If the salesman starts from city A and has to come back to city
9. Solve the travelling salesman problem given by the following
A, which route should he select so that total distance travelled
data
by him is minimized? [Delhi Univ., MBA, 2004]
C12 = 20, C13 = 4, C14 = 10, C23 = 5, C24 = 6.
8. The expected times required to be taken by a salesman in C25 = 10, C35 = 6, C45 = 20, where Cij = Cji
travelling from one city to another are as follows: and there is no route between cities i and j if the value for Cij
is not shown.

HINTS AND ANSWERS

1. Calculate layover times when (i) the crew is based in Delhi and 2. (i) A – D, D – B, B – C, C – A; (ii) A – C, C – B, B – D,
(ii) when the crew is based in Jaipur, by considering 15 minutes D – A. Total minimum cost = Rs 19.
= 1 unit. 3. A  C  D  B  E  A; 30 km
(i) 103 – 201; 104 – 202, 101 – 203; 102 – 204 (ii) 103 – 4. A  D  C  B  E  A; 1,165 km
201; 101 – 202; 104 – 203; 102 – 204. 5. P  R  S  Q  P; 98 km
Total minimum layover time is 52 hours and 30 minutes. 6. 1  2  5  3  4  1. Total cost = Rs 47.

CHAPTER SUMMARY
Given a set of tasks to be performed and a set of assignees who are available to perform these tasks (one assignee per task), an
assignment problem deals with the allocation of an assignee to a particular task so as to minimize the total cost of performing all
the tasks. The assignees can be people, machines, vehicles, plants and so on. The formulation of an assignment problem requires
constructing a cost matrix for each possible pair of an assignee to a task.
The aim of this chapter is to enable readers to recognize whether a problem can be formulated and analyzed as an assignment
problem or as a variant of one of these problem types.

CHAPTER CONCEPTS QUIZ

True or False
1. In assignment problem, an optimal assignment requires that the 8. The minimum number of lines required to cover all zeroes can not
maximum number of lines which can be drawn through squares be more than the number of columns/rows in the problem.
with zero opportunity cost be equal to the number of rows. 9. In an assignment problem involving assignment of salesmen to
2. In an assignment problem, if a constant is added or subtracted from different zones to exploit their sales potential fully, if a salesman
every element of any row/column in the given cost matrix, then an cannot be assigned to a particular zone, then the relevant cell value
assignment that minimizes the total cost in one matrix also would be replaced by M for solving the problem.
maximizes the total cost in other matrix. 10. The relevant cost element is replaced by a zero in case a certain
3. The travelling salesman problem can not be solved as an work is not to be assigned to a particular job.
assignment problem.
4. Multiple zeros in columns and rows are all indicative of multiple Fill in the Blanks
optimal solutions.
5. Assignment problem deals with assignment of persons to jobs that 11. An assignment problem that has an objective of maximizing profit
they can perform with varying efficiency. is solved as a minimization problem after converting all cost value
6. If all entries in a row of the cost matrix are increased by a constant, into values.
then it will affect the optimal solution to the problem. 12. When applying Hungarian method, we subtract the smallest element
7. All dummy rows or columns in the assignment problem are from all elements in each column. Then we the smallest element
assumed to be non-zero. in each row from all elements in the row and the remaining elements
represent .
336 Operations Research: Theory and Applications

13. An optimal solution can be found from an opportunity cost table by 26. Maximization assignment problem is transformed into a
drawing lines covering all zero squares when the minimum number minimization problem by
of the lines possible equals the number of . (a) adding each entry in a column from the maximum value in
14. In the assignment problem, the number of allocations in each row that column
and column are . (b) subtracting each entry in a column from the maximum
value in that column
15. The ______ method provides an efficient method of finding the
(c) subtracting each entry in the table from the maximum value
optimal solution without making a direct comparison of every
in that table
solution.
(d) any one of the above
16. If in the assignment problem, there is no assignment in a , then
it implies that the total number of assignments are the number of 27. If there were n workers and n jobs there would be
rows/columns in the square matrix. (a) n! solutions
(b) (n – 1)! solutions
17. In the method, a list of possible assignments among the given (c) (n !)n solutions
assignees and activities is prepared. (d) n solutions
18. The problem of degeneracy makes the transportation method for 28. An assignment problem can be solved by
solving an assignment problem. (a) simplex method
19. The problem is a special case of the balanced transportation (b) transportation method
problem where all rim requirements equals . (c) both (a) and (b)
20. In the assignment problem, the cost is the difference betweenthe (d) none of the above
best possible assignment of an assignee to an activity and thebest 29. For a salesman who has to visit n cities which of the following
possible assignment is an assignment with opportunity cost equal are the ways of his tour plan
to . (a) n!
(b) (n + 1)!
Multiple Choice (c) (n – 1)!
(d) n
21. An assignment problem is considered as a particular case of a 30. The assignment problem
transportation problem because (a) requires that only one activity be assigned to each resource
(a) the number of rows equals columns (b) is a special case of transportation problem
(c) can be used to maximize resources
(b) all xij = 0 or 1
(d) all of the above
(c) all rim conditions are 1
(d) all of the above 31. An assignment problem is a special case of transportation prob-
lem, where
22. An optimal assignment requires that the maximum number oflines
that can be drawn through squares with zero opportunity cost (a) number of rows equals number of columns
(b) all rim conditions are 1
be equal to the number of
(a) rows or columns (c) values of each decision variable is either 0 or 1
(b) rows and columns (d) all of the above
(c) rows + columns 1 32. Every basic feasible solution of a general assignment problem,
(d) none of the above having a square pay-off matrix of order, n should have assign-
ments equal to
23. While solving an assignment problem, an activity is assigned to
(a) 2n + 1 (b) 2n – 1
a resource through a square with zero opportunity cost because
(c) m + n – 1 (d) m + n
the objective is to
33. To proceed with the MODI algorithm for solving an assignment
(a) minimize total cost of assignment
problem, the number of dummy allocations need to be added are
(b) reduce the cost of assignment to zero
(a) n (b) 2n
(c) reduce the cost of that particular assignment to zero
(c) n – 1 (d) 2n – 1
(d) all of the above
34. The Hungarian method for solving an assignment problem can also
24. The method used for solving an assignment problem is called be used to solve
(a) reduced matrix method (a) a transportation problem
(b) MODI method (b) a travelling salesman problem
(c) Hungarian method (c) both (a) and (b)
(d) none of the above (d) only (b)
25. The purpose of a dummy row or column in an assignment 35. An optimal solution of an assignment problem can be obtained only
problem is to if
(a) obtain balance between total activities and total resources (a) each row and column has only one zero element
(b) prevent a solution from becoming degenerate (b) each row and column has at least one zero element
(c) provide a means of representing a dummy problem (c) the data are arrangement in a square matrix
(d) none of the above (d) none of the above

Answers to Quiz
1. F 2. F 3. F 4. T 5. T 6. F 7. F 8. F 9. T 10. T
11. regret 12. subtract; opportunity cost 13. rows or columns 14. equal 15. Hungarian
16. row or columns; less than 17. enumeration 18. inefficient 19. assignment; one
20. job opportunity; zero. 21. (d) 22. (a) 23. (a) 24. (c) 25. (a) 26. (c) 27. (a) 28. (c)
29. (c) 30. (d) 31. (d) 32. (b) 33. (c) 34. (b) 35. (d)
 Assignment Problem 337

CASE STUDY

Case 10.1: Shreya & Sons

Shreya & Sons are planning to develop three products 1, 2 and 3 in its three plants A, B and C. Only a single product
is decided to be introduced in each of the plants. The unit cost of producing one product in a plant, is given in the following
matrix.
Plant
A B C
1 8 12 –
Product 2 10 6 4
3 7 6 6
The management is interested to understand how should the product be assigned so that the total unit cost is
minimized? Also
(a) If the quantity of different products to be produced is as follows, then what assignment shall minimize the
aggregate production cost?

Product Quantity (in units)

1 2,000
2 2,000
3 10,000

(b) If the three products were to be produced in equal quantities, then what is consequence?
It is expected that the selling prices of the products produced by different plants would be different. The prices are
shown in the following table:
Plant
A B C
1 15 18 –
Product 2 18 16 10
3 12 10 8
Assuming the quantities mentioned in (a) above would be produced and sold, how should the products be assigned
to the plants in order to obtain maximum profits?

Case 10.2: City Corporation

A city corporation has decided to carry out repairs on four main dispensaries of the city. The government has agreed
to make a special grant of Rs 50 lakh towards the costs, with a condition that the repairs must be done at the lowest
cost and the quickest time. If conditions warrant, then a supplementary token grant will also be considered favourably.
The corporation has floated tenders and five contractors have sent in their bids. In order to expedite work, one dispensary
will be awarded to only one contractor.
Cost of Repairs (Rs in lakh)
Contractor Dispensary
D1 D2 D3 D4

C1 9 14 19 15
C2 7 17 20 19
C3 9 18 21 18
C4 10 12 18 19
C5 10 15 21 16

You as a consultant suggest to the corporate the best way of assigning the repair work to the contractors. Also
(a) If it is necessary to seek supplementary grants, then what would be the amount sought?
(b) Which of the five contractors will be unsuccessful in his bid?
338 Operations Research: Theory and Applications

Case 10.3: Kamal Transport

A trip from Chennai to Bangalore takes six hours by bus. A typical time-table of the bus service prepared by the
transport authorities in both directions is given below:

Departure from Route Number Arrival at Arrival at Route Number Departure from
Chennai Bangalore Chennai Bangalore

06.00 a 12.00 11.30 1 05.30

07.30 b 13.30 15.00 2 09.00
11.30 c 17.30 21.00 3 15.00
19.00 d 01.00 00.30 4 18.30
00.30 e 06.30 06.00 5 00.00

The cost of providing this service by the transport company depends upon the time spent by the bus crew (driver
and conductor) away from their places in addition to service time. There are five crew. There is, however, a constraint
that every crew should be provided with more than 4 hours of rest before the return trip again and should not wait
for more than 24 hours for the return trip. The company has residential facilities for the crew at Chennai as well as
at Bangalore. Suggest how crew should be assigned with which line of service or which service line should be connected
with which other line, so as to reduce the waiting time to the minimum.

APPENDIX: IMPORTANT RESULTS AND THEOREMS

Theorem 10.1 In an assignment problem, if a constant is added to or subtracted from every elementof
any row or column in the given cost matrix, then an assignment that minimizes the total cost in one matrixalso
minimizes the total cost in the other matrix.
n n n n
Alternately Let xij = x ij , Minimize Z =   cij xij such that  xij =  xij = 1, for all xij = 0 or 1,

i 1 j 1 i 1 j 1
n 

then x  also Minimize Z  = n c x , where c = c – u – v , for i, j = 1, 2, . . ., n and where u
ij   ij ij ij ij i j i
i 1 j 1
and vj are some real numbers.
Proof Given that the assignment xij = x ij minimizes the total cost,
Z = n c x = n (c  u  v ) x
n n
  ij ij   ij i j ij
i 1 j 1 i 1 j 1

n n n n n n
=   cij xij   ui  xij   vj  xij
i 1 j 1 i 1 j 1 j 1 i 1
n n n n n n

=   cij xij   ui   vj = Z –  ui   vj

i 1 j 1 i 1 j 1 i 1 j 1

As the terms n u and n

v , which are subtracted from Z to give Z  are independent of x , therefore
  j ij
i
i 1 j 1

Z  is also minimum for xij = x ij .

n n
Theorem 10.2 If all cij > 0 and it is possible to find a set xij = x ij so that   cij xij = 0, then this

i 1 j 1
assignment is optimal.

Proof Left to the reader as an exercise. A state of nature can be a state of economy (e.g. inflation), a weather
condition, a political development, etc. The states of nature are usually not determined by theaction of
an individual or an organization. These are the result of an ‘act of God’ or result of many situationspushing
in various directions.
The most relevant states of nature may be identified through some technique such as scenario analysis,
i.e. there may be certain possible states of nature that may not have a serious impact on the decision, while
others could be quite serious. In scenario analysis, various knowledgeable section of people are interviewed –
stakeholders, long-time managers, etc., in order to determine the most relevant states of nature that affect the
decision.
 13
“Successful project management requires more front and back end resources (and less middle)
than are usually allocated.”
– Anonymous

The objective of project management study is to schedule activities associated with any project in an
efficient manner so as to complete the project on or before a specified time limit and at the minimum cost
with specified quality standard.
Two project management techniques – PERT and CPM are commonly used to show the logical
sequence of activities to be performed in any project in order to achieve project objectives.

LEARNING OBJECTIVES
After studying this chapter you should be able to
 understand the significance of using PERT and CPM techniques for project management.
 know the basic difference between PERT and CPM network techniques.
 understand phases of any project and various activities that need to be done during these phases.
 construct network diagrams with single and three time estimates of activities involved in a project.
 determine critical path and floats associated with non-critical activities and events along with total project
completion time.
 determine the probability of completing a project on or before the schedule date.
 know how to update a project along with resource leveling and smoothing.
 crash project schedule time and establish a time-cost trade-off for the completion of a project.

CHAPTER OUTLINE
13.1 Introduction  Conceptual Questions C
13.2 Basic Difference Between Pert and Cpm  Self Practice Problems C
13.3 Phases of Project Management  Hints and Answers
13.4 Pert/Cpm Network Components and 13.7 Project Time-Cost Trade-Off
Precedence Relationships  Self Practice Problems D
 Conceptual Questions A  Hints and Answers
 Self Practice Problems A 13.8 Updating of the Project Progress
13.5 Critical Path Analysis 13.9 Resource Allocation
 Conceptual Questions B  Self Practice Problems E
 Self Practice Problems B  Chapter Summary
 Hints and Answers  Chapter Concepts Quiz
13.6 Project Scheduling with Uncertain Activity  Case Study
Times
418 Operations Research: Theory and Applications

13.1 INTRODUCTION
A project involves a large number of interrelated activities (or tasks) that must be completed on or before
a specified time limit, in a specified sequence (or order) with specified quality and minimum cost of using
resources such as personnel, money, materials, facilities and/or space. Examples of projects include,
construction of a bridge, highway, power plant, repair and maintenance of an oil refinery or an air plane;
design, development and marketing of a new product, research and development work, etc. Since a project
involves large number of interrelated activities, therefore it is necessary to prepare a plan for scheduling
and controlling these activities (or tasks). This approach will help in identifying bottlenecks and even
discovering alternate work-plan for the project.
Network Analysis, Network Planning or Network Planning and Scheduling Techniques are used for
planning, scheduling and controlling large and complex projects. These techniques are based on the represen-
tation of the project as a network of activities. A network is a graphical presentation of arrows and nodes for
showing the logical sequence of various activities to be performed to achieve project objectives. In this
chapter, we shall discuss two of these well-known techniques – PERT and CPM.
PERT/CPM is a PERT (Programme Evaluation and Review Technique) was developed in 1956–58 by a research team to
technique used for help in the planning and scheduling of the US Navy’s Polaris Nuclear Submarine Missile project involving
assisting project thousands of activities. The objective of the team was to efficiently plan and develop the Polaris missile
managers carry out
their responsibilities.
system. This technique has proved to be useful for projects that have an element of uncertaintyin the
estimation of activity duration, as is the case with new types of projects which have never been taken up
before.
CPM (Critical Path Method) was developed by E.I. DuPont company along with Remington Rand
Corporation almost at the same time, 1956-58. The objective of the company was to develop a technique
to monitor the maintenance of its chemical plants. This technique has proved to be useful for developing
time-cost trade-off for projects that involve activities of repetitive nature.

13.2 BASIC DIFFERENCE BETWEEN PERT AND CPM

Both PERT and CPM share in common the determination of a critical path and are based on the network
representation of activities and their scheduling that determines the most critical activities to be controlled so
as to meet the completion date of a project. However, the following are some of their major differences.

PERT
1. In PERT analysis, a weighted average of the expected completion time of each activity is calculated
given three time estimates of its completion. These time estimates are derived from probability
distribution of completion times of an activity.
2. In PERT analysis emphasis is given on the completion of a task rather than the activities required to
be performed to complete a task. Thus, PERT is also called an event-oriented technique.
3. PERT is used for one time projects that involve activities of non-repetitive nature (i.e. activities that may
never have been performed before), where completion times are uncertain.
4. PERT helps in identifying critical areas in a project so that necessary adjustments can be made to meet
the scheduled completion date of the project.

CPM
1. In CPM, the completion time of each activity is known with certainty that too unique.
2. CPM analysis explicitly estimate the cost of the project in addition to the completion time. Thus, this
technique is suitable for establishing a trade-off for optimum balancing between schedule time and cost
of the project.
3. CPM is used for completing of projects that involve activities of repetitive nature.

13.2.1 Significance of Using PERT/CPM

1. A network diagram helps to translate complex project into a set of simple and logical arranged activities
and therefore,
 helps in the clarity of thoughts and actions.
 Project Management: PERT and CPM 419

 helps in clear and unambiguous communication developing from top to bottom and vice versa
among the people responsible for executing the project.
2. Detailed analysis of a network helps project incharge to peep into the future because
 difficulties and problems that can be reasonably expected to crop up during the course of execution,
can be foreseen well ahead of its actual execution.
 delays and holdups during course of execution are minimized. Corrective action can also be taken
well in time.
3. Isolates activities that control the project completion and therefore, results in expeditious completion
of the project.
4. Helps in the division of responsibilities and therefore, enhance effective coordination among different
departments/agencies involved.
5. Helps in timely allocation of resources to various activities in order to achieve optimal utilization of
resources.

13.3 PHASES OF PROJECT MANAGEMENT

In general, project management consists of three phases: Planning, Scheduling and Control.

1. Project planning phase In order to understand the sequencing or precedence relationship among
activities in a project, it is essential to draw a network diagram. The steps involved during this phase are
listed below:
(i) Identify various activities (tasks or work packages/elements) to be performed in the project, that
is, develop a breakdown structure (WBS).
(ii) Determine the requirement of resources such as men, materials, machines, money, etc., for carrying
out activities listed above.
(iii) Assign responsibility for each work package. The work packages corresponds to the smallest work
efforts defined in a project and forms the set of tasks that are the basis for planning, scheduling
and controlling the project.
(iv) Allocate resources to work packages.
(v) Estimate cost and time at various levels of project completion.
(vi) Develop work performance criteria.
(vii) Establish control channels for project personnel.
2. Scheduling phase Once all activities have been identified and given unique codes, the project
scheduling (when each of the activities is required to be performed) is taken up. Prepare an estimate
of the likelihood of the project to be completed on or before the specified time. The steps involved during
this phase are listed below:
(i) Identify all people who will be responsible for each task.
(ii) Estimate the expected duration(s) of each activity, taking into consideration the resources required
for their execution in the most economic manner.
(iii) Specify the interrelationship (i.e. precedence relationship) among various activities.
(iv) Develop a network diagram, showing the sequential interrelationship between various activities. For
this, tips such as; what is required to be done; why it must be done, can it be dispensed with; how to
carry out the job; what must precede it; what has to follow; what can be done concurrently, may be
followed.
(v) Based on these time estimates, calculate the total project duration, identify critical path; calculate
floats; carry out resources smoothing (or levelling) exercise for critical (or scare) resources, taking
into account the resource constraints (if any).

3. Project control phase Project control refers to the evaluation of the actual progress (status) against
the plan. If significant differences are observed, then remedial (modifying planning) or reallocation of
resources measures are adopted in order to update and revise the uncompleted part ofthe project.
420 Operations Research: Theory and Applications

Fig. 13.1
Relationship
between Phases
of Project
Management

The relationship among phases of project management is shown in Fig. 13.1. It indicates how the actual
task performance data is used to track deviations from the original plan and schedule. These adjustments
are helpful to aggregate work packages into subsystems and track the progress of these subsystems as
part of the reporting and review procedures.

13.4 PERT/CPM NETWORK COMPONENTS AND PRECEDENCE

RELATIONSHIPS

PERT/CPM network consists of two major components. These are discussed below:
Events Events in the network diagram represent project milestones, such as the start or the completionof
an activity (task) or activities, and occur at a particular instant of time at which some specific part ofthe
project has been or is to be achieved. Events are commonly represented by circles (nodes) in the
network diagram.
The events can be further classified into the following two categories:
(i) Merge Event : An event which represents the joint completion of more than one activity is knownas
a merge event. This is shown in Fig. 13.2(a).

Fig. 13.2
Types of Events

(ii) Burst Event : An event that represents the initiation (beginning) of more than one activity is
known as burst event. This is shown in Fig. 13.2(b).
Events in the network diagram are identified by numbers. Each event should be identified by a number
higher than that the one allotted to its immediately preceding event to indicate progress of work. The
 Project Management: PERT and CPM 421

numbering of events in the network diagram must start from left (start of the project) to the right (completion
of the project) and top to the bottom. Care should be taken that there is no duplication in the numbering.
Activities Activities in the network diagram represent project operations (or tasks) to be conducted. As such
each activity except dummy activity requires resources and takes a certain amount of time for
completion. An arrow is commonly used to represent an activity with its head indicating the direction of
progress in the project.
Activities are identified by the numbers of their starting (tail or initial) event and ending (head, or
terminal) event, for example, an arrow (i, j) between two events; the tail event i represents the start of the
activity and the head event j represents the completion of the activity as shown in Fig. 13.3(a). The activities Activity is a distinct
can be further classified into the following three categories: task that needs to
be performed as
(i) Predecessor Activity: An activity which must be completed before one or more other activities part of a project.
start is known as predecessor activity.
(ii) Successor Activity: An activity which starts immediately after one or more of other activities are
completed is known as successor activity.
(iii) Dummy Activity: An activity which does not consume either any resource and/or time is known
as dummy activity.
A dummy activity in the network is added only to establish the given precedence relationship among
activities of the project. It is needed when (a) two or more parallel activities in a project have same head
and tail events, or (b) two or more activities have some (but not all) of their immediate predecessor activities
in common. A dummy activity is shown by a dotted line in the network diagram as shown in Fig. 13.3(b).

AON project
network where
each activity is
represented by a
node (circle or
rectangle); the
arrows (arcs) show
the precedence
relationships
between the
activities.

Fig. 13.3
Activity-Node
Relationship in
Network Diagram

Network models use the following two types of precedence network to show precedence requirements
of the activities in the project.
Activity-on-Node (AON) network In this type of precedence network each node (or circle) represents a
specific task while the arcs represent the ordering between tasks. AON network diagrams place the activities
within the nodes, and the arrows are used to indicate sequencing requirements. Generally, these diagrams
have no particular starting and ending nodes for the whole project. The lack of dummy activities in these
diagrams always make them easier to draw and to interpret.

Activity-on-Arrow (AOA) network In this type of precedence network at each end of the activity arrow
is a node (or circle). These nodes represent points in time or instants, when an activity is starting or ending.
The arrow itself represents the passage of time required for that activity to be performed.
These diagrams have a single beginning node from which all activities with no predecessors may start.
The diagram then works its way from left to right, ending with a single ending node, where all activities with
no followers come together. Three important advantages of using AOA are as follows:
422 Operations Research: Theory and Applications

(i) Many computer programs are based on AOA network.

(ii) AOA diagrams can be superimposed on a time scale with the arrows drawn, the correct length to
indicate the time requirement.
(iii) AOA diagrams give a better sense of the flow of time throughout a project.
In AOA project
network each In this chapter, only AOA network diagrams will be used. Fig. 13.4 illustrates AON and AOA diagramming.
activity is
represented by
an arrow (or arc). 13.4.1 Rules for AOA Network Construction
Following are some of the rules that have to be followed while constructing a network:

AOA network AON network

1. Activity A A
A

A B
2. B must follow A A B

B
3. B and C must follow A
B
A
A

C
C

4. C must follow A and B A

A

C
C
B

5. C must follow A, and D must A C

Fig. 13.4 follow A and B A C
AOA and AON
Diagramming
B D
B D

1. In network diagram, arrows represent activities and circles the events. The length of an arrow is of no
significance.
2. Each activity should be represented only by one arrow and must start and end in a circle called event.
The tail of an activity represents the start, and head the completion of work.
3. The event numbered 1 denotes the start of the project and is called initial event. All activities emerging
(or taking off) from event 1 should not be preceded by any other activity or activities. An event carrying
the highest number denotes the completion event. A network should have only one initial event and only
one terminal event.
4. The general rule for numbering the event is that the head event should always be numbered larger than
the number at its tail. That is, events should be numbered such that for each activity (i, j), i < j.
5. An activity must be uniquely identified by its starting and completion event, which implies that:
(a) An event number should not get repeated or duplicated.
(b) Two activities should not be identified by the same completion event.
(c) Activities must be represented either by their symbols or by the corresponding ordered pair of
starting-completion events.
6. The logical sequence (or interrelationship) between activities must follow following rules:
(a) An event cannot occur until all its incoming activities have been completed.
(b) An activity cannot start unless all the preceding activities, on which it depends, have been completed.
 Project Management: PERT and CPM 423

(c) Though a dummy activity does not consume either any resource of time, even then it has to follow
the rules 6(a) and (b).

13.4.2 Errors and Dummies in Network

Looping and Dangling Looping (cycling) and dangling are considered as faults in a network. Therefore,
these must be avoided.
(i) A case of endless loop in a network diagram, which is also known as looping, is shown in Fig. 13.5(a),
where activities A, B and C form a cycle.
Due to precedence relationships, it appears from Fig. 13.5(a) that every activity in looping (or cycle) is a
predecessor of itself. In this case it is difficult to number three events associated with activity A, B and
C so as to satisfy rule 6 of constructing the network.

Fig. 13.5
Looping and
Dangling in the
Network Diagram

(ii) A case of disconnect activity before the completion of all activities, which is also known as
dangling, is shown in Fig. 13.5(b). In this case, activity C does not give any result as per the rules
of the network. The dangling may be avoided by adopting rule 5 of constructing the network.
Dummy (or Redundant) Activity The following are the two cases in which the use of dummy activity
may help in drawing the network correctly, as per the various rules.
(i) When two or more parallel activities in a project have the same head and tail events, i.e. two events
are connected with more than one arrow.

Fig. 13.6
Parallel Activities
and Dummy
Activity

In Fig. 13.6(a), activities B and C have a common predecessor – activity A. At the same time, they
have activity D as a common successor. To arrive correct network, a dummy activity for the ending
event B to show that D may not start before B and C, is completed. This is shown in Fig. 13.6(b).
(ii) When two chains of activities have a common event, yet are completely or partly independent of each
other, as shown in Fig. 13.7(a). A dummy which is used in such a case, to establish proper logical
relationships, is also known as logic dummy activity.
In Fig. 13.7(a), if head event of C and D do not depend on the completion of activities A and B, then
the network can be redrawn, as shown in Fig. 13.7(b). Otherwise, the pattern of Fig. 13.7(a) must be followed:
424 Operations Research: Theory and Applications

Fig. 13.7
Dependent and
Independent
Events

Example 13.1 An assembly is to be made from two parts X and Y. Both parts must be turned on a lathe.
Y must be polished whereas X need not be polished. The sequence of activities, together with their
predecessors, is given below.

Activity Description Predecessor Activity

A Open work order –
B Get material for X A
C Get material for Y A
D Turn X on lathe B
E Turn Y on lathe B, C
F Polish Y E
G Assemble X and Y D, F
H Pack G

Draw a network diagram of activities for the project.

Solution The network diagram for the project is shown in Fig. 13.8.

Fig. 13.8
Network Diagram

Example 13.2 Listed in the table are the activities and sequencing necessary for a maintenance job on
the heat exchangers in a refinery.
 Project Management: PERT and CPM 425

Activity Description Predecessor Activity

A Dismantle pipe connections —

B Dismantle heater, closure, and floating front A
C Remove tube bundle B
D Clean bolts B
E Clean heater and floating head front B
F Clean tube bundle C
G Clean shell C
H Replace tube bundle F, G
I Prepare shell pressure test D, E, H
J Prepare tube pressure test and reassemble I

Draw a network diagram of activities for the project.

Solution The network diagram for the project is shown in Fig. 13.9.

Fig. 13.9
Network Diagram

Example 13.3 Listed in the table are the activities and sequencing necessary for the completion of a
recruitment procedure for management trainees (MT) in an organization.

Activity Description Predecessor Activity

A Asking units for requirements –
B Ascertaining management trainees (MTs) requirements for
commercial functions A
C Ascertaining MTs requirement for Accounts/Finance functions A
D Formulating advertisement for MT(A/C) C
E Formulating advertisement for MT (Commercial) B
F Calling applications from the successful candidates passing
through the Institute of Chartered Accountants (ICA) C
G Releasing the advertisement D, E
H Completing applications received G
I Screening of applications against advertisement H
J Screening of applications received from ICA F
K Sending of personal forms I, J
L Issuing interview/regret letters K
M Preliminary interviews L
N Preliminary interviews of outstanding candidates from ICA J
O Final interview M, N

Draw a network diagram of activities for the project.

426 Operations Research: Theory and Applications

Solution The network diagram for the project is shown in Fig. 13.10.

Fig. 13.10
Network Diagram

CONCEPTUAL QUESTIONS A
1. Briefly mention the areas of application of network techniques. 8. How does the PERT technique help a business manager in
2. Compare and contrast CPM and PERT. Under what conditions decision-making? [Delhi Univ., MCom, 2000]
would you recommend the scheduling by PERT? Justify your 9. Critically comment on the assumptions that are mode for PERT/
answer with reasons. CPM analysis of projects.
3. Discuss the various steps involved in the applications of PERT and 10. PERT now has proven to be an effective management tool that can
CPM. be utilized by companies of all sizes in almost every industry.
4. Highlight the difficulties encountered in using network techniques. Discuss. [Delhi Univ., MBA, 2001]
5. Explain the basic logic of arrow networks. 11. ‘PERT provides the framework with which a project can be
described, scheduled and then controlled?’ Discuss.
6. Explain the reasons for incorporating dummy activities in a network
diagram. In what way do these differ from the normal activities? [Delhi Univ., MBA, 2003]
12. What are the major limitations of the PERT model? Discuss.
7. State the circumstances where CPM is a better technique ofproject
[Delhi Univ., MBA, 2004]
management than PERT. [Delhi Univ., MCom, 2002]

SELF PRACTICE PROBLEMS A

1. Draw a network diagrams from the following list of activities: 3. A new type of water pump is to be designed for an automobile.
Its major specifications are given in the table below. Draw the
Activity Predecessor Activity network diagram of activates involved in the project.
Set 1 Set 2 Set 3 Set 4 Activity Description Predecessor
Activity
A – – – –
B – – – A A Drawing prepared and approved –
C – – – A B Cost analysis A
D A A A B C Tool feasibility (economics) A
E B A, B A, B B D Tool manufactured C
F B, C A, B, C B, C D, E E Favourable cost B, C
G D, E, F D, E, F C D F Raw materials procured D, E
G Subassemblies ordered E
H E, F F D, E,F C, F, G
H Subassemblies received G
2. Following is the list of activities associated with the assembly of I Parts manufactured D, F
J Final assembly I, H
a space module for an upcoming mission. Draw the network
diagram of activities involved in the project. K Testing and shipment J

4. Listed in the table are the activities and sequencing required in

Activity Description Predecessor
Activity the computerization of a bank branch.
Activity Description Predecessor
A Construct shell of module – Activity
B Order life support system and scientific
experimentation package from supplier – A Preparation/Deliberations in the
C Order components of control and department –
navigation system – B Dialogue with the union A
D Wire module A C Discussion/Approval/Sanction of local
E Assemble control and navigational management A
system C D Customer education B
F Preliminary test of life support system B E Preparing specifications for the system C
G Install life support system in module D, F F Selection of staff B
H Install scientific experimentation G Order/Acquisition of system D, E
package in module D, F H Alterations in the branch premises C
I Preliminary test of control and
I Training of staff F
navigational system E, F
J Wiring/Power-supply setup of the
J Install control and navigational system
in module H, I branch H
K Final testing and debugging G, J K Transition G, I, J
L Parallel run K
 Project Management: PERT and CPM 427

Draw the network diagram of activities associated with the project.

Activity Description Predecessor
5. The medical faculty of a university is considering to hold a faculty Activity
development programme. It has planned the following activities.
Prepare a network diagram, showing the interrelation- ships of the A Assemble crew –
various activities. B Test lights for burned bulbs –
C Obtain needed bulbs B
Activity Description Predecessor D Paint light standards below banks A
E Replace burned bulbs C
A Design conference meetings and F Deactivate system B
theme – G Check all wiring for wear A, F
B Design front cover of the conference H Obtain needed wire G
proceedings A I Clean lenses on lights A, F
C Prepare brochure and send request J Remove worn wire G
for papers A K Cut new wire to needed lengths J, H
D Compile list of distinguished L Check insulators that support wires J
speakers/guests A M Replace worn insulators L
E Finalize brochure and print it C, D N Replace old wire M, K
F Make travel arrangements for O Splice new wire with old N
distinguished speakers/guests D P Insulate splices O
G Despatch brochures E Q Paint light banks P
H Receive papers for conference G R Replace broken lenses E
S Reactivate system Q, D, I, R
I Edit papers and assemble proceedings F, H
T Clean up R
J Print proceedings B, I

[Delhi Univ., MBA (HCA), 2004] Draw the network diagram of activities associated with the
project.
6. Activities and their description associated with the Haryali
Watershed Project are listed below: 8. Draw the network diagram of activities involved in the project of
introduction of a new product.
Activity Description Predecessor
Activity Description Predecessor
Activity
Activity
A Consult General Water Board (GWB) –
A Develop plan for introduction of the
B Motivate farmers A
project –
C Identify beneficiaries spots B
B Prepare product drawings A
D Obtain demand certificate from GWB C
C Test and approve packages B
E Obtain no objection certificate from
D Approve unit costs C
GWB C
E Prepare and send questionnaires in
F Apply to DPAP C
the field D
G Forward to bank D, E, F
F Evaluate questionnaires E
H Process application G
G Finalize labels, shipping boxes and
I Requisition drilling high G
cartons F
J Release loans H
H Plan media advertising schedule and
K Apply for power H
sales literature F
L Purchase pump sets I, J
I Assess manpower needs, and hire
M Drilling of walls I, J
and train employees F
N Construct storage tank I, J
J Design and develop production
O Prepare field channels I, J
equipment B
P Test water for quality M
K Arrange and instal production equipment J
Q Install pump sets L, P
L Debug equipment K
R Power connections K
M Design and develop product and
S Tests R quality standards J
N Manufacture labels, shipping boxes
Draw the network diagram of activities associated with the project.
and cartons G
7. Listed in the table are the activities and sequencing requirement O Order, receive and check raw materials I, L
necessary for light system maintenance project of a particular P Prepare and check warehouse space F
stadium. Q Conduct trial run O
R Manufacture new product,
first production run P, Q
S Conduct sales meetings H, I
T Contact customers and receive orders O, S
U Process orders and send the product R, T

[Delhi Univ., MBA, 2004]

428 Operations Research: Theory and Applications

13.5 CRITICAL PATH ANALYSIS

The objective of critical path analysis is to estimate the total project duration and to assign starting and
finishing times to all activities involved in the project. This helps to check the actual progress against the
scheduled duration of the project.
The duration of individual activities may be uniquely determined (in case of CPM) or may involve the
three time estimates (in case of PERT), out of which the expected duration of an activity is computed. Having
done this, the following factors should be known in order to prepare the project scheduling.
(i) Total completion time of the project.
(ii) Earlier and latest start time of each activity.
(iii) Critical activities and critical path.
(iv) Float for each activity, i.e. the amount of time by which the completion of a non-critical activity
can be delayed, without delaying the total project completion time.
Consider the following notations for the purpose of calculating various times of events and activities.
Ei = Earliest occurrence time of an event, i. This is the earliest time for an event to occur when all
the preceding activities have been completed, without delaying the entire project.
Li = Latest allowable time of an event, i. This is the latest time at which an event can occur without
causing a delay in project’s completion time.
ESij = Early starting time of an activity (i, j). This is the earliest time an activity should start without
affecting the project completion.
Critical path is the LSij = Late starting time of an activity (i, j). This is the latest time an activity should start without
longest path through delaying the project completion.

the project network;

the activities on the EFij = Early finishing time of an activity (i, j). This is the earliest time an activity should finish without
path are the critical affecting the project completion.

activities, therefore LFij = Late finishing time of an activity (i, j). This is the latest time an activity should finish without
any delay in their
completion must be
delaying the project completion.
avoided to prevent tij = Duration of an activity (i, j).
delay in project As mentioned earlier, a network diagram should have only one initial event and one end event. The
completion. other events are numbered consecutively with integer 1, 2, . . ., n, such that i < j for any two events i and
j connected by an activity, which starts at i and finishes at j.
For calculating the earliest occurrence and latest allowable times for events, following two methods:
Forward Pass method and Backward Pass method are used:
13.5.1 Forward Pass Method (For Earliest Event Time)
In this method, calculations begin from the initial event 1, proceed through the events in an increasing order
of event numbers and end at the final event, say N. At each event, its earliest occurrence time ( E) and earliest
start and finish time for each activity that begins at that event is calculated. When calculations endat the
final event N, its earliest occurrence time gives the earliest possible completion time of the project. The
method may be summarized as follows:
1. Set the earliest occurrence time of initial event 1 to zero. That is, E1 = 0, for i = 1.
2. Calculate the earliest start time for each activity that begins at event i (= 1). This is equal to the earliest
occurrence time of event, i (tail event). That is:
ESij = Ei , for all activities (i, j) starting at event i.
3. Calculate the earliest finish time of each activity that begins at event i. This is equal to the earliest
start time of the activity plus the duration of the activity. That is:
EFij = ESij + tij = Ei + tij , for all activities (i, j) beginning at event i.
4. Proceed to the next event, say j; j > i.
5. Calculate the earliest occurrence time for the event j. This is the maximum of the earliest finish times
of all activities ending into that event, that is,
Ej = Max {EFij } = Max {Ei + tij }, for all immediate predecessor activities.
6. If j = N (final event), then earliest finish time for the project, that is, the earliest occurrence time EN
for the final event is given by
EN = Max { EFij } = Max { EN – 1 + tij }, for all terminal activities
 Project Management: PERT and CPM 429

13.5.2 Backward Pass Method (For Latest Allowable Event Time)

In this method, calculations begin from the final event N. Proceed through the events in the decreasing
order of event numbers and end at the initial event 1. At each event, latest occurrence time ( L) and latest
finish and start time for each activity that is terminating at that event is calculated. The procedure continues
till the initial event. The method may be summarized as follows:
1. Set the latest occurrence time of last event, N equal to its earliest occurrence time (known from forward
pass method). That is, LN = EN , j = N.
2. Calculate the latest finish time of each activity which ends at event j. This is equal to latest occurrence
time of final event. That is:
LFij = Li , for all activities (i, j) ending at event j.
3. Calculate the latest start times of all activities ending at j. This is obtained by subtracting the duration
of the activity from the latest finish time of the activity. That is:
LFij = Lj
and LSij = LFij – tij = Lj – tij, for all activity (i, j) ending at event j.
4. Proceed backward to the event in the sequence, that decreases j by 1.
5. Calculate the latest occurrence time of event i (i < j). This is the minimum of the latest start times of
all activities from the event. That is:
Li = Min {LSij } = Min {Lj – tij }, for all immediate successor activities.
6. If j = 1 (initial event), then the latest finish time for project, i.e. latest occurrence time L1 for the initial
event is given by:
L1 = Min {LSij} = Min{Lj – 1 – tij}, for all immediate successor activities.

13.5.3 Float (Slack) of an Activity and Event

The float (slack) or free time is the length of time in which a non-critical activity and/or an event can be
delayed or extended without delaying the total project completion time.

Slack of an Event
The slack (or float) of an event is the difference between its latest occurrence time (Li) and its earliest
occurrence time (Ei). That is:
Event float = Li – Ei
It is a measure of how long an event can be delayed without increasing the project completion time.
(a) If L = E for certain events, then such events are called critical events.
(b) If L  E for certain events, then the float (slack) on these events can be negative (L < E) or positive
(L > E).

Slack of an Activity
It is the amount of activity time that can increased or delayed without delaying project completion time. This
float is calculated as the difference between the latest finish time and the earliest finish time for the activity.
There are three types of floats for each non-critical activity in a project.

(a) Total float: This is the length of time by which an activity can be delayed until all preceding activities
are completed at their earliest possible time and all successor activities can be delayed until their latest
permissible time.
For each non-critical activity (i, j) the total float is equal to the latest allowable time for the event at
the end of the activity minus the earliest time for an event at the begining of the activity minus the activity
duration. That is:
Total float (TFij) = ( Lj – Ei ) – tij = LSij – ESij = LFij – EFij
(b) Free float: This is the length of time by which the completion time of any non-critical activity can be
delayed without causing any delay in its immediate successor activities. The amount of free float timefor
a non-critical activity (i, j) is computed as follows:
Free float ( FFij ) = (Ej – Ei ) – tij
= Min {ESij , for all immediate successors of activity (i, j)} – EFij
430 Operations Research: Theory and Applications

(c) Independent float: This is the length of time by which completion time of any non-critical activity (i,
j) can be delayed without causing any delay in its predecessor or the successor activities. Independent float
time for each non-critical activity is computed as follows:
Independent float (IFij) = ( Ej – Li) – tij = {ESij – LSij } – tij
The negative value of independent float is considered to be zero.

Remarks 1. Latest occurrence time of an event is always greater than or equal to its earliest occurrence
time (i.e. Li  Ei),
TFij  (Lj – Ei) – tij
This implies that the value of free float may range from zero to total float but will not exceed total float
value. That is, Independent float  Free float  Total float.
2. The calculation of various floats can help the decision-maker in identifying the underutilized resources,
flexibility in the total schedule and possibilities of redeployment of resources.
3. Total float for a non-critical activity may be viewed as follows:
(a) Negative (i.e. L – E < 0): Project completion is behind the schedules date, i.e., resources are not
adequate and activities may not finish in time. This needs extra resources or certain activities need
crashing in order to reduce negative float value.
(b) Positive (i.e. L – E > 0): Project completion is ahead of the schedule date, i.e., resources are surplus.
These resources can be deployed elsewhere or execution of the activities can be delayed.
(c) Zero ( i.e. L = E): Resources are just sufficient for the completion of activities in a project. Any
delay in activities execution will necessarily increase the project cost and time.

13.5.4 Critical Path

Certain activities in any project are called critical activities because delay in their execution will cause further
delay in the project completion time. All activities having zero total float value are identified as critical
activities, i.e., L = E
The critical path is the sequence of critical activities between the start event and end event of a
project. This is critical in the sense that if execution of any activity of this sequence is delayed, then
completion of the project will be delayed. A critical path is shown by a thick line or double lines in the
network diagram.
The length of the critical path is the sum of the individual completion times of all the critical activities and
defines the longest time to complete the project. The critical path in a network diagram can be identified as:
(i) If Ei-value and Lj-value for any tail and head events is equal, then activity (i, j) between such events
is referred as critical, That is, Ej = Lj and Ei = Li.
(ii) On critical path Ej – Ei = Lj – Li = tij.

Example 13.4 An established company has decided to add a new product to its line. It will buy the product
from a manufacturing concern, package it, and sell it to a number of distributors that have been selected on
a geographical basis. Market research has already indicated the volume expected and the sizeof sales force
required. The steps shown in the following table are to be planned.

Activity Description Predecessors Duration (days)

A Organize sales office – 6
B Hire salesmen A 4
C Train salesmen B 7
D Select advertising agency A 2
E Plan advertising campaign D 4
F Conduct advertising campaign E 10
G Design package – 2
H Setup packaging facilities G 10
I Package initial stocks J, H 6
J Order stock from manufacturer – 13
K Select distributors A 9
L Sell to distributors C, K 3
M Ship stocks to distributors I, L 5
 Project Management: PERT and CPM 431

The precedence relationship among these activities are shown in the following figure.

As the figure shows, the company can begin to organize the sales office, design the package, and order
the stock immediately. Also the stock must be ordered and the packing facility must be set up before the
initial stocks are packaged.
(a) Draw an arrow diagram for this project.
(b) Indicate the critical path.
(c) For each non-critical activity, find the total and free float. [Delhi Univ., MBA, Dec. 2005]
Solution (a) The arrow diagram for the given project, along with E-values and L-values, is shown in
Fig. 13.11. Determine the earliest start time – Ei and the latest finish time – Lj for each event by proceeding
as follows:
Forward Pass Method
E1 = 0 E2 = E1 + t1, 2 = 0 + 6 = 6
E3 = E1 + t1, 3 = 0 + 2 = 2 E4 = Max {Ei + ti, 4}
E5 = E2 + t2, 5 = 6 + 4 = 10 = Max {E1 + t14; E3 + t34}
i = 1, 3
= Max {0 + 13, 2 + 10} = 13
E6 = Max {Ei + ti, 6} = Max {E2 + t2, 6 ; E5 + t5, 6} E8 = E7 + t7, 8 = 8 + 4 = 12
i = 2, 5 E10 = Max {Ei + ti,10}
= Max {6 + 9; 10 + 7} = 17 i = 8, 9
E7 = E2 + t2, 7 = 6 + 2 = 8 = Max {E8 + t8, 10 ; E9 + t9, 10}
E9 = Max {Ei + ti, 9} = Max {E4 + t4, 9; E6 + t6, 9} = Max {12 + 10; 20 + 5} = 25.
i = 4, 6
= Max {13 + 6; 17 + 3} = 20

Fig. 13.11
Network Diagram
432 Operations Research: Theory and Applications

Backward Pass Method

L10 = E10 = 25 L9 = L10 – t9, 10 = 25 – 5 = 20
L8 = L10 – t8, 10 = 25 – 10 = 15 L7 = L8 – t7, 8 = 15 – 4 = 11
L6 = L9 – t6, 9 = 20 – 3 = 17 L5 = L6 – t5, 6 = 17 – 7 = 10
L4 = L9 – t4, 9 = 20 – 6 = 14 L3 = L4 – t3, 4 = 14 – 10 = 4
L2 = Min {Lj – t2, j} L1 = Min {Lj – t1, j}
j = 5, 6, 7 j = 2, 3, 4
= Min { L5 – t2, 5; L6 – t2, 6; L7 – t2, 7} = Min {L2 – t1, 2; L3 – t1, 3; L4 – t1, 4}
= Min {10 – 4; 17 – 9; 11 – 2} = 6 = Min {6 – 6; 4 – 2; 14 – 13} = 0
(b) The critical path in the network diagram (Fig. 13.11) has been shown. This has been done by double
lines by joining all those events where E-values and L-values are equal.
The critical path of the project is: 1 – 2 – 5 – 6 – 9 – 10 and critical activities are A, B, C, L and M.
The total project completion time is 25 weeks.
(c) For each non-critical activity, the total float and free float calculations are shown in Table 13.1.

Activity Duration Earliest Time Latest Time Float

( i, j ) ( tij ) Start Finish Start Finish Total Free
( Ei ) ( Ei + tij ) ( Lj – tij ) Lj ( Lj – tij – Ei ( Ej – Ei ) – tij
1 – 3 2 0 2 2 4 2 0
1 – 4 13 0 13 1 14 1 0
2 – 6 9 6 15 8 17 2 2
2 – 7 2 6 8 9 11 3 0
3 – 4 10 2 12 4 14 2 1
4 – 9 6 13 19 14 20 1 1
Table 13.1
7 – 8 4 8 12 11 15 3 0
Total Float and
Free Float 8 – 10 10 12 22 15 25 3 3

Example 13.5 An insurance company has decided to modernize and refit one of its branch offices. Some of
the existing office equipments will be disposed of but the remaining will be returned to the branch after the
completion of the renovation work. Tenders are invited from a number of selected contractors. The
contractors would be responsible for all the activities in connection with the renovation work excepting the
prior removal of the old equipment and its subsequent replacement.
The major elements of the project have been identified, as follows, along with their durations and
immediately preceding elements.

Activity Description Duration (weeks) Immediate Predecessors

A Design new premises 14 –

B Obtain tenders from the contractors 4 A
C Select the contractor 2 B
D Arrange details with selected contractor 1 C
E Decide which equipment is to be used 2 A
F Arrange storage of equipment 3 E
G Arrange disposal of other equipment 2 E
H Order new equipment 4 E
I Take delivery of new equipment 3 H, L
J Renovations take place 12 K
K Remove old equipment for storage or disposal 4 D, F, G
L Cleaning after the contractor has finished 2 J
M Return old equipment for storage 2 H, L

(a) Draw the network diagram showing the interrelations between the various activities of the project.
(b) Calculate the minimum time that the renovation can take from the design stage.
 Project Management: PERT and CPM 433

(c) Find the effect on the overall duration of the project if the estimates or tenders can be obtained in two
weeks from the contractors by reducing their numbers.
(d) Calculate the ‘independent float’ that is associated with the non-critical activities in the network
diagram.
Solution (a) The network diagram for the given project, along with E-values and L-values, is shown in
Fig. 13.12.

Fig. 13.12
Network Diagram

The critical path in the network diagram (Fig 13.12) has been shown by double lines joining all those events
where E-values and L-values are equal.
(b) The critical path of the project is: 1 – 2 – 3 – 4 – 7 – 8 – 9 – 10 – 12 and critical activities are A, B,
C, D, K, J, L and I. The total project completion time is 42 weeks.
For non-critical activities, the total float, free float and independent float calculations are shown in
Table 13.2.

Earliest Time Latest Time Float

Activity Duration Start Finish Start Finish Total Free Independent
(i, j) (tij) ( Ei ) ( Ei + tij ) ( Lj – tij ) Lj ( Lj – tij ) – Ei ( Ej – Ei ) – tij ( Ej – Li ) – tij

2 – 5 2 14 16 16 18 2 0 0
6 – 7 3 16 19 18 21 2 2 0
5 – 7 2 16 18 19 21 3 3 1 Table 13.2
5 – 10 4 16 20 35 39 19 19 17 Calculation of
10 – 11 2 39 41 40 42 1 0 0 Floats

(c) The effect on the overall project duration, if the time of activity B is reduced to 2 weeks instead of 4
weeks, is shown in Table 13.3.

Path Duration
(i) A – E – H – I 23
(ii) A – E – H – M 22
(iii) A – B – C – D – K – J – L – I
(Critical path 42 weeks) 40 (New critical path)
(iv) A – B – C – D – K – J – L – M
(Critical path 41 weeks) 39
(v) A – E – G – K – J – L – I 39
(vi) A – E – G – K – J – L – M 39
(vii) A – E – F – K – J – L – I 40 (New critical path)
(viii) A – E – F – K – J – L – M 39 Table 13.3
434 Operations Research: Theory and Applications

CONCEPTUAL QUESTIONS B
1. Explain the following terms in PERT/CPM. 3. What is meant by the term critical activities, and why is it
(i) Earliest time, (ii) Latest time, (iii) Total activity time, (iv) Event necessary to know about them?
slack, and (v) Critical path. 4. Explain the significance of ‘working out of float’ in the network
2. (a) What is float? What are the different types of floats? of project activities. Discuss, in brief, the different types of floats.
(b) Discuss in brief: (i) total float, and (ii) free float. Also explain [Delhi Univ., MBA, 1998 ]
their uses in network. 5. Explain the following terms in the context for project management:
(i) Resources float, (ii) Activity variance, (iii) Project variance.

SELF PRACTICE PROBLEMS B

1. An architect has been awarded a contract to prepare plans for G Clean shell C 6
an urban renewal project. The job consists of the following activities H Replace tube bundle F, G 8
and their estimated times: I Prepare shell pressure test D, E, H 24
J Prepare tube pressure test
Activity Description Immediate Time and make the final reassembly I 16
Predecessors (days)

A Prepare preliminary sketches – 2 (a) Draw a network diagram of activities for the project.
(b) Identify the critical path. What is its length?
B Outline specifications – 1
(c) Find the total float and free float for each task.
C Prepare drawings A 3 [Delhi Univ., MBA, 2004]
D Write specifications A, B 2
4. The sales manager of Domestic Products Limited was informed
E Run off prints C, D 1 by the company’s R&D department about the completion of the
F Have specification B, D 3 prototype of a particular product. He consulted the production
G Assemble bid packages E, F 1 manager on the time taken to produce the first batch of theproduct,
which is needed for demonstration in his sales promo- tion
(a) Draw the network diagram of activities for the project. programme. He also decided to invite a few industrial
(b) Indicate the critical path, and calculate the total float and representatives to the demonstration of this new product and
free float for each activity. through them to launch it in the market. The various activities
involved in this marketing project, their descriptions, estimated
2. A research and development department is developing a new
duration (in days) and immediate predecessors are given in the
power supply for a console television set. It has broken the job
following table:
down into the following:

Job Description Immediate Time Activity Description Time Immediate

Predecessors (days) (days) Predecessors

A Determine output voltages – 5 A Collect data on specifications

B Determine whether to use solid and capabilities 4 –
state rectifiers A 7 B Prepare operation manual 4 A
C Choose rectifier B 2 C Chart out promotion programme 4 B
D Choose filters B 3 D Make copies of manual and
E Choose transformer C 1 promotion material 9 B
F Choose chassis D 2
E Produce first batch for
G Choose rectifier mounting C 1
demonstration 16 B
H Layout chassis E, F 3
F Prepare list of press
I Build and test G, H 10
representatives 2 C
G Chief executive’s conference
(a) Draw the network diagram of activities involved in the
with Managers 1 C
project and indicate the critical path.
(b) What is the minimum completion time for the project? H Press representatives reach
Bombay 2 F, G
3. The following maintenance job has to be performed periodically
on the heat exchangers in a refinery: I Promotional meetings 4 D, H
J Product demonstration 2 E, I
Task Description Immediate Time K Press representatives return
Predecessors (days) home 2 J

A Dismantle pipe connections – 14 (a) Draw the network diagram for the project.
B Dismantle header, closure, and (b) Identify the critical path. What is the maximum time required
floating head front A 22 to complete the project.
C Remove tube bundle B 10 (c) Find the total float and free float (if any) for all the non-
D Clean bolts B 16 critical activities.
E Clean header and floating head front B 12 5. Listed in the table are the activities and sequencing require-
F Clean tube bundle C 10 ments necessary for the completion of a research report.
 Project Management: PERT and CPM 435

H Receive papers for conference G 25

Activity Description Time Immediate
I Edit papers and assemble
(days) Predecessors
proceedings F, H 10
A Literature search – 6 J Print proceedings B, I 20
B Formulation of hypothesis – 5
C Preliminary feasibility study B 2 (a) Prepare a network diagram showing the interrelationships
D Formal proposal C 2 of the various activities.
E Field analysis A, D 2 (b) Find the total time required to hold the conference.
F Progress report D 1 (c) Compute the total float for the non-critical activities.
G Formal research A, D 6 [Delhi Univ., MBA (HCA), 2002 ]
H Data collection E 5 8. The following table gives a list of various activities and their
I Data analysis G, H 6 immediate predecessors involved in installation of CAT scanner
J Conclusions I 2 in a hospital:
K Rough draft G 4
L Final copy J, K 3 Activity Description Immediate Expected
M Preparation of oral Predecessors Duration
presentation L 1 (days)
(a) Draw a network diagram for the project.
(b) Find the critical path. What is its length? A Finalization of the layout plans – 2
(c) Find the total float and the free float for each non-critical B Demolition of the structures A 6
activity. C Walls erection B 12
D Flooring B 8
6. Listed in the table are the activities and their simplified sequenc- E Electrical wiring C 6
ing requirements for publishing a textbook. F Air conditioning ducting C 4
G Fire alarm installation C 3
Activity Description Preceding Expected H False ceiling and light fittings E, F, G 10
Activity Completion I Wall plastering and painting H 9
Time
J Equipment installation D, I 6
(months)
K Calibration and testing J 3
A Write book – 12 L Final finishing K 2
B Design book A 1 M Handing over L 1
C Edit manuscript A 6
D Check editing C 2 (a) Prepare a network diagram for the project.
E Accept design B 1 (b) Find the total time required to install CAT scanner.
F Copy edit D, E 2 (c) Calculate the total float for various non-critical activities.
[Delhi Univ., MBA (HCA), 2000, 2003 ]
G Prepare artwork D, E 4
H Accept and correct artwork G 1/2 9. A medical association prepares an annual programme each year,
I Set galleys F 4 giving the monthly meeting dates, background on the speakers, an
J Check and correct galleys I 1 abstract of their talks, and an alphabetic listing,both by name and
K Full page proofs H, J 2 medical college/hospital affiliation, of all dues paying members.
L Check and correct pages K 1 The programme is mailed to these members as well as to
M Prepare index K 1 selected individuals and organizations. The activi- ties to be
N Set and correct index M 1/2 performed are listed as follows:
O Check camera-ready copy L, N 1/2
P Print and bind book O 1 Activity Description Preceding Estimated
Activities Time
(a) Construct the network of activities and find the critical path. (weeks)
(b) For each non-critical activity, find total the float and free float.
A Decide on general orientation
7. Delhi Medical Association is considering to hold a conference. for this year’s programme – 1
The following table gives the list of activities involved, their B Get commitments from speakers
immediate predecessors, and their duration (in days): and abstracts of their talks A 4
C Solicit advertising to appear in
Activity Description Predecessor Duration the programme A 3
(days) D Mail out dues notices and
A Design conference meetings wait for response – 6
and theme – 3 E Prepare list of dues-paying
B Design front cover of the members D 1
conference proceedings A 2 F Give copy to printer before
proofreadng B, C, E 2
C Prepare brochure and send
request for papers A 6 G Get programme printed and
assembled F 2
D Compile list of distinguished
H Prepare final mailing list E 1
speakers/guests A 3
I Staff envelopes and mail
E Finalize brochure and print it C, D 7
programmes G, H 1
F Make travel arrangements for
speakers/guests D 4 (a) Develop a network diagram to show the relationship be-tween
G Despatch brochures E 3 all activities. Specify the activities on the critical path and the
project completion time.
436 Operations Research: Theory and Applications

(b) The programme chairman now claims that it will take him H Manufacture from raw materials
6 weeks to get commitments and abstracts from the speakers. for test and first production runs G 4
Will this delay the project completion time? How will it affect I Receive containers and
the critical path? packaging supplies P 1
[Delhi Univ., MBA (HCA), 2005] J Have personnel available for
10. A reactor and storage tank are interconnected by a 3 insulated first production run F 0
process line that needs periodic replacement. There are valves K Run manufacturing test D, J, H, I, T 3
along the lines and at the terminals and these need replacing as L Run first production K 7
well. No pipe and valves are in stock. Accurate, as built, drawings Marketing activities
exist and are available. This line is overhead and requires
scaffolding. Pipe sections can be shop-fabricated at the plant. M Price product B, S 2
Adequate craft labour is available. N Prepare artwork for advertising M 4
You are the maintenance and construction superintendent O Send out advertising materials
responsible for his project. The works engineer has requested your and packaging orders to suppliers N 5
plan and schedule for a review with the operating super- vision. P Produce advertising and
The plant methods and standards section has furnished the packaging materials O 1
following data. The precedent for each activity have been Q Hold sales meeting K, S 2
determined from a familiarity with similar projects. R Training salespeople Q 3
Accounting activities
Activity Description Time Precedents
S Determine cost of new product B 2
(hrs)
T Determine cost of the new
A Develop required material list 8 – product inventory S 3
B Procure pipe 200 A
(a) Draw a network diagram for the project.
C Erect scaffold 12 –
(b) Identify the critical path and determine its length.
D Remove scaffold 4 H, M
12. In Border Roads Organization, the vehicles and equipment are
E Deactivate line 8 –
being overhauled in base workshop. The overhauling of vehicles
F Prefabricate sections 40 B
involves a number of jobs. These are given in the following table.
G Place new pipes 32 F, K
H Fit up pipe and valves 8 G, J Job Description Immediate Expected
I Procure valves 225 A Predecessors Time
J Place valves 8 I, K (days)
K Remove old pipe and valves 35 C, E
L Insulate 24 G, J A Dismantling of vehicle – 3
M Pressure test 6 H B Inspection and preparation
N Clean-up and start-up 4 D, M of visual inspection on report A 2
C Stripping minor assys B 2
(a) Sketch the network diagram of the project plan. D Stripping major assys B 3
(b) Make the forward pass and backward pass calculations on this
E Overhauling engine C 40
network, and indicate the critical path and its length.
(c) Calculate the total float and free float for each of the non- F Cleaning chassis C, D 2
critical activities. G Cleaning body and cabin C, D 2
11. A company has decided to market a new product for the consumer H Repairing gear box C 2
market. The problems of how to plan and control the various I Repairing break system D 1
phases of this project – sales promotion, training of salespeople, J Repairing axle and propeller shaft D 6
pricing, packaging, advertising, and manufacturing K Repairing road spring D 2
– are obvious to the management of this firm. They have asked you
L Repairing chassis F 2
to guide them through this difficult venture using CPM. The first firm
to market this type of product will reap substantial profits and will M Repairing body and cabin G 5
enhance its image by marketing such a revolu- tionary product. A N Placing chassis and fitment of
list of the activities, with the expected time duration for each, is engine E, L 2
given in the table, in terms of weeks. O Fitting major assys N, H 1
P Fitting minor assys and cabin N, I, J, K, M 2
Activity Description Precedence Time
Q Checking and rectification O, P 1
(weeks)
R Mechanical inspection for passing Q 1
S Painting R 1
Manufacturing activities T Checking and rectification before
A Study equipment requirement – 1 final inspection S 1
B Select supplier of equipment A 1 U Final inspection and passing T 1
C Study manufacturing procedures B 3 V Closing the job card U 1
D Study quality control procedures C 3
E Study purchasing and inventory
rules B 3 (a) Draw a network diagram for the overhauling project.
F Receive and install equipment B 8 (b) Identify the critical path. What is its length?
G Place order for raw materials E 2 (c) Find the total float and free float for each task.



Project Management: PERT and CPM 437

HINTS AND ANSWERS

1. A – D – F – G 4. A – B – E – J – K; 28 days
2. A – B – D – F – H – I; 30 days 5. B – C – D – E – H – I – J – L – M; 28 days
3. A – B – C – F – H – I – J; 104 days 9. A – B – C – E – N – P – Q – R – S – T – U – V; 56 days

13.6 PROJECT SCHEDULING WITH UNCERTAIN ACTIVITY TIMES

PERT was developed to handle projects where the time duration for each activity is not known with certainty
but is a random variable that is characterized by  (beta)-distribution. To estimate the parameters: mean and
variance, of the -distribution three time estimates for each activity are required to calculate its expected
completion time. The three-time estimates that are required are as under.
(i) Optimistic time ( to or a ) The shortest possible time (duration) in which an activity can be performed
assuming that everything goes well.
(ii) Pessimistic time (tp or b ) The longest possible time required to perform an activity under extremely
bad conditions. However, such conditions do not include natural calamities like earthquakes, flood,
etc.
(iii) Most likely time ( tm or m ) The time that would occur most often to complete an activity, if the
activity was repeated under exactly the same conditions many times. Obviously, it is the completion
time that would occur most frequently (i.e. model value).
The -distribution is not necessarily symmetric, the degree of skewness depends on the location of
tm to to and tp. The range of optimistic time (to) and pessimistic time (tp) is assumed to enclose every possible
duration of the activity. The most likely completion time (tm) for an activity may not be equal tothe
midpoint (to + tp)/2 and may occur to its left or to its right as shown in Fig. 13.13.
In Beta-distribution the midpoint (to + tp)/2 is given half weightage than that of most likely point (tm).
Thus, the expected or mean (te or ) time of an activity, that is also the weighted average of three time
estimates, is computed as the arithmetic mean of (to + tp)/2 and 2 tm. That is:

 Fig. 13.13
An Example of
 Beta Distribution

( to  t p )/2  2 tm to  4 tm  t p
Expected time of an activity ( te )  
3 6

If duration of activities associated with the project is uncertain, then variance describes the dispersion
(variation) in the activity time values. The calculations are based on the concept of normal distribution where
99 per cent of the area under normal curve falls within  3 from the mean or fall within the range
approximately 6 standard deviation in length. Therefore, the interval (to, tp) or range (tp – to) is assumed
to enclose about 6 standard deviations of a symmetric distribution. Thus, if i is the standard deviation
of the duration of activity i, then
6  t t or t p  to
i p o
i 
6
2 1 2
Variance of activity time, i
  (t p  to ) 
6 




 438 Operations Research: Theory and Applications

If duration of activities is a random variable, then the variance of the total critical path’s duration is
obtained by adding variances of individual critical activities. Suppose c is the standard deviation of the
critical path. Then:
 2 =  2 and  =

2
c i c

13.6.1 Estimation of Project Completion Time

Due to variation in the activity completion time, there are chances of variation in the scheduled completion
time of the project. Thus, decision-maker needs to understand the probability of actually meeting the
scheduled time.
The probability distribution of times for completing an event can be approximated by the normal
distribution using statement of central limit theorem. Thus, the probability of completing the project on the
schedule (or desired) time, Ts is given by:
 Ts  Te 
Prob  Z  
i
 
where, Te = expected completion time of the project
Z = number of standard deviations, the scheduled completion time is away from the expected
(mean) time.
2 =  2 +  2 + . . . +  2 is the sum of variances of critical activities.
i 1 2 n

Hence, revised scheduled completion time of the project can be calculated as: Ts = Z i + Te , where value
of Z corresponds to the probability of project completion time. The computation of Ts enables a decision-
maker to make certain commitments, knowing the degree of risk. The expected completion time (Te) of the
project is obtained by adding the expected time of each critical activity.
Example 13.6 The following network diagram represents activities associated with a project:
Activities : A B C D E F G H I
Optimistic time, t0 : 5 18 26 16 15 6 7 7 3
Pessimistic time, tp : 10 22 40 20 25 12 12 9 5
Most likely time, tm : 8 20 33 18 20 9 10 8 4

Determine the following:

(a) Expected completion time and variance of each activity
(b) The earliest and latest expected completion times of each event.
(c) The critical path.
(d) The probability of expected completion time of the project if the original scheduled time of completing
the project is 41.5 weeks.
(e) The duration of the project that will have 95 per cent chance of being completed.
 Project Management: PERT and CPM 439

Solution (a) Calculations for expected completion time (te) of an activity and variance (2), using
following formulae are shown in Table 13.4.
1 2 1 2
te 
( t  4 tm  t p ) and i   (t p  to )
6 o 6 
(b) The earliest and latest expected completion time for all events considering the expected completion time
of each activity are shown in Table 13.4.

1 1
Activity to tp tm te = (t + 4tm +tp) 2 = [ (tp – to)]2
6 o 6
1 – 2 5 10 8 7.8 0.696
1 – 3 18 22 20 20.0 0.444
1 – 4 26 40 33 33.0 5.429
2 – 5 16 20 18 18.0 0.443
2 – 6 15 25 20 20.0 2.780
3 – 6 6 12 9 9.0 1.000
4 – 7 7 12 10 9.8 0.694
5 – 7 7 9 8 8.0 0.111
6 – 7 3 5 4 4.0 0.111 Table 13.4

Forward Pass Method

E1 = 0 E2 = E1 + t1, 2 = 0 + 7.8 = 7.8
E3 = E1 + t1, 3 = 0 + 20 = 20 E4 = E1 + t1, 4 = 0 + 33 = 33
E5 = E2 + t2, 5 = 7.8 + 18 = 25.8 E6 = Max {Ei + ti, 6} = Max {E2 + t2, 6 ; E3 + t3, 6}
E7 = Max {Ei + ti,7} = Max {7.8 + 20; 20 + 9} = 29
= Max {E4 + t4, 7 ; E5 + t5, 7 ; E6 + t6, 7}
= Max {33 + 9.8 ; 25.8 + 8; 29 + 4} = 42.8

Backward Pass Method

L7 = E7 = 42.8 L6 = L7 – t6, 7 = 42.8 – 4 = 38.8
L5 = L7 – t5, 7 = 42.8 – 8 = 34.8 L4 = L7 – t4, 7 = 42.8 – 9.8 = 33
L3 = L6 – t3, 6 = 38.8 – 9 = 29.8 L1 = Min {Lj – t1j}
L2 = Min {Lj – t2, j} = Min {L2 – t1, 2; L3 – t1, 3 ; L4 – t1, 4}
= Min {L5 – t2, 5 ; L6 – t2, 6 } = Min {16.8 – 7.8 ; 29.8 – 20 ; 33 – 33} = 0
= Min {34.8 – 18 ; 38.8 – 20} = 16.8
The E-value and L-values are shown in Fig. 13.14.

Fig. 13.14
Network Diagram
440 Operations Research: Theory and Applications

(c) The critical path is shown by thick line in Fig. 13.14 where E-values and L-values are the same. The
critical path is: 1 – 4 – 7 and the expected completion time for the project is 42.8 weeks.
(d) Expected length of critical path, Te = tC + tG = 33 + 9.8 = 42.8 weeks (Project duration).
Variance of critical path length, 2 =  2 +  2 = 5.429 + 0.694 = 6.123 weeks.
C G

Since Ts = 41.5, Te = 42.8 and   6.123  2.474 , the probability of meeting the schedule time is given by:
 T  Te   41.5  42.8
Prob Z  s = P Z = Prob (Z  – 0.52)
   2.474 
  
= 0.5 – 0.1952 = 0.3048 (from normal distribution table)
Thus, the probability that the project can be completed in less than or equal to 41.5 weeks is 0.3048. In
other words, the probability that the project will get delayed beyond 41.5 weeks is 0.6952.
 T  Te 
Given that P Z  s  0.95
  

But Z0.95 = 1.64, from normal distribution table. Thus,
T  42.8
1.64  s or Ts = 1.64 × 2.474 + 42.8 = 46.85 weeks.
2.474

Example 13.7 A small project involves 7 activities, and their time estimates are listed in the following
table. Activities are identified by their beginning (i ) and ending ( j) node numbers.

Activity Estimated Duration (weeks)

(i– j) Optimistic Most Likely Pessimistic

1 – 2 1 1 7
1 – 3 1 4 7
1 – 4 2 2 8
2 – 5 1 1 1
3 – 5 2 5 14
4 – 6 2 5 8
5 – 6 3 6 15

(a) Draw the network diagram of the activities in the project.

(b) Find the expected duration and variance for each activity. What is the expected project length?
(c) Calculate the variance and standard deviation of the project length. What is probability that the project
will be completed:
(i) at least 4 weeks earlier than expected time.
(ii) no more than 4 weeks later than expected time.
(d) If the project due date is 19 weeks, what is the probability of not meeting the due date.
Given: Z : 0.50 0.67 1.00 1.33 2.00
Prob. : 0.3085 0.2514 0.1587 0.0918 0.0228 [Gujarat, MBA, 2001]
Solution The network diagram of activities in the project is shown in Fig. 13.16. The earliest and latest
expected time for each event is calculated by considering the expected time of each activity, as shown in
Table 13.5.
1 1
Activity to tm tp te = (t + 4tm +tp ) 2 = [ (tp – to)]2
6 o 6
1 – 2 1 1 7 2 1
1 – 3 1 4 7 4 1
1 – 4 2 2 8 3 1
2 – 5 1 1 1 1 0
3 – 5 2 5 14 6 4
4 – 6 2 5 8 5 1
Table 13.5 5 – 6 3 6 15 7 4

The E-values and L-values based on expected time (te) of each activity are shown in Fig. 13.15.
 Project Management: PERT and CPM 441

Fig. 13.15
Network Diagram

(a) Critical path is: 1 – 3 – 5 – 6.

(b) The expected duration and variance for each activity is shown in Table 13.5. The expected project
length is the sum of the duration of each critical activity:
Expected project length = 1 – 3 – 5 – 6 = 4 + 6 + 7 = 17 weeks
(c) Variance of the project length is the sum of the variances of each critical activity:
Variance of project length = 1 – 3 – 5 – 6 = 1 + 4 + 4 = 9 weeks

Therefore, Standard deviation,   9  3

(i) Probability that the project will be completed at least 4 weeks earlier (i.e. 13 weeks) than the expected
project duration of 17 weeks is given by

Prob. Z Ts  Te 13 17  = Prob. {Z  – 1.33} = 0.5 – 0.4082 = 0.0918
 
 
 3
 
Thus the probability of completing the project in less than 13 days is 9.18 per cent.
(ii) Probability that the project will be completed in 4 weeks later (i.e. 21 weeks) than expected project
duration of 17 weeks is given by
 21  17 
P Z = P(Z  1.33) = 0.5 + 0.4082 = 0.9082
 
3
 

CONCEPTUAL QUESTIONS C

1. Explain the following in the context of project management: (i) 4. ‘PERT takes care of uncertain durations.’ How far is this statement
Activity variance, and (ii) Project variance. correct? Explain with reasons.
2. Explain the following terms in PERT: (i) Three time estimates, 5. How does PERT provide for uncertainty in activity time esti- mates?
(ii) Expected time, and (iii) Activity variance. What is the rationale for using beta probability distribu- tion?
3. What are the basic assumptions underlying the ‘expected time’ [Delhi Univ., MBA, 2005]
estimate?

SELF PRACTICE PROBLEMS C

1. A project has the following activities and other characteristics: G B 3 9 27

H E, F 1 4 7
Activity Preceding Time Estimates (weeks) I G 4 19 28
Activity Optimistic Most Likely Pessimistic
(a) Draw the network diagram for the project.
A – 4 7 16 (b) Identify the critical path.
B – 1 5 15 (c) Prepare the activity schedule for the project.
C A 6 12 30 (d) Determine the mean project completion time.
D A 2 5 8 (e) Find the probability that the project is completed in 36
E C 5 11 17 weeks. [Delhi Univ., MCom, 2001 ]
F D 3 6 15
442 Operations Research: Theory and Applications

2. A small project consists of seven activities, the details of which

are given below: Activity Immediate Activity Time (weeks)
Predecessor Most Most Most
Activity Duration (days) Immediate Optimistic Likely Pessimistic
Most Likely Optimistic Pessimistic Predecessor A – 3 4 5
B – 4 8 10
A 3 1 7 – C B 5 6 8
B 6 2 14 A D A, C 9 15 10
C 3 3 3 A E B 4 6 8
D 10 4 22 B, C F D, E 3 4 5
E 7 3 15 B G D, E 5 6 8
F 5 2 14 D, E H D, E 1 3 4
G 4 4 4 D I G 2 4 5
(a)
Draw the network, number the nodes, find the critical path, the J F, I 7 8 10
expected project completion time and the next most critical K G 4 5 6
path. L H 8 9 13
(b) What project duration will have 95 per cent confidence of M J, K, L 6 7 8
completion?
(a) Draw the network diagram for the project.
3. The owner of a chain of fast food restaurants is considering a
(b) Compute the expected project completion time.
new computer system for accounting and inventory control. A
(c) What should be the due date to have 0.90 probability of
computer company sent the following information about the
completion?
computer system installation:
(d) Find the total float and free float for all non-critical activities.
Activity Activity Immediate Times (days) [Agra, Univ., MCA, 2000]
Description Predecessor 6. A sociologist plans a questionnaire survey, consisting of the
Optimistic Most Pessim- following tasks:
Likely istic
A Select the computer model – 4 6 8 Activity Description Immediate Duration (days)
B Design input/output system A 5 7 15 Predecessor Likely Mini- Maxi-
C Design monitoring systems A 4 8 12
mum mum
D Assemble computer
hardware B 15 20 25 A Design of questionnaire – 5 4 6
E Develop the main B Sampling design – 12 8 16
programmes B 10 18 26 C Testing of questionnaire
F Develop input/output and refinements A 5 4 12
routines C 8 9 16 D Recruiting for interviewers B 3 1 5
G Create database E 4 8 12
E Training of interviewers D, A 2 2 2
H Install the system D, F 1 2 3
F Allocation of areas to
I Test and implement G, H 6 7 8
interviewers B 5 4 6
(a)
Construct the network diagram for the project. G Conducting interviews C, E, F 14 10 18
(b)
Determine the critical path and compute the expected H Evaluation of results G 20 18 34
completion time.
(c) Determine the probability of completing the project in 55 days. (a) For this PERT network find the expected task durations and
4. A company manufacturing plant and equipment for chemical the variances of task durations.
processing is in the process of quoting a tender called by apublic (b) Draw a network for this project and find the critical path.
sector undertaking. The delivery data, once promised, is crucial What is the expected length of the critical path? What is the
and a penalty clause is applicable. The project manager has listed variance of the length of the critical path?
down the activities in the project as under: (c) What is the probability that the length of the critical path
does not exceed 60 days? [Delhi Univ., MBA, 2002]
Activity Immediate Activity Time (weeks)
Predecessor 7. A management student identifies the following list of activities
Optimistic Most Likely Pessimistic and sequencing requirements along with the time estimates for
A – 1 3 5 various activities related to the completion of his project work.
B – 2 4 6
C A 3 5 7 Activity Description Immediate Activity Time (days)
D A 5 6 7 Predec-
Optimistic Likely Pessi-
E C 5 7 9 essor
mistic
F D 6 8 10
G B 7 9 11 A Search of list – 3 6 9
H E, F, G 2 3 4 B Dedicating the project – 2 4 12
C Preliminary work B 1 1.5 5
(a) Find out the delivery week from the date of commencement D Formal proposal C 1 2 3
of the project, and E Project committee’s
(b) Find the total float and free float for each of the non-critical approval A, D 1.5 2 4.5
activities. F Progress report E 0.5 1 1.5
G Format research A, D 4.5 5 11.5
5. Consider a project having the following activities and their time H Data collection E 2 5 8
estimates: I Analysis G, H 4 5.5 10
J Conclusion I 1.5 2.5 4.5
 Project Management: PERT and CPM 443

K Draft I, F 2 3.5 8 J System debugging and

L Final draft J, K 2.5 3 1.5 installation F, G 2 2 2
M Presentation L 0.5 1 1.5 K Equipment changeover G, I 5 8 11
With the help of a network diagram, determine the minimum time
(a) Draw the network diagram for this project and find the
required to complete the project. From the network, identify the
critical path and its length.
activities that can be delayed without affecting the project duration
(b) Calculate the total and free floats for non-critical activities.
and the extent of delay that is possible for such activities.
[Delhi Univ., MBA, 2004] (c) What is the probability that the length of the critical path
does not exceed 40 days? [Delhi Univ., MBA, 2003]
8. A publisher is preparing to produce the second edition of a
textbook. The activities required and their estimated time are as 10. A promoter is organizing a sports meeting. The relationship among
follows: the activities and time estimates in days are shown below in the
table:
Activity Description Immediate Activity Time (days)
Predec- Optimistic Likely Pessi- Activity Description Predecessor Activity Time (days)
essor mistic
Optimistic Likely Pessi-
A Assess market – 1 3 2 mistic
B Get reviews from users A 1 2 1.5
A Prepare draft programme – 3 7 11
C Revamp old material
and add new material B 3 9 5 B Send to sports
D Obtain review and organizations and wait
prepare final draft C 4 12 6 for comments A 14 21 28
E Revise and expand C Obtain promoters A 11 14 17
problems B 2 7 4
D Prepare and sign
F Copy edit final draft D 1 2.5 1.5
documents for stadium
G Copy edit problems E 0.5 1.5 1
hire A, C 2 2 2
H Set type, proof and
print book F, G 5 9 6 E Redraft programme and
I Prepare instructor’s request entries B 2 3.5 8
manual E 2 4 3 F Enlist officials D, E 10 14 21
J Produce instructor’s G Arrange accommodation
manual I 1 2 1.5 for touring teams E 3 4 5
K Complete book and
H Prepare detailed
instructor’s manual H, J 0 0 0
programme E, F 4 4.5 8
(a) Draw a PERT network diagram and determine the critical I Make last-minute
path. arrangements G, H 1 2 4
(b) What is the probability that the project will be completed within
(i) 21 months? (ii) 27 months? (a) Draw the network diagram for the project and compute the
9. A nationalized bank wishes to plan and schedule the develop- ment expected completion time of the project.
and installation of a new computerized cheque processing system. (b) What should be the due date to have 0.90 probability of
The changeover in cheque-processing procedures requires project completion?
employment of additional personnel to operate the new system, (c) Find the total float and free float for all non-critical activities.
development of new systems (computer software), and (d) What is the probability that the length of the critical path
modification of existing cheque sorting equipment. The activities does not exceed 56 days? [Delhi Univ., MBA, 2005]
required to complete the project, along with three time estimates 11. A management student identifies the following list of activities
and the precedence relationship among the activities, have been and sequencing requirements along with the time estimates for
determined by the bank management and are given in the following various activities related to the completion of his project:
table:
Activity Description Immediate Activity Time (days)
Activity Description Predecessor Time Estimates Predec Optimistic Likely Pessi-
(days) essor mistic
Optimistic Most Pessi- A Search of Literature – 3 6 9
Likely mistic B Deciding the project – 2 4 12
C Preliminary work B 1 1.5 5
D Formal proposal C 1 2 3
A Position recruiting – 5 8 17
E Project committee’s
B System development – 3 12 15 approval A, D 1.5 2 4.5
C System training A 4 7 10 F Progress report E 0.5 1 1.5
D Equipment training A 5 8 23 G Formal research A, D 4.5 5 11.5
E Manual system test B, C 1 1 1 H Data collection E 2 5 8
I Analysis G, H 4 5.5 10
F Preliminary system
J Conclusion I 1.5 2.5 4.5
changeover B, C 1 4 13
K Draft I, F 2 3.4 8
G Computer-personnel L Final Draft J, K 2.5 3 1.4
interface D, E 3 6 9 M Presentation L 0.5 1 1.5
H Equipment modification D, E 1 2.5 7
I Equipment testing H 1 1 1
444 Operations Research: Theory and Applications

With the help of an arrow diagram, determine the minimum time L Pricing decision J, F 1 1.5 3
required to complete the project. From the network, identify the M Marketing mix finalization L 2 2.5 3
activities that can be delayed without affecting the project duration N POS development M 2 4 5
and the extent of delay that is possible for such activities. O Launch plans M 1 1 1.5
[Delhi Univ., MBA, 2001] P Branch communication O 0.5 0.5 1
Q Supplier’s delivery of
12. A computer software company has broken down the processof
packaging H 4 5 8
integrating a computer system into its operation into several steps.
R Production trail D, Q 1.5 2 3
Some of the steps cannot begin until the others are completed.
S Management final
These relationships are shown in the accompanying table. In
approval R 0.5 1 1.5
addition, estimates of the most likely, optimistic, and pessimistic
T Final production S 1 2.5 3
times required for each are listed below:
U Stock movement T 0.5 1.5 2
Activity Immediate Expected Time (weeks) V Position movement T 0.5 1 1.5
W Launch U, V 1 2 4
Optimistic Most Likely Pessimistic
The management of the company desires to know the realistic
A – 2 3 4
completion time for this project and detailed analysis of float times
B – 3 4 11 (if any). [Delhi Univ., MBA, 2000, 2004]
C A 2 5 8 14. Following is the list of various activities involved in the launch
D A 1.5 3.5 8.5 of a new Credit Card service by a company, their immediate
E B 5 7 9 predecessors and their expected durations (in days):
F B 2 5.5 6
Activity Description Immediate Activity Time (days)
G C 1.5 2.5 6.5
Predec- Optimistic Likely Pessi-
H C, D 3 4 11 essor mistic
I G 4 6 8 A Conduct market research to
J H, E 3 4.5 9 determine size, potential and
competition in the market – 10 12 14
K F 5 6 7 B Define marketing strategy in
L I, J, K 1 3 11 terms of positioning and
product features A 14 15 17
(a) Draw a PERT chart for this project, and calculate the critical C Estimate expected volumes B 2 3 4
path. D Estimate additional manpower
required C 4 6 8
(b) By how much time can activity F be delayed without delaying E Identify modifications required
the project as a whole? to the present computer system
(c) If labour costs Rs 1,500 per week find the probability that based on expected volumes C 10 12 14
the labour costs for this project will exceed Rs 36,000. F Implement computer system
(d) Company wishes to budget an amount for labour costs that modifications E 20 25 27
G Update physical facilities C 10 17 20
will be sufficient with 95 per cent probability. How much H Prepare instruction manuals
should they budget? for modified system F 5 6 7
13. A multinational FMCG company wishes to launch a new Fruit I Hire additional sales force
and operations staff D 7 12 14
Yogurt in the coming season. A brief description of the activities J Train sales force and
associated with this project, their expected durations (in weeks) operations staff H, I 14 17 20
and their immediate predecessor(s) are given in the following table: K Advertise for suppliers C 1 2 3
L Check supplier samples K 10 15 20
Activity Description Predecessor Expected Time M Identify key suppliers and
(weeks) define quality standards L 3 5 7
N Roll-out basic product to a few
Optimistic Likely Pessi- employees as a method of
mistic pre-testing M, J 13 15 17
O Generate and assess
A Management approval — 2 2.5 4 feedback N 20 21 22
B Product concept test A 3 4.7 5 P Modify and make changes
C Technical feasibility A 2 2 3 wherever necessary O 7 9 14
Q Update procedures manual P 2 3 4
D Recipe finalization C, B 1 1 2 R Retain commercial staff Q 2 2 2
E Shelf life trials D 8 12 15 S Design communications to
F Brand positioning study B 4 5 7 prospective card-holders P 7 10 13
G Packaging key lines F 1 1 2.5 T Design advertising schedule S 5 7 9
H Agency advertisement U Roll out final product-Launch T, R, G 4 8 12
development F, E 4 8 10
I Agency: layouts artworks G 2 3 5 (a) Draw the network diagram for the project.
J Advertisement test (b) Find the expected project completion time of the project.
research H 3 4 5 (c) Determine the probability of completing the project in 165
K Cost finalization D 1 1 2 days. [Delhi Univ., MBA, 2000]
 Project Management: PERT and CPM 445

HINTS AND ANSWERS

1. (i) Critical path: A – C – E – H. Expected project duration is (ii) Project duration which has 95 per cent probability of
37 weeks with its variance 25 weeks. completion is

Ts  Te 36  37
(ii) Z     0.2 ; P(Z  – 0.2) = 0.5 – 0.0793 Z (= 1.64) = Ts  27

 5 18
= 0.4207. or Ts = 1.64 × 18 + 27 = 34 days (approx)
Ts  37
(iv) Z(= 2.33) = or Ts = 37 + 2.33 × 5 = 48.65 weeks. 3. (i) Critical path: A – B – E – G – I with project duration of
5 47 days and its variance is 110/9.

2. (i) Critical path A – B – D – F. Expected project duration is (ii) Probability of completing the project in 55 days is
27 days. Next critical path: A – B – D – G with duration  Ts  Te 55  47 
of 25 days. P Z    = P(Z  2.29) = 0.9634
  3.49 

13.7 PROJECT TIME-COST TRADE-OFF

While discussing the scheduling of project activities in previous section, the cost of resources consumed
by activities were not taken into consideration. The project completion time can be reduced by reducing
(crashing) the normal completion time of critical activities. The crashing of normal completion time of critical
activities will increase the total cost of the project. But, the decision-maker will always look for trade-off Crashing the
between the total cost of project and the total time required to complete it. project means
crashing a number
13.7.1 Project Crashing of activities to
reduce the duration
Crashing of completion time of critical activities to reduce the project completion time would certainly of the project, below
its normal time.
require extra resources (cost). However, as shown in Fig. 13.16, beyond point A project cost increases when
time is reduced. Similarly, beyond point B, the time increases while the project cost decreases. Obviously,
an activity time may not be reduced beyond a certain limit and this limit is referred as crash point. Also,
extending an activity time beyond normal point (cost efficient) may increase cost of executing that activity.
For simplicity, the relationship between normal time and cost as well as crash time and cost for an
activity is assumed to be linear. Thus, the crash cost per unit of time can be estimated by computing the
relative change in cost (cost slope) per unit change in time.
As shown in Fig. 13.16, a decision-maker must be interested in the central region of the curve between
points A and B. This will help in developing a trade-off between time and direct cost of completing an activity.

Fig. 13.16
Time cost Trade-
off

13.7.2 Time-Cost Trade-Off Procedure

The starting point for crashing is when all critical activities are completed with their normal time, and
crashing is stopped when all critical activities are crashed.
446 Operations Research: Theory and Applications

The method of establishing time-cost trade-off for the completion of a project can be summarized as
follows:
Step 1: Determine the normal project completion time and associated critical path.
Step 2: Identify critical activities and compute the cost slope for each of these by using the relationship
Crash cost– Normal cost
Cost slope =
Normal time – Crash time
The values of cost slope for critical activities indicate the direct extra cost required to execute an activity
per unit of time.
Crashing an Step 3: For reducing the total project completion time, identify and crash an activity time on the critical path
activity means with lowest cost slope value to the point where
taking special costly
measures to reduce (i) another path in the network becomes critical, or
the duration of an (ii) the activity has been crash to its lowest possible time.
activity below its
normal time. Step 4: If the critical path under crashing is still critical, return to Step 3. However, if due to crashingof
an activity time in Step 3, other path(s) in the network also become critical, then identify and crash the
activity(s) on the critical path(s) with the minimum joint cost slope.
Step 5: Terminate the procedure when each critical activity has been crashed to its lowest possible time.
Determine total project cost (indirect cost plus direct cost) corresponding to different project durations.
Example 13.8 The data on normal time, and cost and crash time and cost associated with a project are
shown in the following table.

Activity Normal Crash

Time(weeks) Cost (Rs) Time (weeks) Cost (Rs)
1 – 2 3 300 2 400
2 – 3 3 30 3 30
2 – 4 7 420 5 580
2 – 5 9 720 7 810
3 – 5 5 250 4 300
4 – 5 0 0 0 0
5 – 6 6 320 4 410
6 – 7 4 400 3 470
6 – 8 13 780 10 900
7 – 8 10 1,000 9 1,200
4,220

Indirect cost is Rs 50 per week.

(a) Draw the network diagram for the project and identify the critical path.
(b) What are the normal project duration and associated cost?
(c) Find out the total float associated with non-critical activities.
(d) Crash the relevant activities and determine the optimal project completion time and cost.
Solution (a) The network for normal activity times is shown in Fig. 13.17. The critical path is: 1 – 2 –5
– 6 – 7 – 8 with a project completion time of 32 weeks.

Fig. 13.17
Network Diagram
 Project Management: PERT and CPM 447

(b) The normal total project cost associated with normal project duration of 32 weeks is as follows:
Total cost = Direct normal cost + Indirect cost for 32 weeks
= 4,220 + 50 × 32 = Rs 5,820
(c) Calculations for total float associated with non-critical activities are shown in Table 13.6.

Activity Total Float

(Lj – Ei) – tij

2 – 3 (7 – 3) – 3 = 1
2 – 4 (12 – 3) – 7 = 2
3 – 5 (12 – 6) – 5 = 1
4 – 5 (12 – 10) – 0 = 2
6 – 8 (32 – 18) – 13 = 1 Table 13.6
Total Float
(d) For critical activities, crash cost-slope is given in Table 13.7.

Critical Activity Crash Cost per Week (Rs)

400  300
1 – 2  100
32
810  720
2 – 5  45
97
410  320
5 – 6  45
64
470  400
6 – 7  70
43
1200  1000 Table 13.7
7 – 8  200
10  9 Crash Cost Slope

The minimum value of crash cost per week is for activity 2 – 5 and 5 – 6. Hence, crashing activity2
– 5 by 2 days from 9 weeks to 7 weeks. But the time should only be reduced by 1 week otherwise
another path 1 – 2 – 3 – 5 – 6 – 7 – 8 become a parallel path. Network, as shown in Fig. 13.18, is developed
when it is observed that new project time is 31 weeks and the critical paths are 1 – 2 – 5 – 6 – 7 – 8 and
1 – 2 – 3 – 5 – 6 – 7 – 8.
With crashing of activity 2 – 5, the crashed total project cost becomes:
Crashed total cost = Total direct normal cost + Increased direct cost due to crashing of
activity (2 – 5) + Indirect cost for 31 weeks
= 4,220 + 1 × 45 + 50 × 31 = 4,265 + 1,550 = Rs 5,815
For revised network shown in Fig. 13.18, new possibilities for crashing critical activities are listed in
Table 13.8.

Fig. 13.18
Network Diagram
448 Operations Research: Theory and Applications

Critical Activity Crashed Cost per Week (Rs)

400  300
1 – 2  100
32
2 – 5 × (Crashed)
2 – 3 0 (Crashing is not required)
300  250
3 – 5  50
54
410  320
5 – 6  45
64

470  400
6 – 7  70
43
Table 13.8 1200  1000
7 – 8  200
Crash Cost Slope 10  9

Since crashed cost slope for activity 5 – 6 is minimum, its time may be crashed by 2 weeks from 6 weeks
to 4 weeks. The updated network diagram is shown in Fig. 13.19.

Fig. 13.19
Network Diagram

It may be noted in Fig. 13.19, that the critical paths shown in Fig. 13.18 remain unchanged because
activity 5 – 6 is common in both. With crashing of activity 5 – 6 by 2 weeks, the crashed total cost becomes:
Crashed total cost = Total direct normal cost + Increased direct cost due to crashing of 5 – 6 +
Indirect cost for 29 weeks
= 4,220 + (1 × 45 + 2 × 45) + 50 × 29 = Rs 5,805
For revised network given in Fig 13.19, new possibilities for crashing in the critical paths are listed in
Table 13.9.

Critical Activity Crashed Cost per Week (Rs)

400  300
1 – 2  100
32
2 – 3 0 (Crashing is not required)
2 – 5 × (crashed)
5 – 6 × (crashed)
470  400
6 – 7  70
43
Table 13.9 1200  1000
7 – 8 = 200
Crash Cost Slope 10  9

The further crashing of 6 – 7 activity time from 4 weeks to 3 weeks will result in increased direct cost
than the gain due to reduction in project time. Hence, terminate crashing. The optimal project duration is
29 weeks with associated cost of Rs 5,805 as shown in Table 13.10.
 Project Management: PERT and CPM 449

Project Crashing Direct Cost (Rs) Indirect Total Cost

Duration Activity Cost (Rs) (Rs)
Normal Crashing Total
(weeks) and Weeks

32 – 4,220 – 4,220 32 × 50 = 1,600 5,820 Table 13.10

31 2 – 5(1) 4,220 1 × 45 = 45 4,265 31 × 50 = 1,550 5,815 Crashing
29 5 – 6(2) 4,220 45 + 2 × 45 = 135 4,355 29 × 50 = 1,450 5,805 Schedule of
28 6 – 7(1) 4,220 135 + 1 × 70 = 205 4,425 28 × 50 = 1,400 5,825 Project

Example 13.9 The data on normal time and cost along with crashed time and cost associated with a project
are shown in the following table:
Acltivity Immediate Normal Crash
Predecessor Time (weeks) Cost (Rs ’000) Time (weeks) Cost (Rs ’000)

A – 10 20 7 30
B – 8 15 6 20
C B 5 10 4 14
D B 6 11 4 15
E B 8 9 5 15
F E 5 5 4 8
G A, D, C 12 3 8 4
71,000

The indirect cost is Rs. 400 per day. Find the optimum duration and the associated minimum project cost.
Solution The network diagram with the normal time of the project activities is shown in Fig. 13.20. The
critical path in the network diagram is shown by thick lines. The critical path is: 1 – 2 – 3 – 4 – 6 with normal
project duration of 26 weeks.
Project cost associated with project duration of 26 weeks is Rs 1,43,800:

Fig. 13.20
Network Diagram

Total cost = Direct normal cost + Indirect cost for 26 weeks

= 71,000 + 26 × 7 × 400 = Rs 1,43,800
To begin crash analysis, the crash cost slope values for critical activities is shown in Table 13.11.

Critical Activity Crash Cost per Week (Rs)

20  15
B (1 – 2)  2.5
86
15  11
 2
D (2 – 3) 64
43 Table 13.11
G (4 – 6)  0.25 Crash Cost Slope
12  8
450 Operations Research: Theory and Applications

Fig. 13.21
Network Diagram
The critical activity G with cost slope of Rs (0.25 × 1,000) = Rs 250 per week is lowest and can be
crashed by 4 weeks (being last activity of critical path). The revised network is shown in Fig.
13.21. The new critical path is still 1 – 2 – 3 – 4 – 6 with total project duration of 22 weeks.
With crashing of activity G (4 – 6), the total project cost becomes:
Total crashed cost = Direct normal cost + Increased cost due to crashing of G + Indirect cost for
22 weeks
= 71,000 + 0.25 × 1000 × 4 + 22 × 7 × 400
= 71,000 + 1000 + 61,600 = Rs. 1,33,600
In Fig. 13.21, the path 1 – 2 – 3 – 4 – 6 is still the critical path. As shown in Table 13.11, the next least
expensive activity on the critical path is activity D (2 – 3). It can be crashed by one week. The crashed
network is shown in Fig. 13.22 with total project duration of 21 weeks.

Fig. 13.22
Network Diagram

With the crashing of activity D, the total project duration is reduced to 21 weeks with the critical paths
each with project duration of 21 weeks are shown in Fig. 13.22.
(i) 1 – 2 – 5 – 6
(ii) 1 – 2 – 3 – 4 – 6
(iii) 1 – 2 – 4 – 6
The new total project cost involved with network shown in Fig. 13.22 becomes:
Total crashed cost = Direct normal cost + Increased direct cost due to crashing of D + Indirect cost
for 21 weeks
= 71,000 + (1000 + 2 × 1000 × 1) + 21 × 7 × 400
= 71,000 + 1000 + 2000 + 58,800 = Rs 1,32,800
 Project Management: PERT and CPM 451

Since activity B is common in all three critical paths, as shown in Fig. 13.22, therefore it can be crashed
to a maximum of 2 weeks. The new total project cost with project duration of 19 weeks becomes:
Total crashed cost = Direct normal cost + Increased direct cost due to crashing of B, D and E +
Indirect cost for 19 weeks
= 71,000 + (1,000 + 2,000 + 2.5 × 1,000 × 2) + 19 × 7 × 400
= 73,000 + 5,000 + 53,200 = Rs 1,31,200

Fig. 13.23
Network Diagram

With the crashing of activity B by 2 weeks, the project duration is reduced to 19 weeks with the
following critical paths each with project duration of 19 weeks as shown in Fig. 13.23.
(i) 1 – 2 – 5 – 6
(ii) 1 – 2 – 3 – 4 – 6
(iii) 1 – 2 – 4 – 6

Fig. 13.24
Network Diagram

For revised network shown in Fig. 13.24, the crashing possibilities on the critical path are listed in Table 13.12.
Crashed Creashed Cost
Activity per Week (Rs)

14  10
2 – 4 4
54
15  11
2 – 3 2
64

2 – 5 15  9 Table 13.12
2 Crash Cost slope
85
452 Operations Research: Theory and Applications

Since critical activities B and G cannot be crashed further, therefore crashing activity C in path1
– 2 – 4 – 6, activity D in path 1 – 2 – 3 – 4 – 6 and activity E in path 1 – 2 – 5 – 6 each by 1 weekin
order to reduce the project duration to 18 weeks which is equivalent to the project time of non-critical path 1
– 4 – 6. The reduced project duration along with critical paths is shown in Fig. 13.24.
The total project cost with project duration of 18 weeks becomes:
Total crashed cost = Direct normal cost + Increased direct cost due to crashing of activities
C, D and E + Indirect cost for 18 weeks
= 71,000 + (1,000 + 2,000 + 5,000) + (4 × 1,000 × 1 + 2 × 1,000 × 1 + 2
× 1,000 × 1) + 7 × 400 × 18 = 1,37,400
Since total project cost for 18 weeks is more than the cost for 19 weeks, therefore, further crashing
is not desirable. Hence, the optimum pair of time and cost associated with the project is 19 weeks and
Rs 1,32,200, as shown in Table 13.13.
Project Crashing Direct Cost (Rs) Indirect Total Cost
Duration Activity Cost (Rs) (Rs)
Normal Crashing Total
(weeks) and Weeks

26 – 71,000 – 71,000 26 × 2,800 = 72,800 1,43,800

22 4 – 6(4) 71,000 4 × 250 = 1,000 72,000 22 × 2,800 = 61,600 1,33,600

Table 13.13 21 2 – 3(1) 71,000 1,000 + 2,000 = 3,000 74,000 21 × 2,800 = 58,800 1,32,800
Crashing 19 1 – 2(2) 71,000 3,000 + 5,000 = 8,000 79,000 19 × 2,800 = 53,200 1,32,200
Schedule of the 18 2 – 4(1)
Project 2 – 3(1), 2 – 5(1) 71,000 8,000 + 8,000 = 16,000 87,000 18 × 2,800 = 50,400 1,37,400

Example 13.10 The following table gives the activities in a construction project and also gives other
relevant information:

Activity Immediate Time (months) Direct Cost (Rs ’000)

Predecessor Normal Crash Normal Crash
A – 4 3 60 90
B – 6 4 150 250
C – 2 1 38 60
D A 5 3 150 250
E C 2 2 100 100
F A 7 5 115 175
G D, B, E 4 2 100 240
713

Indirect costs vary as follows:

Months : 15 14 13 12 11 10 9 8 7 6
Cost (Rs) : 600 500 400 250 175 100 75 50 35 25
(a) Draw an arrow diagram for the project.
(b) Determine the project duration that will result in minimum total project cost. [Osmania Unit., MBA, 2002]
Solution The network for normal activity times indicates a project duration of 13 months with critical path:
A – D – G (1 – 2 – 4 – 5) as shown in Fig. 13.25. The crash cost slope for various critical activitiesof the
project is given in Table 13.14.

Critical Crash cost/month

Activity (Rs)
90  60
A(1 – 2)  30
43
250  150
D(2 – 4)  50
53
Table 13.14 240  100
G(4 – 5) = 70
Crash Cost Slope 42
 Project Management: PERT and CPM 453

Among the critical activities A, D and G, the least expensive activity is A. This means that the duration of
this activity should be crashed by maximum of one month. Thus, project duration reduces to 12 months
and new total project cost is as follows:
Total crashed cost = Total direct cost + Indirect cost due to crashing of activity A by one week
+ Indirect cost for 12 months
= 713 + 1 × 30 + 250 = Rs 993

Fig. 13.25
Network Diagram

Since existing critical path remains same even after crashing critical activity A by one month, therefore
next lowest cost slope activity D is crashed by 2 months. Thus, project duration reduces to 10 months as
shown in Fig. 13.26.

Fig. 13.26
Network Diagram

In Fig. 13.26, there are three critical paths: 1 – 2 – 5; 1 – 2 – 4 – 5 and 1 – 4 – 5, each having a duration
of 10 months. The new total project cost becomes:
Total crashed cost = Total direct cost + Indirect cost due to crashing of activity D by 2 months
+ Indirect cost for 10 months
= 713 + (30 + 2 × 50) + 100 = Rs 943
454 Operations Research: Theory and Applications

Fig. 13.27
Network Diagram

Since the critical activities A and D cannot be further crashed, therefore activities F and G may be crashed
by 2 months each. The revised project network is shown in Fig. 13.27 with total project durationof 8
months. The new total project cost then becomes:
Total crashed cost = Total direct cost + Increased direct cost due to crashing of activities F and G
+ Indirect cost for 8 months
= 713 + (30 + 100 + 2 × 30 + 2 × 70) + 50 = Rs 1093
Since the total project cost for 8 months is more than the cost for 10 months, therefore, further crashing
is not desirable. Hence, optimum pair of time and cost associated with the project is 10 months and Rs 993,
as shown in Table 13.15.

Project Crashing Direct Cost (Rs ’000) Indirect Total Cost

Duration Activity with Cost (Rs ’000) (Rs ’000)
(months) Months Normal Crashing Total

13 – 713 – 713 400 1,113

Table 13.15 12 1 – 2(1) 713 30 743 250 993
Crashing 10 2 – 4(2) 713 30 + 2 × 50 = 130 843 100 943
Schedule of the 8 2 – 5(2) 713 130 + 2 × 30 1,043 50 1,093
Project 4 – 5(2) + 2 × 70 = 330

SELF PRACTICE PROBLEMS D

1. The following table gives the activities in a construction

project and other relevant information:

Activity Description Immediate Normal Crash Cost Rank

Predecessor Time Cost Time Cost Slope
A Dismantle and
inspect – 2 100 1 140 40 II
B Sub-contract repairs A 6 200 2 500 75 IV
C Buy – spares A 3 100 1 140 20 I
D Rebuilt/repair A 4 500 2 800 150 V
E Assemble B, C, D 2 200 1 250 50 III
F Trial run E 1 50 – 250 – –

Fixed-cost data for the different project durations are given 7 700
below: 6 600
5 500
Duration (in days) Cost in (Rs) 4 400
11 1,100 (a) Draw the network diagram for the project.
10 1,000 (b) Determine the project duration which will return in minimum
9 900 total project cost.
8 800
 Project Management: PERT and CPM 455

2. The required data for a small project consisting of different a bonus being given for early delivery. The marketing depart-
activities are given below: ment has established the following cost and time information:

Activity Predecessor Normal Crash

Normal Time (weeks) Normal Crash Crash
Activities Duration Cost Duration Cost Activity Cost Time Cost
(days) (Rs) (days) (Rs) Optim- Pessim- Most
istic istic Likely (Rs) (weeks) (Rs)
A – 6 300 5 400
1 – 2 1 5 3 15,000 1 19,000
B – 8 400 6 600
2 – 3 1 7 4 18,000 3 24,000
C A 7 400 5 600
2 – 4 1 5 3 14,000 2 16,000
D B 12 1,000 4 1,400
2 – 5 5 11 8 15,000 7 16,000
E C 8 800 8 800
3 – 6 2 6 4 13,000 2 15,000
F B 7 400 6 500
4 – 6 5 7 6 12,000 4 13,000
G D, E 5 1,000 3 1,400
5 – 7 4 6 5 20,000 4 24,000
H F 8 500 5 700
6 – 7 1 5 3 17,000 1 20,000
(a) Draw the network diagram for the project and find the normal
and minimum project length. The normal delivery time is 16 weeks for a contract price of
(b) If the project is to be completed in 21 days with minimum crash Rs 1,24,000. Based on the probability for each of the following
cost which activities should be crashed to how many days? specified delivery time, recommend the delivery schedule thatthe
Patel Machinery Co. should follow:
3. The time and cost estimates and precedence relationship of the
different activities constituting a project are given below: Contract Delivery Time Contract Amount
(weeks) (Rs)
Activity Predecessor Time (in weeks) Cost (in Rs)
Activities Normal Crash Normal Crash 15 1,42,500
14 1,45,000
A – 3 2 8,000 19,000
13 1,50,000
B – 8 6 600 1,000
C B 6 4 10,000 12,000 12 1,52,500
D B 5 2 4,000 10,000
6. The President of ABC Manufacturing Company has an oppor-
E A 13 10 3,000 9,000
tunity to participate in a project that has a sales price of Rs
F A 4 4 15,000 15,000
90,000 but the project must be completed within 8 weeks. This
G F 2 1 1,200 1,400
letter of intent was received on Friday afternoon. Both the
H C, E, G 6 4 3,500 4,500
superintendent of production and the cost accountant came in
I F 2 1 7,000 8,000
on Saturday and completed the appropriate time and cost for
(a) Draw a project network diagram and find the critical path. you based upon past jobs. Since the president needs an answer
at 8.30 a.m. on Monday (start of the 8th week), you, have been
(b) If a dead line of 17 weeks is imposed for completion of the requested to determine the profitability of the project on an 8
project, what activities will be crashed, what would be the week basis. An answer at 8.30 a.m. Monday allows the firm to start
additional cost and what would be critical activities of the the production order at 10.00 a.m. in order to stay within the 8
crashed network after crashing? weeks demanded by the customer. The time and cost under
4. The activities of a project are tabulated below with immediate normal conditions without crashing the project is based upon an 11
predecessors and normal and crash time cost. week basis. What answer should the president give the customer
on Monday morning. A table of times and cost is given below:
Activity Immediate Normal Crash
Predecessor Cost Time Cost Time Event Preceding Normal Crash
(Rs) (days) (Rs) (days) Event Time Cost Time Cost
A – 200 3 400 2 (weeks) (Rs) (weeks) (Rs)
B – 250 8 700 5 4 1 2 8,000 1 13,000
C – 320 5 380 4 2 1 2 7,000 1 19,000
D A 410 0 800 4 3 1 6 11,000 5 13,000
E C 600 2 670 1 4 2 4 6,000 3 10,000
F B, E 400 6 950 1 3 2 2 9,000 1 10,000
H B, E 550 12 1,000 6 5 2 7 8,500 6 11,000
G D 300 11 400 9 5 4 4 10,000 3 16,000
5 3 3 5,000 2 7,000
(a) Draw the project network diagram using normal time of
activities. 7. The time-cost estimates for the various activities of a projectare
given below:
(b) Determine the critical path and the normal duration and cost
of the project.
Activity Preceding Time (weeks) Cost (Rs)
(c) Suitably crash the activities so that the normal duration may Activities
be reduced by 3 days at minimum cost. Also find the project Normal Crash Normal Crash
cost for this shortened duration if the indirect costper day A – 8 6 8,000 10,000
is Rs 25. B – 7 5 6,000 8,400
5. Patel Machinery Co. has been offered a contract to build and C A 5 4 7,000 8,500
deliver nine extruding presses to the ABC Bottling Co. The contract
price is contingent on meeting a specified delivery time, (Conted..)
456 Operations Research: Theory and Applications

500 G Prepare bottlemaker 6 3 E

Activity Preceding Time (weeks) Cost (Rs) H Run test production 4 4 F, G
Activities Normal Crash Normal Crash 500 I Examine test quantity
and make adjustments 4 2 H
D B 4 3 3,000 3,800 J Refill oven with glass 2 2 H
E A 3 2 2,000 2,600
F D, E 5 3 5,000 6,600 10. A company has recently won a contract for the installation of a
G C 4 3 6,000 7,000 die casting machine and its associated building construction work
at a local factory of a large national firm of electronic engineers.
The following table gives the various activities in- volved in this job,
The project manager wishes to complete the project in the their normal time and cost estimates and their crash time and their
minimum possible time. However, he is not authorized to spend cost estimates.
more than Rs 5,000 on crashing.
Suggest the least-cost schedule for achieving the objective of the Activity Description Predec Normal Crash
project manager. Assume that there is no indirect or utility cost. essor
Time Cost Time Cost
[Delhi Univ., MCom, 2000] (days) (Rs) (days) (Rs)
8. An electronics firm has signed a contract to install an instrument A Prepare foundations
landing device at the local airport. The complete installation can and underground services
be broken down into fourteen separate activities. Each activity and erect building frame
(labelled A through N), its predecessor activities, normal timeand structure – 30 90,000 25 1,05,000
cost and crash time and cost are given in the following table. The B Fabricate part and asse-
contract specifies that the installation will be completed within 18 mble steel frames to sup-
days. There is a penalty of Rs 100 per day beyondthe specified port the machine – 25 1,80,000 20 1,90,000
completion time. C Collect die casting machine
and its associated gear
Activity Preceding Normal Normal Crash Crash from the manufacturers – 10 50,000 8 54,000
Time Cost Time Cost D Assemble and check
(days) (Rs) (days) (Rs) control gear C 10 7,500 7 9,000
E Fit control gear on steel
A – 3 320 2 360 frames and instal B, D 10 4,200 10 4,200
B – 5 550 4 600 F Fit aluminium sheet wall
C – 6 575 4 700 claddings A, E 20 20,000 16 30,000
D A 7 750 5 850 G Erect assembled plant on
to prepared foundation
E A 4 420 3 490
and framework and
F B, D 2 180 2 280 connect services A, E 35 28,000 30 35,500
G C 4 425 3 485 H Erect mechanical
H A 8 850 5 900 handling plant B, D 20 12,000 18 15,000
I C 5 475 4 535 I Fit ventilation and fire
J C 7 675 5 735 protection system F 20 14,000 15 24,000
K E, F, G 4 400 3 440 If the variable overhead costs are Rs 5,000 per day, determine
L H, I 6 650 4 750 the optimal project duration. [Delhi Univ., MBA, Dec. 1995]
M L 3 280 2 335 11. A small marketing project consists of the jobs given in the table
N J, K 5 525 4 575 given below. Next to each job is listed its normal time and a
minimum or crash time (in days). The cost (in Rs per day) of
The project will require the positioning of one engineer till the crashing each job is also given.
completion of work. The monthly cost of each engineer for such job
is Rs 10,000. This will be considered as indirect cost.
Job Normal Duration Crash Cost of Crashing
9. American Bottle Company (ABC) produces several types of glass (days) Duration (days) (Rs per Day)
containers. They have recently reduced capacity atseveral of
their plants. The manufacturing of these grass containers involves 1 – 2 9 6 20
large, expensive machines (including ov- ens), several of which 1 – 3 8 5 25
were shut down during capacity reduc- tion. The machines are 1 – 4 15 10 30
hard to shut down and to start up. Inthe event of a surge in 2 – 4 5 3 10
demand, they want to know how quickly they could start one back 3 – 4 10 6 15
again. How quickly can they start a new oven using normal 4 – 5 2 1 40
times? What is the fastest time in which a new oven can be
started, and how much additional cost would this involve? (a) What is the normal project length and the minimum project
length?
Cost (Rs) Activity Normal Crash Predecessor (b) Determine the minimum crashing costs of schedules rang- ing
per Unit Time Time from normal length, down to, and including, the minimum
Time (hour) length schedule, i.e. if L is the length of the normal schedule,
Reduction find the costs of schedules which are L, L – 1, L – 2 days
long and so on.
– A Preheat glass 8 8 C
– B Preheat oven 12 12 D The overhead cost for the project is Rs 60 per day. What is
400 C Obtain materials 4 2 – the optimal length schedule duration of each job for your solution?
200 D Check valves 4 2 – 12. The following table gives the list of various activities involved in
200 E Check pressure seals 2 1 B the production of a wireless communication equipment, their
– F Add glass to oven 2 2 A, E immediate predecessor(s), their normal time and cost estimates
and their crash time and cost estimates:
 Project Management: PERT and CPM 457

C Chart out promotion programme 4 B

Activity Description Time (months) Cost (Rs) Predecessor D Make copies of manual and promotion
Normal Crash Normal Crash material 9 B
E Produce first batch for demonstration 16 B
A System calculations 3 1 45,000 63,000 – F Prepare list of press representatives 2 C
B Release of drawings 2 1 15,000 23,000 – G Chief executives conference with
C Procurement of PSU 12 10 85,000 1,05,400 A, B managers 1 C
D Procurement of raw H Press representatives reach Bombay 2 F, G
material 8 6 4,50,000 5,00,000 A, B I Promotional meetings 4 D, H
E Production J Product demonstration 2 E, I
documentation 4 3 40,000 49,500 A, B K Press representatives return home 2 J
F Time study/shop order 3 1 25,000 41,000 E
G PCB Manufacture 6 5 50,000 59,000 D, F (a) Determine the maximum time required to complete the above
H Mechanical parts project, list the critical activities and find the total float and free
manufacturers 7 6 2,00,000 2,20,000 D, E, F
float, if any, for all the non-critical activities.
I Electronic assembly 2 2 43,000 43,000 C, G, H
J Testing 4 2 50,500 75,000 I (b) If the indirect cost per day for the project under consider- ation
is Rs 300, the normal and crash time and cost estimate for
various activities are as given in the following table.
The indirect costs per month are Rs 20,000. Determine the optimal Determine the optimal project duration.
time versus the cost schedule.
13. The Sales Manager of Domestic Products Limited, Mumbai, was Activity Normal Crash
informed by the R&D department about the completion of the proto- Time (days) Cost (Rs) Time (days) Cost (Rs)
type of a particular product. He consulted the production manager
on the time taken to produce the first batch of the product, which A 4 100 3 450
was needed for demonstration during his sales promotion B 4 150 2 510
programme. He also decided to invite a few industrial representa- C 4 200 4 200
tives to the demonstration of this new product and through them to D 9 500 4 1,000
launch it in the market. The various activities involved in this mar- E 16 2,000 8 2,960
keting project, their descriptions, estimated durations (in days) and F 2 60 1 140
immediate predecessors are given in the following tables: G 1 100 1 100
H 2 2,500 1 6,000
Activity Description Duration Predecessor I 4 2,200 3 2,340
(days) J 2 700 2 700
K 2 2,500 1 6,000

A Collect data on specifications and

capabilities 4 –
B Prepare operation manual 4 A

HINTS AND ANSWERS

1. Project Critical Crashed Total cost 18 A – E – H E 54,300 + 1 × 2,000

Length Path(s) Activities Direct + Increment + Indirect B – C – H = 56,300
17 A – E – H B, E 56,300 + 1 × 200
10 A–B–F–G A 1,150 + 1 × 40 + 1,000 = 2,190 B – C – H + 1 × 2,000 = 58,500
9 A–B–F–G F 1,190 + 2 × 50 + 900 = 2,140
7 A–B–F–G B 1,140 + 2 × 75 + 700 = 2,090 4. Critical path A – H – D. Reduce activity H by 2 days at a cost
of Rs 100 and activity D by one day at a cost of Rs 78. Thus
2. (a) First crash activity A by one day at a cost of Rs 100 and then
Total cost = Direct cost + Increase in direct cost due to
activity G at a cost of Rs 400.
crashing + Indirect cost for 20 days project
(b) Crash activity H by 2 days at a cost of Rs 400/3, activityC duration
by 2 days at a cost of Rs 200 and activity D by 2 daysat = (3,030 + 2 × 50 + 1 times 78) + 20 × 25
a cost of Rs 400.
= Rs 3,708.
Total additional cost for minimum project duration of 21 days is
Rs (100 + 400 + 400/3 + 200 + 400) = Rs 1,233.33. 5. Critical path: 1 – 2 – 5 – 7 with project duration of 16 weeks
and cost of Rs 1,24,000.
3.
Project Critical Crashed Total cost
Weeks Contract Contract Project
Length Path(s) Activity Direct + Increment Amount (Rs) Cost (Rs)

21 A – E – H H 52,300 + 1 × 500 16 1,24,000 1,24,000 –

= 52,800 15 1,42,000 1,25,000 17,500
20 A – E – H H 52,800 + 1 × 500 14 1,45,000 1,27,000 18,000
= 53,300 13 1,50,000 1,29,000 21,000
19 A – E – H A 53,300 + 1 × 1,000
A – E – H = 54,300
458 Operations Research: Theory and Applications

13.8 UPDATING OF THE PROJECT PROGRESS

While executing a project, there are certain unexpected delays and difficulties in terms of supply of materials,
availability of machines and/or breakdown of machines, availability of skilled manpower, natural calamity,
etc. In such cases, it is necessary to review the progress of planning and scheduling to take stock of the
progress that has been made. Such review is necessary to take remedial action in terms of time and resources
required for the uncompleted activities in the project.
As such there is no rule about the specific time required to review the project progress. However, to
add dynamism to the nature and progress of work, updating may be carried out as frequently as economically
possible.
The project progress can be updated in two ways:
(i) Use the revised time estimate of incomplete activities and from the initial event calculate the earliest
and latest completion time of each event in order to know the project completion time.
(ii) Change the complete work to zero duration and represent all the activities already finished by an
arrow called the elapsed time arrow. Events in the revised network diagram are renumbered. The
completion times of remaining activities are taken as the revised time.
Example 13.11 The network diagram for a project is shown below. A review of the project after 15 days
reveals that:

(i) Activities 1 – 2, 1 – 3, 2 – 3, 2 – 4 and 3 – 4 are completed.

(ii) Activities 3 – 5 and 4 – 6 are in progress and need 2 and 4 days more, respectively.
(iii) The revised estimate shows that activity 8 – 9 will take only 8 days but 7 – 9 will need 10 days.
Draw a new network diagram after updating the project and determine the new critical path. The work
completed may be shown by an elapsed time activity.
Solution The progress of the work noted at the end of 15th day from the start of the project may be
summarized as shown in Table 13.16.

Activity Time Required (days) Job Status

1 – 2 0 Complete
1 – 3 0 Complete
2 – 3 0 Complete
2 – 4 0 Complete
3 – 4 0 Complete
3 – 5 (20 – 25) 2 In progress
4 – 6 (20 – 26) 4 In progress
5 – 8 (25 – 28) 9 Not started
5 – 7 (25 – 27) 5 Not started
6 – 8 (26 – 28) 8 Not started
6 – 9 (26 – 29) 12 Not started
Table 13.16
Status of 7 – 9 (27 – 29) 10 Not started
Progress 8 – 9 (28 – 29) 8 Not started
 Project Management: PERT and CPM 459

In the network diagram, as shown in Fig. 13.28, activity 1 – 20 shows the elapsed time of 15 days. Other
activities are assigned the time that is required for their completion after review. This is given in Table
13.16. Based on these revised time estimates, the critical path: 1 – 20 – 26 – 28 – 29 is shown bythick
lines in Fig. 13.28. The total project duration has increased by one day.

Fig. 13.28
Network Diagram

13.9 RESOURCE ALLOCATION

Resources such as men, money, material, machinery, etc., are limited and conflicting demands are made for
the same type of resources as a project progresses. A systematic method for the allocation of resources
therefore becomes essential. The aim is to prevent the day-to-day fluctuation in the level of the required
resources and obtain a uniform resource requirement during the project duration. There are two approaches
of optimum utilization of resources that are required to complete a project.

13.9.1 Resource Levelling

The analysis for stabilizing the rate of resource utilization by various activities at different times without
changing the project duration, is called resource levelling.
The amount of total float of non-critical activities is used to stabilize the use of the existing level of
resources. The resource requirement may be minimized by shifting a non-critical activity between its earliest
start time and latest allowable time. The following two general rules are normally used in scheduling non-
critical activities:
(i) If the total float of a non-critical activity is equal to its free float, then it can be scheduled anywhere
between its earliest start and latest completion times.
(ii) If the total float of a non-critical activity is more than its free float, then its starting time can be
delayed relative to its earliest start time by no more than the amount of its float, without affecting
the scheduling of its immediately succeeding activities.

13.9.2 Resource Smoothing

The efforts to reduce the peak demand for resources and to reallocate them among the activities of a project
so that the total project duration remains the shortest is known as resource smoothing (or loading).
The procedure of carrying out resource smoothing can be summarized in the following steps.
Step 1: Calculate the earliest start and latest finish times of each activity and then draw a time scaled version
(or squared) of the network. A critical path is drawn along a straight line and non-critical activities on both
sides of this line. Resource requirement of each activity is given along the arrows.
Step 2: Draw the resource histogram by taking earliest start times or latest start times of activities on
the x-axis and required cumulative resource on y-axis.
Step 3: Shift the start time of those non-critical activities that have the largest float in order to smoothen the
resources.
Example 13.12 A network with the following activity durations and manpower requirement is given. Analyse
the project form point of view of resource to bring out the necessary steps involved in the analysis and
smoothing of resources.

Activity : 1–2 2–3 2–4 3–5 4–6 4–7 5–8 6–8 7–9 8 – 10 9 – 10
Duration (weeks) : 2 3 4 2 4 3 6 6 5 4 4
Manpower required : 4 3 3 5 3 4 3 6 2 2 9
460 Operations Research: Theory and Applications

Solution The critical path is drawn through events where E-values and L-values are equal by a thick line.
The critical path is: 1 – 2 – 4 – 6 – 8 – 10, with the project duration of 20 weeks. The network diagram in the time-
scaled form, also called squared network is shown in Fig. 13.29 where critical path is drawn along a horizontal
line. The dotted lines show the total float of an activity

(a) Square Network

Fig. 13.29
Squared Network (b) Loard Chart

The squared network is based on the earliest start times, and has been obtained by vertically summing up
the manpower requirements for each week. The maximum demand of 15 men occurs in the 15th and 16th weeks.

Fig. 13.30
Resource
Smoothing
 Project Management: PERT and CPM 461

As shown in Fig. 13.30, activities are shifted depending upon the floats. The activity path 4–7–9–10 has a
float of two weeks, and the activities 7–8 and 9–10 are shifted to the right so that the start of each is delayed by
two weeks. Similarly, activity 5–8 can be shifted to the right so that it starts on 11th day instead of starting on
8th day. The loading chart indicates that the maximum manpower required is 11 men. Thus, the project can be
completed in 20 weeks by 11 men as compared to 15 men for the previous schedule.
Example 13.13 Consider the following network diagram. Durations of the activities are marked along their
arrows. The requirement of masons (M) and labourers (L) for each activity is given the table below. Analyse the
project and smoothen the requirement of the resources.

Activity Duration Masons (M) Laboures (L)

(weeks) (weeks)
1–2 2 1 2
2–3 3 2 2
2–4 4 3 2
2–5 2 1 3
3 – 10 4 2 2
4–6 2 3 2
4–7 4 3 3
5–9 4 5 3
6–8 2 1 2
7–9 5 1 3
8–9 3 – 4
9 – 11 2 1 1
10 – 11 3 1 2
11 – 12 2 1 2

Solution The critical path is identified by first calculating E-values and L-values and then joining those
events by a thick line where these values are equal. The critical path is: 1 – 2 – 4 – 7 – 9 – 11 – 12 with the project
duration of 19 weeks.
The network diagram is represented in the time scaled form as shown in Fig. 13.31(a). The critical path is
drawn along a horizontal line. The dotted lines show the total float of an activity. Durations and requirements
of masons (M) and labourers (L) for each activity are marked along the activity arrows. The loading chart shows
the total numbers of masons and labourers required each day.
462 Operations Research: Theory and Applications

(a) Square Network

Fig. 13.31
Resource
Smoothing
(b) Loard Chart
Figure 13.31(a) shows that activities 2–5 and 5–9 have a total float of 7 days. The start time of activity 5–9
can be shifted by 7 days so that it starts on 12th day instead of 5th day. Figure 13.32(a) represents the modified
squared network along with the loading chart.
Figure 13.32(a) indicates that the demand for masons has decreased from 13 to 8 on the 7th and 8th days.
However, the demand of labourers has increased from 9 to 12th day.

(a) Square Network

Fig. 13.32
(b) Loard Chart

As shown in Fig. 13.33(a), the start time of activity 8 – 9 can be shifted from 11th day to 13th day by utilising
the float available for 2 days, wherein the requirement of labourers has also decreased from 12 to 10. The
requirement of resources is smoothed without affecting the project duration.
 Project Management: PERT and CPM 463

(a) Square Network

Fig. 13.33
Resource
Smoothing

(b) Loard Chart

Example 13.14 Following are the manpower requirements for each activity in a project.
Activity Normal Time Manpower Required
0–1 2 4
1–2 3 3
1–3 4 3
2–4 2 5
3–5 4 3
3–6 3 4
4–7 6 3
5–7 6 6
6–8 5 2
7–9 4 2
8–9 4 9
(a) Draw the network diagram of the project activities.
(b) Rearrange the activities suitably for reducing the existing total manpower requirement.
(c) If only 9 men are available for the execution of the project, then rearrange the activities suitable for levelling
the manpower resource.
Solution (a) The network diagram of the given project is shown in Fig. 13.34. Numbers in bracket with each
activity in Fig. 13.34 is representing manpower requirement of the activity.
The earliest and latest completion times of each event has been calculated in the usual manner as described
earlier and are indicated along the nodes in Fig. 13.34. The critical path: 0 – 1– 3– 5– 7 –9 is shown with thick
lines. The total project duration is of 20 days.

Fig. 13.34
Network Diagram
464 Operations Research: Theory and Applications

In order to exhibit activities on a time scale so that ranges of time allowable for various activities may be
known, a time scaled version of the given network (also squared network) has been drawn. For this, first of all
critical path is drawn along straight line and non-critical activities are added with it as shown in Fig. 13.35. Here,
it is assumed that the various activities may be started as soon as possible, at their earliest start (E) times.

Fig. 13.35
Time Scaled
Graph

The performance of some of the non-critical activities is shifted so that requirement of manpower may be
rearranged. Because these non-critical activities have slack. To reduce the manpower requirements in the
earlier weeks and to raise it in the last weeks, we may postpone the beginning of activity 3 – 6,6
– 8 and 8 – 9 by two weeks each to start all at the beginning of 11th and 16th day, respectively and that of
activity 4 – 7 by 3 days to start at the 10th day. The result of such rearrangement is shown in Fig. 13.36.
In Fig. 13.36, it may be observed that the requirement is less uneven in comparison with the previous
situation and now 11 men are required compared to 15 men. Figures 13.35 and 13.36 indicate that there is no
need of employing 11 men at a time. Men may be employed as per need.
Since only 9 men are available instead of 11, therefore activities have to be rearranged and expanded in terms
of their completion time. This would increase the total project duration from 20 days to 25 days.

Fig. 13.36
Time Scaled
Graph

Example 13.15 Following are the manpower requirement for each activity in a project.
Activity Normal Time Manpower Required
(Days) per Day
1–2 10 2
1–3 11 3
2–4 13 4
2–6 14 3
3–4 10 1
4–5 7 3
4–6 17 3
 Project Management: PERT and CPM 465

5–7 13 5
6–7 9 8
7–8 1 11

(a) Draw the network and find out total float and free float for each activity.
(b) The contractor stipulates that during the first 26 days only 4 to 5 men and during remaining days 8 to 11
men only can be made available. Rearrange the activities suitably for levelling the manpower resources,
satisfying the above condition.
Solution The network diagram of the given project is shown in Fig. 13.37. The value of total float and free
float for non-critical activities are given in Table 13.16.

Activity Normal Time Earliest Time Latest Time Total Free

(Days) Float Float
Start Finish Start Finish
1–2 11 0 11 2 13 2 0
2–6 14 10 24 26 40 16 16 Table 13.16
3–4 10 11 21 13 23 2 2 Activity
4–5 7 23 30 29 36 6 0 Scheduling Times
5–7 13 30 43 36 49 6 6 and Float

The total project duration is 50 days. Now it is required to rearrange activities so that they can be performed
within 50 days (project duration) with the available manpower at different intervals.
The critical activities 1 – 2, 2 – 4, 4 – 6, 6 – 7 and 7 – 8 are scheduled first. Since non-critical activity 1 – 3 is
scheduled at t = 0 followed by the non-critical activity 3 – 4, therefore, on 11th day (earliest finish time of 1 – 3)
the manpower requirement would rise to 7, which is more than the available on 11th day. The total float of
activity 1 – 3 is more than its free float, even than rule (b) is not applicable because its earliest start time as well
as free float are both zero. Thus, it has to be scheduled otherwise it would result in a delay of the project.

Fig. 13.37
Network Diagram

Starting time of non-critical activities 2 – 6 and 4 – 5 can be delayed both to t = 26 because more men would
then be available (activity 4 – 6 can also be delayed upto t = 29). The last non-critical activity5
– 7 is scheduled at t = 33. The total daily requirement of men is shown in Table 13.17 and manpower
requirement of various activities is shown in bracket in Fig. 13.38.

Fig. 13.38
Time Scaled
Graph
466 Operations Research: Theory and Applications

Interval Day(s) Men

0 – 10 1 to 10th 5
10 – 11 11th 7
11 – 21 12th to 21st 5
21 – 23 22nd to 23rd 4
23 – 26 24th to 26th 5
Table 13.17 26 – 46 27th to 46th 11
Manpower 46 – 49 47th to 49th 8
Requirement 49 – 50 50th 11

Example 13.16 A project with the following activities duration and manpower requirement is given:
Activity : 1–2 1–3 1–4 2–5 2–6 3–7 4–8 5–9 6–9 7–8 8–9
Duration (days) : 2 2 0 2 5 4 5 6 3 4 6
Manpower
required : 5 4 0 2 3 6 2 8 7 4 3

(a) Draw the network diagram of the project indicating the earliest start, earliest finish, latest finish and float of
each activity.
(b) There are 11 persons who can be employed for this project. Carry out the appropriate manpower levelling
so that the fluctuation of work force requirement from day-to-day is as small as possible.
Solution Network diagram based on activities schedule has been drawn as shown in Fig. 13.39. The critical
path: 1 – 3 – 7 – 8 – 9 has been shown with thick lines.

Fig. 13.39
Network Diagram

Table 13.18 illustrates man-days required for each activity and floats.

Activity Duration Manpower Man-days Earliest Time Latest Time Total

(D) (M) (M × D) Start Finish Start Finish Float
(Ei) (Ei + tij) (Lj – tij) (Lj) (Lj – tij) – Ei

1–2 2 5 5 × 2 = 10 0 2 6 8 6
1–3 2 4 4× 2 = 8 0 2 0 2 0
1–4 0 0 0× 0 = 0 0 0 5 5 5
2–5 2 2 2× 2 = 4 2 4 8 10 6
2–6 5 3 3 × 5 = 15 2 7 8 13 6
3–7 4 6 6 × 4 = 24 2 6 2 6 0
4–8 5 2 5 × 2 = 10 0 5 5 10 5
5–9 6 8 8 × 6 = 48 4 10 10 16 6
6–9 3 7 7 × 3 = 21 7 10 13 16 6
7–8 4 4 4 × 4 = 16 6 10 6 10 0
Table 13.18 8–9 6 3 3 × 6 = 18 10 16 10 16 0
 Project Management: PERT and CPM 467

Table 13.19 represents the resource required to be allocated along with loaded chart as shown in Fig. 13.40.
This figure indicates that five persons remain idle from 8th to 10th day and four persons remain idle from 17th
to 19th day. The project cannot be completed in the normal time duration of 16 days and has to be delayed by
3 days if 11 persons only are to be employed.

Halting Available Activities in the Queue Resource

Time Resource Activity Man days Allocated
(M × D) Float Priority

0 11 M 1–2 5 × 2 = 10 6 III 5M
1–3 4× 2 = 8 0 I 4M
4–8 2 × 5 = 10 5 II 2M
2 9M 2–5 2× 2 = 4 6 III —
2–6 3 × 5 = 15 6 II 3M
3–7 6 × 4 = 24 0 I 6M
5 2M 2–5 2× 2 = 4 3 I 2M
6 6M 7–8 4 × 4 = 16 0 I 4M
7 5M 5–9 8 × 6 = 48 3 I —
6–9 7 × 3 = 21 6 II —
5–9 8 × 6 = 48 0 I 8M
10 11 M 6–9 7 × 3 = 21 3 III — Table 13.19
8–9 3 × 6 = 18 0 II 3M Resource
16 11 M 6–9 7 × 3 = 21 – I 7M Allocation

Fig. 13.40
Loading Chart

Example 13.17 For a project consisting of several activities, the durations and required resources for
carrying out each of the activities and their availabilities are given below:

Activity Equipment Operators Duration (days)

1–2 X 30 4
1–3 Y 20 3
1–4 Z 20 6
2–4 X 30 4
2–5 Z 20 8
3–4 Y 20 4
4–5 Y 20 4
4–5 X 30 6

Resources availability: (a) Operators = 50 (b) Equipment X = 1, Y = 1 and Z = 1.

(a) Draw the network, identify critical path and compute the total float for each of the activities.
(b) Find the project completion time under the given resource constraints.
Solution Network diagram based on activities schedule has been shown in Fig. 13.41. The critical path:
1 – 2 – 4 – 5 has been shown with thick lines
468 Operations Research: Theory and Applications

Fig. 13.41
Network Diagram

Table 13.20 illustrates man-days required for each activity and floats occurred.

Activity Duration Manpower Man-days Equip- Earliest Time Latest Time Total
(D) (M) (M × D) ment Start Finish Start Finish Float
(Ei) (Ei + tij) (Lj – tij) (Lj) (Lj – tij) – Ei

1–2 4 30 30 ×4= 120 X 0 4 0 4 0

1–3 3 20 20 ×3= 60 Y 0 3 1 4 0
1–4 6 20 20 ×6= 120 Z 0 6 2 8 0
2–4 4 30 30 ×4= 120 X 4 8 4 8 0
2–5 8 20 20 ×8= 160 Z 4 12 6 14 0
3–4 4 20 20 ×4= 80 Y 3 7 4 8 0
Table 13.20 3–5 4 20 20 ×4= 80 Y 3 7 10 14 6
Network Analysis 4–5 6 30 30 ×6= 180 X 8 10 8 14 2

Table 13.20 represents the resource required to be allocated along with loaded chart as shown in Fig. 13.41.
Activities 1–2, 1–3 are taken up at time 0 while 1 – 4 is delayed to start at time 3. Float of activity 1 – 4 becomes
negative and therefore float will not be the criterion for finding priorities. At time 3, activities 3 – 4 and 3–5 have
the same man-days, manpower size and activity sequence, therefore, priority II is assigned to activity 3 – 4 and
priority III to activity 3–5.
At time 4, available resources are 30M, X and Y. Activity 2–5 has the highest man-days and is assigned priority
I. However, since it requires equipment Z, it cannot be taken up and, instead, activity 2–4 with priority II is
selected for execution. Figure 13.42 indicates that the project requires 21 days for completion and is, therefore,
delayed by 7 days beyond normal completion time. The idle man-days are shown shaded. It may be observed that
during 19th, 20th and 21st days only 20 persons are required. The remaining 30 persons can be relieved/shifted
to other project after 18 days. Daily requirement of operators and equipment is shown in Fig. 13.42.
Halting Available Activity Man-days Float Priority Resources
Time Resources Equipment (M × D) Allocated

0 50 M; X, Y, Z 1–2 X 30 × 4 =120 0 I 30 M, X
1–3 Y 20 × 3 = 60 I II 20 M, Y
1–4 Z 20 × 6 =120 2 III
3 20 M; Y, Z 1–4 Z 20 × 6 =120 – I 20 M, Z
3–4 Y 20 × 4 = 80 – II
3–5 Y 20 × 4 = 80 – III
4 30 M; X, Y 2–4 X 30 × 4 =120 – II 30 M, X
2–5 Z 20 × 8 =160 – I
3–4 Y 20 × 4 = 80 – III
3–5 Y 20 × 4 = 80 – IV
8 30 M; X, Y 2–5 Z 20 × 8 =160 – I –
3–4 Y 20 × 4 = 80 – II 20 M, Y
3–5 Y 20 × 4 = 80 – III
 Project Management: PERT and CPM 469

9 30 M; X, Z 2–5 Z 20 × 8 = 160 – I 20 M,Z

3–5 Y 20 × 4 = 80 – II
12 30 M; X, Y 3–5 Z 20 × 4 = 80 – II
Table 13.21
4–5 X 30 × 6 = 180 – I 30 M, X Resource
17 20 M; Y, Z 3–5 Y 20 × 4 = 80 – I 20 M, Y Allocation

Fig. 13.42
Loading Chart

SELF PRACTICE PROBLEMS E

1. The following table gives, for each activity of a project, its duration C – 8 2
and corresponding resource requirements as well as the D A 5 3
availability of each type of resource. E C 4 1
F B, E 4 2
Activity Duration Resource Required
G C 11 2
(days) Machines Men H G, Fs 4 1
1 – 2 7 2 20
1 – 3 7 2 20 (a) For the above project draw the network. Determine the critical
2 – 3 8 3 30 path and its duration.
2 – 4 6 4 30 (b) If there were only three men available at any one time how
3 – 6 9 2 20 long would the project take and how would you allocate the
4 – 5 3 2 20 men to the activities?
(c) If there were no restrictions on the amount of labour available,
5 – 6 5 2 20
explain how you would schedule the activities.
Minimum available resources 4 40 3. The activities, activity durations and manpower requirements of
(a) Draw the network, compute the earliest occurrence time,and a project are given below.
the latest occurrence time for each event, the total float for
each activity and identify the critical path, assuming that there Activity Duration (days) No. of Men Required
are no resource constraints.
(b) Under the given resource constraints find out the minimum 1–2 2 5
duration to complete the project and compare the utilization of 1–3 2 4
the resources for that duration. 1–4 0 0
2. The activities involved in a certain project have been identified 2–5 2 2
as follows: 2–6 5 3
3–7 4 6
4–8 5 2
Activity Preceding Activity Duration No. of Men 5–9 6 8
(weeks) Required 6–9 3 7
7–8 4 4
A – 4 1 8–9 6 3
B – 7 1
470 Operations Research: Theory and Applications

There are eleven persons who can be employed for this project. D 2 A, B
Carry out the appropriate manpower levelling so that the E 4 B, C
fluctuations of workforce requirement from day-to-day are as small F 1 C
as possible. G 4 D
4. A project has the following activities and their durations. H 8 G, E, F

(a) Draw the project network and indicate the critical path.
Activity Duration (days)
(b) What is the minimum completion time?
12 13 (c) During the second day of work it is discovered that activity
13 15 F would take 4 days instead of 1. Will this delay the project?
If this activity takes 6 days, will the project be delayed?
14 9
(d) The company has limited number of men available to work
24 10 on the project. Only two activities can be under way at the
25 27 same time. Will this delay the project when compared to the
34 7 time that unlimited resources would have been? (Activity F
36 18 takes 6 days to complete).
47 30 6. The following table gives the activities in a small project andother
57 12 relevant information:
67 10
68 10 Activity Duration Immediate Resource Required
78 9 (days) Predecessor Operators Mechanics

A 3 – 2 –
(a) Draw the network of the project and find its duration.
B 2 A 2 2
(b) At the end of 25 days it is observed that
C 4 A 4 4
(i) activities 12, 13, 14 have been completed. D 6 A 5 5
(ii) activity 24 is in process and will be completed in 5
E 3 B 2 2
more days.
F 2 E 2 –
(iii) activity 36 is in progress and will need 20 more days
for completion. G 6 C – 2
(iv) activity 67 is presenting some problem and will take H 4 D 2 2
15 days. I 4 G 4 2
Draw the updated network and find out its revised duration. J 2 D 2 –
Number 12 denotes activity 1 – 2 and same is true for other K 2 J 2 2
activities. L 4 F, H, I 4 4
5. A project has the following activities and their durations: (a) Draw the network, compute the earliest start time and latest
finish time for each of the activities and find out the project
Activity Time Preceding Activity completion time. Also identify the critical path.
(days) (b) Draw the time-scaled diagram with resource accumulation
table. Comment on the demand for the operations and
A 1 – mechanics for the entire project duration and suggest the
B 2 – method of smoothening the resources.
C 2 –

CHAPTER SUMMARY
Ever since their inception in the late 1950s, PERT (Program Evaluation and Review Technique) and CPM (Critical Path Method)
have been used extensively to assist project managers in planning, scheduling, and controlling their projects.
The application of PERT/CPM begins by breaking down the project into its individual activities, identifying the immediate
predecessors of each activity, and estimating the duration of each activity. The next step is to construct a project network to display
this information.
PERT/CPM generate scheduling information for the project, including the earliest start time, the latest start time, and the slack
for each activity. It also identifies the critical path of activities such that any delay along this path will delay project completion.
Since the critical path is the longest path through the project network, its length determines the duration of the project, assuming all
activities remain on schedule.
However, it is difficult for all activities to remain on schedule because there is often a considerable uncertainty about what
the duration of an activity will turn out to be. The three-estimate approach in PERT deals with this situation by obtaining three
different types of estimates (most likely, optimistic, and pessimistic) for the duration of each activity. This information is used
to approximate the mean and variance of the probability distribution of this duration. It is then possible to approximate the
probability that the project will be completed by the deadline.
The time-cost trade-offs approach in CPM enables the project manager to investigate the effect on total cost of changing
the estimated duration of the project to various alternative values. The data needed for this activity are the time and cost for
each activity when it is done in the normal way and then when it is fully crashed (expedited).
 Project Management: PERT and CPM 471

CHAPTER CONCEPTS QUIZ

True or False
1. The objective of network analysis is to minimize total project cost. 25. Generally the PERT technique deals with the project of
2. A project is an endeavour to create a unique product or service. (a) repetitive nature (b) non-repetitive nature
3. The techniques of operations research used for planning, (c) deterministic nature (d) none of the above
scheduling and controlling projects are referred to as network 26. In PERT the span of time between the optimistic and pessimistic
analysis. time estimates of an activity is
4. The CPM is sued for completing the projects that involves activities (a) 3 (b) 6 
of repetitive nature. (c) 12  (d) none of the above
5. PERT is referred to as an activity oriented technique. 27. If an activity has zero slack, it implies that
6. PERT is a tool for planning and control of time. (a) it lies on the critical path (b) it is a dummy activity
7. Project planning phase allocates resources to work packages. (c) the project is progressing well (d) none of the above
8. Scheduling phase identify manpower that will be responsible for 28. A dummy activity is used in the network diagram when
each task. (a) two parallel activities have the same tail and head events
9. A network in which the activities are represented by an arrow (b) the chain of activities may have a common event yet be
is referred to as activity-on-node network. independent by themselves
10. Beta probability distribution is often used in computing the (c) both (a) and (b)
expected activity completion times and variances in networks. (d) none of the above

Fill in the Blanks 29. While drawing the network diagram, for each activity project, we
should look
11. Resource leveling is the process of the utilization of (a) what activities precede this activity?
resources in a project. (b) what activities follow this activity?
12. Crashing is the process of reducing the total time that it takes (c) what activities can concurrently take place with this activity?
to complete a project by expending . (d) all of the above
13. is the time consuming job or task that is a key subpart
30. In the PERT network each activity time assumes a Beta-
of the total project.
distribution because
14. Earliest finish time that an activity can be finished without (a) it is a unimodal distribution that provides information regard-
of precedence requirements.
ing the uncertainty of time estimates of activities
15. is the point in time that marks the beginning or ending (b) it has got finite non-negative error
of an activity. (c) it need not be symmetrical about model value
16. A bar chart indicating when the activities in a project will be (d) all of the above
performed is referred to as .
31. The critical path satisfy the condition that
17. Network is the graphical display of a project that contains both
and . (a) Ei = Li and Ej = Lj
(b) Lj – E i = Li – Lj
18. A small circle or rectangle that is known as serves as
a junction point in the project network. (c) Lj – Ei = Li – Ej = d (constant)
(d) all of the above
19. Latest finish time that an activity can be finished without
the entire project. 32. Float or slack analysis is useful for
20. The amount of time that is expected to complete the activity is (a) projects behind the schedule only
called . (b) projects ahead of the schedule only
(c) both a and b
Multiple Choice (d) none of the above
21. The objective of network analysis is to 33. The activity that can be delayed without affecting the execution
(a) minimize total project duration of the immediate succeeding activity is determined by
(b) minimize total project cost (a) total float (b) free float
(c) minimize production delays, interruption and conflicts (c) independent float (d) none of the above
(d) all of the above 34. In time cost-trade-off function analysis
22. Network models have advantage in terms of project (a) cost decreases linearly as time increases
(a) planning (b) scheduling (b) cost at normal time is zero
(c) controlling (d) all of the above (c) cost increases linearly as time increases
23. The slack for an activity is equal to (d) none of the above.
(a) LF – LS (b) EF – ES 35. Activity-on-Arrow (AOA) diagram is preferred over Activity-on-
(c) LS – ES (d) none of the above Node (AON) diagram because
24. The another term commonly used for activity slack time is (a) AOA diagrams are simple to construct
(a) total float (b) free float (b) AOA diagrams give a better sense of the flow of time
throughout a project
(c) independent float (d) all of the above
(c) AOA diagrams do not involve dummy activities.
(d) all of the above
472 Operations Research: Theory and Applications

Answers to Quiz
1. F 2. T 3. T 4. T 5. F 6. T 7. T 8. T 9. F 10. T
11. smoothing out 12. additional funds 13. activity 14. violation 15. event
16. gantt chart 17. activities, events 18. node 19. delaying 20. most likely time
21. (a) 22. (d) 23. (c) 24. (d) 25. (d) 26. (b) 27. (a) 28. (c) 29. (d) 30. (a)
31. (a) 32. (a) 33. (b) 34. (a) 35. (b)

CASE STUDY
Case 13.1: Kaushik Mills
Kaushik Mills decided to increase its production capacity by building a new feed mill. The project consisted of a number
of separate tasks, some of which could not be started before others were complete. Exhibit 1 lists the activities, along
with the times expected for each, as ‘agreed upon by management and the precedence relationships’.
The management wanted to advance the schedule as much as possible in order to save valuable time in getting the
new mill into operation. The president of the mills commented: “Every week saved is worth Rs 70,000 in lost contribution
if we can get going.”
Some of the construction activities could be sped up. For example, the firm’s architects could by working overtime,
design the new plant in 10 weeks instead of the originally estimated 12 weeks. This advancement would cost the mills
an additional Rs 25,000 per week that is advanced. The following table shows the maximum amount each activity
could be crashed as well as the crash cost per week.
The president of mills had already held discussions with mill contractors, an independent firm which was a potential
contractor for one of the projects major tasks, for building the plant. Kaushik Mills intended to do the other tasks
either itself or through its agents. During the talks with mill contractors, the management had explored a number of bonus
and penalty clauses. One of these was that for every week that the plant was built ahead of 10 weeks, themills
would pay contractors an additional Rs 75,000.

Activity Description Expected Precedent Minimum Crash

Time (weeks) Activities Time (weeks) (Rs/week)
A Degin plant 12 – 10 24,000
B Select plant site 8 A 18 –11
C Select plant builder 6 A 11 3,000
D Select operating personnel 13 A 13 –11
E Prepare the building site 4 B, C 4 –11
F Make or buy mill equipment 101 C 8 30,000
G Prepare mill operations manual 6 C 4 1500
H Build the plant 101 E, F 6 75,000
I Train plant operators 8 D, G 8 –11
J Test the plant 8 H, I 8 –11
K Obtain a production licence 4 J 4 –11

The management of Kaushik Mills desires to know which activities it would crash and how it should schedule its workers.

Case 13.2: Krishna Foods

Krishna Foods has broken down the process of launching a new Fruit Yogot in the market into several steps. Some
of these steps cannot begin until the others are completed. These relationships are shown in the following table. The
estimates of the completion time (in weeks) of each activity and its immediate predecessor(s) are listed in the following
table:
 Project Management: PERT and CPM 473

Activity Description Predecessor Time (weeks)

Optimistic Likely Pessimistic
A Management approval – 2 2.5 4
B Product concept test A 3 4.7 5
C Technical feasibility A 2 2 3
D Recipe finalization C, B 1 1 2
E Shelf life trials D 8 12 15
F Brand positioning study B 4 5 7
G Packaging key lines F 1 1 2.5
H Agency advertisement development F, E 4 8 10
I Agency: layouts artworks G 2 3 5
J Advertisement test research H 3 4 5
K Cost finalization D 1 1 2
L Pricing decision J, F 1 1.5 3
M Marketing mix finalization L 2 2.5 3
N POS development M 2 4 5
O Launch plans M 1 1 1.5
P Branch communication O 0.5 0.5 1
Q Supplier’s delivery of packaging H 4 5 8
R Production trail D, Q 1.5 2 3
S Management final approval R 0.5 1 1.5
T Final production S 1 2.5 3
U Stock movement T 0.5 1.5 2
V Position movement T 0.5 1 1.5
W Launch U, V 1 2 4
The management of the company desires to budget an amount for labour costs that will be sufficient with 95 per
cent probability of completing the project. How much delay is possible in the project completion.
 C h a p t e r 20
“When everything seems to be going against you, remember that the airplane takes off
against the wind, not with it.”
– Henry Ford

The sequencing techniques deals with the problem of preparing optimal timetable for jobs, equipment,
people, materials, facilities and all other resources that are needed to support the production schedule. The
objective is the minimization of the total elapsed time between the completion of first and last job in a
particular order.

LEARNING OBJECTIVES
After studying this chapter you should be able to
 appreciate the aims to study sequencing techniques.
 use Johnson’s rule of sequencing or scheduling.
 solve some specific problems of scheduling jobs on one, two or three machines.
 see how to extend Johnson’s rule to more complicated problems.

CHAPTER OUTLINE
20.1 Introduction  Self Practice Problems B
20.2 Notations, Terminology and Assumptions  Hints and Answers
20.3 Processing n Jobs Through Two Machines
 Conceptual Questions A
 Self Practice Problems A
 Hints and Answers
20.4 Processing n Jobs Through Three Machines
20.5 Processing n Jobs Through m Machines
20.6 Processing Two Jobs Through m Machines
 Conceptual Questions B
 Self Practice Problems B
 Hints and Answers
 Chapter Summary
 Chapter Concepts Quiz
 Sequencing Problem 709

20.1 INTRODUCTION
The optimal order (sequence) shows the minimum time in which jobs, equipment, people, materials, facilities
and all other resources are arranged to support the production schedules to give low costs and high
utilizations. Other objectives of calculating optimal production schedule are minimizing customers waiting
time for a product or service, meeting promised delivery dates, keeping stock levels low providing preferred
working pattern, and so on.
If n jobs are to be performed, one at a time, on each of m machines, where sequence (order) of the
machines in which each job should be performed, and the actual (or expected) time required by the jobs
on each of the machines are given, then the general sequencing problem is to find a sequence out of (n!)m
possible sequences, which minimize the total elapsed time between the start of the job on first machine and
the completion of the last job on the last machine.
In particular, if there are n = 3 jobs to be performed and m = 3 machines are to be used, then the total
number of possible sequences will be (3!)3 = 216. Theoretically, it may be possible to find the optimum
sequence but this would require a lot of computational time. Thus, one should adopt the sequencing
technique.
To find the optimum sequence, we first need to calculate the total elapsed time for each of the possible
sequences. As stated earlier, even if the values of m and n are very small, it is difficult to get the desired
sequence with the total minimum elapsed time. However, due to certain rules designed by Johnson, the task
of determining an optimum sequence has become quite easy.

20.2 NOTATIONS, TERMINOLOGY AND ASSUMPTIONS

Notations
tij = Processing time (time required) for job i on machine j. Sequencing
problem is the
T = Total elapsed time for processing all the jobs. This includes idle time, if any. problem of finding
Iij = Idle time on machine j from the end of job (i – 1) to the start of job i. an optimal sequence
of completing certain
number of jobs so
Terminology as to minimize the
 Number of Machines The number of machines refer to the number of service facilities through which total elapsed time
between completion
a job must pass before it is assumed to be completed. of first and last job.
 Processing Time This is the time required by a job on each machine.
 Processing Order This refers to the order (sequence) in which machines are required for completing the
job.
 Idle Time on a Machine This is the time during which a machine does not have a job to process,
i.e. idle time from the end of job (i – 1) to the start of job i.
 Total Elapsed Time This is the time interval between starting the first job and completing the
last job, including the idle time (if any), in a particular order by the given set of machines.
 No Passing Rule This refers to the rule of maintaining the order in which jobs are to be processed on
given machines. For example, if n jobs are to be processed on two machines, M1 and M2 in the order
M1M2, then each job should go first to machine M1 and then to M2.

Assumptions
1. The processing time on different machines are exactly known and are independent of the order of the
jobs in which they are to be processed.
2. The time taken by the job in moving from one machine to another is negligible.
3. Once a job has begun on a machine, it must be completed before another job can begin on the same
machine.
4. All jobs are known and are ready for processing before the period under consideration begins.
5. Only one job can be processed on a given machine at a time.
6. Machines to be used are of different types.
7. The order of completion of jobs are independent of the sequence of jobs.
710 Operations Research: Theory and Applications

20.3 PROCESSING n JOBS THROUGH TWO MACHINES

Let there be n jobs, each of which is to be processed through two machines, M1 and M2 in the order M1M2,
i.e. each job has to pass through the same sequence of operations. In other words, a job is assigned on
machine M1 first and once processing is over on machine M1, it is assigned to machine M2. If the machine
M2 is not free for processing the same job, then the job is placed in waiting line for its turn on machine
M2, i.e. passing is not allowed.
Since passing is not allowed, therefore, machine M1 will remain busy in processing all the n jobs one-
by-one, while machine M2 may remain idle waiting for the jobs to come from M1. The idle time for both
M1 and M2 may be reduced by determining an optimal sequence of n jobs to be processed on two machines
M1 and M2. The procedure suggested by Johnson for determining the optimal sequence is summarized as
follows:

20.3.1 Johnson’s Procedure

Step 1: List the jobs along with their processing times on each machine in a table, as shown below:
Processing Time Job Number
on Machine 1 2 3 ... n
M1 t11 t12 t13 .. . t1n
M2 t21 t22 t23 .. . t2n

Step 2: Examine the processing times on machines M1 and M2, in each column of the table and find the
shortest processing time, i.e. find out, min. (t1j, t2j) for all j.
Step 3(a): If the shortest processing time is on machine M1, then place the job in the first available position
in the sequence. If the processing time is on machine M2, then place the job in the last available position in
the sequence.
(b) If there is a tie in selecting the minimum of all the processing times, then the following threesituations
may arise:
(i) If minimum processing time is same on both machines, i.e. min (t1j, t2j) = t1k = t2r , then process
the kth job first and the rth job last.
(ii) If minimum processing times, t1j on machine M1 are same for more than two jobs, then select the
job corresponding to the smallest job subscript, j first.
(iii) If minimum processing times, t2j on machine M2 are same for more than two jobs, then select the
job corresponding to the largest job subscript, j last.
Step 4: Cross off assigned jobs from the table. If no job remains to be assigned, then stop the procedure and
go to Step 5. Otherwise, go to Step 2.
Step 5: Calculate the idle time for machines M1 and M2:
(a) Idle time for machine M1 = (Total elapsed time) – (Time when the last job in a sequence finishes
on machine M1)
(b) Idle time for machine M2 = Time when first job in a sequence finishes on machine M1
n
+  {(Time when the jth job in a sequence starts on machine M2)
j2

– (Time when the ( j –1)th job in a sequence finishes on machine M2)}.

Step 6: The total elapsed time to process all jobs through two machines is given by:
Total elapsed time = Time when the nth job in a sequence finishes on machine M2.
n n
=  M2j   I2 j
j 1 j 1

where M2j = Time required for processing jth job on machine M2.
I2j = Idle time for machine M2 after processing (j – 1)th job and before the start of jth job
processing.
 Sequencing Problem 711

Example 20.1 A book binder has one printing press, one binding machine and manuscripts of 7 different
books. The times required for performing printing and binding operations for different books are shown
below:

Book : 1 2 3 4 5 6 7
Printing time (hours) : 20 90 80 20 120 15 65
Binding time (hours) : 25 60 75 30 90 35 50

Decide the optimum sequence of processing of books in order to minimize the total time required to
bring out all the books.
Solution Examine both printing and binding operations time columnwise. The smallest value is 15 printing
hours for book 6. The book 6 is placed first in the sequence and cross off column 6 of the table. Repeat
the procedure of finding smallest processing time columnwise. Books 1 and 4 are placed in the sequence
as shown below:

6 1 4

Cross off columns 1, 4 and 6 of the table. For the remaining columns 2, 3, 5 and 7 the smallest time
is 50 binding hours for book 7, so it should be sequenced in the last as shown below:

6 1 4 7

After this sequence, remaining set of printing and binding times gets reduced as follows:

Book : 2 3 5
Printing time : 90 80 120
Binding time : 60 75 90

The smallest time in this reduced data is 60, which corresponds to binding. So book 2 is placed in the
second cell from right. Proceeding in the same manner, the optimal sequence so obtained is shown below:

6 1 4 5 3 2 7

The total minimum elapsed time for printing and binding is given in Table 20.1.

Printing Binding
Book Time In Time Out Time In Time Out Idle Time
6 0 0 + 15 = 15 15 15 + 35 = 50 15
1 15 15 + 20 = 35 50 50 + 25 = 75 —
4 35 35 + 20 = 55 75 75 + 30 = 105 —
5 55 55 + 120 = 175 175 175 + 90 = 265 70
3 175 175 + 80 = 255 265 265 + 75 = 340 — Table 20.1
2 255 255 + 90 = 345 345 345 + 60 = 405 5 Minimum Elapsed
7 345 345 + 65 = 410 410 410 + 50 = 460 5 Time

In Table 20.1, the minimum elapsed time, i.e. time from start of printing book 6 to binding the last book
7 is 460 hours. During this time the printing machine remains idle for 460 – 410 = 50 hours. The idle time
for binding machine is shown in Table 20.1.
Example 20.2 A manufacturing company processes 6 different jobs on two machines A and B. Number of
units of each job and its processing times on A and B are given in the following table. Find the optimum
sequence, the total minimum elapsed time and idle time for each machine.
712 Operations Research: Theory and Applications

Job Number No. of Units Processing Time (hours)

of Each Job Machine A Machine B
1 3 5 8
2 4 16 7
3 2 6 11
4 5 3 5
5 2 9 7.5
6 3 6 14

Solution Examine processing time of machines A and B. The smallest value is 3 minutes for job 4 on
machine A. Thus job 4 is scheduled first in the sequence as shown below:

Repeating this procedure, the optimal sequence so obtained is shown below. The calculations for total
elapsed time is shown in Table 20.2.

4 1 3 6 5 2

3 4 2 5 2 3  Number of units of each job.

Job Unit Number Machine A Machine B

Number of the Job Time In Time Out Time In Time Out
4 1 0 3 3 8
2 3 6 8 12
3 6 9 13 18
4 9 12 18 23
5 12 15 23 28
1 1 15 20 28 36
2 20 25 36 44
3 25 30 44 52
3 1 30 36 52 63
2 36 42 63 74
6 1 42 48 74 88
2 48 54 88 102
3 54 60 102 116
5 1 60 69 116 123.5
2 69 78 123.5 131
2 1 78 94 131 138
Table 20.2 2 94 110 138 145
Total Minimum 3 110 126 145 152
Elapsed Time 4 126 142 152 159

The total elapsed time for all the jobs including the number of units is 159 minutes. However, machine
A remained idle for 17 (= 159 – 142) minutes and machine B remained idle for 3 (3 + 152 – 152) minutes.
Example 20.3 There are seven jobs, each of which has to go through the machines A and B in the order
AB. Processing times in hours are as follows:
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. Also find T and idle time
for machines A and B. [Karn. Univ., BE (Mech.) 2000, G.J. Univ., BE 2001]
Solution The smallest processing time is 1 hour for job 6 on machine B. Thus job 6 will be processed last
on machine A as shown below:
 Sequencing Problem 713

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 minimal values: processing time of 3 hours for job 1 on machine A and processing
time of 3 hours for job 7 on machine B. According to Johnson’s rules, job 1 is scheduled first and job 7
next to 6 as shown below:

1 7 6

The reduced set of processing times becomes

Job : 2 3 4 5
Machine A : 12 15 6 10
Machine B : 10 10 6 12
Again there are two equal minimal values: processing time of 6 hours for job 4 on machine A as well
as on machine B. We may choose arbitrarily to process (schedule) job 4 next to job 1 or next to job 7 as
shown below:

1 4 7 6

or 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 jobs
2 and 3 on machine B. According to rules: job 5 is scheduled next to job 4 in the first schedule or next
to job 1 in the second schedule. Job 2 then is scheduled next to job 7 in the first schedule or next to job
4 in the second schedule. The optimal sequences are shown below:

1 4 5 3 2 7 6

or 1 5 3 2 4 7 6

The calculations for total elapsed time and idle times for machines A and B are shown in Table 20.3.

Machine A Machine B Idle for

Job Time in Time out Time in Time out Machine 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 Table 20.3
7 46 55 56 59 0 Total Elapsed
6 55 66 66 67 7 and Idle Time
714 Operations Research: Theory and Applications

The minimum elapsed time is 67 hours. Idle time for machine A is 1 hours (66th–67th hour) and for
machine B is 17 hours.
Example 20.4 A manufacturing company processes 6 different jobs on two machines A and B. Number of
units of each job and its processing times on A and B are given below. Find the optimal sequence, the total
minimum elapsed time and idle time for either machine.

Job Number Number of Units Processing Time (in Minutes)

of Each Job
Machine A Machine B

1 3 5 8
2 4 16 7
3 2 6 11
4 5 3 5
5 2 9 7.5
6 3 6 14

[Punjab Univ., BCom 2004; Karn. Univ., BE (Mech.) 2001]

Solution The smallest processing time is 3 minutes for job 4 on machine A. Thus job 4 is scheduled first on
machine A as shown below:

(i) 4

Repeating the procedure as discussed in earlier examples, we get the following sequence of job scheduling

(ii) 4 1

(iii) 4 1 3 6

(iv) 4 1 3 6 5 2
5 3 2 3 2 4  Number of units of each job

The optimal sequence as well as the number of units of each job are shown in (iv). The optimal
sequence is:
First : 5 units of job 4, Fourth : 3 units of job 6,
Second : 3 units of job 1, Fifth : 2 units of job 5,
Third : 2 units of job 3, Sixth : 4 units of job 2.
The calculations for total elapsed time are shown in Table 20.4

Job Unit Number of Machine A (in Minutes) Machine B (in Minutes) Idle Time
Number the Job of Machine
Time In Time Out Time In Time Out
(in minutes)

4 1 0 3 3 8 3
2 3 6 8 13 —
3 6 9 13 18 —
4 9 12 18 23 —
5 12 15 23 28 —

1 1 15 20 28 36 —
2 20 25 36 44 —
3 25 30 44 52 —

3 1 30 36 52 63 —
2 36 42 63 74 —
Table 20.4 6 1 42 48 74 88 —
Total Elapsed 2 48 54 88 102 —
Time 3 54 60 102 116 —
 Sequencing Problem 715

Job Unit Number of Machine A (in Minutes) Machine B (in Minutes) Idle Time
Number the Job of Machine
Time In Time Out Time In Time Out
(in minutes)

5 1 60 69 116 123.5 —
2 69 78 123.5 131 —

2 1 78 94 131 138 —
2 94 110 138 145 —
3 110 126 145 152 —
4 126 142 152 159 —

Thus, the total minimum elapsed time is 159 minutes, idle time for machine A is 17 minutes and for
machine B is 3 minutes.

CONCEPTUAL QUESTIONS A

1. Explain the four elements that characterize a sequencing problem. 7. Give a mathematical formulation of the optimal assignment and
2. Explain the principal assumptions made while dealing with travelling salesman problems. In what ways do the feasible solution
sequencing problems. matrices in the two cases differ?
3. Give Johnson’s procedure for determining an optimal sequence 8. Explain briefly, the solution procedure of processing two jobs
for processing n items on two machines. Give justification of the through the machines when the technological ordering of each
rule used in the procedure. of the jobs through the machines is prescribed in advance.
Establish the following rule: If machine A precedes machine B
4. What is no passing rule in a sequencing algorithm? Explain the
for job 1 and machine B precedes machine A for job 2, then no
principal assumptions made while dealing with sequencing
programme that contains both the decisions: (a) job 2 before job
problems. [Meerut, MSc (Maths), 2002]
1, and (b) job 1 before job 2 on machine B, is technologically
5. Give three different examples of sequencing problems from your feasible.
daily life.
9. What do you understand by the problem of sequencing? Discuss
6. Write a short note on the ‘sequencing decision problem for n the various aspects of data required to formulate the problem
jobs on two machines’. of sequencing two jobs on m machines.

SELF PRACTICE PROBLEMS A

1. We have five jobs, each of which must be processed on the two Determine the optimum sequence of jobs that minimizes the total
machines A and B, in the order AB. Processing times in hours are elapsed time to complete the jobs. Also compute the minimum time.
given in the table below: 4. The following table shows the machine time (in hours) for 5 jobs
Job : 1 2 3 4 5 to be processed on two different machines:

Machine A : 5 1 9 3 10 Job : 1 2 3 4 5
Machine B : 2 6 7 8 4 Machine A : 3 7 4 5 7
Determine a sequence for the five jobs that will minimize the Machine B : 6 2 7 3 4
elapsed time T. Passing is not allowed. Find the optimal sequence in which jobs
2. A book binder has one printing press, one binding machine, and should be processed.
manuscripts of a number of books. The time required to perform 5. Find the sequence that minimizes the total elapsed time and
the printing and binding operations on each book are shownbelow. processing time in hours required to complete the following jobs:
The binder wishes to determine the order in which the books should
be processed, so that the total time required to process all books Job : 1 2 3 4 5 6
is minimized. Machine A : 4 8 3 6 7 5
Book : 1 2 3 4 5 6 Machine B : 6 33 7 2 8 4

Printing time (Hours) : 30 120 50 20 90 110 6. Six jobs go over machine I first and then over II. The order of
Binding time (Hours) : 80 100 90 60 30 10 the completion of jobs has no significance. The following table
gives the machine times in hours for six jobs and the two machines.
[Delhi, MCom, 2000]
Job : 1 2 3 4 5 6
3. Five jobs are performed, first on machine X and then on machine
Y. The time taken, in hours by each job on each machine is given Machine I : 5 9 4 7 8 6
below: Machine II : 7 4 8 3 9 5
Job : A B C D E Find the sequence of the jobs that minimizes the total elapsed time
for completing the jobs. Find the minimum time by using Gantt
Time on machine X : 12 4 20 14 22 Chart or by any other method.
Time on machine Y : 6 14 16 18 10
716 Operations Research: Theory and Applications

7. We have five jobs, each of which must go through two machines

in the order AB. Their processing times are given below: Job Time of Turning Time of Threading
(minutes) (minutes)
Job : 1 2 3 4 5
1 3 8
Machine A : 10 2 18 6 20 2 12 10
Machine B : 4 12 14 16 18 3 5 9
4 2 6
8. A machine operator has to perform two operations, turning and 5 9 3
threading, on a number of different jobs. The time required to 6 11 1
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.

HINTS AND ANSWERS

1. Optimal sequence: 2 – 4 – 3 – 5 – 1; Elapsed time = 30 hours. Idle 5. Optimal sequence: 3 – 1 – 5 – 6 – 2 – 4; Elapsed time = 35
time for machine A = 2 hours, and for machine B = 3 hours. hours.
2. Optimal sequence: 4 – 1 – 3 – 2 – 5 – 6; Elapsed time = 430 hours. 6. Optimal sequence: 3 – 1 – 5 – 6 – 2 – 4; Elapsed time = 35
Idle time for printing machine = 10 hours, and for binding hours.
machine = 40 hours. 7. Optimal sequence: 2 – 4 – 3 – 5 –1; Elapsed time = 60 hours.
3. Optimal sequence: B – D – C – E – A; Elapsed time = 48 hours. 8. Optimal sequence: 4 – 1 – 3 – 2 – 5 – 6; Elapsed time = 43
Idle time is 12 hours for machines X and Y each. minutes. Idle time for turning operation = 1 minute, and for
4. Optimal sequence: 1 – 3 – 5 – 4 – 2; Elapsed time = 28 hours. Idle threading operation = 6 minutes.
time for machine A = 2 hours, and for machine B = 3 hours.

20.4 PROCESSING n JOBS THROUGH THREE MACHINES

An extension of Johnson’s procedure for scheduling jobs on two machines M1 and M2 in the order M1
M2 has been discussed in this section. The list of jobs with their processing times on three machines M1,
M2 and M3 is given below. An optimal solution to this problem can be obtained if either or both of the
following conditions hold good:
Processing Time Job Number
on Machine 1 2 3 .. . n
M1 t11 t12 t13 .. . t1n
M2 t21 t22 t23 .. . t2n
M3 t31 t32 t33 .. . t3n

1. The minimum processing time on machine M1 is at least as great as the maximum processing time on
machine M2, that is, min t1j  max t2j, for j = 1, 2, . . ., n
2. The minimum processing time on machine M3 is at least as great as the maximum processing time on
machine M2, that is, min t3j  max t2j, for j = 1, 2, . . ., n
If either or both the above conditions hold good, then the steps of the algorithm can be
summarized in the following steps:

20.4.1 The Procedure

Step 1: Examine the processing times of the given jobs on all three machines and if either one or both
the above conditions hold, then go to Step 2, otherwise the algorithm fails.
Step 2: Introduce two fictitious machines, say G and H with corresponding processing times given by:
(i) tGj = t1j + t2j, j = 1, 2, . . ., n
that is, the processing time on machine G is the sum of the processing times on machines M1 and M2,
and
(ii) tHj = t2j + t3j, j = 1, 2, . . ., n
that is, processing time on machine H is the sum of the processing times on machines M2 and M3.
 Sequencing Problem 717

Step 3: Determine the optimal sequence of jobs for this n-job, two machine equivalent sequencing
problem with the prescribed ordering GH in the same way as discussed earlier.
Example 20.5 Find the sequence that minimizes the total time required in performing the following jobs
on three machines in the order ABC. Processing times (in hours) are given in the following table:
Job : 1 2 3 4 5

Machine A : 8 10 6 7 11
Machine B : 5 6 2 3 4
Machine C : 4 9 8 6 5
[IAS 1995; Anantpur Univ., MSc (Maths), 2000]

Solution From data of the problem, we know that min (tAj) = 6; min (tCj) = 4; max (tBj) = 6. Since min (tAj) 
(tBj) for all j is satisfied, the given problem can be converted into a problem of 5 jobs and two machines. The
processing time on two dummy machines G and H can be determined by the following relationships:
tGj = tAj + tBj ; and tHj = tBj + tCj ; j = 1, 2, . . ., n
The processing times for the new problem are given below:
Job : 1 2 3 4 5
Machine G : 8 + 5 = 13 10 + 6 = 16 6 + 2 = 8 7 + 3 = 10 11 + 4 = 15
Machine H : 5 + 4 = 9 6 + 9 = 15 2 + 8 = 10 3 + 6 = 9 4 + 5 = 9
When the procedure described for n jobs on two machines is applied to this problem, the optimal
sequence, so obtained, is given by:

3 2 5 1 4

The total minimum elapsed time is given in Table 20.5.

Job Machine A Machine B Machine C

Sequence Time In Time Out Time In Time Out Time In Time Out

3 0 6 6 8 8 16
2 6 16 16 22 22 31
5 16 27 27 31 31 36 Table 20.5
1 27 35 35 40 40 44 Minimum Elapsed
Time
4 35 42 42 45 45 51

Table 20.5 indicates that the minimum total elapsed time is 51 hours. The idle time for machines A, B
and C is 9 (= 51 – 42) hours, 6 (= 51 – 45) hours and 9 {= (8 – 0) + (45 – 44)} hours, respectively.
Example 20.6 A readymade garments manufacturer has to process 7 items through two stages of
production, viz., cutting and sewing. The time taken for each of these items at the different stages are given
below in appropriate units:
Item : 1 2 3 4 5 6 7
Cutting : 5 7 3 4 6 7 12
Process time 
Sewing : 2 6 7 5 9 5 8
(a) Find an order in which these items are to be processed through these stages so as to minimize the
total processing time.
(b) Suppose a third stage of production is added, viz. pressing and packing, with processing time for these
items as follows:
Item : 1 2 3 4 5 6 7
Processing time
(Pressing and Packing) : 10 12 11 13 12 10 11
Find an order in which these seven items are to be processed so as to minimize the time taken to
process all the items through all the three stages.
Solution Using Johnson’s optimal sequencing procedure, the optimal sequence so obtained is shown below.
The calculations for total elapsed time is shown in Table 20.6.
718 Operations Research: Theory and Applications

3 4 5 7 2 6 1

Item Cutting Sewing Idle Time for

Time In Time Out Time In Time Out Sewing

3 0 3 3 10 3
4 3 7 10 15 —
5 7 3 15 24 —
7 13 25 25 33 1
Table 20.6 2 25 32 33 39 —
Total Minimum 6 32 39 39 44 —
Elapsed Time 1 39 44 44 46 —

The total elapsed time is 46 hours. Idle item on cutting machine is 2 (= 46 – 44) hours and on sewing
machine 4 hours.
(b) When the third stage of production (viz., pressing and packing) is added, the problem is that of
seven jobs on three machines with processing time (in hours) given below:
Item : 1 2 3 4 5 6 7
Cutting, Ci : 5 7 3 4 6 7 12
Sewing, Si : 2 6 7 5 9 5 8
Pressing, Pi : 10 12 11 13 12 10 11
Given that min (Ci) = 3, min (Pi) = 10 and max (Si) = 9. Since the condition min (Pi)  max (Si) for alli
is satisfied, the problem can be converted into that of 7 jobs and 2 machines. If G and H are two fictitious
machines such that Gi = Ci + Si and Hi = Si + Pi for all i, then the given problem can be re-written as
the problem of two machines and 7 items with their processing times given below:
Item : 1 2 3 4 5 6 7
Gi : 5+2= 7 7 + 6 = 13 3 + 7 = 10 4+5= 9 6 + 9 = 15 7 + 5 = 12 12 + 8 = 20
Hi : 2 + 10 = 12 6 + 12 = 18 7 + 11 = 18 5 + 13 = 18 9 + 12 = 21 5 + 10 = 15 8 + 11 = 19
Using the optimal sequence algorithm, the optimal sequence so obtained is as follows:

1 4 3 6 2 5 7

The total elapsed time for three processes is given in Table 20.7.

Item Cutting Sewing Printing and Packing

Time In Time Out Time In Time Out Time In Time Out
1 0 5 5 7 7 17
4 5 9 9 14 17 30
3 9 12 14 21 30 41
6 12 19 21 26 41 51
Table 20.7
2 19 26 26 32 51 63
Minimum Elapsed
5 26 32 32 41 63 75
Time
7 32 44 44 52 75 86
The minimum total elapsed time is 86 hours with idle time of 42 hours for cutting, 44 hours for sewing
and 7 hours for pressing and packing.

SELF PRACTICE PROBLEMS B

1. We have six jobs, each of which must go through machines A, Job : 1 2 3 4 5 6

B and C in the order ABC. Processing time (in hours) are given
in the following table: Machine A : 8 3 7 2 5 1
Machine B : 3 4 5 2 1 6
Machine C : 8 7 6 9 10 9
 Sequencing Problem 719

Determine a sequence for the five jobs that will minimize the Job : 1 2 3 4 5 6 7
elapsed time t.
Machine A : 12 6 5 11 5 7 6
2. Determine the optimal sequence of jobs that minimize the total Machine B : 7 8 9 4 7 8 3
elapsed time, based on the following information. Processing time Machine C : 3 4 1 5 2 3 4
on machines is given in hours and passing is not allowed.
5. A machine operator has to perform three operations, turning,
Job : A B C D E F G threading and knurling on a number of different jobs. The time
required to perform these operations (in minutes) for each job
Machine (M 1) : 3 8 7 4 9 8 7
is known and is give below:
Machine (M2) : 4 3 2 5 1 4 3
Machine (M3) : 6 7 5 11 5 6 12 Job Time for Time for Time for
3. We have five jobs, each of which must go through the machines Turning Threading Knurling
A, B and C in the order ABC. Processing times (in hours) is as (minutes) (minutes) (minutes)
follows: 1 3 8 13
Job : 1 2 3 4 5 2 12 6 14
3 5 4 9
Machine A : 5 7 6 9 5 4 2 6 12
Machine B : 2 1 4 5 3 5 9 3 8
Machine C : 3 7 5 6 7 6 11 1 13
Determine the sequence for the jobs that will minimize the total
Determine the order in which the jobs should be processed in
elapsed time.
order to minimize the total time required to turn out all the jobs.
4. Find the sequence that minimizes the total elapsed time required
to complete the following tasks. Each job is processed in the
order ABC.

HINTS AND ANSWERS

1. Optimal sequences: (i) 4 – 5 – 2 – 6 – 1 – 3, (ii) 4 – 5 – 2 – is 8 hours for machine A; 25 hours for machine B and 12 hours
6 – 1 – 3; Elapsed time = 53 hours. Idle time is 27 hours for for machine C.
machine A; 32 hours for machine B and 4 hours for machine C. 4. Optimal sequences: (i) 3 – 5 – 2 – 6 – 2 – 4 – 7, (ii) 3 – 5 – 6 – 2 – 1
2. Optimal sequences: D – G – F – B – C – E; Elapsed time = – 4 – 7; Elapsed time = 59 hours; Idle time is 7 hours for machine A;
59 hours. Idle time is 13 hours for machine M1; 37 hours for 13 hours for machine B and 37 hours for machine C.
machine M2 and 7 hours for machine M3. 5. Optimal sequences: (i) 4 – 3 – 1 – 6 – 2 – 5; Elapsed time = 77
3. Optimal sequences: (i) 2 – 5 – 4 – 3 – 1, (ii) 5 – 4 – 3 – 2 – minutes; Idle time is 35 minutes for turning operation; 49 minutes
1, (iii) 5 – 2 – 4 – 3 – 1; Elapsed time = 40 hours; Idle time for threading operation and 8 minutes for knurling operation.

20.5 PROCESSING n JOBS THROUGH m MACHINES

Let there be n jobs, each of which is to be processed through m machines, say M1, M2, . . ., Mm in the order
M1 M2 . . . Mm. The optimal solution to this problem can be obtained if either or both of the following
conditions hold good.
(a) Min {t1j}  Max {tij} ; j = 2, 3, . . ., m – 1 and/or
(b) Min {tmj}  Max {tij} ; j = 2, 3, . . ., m – 1
that is, the minimum processing time on machines M1 and Mm is as great as the maximum processing time
on any of the remaining (m – 1) machines.
If either or both these conditions hold good, then the steps of the algorithm can be summarized in the
following steps:
Step 1: Find, Min {t1j}, Min {tmj} and Max {tij} and verify the above conditions. If either or both the
conditions mentioned above hold, then go to Step 2. Otherwise the algorithm fails.
Step 2: Convert m-machine problem into 2-machine problem by introducing two fictitious machines, say
G and H with corresponding processing times given by:
(i) tGj = t1j + t2j + . . . + tm – 1j ; j = 1, 2, . . ., n
i.e. processing time of n-jobs on machine G is the sum of the processing times on machines M1, M2, . . .,
Mm–1, j
 (ii) tHj = t2j + t3j + . . . + tmj ; j = 1, 2, . . ., n
i.e. processing time of n-jobs on machine H is the sum of the processing times on machines M2, M3, . . ., Mm.
Step 3: The new processing times, so obtained, can now be used for solving n-job, two-machineequivalent
sequencing problem with the prescribed ordering HG in the same way as discussed earlier.
Remarks 1. In addition to the conditions given in Step 2, if:
t2j + t3j + . . . + tm – 1, j = k (constant)
for all j = 1, 2, . . ., m – 1, then the optimal sequence can be obtained for n-jobs and two machines M1
and Mm in the order M1Mm as usual.
2. If t1j = tmj and tGj = tHj, for all j = 1, 2, . . ., n, then the total number of optimal sequences will be n
and total minimum elapsed time in these cases would also be the same.
3. The method described above for solving n-jobs and m-machines sequencing problem is not a general
method. It is applicable only to certain problems where the minimum cost (or time) of processing the
jobs through first and/or last machine is more than or equal to the cost (or time) of processing the
jobs through the remaining machines.
Example 20.7 Find an optimal sequence for the following sequencing problems of four jobs and five
machines, when passing is not allowed. Its processing time (in hours) is given below:

Job Machine

M1 M2 M3 M4 M5

A 7 5 2 3 9
B 6 6 4 5 10
C 5 4 5 6 8
D 8 3 3 2 6

Also find the total elapsed time.

Solution Here, Min (t M 1 , j ) = 5 = t M ,C ; Min (t M ,j) = 6 = tM ,D

1 5 5
and Max {t M , j, tM , j, tM , j}= {6, 5, 6} respectively.
2 3 4

Since the condition of Min (t M 5, j )  Max t M 2 , j , t M o , j, t M4 , j t is satisfied, therefore the given

3

problem can be converted into a four jobs and two machines problem as G and H. The processing times
of four jobs denoted by tGj and tHj on G and H, respectively are as follows:
Job : A B C D
Machine G : 17 21 20 16
Machine H : 19 25 23 14
m1 m
where tGj   tij and t Hj   tij .
i 1 i2

Time

Table 20.8
Minimum Elapsed
720 Operations Research: Theory and Applications

Now, using
A C B D
the optimal
sequence The total elapsed time, corresponding to the optimal sequence can be calculated as shown in Table
algorithm, the 20.8, using the individual processing times given in the original problem.
following Table 20.8 shows that the minimum total elapsed time is 51 hours. The idle time for machines M1, M2,
optimal M3, M4 and M5 is 25, 33, 37 and 18 hrs, respectively.
sequence can
be obtained Job Sequence Machine

M1 M2 M3 M4 M5

A 0–7 7–12 12–14 14–17 17–26

C 7–12 12–16 16–21 21–27 27–35
B 12–18 18–24 24–28 28–33 35–45
D 18–26 26–29 29–32 33–35 45–51
 Sequencing Problem 721

Example 20.8 Solve the following sequencing problem, giving an optimal solution when passing is not
allowed.

Machine Job

A B C D E

M1 11 13 9 16 17
M2 4 3 5 2 6
M3 6 7 5 8 4
M4 15 8 13 9 11

[Meerut, MSc (Maths), 2001]

Solution From the data of the problem it is observed that
Min (t M , j )  9  t M , C ; Min (t M , j )  8  t M ,B
1 1 4 4

and Max {t M2 , j }  6  t M 2, E ; Max {t M 3 , j }  8  t M 3 ,D

Since both the conditions
Min (t M 1 , j }  Max { t M 2, j ; t M 3, j }

Min ( t M , j}  Max { t M , j ; tM , j} ; j  1, 2, ..., 5

4 2

3
are satisfied, therefore the given problem can be converted into a 5-jobs and 2-machine problem as G and H.
Further, it may be noted that, t M 2 , j  t M3 , j  10 (a fixed constant) for all j ( j = 1, 2, . . ., 5). Thus, the
given problem is reduced to a problem of solving 5-jobs through 2-machines M1 and M4 in the order M1M4.
This means machines M2 and M4 will have no effect on the optimality of the sequences.
The processing times of 5 jobs on machine M1 and M4 is as follows:
Job : A B C D E
Machine M1 : 11 13 9 16 17
Machine M4 : 15 8 13 9 11
Now, using the algorithm described earlier, the optimal sequence so obtained is as follows:

C A E D B

The total elapsed time corresponding to the optimal sequence is 83 hours as shown in Table 20.9, using
the individual processing times given in the original problem:

Job Machine

M1 M2 M3 M4

C 0–9 9–14 14–19 19–32

A 9–20 20–24 24–30 32–45
E 29–36 36–42 42–46 46–57 Table 20.9
D 36–52 52–54 54–62 62–71 Minimum Total
B 52–65 65–68 68–75 75–83 Elapsed Time

20.6 PROCESSING TWO JOBS THROUGH m MACHINES

Let there be two jobs A and B, each of which is to be processed on m machines say M1, M2, . . ., Mm, in
two different orders. The technological ordering of each of the two jobs through m machines is known in
advance. Such ordering may not be same for both the jobs. The exact or expected processing times on the
given machines are known. Each machine can perform only one job at a time. The objective is to determine
an optimal sequence of processing the jobs so as to minimize total elapsed time.
The optimal sequence in this case can be obtained by using a graph. The procedure can be illustrated
by taking examples.
722 Operations Research: Theory and Applications

Example 20.9 Use the graphical method to minimize the time needed to process the following jobs on the
machines shown, i.e. for each machine find the job that should be done first. Also, calculate the total elapsed
time to complete both jobs.
Machine
R
ST
Sequence: A B C D E
Job 1 3 4 2 6 2
Time (hrs)
Machine
R Sequence:
Job 2 ST Time (hrs) B
5
C
4
A
3
D
2
E
6
[Karnataka Univ., BE, 2000; Meerut Univ., MSc (Maths), 2002]

Solution The solution procedure for solving the above problem can be summarized in the following steps:
1. Draw a set of axes at right angle to each other where x-axis represents the processing time of job 1 on
different machines while job 2 remains idle and y-axis represents processing time of job 2 while job
1 remains idle.
2. Mark the processing times for jobs 1 and 2 on x-axis and y-axis, respectively, according to the given
order of machines as shown in Fig. 20.1.

Fig. 20.1
Graphical
Solution of 2-Jobs
and m-Machines
Sequencing
Problem

For example, machine A takes 3 hours for job 1 and 3 hours for job 2. Construct the rectangle for machine
A, as shown in Fig. 20.1. Similarly, construct other rectangles for machines B, C, D and E.
3. Construct various blocks starting from the origin by pairing the same machines until a point marked
‘finished’ is obtained.
4. Draw a line starting from the origin to the point marked ‘finish’ by moving horizontally, vertically and
diagonally along a line that makes an angle of 45º with the horizontal axis. Moving horizontally along this
line indicates that the first job is under process while the second job is idle. Similarly, moving vertically along
this line indicates that the second job is under process while the first job is idle. The diagonal movement
along this line shows that both the jobs are simultaneously under process.
Since simultaneous processing of both jobs on a machine is not possible, therefore, a diagonal movement
is not allowed. In other words, diagonal movement through rectangle areas is not allowed.
5. An optimal path is one that minimizes the idle time for both the jobs. Thus, we must choose the path
on which diagonal movement is maximum as shown in Fig. 20.1.
6. Total elapsed time is obtained by adding the idle time for either job to the processing time for that
job. In this example, the idle time for the chosen path is found to be 5 hrs and 2 hrs for jobs 1 and 2,
respectively. The total elapsed time is calculated as follows:
 Sequencing Problem 723

Elapsed time, Job 1 = Processing time of job 1 + Idle time for job 1
= 17 + (2 + 3) = 22 hours
Elapsed time, Job 2 = Processing time of job 2 + Idle time for job 2
= 20 + (17 – 15) = 20 hours.
Example 20.10 Using the graphical method, calculate the minimum time needed to process jobs 1 and2
on five machines A, B, C, D and E, i.e. for each machine find the job that should be done first. Also, calculate
the total time needed to complete both jobs.
Machines
RS Sequence : A B C D E
T Time (hrs)
Job 1
: 6 8 4 12 4

Job 2 RS Sequence : B C A D E
T Time (hrs) : 10 8 6 4 12
[Punjab Univ., BE (E&CE), 2006]
Solution Draw two axes at right angle to each other where x-axis represents the processing time of job 1 on
different machines while job 2 remains idle and y-axis represents the processing time of job 2 on different
machines while job 1 remains idle.

Fig. 20.2
Graphical
Solution of
2 Jobs and m-
Machines
Sequencing
Problem

Mark the processing times for both the jobs 1 and 2 on x-axis and y-axis, respectively according to the
given order of machines, as shown in Fig. 20.2. For example, machine A takes 6 hours for job 1 as well as
for job 2. Construct the rectangle for machine A as shown in Fig. 20.2. Similarly, construct other rectanglesfor
machines B, C, D and E.
Draw a line starting from the origin to the point marked ‘finished’ by moving horizontally, vertically and
diagonally along a line which makes an angle of 45º with the horizontal axis. A horizontal move represents
processing of job 1 while job 2 remains idle; a vertical move represents processing of job 2 while job 1 remains
idle. The diagonal movement (i.e. a 45º line) shows that both the jobs are under process simultaneously. Since
simultaneous processing of both the jobs on a machine is not possible, therefore, diagonal movement is not
allowed.
An optimal path is one that minimizes idle time for both jobs 1 and 2. This means that this path should
coincide with 45º line to the maximum extent. Thus, we must choose the path on which diagonal movement
is maximum, as shown in Fig. 20.2. The total elapsed time is obtained by adding idle time for either job
to the processing time for that job.
The idle time for the chosen path is found to be 10 (= 4 + 6) hours for job 1 and 4 hours for job 2.
The total elapsed time is calculated as follows:
Elapsed time, Job 1 = Processing time + Idle time = 34 + (4 + 6) = 44 hours. Elapsed
time, Job 2 = Processing time + Idle time = 40 + (32 – 28) = 44 hours.
724 Operations Research: Theory and Applications

CONCEPTUAL QUESTIONS B

1. A job consists of N steps. Step i takes time ti. If these jobs are (i) Job arrival pattern
grouped somehow into station systems, then twice as many units (ii) Number of machines
can be produced each day. Also two set-ups in parallel can double (iii) The flow pattern in the shop
the production rate. Critically examine the advantages of these two (iv) The criteria of evaluating the performance of a schedule
approaches. 3. By using appropriate notion, obtain formulae for the following:
2. What do you understand by the following terms in the context (i) Waiting time of a job (ii) Completion time of a job
of sequence of jobs: (iii) Mean flow time (iv) Mean lateness

SELF PRACTICE PROBLEMS B

1. Find an optimal sequence for the following sequencing problem Machines

of four jobs and five machines (when passing is not allowed) of R Sequence
ST Time (hrs)
: A B C D E
which processing time (in hrs) is as follows. Job 1
: 2 3 4 6 2
Job : 1 2 3 4
R Sequence : C A D E B
Machine M1 : 6 5 4 7 Job 2 ST Time (hrs) : 4 5 3 2 6
Machine M2 : 4 5 3 2
Machine M3 : 1 3 4 2 4. Two jobs are to be processed on four machines A, B, C and
Machine M4 : 2 4 5 1 D. The technological order for these jobs on machines is as
Machine M5 : 8 9 7 5 follows:
Job 1 : A B C D
Also find the total elapsed time. Job 2 : D B A C
2. Five jobs have to be processed on the same machine. The set- Processing times are given in the following table
up time for each job depends on the job processed earlier. A Machines
table of the set-up time is shown below. Find a sequence for
A B C D
processing all jobs that minimizes the total set-up costs.
Job 1 : 4 6 7 3
Predecessor Follower Job Job 2 : 4 7 5 8
Job A B C D E Find the optimal sequence of jobs on each of the machines.
A 0 29 20 18 24 5. A machine shop has four machines, A, B, C and D. Two jobs
B 0 0 14 19 16 must be processed through each of these machines. The time
C 0 35 0 37 26 (in hrs) taken on each of the machines and the necessary
D 0 15 10 0 10 sequence of jobs through the shop are given below:
E 0 18 16 40 0 Sequence : A B C D E
Job 1
Time (hrs) : 2 4 5 1 2
3. Use the graphical method to find the minimum elapsed total time Sequence : D E A C B
Job 2
sequence of 2 jobs and 5 machines, when we are given the Time (hrs) : 6 4 2 3 6
following information: Use the graphic method to obtain the total minimum elapsed time.

HINTS AND ANSWERS

1. Optimal sequence: 1 – 3 – 2 – 4; Minimum total elapsed time 4. Idle time is 4 hours for job 1 and zero hour for job 2; Elapsed
= 43 hours. time for job 1 is 20 + 4 = 24.
3. Idle time is 3 hours for job 1 and zero hour for job 2; Elapsed 5. Total elapsed time is 15 hours.
time for job 1 is 17 + 3 = 20 hours.

CHAPTER SUMMARY
The short-term schedules show an optimal order (sequence) and time in which jobs are processed. They also show timetables
for jobs, equipment, people, materials, facilities and all other resources that are needed to support the production plan. The schedules
should use the resources efficiently to give low costs and high utilizations. Other purpose of scheduling are, minimizing customers
waiting time, meeting promised delivery dates, keeping stock levels low, giving preferred working pattern, minimizing waiting
time of patients in a hospital for different types of tests, and so on.
 If there are n jobs to be performed, one at a time, on each of m machines and the actual or expected time required by the
jobs on each of the machines is also given, then the general sequencing problem is to find the sequence out of (n!)m possible
sequences, which minimize the total elapsed time between the start of the job in the first machine and the completion of the last job
on the last machine.

CHAPTER CONCEPTS QUIZ

1. If there are in jobs to be performed, one at a time, on each of (a) Min t1j = Max t2j (b) Min t1j  Max t2j
m machines, the possible sequences would be (c) Min t1j  Max t2j (d) Min t2j  Max t1j
(a) (n !)m (b) (m !)n
m
4. You would like to assign operators to the equipment that has
(c) (n) (d) (m)n
(a) most jobs waiting to be processed
2. Total elapsed time to process all jobs through two machines is (b) job with the earliest due date
given by (c) job which has been waiting longest
n M  n M n n (d) all of the above

(a)  1j  2j (b)  M 2 j   M1 j
j 1 j 1 j 1 j 1 5. Unforeseen factors that prevent the plans from actually happen-
n ing are

(c)  (M1 j  I1 j ) (d) none of the above (a) equipment may develop a fault
j 1 (b) additional order may arrive to be added to schedules
3. The minimum processing time on machine M1 and M2 are related (c) specifications may be changed
as (d) all of the above
 Sequencing Problem 725

Answers to Quiz
1. (a) 2. (b) 3. (c) 4. (d) 5. (d)