Indian Institute of Technology Madras

Department of Humanities & Social Sciences

Semester : Jul-Nov 2016
Course Name :: Decision Modelling
Course No ::Worksheet
HM4001 1: Problems on Linear Programming Formulations
Worksheet :: Linear Programming Models
1. <Product Mix>A plant that manufactures specialty metal products has recently discontinued
sales of a particular product line that was no longer profitable. This created excess production
capacity that can be used to manufacture more of the remaining lines. Management of the firm
has identified three existing profitable products and wishes to know how it should allocate the
new capacity among these three so that they enhance their profits as much as possible.
The production routing sheets for these three products all specify process on a numerical
controlled milling machine, a lathe and a grinder. The time now available on these machines is
as follows:
Table 1.1
Resource
Available Time
Type
(Machine Hours per Week)
NC machine
520
Lathe
210
Grinder
125
The routing sheets also provide the following standard time information (in machine hours per
unit)
Table 2.2
Machine
Product 1
Product 2
Product 3
NC machine
16
5
8
Lathe
10
5
1
Grinder
2
1
1
The firms marketing department forecasts that the sales potential for products 1 and 2 exceed
the maximum production rate, but the potential for product 3 is limited to 50 units per week. At
the same time, the industrial engineering department has estimated that, at the volume likely to
be achieved and the productivity levels being projected by management, the unit profit for the
products would be $ 45, $20, $ 25 respectively.
Formulate the above problem as an Linear Programming model.
<Source: Industrial Engineering & Operations Research by Miller & Schmidt, John Wiley & Sons, 1990>

2. <Blending> A refinery has four different crudes which are to be processed to yield four
products: gasoline, heating oil, jet fuel and lube oil. There are maximum limits both on product
demand (what can be sold) and crude availability. A schematic of the processing operations is
given below.

Faculty: S Seshasayee

Fig 2.1
Maximum Product Demand
(barrels/week)

Crude Availability
(barrels/week)

Gasoline
100000

100000

100000

170000
Fuel
Chain

Heating Oil 85000

Jet Fuel

200000

Lube Chain

Lube Oil

85000

20000

Given the tabulated profits, costs and yields in Table 2.1 formulate an appropriate LP model for
scheduling the refinery operations to maximize total profit.
Table 2.1
Profits, Costs & Yields
Crude 4
-----------------------Fuel
Lube
Process
Process
Gasoline
Heating Oil
Jet Fuel
Lube Oil
Others*
Crude cost $/bbl
Operating cost $/bbl

Crude 1
0.6
0.2
0
0.1
0.1
15.00
5.00

Crude 2
0.5
0.2
0.2
0
0.1
15.00
8.50

Crude 3
0.3
0.3
0.3
0
0.1
15.00
7.50

0.4
0.3
0.2
0
0.1
25.00
3.00

*Refers to losses in processing

<Source: Operations Research: Principles & Practices By Ravindran et all John Wiley 1988>

0.4
0.1
0.2
0.2
0.1
25.00
2.50

Product
Value
$/bbl
$45
$30
$15
$60

3.<Production Planning -Make Or Buy Decisions>

Case: MacNeil Chemicals
The Toronto plant of Macneil Chemicals had a small process which produced methyl amines.
These amines were used with further processing in the manufacture of plastics and fibre glass.
The amine production system was a vapour phase reaction. Ammonia and Methanol were
vapourized in the presence of a catalystto convrt them into monomethyl amine (MONO),
dimethyl amine (DI) and trimethyl amine (TRI). Any or all of the three amines and products
recycled through the process if desired. Almost all of the MONO and recycled because it has
very little market potential.
The plant manager, Brian Casey, estimated the raw material costs to be $ 0.05 for DI and $
0.06 for TRI. He considered labour and all other costs to be fixed.
The nature of the process restricted the output at the Toronto plant.
Mr Casey had attempted to increase the DI production but had been unable to increase daily
output above 8950 pounds. The combined output of DI and TRI was limited to 13520 pounds.
Serious inefficiencies resulted in the process if the daily production of TRI fell below 3750
pounds. For similar reasons DI output could not be less than TRI output.
The plastic and fiberglass manufacturing process required 12, 350 pounds of DI and 5,180
pounds of TRI each day. To meet this internal demand, Mr Casey could purchase DI and TRI
from Ridd Brothers Ltd. and Kenra Chemicals Company. Ridd Brothers produced DI and TRI
for its own use but was able to produce an excess of up to 3,450 of DI and upto 400 pounds of
TRI. Because of these small volumes, Ridd would sell either product for $ 0.1 a pound. Kenra
could supply any amount of DI and TRI for $ 0.11 and $ 0.12 respectively.
Mr Casey wondered what quantities of DI and TRI he should produce and what amounts he
should order from his two suppliers.
<Source: Management Science and the Manager by E F Peter Newson, Prentice Hall, 1977>

4. <Financial Planning: Short Term>

CASE

Racy's Departmental Stores

The treasurer of Racy's Departmental Store is performing his financial planning for the next six
months, September through February. Because of the Christmas season, Racy's had need for
large amount of cash, particularly in the months of November and December and a large cash

inflows occur in January and February where customers pay their Christmas bills. These
requirements are summarized in the table given at the end (all figures in '000 $).
The treasurer has three sources of short-term funds to meet Racy's needs. These are:
1.

Pledge Accounts Receivable:

A local bank will loan Racy's funds on a month-by-month basis against a pledge on the
accounts receivable balance at the beginning of the given month. The maximum loan is
75% in the given month. The cost of this loan is 1.5% per month of the amount
borrowed.

2.

Stretch Payment of Purchases:

Payment of purchases can be delayed one month. Thus, for example, the $100,000
planned for payment for November could be delayed until December and Racy's could
use the funds to meet November needs. When purchase payments are thus stretched,
Racy's loses the 3% discount it normally receives for prompt payment.

3.

Use short-term loan:

A bank is willing to lend Racy's any amount from $40,000 to $100,000 on a 6-month
basis. The loan would be taken out in full in the beginning of September for a fixed
amount and paid back at the end of February. It would not be possible to add to the
loan or pay off part of the loan during the period. The cost of the loan would be 1% per
month, payable each month.

In any period, if the firm has excess funds, they can be invested in short term government
securities that return 0.5% per month.
The objective of the treasurer is to minimize the net interest cost to Racy's while meeting the
firm's cash needs.
Table 4.1

Accounts receivable balance

Sep

Oct

Nov

Dec

Jan

Feb

70

50

70

120

100

50

80

90

100

60

40

50

(at the beginning of the month)

Planned payments of purchases
(on assumption that discount is taken)

Cash needs for operation

Cash surplus from operation

30

60

90

20

30

150

Note:
1.
2.

Cash needs / surplus indicates the funds in need or in surplus, as the case may be,
the NETT AMOUNT ARRIVED AFTER TAKING INTO ACCOUNT
ACCOUNTS RECEIVABLES & PLANNED PAYMENTS.
Pledging against Accounts receivables in other words means 'Factoring of
Accounting Records'
<Source: Management Science and the Manager by E F Peter Newson, Prentice Hall, 1977>

5. < Financial Planning: Long term>

CASE

MITCHELL ENTERPRISES

Mr.Gordon Mitchell, President of Mitchell Enterprises, had called a special meeting of the
company's Investment Review Committee in early December 1975.
Members of that
committee were Mr.Charles Gilbert, the Treasurer; Ms.Roberta Phillips, the Controller; and
Mr.Paul Chesler, Special Assistant to Mr.Mitchell. The committee had spent its last meeting
reviewing different methods for evaluating investment projects.
At that time, Mr.Gilbert had felt that it would be appropriate to use some method that took into
account the value of funds over the entire project life and Ms.Phillips had suggested the use of
discounting to accomplish this. Unfortunately, the committee could not agree on a hurdle rate
that reflected the company's current financial position. Indeed, they had found it hard to accept
any rate that would stay constant over the life of projects that might be considered. They had
considered using a 10% hurdle rate for after-tax cash flows, typical of practice in their industry,
but had not felt comfortable about that figure. Mr.Chesler had suggested that use of Linear
Programming they could determine a portfolio of projects and decide on the amount to be
invested in each. In addition, he knew that an LP solution would help determine what hurdle
rate or rates would be appropriate for the company. Overall, there had been much argument
and little progress because they had been discussing the methods in abstract terms and finally
Mr.Mtchell suggested that Mr.Chesler prepare a list of projects they had evaluated that year.
This list was to be circulated and used to focus the discussion at the next meeting.
Subsequently, Mr.Chesler had circulated the following memoranda.
MEMORANDA
To: Investment Review Committee
From: Paul Chesler
Subject: Hypothetical Investment Projects

December 5, 1975

As you know, Mr.Mitchell has asked me to prepare a list of typical investment projects
for us to consider in our next meeting. The table below describes five projects that might
compete for our investment dollar. The table below shows the cash flow that will result from
investing one dollar. Project A is a two-year investment available at the beginning of 1976,
which pays 30 cents per dollar invested at the end of the first year and returns an additional
dollar per dollar invested at the end of the second. Almost $500,000 can be invested in A.
Project B is identical to A except that it is available a year later. C is a one-year investment
available only at the beginning of 1976, which pays $1.10 per dollar invested at the end of that
year. Project D is a three year investment available at the beginning of 1976, which pays
$1.75 per dollar invested at the beginning of 1979. E will become available at the beginning of
1978 and will, after, pay $1.40 per dollar invested. Project E is limited to a maximum
investment of $750,000. Of course, the cash we receive from any of the projects may be
reinvested in others that are available at the time. In addition, we could obtain 6% visa shortterm bank accounts for any money not invested in a given year.
Table 5.1
Cash Flow per Dollar Invested
Year

Project
A

1976

-1.00

-1.00

-1.00

1977

+0.30

-1.00

+1.10

1978

+1.00

+0.30

-1.00

1979

+1.00

+1.75

+1.40

$500,000

$500,000

LIMIT

NONE

NONE

$750,000

For the purpose of discussion I am assuming we want to put $1,000,000 of our money into
some mix of these projects at the beginning of 1976 but no more thereafter, although we will
reinvest throw offs. All cash received on January 1 1979, will be withdrawn.
The table below gives the results of discounting these projects as we had considered in our last
meeting.
Project

Net Present value

@ 10%

Internal Rate of
Return (%)

A OR B

$0.099

16.1

$0.000

10

$0.315

20.5

$0.273

40

I would suggest that the portfolio problem be modeled in terms of Linear Program. Notice that
since there are no non-cash expenses, post-tax cash flow is strictly proportional to pre-tax flow.
For this reason we can justify using pre-tax accumulated cash as the quantity to be maximized;
this is equivalent to maximizing after-tax cash.
Discussion Questions:
What do you suggest as the optimal investment policy? Clearly indicate the rational behind
your answer.
<Source: Introduction to Management Science by Cook & Russel, Prentice Hall, 1994>

6. <Blending>A coffee manufacturer blends three component coffee beans into three blends of
coffee. Although the recipes for the three final blends are imprecise, certain restrictions must be
satisfied when combining the three components:
Component 1 should constitute no more than 30 percent of final blend 1 by
weight
Component 2 should constitute at least 20 percent of final blend 3 by weight
Component 2 and 3, combined, should constitute at least 80 percent of final
blend 2 by weight
In addition to the recipe restrictions, there is limited availability of the three components. The
maximum weekly availabilities are 60,000, 25,000 and 50,000 lbs, respectively. Weekly
capacity for the plant is 1,25,000 lbs. To satisfy the needs of a favoured customer, weekly
production should include at least 40,000 lbs of final blend 1.
Given that the three components currently cost the manufacturer is $1, $1.20 and $1.35 per
pound and the three final blends are sold at wholesale prices of $ 1.75, $2 and $ 1.90 per
pound, management wishes to determine what number of pounds of each component should be
used so as to maximize total weekly profit margin. (other weekly operating costs are excluded
from consideration)
<Source: Principles of Operations Research for Management by Frank S Budnick et all, Irwin, 1988>

7. <Resource Allocation-Manpower Planning> ORPIW Technologies is one of the recent

Indian companies who have got into the bandwagon of BPO. For one of their call center
divisions, the requirement of Customer Relations Officers during various time slots of the day
has been studied carefully and the data collected is presented below:
Time Period
06::00 hrs - 10:00 hrs
10::00 hrs - 14: 00 hrs
14::00 hrs - 18:00 hrs

Table 9.1
No of Officers Required
20
25
28

18:00 hrs - 22:00 hrs

22:00 hrs - 02:00 hrs
02::00 hrs - 06:00 hrs

18
12
22

The join duty at 06:00 hrs, 10:00 hrs, 14:00 hrs 2:00 hrs officers work on a 8 hour shift,
without any break. The management wants to know what is the least number of officers to be
employed so that the requirement of the manpower is met for each of the time slots.
<Source: Similar to the above problem can be found in may standard text books on OR.>

8. <Resource Allocation-Cargo Loading> <Also known as knapsack problem>

Four types of merchandise are available loaded into three holds of a freighter. These represent
the maximum quantities that can be shipped. Relevant data are given in Tables 10.1 & 10.2
Table 10.1
Freighter Capacities
Hold
Weight
Volume
(tons)
(cu.ft)
1. Forward
100
6000
2. Centre
125
9000
3. Aft
75
7500
Table 10.2
Cargo Data
Merchandise
1. Sugar
2. Rice
3. Ore
4. Soyabeans

Weight
(tons)
150
175
600
100

Volume
(cu ft per ton)
48.6
60.0
4.1
55.0

Revenue
($ per ton)
700
750
250
300

The captain needs to know how much weight of each type of merchandise is to be loaded in
each hold such that the total revenue is maximized without violating weight and volume
constraints. Moreover, the merchandise must be loaded such that the trim of the ship is
preserved. This means that the ratio of loaded in a hold to the weight capacity in that must be
identical for all the three holds. Formulate this problem as an LP model.
<Source: Principles of Operations Research for Management by Frank S Budnick et all, Irwin, 1988>

9. <Cutting Stock or Trim Loss>

<This classical LP scenario includes such applications as paper slitting, cable cutting, photo film cutting
and textile cutting, whereby jumbo reels are cut into smaller reels having various widths>

A paper mill produces jumbo reels of paper 60inches wide. The company receives orders for
reels 12 inches wide, 15 inches wide, 18 inches wide and 25 inches wide. The manufacturer
has received orders for 300, 250, 200 and 150 reels, respectively, of 12-, 15-, 18- and 25- inch
reels. The firm wishes to determine how to meet these orders so as to minimize total waste.
Waste is defined as any leftover portions of a jumbo reel that cannot be used to met demand.
There are two sources of waste: trim loss and surplus. For example, if a jumbo reel is slit into

two 25-inch reels, there will be leftover paper (trim loss) having a width of 10 inches. Since
there is no use for 10-inch reels (as there is no demand for it) , the 10 inches is a measure of
waste. Note that this cutting pattern yields two 25-inch reels for each jumbo reel that is cut.
Because different cutting patterns can create multiple reels out of jumbo reels, there is
likelihood that surplus or excess will be cut. For example, if a 210 12-inch reels result from the
cutting process and only 200 are required, then (210-200) x 12 is a measure of surplus waste.
(Hint: The following table gives you the list of all the cutting patterns and the associated trim loss)
Patterns of 60-Inch Reels
Required
Width
(inches)

1
12
0
15
0
18
0
25
2
Trim Loss 10

Cutting Pattern

2
0
1
1
1
2

3
1
0
1
1
5

4
0
2
0
1
5

5
1
1
0
1
8

6
2
0
0
1
11

7
0
0
3
0
6

8
0
1
2
0
9

9
2
0
2
0
0

10
1
2
1
0
0

11
2
1
1
0
3

12
3
0
1
0
6

13
0
4
0
0
0

14
1
3
0
0
3

<Source: Principles of Operations Research for Management by Frank S Budnick et all, Irwin, 1988>

15
2
2
0
0
6

16
3
1
0
0
9

17
5
0
0
0
0

10. A dietician wishes to design a minimum-cost diet to meet minimum daily requirements for
calories, protein, carbohydrate, fat, vitamin A and vitamin B dietary needs. Several different
foods can be used in the diet, with data as specified in the following table.
Content and costs per pound consumed

Food
1

Food
2

Food
3

a11

a12

a13

a21

a22

Carbohydrate
(grams)

a31

Fat
(grams)

Calories
Proteins
(grams)

Food
n
a1n

b1

a23

a2n

b2

a32

a33

a3n

b3

a41

a42

a43

a4n

b4

Vitamin A
(milligrams)

a51

a52

a53

a5n

b5

Vitamin B
(milligrams
)

a61

a62

a63

a6n

b6

Costs
($)

c1

c2

c3

Daily
Requirement

cn

Formulate the above minimum-cost diet problem as an LP.

{The above problem is known as the classical Diet Problem which has its origin. Pl refer to

http://www.stiglerdiet.com/category/or-example/}11.
Case

Controlling Air Pollution At Noori & Leets

The Nori & Leets Co., one of the major producers of steel in its part of the world, is located in
the city of Steeltown and is the only large employer there. Steeltown has grown and prospered
along with the company, which now employs nearly 50,000 residents. Therefore, the attitude of
the townspeople always has been "What's good for Nori & Leets is good for the town."
However, this attitude is now changing; uncontrolled air pollution from the company's furnaces
is ruining the appearance of the city and endangering the health of its residents.
A recent stockholders' revolt resulted in the election of a new enlightened board of directors for
the company. These directors are determined to follow socially responsible poliviers, and they
have been discussing with Steeltown city officials and citizens' groups what to do about the air
pollution problem. Together they have worked out stringent air quality standards for the
Steeltown airshed.
The three main types of pollutants in this airshed are particulate matter, sulfur oxides, and
hydrocarbons. The new standards require that the company reduce its annual emission of these
pollutants by the amounts shown in the Table 1. The board of directors has instructed
management to have the engineering staff determine how to achieve these reductions in the
most economical way.
Table 1
Clean Air Standards for Nori & Leets Co.
Required Reduction in Annual Emission Rate Pollutant (Million pounds)
Particulates 60
Sulfur Oxides 150
Hydrocarbons 125

The steelworks has two primary sources of pollution, namely, the blast furnaces for making
pig iron and the open-hearth furnaces for changing iron into steel. In both cases the engineers
have decided that the most effective types of abatements methods are (1) increasing the height
of the smokestacks, (2) using filter devices (including gas traps) in the smokestacks, and (3)
including cleaner, high-grade materials among the fuels for the furnaces. Each of these
methods have technological limits on how much heavily it can be used (eg maximum feasible
increase in the height of the smokestacks) but there is also considerable flexibility for using
fraction the method at a fraction of the of its technological limit.

Table 2 shows how emission can be eliminated from each type of furnace (in millions of
pounds per year) by fully using any type of abatement method its technological limit.
Table 2
Reduction in Emission Rate from Maximum Feasible Use of Abatement Method for Nori &
Leets Co.

Taller Smokestacks

Pollutant

Blast

OpenHearth

Furnaces

Filters

Blast

Better Fuels

OpenHearth

Furnaces
Furnaces

Blast

OpenHearth

Furnaces
Furnaces

Furnaces

Particulates

12

25

20

17

13

Sulfur Oxides

35

42

18

31

56

49

Hydrocarbons

37

53

28

24

29

20

For the purposes of analysis it is assumed that each method can be used less fully to achieve
any fraction (including zero) of the emission rate as shown in this table. Furthermore, the
fractions can be different for blast furnaces and open-hearth furnaces. For either type of
furnace, the emission reduction achieved by each method is not substantially affected by
whether or not the other methods are also used.
After these data were developed, it became clear that no single method by itself could achieve
all the required reductions. On the other hand, combining all three methods at full capacity on
both types of furnaces (which would be prohibitively expensive if the company's products are
to remain competitively priced) is much more than adequate. Therefore, the engineers
concluded that they would have to use some combination of the methods, perhaps with
fractional capacities, based upon their relative costs. Furthermore, because of the differences
between the blast and the open-hearth furnaces, the two types probably should not use the same
combination.
An analysis was conducted to estimate the total annual cost that would be incurred by each
abatement method. In addition to increased operating and maintenance expenses, consideration
was given also to the initial costs (converted to an equivalent annual basis) of the method as

well as any resulting loss in efficiency of the production process. This analysis led to the total
cost estimates (in millions of dollars) given in Table 3 for using the methods at their full
abatement capacities. It also was determined that the cost of a method being used at a lower
level is essentially proportional to its fractional capacity. Thus, for any given fraction used, the
total annual cost would be that fraction of the corresponding quantity in Table 3.

Table 3
Total Annual Cost from Maximum Feasible Use of Abatement Method for Nori & Leets
Co.

Abatement
Method

Blast

Openhearth

Furnaces
Furnaces

Taller
Smokestacks

10

11

Filters
Better Fuels

The stage was now set to develop the general framework of the company's plan for pollution
abatement. This plan would consist of specifying which types of abatement methods would be
used and at what fractions of their abatement capacities for (1) the blast furnaces, and (2) the
open-hearth furnaces. Because of the combinatorial nature of the problem of finding a plan that
satisfies the requirements with the smallest possible cost, is to be formulated and solved.

12.

Acompanywantsahighlevel,aggregateproductionplanforthenext6months.Projected
ordersforthecompany'sproductarelistedinthetable.Overthe6monthperiod,unitsmaybe
producedinonemonthandstoredininventorytomeetsomelatermonth'sdemand.Becauseof
seasonalfactors,thecostofproductionisnotconstant,asshowninthetable.
The cost of holding an item in inventory for 1 month is $4/unit/mo. Items produced and
sold in the same month are not put in inventory.
The maximum number of units that can be held in inventory is 250.
The initial inventory level at the beginning of the planning horizon is 200 units;
the final inventory level at the end of the planning horizon is to be 100.
The problem is to determine the optimal amount to produce in each month so that demand is
met while minimizing the total cost of production and inventory. Shortages are not permitted.

Aggregateplanningdata
Deman
d
Month

(units)

1
2
3
4
5
6

1300
1400
1000
800
1700
1900

Productio
n
cost
($/unit)
100
105
110
115
110
110

13
Case: Optimal Fuel Loading
An important problem with which all airplanes is faced is the determination of how much an
aircraft should load at each stop. Since the efficiency of an aircraft is related to its weight, an
aircraft fully loaded with fuel burns more fuel to take them to them to their next stop.
Unfortunately, since fuel prices vary quite substantially from one location to another, this policy
of minimal fill-ups may be more costly than filling the tank at very inexpensive locations. Airlines
need some way of achieving the appropriate balance, especially since at current fuel prices, fuel
costs are, by far, the largest single item in airlines costs.
For the following case find the optimal fuel loading at each airport.
Flight schedule:
Los Angeles Tampa Miami Fort Lauderdale New York Miami Houston Los
Angeles (back to base)
Fuel requirements for each leg and cost at each location:

City

Fuel
Cost/gallon

Los Angeles

$ 1.236

Tampa

Minimum Fuel
(gallons) needed to
get into next stop

Additional fuel (0.01

gallons) burnt per gallon of
fuel loaded over the
minimum

330

4.604

1.276

42

0.585

Miami

1.299

0.070

Fort Lauderdale

1.204

151

2.105

New York

1.269

151

2.105

Fort Lauderdale

1.204

0.070

Miami

1.024

186

2.593

Houston

1.299

177

2.468

Note: The last column means the data is expressed as one-hundredth of a gallon.