285

B
Q U A N T I T A T I V E M O D U L E
Linear Programming
DISCUSSION QUESTIONS
1. Students may select from eight LP applications given in the
introduction. These include school bus scheduling, police patrol
allocation, scheduling bank tellers, selecting product mix, picking
blends to minimize cost, minimizing shipping cost, developing
production schedules, and allocating space.
2. LP theory states that the optimum lies on a corner. All three
solution techniques make use of the corner point feature.
3. The feasible region is the area bounded by the set of problem
constraints. A feasible solution is any combination of x, y coordi-
nates (or x
1
, x
2
coordinates) that is in or on the feasible region.
4. Each LP problem that has been formulated correctly does
have an infinite number of possible solutions. Any point within
the feasible region is a solution that satisfies all constraints (al-
though it is not necessarily optimal). In addition, for any problem
in which the optimal solution lies on a constraint that is parallel to
the objective function, all points along that constraint are also
both feasible and optimal.
5. The objective function contains the profit or cost information
that enables us to determine whether one solution is better than an-
other solution. Our choice of best depends only on the objective.
6. Before activity values can be placed into the objective, they
must meet the constraints. Notice that the objective function has no
minimum-required profit level unless it is included as a constraint.
7. As long as the costs do not change, the diet problem always
provides the same answer. In other words, the diet is the same
every day. Unlike animals, people enjoy variety, and variety can-
not be included as a linear constraint.
8. The number of feasible solutions is infinite. We only need to
consider extreme pointscorner pointsto find the optimal solu-
tion. If we use isoprofit lines, we only need to examine one corner
point to determine the optimal solution.
9. Shadow price or dual: the value of one additional unit of a
resource, such as one more hour of a scarce labor resource or one
more dollar to invest.
10. The iso-cost line is moved down in a minimization problem
until it no longer intersects with any constraint equation. The last
point in the feasible region that the line touches is the optimal
corner point.
11. The corner point method examines the profit at every corner
point, whereas the iso-profit line method draws a series of parallel
profit lines until one line finally touches the last tip (corner point)
of the feasible region. That last point touched is the optimal solu-
tion, so other corner points need not be tested.
12. When two constraints do not cross at an axis, we use simul-
taneous equationsthere is only one point where two linear equa-
tions (constraints) cross.
13. (a) Adding a new constraint will reduce the size of the feasi-
ble region unless it is a redundant constraint. It can never
make the feasible region any larger.
(b) A new constraint can only reduce the size of the feasible
region; therefore the value of the objective function will
either decrease or remain the same. If the original solu-
tion is still feasible, it will remain the optimal solution.
ACTIVE MODEL EXERCISE
ACTIVE MODEL B.1: LP Graph
1. By how much does the profit on Walkmans need to rise to
make it the only product manufactured?
If the profit per walkman is more than $10 per unit then
it is the only product that should be manufactured.
2. By how much does the profit on Walkmans need to fall to
stop manufacturing it?
At $6.66 and below we should not manufacture any
Walkmans.
3. What happens to the profit as the number of assembly hours
increases by 1 hour at a time? For how many hours does this
hold true?
The profit rises by $.50 per hour until we reach 120 hours
at which point the rise stops.
4. What happens if we can reduce the electronics time for
Watch-TVs to 2.5 hours?
The profit rises by $70.
286 QUANTITATIVE MODULE B LI NEAR PROGRAMMI NG
END-OF-MODULE PROBLEMS
X Y Z = 4X + 6Y
0 0 0
4 0 16
0 4 24
1.33 3.33 25.33 (optimal)
Feasible corner points (x,y): (0,3), (0,10), (2.4,8.8), (6.75,3).
Maximum profit 100 at (0,10).
B.3 (a) The optimal solution of $26 profit lies at the point
X
1
= 2, X
2
= 3.
(b) If the first constraint is altered to 1X
1
+ 3X
2
s 8, the fea-
sible region and optimal solution shift considerably, as
follows.
B.1
B.2
QUANTITATIVE MODULE B LI NEAR PROGRAMMI NG 287
B.4 (a) Corner points (0,50), (50,50), (0,200), (75,75), (50,150).
(b) Optimal solutions: (75,75) and (50,150). Both yield
profit of $3,000.
B.5
X Y Z = 24X + 15Y
0 20 300
11 0 264
3.86 4.54 160.86 (optimal)
B.6 (a) Let x
1
= number of liver flavored biscuits in a package
x
2
= number of chicken flavored biscuits in a package
Minimize x
1
+ 2x
2
Subject to x
1
+ x
2
> 40
2x
1
+ 4x
2
> 60
x
1
s 15
x
1
, x
2
> 0
(b) Corner points are (0,40) and (15,25). Optimal solution is
(15,25) with cost of 65.
(c) Minimum cost = 65 cents.
288 QUANTITATIVE MODULE B LI NEAR PROGRAMMI NG
B.7
Let x
1
= number of air conditioners to be produced
x
2
= number of fans to be produced
Maximize 25x
1
+ 15x
2
Subject to 3x
1
+ 2x
2
s 240 (wiring)
2x
1
+ 1x
2
s 140 (drilling)
x
1
, x
2
> 0 (nonnegativity)
Profit:
@a: (x
1
= 0, x
2
= 0) Obj = $0
@b: (x
1
= 0, x
2
= 120) Obj = 25 0 + 15 120 = $1,800
@c: (x
1
= 40, x
2
= 60) Obj = 25 40 + 15 60 = $1,900*
@d: (x
1
= 70, x
2
= 0) Obj= 25 70 + 15 0 = $1,750
* The optimal solution is to produce 40 air conditioners and
60 fans each period. Profit will be $1,900.
B.8
Let x
1
= number of Model A tubs produced
x
2
= number of Model B tubs produced
Maximize 90x
1
+ 70x
2
Subject to 125x
1
+ 100x
2
s 25,000 (steel)
20x
1
+ 30x
2
s 6,000 (zinc)
x
1
, x
2
> 0 (nonnegativity)
Profit:
@a: (x
1
= 0, x
2
= 200) Obj = 90 0 + 70 200
= $14,000.00
@b: (x
1
= 85.71, x
2
= 142.86) Obj = 90 85.71 + 70 142.86
= $17,714.10
@c: (x
1
= 200, x
2
= 0) Obj = 90 200 + 70 0
= $18,000.00*
* The optimal solution is to produce 200 Model A tubs, and 0
Model B tubs. Profit will be $18,000.
B.9 (a) X
1
= number of mattresses
X
2
= number of box springs
Minimize cost = 20X
1
+ 24X
2
subject to X
1
+ X
2
> 30
X
1
+ 2X
2
> 40
X
1
= 20, X
2
= 10, cost = $640
B.10
Let x
1
= number of Alpha-4 computers
x
2
= number of Beta-5 computers
Maximize: 1200x
1
+ 1800x
2
Subject to 20x
1
+ 25x
2
= 800 (total hours)
x
1
> 10 (Alpha-4s)
x
2
> 15 (Beta-5s)
x
1
, x
2
> 0 (nonnegativity)
Profit:
@ a: (x
1
= 10, x
2
= 24) Obj = 1200 10 + 1800 24
= $55,200*
@ b: (x
1
= 21.25, x
2
= 15) Obj = 1200 21.25 + 1800 15
= $52,500
* The optimal solution is to produce 10 Alpha-4 and 24 Beta-5
computers per period. Profit is $55,200.
(b)
QUANTITATIVE MODULE B LI NEAR PROGRAMMI NG 289
B.11 Let: X
1
= number of pounds of compost in each bag
X
2
= number of pounds of sewage waste in each bag
Minimize cost = 5X
1
+ 4X
2
(in cents)
Subject to X
1
+ X
2
> 60 (pounds per bag)
X
1
> 30 (pounds compost per bag)
X
2
s 40 (pounds sewage per bag)
Corner point a:
(X
1
= 30, X
2
= 40) cost = 5(30) + (4)(40) = $3.10
Corner point b (which is optimal):
(X
1
= 30, X
2
= 30) cost = 5(30) + (4)(30) = $2.70
Corner point c:
(X
1
= 60, X
2
= 0) cost = 5(60) + (4)(0) = $3.00
B.12
The optimal point, a, lies at the intersection of the
constraints:
3x
1
+ 2x
2
> 120
x
1
+ 3x
2
> 90
To solve these equations simultaneously, begin by writing
them in the form shown below:
3x
1
+ 2x
2
= 120
x
1
+ 3x
2
= 90
Multiply the second equation by 3, and add it to the first:
3x
1
+ 2x
2
= 120 3x
1
+ 2x
2
= 120
3(x
1
+ 3x
2
= 90) 3x
1
9x
2
= 270
7x
2
= 150
Therefore, x
2
= 150/7 = 21.43. Given: 3x
1
+ 2x
2
= 120 then
3x
1
= 120 2x
2
= 120 2 21.43
and
= =
1
77.14
25.71
3
x
Thus, the optimal solution is: x
1
= 25.71, x
2
= 21.43
The cost is given by:
C = x
1
+ 2x
2
= 25.71 + 2 21.43 = $68.57
B.13 The last constraint is not linear because it contains the
square root of x and the objective function and first constraint are
not because of the x
1
x
2
term.
B.14 (a) Using software, we find that the optimal solution is:
x
1
= 7.95
x
2
= 5.95
x
3
= 12.60
Profit = $143.76
(b) There is no unused time available on any of the three
machines.
(c) An additional hour of time on the third machine would
be worth $0.26.
(d) Additional time on the second machine would be
worth $0.786 per hour for a total of $7.86 for the addi-
tional 10 hours.
B.15 (a)
290 QUANTITATIVE MODULE B LI NEAR PROGRAMMI NG
(c) If X
1
s profit coefficient was overestimated, but should
only have been $1.25, it is easy to see graphically that
the solution at point b remains optimal.
B.16 (a) Let X
ij
number of students bused from sector i to
school j. Objective:
subject to
700 (number students in sector )
500 (number students in sector )
100 (number students in sector )
800 (number students in sector )
400(number s
AB AC AE
BB BC BE
CB CC CE
DB DC DE
EB EC EE
X X X
A
X X X
B
X X X
C
X X X
D
X X X

tudents in sector )
900 (school capacity)
900 (school capacity)
900 (school capacity)
AB BB CB DB EB
AC BC CC DC EC
AE BE CE DE EE
E
X X X X X
B
X X X X X
C
X X X X X
E

d

d

d
(b) Solution: X
AB
400
X
AE
300
X
BB
500
X
CC
100
X
DC
800
X
EE
400
Distance 5,400 student miles
X
1
$ invested in Treasury notes
X
2
$ invested in bonds
Maximize ROI 0.08X
1
0.09X
2
X
1
t $125,000
X
2
d $100,000
X
1
X
2
$250,000
X
1
, X
2
t 0
Point a (X
1
150,000, X
2
100,000),
ROI $21,000 (optimal solution)
Point b (X
1
250,000, X
2
0), ROI $20,000
(b)
B.17
minimize total travel miles = 5 8 6
0 4 12
4 0 7
7 2 5
12 7 0
AB AC AE
BB BC BE
CB CC C E
DB DC DE
EB EC EE
X X X
X X X
X X X
X X X
X X X

QUANTITATIVE MODULE B LI NEAR PROGRAMMI NG 291
B.18 Problem Data Solution

Period

Time Period
Workers
Required
Hire
Solution 1
Hire
Solution 2
Hire
Solution 3
1 3 AM7 AM 3 0 3 3
2 7 AM11 AM 12 16 9 14
3 11 AM3 PM 16 0 7 2
4 3 PM7 PM 9 9 2 7
5 7 PM11 PM 11 2 9 4
6 11 PM3 AM 4 3 0 0
_ = 30 _ = 30 _ = 30
Let x
i
= number of workers reporting for the start of work
in period i, where i = 1, 2, 3, 4, 5, or 6. The equations
become:
Objective:
x
1
+ x
2
+ x
3
+ x
4
+ x
5
+ x
6
(Minimize staff size)
Subject to:
1 2
2 3
3 4
4 5
5 6
1 6
1 2 3 4 5 6
12
16
9
11
4
3
, , , , , 0
x x
x x
x x
x x
x x
x x
x x x x x x
+ >
+ >
+ >
+ >
+ >
+ >
>
Note that three alternate optimal solutions are provided to
this problem. All solutions could be implemented using
only 30 staff members.
B.19 (a) Let X
1
= wren houses
X
2
= bluebird houses
Maximize Profit = 6X
1
+ 15X
2
4X
1
+ 2X
2
s 60
4X
1
+ 12X
2
s 120
Corner Points
X1 X2 Profit
0 0 0
15 0 90
0 10 150
12 6 162 (Optimal)
The maximum value of the objective is $162, obtained by
producing 12 wren houses and 6 bluebird houses.
B.20 The original equations are:
Objective: 9x
1
+ 12x
2
(maximize)
Subject to: x
1
+ x
2
s 10 (gallons, varnish)
x
1
+ 2x
2
s 12 (lengths, redwood)
where: x
1
= number of coffee tables/week
x
2
= number of bookcases/week
Optimal: x
1
= 8, x
2
= 2, Profit = $96
B.21 (a) Using the isoprofit line or corner point method, we see
that point b (where X
1
= 37.5 and X
2
= 75) is optimal if
the profit = $3X
1
+ $2X
2
.
(b) If the profit changes to $4.50 per unit of X
1
, the opti-
mal solution shifts to point c.
(c) If the objective function becomes P = $3X
1
+ $3X
2
, the
corner point b remains optimal.
B.22
(b)
292 QUANTITATIVE MODULE B LI NEAR PROGRAMMI NG
The original equations are:
Objective: 4x
1
5x
2
(minimize)
Subject to: x
1
2x
2
t 80
3x
1
x
2
t 75
The optimal solution is found at the intersection of the two
constraints:
x
1
2x
2
t 80
3x
1
x
2
t 75
To solve these equations simultaneously, begin by writing
them in the form shown below:
x
1
2x
2
80
3x
1
x
2
75
Multiply the second equation by 2 and add it to the first:
1 2 1 2
1 2 1 2
2 80 2 80
2(3 75) 6 2 150
5 70 z
x x x x
x x x x
x
o
o

Thus, x
1
70/5 14. Given: x
1
2x
2
80,
2x
2
80 x
1
80 14 or x
2
66/2 33.
The cost is given by:
C 4x
1
5x
2
4 u 14 5 u 33 221
B.23 Let x
1
the number of class A containers to be used
x
2
the number of class K containers to be used
x
3
the number of class T containers to be used
The appropriate equations are:
Maximize: 9x
1
7x
2
15x
3
Subject to: 2x
1
x
2
3x
3
d 130 (Material)
2x
1
6x
2
4x
3
240 (Time)
x
1
, x
2
, x
3
t 0 (nonnegativity)
Using software we find that the optimal solution is:
x
1
0, x
2
14.29, x
3
38.57
and
Profit $678.57
B.24 (a) The unit profit of the air conditioner must fall in the
range $22.50$30.00
(b) The shadow price for the wiring constraint is $5.00,
and it holds within the range 210280 hours.
Let: X
1
number of TV spots
X
2
number of newspaper ads
Maximize exposures 35,000X
1
20,000X
2
2
1
1
1
2
Subject to: 3,000 1, 250 $100,000
5
25
10
X X
X
X
X
d
t
d
t
1 2
1 2
1 2
Point ( 5, 10), exposure = 375,000
Point ( 5, 68), exposure = 175,000 + 1,360,000
= 1,535,000 (this is optimal)
Point ( 25, 20), exp
a X X
b X X
c X X



1 2
osure 875, 000 400, 000
1, 275,000
Point ( 25, 10), exposure 875,000 + 200,000
1,075,
d X X

000
B.26 (a) Minimize:
1 1 1 2 2 2 3 3 3
6 5 3 8 10 8 11 14 18
a b c a b c a b c
x x x x x x x x x
1 2 3
1 2 3
1 2 3
1 1 1
2 2 2
3 3 3
Subject to: 7
12
5
6
8
10
All variables 0
a a a
b b b
c c c
a b c
a b c
a b c
x x x
x x x
x x x
x x x
x x x
x x x



d
d
d
t
(b) Solution:
1
2
2
3
3
6
3
5
7
3
Cost $219
b
b
c
a
b
x
x
x
x
x

An alternate solution, at the same $219 ($219,000) cost, is

1
1
2
3
3
1
5
8
7
3
b
c
b
a
b
x
x
x
x
x

B.27 a) V1 Fertilizer shipped to Customer A from Warehouse W1

V2 Fertilizer shipped to Customer A from Warehouse W2
V3 Fertilizer shipped to Customer A from Warehouse W3
V4 Fertilizer shipped to Customer B from Warehouse W1
V5 Fertilizer shipped to Customer B from Warehouse W2
V6 Fertilizer shipped to Customer B from Warehouse W3
b) Z 7.5V1 6.25V2 6.5V3 6.75V4 7V5 8V6
B.25
QUANTITATIVE MODULE B LI NEAR PROGRAMMI NG 293
d) How many of the constraints are binding? 4 (tell by non-
zero shadow prices)
e) How much slack/surplus is there with the non-binding
constraint(s)? 50 tons (600550)
f) The range of optimality on Variable V3? $6.25 to $7.25
(to get this look at Allowable Increase (.75) and De-
crease (.25) for V3. Add/subtract from C3 (6.5): 6.25
6.5 .25 & 7.25 6.5 .75. If we could ship 10 tons
less to Customer A, we might be able to save: $65
(shadow price $6.5/ton)
g) Cust B Shadow price is higher for filling Customer Bs
demand $7.25 > $6.5 and we are trying to minimize
costs.
B.28 Let x
1
number of medical patients
x
2
number of surgical patients
The appropriate equations are:
Maximize 2280x
1
1515x
2
1 2
1 2
1 2
2
1
Subject to: 8 5 32,850 (patient days available)
3.1 2.6 15,000 (lab tests)
1 2 7,000 (X rays)
2,800 (operations)

x x
x x
x x
x
x
d
d
d
d
2
, 0 (nonnegativity) x t
Optimal: x
1
2790, x
2
2104, Profit $9,551,659
or: 2790 medical patients, 2104 surgical patients,
Profit $9,551,659
Beds required:
Use: Medical: 8 2790 22, 320
Surgical: 5 2104 10,520
32,840
u
u
Medical uses:
22, 320
68% 61 beds
32,840
o
Surgical uses:
10,520
32% 29 beds
32,840
o
Here is an alternative approach that solves directly for the
number of beds:
Maximize revenues 104,025x
1
110,595x
2
d
d
d
d
d
1 2
1 2
1 2
1 2
2
Subject to: 90 beds
8 5 32,850 (patients/yr)
141.44 189.8 15, 000 (lab tests)
45.63 146 7,000 (x-rays)
73 2,800 (operations)
x x
x x
x x
x x
x
where x
1
no. of medical beds 61.17
x
2
no. of surgical beds 28.83
Revenue is $9,551,659, as before.
B.29 (a) Let X
1
no. of units of internal modems produced
per week
X
2
no. of units of external modems produced per week
X
3
no. of units of circuit boards produced per week
X
4
no. of units of floppy disk drives produced per week
X
5
no. of units of hard drives produced per week
X
6
no. of units of memory boards produced per week
Objective function analysis: First find the time used on
each test device:
hours on test device 1
1 2 3 4 5 6
7 3 12 6 18 17
60
X X X X X X

hours on test device 2

1 2 3 4 5 6
2 5 3 2 15 17
60
X X X X X X

hours on test device 3

1 2 3 4 5 6
5 1 3 2 9 2
60
X X X X X X

Thus, the objective function is

maximize contribution per unit revenue material
cost test cost
1 2 3 4 5 6
1 2 3 4 5 6
1 2 3 4 5 6
1 2 3 4 5 6
1 2 3 4 5 6
200 120 180 130 430 260
35 25 40 45 170 60
7 3 12 6 18 17
15
60
2 5 3 2 15 17
12
60
5 1 3 2 9 2
18
60
X X X X X X
X X X X X X
X X X X X X
X X X X X X
X X X X X X

This can be rewritten as

1 2 3
4 5 6
maximize contribution
= $161.35 92.95 135.50
per unit
82.50 249.80 191.75
X X X
X X X


subject to
1 2 3 4 5 6
1 2 3 4 5 6
1 2 3 4 5 6
7 3 12 6 18 17
120 hours
60
2 5 3 2 15 17
120 hours
60
5 1 3 2 9 2
100 hours
60
X X X X X X
X X X X X X
X X X X X X

d

d

d
c) -description- variables and coefficients? what type? RHS?
C1: Cust As demand formula: V1 V2 V3 >| 650 (either > or okay)
C2: Cust Bs demand formula: V4 V5 V6 >| 800 (either > or okay)
C3: W1s Supply formula: V1 V4 < 400
C4: W2s Supply formula: V2 V5 < 500
C5: W3s Supply formula: V3 V6 < 600
294 QUANTITATIVE MODULE B LI NEAR PROGRAMMI NG
= = = + =
= = = + =
= = = + =
1 2
1 2
1 2
@ : ( 0, 100) Obj 9 0 20 100 $2, 000.00
@ : ( 262.5, 25) Obj 9 262.5 20 25 $2,862.50*
@ : ( 300, 0) Obj 9 300 20 0 $2, 700.00
a x x
b x x
c x x
(b) The solution is
X
1
= 496.55 internal modems
X
2
= 1,241.38 external modems
X
3
through X
6
= 0
Profit = $195,504.80
(c) The shadow prices, as explained in Module B, for addi-
tional time on the three test devices are $21.41, $5.75,
and $0, respectively, per minute.
B.30 (a) Minimize total cost = $0.60X
1
+ 2.35X
2
+ 1.15X
3
+ 2.25X
4
+ 0.58X
5
+ 1.17X
6
+ 0.33X
7
1 2 3 4 5 6 7
1 2 3 4 5 6 7
1 2 3 4 5 6 7
1 2 3 4 5 6 7
1 2
subject to
295 1,216 394 358 128 118 279
1,500
295 1,216 394 358 128 118 279
900
0.2 0.2 4.3 3.2 3.2 14.1 2.2 4
16 96 9 0.5 0.8 1.4 0.5 50
16 81
X X X X X X X
X X X X X X X
X X X X X X X
X X X X X X X
X X
+ + + + + +

+ + + + + +

+ + + + + +
+ + + + + +
+ +
3 4 5 6 7
1 5 6 7
74 83 7 14 8 26
22 28 19 63 50
All 0
i
X X X X X
X X X X
X
+ + + +
+ + +

(b) Problem solved by computer:

The meal plan for the evening is
No milk (X
1
= 0)
0.499 pound of ground meat (X
2
)
0.173 pound of chicken (X
3
)
No fish (X
4
= 0)
No beans (X
5
= 0)
0.105 pound of spinach (X
6
)
0.762 pound of white potatoes (X
7
)
Each meal has a cost of $1.75.
(c) The meal is fairly well-balanced (two meats, a green vege-
table, and a potato). The weight of each item is realistic.
(d) This problem is very sensitive to changing food prices.
Sensitivity analysis when prices change:
Milk increases 10 cents/lb: no change in price or diet
Milk decreases 10 cents/lb: no change in price or diet
Milk decreases 30 cents/lb (to 30 cents): potatoes drop
out and milk enters, price = $1.42/meal
Ground meat increases from $2.35 to $2.75: price = $1.93
and spinach leaves the optimal solution
Ground meat increases to $5.25/lb: price = $2.07 and
meat leaves; milk, chicken and potatoes in solution
Fish decreases from $2.25 to $2.00/lb: no change
Chicken increases to $3.00/lb: price = $1.91 and meat,
fish, spinach, and potatoes in solution.
INTERNET HOMEWORK PROBLEMS
These problems (B.31B.40) appear on our companion
web site at www.prenhall.com/heizer
B.31 Let x = number of standard model to produce
y = number of deluxe model to produce
Maximize 40x + 60y
Subject to 30 30 450
10 15 180
6
, 0
x y
x y
x
x y
+
+

Feasible corner points (x,y): (6,0), (15,0), (6,8), (9,6). Max-

imum profit is $720 by producing either 6 standard and
8 deluxe or 9 standard and 6 deluxe.
B.33
Let x
1
= number of benches produced
x
2
= number of tables produced
Maximize 9x
1
+ 20x
2
Subject to 4x
1
+ 6x
2
1,200 (hours)
10x
1
+ 35x
2
3,500 (board-feet)
x
1
, x
2
0 (nonnegativity)
Profit:
* The optimal solution is to make 262.5 benches and 25 tables per
period. Profit will be $2,862.50. Because benches and tables may
be matched (two benches per table), it may not be reasonable to
maximize profit in this manner. Also, this problem brings up the
proper interpretation of the statement that One should make 262.5
(a fractional quantity) benches per period.
B.32
QUANTITATIVE MODULE B LI NEAR PROGRAMMI NG 295
Note that this problem has one constraint with a negative
sign. This may cause the beginning student some confu-
sion in plotting the constraint.
This would probably be a good time to stress the plot-
ting of equations of the form:
x
1
2x
2
s 10
found in this problem, and of the form:
3x
1
2x
2
s 0
The optimal point, a, lies at the intersection of the
constraints:
3x
1
+ 5x
2
s 150
5x
1
+ 3x
2
s 150
To solve these equations simultaneously, begin by writing
them in the form shown below:
3x
1
+ 5x
2
= 150
5x
1
+ 3x
2
= 150
Multiply the first equation by 5, the second by 3, and add
the two equations.
+ = + =
+ = =
=
1 2 1 2
1 2 1 2
2
5 (3 5 150) 15 25 750
3 (5 3 150) 15 9 450
16 300
x x x x
x x x x
x
Therefore, x
2
= 300/16 = 18.75.
Given: 3x
1
+ 5x
2
= 150 then
3x
1
= 150 5x
2
= 150 5 18.75
and
1
56.25
18.75
3
x = =
Thus, the optimal solution is: x
1
= 18.75, x
2
= 18.75
The profit is given by:
C = 4x
1
+ 4x
2
= 4 18.75 + 4 18.75 = $150
B.35 Because the decision centers about the production of the
two different cabinet models, let:
x
1
= number of French Provincial cabinets produced
per day
x
2
= number of Danish Modern cabinets produced each day
The equations become:
Objective: 28x
1
+ 25x
2
(Maximize revenue)
1 2
1 2
1 2
1
2
1 2
Subject to 3 2 360 (hours, carpentry)
1.5 1 200 (hours, painting)
0.75 0.75 125 (hours, finishing)
60 (units, contract)
60 (units, contract)
, 0 (nonnegativity)
x x
x x
x x
x
x
x x
+ s
+ s
+ s
>
>
>
The solution is:
x
1
= 60, x
2
= 90, Revenue = $3930/day
Define the following variables:
x
1
= thousands of round tables produced per month
x
2
= thousands of square tables produced per month
The appropriate equations then become:
Objective: 10x
1
+ 8x
2
(minimize handling and storage costs)
Subject to x
2
> 5 (square tabletop contract)
2x
1
+ 1x
2
> 20 (total labor capacity)
x
1
, x
2
> 0 (nonnegativity)
Cost:
= = = + =
= = = + =
1 2
1 2
@ : ( 7.5, 5) Obj 10 7.5 8 5 $115*
@ : ( 0, 20) Obj 10 0 8 20 $160
a x x
b x x
* The optimal solution is to produce 7500 round tables and 5000
square tables, for a cost of $115,000.
B.34
B.36
B.37
296 QUANTITATIVE MODULE B LI NEAR PROGRAMMI NG
The original equations are:
Objective: 3x
1
+ 5x
2
(maximize)
Subject to: x
2
6
3x
1
+ 2x
2
18
x
1
, x
2
0 (nonnegativity)
The optimal solution is found at the intersection of the two
constraints. Solving for the values of x
1
and x
2
at the inter-
section, we have:
x
2
= 6
2
1
18 2 18 2 6 6
2
3 3 3
x
x

= = = =
Profit = 3x
1
+ 5x
2
= 3 2 + 5 6 = 6 + 30 = $36
B.38 Let x
1
= number of newspaper ads placed
x
2
= number of TV spots purchased
Given that we are to minimize cost, we may develop the
following set of equations:
Minimize: 925x
1
+ 2000x
2
Subject to: 0.04x
1
+ 0.05x
2
0.40 (city exposure)
0.03x
1
+ 0.03x
2
0.60 (exposure in
NW suburbs)
Note that the problem is not limited to unduplicated expo-
sure (for example, one person seeing the Sunday newspa-
per three weeks in a row counts for three exposures).
Solution:
x
1
= 20 ads, x
2
= 0 TV spots, cost = $18,500
Maximize: 57x
1
+ 55x
2
Subject to: x
1
+ x
2
390
2.5x
1
+ 2.4x
2
960
= = = +
=
= = = +
=
= = = +
=
1 2
1 2
1 2
@( 384, 0) Obj 57 384 55 0
$21,888
@( 0, 390) Obj 57 0 55 390
$21, 450
@( 240, 150) Obj 57 240 55 150
$21,930*
x x
x x
x x
* The optimal solution occurs at x
1
= 240 boys bikes,
x
2
= 150 girls bikes, producing a profit of $21,930.
You should point out that the possible optimal solutions
are so close in profit that sensitivity analysis becomes an
important issue.
B.40 (a) maximize: 18.79x
1
+ 6.31x
2
+ 8.19x
3
+ 45.88x
4
+ 63x
5
+ 4.1x
6
+ 81.15x
7
+ 50.06x
8
+ 12.79x
9
+ 15.88x
10
+ 17.91x
11
+ 49.99x
12
+ 24x
13
+ 88.88x
14
+ 77.01x
15
const 1: 4x
2
+ 6x
3
+ 10x
4
+ 12x
5
+ 10x
7
+ 5x
8
+ 1x
9
+ 1x
10
+ 2x
12
+ 10x
14
+ 10x
15
980
const 2: 0.4x
1
+ 0.5x
2
+ 0.4x
4
+ 1.2x
5
+ 1.4x
6
+ 1.4x
7
+ 1x
8
+ 0.4x
9
+ 0.3x
10
+ 0.2x
11
+ 1.8x
12
+ 2.7x
13
+ 1.1x
14
400
const 3: 0.7x
1
+ 1.8x
2
+ 1.5x
3
+ 2x
4
+ 1.2x
5
+ 1.5x
6
+ 7x
7
+ 5x
8
+ 1.5x
12
+ 5x
13
+ 5.8x
14
+ 6.2x
15
600
const 4: 5.8x
1
+ 10.3x
2
+ 1.1x
3
+ 8.1x
5
+ 7.1x
6
+ 6.2x
7
+ 7.3x
8
+ 10x
9
+ 11x
10
+ 12.5x
11
+ 13.1x
12
+ 15x
15
2500
const 5: 10.9x
1
+ 2x
2
+ 2.3x
3
+ 4.9x
5
+ 10x
6
+ 11.1x
7
+ 12.4x
8
+ 5.2x
9
+ 6.1x
10
+ 7.7x
11
+ 5x
12
+ 2.1x
13
+ 1x
15
1800
const 6: 3.1x
1
+ 1x
2
+ 1.2x
3
+ 4.8x
4
+ 5.5x
5
+ 0.8x
6
+ 9.1x
7
+ 4.8x
8
+ 1.9x
9
+ 1.4x
10
+ 1x
11
+ 5.1x
12
+ 3.1x
13
+ 7.7x
14
+ 6.6x
15
1000
const 7: 1x
1
0
const 8: 1x
2
20
const 9: 1x
3
10
const 10: 1x
4
10
const 11: 1x
5
0
const 12: 1x
6
20
const 13: 1x
7
10
const 14: 1x
8
20
const 15: 1x
9
50
const 16: 1x
10
20
const 17: 1x
11
20
const 18: 1x
12
10
const 19: 1x
13
20
const 20: 1x
14
10
const 21: 1x
15
10
Solution Value = 9683.229
Optimal Reduced Original Lower Upper
Value Cost Coefficient Limit Limit
x1 0.00 1.32373 18.79 Infinity 20.11373
x2 20.00 0.00 6.31 Infinity 52.49662
x3 10.00 0.00 8.19 Infinity 34.64475
x4 10.00 0.00 45.88 Infinity 48.41532
x5 11.507 0.00 63.00 59.01673 77.38298
x6 20.00 0.00 4.10 Infinity 31.47013
x7 10.00 0.00 81.15 Infinity 115.1911
x8 20.00 0.00 50.06 Infinity 82.73582
x9 50.00 0.00 12.79 Infinity 24.539
x10 20.00 0.00 15.88 Infinity 26.7216
x11 20.00 0.00 17.91 Infinity 27.28386
x12 54.945 0.00 49.99 47.34072 56.43208
x13 20.00 0.00 24.00 Infinity 53.243
x14 12.202 0.00 88.88 80.82936 126.4764
x15 10.00 0.00 77.01 Infinity 125.8801
B.39
QUANTITATIVE MODULE B LI NEAR PROGRAMMI NG 297
Solution Value = 9683.229
Shadow Slack or Original Lower Upper
Prices Surplus RHS Limit Limit
const 1 2.711812 0.00 980.00 861.5504 1024.236
const 2 0.00 113.866 400.00 286.1337 Infinity
const 3 10.6486 0.00 600.00 587.7851 608.5712
const 4 2.182708 0.00 2500.00 1889.72 2534.683
const 5 0.00 258.885 1800.00 1541.115 Infinity
const 6 0.00 8.52954 1000.00 991.4705 Infinity
const 7 0.00 0.00 0.00 Infinity 0.00
const 8 46.1866 0.00 20.00 17.91737 41.84552
const 9 26.4548 0.00 10.00 5.041353 19.9601
const 10 2.53532 0.00 10.00 0.00 16.993
const 11 0.00 11.5072 0.00 Infinity 11.50722
const 12 27.37 0.00 20.00 16.50255 37.096
const 13 34.041 10.00 10.00 3.532913 12.01538
const 14 32.6758 0.00 20.00 17.09391 23.00434
const 15 11.75 0.00 50.00 39.20661 116.4478
const 16 10.8416 0.00 20.00 14.30611 79.923
const 17 9.37385 0.00 20.00 15.88757 68.822
const 18 0.00 44.95 10.00 Infinity 54.94591
const 19 29.243 0.00 20.00 15.45261 22.44298
const 20 0.00 2.20215 10.00 Infinity 12.20215
const 21 48.87 0.00 10.00 8.355577 12.84913
The optimal solution provides a profit of $9683. Note that
only product A158 is not produced.
(b) The shadow prices are given in the table above.
(c) There is no value to adding more workers because those
available are not yet fully occupied.
(d) Two tons of steel at a total cost of $8,000 implies a cost per
pound of $2.00. Forty-four pounds of it should be pur-
chased because the shadow price is $2.71 and the upper
limit is 1,024 pounds.
(e) Change coefficient for variable x
14
in objective function
from 88.88 to 8.88. The following results:
Solution Value = 8865.5
Optimal Reduced Original Lower Upper
Value Cost Coefficient Limit Limit
x1 0.00 1.23911 18.79 Infinity 20.02911
x2 20.00 0.00 6.31 Infinity 51.68507
x3 10.00 0.00 8.19 Infinity 32.86478
x4 16.993 0.00 45.88 38.60427 60.25521
x5 7.056 0.00 63.00 42.61333 71.26741
x6 20.00 0.00 4.10 Infinity 30.43315
x7 10.00 0.00 81.15 Infinity 106.3944
x8 20.00 0.00 50.06 Infinity 76.83481
x9 50.00 0.00 12.79 Infinity 26.18145
x10 20.00 0.00 15.88 Infinity 28.52473
x11 20.00 0.00 17.91 Infinity 29.29111
x12 57.697 0.00 49.99 47.574 91.778
x13 20.00 0.00 24.00 Infinity 45.986
x14 10.00 0.00 8.88 Infinity 80.82936
x15 10.00 0.00 77.01 Infinity 119.6576
298 QUANTITATIVE MODULE B LI NEAR PROGRAMMI NG
Solution Value = 8865.5
Shadow Slack or Original Lower Upper
Prices Surplus RHS Limit Limit
const 1 2.74856 0.00 980.00 913.6641 993.1374
const 2 0.00 113.879 400.00 286.1211 Infinity
const 3 9.197201 0.00 600.00 587.7851 601.577
const 4 2.343288 0.00 2500.00 2342.00 2512.443
const 5 0.00 266.934 1800.00 1533.066 Infinity
const 6 0.00 2.36523 1000.00 997.6348 Infinity
const 7 0.00 0.00 0.00 Infinity 0.00
const 8 45.3751 0.00 20.00 19.45971 41.84552
const 9 24.6748 0.00 10.00 8.988791 19.9601
const 10 0.00 6.993 10.00 Infinity 16.993
const 11 0.00 7.05643 0.00 Infinity 7.056433
const 12 26.3331 0.00 20.00 19.15507 37.096
const 13 25.2444 0.00 10.00 9.459686 12.01538
const 14 26.7748 0.00 20.00 19.5257 23.00434
const 15 13.3914 0.00 50.00 39.20661 62.76064
const 16 12.6447 0.00 20.00 17.28464 31.80706
const 17 11.3811 0.00 20.00 18.28127 32.64
const 18 0.00 47.70 10.00 Infinity 57.69793
const 19 21.986 0.00 20.00 19.46232 22.44298
const 20 71.9494 0.00 10.00 9.155032 12.20215
const 21 42.6476 0.00 10.00 9.67822 12.84913
Note that the profit declines to $8,865, with the reduction
in contribution to $8.88.
(f) Constraints 7 through 11 become: x
1
t 0, x
2
t 0, x
3
t 0,
x
4
t 0, x
5
t 0. The following results:
Solution Value = 9380.23
Optimal Reduced Original Lower Upper
Value Cost Coefficient Limit Limit
x1 0.00 7.90441 18.79 Infinity 26.69441
x2 0.00 16.81 6.31 Infinity 23.1219
x3 0.00 10.9491 8.19 Infinity 19.1391
x4 0.00 2.75734 45.88 Infinity 48.63734
x5 28.72255 0.00 63.00 61.75618 63.859
x6 20.00 0.00 4.10 Infinity 12.95034
x7 10.00 0.00 81.15 Infinity 86.86531
x8 37.51722 0.00 50.06 49.69948 71.07961
x9 50.00 0.00 12.79 Infinity 23.18852
x10 20.00 0.00 15.88 Infinity 20.73238
x11 33.94098 0.00 17.91 17.22904 18.570
x12 37.485 0.00 49.99 48.67592 51.016
x13 20.00 0.00 24.00 Infinity 24.49456
x14 10.00 0.00 8.88 Infinity 70.86956
x15 10.27741 0.00 77.01 75.18908 77.47366
QUANTITATIVE MODULE B LI NEAR PROGRAMMI NG 299
Solution Value = 9380.234
Shadow Slack or Original Lower Upper
Prices Surplus RHS Limit Limit
const 1 1.494825 0.00 980.00 969.9414 1202.002
const 2 0.00 120.755 400.00 279.2448 Infinity
const 3 0.7247843 0.00 600.00 598.0171 811.0541
const 4 0.8810187 0.00 2500.00 2492.973 2917.931
const 5 0.0234673 0.00 1800.00 1530.888 1805.481
const 6 6.716568 0.00 1000.00 918.2866 1002.674
const 7 0.00 0.00 0.00 Infinity 0.00
const 8 0.00 0.00 0.00 Infinity 0.00
const 9 0.00 0.00 0.00 Infinity 0.00
const 10 0.00 0.00 0.00 Infinity 0.00
const 11 0.00 28.7226 0.00 Infinity 28.72255
const 12 8.85034 0.00 20.00 17.19764 40.10845
const 13 5.71531 0.00 10.00 0.00 25.09986
const 14 0.00 17.5172 20.00 Infinity 37.51723
const 15 10.3985 0.00 50.00 42.69018 75.98374
const 16 4.85238 0.00 20.00 0.00 38.00887
const 17 0.00 13.94 20.00 Infinity 33.94098
const 18 0.00 27.4846 10.00 Infinity 37.485
const 19 0.494562 0.00 20.00 1.392963 21.02138
const 20 61.9896 0.00 10.00 0.7036638 10.96196
const 21 0.00 0.2774 10.00 Infinity 10.27741
1
Profit increases to $9,380, and none of the products
beginning with AD remain.
Previously, only A158 was not produced.
CASE STUDY
GOLDING LANDSCAPE AND PLANTS, INC.
Let x
1
pounds of C - 30
x
2
pounds of C - 92
x
3
pounds of D - 21
x
4
pounds of E - 11
Given that we are to produce a 50-pound bag, we may develop the
following set of equations:
Maximize: 12x
1
9x
2
11x
3
4x
4
1 2 3 4
4
1 2
2 3
Subject to: 50.0 (50 pounds)
7.5 (E - 11)
22.5 (C - 92, C - 30)
15.0 (D
x x x x
x
x x
x x

t
t
d
1 2 3 4
- 21, C - 92)
, , , 0 (nonnegativity) x x x x t
Note that if we wish to change the size of the bag of fertilizer, we
must change all four constraints to reflect the new size.
A more general set of equations can be developed:
Maximize: 12x
1
9x
2
11x
3
4x
4
1 2 3 4
4 1 2 3 4
1 2 1 2 3 4
Subject to: 50 (50 pounds)
0.15( ) (E - 11)
0.45( ) (C - 92,

x x x x
x x x x x
x x x x x x

t
t
2 3 1 2 3 4
C - 30)
0.30( ) (D - 21,

x x x x x x d
1 2 3 4
C - 92)
, , , 0 (nonnegativity) x x x x t
These equations can be rewritten as:
Maximize: 12x
1
9x
2
11x
3
4x
4
1 2 3 4
1 2 3 4
1 2 3 4
1 2 3 4
Subject to: 50.0 (50 pounds)
0.15 0.15 0.15 0.85 0 (E - 11)
0.55 0.55 0.45 0.45 0 (C - 92, C - 30)
0.30 0.70 0.70 0.30 0 (D - 21, C - 92)

x x x x
x x x x
x x x x
x x x x

d
d
t
1 2 3 4
, , , 0 (nonnegativity) x x x x t
Now, if the size of the bag is changed, only the first constraint
must be modified.
The solution requires 4 iterations:

1
2
3
4
Optimal solution: 7.5 (C - 30)
15.0 (C - 92)
0.0 (D - 21)
27.5 (E - 11)
Cost $3.35
x
x
x
x
INTERNET CASE STUDIES*
MEXICANA WIRE WORKS
1. Maximize P = 34W75c 30W33c 60W5x 25W7x
d
d
d
d
d
d
d
d
t
t
Subject to: 1 75 1400
1 33 250
1 5 1510
1 7 1116
1 75 2 33 0 5 1 7 4, 000
1 75 1 33 4 5 1 7 4,200
1 75 3 33 0 5 0 7 2, 000
1 75 0 33 3 5 2 7 2, 300
1 75 150
1 7 600
W c
W c
W x
W x
W c W c W x W x
W c W c W x W x
W c W c W x W x
W c W c W x W x
W c
W x
*Solutions to cases that appear on our companion web site
(www.prenhall.com/heizer).
300 QUANTITATIVE MODULE B LI NEAR PROGRAMMI NG
2
3
Solution:
Produce: 1100 units of W75cbackorder 300 units
250 units of W33cbackorder 0 units
0 units of W5xbackorder 1,510 units
600 units of W7xbackorder 516 units
Maximized profit will be $59,900.
By addressing quality problems listed earlier, we could increase
our capacity by up to 3 percent reducing our backorder level.
2. Bringing in temporary workers in the Drawing Department
would not help. Drawing is not a binding constraint. However, if
these former employees could do rework, we could reduce our
rework inventory and fill some of our back orders, thereby in-
creasing profits. We have about a third of a months output in
rework inventory. Expediting the rework process would also free
up valuable cash.
3. The plant layout is not optimum. When we install the new
equipment, an opportunity for improving the layout could arise.
Exchanging the locations of Packaging and Extrusion would cre-
ate a better flow of our main product. Also, as we improve our
quality and reduce our rework inventory, we could capture some
of the space now used for rework storage and processing and put
it to productive use.
Our machine utilization of 63% is quite low. Most manufac-
turers strive for at least an 85% machine utilization. If we could
determine the cause(s) of this poor utilization, we might find a key
to a dramatic increase in capacity.
COASTAL STATES CHEMICAL
This case illustrates the use of linear programming in production
planning.
1. To develop the model:
Let: x
1
= tons of phosphoric acid produced per day
x
2
= tons of urea produced per day
x
3
= tons of ammonium phosphate produced per day
x
4
= tons of ammonium nitrate produced per day
x
5
= tons of chlorine produced per day
x
6
= tons of caustic soda produced per day
x
7
= tons of vinyl chloride monomer produced per day
x
8
= tons of hydrofluoric acid produced per day
The appropriate model equations then become:
Maximize: 60x
1
+ 80x
2
+ 90x
3
+ 100x
4
+ 50x
5
+ 50x
6
+ 65x
7
+ 70x
8
Subject to: Demand Constraints:
1
2
3
4
5
6
7
8
320
200
270
300
480
600
300
320
x
x
x
x
x
x
x
x
s
s
s
s
s
s
s
s
Supply constraint (with no curtailment):
5.5x
1
+ 7x
2
+ 8x
3
+ 10x
4
+ 15x
5
+ 16x
6
+ 12x
7
+ 11x
8
s 36,000
(a) Supply constraint (20 percent gas curtailment):
5.5x
1
+ 7x
2
+ 8x
3
+ 10x
4
+ 15x
5
+ 16x
6
+ 12x
7
+ 11x
8
s 28,800
(Note: 36000 cu. ft. 10
3
0.80 = 28,800 cu. ft. 10
3
)
(b) Supply constraints (40 percent gas curtailment):
5.5x
1
+ 7x
2
+ 8x
3
+ 10x
4
+ 15x
5
+ 16x
6
+ 12x
7
+ 11x
8
s 21,600
(Note: 36000 cu. ft. 10
3
0.60 = 21,600 cu. ft. 10
3
)
With a 20 percent natural gas curtailment, the optimal pro-
duction schedule, in tons/day, is:
1
2
3
4
5
6
7
8
320
200
270
300
Profit: $174,650
480
385
300
320
x
x
x
x
x
x
x
x
=

`
=

=
)
With a 40 percent natural gas curtailment, the optimal
production schedule, in tons/day, is:
1
2
3
4
5
6
7
8
320
200
270
300
Profit: $151,933.50
410.67
0
300
320
x
x
x
x
x
x
x
x
=

`
=

=
)
2. Obviously, those products that have high energy consumption
factors must undergo extensive scrutiny to conserve energy. These
products include chlorine (15.0) and caustic soda (16.0). Energy
consumption is high for these chemicals because they are pro-
duced by electrolytic cell operation. The caustic soda also requires
extensive steam heated evaporation beyond production at the cells.
Energy conservation measures include elimination of steam leaks
as well as electrical ground and proper insulation of equipment.
3. These products are all produced by large-volume, capital-
intensive plants. Emergency shutdowns often result in loss of raw
materials, pollution, potential personnel hazards, and equipment
damage. These plants are staffed for normal operations, and
emergency rate reductions often require operator manpower be-
yond normal staffing. Consequently, these plants may be out of
control during unplanned and drastic rate reductions.
4. Normal profit: $185,400/day
Profit with a 20 percent curtailment: $174,650/day
Profit with a 40 percent curtailment: $151,933/day
CHASE MANHATTAN BANK
This scheduling problem can be solved most expeditiously using
linear programming. Let F denote the number of full-time employ-
ees. Some number, F1, of them will work one hour of overtime
between 5 PM and 6 PM each day and some number, F2, of the full-
time employees will work overtime between 6 PM and 7 PM. There
will be seven sets of part-time employees who begin their work
day at hour j, j = 1, 2, . . . ,7, with P1 being the number of workers
beginning at 9 AM, P2 at 10 AM, . . . , P7 at 3 PM. Note that because
part-time employees must work a minimum of four hours, none
QUANTITATIVE MODULE B LI NEAR PROGRAMMI NG 301
can start after 3 PM because the entire operation ends at 7 PM.
Similarly, some number of part-time employees, Qj, leave at the
end of hour j, j = 4, 5, . . . , 9.
The workforce requirements for the first two hours, 9 AM and
10 AM, are:
F + P1 14
F + P1 + P2 25
At 11 AM half of the full-time employees go to lunch; the re-
maining half go at noon. For those hours:
0.5F + P1 + P2 + P3 26
0.5F + P1 + P2 + P3 + P4 38
Starting at 1 PM, some of the part-time employees begin to
leave. For the remainder of the straight-time day:
1 2 3 4 5 4 55
1 2 3 4 5 6 4 5 60
1 2 3 4 5 6 7 4 5 6 51
1 2 3 4 5 6 7 4 5 6 7 29
F P P P P P Q
F P P P P P P Q Q
F P P P P P P P Q Q Q
F P P P P P P P Q Q Q Q
+ + + + +
+ + + + + +
+ + + + + + +
+ + + + + + +
For the two overtime hours:
1 2 3 4 5 6
7 4 5 6 7 8 14
1 2 3 4 5 6
7 4 5 6 7 8 9 9
F P P P P P P
P Q Q Q Q Q
F P P P P P P
P Q Q Q Q Q Q
+ + + + + +
+
+ + + + + +
+
If the left-hand sides of these 10 constraints are added, one
finds that 7F hours of full-time labor are used in straight time
(although 8F are paid for), F1 + F2 full-time labor hours are used
and paid for at overtime rates, and the total number of part-time
hours is:
10 1 9 2 8 3 7 4 6 5 5 6 4 7
6 4 5 5 4 6 3 7 2 8 9 128.4
P P P P P P P
Q Q Q Q Q Q
+ + + + + +

which is 40% of the days total requirement of 321 person hours.
This also leads to the objective function. The total daily labor
cost, which must be minimized, is:
8(10.11) 8.08( 1 2)
7.82(10 1 9 2 8 3 7 4 6 5 5 6 4 7
6 4 5 5 4 6 3 7 2 8 9)
Z F F F
P P P P P P P
Q Q Q Q Q Q
= + +
+ + + + + + +

Total overtime for a full-time employee is restricted to five
hours or less, an average of one hour or less per day per employee.
Thus, the number of overtime hours worked per day cannot exceed
the number of full-time employees:
F1 + F2 F
Because part-time employees must work at least four hours
per day,
Q4 P1
for those leaving at the end of the fourth hour. At the end of the
fifth hour, those leaving must be drawn from the P1 Q4 remain-
ing plus the P2 that arrived at the start of the second hour:
Q5 P1 + P2 Q4
Similarly, for the remainder of the day:
6 1 2 3 4 5
7 1 2 3 4 4 5 6
8 1 2 3 4 5 4 5 6 7
9 1 2 3 4 5 6 4 5 6 7 8
Q P P P Q Q
Q P P P P Q Q Q
Q P P P P P Q Q Q Q
Q P P P P P P Q Q Q Q Q
+ +
+ + +
+ + + +
+ + + + +
To ensure that all part-timers who began at 9 AM do not work
more than seven hours:
Q4 + Q5 + Q6 + Q7 P1
Similarly:
Q4 + Q5 + Q6 + Q7 + Q8 P1 + P2
Q4 + Q5 + Q6 + Q7 + Q8 + Q9 P1 + P2 + P3
Finally, to ensure that all part-time employees leave at some time:
P1 + P2 + P3 + P4 + P5 + P6 + P7 = Q4 + Q5 + Q6 + Q7
+ Q8 + Q9
The resulting problem has 16 variables and 22 constraints. If inte-
ger programming software with sufficient capacity is not avail-
able, the linear programming problem can be solved and the solu-
tion rounded, making certain that none of the constraints have
been violated. Note that the integer programming solution might
also need to be adjustedif F is an odd integer, 0.5F will not be
an integer and the requirement that half of the full-time employ-
ees go to lunch at 11 AM and the other half at noon will have to be
altered by assigning the extra employee to the appropriate hour.
1. The least-cost solution requires 29 full-time employees, 9 of
whom work two hours of overtime per day. In actuality, 18 of the
full-time employees would work overtime on two different days and
9 would work overtime on one day. Fourteen of the full-time work-
ers would take lunch at 11 AM and the other 15 would take it at
noon. Eleven part-timers would begin at 11 AM with 9 of them leav-
ing at 3 PM and the other 2 at 4 PM. Fifteen part-time employees
would work from noon until 4 PM and 5 would work from 2 PM until
6 PM. The resulting cost of 232 hours of straight time, 18 hours of
overtime, and 126 hours of part-time work is $3,476.28 per day.
This solution is not uniqueother work assignments can be
found that result in this same cost.
2. The same staffing would be used every day. In fact, one
would expect different patterns to present themselves on different
days; for example, Fridays are usually much busier bank days
than the others. In addition, the person hours required for each
hour of the day are assumed to be deterministic. In a real situation,
wide fluctuations will be experienced in a stochastic manner.
The optimal solution results in a considerable amount of idle
time, partly caused by the restriction that employees can start at the
beginning of an hour and leave at the end. Eliminating this restriction
might yield better results at the risk of increasing the problem size.