Graphical Sensitivity Analysis

Luis A. Moncayo–Martı́nez

Problems based on the web page of J. E. Beasley from Brunel University (UK)

1 Two mines
The Two Mines Company own two different mines that produce an ore which, after being crushed,
is graded into three classes: high, medium and low-grade. The company has contracted to provide a
smelting plant with 12 tons of high-grade, 8 tons of medium-grade and 24 tons of low-grade ore per
week. The two mines have different operating characteristics as detailed below.

Production (ton/day)
Mine Cost per Day (x1000) High Medium Low
1 180 6 3 4
2 160 1 1 6

How many days per week should each mine be operated to fulfil the smelting plant contract?

Solution:
xi = the number of days mine i works, i = x, y

12
A
11

10
Min z = 180x1 + 160x2
9 Feasible region
subject to E
8
6x1 + x2 ≥ 12
3x1 + x2 ≥ 8 7
4x1 + 6x2 ≥ 24
x2

x1 , x2 ≥ 0 5
C
Point G is the optimum 4
H
solution which lies on the
3 D G
intersection of constraints
3x1 + x2 ≥ 8 and 2
4x1 + 6x2 ≥ 24.
1 z
O B F I
0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6
x1

Changes in the objective function’s coefficients

If constraint 3x1 + x2 ≥ 8 replaces the objective function z = 180x1 + 160x2 (i.e. z = 3x1 + x2 ), is
the optimum solution on point G? The answer is yes. if z = 4x1 + 6x2 , the answer is the same (verify
this). Thus, it is necessary to compute the slope of two constraints as follow:

1
  
dx2 d 3 3
For constrain 3x1 + x2 ≥ 8, the slope is = 8 − x1 = − and for the another is
  dx1 dx1 1 1
dx2 d 24 4 4
= − x1 = − .
dx1 dx1 6 6 6
Therefore

if z = 180x1 + c2 x2 ,
if z = c1 x1 + 160x2 , dx2 d

z 180

180
= − x1 = − .
dx2 d  z c1  c1 dx1 dx1 c2 c2 c2
= − x1 = − .
dx1 dx1 160 160 160
3 180 4
− ≤− ≤−
3 c1 4 1 c2 6
− ≤− ≤−
1 160 6 1 c2 6
− ≥− ≥−

3 c1 4
 3 180 4
− ≤− ≤− (−160)
1 160 6 
1 c2 6

− ≥− ≥− (−180)
3 180 4
480 ≥ c1 ≥ 106.66
60 ≤ c2 ≤ 270

Changes in the RHS (Dual prices)

The optimum value is in the intersection of constraints 3x1 + x2 ≥ 8 and 4x1 + 6x2 ≥ 24; thus, we
can move all the constraints (one at a time parallel) until the optimum solution STILL lies on the
intersection of these two constraints. Hence:

3x1 + x2 ≥ 8 moves from D(1.5, 3) to I(6, 0).

bD = 3(1.5) + 3 = 7.5
Constraint 6x1 + x2 ≥ 12 can move from −∞ to
G(1.71, 2.86). bI = 3(6) + 0 = 18

bG = 6(1.33) + 2.86 = 13.14 • 7.5 ≤ b2 ≤ 18

• − inf ≤ b1 ≤ 13.14 zD = 180(1.5) + 160(3) = 750

• D1 = 0 zF = 180(6) + 160(0) = 1080

1080 − 750
• D2 = = 31.42
18 − 7.5
4x1 +6x2 ≥ 24 moves from C(1.33, 4) to F (2.66, 0).

bC = 4(1.33) + 6(4) = 29.33

bF = 4(2.66) + 6(0) = 10.64

• 10.64 ≤ b3 ≤ 29.33

zC = 180(1.33) + 160(4) = 879.4

zF = 180(2.66) + 160(0) = 478.8

879.4 − 478.8
• D3 = = 21.43
29.33 − 10.64

2 A company
A company makes two products (x and y) using two machines (A and B). Each unit of x that
is produced requires 50 minutes processing time on machine A and 30 minutes processing time on

2
machine B. Each unit of y that is produced requires 24 minutes processing time on machine A and 33
minutes processing time on machine B.

At the start of the current week there are 33 units of x and 90 units of y in stock. Available processing
time on machine A is forecast to be 40 hours and on machine B is forecast to be 30 hours.

The demand for x in the current week is forecast to be 75 units and for y is forecast to be 95 units.
Company policy is to maximise the combined sum of the units of x and the units of y in stock at the
end of the week.

• Formulate the problem of deciding how much of each product to make in the current week as a
linear program.

Solution:
Let be x = the number of products x to produce and y = the number of products y to produce. To
solve this problems, it is necessary to compute the inventory left for each product such as inventory
already in stock + number of units to produce − forecast demand, e.g. the inventory left for product
x is 33 + x − 75 and for product y is 90 + y − 95. As a results, the model must maximise the number
of inventory left as z = (33 + x − 75) + (90 + y − 95). The model must warranty that the units to
produce must satisfy the demand, so x ≤ 75 − 33 and y ≤ 95 − 90.

100
A
90 Feasible region

Max z = x + y − 47
80

subject to 70
50x + 24y ≤ 2400 min
30x + 33y ≤ 1800 min 60
F
x ≥ 42
50
y

y≥5
z
40
x, y ≥ 0
30
Point D is the optimum
solution which lies on the C
20 H
intersection of constraints
D
50x + 24y ≤ 2400 and x ≥ 42. 10
L K E I
O J B G
0
0 5 10 15 20 25 30 35 40 45 50 55 60
x

Changes in the objective function’s coefficients

If constraint 50x + 24y ≤ 2400 replaces the objective function z = x + y − 47 (i.e. z = 50x + 24y), is
the optimum solution on point D? The answer is yes. But if z = x (second constraint) the optimum
solution changes to point E.  
dy d 2400 50 50
The slope of the first constraint is = − x = − . For the second constraint, ∞ is
dx dx 24 24 24
used not because the slope of x ≥ 42 but the change in the optimum solution.

Therefore

3
 if z = x + c2 y − 47,
if z = c1 x + y − 47, dy

d z + 47 1

1
= − x =− .
dy d dx dx c2 c2 c2
= (z + 47 − c1 x) = −c1 .
dx dx
50 1
− ≤− ≤∞
50 24 c2
− ≤ −c1 ≤ ∞
24 24
− ≥ −c2 ≥ −∞

50
 50
− ≤ −c1 ≤ ∞ (−1)
24 
24

− ≥ −c2 ≥ −∞ (−1)
50
2.083 ≥ c1 ≥ −∞
0.48 ≤ c2 ≤ +∞

Changes in the RHS (Dual prices)

The optimum value is on the intersection of constraints 50x + 24y ≤ 2400 and x ≥ 45; thus,

Constraint 50x + 24y ≤ 2400 can move from

K(42, 5) to H(42, 16.36).

bK = 50(42) + 24(5) = 2200 Constraint 30x + 33y ≤ 1800 can move from
D(45, 6.25) to +∞.
bH = 50(42) + 24(16.36) = 2492.64
bD = 30(45) + 33(6.25) = 1556.25
• 2200 ≤ b1 ≤ 2492.64
• 1556.25 ≤ b2 ≤ +∞
zK = 1(42) + 1(5) = 47
• D2 = 0
zH = 1(42) + 1(16.36) = 58.63
58.63 − 47
• D1 = = 0.0397
2492.64 − 2200

Constraint x ≥ 45 can move from C(38.71, 19.35)

to E(45.6, 5).

bC = 38.71 Constraint y ≤ 5 can move from −∞ to

D(45.6, 6.25).
bE = 45.6
bD = 6.25
• 38.75 ≤ b3 ≤ 45.60
• −∞ ≤ b4 ≤ 6.25
zC = 38.71 + 19.35 = 58.06
• D4 = 0
zE = 45.6 + 5 = 50.6
58.06 − 50.6
• D3 = = −1.08
38.71 − 45.6

3 The demand of two products

The demand for two products is 37 and 14 units, respectively. These products are produced using
two machines, X and Y. Each unit of product 1 that is produced requires 15 minutes processing on
machine X and 25 minutes processing on machine Y. Each unit of product 2 that is produced requires
7 minutes processing on machine X and 45 minutes processing on machine Y. The available time on
machine X is forecast to be 20 hours and on machine Y in week 5 is forecast to be 15 hours. Each
unit of product 1 sold in week 5 gives a contribution to profit of $10 and each unit of product 2 sold
in week 5 gives a contribution to profit of $4.

4
It may not be possible to produce enough to meet your forecast demand for these products and each
unit of unsatisfied demand for product 1 costs $3, each unit of unsatisfied demand for product 2 costs
$1.

• Formulate the problem of deciding how much of each product to make in week 5 as a linear
program

Solution:

Be x1 is the number of units of product X and x2 of product Y. The aim is to maximise the revenue
computed as profits minus cost of unmet demand. The number of unmet demand is 37−x1 for product
X; i.e. the forecast minus the manufactured X products. For those Y, unmet demand is 14 − x2 . The
revenues are z = 10x1 + 4x2 − 3(37 − x1 ) − 1(14 − x2 )

180
A
160
Max z = 13x1 + 5x2 − 125
140 Feasible region
subject to
15x1 + 7x2 ≤ 1200 min 120
25x1 + 45x2 ≤ 900 min
x1 ≤ 37 100 C
x2

x2 ≤ 14
80

x1 , x2 ≥ 0 z
60

Point F is the optimum 40

solution which lies on the
E
intersection of constraints 20 G
J I D
25x1 + 45x2 ≤ 900 and x2 ≥ 0. O F H B
0
0 10 20 30 40 50 60 70 80
x1

Changes in the objective function’s coefficients

if z is replaced by x2 ≥ 0, (z = x2 ), the optimum solution is not point F . The new optimum solution
lies on the line from point J to G because the problem has multiple solutions. To compute the
intervals, −∞ is used. On the other  hand, if z =
 25x1 + 45x2 , the optimum solution remains in point
dx2 d 900 25 25
G. Hence, its slope is = − x1 = − .
dx1 dx1 45 45 45
if z = 13x1 + c2 x2 − 125
if z = c1 x1 + 5x2 − 125, dx2 d

z + 125 13

13
= − x1 = − .
dx2 d

z + 125 c1

c1 dx 1 dx 1 c2 c2 c2
= − x1 = − .
dx1 dx1 5 5 5
13 25
−∞ ≤ − ≤ −
c1 25 c2 45
−∞ ≤ − ≤ −
5 45 c2 45
∞≥− ≥−

c1 25
 13 25
−∞ ≤ − ≤ − (−5)
5 45 
c2 45

∞≥− ≥− (−13)
13 25
∞ ≥ c1 ≥ 2.77
−∞ ≤ c2 ≤ 23.4

5
Changes in the RHS (Dual prices)
The optimum value is in the intersection of constraints 25x1 + 45x2 ≤ 900 and x2 ≥ 0; thus,

Constraint 25x1 + 45x2 ≤ 15(60) can move from

O(0, 0) to H(37, 0).

Constraint 15x1 + 7x2 ≤ 20(60) can move from bO = 25(0) + 45(0) = 0

F (36, 0) to +∞.
bH = 25(37) + 45(0) = 925
bF = 15(36) + 7(0) = 540
• 0 ≤ b2 ≤ 925
• 540 ≤ b1 ≤ +∞
zO = 13(0) + 5(0) = 0
• D1 = 0
zH = 13(37) + 5(0) = 481
481 − 0
• D3 = = 0.52
925 − 0

Constraint x1 ≤ 37 can move from F (36, 0) to +∞. Constraint x2 ≤ 14 can move from H(37, 0) to +∞.

bF = 36 bH = 0

• 36 ≤ b3 ≤ +∞ • 0 ≤ b4 ≤ +∞

• D3 = 0 • D4 = 0

4 Production of two items (x and y)

A company is involved in the production of two items (x and y). The resources need to produce x and
y are twofold, namely machine time for automatic processing and craftsman time for hand finishing.
The table below gives the number of minutes required for each item:

Item Machine Time Craftsman time

x 13 19
y 19 29

The company has 40 hours of machine time available in the next working week but only 35 hours of
craftsman time. Machine time is costed at $10 per hour worked and craftsman time is costed at $2
per hour worked. Both machine and craftsman idle times incur no costs. The revenue received for
each item produced (all production is sold) is $20 for x and $30 for y. The company has a specific
contract to produce 10 items of X per week for a particular customer.

• Formulate the problem of deciding how much to produce per week as a linear program

Solution:

Let be x the number of products x to produce and y the ones produced of y. The objective is to
maximised
 the revenues
 such as profits
 20x + 30y minus the machine cost of every machine in hours
13x + 19y 19x + 29y
10 +2 .
60 60    
13x + 19y 19x + 29y
Hence, the objective function is to maximise z = 20x + 30y − 10 −2 .
60 60

6
 140

A
C
120

Max z = 17.16x + 25.86y

100
subject to
13x + 19y ≤ 2400 min
19x + 29y ≤ 2100 min 80
D
x ≥ 10

y
F
z
60
x, y ≥ 0

Point E is the optimum

40
solution which lies on the
intersection of constraints
19x + 29y ≤ 2100 and y ≥ 0. 20

Feasible region
O G E B
0
0 20 40 60 80 100 120 140 160 180
x

Changes in the objective function’s coefficients

If the current objective function z = 17.16x + 25.86y is changed for z = y, the optimum solution is
now in point F . The −∞ must be used in this  constraint. 
dy d 2100 19 19
The slope of the second constraint is = − x =− .
dx dx 29 29 29
Therefore

if z = 17.16x + c2 y,
if z = c1 x + 25.86y, dy d

z 17.16

17.16
= − x =− .
dy d  z c1  c1 dx dx c2 c2 c2
= − x =− .
dx dx 25.86 25.86 25.86
17.16 19
−∞ ≤ − ≤−
c1 19 c2 29
−∞ ≤ − ≤−
25.86 29 c2 29
∞≥− ≥−

c1 19
 17.16 19
−∞ ≤ − ≤− (−25.86)
25.86 29 
c2 29

∞≥− ≥− (−17.16)
17.16 19
∞ ≥ c1 ≥ 16.94
−∞ ≤ c2 ≤ 26.20

Changes in the RHS (Dual prices)

The optimum value is in the intersection of constraints 19x + 29y ≤ 2100 and y ≥ 0; thus,

7
 Constraint 19x + 29y ≤ 2100 can move from
G(10, 0) to B(184.6, 0).

Constraint 13x + 19y ≤ 2400 can move from bG = 19(10) + 29(0) = 190
E(110.5, 0) to +∞.
bB = 19(184.6) + 29(0) = 3507.4
bE = 13(110.5) + 19(0) = 1436.5
• 190 ≤ b2 ≤ 3507.4
• 1436.5 ≤ b1 ≤ +∞
zG = 17.16(10) + 25.86(0) = 171.6
• D1 = 0
zB = 17.16(184.6) + 25.86(0) = 3167.73
3167.73 − 171.6
• D2 = = 0.9031
3507.4 − 190

Constraint x ≤ 0 can move from −∞ to

E(110.52, 0).

bE = 110.52

• −∞ ≤ b3 ≤ 110.52

• D3 = 0

5 A carpenter makes tables and chairs

A carpenter makes tables and chairs. Each table can be sold for a profit of £30 and each chair for a
profit of £10. The carpenter can afford to spend up to 40 hours per week working and takes six hours
to make a table and three hours to make a chair. Customer demand requires that he makes at least
three times as many chairs as tables. Tables take up four times as much storage space as chairs and
there is room for at most four tables each week.

Solution:

x1 is the number of tables to produced and x2 the number of chairs. About the constraint of
proportionality, e.g. if x2 = 3x1 and x1 = 1, then x2 = 3(1) = 3 chairs. As the proportion is at
least three times as many chairs as tables, x2 ≥ 3(1) = 3.
16
C

14 F
Max z = 30x1 + 10x2
12
subject to E
3x1 − x2 ≤ 0 10 Optimum solution
1
x1 + x2 ≤ 4 tables
4 B
x2

6x1 + 3x2 ≤ 40 hr 8 Feasible region

A
x1 , x2 ≥ 0 6

Point E is the optimum 4

z
solution which lies on the
intersection of constraints x1 + 2
0.25x2 ≤ 4 and 6x1 +3x2 ≤ 40.
O D G
0
0 1 2 3 4 5 6 7
x1

8
Changes in the objective function’s coefficients
dx2 d
The slope of the two constraints in which the optimum solution E lies is = (4 × 4 − 4x1 ) = −4
  dx1 dx1
dx2 d 40 6 6
and = − x1 = − .
dx1 dx1 3 3 3
if z = 30x1 + c2 x2 ,
if z = c1 x1 + 10x2 , dx2 d

z 30

30
= − x1 = − .
dx2 d  z c1  c1 dx1 dx1 c2 c2 c2
= − x1 = − .
dx1 dx1 10 10 10
30 6
−4 ≤ − ≤ −
c1 6 c2 3
−4 ≤ − ≤ −
10 3 1 c2 3
− ≥− ≥−

c1 6
 4 30 6
−4 ≤ − ≤ − (−10)
10 3 
1 c2 3

− ≥− ≥− (−30)
4 30 6
40 ≥ c1 ≥ 20
7.5 ≤ c2 ≤ 15

Changes in the RHS (Dual prices)

The optimum value is in the intersection of constraints x1 + 0.25x2 ≤ 4 and 6x1 + 3x2 ≤ 4; thus,

Constraint x1 + 0.25x2 ≤ 4 can move from

F (0, 13.33) to B(2.67, 8).

Constraint 3x1 − x2 ≤ 0 can move from bF = 0 + 0.25(13.33) = 3.33

E(1.33, 10.67) to +∞.
bB = 2.67 + 0.25(8) = 4.67
bE = 3(1.33) − 10.67 = −6.68
• 3.33 ≤ b2 ≤ 4.67
• −6.68 ≤ b1 ≤ +∞
zF = 30(0) + 10(13.33) = 133.3
• D1 = 0
zB = 30(2.67) + 10(8) = 160.1
160.1 − 133.3
• D2 = = 20
4.67 − 3.33

Constraint 6x1 + 3x2 ≤ 4 can move from

A(2.28, 6.85) to C(0, 16).

bA = 6(2.28) + 3(6.85) = 34.23

bC = 6(0) + 3(16) = 48

• 34.23 ≤ b2 ≤ 48

zA = 30(2.28) + 10(6.85) = 136.9

zC = 30(0) + 10(16) = 160

160 − 136.9
• D3 = = 1.67
48 − 34.23

6 Dorian Auto manufactures luxury cars and trucks

Dorian Auto manufactures luxury cars and trucks. The company believes that its most likely customers
are high-income women and men. To reach these groups, Dorian Auto has embarked on an ambitious

9
TV advertising campaign and has decided to purchase 1-minute commercial spots on two types of
programs: comedy shows and football games.

Each comedy commercial is seen by 7 million high-income women and 2 million high-income men.
Each football commercial is seen by 2 million high-income women and 12 million high-income men.
A 1-minute comedy ad costs $50,000, and a 1-minute football ad costs $100,000. Dorian would like
the commercials to be seen by at least 28 million high-income women and 24 million high-income men.

Use linear programming to determine how Dorian Auto can meet its advertising requirements at
minimum cost.

Solution:

Let be xi is the number 1-minute i add, i = 1 (comedy), 2 (sports).

10 A

Min z = 50x1 + 100x2

8

subject to
7x1 + 2x2 ≥ 28
6 Feasible region
2x1 + 12x2 ≥ 24
x2

x1 , x2 ≥ 0
4
Point C is the optimum
solution which lies on the
D
intersection of the two 2 C
constraints.

O B z E
0
0 1 2 3 4 5 6 7 8 9 10 11 12
x1

Changes in the objective function’s coefficients

 
dx2 d 28 7 7
The slope of the two constraints in which the optimum solution C lies is = − x1 = −
  dx1 dx1 2 2 2
dx2 d 24 2 2
and = − x1 = − .
dx1 dx1 12 12 12
if z = 50x1 + c2 x2 ,
if z = c1 x1 + 100x2 , dx2 d

z 50

50
= − x1 = − .
dx2 d  z c1  c1 dx1 dx1 c2 c2 c2
= − x1 = − .
dx1 dx1 100 100 100
7 50 2
− ≤− ≤−
7 c1 2 2 c2 12
− ≤− ≤−
2 100 12 2 c2 12
− ≥− ≥−

7 c1 2
 7 50 2
− ≤− ≤− (−100)
2 100 12 
2 c2 12

− ≥− ≥− (−50)
7 50 2
350 ≥ c1 ≥ 16.67
14.28 ≤ c2 ≤ 300

10
Changes in the RHS (Dual prices)
The optimum value is in the intersection of constraints 7x1 + 2x2 ≥ 28 and 2x1 + 12x2 ≥ 24; thus,

Constraint 7x1 + 2x2 ≥ 28 can move from D(0, 2) Constraint 2x1 + 12x2 ≥ 24 can move from B(4, 0)
to E(12, 0). to A(0, 14).

bD = 7(0) + 2(2) = 4 bB = 2(4) + 12(0) = 8

bE = 7(12) + 2(0) = 84 bA = 2(0) + 12(14) = 168

• 4 ≤ b1 ≤ 84 • 8 ≤ b2 ≤ 168

zD = 50(0) + 100(2) = 200 zB = 50(4) + 100(0) = 200

zE = 50(12) + 100(0) = 600 zA = 50(0) + 100(14) = 1400

600 − 200 1400 − 200
• D1 = =5 • D2 = = 7.5
84 − 4 168 − 8

7 Leary Chemical
Leary Chemical manufactures three chemicals: A, B, and C. These chemicals are produced via two
production processes: 1 and 2. Running process 1 for an hour costs $4 and yields 3 units of A, 1 of
B, and 1 of C. Running process 2 for an hour costs $1 and produces 1 unit of A and 1 of B. To meet
customer demands, at least 10 units of A, 5 of B, and 3 of C must be produced daily. Graphically
determine a daily production plan that minimizes the cost of meeting Leary Chemical’s daily demands.

Solution:

Let be xi is the number 1-minute i add, i = 1 (comedy), 2 (sports).

A
10

Min z = 4x1 + x2 8

subject to 7
z
3x1 + x2 ≥ 10
6 Feasible region
x1 + x2 ≥ 5
E
x2

x1 ≥ 3 5

x1 , x2 ≥ 0 4

Point G is the optimum 3

solution which lies on the C
G
2
intersection of constraints
x1 ≥ 3 and x1 + x2 ≥ 5.
1 D
O H B F
0
0 1 2 3 4 5 6
x1

Changes in the objective function’s coefficients

If constraint x1 ≥ 3 is z = x1 , the optimum solution still on point G. Hence, the slope used to compute
dx2 1
the interval is = , i.e. the slope in undefined.
dx1 0
11
 dx2 d
The slope of the second constraint is = (5 − x1 ) = −1.
dx1 dx1
if z = 4x1 + c2 x2 ,
if z = c1 x1 + x2 , dx2 d

z 4

4
= − x1 = − .
dx2 d dx1 dx1 c2 c2 c2
= (z − c1 x1 ) = −c1 .
dx1 dx1
1 4
− ≤ − ≤ −1
1 0 c2
− ≤ −c1 ≤ −1
0 0 c2
− ≥ − ≥ −1

1
 1 4
− ≤ −c1 ≤ −1 (−1)
0 
0 c2

− ≥ − ≥ −1 (−4)
1 4
∞ ≥ c1 ≥ 1
0 ≤ c2 ≤ 4

Changes in the RHS (Dual prices)

The optimum value is in the intersection of constraints x1 + x2 ≥ 5 and x1 ≥ 3; thus,

Constraint x1 + x2 ≥ 5 can move from D(3, 1) to

+∞.

bD = 3 + 1 = 4

• 4 ≤ b2 ≤ +∞ – This IS true

• D2 = 0 – This is NOT true

Constraint 3x1 + x2 ≥ 10 can move from −∞ to Could you figure it out why?
G(3, 2).
The true value is:
bG = 3(3) + 2 = 11 Constraint x1 + x2 ≥ 5 can move from D(3, 1) to
any point on constraint x1 ≥ 3, to compute the
• −∞ ≤ b1 ≤ 11
dual the point G(3, 2) is selected.
• D1 = 0
bD = 3 + 1 = 4

bG = 3 + 2 = 5

zD = 4(3) + 1 = 13

zG = 4(3) + 2 = 14
14 − 13
• D2 = =1
5−4
Constraint x1 ≥ 3 can move from C(2.5, 2.5) to
F (5, 0).

bC = 2.5

bF = 5

• 2.5 ≤ b3 ≤ 5

zC = 4(2.5) + 2.5 = 12.5

zF = c4(5) + 0 = 20
20 − 12.5
• D3 = =3
5 − 2.5
12