Sensitivity Analysis

• In LP, parameters (i/p) of the model can change

certain limits without causing the optimum solution
to change. This is referred to as sensitivity analysis.
• In LP models, the parameters are usually not exact.
With SA, we can ascertain the impact of this
uncertainty on the quality of the optimum solution.
• It reveals that the optimum remains the same for a
±10% change in the unit profit. In this case the
solution is more robust for indifference range of ±1%.
 Graphical Sensitivity Analysis

• It will demonstrates the general idea of sensitivity

analysis. Two cases will be considered:
1. Sensitivity of the optimal solution to changes in the
availability of the resources (RHS of constraint)
2. Sensitivity of the optimum solution to changes in
unit profit or unit cost (coefficients of the objective
function).
 Example – (Changes in RHS)

• JABCO produces two products on two machines. A

unit of product 1 requires 2 hours on machine 1 and
1 hour on machine 2. For product 2, a unit requires 1
hour on machine 1 and 3 hours on machine 2. The
revenues per unit of products 1 and 2 are $30 and
$20, respectively. The total daily processing time
available for each machine is 8 hours.
 Letting x1 and x2 represent daily number of units of
products 1 and 2, respectively, the LP model is given
as –
Maximize z = 30x1 + 20x2
subject to
2x1 + x2 ≤ 8 (M/c 1)
x1 + 3x2 ≤ 8 (M/c 2)
x1,x2 ≥ 0
• (Rate of revenue change result from increasing
machine 1 capacity by 1 hr
[point C to point G])
= zG – zC/capacity change
142 −128
=
9−8
= $14.00/hr
The computed rate provides a direct link between the
model input (resources) and its output (revenue) that
represents the unit worth of a resource (in $/hr) – that
is, the change in the optimum objective value per unit
change in the availability of the resource (m/c
capacity). This means that a unit increase (decrease) in
m/c1 capacity will increase (decrease) revenue by
$14.00
• Looking at graph, we can see the dual or shadow price
of $14.00/hr remains valid for changes (increases or
decreases) in m/c1 capacity that move its constraint
parallel to itself to any point on the line segment BF.
So, the range of dual price can be computed as
follows:
 Range of applicability of dual prices

• Min capacity of M1 at B[0,2.67]

= 2*0 + 1*2.67
=2.67 hr.
• Max capacity of M1 at F[8,0]
=2*8 + 1*0
=16 hr
• Validity of dual prices
2.67 hr ≤ capacity of M1 ≤ 16 hr
• Similarly for mc2:
Min m/c2 capacity[atD = (4,0)] = 1*4 + 3*0 = 4hr
Max m/c2 capacity[atE = (8,0)] = 1*0 + 3*8 = 24hr
Thus range  4hrs to 24 hrs
• The computed limits for m/c 1 and 2 are referred to as
the feasibility ranges.
• The dual prices allow making economic decisions about
LP problem, as the following questions demonstrate:
Q1 – If JOBCO can increase the capacity of both
machines, which machine should receive highest priority?
 The dual prices for m/c 1 and 2 are $14.00/hr and
$2.00/hr. Thus priority should be given to m/c 1.
Q2 – A suggestion is made to increase the capacities of
m/c 1 and 2 at the additional cost of $10.00/hr. Is this
advisable?
 For m/c1, the additional net revenue per hour is 14.00
– 10.00 = $4.00 and for m/c 2 is 2.00 – 10.00 = -$8.00.
Hence, only capacity of m/c 1 should be increased.
Q3 – If the capacity of m/c 1 is increased from the
present 8 hours to 13 hours, how will the increase
impact the optimum revenue?
 The dual price of m/c1 is $14.00 and is applicable in
range (2.6,16) hrs. The proposed increase to 13
hours falls within the range. Hence, the increase in
revenue is 14.00*(13-8) = $70.00, which means that
the total revenue will be increased to (current
revenue + change in revenue) = 128 + 70 = $198.00
Q4 – Suppose that the capacity of m/c1 is increased to
20 hours, how will it impact the optimum revenue?
 The proposed change is outside the range (2.67,16).
Thus, we can only make an immediate conclusion
regarding an increase upto 16 hours. Beyond that,
further calculations are required.
Q5 – We know that the change in the optimum
objective value equals (dual price x change in resource)
so long as the change in the resources is within the
feasibility range. What about the associated optimum
values of the variables?
 The optimum values of the variables will definitely
change. However, the level of information we have
from the graphical solution is not sufficient to
determine the new values.
 Changes in O.F coefficients
• Max z = c1x1 + c2x2
• The optimum remains at C as long as it lies at
intersection of BF & DE
• The optimum remains at C as long as z = c1x1 +
c2x2lies between x1 + 3x2 =8 & 2x1 + x2 =8

• 1≤c1 slope
≤2
3 c2 1
 Example : Changes in the Objective
Coefficients
• Figure shows the graphical solution space of
JOBCO problem. The optimum occurs at point
C (i.e. x1= 3.2, x2= 1.6, z= 128). The changes in
revenue units (i.e. objective function coeff.)
will change the slope of z. however as can be
seen from fig. the optimum constraints that
define the optimum point. This means that
there is a range for the coeff. Of the objective
function that keep the optimum solution
unchanged at C.
• Max z = c1x1 + c2x2
– 1 ≤c1 ≤ 2
3 c2 1
– This provide immediate answer to following
questions:
Q.1 : Suppose the unit revenues for product 1 &
2 are changed to $35 and $25, resp. will the
current optimum remain the same?

– New obj fun.

– Max z = 35x1 + 25 x2
– c1 = 35 = 1.4 [ Remains within range 0.33 to 2]
c2 25
 – Z = 35*(3.2) + 25*(1.6) = $152
Q.2 : Suppose unit revenue of the product 2 is
fixed at its current value of C2 = $20.00 what is
associate range for C1 ,
C2 = 20
6.67≤ C1 ≤ 40
Optimality range for C2 can be found by fixing C1.
Practice question 1
Solution
Practice Question 2
Solution