7

Exponential Distribution
Most widely used probability distribution in reliability assessment.
Requirements:
-Events must be random
-Hazard rate must be constant
R(t) = e

Q(t) = 1 -

e t

Only applies to the useful life

period of a system component
f(t) =

Mean or Expected value of f(x)

E(x) = x.f(x)dx

Mean Time to Failure, MTTF = t.f(t)dt = 1/

dQ(t)
=
dt

dR(t)
dt

MTTF = - t.dR(t) = R s ( t )dt

f(t)

Q(t)

But, f(t) =

-t
dQ(t) d(1 - e )
t
=
= e
dt
dt

R(t)
0.37

0
0

1/

time

11

Example: Exponential Distribution

Find the mean time to failure of a component which has a failure rate of
2 failures per year. Calculate its reliability for different mission times,
e.g. 10, 1000, 10000 hours.
MTTF = 1/ = = 0.5 yrs = 0.5 x 8760 = 4380 hrs
R(t) = e

R(10)=0.997719, R(1000)=0.795877, R(10000)=0.101967

1.0
0.8

Rt)

0.6
0.4
0.2f/yr

0.2

1f/yr
2f/yr

0.0
0

12

10

time in 1000 hrs

Failure Probability in a Time Interval

A priori probability of failure in time interval t, Q(t) = 1 -

e t

A Priori Probability:
probability calculated by logically examining existing information
A Posteriori Probability:
conditional probability that is assigned after relevant information
is taken into account.
The probability of failure in the next interval t actually depends conditionally
upon its behavior preceding that interval.
e.g. it cannot fail in that interval if it already failed prior to that interval
It is, therefore, required to determine (a posteriori) probability of a
component failing in an interval t given that it has survived prior to that
interval.
13

A Posteriori Probability
Probability of component failing during t given that
it has survived up to T, Qc(t)
T+t

P(A B)
P(A|B) =
=
P(B)

f(t)dt
f(t)dt
T

But, f(t) = e

T+t

Q (t) =

Event
A: failure during t (shaded area)
B: surviving up to T (colored area)

e- t dt

e - t dt

e - T e - (T+ t)
= 1 - e t = a priori probability Q(t)
- T
e

Reliability evaluation in the useful life of a component is, therefore, relatively

simple as exponential distribution is applicable.
In the wear-out phase, conditional probability must be used.

14

Exponential Distribution Applications

Series Systems
1

Parallel Systems

product rule of unreliability, Qs = Q1 . Q2

product rule of reliability

Qs(t) = [1 - e

R s = R 1R 2

R S = e 1t + e 2t e -( 1 + 2 )t
n

it

= e

R s = R1 + R 2 R1R 2

= e - ( 1 + 2 )t
t
=e i
Rs(t) = e

it

i =1

= e 1t e 2 t

it
i =1

= e

et

i =1

i.e. resulting distribution for the

system is also exponential

cannot obtain equivalent hazard rate

for exponential distribution
resulting distribution for the
system is non-exponential
i.e. resulting hazard rate for the system
is no longer constant, but a function of time
15

Partially Redundant (m out of n) Systems

Apply Binomial Expansion
n

[R(t) +

Q(t)] n

= nCr R(t) n-r Q(t) r

r=0

During useful life period when component failures are

exponentially distributed
R(t) = e

and Q(t) = 1 - e t

16

Example:
A simple electronic circuit consists of 6 transitors each having a failure rate of
10-6 f/hr, 4 diodes each having a failure rate of 0.5 x 10-6 f/hr, 3 capacitors each
having a failure rate of 0.2 x 10-6 f/hr, 10 resistors each having a failure rate of
5 x 10-6 f/hr and 2 switches each having a failure rate of 2 x 10-6 f/hr. Assuming
connectors and wiring are 100% reliable, evaluate the equivalent failure rate
of the system and the probability of the system surviving 1000 and 10000 hours
if all components must operate for system success.
n

Equivalent failure rate of the system, e

=

6(10-6 )

+ 4(0.5 x

10-6 )

+ 3(0.2 x

t
6.26 x 10
Rs(1000 hr) = e e = e

10-6 )
6

i =1

+ 10(5 x 10-6 ) + 2(2 x 10-6 ) = 6.26 x 10-5 f/hr

x1000

t
6.26 x 10
Rs(10,000 hr) = e e = e

= i

= 0.9393

x10000

= 0.5347
17

Example:
Consider a system comprising of 4 identical units each having a failure rate of
0.1 f/yr. Evaluate the probability of the system surviving 0.5 years and 5 years
if at least two units must operate successfully.
Using Binomial Expansion,
[R(t) + Q(t)] 4 = R4(t) + 4 R3(t)Q(t) + 6 R2(t)Q2(t) + 4 R(t)Q3(t) + Q4(t)
where, R(t) = e

and Q(t) = 1 - e

For 2 out of 4 system,

Rs(t) = R4(t) + 4 R3(t)Q(t) + 6 R2(t)Q2(t)
4 t
3 t
t
2 t
t 2
=e
+4e
(1 - e ) + 6 e
(1 - e
)

For t = 0.5 years, t = 0.1 x 0.5 = 0.05

For t = 5 years,

t = 0.1 x 5 = 0.5

Rs(0.5) = 0.9996
Rs(5) = 0.8282

18