Risk, Reliability and Safety

GENG5507
Probability and Statistics
Lecture Week 1

Introduction

Dr Gopalan Nair
School of Mathematics and Statistics
1
Revised 8/18GN
 Content

1. What this unit is about

2. Events and Probabilities
3. Random variables and their properties

2
1. What this unit is about
Case Study 1 – Reliable water supply

Source: UWA Honours thesis by Emily Briggs

3
Components

Source: UWA Honours thesis by Emily Briggs

4
 What can go wrong?

Cost-Likelihood Risk Matrix

5
Source: UWA Honours thesis by Emily Briggs
Varanus Island Case Study
[GENG5507 Exam – Semester 1 2016]
• Rupture of a gas pipeline at Varanus Island off the
north coast of WA in 2008.
• Resulted in about A$60 million in damage to the
gas processing facility.

• 3 of 6 pipelines
ruptured

• A third WA gas supply

disrupted

6
 Some of the quantities of interest

Pipe Corrosion
• How to prevent?
• When to repair?
• When to replace?
• Replace after failure is not an option! 7
2. Events and Probabilities
Text book reading

Modarres, Kaminskiy & Krivtsove (2017): Relaibility

Engineering and Risk Analysis, 3rd edition
Section 2.2 [pages 15-28]

Other reading:
Johnson, R. A. Miller & Freund's Probability and
Statistics for Engineers: International 8th edition, 2011
Chapter 3. [Pages 49-79]
Test your knowledge by attempting odd numbered
Exercises in the above sections
8
Events and Probabilities
What are we interested in?
• Type of failure
v the list of possible types of failures is denoted as
S (Sample space)

• Probabilities of certain types of failures

v Event A and its probability P(A)

• Life time of a part or a component, and chance of it

lasting beyond certain time
v Random Variable X , and its distribution 9
Disjoint and exhaustive events

• If B1 and B2 are two sets of failure types, then they

are disjoint (or mutually exclusive) if
B1∩B2 =Φ, empty set

• B1, B2, …, Bn events are exhaustive if

n
! Bi = S
i =1

• mde: Mutually disjoint and exhaustive (partition of S)

10
Some properties
• If A and B are any two events then
P(AUB)=P(A)+P(B)-P(A∩B);

• Conditional probability. For P(B)>0

P( A Ç B)
P( A B) =
P( B)
• If B1, B2, …, Bn are mde and A is any event
then,

P( A ) = P( A ! B 1 ) + P( A ! B 2 )+"+ P( A ! B n )

• If B1, B2, …, Bn is a partition (mde) of S with P(Bi )>0 and A is any

event then,
P ( A ) = P ( A B1 ) P ( B1 ) + P ( A B 2 ) P ( B 2 ) +!+ P ( A Bn ) P ( Bn )
11
Independent events:
Events A & B are independent if
P(A Ç B) = P(A) P(B),
Properties: If A and B are independent, then

a) P(A|B) = P(A) if P(B)>0;

b) P(B|A) = P(B) if P(A)>0;

A collection of events A1, A2, …, An are independent if every

sub collection Ai1, Ai2, …, Aim, for 1<m≤n, satisfies

P(Ai1ÇAi2Ç…ÇAim) = P(Ai1) P(Ai2) … P(Aim). 12

 Some important formula

P(A Ç B Ç C) = P(C|B Ç A) P(B|A) P(A)

Bayes Theorem: If B1, B2, …, Bn is a partition (mde) of S

with P(Bi )>0 and A is any event then, for any j

P(A | B j )P(B j )
P(B j | A) =
P ( A)
P(A | B j )P(B j )
=
P ( A B1 ) P ( B1 ) + P ( A B 2 ) P ( B 2 ) +!+ P ( A Bn ) P ( Bn )

13
Interpretations of the rules

14
3. Random variables and their properties
Text book reading
Modarres, Kaminskiy & Krivtsove (2017): Relaibility
Engineering and Risk Analysis, 3rd edition
Sections 2.3.1 and 2.4 [pages 15 & 47-52]

Other reading:
Johnson, R. A. Miller & Freund's Probability and
Statistics for Engineers: International 8th edition, 2011
• Browse through Chapter 5 (pages 87-124). This
chapter these will discrete distributions.
• 5.1 Continuous Random variables (pages 125-130)
• 16.2 Failure-time distributions (pages 511-515) 15
Random variables and their properties

Time to failure - how long it lasts?

X = time to failure

• X is a random variable

• What is P(X>c)?
Probability it last lasts at least c units of
time?

• What is mean life time?

• How much does it vary?

• What are possible models for X?

16
pdf & cdf
• f(x): probability density function (pdf) of a continuous
random variable X, if
¥
f(x) ³ 0 for all x; ò f ( x)dx = 1
¥
b
P(a < X < b ) = ò f ( x )dx = area under f ( x) between a & b
a

• The cumulative distribution function, cdf, F(x):

x
F (x ) = P( X £ x) = ò f ( y )dy
-¥
17
Graph of Pdf & Cdf
x
F ( x) = P{ X £ x} = ò f X (t )dt
-¥

f(x) F(x)
1
0.9
0.8
0.7
f(x) 0.6
0.5
0.4
0.3
0.2
F(x)
0.1
0

0 x x x x
F(x) 18
Mean and Variance
• For a continuous rv X whose pdf is defined by fX(x), its
mean is given by
¥
µ X = ò xf X ( x)dx
-¥

• Its variance, var(X) or σX2 is given by

¥
var( X ) = s = ò ( x - µ X ) f X ( x)dx
2
X
2
-¥

Var(X)=E(X2) – [E(X)]2
19
 Reliability Function R(x)
Note: f(x) is
• R(x): Probability of survival to at least time x not a
probability
• Also called Survival Function value, it is a
1
probability
0.9
fX(x) density
0.8
¥
0.7
R( x) = P{ X ³ x} = ò f X (t )dt
0.6 x
0.5

F(x1)
R( x) = 1 - F ( x)
0.4

0.3
R(x1)
0.2

0.1

x1 x
20
 R(x): Probability of survival
R(x) to at least time x.
1
0.9
0.8
0.7
0.6
0.5
0.4 F ( x) + R( x) = 1
0.3
0.2
F(x)
0.1
0
x
F(x): Probability of failure
before time x 21
Important INFO

You are now ready to solve problems in Practice

sheet 1!

Quiz 1 –baseline testing is already released; closes at

6pm Monday 6 August

Quiz 2- on today’s topic will be released on Sunday 6

August at 6pm (closes on following Monday at 6pm)

22