Table of Contents Table of Contents
Previous Page  351 / 648 Next Page
Information
Show Menu
Previous Page 351 / 648 Next Page
Page Background

CDOIF

Chemical and Downstream Oil

Industries Forum

CDOIF is a collaborative venture formed to agree strategic areas for

joint industry / trade union / regulator action aimed at delivering

health, safety and environmental improvements with cross-sector

benefits.

Guideline – Demonstrating Prior Use v4

Page 24 of 30

Calculations

There are many sources and techniques for performing reliability calculations. Various formulae

and techniques can be found in BS EN 61508 and BS EN ISO 14224, to name just two. There are

also many technical publications on Reliability Assessments which provide further calculations.

For the purpose of this example, two techniques have been demonstrated, the first example

utilising a Mean Time between Failure calculation and the second utilising formulae within BS EN

61508 for a 1oo1 system.

Calculation Method 1 - Failure Data MTBF Calculation for 1oo1

Total number of dangerous failures = 1

See Footnote

7

Thus the MTBF is

!"# $

years = 960 years

Basic Formula for converting MTBF to failure rate

λ

=

$ &'()

Converting MTBF to dangerous failure rate (

λ

DU

) =

$ &'()

=

$ !"#

per year = 1.04 x 10

-3

per year

Simplified PFD Calculation

See Footnote

8

Approximate PFD

(avg)

=

λ

DU

x

'* +

(Where TI is the proof test interval in years)

For this case, assuming TI = 1 year then:

PFD

(avg)

= (1.04 x 10

-3

) x (

$ +

)

PFD

(avg)

= 5.2 x 10

-4

Complete PFD Calculation

PFD

(avg)

= 1 - (

$ ,-.

x TI ) x (1-e

- (

λ

DU x TI)

)

= 1 – (

$ $.#0

x 10

-3

) x (1 - e

-(1.04 x 10-3)

)

= 1 – (961.5 x (1-0.99896))

PFD

(avg)

=

5.19 x 10

-4

7

Note – Calculation will not work if there are no dangerous failures recorded and is only valid when repair

rate is much greater than failure rate.

8

The theory behind the formula is developed in Reliability, Maintainability and risk by David J Smith and

based upon algebraic simplifications of Markov models.