Table of Contents Table of Contents
Previous Page  485 / 1143 Next Page
Information
Show Menu
Previous Page 485 / 1143 Next Page
Page Background

©2016 Engineering Safety Consultants Limited

7 Proof testing categories

7.1 Partial testing

Partial testing is the term used for tests which do not reveal all possible failure modes and are therefore

tests with PTC <100%.

A partial test example is partial stroke checking of a valve, this would involve closing / opening a valve

from its fully open / close position by a small percentage between 10 and 20%. This test would likely be

credited of a PTC of <100%.

The failure modes of the equipment, for example a valve arrangement, must be analysed to see if Partial

Stroke Testing would detect the failure.

Table 4

shows an example for a typical valve arrangement.

Table 4 - Valve Arrangement analysis for partial testing

Failure Mode

Failure Rate

(per year)

Detected by

partial stroke

testing?

Solenoid fails to

vent

6.00E-08

Yes

Valve sticks open 7.20E-07

Yes

Valve doesn’t fully

close or passes

1.80E-07

No

Other unknown

failures

1.08E-06

No

TOTAL

2.04E-06

Without partial testing is can be seen that with a 4-year test based on the failure rate data i

n Table 4

that the PFD of the valve assembly is:

PFD

avg

=

λdu∗Ti

2

PFD

avg

=

2.04E−06∗4

2

Therefore, PFD

avg

= 4.08E

-06

By applying a 3-month partial test on the valve assembly based on the data contained within

Table 4,

the remainder of the test staying at 4 years, the PFD can be seen to be:

PFD

avg

=

λdu(pst)∗Ti(pst)

2

+

λdu(npst)∗Ti(npst)

2

PFD

avg

=

7.2E−07∗0.25

2

+

1.26E−06∗4

2

PFD

avg

= 9.0E

-08

+ 2.52E

-06

Therefore, PFD

avg

= 2.62E

-06

This example demonstrates that with this data that the PFDavg is approximately 1.5 times better with

partial stroke testing

. Figure 6

taken from HSE guidance OG-00054 Appendix 3 shows the effect on

the PFDavg when partial testing is implemented.