![Show Menu](styles/mobile-menu.png)
![Page Background](./../common/page-substrates/page0240.jpg)
Functional Safety 2014
4
th
– 5
th
November 2014
Copyright © 2014 by Cenbee Bullock PFS Consulting Ltd
Page 4 of 14
It is not easy or even impossible to obtain reliability data for systematic failures since the causes of
failure are widespread even within a particular industry. Currently there is a limited amount of
research failure data available for certain typical failures but these should only be used for reference
purposes.
In accordance with IEC61508-2 there are two types of system: Electro-mechanical systems are
classified as type A, i.e. they do not consist of any microprocessor or programmable electronic
functions (see 7.4.4.1.2 for detail); and Programmable Electronic systems are classified as type B (see
clause 7.4.4.1.3 for detail). Both types are subject to Systematic failures, though some types of
Systematic failure will only occur with type B systems (Fig. 1).
Fig.1 Types of failures
Safety Integrity
Safety Integrity is defined as “The probability of a Safety Instrumented Function satisfactorily
performing the required safety functions under all stated conditions within a stated period of time.”
IEC61508 states
“In determining safety integrity, all causes of failure (both random hardware failures and
systematic failures) that lead to an unsafe state should be included.”
With reference to a UK HSE study
1
on why control systems go wrong, most incidents happen because
of errors in than one phase of the safety lifecycle (analysis, realisation and operation). The analysis
also shows that the majority of the incidents were not caused by any failures of a device or control
system but resulted from systematic failures. Fig. 2 shows the percentage of primary causes
attributable to each phase of the lifecycle. The survey also shows that more than 80% of the failure
causes are attributable, in some degree, to human errors.
1
UK Health and Safety Executive, “Out of Control”, 2003