Talk:Safety engineering

Fault trees

We should expand the discussion of Fault Trees, add the concept of Event Trees, give the idea of Cutset and Minimal Cutset. Linuxlad

I think we should create a new article on Fault Trees, perhaps adding event trees into it. It's very widely used and doesn't deserve just one paragraph (no pictures) in a bigger article. --El Pollo Diablo | Talk 10:31, 18 July 2005 (UTC)

A seperate Fault Tree article could explain a lot more as well as build in the cutset concept and methods of calculation. Ultimately the Safety Engineering article would be suited to giving an overview of the topic and the way the different methods can be used to build a ‘safe‘ solution. --Colin H 21:14, 3 November 2005 (UTC)

Inherent safety

Inherent safety is an important principle in and of itself, that's why I created a separate article and made some new redirects. However, the current chapter on Inherently fail-safe systems fits nicely into this article. I suggest we just move the additional examples to the main page (there is already a more… link) and keep the rest of inherent safety material as a duplicate. Arru 17:45, 6 January 2006 (UTC)

There is a big difference between inherent safety and inherent fail-safes, so I have re-written the inherent safety article. I will do something here eventually.Chemical Engineer (talk) 22:19, 2 August 2008 (UTC)

Conflict on number

There seems to be a conflict between the safety certification section of this article, Life-critical system and hazard analysis regarding 10⁹ hours of operation. This article states 10⁹ seconds, which I beleive is incorrect. -Nordby73 17:46, 21 May 2006 (UTC)

I noticed the same thing. On Safety engineering, it says loss of "less than one life per 30 years of operation (10⁹ seconds)." In fact, those numbers match up (10⁹ seconds = 31.7 years). 10⁹ hours is 114,755 years. These numbers are big enough to be hard to imagine. However, many systems (power plants, buildings, battleships) operate for 30 years or more without loss of life, so that number seems a bit low. On the other hand, no man-made systems have gotten anywhere near 114 kilo-years. Could it be in minutes (1903 years)? Ben 02:18, 15 April 2007 (UTC)

This is annoying me, so I browsed around and found this slide show from a course on safety. It describes the highest Safety Integrity Level as a Probability of dangerous failure per hour on the order of 10^-9. I found similar material at [1] and at [2]. So, I am now concluding that the correct answer is hours and that SIL-4 is very, very safe, indeed! I will change the articles. Ben 03:12, 15 April 2007 (UTC)

singling out the nuclear industry?

On the paragraph about late correction of safety designs, it looks as if the nuclear industry was singled out. It would be a good idea to be more specific (which projects, which context, etc) or add additional examples, in other industries, such as the agro-business and the mad-cow disease, for example. Could we strive for a more neutral wording? ThierryQ 15:35, 25 October 2006 (UTC)