There has been some recent group discussions relating to the causes of major process accidents which is encouraging but also disappointing that we are still allowing these incidents to occur (and hence having to analyse them).
Over 25 years ago, the IChemE Loss Prevention Bulletin (LPB - issue 093 June 1990) included an article by Dr Neil Dunsford (HSE Accident Prevention Advisory Unit) entitled:
A Strategy for Plant Management to Prevent Loss - 7 Ways for Managers to Cut Incidents by up to 44%
The crux of this analysis was, by examining (from several hundred incidents) immediate & underlying causes of pipework failure (the principles can be applied to other equipment) and their respective preventive actions, Management can develop and apply a more focused & effective prevention strategy.
We see from this that "interacting" with the pipework (during Maintenance) was the most common cause.
Thereafter, the combination of Human Factors & Task Checking account for over ½ of the preventive measures.
So we know (knew) what goes wrong and how to fix it (or at least a significant part of it with the seven specific areas for management action tabulated above) yet, 25 years on, pipework (and other) failures still occur.
It should come as no surprise to anybody (but seems to get dismissed by nearly everybody) that the majority of incidents have already occurred i.e. these are "Known Knowns" rather than the "The Black Swans".
At Hazards XX in 2008, Trevor Kletz 'predicted' the Accidents of the Next 15 Years based on his (and the wider industry's) experience of the previous 15 years. He concluded his paper with the following wisdom:
The reality is that mission statements have done little to change the corporate world for the better... Telling them what has happened and will happen again unless they learn from it is more effective. Better, let them tell you what they think is the best method of prevention.
To quote the X-Files - "The Truth is Out There" - we have a plethora of incident data and cause analysis; so why aren't we using it?
The AIChE CCPS Guidelines for Chemical Process Quantitative Risk Analysis has (in Appendix A) a useful summary which gives a list (acknowledged not to presume to be exhaustive) of 60 Loss of Containment Causes in the Chemical Industry which can be broken down into the following simple categories:
Opening the (process) envelope [DISCIPLINE].
Creating a weak envelope [DESIGN] or allowing a capable envelope to deteriorate [DEGRADATION].
External impact on the envelope [DAMAGE].
Over-stressing the envelope [DEVIATION].
These are visualised in a series of BowTies below:
This representation includes Risk Management components from the CCPS 20 Elements of Risk Based Process Safety (RBPS) which are shown as simple Prevention or Mitigation barriers.
Again - we know the Underlying Causes and the Preventive Mechanisms - are we doing anything about it?
Capturing/presenting this using bowties provides a simple summary of the possible Threats, potential (actual) Barriers & predicted Consequences to help engage all stakeholders and quickly spot where there are gaps in the current or future protection strategy.
This doesn't replace the systematic rigour of a HAZOP/PHA/Hazard Study but does capture (confirm) the 'obvious' threats at an early stage.
Are we not identifying (acknowledging) the Hazards and/or not assessing (understanding) the Risks; or is it simply the case that we are Pushing the Envelope because we are 'greedy' for Production and in fact are just Pushing our Luck.
If we are hoping to be/stay lucky; the sad reality of 'gambling' is that The House Always Wins !
PS - I'd really like to hear from anybody who knows of an incident (type) that isn't in the 60 cited by Appendix A.