One of the challenges to do with process safety management (PSM) is circular reasoning or logic. Many of the PSM elements tend to become self-referential. Process hazards analyses can illustrate this conundrum with discussions on the following lines.
- Could high temperature in this vessel lead to a serious incident?
- What is high temperature?
- It’s that temperature that could cause a serious incident.
Similar difficulties can be seen when deciding on how to manage a proposed change to the facility. This time the question/answer sequence can go as follows.
- Does this proposed change require implementation of the Management of Change (MOC) program?
- We don’t know — we will have to run it through the MOC system to find out.
The use of risk-based analysis tools can create a related set of problems. For example, many facilities use the concept of ALARP (As Low As Reasonably Practical Risk), say when deciding on whether to use a new material of construction.
But the use of the word “reasonable” in the definition of ALARP requires its own definition, otherwise, once more, a circular discussion is created.
In situations such as these a fundamental problem is that the values being used are qualitative, and often quite subjective — a level of risk that is acceptable to one person may be totally unacceptable to another. There is no right or wrong in situations such as these — merely different opinions.
The only way of breaking the circular logic is to use to use numbers — to move toward quantification. So, for example, it may be decided that “high temperature” is 250C. This means that, if the temperature in the vessel reaches that value, then it action must be taken to reduce the temperature — otherwise there could be a release of flammable or toxic materials from it.
With regard to risk it is much difficult to assign numerical values due to the inherent subjectivity of the topic, as just discussed. Moreover, there is a public relationships aspect to think about. If a manager states that the target value for fatalities is say one in a thousand years, then that manager has said that fatalities, even at a very low level, are acceptable. This is something that he or she probably wishes to avoid doing.
There is no definitive answer to the difficulties just raised. Risk is fundamentally a subjective matter. All that can be said is that, wherever possible, quantitative limits and goals need to be defined.