Search
Search

A small improvement in the method for assessment of probability

1. Risk assessment is an important exercise in any OH&S Management System for taking preventive action and assuring safety of people.

2. The procedure for Risk Assessment is freely available from many sources.For example the following sites provide good guideline.

http://www.hse.gov.uk/risk/fivesteps.htm

3. Risk is defined as the net result of probability and consequence; r=pc.

4. As part of assessment, what is suggested in the procedure for assessing probability is to select from a table such as the following:

Probability :   (as per www.uq.edu.au/ohs/pdfs/ohsriskmgt.pdf)

Almost certain                            The most likely and expected result if the hazard –

event takes place.

Quite possible                            Quite possible, would not be unusual, even 50/50

chance

Unusual but possible                    Unusual but possible sequence or coincidence

Remotely possible                        Remotely possible coincidence

Conceivable but unlikely                Has never happened after may years of exposure,

but is conceivably possible

Practically impossible.                   Practically impossible, has never happened before

5 My view

P being one of just two items in the equation, and the assessment having a strong bearing on assurance for people’s safety, it is considered desirable to make the assessment a little more objective.It is true that arbitrariness could not be completely eliminated. But it could be reduced to some extent by taking the analysis a bit further.

6. In view of the above, the line of thought is as follows: An assessor is making the assessment at certain point of time.The assessor is making an assessment with respect to certain situation/location. The likelihood of an undesirable event happening would depend on certain factors existing in the situation/location.

Hence it is felt that the assessor must list the general and specific factors having influence on an accident,    visualize the various states in which the factors likely to exist anywhere, identify each factor-wise the state which is prevalent in the location and, based on the combination of the states of the factors, arrive at conclusion regarding probability.

7. The following Table is provided herewith as an example:

 Sl.No Factors that have influence for an accident Probability of Occurrence and Numerical value assignable Will Happen May Happen Equal Chance May not happen Will not happen Numerical value Assignable 5 4 3 2 1 State of the Factor 1 Occurrence of Incidents Frequent * Noticeable Rare Nil 2 Occurrence of Accident Every Year * Last year * Never 3 Condition of work environment Bad Not satisfactory Satisfactory Scope for improvement Good 4 Exposure of persons Long * Medium * Short 5 Permit system(Hot permit, Height permit etc.,) Does not exist * Exists not followed always * Exists and followed 6 Supervision Does not exist Short period Intermittent frequent Continuous 7 Education/ Experience/ Training Less Less Nil * Qualified Minimum Not trained * Educated Experienced Trained 8 Engineering control Nil Exists but Not reliable * * Exists and reliable 9 Use of PPE Not a practice * * * Completeadherence

* To be described similarly

8. We can also add ‘weight’ to each factor, assign numerical values for the ‘states’ and arrive at a ‘score’ for the overall assessment.

9. The above method involving factorization, factor-wise description of the likely scene in general and selecting the scene that exists in the location at the assessment time , it is felt, may be adopted for any other assessments also, by adapting it suitably.

10. A small step to make the risk assessment a bit more reliable! Or shall we call it a step for ‘process refinement’?

Sam Anbazhagan

To report this post you need to login first.

You must be Logged on to comment or reply to a post.

1. Vijay Vijayasankar
Hey Sam, I see the rationale for using the factors, and also for using a weighted score. But how does one ensure that the factors themselves are as close to mutually exclusive and collectively exhaustive? Unless we do that, wouldn’t the risk calculation be flawed?

cheers
Vijay

(0)
1. Anbazhagan Sam Venkatesan Post author
Hi Vijay,

The resulting assessment is expected to be better than only thoughtfully selecting from the Table at para4.

True.
It is better to have mutually exclusive factors only and limiting it to a reasonable number, say 7.

I also wanted to add a line that exaustive number of factors may not lead to nearness to true probability.It may burden the Assessor without adding much value.

It is a small step for improvement. It can certainly be improved further.

Thanks.

Sam Anbazhagan

(0)