Risk assessment processes have been around for a number of years. The first time I can remember using one was in the early 1960's. Yes, I'm aware this is almost ancient history but it has a place in this discussion. We were looking at all the ways one could be hurt while performing welding and assembly operations while building an agricultural tool called a roller Harrow. You would have a hard time finding a roller harrow today because technological advances have made it all but obsolete. Back then there was a big demand for the product and it was a difficult product to produce. Lots of heavy lifting, difficult weld positions and so many pinch points you couldn't count them all. We had few safety regulations but not many and OSHA was still years away. We had many injuries and we knew we had a problem. We had to do something. So, we sat down as a team and listed each step of the job and tried to determine any exposure that could cause an injury. Exposures like sharp edges, pinch points, hot material, heavy lifts, etc. Much like an early version of a Job Hazard Analysis or a job safety analysis. The basic risk matrix might look like Exhibit #1 below.
I've taken the liberty to have five levels of risk represented, signified by the colors red, orange, yellow, beige, lime green and green. Most of the risk assessments I've seen have only three colors and of course 3 risk levels.
Essentially, we looked at the likelihood of something happening (Very Likely, Likely and Not Likely). Then we looked at the severity possibility if that something happened. We wanted to know how serious the injury might be (Very Serious, Serious or Not Serious). While this isn't rocket science, it is a system intended to create better awareness, and eliminate a situation we termed unacceptable risk. We use this system even today and for good reason. This system is simple and it gets us to increase our awareness and correct obvious issues.
The logical conclusion is that high probability and high severity required significant action to lower both severity and probability, while low probability and low severity would likely not require any or very little corrective action. Well, this all sounds good but there is an effectiveness issue because it is based only on two dimensions – likelihood and severity. Even back then I knew something was missing in the way the risks were assessed. What about miss-calculation, errors in judgement, issues with skill deficiencies or even simple human error. The risk assessment ignores any of these variables or others for that matter.