KAJ Some organisations use RAs where the score is identified prior to the implementation of controls and then with controls. Personally I do not like this method because it becomes overly complex and as far as I'm concerned the score/risk is not really relevant to the eventual outcome. I would argue against such a methodology and there is no legal requirement to do so.

I have also had HSE inspectors saying that not only are they not interested in the scores; they are not interested in the risk assessments either. They should be focusing on the controls not the paperwork. One again it might be time for the H&S professional to tell the enforcer to wind their neck in.

Clairel, the whole 'before/after' thing is something that can be quibbled over but surely 'evaluating the risk' involves some kind of scoring (1,2,3...High, Medium, Low). You have to meld likelihood and severity, right?

Put simply 'no'! The HSE don't require it, don't like it, and it's not necessary. You need to know what the hazard is and how you control that hazard. Numbers IMO confuse an issue. One number lower or higher and suddently someone is saying well look my number says it's only a medium risk instead of just asking themsleves whether they can do more or less and whether it is reasonably practicable to more or less. People can and do become fixated on the numbers instead of the hazards and the controls.

There is no legal requirement for quantitative risk assessments with a scoring matrix. However, they have been universally adopted within industry as an 'idiots guide' with low/medium/ high risks identified from the scoring matrix.

Not sure anyone's saying a scoring matrix is a legal requirement, Ray. But evaluating risk, in whatever way, is (via ACoP)

Fair point safetyamateur, but in the context of the original query about evaluating risks before and after controls are applied the thread has slightly deviated. I did not mean to imply that hazards, risks and controls do not need to be identified in a RA - that's a given.

Before anyone else adamantly support the argument in favour of the use of risk matrixes within risk assessment, please look at the following paper: http://www.evira.fi/atta...essment/riskmatrices.pdf

KAJ, to get back to your original question: No it's not normal although some people do it. Suggest you ask the inspector "Why?" and then use the arguments put by previous posters.

That said... they can be a useful tool - just wanted to ensure I covered both sides :)

MEden, my practice suits the requirements fine. If a hazard is not significant, I'm not even going to [formally] assess the risk. And this is where we get back to the original question. The inspector(?) is suggesting that the intial judgement of 'significant' is recorded. Not something I'd do willingly as, to my mind, the fact that there's a formal risk assessment proves it's been judged significant. And in reality that judgement's based on the fact that it's likely to happen and there's some severity involved. Life's too short to have a piece of paper for every hazard. "Insignificant risks can usually be ignored, as can risks arising from routine activities associated with life in general". We then move to the method by which risk and the controls are judged: Likelihood and Severity. If you don't have a standardised way of judging this, then it's down to individual perceptions which has its weaknesses. Whether there's a bunch of numbers or colour codes or equations doesn't matter; it gives you a way of arriving at 'reasonably practicable'. This is good for your organisation AND good for explaining your action/inaction when an inspector calls. But, going back to the OP, no, I don't think recording that particular judgement is necessary.

Just for the record, it was an HSE Inspector (25 years service). the comments have been good and as Nlivesey states - a simple but key area has so many differnent views. We put the number in to keep them happy but it frustrates me when they come up with ideas that don't come from their own guidance. thanks for the comments

KAJ Safe, Do what feels best for the end user and use the risk assessment as a poster for controls. I think the numbers are a guide in most risk assessments that are carried out. The only key thing I tell people who I train, is not to be a death person !!! eg dont rate everything to kill its not realistic. Not sure this helps but thought I would contribute.

As Clairel and other responders have already pointed out, numerical scoring systems are unnecessary and inappropriate because people tend to get bogged down with the numbers and ose sight of what they are supposed to be doing i.e. making informed judgements/decisions about the risks posed by activities, processes or foreseeable circumstances and what precautions, if any, are provided or needed. Furthermore, as imwaldra states at #15, allocating numerical scores inherently involves qualitative judgements or guesses. Over the years I've provided guidance and training about risk assessment & management to various groups of people including numerous headteachers and other managers. I've never advocated or used numerical rating systems and have explained that they are neither compulsory nor recommended in my opinion. However, from discussion and experience I have found that, following basic explanation, people can readily understand and use a simple matrix diagram to assess overall risk based on a) worst likely harm and b) the degree of likelihood of such harm occurring. Another point: If the inspector mentioned above has been with HSE for 25 years, he/she would have started with HSE before 'risk assessment' arrived in general OS&H legislation and parlance. Several years ago I had a discussion with a retired senior HSE inspector at a large disused slate quarry in North Wales (as one does from time to time) about when risk assessment did arrive. We agreed that its first appearance in legislation was in the COSHH Regulations of 1988 regarding hazardous substances. Several years later, as the cornerstone of the Management of Health and Safety at Work Regulations 1992, it became a comprehensive requirement for all workplaces and activities. Even though we couldn't recall encountering risk assessment as a phrase or formal process before 1988 (other than in relation to specialist analysis about the likelihood of major incidents at large chemical plants and similar premises), we agreed that inspectors and others in OS&H had long used mental and basic versions of what are now fully established principles and practices for risk asessment and management.

With respect Graham but there are times where it's necessary to put a number on severity and likelihood to effectively 'rank' a risk. What is important is that where you do apply a risk score/rank then there are some very clear definitions of what constitutes each number (e.g. 1= extremely unlikely/1 occurrence in 1000/minor injury; 5= will happen/1 occurrence in 10/multiple fatalities). If a numerical system is used in this manner then it will reduce the overall guesstimate of the assessor and can be applied throughout an organisation to a pretty robust standard. In this way you can also manage the perception of risk to a greater degree and ensure that people are singing from the same hymn sheet. The manner in which a RA can and will differ depending on what is required from the output. True, the key aim is to enable users to identify the control measures required for an activity and should be kept simple/user friendly, but there may be other requirements from the output, e.g. to prioritise funding/work to reduce risk. If this is the case then the simpler method wouldn't supply this level of information.

"lets not get away from the fact the risk assessment is a piece of paper". Therein lies a problem. Assessment of risk (as per MHAS regs) is a decision-making process. Recording the significant findings from that process can be done on paper or tablets of stone I guess - but this view that 'a risk assessment' is a physical thing, piece of paper / form whatever, is like someone scratching fingernails down a blackboard to me. However, making decisions can be simple (cheese or ham?) or complicated (what are the pros & cons of this car/house/school - let's put them all into columns and score the important attributes of each). Evidence of how you reached your decision? Not needed for sarnie, but there's the matrix justifying why a Ferrari made more sense than a Ford Ka. If the risk merits a great deal of consideration and many inputs, and it's important to be able to retrace the process, choose a tool for that. If it's simple, just solve it - get the step-ladder or the sandwich and get on with it. More than one tool exists.

Staple diet having spent all my dosh on Ferrari . . .

Not necessarily so. As I've already mentioned, the numerical output from quantative RA's can be used for a variety of purposes and, more often than not, won't mean a thing to someone who's using the RA to understand what controls need applying to manage the risk. It's worth remembering that RA's aren't just produced for workplace activities, so to use the term RA can cover a multitude of sins, so to speak. The key is ensuring the output suits the requirements of the audience.

Necessary? Really? Why?