I am semi-retired, working only one day a week at my current employer. When I came in I found that they were using the standard 5x5 scoring system in their Risk Assessments. As this was the system they were familiar, and comfortable with I saw no reason to change this, and so have continued using this in their RAMS. 

I have found, over the years, that it is quite good when used to compare risk scores with/without controls, giving some indication of the effectiveness of your control measures. The problem often occurs when a control which reduces likelyhood brings down the score for consequence, and vice versa. However this can be addressed through eduaction. As a side note: HSE Guidance no longer uses a scoring system, IOSH Managing Safely does.

So it would appear, as is often the case with H&S, that it is down to your individual needs/wants. 

Hi Mike

Welcome to the Forums.

Personally I think that for most organisations using numeric risk matrices is an unnecessary distraction, with many problems that tend to come up time after time:

First there is the problem that they give the perception of a scientific approach but are far from being a Quantified Risk Assessment for the simple reason that it is very rare that there is sufficient data to come up with anything like accurate assesments of the probability of something going wrong with the result of severity X happening.

As example, it used to be the case that about 10 people a year fell to their deaths whilst doing roofwork. But to work out what is reasonably practicable STATISTICALLY you need to be able to work out what the Individual Risk (IR) of death is per annum - See HSE "R2P" for explanation of the concept of IR within a discussion of the tolerability of risk.

But we don't know how many people do "roofwork". Might have a reasonably accurate idea of the number of people whose job title is "roofer" (or slater or something similar) BUT it's not just roofers who do roofwork.

I have investigated 10 fatal accidents during roofwork. Only about half of the victims would be described as "roofers".

So, you don't know the population at risk, you don't know how many hours on average per year that population work on roofs and hence you cannot work out the IR with any level of certainty.

Hence, it is more effective to look at the task (with any mitigations) and assess the level of risk than to try and play a numbers game which is pseudoscience.

Next the assessors almost routinely apply a worst case scenario to accident severity whilst not doing the same for occupational health risks.

So, back to those roofers, statistically they are much more likely to do die prematurely from occupational ill health (e.g. due to exposure to asbestos) than from dying from a fall. Just takes longer.

.....and the assessors determine severity according to the hazard rather than risk.

So, as example if using a scale of say 1-5, where 5 is one or more fatalities, then a 5 for a driver on a high speed road but a lower number if the speed limit is 20 or 30 mph. This is statistically nonsensical as per passenger kilometre it is MUCH more likely that A death will happen on a low speed than on a high speed road - even if that death is more likely to be to another road user (pedestrian, cyclist etc) than the driver or the passenger in their vehicle.

Then, amongst all the other issues, there is the question of volume of risk.

Back to the roofer, does the matrix recognise that only ONE person is only on the roof for two hours and reduce the risk compared to a scenario where there are 10 people working on the roof for a week?

Similarly if you have a construction site with a crane you can work out a number for the risk. What do you do if there are five cranes. Do you write out a separate line for each as the risk MUST be about five times higher. If you only have one line it follows that your calculation will underestimate the total risk.

So, you go for five lines in your matrix to solve that problem!!

If you have 100 people hitting nails with hammers will you write out 100 lines on your matrix to adequately assess the total risk faced by those hitting nails with hammers or will you take a sensible approach and dump the numbers game?

As has been said, HSE does NOT endorse these matrices and I don;t think it ever has.

What happens far too often is that people fiddle the numbers to make them work and cheats shouldn't prosper.

My whole beef with these (and I mention this whenever I teach the IOSH Managing Safely course) is that the whole point a risk assessment  process is to enable you decade if a particular course of action at a particular time is safe (the risk is acceptable) or not. But for the purpose of making it look more science like   instead of just stopping at RED and GREEN  people add various shades of yellow which means at the end of the fiddling of the figures you still can’t tell if something is actually safe to do.

Originally Posted by: Mark-W

So despite the HSE not officially advocating them, some inspectors do.

The trouble with pandering to the whims of individuals is that they become contradictory.  I've not had it from HSE inspectors, but I get it from auditors - they (personally) want something one way, if you change it to that way the next one is liable to want it different and you change it back again.

Personally, advocating a risk assessment format that is contrary to every piece of HSE guidance on the topic is something I wouldn't do.

(I also don't like 1-5 x 1-5 for all the reasons set out up the thread. We have some clients that demand it, however, and then we do it 'our' way and also prepare something to let them tick their boxes).

Originally Posted by: Kate

I had one auditor who said "Don't you think it would be easier to score the risk both before and after controls instead of just after them?"

To which I said "No, I don't think that would be easier" and ignored the suggestion.

Interestingly key to my question that is what is currently being done in my organisation. Having just completed NEBOSH i  find it hard as the standard you are taught is as per the HSE guidance with no scoring matrix.

When working as an inspector i never bothered if scored or not - as long as the controls looked sensible thats all i was looking at. There is no right or wrong was to document a risk assessment. For very small companies we were even told to accept method statements if all the information looked to be there without any thing entitled Risk Assessment. 

Personaly i dont like scoreing - for all the reasons others have said. 

Re my comments at #4, an illustration of how difficult it is to put a number to the population at risk and how long they may be at risk for:

Teenager fell more than 20 feet through roof of Bodmin dairy farm (shponline.co.uk)

OK this person survived but the narrative reminds me of one of the first roofing fatalities I investigated.

Some of you don't remember the days before a PC, laptop or tablet was a ubiquitous feature of the working environment, but in ye olden days the new fangled very expensive computer in a factory was getting wet due to a leak in the roof.

So, they sent a teenager up on the roof at 10pm in October to sort out the problem, which in simple terms was pigeon guano and feathers blocking the drains.

So, up he goes armed with bin bag and gloves and down he fell straight through a sheet of asbestos cement.

This person was never in a month of Sundays  going to be be officially classified as a "roofer", any more than the Devon farmworker.

Just another person happening to find themselves working on a roof to do something entirely different from their normal job.

In BOTH cases a simple risk assessment WITHOUT playing numbers games would have said "BAD NEWS".

P

Originally Posted by: O'Donnell54548

I have found, over the years, that it is quite good when used to compare risk scores with/without controls, giving some indication of the effectiveness of your control measures. The problem often occurs when a control which reduces likelyhood brings down the score for consequence, and vice versa. However this can be addressed through eduaction. 

So it would appear, as is often the case with H&S, that it is down to your individual needs/wants. 

Aargh!  Too right about the problem.  I have wasted a fair amount of time in addressing just this issue of reducing the consequence instead of the likelihood, even though I don't care what the risk rating is!

The conversation with the HSE inspector went along the following lines:

Inspector: “Oh, where is the control panel?”

Us: “There isn’t one, we  control it from a laptop in the office”

Inspector: “So why don’t you have a control panel?”

Us: “We don’t need one we can control everything from the laptop. Look is does everything that the control panel would do.”

Inspector: “What if the laptop breaks down?”

Us: “What if  the panel fails?”

Inspector: “Hmmm…”

So, we left it there. No indication as to why he thought the panel was better than a laptop or exactly what legal requirements we had not complied with. There was just this impression that he didn’t like the laptop.

We were surprised that he mentioned it in his report and since we had spent all of budget we decided not to waste taxpayer’s installing an unnecessary item.