Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
What is a RISK?
#1
We are often asked to define a risk, to which the simple answer is:

risk = likelihood * severity

You can flower it up by talking about the likelihood of a hazard - once exposed - leading to an accident, and the magnitude of the 'loss' arising as a result of the accident. You may even want to talk about the fact that in railway business we are interested in two types of 'loss': human life, & financial.

So that question is easy (or so we think) and we've, got our two marks. The study group then progressed to part B of 2005 Q6

b) Identify the 5 most significant safety risks which exist under the present arrangement, justifying your selection. [10 marks]

And our study group got bogged down: what is a safety risk; are we supposed to define the accidents which are the consequence, the hazard that led to the accident or the cause/precursor?

In the end I concluded that I've never read anywhere what a risk is, other than the definition provided above (and variations thereof). So how should we answer part B of 2005 Q6?

I summarised that we should talk about the chain as the risk: cause-hazard-accident; but can't find documentation anywhere that confirms this as either right or wrong . . .

??? Your thoughts ???
Reply
#2
(17-07-2009, 05:06 PM)Douglas Wrote: We are often asked to define a risk, to which the simple answer is:

risk = likelihood * severity

what is a safety risk; are we supposed to define the accidents which are the consequence, the hazard that led to the accident or the cause/precursor?

In the end I concluded that I've never read anywhere what a risk is, other than the definition provided above (and variations thereof). So how should we answer part B of 2005 Q6?

I summarised that we should talk about the chain as the risk: cause-hazard-accident; but can't find documentation anywhere that confirms this as either right or wrong . . .

??? Your thoughts ???

Interesting the coincidence and similarity between this and yesterday's question

Doug, if you don't know what is wanted re a hazard / pivotal event / risk, then I don't know who does.......

However for my twopennth, I think what I'd do for the IRSE Exam is to concentrate on the "potential accident".
However I do think that it is very sensible to include "the potential secondary accident".
A classic case is Clapham 1988. Rear end collision not even a particularly high speed even for conventional line- generally considered reasonably low risk; we don't provide TPWS protection even for New Works in 2009. However death toll high because another train ploughed into wreckage; thinking about the risks, the likelihood of the 2nd event on a Southern Region commuter line in rush hour following rapidly after a first accident is pretty high.

You can often present mitigations as a means of reducing the likelihood of a more major accident (i.e. one having higher consequence), rather than a means of reducing the consequences of the first event. OK it is "playing with words" but perhaps that is what you need to do in the exam so you show examiner that you do really know the meaning of "risk".

Hence perhaps we need to answer:

1. Collision between mainline train and tourist train (I don't see any significant difference between rear endd and head-on in these circumstances)
2. Derailment of tourist train
3. Secondary accident- train on top line runs into previous accident
4. Tourist train catches fire
5. Tourist who has left the train hit by mainline train

I do share your discomfort and actually wonder whether the examiners really meant risks. I think that instead I might go for hazardous events such as:
1. Tourist train set alight by sparks from loco exhaust
2. Tourist train fails in tunnel, passengers decide to escape from smokey environment by disembarking
3. Tourist train derails in tunnel due to gauge spread
4. Main line signal clears invalidly due to tourist train becoming lost to train detection when in tunnel
5. Main line train unable to brake effectively due to rail surface contamination left by tourist train.

Item 3 looks a bit too much like the accident but to me it is what happens next that would be significant- falling between the rails at slow speed would really be the equivalent of item 2, but could topple and fall foul of the other running line or crush people in flimsey carriages against the tunnel wall, result in item 4 etc.

In the exam it is more important to be able to show a diversity of different scenario (because it gives you plenty to talk about to display the width and breadth of knowledge) than necessarily actually get the "correct" top 5 risks. Hence item 5 could result in a mainline accident some distance beyond the tunnel not involving the tourist train at all- except for what it had left behind during its last passage.

In the discussion justifying why consider a high risk, your answer can work in all the other factors that add to the risk (e.g. all the foreign tourists on board who don't understand French, lack of visibility in the tunnel exacerbated by smoke, low crash resistance of rollingstock etc etc) so if those were the things the examiners were looking for you have worked them into your answer.

If I did decide to go for the "hazardous events" approach though, I'd caveat this to inform the examiners that this was my interpretation of the question- I'd certainly start off by briefly stating the actual risks (just as I have here) but then go on to describe the event sequence- indeed I suspect that this was somewhat similar to what Doug was describing.

No idea if I am right of course, but that is the way I'd have tackled.
PJW
Reply
#3
Yes I agree this can be complex (though not as complex as me trying to post a reply to this question on thsi website.....). You clearly have the official definiton of risk but its not helping conceptually. I am only interested here in "safety risk" (which in its broadest sense is jsut another business risk but not what the IRSE exam is all about i think). I find it simplest to think of risk as being the opposite of safety - a measure of the absence of safety if you like. Nothing in life is risk free (i.e. totally safe) so we need to manage risk to achive an affordable balance between enterprise and harm.

So what are the biggest risks means what are the biggest safety issues. When you start taling about hazards it all gets more comlicated because hazards (or hazardous events) are not absolute. What is my hazard is not yours and vice versa. A hazard is only meaningful if it is defined with respect to a certain system or subsystem or product. Hazards exist at boundaries (this is a key YB theme) - causes - things that can lead to thema re within the system and therefore under your control to a greater or lesser extent. Events that arise from them are by defintion outside the direct control of your system and you lay yourslef at the feet of other mitigating measures that may get you off the hook (phew that was close) or limit the safety harm that you cause (good emergency response will reduce severity of casualties) i.e. the consequences, some, but not all, of which may be accidents. Once established it is then possible to construct credible accident sequences (little stories) that describe something happening and these can then be managed away to the extent necessary and reasoanbly practicable. In doing so youa re developign an approximate (remember that) model of how the world works under certain circumstances. This enables you to reason about risk and see waht happens if....

We need to do all this beacuse we are dealing with rare events and a system that ahs years of expericene built into it. Yet we have to change and modify that system all the tiem so we need a way fo makign sure that we arent affecting the risks unknowingly.

Going back to the defintion all it is tryign to do is allow us to balnce high frequency, low consequence events in a common currency with low frequency, high consequnce events (and all those less obviosu ones in between). So serious train crashes are rare but slips trips and falls on escalators are frequent. The first is low frequecy but high consequence the second is low consequence (usually) and high frequency. How do we know how mcuh effort we need to put in to protect agaisnt these - by using the concepts of risk to measure them agaisnt each other. BEWARE however that we are not makign an acceptabiltiy judgement - "risk tolerabiltiy" as it is called is something else all together - this is a givernment decision (hopefully made through some sort of consultation with the public) about how mcuh certain risks matter - so a serious train crash may not in fact have a higher risk as such but it may be less tollerable and therefore things have to be done to prevent or reduce that risk. In the UK we have seen this through the introducton of TPWS, elimitaion of mark1 rolling stock etc. They weren't the top risks in frequencyxseverity tersm but they were judged as being unacceptable to society and money was spent to elimiante or reduce them when actually the same amunt of money coudl have saved more lives spent elsewhere (on the railway or more controversially in the health service).

Oh, but dont be suprised if the examiners use risk in a more colleqial way in the exam - like the weahter forecasters with their risk of showers - you jsut have to work out when they mean real risk and when they jsut mean likelihood. Sorry!
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)