Risky Business

Danger is all around us, but what is it really and how can we objectively judge whether or not it's significant?

Welcome to the word of risk assessment. It's probably a term you've often heard being bandied about;

How did that pass a risk assessment? or

Has anyone done a risk assessment on that?

Well, something doesn't pass or fail a risk assessment and the process of undertaking a risk assessment doesn't necessarily make something safer. 

Whether something is 'dangerous' or 'safe' are probably the two extremes used when people are actually talking about risk without knowing it and so it's worth familiarising ourselves with the basics.

Essentially, 'risk' is the exposure to danger or the likelihood of a hazard to cause harm. I know what you're thinking - more buzzwords. OK, a 'hazard' is a thing that could cause harm, whereas 'harm' is the consequence of being exposed to the hazard. Let's look at a couple of examples;

In the kitchen, a knife is a hazard and the harm from an interaction could be a cut. A puddle of water on the floor could be a hazard and a broken arm from slipping on the floor could be the harm. Lots of things are hazards, but they only have the capacity to do harm when there's an interaction - a poisonous mushroom is a hazard, but so long as you don't eat it, there's no chance of harm.

What is a hazard for one person might not be a hazard for another. For example, a long crack in the road could be a hazard for someone cycling (and the harm could be a serious injury), whereas the same defect might not create a hazard for a lorry driver whose vehicle has wide tyres. 

Hazards don't necessarily have to be something which can cause immediate likelihood of harm, prolonged exposure might be the issue. For example, an air pollution event might not be an immediate hazard, but long term exposure could cause considerable harm

So, how do we assess these risks? Unknowingly, we do this all the time (often called 'dynamic risk assessment'). When I cycle, do I chance it on the dual-carriageway or do I go the long way round through the park? Do I cross the road at the traffic lights or nip across here? Do I drive within the speed limit, or stick my foot down? We make decisions all the time as we travel around, but we don't think in depth, we simply consider the hazards and the likelihood of harm occurring.

When it comes to design (and it's applicable to temporary as well as permanent situations) how do we go about undertaking a risk assessment? Well Managed Highway Liability Risk from the Institute of Highway Engineers is a useful resource on managing risk, but I want to pull out the risk assessment process and show how anyone could do it.

In looking at risk, we are concerned with identifying it, analysing at and evaluating it. In terms of a design the evaluation may lead to a change in the design. The guidance proposes a risk assessment matrix which I reproduce below;

Personally, I prefer a 4 x 4 matrix because it stops you picking the middle ground; but, the red/ green/ amber approach and the weighting of the options here work fine. This matrix talks about "events" rather than hazards, but it's the same thing as far as we are concerned here and in fact, the language may be a little easier to understand.

Let's try something simple like crossing the road. The scenario is a dual carriageway with three 70mph lanes in each direction and a central reservation. We are looking to add some dropped kerbs to help people cross.

So the event (or hazard) is a person getting hit by a car. What is the likelihood? It's a fast and wide road and so people are going to be careful, but if we thought about the whole community, might there be a chance that someone gets hit (it's tricky). Let's go with 'low' likelihood (the column on the left). What's the consequence of someone being hit by a car? Clearly, it's not going to end well, perhaps high or even severe? (the row across the stop). Let's take low likelihood and high consequence. That puts us in the amber area which is a medium risk (a score of 12, but don't worry about scoring).

Does this mean we are OK to proceed? I'd say anything in amber means we need to have another think. What could we do reduce the likelihood of someone getting hit? Let's signalise the crossing because everyone obeys traffic signals don't they! Perhaps we can get the likelihood down to very low, but not negligible because people don't always obey the signals. The consequence is still going to be high and so reading the chart, we are still amber (the box with a score of 8).

Amber, so we need to think again. What if we want to deal with the consequences and reduce the severity. We know that for pedestrians, if they are hit by a driver going more than 20mph, the chances of being killed increase rapidly with speed and so let's drop our speed limit to 20mph on the basis that someone getting hit is much more likely to survive than being hit at 70mph (where they are almost certainly dead). Keeping the very low likelihood and perhaps choosing a low consequence gets us into the green (the box with a score of 4) and so job done. We'll stick in some signals and a 20mph speed limit.

Except we've not taken any wider context into consideration. If our dual carriageway is a main route into a large city as it leaves the countryside, then practically, 20mph is not going to be rational and unless there is heavy enforcement with average speed cameras, we are going to get lots of speeding - enforcement will not guarantee 20mph either because the enforcement threshold will be higher (because that is how it works).

How about separating people from traffic completely? Let's build a footbridge and leave the speed limit at 70mph. The likelihood of getting hit has to be negligible, but the consequence of someone being hit still has to be high (it's a theoretical issue) and so we have a low risk. Of course, the footbridge needs to have a convenient layout otherwise people might still cross at road level and perhaps drivers not expecting people to cross leads to less attention so the likelihood could rise against the original idea of putting in dropped kerbs.

It is not a precise science. If my example was followed every time, we wouldn't be expecting people to cross busy and fast roads at road level, there would be footbridges and underpasses everywhere. Risk assessment is a tool to help design, it is not a tool which does the design for us.

Of course, we could take a leaf out of some parts of the United States and adopt control measures. By this I mean that rather than designing to protect users crossing the road, we make the user responsible for their own safety. There are plenty of stories from the US where people crossing the road are expected to cross the road with a flag to highlight themselves to drivers. It's of course nonsense and doesn't reduce the risk.

This is just a short introduction to the subject of risk management, but I hope I have been able to respond to the point that we don't pass a risk assessment and even undertaking one doesn't mean the design which provides least harm is the chosen one.

We also need to think holistically and that means design principles. A system which separates people by their ability do do harm would help us keep risk on our streets in the low category.