First published in Health and Safety at Work Magazine, October 2012
In her series exploring the foundation concepts of safety management Bridget Leathley looks at the traps we fall into when we talk about hazards.
We all know what a hazard is, don’t we? It’s fundamental to understanding how we keep people safe and healthy. Yet start talking to your health and safety colleagues and you’ll soon find significant differences in how they decide what constitutes a hazard in any given situation.
The most commonly cited definition of hazard, used in IOSH and NEBOSH courses, is “something with the potential to cause harm”.
The HSE, however, seems to have dropped the word “potential”. When I wrote this article in 2012, the 2011 revision of INDG 163 defined a hazard as “anything that may cause harm”, giving examples of chemicals, electricity, working from ladders, noise or an open drawer.
Both these definitions of hazard have three elements in common:
This breakdown pinpoints one of the first problems. “Identify hazard” is supposed to be the first of the five steps, before “identify who will be harmed” and “evaluate the risks”.
But characterising something as a hazard assumes both some degree of possibility and a level of consequence to a population of people who can be harmed; in other words, you need to carry out elements of the second and third steps before you identify something as a hazard.
In this article, we will look in more detail at these three elements — the thing, the possibility, the harm — and ask what impact this has on how we identify and control hazards.
I’ve amended the text to make clear that the quote was from the revision of INDG 163 current at the time the article was written. The published version of INDG 163 was updated in 2014. Curiously, it contains no definition for hazard, and the definition previously in the FAQ on the HSE website was also removed. There was a 2016 draft of INDG 163 released, but never published. This also had no definition of hazard. The best you’ll find on the HSE website at the time of writing is the incredibly vague explanation under the heading ‘2. Steps needed to manage risk’:
‘Look around your workplace and think about what may cause harm (these are called hazards).’
I thought in 2012 that the concept of a hazard was wildly inconsistent in how it was applied. Today I fear the situation is worse, something I am writing about in my book.
Hazards are perhaps easiest to identify when they are solid objects, such as vehicles, machinery or tools. But things, and therefore hazards, can also include places, occurrences and people.
The HSE’s examples include “working from ladders” not just “ladders”, suggesting that an activity can be a hazard.
In his book, The human contribution to error (2008), James Reason starts with the suggestion that the human is a hazard: “a system component whose unsafe acts are implicated in the majority of catastrophic breakdowns”. But Reason also goes on to place the human as hero; as one of the control measures available to reduce the impact of unsafe events.
“Anger”, “tiredness”, “curiosity” and “laziness” are also all things that we could classify as hazards in some contexts. If you consider violence against shop staff, customer anger might be the hazard, not customers; for air- traffic controllers, the human is the control, but the boredom that leads them to become unobservant could be treated as the hazard.
One method for considering hazard types is to think about unsafe conditions and unsafe actions. Rather than simplifying the definition, however, this leads to further debate: Is the kettle the hazard or is the hazard the fact that it is switched on? Or is it the action of leaving it to boil unsupervised that is the hazard? Is the pillar drill a hazard before a drill bit is inserted and it’s used?
The answer will depend on context, which requires you to understand the state of the kettle or pillar drill, how it is uses, and which people are likely to come into contact with it.
Categorising unsafe actions as hazards can lead to duplication. A kerb might not be a hazard if it is easily visible and all step users behave accordingly. But where someone cannot see the kerb because they are running or carrying a large load, is the kerb now a hazard, or is the unsafe action the hazard?
Include both and you have an unwieldy risk assessment to manage.
These categories all offer ways of prompting consideration of different types of hazard. Of course, there will be overlaps and omissions in the categories, but the right schema for your context will reduce your chances of failing to identify a hazard.
Another area of confusion and possible duplication is whether the lack of an object is a hazard or a failed control. A missing toe board or scaffold rail, for example, could be a hazard. But is missing personal protective equipment (PPE) really the hazard? The hazard is the flying debris; missing PPE is surely the failed control, not in itself a hazard. And could the same be said about the toe board?
Fire causes similar misunderstandings. The first part of the NEBOSH fire certificate practical asks candidates to list “fire hazards” (those things that can cause a fire); however, the second part refers to “issues” related to the spread of fire, escape and rescue. Hence, the lack of a smoke detector is not a hazard, but an issue that needs fixing by applying control measures.
Roger Bibbings, occupational safety adviser at the Royal Society for the Prevention of Accidents (RoSPA), has noticed the tendency for mislabelling omissions. “Things such as lack of safety information or poor training or failure to communicate are often referred to as hazards,” he says. “These are better described as ‘risk factors’.”
A further difficulty with defining the thing that is the hazard is that there is usually a chain of events leading to the injury. Falling off a roof is not really that dangerous; landing heavily on the ground is the harmful bit. Yet, if you had not fallen from the roof, you would not have landed. So is falling or landing the hazard? Or is the hazard working at height, such that there is the potential to fall (and land)?
Again, fire provides a good example where there is much confusion between the “thing” that is the hazard and the consequence, explains John Norton-Doyle, now an independent safety consultant, but previously with the London Fire Brigade for 12 years. “It seems to me that the RRO [the Regulatory Reform (Fire Safety) Order] requires two risk assessments,” he says. “In the first, fire is a consequence of ignition, fuel and oxygen hazards. In the second, the fire is the hazard, and the consequences are harm to life if fire detection and suppression, and evacuation, escape and rescue are inadequate.”
Another way of looking at this is that when designing and fitting out a building, or carrying out maintenance and inspection, fire must be considered as a hazard, but in daily use the hazards are the things that might lead to a fire. All these points mean our definition of hazard is looking more complicated, since it is clear the “thing” could include objects, people, places and actions.
Just as philosophers ask whether a tree makes a noise if it falls in a forest with no one listening, so we can ask if a hazard is a hazard if no one is exposed? If asbestos fibres are contained in undamaged tiles on the ceiling of a cupboard no one uses, there is no one to be harmed. Here, you might say the fibres are not a hazard; but since there is still the possibility that someone will open the cupboard or worse, drill into the tiles, the asbestos is still a hazard to be controlled.
As we mentioned earlier, the HSE’s five steps process puts assessing who will be harmed as step two, but in reality you will need to consider this aspect at the point of identifying a hazard. By considering harm to vulnerable people — such as expectant mothers, young people, the disabled — you can identify hazards that may have been considered too trivial to be worthy of assessment when carried out by an “average worker”.
Think of a 14-year-old on work experience making and carrying a tray of hot drinks; a worker who has already had six months off work from back problems picking up and carrying a box of photocopy paper; a pregnant woman suffering from morning sickness asked to drive for two hours to get to a 9am meeting.
Timescales also affect our judgement of the seriousness of a hazard. For the inexperienced, a harmful outcome, such as an injury resulting from a fall from height, is more immediate and therefore seems more likely than a long-term health problem caused by exposure to asbestos or manual handling.
“Hazards with delayed effects to health are more frequently missed than visible and immediate safety hazards,” notes RoSPA’s Bibbings.
Since the definition of hazard includes the concept of possibility or potential, the term “potential hazard” is as accurate (and unfortunately as widely used) as “PIN number” or “PAT test”.
What the definition of hazard does not make clear is how high that potential needs to be before we consider something as a hazard. The well-illuminated step has such a low potential for leading to harm under normal use that you can consider it trivial. But under inappropriate use, the potential could increase to a non-trivial point.
The level at which a hazard might be thought of as too trivial to manage is one that causes the most publicity. You only have to look at the “conkers bonkers” type stories where a school decides that the conker (or is it the action of playing with conkers?) is a hazard that it must eliminate. On the issue of triviality, Duncan Spencer, safety manager for Waitrose, recommends using reasonable judgement: “For a paper cut in an office, the reasonably foreseeable worst case injury would be an annoying but otherwise trivial injury; within a laboratory dealing with biological hazards, it is reasonably foreseeable that the same incident could result in a serious infection,” he explains. “In the office, the paper cut is trivial and doesn’t demand control; in the laboratory, the context is different and demands good controls.”
Reason (2008) uses the terms “expected hazard” and “unlikely but possible hazards” to distinguish between the hazards we should have planned for and the hazards that need heroes to deal with them. He regards the Apollo 13 explosion, for example, as “unlikely but possible”, making it harder to provide detailed plans in advance.
Each organisation has to decide on its risk tolerance when judging the cut-off point for triviality, and hence the definition of hazard has to include an idea of a sufficiently high potential for a significant level of harm.
“For a paper cut in an office, the reasonably foreseeable worst case injury would be an annoying but otherwise trivial injury; within a laboratory dealing with biological hazards, it is reasonably foreseeable that the same incident could result in a serious infection, In the office, the paper cut is trivial and doesn’t demand control; in the laboratory, the context is different and demands good controls.”
Duncan Spencer, Waitrose (at the time of the original article)
What all this tells us is that, in practice, we use a more complicated definition of hazard. Perhaps the following definition is more useful and realistic: “A hazard is something (including the idea of someone, somewhere, some action) with a sufficiently high potential for a sufficiently high level of harm (to people who should be defined) that we need to do something to reduce the potential for harm or the level of harm — or both.”
The next interesting question to ask is whether some of the examples of hazards provided by organisations we look to for advice fit this definition. Looking at the HSE’s example risk assessments, the hazard columns includes falling from height, slips and trips, workplace transport, manual handling, hazardous substances, electricity and welfare.
Though workplace transport, manual handling, hazardous substances and electricity all fit the definition of hazard, falling, slipping and tripping are all consequences (albeit intermediate consequences) of a hazard, not themselves hazards. And welfare is surely a control measure, not a hazard.
The IOSH Managing Safely course lists 17 hazards in its module on identifying hazards. Some, such as chemicals and electricity, fit our definition of hazard. But is “getting in and out” a hazard? The locked fire exit or the abandoned boxes blocking an entrance are the hazards. “Housekeeping” is not the hazard, it is the control, and “stress” is the outcome of hazards such as bullying, work pressure and poor job design. Perhaps what these bodies are trying to do is suggest categories under which assessors should try to identify hazards. But such examples do not help clarify what the word “hazard” really means.
Once we have a clear understanding of what we are looking for, we need to know where we are looking, and which locations and work activities are to be included. Looking only at the tasks may miss a location-based hazard, such as friable asbestos. Thinking about only the location-based hazards focuses on static hazards, and could miss those associated with temporary conditions or activities.
The scope should also define who will be covered, both as sources of hazard and as receptors of harm. The workforce is always included, but how much consideration do we give to contractors, customers, visitors and perhaps people passing by a workplace?
The HSE’s five steps recommend identifying hazards by walking around; asking employees; referring to information from the HSE, trade associations and manufacturers; looking at accident and ill health records; and thinking.
Duncan Spencer and Chris Jerman, corporate safety manager at the John Lewis Partnership, have described a task-based method for identifying hazards which starts by looking in detail at the activities carried out in the workplace. Other methods rely on categories of hazard, used as prompts, but with the scope for omissions and overlap. See Schemas for hazard identification for some category schemas used to prompt hazard identification.
In assessing how you can eliminate or control a hazard, clarity about what that hazard is must be always be the first step. If the hazard is an open drawer of a filing cabinet, you can eliminate it by closing the drawer. Such a simplistic approach, however, may actually miss the real problem. Define the hazard as having the cabinet in a position where the drawers open onto a busy corridor, and a more effective control would be to move the cabinet to a quieter place.
If you are to select the most effective risk controls, understanding the answer to the basic question “What is a hazard?” is always essential. In the second article in our “What is…?” series, we will look at near misses.