Chapter 3: Identify the hazards - doing it differently
3.1 Risk factors
3.2 Learning from fire risk assessments
3.3 Application to other hazard types: slips and trips
3.4 Application to health, welfare and wellbeing
3.5 Application to complex problems
3.6 Application to COVID-19 (other pandemics are available)
3.7 Better hazard identification – beyond I-spy
3.8 Conclusion
3.1 Risk factors
In Chapter 1, Figure 1.1 made the distinction between a hazard, a hazardous event and the consequences. For this chapter we need another term – risk factors. The HSE uses the term a lot, particularly in relation to the assessment of manual handling tasks, but without defining it. The HSE gives examples which hint at the intended meaning:
- Physical risk factors such as force, posture and repetition
- Psychosocial risk factors including high workloads, tight deadlines, lack of control of the work.
Most definitions of “risk factor” come from medical sources, where it is defined as something that increases the likelihood of developing a particular disease. For our purposes a useful definition includes those factors that increase, or decrease, the likelihood or severity of a harmful consequence. As such, the risk factors can apply in multiple places:
- In making it more or less likely that, given there is a hazard, a hazardous event occurs (pre-hazardous event risk factors).
- In affecting what happens once the hazardous event has occurred (post-hazardous event risk factors).
Figure 3.1 illustrates this in principle, and examples in this chapter provide concrete examples. While risk factors have an impact on the consequences, they are not the hazards themselves – although they are often labelled as such in risk assessments. Understanding the risk factors – and whether they are relevant before or after the hazardous event occurs – is important when we later try to control the risk (Chapter 11 onwards). Note that in the diagrams in this chapter the risk factors can often be considered in increasing detail, but have been simplified for our purposes.
Figure 3.1: Adding risk factors to the model
3.2 Learning from fire risk assessments
A former head of health and safety for the London Fire Brigade, John Norton-Doyle, explained to me some years ago that he sees fire risk assessments as two risk assessments. He said:
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.
The first FRA is a reasonably straightforward exercise, which in a low-risk environment can be carried out by someone with a reasonable grasp of risk assessment and ability to recognise things that might burn, things that might start a fire, doors and windows, and things marked with an oxidiser symbol.
The second FRA often needs more expertise. How do detection and alarm systems work? How often should you check that alarms and emergency lighting are working? How much emergency lighting, and how many signs do you need? There are British, European and ISO standards covering many technical aspects. Understanding which standards you need requires experience and a higher competence level, and purchasing the relevant ones a sizeable budget.
With the first risk assessment hazards are all the things that can burn (oil, paper, plastic), all the things that could ignite (electrical appliances, cigarettes, hot surfaces) and where relevant, anything that will increase the oxygen content (solvents containing oxidising chemicals, open windows and doors). In the second FRA, a column marked “hazard” could technically contain only the word “Fire”. Since that doesn’t help to structure the rest of the risk assessment, the first column is often relabelled as “issues” or “fire protection measures.” The lack of a smoke detector is not a hazard, but a fire protection measure that needs to be assessed. More generically, the presence or absence of a smoke detector is a post-hazardous event risk factor.
Figure 3.2 uses the model from Figure 3.1 for fire. In this case, three hazards have to occur together for a fire to start. Once a fire has started, the three darker green boxes indicate the areas where we would look for post hazardous event risk factors that individually or in combination could increase the chance that people will be harmed by the fire. Not evacuating or not knowing about fire are not hazards. They are only a problem if threatened by the hazard of fire.
Figure 3.2: Simplified structure of a Fire Risk Assessment (FRA)
Ask yourself:
Table 3.1 below shows the table we used in Chapter 1 with an extra column for the risk factors. What risk factors would you add for each hazardous event? See Appendix 2 for suggestions.
Table 3.1: Including risk factors to understand hazards
| Hazard | Hazardous event | Post-event risk factors | Consequence |
|---|---|---|---|
| Oil on floor | Slip and fall | Injury eg broken bones | |
| Uneven floor surface | Trip and fall | Injury eg broken bones | |
| Vibrating tools | Prolonged use resulting in excessive exposure to vibration | Numbness, reduced grip strength, HAVS | |
| UV from welding | Eyes exposed to UV light | Eye damage |
3.3 Application to other hazard types: slips and trips
This idea can help us with other common hazard groups.
Rather than the lazy approach of sticking “slips and trips” in the hazard column, and treating the twin imposters just the same, the flowchart in Figure 3.3 takes a similar approach to the FRA in Figure 3.2 – what could cause the slip or the trip, and what happens once some starts to slip or trip?
Figure 3.3: What leads to a slip or trip (the hazardous event), and what happens next?
In Figure 3.2 we needed all three hazards to start a fire, but in Figure 3.3 either hazard could lead to an accident. The model reminds us to look in more detail for both things that might leak or spill (drinks, oil from machinery, water from a sink) and all the things that might be obstacles (tool boxes, personal belongings, deliveries). This provides a better understanding of the hazards I need to manage than the vague label “slips and trips”.
Then, assuming the hazardous events (slipping or tripping) occur, consider all the risk factors that make someone more likely to lose their balance, and to hurt themselves if they fall. Poor lighting or carrying a bulky load might make someone less stable. Individual factors like balance and infirmity, and environmental factors like the presence of a handrail or grab rail, will impact whether or not they fall.
3.4 Application to health, welfare and wellbeing
As we identified in Chapter 1, the HSE instruction to “Look around your workplace” encourages the identification of physical objects as hazards. Putting “health, welfare and wellbeing” into the hazard column will not overcome this.
You could instead put the negative consequences “wellbeing suffers” at the end of a flow and work backwards to identify the risk factors and the hazards that kick the process off. Figure 3.4 provides an example for one aspect of wellbeing, reminding us that identifying the causes of stress is more useful than offering counselling and free fruit (more on this in Chapter 11).
The three psychosocial hazards shown as leading to stress in Figure 3.4 are from the longer list of factors suggested by the HSE (see Box: HSE management strandards). You could consider all of them and document the ones which are relevant in your organisation. The key point is that having some structure will help you to identify hazards you can manage – putting “stress” into the hazard column will not.
Figure 3.4: What can lead to stress, and what happens next?
Box 3.1: HSE management standards
The HSE management standards provide a good checklist for identifying hazards that can lead to stress-related conditions. These are listed here, but I encourage you to look at the detail on the HSE website:
Demands: look for excessive or insufficient workload (being bored can be stressful too), work patterns and the work environment
Control: people are better at coping with demands they have some control over, than demands placed on them by others.
Support: knowing that our boss and our colleagues are on our side makes it easier to cope. What factors in your organisation prevent people from supporting each other? For example, competitive targets, or lone-working. Has this become more of a problem with a pandemic-led increase in homeworking?
Relationships: this overlaps with support, but includes how organisations deal with unacceptable behaviour such as harassment and bullying.
Role: asking someone to do a job quickly “but do it safely” sets up a conflict in their role. Not being told what you need to do is equally stressful.
Change: I’ve seen so many organisations get this wrong, from large corporates, through government departments, to small owner-manager businesses. Senior management seem to think no one will notice that change is being discussed, so they choose not to communicate with workers. People would rather know which jobs are at risk, even if it’s their job, than be left in the dark.
For other health hazards it is useful to construct your own checklist of what might be relevant in your organisation. For every task or location you could remind your self to look for hazards in those categories. As a starting point, consider exposure to vibration and noise and look for hazards that can lead to respiratory, skin or musculoskeletal disorders.
3.5 Application to complex problems
While stress, noise, vibration and some other health issues involve hazards that are not easily seen with the I-spy approach, the topics are well-documented, and there is plenty of guidance available to help you structure your hazard identification. Some abstract hazards are less well documented and will be very dependent on your circumstances. In our structured “what if” example for making tea in Chapter 2 we could have considered the impact of emotions. What if someone was too angry, too tired, too lazy or even too curious? Anger and fatigue can lead to errors and carelessness; being lazy might result in missed steps. Being too curious might lead a worker to experiment with a process. That might be alright when making tea, but what would the impact be on a safety-critical process? See Uddin v Associated Portland Cement Ltd (1965) for where curiosity led to disaster.
Figure 3.5: What makes people angry, and what happens next?
Figure 3.5 considers anger within hazard identification. People making a claim for benefits sometimes become aggressive towards the claims handling staff. If you identify the claimants as the hazard, an engineering solution is to put physical barriers between the claimants and the staff. In practice, this tends to make relationships worse.
If you treat “anger” like “fire” you start by asking what hazards create anger – for example, being kept waiting without any information when you are already worried about how you are going to pay the heating bill this month.
The second part of the assessment assumes the first part has failed – that someone is angry, but that mitigation is still possible before someone is harmed.
The better we understand the hazard path, the better we can control the hazards when we get to Chapter 11.
3.6 Application to COVID-19 (other pandemics are available)
There was a virtual avalanche of new risk assessments in 2020. Impressively, rather than the annual token review, these were updated regularly – one colleague proudly explained they were on version 23, within a year of the first risk assessment. These were the risk assessments that helped us to get back to work, to school, to scouting, to church and back to living life in the new normal of the COVID-19 pandemic.
We knew what the hazard was: it was COVID-19. We knew that it could kill, but we didn’t know for any cohort of people how likely it was that anyone had it already and could transmit it, nor did we know how resilient people would be to the virus if they did catch it. We had to assume that it was possible, and that although many of the measures needed under normal circumstances would not have been “reasonably practicable” (see Chapter 5) they had to be done anyway. So restaurants had to open with fewer customers than was economically viable, public transport had to run with fewer travellers, and workplaces had to operate with fewer staff.
The UK HSE soon produced a risk assessment example for organisations to use as a model. Disappointingly, the HSE defined the hazards as the lack of a fixed list of controls. For organisations following this template, the conclusion of their risk assessment was pre-determined – they needed to apply social distancing, physical separation, hand washing and workplace cleaning regimes, without regard to practicability.
Since putting “COVID-19” in the hazard column wouldn’t provide sufficient structure, more experienced risk assessors looked at other ways to structure their risk assessments. So the first column might be locations around a building where people could come into contact with each other. Or it could be steps in a process, not dissimilar to a task analysis suggested in Chapter 2.
The flowchart model suggested in this chapter here might have provided the structure needed – see Figure 3.6.
Figure 3.6: Why would an infected person come into the workplace, and what happens next?
In Figure 3.6 the hazard is the COVID (or other) virus, and the hazardous event (the equivalent to fire in our FRA example) is the presence of an infected worker in the workplace. If you could be sure no one infected was in the workplace, you wouldn’t need any further precautions.
This prompts us to consider two sets of risk factors:
- What might cause someone to come into work with the infection? For example, they might not be able to afford to stay at home if they are on minimum statutory sick pay, or have no sickness benefits at all. Before testing was available, the pre-hazardous event risk factors would have been different, focussing on awareness of symptoms, and knowledge of being in contact with someone else infected. This should help you to consider the support you provide to workers to test, and to stay away from work when infected.
- Once the hazardous event has occurred – an infected worker is in the workplace – how would someone else contract the same infection? Most sources suggest there are three routes for infection – via contact surfaces, via the air (where the virus is assumed to remain in the air for a period) and directly, for example when someone infected coughs particles onto another person. Now you have a structure to identify specific workplace hazards.
With this information you can identify where controls such as cleaning and ventilation are needed:
- Surfaces which are likely to be touched by two people in succession,such as door handles and light switches.
- Locations where people are breathing in the same air, particularly in smaller, less well-ventilated rooms.
- Situations where people are in close proximity For example, a team manual handling task, or corridors at busy times of day.
Visualising this structure makes it more likely we can identify specific hazards in our workplaces. As we’ll see in Chapter 11 onwards, it also makes it easier to see where controls will be most effective.
3.7 Better hazard identification - beyond I-spy
Too often I’ve been told “it hasn’t happened before, so it can’t happen.” While something not having happened before can be some defence in court (as in the case of the child jumping off the steps discussed in Chapter 1) it is short-sighted if not tested further. To encourage better hazard identification, get a team together (or at least one other person), and try some of these questions:
- What’s the worst that could happen when you do this task?
- What would have to go wrong for someone to get killed or injured?
- What does successful performance of this task look like?
- What are the things that stop something bad happening?
Another approach is to carry out a “pre-mortem”. Imagine the worst has already happened, and ask “If someone was injured, what questions would we be asking?” For a task like using a set of steps to reach a shelf in a storeroom, questions might include:
- Was the ladder damaged? Is there a process in place to check for damage to ladders?
- Was the floor uneven? Could the floor become uneven?
- Did someone (accidentally or deliberately) push the ladder? What measures are in place to stop the ladder being pushed?
Beyond I-spy - the boxes near the staircase are an obvious hazard, but I-spy won't help you to work out if the evacuation arrangements are sufficient
3.8 Conclusion
In Hazop and Hazan Trevor Kletz defines a hazard as “a substance, object or situation that can give rise to injury or damage.” Immediately, there is a clearer emphasis on substances, which often result in health hazards, over objects that more often relate to safety hazards. The inclusion of the word “situation” also prompts the identifier of hazards to look beyond what they can currently see. Where a hazard could be an object or a situation, which you put in the risk assessment depends on the context. For a risk assessment of a location, the hazard might be an object such as a ladder or a roof. For a risk assessment of a job of work, the hazard might be a situation, such as working from a ladder or climbing onto a roof.
The key message here is to recognise that hazard identification is not a single step. You must be clear about what you are assessing, for example using a task description or inventory. You need to consider the hazardous events and possible consequences, iteratively evaluating the severity and likelihood of those consequences to determine if there is a non-trivial hazard to control. And in all of this, consider the full range of people who could be harmed, something we look at in more detail in Chapter 4.
You can use the Contact form to send me feedback. If you’d like to receive an email when I add or update a chapter, please subscribe to my ‘book club’
Alternatively, go back to the book contents page
Appendix 1: Case studies by year
Appendix 2: Answers to questons posed in each chapter
Appendix 3: Lost HSE references