The borrowed risk assessment
Northgate Distribution runs a 1,200 m2 ambient warehouse with 38 staff plus agency pickers. In March the company built a mezzanine pick floor served by a steel staircase and installed a new conveyor. The operations manager, three months into the job, produced a risk assessment by taking the document from his previous employer (a food packing plant), changing the logo and site name, and leaving the hazard list untouched. Every hazard is scored likelihood 3 x severity 3 = 9, described as 'medium, monitor'. The document is signed by the operations manager alone and filed. Three weeks later an agency picker trips on a loose conveyor cable at the foot of the stairs. She is unhurt, but the cable has been loose since installation. Her supervisor says: 'Trips are on the assessment, it is rated medium, so we are covered.'
1. Is this risk assessment fit for purpose? Justify your judgement and state what a suitable and sufficient assessment for this mezzanine would need to contain.
Read the worked answer
No, and the justification matters more than the verdict.
- Hazards are not this workplace's hazards. A food plant's list will not carry mezzanine edge protection, falling objects onto pickers working below, conveyor nip points, stair traffic, or the loading pattern of the new floor. Copying the document ensures the obvious is present and the relevant is absent.
- Nobody has been identified as exposed. Agency pickers are a distinct group with different induction, supervision and language needs; the assessment shows no awareness of them.
- The controls are generic. 'Monitor' is not a control. A control has to be a physical or procedural thing someone can check.
- The scoring shows no thought. Identical 3 x 3 for everything means likelihood and severity were never genuinely considered, so the document cannot rank anything.
- It was produced and signed by one person, in isolation. No consultation, no second pair of eyes, no competence check recorded.
- It is not connected to the floor. A cable loose since installation and an accident three weeks later shows the assessment is a filing exercise.
A suitable and sufficient assessment would identify the hazards of this installation, who might be harmed and how (including agency staff and anyone working beneath the mezzanine), evaluate the risks, set proportionate controls, record the significant findings in a form the workforce can use, name an owner and a review trigger, and be communicated to those exposed. Suitable and sufficient does not mean perfect, but it does mean it was made for this building by people who looked at it.
2. What is wrong with scoring every hazard 3 x 3 = 9, and how would you get a 38-person site with very little management spare time to a prioritised action plan?
Read the worked answer
A score earns its place only by discriminating. When everything scores 9, the matrix has stopped doing the one job it has: telling you where to spend the next hour. It also flattens severity, which is the half of the sum that matters most. A fall from the mezzanine is a potential fatality regardless of how likely it is; a stubbed finger on the stair is not. Averaging them into the same number is a design fault, not a data problem.
How to fix it without a three-month project:
- Re-anchor the scale. Agree in plain words what severity 1, 3 and 5 mean here (first aid / over-7-day injury / fatal or disabling), so two people scoring the same hazard land in the same place.
- Score with the people who do the job, not about them. Twenty minutes with two pickers and the shift supervisor will surface more real hazards than an afternoon at a desk.
- Narrow the scope. Do not re-score the world. Take the five to eight hazards with genuine potential for serious harm, plus anything the accident has exposed.
- Sort, then override with judgement. Rank by score, but never let a number talk you out of a known serious hazard.
- Convert to an action list, each line with a control, an owner and a date, working down the hierarchy: re-route and clip the cable in trunking (eliminate/engineering) before 'remind staff to look where they are walking'.
- Set review triggers as well as a date: new equipment, new agency cohort, any incident, any change of process.
3. What should happen to the existing signed document and the loose cable, and who needs to be involved in the replacement assessment?
Read the worked answer
The cable is dealt with today, physically: clip it, trunk it, or take it out of the walkway. An assessment is not a control and cannot be cited as one. Temporary measures are fine for a day, not for three weeks.
The document is withdrawn, not shredded. Tell the people who have seen it that it is being replaced and why, so that a second document appearing does not erode trust further. Keep the old one as a record of what the organisation knew and when, because that is exactly what it is.
The replacement should involve the shift supervisor, two or three pickers including an agency worker, and the maintenance technician who installed the conveyor; someone with risk assessment competence should facilitate and sign it off, and the operations manager should own it rather than author it alone. Consultation is not decoration: the picker who found the cable loose knew about it, and would have said so in a ten-minute conversation.
Finally, communicate the significant findings to everyone exposed, agency workers included, through the induction and the shift briefing, and set a review date with a named owner. The test of the new document is not that it exists but that the supervisor can point to the control it demands when standing in front of the cable.