The scenarios below are composites. They are not real projects and no detail identifies anybody — they are assembled from the patterns that recur across submitted documents and published enforcement, and they are written the way these things actually unfold rather than the way a cautionary tale usually does.
Each one follows the same shape, because these failures do: a small documentary gap, a plausible reason it happened, a consequence out of proportion to the gap, and a check that would have caught it in seconds.
Case one: the control that was never written down
The gap
A fit-out contractor's RAMS for second-fix mechanical work identified "manual handling of ductwork sections at height" as a hazard, scored it 4 × 4, and listed the control as "two-person lift, mechanical aids where practicable, operatives trained in manual handling".
Nowhere did it say what happened to a section while the second operative repositioned the tower. Nothing described how a 1.8-metre length was secured between being lifted and being fixed.
How it happened
The document was adapted from a ground-level installation job where the question did not arise. The hazard line was updated to say "at height"; the control line was not. Both authors and reviewers read the control as adequate because it is adequate — for the job it was written for.
The consequence
A section slipped while being held one-handed and struck an operative below the platform, causing a fractured wrist. A specified injury, reportable under RIDDOR, an investigation, a week of disruption on the floor, and a document that identified the hazard and did not control it — which is the worst combination available, because it establishes the risk was foreseen.
What would have caught it
The mechanical check: does every hazard have a control that addresses the hazard as stated? "At height" appears in the hazard and nowhere in the control. That mismatch is visible without knowing anything about ductwork, and it is exactly the kind of finding that comes back as critical — a hazard whose control does not answer it. See what critical means in a report.
Case two: the muster point from the last project
The gap
A groundworks RAMS for a school extension gave the muster point as "the visitor car park to the front of the main building". The school's fire assembly point, and the one on the site plan, was the playing field at the rear. The car park was inside the works cordon and, at that stage of the programme, held the site cabins and a diesel bowser.
How it happened
Copied from the previous project, where it was correct. The document had been through two revisions, both of which changed the sequence of works and neither of which touched section 10. Nobody had read the emergency arrangements since the day they were pasted in, because nothing had gone wrong.
The consequence
An HSE inspector on a proactive visit asked an operative where they would go if the alarm sounded. The operative gave the site induction's answer — the playing field — which was correct, and the inspector then read the RAMS, which was not. That mismatch turned a look at the excavation into a look at the contractor's document control: three more RAMS were requested, two carried the same muster point, and the visit ended in a notification of contravention and Fee For Intervention costs. No injury, no notice, and several thousand pounds of chargeable inspector time plus a client conversation nobody wanted.
What would have caught it
Reading section 10 as part of the review rather than assuming it. A copied muster point is one of the most common findings there is, which is precisely why it is checked — and why a document whose emergency section has survived two revisions untouched deserves suspicion. Generic versus site-specific RAMS lists the other five tells that travel with it.
Case three: the scope that changed and the register that did not
The gap
A demolition and strip-out package was approved with a scope covering soft strip of internal partitions and ceilings. Four weeks in, the client added removal of a redundant plant room, including breaking out a concrete plinth. The scope section was updated by email agreement and the RAMS reissued at revision 3 with the new paragraph in section 2.
Section 6 was untouched. No hazard for breaking out concrete, so no entry for respirable crystalline silica, no hand-arm vibration exposure, no COSHH assessment reference, and no control beyond the general dust wording written for soft strip.
How it happened
The most ordinary way in construction. The change was commercial in origin, agreed under time pressure, and the person who updated the scope was not the person who wrote the hazard register. Revision 3 looked complete because the visible thing — the scope — had changed.
The consequence
Two weeks of dry breaking out with an FFP2 mask and no on-tool extraction, in a partially enclosed space, before a client-side safety advisor stopped it. No injury and no enforcement — the harm here is the kind that turns up two decades later in an occupational disease claim, which is why it was allowed to run for two weeks and why it is the most expensive of the three. The immediate costs were a stopped activity, a health surveillance referral for two operatives, and a subcontractor removed from the client's approved list.
What would have caught it
The scope-to-hazard cross-check, run again at the revision rather than at first issue. Concrete breaking appears in section 2 and generates no hazard in section 6 — a mismatch that takes seconds to find if anyone looks, and that nobody looks for when a revision is treated as a paperwork step. How often a RAMS should be reviewed covers the triggers, and a scope change is the first of them.
What the three have in common
None was caused by ignorance. In all three the contractor knew perfectly well how to control the risk — the ductwork crew knew how to secure a section, the groundworkers knew where the muster point was, the demolition team knew what silica does. The failure was that the document stopped describing the work, and once that happens the document cannot be briefed, cannot be supervised against, and cannot be defended.
They also share a shape: each was findable mechanically. A hazard whose control does not address it. An emergency section unchanged across revisions. A scope entry with no corresponding hazard. None of those needs a site visit or an engineering opinion to spot — they need somebody, or something, to compare one part of the document against another, every time, without getting bored.
That is the specific thing tools like Valify do: the cross-checks run on every submission, in the same way, and the mismatches come back listed by severity rather than waiting to be noticed. It does not judge whether the method is right — that stays with your competent person, as it must — but a hazard with no matching control, a substance with no COSHH reference and a scope activity that produces no hazard are all findings it returns automatically, which is precisely the class of gap all three of these were.
The cheapest of the three consequences was several thousand pounds. The check that would have prevented it took under a minute.
This article is general guidance on UK construction documentation practice. It is not legal advice, a compliance certification or a substitute for review by your own qualified health and safety personnel, who remain responsible for deciding whether any particular document is adequate. See our disclaimer.