Zero RIDDOR Is a Lagging Number

The leading one I watch is how often work is paused without a row.

Zero RIDDOR looks tidy on a client report. It is a lagging number. It tells you what has not been recorded yet. It does not tell you whether the next interface will be stopped before it becomes an event.

The leading measure I actually watch on commissioning and multi-contractor sites is simpler: how often work is paused without a row.

A commissioning afternoon, not a dashboard

Late in a recycling-line commissioning package two contractors were working adjacent scopes. One was fitting and proving guards on a conveyor. The other was preparing a pressure test on connected pipework. Both method statements had been accepted. The SIMOPS register existed. The programme was tight.

A banksman on the conveyor side saw the test team moving into a position that closed the exclusion the briefing had described. He stopped his own crew and called it.

Twelve minutes. Zone reset. Briefing updated. Test restarted under the same permit with the change written on it. Nobody raised their voice. The client who happened to be walking the area asked what had changed and accepted the hold.

That pause will never appear on a RIDDOR return. It is exactly the signal I look for. If the same call had been talked down because the crane slot was booked or “we’re nearly there”, the lagging chart would still have read zero — until the week it didn’t.

Why a clean lagging chart is a poor control

RIDDOR, AFR and lost-time rates are outcomes. They are easy to graph and easy to celebrate. They move after the fact. By the time they deteriorate the system has already failed or luck has expired.

On live commissioning the useful data arrives earlier and looks like friction if the only story you tell is progress:

• How often work is paused and the restart is treated as a second decision, not an argument.

• How often a permit is held or refused at the point of work because the method in front of people does not match the plant in front of them.

• Whether briefing challenges come from the people who will do the task, or only from the person running the huddle.

Those numbers look like delay if you only report output. They look like control if you have watched a plant that ran on paper and then had a serious event in week three of commissioning.

Three leading indicators that actually predict behaviour

Cornerstone Safety Limited · Professional discussion only — not legal advice · CDM 2015

1. Pauses without a row

Not theatrical stand-downs that become investigations. Short, specific holds that are accepted, fixed and restarted under control. Count them. Listen to the tone. If the first pause of the month produces a sigh from the client’s representative, you do not have stop-work. You have a slogan.

2. Permits held or rejected at the board

A permit that never comes back is either perfectly written or nobody is reading the job in front of them. A healthy board has occasional “not yet”. That is competence being exercised, not bureaucracy.

3. Challenges from the tools, not only the supervisor

If the briefing is a monologue, you have compliance theatre. If an operative asks “what about the interface with the other package?” and the question is welcomed and the method changes, you have a live system. That is the same discipline high-hazard plants treat as normal.

The contrast I keep on one page

Lagging (what gets reported upwards): RIDDOR = 0. AFR trending down. Audit score 92 percent.

Leading (what I actually watch): four pauses this week, two permits held overnight, three briefing challenges that changed the written method.

The first set is comfortable. The second set is the control. Sites that only celebrate the first set train people to keep going.

How to use it without another dashboard

You do not need software. A weekly note from the supervisor or package lead is enough: number of holds, number of permit queries at the board, one sentence on the most useful challenge in a briefing.

Review it in the same meeting where you look at progress photographs. If the leading numbers sit at zero for three consecutive weeks while the programme is aggressive, that is the risk signal — not a success story.

Do not add a new KPI to the construction phase plan and then ignore the first pause because it is inconvenient. The measure only works if the pause is treated as the job, not as a problem.

What this is not

This is not an argument against reporting RIDDOR or tracking AFR. Those remain legal and contractual requirements. It is an argument against using a clean lagging chart as evidence that the system is under control.

Zero events can be the product of good control. It can also be the product of silence, luck and a programme that has not yet been tested by an interface.

A clean lagging chart is comforting. It is not the same as control.