Shutdown Planning: The 8 Things Most Sites Get Wrong

Why we keep seeing the same failures

A shutdown that overruns is not just an operational headache. By the time you factor in the production loss, the overtime, the contractors standing around waiting and the snag list that follows you for the next three months, the cost is usually multiples of the original budget. And yet the same mistakes keep happening on sites that have been running shutdowns for years.

The reason is not incompetence. It is that the maintenance team is usually managing normal operations right up until the shutdown starts, the planning gets squeezed into the gaps, and the problems only become visible when it is too late to fix them. Here are the eight we see most often.

1. The scope was never actually frozen

This one causes more shutdown overruns than anything else. A scope gets defined, then over the following weeks jobs keep getting added — "while we're in there" becomes the most expensive phrase in engineering. By the time execution starts, there is 90 hours of work in a 60-hour window and the team is behind before they have started.

You need a hard scope freeze, typically four to six weeks out, after which new work only gets added if something else comes out or extra resource is confirmed. It feels rigid. It is what makes the difference between a shutdown that finishes on time and one that does not.

2. The contractors were booked too late

Specialist contractors — crane teams, electrical testing, inspection services, specialist fabricators — have busy diaries. If you are calling them five weeks before your shutdown, you will either get whoever is available, pay a premium, or be told they cannot make your dates. We have seen shutdowns rescheduled because the right contractors were not available. Book critical specialist resource at the same time you fix the date.

3. Nobody checked whether the parts were actually there

This one sounds too basic to include. We include it because we have seen it happen on sites that absolutely should have known better. The shutdown starts, someone goes to pull the seal kit for a pump that has been on the scope for eight weeks, and it is not there. Lead time is three weeks. The pump does not get done. The shutdown window is partly wasted.

Spares and materials management for a shutdown is a separate workstream. There should be a named person, a confirmed kitting list, and a physical check that everything is on site before execution begins. Not the day before — at least a week out.

4. The risk assessments were written the day before

We understand why this happens. The planning is frantic, there is always something more urgent, and safety documentation gets pushed to the end of the list. But a risk assessment written in a hurry the night before a shutdown is worth very little. It will not cover the actual complexity of the work. It will not reflect the specific isolation arrangements. The people doing the work will not have seen it.

Safety documentation for a shutdown needs to be part of the planning process, not a box-ticking exercise at the end. High-risk tasks — confined space entry, work at height, high-voltage isolation, hot work — need to be identified early and planned properly. Shortcuts here are where serious incidents happen.

5. There is nobody solely responsible for coordination during execution

On a busy shutdown with multiple contractors working in parallel, you need one person whose entire job for that window is managing the execution. Watching the schedule, sequencing the trades, resolving the conflicts that will inevitably arise and making decisions quickly when things change.

That cannot be the maintenance manager who is also fielding calls from operations about normal plant issues. It cannot be the site engineer who has their own scope of work to deliver. It needs to be a dedicated coordinator. Without one, you get drift — work runs in the wrong order, contractors wait on each other, problems escalate slowly. Hours disappear.

6. The schedule has no contingency and everybody knows it

When a shutdown schedule is built on optimistic task durations with no float, two things happen. The team knows it is unrealistic from the start, which affects morale and planning discipline. And when the first unexpected issue hits — and there will be one — there is nothing to absorb it.

Every shutdown we have ever been involved in has had at least one significant unexpected finding once plant was opened up. Build that expectation into the schedule. How much contingency depends on the age and condition of the plant, the quality of your condition data, and how long it has been since you were last inside these assets. If the honest answer is "we do not really know what we are going to find," build in more.

7. Isolation management is being handled informally

On a single-trade, single-task shutdown this is less of an issue. On anything with multiple contractors working in parallel on the same or adjacent systems, informal isolation arrangements — where individuals are making their own assumptions about what is isolated and what is not — create serious risk.

We are quite blunt about this when we see it. A formal lock-out/tag-out system is not optional when you have three different contractors working in the same area. The paperwork feels like overhead right up until it prevents someone from getting seriously hurt.

8. The shutdown ends and everyone moves on

There is always pressure to get back to normal as quickly as possible after a shutdown. We get it. But closing out without capturing what you learned means you will have the same conversation about the same problems next time round.

A shutdown review does not have to be long. An hour with the key people, a one-page summary of what went to plan and what did not, and three or four specific things to do differently next time. Do that consistently and your shutdowns will genuinely improve. Skip it and you are starting from scratch every time.

What good preparation actually looks like

The sites that run the best shutdowns we have seen all do the same things. They start planning early — eight to twelve weeks out for anything significant. They freeze the scope and defend it. They treat spares, contractors and safety documentation as parallel workstreams, not afterthoughts. And they have one person who owns the execution from start to finish.

None of that is complicated. The challenge is applying that discipline when you are already stretched running normal operations — which is exactly why having external support for shutdown planning and coordination delivers real value.