Making a shutdown window productive

A shutdown — a weekend, a holiday period, a planned outage — is the only opportunity to do work that cannot happen while the building is in use. It is also a hard deadline: the building reopens whether the work finished or not. For a separate product-side perspective on the difference between output and efficient delivery, this comparison provides additional detail.

Shutdowns overrun for consistent and avoidable reasons, and almost all of them are decided before the window opens.

Freeze the scope, early

Scope arrives continuously in the run-up to a shutdown, because a shutdown is where everyone's deferred item wants to go. Each addition is individually small and the accumulation is what causes the overrun. For a wider operational and compliance reference, consult OSHA lockout/tagout guidance.

Set a date after which nothing is added except genuine emergencies, far enough ahead that materials can be ordered against the frozen scope. Two to four weeks is typical for a small site. Late additions do not just add their own duration; they disrupt a sequence that was planned around a different set of tasks.

Stage everything before the window opens

Time spent during a shutdown collecting parts, finding tools or waiting for a delivery is time that cannot be recovered. Every item should be on site, checked and staged in the right place before the window starts.

Checked is the operative word. A part that arrives in the wrong specification is discovered at the point of installation, which in a shutdown means the task does not complete.

Sequence by dependency and by isolation

Two constraints govern the order. Dependencies — what must finish before something else can start — and isolations, since work sharing an isolation should be grouped so the system is drained, locked off and recommissioned once rather than three times.

Recommissioning is the step most often underestimated. Refilling, venting, pressure testing, restarting and proving a system takes substantial time, and it must complete before reopening. Work backwards from the reopening time through recommissioning to find the real deadline for tools-down, which is usually several hours earlier than people assume.

Decide the abort points in advance

For each significant task, agree beforehand the time at which, if it has not reached a defined stage, it will be reversed and rescheduled rather than pushed.

This is the discipline that prevents the classic shutdown failure: a task half-completed at the reopening deadline, with the system neither in its old state nor its new one. Deciding the abort point calmly in advance is far easier than deciding it at four in the morning.

Plan the hours as well as the work

Shutdowns are worked by people who are also working the rest of the week. A weekend shutdown followed immediately by a normal Monday produces a tired team, and rest requirements are a real constraint rather than a preference — in many jurisdictions a legal one.

Roster the shutdown as a shift pattern with defined breaks and a defined finish, and adjust the surrounding week. Shutdowns staffed by everyone working until it is done reliably produce both an overrun and a poor week afterwards.

Debrief within the week

What ran over, what was found that nobody expected, what was staged badly, what should not have been in scope. Held immediately, this produces specifics; held a month later it produces impressions.

The output is a better estimate for next time, which is the only mechanism by which shutdown planning improves. Most organisations run the same shutdown annually and re-learn the same lessons each year.