Most turnaround schedule failures are predictable. They are the result of conditions that were present and visible weeks before the first wrench turned. By the time the project is running behind, the window to do anything about it has usually already closed.
Here are the five signs we look for when we arrive on a new turnaround.
1. The schedule was built top-down, not from field commitments
A schedule built by a planner and validated by managers is a set of assumptions in Gantt format, not a field schedule.
The test is simple: ask a foreman or lead craftsperson whether they were in the room when their work scope was estimated. If the answer is no (if they were told what the schedule says and asked to confirm it), the durations are guesses. Confident ones, often, but guesses nonetheless.
Commitment-based scheduling starts from the opposite end. The people who will do the work estimate their own scope, in specific terms, against their actual crew and material status. That process surfaces blockers that top-down schedules never see, because the people who know about the blockers were never asked.
2. The critical path is invisible to the field crews
If you ask a supervisor what their float is, and they can’t tell you, the schedule is already in trouble.
A critical path that lives in the planner’s software and gets reported out in a weekly status slide is just a document. For the critical path to function as a management tool, the people who own the critical tasks need to know they own them, and understand what a one-day slip in their area costs the overall schedule.
This has nothing to do with distributing a PDF. What matters is whether the field understands which work has zero tolerance for variance, and why. When that understanding is absent, normal daily decisions (sequencing adjustments, crew reallocations, deferred punch items) get made without any awareness of their downstream cost.
3. The risk register is managed but not used as a live tool
Every turnaround has a risk register. The question is whether anyone opens it between the pre-job kickoff and the post-job lessons-learned.
A risk register that is updated weekly by a project controls analyst and presented in a monthly report is a compliance artifact, not a field risk management tool. By the time a risk in that register has become a problem in the field, it has typically been visible to foremen and supervisors for days, who had no mechanism to escalate it and no reason to believe anyone would act on it if they did.
The distinction is between a register that tracks risk and a process that uses risk information to drive daily decisions. The second version gets discussed in standup. Items that move toward likelihood or impact get flagged and owned before they become schedule events.
4. Permit status and material status are not on the daily agenda
The two most common reasons work cannot start when it is scheduled to start: the permit isn’t in hand, or the material isn’t on site. Both are knowable in advance. Neither is consistently on the daily agenda in most turnarounds we have walked into.
A permit that has been submitted but not yet approved does not let work start. Neither does a part that has been ordered but is still in transit. Any scheduled task that depends on either is a task with a hidden constraint, and that constraint will become a delay the moment the crew shows up to work.
The daily standup should have a live view of open permits and critical material status for the next 72 hours. If that information is not being tracked at that cadence, the first sign of trouble will be a crew standing idle while the permit is expedited.
5. There is no recovery plan before work starts
If the question “what happens if we are three days behind at the midpoint?” does not have an answer before the turnaround starts, it will get answered under the worst possible conditions: mid-execution, with full crews on-site and commercial pressure bearing down.
Recovery planning is not pessimism, it is what separates a team that can respond to variance from a team that can only react to it. What tasks can be done in parallel if schedule margin disappears? Which work can be deferred without affecting mechanical completion? What crew reallocation options exist? What is the overtime authorization threshold and who approves it?
These decisions should be made cold, before the work starts, not hot, at the crisis point.
None of these signs require a failed project to appear. They are visible in the pre-job schedule, in the kickoff meeting, in a 30-minute conversation with the field supervisors who own the critical path. Finding them early is what makes them fixable.
The same conditions that produce schedule failure also produce cost overruns and extended unit downtime. None of it is inevitable. It all traces back to a planning process that never pressure-tested its assumptions against the people who would have to execute them.