Some plants sign the labor contract for their outage a full year ahead.
“I know facilities that are booking their outage personnel,” Jason Adkins says, “signing a contract for 90 guys a year before the outage, saying you will provide 90 people one year from now for these three weeks.”
That fixes the dates, the headcount and the window. What is still open is the order of the work, and how much of it has to happen inside the window at all. That is outage sequencing engineering. Every one of those ninety people is paid whether their work is ready or not, and on a coal unit that is staying online longer than anyone planned, the sequence is where the schedule is won or lost.
In power generation, the currency is outage days
Jason Schwartz has spent thirty years in bulk material handling, and he is direct about what power generation buyers weigh. “Most of my power gen people don’t really care about losing money. It’s all about turnaround and schedule. How fast can we get the job done? How much rework can we avoid?”
That is not carelessness about cost. It is an accurate reading of where the cost lives. A unit that comes back late is the cost the whole plant feels. Rework inside the window is the worst kind, because it burns days the plant cannot buy back and crews the plant has already paid for.
Coal units that were headed for retirement are now being kept running — DOE’s September 202(c) orders holding Indiana units available past their planned retirement were the fourth set since December 2025, according to the Indiana Capital Chronicle. A plant with a longer future has more capital work to put through the same short outage windows. The windows do not get longer. The sequence has to get better.
The first domino is the measurement
Every outage schedule has a chain behind it, and it starts long before the unit comes down.
Measure the existing conditions. Finish and approve the design. Order the material. Fabricate. Deliver. Install. Each link waits on the one before it, and long-lead material stretches the whole chain.
An OEM project manager on a coal unit retrofit put it plainly during a model review this fall, with material that could be sixteen weeks out: “If we can get these guys on site next week, I think we got a chance of having material on site by March 1st. But we can’t get these guys on site until… the end of October, then that’s going to be in jeopardy, I think.”
A slip of a few weeks in field capture became a threat to the outage itself. Not because the measurement took long. Because everything after it was waiting.
That is the first rule of outage sequencing: the schedule starts at the measurement, not at the shutdown. Treating the verified basis as an early milestone, rather than a task to squeeze in when the design team asks for it, is the cheapest schedule protection a project has.
Move work out of the window
Once the existing conditions are measured and modeled accurately, a different question opens up. Adkins frames it the way a planner should: “If I know I have a shutdown that’s coming up and I’ve only got one week, two weeks, three weeks, whatever, to accomplish all this work — how much work can I do ahead of time?”
A lot, usually. With an accurate model of the area, fire protection and piping that sit in the way of the new work can often be relocated before the shutdown. Supports can be set, clearances confirmed and interferences resolved while the unit is still running, when crews are easier to find and the clock is not running on the whole plant.
Every hour of pre-work is an hour that does not have to be done by the most expensive, hardest-to-book labor on the project. When the ninety people show up, the scope in front of them should be the work that truly requires the unit down, and nothing else.
At the end, a rescan of the finished area replaces the redlines. The next outage starts from a record that matches the plant.
Outage sequencing engineering: the order of operations is a deliverable
Inside the window, the sequence itself needs engineering. Not a schedule of durations, but a physical order of operations based on what the structure and the systems will allow.
- What comes out first, and what has to stay standing while it does. We have seen what happens when a demo crew is handed drawings that show what goes but not what stays — that story is in conveyor transfer point retrofit engineering.
- What temporary support carries the load in the hours between removal and installation.
- What the rigging path allows, in what order, through the openings that actually exist once demolition is done.
- Which crafts can work in the same area at the same time, and which tasks must finish before the next one can start.
- Where the installation is simply the demolition run backward, so the crew sees the same steps in reverse instead of solving a new puzzle on the deck.
All of it depends on the same thing: a basis measured from the plant as it is. A sequence built on the record drawings inherits every change nobody drew, and the first person to find the difference is a crew lead with a crane waiting. We covered why record drawings drift in Engineering Against a Plant That Moved.
Four weeks’ notice is not a plan
The biggest sequencing problem is usually not technical. It is when the engineering gets called.
Adkins has spent years retraining utility customers on this. “Early on I had a lot of customers that would call me a month before the project starts, and it’s like, geez man, thanks for the heads up.” His ask now is simple: “Don’t call me four weeks before the project starts. I need to know what you’re doing in the spring, three and four months ahead of time.”
His target is six to eight months ahead, and at least one utility has started pulling dollars forward to get areas scanned and modeled before the outage rather than during it. The payoff is not only a cleaner schedule. It is that the pre-work gets done months early, which frees outage manpower — and outage manpower is always the constraint.
Plants that book ninety people a year out already understand that labor has to be planned early. The engineering that tells those ninety people what to do, in what order, belongs on the same calendar.
The window is fixed. The order of work is the part you still control.
Engineer the retrofit around reality.
→ Fabrication and installation support · Industrial retrofit engineering
Acensium is an industrial engineering firm for complex brownfield retrofits — bulk material handling, high-energy piping, and equipment replacement in operating facilities. Founded in 2003, 1,300+ projects delivered. We connect verified field conditions with practical design, analysis, fabrication control and installation planning.