The Screws That Told the Real Story
What two "simple" shutdown projects taught me about the due diligence we skip when we're moving fast
I want to tell you about two projects that ran side by side inside a food manufacturing plant during a shutdown. Two weeks. That was the window. Shut the plant down, get the work done, bring it back online, and don't be the reason production slips.
If you've worked capital projects in an operating facility, you know that kind of schedule pressure. It shapes everything, starting with how the work gets scoped before a single bidder ever walks the site.
Both projects were infrastructure upgrades. Not new equipment, not a expansion. Improvements to existing assets. One was a wall coating replacement in a kitchen area. The other was a full roof replacement, TPO membrane, over a large section of the facility. On paper, neither one sounded complicated. Two straightforward scopes, two weeks, two crews. What could go wrong.
Project One: The One That Went Right
The wall coating project is worth mentioning because of what we did well, not because anything dramatic happened.
We used hand grinders and diamond blades to remove the old coating and prep the concrete for the new one. We built proper containment, sealed it with plastic, and ran ventilation to the outside so we were actually cleaning the air as it moved through the space instead of just pushing dust around.
The part that mattered most wasn't the technical execution. It was the communication. I took photos every day. I wrote daily reports. I sent progress updates that told plant leadership and the corporate team three things: how we were tracking against the plan, what the risks were to hitting our startup date, and what we were doing about them.
That transparency did a lot of quiet work. Nobody upstairs was wondering what was happening behind the plastic sheeting. When you're asking a facility to trust you with two weeks of downtime, that kind of visibility is not optional. It is the job.
Project Two: Where the Plan Met the Roof
The roof replacement started the same way. Cones out, barriers up, access points defined. We demoed the existing membrane and started pulling off the layer underneath, some kind of wood sheathing.
That's when things got interesting.
As we pulled the old sheathing free, we started backing out eight inch long screws. Long enough that I already suspected this job was going to outlast our two week window. We didn't panic about that part. The roof was external to the building, sitting over a warehouse section rather than an open product area, so the schedule risk felt manageable. Low consequence, we figured, if it ran long.
What I hadn't accounted for was what was happening underneath those screws, not just around them.
I climbed up onto one of the lines to get a better look at what was going on below the roof deck. That's when I noticed it. Those eight inch screws weren't just anchoring the sheathing. Some of them had pierced electrical conduit from the original installation. And as we backed them out, the vibration was shaking loose dust and debris, sending it falling into spaces we hadn't planned for at all.
Most of that fallout landed over the warehouse, which was a problem but not the worst version of the problem. The worst version was directly below a can de-palletizer, a piece of equipment feeding open, unlidded cans toward the filling line. Cans that would become finished product. Cans with nothing between them and whatever was now raining down from the roof deck above.
That is the moment a routine infrastructure project turns into a food safety event waiting to happen.
The Scramble, and What It Cost Us
We didn't have the luxury of redesigning the sequence. The facility needed to come back online. So we improvised an emergency mitigation on the spot: draped plastic sheeting above the de-palletizer and built a false ceiling to catch anything falling through before it reached the open cans.
It worked. But it was scope nobody had planned for, priced for, or scheduled for. It drove a change order, and it drove another one right behind it. Because once we found screws piercing conduit, we couldn't just leave it. Some of that conduit turned out to be abandoned, dead and safe to demo. Some of it was still live, still active, and had to be properly replaced.
Two small discoveries. Two change orders. Neither one catastrophic on its own. But both entirely avoidable if the front end investigation had gone one layer deeper before the bid package went out.
Where the Real Lesson Lives
Here's what I keep coming back to. Both of these projects were "simple" upgrades to existing assets. No new equipment, no unfamiliar systems, nothing that should have surprised anyone with real field experience. And yet the surprises still found us, because the thing we were replacing was not the thing that mattered. What mattered was everything attached to it, buried beneath it, or operating underneath it while we worked.
I climbed up on that roof. I did look. And I still missed the conduit until the screws started coming out. That's not a confession of incompetence, it's an honest look at how easy it is to under scope the investigation phase when a project looks routine and the clock is already running.
So how do we, as the people scoping these jobs on the front end, actually get better at this? I've started thinking about it as three questions that should get answered before a single line item goes into an RFP.
A Framework for Front-End Risk Discovery
Look Up. What is above this work, and what happens if something falls, leaks, or drops through during execution? On the roofing project, that meant asking what sat directly below every square foot of deck we were about to disturb, not just what sat below the building in general.
Look Down. What is buried in, attached to, or running through the asset you're removing? Conduit, piping, abandoned infrastructure from a decade of prior work, it's rarely visible until you're mid-demo. Assume it's there and verify, rather than assuming it isn't and being surprised.
Look Around. What is still operating adjacent to this work while you're doing it? A can de-palletizer running beneath a roof deck is a food safety risk. The same work above an empty storage bay is a Tuesday. Context changes the entire risk profile of identical work.
None of these questions require new tools or new budget. They require time, on the front end, before the RFP goes out and the schedule gets locked. That's the hard part. Time is exactly what feels most expendable when you've got a two week shutdown window and a plant waiting to restart.
But the change orders, the emergency mitigations, the scrambling to protect open product with plastic sheeting and plywood, all of that costs more time than the investigation would have. It just costs it later, when you have less room to absorb it.
If I'd walked that roof with those three questions in mind before we ever started pulling screws, I'd have caught the conduit risk in the estimate instead of in the field. I'd have planned for the de-palletizer exposure instead of improvising around it. The work still would have taken the same two projects, the same two weeks of pressure. But it would have looked a lot less like discovery, and a lot more like execution.
That's the standard I hold myself to now. Look up, look down, look around, before the bid goes out, not after the screws start coming loose.
