Breton Logo
Castello di Godego, Italy · Global quartz plant support [email protected] · +39 0423 7691
Manufacturing

Prefabricated Steel Roof Trusses: Why the Real Problems Show Up After Delivery

2026-10-16 · Nisha Kapoor

Article visual for Breton process topic

The Confession I Make to Every New Steel Supplier

I manage procurement for a mid-sized commercial contractor—roughly $4M in annual spend across about a dozen vendors. Steel packages, concrete, roofing, MEP. Been doing this since 2019, and I've placed enough orders for prefabricated steel roof trusses and rigid frame buildings to know one thing for certain:

The steel itself is almost never the problem.

I know how that sounds. You ordered a rigid frame structure. It showed up. Dimensions checked out on the packing list. The structural welding certificates were in the folder. Someone in the office signed off, the truck left, and everyone moved on to the next thing. Then three weeks later your site superintendent calls you at 7:20 in the morning and says, "We're stopped. The knee bracing connection doesn't line up with the embed plate. Nobody caught it in review."

And you're standing there in your kitchen holding a coffee, thinking: how did this get past everyone?

That's the moment this article is about.

What You Think the Problem Is vs. What It Actually Is

When a steel package goes sideways, the instinct is to blame the fabricator. Wrong weld, wrong cut, wrong hole placement. And yeah—sometimes that's it. But after tracking roughly 40 steel deliveries across the last three years, I can tell you the pattern is different. The deliveries were fine. What fell apart was the handoff.

Three things specifically:

  1. Nobody agreed on what "standard" meant for the structural welding scope
  2. Knee bracing construction was treated as field-fix-it work instead of engineered work
  3. The purlin types specified on paper didn't match what the erector assumed he was getting

Each of these looks like a small clerical thing. None of them are.

Deep Cause #1: Structural Welding Is a Scope Problem, Not a Quality Problem

Here's what I didn't understand until roughly my third year doing this: when a steel package says "structural welding per AWS D1.1," that's not a spec. That's a reference. The actual scope—which welds, what inspection level, whether visual only or with ultrasonic testing—gets decided somewhere between the fabricator's shop and the erector's field crew. And if nobody writes it down, both sides assume the other one is handling it.

I'm not a certified welding inspector, so I can't tell you whether a specific weld passes or fails. What I can tell you from the procurement side is that the gap between "welded per code" and "welded per what the engineer actually intended" is where most of the expensive rework lives. And it almost always surfaces during erection, not during shop inspection.

According to the American Welding Society (AWS D1.1 Structural Welding Code—Steel), visual inspection is the minimum requirement for most structural welds. But the code also allows project specifications to require additional nondestructive examination (NDE)—and if the spec doesn't call it out explicitly, it often doesn't happen.

So the question isn't "did the fabricator weld to code." The question is: which code, at what inspection level, verified by whom, and documented where? If you can't answer all four, you don't have a welding scope. You have a hope.

Deep Cause #2: Knee Bracing Construction Gets Treated as an Afterthought

Knee bracing—the diagonal members that stiffen the connection between column and rafter in a rigid frame—is one of those things that looks simple on a drawing and gets complicated fast in the field. The brace itself is straightforward. What's not straightforward is the connection geometry, because the column-to-rafter moment connection has tolerances, and the knee brace has to fit inside them.

What I've seen happen—more than once—is that the fabricator ships knee braces loose or "field-fit," the erector installs them after the frame is plumbed, and by then the bolt holes don't align because the frame moved slightly during erection. Now you're field-drilling, which means you're modifying a moment connection in the field, which means someone has to sign off on that modification, which means a structural engineer has to be called, and the schedule stops while everyone argues about who pays for it.

The deeper problem: nobody owns knee bracing construction in the contract language. It falls in the seam between fabrication scope and erection scope. And anything that falls in a seam gets missed.

Deep Cause #3: "Types of Purlins" Is Not a Detail—It's a Design Decision

C purlins, Z purlins, and structural channel—each one behaves differently under load, each one nests differently for shipping, and each one has a different lap condition at the splice. If your spec says "Z purlins" but the fabricator ships C purlins because that's what was on the rack, the building might still stand. But the load path just changed, and the diaphragm bracing might no longer be adequate.

On a project last year, we had a supplier swap C for Z mid-order because of a mill availability issue. They didn't flag it. The erector installed them. Two months later, a roofing consultant doing a warranty inspection flagged the diaphragm detail as inconsistent with the stamped drawings. We had to commission a structural review. That review cost us roughly $6,800, and the schedule slipped 11 days.

All for a purlin substitution that probably wouldn't have failed. But "probably" isn't a word you want in a stamped submittal.

What This Actually Costs You

Let me run the numbers from our own records.

  • Steel that fails a field weld inspection: average rework cost $3,200–$5,500 per connection group, plus 3–5 days of schedule per occurrence
  • Knee bracing field modifications: $1,800–$4,000 per frame line, plus structural engineer review time
  • Purlin substitution discovered post-erection: $6,000–$10,000 in review costs and potential retrofit

And that's just direct cost. The indirect cost is worse. When the site superintendent calls you at 7:20 AM, you're not just solving a technical problem. You're defending your procurement decisions in front of the PM, who's defending them in front of the client.

I had one project where a knee bracing issue put us three weeks behind on a tenant buildout. The tenant had a rent commencement date tied to the schedule. Our company ate $22,000 in liquidated damages. My name was on the purchase order. That's a conversation I don't want to have again.

The Fix Is Embarrassingly Simple (and Almost Nobody Does It)

Looking back, I should have required a pre-shipment alignment call with the fabricator, erector, and structural engineer on the same line—30 minutes, three specific questions:

  1. What's the weld scope by category, and who's verifying it before shipment?
  2. Is knee bracing pre-assembled, loose-shipped, or field-fit—and who owns the connection if it doesn't line up?
  3. Are the purlin types on the shipping list the same as the stamped drawings, and what's the substitution protocol if they're not?

That's it. No new technology. No expensive consultant. Just forcing the conversation before the truck leaves the yard.

The fundamentals of steel construction haven't changed—gravity still works the same way it did 50 years ago. What's changed is the number of hands a single rigid frame structure passes through before it's standing. More hands means more translation loss. A 30-minute call before shipment is a lot cheaper than three weeks of rework after.

I've never fully understood why this isn't standard practice. My best guess is that everyone assumes someone else is already doing it. That assumption costs more than any of us want to admit.

Nisha Kapoor
Nisha Kapoor

Nisha Kapoor is a construction-systems analyst covering formwork, scaffolding, shoring, prefabricated buildings, modular rooms, panelized assemblies, geotextiles, geogrids, and drainage composites. She uses ASTM D4595 and ASTM D6637 for their respective geotextile and geogrid tensile evaluations while tracking platform load, form deflection, prop spacing, module tolerance, lifting points, seam strength, aperture stability, and installation damage. Her technical guides help contractors, engineers, and developers coordinate temporary works, factory tolerances, site logistics, ground reinforcement, inspection hold points, and documented acceptance criteria.