Steel Structures Q&A: What I've Learned Reviewing 200+ Projects a Year

The questions I actually get asked

I review steel structure specifications and completed projects before they go out the door—roughly 200 unique builds a year across sports venues, warehouses, truss bridges, and multi-storey projects. Most of the questions I field from contractors and project owners fall into the same seven buckets. So here they are, answered straight, based on what the inspections and the reworks taught me.

1. What types of steel structures are we actually choosing between?

You're usually picking from five families: hot-rolled sections (I-beams, H-columns), cold-formed light gauge, tubular sections (CHS, SHS, RHS), cable and tension structures, and fabricated plate girders. Each has a different cost-per-span curve.

For a warehouse, cold-formed or hot-rolled portal frames usually win on cost. For a sports venue with a 60m clear span and a roof that needs to look like something, you're into trusses or plate girders—and the budget jumps. For a multi-storey steel structure, you're almost always in composite construction with hot-rolled beams and a concrete deck.

The mistake I see most: someone specifies tubular sections for a warehouse because they look clean, then discovers the connections cost 30% more to fabricate and inspect. Tubular is great—but it's great for the projects where it's actually necessary.

2. When does a multi-purpose steel building make sense over a single-use one?

A multi-purpose steel building makes sense when the occupancy changes faster than the structure does. I've reviewed builds for a municipal client that runs a sports venue nine months a year and a trade exhibition hall the other three. The steel frame didn't change—the floor plates and the egress configuration did.

Here's the thing, though: a multi-purpose building usually costs 10–15% more upfront than a single-use one. That premium buys you clear spans, movable partitions, and services designed for reconfiguration. If your occupancy plan is genuinely stable for the next decade, I'd question whether you're paying for flexibility you'll never use.

The most frustrating part of my job: watching a project pay the multi-purpose premium and then lock the floor plan in concrete and drywall. You'd think the design review would catch it, but it doesn't always.

3. What should I actually check in a steel warehouse spec?

Five things, in this order: clear height, column spacing, floor slab load capacity, crane provisions (even if you don't need one yet), and the roof drainage design.

Column spacing is where budgets quietly die. A 24m bay sounds efficient until you price the frames. A 12m bay is cheaper per square metre but eats into your usable racking layout. I've seen clients save $80,000 on structure and lose $200,000 in racking density. That's not a hypothetical—that was a 2023 project I signed off on, and I still think about it.

On crane provisions: if there's any chance you'll add a crane in ten years, spec the columns and the runway beam connections now. Retrofitting a crane into a warehouse that wasn't designed for one is the kind of project that makes contractors swear.

4. What's different about steel structures for sports venues?

Three things: span, dynamic loading, and sightlines.

Sports venues push you into long-span roof systems—trusses, space frames, cable-stayed roofs. The structural design is one problem. The other is that a crowd of 10,000 people generates vibration and dynamic loading that a warehouse never sees. Deflection limits are tighter, and the analysis is more involved.

Sightlines are the part people forget. The roof structure determines where you can put the primary steel, and that determines where the column lines fall, which determines whether the corner seats can actually see the pitch. I've reviewed a venue design where the roof geometry was structurally elegant and functionally wrong—the back four rows had a column blocking the view. The frame was fine. The building wasn't.

Honestly, I'm not sure why sightline coordination falls through the cracks so often in early design. My best guess is that the structural and the architectural teams are optimizing different things, and nobody owns the intersection until it's expensive to change.

5. Where do steel truss bridges fit, and what's the usual mistake?

Steel truss bridges are the right call for spans roughly between 30m and 120m where you want a shallow structure, fast erection, or a cost-effective alternative to concrete. They're common for pedestrian and light vehicle crossings, and they're increasingly specified in industrial and park settings where the aesthetic works.

The usual mistake is under-specifying the connections and the corrosion system. Truss bridges look simple in elevation. In detail, you've got hundreds of gusset plates, bolted or welded joints, and a maintenance regime that has to be realistic. I've seen a bridge spec that called for a 'high-performance coating system' without defining it. Three years later, the maintenance team was repainting at twice the expected frequency. Coatings are not the place to leave a spec vague.

6. Can I get design, fabrication, and erection from one supplier?

Often, yes—design-build is common in steel structures, and it can save you coordination pain. But I'd push back gently on the idea that one supplier should do everything.

A supplier who's genuinely strong in design, fabrication, and erection is rare. Most are strong in two and passable in the third. The vendor who says 'erection isn't our strength—we'll work with a specialist erector and coordinate' earns more of my trust than the one who claims to do it all in-house at a competitive price. That's not a knock on generalists—it's an observation from reviewing enough projects to see where the quality drop-off happens.

I'd rather work with a specialist who knows their limits than a generalist who overpromises. And the good generalists will tell you which parts they subcontract and why.

7. What's the one thing that most often gets rejected in a steel structure inspection?

Dimensional tolerance on fabricated members. It's not glamorous, but it's consistent. In our Q1 2024 audit, about 12% of first deliveries had at least one member outside the specified tolerance on straightness or hole position. Normal tolerance for fabricated steel members under our spec is L/1000 or 5mm, whichever is greater—and the misses are usually in the 6–10mm range.

The frustrating part is that it's usually not a capability problem. It's a QA culture problem. The fabricator can hit tolerance; they just don't always check. Now every contract we touch includes a first-article inspection requirement and a sampling protocol, and the defect rate dropped to under 4% within two quarters.

This worked for us, but our situation was a mid-size operation with a steady flow of projects. If you're building two structures a year, the calculus of negotiating a first-article inspection clause might be different—you have less leverage, and the overhead may not be worth it. I can only speak to the volume we handle.

One boundary worth stating

I'm a quality reviewer, not a structural engineer. I can tell you what fails inspection, what gets rejected, and where specs go vague. I can't tell you what your specific bridge or venue should be designed to—that's an engineer's job, and honestly, you want them on the project. The best project teams I've reviewed know exactly what they're good at and bring in the right specialist for the rest. That's not a weakness. That's how buildings stay standing.