The hidden cost drivers and overlooked details of structural steel fabrication

FIGURE 1. This complex, cover-plated, wide-flange column has extensive bracing, designed for blast resistance. All this work adds significant cost, both for labor in the shop and in the field. Sometimes such complexity just isn’t avoidable, but the cost of that complexity needs to be factored into the job.

Tim Bradshaw is all about giving shop tours, especially to structural engineers. As vice president of project delivery at Owen Steel Co. , Columbia, S.C., Bradshaw has been designing and fabricating structural steel his entire career. For more than 25 years, he has worked as a professional engineer and manager in the industry. He knows when metal fits together well—and when it doesn’t. Floating Wall Shelves

The hidden cost drivers and overlooked details of structural steel fabrication

“Three factors influence cost in structural steel. First is the cost of material, second is the shop labor cost, and third is the erection cost. A long-held rule of thumb has held that each comprises about a third of overall cost. That is, a job’s cost is one-third material, one-third shop labor, and one-third erection.”

Bradshaw made this statement in March 2022 during a presentation he gave in Denver at NASCC: The Steel Conference, organized by the American Institute of Steel Construction (AISC). He added that the rule fluctuates over time with changing material and labor costs. Sometimes, however, the rule swings out of balance because of certain design choices the engineer of record (EOR) made or factors the EOR simply overlooked.

Of all the sectors in metal fabrication, structural steel fabrication stands apart. The construction supply chain is less of a chain and more of a complex web of interconnected parties. A structural steel fabricator can be just a cog in the wheel, fabricating what it’s given, or it can be an active communicator, showing the EOR (and anyone else) what truly drives costs and efficiencies on the fab shop floor (see Figures 1 and 2). It’s why Bradshaw, who serves on three AISC committees, always welcomes the chance to give a shop tour. In fact, his NASCC presentation could be seen as a virtual shop tour of sorts, during which he pointed out various factors that throw the costing rule of thumb off-kilter.

Bradshaw pointed to a simple beam with straightforward connections. “In this case, less weight really does equal less cost. It’s a simple beam, there’s not a lot of shop labor, and it’s not expensive to install in the field.”

He then pointed to a spandrel beam (edge beam constructed along the exterior wall of each floor) with a long, cantilevered edge. Its lightweight construction looks great in isolation, but that light weight also requires steel reinforcement. “You just quadrupled the weight of the beam because of all the reinforcing steel you need to add to that beam to handle the cantilevered edge. It might have been better to make the beam a little heavier or use a tubular section for your spandrel, or to not have that long cantilevered edge.”

He then pointed to some shop drawings involving stiffeners and doublers. Depending on the design, lighter columns might require more robust stiffeners and doublers, and welding them isn’t free. In his presentation, he pointed to a lightweight W14 × 159 column with a 6-in.doubler and a W14 × 342 with a ½-in. doubler. From a pure material cost perspective, W14 × 159 seems the way to go, until you look at all the welding those doublers require. The 6-in. doubler plates on W14 × 159 increased overall weld volume by 72 times. Much of this has to do with geometry; a full-penetration weld is essentially a triangle, and its volume increases exponentially with its size. With increased weld volume comes multiple weld passes and more costs.

“In this case, it might be less expensive to go with a W14 × 342 in place of that W14 × 159 column,” he said, “especially if you’re making moment connections at multiple levels coming in from multiple directions.”

Detailing, connection engineering, fabrication—sometimes all three are done under one roof; other times they’re done by different firms. Regardless, open communication is key (see Figure 3).

FIGURE 2.A lightweight beam might require web reinforcements like this—another factor that affects weight versus cost considerations.

Some fabricators, especially those with plenty of punching and drilling equipment, are set up to do bolted work very efficiently; others weld efficiently; some excel at both. Of course, engineers often develop drawings without knowing which fabricator will perform the work. “If you don’t know whether the fabricator will be set up to do welded work, bolted work, or both, try to allow for either bolted or welded connections.”

Single-sided connections work best for the erector. “You of course can’t use a single-sided connection all the time, but use as many as you can,” he said. “There’s a lot of research in the latest Steel Construction Manual [published by AISC] about not only standard shear tabs but also extended shear tabs. [Using these] can go a long way, especially if you’re connecting to the web of a column, or if you’re coping into a web of a girder that has a really wide flange. If you can use an extended shear tab connection, you don’t have to worry about coping the beam as much.”

He added that when it comes to design loads, specifics matter. When given a specific load requirement—this W12 has a minimum connection load of 20 kips; for that W14, it’s 24 kips—connection engineers check every connection against that load.

Sometimes, though, connection engineers are just given a tabulated list that shows the minimum number of a specific type and size of bolt. “That doesn’t tell us much from a connection standpoint,” he said. “We can give you a W16 with four bolts in it, but what does that mean for block shear? What does that mean for bolt bearing? What does that mean for bending on the net section? We really need specific load information in some form.”

“When you consider weld sizes, consider the type of weld,” Bradshaw said. “We would ask you to size the weld based on actual load demand or to meet code requirements. It’s easy to go on a set of plans that say ‘full-penetration weld.’ But is that really needed? If it is, great. We’re happy to do it. But it’s not always necessary.”

Full-penetration welds require weld access holes, at least in most cases. “Think about what those will look like, and if your architect will even allow them to remain open. Also, never fill a weld access hole with weld metal after the joint is complete. If you have a weld access hole and you need to fill it, try body filler putty. If you outsource painting, let the painter do it.”

Weld position matters too. A fabricator can position a weld for easy access in the shop, but in the field not so much. As Bradshaw explained, “Based on the drawings, are you using a downhand weld technique? A vertical weld? Or is it overhead, which is the most difficult to make?”

Again, weld volume matters, and fillet welds especially can take up a lot of volume (see Figure 4). “In some cases, a fabricator may see the drawings that ask for a very large fillet and ask if the weld can be a partial- or even full-penetration weld. Two very large fillet welds on either side of the plate may actually have more weld volume than a full-penetration weld.”

Also consider the size of weld a shop’s welders can perform in one pass with wire welding. These days, some shops (depending on the equipment and wire diameter) can perform a single-pass weld up to a 5/16-in. weld size. Other fabricators lay down ¼ in. at a time. “Regardless, anything over 5/16 in. is going to be a two-pass weld,” Bradshaw said, adding that the number of passes goes up exponentially with weld size. “A 5/8-in. weld becomes six passes, a ¾-in. weld becomes 10 passes, a 7/8-in. weld becomes 15 passes, and a 1-in. fillet weld becomes 21 passes. That’s why we’d rather do a full- or partial-penetration weld in some cases.”

He added that robotic welding cells can lay down very large welds, but they’re still limited by the wire size. And yes, submerged arc welding (SAW) can weld very large joints in a single pass, but not every weld can be designed for SAW’s mechanized setup.

FIGURE 4. Multipass fillet welds like these can require a lot of weld metal, which adds to project costs.

The building façade contractors often are the last to get involved in a project. That can be a challenge, as Bradshaw explained, since early coordination is essential to decide how the façade will connect to the edge beams and slab.

Complicated sequencing of the work can also lead to added costs. For instance, Bradshaw described one project that required façade angle bracing to be expansion anchored to the underside of the slab—bracing that couldn’t be installed until after the slab was poured. “So in this case, the steel was erected, the slab was poured, and someone had to go back and put in the angles.”

In some cases, coordinating façade connections from the get-go—perhaps by moving the beam or altering the slab position a fraction of an inch—can lead to a simple, elegant solution developed in the design phase. That solution will almost surely be less costly than having to scramble fabrication and erection schedules during the building phase.

Bradshaw pointed to a drawing with a moment connection where a W40 beam was supported by a W30 beam. “In this case, we had to weld a WT on the bottom of the W30 to make it the same depth as the W40, to put the flange plates in place.

“That’s an extreme example,” Bradshaw continued, “but think about framing. We’ve seen a W24 framing into a W12. What does that coped section look like? You have to cope away half the W24 to frame it into a W12. Many times, you need to cope that W24 so much that you now have to reinforce it or put some kind of extended plate on the bottom. The bottom line: Try not to frame a deeper section into a substantially shallower section. It’s easy to catch.”

In areas where multiple moment connections come into a column, like in a moment frame, “it’s always good to use the same nominal depth for the moment beams, so you don’t end up with multiple layers of stiffeners,” Bradshaw said. “If you can keep your moment beams the same nominal depth on either side of the column, that reduces the amount of labor that goes into installing multiple layers of stiffeners.”

Fillet welding of stiffeners usually is straightforward and cost effective, but it’s not always possible if stiffeners are too close together. Sometimes welders end up performing full-penetration welds on opposite sides. “It gets complicated,” Bradshaw said, “and it compounds the hours needed to fabricate a piece.”

Structural fabrication isn’t just about how many tons of steel a shop can produce in a given time. Labor plays a huge role. “We like to measure the cost in terms of how many man hours per ton,” Bradshaw said.

Automation can decrease those man hours per ton, of course, but so can clarity and constant communication. Do documents specify the exact version of the code? Specifying “latest version” sometimes can cause confusion, since different codes from different organizations have varying release schedules.

Are project-specific tolerance requirements specified? While common fabrication tolerances are specified in ANSI/AISC 303-16, Code of Standard Practice for Steel Buildings and Bridges, additional tolerances may need to be specified in the design documents to accommodate the work of other trades, like the façade installation or other architectural finishes.

Wherever structural fabricators gather—at NASCC, FABTECH, or anywhere else—many sessions cover technological advancements, but a good portion usually focus on communication and breaking down the silos among all the parties involved in a construction project. That, Bradshaw said, is why he always welcomes EORs, detailers, erectors, connection engineers (which Owen Steel employs), mechanical contractors, façade contractors, and anyone else associated with building construction to come pay a visit. Efficient structural fabrication requires the spreading of knowledge, and for more than two decades, Bradshaw and his colleagues—both within Owen Steel and fellow members of AISC—are working to do just that.

FIGURE 3. Owen Steel’s connection engineers designed this bracing end connection in-house. Whether connection engineering is done in house or outsourced, open communication is key.

See More by Tim Heston

Tim Heston, The FABRICATOR's senior editor, has covered the metal fabrication industry since 1998, starting his career at the American Welding Society's Welding Journal . Since then he has covered the full range of metal fabrication processes, from stamping, bending, and cutting to grinding and polishing. He joined The FABRICATOR's staff in October 2007.