Shop fabrication and prefabrication: why it fails in the field and what makes it work

Where the savings come from

Understanding the mechanism prevents overpromising. The gains come from specific places:

• Labor productivity. A worker at a bench with materials at hand produces far more than the same worker on a lift in a congested area.
• Schedule compression. Fabrication happens in parallel with site work rather than after it.
• Reduced rework. Controlled conditions and repeatable jigs reduce errors.
• Safety. Less work at height, fewer trades stacked in the same space, less field cutting.
• Material waste. Shop cutting optimizes yield in a way field cutting never does.
• Weather independence. Fabrication continues regardless of site conditions.

What it does not automatically produce is a lower unit cost of material, and contractors who sell prefabrication on material savings usually end up disappointed.

The decision has to come early

This is the point that determines everything else. Prefabrication requires design information to be complete earlier than a traditional sequence demands, because a shop cannot build from a drawing that is still evolving.

Practically, that means the decision belongs in preconstruction, not in the field when the schedule starts slipping. It means coordination and shop drawings have to be finished ahead of the normal curve. It means the owner and designer need to understand that late changes carry a different cost, because a change after fabrication is scrap rather than an adjustment.

Projects that decide to prefabricate mid-stream, as a recovery measure, almost always find that the information is not ready — and then conclude that prefabrication does not work.

Coordination is the actual product

For multi-trade assemblies, the value comes from the coordination effort rather than from the shop itself. When mechanical, electrical, plumbing and fire protection are modeled together and clashes are resolved before fabrication, the field install becomes fast and predictable. When they are not, prefabricated assemblies collide in the ceiling exactly as field-built ones would, except now the conflict is in a rigid module nobody can adjust.

That makes clash detection and model coordination a prerequisite rather than an enhancement. It also means a contractor without that capability should start with single-trade assemblies, where coordination risk is contained, rather than with multi-trade racks.

Selling it to the owner and the design team

Prefabrication requires something from the owner and the designer that a traditional sequence does not: earlier decisions. Selling that well is part of the work, because a project team that agrees in principle and then continues making changes on the old schedule produces the worst version of both approaches.

The argument that lands with owners is rarely about your labor productivity, which is your problem rather than theirs. It is about schedule and disruption. Twelve days off a corridor rough-in is twelve days earlier occupancy, or twelve fewer days of a department relocated. Fewer workers overhead in a live facility is less disruption to operations. Fewer field cuts is less dust, noise and cleanup. These are outcomes an owner feels directly.

With designers, the conversation is about specific decision deadlines rather than general cooperation. A dated list of what must be finalized, and by when, for each prefabricated assembly is far more useful than a request to finish drawings early. It also creates a record, which matters if the information arrives late and the schedule has to be reset.

The honest part of this conversation is the trade-off: reduced flexibility. An owner who wants the ability to change layouts during construction is buying something incompatible with fabrication, and saying so plainly at the start is much better than discovering it in month five.

Dimensional control: the failure mode

The single most common way prefabrication goes wrong is a mismatch between the assumed dimension and the built condition. A structure that is out of tolerance, a slab that is not level, a wall that moved during construction — any of these can turn an accurate assembly into an expensive misfit.

The controls that prevent it: field verification with actual measurement before fabrication rather than reliance on design drawings, laser scanning on projects where it is justified, deliberate tolerance allowances built into connections, and a rule that fabrication only begins after the relevant field conditions are confirmed.

Designing connections with adjustability is the cheapest insurance available here, because it converts a fit problem into a field adjustment instead of a rebuild.

Delivery, access and sequencing

An assembly that cannot get to its final position is worthless, and this is discovered embarrassingly often. The questions that must be answered before fabrication:

• What are the dimensions and weight, and what is required to lift and move it?
• What openings does it pass through, and will those openings still exist at delivery time?
• Where will it be staged, and for how long?
• What has to be complete in the building before it arrives, and what must not be complete yet?
• Is crane or lift time available on the required date?
• Can the truck reach the site, and what are the route restrictions?

The fourth question is the subtle one. Many prefabricated assemblies must be installed before a wall closes or a floor above is poured, which means a delivery delay does not just push that scope — it stops the sequence behind it.

Your own shop or a fabricator

Both models work and the choice depends on volume and repetition rather than preference.

An in-house shop makes sense with consistent volume across projects, work that repeats enough to justify jigs and tooling, and enough backlog to keep the space busy. It offers control and captures the margin, and it also creates a fixed cost that continues when the backlog thins.

Using a specialized fabricator makes sense for occasional work, for assemblies requiring capabilities you do not have, and for testing the approach before committing capital. It costs more per unit and much less in risk.

A common progression that works is subcontracting fabrication for two or three projects, learning where the value actually is, and then building an internal capability around the specific assemblies that proved out.

What it does to your workforce

Prefabrication changes the labor model, and ignoring that is how good programs die. Shop work is different from field work: it is indoors, more repetitive, often steadier, and frequently pays differently. Some workers prefer it strongly and others dislike it.

There are also representation and classification considerations in union environments that vary by agreement and region, and they need to be understood before the first module is built rather than after. Beyond the formal questions, there is a practical one: a shop that becomes a place for workers who cannot perform in the field will produce poor assemblies, while a shop staffed with your most precise people will produce work the field can install quickly.

Quality control in the shop and the field

A shop produces consistency, and consistency amplifies whatever process it is given. A well-controlled shop produces many accurate assemblies; a poorly controlled one produces many identical defects, which is considerably worse than the same error appearing once in the field where somebody would have caught it.

The controls that prevent that are unglamorous. A first-article inspection on the initial unit, checked against the drawing and against actual field conditions before the rest are built, catches the systematic error while it is still one unit. A documented checklist per assembly type, signed by whoever built it, creates accountability and traceability. Photographs of each completed assembly with its identification number give the field a reference and the project a record.

Labeling deserves particular attention because it is where simple programs fall apart. Every assembly should carry a unique identifier tied to its location in the building, and the installation drawings should reference the same identifiers. Without that, a delivery of forty similar-looking racks becomes a puzzle solved by trial and error on a lift.

Finally, the field crew installing the assemblies should be involved in reviewing the design before fabrication begins. They know what will be awkward to lift, where a connection will be unreachable, and what will not clear an existing condition. Twenty minutes of their input before the first unit is built routinely prevents the problem that would otherwise repeat forty times.

The economics on a real scope

Overhead multi-trade corridor racks on a four-floor medical office fit-out:

• Field-built estimate: 1,480 labor hours at 82 dollars burdened = 121,360 dollars
• Prefabricated: 610 shop hours at 68 dollars = 41,480 dollars, plus 340 field hours at 82 dollars = 27,880 dollars
• Coordination and shop drawing effort added: 180 hours at 95 dollars = 17,100 dollars
• Shipping, staging and lifting: 6,400 dollars
• Total prefabricated: 92,860 dollars | Savings: 28,500 dollars
• Schedule impact: corridor rough-in reduced from 19 working days to 7

The coordination line is the one people forget to include, and it is substantial. The schedule line is the one that usually matters more than the money, because twelve days of corridor access returned to other trades has a value that does not appear in this table at all.

Contract and money questions

Prefabrication creates financial questions that traditional contracts handle poorly. Materials are purchased earlier, work is performed off site, and value accumulates somewhere other than the project.

The issues to address explicitly: whether stored materials and off-site work are payable under the payment application terms, what insurance covers assemblies in a shop and in transit, who holds title and when it transfers, what happens to fabricated work if the project is cancelled or suspended, and how a change order is priced after fabrication has begun.

These should be negotiated at contract stage. Raised for the first time in a payment application, they become a dispute at exactly the moment cash flow is tightest.

Starting without betting the company

The realistic path into prefabrication is narrow and repetitive rather than ambitious. Pick one assembly that repeats many times on a single project, verify that the design information will be available, build a jig, measure the hours honestly against a field-built baseline, and document what went wrong.

The first attempt usually yields a modest gain and a long list of process lessons. The second attempt on the same assembly type is where the numbers get interesting, because the coordination effort is reusable and the jig already exists. Contractors who judge the approach on a single first attempt almost always abandon it one project too early.

The mistakes that cost the most

• Deciding to prefabricate mid-project as a schedule recovery move.
• Fabricating from drawings that are not final.
• Skipping field verification of actual dimensions.
• Designing connections with no adjustability.
• Not confirming access, openings and lift availability before building.
• Leaving coordination hours out of the comparison.
• Ignoring stored-material and title questions until the pay application.
• Judging the whole approach on one first attempt.

Where to start

Choose one repetitive assembly on your next project and raise it in preconstruction, not in the field. Confirm that the design information can be finalized early enough, and get the owner and designer to acknowledge what late changes will cost. Field-verify dimensions before anything is built. Include the coordination hours in the comparison so the number is honest. Then measure the result against a field-built baseline and do it again on the next job — the second time is where prefabrication starts paying what it promises.

Frequently asked questions

What kinds of construction work prefabricate well?

Five characteristics predict success. Repetition, meaning the same assembly many times, as in hotel bathrooms, patient rooms, apartment units or identical bays. Complexity concentrated in a small area, like mechanical rooms, overhead multi-trade racks and riser assemblies. Difficult field conditions such as high ceilings, tight access, live environments or extreme weather. Dimensional certainty, meaning the work can be measured or controlled precisely before fabrication. And transportability, so the assembly fits through openings and onto trucks. Work that is genuinely custom, dimensionally uncertain until the building is up, or too large to move gains little and frequently loses money.

Why do prefabrication attempts fail?

Usually for one of three reasons, and none of them is that fabrication itself is difficult. The first is timing: the decision was made mid-project as a schedule recovery move, when the design information was not final and could not support fabrication. The second is dimensional control, where the assumed dimension did not match the built condition because field verification was skipped or connections were designed with no adjustability. The third is logistics, where the assembly could not reach its final position because access, openings, lift availability or delivery sequencing were not confirmed before building. All three are preventable, and all three are decided long before anyone picks up a tool.

Does prefabrication actually save money?

It can, and the honest comparison has to include the coordination effort that people leave out. On a four-floor medical office fit-out, field-built overhead corridor racks estimated at about fourteen hundred eighty labor hours came to roughly one hundred twenty-one thousand dollars, while the prefabricated version totaled about ninety-three thousand including shop hours, field install hours, an added one hundred eighty hours of coordination and shop drawing effort, and shipping, staging and lifting. That is roughly twenty-eight thousand dollars saved. The schedule impact mattered more: corridor rough-in dropped from nineteen working days to seven, returning twelve days of corridor access to every other trade.