MEP Coordination for Constructibility, Not Just for Clash Counts

Executive summary

A clash report with zero conflicts can still describe an unbuildable ceiling. Geometric clash detection proves two objects don't occupy the same space; it says nothing about whether a duct route can actually be installed, maintained, or prefabricated. The shift underway is from counting clashes to qualifying them: prioritizing by installation sequence, access for maintenance, and fit for offsite fabrication. Bad project data was estimated to cost the global construction industry $1.85 trillion in 2020, and a joint PlanGrid/FMI study attributes 26% of rework to poor communication and another 22% to inaccurate or incomplete project information [1]. Constructibility-driven coordination targets exactly that gap — not the geometry, but what the geometry doesn't say.

The real problem and what's changing

A model can be "clash-free" and still fail on site: a valve with no clearance to operate it, a duct route that blocks a required access panel, a rack that fits in the ceiling but not through the door it needs to pass through during installation. None of these show up as a geometric intersection. They show up as a change order.

Constructibility, as defined by the Construction Industry Institute (CII), is "the optimum integration of construction knowledge and experience into project planning, design, procurement, and field operations to achieve overall project objectives" [2]. CII's own benchmarking associates disciplined constructability programs with a 6.1% improvement in cost performance and 7.1% in schedule performance [2] — evidence that the practice isn't a formality, but a driver you can measure.

Applied to MEP coordination, constructibility means the coordination model has to answer questions clash detection alone doesn't ask: Can this be installed in sequence without removing something already placed? Can it be reached for maintenance without demolishing a wall? Can it be broken into shippable, liftable, weldable pieces that match how the trade actually fabricates and installs?

Why now

Two pressures are converging. First, the data problem: most rework isn't caused by a design error nobody caught, it's caused by information that was inconsistent, late, or incomplete when the decision was made [1]. Second, the shift to offsite fabrication changes what "coordinated" needs to mean. BIM usage for prefabrication was already expected to grow from 44% to 75% of contractors within three years, according to Dodge Data & Analytics' SmartMarket Report on Prefabrication and Modular Construction [3] — and once a model feeds spool sheets and fabrication shops directly, an uncoordinated model doesn't just cause an RFI, it produces a physically wrong part.

Construction productivity overall grew only about 10% between 2000 and 2022, compared with roughly 90% in manufacturing [4]. Constructibility-driven coordination is one of the few levers that acts directly on that gap, because it front-loads decisions that would otherwise be made — expensively — in the field.

How it works and where it applies

Traditional clash detection ranks conflicts by count or by discipline pairing. A constructibility-driven workflow ranks them by consequence:

• Installation sequence: does resolving this clash require re-sequencing trades, or does the current routing block something that has to go in first?
• Access and maintenance: does the routing leave clearance for the equipment's required service envelope, not just its physical footprint?
• Fabrication and transport constraints: does the assembly fit standard module sizes, shipping dimensions, and lift capacity, or does it force field welding that wasn't planned for?
• Tolerance stacking: do accumulated tolerances across trades still leave a workable margin at the point of final connection?

Concrete applications:

• Prioritized clash resolution: routing conflicts through a severity matrix based on cost, schedule, and installation impact instead of resolving in the order they were found.
• Fabrication-ready modeling at LOD 400: BIMForum's Level of Development framework defines LOD 400 as model detail sufficient for shop drawings and fabrication, not just design intent [5] — the level at which a model can actually drive a spool sheet or panel cut list.
• Prefabricated MEP assemblies: racks, headwalls, and skid-mounted equipment validated for constructibility before fabrication, since a rack that doesn't fit through a stairwell is a field problem no amount of clash detection would have caught without a constructibility check.
• Access-panel and maintenance-zone modeling: treating equipment service clearances as real, dimensioned model elements instead of a note on a drawing.

Business and team impact

The direct financial case is rework avoidance: information-related rework alone accounts for roughly half of documented rework causes in the PlanGrid/FMI data [1], and CII's benchmarking puts the schedule and cost benefit of formal constructability programs in the single digits per project — modest per project, compounding across a firm's full pipeline [2].

The organizational shift is in who owns the coordination decision. Geometric clash resolution can be delegated to whoever's model touches whoever else's. Constructibility review requires someone who has actually installed MEP systems in the field — a role that blends BIM coordination with trade experience, and one that many firms staff too late, after the model is already "coordinated" in the geometric sense.

Barriers, risks, and maturity level

The main barrier isn't technical, it's sequencing: constructibility review works best early, when design changes are cheap, but the people with the field knowledge to do it are often the same people who are busiest once construction starts. Bringing constructability input into design development — rather than treating it as a preconstruction checkpoint — requires a deliberate staffing decision most firms haven't made.

A second risk is over-indexing on the model. A clash-free, constructibility-reviewed model is still an intent, not a guarantee — field conditions, substitutions, and sequencing changes during construction can reintroduce exactly the problems the review was meant to prevent. Constructibility review reduces risk; it doesn't eliminate the need for field verification.

Maturity is uneven: geometric clash detection is universal practice; constructibility-driven prioritization is common on large, prefabrication-heavy projects and still inconsistent on smaller ones, where it's often left to individual coordinators' judgment rather than a documented process.

How to prepare

1. Define a severity matrix before coordination starts, not during it — installation sequence, access, and fabrication fit as explicit ranking criteria, not implicit judgment calls.
2. Bring trade and fabrication knowledge into design development, not just preconstruction review, so constructibility issues surface while changes are still cheap.
3. Model equipment service clearances as real geometry, not annotation, so access conflicts show up in clash detection instead of on site.
4. Match model detail to how the work is actually fabricated, using LOD 400 deliberately for assemblies headed to a shop, not uniformly across the whole model [5].
5. Track rework causes, not just clash counts, so the firm can tell whether coordination effort is actually reducing field problems or just producing cleaner-looking reports.

Future outlook and conclusion

The direction is toward coordination models that double as fabrication inputs — where a resolved clash isn't the end state, but a checkpoint on the way to a spool sheet or a panel cut list. As offsite and robotic panelized manufacturing capacity grows [6], the cost of an uncoordinated model shifts from a field change order to a rejected fabrication run, which raises the stakes for getting constructibility right before the model leaves design.

The strategic point is that clash detection and constructibility review answer different questions, and only one of them predicts what actually happens on site. A model with zero clashes tells you the geometry is consistent. A constructibility review tells you whether the building can actually be built the way the model says it will be — and that's the review that prevents the RFI.

Frequently asked questions

Does constructibility review replace clash detection? No. Geometric clash detection stays necessary and deterministic; constructibility review adds a layer of judgment about installation, access, and fabrication that geometry alone can't capture.

Is constructibility review only relevant for prefabricated or modular projects? No, but the stakes are higher there. On a traditionally field-built project, an uncoordinated route causes rework. On a prefabrication-heavy project, it can mean rejected parts and a broken fabrication schedule.

Who should lead a constructibility review? Someone with field installation experience, not only modeling experience. Firms that treat it purely as a BIM coordination task tend to miss the access, sequencing, and tolerance issues that only show up to someone who has installed the system before.

---

Sources and references

[1] Autodesk & FMI. "Harnessing the Data Advantage in Construction." 2021; PlanGrid & FMI, "Construction Disconnected" rework-cause data cited within.

[2] Construction Industry Institute (CII). "Constructability: A Primer" and constructability benchmarking research.

[3] Dodge Data & Analytics / Dodge Construction Network. "Prefabrication and Modular Construction SmartMarket Report," 2020.

[4] Mischke, J.; Stokvis, K.; Vermeltfoort, K.; Biemans, B. "Delivering on construction productivity is no longer optional." McKinsey & Company, August 9, 2024.

[5] BIMForum. "Level of Development (LOD) Specification 2024" — LOD 400 definition for fabrication-level detail.

[6] GlobeNewswire / market research. "Global Robotic Panelized Home Builders Market" report, January 2026.