MEP Coordination Process: A Step-by-Step Guide

MEP coordination is where most construction conflicts hide — and where the cost of catching them late is highest.

Mechanical, electrical, and plumbing scope cross every floor plate and touch every other discipline. When those trades coordinate well, the project runs predictably. When they don’t, the failures land as change orders, RFIs, and rework — often traced back to a conflict that was sitting in the drawings the whole time.

This guide walks through the MEP coordination process step by step: what each phase covers, where the seams break down, how to classify the conflicts you find, and where document-level review fits into the sequence — including before a BIM model ever exists.

What MEP coordination actually covers (and why it fails)

MEP coordination is the process of reconciling mechanical, electrical, and plumbing scope with each other — and with architecture and structure — so that the trades can build without running into each other. In practice, it means aligning duct routing, pipe runs, conduit layouts, cable trays, structural framing, ceiling heights, and fire-protection systems across a set of drawings that were produced by separate design teams on separate schedules.

The failure mode is structural to how projects are designed. Each discipline produces its own set of drawings, solving its own scope. The MEP engineer routes ductwork with the floor-to-floor clearance in mind. The structural engineer sizes the beams and frames the shear walls. The architect carries the ceiling heights and the room programs. Nobody draws on the same sheet. The conflicts are in the seams — and the seams are where the documents don’t overlap.

The costliest seams are:

• Structure vs. MEP: A shear wall, beam, or footing lands exactly where a duct main, pipe, or conduit run needs to go. Neither drawing is wrong in isolation; they just can’t coexist in the field.
• Spec vs. drawing: The mechanical spec calls one product; the drawing details another. The conflict is invisible to either discipline’s reviewer working alone.
• Cross-sheet conflicts: An elevation or dimension on the architectural sheets contradicts what the mechanical drawings assume for ceiling height or plenum depth.
• Missing-coordination gaps: A discipline simply doesn’t appear where it should. Fire protection isn’t extended into a new wing. Electrical hasn’t coordinated with the revised wall layout.

These aren’t exotic edge cases. On one real project, more than 40 change orders traced back to conflicts entre 2D plan sheets — conflicts that existed in the documents long before ground broke. A single structural-MEP conflict — a shear wall colliding with HVAC duct routing — carried $28,000–$45,000 in rework and 14–21 days of schedule. And errors discovered during construction cost roughly diez veces más para arreglar than the same errors caught at plan review.

Step by step: the MEP coordination workflow

Step 1 — Trade input and constructability review

Each MEP trade submits its design-intent scope: the mechanical engineer’s drawings, the electrical engineer’s drawings, the plumbing engineer’s drawings, and the fire-protection design (if separate). At this stage, the inputs are typically discipline-specific and haven’t been overlaid. A revisión de constructibilidad at this point checks whether each discipline’s scope is buildable on its own terms — adequate clearances, code-compliant routing, sequencing logic — before the overlay begins.

What to catch here: incomplete scope (a discipline that hasn’t resolved its routing in a critical area), missing details, or spec documents that don’t yet align with the drawing set.

Step 2 — Document-level overlay and conflict identification

Before a 3D model exists — and sometimes instead of one — the coordination process starts with a disciplined overlay of the 2D drawing sets. This is where most coordination conflicts are actually identified on projects that don’t have fully federated BIM: reading the architectural, structural, and MEP sheets against each other, sheet by sheet and spec section by spec section.

This is the step that manual coordination most consistently shortchanges. Reading hundreds of sheets against each other, without missing the seam where two disciplines intersect, is exactly the kind of work that fatigue and deadline pressure defeat. Plan de revisión de construcción con IA runs this overlay systematically — catching the cross-discipline conflicts, spec mismatches, and missing-coordination gaps that distributed human review misses. On projects without BIM coordination, this is the only comprehensive conflict check that happens before bid.

What to catch here: structural-MEP clashes, spec-vs-drawing mismatches, cross-sheet dimension conflicts, missing fire protection or electrical coverage, ceiling height conflicts.

Step 3 — Conflict classification

Not every conflict has the same urgency or resolution path. A workable classification keeps the coordination log actionable:

Hard conflicts are physical impossibilities. Two objects cannot occupy the same space: the duct main cannot pass through the shear wall as drawn. Resolution requires a design change — rerouting, resizing, or revising the structural element — before the job can proceed. These are the $28,000–$45,000 items.

Soft conflicts are clearance problems. Two elements can technically coexist but don’t meet minimum maintenance access, code clearance, or installation-sequence requirements. A pipe run that leaves inadequate maintenance access above a ceiling tile is buildable as drawn but will cause problems for the life of the building.

Coordination conflicts are sequencing or priority questions. Duct, conduit, and pipe all want to run in the same interstitial space. There’s room for all three if someone decides who runs high and who runs low — but that decision has to be made explicitly, not left to the field.

Document conflicts are disagreements within the written record — spec vs. drawing, sheet vs. sheet, note vs. detail. These don’t create a physical problem until someone in the field has to decide which document controls, and that decision has real cost (an RFI, a stoppage, a potential change order). See cuánto cuesta una RFI for what that decision-by-field looks like in dollars.

Step 4 — Resolution and design coordination

Identified conflicts go back to the responsible design disciplines for resolution. The process here is typically: conflict log → RFI to design team → revised drawing or spec clarification → updated set issued. For hard conflicts, this often requires a coordination meeting where structure, MEP, and architecture are all at the table. For document conflicts, a written clarification or spec addendum usually suffices.

The coordination log is the record of what was found, who owns it, and what the resolution is. Without it, resolved conflicts can reopen — a revised sheet that doesn’t carry the coordination note, a sub who works from an old drawing, a late-issued addendum that doesn’t reach the field.

Step 5 — Sign-off and issue-for-construction confirmation

Before the set goes issue-for-construction, the coordination log should be closed: every identified conflict has a resolution, a revised drawing or written clarification, and a responsible party. This is the pre-construction QA gate — and it’s where the pre-construction QA checklist is most useful as a structured final check.

The sign-off step isn’t just administrative. It’s the moment when the team confirms that what’s on the drawings is actually buildable. Without it, coordination work done in Steps 2–4 can leak back into the field as unresolved conflicts because the updated scope was never reconciled across all discipline packages.

Hard, soft, and 4D clashes: the full taxonomy

The classification above (hard/soft/coordination/document) is a practitioner’s working taxonomy. The formal BIM taxonomy adds a fourth term: 4D clashes, which are sequencing conflicts — elements that don’t interfere in the finished building but do interfere during construction (a structural pour that blocks MEP access for a required installation window, for example). 4D clash detection is a BIM-native capability and requires a scheduled 3D model to run.

For the 2D-PDF coordination process, the relevant distinction is simpler: design conflicts (what the documents show can’t be built as drawn) vs. coordination conflicts (what the documents show is technically buildable but hasn’t been sequenced or prioritized). Both categories show up in the document set. Design conflicts require redesign; coordination conflicts require explicit decisions.

Where 2D AI review fits before and without BIM

The full MEP coordination process described above assumes that someone is doing Step 2 — the overlay — comprehensively. In practice, that step is often partial: one reviewer working through a critical path of known-risky areas, a coordination meeting that covers the major MEP rooms but not every floor plate, a visual pass that catches the obvious spatial clashes but misses the spec discrepancies and missing-coordination gaps.

Clash detection without BIM is not only possible — for most projects at design review, it’s the only form of coordination review that happens before bid. AI plan review runs the full overlay on the 2D PDF set you already have, at design review speed, before a 3D model exists or is needed. It catches the hard structural-MEP clashes, the spec-vs-drawing discrepancies, and the missing-coordination items that manual review at pace misses.

Where does BIM fit? If you have a coordination-ready federated model, run BIM clash detection for the hard spatial geometry — that’s what it’s built for. But BIM doesn’t read your specs. It doesn’t catch the cross-sheet dimension conflicts or the fire-alarm coverage omission. And it arrives late, after a model has been built and federated. The two tools catch different things at different moments in the same process.

Multifamily-specific pitfalls

Multifamily MEP coordination has a particular failure pattern: high repetition creates overconfidence. A 10-story residential tower runs the same floor plan 10 times, so the coordination logic on floors 3–9 feels like it follows from floors 1–2. But unit stacks shift. Mechanical rooms are on different floors than originally assumed. A change to the structural layout on level 4 doesn’t automatically propagate to the MEP drawing for level 4.

The MEP coordination failures in multifamily post covers this in detail, but the short version is: repetition masks drift. The conflicts that surface on a multifamily project are frequently in the difference between a typical floor and a non-typical floor — and they’re caught either by a systematic overlay or in the field as a change order.

Flikt’s drawing-quality benchmark — conflictos por cada 100 hojas — is particularly useful on multifamily because it reveals whether the conflict rate climbs on the non-typical floors and transition levels. That signal is invisible to a coordination process that samples rather than reads the full set. See the conflictos por cada 100 hojas post for how the metric is constructed and what “good” looks like.

El resultado final

MEP coordination doesn’t fail because the design teams aren’t competent. It fails because the seams between disciplines are distributed across hundreds of sheets that no individual reviewer reads comprehensively against each other. The shear-wall-vs.-duct conflict sits in the drawings from the moment both drawings are issued. The question is whether someone finds it before the field does.

The step-by-step process above creates the structure for finding it: systematic trade input, disciplined overlay, classified conflict log, tracked resolution, and a signed-off issue-for-construction set. The gap, on most projects, is the overlay step — and that’s where document-level AI review closes it, without requiring a model, without requiring BIM staff, and at the moment in the schedule when catching a conflict is still cheap.

If you coordinate MEP on your projects, Flikt reads your 2D drawing sets for cross-discipline conflicts before construction — no BIM required. Ver la evidencia o contact us to walk through a real set.