When Buildings Outlive Their Documentation
Why facility owners keep making expensive decisions with outdated building information, and what to do about it
Authored by: Reetie Lubana
Most as-built drawings are wrong. The money is in knowing exactly where.
After more than 1,000 existing-conditions assessments over twenty years, I’ve stopped treating any drawing set made before 2000 as anything more than a working hypothesis. That’s not a critique of the architects or engineers who produced them it’s just the reality of what happens to documentation when buildings keep living and records don’t.
Most buildings carry two versions of reality: the structure itself, and the drawings that claim to describe it. In older facilities, these two rarely match. But here is the part most articles skip: the gap is never uniform. A set of drawings can be wrong in a dozen places, and only two of them will cost you anything. The skill that separates an expensive renovation from a predictable one is not “keep your documents current.” It is knowing which discrepancies quietly drain a budget and which are harmless.
Why the records fall behind
Documentation drifts for boring reasons. Projects close out, as-builts get filed, and ownership of those records blurs. Tenants modify spaces. Equipment is swapped for units with different footprints. Emergency repairs reroute the conduit. None of it feels significant enough to trigger a formal update, and owners aren’t wrong to skip most of it. The problem is that the same instinct gets applied to changes that do matter, and nobody sorts one from the other until a project is already underway.
My rule of thumb: on any building that has changed hands, changed tenants, or changed use in the last fifteen years, I assume the drawings are wrong about mechanical routing, electrical panel locations, and anything structural or load-dependent until a scan proves otherwise. I’ve rarely been wrong in the first two.
What I’ve assessed firsthand
A few years ago we were engaged on a federal defense facility at the Washington Navy Yard in DC, a government renovation project in an active operational environment where the existing drawings hadn’t kept pace with years of infrastructure changes. The client needed comprehensive as-built documentation of the MEPF systems as a verified baseline before any new design work could begin. What the drawings showed and what was actually installed were two different things.
We developed a LOD 350 MEPF BIM model using Autodesk Revit and Navisworks, capturing the real routing paths, connection details, and system configurations across the facility. The as-built model became the foundation for clash coordination and shop drawing production. The project detail is here. Had the design team proceeded against the original records, the clash issues would have surfaced during construction, in a federal facility where operational disruption isn’t just a schedule problem.
Which gaps actually cost money
The discrepancies that blow up budgets cluster in a few categories:
- anything load-bearing or load-dependent structural members, raised-floor capacity, floor loading
- anything routed through walls and ceilings like mechanical, electrical, plumbing.
Get those wrong and the design itself has to change. Cosmetic gaps a missing partition, an outdated finish schedule are annoying but rarely move the number.
Across the assessments we’ve run, ceiling heights and MEP routing are the two things as-builts get wrong most consistently. They are also the two things most likely to force a design revision mid-construction if they’re not verified beforehand.
In data center environments the stakes are higher still. Power capacity, cooling loads, and floor loading all depend on precise spatial and systems of information. An outdated floor plan cannot tell you whether a raised-floor section handles new equipment density, or whether cable pathways have room to grow. Get that wrong and it is not a cost overrun it is a service disruption.
What to do about it
Three moves worth making. First, before any project, do a targeted reality check against the high-cost categories not the whole building, just the systems and structural elements that would force a design change if they’re not where the drawings say.
Second, treat existing-conditions capture as a non-negotiable scope item whenever the building has undocumented changes, whenever the project touches anything load-bearing or routed, and whenever the cost of a wrong assumption mid-construction exceeds the cost of verification upfront. That condition is met on nearly every renovation of any substance.
Third, keep the model current after closeout but only for what matters. Trying to maintain everything is how the process collapses under its own weight.
Documentation is not a deliverable file at closeout. It is an asset, and like any asset, the only question worth asking is which parts are worth maintaining. The barrier was never the technology. It is the assumption that old drawings are good enough to plan from. They rarely are, and the projects that learn that mid-construction finish late, over budget, and with everyone blaming everyone else.
________________________________________________________________________
Author Bio: Reetie Lubana is Director of Operations at Tejjy Inc., an 8(a) certified BIM and digital engineering firm specializing in existing-conditions documentation, Scan-to-BIM, and facility data management for commercial, institutional, heritage, and data center projects. tejjy.com
________________________________________________________________________