How to perform an MEP existing conditions survey

Agree the required level of detail before the visit. A survey supporting capital planning needs age, capacity and condition, and needs electrical detail appropriate to the planning decision. A survey supporting retrofit design needs the electrical ratings and the physical constraints, because feeder sizes and rigging routes come off them. A due diligence survey needs coverage and defensibility above all — the reader is deciding whether to buy the building.

Write the purpose down and derive the field list from it. The alternative is capturing whatever is convenient and discovering the gap when someone asks a question the data cannot answer.

Work by system and by floor, not by whatever door opens next. Start at the primary plant — chillers, boilers, main air handlers, main switchgear — because those units anchor the relationships everything else hangs from. Then follow the distribution outward: risers, floor plant, terminal units, local equipment.

Surveying in system order means that when a terminal unit appears, you already know what probably feeds it. Surveying in door order means reconstructing all of that later from photographs, which is where hours disappear.

This is the step that separates a survey from a photo album, and it has to happen while the building is still in front of you. Four questions:

Which captures could not be read? Which schedule rows were never matched to anything found? Which captures could not be tied to a room? Which areas were refused or inaccessible?

Some gaps can be resolved during the visit. When access remains unavailable, record the limitation precisely — which room, why, and what would be needed — so a return visit can be scoped.

Back at the desk, compare what was found against what was drawn and scheduled. The differences fall into a handful of classes: equipment on the schedule that was never found, equipment found that appears on no schedule, tags that match but models that do not, renumbered tags, and location disagreements.

Each class means something different for the design, so report them as distinct findings. The reconciliation guide works through each one. Where a photograph and a drawing disagree, keep both — never average them into a single number whose origin nobody can reconstruct.

The deliverable is an equipment schedule plus its boundary. Each unit carries identity, ratings, age, location, condition and its place in the system. Alongside it sits the list of what is unresolved and why.

Keep the evidence reachable. Existing conditions records get read during claims and value engineering exercises, and the useful question is never what the spreadsheet says but how you know. Being able to point at the photograph a capacity was read from is a materially different position than relying on recollection.

Most of the steps above are unavoidable — the walking, the access, the judgement. What changes is where the checking happens. If captures become records as they are taken, step five stops being a hopeful review of a camera roll and becomes a list the system already knows: these three plates did not read, these two schedule rows were never matched, this capture has no room.

That is the difference MEP field survey software is aiming at. Organizing captures and findings during the visit can reduce later transcription and help focus the office review.