01: Consolidate the field record first
01
Field vs drawing
Reconciling field conditions with drawings and schedules
The drawing is a claim about the building; the field is the record. Reconciliation is the process of establishing exactly where those two disagree, and reporting the disagreement rather than resolving it away.
Reconciliation compares what a survey found against what the record drawings and design schedules say should be there. Match on tag, verify on model and capacity, then classify whatever does not resolve: missing, undocumented, altered, renumbered, relocated or a quantity mismatch. Each class carries a different consequence for the design, so each is reported as its own finding.
Why the two sources drift
A building runs for decades after its drawings are filed. Chillers fail and are replaced with whatever was available and affordable that quarter, usually at a slightly different capacity. A tenant fit-out adds eight fan coils that never reach the archive. A contractor renumbers a run of terminal units during commissioning and the numbering lives on nobody else's drawing. Somebody moves a pump.
These are examples of how records can drift from site conditions. Reconciliation checks the differences without assuming their cause.
Match on tag, verify on substance
Tags are the natural join because both documents are organized around them. But a tag match is a hypothesis, not a result. The verification is the model and the capacity: if RTU-3 on the schedule is a fifteen-ton unit and RTU-3 on the roof is a twelve-and-a-half ton unit from a different manufacturer, the mismatch needs investigation. Check the equipment identity, rating basis and document revision before concluding that a replacement occurred.
Where site tags follow a different convention entirely, the match has to be made on location and type instead, and it should be recorded as a lower-confidence binding rather than presented with the same certainty as a serial match.
The discrepancy classes
01: Missing in field
02: Undocumented in field
03: Altered
04: Renumbered
05: Relocated
06: Quantity mismatch
| Class | What you see | Usual cause | How to resolve |
|---|---|---|---|
| Missing in field | Scheduled unit no capture matches | Removed, replaced under a different tag, or in an area not surveyed | Confirm removal before it is priced as existing plant |
| Undocumented in field | Capture matching no schedule row | Added during a later fit-out and never drawn | Add to the record — it is load and maintenance the design must carry |
| Altered | Tag matches, model or capacity does not | Replaced in kind at some point, often at a different capacity | Verify both values and record the reason for any chosen design basis |
| Renumbered | Same unit, different tag on site than on the drawing | Contractor renumbering that never reached the archive | Bind both tags so each remains traceable |
| Relocated | Unit found, location disagrees | Moved during works, or the drawing was wrong originally | Verify field placement and retain the drawing reference |
| Quantity mismatch | Counts differ for an otherwise matching type | Partial replacement, or duplicate rows in one of the sources | Resolve duplicates first — the count is usually the symptom |
Working through a reconciliation
02: Bind the tags
02
Bind the tags
03: Verify the bindings
03
Verify the bindings
04: Classify the remainder
04
Classify the remainder
05: Separate unresolved from absent
05
Separate unresolved from absent
06: Report the disagreement, not an average
06
Report the disagreement, not an average
Never average a disagreement
The most damaging thing a reconciliation can do is produce a single tidy number where two sources disagreed. It looks like a resolved survey and it destroys the finding. Six months later nobody can tell that the drawing said fifteen tons, the plate said twelve and a half, and someone chose.
Carrying both values, with their sources, costs a column and preserves the only thing the reconciliation was for. It also means the eventual decision is made by the person with the design responsibility.
Doing this as you go
Reconciliation is usually a desk exercise performed weeks after the visit, which is why “scheduled but never found” so often turns out to mean “nobody looked in that room”. Running the comparison while the survey is live changes the class of the finding: a schedule row with no matching capture becomes something to go and check before leaving the building.
That is what existing conditions survey software is aiming at, and the survey process guide covers where it fits in the sequence.
When the two disagree
How do you compare field conditions with drawings and schedules?
Start with tags, then verify equipment type, model, location and comparable ratings. A model or capacity mismatch is a finding to investigate, not proof of replacement. Record unmatched or conflicting information separately and state the evidence needed to resolve it.
Which source do you trust when they disagree?
The field, for what exists today; the drawing, for what was intended. They are not competing claims about the same thing. The mistake is resolving the disagreement into a single value — you lose the fact that a discrepancy existed, which is often the finding the client most needs.
What if there are no drawings at all?
Build a field record from the available observations and documents. Record missing information and uncertain relationships. Comparison against design intent must wait until a suitable drawing or schedule is available.