Geotech Report vs Footing Design: The Reconciliation Nobody Does
AS 2870 classifies residential sites by expected ground movement — A, S, M, H1, H2, E and P in ascending severity — and ties permissible footing and slab systems to that classification. The reconciliation question is brutally simple: does the site classification stated on the structural footing drawings match the one assigned in the final geotechnical report, and do the drawings implement the report's specific recommendations? It is a comparison of one letter and a page of recommendations across two documents. It is also, in practice, a check that falls between every stool on the project: the geotech has finished, the structural engineer assumes their brief was current, the builder assumes the engineers talked, and the certifier assumes the consistency was someone's job.
Why two documents disagree about one site
- Sequence. Structural design often starts before the geotech investigation is complete, on an assumed classification ("design for M, confirm on receipt of report"). The report arrives assigning H1; the assumption is never revisited, and "Class M" is printed in the footing drawing's notes for the life of the project.
- Templates. Volume builders and small structural offices reuse general notes sheets. A classification carried in a template note survives from the last job — the drawings are internally consistent and confidently wrong.
- Report revisions. Additional boreholes, a cut/fill change, a moved footprint — geotech reports revise like any document, and a reclassification in Rev 2 of the report has to chase a footing design already issued for construction. If the report is not in the drawing register with a paired revision history, nobody is even tracking that it changed.
- Recommendations that never land. Beyond the class letter, reports carry site-specific instructions: found footings below the fill layer, pier through the uncontrolled fill to natural ground, maintain site drainage away from the slab, allow for trees within influence distance. These live in report prose; the drawings are where they become instructions. The gap between "the report said" and "the drawings show" is precisely where post-handover slab claims are argued.
The check, step by step
- Classification match. The class stated in the final geotech report (current revision) versus the class stated on the structural general notes and footing plans (current revision). One value, two documents. Any mismatch is a finding with both citations, resolved by RFI to the structural engineer — never by assuming the drawings "must have been designed for the higher class anyway".
- Class P handling. A P classification means the standard's standard solutions don't apply and site-specific engineering is required. Drawings responding to a P site with a catalogue slab type, with no reference to the specific engineering, are a finding of a different order.
- Recommendation flow-through. Walk the report's recommendations section line by line against the drawings: founding material and minimum depths versus the drawn footing depths; fill treatment versus what the sections show; pier requirements versus the pier schedule; drainage and moisture-management conditions versus the site works documentation. Each recommendation either appears on a drawing, is explicitly superseded by the engineer's design, or is a gap.
- Assumption audit. Reports state the conditions they relied on — footprint location, finished levels, cut/fill extents. If the architectural or civil set has since moved the building or changed the earthworks, the report's basis may be stale even though nothing "disagrees" yet. This is where revision tracking on the architectural set feeds a geotech re-check — the cross-discipline trigger logic that makes automated discipline checks more reliable than calendar-based reviews.
- Re-check on either side revising. New report revision, new footing issue, or a footprint/levels change: the comparison re-opens. The check takes minutes; the discipline is remembering to run it, which is exactly what should not depend on memory.
Making it someone's job
This reconciliation has no natural owner, which is why it doesn't happen. The fix is administrative, not technical: put it in the CA's or building surveyor's documentation review checklist at "for construction" issue, record the check as a finding set (match confirmed, or discrepancies raised) rather than a mental note, and keep the geotech report in the controlled document set so its revisions trigger review like any drawing's would. Where slab preparation revealed conditions differing from the report — the classic latent conditions moment — the same paper trail is what a claim under the contract stands on. One letter, two documents, minutes to check: there is no better ratio of effort to avoided consequence anywhere in the compliance checking landscape.
FAQ
What is an AS 2870 site classification?
A classification of residential sites by expected ground movement — A, S, M, H1, H2, E, P — assigned by the geotechnical engineer, which drives the footing and slab systems the standard permits.
Who assigns it?
The geotechnical engineer in the site investigation report. The structural engineer designs to it — and must be checked against it.
How do the documents end up inconsistent?
Structural design starting on an assumed class, template notes carrying old classes, report revisions arriving after footing issue, and report recommendations that never land on drawings.
What happens when they conflict?
Pre-pour: an RFI and a redesign. Post-handover: slab movement, a defect claim, and two contradictory exhibits with your project's name on them.
Run it through ParitySense alongside your manual review and measure the delta.
See how it works