Geotech Report Review

A geotechnical report is read once, by the structural engineer, for the bearing value. Everything else in it — the water table, what has to be removed, what cannot be built in winter, what has to be observed during construction — becomes an architectural problem later, usually at the worst moment.

Workflow

Step 1. Establish the scope of the investigation

Ask, in one message:

1. What was the report commissioned for, and by whom? A report for a previous scheme is not a report for this one.
2. How many boreholes, how deep, and where were they relative to what is now proposed?
3. When was the fieldwork done, and in what season?
4. Does the scheme now extend beyond where the investigation went, in plan or in depth?
5. Is there a basement, and how deep?

Question 4 is the one that matters most. A building that moved fifty feet after the boreholes were drilled is a building on inferred ground, and that is a finding rather than a detail.

Step 2. Separate investigated from inferred

Plot, in words, what was actually tested against what the scheme now occupies:

Location	Investigated?	Depth reached	Under what part of the scheme

Report the parts of the footprint with no borehole under or near them, and the depths the scheme reaches that the investigation did not. Everything between boreholes is interpolation, and the report will say so somewhere; quote it.

Step 3. Pull out every recommendation that becomes an obligation

The report is written as advice. Once the design follows it, each item is a commitment. List them:

Recommendation	Quoted	Who it binds	What it costs the design

Work through at least: bearing stratum and depth, whether any material has to be removed and replaced, groundwater and what has to be done about it, slab support, lateral earth pressures, whether the ground is expansive or collapsible, seismic site class, corrosivity or chemical attack on concrete and buried metal, slope stability, and any requirement for ground improvement.

Step 4. Find the contingencies and the observation requirements

The two categories that surprise projects:

• Contingent recommendations. Anything phrased as "if X is encountered" or "subject to confirmation during construction". Each one is a decision deferred to the site, and each needs somebody who will make it, a budget, and time in the programme. List every one.
• Observation requirements. Where the report requires the geotechnical engineer to observe or test during construction, that is a scope item, a fee and a hold point. Every one gets named, with what happens if it is skipped.

Both belong in the specification and in the programme, and both are routinely neither.

Step 5. Translate to the building

What the report means for the architecture, not for the foundation:

• Does the ground floor level work, given the bearing depth, the water table and what has to be removed?
• Is the basement viable, and what does the water condition do to it? Point at design-below-grade-assembly.
• What does the site work look like: cut, fill, retaining, and what the report says about reusing excavated material.
• Programme: seasonal restrictions, dewatering, surcharge and waiting periods, ground improvement.
• What this does to the neighbours: vibration, dewatering drawdown, shoring against a party line, underpinning.
• Cost signals worth telling the client now rather than at tender.

Step 6. Report

• The investigated-versus-inferred finding, first.
• The obligations table.
• Every contingency, with who decides and when.
• Every observation requirement, with whether it is in anybody's scope yet.
• What it means for the building.
• Every inconsistency found, as a question for the geotechnical engineer.
• What further investigation is warranted, and when it has to happen to be useful.

Step 7. Save, if asked

Ask whether to write the review to a file and where.

Rules

• Never supply a soil value or a design parameter. Quote the report.
• Never verify or contradict a recommendation. Inconsistencies are questions.
• Compare the investigation's coverage against the scheme's footprint and depth, every run.
• List every contingent recommendation and every observation requirement.
• Say what the report means for the building, not only for the foundation.
• Never treat a report commissioned for another scheme as current.

Anti-patterns

• Reading the report for the bearing value and stopping.
• Treating interpolation between boreholes as investigated ground.
• Using a report from a previous scheme without checking the footprint moved.
• Leaving contingent recommendations for the contractor to discover.
• Omitting the observation requirements from the specification and the fee.
• Reporting to the structural engineer's audience rather than the design team's.
• Second-guessing a geotechnical recommendation.

Resources

None. This skill is one file. Output is written directly at the path you choose.