The ground model and Eurocode categories in a geotechnical report

A geotechnical assessment is built around a ground model: a layered description of what sits beneath the site, how thick each layer is, where water stands within it, and how each layer behaves when you load it or cut into it. The Eurocode 7 geotechnical category, the scope of ground investigation and the foundation solution all follow from that model. For a self-builder assembling a minor application of one to nine dwellings, the practical question is how much of that work the council expects before permission and how much can wait for a condition.

Usually, not much before permission. Councils validating a small residential scheme want evidence that a known ground hazard has been identified and that the scheme can be built without harming neighbours, highways or controlled waters. Full intrusive investigation normally follows under a condition discharged before work starts, which is why a geotechnical assessment at application stage is mostly a desk exercise with a clear statement of what remains to be proven.

What the ground model is and what it rests on

A ground model is a set of statements about strata, each one traceable to a source. Made ground 1.2 m thick over firm brown London Clay, groundwater struck at 3.5 m, no evidence of solution features: that is a model. It is not a borehole log, and it is not a soil classification table. It is the interpretation that sits on top of those things.

Three inputs build the preliminary version. Published geology from the British Geological Survey gives the expected superficial deposits and bedrock. Historical mapping shows what the site was before, which is where made ground, backfilled pits, former ponds and old foundations come from. And the walkover tells you what the published data cannot: cracked boundary walls, a slope that has crept, standing water in October, a neighbouring extension that has clearly been underpinned.

Where the three disagree, say so in the report. A model that admits uncertainty is far more useful to a reviewing engineer than one that quietly smooths it over.

Desk study, walkover and the preliminary model

The desk study costs little relative to everything that follows, and it is the stage where most money is saved or wasted. For a small site the work runs roughly in this order.

1. Collect the BGS geology, historical Ordnance Survey editions, any available mining or quarrying records, radon potential, and a check of whether the site falls in a Development High Risk Area for coal mining legacy.
2. Review any ground information from neighbouring developments, which for infill plots is often the single most useful source.
3. Walk the site, photograph boundary structures, record levels and note trees, drainage and any sign of movement.
4. Write the preliminary ground model with an explicit list of what is assumed rather than known.

That last point matters at planning stage. A sentence such as "made ground thickness assumed at 1.5 m, to be confirmed by trial pits before design" is the honest version and it tells the case officer that the risk has been scoped rather than ignored. Where a ground condition interacts with contamination, the two reports stay separate: a contaminated land assessment deals with sources, pathways and receptors, while the geotechnical work deals with strength, settlement and stability.

Geotechnical categories and the investigation they imply

Eurocode 7, published in the UK as BS EN 1997, sorts structures into three geotechnical categories by the consequence of failure and the complexity of the ground. The engineer decides the category; the council does not. It remains the clearest shorthand for how much investigation a scheme needs.

Most two-storey houses on ground the BGS already describes well sit in Category 2. A single garden plot with no history and no slope can sit in Category 1. A deep basement under a terrace is Category 3 and needs the full treatment, which is where a basement impact assessment takes over the hydrogeology and movement analysis.

The category also drives who signs the report. Category 2 and 3 work is engineering judgement with liability attached, and councils expect a suitably qualified author. Professional review is strongly advised for any geotechnical document that will support a decision.

Sampling, in situ testing and laboratory scheduling

A scope of investigation on a small site is a short, specific list, and vagueness here is what produces useless data. Name the technique, the depth, the number of positions and the tests.

• Trial pits to 3 m to log made ground, prove founding stratum and inspect existing foundations, with one pit per plot as a minimum.
• Window sample or cable percussion holes to 6 to 10 m where piling or a retaining structure is contemplated, with standard penetration tests at 1 m intervals in granular soils.
• Hand shear vane or undrained triaxial tests on cohesive samples for bearing capacity.
• Atterberg limits and moisture content for shrinkage assessment in clay.
• Soakaway infiltration testing to BRE Digest 365 where surface water is going to ground, which the drainage strategy will rely on.

Schedule the laboratory work after the logs are available rather than before. Blanket test suites ordered in advance are the commonest way a small ground investigation budget disappears into results nobody uses.

Shrink swell clay and foundation depth

On firm to stiff clays of high or very high volume change potential, foundation depth is set by desiccation and tree influence rather than bearing capacity. Clay shrinks as it dries and heaves as it rewets, and a foundation sitting in that zone moves with it.

The method is well established. Classify the clay by plasticity index into low, medium, high or very high volume change potential, identify the mature and proposed vegetation within influencing distance, apply the water demand category of each species, and read off the minimum depth and any requirement for heave precautions such as compressible fill or a suspended floor. Where trees have been removed, long-term heave can be the governing case, and that is a design condition builders routinely miss.

Illustrative example: a pair of semi-detached houses on high volume change clay with a mature oak 8 m from the rear elevation. The bearing capacity at 0.9 m is perfectly adequate. The foundation still goes to 2.4 m on that elevation because of the oak's water demand, and the trench fill cost roughly doubles. Figures here are illustrative and depend on the plasticity index and species measured on site; the point is that the controlling factor is vegetation, not load.

Groundwater, slope stability and mining legacy

Three hazards turn a routine report into a phased one. Groundwater matters for buried structures, for soakaway viability and for excavation stability, and a single autumn dip is weak evidence. Where it matters, monitor through a winter.

Slope stability needs attention on any site with more than a gentle fall, on old landslip ground, or where a cut or fill platform is proposed. Mining legacy needs a dedicated coal mining risk assessment in a Development High Risk Area, and the report should state plainly whether shallow workings, shafts or adits are suspected and what treatment follows. None of these can be resolved by desk study alone, and a report that implies otherwise will not survive review.

What planning needs against what building control needs

The two regimes want different things from the same investigation, and conflating them wastes money.

The planning authority is interested in whether the development is acceptable in land use terms: that ground instability has been identified, that neighbouring buildings and the highway are protected, that controlled waters are not put at risk, and that any resulting mitigation does not change the scheme beyond what has been applied for. That is usually satisfied by a desk study plus a condition requiring investigation and remediation before works start, drafted along the lines the planning guidance on conditions describes. A structural survey of adjoining property may be asked for where excavation comes close to a boundary.

Building control wants design values. Bearing capacity, settlement limits, sulphate class for concrete, gas protection measures, foundation depth and detail. That evidence has to be real and site-specific, and it is checked against the submitted calculations.

Write one investigation that answers both, and sequence it so the planning submission carries the desk study and the conditional commitment while the design stage carries the intrusive work.

PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval. For specialist technical reports, qualified professional input may be required. Enter your postcode on PlanScope to see which reports your site is likely to need.