Geotechnical investigation scope for a small site

A ground investigation on a single plot or a handful of units has one job: give the foundation designer enough information to design a foundation that will not move, and give the drainage designer enough to know whether water will soak away. Everything else is optional. Get the scope right and a small site investigation is a couple of days of site work and three weeks of laboratory turnaround. Get it wrong and you pay twice, because the engineer asks for a second visit after the rig has left.

The mistake self-builders make is buying a "soil test" without specifying what it has to answer. The mistake the industry makes is selling a standard package of two window samples regardless of what the site is.

What the investigation has to answer

Write the questions down before asking for a quotation. For a typical small site they are:

• What is the soil profile, and at what depth is a stratum capable of carrying the foundation?
• Is the near-surface soil shrinkable clay, and if so what is its volume change potential?
• Where is the groundwater table, and does it vary?
• Will water soak away at a rate that makes a soakaway viable?
• Are there obstructions, made ground, old foundations or fill?
• Is there anything that suggests contamination, mining, or dissolution features?

Those six questions determine the entire specification. A site on chalk with no trees nearby needs a very different investigation from a plot on London Clay with three mature oaks on the boundary.

Contamination is deliberately framed as a screening question here. A full contamination investigation to BS 10175 is a separate exercise, and where the desk study flags a former industrial use, a landfill, a filling station or made ground of unknown origin, the local planning authority will expect reports prepared by suitably qualified and experienced persons. A Land Contamination assessment is professionally required work, and an AI-assisted draft is a starting point for that professional rather than a replacement.

Desk study before any intrusive work

The desk study costs a fraction of the intrusive work and often changes it completely. On a small site it should cover:

1. Historical Ordnance Survey maps across the full sequence available, looking for ponds, quarries, watercourses, buildings and industry.
2. British Geological Survey mapping and borehole records, which frequently include logs from neighbouring sites.
3. Coal mining status, and a mining report where the site is in a coalfield referral area.
4. Radon potential, because it changes the construction detail rather than the foundation.
5. Flood and groundwater information, including any source protection zone.
6. A walkover recording levels, trees, existing structures, evidence of fill, cracking in nearby buildings and vegetation patterns.

The walkover is the part people skip and the part that pays. Cracking in a neighbour's bay window, a line of poplars along a boundary, or a distinctly flat platform in a sloping field all tell you something before a rig arrives.

Trial pits against boreholes on a small plot

Trial pits give you what a borehole cannot: a face you can look at, describe and photograph. On a small site with shallow foundations they are usually the better tool, because you see the fabric of made ground, the base of a former slab, root distribution and any perched water arriving.

Their limit is depth. A machine-excavated pit is practical to about 3 to 4 metres in stable ground, less in loose granular soil or below the water table, and the sides must not be entered without support.

Window sampling and light percussion boreholes take you deeper with a much smaller hole and allow undisturbed samples for laboratory testing. Where the investigation has to establish the depth of compressible alluvium, prove rockhead, or test at 6 metres or more, boreholes are the answer.

The sensible pattern on many small sites is a mix: two or three trial pits to characterise the near surface and confirm what is where, plus one or two deeper holes at the positions that matter for foundation design.

Number and depth of exploratory holes

There is no national rule for a single plot, and anyone who quotes one is applying a rule of thumb. What BS 5930 and Eurocode 7 do is set the principle: enough exploratory positions to establish the ground model with confidence, at depths sufficient to cover all strata affected by the works.

Practical rules of thumb an engineer will recognise:

• At least two exploratory positions on a single plot, and three or more where the ground looks variable or the footprint is spread out.
• Positions near the corners and at any point where the profile is expected to change, not all clustered in the middle of the easiest access.
• Depth to at least 1.5 times the foundation width below founding level for a strip footing, and deeper where compressible soil, fill or trees are involved.
• Where trees are a factor, deep enough to sample the clay below the likely influence of root-induced desiccation, which is usually deeper than people expect.
• One position dedicated to infiltration testing if a soakaway is proposed, at the depth and location of the proposed feature.

Plasticity testing is what turns clay into a foundation depth. Warranty provider standards, including the NHBC Standards, band volume change potential using the modified plasticity index and then set foundation depths against distance to trees and species water demand. That is why the investigation has to include Atterberg limits on samples from the right depths, and why the tree survey and the ground investigation have to be read together.

Testing schedule and turnaround

Specify the laboratory schedule rather than leaving it to the driller. For a small residential site it commonly includes classification tests such as moisture content and Atterberg limits, shear strength by hand vane in the field and triaxial or unconfined tests on undisturbed samples, sulphate and pH for concrete class selection to the relevant design guidance, and a screening suite for contaminants where the desk study justifies it.

Allow two to three weeks for laboratory results after site work, longer for a full contamination suite. That single fact reorders a self-build programme more often than anything else here: if you want foundation design for a January start, the rig needs to be on site in October, not December.

Infiltration testing to the BRE Digest 365 method takes its own time, because the test involves repeated fills and timing of the fall from three-quarters to one-quarter full, and it has to be done at the right season to avoid overstating performance in dry ground.

Access constraints and their cost effect

Access decides the plant, and plant decides the price.

A plot reached through a 900 millimetre gate needs hand-dug pits or a small tracked rig that can be dismantled. A rear garden with a single-storey extension in the way may need a hand-held window sampler with reduced depth capability. Low headroom under trees rules out a mast. Every one of those constraints reduces what the investigation can achieve and increases the day rate.

Two practical matters worth budgeting. Service clearance before any excavation, using a cable avoidance tool and hand-dug inspection to a safe depth, which is a requirement of safe working rather than an optional extra. And reinstatement, especially where pits cross a driveway or a lawn the neighbours can see.

An illustrative sequence for a single new dwelling on a sloping garden plot with two mature limes on the south boundary. Desk study first, which flags shrinkable clay and no mining. Then three trial pits at the corners of the footprint and one window sample to 6 metres at the downslope corner, plus one infiltration test position where the soakaway is proposed. Laboratory schedule: Atterberg limits on four samples, triaxial on two, sulphate and pH on two. Site work takes a day and a half. Results arrive at week three, and the engineer designs trench fill to a depth set by the plasticity result and the distance to the limes.

Using the results in the planning submission

Ground investigation is often treated as a building control matter, and on a constrained site it belongs in the planning submission too.

The results feed four documents. A Geotechnical Assessment summarising the ground model, groundwater and foundation recommendations. The drainage strategy, because infiltration rates decide whether a soakaway is acceptable and whether attenuation and a controlled discharge are needed instead. The arboricultural submission, because foundation type and depth near retained trees is exactly what a tree officer asks about, and the Tree Survey has to be read alongside the ground model. And, on sites with excavation below adjoining ground level, a Basement Impact Assessment where the local authority requires one, which our note on cellar conversion against new basement excavation looks at in more detail.

State clearly in the report what was not investigated and why. A small-site investigation is proportionate by design, and an honest limitations section protects everyone, including the designer who reads it in two years' time.

This is general guidance, not legal advice; check the position with your local planning authority, your building control body and a qualified geotechnical engineer before relying on it. For specialist technical reports, qualified professional input may be required.

Browse the report types on PlanScope and draft the documents your small site needs. PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval.