Using geological data for a desk study

Most of the geological data you need for a desk study is free, and the limiting factor is knowing what the records can and cannot tell you. The British Geological Survey publishes mapped bedrock and superficial deposits, a national archive of scanned borehole logs and a set of hazard datasets covering shrink-swell, landslide potential, soluble rocks and compressible ground. Used properly, geological data for a desk study at a site will tell you what to expect, where the uncertainty sits and what the intrusive investigation needs to target. It will not tell you a foundation depth.

What the national survey publishes free of charge

Two viewers do most of the work. The Geology of Britain viewer shows mapped bedrock and superficial geology at a scale suitable for a site appraisal, with the formation names you will need to quote. GeoIndex Onshore is the more detailed tool, with layers for borehole locations, mapped artificial ground, mine entries where recorded and the hazard datasets.

Start with the two geology layers rather than one. Bedrock is the solid formation beneath the site. Superficial deposits are everything laid over it since: alluvium along a watercourse, river terrace sand and gravel, glacial till across much of the Midlands and the north, head deposits on slopes. The superficial layer usually governs foundation and infiltration behaviour, and it is the layer people skip.

Artificial ground is mapped separately and patchily. Where the survey has recorded made ground it is worth knowing, and absence of mapped made ground means nothing at all on a site with any industrial history.

Borehole records and their reliability

The borehole archive is the most useful free resource in UK ground engineering and the most frequently over-read. Records are indexed by location and many are available as scanned logs you can open in the browser, including historic water-well records, site investigation boreholes, trial pits and window samples.

Four things limit what a record proves.

1. Location accuracy. Older records were positioned by hand on a paper map, and an error of tens of metres is common. A log 40 metres away may be describing a different deposit.
2. Who wrote it. A driller's log describing "brown clay" is not an engineer's log with plasticity indices and undrained shear strengths, and most of the archive is the former.
3. Depth of interest. A water well sunk in 1908 cared about the aquifer and recorded the first 3 metres in a single line.
4. Date. Ground that has since been developed, filled, cut or remediated will not match a record from before that happened.

Use the archive to build an expectation of the stratigraphy and to pick the depths your own investigation should reach. Treat a single nearby log as a hypothesis, and two or three logs that agree with each other as a reasonable working model.

Superficial deposits against bedrock layers

The distinction matters because it determines which questions you have to ask. Bedrock tells you about long-term settlement, rock head depth, aquifer status and whether rock excavation is on the cards. Superficial deposits tell you about variability.

Alluvium is the one to watch on any site near a river. It is typically soft, often organic, highly variable over short distances, and frequently sits over gravel at an unpredictable level. A site with mapped alluvium and a flood zone designation needs the two investigations planned together, because the ground model feeds directly into the levels work described in a Flood Risk Assessment.

Glacial till behaves differently again, with cobbles and boulders that defeat small-diameter sampling and lenses of sand that carry water. And on sloping sites, head deposits and solifluction material can mask relict slope instability that the mapped landslide layer never picks up at site scale.

Shrink swell and other hazard datasets

The survey's hazard datasets rate susceptibility rather than predicting an outcome. Shrink-swell is the one most householders meet, because the dataset drives the National House Building Council foundation depth guidance for clay sites and the insurance industry's view of subsidence risk.

A high susceptibility rating is not a constraint in planning terms. It is an instruction about what the ground investigation has to answer, and a scheme that acknowledges it and scopes the work accordingly will read far better to a case officer than one that ignores it.

Mining and dissolution indicators

Coalfield sites need a separate check. The Coal Authority maintains the definitive record of coal mining legacy and publishes referral areas where a coal mining risk assessment is expected with a planning application. The geological mapping will show the coal measures; it will not show the shafts.

Non-coal mining is harder. Metalliferous mining, chalk and clay extraction, brickearth pits and limestone quarrying leave features that appear on historic maps long before they appear on any dataset, which is why map regression and the geological record have to be read together. The approach to that evidence is set out in using historic maps to evidence a former land use, and aerial imagery adds a further layer, as reading aerial photography as planning evidence explains.

Dissolution is the quieter hazard. Gypsum in parts of Nottinghamshire and North Yorkshire, chalk solution features in Kent and Hertfordshire, and limestone karst in the Mendips and the Peak all produce ground conditions that mapped geology hints at and only investigation confirms.

Where intrusive work is unavoidable

A desk study tells you what to expect. It cannot provide the parameters anyone designs from.

Intrusive investigation is unavoidable for foundation design to Eurocode 7, for infiltration rates if you want a soakaway accepted, for any contamination assessment, and wherever the desk study identifies a hazard the mapped data only rates as possible. A Geotechnical Assessment built purely on archive records will carry a long list of assumptions, and the authority's own consultant will say so.

Land contamination has its own rule. The local planning authority expects investigation carried out in accordance with BS 10175 by suitably qualified and experienced persons, and a contaminated land assessment is a report type where a competent professional is legally required rather than advisable. A draft prepared from desk data is a starting point for that professional and never a substitute.

Citing records in a report

Vague sourcing is what gets a desk study sent back. Record each source so that someone else can retrieve it.

For mapped geology, give the formation name, the mapped scale and the date the data was accessed. For boreholes, give the archive reference number for each log you relied on, its recorded location, its total depth and its date. For hazard datasets, name the dataset and the susceptibility class rather than paraphrasing it as "moderate risk".

Then state what you did not have. A desk study that says plainly which questions remain open, and what investigation will answer them, is more useful to a case officer and to your own engineer than one that smooths over the gaps. Honest uncertainty about rock head depth is a normal finding. An invented figure is not.

Enter your postcode on PlanScope to see which ground and site reports a project is likely to need.

This is general guidance, not legal advice; check the position with your local planning authority or a qualified professional before relying on it.

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.