Ora LaboraOra Labora Book a consultation

Site Investigation: What Happens on Site and at the Lab

By · November 16, 2026 · Structural Surveys · Written by Munyovi Victor Indiazi

Window sampling rig driving a borehole during a site investigation, with an engineer in orange hi-vis at the controls

Nobody designs the ground. It was there first. It was laid down over tens of thousands of years by ice, rivers and weather, it varies across a single garden, and it has no interest whatsoever in your drawings. A site investigation is the process of going out and finding what it already decided — before anyone commits to a foundation built on an assumption.

What follows is a real one, start to finish: what happens on site, what goes to the laboratory, and what the results actually change about the design.

What is a site investigation?

A site investigation is the work of establishing what the ground beneath a site is made of, how strong it is, how it behaves when it gets wet or dries out, and whether it is chemically aggressive to concrete — so that a foundation can be designed for the ground that is genuinely there rather than the ground someone expected.

In the UK it is carried out in line with BS 5930, the code of practice for ground investigations, and the results feed straight into geotechnical design under Eurocode 7. On a house, an extension or a small commercial building it usually comes down to a small number of holes, a careful record of what comes out of each one, and a set of laboratory tests on the samples recovered.

It is not the same thing as a structural survey. A structural survey looks at the building. A site investigation looks at what the building is standing on. If you are trying to work out which of the two you need, we have set out the difference in detail in our guide to a structural survey versus a structural appraisal.

Why the ground is the one element nobody gets to specify

Every other part of a building arrives with a specification. The steel has a grade. The concrete has a mix. The timber has a strength class and a certificate to prove it. You can look any of them up, and if you do not like what you read you can order something else.

The ground is the exception. You inherit it. You cannot order a different one, and the only way to find out what you have been given is to go and look — which, inconveniently, means putting holes in it.

That is the whole argument for a site investigation in one sentence. Everything above ground level is a known quantity. Everything below it is a question, and a foundation is the answer to that question whether or not anybody has bothered to ask it.

It starts with a walkover, not a drill

The first thing that happens on site is not drilling. It is walking.

A walkover survey decides where the holes go, and getting that wrong wastes the whole exercise. The engineer is reading the surface for what it says about the depth: the fall of the land and where water will collect, trees and large shrubs and how close they sit to the proposed footprint, soft or boggy patches, made ground, old slabs, filled hollows, the line of drains, and any sign that a previous building stood there.

Trees matter enough to change a foundation on their own. On a shrinkable clay, a mature high-water-demand species close to a building can require a foundation two or three times deeper than the same house would need on an open plot. We have written separately about how close is too close when it comes to trees and subsidence, and about the shrubs and leaking drains that cause the same damage without anyone blaming them.

Made ground matters just as much and is harder to spot. A plot that was levelled twenty years ago with whatever came off the last job can look like a perfectly ordinary field until the first hole goes down.

What is window sampling?

Window sampling is a method of driving a series of steel tubes into the ground with a percussive hammer and recovering the soil inside them in one continuous column. The tubes have long slots — the “windows” — cut down one side, so the soil can be seen, described and sampled without being tipped out and jumbled up.

The rig is small, tracked, and will fit through a side gate, which is the reason it dominates domestic and small commercial work. Tubes are driven in roughly one-metre runs, stepping down in diameter as the hole deepens, and a typical investigation reaches somewhere between four and ten metres depending on what it finds. Dense gravel, obstructions or rock will stop it, and when that happens the hole is logged for what it did reach and another method is considered.

Two things are being recorded while it works. The first is the material itself. The second is how hard it is to drive — the blow counts as each tube goes down give a continuous read on the ground getting stiffer or softer with depth, which is often the first clue that competent ground has been reached.

Why the soil is logged by depth rather than by the bucketful

Everything that comes out of the hole is laid out in the order it came out, and every change is recorded against the depth it occurred at.

This is the part that looks like gardening with a clipboard, and it is the part that carries most of the value. A pile of soil tells you almost nothing. The same soil, laid out in sequence with depths written against it, tells you that there is 0.4 m of topsoil, then 1.1 m of soft brown clay, then firm grey clay to 4.2 m, then gravel — and that sequence is what a foundation is designed from.

Depth is the variable that decides the cost of the foundation. Knowing that competent ground starts at 1.6 m rather than 2.8 m is the difference between a trench fill foundation and a piled one, and that difference is usually larger than the entire cost of the investigation.

Each sample is then bagged, sealed and labelled with its hole reference and its depth, and signed over to the laboratory under a chain of custody. Sealing matters more than it sounds: moisture content is one of the things being measured, and a sample that has sat open in the back of a van has already changed its answer.

What the laboratory actually tests

The samples go to a geotechnical laboratory, and what is asked for depends on what the holes found. Three groups of tests cover most low-rise work.

Classification and plasticity

Classification establishes what the soil is: particle size distribution for granular soils, and the Atterberg limits — liquid limit and plastic limit — for cohesive ones. The gap between those two is the plasticity index, and the plasticity index is what tells you how much a clay will shrink in a dry summer and swell again over winter.

That single number drives foundation depth on clay sites. NHBC guidance sorts clays into high, medium and low volume change potential on the basis of it, and then reads foundation depth off that classification together with the species and distance of nearby trees. Get the plasticity wrong and every depth that follows is wrong.

Strength and bearing capacity

Strength testing establishes how much load the ground will carry without failing or settling unacceptably. In cohesive soils that usually means undrained shear strength, measured with a hand vane on site and confirmed by triaxial testing on undisturbed samples. In granular soils it is inferred from the penetration resistance recorded while the hole was being driven.

Bearing capacity is not a single property of a soil. It depends on the strength of the ground, the depth and width of the foundation and how much settlement the building can tolerate, which is why the laboratory produces the ingredients and the engineer produces the answer.

Chemistry — sulfates, pH and contamination

Concrete put into aggressive ground will deteriorate, so sulfate content and pH are tested and the result is classified under BRE Special Digest 1. That classification sets the concrete specification for the buried elements. It is routine, it is cheap, and skipping it means guessing a concrete mix that has to last a century underground.

On brownfield sites or anywhere with a history of industrial use, contamination testing is added — metals, hydrocarbons, asbestos — because the planning condition will usually require it and because it changes how the spoil can be handled.

What the results actually change

A completed site investigation hands the designer four things that a foundation cannot honestly be designed without.

  • Bearing capacity — how much load the ground will take, and therefore how wide the foundation has to be.
  • Depth of competent ground — where the soft and variable material stops and the reliable material starts, and therefore how deep to go.
  • Shrink and swell potential — how much seasonal movement the clay is capable of, which sets the minimum depth and decides whether heave precautions are needed.
  • Groundwater — where the water sits, which affects the excavation, the temporary works and the buildability of the whole sequence.

Measured, those four numbers usually make a foundation cheaper, not dearer. A designer working without them has to cover the worst credible case, and covering the worst credible case means concrete that was never needed. Measured, the design is sized for what is actually under the building.

Assumed, they do not disappear. They turn up later as a variation, with the groundworker standing in a trench waiting for an answer and the programme stopped around him.

Site investigation, trial pit or structural investigation?

These three get used interchangeably and they are not the same job.

A site investigation samples the ground across a site, usually by boring, and is aimed at designing something new or explaining ground-related movement. A trial pit is a machine-dug hole next to an existing foundation that lets you see the foundation itself — its depth, its width, what it is made of and what it is bearing on. Trial pits go shallower, but they show you things a borehole never will; we have covered what they reveal in our guide to trial pits.

A structural investigation is about the building rather than the ground — opening up finishes to see what the structure is doing behind them. That is a different exercise again, described in what happens in a structural investigation.

On a building that is already cracking, the honest answer is often that you need more than one of them. The ground tells you what is moving and the fabric tells you how much it has moved.

When do you actually need one?

New builds

Effectively always. Building Control and warranty providers will expect ground information, and on a plot with no building history there is nothing else to go on. This is the case where the investigation is least negotiable and most useful.

Extensions

It depends on the site. On a straightforward plot with no trees, no made ground and a neighbouring house that has stood still for eighty years, trial pits next to the existing foundation are often proportionate. Add a mature oak, a filled hollow, a nearby watercourse or a sloping garden and a proper investigation earns its fee immediately.

It is worth remembering that the junction between an extension and an existing house is one of the most common places for cracking to appear, frequently because the two are founded at different depths on ground that moves. We have written about cracks where an extension meets the house and why they so often trace back to the ground.

A building that is already moving

When a property is cracking and subsidence is suspected, ground investigation is how the cause is established rather than guessed. Insurers and loss adjusters expect it, and a structural report for an insurance claim that is not backed by ground data tends not to survive scrutiny.

Investigation is usually paired with monitoring, because a single set of readings cannot tell you whether movement is still happening — something we set out in why one crack measurement is never enough.

What does a site investigation cost?

It is priced per site rather than off a list, and four things move the figure: how many holes are needed and how deep they have to go, how easy the rig is to get to the positions, which laboratory tests the ground turns out to warrant, and whether contamination testing is required.

The useful way to think about it is proportion. On almost every residential scheme the investigation is a small percentage of the foundation cost alone, and the foundation is a fraction of the build. Against that, the two outcomes it exists to prevent — a foundation designed deeper than it needed to be, and a foundation designed shallower than it needed to be — are both measured in thousands.

The second of those is the expensive one. A foundation that is too conservative wastes concrete once. A foundation that is too optimistic produces movement, cracking, an insurance claim and underpinning, and it does so slowly enough that nobody connects it to a decision made before the first course of bricks went down.

What it looks like when it is skipped

The pattern is consistent enough that we can usually predict it before arriving.

A building goes up on an assumption. The assumption holds for a few years, because assumptions usually do. Then a dry summer arrives, the clay under one corner shrinks further than the clay under the other, and the building finds out about the difference by cracking. By that point the options are investigation, monitoring, a report and a repair — which costs considerably more than the investigation would have, and takes a year rather than a week.

None of this is exotic. It is the most ordinary structural failure there is, and almost all of it is a ground problem that was never looked at.

Site investigations in Peterborough and London

Ora Labora Structural Engineers carry out site investigations and ground investigations across Peterborough, Cambridgeshire and the surrounding towns, and from our London office across north London and Hertfordshire.

The ground differs between the two, and so does the work. Around Peterborough we see Oxford Clay, river terrace deposits along the Nene and a great deal of made ground on former industrial land. Across north London and Hertfordshire it is the London Clay Formation and the boulder clays above it — high plasticity, strongly seasonal, and behind a large share of the subsidence claims in that part of the country.

If you are planning a new build, designing an extension on a difficult plot, or dealing with a building that has started to move, a site investigation is where the answer comes from. Our team are chartered, we report in plain language, and the recommendations are written so that a groundworker can build from them.

For more on what we do and how we work locally, see our guide to structural engineers in Peterborough, or get in touch to discuss a site.

Have a question about your own project?

Send us a drawing or a photo and an engineer will tell you what is possible.

Book a consultation
Google Reviews 4.9 31 reviews
Write a review
Andrew Vear 6 months ago

Stella did a great job for us. She liaised with building control (quite a challenge) and thought of things we would have never considered. Overall she did a great job for us and we are extremely happy with the outcome.

Read on Google
Luke Sowinski a year ago

We used Ora Labora to get structural drawings for the removal of 5 load bearing walls. I had contacted many other structural engineer firms but they were hesitant to take on the project. When I rang Ora Labora there were no issues and they were able to come out quickly to survey the property. Everyone was very professional and I was able to get the final drawings very quickly at a good price. I handed these over to my builder and they were able to complete the work without any issues. I would highly recommend and would use them again!

Read on Google
Graham Stewart 3 years ago

We were referred to Ora Labora and it was a great referral. Stella came around and checked the property and had given us enough notice of where we needed to excavate for her to get a better understanding of the existing building. Drawings were submitted for our approval before final drawings completed. All done in great time, to me. Thank you Ivy for all your effort and communication as well. Well deserved 5 plus stars.

Read on Google
Jay Shah 4 years ago

Highly recommended structural engineer. I was introduced by a friend to a very good and efficient structural engineer, Stella Macharia. She analysed the project and supported me during the process of new build on the 3rd floor. Stella attended at the site time stated and identified the issue straight away, and was very knowledgeable. I would highly recommend Ora Labora and her team and wish to continue to work with them for any future projects.

Read on Google
derek elding 6 months ago

Excellent service from all at Ora Labora, they work with you to achieve the best result for your requirements.

Read on Google
Harry Gresham 7 months ago

Smooth process and trusted team to work with for our internal renovations. Would use again.

Read on Google
yasamine watts 6 months ago

Helped us very quickly in an urgent situation, excellent work. Would definitely recommend.

Read on Google
James Ekberg 8 months ago

Really impressive service - prompt, professional, and helpful.

Read on Google
S.A.M 5 months ago

Smooth experience working with Ora Labora Ltd. Will definitely use again.

Read on Google

See all reviews on Google →