Made ground and fill: the structural consequences

Made ground does not usually fail a site. It changes the foundation solution, and the foundation solution changes the budget, the programme and sometimes the layout. The reason it catches developers is that the problem is not strength. A fill that will carry the load perfectly well across most of a plot can still be unbuildable on conventional foundations because its depth varies by three metres across 40 metres, and differential settlement between two corners of a building is what cracks it.

So the question to ask a geotechnical engineer is not "will it take the load". It is "how much does it vary, and over what distance".

Recognising made ground from a desk study

Almost all of it is visible in the desk study if you look at the right sources.

Historic map regression is the first and most productive. Sequential editions of the Ordnance Survey plans show quarries, claypits, brickworks, gravel pits, ponds, watercourses, railway cuttings and embankments, docks and basins, and the point at which each disappeared. A pond shown in 1890 and absent by 1930 is a filled void, and the fill is whatever was to hand at the time.

Geological mapping identifies artificial ground as a mapped unit, distinguishing made ground, infilled ground, worked ground and landscaped ground, and the published records of boreholes in the area often give thicknesses.

Aerial photography fills the gaps between map editions and shows tipping, earthworks and wartime damage.

Then the local knowledge: urban sites with a gap in the terrace are frequently bomb-damage sites backfilled with rubble, coalfield areas carry colliery spoil, and sites adjacent to former industry often carry process waste, ash and clinker.

What the desk study should produce is not a statement that made ground is present. It should produce a predicted thickness and distribution, with the evidence for each, so the investigation can be designed to test that prediction rather than to discover it.

Why depth and variability matter more than strength

Three mechanisms do the damage, and all three are about variability.

Differential settlement. A building founded partly on competent natural ground and partly on 2.5 metres of fill will settle unevenly, and the differential movement is what causes structural distress. The total settlement can be modest and still unacceptable.

Ongoing settlement from degradation. Where the fill contains timber, paper, domestic refuse, ash or other degradable material, it continues to settle for decades as the material breaks down, irrespective of load. Self-weight collapse settlement of poorly compacted granular fill on first wetting belongs in the same category.

Voids and obstructions. Uncontrolled fill contains voids, old foundations, slabs, basements, tanks and services, which is both a settlement risk and a buildability problem for piling.

The practical consequence is that engineered fill placed and compacted to a specification, with records, is a usable founding stratum, and uncontrolled fill of similar material usually is not. The difference is documentation, not soil mechanics.

Investigation that actually characterises it

The design of the investigation matters more than the number of holes, and the common error is to specify boreholes on a grid.

Trial pits are better than boreholes for characterising made ground, because the engineer can see the material, its structure, its inclusions and the base of the fill directly, and can log the contact with natural ground rather than inferring it from a sample. Pits are limited in depth, so they work alongside rather than instead of deeper holes.

Dynamic probing or cone penetration testing between the pits and boreholes gives a low-cost, dense profile of fill thickness across the site, which is exactly the variability information the foundation design needs. On a site where the issue is distribution rather than properties, a close grid of probes with a smaller number of logged pits and boreholes is a better investment than a sparse grid of deep boreholes.

Design the whole thing to the site investigation standard and plan the geotechnical design to the Eurocode framework, which expects a ground investigation report and a geotechnical design report, with the ground model stated explicitly. A ground model is the deliverable to insist on: a statement of what is where, with the uncertainty named.

Two further points. Specify enough holes to cover the building footprints rather than the site boundary, because a probe in the middle of the car park tells you little about the corner of the block. And where an existing building is to be retained or extended, inspect and record the existing foundations, because a Structural Survey that does not establish founding depth leaves the extension design unresolved.

Foundation options and their cost consequences

Four routes, with different cost profiles and different planning implications.

Deepen the foundations to natural ground. The simplest answer where the fill is shallow and reasonably consistent, using trench fill to the competent stratum. Cost rises steeply with depth, and beyond about two and a half metres the excavation support and the volume of concrete usually make it uncompetitive against piling.

Excavate and replace with engineered fill. Works where the fill is shallow and there is somewhere to put the arisings or a use for them on site. The waste classification of the excavated material becomes a cost driver in its own right.

Ground improvement. Vibro stone columns, dynamic compaction or similar treatment, designed and verified by a specialist, which can allow conventional footings and ground-bearing slabs. Attractive on cost where it works, and it needs the fill to be treatable, which degradable material often is not. Verification testing is part of the price.

Piling. The default for deep or variable fill, with the slab suspended on the piles rather than bearing on the ground. More expensive per square metre and far more predictable, which on a funded scheme is often worth more than the saving.

The planning-relevant consequences are worth noting at design stage. Piling generates noise and vibration that will attract a condition requiring a piling method statement. Deep excavation near trees conflicts with root protection areas. Raising levels to avoid excavation affects drainage, access gradients and the relationship to neighbouring land. A Geotechnical Assessment is the report type that addresses this, and councils usually expect a suitably qualified author for it with professional review strongly advised.

Overlap with contamination sampling

Made ground is both a geotechnical and a contamination issue, and the investigation should serve both from the same holes.

The practical integration points: schedule chemical analysis on samples from the same pits and boreholes, install gas monitoring wells during the geotechnical work rather than returning for a second visit, and specify the analytical suite to cover both the risk assessment for the end use and the waste classification for any material leaving site.

Two specific hazards recur in made ground and both affect cost and programme. Asbestos is common in demolition-derived fill, including as fragments and in cement-bound products, and its presence changes how the material is handled, who may work in it and where it can go. And ground gas generated by degradable fill drives the gas protection design, which needs a monitoring period long enough to capture falling atmospheric pressure, which is months rather than weeks.

Start the gas monitoring at the first investigation. A scheme held up waiting for a monitoring programme that could have begun six months earlier is the most avoidable delay in this field.

What the planning application should state

Be straightforward about ground conditions in the application. Officers do not expect a full geotechnical design at planning stage, and they do expect to see that the issue has been identified.

The planning submission should state the ground conditions anticipated and the evidence for them, the investigation carried out so far, the foundation approach assumed, and any consequence of that approach that shows on the drawings: finished levels, retaining structures, the relationship to trees and to neighbouring foundations, and any piling implied.

Where the foundation solution is still open, say which options remain and confirm that none of them would require a change to the approved layout or levels. That sentence is what allows the officer to condition the detail rather than ask for it before determination.

What not to do is submit a scheme whose levels and layout only work if the ground turns out to be better than the desk study suggests. That is the version that comes back as an application to vary conditions during construction.

Conditions and staged investigation

Expect a staged condition structure, and read the wording before accepting it.

The usual sequence is a condition requiring a site investigation and risk assessment to be submitted and approved, then a remediation or foundation strategy where the investigation identifies a need, then a verification report on completion. Where gas protection is required, the verification condition normally includes integrity testing of the membrane.

Separately, a piling method statement condition is common where piling is proposed near existing buildings or over a sensitive aquifer, and where groundwater is a concern the Environment Agency may ask for a piling risk assessment as a consultee.

Three drafting points to negotiate.

Tie the trigger to the right stage. A condition requiring approval before any development begins will stop you doing the investigation itself, which requires access and sometimes demolition. Ask for the trigger to exclude investigation, demolition and site clearance.

Keep the verification requirement proportionate to what was found, rather than accepting a standard remediation and verification package on a site where the investigation shows no significant risk.

And where the foundation solution may evolve, ask for the condition to approve a strategy and a performance requirement rather than a specific foundation type, so a change from piling to ground improvement does not need a separate application.

The report index sets out the ground-related report types, and the adjacent structural and basement issues are covered in existing foundations and basement depth and settlement monitoring conditions on a basement permission.

This is general guidance, not legal advice; check the position with your local planning authority or a qualified professional before relying on it. For specialist technical reports, qualified professional input may be required. PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval.