Heave in shrinkable clay after removing a tree

Clay heave after tree removal is the slow problem. A mature tree growing in shrinkable clay keeps a persistent soil moisture deficit beneath and around it, and the clay sits compressed. Take the tree out and that deficit recovers. The clay rehydrates and swells upwards, and on a high plasticity soil with a large tree the movement can continue for a decade or more.

So felling a tree before building does not remove the foundation problem. It replaces subsidence risk, which acts downwards and can be managed by depth, with heave risk, which acts upwards and has to be managed by detailing.

Why removal causes movement in the opposite direction

A tree draws water from the soil through the growing season, and on clay the effect persists through winter rather than fully recovering. Over years, this produces a zone of desiccated clay extending below and around the root system.

Build a conventional strip footing in that desiccated zone and leave the tree in place, and the risk is seasonal shrinkage and subsidence as the deficit deepens in dry summers. Remove the tree and the clay slowly takes up water again, expands, and lifts whatever sits on it.

The two failure modes look quite different on a building. Subsidence tends to produce tapering cracks widening upwards, often with distortion towards the tree. Heave produces lifting at the perimeter, cracking that widens downwards, doors binding at the head, and floor slabs doming. Heave is also harder to remedy, because the pressure is sustained and considerable.

Shrinkable clay classification and plasticity

Volume change potential is classified from the modified plasticity index of the soil, determined by laboratory testing of samples from trial pits or boreholes. The NHBC Standards chapter on building near trees sets out the banding used across the industry, and foundation depth charts are read against it.

Do not take the classification from a geological map alone. The map tells you the formation. The plasticity index tells you what the clay on your plot actually does, and the two diverge more often than people expect, particularly on made ground and on weathered upper horizons. One trial pit with samples tested for Atterberg limits is inexpensive against the cost of getting the foundation wrong.

Influence distances by species and mature height

The industry approach relates the distance from the tree to its mature height, with species grouped by water demand. Oak, poplar, willow, elm and eucalyptus are high demand. Birch, sycamore, lime and most fruit trees sit in the moderate group. Many conifers are lower.

For a given soil classification and water demand, the charts give a required foundation depth at a stated ratio of distance to mature height. Close to a high demand tree on high plasticity clay, depths of two to three metres are routine, and that is before heave precautions are added.

For a removed tree, the geometry changes. What matters is the extent of the desiccated zone left behind, which broadly follows the former root spread, and how much deficit remains at the time of construction. A tree felled ten years ago on medium plasticity clay may have left little residual effect. One felled last winter has left all of it.

The practical consequence is that the date of removal is a design input. Record it, with evidence, because a structural engineer asked to design for heave with no information on timing will assume the worst case.

Foundation depth and heave precautions

Two things are needed: enough depth to found below the zone of seasonal and residual movement, and detailing that lets the clay swell without loading the structure.

• Trench fill or piles taken to a depth determined from the soil classification and the tree data.
• Compressible material against the vertical faces of the foundation, so swelling clay does not grip and lift it.
• A void or compressible layer beneath ground beams and suspended slabs, sized for the predicted heave.
• Slip membranes on piles to reduce uplift adhesion, with piles designed in tension as well as compression.
• A suspended floor rather than a ground-bearing slab where heave is predicted.

Drainage and services need the same thinking. Rigid connections at the point where a service enters a building founded below a heaving layer will fail first, so flexible joints and a rocker pipe arrangement belong in the design.

Eurocode 7 governs the geotechnical design and BRE Digest 298 remains the standard reference on tree influence in clay soils. A Geotechnical Assessment sets out the ground model, and councils expect a suitably qualified author for it, with professional review strongly advised.

Interaction with root protection areas

Retaining a tree brings its own constraint. Under BS 5837:2012 the root protection area for a single stem tree is calculated as a circle with a radius of 12 times the stem diameter, capped at 707 m², and construction within it needs justification and protective measures.

Which produces the familiar squeeze on small plots. The foundation needs depth because of the tree, and the tree needs its root protection area kept clear, and the two together push the building envelope away from where the client wanted it.

Removal is not always within your gift either. A tree covered by a Tree Preservation Order needs the council's consent for felling, and a tree in a conservation area needs six weeks' written notice to the council under section 211 of the Town and Country Planning Act 1990, during which the council can make an Order. Felling a protected tree without consent is an offence.

A Tree Survey to BS 5837 gives the stem diameters, the root protection areas and the condition assessment that the foundation design and the planning application both need. Protection during the works is a separate exercise, covered in temporary works and tree protection on tight sites.

Where retaining the tree is the better answer

Often it is, and on several grounds at once.

Retention avoids the heave problem entirely, leaving only a subsidence design that is handled by depth. It avoids the consent risk where the tree is protected. It usually helps the biodiversity position. And it tends to go down better with a case officer than an application that opens with the removal of a mature oak.

The counter-argument is footprint. If the root protection area occupies the only buildable part of the plot, retention may mean a smaller house.

Run the comparison properly before deciding. Price the retained-tree foundation, with its depth, against the removed-tree foundation with piles, heave precautions, suspended floor and flexible services. On high plasticity clay next to a high water demand tree, the removal option is frequently the more expensive of the two, which surprises self-builders who assumed felling was the low-cost route.

Evidence for the application

Councils look at this through the tree and design lenses rather than the structural one, so the application has to connect them.

Include the BS 5837 survey with the root protection areas plotted on the proposed layout, an arboricultural impact assessment explaining what is being removed and why, and a short statement from the engineer confirming that the foundation solution respects the root protection areas of retained trees. Where removal is proposed, set out the replacement planting.

Where demolition is also involved, the surveys overlap, and our note on demolition and asbestos surveys in the planning submission covers what else is needed. The structural position itself belongs in a Structural Survey, for which councils usually expect a suitably qualified author.

This is general guidance rather than legal advice, and ground conditions are site-specific, so take advice from a chartered structural or geotechnical engineer and a qualified arboriculturist before relying on any of 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.