Building on Adelaide's reactive clay: what it demands of design and construction.

Much of Adelaide and the surrounding plains sit on expansive — reactive — clay. These soils take on water and swell through the wet months, then dry out and shrink through summer. The ground does not stay still under a house; it lifts and settles on an annual cycle, and the drier the season, the deeper that movement reaches.

The movement is rarely uniform. A slab edge under a wide eave stays drier than one exposed to rain. A garden bed, a downpipe or a leaking service feeds water into the soil at one point and not another. The result is differential movement — one part of the footing rises or falls relative to another — and it is that difference, not the movement itself, that loads the structure.

Left unaccounted for, the consequences are structural and visible. Footings and slabs that are too light for the reactivity flex and crack. The movement carries up into the building as cracked masonry, racked door and window frames, drummy or split tiling, and separation at the junctions where one material meets another. On reactive ground, a footing that ignores the soil is a defect waiting for a dry summer.

Reactive ground is a design input, not a construction afterthought. The site's reactivity governs the footing system, the finished floor levels, where water is taken and how it is kept away from the soil, and how rigid the slab has to be to ride the movement without passing it into the finishes. These are not details to be settled once the foundations are poured — they shape the building from the first sections.

When design and construction are split, that input arrives late. An architect resolves the design; a builder is appointed afterward and confronts the soil at excavation. The footing is then sized against a construction budget rather than the design, levels are adjusted on site to suit it, and material junctions that were drawn to be flush are reopened to absorb movement nobody planned for. The ground forces a renegotiation, and the design loses the argument.

Resolved inside one process, the order reverses. The footing system, the levels, the drainage strategy and the material junctions are decided together, against the actual reactivity of the site, and committed once. The structure is designed to hold the movement and the finishes are detailed to tolerate what remains — so the drawing commits only to what the ground and the construction will carry through.

In Australia, residential footings and slabs are designed to AS 2870, Residential slabs and footings. The standard begins with a site classification — set from a geotechnical investigation of the soil — that grades the ground by how much it is expected to move. The classes run from stable sites through slightly, moderately and highly reactive clay to extremely reactive and problem sites (A, S, M, H1, H2, E and P). That class is what a compliant footing system is sized against.

In an integrated firm, the classification is not a document handed across a gap. It informs the design directly. The footing type, slab stiffness and reinforcement are resolved against the site class while the building is still being drawn, alongside the levels and the drainage that keep moisture away from the soil and limit how far it moves. Tree positions, paving and the path of stormwater and plumbing are treated as part of the same problem, because each one changes the soil moisture the footing has to answer to.

Then the same firm builds it. The team that classified the site and engineered the response is the team on site when the ground is opened up, so what was drawn against the soil is what gets poured into it. There is no second party to argue the footing down against a budget, and no point where the structural intent changes hands. The response to the ground is carried through to built form, and the firm that drew it is accountable for how it stands.