What can cause a building to settle and how are the causes investigated?
Homeowners, buyers and builders
A building can settle because its supporting ground compresses under load, collapses when wetted, loses material or contains a void. Those are different processes. Calling all of them 'the house sinking' hides the questions that determine a useful investigation: what is supporting the building, where has it changed, and how does that change fit the damage? These alternatives need investigation rather than assigning a cause from a wall crack.
Compression is not always an immediate event
Loose fill can compress under its own weight or the weight of construction above it. A filled site may also vary in depth, so different parts of a building experience different movement. A level surface does not reveal the original ground profile or how the fill was placed.
Soft saturated ground can respond in three different stages. Immediate deformation can occur under loading before appreciable water drainage. Consolidation is the time-dependent compression as excess water pressure from loading dissipates and water leaves the soil. Secondary compression is continuing deformation after that excess pressure has largely dissipated; it can be important in peat and other highly organic ground. These responses are distinct from the decomposition of organic material in a landfill: compression of the ground and biological breakdown are not the same explanation.
The support distinction is illustrated by reported examples from the Meadowlands, a marshy area in New Jersey. A floor slab may break away from pile caps or deform around them as the surrounding peat settles. The piles and slab are not necessarily following the same movement, even though they belong to the same building.
Water can change support without adding another storey
Some dry, relatively open soil structures lose volume when water enters them, even without a new building load. This is wetting collapse. Other materials may weaken, lose soluble particles or permit soil to migrate into underlying openings. Reactive clay can instead shrink or swell as its moisture changes. The direction of movement therefore cannot be decided from 'the ground became wet'. The mechanism and actual material matter.
A leak, irrigation change or altered drainage can be relevant, but timing alone does not establish the sequence. Ground movement may damage a service, which then introduces more water. The investigation needs evidence of the pipe condition, water route, ground response and building movement, not just a leak report beside a crack. Wetting collapse and other settlement mechanisms need to be distinguished.
The history may begin before the house
Old land use can leave a support problem that the current owner cannot see. One reported case involved a motel constructed unknowingly over spoil used to fill a former strip-mining operation. Movement began during construction and the motel was never put into service. This reported case distinguishes settlement of deposited spoil from collapse of an underground mine opening; both require more than a current photograph to investigate.
On a different scale, regional extraction of groundwater or oil can change ground conditions over a wide area. These historical examples do not establish that a particular Australian property is affected. They show why a site explanation sometimes needs geological and regional records as well as the individual building's history.
Match the investigation to the proposed process
Useful records include the original ground and fill profile, available soil reports, foundation information, alterations, drainage changes and dated damage observations. A professional may need sampling or testing that addresses a particular question: compressibility, collapse on wetting, the presence of fill or a loss of support. A soil name on an old report is not a substitute for establishing the material and conditions relevant to the damaged area.
The building response also needs checking. Uniform tilt, local bending and displacement between separately supported parts can produce different damage. One floor-level survey records the shape found on that date; it does not, by itself, supply the original level or a movement rate.
Before commissioning more work, ask which settlement mechanisms remain plausible, what supports each one and what the next test would distinguish. Keep observation and inference separate in your own file. Sudden loss of support, apparent instability or rapid change needs urgent professional attention rather than a longer monitoring record.