Holding Water In Versus Keeping Water Out

What each structure is trying to avoid says most of it. A basement wall exists to keep water out of an occupied space, and its failure shows as damp inside. A tank wall exists to keep water in, and its failure shows as water going missing. Those are different symptoms, different investigations and different remedies.

It also changes who notices. Damp inside a room is reported quickly by whoever uses it; a falling level in a tank can be read as evaporation for a long time, and a damp patch in the ground outside may never be connected to the structure at all.

A basement structure is the whole below-ground enclosure treated as one object: perimeter walls holding back ground and groundwater, the base slab they stand on, and the floor plates spanning between them. Drawn on separate sheets those look like ordinary walls and slabs; built, they behave as a closed box whose faces restrain each other.

Ground pushes on a perimeter wall, the wall leans on the base slab at its foot and on the floor plate near its head, and the plate carries that push across the building into the opposite wall. Remove the plate and the wall has to find restraint elsewhere, usually by bending far more or by being propped.

In a water-retaining structure the structure and the containment are the same element. There are no floor plates to prop the walls, and both load conditions are designed for at once because the structure moves between them in service rather than during construction.

That is also why flotation appears in one set of questions and not the other. A base that the contained water was holding down is trying to lift once that water is gone and the ground around the shell is wet, which makes what the design assumed about groundwater a live question rather than a background one.

Joints matter in a basement because positions that relieve the structural engineer of shrinkage may be exactly where the water-resisting designer least wants an interruption, and joints chosen independently produce a box that is structurally reasonable and hydraulically awkward.

In a tank shell joints are closer to the whole point of the exercise. The shell is an element that must not leak, so every place a pour was interrupted and every place movement was deliberately allowed is a planned discontinuity through the containing material, and waterstops, reinforcement continuity and the wet-face finish are all set out from those positions.

A basement asks what water regime it has been designed against, because water reaches it as groundwater under pressure, as water perched in backfill and as vapour from the ground and from use of the space. Those are different problems and a measure aimed at one may not address another.

A tank asks the same question from the other side, and adds one of its own: what the flotation check is based on and what it assumes about the ground around the structure. Both questions are answered from ground investigation rather than from the appearance of the site.

The practical route is to name the structure before designing anything. If people will occupy the space and water is to be excluded, the basement questions apply: which plates prop which walls, where the joints fall, and which approach to water resistance is being relied on. If water is to be contained, the tank questions apply: what the empty case assumes, who may drain it, and where the joints and penetrations sit.

Structures that do both, such as a pool within a basement, inherit both sets rather than averaging them, which is a reason to have that combination recognised explicitly at the outset.