For an ordinary beam or slab, strength is the headline check. For a water-retaining structure - a tank, a reservoir, a treatment basin - strength is rarely the issue. The wall is usually thick and lightly stressed, and it passes the ultimate-limit-state check with room to spare. What decides whether the structure succeeds or fails is something far less glamorous: crack width, and the limit on it is severe.
Why leakage, not collapse, governs
A water tank does not need to be strong so much as tight. A crack that would be cosmetically irrelevant in a building becomes a leakage path here, and a leaking tank fails its basic purpose long before it is anywhere near structural distress. So the governing design criterion shifts from stress to serviceability crack width, and the whole reinforcement design is driven by keeping cracks fine enough to stay watertight.
The limit
EC2 Part 3 (the water-retaining code) ties the allowable crack width to the hydraulic gradient - essentially the ratio of water depth to wall thickness, a measure of how hard the water is being pushed through any crack. For many tanks that means a limit of around 0.2 mm or tighter, well below the roughly 0.3 mm tolerated in ordinary structures.
Self-healing helps, but only within limits: very fine cracks seal themselves autogenously under sustained hydrostatic pressure, as calcium carbonate and continued hydration block the path. Wide cracks do not - which is exactly why the limit is set where it is.
That self-healing mechanism is the physical reason the limit is a width, not zero. Below a certain width the crack closes itself; above it, the leak persists and worsens.
What actually controls the crack width
The surprise for many engineers is that the governing cracks are usually not from the water load at all - they are from early-age thermal effects and restraint. As a freshly cast wall hydrates it heats up, then cools and tries to shrink; the older base slab restrains it, tension builds, and once it exceeds the concrete's tensile strength the wall cracks at roughly regular spacing before it has ever held water.
Reinforcement does not prevent this cracking. Its job is to control crack width - to force many fine, self-sealing cracks instead of a few wide, leaking ones. The steel is sized for distribution, and the spacing of cracks follows from the bond and the reinforcement ratio.
Designing for it
Three things matter more here than in ordinary concrete:
- Pour sequencing and size are part of the structural design, not a site afterthought. Smaller pours and well-placed movement or construction joints reduce restraint and the thermal cracking it drives.
- Curing and temperature control limit the peak hydration temperature and the thermal gradient, attacking the cracking at its source.
- Distribution steel governs, often more than the flexural steel. The reinforcement that keeps cracks fine is the reinforcement that keeps the tank tight.
The related shrinkage cracking article goes deeper into the restraint mechanism, and the EC2 crack-width provisions are in the standards reference. The headline to carry away: a water tank lives or dies by a 0.2 mm limit, and that limit is met with pour planning and distribution steel, not with extra strength.