Self-healing concrete is concrete engineered or inherently able to repair its own cracks once damage occurs, without external intervention from a contractor. The repair mechanism relies either on the material's natural chemistry or on additives built into the mix. In UK design practice, this property is directly tied to crack-width limits: historic British Standards allowed a maximum crack width of 0.2mm for self-healing to be relied upon, while the Eurocode regime now sets far tighter and pressure-dependent limits, down to 0.05mm for the most onerous cases. Understanding how the healing process actually works on site is essential for anyone specifying concrete for water-retaining or underground structures, where leakage through cracks is the defining durability risk.

What causes concrete to crack in the first place?

Cracking is not a sign of material failure by default — reinforced structures are designed to crack so the embedded reinforcing steel can take up tensile load. Problems arise when cracking becomes excessive or uncontrolled. Typical crack categories include:

  • Settlement fractures and plastic shrinkage cracks
  • Shear and flexural cracks from structural loading
  • Splits triggered by reinforcement corrosion
  • Weathering-related cracks and thermal strain cracks
  • Cracks from improper mixing, agitation or curing, or an incorrect water-to-cement ratio

The underlying driver is a loss of durability: as the material's resistance to physical and chemical attack decreases, the likelihood of cracking rises correspondingly.

How does concrete heal itself without additives?

Autogenous (natural) self-healing uses unhydrated cement particles already present in the matrix as the repair agent. When water reaches a fine crack, minerals in the clinker hydrate and the resulting calcium dissolution process seals the gap. This mechanism works reliably only for very minor damage — for anything beyond hairline cracking, autonomous methods with external healing agents are needed.

Which autonomous healing mechanisms are used?

  • Vascular self-healing: a network of hollow, chemically inert ducts embedded in the concrete carries a healing agent. Hydrostatic pressure, capillary action and gravity push the agent through the crack to seal it; clay ducts combined with inorganic phosphate cement are a common pairing.
  • Capsule-based self-healing: spherical or cylindrical microcapsules — made from ceramics, silica, glass, urea-formaldehyde or polystyrene — rupture when a crack passes through them, releasing healing agents such as epoxy resin, polyurethane or methyl methacrylate monomers.
  • Shape memory alloys (SMAs): pre-tensioned SMA wires embedded in a concrete beam are heated with an electrical current; the heat transferred to the surrounding concrete has been shown in testing to close cracks formed under midspan loading.
  • Bacterial (microbial) self-healing: microbial spores precipitate calcium carbonate (CaCO3) within microcracks, patching them from the inside.

What crack-width limits apply under UK and Eurocode design?

Under the older British Standards BS 8110 and BS 8007, self-healing could be assumed for crack widths up to 0.2mm, provided there was no critical aesthetic appearance requirement — a flat limit regardless of the water pressure acting on the structure. CIRIA R91 covered early-age thermal (EAT) design in that era.

Since the shift to Eurocode — with BS 8007 replaced by EC2-3 and CIRIA R91 updated to C660 — the rules became more detailed and considerably stricter. Eurocode 2 Part 3 distinguishes between through cracks, which penetrate the entire section and create a direct water path, and non-through cracks on the tension side of a bending element, which do not compromise water resistance. For through cracks, the allowable design width ranges from roughly 0.2mm down to 0.05mm depending on the pressure acting on the structure: the higher the pressure, the smaller the permitted crack, because high-pressure cracks are harder to self-heal.

At the 0.05mm end of that range, meeting the requirement through reinforcement alone can demand so much steel that the design becomes unbuildable — and overcrowded reinforcement brings its own risk, since fresh concrete can no longer compact properly around it, leaving voids that are worse than the cracks they were meant to prevent. CIRIA's subsequent update, C766 — retitled "Control of cracking caused by restrained deformation" to reflect that early-age thermal effects are only one contributing factor — relaxed the allowable crack width for deep structures from 0.05mm to 0.1mm, provided the water pressure on the structure remains stable. Although the numerical change looks small, it represents a 100% increase in the permitted crack width and can significantly reduce the reinforcement required in deep underground structures.

Practical takeaway for specifiers

When specifying crack-width limits for water-retaining or underground concrete, check which design generation governs the project — BS 8007's flat 0.2mm limit, early Eurocode/C660 practice, or the relaxed C766 approach — since the reinforcement quantity, buildability and self-healing reliance can differ substantially between them. For structures exposed to stable, moderate water pressure, the 0.1mm limit under C766 can avoid the unbuildable steel densities associated with 0.05mm detailing, while still relying on documented self-healing behaviour rather than reinforcement alone.