Concrete cracks almost always — it is a material that is brittle in tension, and capillary cracks under 0.3-0.4 mm are generally considered acceptable by the Eurocodes themselves, because they don't compromise structural function in the short term. The problem is what happens afterwards: a crack, even a small one, is a privileged entry route for water, carbon dioxide and chlorides, the same agents that trigger reinforcement corrosion and the long-term degradation described in the chapter on concrete. Self-healing concrete exists to answer exactly this problem, without requiring human maintenance, by triggering an autonomous chemical reaction the moment the crack opens.
The starting problem: which organism survives inside concrete
The central technical challenge is not so much finding a bacterium able to produce calcium carbonate — there are many, and bacterial calcite precipitation has been known in geomicrobiology for decades — but finding one able to survive, in a dormant state, inside a hostile environment like fresh and hardened concrete: a strongly alkaline pH between 12 and 13, an almost total absence of free water once curing is complete, and the mechanical shear stresses of mixing the batch, which would kill most microorganisms. The research group led by Henk Jonkers at TU Delft, in the Netherlands, solved the problem from the mid-2000s onward by selecting spore-forming alkaliphilic bacterial strains of the genus Bacillus — in particular Bacillus pseudofirmus and Bacillus cohnii — organisms naturally found in extreme environments (alkaline lakes, saline soils) and able to form endospores, a metabolically inert resistance form that can stay viable for decades, and by some estimates even centuries, waiting for favourable conditions.
Encapsulation: protecting the bacterium from the pour
Even a spore-forming alkaliphilic bacterium would not survive direct mixing into the fresh batch without protection: the mixer's shear forces and direct contact with the cementitious environment during setting would still be lethal for a significant part of the bacterial population. The solution adopted by Jonkers and developed commercially in the product Basilisk consists of encapsulating the spores, together with a nutrient reserve, inside porous granules of expanded clay (LECA), typically 2-4 mm in diameter. The expanded clay serves a double function: it physically protects the spores during mixing and vibration of the pour, and it acts as an internal water reserve that helps the concrete's own maturation in its early stages (a principle similar to internal curing, already known independently of the biological context).
The chosen nutrient is typically calcium lactate, a molecule the bacteria can metabolise aerobically when oxygen is available — a condition that arises precisely when a crack opens, putting the material's interior in contact with the outside air. It's an elegant design detail: the system is meant to stay completely inert as long as the concrete is intact, and to activate only at the point and moment the damage it must repair appears.
The reaction: from dormant spore to calcite
When a crack passes through a granule of expanded clay containing spores and nutrient, rainwater or groundwater seeping into the crack reaches the granule and triggers the germination of the dormant spores. The bacteria, now metabolically active, consume the available calcium lactate in the presence of the oxygen that has just entered through the crack, and produce calcium carbonate (CaCO₃) as a metabolic by-product, in the form of crystalline calcite, which precipitates exactly at the point of reaction — that is, inside and around the crack itself. The process requires a constant presence of moisture to proceed: under laboratory experimental conditions documented by the Delft group, cracks up to about 0.8 mm wide closed completely within roughly three weeks of intermittent exposure to water, a timeframe compatible with natural rainfall cycles in a temperate climate.
It's important to be precise on one point: the material does not "heal" in the sense a biological tissue heals, reconstituting a structure mechanically identical to the original. The calcite produced seals the crack, preventing further ingress of water and aggressive agents, and partially restores — not necessarily 100% — the local load-bearing capacity. The main documented and measurable benefit is the drastic reduction in the crack's permeability — the real mechanism through which most long-term degradation of reinforced concrete originates.
The non-biological alternative: microcapsules
In parallel with the bacterial route, other research groups — including those at the University of Bath and Cardiff University in the UK, as part of the multi-university Materials for Life project — have developed self-healing systems based on non-biological microcapsules, containing synthetic sealing agents such as polyurethane resins or sodium silicate. The principle is purely mechanical: when a crack passes through a capsule, it ruptures under the propagating fracture and releases its liquid contents, which react with the surrounding environment (air, moisture, free lime in the cement) by solidifying and sealing the crack. Compared with the bacterial approach, the microcapsule system is faster — the reaction is almost instantaneous, it doesn't take weeks — but it is also single-use: once broken, the capsule can no longer act on a later crack at the same point, whereas dormant bacterial spores can, in theory, remain available for repeated repairs over time, until the nutrient reserve in the granule runs out.
Where it's used today, and why not everywhere
The added cost of bacterial encapsulation or microcapsules — estimated at a significant premium over ordinary concrete per cubic metre — for now limits its use to contexts where lifecycle maintenance or replacement costs are particularly high: underground infrastructure (galleries, tunnels), hydraulic and marine works, industrial floor slabs prone to shrinkage cracking, and a handful of pilot projects on cycle paths and small public works in the Netherlands, used explicitly as real-world test beds, not just laboratory ones. It's the same reason, discussed in the regulation piece in this section, that the absence of a harmonised standard quantifying self-healing efficiency slows adoption more than the material's scientific maturity does: without a recognised reference metric, every client has to judge case by case whether the upfront premium is justified by future maintenance savings.