Self-healing concrete and lessons from Rome

Self-Healing Concrete, Roman Recipes, and Infrastructure That Outlives Its Builders

Most modern concrete is treated as a consumable: a bridge or a seawall is designed for a few decades, then patched, then torn down and poured again. This page argues that the most underrated path to abundance in construction is durability itself. Concrete that seals its own cracks, as some ancient Roman concrete appears to have done for two thousand years, turns infrastructure from a recurring bill into an inheritance. The page serves the futurism and abundance goals of this wiki by asking a concrete research question: how much of the scarcity people feel in public budgets, housing, and maintenance comes from materials that wear out too soon, and what would change if the materials repaired themselves?

The puzzle of the white lumps

Rome built harbors, aqueducts, and domes that still stand. The Pantheon, dedicated in 128 CE, still has the largest unreinforced concrete dome in the world, and some Roman aqueducts still deliver water to the city today.[1] Many modern concrete structures, by contrast, begin to crumble within a few decades.

For a long time the usual explanation was the volcanic ash from the area around Pozzuoli on the Bay of Naples, which Romans shipped across the empire for building. Researchers also kept finding small, bright white mineral lumps called lime clasts in Roman samples. These were long dismissed as signs of sloppy mixing or poor raw materials.[1]

A 2023 study in Science Advances by a team from MIT, Harvard, and laboratories in Italy and Switzerland proposed a different reading. The Romans appear to have used quicklime, the more reactive form of lime, in a process called hot mixing. The heat of that reaction left behind brittle, reactive lime clasts. When a tiny crack forms, it tends to run through a clast. Water entering the crack dissolves calcium from the clast, and that calcium recrystallizes as calcium carbonate and fills the gap.[1]

The team tested the idea directly. They made hot-mixed samples, cracked them on purpose, and ran water through the cracks. Within two weeks the cracks had healed and water stopped flowing. A matching sample made without quicklime never healed, and water kept running through it.[1]

Evidence from a building site frozen in 79 CE

A laboratory test shows a mechanism is possible. It does not prove that Roman builders actually worked that way. In 2025, researchers published an analysis in Nature Communications of an unfinished construction site in Pompeii, buried in the eruption of Mount Vesuvius in 79 CE. The site held finished walls, walls still under construction, and piles of raw material waiting to be used. The dry piles contained intact quicklime fragments already mixed with volcanic ash before any water was added, which is the first step of hot mixing.[2][3] The authors describe their findings as a scientific basis for developing more durable concretes and restoration materials inspired by ancient practice.

Concrete that grows stronger in the sea

Roman harbor concrete adds a second surprise. Pliny the Elder wrote in the first century CE that volcanic ash, once submerged in the sea, becomes "a single stone mass, impregnable to the waves and every day stronger." A 2017 study in American Mineralogist examined drill cores from Roman breakwaters and piers and found that seawater percolating through the concrete dissolved parts of the volcanic ash and allowed new minerals, aluminous tobermorite and phillipsite, to grow at ordinary seawater temperatures long after the concrete first hardened.[4] The platy crystals interlock and increase resistance to brittle fracture.[5]

Why modern concrete fails sooner

The comparison is not as simple as "Romans were better." Modern concrete is far stronger in compression and cures much faster, and slow hydration like that of Roman mortar is not considered an advantage in modern structural concrete.[6] Roman builders also did not use steel reinforcing bars, so they relied on massive walls, arches, and domes that keep concrete mostly in compression.

Steel reinforcement is what lets modern builders make thin beams, tall towers, and long spans. Steel also rusts. Rust takes up more space than the steel it replaces, which pushes the surrounding concrete apart, causing cracks and spalling. Corrosion of embedded steel is the most common form of concrete deterioration.[7] Every crack that lets water and salt reach the steel speeds the cycle. That is why self-healing matters so much: a crack sealed early may never become a path to the steel.

The scarcity hidden in maintenance

The cost of short-lived materials shows up in public budgets. The 2025 Infrastructure Report Card from the American Society of Civil Engineers counts 623,218 bridges in the United States with an average age of about 47 years. Many are approaching or past the 50-year life they were designed for. About 6.8 percent are rated poor and 49.1 percent fair, and bridge rehabilitation needs are estimated at 191 billion dollars. The report notes that bridges in fair condition can be preserved at lower cost than bridges that have slipped to poor.[8]

Seen through an abundance lens, a maintenance backlog is a form of scarcity that a society creates for itself. Money and skilled labor spent rebuilding the same structure every few decades are not available for new housing, transit, or research. A lead researcher on the Roman concrete work put the goal modestly: concrete does not need to last as long as Roman structures to matter, and adding 50 or 100 years to its lifespan would mean less demolition, less maintenance, and less material over time.[1] Material is also a climate question, since cement production accounts for roughly 8 percent of global greenhouse gas emissions.[1][6] Pouring less often is one of the simplest ways to emit less.

Other routes to self-repair

Roman-inspired hot mixing is one approach. Researchers at Delft University of Technology have spent over a decade developing bacteria-based self-healing concrete. Dormant spores of bacteria that tolerate the high alkalinity of concrete are added with a lactate-based nutrient. When water enters a crack, the spores become active and produce limestone that seals it. In laboratory tests, cracks up to 0.8 millimeters wide sealed at a dosage of 15 kilograms of healing agent per cubic meter of concrete, and cracks up to 0.4 millimeters at 5 kilograms.[9] Early work found that spores added without protection stayed viable for only two to four months as the cement pores closed, which pushed researchers toward protected capsules.[10]

Another route borrows the ash itself. Natural pozzolans such as finely ground volcanic glass can replace part of the Portland cement in a mix. One account describes structural concretes in northern California replacing up to 35 percent of the cement by weight with volcanic glass while still exceeding 27 megapascals of strength after 28 days.[6]

Honest limits

Self-healing concrete is not a cure for every kind of damage. The mechanisms described here seal narrow cracks in the presence of water; they do not repair a collapsed beam, a failed foundation, or damage from overloading. Laboratory healing over two weeks is not the same as proven performance over a century in a real bridge deck exposed to road salt and freezing. Roman structures that survive are also the ones that happened to survive; many Roman buildings did not. Healing agents add cost and must pass building codes, which change slowly for good reason. Finally, longer-lasting materials do not fix weak maintenance planning or underfunded public works on their own.

Research questions worth pursuing

  • How much would the lifetime cost of a typical bridge fall if its deck needed major repair every 60 years instead of every 30?
  • Can hot-mixed or bacteria-based concrete protect steel reinforcement in practice, or only seal cracks in unreinforced sections?
  • Which regions have local volcanic ash or calcined clay that could replace part of their cement and reduce import costs?
  • How would self-healing concrete change the design of structures in places that are hard to maintain, such as offshore platforms, remote towns, or future settlements on the Moon?
  • Explain why a material that lasts twice as long can be a bigger gain for abundance than a material that costs half as much.
  • Describe how the lime clasts in Roman concrete were misread as a mistake for so long, and what that suggests about studying ancient technology with fresh eyes.
  • Compare hot-mixed Roman-style concrete with bacteria-based self-healing concrete in terms of how each one seals a crack and where each one might fit best.
  • Argue whether public agencies should pay more up front for longer-lasting materials when the savings arrive after current officials have left office.
  • Trace how rust in steel reinforcement turns a small crack into a structural problem, and why sealing cracks early changes that chain.
  • Imagine a town that builds its main bridge to last 300 years, and describe what that town could do with the money and labor it no longer spends on replacement.
  • Evaluate what self-healing materials could mean for building habitats on the Moon or Mars, where repair crews and supplies are scarce.

See also

References

  1. ↑ 1.0 1.1 1.2 1.3 1.4 1.5 MIT News, "Riddle solved: Why was Roman concrete so durable?", January 6, 2023. https://news.mit.edu/2023/roman-concrete-durability-lime-casts-0106
  2. ↑ Vaserman, E., et al., "An unfinished Pompeian construction site reveals ancient Roman building technology", Nature Communications, December 9, 2025. https://www.nature.com/articles/s41467-025-66634-7
  3. ↑ MIT News, "Pompeii offers insights into ancient Roman building technology", December 9, 2025. https://news.mit.edu/2025/pompeii-offers-insights-ancient-roman-building-technology-1209
  4. ↑ Jackson, M. D., et al., "Phillipsite and Al-tobermorite mineral cements produced through low-temperature water-rock reactions in Roman marine concrete", American Mineralogist 102(7), 2017. https://doi.org/10.2138/am-2017-5993ccby
  5. ↑ University of Utah, "How seawater strengthens ancient Roman concrete". https://unews.utah.edu/articles/roman-concrete/
  6. ↑ 6.0 6.1 6.2 Jackson, M. D., et al., "Extreme durability in ancient Roman concretes" 2018. https://www.ancientportsantiques.com/wp-content/uploads/Documents/ETUDESarchivees/Ma%C3%A7onnerie/B%C3%A9tonRomain/RomanConcrete-Jackson2018.pdf
  7. ↑ Practical Engineering, "Was Roman Concrete Better?", March 9, 2019. https://practical.engineering/blog/2019/3/9/was-roman-concrete-better
  8. ↑ American Society of Civil Engineers, "Bridges", 2025 Infrastructure Report Card. https://infrastructurereportcard.org/cat-item/bridges-infrastructure/
  9. ↑ Mors, R. and Jonkers, H. M., "Bacteria-based self-healing concrete: evaluation of full scale demonstrator projects", RILEM Technical Letters 4, 2019. https://doi.org/10.21809/rilemtechlett.2019.93
  10. ↑ Wang, J. Y., et al., "Bacteria-based self-healing concrete: An introduction", TU Delft repository. http://resolver.tudelft.nl/uuid:63f20865-ed50-4613-a181-ca7af2f01aab