Roman Concrete: The Lost Technology
Roman concrete has been underwater for two thousand years in some locations and is stronger now than when it was poured. This is not a figure of speech. The concrete used in Roman harbor structures — the piers, breakwaters, and seawalls built along the Mediterranean coast during the Republic and Empire — has been studied by geologists and materials scientists who have found that it has been gaining strength over time rather than degrading, a property that modern Portland cement concrete does not share. Understanding why this happens has become one of the more productive intersections of archaeology, geology, and materials science in recent decades, and the answer reveals something important about Roman empirical knowledge and its limits.
The key ingredient was volcanic ash — specifically, pozzolana, the fine ash produced by volcanic activity in the region around Pozzuoli (ancient Puteoli) on the Bay of Naples. Mixed with lime and seawater, pozzolana produces a hydraulic cement — one that hardens in water rather than requiring air to cure — through a chemical reaction that has been studied intensively since samples from ancient Roman harbor structures were analyzed in the 2010s. The reaction between seawater, lime, and the volcanic ash produces crystals of aluminous tobermorite that grow into the spaces between the aggregate over time, progressively strengthening the material. Two thousand years of mineral growth have left some Roman harbor concrete structures substantially harder than when they were constructed.
The Romans did not understand the chemistry. They did not need to. What they had was empirical knowledge accumulated over generations of construction: this material, from this region, mixed in these proportions, with this aggregate, produces a material that hardens in water and proves durable over time. The knowledge was practical and effective without requiring the theoretical understanding that would explain why it worked. Roman engineering throughout its history operated on this basis: systematic observation and accumulated practice rather than theoretical principles derived from first scientific premises. This is not a failure of intelligence; it is a description of how technical knowledge actually develops in most cultures before the scientific revolution.
The volcanic ash was not universally available — it was specific to certain Italian geological regions — and Roman builders working in areas without access to pozzolana used other approaches. In Britain, where neither volcanic ash nor the warm Mediterranean climate was available, Roman concrete was made differently and performed differently. The remarkable durability of Italian Roman concrete is partly a function of its specific ingredients and partly of the Mediterranean climate that allowed slow curing without the freeze-thaw cycles that crack concrete in colder regions. The technology that produced the Pantheon’s dome and the harbor structures of Caesarea Maritima was not straightforwardly transferable to every location where the Romans built.
The loss of Roman concrete technology after the Western Empire’s collapse is a genuine case of knowledge loss with significant consequences. Medieval European builders did not have access to hydraulic concrete and could not build the harbor structures, underwater foundations, and massive vaulted spaces that Roman concrete had made possible. The rediscovery of hydraulic cement in the eighteenth century — John Smeaton’s work with natural cements in Britain in the 1750s, leading eventually to Portland cement’s development in the nineteenth century — produced a material with different properties than Roman concrete, notably including the long-term strength degradation that Roman concrete avoided. Modern harbors built from Portland cement degrade over decades in saltwater; Roman harbors built from pozzolanic concrete remain structurally sound after two millennia. The ancient material was, by this specific measure, superior.
The coffers of the Pantheon dome, the arched vaults of the thermae, the massive foundations of the Colosseum — all of these were made possible by a material that could bear loads across spans and in configurations that cut stone masonry could not accommodate without the scaffolding and precise fitting that made complex curves impractical. Roman concrete freed construction from the structural logic of the post and lintel and the simple arch, making possible the complex geometric spaces of Roman public architecture. The dome is the most dramatic example: the Pantheon’s coffered dome is a single pour of concrete — tons of material suspended across a 43-meter span — that has not required structural repair in nineteen hundred years. The material was not perfect; Roman concrete structures have failed, and the physics of concrete aging are complex. But the specific properties of pozzolanic concrete, in the right applications and environments, produced structures of extraordinary durability that modern construction has not reliably equaled.
The scientists studying Roman concrete are not, in the main, trying to revive the ancient technology — modern construction has different requirements and different material options. What they are trying to understand is the mechanism by which a material can gain strength over time rather than losing it, with the hope that the chemistry of Roman harbor concrete can inform the development of more durable and environmentally sustainable modern cements. The ancient building material, investigated with twenty-first-century analytical techniques, may contribute to solving twenty-first-century problems with infrastructure durability and carbon emissions. That the Romans would find this posthumous usefulness thoroughly characteristic of how Roman things tend to work is probably not a coincidence.