Roman Concrete Was Built to Last. Modern Wasn't.

The Pantheon has been standing since the reign of Emperor Hadrian, sometime in the early second century CE. It has survived the fall of Rome, the medieval period, the Renaissance, two world wars, and roughly six million tourists per year. Its dome — 43 meters across, open to the sky through a nine-meter oculus, containing not a single piece of rebar — remains the largest unreinforced concrete dome in the world.

The road outside your house probably needs repaving every decade.

That contrast, blunt and a little embarrassing, is the starting point for a body of scientific research that has been quietly accelerating over the past twenty years. What did the Romans know about concrete that we lost, or never bothered to learn? The answer, it turns out, is not one thing but several — and they compound each other in ways that modern materials science is still working to fully describe.

Not Stronger. Tougher.

The first thing to get straight is what Roman concrete is not. In terms of raw compressive strength, it is not better than what we pour today. Modern ordinary Portland cement, or OPC, is roughly ten times stronger under direct pressure than its Roman equivalent. If you dropped a column of OPC-based concrete and a column of Roman concrete from the same height, the Roman one would fare worse.

So how are Roman harbor piers, submerged in the Mediterranean for two thousand years, not just intact but measurably more structurally sound than when they were built?

The distinction is between strength and resilience — between how much force a material can resist before it breaks, and how well it holds together over time in hostile conditions. Modern concrete is strong but brittle and chemically vulnerable. Roman concrete is weaker but, under real-world conditions, vastly more durable. It does not just resist degradation. In certain environments, it reverses it.

The Seawater Problem — and Its Solution

Seawater is ordinarily lethal to concrete. The chlorides and sulfates it carries attack the cement matrix and cause it to swell, crack, and eventually disintegrate. Most modern marine concrete structures deteriorate within decades. This is not an edge case — it is a central engineering challenge with enormous infrastructure costs.

The harbor piers at Caesarea Maritima, built by Herod the Great on the coast of what is now Israel, have been underwater for approximately two millennia. In 2013, a research team led by Paulo Monteiro, a professor of civil and environmental engineering at UC Berkeley, analyzed concrete samples from those piers and found something unexpected: crystals of a mineral called tobermorite, a calcium silicate hydrate that does not appear in Roman concrete's original recipe. It had formed in place, grown within the concrete matrix as seawater percolated through and reacted with phillipsite, a mineral in the volcanic ash used as a key ingredient.

Tobermorite is harder than the concrete's original constituents. More importantly, it grows in ways that trap the very molecules — chlorides and sulfates — that would ordinarily cause damage. The seawater was not destroying the concrete. It was, in a slow and extraordinary chemical process, improving it.

That volcanic ash — pozzolana, a fine glassy material found mainly around the Bay of Naples — was the ingredient the Romans considered so essential that it was traded across the Mediterranean. Research by scholars including Brandon, Hohlfelder, and Jackson has documented how extensively this material moved through Roman supply networks, reaching as far as Alexandria, Egypt. The chemistry it enabled, in combination with quicklime, produced a cement that could set underwater without the need for dry cofferdams — a construction capability that shaped the entire Roman approach to harbor engineering.

The Lime Clasts Problem — Which Wasn't a Problem

For decades, archaeologists noticed something odd in Roman concrete samples: scattered white flecks, brittle calcium-rich inclusions called lime clasts, distributed throughout the matrix. The standard interpretation was that this represented poor quality control — inadequately mixed raw materials, the kind of sloppiness you might expect at the edges of a vast empire.

That interpretation never sat well with researchers who looked closely at the documentary record. Vitruvius, writing in 25 BCE, laid out precise concrete recipes in his De Architectura. Emperor Augustus, who reigned from 27 BCE to 14 CE, oversaw the systematic standardization of Roman construction practice. The idea that a civilization that meticulous would have casually tolerated manufacturing defects in its signature building material seemed, to put it plainly, unlikely.

In 2022, a team led by Admir Masic, a professor of civil and environmental engineering at MIT, conducted spectroscopic analysis of concrete samples from the mausoleum of Caecilia Metella, built between 30 and 10 BCE. The analysis offered a different explanation for the lime clasts entirely.

When Roman concrete shifts — due to ground settling, seismic activity, thermal expansion, or any of the other forces that work on a structure over centuries — cracks form. Specifically, they form preferentially through the lime clasts, which are more brittle than the surrounding cement matrix. So far, this sounds like a weakness. But here is the mechanism that changes everything: when water percolates through those cracks, it reacts with the lime clasts and promotes the crystallization of calcium carbonate, which gradually fills and reinforces the crack from within. The concrete, in effect, heals itself.

To test this, Masic's team cracked samples of both Roman and modern concrete and ran water through them. Within two weeks, the cracks in the Roman samples had sealed. The cracks in the modern samples had not.

The lime clasts were not a defect. They were a design feature — one baked into the material by the specific method Romans used to make it.

Hot Mixing and the Dry Layup

Modern concrete production pre-slakes the quicklime before it ever touches the mix — quicklime reacts with water first to produce calcium hydroxide, which is then added to the aggregate. The Romans did something different: they added quicklime directly to the mixture in a process now called hot mixing. The reaction generates significant heat. That heat is what produces the lime clasts in the first place.

The layup method compounded the effect. Roman concrete was not poured wet, the way modern concrete is. According to Vitruvius, it was mixed as dry as possible, troweled into place in layers, with large aggregate — fist-sized chunks of volcanic tephra, brick fragments, recycled building stone — laid in by hand between courses. The whole assembly was then compacted with specialized tools. Less water meant fewer voids; fewer voids meant a denser matrix with less room for the microscopic weaknesses that eventually become cracks.

The result was a material built slowly, deliberately, and with an understanding — even if empirically derived rather than scientifically explained — of how chemistry, time, and environment would interact within it.

The Uncomfortable Question for Modern Practice

Concrete is not a marginal material. Global production runs to approximately 19 billion tons per year, and the manufacture of Portland cement clinker — which requires kiln temperatures around 1,450 degrees Celsius — accounts for roughly 8 percent of global carbon dioxide emissions. That is an outsized environmental footprint for something that, by Roman standards, does not last very long.

The Roman system, by contrast, fired its ingredients at substantially lower temperatures, produced structures that needed far less frequent replacement, and relied on a volcanic ash that, while not universally available, has functional analogues — including fly ash, a coal combustion byproduct that is widely available as industrial waste. According to Wikipedia's overview of Roman concrete research, fly ash has shown promise as a pozzolana substitute in modern formulations, potentially reducing both costs and the carbon footprint of production.

None of this means modern engineers should simply adopt Roman methods wholesale. Roman concrete's lower compressive strength makes it unsuitable for many structural applications where OPC is now standard. The timescales involved in Roman-style chemical strengthening are not compatible with projects that need to be load-bearing within days rather than decades. There are real reasons the industry moved where it moved.

But the research coming out of Berkeley, MIT, and institutions across Europe and the Middle East suggests that the Roman approach encodes solutions to problems that are becoming more urgent — not less — as global infrastructure ages, coastal structures contend with increasingly corrosive conditions, and the construction industry faces pressure to reduce its environmental impact.

The Pantheon has been standing for nearly two thousand years. It was not built with the best material available. It was built with the most carefully considered one — mixed hot, laid dry, composed of ingredients that were still reacting and improving long after the builders were gone.

That is a different engineering philosophy than the one that paved your street. Whether the industry finds a way to close that gap is one of the more consequential material science questions of the next few decades.

---

By Margaret "Maggie" Holloway, History & Ideas Correspondent