A Roman solution to Raac’s ruin – is self-healing concrete the answer?

<span>Photograph: Rik Hamilton/Alamy</span>
Photograph: Rik Hamilton/Alamy

Concrete research gets caricatured as the epitome of dull – until the roof falls in. The dangerous state of many British schools built partly from reinforced autoclaved aerated concrete (Raac) has dominated headlines, alarmed parents and embarrassed the government, leading to emergency closures just as the new school term began. The crisis highlights that, however boring concrete might seem, our civilisation almost literally stands or falls on it.

Far from being prosaic, concrete is a hi-tech substance at the forefront of materials research. One dream is to make concrete self-healing: able to repair its own cracks automatically. And modern research is drawing inspiration from an ancient source – the unassailable concrete of the monuments, aqueducts and harbours built by the Romans more than 2,000 years ago. Couple this with ingenious ploys such as entombing live, crack-sealing bacteria inside concrete, and research in this area could transform the way we build.

Scientists are exploring the idea of trapping live bacteria inside the concrete to conduct the repairs

As the Raac saga shows, concrete failure can pose lethal hazards and be hideously expensive to put right. Self-healing concrete would solve that problem without anyone lifting a finger. It could become foundational (see what I did there?) for the vision of sustainable cities – not least because concrete production has a massive carbon footprint, so extending its lifetime would have green dividends too.

Raac was invented about a century ago. But although it has been portrayed in the press as a shoddy, outdated material, that’s unfair, says concrete expert Prof Chris Goodier of Loughborough University: “It has many advantages and has long been a respectable building material.”

It is made by adding aluminium powder to a standard concrete mix, causing chemical reactions that produce bubbles of hydrogen gas as the concrete sets. This gives the material a foamy texture, making it lightweight – just a quarter to a third of the density of ordinary concrete – and a good thermal and sound insulator. Like most concrete used in construction, it is laced with steel reinforcement (usually bars) so that the material can withstand bending or stretching forces, allowing it to be used for load-bearing applications.

The older varieties of Raac installed from the 1950s to the 1970s – and now causing problems – have been steadily improved since then. But even those older panels “will last if looked after well and installed properly”, says Goodier. “Much of it is 50 years old and doing fine.” Raac is still used without apparent problems in many countries, including Germany, the US and Australia.

The real issue is maintenance. Because it is so porous, Raac will let in moisture – and if the water reaches the steel reinforcement, it corrodes. Raac exposed to the elements therefore needs to be waterproofed, for example by coating it with bitumen – especially in panels used for flat roofs on which rainwater can pool. But many panels in schools have been neither replaced nor adequately maintained, largely because of lack of funding.


“We have known about issues with Raac since about 1992,” materials scientist Phil Purnell of the University of Leeds told Chemistry World magazine. He too stresses that we can’t blame the material: the problem is poor building maintenance because of squeezed budgets. After the Raac roof of a school building in Kent collapsed without warning (luckily over a weekend) in 2018, a committee of the Institution of Structural Engineers issued an industry-wide alert. But still no significant action was taken.

Hospitals, where Raac was also widely used, have done somewhat better, though there are still issues. The NHS has funded Goodier’s team at Loughborough to study the signs and mechanisms of failure. “Hospitals are the leaders in the field in making Raac safe,” he says. “They invested millions, and in hindsight this has proved an extremely wise investment.”

A volcanic mix

Our concrete buildings at risk of collapse after just a few decades are put to shame by some of the concrete structures erected by the Romans, which are still standing today. People had been making plasters and cements since the Neolithic period around 14,000 years ago, typically from lime: calcium oxide, made by heating up limestone (calcium carbonate) in a kiln. Adding water to powdered lime creates a paste of “slaked lime” which can be mixed with sand, clay or lumps of rubble or rock (now called aggregate) to make mortars and concrete. As the material dries, slaked lime reacts with carbon dioxide in the air to regenerate calcium carbonate crystals that bind the aggregate together.

Lime plaster and cement was used from the Andes to China, but the Romans were the first true masters of concrete. They mixed lime with volcanic ash known as pozzolan, particularly abundant around Mount Vesuvius near Naples, along with an aggregate of volcanic rock. This pozzolanic concrete was used in structures such as the Pantheon in Rome that remain sturdy today.

The Baths of Caracalla in Rome used pozzolanic concrete.
The Baths of Caracalla in Rome used pozzolanic concrete. Photograph: Stanislav Halcin/Alamy

The production of volcanic ash-lime cement was described in the first century BC by the Roman engineer Vitruvius. Pliny the Elder claimed not only that the Roman pozzolanic concrete was “impregnable to the waves” – it will harden underwater, making it ideal for harbour engineering – but also that it is “every day stronger”. And that’s literally true, as shown in a study by civil engineer Marie Jackson of the University of Utah in Salt Lake City and her colleagues in 2014.

The key is the volcanic ash, which contains minerals called aluminosilicates. These react with slaked lime to make a hard calcium aluminosilicate material that binds the aggregate. “Roman concrete becomes rock,” Jackson says. The Romans typically used big lumps of aggregate, so the result was more like a kind of rubble pile with strong mortar filling all the gaps.

Crucially, the mortar goes on changing for years. As moisture penetrates the concrete structure, some of the remaining volcanic material dissolves and then crystallises into little mineral plates that reinforce the material, blocking the progress of cracks. This even lets the concrete heal itself. “If a fracture forms, new mineral cements can form and repair the fracture,” Jackson says. In this way, “The Romans built huge longevity into their structures.” The concretes used for submerged structures in harbours have not been repaired for more than 2,000 years.

A bioconcrete future

The Roman methods didn’t travel well, because there weren’t many known natural deposits of the volcanic materials it used. In the 19th century, Portland cement became the key ingredient of concrete, made by heating limestone with other minerals (typically clay) and grinding the product to a powder. Portland cement (used in Raac) is cheap and sets fast and hard: well-made concrete structures can now last a century or so. But they still need repair, which can consume up to half of the annual construction budget of government agencies.

That’s one reason why the principle of a self-healing concrete, invented serendipitously by the Romans, is something researchers are now exploring. With the steel-reinforced concretes generally in use today, the problem is not that cracks weaken the concrete itself but that they allow water to reach and corrode the steel. “It’s actually the steel that’s the problem,” says structural engineer Prof Kevin Paine of the University of Bath. The aim is thus to make small cracks self-sealing so that the concrete continues to protect the steel. “We’re not trying to get a recovery of strength but of impermeability,” he explains.

One approach is to pepper the concrete mixture with tiny plastic capsules that contain a substance capable of healing a crack once the crack splits the capsules open. Some of these efforts encapsulate a mineralising chemical such as sodium silicate, or tough resins and glues. Others deliver the healing agent through a network of tubes embedded in the concrete, a bit like the capillaries of our blood supply.

Cheddar Gorge in Somerset.
Bacteria that produce carbonate ions are common in limestone environments such as Cheddar Gorge in Somerset. Photograph: Joe Daniel Price/Getty Images

One idea that Paine and others are exploring is that of trapping live bacteria inside the concrete to conduct the repairs. Some bacteria produce carbonate ions during their metabolism, and if there’s calcium around (as there is in concrete), this can then crystallise as calcium carbonate outside the bacterial cell. Such bacteria are common in limestone environments – Paine collects some of his from nearby Cheddar Gorge.

The bacteria are encapsulated in the concrete as dried-out spores, remaining dormant until a crack opens the capsule and water triggers the spores to germinate. Then the bacteria start generating the carbonate that will seal the crack within a week or so. They need nutrients to grow, and often these are supplied as yeast extract mixed directly into the concrete, as if seasoning it with Marmite.

In a recent collaboration with the University of Newcastle, Paine’s team found that concrete exposed to wastewater might not need added bacteria: there are carbonate-producing bugs already present in the waste. It’s enough then simply to lace the concrete with nutrients and some extra calcium, which the bacteria can access if cracks open up.

Although there have been a few field tests of bacterial self-healing concrete in the UK, the Netherlands and China, Paine says that “very few site trials have ever demonstrated self-healing”. The problem, he explains, is that if you leave a sample and come back several years later and find there are no cracks, it can be hard to know if that’s because cracks have self-healed or because they never formed. “You’re looking for something you don’t want to happen.”

In 2020, researchers at the University of Colorado at Boulder enlisted the help of bacteria to actually make concrete. They unleashed calcium carbonate-forming bacteria within a soft, jelly-like mixture of sand and gelatin and found that the bugs glued the sand grains together with carbonate. Once dry, the material became as hard and strong as some cement mortars. The researchers could switch the bacterial activity on and off by controlling the temperature and humidity, so that this “living material” could be generated on demand. Better still, the final product could be ground up and used to seed the growth of another generation of this “bioconcrete”, and another: the material could be regenerated again and again.

Self-healing won’t be the answer to all the challenges in concrete construction – for one thing, it is likely to be too expensive for use on a massive scale. But Paine says it could be most valuable for infrastructure where repairs could cause a lot of disruption – bridges for railways, say – or those that are hard to reach such as underground sewer pipes.

Even so, he says that the problems with Raac have made him re-evaluate the idea. “My view used to be that not all concrete needs to be self-healing,” he says. “Given the recent problems, I’m starting to reassess that point of view… Maybe self-healing concrete could have more use than I thought.”