Introduction and Overview.
A recent study has provided further insight into the cause of the remarkable durability of Roman concrete1. As is stressed in the paper1, the Ancient Romans were well aware of the robustness of their concrete, which they named opus caementicium: thus, Pliny the Elder recorded in his Naturalis Historia, “that as soon as it comes into contact with the waves of the sea and is submerged becomes a single stone mass (fierem unum lapidem), impregnable to the waves and every day stronger”; while the Roman engineer and architect, Vitruvius, wrote in the first century BCE of “a kind of powdery earth (pulvis) that by its nature produces wonderful results”, and that when “...pumiceous ash (pulvis), lime (calx), and tuff (tofus) ...are brought together in one mixture, and suddenly they are put into contact with [sea-] water, they cohere into a single mass, quickly solidifying, hardened by the moisture, and neither the effect of the waves nor the effect of water can dissolve them.”
The fact that some of their marine structures (Figure 1), including seawalls and harbour piers, still remain standing, and in good condition after 2,000 years, attests to the remarkable strength of Roman concrete, which might be used for the construction of seawalls, and for such tidal electricity generation facilities as the proposed Swansea Bay tidal lagoon power station2, rather than the kinds of concrete that are now typically employed, which are eroded by the impact of the waves over far shorter periods of time. It is clear that the Romans exercised a considerable amount of effort and ingenuity in arriving at the correct formulation of their concrete, whose remarkable durability appears to be due to the presence of poorly crystalline, calcium-aluminum-silicate-hydrate (C-A-S-H binder) in the cementing matrix of the mortar3, the sequestration of chloride and sulfate ions in discrete crystalline microstructures4, and, as recently identified1, the pervasive crystallization of Al-tobermorite (Figure 2) throughout the fabric of the material, frequently in association with the zeolite, phillipsite (Figure 3). Al-tobermorite is a rare, layered, calcium-silicate hydrate mineral composed of aluminosilicate chains bounded by an interlayer region and a calcium oxide sheet, and it is the widespread presence of tobermorite which prevents cracks from extending through the structure of the concrete, and maintains its integrity over time1. The Romans formed a mortar for their concrete by mixing volcanic ash with quicklime and seawater, and then added fist-sized pieces of volcanic rock to this to act as the "aggregate". Once formed, this combination undergoes a pozzolanic reaction. Pozzolans (named after the city of Pozzuoli, in the Bay of Naples) have been defined5 as materials “which, in themselves, possess little or no cementitious value but which will, in finely divided form and in the presence of water, react chemically with calcium hydroxide [portlandite, Ca(OH)2] to form compounds possessing cementitious properties.” The Romans may have been inspired in arriving at the correct formulation by the naturally cemented volcanic ash deposits which are commonly found in the area, and referred to as (volcanic) “tuff” 6,7.
Such concrete was used in the fabrication of the Pantheon and Trajan's Market, now Museo Fori Imperiali, in Rome in Rome (Figure 4), while massive marine constructions served to shield harbours from the open sea and provided anchoring points for ships and underpinning for warehouses. While Portland cement also contains aggregate in the form of sand and gravel, these components need to be unreactive toward the cement paste components, so that gels do not form, which can cause the concrete to crack, as a result of expansion. Marie Jackson, who is a geologist at the University of Utah is quoted7 as saying that:
"This alkali-silica reaction occurs throughout the world and it's one of the main causes of destruction of Portland cement concrete structures."
In contrast with the case of Portland cement where the inert aggregate particles only cause any cracks to lengthen, in the Roman concrete, mineral inter-growths are formed between the aggregate and the mortar, which prevent them from extending further7. Aluminous tobermorite (Al-tobermorite) has also been found in ancient Roman cement structures, which is normally formed in lime particles at relatively elevated temperatures. However, it is thought that the material can be formed at much lower temperatures, as a result of seawater percolating through the concrete in seawalls and in piers, dissolving minerals from the volcanic ash and creating a strongly alkaline environment from which Al-tobermorite and phillipsite can crystallize. Significantly, Al-tobermorite forms crystals with platy shapes that interlock to reinforce the cementing matrix and thus increase the concrete's resistance to brittle fracture7.
Reinforced concrete (rebar) has underpinned the massive structures which characterise the modern world; however, despite having a far higher tensile strength, its longevity is one to two orders of magnitude less than that of the concrete used by the Romans. A main reason for this is that when the surrounding concrete cures, oxidation takes place and the reinforcing bars rust over decades, causing sufficient expansion that cracks are formed in the concrete. In contact with seawater, rebar has a life-expectancy of perhaps just 50 years, due to corrosion of the steel, and reactions with calcium hydroxide that cause expansion within the concrete structure8. This vulnerability is absent in Roman concrete, since there is no steel reinforcement in the ancient structures, and furthermore, the volcanic ash mortars are more ductile, and the concrete is reinforced at the structural scale within a conglomeratic rock framework9-11.
Massive harbour structures throughout the Mediterranean region, dating from Roman times, that have been exposed to seawater for two thousand years, are found to contain reactive, alkaline, pumiceous ash aggregate (ranging in size from fine-sand to gravel) commonly with zeolitic surface coatings. However, the nature of those processes that might occur well after the consumption of calcium hydroxide, through pozzolanic reaction with the volcanic ash aggregate, is not at all clear1. This pozzolanic reaction was known to Vitruvius, who noted1 that “latent” heat was released when tuff, pumiceous ash, and lime (CaO) (tofus, pulvis, and calyx) from the Campi Flegrei and Vesuvius volcanic districts “come into one mixture and suddenly take up water and cohere together”. Jackson et al. have reported3 an adiabatic model of exothermic heat being evolved during the hydration of lime and the formation of pozzolanic C-A-S-H binder in a Baianus Sinus breakwater, in the Bay of Pozzuoli, Italy that suggests that temperatures in the range 65–95 °C were maintained for 2–3 years3. In partially dissolved, relict lime clasts, it is likely that this pozzolanic phase of reaction involved crystallization of Al-tobermorite associated with C-A-S-H, a process that was probably complete fairly early on following the building of the concrete marine structures, since it has been shown that that portlandite [Ca(OH)2] was fully consumed after 5 years of hydration in seawater, in a mock-up of a Roman concrete breakwater11-13.This result is similar to those obtained from other experiments in which concretes with volcanic ash aggregates, in contact with seawater8, were investigated.
Jackson et al.1 used a combination of synchroton-based X-ray microdiffraction) to investigate the textures of zeolite and Al-tobermorite in situ within the Portus Cosanus subaerial pier and the Baianus Sinus and Portus Neronis submarine breakwaters, and Raman spectroscopy to determine the bonding environments in Baianus Sinus Al-tobermorite from various crystallization environments. On this basis, insight is provided into the nature of the low temperature crystallization and stability of phillipsite and Al-tobermorite in alkaline aqueous environments, and how cycling of authigenic minerals [minerals that form in situ within the depositional site by precipitation or recrystallization, instead of being transported from elsewhere (allogenic) by water or wind] occurs over very long service lifetimes, and beneficially, within construction materials that contain natural, volcanic pozzolan.
Using the X-ray synchrotron at the Lawrence Berkeley National Laboratory in Berkeley, California, Jackson et al.1 were able to determine the distribution of minerals in samples of Roman harbour concrete. A silicate mineral called phillipsite, which is frequently found in volcanic rocks, was identified, but interestingly, they further observed that crystals of aluminium tobermorite were growing from it. Thus it is concluded that tobermorite had grown from the phillipsite due to the action of seawater washing through the concrete, which increased its degree of alkalinity. Such kinds of crystallization are very unusual events, and indeed are only found in such places as the volcanic island of Surtsey, which is located in the Vestmannaeyjar archipelago off the southern coast of Iceland14. It is thought that, as the tobermorite grows throughout the concrete, it may provide strength to the due to its long, plate-like crystals which allow the material to flex when stressed, and not shatter14. The formation of Al-tobermorite in geologic settings is thought to be a result of hydrothermal processes1: a view that is supported by various laboratory syntheses of the material, although these have all required temperatures in excess of at ≥80 °C 15. However, phillipsite and Al-tobermorite have been documented to form within the pores of a Portland cement paste which was in contact with a claystone interface, one year after installation, and at a lower temperature of 70 °C 16. In addition, examination of a 181 metre core that was drilled-out in 1979, on the volcanic island of Surtsey, showed the pervasive presence of Al-tobermorite, and which spanned a temperature range from 25 °C in surficial deposits to 140 °C in hydrothermally altered tuff17. The significance of the Roman marine mortar system is that it evidences authigenic mineral cycling processes that occur at low temperatures, in which an interplay of different processes, including dissolution and reactions of minerals originally present in volcanic ash, the production of micro-encapsulated alkaline fluids, and the precipitation of new minerals therein, primarily phillipsite, leading on to chemical changes that cause the crystallization of Al-tobermorite crystallization throughout the structure of the concrete. Although it is only though the agency of sophisticated modern analytical methods that the detailed nature of Roman marine concrete technologies are finally being realised, it is clear that the Roman builders intended to produce pozzolanic cementitious systems, whose evolution occurred via the recycling of authigenic minerals, similar to those that occur in geological processes and which could offer a practically geological longevity. Put simply, the Romans designed their marine and other concrete structures to stand the test of time! In contrast with modern rebar concrete based on Portland cement, exposure to seawater causes a “corrosion” that actually strengths the material, and extends its longevity, rather than degrading it, so that it may last for only five decades or so18.
Concrete containing fly ash.
Fly ash is produced by the combustion of coal in power plants, and consists of particles that are sufficiently fine to be carried out with the flue gases. The particles are captured, e.g. using electrostatic separators, before the flue gases reach the chimneys. The exact composition of fly ash depends on the type and origin of the coal-fuel, but the major components are always silicon dioxide (SiO2), aluminium oxide (Al2O3) and calcium oxide (CaO). That fly ash can be used to partially substitute for portland cement in concrete, was discovered as early as 1914, although it was not until 1937 that a significant such use was first demonstrated, and around 15 million tonnes are used annually as a component of concrete19. Due to its broadly similar chemical composition, fly ash has similar (pozzolanic) properties to the volcanic ash that Romans used to make their concrete, which as we have seen, greatly improves the strength and durability of concrete and is a critical factor in the remarkable degree of preservation of buildings and marine structures dating from the time of the Roman Empire1.
There is ongoing research at Ghent University in Belgium, in the use of fly ash to create a kind of concrete which causes cracks to closes up, in a kind of “self-healing” process20. Fly ash is often used to replace up to 30% of the mass of Portland cement, but greater proportions are sometimes used, and there are methods available to permit a 50% degree of substitution. In the construction of the Ghatghar dam project in Maharashtra, India, a 70% replacement level has been attained using processed fly ash, and roller-compacted concrete, a material widely employed in dam-construction21. The workability of cement can be enhanced while using a reduced water-cement ratio, in consequence of the spherical shape of the fly ash particles. The term “Green Concrete” is sometimes applied to concrete that contains fly ash, since it is argued that replacing Portland cement with fly ash reduces the “carbon footprint” of concrete, because there is no new CO2 generated from the fly ash (this has already been produced in the combustion of the coal it is derived from), while the production of each tonne of Portland cement occasions practically one tonne of CO2. In the initial production of the fly ash (i.e. burning coal), around 20 to 30 tonnes of CO2 are produced per tonne of fly ash21. Since the worldwide production of Portland cement was 4.2 billion tons in 2016 22 – a figure that is expected to rise along with the increasing urbanisation of global civilization – if a significant proportion can be substituted for by fly ash, a substantial reduction in the CO2 emissions from the construction industry might be achieved.
Reducing both CO2 emissions and water consumption.
The production of concrete is responsible for around 5% of global CO2 emissions23, and is reckoned by the IPCC (Intergovernmental Panel on Climate Change), to be the largest source of global carbon dioxide emissions beyond the energy sector. It is the production of clinker [in which a mixture of limestone and clay is sintered, by heating it to around 1,450 oC 24, such that the limestone (CaCO3) is converted to lime (CaO) in situ] that is the major contributor to the overall CO2 emissions from Portland cement manufacture. Since this process consumes a large amount of energy23 (5 GJ per tonne, which is equivalent to about 180 kg of coal), there is a further CO2 burden, and for each tonne of cement that is produced, overall around one tonne of CO2 is released. Since the Romans burned their limestone in a lime kiln, which can be achieved at a lower temperature of 900 oC (1,000 oC is typically used to accelerate the reaction sufficiently)25, the fuel requirements and energy emissions are accordingly less. There are additional CO2 emissions to be accounted for, arising from the need to transport the various component materials, and to mix the cement with gravel, sand, and water, to pour it into the particular structures it is intended to fabricate, and as produced by the various chemical reactions that occur as part of the hardening process for concrete. As an additional factor, large volumes of water are consumed by concrete manufacture23: in total, around 500 litres of water are required to make a cubic metre of concrete, along with the processes of curing, cooling, and cleaning. Thus, the global annual production and application of 4 billion cubic meters of concrete consumes about two cubic kilometers of water, mostly sufficiently pure for drinking. The Romans employed a lower water/cement ratio to make their concrete, which suggest that there is scope for reducing the water demand in global concrete manufacture. Moreover, if seawater could be used in place of drinking water, to produce concrete in coastal regions, where most of the people live across the globe, and using locally available volcanic rocks, considerable savings on both CO2 emissions and water use might be possible.
Paulo Monteiro of the University of Berkley has pointed out that although the Roman concrete exhibits a remarkable durability over very long timescales, it is unlikely to become a wholesale replacement for the construction of many modern buildings, which require the concrete to set more rapidly24. The Berkley group are researching into using volcanic ash as a high volume substitute for fly ash, for use particularly in countries where the amount of available fly ash is limited:
Montario is quoted as saying24: “There is not enough fly ash in this world to replace half of the Portland cement being used. Many countries don’t have fly ash, so the idea is to find alternative, local materials that will work, including the kind of volcanic ash that Romans used. Using these alternatives could replace 40 percent of the world’s demand for Portland cement.”
Since the production of the Roman type concrete requires lower temperatures, less energy and accordingly smaller carbon dioxide emissions than for the kind of concrete typically used today, clear environmental advantages are offered, along with the fact that it may serve to provide a highly robust and durable material for the fabrication of seawalls and other marine structures. However, Roman concrete hardens more slowly and takes time for the seawater to strengthen it; furthermore, the final material is compressively weaker than Portland cement. Hence, there are limitations to where Roman concrete could be used, and a wholesale substitution is unlikely, e.g, for constructing tall buildings and bridges, although niche applications can be envisaged7. For example, it has been proposed7 that Roman concrete should be used to build the seawall for the Swansea Tidal Lagoon power plant, which is intended to operate for 14 hours per day with a maximum output of 320 MW, enough to power around 155,000 homes. The structure is to be made sufficiently strong to withstand the most destructive of storms - such as occur every 500 years - and to provide a coastline protection against storms and floods2. In order to recoup the costs of building it, the Lagoon would need to generate power for 120 years, by when the reinforced concrete from which it was constructed “would be a mass of corroding steel." 7 In contrast, a structure made from Roman concrete might preserve its integrity for centuries. Clearly the material might also be used to construct other barrage structures for low-carbon energy generation, e.g. the vexed Severn Barrage tidal power station26.
The material also offers the potential to furnish robust defences against sea level rise, which is expected to become a more severe problem, as global temperatures rise. Indeed, sea ice is melting and causing the sea level to rise at triple the pace it did in 1900 27. Although there is an element of uncertainty as to exactly what the sea level will be in the future, it appears that it will be necessary to build barriers to defend coastal cities, in particular seawalls. These serve to prevent cities from being flooded during storms and high tides, and are used around the world to defend vulnerable, coastal locations. Hence, highly durable Roman cement formulations may prove invaluable for the fabrication of such fortifications.
(1) Jackson, M.D., Mulcahy, S.R., Chen, H. et al. (2017) American Minerologist, 102, 1435-1450.
(2) Waters, S., Aggidis, G. (2016): A World First: Swansea Bay Tidal lagoon in review. In: Renewable and Sustainable Energy Reviews 56, 916–921.
(3) Jackson, M.D., Chae, S.R., Mulcahy, S.R. et al. (2013) American Mineralogist, 98, 1669–1687.
(4) Jackson, M.D., Vola, G., Všianský, D. et al. (2012) Cement microstructures and durability in ancient Roman seawater concretes. In J. Válek, C. Groot, and J. Hughes, Eds., Historic Mortars, Characterisation, Assessment and Repair, p. 49–76, Springer, Berlin.
(5) Mehta, P., and Monteiro, P.J.M. (2015) Concrete: Microstructure, Properties, and Materials, 4th ed. McGraw Hill, New York.
(6) Wikipedia (2017) https://en.wikipedia.org/wiki/Tuff [Accessed December 11th, 2017].
(7) University of Utah Website (2017), July 3rd. https://unews.utah.edu/roman-concrete/ [Accessed December 11th, 2017]
(8) Massazza, F. (1985) Il Cemento, 82, 26–85.
(9) Brune, P.F., Ingraffea, A.R., Jackson, M.D., and Perucchio, R. (2013) Engineering Fracture Mechanics, 102, 65–76.
(10) Jackson, M.D., Landis, E.N., Brune, P.B. et al. (2014) PNAS, 111, 18485–18489.
(11) Jackson, M.D. (2014) Seawater concretes and their material characteristics. In
J.P. Oleson, Ed., Building for Eternity: the History and Technology of Roman
Concrete Engineering in the Sea, p. 141–187, Oxbow Books, Oxford.
(12) Oleson, J.P., Bottalico, L., Brandon, C., et al (2006) J. Roman Archaeol., 19, 29–52.
(13) Gotti, E., Oleson, J.P., Bottalico, L. et al. (2008) Archaeometry, 50, 576–590.
(14)Witze, A. (2017) Nature News July 3rd. https://www.nature.com/news/seawater-is-the-secret-to-long-lasting-roman-concrete-1.22231. [Accessed December 11th, 2017]
(15) Komarneni, S., and Roy, D. (1983) Science, 221, 647–648.
(16) Lalan, P., Dauzères, A., DeWindt, L. et al (2016) Cement and Concrete Research, 83, 164–178.
(17) Jakobsson, S., and Moore, J.G. (1986) Geological Society of America Bulletin, 97, 648–659.
(18) Dean, A.M. (2017) International Society for Concrete Pavements, October 18th. https://www.concretepavements.org/2017/10/18/rock-solid-history-of-concrete-how-limestone-rocks-and-volcanic-ash-built-the-modern-world/ [Accessed December 11th, 2017].
(19) Thomas, M. (2007) http://www.cement.org/docs/default-source/fc_concrete_technology/is548-optimizing-the-use-of-fly-ash-concrete.pdf [Accessed December 11th, 2017].
(20) Snoeck, D., Van Tittelboom, K., Wang, J. et al. (2017) Ghent University Website, https://www.ugent.be/ea/structural-engineering/en/research/magnel/research/research3/selfhealing [Accessed December 11th, 2017].
(21) Wikipedia (2017) https://en.wikipedia.org/wiki/Fly_ash [Accessed December 11th, 2017]
(22) Statistica (2017) https://www.statista.com/statistics/219343/cement-production-worldwide/ [Accessed December 11th, 2017].
(23) Duplan, N. (2017) Medium, July 22nd. https://medium.com/@nenoduplan/from-the-foundations-of-rome-to-global-carbon-emissions-reduction-9a2957260566 [Accessed December 11th, 2017].
(24) Yang, S. (2013) Berkley News, http://news.berkeley.edu/2013/06/04/roman-concrete/ [Accessed December 11th 2017].
(25) Wikipedia (2017) https://en.wikipedia.org/wiki/Lime_kiln [Accessed DEcember 11th, 2017[.
(26) Wikipedia (2017) https://en.wikipedia.org/wiki/Severn_Barrage [Accessed December 11th, 2017].
(27) Dangendorf, S., Marta Marcos, M., Wöppelmann, G. et al. (2017) PNAS, 114, 5946–5951.
Captions to figures.
Figure 2. Sample of tobermorite from Mexico. Credit: Rob Lavinsky. https://upload.wikimedia.org/wikipedia/commons/5/51/Tobermorite-t08-76a.jpg
Figure 3. Complex sample of phillipsite along with Herschelite (large crystal toward top of image) from Mount Etna, Sicily, Italy. Credit: Didier Descouens https://upload.wikimedia.org/wikipedia/commons/6/6b/Herschelite.jpg
Figure 4. Trajan’s Market, in Rome: thought to be the world’s oldest shopping mall. Credit: Zello. https://upload.wikimedia.org/wikipedia/commons/7/7e/Trajansmarket2.JPG