Balancing the impact and utility of concrete
Professor Phil Purnell of the University of Leeds, UK, shares his insight into today’s use of concrete.
Roads, rail, air and sea ports, energy, communications, flood defence, water and housing – our economic infrastructure – are all enabled by reinforced concrete, yet concrete’s social status remains low. In the post-war period, Le Corbusier’s ‘cities in the skies’ and inner-city ring roads were set to revolutionise urban living, exploiting the seductive new wonder material. But within a decade, some of the hastily built structures that had been erected began to decay, owing to the then poor understanding of reinforced concrete’s durability. Their continued deterioration in the 1970s and 1980s – especially those structures associated with social housing – seemed to mirror the political and economic demises of the post-war consensus and traditional communities. Concrete became a trope, a stock image used as proxy for urban decay. To this is now added the green shibboleth that concrete is a carbon-intensive material. Let’s put this into context.
In 2010, around 40 gigatonnes (1012kg, Gt) of products were manufactured worldwide. Of this, at least 22Gt was reinforced concrete. In the same year, 37Gt of CO2 was emitted, with 3.4Gt (9%) attributable to reinforced concrete. In other words, concrete accounted for over half of all products, but less than a tenth of CO2 emissions. By comparison, steel manufacture (0.95Gt, 2.4%) accounted for 8% (3.0Gt) of CO2 emissions. Timber manufacture (2.2Gt, 5.6%) was arguably associated with the 14% (5.1Gt) of CO2 emissions attributable to forestry operations. On this basis, concrete is less carbon-intensive than almost any other material.
However, a tonne of steel or timber does not do the same job as a tonne of concrete. Figure 1 compares beams on the basis of CO2 emitted per unit of structural performance i.e. span and bending capacity. At the residential scale (5m span, low bending capacity) timber performs best, but it’s not possible to design small enough concrete beams so ‘oversized’ beams must be used (dashed lines). This is why lifecycle analyses of putative concrete versus timber houses favour timber – no sensible designer would use concrete at this engineering scale. As loads and spans increase, concrete rapidly matches or surpasses timber in terms of CO2 per unit of structural performance. Carbon intensity is a function of structural form and scale, not material choice.
Fig 1 above: Embodied CO2 per unit of structural performance for beams. HS = high-strength. PFA = concrete with 40% replacement of cement with PFA. Timber = glulam beams. Steel = UB sections.
Nonetheless, manufacture of the cement that binds concrete together emits huge quantities of CO2. Compared with iron or aluminium manufacture, best-practice energy use in the cement industry (3GJ/t) is close to the thermodynamic minimum (2GJ/t), and secondary approaches such as heat recovery for district heating or use of low-carbon fuels need to be employed to make systemic energy savings.
In any case, 60% of the CO2 emissions associated with calcium silicate construction cements come from the process chemistry (calcining limestone: CaCO3 + heat = CaO + CO2). Non-carbonate calcium sources are available but not on the scale required, according to US Geological Survey data. Around 5Gt of limestone is consumed annually, producing 3Gt of cement, the next two most consumed calcium minerals are gypsum (<0.3Gt) and fluorite (<0.1Gt). Both have alternative higher value uses and would emit sulphates and fluorides on a similar scale. Alternative cement chemistries (e.g. based on magnesium) are used on a limited scale, but most geological sources are also carbonates.
Replacing some of the cement with wastes from iron manufacture, such as ground granulated blast furnace slag (GGBS) and coal burning during electricity production of pulverised fly ash (PFA) that contain chemically active silicates, has been practiced for many years. These wastes are nominally zero carbon, as the CO2 emissions associated with their production are attributed to the primary product (i.e. steel or electricity), so their use reduces concrete’s carbon footprint. Current supplies of PFA and GGBS are 0.7Gt and 0.4Gt respectively and even if these were fully utilised, CO2 savings would be marginal. Their potential supply is declining, ironically because of efforts to decarbonise their parent industries. Use of biomass to replace coal, and of electric arc furnaces to replace blast furnaces, renders PFA and GGBS chemically unsuitable for use in concrete.
Much research focuses on new, low-carbon cement formulations. Yet simply reducing the amount of new concrete we manufacture – using less for the same function, designing better concrete, and keeping it in use for longer – could provide substantial carbon savings. This could be started tomorrow, within existing design practices, rather than waiting years for new cements to come to the market.
The design of standard reinforced concrete beams is a useful example. The traditional approach, familiar to undergraduates, derives a section to resist the maximum bending moment – usually in the centre of the span – and then specifies a prismatic beam based on this section. Yet much of the concrete serves no structural function in such a beam. If we keep the reinforcement area constant and vary the section depth to match the bending moment along its length, we can reduce the concrete volume by nearly 30%. Why don’t we routinely do this, when 19th Century iron structures often display such shape optimisation? The answer lies in economics, not engineering. The ratio between the prices of skilled labour and materials has increased 20-fold over the last two centuries. To cut costs we now save on design hours, leading to shapes that waste materials.
Careful choice of mix designs that avoid extremes of strength and make correct use of admixtures can be very effective at reducing CO2 emitted per unit of structural performance. Restricting the use of mixes above 50MPa, for which Eurocode 2 mandates extra steel areas for a given load resistance, can also provide considerable carbon savings.
Fig 2 above: Embodied carbon per unit of structural performance versus concrete strength grade. Left: reinforced concrete beam (low-slump concrete, crushed aggregate). Right: plain concrete, thin/bold lines = high/low slump concrete, black/red lines = with/without superplasticiser.
Preserving concrete structures avoids the need to make new concrete. The legacy of decaying post-war concrete has given the material a reputation for poor durability. Analysis of 144 notable bridge failures since 1950, of which 23 were concrete, suggests that 22%, 12% and 9% of timber, steel and mixed design bridges respectively failed owing to deterioration, whereas only one concrete bridge was thus affected – the other 22 failed owing to crashes, natural disasters, construction incidents etc. Therefore, this reputation seems undeserved.
Case studies of concrete failures rarely mention inherent degradation, but point to maintenance deficiencies. Maintenance is a cultural problem throughout the built environment. In the paper, From bridges to education: best bits for public investment, published by The Brookings Institution, Harvard Economist Larry Summers describes deferred maintenance as ‘a debt burden on the next generation’ and refers to the ‘tyranny of the ribbon’, where spend on new projects, owing to its greater political value, is prioritised over maintenance. Annual USA state spending on roads is split 55:45 between expansion and repair. Put another way, over half the budget is spent on about 1% of the asset. Highways England figures show £0.3bln – 0.26% of the total asset value – spent on routine annual maintenance, equivalent to expecting to maintain a £40,000 car for £100 per year.
Fixing our socioeconomic blind spot regarding maintenance – especially for reinforced concrete, which with care will last almost indefinitely – would greatly help us reduce carbon emissions. Proponents of steel and aluminium laud their recyclability, yet recycling each tonne requires energy equivalent to that consumed annually by two to six average UK households respectively. Recycling concrete provides little if any saving over making new concrete. Recovering the function of materials by maintaining and reusing structures or their components (beams, columns etc.) offers a better way of drastically cutting emissions. Analyses of the lifetime performance of new low-energy houses compared with refurbishing existing houses shows that there is often little difference in carbon savings. For the former, savings in operational CO2 (heating, lighting etc.) are negated by the embodied CO2 associated with the new materials used. For the large structures and infrastructure to which concrete is well-suited, embodied emissions normally outweigh operational emissions and so the savings associated with reuse will be higher.
There are many issues with reusing concrete components. They are often cast in situ without joints, making structures difficult to dismantle. Key properties such as section capacity are usually bespoke and cannot be inferred from the section dimensions. However, these are design issues, not material limitations. Concrete structures could be designed to be dismantled. Modern digital systems such as BIM could record section properties, and this information could be augmented by loading and environmental histories gathered using embedded sensors integrated via the Internet of Things that would allow residual properties and hence reuse potential to be determined.
So which of these approaches should we adopt? A practical combination of all approaches could cut the embodied carbon of concrete by over 70%. This and not searching for magic cement formulations or energy sources, will help us drastically reduce the carbon footprint of concrete and allow it to gain a place at the table of low-carbon solutions appropriate to its unparalleled utility.