In this part, I give an overview of the problem of embodied emissions, i.e. those emitted in the construction of buildings and infrastructure (section 7.1); then some details about concrete and steel (section 7.2), and cement recarbonation (section 7.3); and about roads (section 7.4).
7.1. Overview
This graphic shows the sources of the built environment’s embodied CO2 emissions for 2019, including emissions from steel manufacture.[1] Each row represents a different breakdown of the same total – the 6.6 GtCO2 of embodied, energy-related emissions. The second row, unlike the other two, also shows the process emissions from the production of steel and cement.
 |
| Sources: * IEA (2020); △ Robbie M. Andrew (2022); § IEA / UNEP (2020), IEA / UNEP (2021) |
The vast majority of the built environment’s embodied emissions come from the burning of fossil fuels during the manufacture of building materials.
For example, in the case of buildings construction, in 2019 just 0.13 Gt CO2 emissions globally came from the buildings construction stage – a comparatively tiny proportion of the roughly 4.45 Gt total embodied emissions.[2] The rest came from the manufacture of building materials prior to construction.
 |
| The Seagram building in New York. Source: Creative Commons |
Of the carbon footprint of those materials that went into buildings construction, around 60% of emissions came from cement and steel manufacture, and 40% from the manufacture of other buildings materials. For the construction sector as a whole, the ratio is something like 50:50 cement and steel emissions to other emissions.
This underlines the point, emphasised in part 3: steel and cement (and concrete made from cement) are the high-energy ingredients of choice for fossil-fuelled global construction.
Sand and gravel are also major inputs. Indeed, the construction sector is driving an impending sand crisis. The main emissions cost of these is the energy of extraction, processing and transport.
In addition, construction consumes 26% of global aluminium output and 19% of all non-fibre plastics.
The levels of embodied emissions in common construction materials can be seen in the “Construction Material Pyramid”, shown below, designed by the Centre for Industrialised Architecture in Denmark. The values given are averages that include direct and indirect emissions footprints by point of sale (cradle-to-gate).
At the base of the pyramid are materials that have a low emissions intensity, i.e. that typically require just a small input of energy or other sources of emissions in their production: rammed earth walls, plywood, construction timber. (Wood here even gets a negative value as a material that “sequesters” carbon, although I think that framing can be misleading. See section 8.4 below.)
Grouped around the top of the pyramid are materials with a high emissions-intensity – those that require either a high expenditure of carbon independently of their energy footprint, or those with a large energy footprint that is entirely or overwhelmingly carbon-based: these include conventional C20 and C25 grades of concrete, structural steel, galvanised steel, and at the peak, aluminium sheetwork.[3]
 |
| Source: Centre for Industrialised Architecture |
7.2. Concrete and steel
Because they are so dominant, I will focus here on concrete and steel emissions.
In part 3 I mentioned the aesthetic and ideological role played by concrete and steel in modern and contemporary architecture.
The Seagram Building in New York City is a good example of this. Designed by Ludwig Mies van der Rohe, and completed in 1958, it is an icon of architectural modernism. However, specified as it was in the Cold War era of fossil capital, its minimal aesthetics hide an incredibly dirty material reality.
Barnabas Calder and Florian Urban, two architectural historians, write that concrete accounts for 79% of the Seagram Building’s mass – most of the rest coming from high-energy steel and glass. Moreover, it spews out vast volumes of operational emissions (see part 9).
The combination of steel and concrete can perform incredibly well in structural terms.[4]
Concrete, and its crucial binding agent, cement, are consumed almost entirely by the construction sector. Steel, by contrast, is used in all sorts of manufactured goods.
By mass, concrete is roughly 10-15% cement (depending on its specification). The other components are aggregates such as sand, gravel and crushed stone, plus water and chemical additives. The cement produces most of the emissions.
The bar chart, below right, shows that in 2019, the global construction industry used around 4.1 Gt of cement. As well as being used in concrete, cement is a crucial component of mortar.
The graph shows that around 31% of the world’s cement went on residential buildings; 17% on commercial buildings; 42% on infrastructure; and 10% is lost or wasted.
 |
| Global cement production forecasts. Source: IEA (2020) |
According to the World Steel Association (WSA), construction consumes 52% of the world’s steel products by weight – around 0.92 Gt out of the total 1.77 Gt in 2019. To make those products, 1.87 Gt of crude steel was used. Somewhere between the crude steel production and the various finished steel outputs, there was a loss of about 0.4 Gt (~21%). This is shown in the next graph, below right.
 |
| Global steel production forecasts. Source: IEA (2020) |
Of the finished steel products that went to infrastructure, almost all of it went towards transport infrastructure (e.g. rail tracks and bridges). This graphic shows the WSA’s estimates of the final uses of steel globally.
 |
| Global steel use. Source: World Steel Association |
The cement industry was responsible for around 2.5 Gt CO2 of emissions in 2019, or about 4.2% of all sociogenic greenhouse gas emissions. (See the graphic at the start of part 7, the second row.)
The iron and steel sector produced a total of 3.6 Gt CO2 of emissions in 2019.
Assuming all end-uses for steel have the same carbon intensity, the iron and steel that goes to the construction sector produces about 1.9 Gt CO2 emissions annually.
Why do cement and steel manufacture for the built environment produce so much greenhouse gas emissions?
In part, because so much is produced (the 4.1 Gt of cement and 0.92 Gt of steel, in 2019, mentioned above). But the main thing is that each of those tonnes is so emissions-intensive.
The production of 1 tonne of cement emits 0.5-0.6 tonnes of CO2, according to the IEA. Steel is even worse: for each tonne of finished steel products, on average 2.0 tonnes of CO2 emissions are dumped into the atmosphere!
Estimates of the emissions from steel and cement manufacture are shown in the table.
“Process” emissions are those that come from the immediate chemical reactions that take place during production. They are distinct from energy-related emissions, e.g. from electricity, or from direct fossil fuel combustion.
Cement is usually made by heating limestone (calcium carbonate, CaCO3) in a kiln with other minerals, such as clay, so that it breaks down into quicklime (calcium oxide, CaO) and CO2. This is called calcination (or decarbonation), and produces clinker as an intermediary product. The CO2 process emissions of cement manufacture are those from calcination.
The heat to produce clinker is incredibly energy-intensive, and for reasons of economy tends to come from burning coal. Those energy-related combustion emissions (categorised as CO2 FFI), alongside other processes that use electricity – such as grinding, milling and loading ingredients – comprise the other one-third of cement’s CO2 emissions.
The weight of cement’s CO2 process emissions comes from the quicklime that goes into cement manufacture, whereas the weight of cement combustion emissions comes mostly (~73%) from oxygen in the air. (See also Appendix 3, in the PDF version.)
Steel production is dominated by energy-related emissions (89%) – although some process emissions come from the use of lime fluxes, graphite, and ferroalloy production. There are many manufacturing pathways, but the main energy-related emissions come from heating a blast furnace to produce molten pig iron at temperatures of up to 1400-1500°C. Most steel manufacture uses coal to provide that heat.
Global cement production nearly doubled from 1.7 Gt in 2000 to 3.3Gt in 2010, growing to 4.1 Gt in 2019 (United States Geological Survey). As you would expect, annual CO2e emissions due to cement manufacture over the 2000-2020 period have also almost doubled. Steel emissions have more than doubled too.
Crucially, the doubling of both concrete and steel emissions since 2000 has massively outweighed some emissions-saving changes in their manufacture.
This is reflected in the bar charts below, which show the embodied CO2 emissions of the cement and steel content of only buildings construction, in 2000 and 2019. The IEA’s decomposition analysis shows the proximate drivers of that growth dominated by demand for more floor space. The red bars show emissions savings per tonne of output – moderate for concrete, tiny for steel. (Projections for 2020, made before the Covid pandemic, are also included.)
 |
| Source: IEA (2020) |
The expansion of cement’s greenhouse gas emissions footprint since 2000 is almost entirely due to construction in China (see part 4).[5] Steel’s carbon footprint enlargement is also mainly driven by demand for steel in China, both for construction and other manufacturing.
7.3 Cement recarbonation
One further point about cement is that it also has an important role as a carbon sink. Cement re-absorbs CO2 from the atmosphere like a sponge, during curing, through a process called cement carbonation or recarbonation. CO2 in the air recombines with calcium hydroxide (Ca(OH)2) in the cement to form calcium carbonate (CaCO3) – i.e., limestone, the main production ingredient of cement. Recarbonation occurs with concrete, or in any similar cementitious material.
But this re-absorption happens very slowly indeed: it is thought that 10-30% of cement’s production emissions are typically reabsorbed over the next 50-100 years. So just using more cement is hardly a carbon-capture solution.
Carbonation happens from the surface inwards, and the amount of CO2 reabsorbed depends on the surface area and physical characteristics of a cement-containing material. For example, carbonation is reduced by surface coatings such as paint. Cement in demolition waste carbonises quicker, so long as it is above ground, because more of its surfaces become exposed to air.
However, when carbonation happens in built structures, it has the disadvantage that, over time, it can compromise the structure of cement-based materials – causing cracking, for example, and the eventual rusting of steel rebar.
Carbonation increases as built cement stocks increase – and is greater if a cement structure remains in place for longer. By one recent estimate world cement carbonation has risen from an average of 0.07 Gt CO2 per year in the 1960s, to an average of 0.7 Gt CO2 per year during the decade 2010-2019. A separate study calculated that, by 2019, about 21 Gt CO2 had been absorbed into cement produced between 1930 and 2019. The 2019 cement carbonation sink alone was roughly 0.89 Gt CO2, about a third of the cement-related emissions for 2019.
 |
| Source: Pierre Friedlingstein et al. (2022) / Global Carbon Budget 2021 |
However, a more apt way to look at it is that the quantity of cement emissions from new construction alone in 2019 was double the uptake of CO2 by the world’s entire accumulated cement stock.
The comparatively high rates of cement carbonation simply echo the devastating extent of historical cement-related emissions. They certainly do not stand in cement’s favour from the perspective of decarbonising the built environment.
7.4 Roads
Finally, I will touch on the material footprint and carbon footprint of roads and their associated infrastructure – also important components of the contemporary built environment.
Asphalt road surfaces comprise asphalt concrete (AC, commonly referred to as asphalt, tarmac, or bitumen macadam), which consists of aggregates of a given size grade (sand, gravel), bound together with heated bitumen. Bitumen is a semi-solid form of petroleum found naturally, or manufactured, which, confusingly, is also called asphalt. Typically, AC is about 95% aggregates to 5% bitumen binder.
The world’s total material stock of asphalt concrete was estimated by researchers in 2015 at around 115.5 Gt.[6] In 2013, about 18 Gt of the stock of asphalt concrete were contained in roadways in the US, according to the US National Asphalt Pavement Association.
Global consumption of asphalt concrete was estimated in 2016 at about 2.1 Gt per year,[7] for new construction, plus large amounts of repair, and maintenance related to subsurface infrastructure such as water pipes and communications cables.
Naturally, the ratio of asphalt cement used in new road construction versus maintenance varies geographically, according to the relative volumes of new road-building versus legacy stocks.
A recent case study of road construction in Vienna, Austria, for example, found that in the period 2011-15, road maintenance comprised ~58% of road construction mineral inputs; maintenance of subsurface infrastructure was the cause of a further ~32%; whereas newly built roads consumed only ~10%. (See graphic below).
 |
| Road construction & maintenance material use in Vienna. Source: Andreas Gassner et al (2020) |
Of course, many roads are not made of tarmac. The structure and sub-structure of tarmacked roads also depends on the situation: for example, urban roads will often be supported by ancillary structures made of reinforced concrete.
In terms of greenhouse gases: a recent case study of the construction of the La Abundancia-Florencia highway in Costa Rica estimated that the direct and indirect embodied emissions from the construction of the road surface, consisting of hot mix asphalt concrete, was 65.8 kg CO2e per lane per kilometre. This does not include subsequent maintenance and upkeep.
Another 2017 case study of the area inside the 5th Ring Road in Beijing (~670 km2) estimated that, on average, around 73% of lifecycle emissions were due to the use of concrete in large urban ancillary structures such as bridges, and other cement products such as pre-cast raised kerbs. Emissions during the production stage of the average road were 1,850 tonnes CO2e/km across the study area. Emissions during a lifetime of maintenance were 1,760 tons CO2e/km. The net contribution of recycled materials to these sums was -200 tonnes CO2e/km.
All these embodied emissions are besides the induced carbon load associated with a greater use of private transport on roadways. Tarmac roads and reinforced concrete flyovers can be reserved for pedestrians, cyclists and e-scooters, but that does not usually happen. Most new roadways go to private motor vehicles and haulage traffic.
That induced carbon load (classified by researchers under transport emissions) is both embodied in the means of transportation, and operational in the use of fuel. For the time being, motor vehicles mostly still run on fossil fuels. Setting aside heavy goods vehicles and looking just at cars: for an average 200,000 km of lifetime mileage, emissions are around 36 tonnes CO2e per vehicle, according to the IEA in 2021.
But even if an electric vehicle (EV) is run entirely on renewables (for now, a big if), its average embodied emissions are still around 8-9.5 tonnes CO2e per vehicle lifetime, compared to around 6 tonnes CO2e for a fossil-powered car. As and when the carbon footprints of manufacturing motor vehicles decline, those numbers will also move – but again, that seems a long time coming.
In part 8, I will consider how embodied emissions in the built environment can be reduced.
🔥 Go to part 8
🔥 Go to Contents and Introduction
Download the whole series as a PDF here
[1] This is an update from the 2018 figures in part 4. (I have switched to 2019, because other data relevant to this section are not readily available for 2018.) Percentages in the graph are the share of total global greenhouse gas emissions, based on a provisional estimate for 2019 of 59.1 Gt CO2e (±5.9 Gt) (EDGAR dataset, cited in UNEP, 2020)
[2] Statistics from the 2020 GlobalABC report by the IEA and UN Environment Programme (UNEP)
[3] For built environment professionals, there are tools like the Embodied Carbon in Construction Calculator (EC3)
[4] On the other hand, if they are mis-specified or under-engineered, or if substandard materials are used to save on costs, over-confidence in these materials can be deadly, as shown by appalling cases of collapsing buildings and infrastructure.
[5] See the IEA web site
[6] According to a study I quoted in part 4 (Dominik Wiedenhofer et al, 2021)
[7] According to another study already cited (Barbara Plank et al, 2022)
No comments