Cement and Concrete Research 114 (2018) 2–26

Contents lists available at ScienceDirect

Cement and Concrete Research

journal homepage: www.elsevier.com/locate/cemconres

Eco-efficient cements: Potential economically viable solutions for a low-CO2 T

cement-based materials industry☆
UN Environment, Karen L. Scrivenera, Vanderley M. Johnb, , Ellis M. Gartnerc
⁎

a
École Polytechnique Fédérale de Lausanne EFPL, Lausanne, Switzerland
b
University of São Paulo, São Paulo, Brazil
c
Imperial College (formerly at Lafarge, France), London, UK

ABSTRACT

The main conclusions of an analysis of low-CO2, eco-efficient cement-based materials, carried out by a multi-stakeholder working group initiated by the United
Nations Environment Program Sustainable Building and Climate Initiative (UNEP-SBCI) are presented, based on the white papers published in this special issue.
We believe that Portland-based cement approaches will dominate in the near future due to economies of scale, levels of process optimisation, availability of raw
materials and market confidence. Two product-based approaches can deliver substantial additional reductions in their global CO2 emissions, reducing the need for
costly investment in carbon capture and storage (CCS) over the next 20–30 years:
1. Increased use of low-CO2 supplements (SCMs) as partial replacements for Portland cement clinker.
2. More efficient use of Portland cement clinker in mortars and concretes.
However, other emerging technologies could also play an important role in emissions mitigation in the longer term, and thus merit further investigation.

1. Introduction research programs undertaken by different sectors including the cement

one. Research is also being done to find ways of utilising captured CO2
Cement is the largest manufactured product on Earth by mass. (CCU). However these are technologies that require energy and are still
Combined with water and mineral aggregates it forms cement-based very expensive at this stage. As most cement is and will be produced
materials (e.g. concrete). It is the second most used substance in the and used in developing countries, lower cost alternatives to CCS/U
world after water. These materials make up a substantial proportion of would be highly desirable.
the built environment. Impressively engineered bridges and dams, ar- To seek solutions to this dilemma, in 2015, the United Nations
chitecturally innovative skyscrapers, roads and railways, high-rise Environmental Program Sustainable Building and Climate Initiative
apartments and single-family homes — none of these would be possible (UNEP-SBCI) established a technical working group to review practical
without cement. Whether handled by highly skilled tradespeople or do- alternative technologies for reducing CO2 emissions and increasing
it-yourselfers, concrete and mortar remain cost- and energy-efficient materials efficiency throughout the cement industry value chain, which
construction materials. Fulfilling our ambition for a more equitable and could reduce the need for CCS. The aim of this report is to demonstrate
sustainable world will require substantial expansion of our built en- that there are several potential solutions for CO2 mitigation that are far
vironment, which will in turn increase demand for cement-based ma- less expensive than CCS, that could be used in the short-to-medium
terials. Tackling this objective according to business-as-usual practices term, while significantly impacting net global CO2 emissions — without
would involve an unacceptable increase in CO2 emissions. To mitigate making concretes and mortars too expensive for ordinary consumers,
these emissions, the main solution proposed by the International Energy particularly those in developing countries where the vast bulk of ce-
Authority on 2009 was CO2 capture and storage, (CCS). The technology ment will be produced. Implementing existing-and developing in-
to implement CCS has improved since 2009, thanks to extensive novative-mitigation technologies will require significant funding. This

☆
This overview paper is also, in a different format, published as a standalone report cited as: United Nations Environment Programme, 2016. Eco-efficient cements:
Potential economically viable solutions for a low-CO2 cement-based materials industry [core writing group: K. Scrivener, V. John, & E.M. Gartner]. United Nations
Environment Programme, Paris, France. pp. 64. © United Nations Environment Programme, it is reprinted in this special issue by permission. However it does not
constitute an endorsement of this special issue by the United Nations Environment Programme. Moreover, the views expressed in the overview paper and in this
special issue do not necessarily represent the decision or the stated policy of the United Nations Environment Programme, nor does citing of trade names or
commercial processes constitute endorsement. United Nations Environment Programme.
⁎
Corresponding author.
E-mail address: [email protected] (V.M. John).

https://doi.org/10.1016/j.cemconres.2018.03.015
Received 4 February 2018; Received in revised form 8 March 2018; Accepted 22 March 2018
Available online 28 June 2018
0008-8846/ © 2018 Published by Elsevier Ltd.
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

will in turn require R&D and technical transfer incentives, a greater 5000 8000

portion of which must probably come from developed economies. The

Cement
4500
ultimate long-term benefits, however, will be shared by everyone.
Steel 7000

More than 20 international experts from academia and industry

World Popula!on
4000
took part in this working group (see annex 1). Our focus was materials 6000

Materials Production (Mt)

technology, since renewable fuels and energy have already been ana-
3500

lysed in great detail. The group explored scientifically informed options 5000

Populaon (M)
3000
with the best potential to be scaled up and make a real contribution, as
opposed to niche solutions, and approaches that lack a solid scientific 2500 4000
justification.
2000
This overview is intended for policy-makers from individual coun- 3000
tries and multilateral organisations, industry leaders, research agencies 1500
and NGOs, as well as for researchers.1 It summarises the work of the 2000
group members, which is presented in the other papers of this special 1000

issue. 500
1000
We hope that our findings will motivate the cement industry to
surpass its CO2 mitigation goals without greatly increasing costs or risks 0 0
to the end-user. Our work should also help broaden multilateral agen- 1950 1960 1970 1980 1990 2000 2010 2020

cies' focus on the mitigation potential for the cement industry; inspire Fig. 1. Comparison of cement (data derived from CDIAC [3] and CEMBUREAU
technical education for architects and civil engineers; improve research [4,5]) and crude steel (World Steel Association [6]) production with population
effectiveness by helping researchers to address the most promising is- (UN Population Division [7]).
sues; and guide funding agencies to support priority research.

Copper

2. Cement and modern society Aluminium

Modern developed societies require a built environment that is

Glass

unimaginable without the widespread use of cement-based materials Asphalt

that allow construction anywhere, at low cost, of complex and massive
Lime
shapes from water, gravel, sand and cementitious powder. In the last
65 years, the amount of cement produced increased almost 34-fold, Iron
[1,2], meanwhile the population has increased < 3-fold [3]. This
Ceramic
growth rate is much higher than other commodities such as steel (see
Fig. 1) [4]. The larger per-capita availability of cement is related to Wood
discernibly improved living standards in most of the world.
Cement
In 2015, the total mass of cement produced was 4,6 billion tonnes
[5,6]. This is equivalent to about 626 kg/per capita, a value higher than Cemenous
the amount of human food consumption [7]. With this cement we can 0 2000 4000 6000 8000 10000 12000 14000 16000 18000
produce around 2.1–2.3 m3 or 4.8–5.5 t per capita of cement-based Materials producon (Mt/year)
materials.2 In 2005, cement-based materials represented about 30% of
the total global materials use including fossil fuels [8,9]. In 1950, it Fig. 2. Estimated consumption of common materials 2002–2005. Data from
[10,11]. Except from ceramics, asphalt and cementitious, all other materials are
represented only 7%. This is a 4-fold increase in proportion in only
shared with other industries. Part of wood is used for fuel and furniture.
55 years.
Fig. 2 illustrates how the amounts of other common building ma-
terials — such as wood, steel, asphalt, and brick — are very small in makes it difficult to modify cements to any great extent to adapt them
comparison to concrete. These materials generally have a higher en- to local economic and environmental needs. This is a significant ob-
vironmental footprint than cementitious materials. But beyond en- stacle to maximising sustainability. If cements could be more easily
vironmental considerations, the production of these materials could not adapted to local raw materials and to specific applications, this could
be sufficiently increased to replace concrete to any significant extent. considerably reduce environmental impact.
Additionally, the current production cost of cement is relatively low; a Our civilisation has become dependent on the availability of this
tonne of bulk cement in Europe and North America typically costs USD cheap mineral binder that hardens quickly in almost every habitable
$100–120, and less than $50 in China. This comparatively low cost and environment to produce a wide variety of 3-D structures, and is simple
high volume means that it is essentially a local material produced close enough to use by unskilled, even illiterate, self-help builders. Cement is
to the site where it is used, as the cost of overland transportation rapidly so common that we take it for granted. Without decisive action, the
becomes uneconomic.3 However, unlike other locally produced mate- beneficial impact of this vital material could fail to be realised.
rials, cements are treated as commodities and subject to stringent, The World Bank [10] shows that in low-income countries 65% of
sometimes inappropriate, national and international standards. This the urban population lives in slums, > 60% do not have access to sa-
nitation, and 35% do not have a safe water supply. About 40% of the
world's population lives in these low-income countries, where the
1
The overview report is also available as a separate document at UN growth in population is also fastest. A more equitable world demands
Environment home-page (www.unenvironment.org). these deficiencies be remedied and this will require a substantial in-
2
Assuming 1 kg of cement is mixed with 6–7 kg of aggregates, 0.6–0.7 kg or
crease in cement production. It is thus important that the price of ce-
water and neglecting the small amount of organic admixtures. Average density
ment is affordable for poor communities.
was assumed to be 2.3 t/m3.
3 Fig. 3 shows the IEA's forecast of cement demand up to 2050 [11].
To give an idea, the zone around a cement plant which can be economically
supplied if around 100 km by road, 300 km by rail, but cement can be trans- Even in the high-demand variant, the growth in cement production is
ported by sea across and ocean for roughly similar cost. only about the same as the rate of population increase. Considering that

3
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

7000 100%

6000 22
80% 37
5000 Other 8
Producon (Mt)

4000 60%
22
India
3000
40% 59

2000
China 32
20%
1000

OECD 11 10
0 0%
2015 2020 2025 2030 2035 2040 2045 2050 2015 2050

Fig. 3. IEA high-consumption scenario for future cement consumption by region [12]. The lower section shows the evolution of cement production distribution
among different regions.

demand grew at a rate 10 times higher from 1950 to 2015, it is possible, emissions has been rising steadily and is now estimated by some sources
and perhaps even socially desirable, that future production will surpass to be around 10% [13], or about 6% of the total anthropogenic
these values. About 90% of cement is currently, and will continue to be, greenhouse gases (GHG) [15]. This has occurred despite the important
produced in non-OECD countries. The proportion of world production improvements in production efficiency and emissions mitigation efforts
in China has already peaked and is expected to diminish from > 50% of the cement industry since the 1970's.
today to around 30% by 2050. Cement production has to grow to meet the demand for decent built
Growth is forecast to be concentrated mainly in the developing non- environment from citizens in developing, low-income countries.
OECD countries that combine population growth with a quantitative However, according to the WWF/Lafarge Report [14], in a business-as-
and qualitative deficit in the built environment. If new solutions for usual (BAU) scenario, CO2 emissions from cement production were
cementitious materials are to be adopted on a significant scale, they expected to increase 260% between 1990 and 2050. The 450 ppm IPCC
must be low cost and used easily by people with minimal training and mitigating scenario (IEA blue scenario) requires a 50% reduction in
scientific knowledge. anthropogenic CO2 emissions by 2050 [15]. If new methods are not
The traditional form of cement — ordinary Portland cement (OPC) implemented for reducing CO2 emissions from cement production, this
containing > 90% Portland cement clinker — is made from abundant would leave the cement industry responsible for about one third of this
raw material cheaply available almost everywhere. The production target amount in 2050.
process requires grinding and calcining (heating to high temperature) a
mixture of clay and limestone. The resulting intermediate material, 3. Cement industry mitigation strategies, their limitations: the
known as clinker, is ground to a fine powder with 3–5% gypsum added IEA/WBCSD: CSI 2009 roadmap
to form OPC. The production of OPC generates on average 842 kg CO2/t
of clinker [12]. Fossil fuel combustion is responsible for < 40% of total The cement industry was active in pursuing strategies to reduced
CO2 emissions, while limestone decomposition (CaCO3 or CaO·CO2) CO2 emissions long before global warming became a priority. Since
during calcination is responsible for the remainder. This is what makes 1999, with the launch of the Cement Sustainability Initiative (CSI) at
CO2 emissions from cement manufacture so different from the emis- the World Business Council for Sustainable Development (WBCSD), the
sions produced simply by burning fossil fuels for energy production. industry has systematically collected evidence and improved its stra-
Increasing energy efficiency is not enough to significantly impact tegies. In 2009 [15], the IEA/WBCSD Roadmap proposed several CO2
emissions. Calcination of limestone must also be minimised, which will emissions and mitigation scenarios [16]. The IEA study found that the
change the composition of the cementitious products. target 50% global emissions reduction goal to keep global warming
Due to the enormous growth in cement demand in the developing at < 2 °C of pre-industrial levels would require an overall reduction of
world, the share of cement production in total anthropogenic CO2 18% in the CO2 emission of the cement sector (compared to a 2006

1400 1400

1200 1200

1000
CO2 Migaon (Mt)

1000
Fuels
800 800
Energy efficiency
600 600
Clinker Substitutes
400 400
CCS

200 200

0 0

Fig. 4. Cement sector CO2 emissions reductions below base line 2006–2050, for low demand (right) and high demand scenarios. Adapted from [Tam, C and Van der
Meer, R, IEA/CSI technology roadmap for the cement industry, 6th International VDZ Congress 2009, pp. 155–157].

4
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

baseline) by 2050. Fig. 4 shows the CO2 emission reduction scenarios 25%

from the Roadmap and the contribution of each of the mitigation

strategies, grouped under the headings of fuels; Energy efficiency;
Clinker substitutes; Carbon capture and storage (CCS). The Roadmap 20%

also discussed other technologies, such as novel cement types, which at 4%

the time were considered too far from practical application to be in-
cluded in the quantitative model.
Others
15%
5% Puzzolana
Fly ash

3.1. Energy efficiency Slag

10% Limestone
Gypsum
Major efforts to increase energy efficiency began after the energy 7%
crisis of the 1970's. A state-of-the art dry-kiln with pre-calciner con-
sumes about 50% less energy than a long wet kiln typically used at that 5%

time [1,17]. The theoretical minimum energy consumption is ~1.9 GJ/t

[15], which means state-of-the-art kilns already achieve about 63%
efficiency, making such kilns probably today's most efficient thermal 0%
1990 2000 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
machine in wide-scale industrial use. It is unlikely there will be sig-
nificant gains in best available technology (BAT) [16], but the pro- Fig. 5. Clinker substitution evolution from companies from the CSI WBCSD
gressive upgrade of old technology, where economic, was estimated in [14]. The evolution of the clinker fraction can be deduced by adding 4–5%
(added as calcium sulfate, e.g. gypsum) in the total fraction of SCMs.
the IEA/WBCSD Roadmap to provide about 10% (1.8% absolute) of the
targeted 18% CO2 emission reduction [17]. This strategy would require
heavy investment, but, since it reduces energy costs, it should not in-
crease the cost of cement. 3.4. Carbon capture and storage or use

The IEA/WBCSD Roadmap introduced carbon capture and storage

3.2. Fuels (CCS) as the main strategy to reduce CO2 emissions in an industry that
it was expected to be growing: 56% of the planned CO2 reduction by
The modern cement kiln is a very flexible machine, which allows the 2050 would be due to such a strategy (Fig. 4). The report estimated it
cement industry to change fuels relatively simply. The Brazilian cement would require between US$ 321 to 592 billion [17] in investment to
industry, for instance, changed from almost 100% fuel oil in the 1970’s capture that fraction of the CO2 from the cement industry. Since 2009
to a mixture of charcoal (~40%) and coal (~50%) in 1984 and now the feasibility of CCS has been extensively studied [18] including sev-
relies almost entirely on petroleum coke [18]. In Europe, the use of eral pilot schemes. Fig. 6 shows the estimated cost of CCS over the next
wastes as fuel can be as high as 80% of the thermal demand. Increased few decades. These costs would increase the marginal cost of clinker
use of waste fuels is an efficient means for their disposal, providing a production 2–3 times. Much of the expense results from the large
useful and ecologically responsible service to society. This flexibility is amount of energy required to drive the CCS process.
an opportunity to reduce CO2 intensity; there are opportunities to ex- Carbon Capture and Use (CCU) is an alternative to CCS that replaces
pand the use of biomass and alternative fuels worldwide. The Roadmap underground storage by another industrial step that transforms CO2
expected the worldwide use of “alternative fuels” to grow from 3% in into commodity chemicals or construction products. CCU has the po-
2006 to about 37% in 2050 and deliver around 15% of the targeted tential to be significantly less expensive than CCS, provided there is an
overall reduction in CO2 emissions. adequate market for the resulting products. Global demand for specific
commodity chemicals requiring CO2 as a raw material (e.g. formic acid)
3.3. Clinker substitution by mineral additions/supplementary cementitious is often small enough to be satisfied by the emissions from only one
materials cement or power plant. Another method being researched is trans-
forming captured CO2 into methane (methanisation) using renewable
Another important and well-established strategy is the replacement energy. However, this technology is a long way from being economic-
of clinker with other materials. This strategy has the advantages of ally viable at present. Another possibility is the mineral capture of CO2
reducing energy consumption as well as increasing production without that has the potential to permanently capture globally significant vo-
requiring new kilns. lumes of CO2 to make useful construction products. Global demand for
The most common clinker substitutes (Fig. 5) are reactive by-pro-
ducts from other industries: granulated blast furnace slag (GBFS), a by- 120
product of pig‑iron production in blast furnaces, whose use in cement post-combus!on
dates from before 1900; and fly ash (FA), generated by burning coal to 100 capture
produce electricity. However, the most common supplementary ce-
mentitious material is the almost inert limestone filler.
Fig. 5 shows the evolution of clinker substitutes over the past
80

25 years for companies in the CSI's GNR database. It shows that the oxy-fuel
€/tCO2

sucessfull demonstra!on

process
level of clinker substitution is levelling off. This corresponds to the low 60

estimate of the contribution of clinker substitutes to further CO2 re-

duction shown in Fig. 5. This arises from the fact that the supply of the 40
most desirable clinker substitutes — particularly blast furnace slags and
coal fly ash of adequate quality — is rather modest compared to total
projects

20
cement production. In 2006, (the baseline for the IEA study) a very high
proportion of these substitutes were already used in cement or concrete. 0
If new sources of good quality SCMs become available this picture 2020 2025 2030 Long term
would change significantly. Given the importance of this strategy it is
discussed more extensively in Section 8. Fig. 6. Estimated cost of carbon capture and Storage (CCS) source ECRA.

5
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

such products far exceeds that of commodity chemicals. Solidia cements review by all members. However, ultimately they express the views of
are an example of this approach. the authors not necessarily shared by all group members. The papers
Regardless of the technological challenges, CCS and CCU would are published in this special issue of Cement and Concrete Research, to
significantly impact cement production costs, affecting the price of broaden discussion within the scientific and technical community.
cement-based materials and the structures built with them — including The white papers have been consolidated in this overview to inform
housing and infrastructure. These costs would have serious social im- a wider readership. The report and the white papers were discussed and
plications in developing countries. commented on by WBCSD CSI member companies and UNEP-SBCI of-
ficials. Our objective was to produce a document that reflects a broad
3.5. Beyond the 2009 roadmap consensus. The resulting report represents the group's majority view-
point.
In 2013, the CSI launched its first regional roadmap in India. This By including a wide cross section of contributors we believe that the
targeted 210 Mt of CO2 reductions compared to a business-as-usual technologies considered here cover all imaginable solutions. Unlike the
scenario, and further projected a target clinker factor of 58% by 2050. WBCSD/IEA study, we did not restrict ourselves to technologies which
In 2014, it initiated the Brazilian roadmap. could be implemented at the cement level, but also considered solutions
In 2015, at the COP21 in Paris, a landmark government level at the concrete level. Broadly speaking, solutions at the cement (or
agreement was reached aimed at restricting increases global average binder) level are the most practical to implement as they can be put in
temperature to well below 2° above pre-industrial levels and at pur- place by individual cement companies. The solution is “in the bag” — it
suing efforts to limit the temperature increase to 1.5 °C. The WBCSD has can be used by anyone, even in the most rustic situations. Concrete
issued a Global Statement of Ambition calling for collaborative efforts solutions generally demand more sophisticated methods for im-
from all cement companies to reduce CO2 emissions by 20–25% by plementation — for example production of concrete in a ready-mix
2030 — beyond business-as-usual. The global roadmap, discussed plant. There are also solutions at the structural level, but these ne-
above, will be updated. cessitate the cooperation of designer and builder. While we do not ig-
While recognising these on-going efforts, this working group takes nore the potential for CO2 savings at this level, this has not been ana-
the quantitative estimations of the 2009 Roadmap as a starting point. lysed in detail for this report. The working group's key findings at the
We recognise, but do not discuss further, the process related mitigation binder- and concrete-levels are discussed in the following sections.
strategies of alternative fuels and energy efficiency. Our efforts have A CO2 mitigation model was developed, using data from the white
instead focused on the possibilities to increase the range and supply of papers and report findings along with relevant public data for emission
clinker substitutes and at any other technologies that have potential to factors and projections for cement production.
reduce the need for CCS or CCU and the high costs these may entail.

4. Our working method 5. Overview of cementitious materials use

Our group of experts in the field of cementitious materials and en- We tend to assume that concrete is the principal material made from
vironmental assessment began by identifying promising materials-re- cement, but analysis of the data indicates that its use in concrete ac-
lated approaches. Rather than focusing only on the end-of-pipe aspect counts for less than half of cement consumption. It is possible to make a
of the cement production process, it was decided to look for opportu- relatively robust estimate of the amount of cement used in reinforced
nities over the entire life cycle of cement, including, cement applica- concrete; we have good figures for the global production of reinforcing
tions and recycling. Some of the aspects have been explored system- steels and for the average quantity of reinforcement used in concrete.
atically for the first time in this report. Fig. 7 shows the global figures with some breakdown by regions. The
In most of these fields a scientific, state-of-the-art white paper (list regional figures are less reliable as the figures for cement and steel
in Section References) was written by group members and their colla- relate to production rather than consumption. They do not account for
borators. The white papers primarily consolidate the available scientific importation and exportation to other regions between production and
knowledge, and when possible, integrate market knowledge to develop use. Nevertheless, it is clear that the proportion of cement used in re-
a consensus for estimating each technology's mitigation potential. To inforced concrete globally is only around 25% of the total.
ensure a comprehensive overview of the challenges for introducing new It is difficult to get precise information about the other 75% of ce-
technologies in the typically conservative construction sector, a ment use. Fig. 8 shows the breakdown for Brazil; and South Africa is
common template was employed addressing these criteria: similar. While it was not possible to obtain reliable figures for all

0.4
− Description of the technology, its degree of development, scope of
Cement in Reinforced Concrete (t/t)

application and robustness when used in different climates and by

people without formal training; 0.35
− Overview of the durability of the solution, a crucial aspect for in-
creasing the resource efficiency of construction; 0.3
− State of the development and research needs;
− Assessment of the scalability potential, including raw material 0.25
availability. Mitigating substantial amounts of CO2 will require so-
lutions that can potentially be deployed on a large scale in different
regions; 0.2

− Evaluation of investment and production costs in comparison with Global Brazil

Portland cement; 0.15
China EU
− CO2 mitigation potential;
− Barriers and incentives predicted for the introduction of the tech- 0.1
USA
nology in relevant markets; 2002 2004 2006 2008 2010 2012 2014
− Research priorities to further develop the technology.
Fig. 7. Fraction of cement used in reinforced concrete. Worldsteel 2014,
These papers were discussed during the group meetings and open to Cembureau 2014.

6
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

materials to calcium carbonate. Some researchers [e.g. [21]] have ad-

Mortar
vocated that this process should be considered to offset the production
40%
emissions. This has not been widely supported because the rate of
Reinforced carbonation of normal concrete is very slow — complete carbonation of
a typical wall would take decades. However, this process would be
Concrete
25%
Concrete much more significant for mortars used in thin layers, such as renders,
40% which will carbonate in a few years.
Plain Only a minority of the 25% of cement used with steel will have any
Concrete significant risk of carbonation-induced corrosion. Carbonation rates
reach a maximum at relative humidities around 60%. This RH range,
15%

typical of indoor concrete, is much too low for active steel corrosion to
Other be problematic. On the other hand, carbonation is slow at the very high
20% RHs needed for active corrosion. From an environmental standpoint it
Fig. 8. Cement use in different sectors in Brazil. would be much better to use concretes containing higher calcium with
higher CO2 emissions only in situations where carbonation corrosion is
a serious risk. For instance, where periods of intermediate humidity
countries, Brazil and South Africa exemplify countries at an inter-
(favouring carbonation) are interspersed with periods of high humidity
mediate development level. While cement use in reinforced concrete is
(favouring corrosion).
undoubtedly higher in developed countries‡, it is also probably lower in
Despite the fact that only a small proportion of cement is used with
countries with a low level of development.
reinforcement, there is strong market attachment to the idea of “general
A comparable amount of cement is used in mortar to that used in
purpose” cement — meaning a cement which can be used in all ap-
concrete. The term “mortar” covers cement-based materials prepared
plications because it makes life simpler for the suppliers and requires a
on site, such as renders and mortars for bricklaying. The sector marked
lower level of care and knowledge on the part of the user. The “general
“other” in Fig. 8 is presumed to be largely accounted for by what can be
purpose” concept, especially with respect to carbonation risk, is built
referred to as “cement-based products,” including blocks, pavers, and
into most modern cement standards. The main exceptions to this are
roof tiles. In developing countries, such cement-based products are used
applications in relatively rare environments — such as high sulfate soils
extensively in social housing, so increased demand will be needed to
— where specially adapted cements are still available. We believe that
meeting development goals.
this “general purpose” approach to cement standards is a serious barrier
This breakdown of cement use is relevant to concerns about car-
to environmental optimisation because it requires that the majority of
bonation resulting from using materials with lower CO2 emissions to
cements be carbonation resistant, when this represents a rather small
protect reinforcement. Normally, the high alkaline (pH) environment
fraction of real-world applications.
inside concrete protects steel from corrosion. Two factors may change
If cement standards could clearly designate a specific category for
this protective situation — the ingress of chloride ions and the lowered
use with steel in concrete, such cements would almost certainly sell at a
pH resulting from carbonation. The former is by far the most wide-
premium price because of the higher energy cost for making them. This
spread problem facing reinforced concrete worldwide.
would discourage their use in the 75% of applications for which do not
Carbonation is not problematic for the concrete itself, only for the
need this characteristic, and for which cements with a much lower
steel inside it. Indeed, carbonation of OPC concrete increases its
carbon footprint might be better suited.
strength. There are concerns that alternative materials containing in-
creased amounts of ettringite would be weakened by carbonation, due
6. The limitations of earth chemistry
to decreased volume of solid products. Experimental studies on BYF
cement indicate this is not the case [19].
Portland cement did not become the earth's most used material by
CO2 absorption by cement-based materials can actually be con-
chance. The processes of nuclear fusion in stars and in planet formation
sidered a natural form of carbon capture and storage [20]. The cement
have yielded the 8 elements — oxygen, silicon, aluminium, iron, cal-
carbon cycle is shown in Fig. 9. “Chemical” CO2 is emitted during
cium, sodium, potassium and magnesium — that make up > 98% of the
production from limestone decomposition (CaCO3 → CaO + CO2); but
earth's crust (Fig. 10). The vital elements of hydrogen and carbon can
the resulting hydration products can then react with the same amount
be added to this list. These are very abundant in the seas and atmo-
of atmospheric CO2, restoring the calcium in the cement-based
sphere, and so are also commonly found in surface minerals. Minerals
containing other elements are not available in the quantities needed to
supply the global demand for cementitious materials. Phosphorus is one
Clinker phases + Other
important example; it is an element essential for life and also chemi-
containing hydrates cally suitable for use in hydraulic cements. Global reserves of phos-
Hydra on phorus ores are barely sufficient for its primary use as a component of
CaO Ca(OH)2 fertilisers. So, any significant diversion of phosphorus into the con-
struction sector would present an enormous problem for the sustain-
ability of modern agriculture!
Because of the high volumes of cement used, the limited availability
CO2 of most elements is a major constraint to practically viable cement
CO2 chemistries. However, it also means that an exhaustive analysis and
exploration of alternatives can be made. Since cements are basically
composed of oxides we can consider the oxides of silicon, aluminium,
iron, calcium, sodium, potassium and magnesium, and their potential to
CaCO3 form hydrates with cementing properties. The basic principle of hy-
draulic cements is shown schematically in Fig. 11. First, discrete cement
Fig. 9. The Cement/Concrete carbon cycle, “chemical” CO2 emitted during particles are dispersed in water. Cement particles (grey) dissolve in the
production from limestone breakdown is proportional to CO2 reabsorption ca- water and then hydrates (shown in red) are precipitated from the
pacity during carbonation. aqueous solution. To work as cements these criteria must be met:

7
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

Mg rest <2% section). From the standpoint of CO2 emissions, the most important
characteristic of these minerals is the calcium content. The calcium
Na comes from calcium carbonate (limestone) and the first step of produ-
K cing clinker is the decarbonation of the limestone:

Ca CaCO3 CaO + CO2 .

Fe This is the chemical reaction that accounts for some 60% of CO2
emissions from the manufacture of traditional Portland cement. Since
no large-volume concentrated sources of calcium exist other than
limestone, the manufacture of calcium-based cements inevitably leads
Al O to substantial “chemical” CO2 emissions associated solely with the
decarbonation reaction, and not with the fuel burned in the process. It
is exactly for this reason that the cement industry is such a significant
CO2 emitter. This also a cause for optimism — it is clearly possible to
reduce these emissions in a relatively inexpensive way, simply by
changing the composition of cements.

7. Cements made from “alternative clinkers”

Si
The white paper by Gartner and Sui, offers a comprehensive dis-
cussion of the alternatives to Portland cement clinker as the basis for
hydraulic cements. This paper's central conclusions are presented here.
Fig. 10. The abundance of elements in the earth's crust As previously explained, the most viable chemistries for practical hy-
[Source: Wikipedia]. draulic cements derive from the CaO-SiO2-Al2O3 system. The only mi-
nerals in this system with significant hydraulic activity are the high-
lime calcium silicates, C3S (alite) and C2S (belite), and the calcium
aluminates, C3A, C12A7, CA and C4A3$ (ye'elimite). Assuming that the
− The hydrates must have a higher volume than the dissolving cement.
main calcium source is limestone, the amounts of chemical CO2 re-
− The ions forming the hydrates must be able to migrate from the
leased by raw materials' calcination to create the clinker minerals found
original particles into the previously water-filled space.
in Portland cement clinker, or other clinkers discussed in this section,
− The hydrates themselves must have low solubility to remain stable
are compared in Table 1, which reveals the very significant differences
over long time periods.
among them.

Hydrates formed primarily from the alkali metals sodium and po-
tassium have very high solubility. These ions stay in solution and 7.1. Belite-rich Portland cement clinkers
contribute little to the strength-giving hydrates. On the other hand, the
ions of iron and magnesium have low mobility in alkaline solutions, so Belite-rich Portland clinkers are produced with the same process as
hydrates from these elements are mainly precipitated within the ordinary Portland cement clinkers, but with less limestone in the
boundaries of the original cement grains and make little contribution to clinker raw material mix, so CO2 generation is reduced. However, this
filling the previously water-filled space. From the standpoint of hy- emission reduction of around 10% is rather modest relative to OPC.
draulic cement, this means that the most important oxides are those of Belite-rich Portland cements and clinkers are chiefly covered by ex-
silicon, calcium and aluminium — which make up about 90% of a ty- isting cements norms and so should not be considered as a new class of
pical Portland cement. In this lime (CaO), silica (SiO2), alumina (Al2O3) cement. A key reason they are not currently widely used is that they
system the only reactive minerals are the calcium silicates and the gain strength much more slowly than most OPCs. Such cements are well
calcium aluminates or sulfo-aluminates (discussed further in the next suited for niche markets where the strength gain after a few days is not
critical. They are mainly employed for reasons of their low heat of

Cement grains Water Hydrates

Fig. 11. Schematic of cement hydration.

8
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

Table 1 equivalent to a filled activated belite cement with good bonding be-
Clinker compound and associated chemical CO2 emission. tween the filler and the hydrates. The overall manufacturing process is
Clinker compound Chemical CO2 complex due to the need for more processing steps than required for
emissions (kg/tonne) OPC production — specifically the preparation of lime, grinding of si-
lica sources, blending, autoclaving, low-temperature drying, and
Alite (C3S) [typically, > 60% of Portland cement clinker] 579
blending/grinding with fillers. Because these approaches are still under
Belite (C2S) 512
Tricalcium aluminate (C3A) 489
development at the laboratory level, no reliable estimates can yet be
Tetracalcium alumino-ferrite (C4AF, “Ferrite”) 362 made for their overall energy- and CO2-efficiencies in a real-world in-
Quicklime (CaO) 786 dustrial context. Simple thermodynamic arguments, however, imply
Wollastonite (CS) [a major component in Solidia clinkers] 379 that manufacturing the reactive belite component is itself unlikely to be
Ye'elimite (C4A3$) [made with CaSO4 as sulphur source] 216
significantly more CO2-efficient than producing an equivalent amount
Periclase (MgO) [made from magnesium carbonate] 1100
Periclase (MgO) [made from basic magnesium silicate rocks] 0 of belite in a belite-rich Portland Cement clinker. Thus, the main in-
terest of this type of binder appears to lie in the very significant increase
in the rate of strength development relative to what is currently pos-
sible with equivalent binders made from belite-rich Portland cement
hydration in the construction of massive concrete dams and founda- clinkers, and the resultant increased level of dilution with low-CO2
tions. Over the past 15 years, belite-rich Portland cements have been fillers that may be made possible by this increased reactivity.
used in concrete engineering projects in China.
7.4. Magnesium-based cements
7.2. Belitic clinkers containing ye'elimite (CSA)
Hydraulic cements based on magnesium oxide (MgO) have recently
A promising lower-carbon alternative to belitic Portland clinkers is been claimed to offer great potential for reducing CO2 emissions.
belitic clinkers containing ye'elimite (also known as calcium sulfoalu- However, most of the research has been done with MgO produced by
minate or CSA). These clinkers can be made in conventional Portland calcination (decarbonation) of magnesium carbonates, for which the
cement plants, requiring only changes in the proportions of the main CO2 emissions are extremely high (see Table 1) so this approach is
raw materials — for example, less limestone and more aluminium clearly unsustainable. Nonetheless, there is still some chance for
sources. The CO2 emissions associated with making such clinkers de- breakthrough in the area of MgO-based clinkers made using globally
crease as their ye'elimite content increases, but unfortunately their cost abundant ultramafic rocks (basic magnesium silicates) instead of
also increases very significantly at the same time because higher limestone as the main raw material. Because these rocks are rich in
ye'elimite contents require more expensive aluminium-rich raw mate- basic MgO but contain no CO2, they have the inherent capacity to
rials. The high cost of high-ye'elimite clinkers is the main reason why capture CO2 as stable magnesium carbonates. In theory, this approach
modern CSA cement technology, developed primarily in China over the has the advantage over limestone-based technologies because it could
last 4 decades, is still restricted to specialty niche applications where be truly carbon-negative if enough magnesium carbonate forms in the
the additional cost can be justified. Recent research in Europe has fo- resulting hardened binder. As yet, no viable energy-efficient industrial
cused on “Belite-Ye'elimite-Ferrite” (BYF) clinkers, in which belite is the manufacturing process has been invented, although there has been
major component, ye'elimite content is below about 35%, and ferrite some recent progress [23]. We consider this area merits further re-
(C4AF) levels are also significant. Manufacture of such clinkers there- search based on its potential for substantial CO2 reduction — but only
fore requires much smaller amounts of the most expensive aluminium- in the very long term, owing to the difficulty associated with developing
rich raw materials than conventional CSA clinkers. BYF clinkers [e.g. and implementing the necessary process methodology.
“Aether” or “Ternocem”] have the potential to replace Portland cement
clinker in cements formulated for many major applications, with CO2 7.5. Carbonation-hardening cements
savings of 20% or greater per unit of clinker in the cement due mainly
to the lower limestone content of the kiln feed. BYF cements could also There has been a considerable research on the manufacture of
replace Portland-slag cements in many applications, thus supple- concrete products by carbonation instead of hydration. Partial carbo-
menting the very limited global supply of granulated blast-furnace slag, nation curing of conventional Portland clinker-based concretes is al-
which is itself a very valuable low-CO2 hydraulic binder. The major ready used in some precast concrete plants — making use of waste flue
barrier to commercialisation of current BYF technology remains its high gases, etc. — providing a small strength boost compared to ordinary
raw materials costs. An exception to this is cases where cheap alumi- humid curing. But minimal CO2 is consumed in this way. What is new is
nium- and iron-rich raw materials, like bauxite wastes, can be used. BYF the development of special calcium silicate clinkers (CCSC) made spe-
cements seem unlikely to be produced on a large scale until the cost of cifically for carbonation curing (Solidia, USA). These clinkers, com-
emitting CO2 becomes significantly higher than it presently is. prising low-lime calcium silicate minerals such as wollastonite, can be
Nevertheless, the BYF approach is still much less expensive than CCS. made in conventional cement kilns using common raw materials
(limestone and silica) and are no more expensive to make than ordinary
7.3. Hydraulic calcium silicate clinkers manufactured by hydrothermal Portland cement clinker; in fact, energy costs and CO2 emissions are
processing lower due to lower limestone contents in the kiln feed. These clinkers
are too unreactive to harden by hydration and can only be cured rapidly
At least two research groups are currently trying to develop CO2- in an atmosphere of almost pure CO2, with controlled relative humidity
efficient approaches to the manufacture of belite-like hydraulic binders well below 100%. This requires some modification of the concrete
by hydrothermal processing [21,22]. At the heart of these approaches is curing chambers typically used for precast products, to allow for CO2
the observation that a hydrated calcium silicate compound (α-C2SH) gas circulation (at atmospheric pressure) and condensation of any
can easily be made by low-temperature autoclaving of lime-silica evaporated water from the fresh concrete. Solidia cements have re-
mixtures. The α-C2SH can then be activated — and at least partially cently been commercialised for fabricating certain non-reinforced pre-
dehydrated — by intergrinding with hard fillers (Celitement) and/or cast cement-based products, but the CO2 gas for curing currently comes
heating at low temperatures to give a material which is very close to from industrial gas suppliers. The long term goal is to use recycled in-
belite (C2S) in composition though far more reactive. In the case of dustrial CO2 from industrial flue gases, which can be considered to be
intergrinding with hard filler, the resulting reactive material is part of the market for CCU, as discussed in Section 3. The global

9
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

Silica fume Used − Widespread availability of raw materials.

Available
− Ease of use enabled by workability time before setting.
Waste glass
− Confidence in long term durability based on wide-ranging and
Vegetable ashes prolonged usage.
Natural Pozzolan

Slag
In the light of this, a very effective strategy to reduce CO2 emissions
is to substitute some of the Portland cement clinker with other materials
Fly ash
(as already discussed in Section 3). These are known variously as mi-
Portland cement neral additions or supplementary cementitious materials (SCMs), and
Filler also include almost inert materials, which may also be called fillers.
Here we will use the term SCM for materials reacting to some extent.
Limestone is usually regarded as a filler, though it is now clear [24,25]
Calcined Clay

0 1000 2000 3000 4000 5000 6000 that it can react with available alumina. For this reason, there is in-
creasing interest in coupled substitutions of limestone with alumina-
Mt/yr

Fig. 12. Use and estimated availability of possible SCMs and fillers. Actual rich SCMs, as discussed later in this section.
possible use will depend on logistics, exact chemical and mineralogical com- As shown previously in Fig. 5, just three materials: limestone,
position, contamination, local availability of other raw materials, etc. granulated blast furnace slag (GBFS), and fly ash (FA) presently con-
stitute the overwhelming majority of mineral additions. The IEA/
effectiveness of this approach will depend on the extent to which a WBCSD Roadmap identified limited potential for further CO2 reduction
circular economy for CO2 develops. Due to the need for specialised by clinker substitutes because of the limited supplies of slag and fly ash
curing procedures, and the fact that the hardened cement does not (Section 3). However, new sources of SCMs would radically change this
protect steel against corrosion, this approach seems best suited for the situation. Fig. 12 shows the estimated availability of possible SCMs and
unreinforced precast cement-based products market — specifically, fillers in comparison to the amount of cement produced.
with products made in factories where curing conditions can be prop-
erly controlled. 8.1. Blast furnace slags
The carbonatable calcium silicate clinker referred to above must
usually be manufactured, because there are insufficient amounts of Granulated blast furnace slag can be substituted up to high levels
naturally occurring carbonatable calcium silicates available. Their (70% is common), but the amount of blast furnace slag available
manufacture requires the decarbonation of limestone in the cement globally is only around 330 Mt/year. This availability has decreased
kiln, so the CO2 captured during curing only represents that part which from 17% of cement production in 1980 to only 8% in 2014. Despite the
was released by decarbonation of the limestone in the kiln, and not that fact that growth in steel production is projected at about the same pace
part resulting from the fuel combustion. There has also been a con- as cement production, blast furnace production of iron and slag is ex-
siderable amount of published research on the carbonation of pected to diminish. This is due to the increased availability of scrap
high‑calcium industrial wastes with poor cementitious properties, in- steel for recycling and the introduction of more efficient steel-making
cluding steel slags. However, simple calculations show that the most technologies. Over the long term, blast furnace slag availability is ex-
energy- and CO2-efficient way to use the uncarbonated calcium in these pected to be below 8% of cement production. Furthermore, iron pro-
wastes is to substitute them for raw material components in the kiln duction is concentrated predominantly in industrialised countries,
feed for making Portland cement clinkers. This can already be done conversely to where the demand for cement is expected to grow most —
today, provided that the wastes can easily be transported to a cement in developing countries.
plant. To be effective as SCMs, blast-furnace slags must be quenched ra-
Additionally, the carbonation-curing approach could be applied to pidly from the liquid state, usually with excess water in a granulator.
MgO-based cements made from abundant natural magnesium silicates Thus, the actual SCM used is a distinct industrial product, “granulated
(MOMS as discussed above) with even greater benefits for net CO2 blast-furnace slag” (GBFS). Granulators are installed and operated in
capture, because in this case the raw materials contain no chemical iron-making factories dedicated to producing GBFS as a by-product sold
CO2, so the CO2 captured during curing would represent a true net CO2 at a profit as an SCM. Because this involves capital investment, GBFS
capture as soon as it exceeds the CO2 emitted by the combustion of the should not be considered a waste product — it often sells for appreci-
fuel required to drive the manufacturing process. ably more on the open market than it costs a cement maker to make
Finally, we mention the possibility of making binders from pre- Portland clinker. Before GBFS became valuable for use as a cement,
cipitated calcium or magnesium carbonates derived from natural brines blast-furnace slags were typically air-cooled, involving minimal in-
— for example, from deep aquifers. An American company (Calera) has vestment, and then often crushed and sold as hard dense aggregates
demonstrated this process. Besides requiring high electrical energy with much less value than GBFS.
input, this method is problematic because the calcium and magnesium Currently, > 90% of blast furnace slag is already used as an SCM
ions in the resulting brines occur almost entirely as chlorides — so it either in cement blended at cement plants or as an addition to concrete
produces an equivalent amount of a chloride-rich waste stream (dilute or other cement- based mixes [12,26]. For these reasons, there is little
hydrochloric acid) which in turn presents a significant disposal problem potential for further CO2 reduction from the use of blast furnace slag.
because there are insufficient large-scale uses for this by-product.
8.2. Fly ash
8. Extending clinker substitution with mineral additions/
supplementary cementitious materials Fly ash results from coal combustion in power plants, and so may be
truly regarded as a waste product. There are greater amounts than slag
The preceding section demonstrates that cements based on Portland available, around 900 Mt/yr, but the quality is very variable, such that
cement clinker will continue to be dominant for the foreseeable future. only about one third of this amount is currently used in cement and
Such cements have the following advantages: concrete. There is probably some scope for increasing this proportion,
through better characterisation and classification. Converting un-
− Economy of scale of production and optimised processing, affecting reactive fly ash into reactive material by adjusting the chemistry is
both cost and energy requirements. unlikely to be economically viable.

10
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

It should also be considered that burning coal to produce electricity ash results in a silica-rich pozzolan, with high chemical reactivity
is by far the largest source of anthropogenic CO2 and in some countries thanks to its high surface area. The performance of RHA may be very
coal fired electricity production is being phased out. On a global-level, good, despite its very high surface area, which in some applications
coal will continue to be a substantial part of the energy mix in the could increase water demand for good flowability.
medium term — largely due to a lack of incentives for using alter- Several obstacles impede more widespread use of RHA and other
natives. However, it bears noting that the recent availability of shale vegetable ashes in cement and concrete — notably, seasonable and
gas in North America has led to a shortage of fly ash there. geographical variability, and the difficulty of producing reactive ashes
Since fly ash and blast furnace slag are by-products, the availability while at the same time exploiting the agricultural wastes for fuel. High
varies regionally. Originally, they were sold at a low price, thereby temperatures and long residence times during burning tend to produce
reducing the cost of cement. This is now changing in many regions due unreactive crystalline quartz. Carbon contamination may also be pro-
to heightened demand. They also were considered to be CO2 free, but blem from incomplete combustion.
allocation of environmental loads is now under discussion. If CO2 re- Furthermore, ashes have other applications, such as soil amendment
allocation discouraged their reuse in cement, this would be counter- (mineral fertiliser), which is generally logistically convenient, and in
productive for total global CO2 emissions. Presently, in many regions, landscaping and other industrial applications [31].
there is a scarcity of these materials and a significant increase in the Globally it is not considered that there is a lot of potential for further
proportion of Portland cement clinker that they can replace in cement is CO2 mitigation through the use of agricultural ashes, although there
not expected. may be some interest on a regional level.

8.3. Natural pozzolans 8.6. Other reactive products

In addition to slag and fly ash, other pozzolans — reactive amor- Any amorphous or imperfectly crystalline material containing silica,
phous or poorly-crystalline siliceous materials from natural sources — alumina and/or lime can be potentially a reactive SCM. Although the
are available in a few regions. The CSI GNR database indicates around dominant sources have been blast furnace slag and fly ash, there are
75 Mt/y of pozzolans are currently used as clinker substitutes. Available others, including natural or industrial residues like slags from steel and
reserves are plentiful, but localised. Reactivity varies considerably. other metals.
Moreover, the angular particle shape and internal porosity of some Among the issues to be considered are available quantities, presence
materials can lead to greater water demand and workability problems. of contaminants, (such as calcium oxide and other minerals that un-
dergo large and slow expansion when exposed to humidity), elevated
8.4. Calcined clays alkali levels, and the existence of competitive uses for the product, as in
the case of waste glass. These issues, combined with logistical costs and
The practice of partially substituting calcined clay for clinker has availability of other alternatives in each region, effectively limits actual
been known for a long time — it was used in 1932 for bridge con- use of such materials. Addressing environmental concerns in the
struction in San Francisco, USA [28], as well as in many of Brazil's large coming decades will necessitate using different SCMs according to local
dams. Since the 1970's Brazil has a constant production of calcined clay availability.
of about 2 Mt annually. Potential clinker replacements have been reviewed by Snellings
Clays, especially those containing some kaolinite, produce reactive [26]. One of the most plentiful is steel slag, of which about 200 Mt is
materials when calcined to around 700–850 °C [27] Clay reserves are so available annually. Slags from steel production differ from those pro-
vast as to be effectively unlimited compared to the amount of cement duced during the reduction of iron in a blast furnace (GBFS, as dis-
produced. In countries such as India and China with established cussed above). Most slag from the LD process is rich in CaO and other
ceramic industries, substantial reserves of suitable clays are currently expansive phases, and so better utilised as a raw material for clinker
stockpiled as waste — the over- or under-burden from existing quar- production. If all CO2 from the decomposition of limestone, added to
rying operations. Exploitation of these reserves represents an enormous the blast furnace charge, is allocated to steel production it could be
potential to increase the global supply of SCMs. Clay reserves in other considered a source of CaO free from “chemical” CO2. The presence of
countries may be less accessible, and the cement industry must also heavy metals in some slags may also restrict their application.
respect the need to preserve natural resources. For the sake of sus- Smaller amounts of non-ferrous slags also exist. Currently, few of
tainability, clay usage should parallel the trend of fly ash and slag — these are quenched or otherwise treated to improve reactivity, because
either by using waste materials from other industries, or carrying-out the required investment is not justified by any demonstrable improve-
onsite calcination in cement plants using local materials and avoiding ment in performance. More research is needed if these are to be used as
long distance transport. SCMs.
High surface area and high water demand, along with colour con-
trol, have been problems that recent technologies are progressively 8.7. Fillers as SCMs
solving.
Calcined kaolinitic clays have the advantage of reacting quite ra- Fillers are fine particulate materials, inert or weakly reactive, pro-
pidly, more rapidly than siliceous fly ashes and even faster than slag. duced by grinding, that can partially replace clinker or other reactive
The high alumina content of calcined kaolinitic clays makes them SCMs. The use of fillers to dilute or extend more valuable raw materials
particularly suitable for co-substitution with limestone [29] as dis- is widespread in other industries, including plastics. Fillers are also a
cussed below and in the white paper by Scrivener. convenient clinker substitute for the cement industry. Because they do
not require calcining, filler use could be very interesting from an eco-
8.5. Vegetable ashes nomic and environmental perspective. Their production needs only
energy for grinding. Since many minerals can be used as fillers, they are
The white paper by Martirena and Monzo discusses the question of available everywhere in effectively unlimited quantities. The use of
biomass or vegetable ashes. The total availability is probably around fillers is discussed further in the white paper by John et al.
100–200 Mt/y assuming an average ash content of 5% [30] on crop The first recorded uses of fillers to replace binder were the
residues reported by [9]. Ash production tends to be dispersed in small Arrowrock and Elephant Butte Dams, built by the US Bureau of
quantities and close to agricultural areas. Reclamation between 1912 and 1916 [27]. Coarse cement was inter-
The most studied is rice husk ash (RHA). If properly processed, this ground with local rock (granite and sandstone), producing a cement

11
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

with 50% filler. A hundred years later these dams are still in use. The fairly high dolomite content, performs well as an additive [35]. There is
technical feasibility and durability of such filler cements were demon- no calcination of dolomite to yield periclase (MgO) and so there is no
strated by a 10-year long investigation conducted by the University of risk of unsoundness. Dedolomitisation may occur slowly, but there is no
California Berkeley [32,33] in 1930's–1940. evidence of deterioration caused by this reaction§. More widespread
After this, it took 40 years for fillers to become widely used in the use of limestone unsuitable for clinker manufacture would lead to more
cement industry. Since the 1980's substituting limestone filler for efficient exploitation of limestone quarries with and significant exten-
clinker has become a common practice in the cement industry. sion of quarry life.
Nowadays, most countries' standards allow filler substitution, with In addition to limestone, any other mineral may be used that is
maximum filler values ranging from 5 to 35% (John et al). Limestone volumetrically stable when exposed to hydrated cement paste and has
filler has become the most widely used clinker substitute (Fig. 5) with no negative influence on long-term durability. This may be important
an average content (among GNR companies) of around 7% that has for regions where the availability of limestone is limited, such as Brazil's
remained constant since 2010. Because a fraction of the limestone does Amazon region and in India [36]. However, for some minerals such as
react with available alumina to form carbo aluminate phases which quartz [37], care must be taken to avoid the health risks of respirable
contribute to strength and durability [24,25], up to 10% limestone can crystalline silica dust.
be added without the negative effect of dilution on properties; lower
levels of addition, typically around 5%, may even improve properties. 8.8. What is the limitation on average clinker factor?
However, increasing filler content using the current typical inter-
grinding technology, will reduce the strength class of the cement and so If we can solve the problem of SCM availability, the question then
will increase the amount of cement needed to achieve a desired arises of what is the technical limit to the average level of clinker
strength. Higher filler contents are typically used in unsophisticated substitution. To reiterate, granulated blast furnace slag, which has la-
applications, where optimisation of the cement content is poor and it tent hydraulic activity and an overall composition similar to calcium
reduces the tendency for over dosage of clinker. silicate hydrate, may be used up to levels of 70% or more. Pozzolans,
The average of limestone filler used in cement varies from 1 to 20% such as fly ash and calcined clays require calcium hydroxide (from the
from country to country (Fig. 13). Several factors influence the uptake hydration of the clinker component) to react and react slower than
of filler: a history of poor-quality, counterfeit high-filler cement, or clinker. Average accepted substitution rates in many applications are
other cultural circumstances may keep standard limits low; the ex- generally defined by the client's requirement for a reasonable early
istence of local over-capacity of clinker productions may make filler less strength. However, the issue of early strength can be partially addressed
attractive to cement producers, etc. There seems to be higher filler at the production level by replacing intergrinding with separate
content in the bagged cement markets — users of bagged cement often grinding.
use cement-rich mix designs and rarely require very high strength Even so, the clinker factor must be minimised in a way that the
concrete. In this situation, high filler contents are an effective way to overall amount of clinker needed to deliver the required performance is
minimise clinker use. Since industrial users prefer high-strength ce- reduced. Current intergrinding technology means that, limestone con-
ments, markets where these users dominate have proportionately lower tents above ~10% will result in a cement of lower strength class be-
filler fractions. Nevertheless, the wide variation indicates that filler cause of clinker dilution. This strength reduction has little practical
levels could certainly be increased to reduce clinker use. implication for the bag cement market (as previously discussed).
Recent developments in Germany and Brazil show that it is possible However, in better optimised concrete mixes, it may increase the
to produce cements with acceptable performance with higher filler amount of cement necessary to achieve the desired concrete or mortar
content than the current 35% maximum in many standards. More so- strength, which may also increase the environmental impact in the final
phisticated grinding processes to optimise particle size distribution, application.
along with adding dispersant chemical admixtures at the cement plant Combining particle size distribution engineering with the use of
to reduce water demand to reach good workability are needed to avoid dispersants (e.g. superplasticisers) allows for filler contents to be in-
the strength reduction caused by dilution [34]. creased to above 50% without losing, and sometimes even gaining,
Recent studies also confirm the possibility of using other minerals as mechanical strength. This can be done if the water demand for ade-
fillers in place of high purity limestone, which currently dominates quate workability is reduced. Less water means there is more solid
standards. Limestone unsuitable for clinker production, such as with (binder + fillers) to fill the space between sand grains so good me-
chanical properties are maintained. This new technology will require
0.25 Africa separate grinding of clinker and fillers and mixing and the addition of
dispersant admixtures. The reactive clinker fraction should be con-
centrated in the fine fraction of the blend, favouring its earlier strength
Lismestone filler (% of grey cement)

Asia - China,
India, CIS
0.2
CIS gain. With such technology, the increase in the cement content needed
to maintain mechanical properties means that the proportional CO2
India reduction will be lower than suggested by the clinker factor. This
0.15
China
technology could also be adopted at the concrete level, as discussed in
Section 10.
La!n America Another promising technology is the coupled addition of SCMs
0.1
Morocco
containing alumina, such as calcined clay, fly ash and slags. The alu-
mina reacts with the calcium carbonate (limestone) to form carbo
Thailand aluminate hydrates, which contribute to space filling, improving
0.05
EU 28
strength and durability. This development is recognised in the recently
proposed European standard extension allowing up to 55% clinker
World substitution. This synergetic addition is particularly effective for fast
reacting calcined clays containing metakaolin — there is no strength
0
1990 1995 2000 2005 2010 2015
decrease going from a 30% solely calcined clay substitution to a 45%
Fig. 13. Limestone filler content in grey cement for selected regions. Data from substitution with 30% clay and 15% limestone [29]. Some aspects of
WBCSD GNR [14]. GNR coverage varies durability, such as resistance to chloride ion penetration may even be
between regions. improved.

12
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

If optimised particle size distribution and combinations of fillers and These aspects are both discussed in more detail in the white paper by
SCMs are fully exploited, we consider that an average clinker sub- John et al.
stitution level of above 40% (clinker factor < 0.6) is realistic world-
wide. CO2 reduction would then be more than currently ascribed to CCS 10.1. Reducing CO2 by improving binder efficiency
in the 2009 IEA/WBCSD Roadmap. However, realising this level of
clinker substitution will require increased research and education ef- It is possible to make considerable improvements in the efficiency of
forts, particularly with users. cement use and this has considerable potential for CO2 emission re-
duction. The decisions and skills of the user in formulating cement-
9. Alkali activated materials based mixtures determine the amount of cement used for a given ap-
plication. An appropriate indicator of cement use efficiency can be ei-
On the research side, alkali activated materials (AAMs) have re- ther “binder intensity,” which signifies the amount of binder (clinker
ceived much attention as materials with lower CO2 emissions. The first and also reactive SCMs but not fillers), or the “CO2 intensity” per m3
alkali activated materials were alkali activated slags, developed and and per strength unit (MPa), a concept proposed by Damineli et al.
widely used in the Ukraine during the 1970s. Subsequently, it has been [38]. Fig. 14, shows an example of the data collected and published by
demonstrated that materials, also known as “geopolymers,” could also those authors, plus examples of recent developments, including data
be obtained from alumina silicate sources such as fly ash. from concretes formulated with up to 70% replacement of binder by
In the white paper by Provis, this technology is described in detail. filler.
Currently, global commercial use of these materials remains extremely The amount of binder used to produce concretes of a given strength
small. Substantial technical obstacles exist for more widespread use. On varies enormously, (a fact that is neglected by typical life-cycle in-
the resource side, there are major limitations. These materials use the ventory databases). This dispersion shows that there is a substantial
same substances as those used to substitute clinker in blends, whose potential for CO2 mitigation by simply improving the cement use effi-
limited availability was discussed in the previous section. In the case of ciency. The minimum binder intensity is 5 kg/m3.MPa above 50 MPa.
slag — the main component of nearly all real-world applications — For lower strengths it follows a line that corresponds to the 250 kg/m3
almost all suitable quality slag is already used in conventional Portland limit present in most reinforced concrete standards. This minimum
based cement or concrete. If slag is diverted from use in Portland based binder content reflects the need for sufficient fine particles to fill the
blends to be used in alkali activated materials, it may be that the CO2 space between aggregates to ensure good rheological behaviour. For
emission per tonne (or m3) of the alkali activated material will be lower concretes with 30 MPa compressive strength the minimum binder in-
than those of an equivalent standard OPC-based material. But there will tensity is around 8 kg/m3·MPa, but the average is around 12 kg/
be no decrease, and likely an increase in the overall global CO2 emis- m3·MPa, a 44% difference. In general, the lower binder intensities are
sions of the cementitious materials sector because (a) the CO2 per tonne for concretes made with pure Portland Cement.
of Portland based materials will increase due to the lack of slag for A consequence of the variation in binder intensity is a variation in
blending, and (b) any CO2 emissions associated with the production of CO2 intensities of more than four times for concrete of the same
the alkali activator must also be factored into the equation. strength — the “worst” concrete results in more than four times the CO2
Therefore, AAMs can only contribute globally to the reduction of emissions of the “best” concrete. The minimum CO2 intensity is around
CO2 emissions in the sector if they primarily use minerals or industrial 2 kg CO2/m3·MPa for concretes above 40–50 MPa and increases ex-
by-products not currently used as clinker substitutes in blended ce- ponentially for lower strengths. Lower CO2 intensities can be achieved
ments. Broadly speaking, the effectiveness of materials in AAMs is by replacing clinker with reactive SCMs — such as granulated blast
roughly correlated to their reactivity in blended cements. The “best” furnace slag and fly ash — assuming they are carbon neutral. The
materials are granulated blast-furnace slags, which can be used to clinker portion of the total binder is not necessarily a good indicator of
produce AAMs at room temperatures. Fly ash with good reactivity the CO2 footprint of the concrete, some concretes made with “pure”
would allow production of alkali activated cements with CO2 footprints cements (~0.95 clinker factor) have lower CO2 intensity than concretes
lower than Portland cement. However, thermal curing is required for made with cement with clinker factors 40% and lower.
the most abundant, class F fly ash, which limit its market penetration to The most effective way to reduce cement dosage, without compro-
precast components and diminishes its mitigation potential. The mising strength, is by selecting the aggregate amounts in different
availability of reactive fly ash is limited to some regions and it is al- fractions to optimise packing, reducing the void space to be filled by
ready diminishing due to climate change related policies. The only cement paste. Reducing the cement dosage while maintaining good
materials with substantial potential to extend the availability of suitable flow and compaction also depends of the use of adequate dispersants
minerals for AAMs are calcined clays. However, at present, consider- (admixtures). Generally speaking, it is only practical to implement such
able quantities of sodium silicate (water glass) are needed to activate technologies with industrialised production (e.g. Ready mix plant).
calcined clays. The total CO2 emissions from clay calcination and large This wide scatter observed for concrete, is also likely to occur in
quantities of sodium silicate may result in cements with high CO2 other concrete applications, like concrete blocks.
emissions. The current global production of sodium silicate worldwide
is < 10 Mt per year (enough to make only around 40–50 Mt of AAMs) 10.2. Potential of filler to reduce binder and CO2 intensities of concrete
and the energy and CO2 emission of the production process used at
present are very high. Materials that could truly reduce CO2 emission in Recent advances in engineering particle size distributions combined
the sector require the invention of a lower energy production process with the use of dispersants allow a binder replacement of up to 70% by
and substantially increased production, requiring significant invest- inert fillers without the negative effects of dilution. Since fillers can be
ment. made from a variety of materials and require no calcination, they can be
cheaper than Portland cement, which makes the technology attractive.
10. CO2 mitigation by improving efficiency of cement use The combination of particle size distribution and dispersant use reduces
the amount of mixing water needed to obtain good flow (Fig. 15). With
In exploring ways to increase the cement use efficiency, we will first less mixing water, fewer hydration products are needed to fill the space
consider improvements in binder efficiency possible in industrialised between particles, improving strength. Because there is less water, there
applications. Secondly, we discuss the mitigation possible through is more solid (binder plus filler), even up to 50% more. Nevertheless,
promoting more industrialised production of concrete and minimising binder intensity can be reduced by > 50%, as indicated by the red line
the inefficient use of cement by untrained and ill-equipped personnel. in Fig. 14. Binder intensities around 4–5 kg/m3·MPa can be achieved for

13
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

20
Int_clinker>95
30 MPa concretes in comparison with values higher than 8 kg/m3·MPa
18
Br_clinker>95 typical with current technology. For strengths above 50 MPa, binder
Int_BFS>5
Br_BFS>5 intensities are about 2 kg/m3·MPa, compared to 5 kg/m3·MPa with the
best current technology. Similar technology can be deployed for mor-
Int_FA>5
Binder Intensity (kg/m³.MPa)

16
Br_FA>5
tars, concrete blocks and other cement-based products.
KTH/CBI (Vogt)
14 Proske et al
Poli USP (Damineli)
Karlsruhe
The refractory castables industry has > 20 years of experience with
12 Poli USP (Qua!rone) this technology [39]. In the Portland cement field, the technology is still
Darmastad 2
10 Darmstad 1 in its infancy, with a limited amount of published research, although
there are some patents in the area [e.g. [40] [41]]. The recent pub-
VDZ
8
lication of ASTM C1797M-16 [41], standardising filler for use in con-
6 crete, and ACI 211.7R-15 [42], a practical guide for proportioning
4 concrete mixtures with fillers attest to growing interest in the market.
The robustness of combining cement with dispersants, and the limited
2 time stability of mixed dispersants and cement, particularly in hot cli-
0 mates, are issues requiring further research. The specification of
10 30 50 70 90 minimum cement contents in current standards is another restriction
Compressive Strength (MPa) [43]. Such requirements are most relevant where carbonation induced
corrosion is an issue (see Section 5). Unreinforced concrete, concrete
20 components such as blocks, levelling, rendering and bricklaying mor-
tars, are applications comprising a large proportion of cement use in
Int_clinker>95
Br_clinker>95
18 Int_BFS>5 some developing markets and offer considerable potential for CO2 re-
Br_BFS>5
16 Int_FA>5 duction.
CO2 Intensity (kg/m³.MPa)

Br_FA>5
14 KTH/CBI (Vogt)
Proske et al 10.3. CO2 reduction through industrialised cement use
Poli USP (Damineli)
12 Karlsruhe
Poli USP (Qua!rone) One of the appeals of cement is its robustness and inherent simpli-
10 Darmastad 2
Darmstad 1 city of use. Untrained personnel can produce concrete using as much as
8 90% locally available sand and gravel. This means that only cement (in
6
bags) needs to be transported significant distances. On the other hand,
such untrained personnel tend to use more cement than necessary, due
4 to non-optimal grading of aggregates, lack of dispersants and low in-
2 tensity mixing. Since industrial clients, in most regions, prefer bulk
delivery, the market share of bagged cement is a rough estimate of
0 inefficient use of cement that is typically higher in the developing world
10 30 50 70 90
(Fig. 19). Cement wastage is also a consequence of poor planning, in-
Compressive Strength (MPa)
appropriate storage or transportation — like exposing cement bags to
20 rain or humidity, storing them for too long, all of which is exacerbated
Int_clinker>95 by the fact that paper is the most common cement-bag material. Fig. 16
Br_clinker>95
shows data from Brazil, illustrating that poorly controlled concrete
18
Int_BFS>5
CO2 Intensity (kg/m³.MPa)

production at building sites leads to higher materials wastage rates,

Br_BFS>5
16 Int_FA>5
Br_FA>5 defined as the percentage of materials actually used at the building site
14 KTH/CBI (Vogt)
Proske et al exceeding the amount in the project design.
For these reasons promoting more industrialised cement usage has
Poli USP (Damineli)
12 Karlsruhe

10
Poli USP (Qua!rone)
Darmastad 2
the potential to reduce cement waste, and therefore overall cement
Darmstad 1 consumption, which would also reduce the environmental impact.
VDZ
8 Other environmental benefits include reducing illegal aggregate ex-
6 traction, which is very common in the developing world [45–49].
It is reasonable to assume that increased industrialised production
4 of cement and cement-based materials could reduce wastage by at least
2 20–30%. Best estimates for the global bag market are around 42%
(Fig. 19). If these saving were realised on this amount, overall cement
0
consumption could be reduced by around 10%. The challenge is making
0.0 0.2 0.4 0.6 0.8 1.0
industrial products competitive compared to informal markets — which
Clinker / binder
avoid costs such as restoration of quarries and payment of social se-
Fig. 14. Potential CO2 mitigation by optimising binder use with current (light curity and taxes for workers — ensuring adequate return rates on the
colours) and high-filler, low-water technologies. Binder intensity (bi) as func- capital needed to build plants and logistical infrastructure. Govern-
tion of 28-day strength (a, top); Estimated CO2 intensity (ci) as a function of ments must be encouraged to make industrialised cement products —
28 days strength (b centre). Estimated CO2 intensity as function of the clinker ready-mixed concrete, dry mix mortar, and precast concrete compo-
fraction on total binder, excluding inert fillers (c, bottom). Benchmark data
nents — more competitive.
from 29 countries (Damineli et al. [27]) are represented in light colours. Red
For example, China included industrialisation of cement use in the
tick lines are the best practice results of current technology. The green line on
Chinese National Climate Change Program [50] by seeking to “dis-
the CO2 intensity graph (b) separates CO2 intensity of cements with < 5% of
SCMs from those with high amount of SCMs. Dots and diamonds in darker courage the production of bagged cement and encourage the develop-
colours are from high-filler, low water concrete formulations from various ment of bulk cement.” In October 2003, the Chinese central government
authors. (For interpretation of the references to colour in this figure legend, the issued Decree 341, banning concrete mixing operations on job sites in
reader is referred to the web version of this article.) 124 cities across the country [51]. In June 2007, mortar mixing op-
erations on job sites were banned in ten large cities, an initiative that

14
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

Fig. 15. The effect of the combination of binder replacement by inert fillers. Without the presence of the dispersant admixture (centre), as is usual in today's cement
industry, ultrafine filler agglomerates may increase the water demand. Dispersants destroy the agglomerates (right), and if the particle size distribution is adequately
engineered, will reduce the amount of water needed to fill intergranular space and separate the particles.

300 higher filler contents and additions such as calcined clays, which
247
can be used to increase the eco efficiency of concrete in many ap-
250 plications. A case study in the white paper by Cheung et al., illus-
Materials Wastage Rate (%)

trates how the use of admixtures with proper mix designs and use of
200 SCMs can even lead to CO2 savings at equal load carrying capacity of
67%.
2. Air entraining agents (chemicals capable of incorporating fine air
150
118 bubbles in mortars or concretes during mixing) improve rheology
and resistance to frost action. In mortar and concrete components
100
where the main requirement is volume filling rather than strength,
56.1
these agents are an effective tool for increased materials efficiency
50 45.2 40.7 and saving materials, including cement.
37.1 23.3
3. Accelerators (chemicals capable of accelerating cement clinker hy-
0 6.4 6.8 8.7 2.4 8.6
dration) enable faster strength development. They allow concretes
Cement Sand Coarse Aggregate Ready-mixed with higher proportions of slower-strength developing SCM's to
concrete reach adequate early strength required for construction by accel-
Fig. 16. Materials wastage rates (max, min and median) in > 40 high-rise erating the clinker phase hydration.
building sites from different Brazilian regions. Data from [44]. 4. Chemicals that address major concrete durability issues, including
freeze thaw protection through air entrainment, reinforcement
corrosion protection through corrosion inhibitors, cracking reduc-
extended to 33 cities in 2008, and 84 cities as of July 2009. The ex- tion through shrinkage reducing admixtures, etc. While these may
pectation was a net savings of 2.4% reduction in cement consumption, a not reduce initial environmental load, prolonging service life in
4.5% savings in reduced materials loss, with the added benefit of critical structures will reduce environmental impact.
avoiding 3.3 million cubic meters of timber use for paper bag produc-
tion. The performance of cement dispersants has improved substantially
Industrialisation is only feasible in medium to large cities, where in recent years, but there may still be compatibility problems with in-
consumption rates are higher and transportation distances are more dividual cements and SCMs, particularly in hot climates. Advances in
favourable between producers and consumers. However, according to recent years in the understanding cement-SCM-admixture compatibility
UN [52,53] the pace of urbanisation, particularly in Africa and Asia, is issues have generated practical solutions. Along with more rapid la-
accelerating. In 2014, 54% of the population was situated in urban boratory- and field-based detection tools, the use of admixtures can
areas; this figure will increase to 66% by 2050. About half the total clearly be better adapted to developing countries, where admixture
urban population lives in cities larger than 500,000 inhabitants and usage is currently limited. These obstacles can be overcome with easy-
one-eighth in cities larger than 10 million. to-use diagnostic tools and disseminating educational material.

10.4. Chemical admixtures 10.5. Recycling cement-based waste

Chemical admixtures are products that used in small amounts are Significant efforts have been made in recent years to recycle con-
capable of improving cement-based materials performance. From the crete and other cement-based waste, particularly by producing recycled
preceding sections, it is clear that dispersants (plasticisers and super- aggregates. This is motivated by the need to reduce construction and
plasticisers) are critical for improving the efficiency of cement use. The demolition waste (CDW) going to land fill, and also by the increasing
white paper by Cheung et al. discusses in detail the role of admixtures scarcity of viable sources of virgin aggregates situated in proximity to
in a more sustainable cement value chain. metropolitan areas. Recycling concrete and mortar yields meaningful
Reduced CO2 emissions can be achieved with four types of ad- environmental benefits, from these perspectives. However, recycling
mixtures: does not contribute significantly to reducing CO2 emissions in the ce-
ment-based materials sector. Naturally occurring virgin aggregates
1. Dispersant-based water reducers (plasticisers and superplasticisers) have a CO2 footprint two orders of magnitude lower than cement [54].
reduce the amount of water needed to make concrete that can be This makes their share of the CO2 footprint very low — typically <
easily placed, and so reduce the amount of cement clinker needed 10%, even taking into account long-distance transport via roads. Pro-
for a given strength and durability. This technology also enables cessing recycled aggregates is generally a bit more energy intensive

15
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

than processing virgin materials, because recycling requires deconta- to about 40 Mt in 2015. Therefore, even is if some reduction in the CO2
minating demolition waste and may generate substantial quantities of emissions are possible, it will not contribute significantly to overall CO2
fines, which typically have no commercial value [55]. Transport and reductions in the sector.
waste disposal are decisive elements in the environmental impact
equation — but generally speaking, if aggregates can be recycled close 12. Structures
to usage sites, there will be a small net reduction in their associated CO2
emissions due to reduced transportation. As discussed in the introduction, this report focusses on technolo-
A more problematic issue for recycled aggregates is that depending gies at the materials level. Nevertheless, it should be mentioned that
on their quality and substitution levels, these materials may increase there is also considerable potential for saving cement at the design
cement demand for a given strength, in turn negatively affecting the level. Not least, it is undoubtedly true that many structures use concrete
overall CO2 footprint of the concrete [54]. At present, most techniques of a higher strength than needed for the design, which amounts to a
that improve aggregate quality by removing cement paste also tend to waste of materials.
have a high CO2 footprint [55]. Fig. 18 shows how recycled aggregate This is not an easy topic, since most of the discussion on sustainable
needs to be sourced much closer to the site of use to compensate for this cement-based materials is focussed on solutions to minimise cement
effect. CO2 footprint by the use of SCMs, etc., and strategies to increase service
Recycling also results in a high amount of CaO rich fines, most of life, especially of steel reinforced concrete structures. However, design
which are non-carbonated. With informed planning and management, decisions, at both architectural and structural design levels, control the
these fines may be recycled as raw materials for clinker production, amount of materials, particularly steel and cement, and therefore, have
thus reducing chemically related CO2 emissions for clinker [56]. An- a direct influence on the minimisation of the environmental impact and
other useful by-product from aggregate recycling with potential market particularly the efficiency of cement use. Designers decide aspects such
value is good quality recycled sand [58]. as the thickness of a wall, of a concrete block component or the number
and size of reinforced concrete columns and beams in a building.
11. Soil concrete There is no international, universally recognised benchmark for the
CO2 footprint or concrete consumption for buildings or other structures
Raw (crude, unbaked) earth (subsoil) is a traditional building ma- allowing producers, consumers and researchers to make informed de-
terial used in a wide range of construction techniques. These are pre- cisions. A unique global initiative in a closely related field is the
sented in more detail in the white paper by van Damme and Houben. Common Carbon Metric for measuring Energy Use & reporting
Stabilised soils remain in significant use for dwelling construction all Greenhouse Gas Emissions from building operations protocol, devel-
over the world. Without industrial additives, it is a material with a oped by UNEP SBCI, but it is focused on building use phase [59].
remarkably low environmental impact. A recent trend has been to The Building and Construction Authority in Singapore edited a
“stabilise” the raw earth with lime, plaster of Paris, Portland cement guide to promote the optimum use of concrete [60], within the scope of
(PC), or supplementary cementitious materials (SCMs). This had led to their BCA Green Mark building rating system (Fig. 20). The guide
the now widespread incorporation, mainly for compressed earth blocks presents a benchmark of the concrete usage index (CUI), which mea-
and in rammed earth, of typically between 5 and 10% PC by weight. sures the amount of concrete (m3) used for each unit of floor area (m2)
This approach is generally used to minimise the high maintenance in- of the building, usually limited to the superstructure (excluding the
volved with pure soil materials, particularly in regions with heavy foundation), with values ranging from 0.35 m3/m2 to 0.7 m3/m2. For
rainfall. 2–3 storey buildings the CUI is much lower, varying from 0.2–0.3 m3/
In comparison with concrete, soil stabilisation can be an extremely m2 (Fig. 19). The guide also includes examples of technologies that help
inefficient way to use large volumes of Portland cement. Fig. 19 shows to save concrete, including pre-stressed slabs, lightweight partition
the CO2 intensity of such cement-stabilised soil-based materials com- walls, hollowed slabs, composite steel-concrete systems and high-
pared to conventional concrete discussed in Section 10 (Fig. 14). The strength concrete. A study conducted by the Lawrence Berkeley La-
data points for stabilised earth follow the exponential curve for con- boratory [61] on concrete buildings in China estimates CUI between
crete. Compressed earth bricks (CEB) behave like a concrete with 0.53 and 0.61 m3/m2 in 2008, which is expected to growth due to the
moderately poor environmental and mechanical performances. They increase in storey numbers to between 0.62 and 0.70 m3/m2 in 2030,
are usually structurally solid masses, while concrete and ceramic blocks with an average steel consumption varying between 81 and 96 kg/m3 of
tend to be hollow. In environmental terms, self-compacting clay con- concrete. In a pioneer study, Warszawski [62] made an estimate of
crete (SCCC) is nearly the worst formulation (in environmental terms), cement consumption for the future in Israel, giving results varying from
even neglecting the impact of the superplasticiser. Rammed earth (RE) 0.39–0.69 m3/m2 of concrete. Data from high-rise buildings in the
and mud bricks (B) have extremely poor environmental performance, Middle East shows results varying between 0.4 and 0.7 m3/m2 for
with CO2 intensity indices 20–25 times larger than the asymptotic value buildings between 20 and 40 floors, but reaching a maximum of 0.9 for
of high performance concretes. Stabilised soil technology seems to offer a building with 85 floors [63]. These values are remarkably coherent.
only moderate mechanical improvement at a high environmental cost, On the other hand, data from 93 buildings in Brazil shows much lower
which can only be positively offset under specific conditions where the CUI, varying from 0.16–0.3 m3/m2 for 4–30 storey buildings and
service life is greatly extended, or as a building material of last resort. 0.07–0.11 for 1–3 storey residential buildings. In all cases scatter is
Van Damme concludes: Stabilisation of soil concrete with OPC is not important, and reflects design decisions such as percentage of steel
advisable. It provides only very moderate benefits while employing reinforcement, which has a high CO2 footprint, differences in loads,
large binder volumes. One acceptable use could be compressed blocks, concrete strength especially for columns, typical spams of beams and
but their strength is still at least three times smaller than the same slabs, the eventual need for earthquake-resistant designs, which is not
amount of OPC would offer for good concrete formulation. He points the case for Brazil. Optimisation of the design of buildings in terms of
out that unstabilised soil construction can have good durability and low environmental impacts is a complex task, because in includes multiple
maintenance if two cardinal rules are followed — “good boots”, functions, including fire resistance, acoustics performance as well as
meaning a water tight foundation, and a “good hat”, meaning a roof to seamless integration with multiple materials, especially the reinforced
keep rain o# the walls. concrete and will require significant research and development effort.
Stabilised soil supplemented with cement or lime is also used for In the literature higher strength concrete is often promoted as a tool
road subbase. Around 5% of cement worldwide is used for road con- to reduce environmental impact, because it is has a lower binder in-
struction, about 20% of which is for stabilising the subbase, amounting tensity and CO2 intensity (see Fig. 14), is capable of carrying more load

16
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

per unit of area, therefore reducing the amount of cement and ag- degrees. The other large cement consuming countries of China, Japan,
gregates needed and increasing the liveable space of the building [64] India, and Russia have their own standards, nevertheless inspired by EN
and have higher durability. However, the actual benefits seem to be or ASTM approaches. Some approaches, such as limestone use and even
smaller than estimated by simple reduction in materials demands and the use of slag and fly ash, can face local resistance resulting from poor
cement use efficiency. Nevertheless, literature shows estimated CO2 commercial or technical understanding.
mitigation as high as 20% for construction of a bridge in France (up to Increased efficiency in cement use requires market segmentation
50% if the uncertain use phase is excluded) [65], with values for and flexibility to exploit local opportunities for raw materials, both for
buildings varying between 4.1 [66] to and 16.5% [64] for the con- clinker and for supplementary materials like fillers. This will allow
struction of buildings. In conclusion, benefits seems be limited for si- products optimised for specific applications, rather than the conven-
tuations where the increase in strength induces a significant volume tional approach of one-type-fits all. This can only be achieved with
reduction, compensating the increased cement amount per cubic meter performance standards specifying properties that must be met, such as
allows a significant increase of service life of the structure [65]. strength, elastic modulus, and durability.
Therefore explains why the amount of concretes with strength above Performance Standards for mechanical properties are relatively easy
the class 35–45 MPa represents < 15% of the European market [67]. to implement based on measurements at relatively short ages, such as
Concepts such as topology optimisation, where the design seeks for 28 days. There is much greater difficulty ensuring durability over the
the shape that reduces the amount of material have been explored for service life of a structure. To be practical performance tests need to give
buildings [68] and certainly can be applied for small components — results in a relatively short time — a few months at most — to reliably
such as concrete blocks to bridge beams. Functionally grading the predict performance in the field over many decades. Considerable ad-
properties of cementitious materials can allow localised reduction of vances have been made in understanding fundamentals of the pre-
cement. This concept has been demonstrated for fibercement [69], but dominant degradation mechanisms, but the simultaneous action of
can also be applied to concrete components. The combination of such various mechanisms remains less well established [70]. Applying fun-
concepts with digital production methods certainly has a potential for damental knowledge gained from short term laboratory testing to long
mitigation. term durability performance is challenging, because of the broad range
It is clear that structural design represents an area where con- of environments in which cement-based materials are used.
siderable savings in CO2 are possible, but this will require substantial Developing a robust scientific understanding and pilot applications
investment in research and development as well as in education to raise in representative conditions could help accelerate the development of
the awareness of engineers to the possibilities. performance standards, especially if the process involves cement pro-
ducers, governmental clients and regulators and large building con-
12.1. Improving durability tractors. Technical approval schemes, where available, can facilitate
pilot applications. This will require international collaboration at a
Cement based materials are typically expected to have a service life scale not yet seen in the field.
of at least several decades. Fifty years is standard, although often the There is also the significant issue of cement versus concrete stan-
expectation is for much longer. Durability is not an intrinsic property; it dards. Cement standards are far easier to enforce because cements are
depends on the interaction between cement-based materials and their generally manufactured by large and easily identifiable companies. It is
usage environment. There is ample scientific and technical expertise well known that concrete performance, especially durability, depends
available to understand degradation, but predicting service life in real on good quality control in the field, for example to ensure the use of the
environments where many factors interact is complex. correct water to cement ratio and avoid cracking. To be truly successful,
The overwhelming majority of problems of concrete durability — the performance standards approach will have to be applied to con-
probably about 90% — are related steel reinforcement corrosion, which cretes and other cement-based end products rather than just to cements.
is related mostly to chloride ingress, and less commonly to carbonation. This objective will be difficult, especially in developing economies,
It is important to realise that only a very small proportion of cement use although this is also where the greatest ecological returns can be rea-
is at risk, because only about 25% of cement use is in reinforced con- lised. We have shown in this report that there are very significant
crete (Section 5 Fig. 8). Only a fraction of this is exposed to conditions ecological returns from minimising clinker use in concrete. These can
posing durability risks. In these applications, longer lasting concrete also be associated with a reduced unit cost on the assumption that
would certainly reduce environmental impact. However, in terms of Portland cement clinker is replaced by less expensive fillers. But such
total cement consumption, the amount of extra cement consumed re- benefits can only accrue if the importance of good concrete mix design
lated to repairing degraded structures is rather small. and quality control is understood by end users.
Every effort should be made to design and build costly infra- Changing our approach to standardisation will therefore also re-
structure, such as bridges, to last as long as possible. However, the using quire a major educational initiative to ensure that the engineers re-
the same concrete for all applications would be an unduly conservative sponsible for designing and executing concrete works fully understand
approach that would also be very wasteful from the standpoint of re- the relevant issues.
sources and CO2 emissions.

13. Standardisation 14. Estimating the mitigation potential

Standardisation can be a major barrier to the introduction of new This study has considered CO2 mitigation material-based technolo-
solutions. It is a consensus-based process, with serious economic im- gies for the cement-based materials industry, which were not quantified
plications for numerous businesses. Standards are further complicated by means of the Cement Technology Roadmap 2009 [1]. The full
by the fact that they operate on (at least) three levels — the cement methodology is detailed in the white paper by Miller et al. The miti-
level, the concrete and at the construction level — often with a lack of gation potential of each technology is uncertain, due to uncertainty on
communication between these levels. CO2 and energy footprints, and future market share, which ultimately
The majority of standards worldwide are prescriptive; they dictate depends on investments in industrial facilities. Actual CO2 emission and
the composition of the materials that can be used. Two sets of standards production costs will also be influenced by the R&D investments in each
dominate — the European EN standards and those originating from technology, both by industry and public agencies. Consequently, the
ASTM, based in North America. Many Latin American, African and mitigation potential presented for each technology simply represents a
Asian countries follow one or the other of these standards to varying desirable possible outcome.

17
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

Table 2
Emission factors considered in the estimate of mitigation potential (kg of CO2 per kg of material produced).
Clinker + gypsum Calcined clay Limestone filler Sodium silicate RBPC BYF CCSC

0,86 0.2–0.35 0.008 0.90–1.8 0.79 0.62–0.66 0.54

14.1. Emission factors considered in the model, the mitigation potential of this option is
overestimated. Calcined clay based AAMs can be scaled up, but this will
Emissions factors for several alternative materials were based on life require investments to increase the production of sodium silicate. To
cycle assessment (LCA) methodology. Production was considered at supply 15%–7,5% of AAM fly ash and 7,5% for calcined clay - of the
regional levels and global emissions factors were estimated by means of cement market of 2050 about 250 Mt of sodium silicate would be
weighted averages. The Getting the Numbers Right (GNR) database needed, an increase by a factor of 42 in comparison with current annual
[12] reports kiln use by region as well as kiln efficiency. Fuel mixes production. However, there are extensive reserves of soda ash and silica
were estimated based on data from GNR and the International Energy that are needed to develop the product [79].
Agency (IEA) [71]; CO2 emissions from fuel were calculated based on For BYF (belite ye'elimite ferrite) clinkers, the critical materials are
values reported by the Intergovernmental Panel on Climate Change bauxite and other high-alumina minerals from which the fraction of
[72]. Raw material derived emissions (“chemical” CO2 from the Al2O3 is 16.4% [80]. Al2O3 content in bauxite is circa 40% [81], and
breakdown of limestone) were considered to be 0.507 kg CO2/kg of bauxite resources are estimated to be 55 to 75 billion tonnes [14]. To-
clinker [73]. For models of Ordinary Portland Cement (OPC), the use of day's 91% of the bauxite production is concentrated in 15 countries,
clinker was modelled with additional processing and the inclusion of mostly Australia, China, Brazil and Malaysia and India; the price of
5% gypsum by mass. Beyond OPC, the production and consumption of exports to USA ports is USD $30–46 [14]. If all of today's bauxite ex-
supplementary cementitious minerals — blast furnace slag, fly ash, traction were diverted to BYF cement, it will be possible to produce
natural pozzolans, and filler — in cement or as cementitious replace- around 650 Mt of cement. Competition with aluminium for bauxite and
ment were considered. Carbon dioxide emissions factors for such ma- regional material availability are limitation factors.
terials were based on calculations by Gursel and Horvath [74], and on Carbonation-hardening CCSC and MOMS require special industrial
consumption data from the GNR database [12]. For the LCAs con- facilities to carbonate cement in a controlled environment. Therefore,
ducted, regional electricity mixes based on calculations by Miller et al. they have a potential to capture a share of the smaller industrial com-
[75] were used for additional processing of clinker and for processing of ponents market. Because there are no published studies on future
other cementitious products. The ranges of values of the emissions market share of industrialised cement-based components their share
factors for the production of calcined clay were taken from various was estimated up to a maximum of 15% in 2050.
sources [76,77]. Carbon dioxide emissions associated with the pro- On the other hand, filler from limestone and other minerals, and
duction of limestone filler were based on methods developed by Gursel calcined clay plus limestone filler are not limited by the resources
and Horvath [74] and were adapted to account for regional electricity available, which are very large in comparison to cement demand.
mixes. However, they are limited by the maximum fraction of clinker sub-
The mitigation potential of Alkali activated materials (AAMs) based stitution. Clinker substitution by filler or calcined clay and filler mod-
on blast furnace slag and fly ash reduces between 40 and 80% of the eling will be developed as an additional to the 10% already consider in
emissions corresponding to Portland cement [78]. CO2 emissions as- the reference mitigation scenario.
sociated with alkali activators, varied from 0.9 t/t — assumed to be
possible with new technology routes — to 1.8 t/t (dry basis), being the 14.3. Mitigation potential
lowest emission factor presented by [78]. Table 2 contains the emis-
sions factors based on regional hydraulic cement production values. The mitigation potential is expressed as the fraction of market share
IEA ETP 2016 [11] low demand, 2 °C scenario was considered as a in the year 2050, forecast to be 4566 Mt [11]. Therefore, we are not
reference for our alternative scenario. For the year 2050 this scenario addressing the time scale of the introduction of each technology, which
foresees a production of cement of 4556 Mt. To meet the 18% target for will not be the same for all technologies and will clearly impact, the
overall CO2 reduction of about 900 Mt, it is assumed 347 Mt of reduc- mitigation that can be realised in practice.
tion in CO2 emissions will come from the solutions (alternative fuels The mitigation potential Mp (tCO2) of a technology is a function of
and raw materials, energy efficiency and conventional SCMs, including the specific mitigation potential, smp (kgCO2/t) multiplied by the
10% of limestone filler as discussed in Section 3) and the remaining amount of cement produced.
552 Mt from capture and storage of CO2. Different technologies may
have inherently different efficiencies for the various typical cement smp = Ct < Cc .
applications, due to differences on mechanical strength and even den- where Ct is the CO2 footprint of the technology; Cc is the CO2 footprint
sities. Despite this being an important factor, it was not considered in of the cement it displaces from the market, in our case, Portland cement
the model. with 10% of filler and 4% of gypsum. However, when the CO2 miti-
gation of the technology diverts SCMs originally used in the production
14.2. Limits for market share gain of Portland cement, such as slag and fly-ash, unless the new technology
increases the SCMs' use efficiency there is no net benefit.
It is estimated that blast furnace slag and fly ash will be around 16% For the technologies based on clinker substitution, namely (a) the
of the cement production in 2050 [15]. Nearly all of this production has combination of calcined clay and limestone filler and (b) fillers, the
already considered to be used for clinker substitution in the IEA miti- market share is limited by the maximum amount of clinker that can be
gation scenario. Therefore, the use of AAMs or any other technology substituted without, affecting strength and durability. Fly ash and
based in such raw materials is, limited. Furthermore, if these materials granulated blast furnace slag are the major SCMs and are available to
are diverted from use in Portland cement clinker based blends, the the production of approximately 740 Mt. These two SCMs groups plus
mitigation potential already assumed will be decreased. The possible the 5% of gypsum result in an average of 70% clinker fraction in the
use of remaining fly ash in some regions is considered, but in most of low demand scenario.
cases it will require thermal curing. Since thermal curing was not It is possible to introduce 10% of filler in all cements produced,

18
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

because at this low rate of substitution, the dilution effect can be easily 14.4. Cement use industrialisation
compensated for by grinding. Assuming a minimum average clinker
factor of 50%, it is possible to estimate that the sources or slag and fly The market share of bagged cement, a proxy for non-industrial,
ash available would allow combinations with, on average, 40% of fly inefficient use of cement, in developing countries is between 50 and
ash and/or slag plus 10% of filler to supply 40% of the cement market, 90% (the latter value is for India), which demonstrates the mitigation
all with 50% of clinker fraction. Without additional sources of SCMs, potential of measures that promote industrialisation. For example, the
the remaining 60% of the cement market would have only 10% of filler. potential for CO2 reduction from the reduction of binder content from
Further CO2 mitigation could then come from the use of other SCMs, industrialisation of concrete can be estimated to be around 20–30%.
limestone and calcined clays or fillers. The engineered filler combined The reduction of materials wastage rate due to industrialisation will add
with dispersant is certainly more challenging to deploy. to this mitigation.
Fig. 21(a) shows the mitigation potential of a combination of Moreover, industrialisation can facilitate an increase in the sub-
25–35% of calcined clay, 15% of limestone filler with minimum 50% stitution rate of binder by filler and allow further mitigation by the
clinker content. Fig. 21(b) presents the mitigation potential of en- increase of aggregates packing, reducing the demand for cement paste.
gineered filler combined with dispersant. These two technologies, These benefits have been proved effective for concrete, but can also be
which are based on minerals and production processes that the cement extended to mortar and other cement-based industrial products.
industry are familiar with, may reach the potential ascribed to CCS in Aspects regarding more efficient design of structures, new concepts
the IEA ETP 2016 2DS scenario. In the high demand scenario, the ce- and digital production must be systematically explored and solutions
ment production will be 30% higher — around 6 Gt. Calcined clay and developed and transferred to the various markets, which will require
fillers can cope with this increased demand, even if slag and blast fur- well-coordinated standardisation efforts.
nace slag production remains constant. The WWF-Lafarge report [14], estimates that increasing use effi-
Fig. 22 summarises the results for the new cements. The CCSC ciency can avoid 15% of cement consumption, which represents about
market is limited to the production of industrial components because 530 Mt of CO2 emissions on the low demand scenario, 96% of the CCS
carbonation hardening requires dedicated industrial facilities. Since the targeted. This will require investment in industrial facilities and mea-
material will be carbonated, protection of steel may be a problem in sures to make industrial products competitive in the highly informal
some environments. To capture uncertainty, the emission factors were markets of developing world. Efficiency gains have much broader en-
varied 15% below and above the nominal CO2 footprint. BYF is also not vironmental and social effects than CO2 mitigation and, therefore, must
fully developed, and its CO2 footprint and cost are subject to changes. be pursued.
Market penetration of this cement will depend on availability and costs
of aluminium-rich minerals in various regions. 14.5. Final remarks
Alkali activated cements are not new. Sodium silicate-based tech-
nology > 60 years old, with very limited market applications. The cal- Our scenario shows that a combination of technologies have a po-
cined clay AAM footprint will be affected by the amount of activator tential to reduce CO2 emissions beyond the current target for CCS, at a
(Na2OxSiO2, where x ~ 1.7) today, this varies between 25 and 55% (dry much lower cost and environmental risks.
basis) of the total cement. A formulation with 55% of sodium silicate Two clinker substitution technologies, calcined clay plus limestone
with 1.1 tCO2/t (dry-basis) will result in no improvement in comparison and engineered filler combined with dispersant, make much higher
with Portland cement with 10% of limestone filler FA activated AAM levels of clinker substitution possible than previously expected. The
can be an option in some regions, but the total amount of fly ash pro- sources of raw materials are almost unlimited and available virtually
duced is already diminishing. It will be feasible in some locations were everywhere. The limits of adoption are related to minimum clinker
sodium silicate is available at competitive cost and fly ash of acceptable imposed by the technical performance in different applications and
quality is available. Further in many practical situations it will require environments. These technologies require modest adaptations in the
thermal curing of the concrete, which limits the market to precast in- existing production lines, are expected to have small or no effect on
dustry. Finally, mitigation potential is overestimated because the CO2 costs. They can be deployed rapidly because they require only modest
from thermal curing was not included in the model. capital investments on plants and allow the clinker fraction to be re-
Belite rich Portland cements have small or zero mitigation potential duced progressively as R&D advances and the industry gains experi-
when compared with Portland cement with 10% of limestone filler and ence. Combined with cement use efficiency gains they may be able to
are not considered further. meet mitigation targets allocated to CCS.
MOMS theoretically provides the possibility of making concretes The potential of new cement technologies is encouraging. They are
with a significantly negative carbon footprint, especially if carbonation expected to cost more than current cement but much less than CCS.
hardening is used. However, reaching this goal will require the devel- Clinkers with ye'elimite as the most reactive phase (BYF: Belite-
opment of an energy-efficient industrial manufacturing process for Ye'elimite-Ferrite and Calcium Sulfoaluminate Cements) can be pro-
MOMS. Magnesium silicate rock seems to be concentrated in some duced in conventional cement kilns but require aluminium rich mi-
areas. This is a potentially promising approach in theory, but with nerals, sulphates and carbonates. CCSC uses widely available not so
current knowledge it is not possible to estimate its mitigation potential. pure limestone and silica, processed in conventional cement kilns. They
None of the new cement technologies are fully developed, but this is require curing in a CO2, rich environment capturing back the process
also the case for CCS and CCU. At this stage of knowledge, BYF and CO2. Therefore, their market is limited to industrialised concrete plants,
CCSC seem to be the most promising technologies from a CO2 mitiga- particularly for thin or porous sections, without steel reinforcement.
tion perspective. The mitigation potential from alkali activated mate- MOMS is a promising approach in terms of CO2 reduction but the
rials (AAMs) is conditional on the availability in the world market of technology is not yet developed.
sodium silicate with much lower CO2 footprint than that currently Alkali activated materials produced with fly ash and blast furnace
produced, and formulations with low amounts of sodium silicate, pre- slag have low CO2 footprint, but their mitigation potential is dubious
ferably with no need for thermal curing. AAMs produced with GBFS and since they will mostly divert slag and fly ash from Portland cement.
FA will be more effective when they use materials that are not in de- AAMs produced with calcined clay are a scalable technology. However,
mand for use in conventional Portland cement clinker blends. they can only contribute to global CO2 reduction if the CO2 footprint of
It worth mentioning that the mitigation potential of these technol- sodium silicate can be at least halved. Scaling up the use of geopoly-
ogies may be enhanced by the partial replacement of the binder by mers is also dependent on significant investments to increase sodium
engineered fillers and adequate dispersants. silicate production.

19
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

An increase of cement use efficiency can be achieved by promoting result in more prestigious publications, it does not contribute in ways
more industrial use of cement such as ready-mix concrete, dry-mix that might be of greater use to society as a whole — for example, de-
mortars and precast components. Industrialisation can accelerate the veloping and advancing the necessary methodologies for improving
introduction of higher amount of fillers in replacement of binders. Use performance standards for cement-based materials.
efficiency should work with both conventional or new cement tech- It is imperative that better research links be built between academia
nologies. New design and processing methods can also offer mitigation and industry. A pioneering example in this area is the Nanocem
opportunities for both concrete structures and components. Industrial-Academic partnership based in Europe (www.nanocem.org).
Clinker-substitution based technologies seem to be more attractive Launched > 12 years ago, it brings together the world's leading con-
for regions where the cement industry is expanding, because they allow struction materials and construction chemical producers with the
mitigation meanwhile preserving investments. In regions in which ex- leading European academic research groups. Because the consortium is
isting kiln facilities are capable of supplying the future demand for pre-competitive, it focuses on fundamental scientific questions under-
cement, with an already highly industrialised market, new cement lying the behaviour of cement and concrete, which are important to
technologies that use existing conventional kilns as well as CCS may be facilitate downstream developments in the field. The network funds
more attractive. “core projects”, which aim to fill gaps in existing research. The inter-
It is possible to achieve the 2 °C mitigation target without CCS, if the action between industrial and academic researchers has an impact
emerging technologies presented are further developed and adopted at which goes beyond these directly funded projects — enabling the aca-
industrial scale. Investment on R&D is therefore a priority. demic groups to secure other funding sources for high quality research
relevant to applications.
15. Research needs This report identifies the two key areas with the greatest potential
for reducing the cement and concrete materials sector's CO2 emissions
This working group's primary objective has been to review tech- over the next 2–3 decades:
nologies proposed for CO2 reduction in the cement and concrete sector
to identify promising areas where research efforts should be en- 1. Extending the use of supplementary cementitious materials (SCMs)
couraged, and also where the investment of research funding is unlikely in cement to further reduce clinker content, chiefly by developing
to be worthwhile for technologies with very low potential. This is im- technology for the combined addition of calcined clay and lime-
portant because many research journal submissions reveal that re- stone.
searchers are often repeating work and obtaining results that are al- 2. Reducing concrete's clinker content by improving mix designs that
ready well known — this is a waste of money and human creativity. allow for increased filler content, which can be added either via the
A considerable amount of research will still be needed to meet the cement or directly during concrete mixing.
challenge of lowering CO2 emissions in the sector. Regrettably, the
sector's existing level of research funding is extremely low. For ex- In both of these cases the main research needs include:
ample, LafargeHolcim, the world's largest cement company, is believed
to spend < 0.2% of its turnover on true R&D — and this percentage is 1. Mastering the workability of fresh concrete through control of par-
probably at the high end for the industry as a whole (Fig. 17). More- ticle packing and the use of appropriate dispersants. The issues of
over, the actual proportion of industry research spending is always robustness relative to variations in cement composition and to
small relative to development part. The concrete and construction in- concrete placing temperatures need to be resolved in a predictable
dustries invest even less on anything that might be called R&D. Much way.
University research is financed by national funding bodies, who often 2. Developing cost-effective methods for producing the necessary
have little understanding of what is realistic in real-world applications. particle size distributions for efficient multi-component binders.
This encourages academic researchers to focus more on scientifically 3. Establishing performance-based approaches between the service life
fashionable topics (e.g. “nanotechnology”). Although this work may and the embodied CO2 of the final products, allowing for specific

100 0

10

R² = 0.7233
80 20

30
India
Cement in Bulk (%)

Cement in Bags (%)

China 60 40
Western Europe
50
North America

North Africa 40 60
Middle East
70
Lan America

Central and Eastern Europe 20 80

Asia without China & India

90
Africa & Sub Saharan Africa

0 10 20 30 40 50 60 70 80 90 100 0 100
0 20000 40000 60000 80000
Bagged Cement (%)
GDP per capita (USD)

Fig. 17. Range of market share of bagged cement in various regions and key countries (top) and market share of bulk versus GDP (bottom). Bagged cement market
share are from LafargeHolcim Corporate Strategy Estimates as of Q1 2016. GDP are from World Bank [13]. Data covers countries responsible for about 90% of total
cement production.

20
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

370 to extend those service lives if needed.

Global Warming (gCO2-eq/m³)

NAC
RAC-50 It will be necessary to develop generic approaches based on un-
350
RAC-100 derstanding the characteristics and interactions of raw materials, their
hydration and microstructural development, which can be adapted to a
RAC-150
wide variety of real raw materials without the need for extensive local
330 empirical testing. This will necessitate the development, validation and
use of improved and advanced modeling tools to expand information
and design heuristics for broader usability, based on data generated for
310 a subset of all possible material compositions and combinations.
New advanced models to design large structures and simple com-
ponents are also of interest, particularly those associated to concepts
290 such as topology optimization and functionally graded materials, which
can be associated to digital production.
0 50 100 150 200
Aggregate Transport Distance (km)
Additionally, there is potentially still value in supporting some
Fig. 18. The effect of transportation distance on associated CO2 emissions of carefully targeted research on other binders that could have a sig-
concretes produced with recycled aggregates and natural aggregates. The nificant longer-term impact:
substantial difference results from tested recycled aggregates requiring much BYF clinkers: While there is as yet no cost-effective alternative to
higher cement content to reach the necessary design strength [46]. Portland cement clinker in the current economic environment, the most
feasible alternative class of hydraulic clinkers is belite-ye'elimite-ferrite
50 (BYF) clinkers, which have already been shown to offer, on an in-
dustrial scale, substantial CO2 reductions relative to Portland cement
Mud bricks clinker, but which suffer from higher raw materials costs. Further R&D
Carbon Intensitycs (kg.m-3.MPa-1)

40
Rammed earth in this area directed at improving the performance/cost ratio seems
justifiable on the grounds that BYF may allow us to progress further in
Self compact clay concrete terms of CO2 reduction than Portland cement clinker-based approaches
30 alone, and at a cost that remains well below that of CCS.
Compressed earth blocks
Technologies for alkali-activated binders that do not require the use
of blast-furnace slags: Alkali-activated binders are compromised by the
20 1000kg/m³
fact that most current practical technology depends on the supply of
granulated blast-furnace slag, (GBFS), a low-CO2 industrial by-product
500kg/m³ that is in limited supply and faces strong demand from other more
10 conventional applications, especially as an SCM for use in PCC-based
250kg/m³ cements designed for special exposure applications. The use of alkali-
100kg/m³ activation to valorise certain under-utilised coal and agricultural ash
0 resources, as well as non-ferrous slags and other industrial by-products,
has in recent years become a prolific field of research, although gen-
0 20 40 60 80 100
Compressive strength (MPa)
erating little generically-transferable information to provide broader
Fig. 19. CO2 intensity index of mud bricks (B), compressed earth blocks (CEB), understanding of underlying factors. This raises opportunities for useful
and rammed earth (RE) stabilised with 5–10% OPC, in comparison with data for and high-impact research. New alkali-activated binder technologies
PC concretes. Also included is the index of self-compacted clay concrete (SCCC) that do not require GBFS-for instance materials obtained by efficient
stabilised by 5% CSA cement. Note that the carbon footprint of the super- calcination of clays which would otherwise be of limited value due to
plasticiser (~40% of that of cement) has not been taken into account (adapted low purity or high iron content-are valued for research and for com-
from Fig. 5 of (Damineli et al. (2010))).
mercialisation if they can be shown to have true industrial potential for
replacing significant amounts of Portland cement clinker in concretes
applications and use-environments. This would entail developing or while retaining a lower net carbon footprint. The durability perfor-
adapting service life prediction test methods for new systems that mance of AAMs in the field also requires significant further in-
may differ from current conventions, as well as supporting research vestigative research.

1 1
Concrete usage index (m³/m²)
Concrete usage index (m³/m²)

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2

0 0
0 5000 10000 15000 20000 25000 0 5000 10000 15000 20000 25000
Cumula!ve Floor Area (m²) Cumula!ve Floor Area (m²)

Fig. 20. Example of concrete usage index for buildings from Singapore. Residential buildings represented in the top chart segment, commercial buildings in the
bottom [51].

21
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

1000 − The use of self-compacting soil concretes, which have a high de-
Total mi!ga!on target mand for expensive admixtures but add little value to more con-
800 ventional soil-based construction technologies.
− Using 3D printing for general cement-based construction applica-
CO2 Mitigation Potential (Mt)

tions. It is a useful technique for highly specialised applications, but

seems unlikely to have much value in reducing global CO2 emis-
600
CCS mi!ga!on target
sions.
400
16. Education and sustainable construction
200
Civil engineering and architecture curricula were defined in simpler
Calcined Clay + OPC times, when concrete was made from simple “pure” Portland cement,
0
Filler + OPC aggregates and water. Most civil engineers would spend their working
time doing structural design and complex calculations without the aid
-200 of computers. Design requirements were primarily concerned with
0 10 20 30 40 50 60 structural safety. The number of materials and components in-
Market Share 2050 (%) corporated in a building or a bridge were limited. Teaching materials is
often just a question of explaining a few standards. The education of
Fig. 21. Mitigation potential as a function of the market share for the 2050 low
demand scenario of (a) a combination of 25–35% calcined clay plus 15% of civil engineers continues to focus on developing the skills needed for
limestone filler; (b) filler substitution, varying between 25 and 35%. Horizontal structural design.
lines present as reference the mitigation targets from the 2DS scenario for CCS In the last 30–40 years the engineering process has changed dra-
and in total. matically. Nowadays, structural design is handled with sophisticated
computer models. Productivity in the field is much higher and a smaller
proportion of engineers are involved in this very specialised field.
1000 Minimising environmental impact over the construction life cycle is
Total mi!ga!on target now an implicit part of the process. New materials and industrialised
800 components are coming to the market at a rapid speed. Variability will
increase faster because minimal environmental impact will require lo-
CO2 mitigation Potential (Mt)

cally optimised solutions. Construction requirements have become

600
CCS mi!ga!on target much more complex, specifying very abstract performance require-
ments reaching as far out as the end of service life. Ethics, social equity,
400 and users' wellbeing are now part of doing business day-to-day. Current
AA Calcined Clay training curricula are not sufficient to equip industry professionals to
200 BFY navigate the unknown future, delivering the high-quality eco-efficient
built environment that society demands. Educating civil engineers and
CCSC
architects has to envision a better common future for humanity. This
0
issue is discussed in detail in the white paper of Schmidt et al.
AA fly ash

The materials dimension of construction is now much more im-

-200 portant, because construction uses > 50% of the materials extracted
0 10 20 30 40 50 60
from nature. Cement based materials alone account for more than one
Market Share 2050 (%) third of raw materials and 5–10% of the anthropogenic CO2. Scientific
Fig. 22. Rough estimative of mitigation potential for new cement technologies understanding — of materials in general and cement-based materials in
as a function of the market share for the 2050 LD scenario. These technologies particular — has advanced significantly, encompassing solutions to si-
are at earlier-stage development and so there is greater uncertainty. Horizontal multaneously improve the built environment while reducing con-
grey line represents the mitigation target from the 2DS scenario for CCS. Dotted struction's environmental impact. This knowledge is not yet reflected in
line represents uncertainty on the estimative. education curricula.
Deploying the technologies described in this report will require
professionals capable of understanding not only the environmental as-
pects faced, but also the fundamentals of the material problem. The
Clinkers made using globally abundant ultramafic rocks instead of education of civil engineers and architects regarding building and ce-
limestone as the main raw material: In theory this approach has the mentitious materials needs to change.
advantage over all limestone-based technologies in that it could be truly Instead of focusing on teaching standard solutions for standard
carbon-negative. No feasible energy-efficient industrial manufacturing problems, university curricula need to teach materials fundamentals to
process has yet been invented, although some progress has been re- prepare future professionals to face future problems and create corre-
ported recently. To date, very little research has been done in this area, sponding solutions. This shift will require a change of in the technical
so we believe it merits further research effort to determine whether or background of university professors and easier access to better educa-
not it is truly feasible, because the potential CO2 payoff could be large. tional materials.
These clinkers would have an even greater impact if used in applica- In short, most construction occurring in the near future will be the
tions where they can be hardened by carbonation rather than by hy- responsibility of professionals that have been educated to solve pro-
dration. blems of the past. Current professionals will be in charge of developing
Specific areas where we believe further fundamental research to be new technologies and updating today's construction standards, which
a poor investment with respect to environmental benefits include: may facilitate the introduction of new, more eco-efficient technologies.
A robust continuing education program must be designed targeting
− Using Portland cement clinker or lime to stabilise rammed earth practitioners in the field. The engagement of professional associations
construction products. This emits more CO2 and gives worse prop- and cement industries of different countries is essential. Governments
erties than conventional concretes. in particular could facilitate the transition to more eco-efficient

22
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

construction technologies by simply educating their staff in more ad- avoid up to 400 million tonnes of CO2 emissions annually, which could
vanced concepts, which in turn would enable public procurement to make the need for CCS less obvious.
accelerate the advances.
In developing countries, a significant part of the construction ac- 17.2. More efficient use of clinker in concrete and mortar
tivity and the cement use is still accomplished by untrained personnel,
most of whom are self-help builders. Basic training on housing essen- In concretes and mortars a similar magnitude of CO2 emissions re-
tials and how to handle construction materials and cement-based ma- ductions of are possible:
terials in particular could have tremendous potential not only for CO2
mitigation, but also for improving their wellbeing and the quality of − Optimising mix design can improve the eco-efficiency, defined in
cities. This topic has so far been neglected by the international com- terms of CO2 per MPa compressive strength, by a factor of 4 when
munity, including building materials producers. This has to consider comparing best practice with worst. Careful optimisation of particle
local problems and solutions; but a core set of fundamentals could be packing, throughout both the coarse and fine fraction of the ce-
developed internationally, with new ways of getting this to the right mentitious materials, coupled with the use of dispersants and the
people. In coming decades, increased cement production is projected to use of fillers can further reduce clinker contents while maintaining
occur primarily in developing regions of the world. This is an oppor- product performance.
tunity to directly apply best practices and increase eco-efficiency. Swift − Using high strength concrete grades, where appropriate in structural
action is required to take advantage of this chance to enhance user applications is more efficient and can reduce overall materials
education and awareness that would result in more sustainable consumption.
building. − Industrialising concrete and mortar production (i.e. ready-mix, dry-
The potential of digital technology — including ubiquitous mobile mortars) compared to poorly controlled on-site mixing, furthers
phones, virtual video channels, e-books and the new MOOCs (Massive substantial savings by avoiding wastage, particularly in urban areas.
Open On-line Courses) — offer new opportunities for a global education
platform. 17.3. New cement technologies could contribute significantly in the longer
The engagement of governments, NGOs, materials producers and term
the stakeholders in an international effort for better education in con-
struction must be a priority. Non-Portland clinkers may offer promising options for the longer
term, but there is as yet no cost-effective alternative to Portland cement
17. Main conclusions clinker in the current economic environment. The most feasible alter-
native class of hydraulic clinkers is belite-ye'elimite-ferrite (BYF) clin-
There are two key areas that can deliver substantial additional re- kers, which present substantial CO2 reductions relative to Portland
ductions in global CO2 emissions related to cement and concrete, cement clinker. Though this approach has higher raw materials costs
minimising the need for costly investment in CCS over the next than SCM and filler approaches it is still significantly less than CCS.
20–30 years: Further R&D in this area is needed to improve the performance to cost
ratio.
1. Increased use of low-CO2 supplements (supplementary cementitious Among non-clinker based cements, alkali-activated binder technol-
materials or SCMs) as partial replacements for Portland cement ogies (AAM) also have the capacity to reduce global CO2 emissions
clinker. beyond what is possible with optimised use of SCMs and fillers.
2. More efficient use of Portland cement clinker in mortars and con- However, many current AAM technologies require the use of GBFS to
cretes. give acceptable performance, and in many locations it is simpler to use
the limited (global) supplies of GBFS as conventional SCMs. The use of
We believe that Portland cement clinker based cements will dom- alkali metal silicates as activators is also both capital- and energy-in-
inate in the near future due to the economy of scale, level of process tensive, if these are produced through conventional process routes. We
optimisation, availability of raw materials and market confidence in therefore believe that the AAM approach requires further R&D if it is to
these products. In the longer term, other emerging alternative tech- have any chance of success in global CO2 mitigation.
nologies could also play a role in emissions mitigation that conse- Newly developed clinker technologies, in which concrete products
quently merit further investigation. are produced by carbonation rather than hydration, can reduce net CO2
emissions up to 70% compared to Portland cement clinker, and are
17.1. Increased use of clinker substitutes (SCMs) in Portland cement clinker already commercially available in some locations. Unfortunately, they
based cements suffer from severe commercial limitations because they require devel-
oping a circular economy for captured CO2, and also because they are
Today's cements contain on average only around 20% of SCMs limited to factory-made products. We therefore believe that they are
substituting Portland cement clinker — mainly fine limestone, granu- unlikely to have a major global CO2 impact as a direct alternative to
lated blast-furnace slags (GBFS) and coal fly ashes (FA). GBFS and FA Portland cement, as the facility to cast cementitious materials on-site is
sources of adequate quality are limited globally to only about 15–25% key to their ubiquitous use in construction.
of cement consumption and are unlikely to increase. A recently devel- Finally, we think that there is still some chance for a breakthrough
oped alternative low-CO2 SCM system uses optimised combinations of in the area of clinkers made using globally abundant ultramafic rocks
calcined clays with ground limestone. Such combinations represent a instead of limestone as the main raw material. In theory, this approach
relatively inexpensive and widely available SCM source capable of re- has the advantage over all limestone-based technologies in that it could
placing up to 50% of clinker while maintaining similar performance to be truly carbon-negative; but no feasible energy-efficient industrial
existing cements. Additionally, a significantly-increased filler content manufacturing process has yet been invented, although recently some
above today's average of 6% is technically feasible by combining par- progress has been made. We consider that this area merits further re-
ticle size control and dispersant admixtures, resulting in cements with search in view of its significant potential for CO2 reduction.
low water demand. In some applications, filler contents as high as 50%
in the cement can offer satisfactory performance. Increasing the 17.4. Requirement for research, coordination and raising awareness
average level of clinker substitution in cement to reach 40%, for in-
stance, through the use of the above-mentioned alternatives could More efficient global use of all possible approaches to low-CO2

23
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

cementitious materials, will need, among other things, flexible and Michael Thomas - University of New Brunswick, Canada
robust performance-based standards for cement and concrete. Sui Tongbo - Sinoma Research Institute, Beijing, China
Developing such standards will require a well-coordinated international Tian Qian - Jiangsu Sobute New Materials Co, Nanjong, China
research effort, as well as strong coordination between the industry, Henri VanDamme - MIT (formerly IFSTTAR, France), Cambridge,
standard making bodies, regulators and society at large to raise USA
awareness and create market acceptance for eco-efficient solutions. Yu Cheng - Jiangsu Sobute New Materials Co, Nanjong, China

17.5. Governmental engagement Acknowledgments

Governmental engagement will be important to the development The authors would like to thank: UNEP SBCI staff and directors,
and implementation of a successful mitigation strategy in the cement World Business Council for Sustainable Development — Cement
industry. The cement value chain makes up a large proportion of all Sustainability Initiative (WBCSD-CSI), International Energy Agency
economies, including a range of stakeholders from large companies to (IEA), LafargeHolcim, State of São Paulo Research Foundation
individuals. Raising awareness in such a complex environment will (FAPESP) 2016/052785, SINOMA, GCP Applied Technologies, and
require commitment from governments. InterCement for their support of this work.
Governments also have influence on educational policies, both in
undergraduate and graduate civil engineering and architecture courses, Annex 1. Glossary
which will have to be reconfigured in ways to make it possible for the
construction industry, including cement-based materials industries, to AAM alkali activated material
cope with the demands of sustainable development. BAT best available technology
Research, development and innovation are strongly influenced by BAU business as usual
governments, not only through funding to academic basic research, but BYF belite ye'elimite ferrite
also by promoting alliances between academy and industry and sti- CCS carbon capture and storage
mulating innovation at the industrial level. Governments are also fre- CCSC carbonatable calcium silicate clinker
quently in a position to influence standardisation processes. CCU carbon capture and usage
Promoting the industrialisation of the cement supply chain most CSA calcium sulfo aluminate
certainly depends on actions of governments. In developing countries CSI Cement Sustainability Initiative (of the WBCSD)
this will require actions to reduce the economic advantage associated FA fly ash
with the use of aggregates from the informal market, which favours the GBFS granulated blastfurnace slag
inefficient use of cement, increasing CO2 footprint. Other options in- GNR getting the numbers right (database of CSI)
clude actions to limit the use of bagged cement, as already done by IEA International Energy Agency
China. IPCC International Panel on Climate Change
Finally, the mitigation potential of each technology will depend on MOMS magnesium oxides derived from magnesium silicates
its success in the market. Governments are among the largest consumers OECD Organisation for Economic Co-operation and Development
of cement based materials, especially when investing in infrastructure. OPC ordinary Portland cement
Therefore, the use of public purchase power can be decisive in accel- PC Portland cement (referred to as OPC in some countries, CEM I
erating market penetration of these mitigation technologies. under European norm EN 197-1)
R&D Research and Development
Group members and Report reviewer SCM supplementary cementitious material
UNEP–SBCI United Nations Environmental Programme–Sustainable
Mark Alexander - University Cape Town, Cape Town, South Africa Buildings and Climate Inititative
Yunus Ballim - University Cape Town, Cape Town, South Africa WBCSD World Business Council for Sustainable Development
Bruce Blair - formerly at Lafarge, USA
Curt Carrigan - UN Environment, Paris, France References
Cristiana Ciaraldi - CSI, World Business Council for Sustainable
Development, Geneva, Switzerland [1] L. Szabó, I. Hidalgo, J.C. Císcar, A. Soria, P. Russ, Energy Consumption and CO2
Anjan K. Chatterjee - Indian Cement & Concrete Industry, Kolkata, Emissions from the World Cement Industry, European Commission Joint Research
Center (DG JRC), 2003, http://www.bvsde.paho.org/bvs-aia/fulltext/energy.pdf.
India [2] SNIC, Relatório Anual 2012–2012 Annual Report, Sindicato Nacional da Indústria
Josephine Cheung -GCP Applied Technologies, Cambridge, USA do Cimento/Brazilian Cement Industry Union, Rio de Janeiro, 2013http://www.
Philippe Fonta - CSI, World Business Council for Sustainable snic.org.br/pdf/relatorio_anual_2012-13_web.pdf.
[3] UN Population Division, World Population Prospects: The 2015 Revision, United
Development, Geneva, Switzerland Nations Department of Economic and social Affairs, New York, 2015.
Ravindra Gettu - Indian Institute of Technology, Madras, India [4] World Steel Association, WorldSteel in Figures 2015, Brussels, (2015).
Jacob Halcomb - UN Environment, Paris, France [5] CEMBUREAU, World Statistical Review 1996–2008 Cement Production, Trade,
Consumption Data, The European Cement Association, Brussels, 2010.
Arpad Horvath - University of California Berkeley, USA [6] CEMBUREAU, Activity Report 2015, Brussels, http://www.cembureau.be/sites/
Edgardo Fabián Irassar - Universidad Nacional del Centro de Buenos default/files/AR2015.pdf, (2016).
Aires, Olavarría, Argentina [7] FAO, Food Wastage Footprint Impacts on Natural Resources: Summary Report,
Food and Agriculture Organisation of the United Nations, FAO, [Rome], 2013.
Fernando Martirena - Universidad Central de Las Villas, Santa Clara,
[8] F. Krausmann, S. Gingrich, N. Eisenmenger, K.-H. Erb, H. Haberl, M. Fischer-
Cuba Kowalski, Growth in global materials use, GDP and population during the 20th
Sabbie Miller - University of California Davis, USA century, Ecol. Econ. 68 (2009) 2696–2705, https://doi.org/10.1016/j. ecolecon.
Sergio Almeida Pacca - University of São Paulo, São Paulo, Brazil 2009.05.007.
[9] J.K. Steinberger, F. Krausmann, N. Eisenmenger, Global patterns of materials use: a
John Provis - University of Sheffield, UK socioeconomic and geophysical analysis, Ecol. Econ. 69 (2010) 1148–1158, https://
Wolfram Schmidt - BAM Federal Institute for Materials Research doi.org/10.1016/j. ecolecon.2009.12.009.
and Testing, Berlin, Germany [10] World Bank, World Development Indicators, World DataBank, http://databank.
worldbank.org/data/reports.aspx?source=world-development-indicators, (2016) ,
Martin Schneider - VDZ - Research Centre of German Cement Accessed date: 11 August 2016.
Industry, Düsseldorf, Germany [11] IEA, Energy Technology Perspectives 2015—Mobilizing Innovation to Accelerate
Surendra Shah - Northwestern University, Evanston, Illinois, USA Climate Action, Energy Technol. Perspect, http://www.iea.org/etp/etp2015/,

24
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

(2015) , Accessed date: 27 July 2016. http://www.google.com/patents/US20100043673, (2010) , Accessed date: 21 July
[12] WBCSD, Cement Sustainability Initiative, Getting the Numbers Right, Project 2016.
Emissions Report 2014, http://www.gnr-project.org/, (2016). [41] ASTM, Standard Specification for Ground-Calcium Carbonate and Aggregate
[13] T. Boden, B. Andres, G. Marland, Global CO2 Emissions From Fossil-Fuel Burning, Mineral Fillers for use in Hydraulic Cement Concrete, ASTM, West Conshohocken,
Cement Manufacture, and Gas Flaring: 1751–2013, Oak Ridge National Laboratory, 2016http://www.astm.org/cgi-bin/resolver.cgi?C1797M-16.
Oak Ridge, 2016http://cdiac.ornl.gov/ftp/ndp030/global.1751_2013.ems. [42] ACI Committee 211, Guide for Proportioning Concrete Mixtures with Ground
[14] N. Müller, J. Harnisch, A Blueprint for a Climate Friendly Cement Industry, http:// Limestone and Other Mineral Fillers, American Concrete Institute, Farmington
assets.panda.org/downloads/english_report_lr_pdf.pdf, (2008). Hills, 2015.
[15] IEA, WBCSD, Cement Technology Road-map 2009 Carbon Emissions Reductions up [43] R. Wassermann, A. Katz, A. Bentur, Mini-mum cement content requirements: a must
to 2050, OECD/IEA:; WBCSD, Paris:; Conches-Geneva, Switzerland, 2009http:// or a myth? Mater. Struct. 42 (2009) 973–982, https://doi.org/10.1617/s11527-
wbcsdcement.org/pdf/technology/WBCSD-IEA_Cement%20Roadmap.pdf. 008-9436-0.
[16] CEMBUREAU, Best Available Techniques for the Cement Industry, http://193.219. [44] C.T. Formoso, L. Soibelman, C. De Cesare, E.L. Isatto, Material waste in building
133.6/aaa/Tipk/tipk/4_kiti%20GPGB/40.pdf, (1999). industry: main causes and prevention, J. Constr. Eng. Manag. 128 (2002) 316 (doi:
[17] IEA, WBCSD, Technology Roadmap: Cement —Foldout, Int. Energy Agency, OECD/ 10.1061/(ASCE)0733-9364(2002)128:4(316)).
IEA; WBCSD, 2009, http://www.iea.org/publications/freepublications/ [45] M. Ramdani, Safi: La Guerre du Sable, L'Economiste, http://www.leconomiste.com/
publication/name,3862,en.html. article/safi-la-guerre-du-sable, (2005) , Accessed date: 6 August 2016.
[18] ECRA, IEAGHG, Deployment of CCS in the Cement Industry, Technical Report, [46] V. Beiser, The Deadly Global War for Sand, WIRED, http://www.wired.com/2015/
(2013). 03/illegal-sand-mining/, (2015) , Accessed date: 6 August 2016.
[19] K. Quillan, Calcium Sulphoaluminate Cements: Reduced CO2, Concrete Properties [47] B. HARTMAN, Police Arrest Five Contractors in Illegal “Sand Mining” Conspiracy,
and Applications. BRE Report, vol. 496, HIS BRE press, UK, 2007. Jerus. Post JPostcom, http://www.jpost.com/Israel-News/Police-arrest-five-
[20] C. Pade, M. Guimares, The CO2 uptake of concrete in a 100 year perspective, Cem. contractors-in-illegal-sand-mining-conspiracy-389760, (2015) , Accessed date: 6
Concr. Res. 37 (2007) 1348–1356. August 2016.
[21] P. Stemmermann, U. Schweike, K. Garbev, G. Beuchle, H. Möller, Celitement—a [48] R. Chevallier, Illegal Sand Mining in South Africa, South African Institute of
sustainable prospect for the cement industry, Cement International 8 (5) (2010) International Affairs, Johannesburg, 2014.
52–66. [49] A.B. Macedo, D.J. de Almeida Mello Freire, H. Akimoto, Environmental manage-
[22] T. Link, F. Bellmann, H. Ludwig, M. Ben Haha, Reactivity and phase composition of ment in the Brazilian non-metallic small-scale mining-sector, J. Clean. Prod. 11
Ca2SiO4 binders made by annealing of alpha-dicalcium silicate hydrate, Cem. (2003) 197–206, https://doi.org/10.1016/S0959-6526(02)00039-2.
Concr. Res. 67 (2015) 131–137. [50] National Development and Reform Commission, China's National Climate Change
[23] E. Gartner, M. Gimenez, V. Meyer, A. Pisch, A. Novel Atmospheric, Pressure ap- Program, China.org.cn, http://www.china.org.cn/english/environment/213624.
proach to the mineral capture of CO2 from industrial point sources, Proceedings of htm, (2007) , Accessed date: 6 August 2016.
the CCUS 2014-Carbon Capture, Utilization and Storage Conference, Pittsburgh, [51] Freedonia, Cement in China (Brochure), Freedonia Group, Cleveland, 2009http://
PA, USA, 2014, https://doi.org/10.13140/2.1.1092.9600. www.freedoniagroup.combrochure/24xx/2430bro.pdf.
[24] B. Lothenbach, et al., Influence of limestone on the hydration of Portland cements, [52] UNIDO, Industrial Energy Efficiency for Sustainable Wealth Creation: Capturing
Cem. Concr. Res. 38 (6) (2008) 848–860. Environmental, Economic and Social Dividends, UNIDO, Vienna, 2011https://
[25] T. Matschei, B. Lothenbach, F.P. Glasser, The role of calcium carbonate in cement www.unido.org/fileadmin/user_media/Publications/IDR/2011/UNIDO%20IDR
hydration, Cem. Concr. Res. 37 (4) (2007) 551–558. %2reprint%20for%20web%20020912.pdf.
[26] R. Snellings, Assessing, Understanding and Unlocking Supplementary Cementitious [53] United Nations, Department of Economic and Social Affairs, Population Division,
Materials. RILEM Technical Letters, [S.l.], vol. 1, (Aug. 2016), pp. 50–55 ISSN World Urbanisation Prospects: the 2014 Revision: Highlights, (2014).
2518-0231. Available at: < https://letters.rilem.net/index.php/rilem/article/ [54] S. Marinković, V. Radonjanin, M. Malešev, I. Ignjatović, Comparative environ-
view/12 > . Date accessed: 20 Oct. 2016 10.21809/rilemtechlett.2016.12. mental assessment of natural and recycled aggregate concrete, Waste Manag. 30
[27] H. Meissner, Pozzolans used in mass concrete, in: T. Stanton, R. Blanks (Eds.), (2010) 2255–2264, https://doi.org/10.1016/j.wasman.2010.04.012.
Symp. Use Pozzolanic Mater. Mortars Concr., ASTM International, 100 Barr Harbor [55] M. Quattrone, S.C. Angulo, V.M. John, Energy and CO2 from high performance
Drive, PO Box C700, West Conshohocken, PA 19428-2959, 1950 pp. 16–16–15 recycled aggregate production, Resour. Conserv. Recycl. 90 (2014) 21–33, https://
http://www.astm.org/doi-Link.cgi?STP39402S , Accessed date: 3 April 2016. doi.org/10.1016/j.rescon rec.2014.06.003.
[28] R. Fernandez, F. Martirena, K.L. Scrivener, The origin of the pozzolanic activity of [56] S. Guignot, S. Touzé, F. Von der Weid, Y. Ménard, J. Villeneuve, Recycling con-
calcined clay minerals: a comparison between kaolinite, illite and montmorillonite, struction and demolition wastes as building materials: a life cycle assessment, J.
Cem. Concr. Res. 41 (1) (2011) 113–122. Ind. Ecol. 19 (2015) 1030–1043, https://doi.org/10.1111/jiec.12262.
[29] M. Antoni, J. Rossen, F. Martirena, K. Scrivener, Cement substitution by a combi- [58] C. Ulsen, H. Kahn, G. Hawlitschek, E.A. Masini, S.C. Angulo, Separability studies of
nation of metakaolin and limestone, Cem. Concr. Res. 42 (2012) 1579–1589, construction and demolition waste recycled sand, Waste Manag. 33 (2013)
https://doi.org/10.1016/j.cemconres.2012.09.006. 656–662, https://doi.org/10.1016/j.wasman.2012.06.018.
[30] F. Biedermann, I. Obernberger, Ashrelated Problems During Biomass Combustion [59] UNEP SBCI, Common Carbon Metric for Measuring Energy Use & Reporting
and Possibilities for a Sustainable Ash Utilisation, WREC 2005, ScotlandElsevier Greenhouse Gas Emissions from Building Operations, UNEP, Paris, 2010http://
Ltd., Aberdeen, 2005, p. 8 http://www.biosbioenergy.at/uploads/media/Paper- www.unep.org/sbci/pdfs/Common-Carbon-Metric-for_Pilot_Testing_220410.pdf.
Biedermann-Ash Related-2005-10-11.pdf. [60] The Centre of Sustainable Buildings and Construction, Building and Construction
[31] I. Obernberger, K. Supancic, Possibilities of ash utilisation from biomass combus- Authority, Sustainable Construction: A Guide on Concrete Usage Index, Building
tion plants, ETA = Renewable Energies, Hamburg, 2009, pp. 2373–2384 http:// and Construction Authority, Singapore, 2012.
www.bios-bioenergy.at/uploads/media Paper-Obern berger-ash-utilisation-2009. [61] Lixuan Hong, David Fridley, Wei Feng, Nina Khanna, Modeling China's Building
pdf. Floor Area Growth and Its Implications for Building Material and Energy Demand,
[32] R. Davis, W. Hanna, E. Brown, Strength, volume changes, and sulfate resistance of ACEEE Sumer Study Energy Effic. Build., American Council for an Energy-Efficient
mortars containing Portland-Pozzolan cements, in: T. Stanton, R. Blanks (Eds.), Economy, Berkeley, (2014), pp. 10-146–10-157 (http://aceee.org/files/proceed-
Symp. Use Pozzolanic Mater. Mortars Concr., ASTM International, West ings/2014/data/papers/10–230.pdf#page=1).
Conshohocken, PA 19428-2959, 1950 pp. 131-131-22 http://www.astm.org/ [62] A. Warszawski, Analysis of cement consumption in a national economy, Build. Sci.
doiLink.cgi?STP39407S , Accessed date: 26 March 2016. 10 (1975) 9–15, https://doi.org/10.1016/0007-3628(75)90003-1.
[33] R.E. Davis, J.W. Kelly, G.E. Troxell, A.H.E. Davis, Properties of mortars and con- [63] A.S.S. Rizk, Structural design of reinforced concrete tall buildings, CTBUH J. (2010)
cretes-containing Portland-Puzzolan cements, J. Proc. 32 (1935) 80–114. 34–41.
[34] C. Müller, S. Palm, C.-A. Graubner, T. Proske, S. Hainer, M. Rezvani, W. Neufert, [64] J. Pacheco, L. Doniak, M. Carvalho, P. Helene, The paradox of high performance
I. Reuken, Cements With a High Limestone Content-Durability and Practicability, concrete used for reducing environmental impact and sustainability increase,
VDZ, https://www.vdz-online.de/fileadmin grup-pen/vdz/3LiteraturRecherche/ Second Int. Conf. Concr. Sustain. ICCS16, Universidad Politécnica de Madrid, 2016,
Betontechnische_Berichte/2004_2006/Dauerhaftigkeit_von_Beton_mit_ pp. 442–453.
Portlandkalkstein.pdf, (2016). [65] G. Habert, D. Arribe, T. Dehove, L. Espinasse, R. Le Roy, Reducing environmental
[35] M. Zajac, S.K. Bremseth, M. Whitehead, M. Ben Haha, Effect of CaMg(CO3)2 on impact by increasing the strength of concrete: quantfication of the improvement to
hydrate assemblages and mechanical properties of hydrated cement pastes at 40 °C concrete bridges, J. Clean. Prod. 35 (2012) 250–262, https://doi.org/10.1016/j.
and 60 °C, Cem. Concr. Res. 65 (2014) 21–29. jclepro.2012.05.028.
[36] J.D. Bapat, Conserving Limestone, Indian Cem. Rev. Mag. http://www.indiance- [66] S. Tae, C. Baek, S. Shin, Life cycle CO2 evaluation on reinforced concrete structures
mentreview.com/News.aspx?nId=jb8EiU9Vpk9v+OMmqmHkTg, (2014) , with high-strength concrete, Environ. Impact Assess. Rev. 31 (2011) 253–260,
Accessed date: 14 August 2016. https://doi.org/10.1016/j. eiar.2010.07.002.
[37] IARC, Working group on the evaluation of carcinogenic risks to humans, I.A. for R. [67] ERMCO, Ready Mixed Concrete Production Statistics 2013, European Ready Mixed
on C. conference, World Health Organisation, Silica Dust, Crystalline, in the Form of Concrete Industry, Brussels, 2014http://www.ermco.eu/document/ermco-
Quartz or Cristobalite, Rev. Hum. Carcinog. International Agency for Research on statistics-2013-pdf/ , Accessed date: 6 August 2016.
Cancer, Lyon, 2012, pp. 355–405. [68] L.L. Stromberg, A. Beghini, W.F. Baker, G.H. Paulino, Application of layout and
[38] B.L. Damineli, F.M. Kemeid, P.S. Aguiar, V.M. John, Measuring the eco-efficiency of topology optimization using pattern gradation for the conceptual design of build-
cement use, Cem. Concr. Compos. 32 (2010) 555–562, https://doi.org/10.1016/j. ings, Struct. Multidiscip. Optim. 43 (2010) 165–180, https://doi.org/10.1007/
cemconcomp.2010.07.009. s00158-010-0563-1.
[39] R.G. Pileggi, V.C. Pandolfelli, Rheology and particle size distribution of pumpable [69] C.M.R. Dias, H. Savastano Jr, V.M. John, Exploring the potential of functionally
refractory castables, Am. Ceram. Soc. Bull. 80 (2001) 52–57. graded materials concept for the development of fiber cement, Constr. Build. Mater.
[40] J.-F. Batoz, M. Behloul, P. Fonollosa, Concrete Compositions, US20100043673 A1, 24 (2010) 140–146.

25
K.L. Scrivener et al. Cement and Concrete Research 114 (2018) 2–26

[70] Y. Xi, K. Willam, D.M. Frangopol Ababneh, A. Nakhi, J.S. Kong, C.L. Nogueira, annual CO2 emissions from the production of concrete by over 20%, Environ. Res.
Accelerated testing and modeling of concrete durability subjected to coupled en- Lett. 11 (2016) 74029, , https://doi.org/10.1088/1748-9326/11/7/074029.
vironmental and mechanical loading, Long Term Durab. Struct. Mater. Elsevier [76] S. Sánchez Berriel, A. Favier, E. Rosa Domínguez, I.R. Sánchez Machado, U. Heierli,
Science Ltd., Oxford, 2001, pp. 45–56 http://www.sciencedirect.com/science/ K. Scrivener, F. Martirena Hernández, G. Habert, Assessing the environmental and
article/pii/B9780080438900500067 , Accessed date: 10 September 2012. economic potential of limestone calcined clay cement in Cuba, J. Clean. Prod. 124
[71] International Energy Agency, Tracking Industrial Energy Efficiency and CO2 (2016) 361–369, https://doi.org/10.1016/j.jclepro.2016.02.125.
Emissions: In Support of the G8 Plan of Action: Energy Indicators, International [77] S.R.S.S. Vieira, J.E. Machado, Production and use of cements with artificial poz-
Energy Agency, Paris, France, 2007. zolan, Cement Plant Environmental Handbook, Second, International Cement
[72] D.R. Gomez, J.D. Watterson, B.B. Americano, C. Ha, G. Marland, E. Matsika, Review, 2015, pp. 156–159.
L.N. Namayanga, B. Osman-Elasha, J.D.K. Saka, K. Treanton, R. Quadrelli, Chapter [78] G. Habert, C. Ouellet-Plamondon, Recent update on the environmental impact of
2: energy: stationary, in: H.S. Eggleston, L.N. Namayanga, Chiky Kanky, Senryaku geopolymers, RILEM Tech. Lett. 1 (2016) 17–23.
Kenky Kikan (Eds.), IPCC Guidel. Natl. Greenh. Gas Invent., Hayama, Kanagawa, [79] W.P. Bolen, Soda Ash, Miner. Commod. Summ. 2016, U.S. Geological Survey,
2006, 2006 http://www.ipcc-nggip.iges.or.jp/public/2006gl/index.htm , Accessed Reston, (2016), pp. 154–155 https://doi.org/10.3133/70140094.
date: 3 October 2016. [80] E.M. Gartner, G. Walenta, V. Morin, P. Termkhajornkit, I. Baco, J.-M. Casabonne,
[73] M.J. Gibbs, P. Soyka, D. Conneely, Chapter 3 industrial processes: CO2 emissions Hydration of a Belite-Calciumsulfoaluminate-Ferrite Cement: AetherTM, 13th Int.
from cement production, in: A. Rosland (Ed.), Good Pract. Guid. Uncertain. Manag. Congr. Chem. Cem., Madrid https://www.research-gate.net/publication/
Natl. Greenh. Gas Invent., Intergovernmental Panel on Climate Change, Montreal, 260421752_Hydration_of_a_Belite-CalciumSulfoaluminate-Ferrite_cement_
2001 http://www.ipcc-nggip.iges.or.jp/public/gp/english/ , Accessed date: 3 AetherTM, (2011) , Accessed date: 23 September 2016.
October 2016. [81] S.K. Ritter, Making The Most of Red Mud, February 24, 2014 Issue, Vol. 92 Issue 8,
[74] A.P. Gursel, A. Horvath, GreenConcrete LCA Webtool, http://greenconcrete. Vol. 92 Chemical & Engineering News, C&en, 2014, http://cen.acs.org/articles/92/
berkeley.edu/concretewebtool.html, (2012) , Accessed date: 13 November 2014. i8/Making-Red-Mud.html , Accessed date: 23 September 2016.
[75] S.A. Miller, A. Horvath, P.J.M. Monteiro, Readily implementable techniques can cut

26