Material Economics Circular Economy

The Circular Economy
a Powerful Force for Climate Mitigation

Transformative innovation for prosperous
and low-carbon industry
TABLE OF CONTENTs
PREFACE
EXECUTIVE SUMMARY
Chapter 1.
THE CLIMATE POTENTIAL OF A CIRCULAR ECONOMY 8
Chapter 2.
STEEL – TOWARDS A CIRCULAR STEEL SYSTEM 54
Chapter 3.
PLASTICS – FINDING A PLACE WITHIN A LOW-CARBON ECONOMY 76
Chapter 4.
ALUMINIUM – REALISING THE POTENTIAL FOR RECYCLING 96
Chapter 5.
MOBILITY – THE PROMISE OF A SHARED CAR SYSTEM 116
Chapter 6.
BUILDINGS AND CEMENT – MORE VALUE FROM LESS MATERIALS 140
THE CIRCULAR ECONOMY – A POWERFUL FORCE FOR CLIMATE MITIGATION (2018)

2
preface
There is intense debate about how to close the gap To our knowledge, this is the first attempt to quanti-
between current climate policy and the long-term goal tatively address this issue across materials and value
of the Paris Agreement: an economy with net zero chains, and for the entire EU economy. As with any first
emissions. Within Europe, policy discussions and pro- effort, there are many uncertainties, and further analysis
posed ‘roadmaps’ to date have focused mainly on in- will be needed. Still, our analysis shows that the poten-
creasing energy efficiency and deploying low-carbon tial for emissions abatement is substantial, and in many
energy sources. Both are crucial, but when it comes to cases can be economically attractive. Indeed, we believe
industry, a major source of emissions, they offer only that ‘materials efficiency’ deserves to be a priority in dis-
a partial solution. Additional strategies are needed to cussions of industrial decarbonisation, just as energy ef-
address the substantial process emissions from the ficiency is a priority in discussions of transforming the
production of materials such as steel, cement, plastics energy system. A more circular economy can reduce CO2
and aluminium. emissions, reduce the scale of the challenge of decar-
bonising materials production, and contain the cost of
Most discussion has focussed on the supply side: the achieving an industrial base compatible with a low-car-
need to develop and deploy new industrial production pro- bon economy.
cesses, turn to non-fossil feedstock and fuels, and use
carbon capture and storage to offset any remaining emis- This project has been carried out by Material Economics,
sions. What is mostly missing from climate and industrial with the support of Climate-KIC, ClimateWorks, Ellen
debates is a discussion of the demand side: Can we make MacArthur Foundation, Energy Transitions Commission,
better use of the materials already produced, and so redu- European Climate Foundation, MAVA Foundation, and
ce our need for new production? SITRA. The Steering Group has comprised Mari Pantsar
(SITRA), Janez Potočnik (International Resource Panel),
The concept of the ‘circular economy’ offers precisely Martin Porter (European Climate Foundation), Adair Turner
this opportunity, through strategies such as recirculating (Energy Transitions Commission), Jocelyn Blériot (Ellen
a larger share of materials, reducing waste in production, MacArthur Foundation), Christoph Wolff and Simon Wolf
lightweighting products and structures, extending the life- (European Climate Foundation) and Anders Wijkman
times of products, and deploying new business models (Climate-KIC). Research guidance have been provided by
based around sharing of cars, buildings, and more. Dr Jonathan Cullen and Prof. Frank Geels. The project
team has included Anna Teiwik, Cornelia Jönsson, Stina
This report takes a first step towards quantifying the Klingvall, and Unn Hellberg. Material Economics is extre-
potential for circular economy opportunities to reduce mely grateful for all the contributions of these organisations
greenhouse gas emissions. It examines the key mate- and individuals, as well as many other researchers, policyma-
rials flows and value chains, identifies relevant circular kers, and business representatives who have contributed
economy approaches, and explores their cost-effective- their knowledge and insight to this project. Partner organi-
ness. It also considers barriers that might stand in the sations and their constituencies do not necessarily endorse
way of achieving a more circular economy, and policies all findings or conclusions in this report. All remaining errors
and initiatives that could overcome them. and omissions are of course the responsibility of the authors.
Per-Anders Enkvist Per Klevnäs
3
The Circular Economy – a Powerful Force for Climate Mitigation / Executive summary
executive summary
Industrial CO2 emissions are a major concern as Euro- The key conclusion is that a more circular economy
pe tries to achieve the deep emission reductions required can make deep cuts to emissions from heavy industry:
for its climate commitments. In the European Commis- in an ambitious scenario, as much as 296 million tonnes
sion’s ‘Roadmap 2050’, one-quarter of the CO2 emis- CO2 per year in the EU by 2050, out of 530 in total – and
sions remaining mid-century were from industry, especi- some 3.6 billion tonnes per year globally. Demand-side
ally from heavy industry producing basic materials. With measures thus can take us more than halfway to net-zero
more ambitious targets after the Paris Agreement, the EU emissions from EU industry, and hold as much promise
must now articulate how to combine net-zero emissions as those on the supply side. Moreover, they are often
with a prosperous industrial base. economically attractive.
So far, discussions of industry emissions have focussed Opportunities for more productive use of materials
on the supply side: reducing the emissions from the produc- therefore deserve a central place in EU climate policy.
tion of steel, cement, chemicals, etc. Far less attention has Much like improving energy efficiency is central to the
been given to the demand side: how a more circular eco- EU’s efforts to achieve a low-carbon energy system, a
nomy could reduce emissions through better use and reuse more circular economy will be key to developing Europe-
of the materials that already exist in the economy. This study an industry while cutting its CO2 emissions. As industry
aims to bridge that gap. It explores a broad range of opportu- associations and the European Commission consider
nities for the four largest materials in terms of emissions (steel, new mid-century ‘roadmaps’ for industry, they should in-
plastics, aluminium and cement) and two large use segments clude circular economy measures for cost-effective ways
for these materials (passenger cars and buildings). to achieve deep emissions cuts.
4
A more circular economy can cut
emissions from heavy industry by 56% by 2050
EU EMISSIONS REDUCTIONS POTENTIAL FROM A MORE CIRCULAR ECONOMY, 2050

MILLION TONNES OF CARBON DIOXIDE PER YEAR
STEEL PLASTICS ALUMINIUM CEMENT
530 178
-56 %
56
62
234
2050 BASELINE MATERIALS PRODUCT MATERIALS CIRCULAR 2050 CIRCULAR

RECIRCULATION EFFICENCY BUSINESS MODELS SCENARIO
MATERIALS PRODUCTION EMISSIONS, 2015 AND 2050

MILLION TONNES OF CARBON DIOXIDE PER YEAR, EU
5
564
530
Demand-side opportunities could reduce EU industri- are wasted in construction. Another opportunity is to use
al emissions by almost 300 Mt per year by 2050, or 56 more advanced materials and construction techniques,
%, with attractive economics. These abatement oppor- such as high-strength steel that can cut materials use by
tunities fall into three major categories: 30%. There also are opportunities to reduce over-speci-
fication, such as the near 100% overuse of steel in buil-
A. Materials recirculation opportunities (178 Mt per dings relative to what is strictly required to meet design
year by 2050). The EU economy is accumulating large specifications. Further gains can be achieved by tailo-
stocks of metals and plastics, and by 2050 could meet ring products better to specific uses; for example, to the
a large share of its need for these materials by recircu- extent that fleets of shared cars can replace individual
lating what has already been produced: 75% of steel, ownership (see below), many of the cars needed will be
50% of aluminium, and 56% of plastics (cement is less smaller, just big enough for a one- or two-passenger trip
amenable to recycling, although it is possible to reuse in the city. Companies already have incentives to use the-
some unreacted cement). Recirculating materials cuts se strategies to some extent, but some opportunities are
CO2 emissions and requires much less energy than new missed through split incentives in complex supply cha-
production does. However, current practice is not set up ins. Many measures will become much more economic
to facilitate these high recycling rates. An influx of new with greater digitalisation in the mobility and buildings
materials is required both to replace metals and plas- value chains and other technological development now
tics that are lost, and to compensate for downgrading underway.
of quality. In some cases, metals are mixed or downg-
raded because materials specialisation requires it, but C. New circular business models in mobility and buil-
there also are many cases where this could be avoided dings, notably through sharing (62 Mt per year by 2050).
or much reduced. For steel, the key is to ensure much This opportunity pivots on making much greater use of
cleaner scrap flows that allow for high-quality secon- vehicles and buildings, which together represent a majo-
dary steel, and less pollution of steel with copper; for rity of European demand for steel, cement and alumini-
aluminium, smaller losses and less mixing of different um. Currently, the utilisation of many of these assets is
alloys will be crucial. Mixing and downgrading effects very low: about 2% for the average European car, and
are particularly serious problems for plastics, making about 40% for European offices, even during office hours.
a large share of used plastics literally worthless. This Sharing enables much more intensive use. For vehicles,
study shows how 56% of plastics could be mechanically this in turn means that higher upfront costs of electric dri-
recycled, with a focus on the five main types of plastic vetrains, more advanced automation technology, or hig-
that account for 70% of volumes. The aim must be to her-performance materials can be paid back over many
move these to a tipping point where recycling is econo- more miles. In addition, professionally managed fleets of
mically viable, driven by the inherent material value. For such higher-value cars are more economical to maintain,
this, as with steel and aluminium, product design and reuse, remanufacture and recycle. Vehicle lifetimes, on
end-of-life disassembly need to change to enable high- a per-kilometre basis, can thus increase drastically. The
value recovery. result is a self-reinforcing loop of incentives for higher
utilisation, lower-carbon energy, and less materials use.
B. Product materials efficiency (56 Mt per year by In a circular scenario, the materials input to mobility falls
2050). These opportunities have in common that they by 75%. It also brings many other benefits, including a
reduce the total materials input to key products. One much lower total cost of travel. Sharing models are taking
strategy is to reduce the amount of materials that are root by themselves, but much more could be done to
lost in production: for example, half the aluminium pro- accelerate their growth, and to find ways to resolve the
duced each year does not reach the final product, but concerns that have arisen with some early iterations of
becomes scrap, while some 15% of building materials such business models.
6
Many of these abatement opportunities are economi- bon economy must be much more circular than today’s.
cally attractive on their own terms, provided that we are While this study has focussed on Europe, an extrapola-
willing to organise the mobility and real estate sectors tion to other world regions suggests that the measures
somewhat differently in the future. Many others cost less identified could contribute 3.6 Gt CO2 per year to global
than 50 EUR per tonne CO2 avoided, less than most other efforts to cut greenhouse gas emissions by 2050. The
ways to reduce these emissions, including supply-side claim on the carbon budget could be reduced by 333 Gt
measures for industry. Circularity is strongly aligned with by 2100. In this setting, the additional supply-side me-
the digitalisation trend that is sweeping across industry; asures required start to look manageable, and a well-
for example, digitalisation means it is ever cheaper to below 2 °C objective within reach.
keep track of complex supply chains and material flows,
optimise sharing business models, and automate mate- Achieving these opportunities is doable and requires
rials handling in construction. A more circular economy ‘energy efficiency-type’ interventions. Many of the abate-
would have many other benefits as well, such as reduced ment opportunities identified are low-cost or even profita-
geopolitical risks, local job creation, lower air pollution, ble, but are held back by multiple barriers. For example,
and reduced water use. They therefore can contribute to product manufacturers lack incentives to enable high-
several of the Sustainable Development Goals. value recycling several steps later in the value chain, and
many externality advantages of sharing business models
A more circular economy is indispensable for meeting are not accounted for. A higher carbon price would help
global material needs without exceeding the available on the margin, but capturing a large share of the oppor-
carbon budget. The Intergovernmental Panel on Climate tunities will require addressing those barriers directly. We
Change has estimated a remaining ‘carbon budget’ for estimate that up to 70–80% of the abatement opportuni-
this century of around 800 billion tonnes (Gt) CO2. This is ties are additional to ones already addressed by existing
the amount of emissions that can be emitted until 2100 climate policy approaches. The situation resembles that
for a good chance of keeping warming below 2°C – with for energy efficiency, where careful analysis of cost-effec-
still less for the ‘well below 2°C’ target set by the Paris Ag- tive potentials and barriers has motivated a range of in-
reement. This study estimates that, on current trends, ma- terventions, from aggregate efficiency targets to product
terials production alone would result in more than 900 Gt standards and labelling schemes. Many of the circularity
of emissions. Energy efficiency and low carbon energy will measures are similarly cost-effective, or could be once
help, but do not resolve this dilemma: emissions add up to scaled-up, but require that barriers are overcome. The
650 Gt even with rapid adoption. This is because so much next task is to explore which policy instruments would be
carbon is either built into the products themselves and most effective in pursuing different opportunities.
then released at their end of life (plastics), or is inherent to
the process chemistry of production (steel, cement). For The priority now should be to firmly embed circular
context, note that 2°C scenarios typically ‘allocate’ about economy measures in the low-carbon agenda. This study
300 Gt CO2 to these sectors for the total world economy. is an early quantitative investigation into the low-carbon be-
nefits of the circular economy. Much more work is required,
Options to get to 300 Gt include a) aggressive sca- but we hope this report nonetheless shows the potential av-
le-up of carbon capture and storage; b) the rapid intro- ailable for European industry. The most urgent priority now
duction of radical process changes that are currently in is to build a solid knowledge base – and then to incorporate
early development stages; and c) reducing demand for circular economy opportunities alongside low-carbon ener-
primary materials through the range of circularity measu- gy supply, electrification of transport and heat, and energy
res discussed above. This report argues that it is almost efficiency as a core part of the transition to a low-carbon
impossible to achieve the cut to 300 Gt without a major economy. A more circular economy could play a key role in
use of category c) – hence our assertion that a low-car- helping Europe and the world to meet our climate targets.
7
The Circular Economy – a Powerful
The Circular Force
Economy – afor ClimateForce
Powerful Mitigation / TheMitigation
for Climate climate potential of a potential
/ The climate circular economy
of a circular economy
1. The climate
potential of a
circular economy
Industrial CO2 emissions are a major con- total – and some 3.6 billion tonnes per year
cern as Europe tries to achieve the deep globally. Demand-side measures thus hold
emission reductions required for its climate as much promise as ones on the supply
commitments. In the European Commissi- side. They are all but indispensable for the
on’s ‘Roadmap 2050’, one quarter of CO2 joint objectives of economic development
emissions remaining mid-century were from and action on climate change. Moreover,
industry, and especially from heavy industry they are often economically attractive.
producing basic materials. With more am-
bitious targets after the Paris Agreement, Opportunities for more productive use of
the EU must now articulate how to combine materials therefore deserve a central place
net-zero emissions with a prosperous in- in EU climate policy. Much like improving
dustrial base. Without strong action, emis- energy efficiency is central to the EU’s ef-
sions from the global production of basic forts to achieve a low-carbon energy sys-
materials alone risk exceeding the available tem, a more circular economy will be key to
‘carbon budget’. developing European industry while cutting
its CO2 emissions. As industry associations
This study shows how a more circular and the European Commission consider
economy can make deep cuts to emissions new ‘roadmaps’ for industry mid-century,
from heavy industry: in an ambitious scena- they should include circular economy me-
rio, as much as 296 million tonnes CO2 per asures for cost-effective ways to achieve
year in the EU by 2050, out of 530 Mt in deep emissions cuts.
8 8
E T T VÄ R D E B E S TÄ N D I G T S V E N S K T M AT E R I A L S Y S T E M / S A M M A N FAT T N I N G
9
1.1 MATERIALS & GHG EMISSIONS
– why a 2°C economy must
be circular
Industry is a vital part of any modern economy. As Zero-carbon energy is a crucial part of the answer,
countries develop and their economies grow, they build but it is not enough. This is because so much carbon is
up an ever-larger stock of basic industrial commodities either built into the products themselves and then rele-
and materials such as steel, cement, aluminium, and ased at their end of life (plastics), or is core to the pro-
plastics to underpin infrastructure, transport systems, cess chemistry of their production (steel, cement). The
buildings and factories, and to produce and package demand side – making more of the materials we have
consumer goods. Materials are the heart of maintaining a already produced – will therefore be key to a materials
high standard of living. A healthy industrial sector is also sector in a low-carbon economy. The good news is that
widely seen as crucial for economic competitiveness. there is large potential. Our analysis shows that circular
approaches can reduce CO2 emissions from materials
Yet industry is also a major source of greenhouse gas production in the EU by 56% by 2050. For the EU, this
emissions: 40% of total emissions in 2014, and growing1. represents a major opportunity to both narrow the emis-
To achieve the Paris Agreement’s long-term goal of a global sions gap, and reassert itself as a trailblazer and global
economy with zero net GHG emissions, we need to sharply leader in industry.
reduce industrial emissions. The challenge for policy-ma-
kers – and for industry – is how to curb emissions while
continuing to meet our economies’ material needs.
Global materials demand is set to increase 2-

to 4-fold as the world economy grows
Global heavy industry is growing fast. Steel production emissions from industry: steel, cement, aluminium and
grew by 40% over the last 10 years, with nearly 95% of this plastics.3 In each case, we developed a detailed sce-
growth in China alone. Cement tripled in just a decade nario, based on long-term population and economic de-
and global plastics demand is doubling about every 20 velopment, and using the state-of-the art approaches in
years. Given the importance of these materials in modern published literature to examine how much of these four
economies, it is no surprise that economic development materials would be needed in different world regions if
around the world has kept increasing demand for these they develop similarly to today’s OECD economies.
commodities. That growth is set to continue. Large parts
of the world are still at the initial stages of urbanisation and The results of this analysis are striking: As shown in
industrialisation. For example, the stock of steel in industri- Exhibit 1.1, steel and cement consumption could roughly
alised countries is typically 10-14 tonnes per capita, but in double, aluminium triple, and plastics quadruple by the
non-OECD countries, the average per capita stock is just 2 end of the century. Precisely how far and how fast de-
tonnes2. Similar gaps exist with other materials. mand would increase is uncertain, of course, but our fin-
dings are in line with other existing research.4 The core
To investigate how future materials demand could in- message is clear: in a business-as-usual development,
fluence CO2 emissions, this study focusses on four primary materials demand would rise rapidly and conti-
materials that, together, account for 75% of direct CO2 nue to do so for many decades.
10
Exhibit 1.1
With current patterns of materials use, global demand

for key materials will increase 2- to 4-fold
STEEL PLASTICS
Gt STEEL PER YEAR Gt PLASTICS PER YEAR
4.0 1.4
1.2
3.0
1.0
x 2.3
0.8 x 4.2
2.0
0.6
0.4
1.0
0.2
0.0 0,0
2015 2050 2100 2015 2050 2100
Steel is used in construction and infrastructure, transportation, industrial Plastics production has grown by 50% in the past decade, to just under 350 million
machinery, and consumer products. Global steel production now stands at 1.6 tonnes per year. In advanced economies, packaging is a major use, followed by
billion tonnes per year, having grown by 40% in the decade to 2015. China alone construction and automotive. In Europe, current annual use of plastics is about
accounted for nearly 95% of this growth. Historically, steel stocks have tended to 100 kg/person, while North America is at about 140 kg/person. Our scenario
grow fast once countries reached incomes of around 5000 USD/person, then tapered illustrates the outcome if all world regions converge to 120 kg/person.
off at higher income levels, at 12–15 tonnes per person. Our scenario derives the
demand resulting if all world regions were to follow this pattern, with convergence
to OECD levels of steel stocks of 13 t per capita.
ALUMINIUM CEMENT
Mt ALUMINIUM PER YEAR Gt CEMENT PER YEAR
300 8.0
250
6.0
200 x 1.7
x 3.4
150 4.0
100
2.0
50
0 0.0
2015 2050 2100 2015 2050 2100
Aluminium is used in packaging, buildings, automobiles and other sectors Global Global cement production has tripled in just a decade and currently stands at just
production of primary aluminium now stands at around 60 million tonnes per over 4 billion tonnes per year. Cement production is closely related to construction
year, with an additional 30 million tonnes of remelted aluminium. Stocks have activity and the build-out of infrastructure. Historically, it has peaked and then
been growing strongly in all advanced economies, though they vary greatly: from declined as GDP per capita grows, but with big variations: China used more
600 kg per person in the United States, to 200–500 kg per person in European cement in three years than the United States did in an entire century. Existing
countries. Our scenario assumes global convergence to 400 kg. scenarios reflect these uncertainties, with some suggesting minimal further growth,
and others predicting an explosion. Our scenario is in the middle, anticipating
cement demand of just over 7 billion tonnes per year by 2100.
11
grow fast once countries reached incomes of around 5000 USD/person, then tapered illustrates the outcome if all world regions converge to 120 kg/person.
off at higher income levels, at 12–15 tonnes per person. Our scenario derives the
demand resulting if all world regions were to follow this pattern, with convergence
to OECD levelsTheofCircular Economy
steel stocks Powerful Force for Climate Mitigation / The climate potential of a circular economy
– acapita.
of 13 t per
ALUMINIUM CEMENT
Mt ALUMINIUM PER YEAR Gt CEMENT PER YEAR
300 8.0
Growth in materials use risks exhausting

250
6.0
the carbon budget and imperil climate objectives

200 x 1.7
x 3.4
150 4.0
Given 100the large emissions from heavy industry, the rise in 2100, and still less for a ‘well-below’ 2°C or 1.5°C target as
2.0
materials
50 use has major implications for CO2 emissions. Look- envisaged in the Paris Agreement.5 This budget must cover all
ing ahead, rapidly increasing materials production therefore major emissions from energy and industry: not just materials
0 0.0
risk laying significant claim to the available ‘CO2 budget’ – i.e., production, but also power generation, transportation, heating,
2015 2050 2100 2015 2050 2100
the total amount of CO2 that can be emitted until 2100, while appliances, manufacturing and more. In fact, the amount ‘al-
limiting the increase
Aluminium in globalbuildings,
is used in packaging, temperatures
automobilestoand
a other
given target
sectors Global located’ for production
Global cement production hasoftripled
the infour
just amaterials
decade and is no more
currently than
stands at just
production of primary aluminium now stands at around 60 million tonnes per over 4 billion tonnes per year. Cement production is closely
6 related to construction
(see box below). For a 2°C scenario, the remaining budget about 300 Gt CO2, as shown in Exhibit 1.2.
year, with an additional 30 million tonnes of remelted aluminium. Stocks have activity and the build-out of infrastructure. Historically, it has peaked and then
for emissions
been growing is around
strongly 800 billion
in all advanced tonnes
economies, though(Gt) of CO
they vary greatly:until
2 from declined as GDP per capita grows, but with big variations: China used more
600 kg per person in the United States, to 200–500 kg per person in European cement in three years than the United States did in an entire century. Existing
countries. Our scenario assumes global convergence to 400 kg. scenarios reflect these uncertainties, with some suggesting minimal further growth,
and others predicting an explosion. Our scenario is in the middle, anticipating
cement demand of just over 7 billion tonnes per year by 2100.
Exhibit 1.2
Existing scenarios imply that at most 300 G t of the remaining

carbon budget could be allocated to the production of materials
CUMULATIVE INDUSTRY CO2 EMISSIONS IN 2ºC SCENARIOS, 2015 – 2100

Gt CO2
500 417
400
~400 ~100
300
~300
200
100
0
INDUSTRY SCENARIOS INDUSTRY TOTAL NON-MATERIALS MATERIALS
350 –500 Gt emissions from industry ~400 Gt is the median for ~100 Gt is required for other ~300 Gt remains available
across ~20 scenarios available 2ºC scenarios industry (including manufact- for materials production
CCS volumes are (extremely) high in uring, pulp & paper, chemicals, (cement, steel, plastics,
these scenarios, reaching 20-40 Gt etc.) assuming these reach net- aluminium)
00
per year zero emissions by 2060
NOTE : INDUSTRY SCENARIOS THAT MEET 2°C TARGET WITH >2/3 PROBABILITY AND WHICH REPORT INDUSTRIAL EMISSIONS
SOURCE : MATERIAL ECONOMICS ANALYSIS OF IPCC AR5 DATABASE. 7
12
CLIMATE TARGETS AND CARBON BUDGETS
Greenhouse gases accumulate in the atmosphere, and as concentrations increase, they cause war-
ming and disrupt the climate. Scientists have estimated roughly what temperature increase is asso-
ciated with different CO2 concentrations, as well as the amount of CO2 emissions that would result
in those concentrations, after accounting for natural carbon sinks and, in some scenarios, future
‘negative emissions’ due to carbon capture and storage (CCS) or other measures.
Subtract emissions to date, and the result is a ‘carbon budget’ for any given climate target – that is,
how much can be emitted between now and the year 2100 if we are to keep warming below 2°C,
for instance.
Several variables can affect the budget, including the target itself, the probability of achieving it (66%
vs. 50%), the use of ‘negative emissions’, and the period covered by emissions to date. Sector-speci-
fic budgets also reflect judgements about the share of remaining emissions that should be allocated
to each sector.
In our analysis, we use scenarios with at least 66% change of meeting a 2°C objective, which are
those closest to the Paris Agreement’s objective of limiting global warming to ‘well below’ 2°C.
13
The Circular Economy – a Powerful Force for Climate Mitigation / The climate potential of a circular economy
Our projections of materials demand, however, imply far To take this one step further, we also investigate how far
larger emissions, as much as 918 Gt CO2. This is true even it would help to switch the energy inputs to production to
though we assume in our calculations that the ‘best avai- low-carbon sources.8 Even if this was completed by 2050,
lable technique’ is rapidly adopted to improve energy and some 649 Gt of CO2 emissions would result by 2100, more
process efficiencies, and that current practices for materials than twice the amount that would put materials production
recycling continues (for example, even in the baseline sce- on a path consistent with climate objectives (Exhibit 1.3). As
nario, the emissions per tonne of steel falls by 40%, mostly shown in Exhibit 1.4, the reason for the continued high emis-
because there is much more scrap-based steel production). sions is that a large share result not from fuel combustion,
Still, current processes are intrinsically very carbon-intensi- but from chemical processes in the production of materials.
ve, resulting in large emissions. Decarbonising energy is therefore not enough.
Exhibit 1.3
Materials production alone risks exceeding the total

remaining carbon budget for a 2°C scenario
CO2 EMISSIONS AND CARBON BUDGET

Gt TONNES CO2
CARBON BUDGET TO 2100 CO2 EMISSIONS FROM MATERIALS PRODUCTION
918
800
649
300
14 % 14 %
2°C CARBON BUDGET CARBON BUDGET MATERIALS MATERIALS

FOR INDUSTRY AVAILABLE FOR EMISSIONS EMISSIONS WITH ENERGY
AND ENERGY MATERIALS WITH ENERGY EFFICIENCY AND
EFFICIENCY ZERO-CARBON ENERGY
SOURCE : MATERIAL ECONOMICS MODELLING AS DESCRIBED IN TEXT. MULTIPLE SOURCES, SEE ENDNOTES. 9
14
2°C CARBON BUDGET CARBON BUDGET MATERIALS MATERIALS
FOR INDUSTRY AVAILABLE FOR EMISSIONS EMISSIONS WITH ENERGY
AND ENERGY MATERIALS Exhibit 1.4 WITH ENERGY EFFICIENCY AND
EFFICIENCY ZERO-CARBON ENERGY
Low-carbon energy is not enough: even with 100% low-carbon

energy by 2050, emissions far exceed the available carbon budget
CUMULATIVE EMISSIONS, 2015 – 2100

Gt CO2
Best available energy efficiency fully

BASELINE 254 79 287 298 918 implemented, leading to 15-25% reduction in
SCENARIO CO2 emissions
Emissions intensity reduced by half through full

implementation of direct reduction, bio-based
STEEL 100 feedstock, and carbon capture by 2050
Zero-carbon electricity for EAF production
achieved by 2050
Half of production emissions eliminated. (e.g.,

PLASTICS 60 by using external, zero-carbon heat) by 2050
Emissions from embedded carbon remain
(continued use of fossil feedstock)
38 Zero-carbon electricity used for electrolysis and

ALUMINIUM low-carbon energy used for alumina refining,
both by 2050
All energy-related emissions (fuel, electricity,

CEMENT 70 transport) eliminated by 2050
LOW-CARBON
PRODUCTION 184 40 226 198 649
BY 2050
Plastic emissions include emissions from production as well as embedded emissions

1
Notes: Plastics emissions include emissions from production as well as embedded emissions, see endnotes.10
Source: Material Economics analysis as described in subsequent chapters of this report.

NOTE : PLASTICS EMISSIONS INCLUDE EMISSIONS FROM PRODUCTION AS WELL AS EMBEDDED EMISSIONS, SEE ENDNOTES . 10
SOURCE : MATERIAL ECONOMICS ANALYSIS AS DESCRIBED IN SUBSEQUENT CHAPTERS OF THIS REPORT.
15
A MORE CIRCULAR ECONOMY IS INDISPENSABLE FOR CLIMATE TARGETS

Given that low-carbon energy will not solve the pro- to deploy the most expensive technologies. Circular app-
blem of CO2 emissions from materials, other measures roaches already require some energy input, but less than
are needed. On the supply side, a broad range of options new materials production, and they shift emissions away
have been proposed, from hydrogen-based steelmaking to from hard-to-abate, costly industrial processes, towards
new routes for plastics synthesis, but all are at early stages activities that are much easier to decarbonise. Notably,
of development. Another option that is often discussed is unlike today’s primary materials production, much of the
to deploy carbon capture and storage (CCS) – which also circular economy can be powered by low-carbon electricity.
lags far behind the levels of deployment required for a 2°C
scenario.11 Both will likely be needed to bring industry in Circular economy strategies also have multiple ‘co-be-
line with the goal of net-zero emissions, but they will not nefits’ beside CO2 reductions. By reducing the need for
suffice. industrial processes associated with substantial air pollu-
tion, for example (e.g. steel production using coal), they
Given the scale of the challenge of low-carbon produc- can reduce disease and mortality.12 By reducing the need
tion, it is high time that climate policy looks not just at how for mining, they reduce soil and water pollution as well as
materials are produced, but at ways to reduce demand for the destruction of ecosystems. By reducing the amount
new materials. In the sections that follow, we show how a of plastic left to decay in landfills and in waterways, they
more circular economy could mean we get much more avoid poisoning wildlife and, through the food chain, hu-
out of the materials we already have, making it possible mans. Many strategies would also create new jobs close
to meet the needs of a modern economy with significantly to where materials are used, and some would make key
lower levels of new primary materials production. services, such as transportation, more accessible and af-
fordable. Thus, along with climate targets, a more circular
Demand- and supply-side strategies can also be mu- economy would also make several Sustainable Develop-
tually reinforcing, as is widely recognised in discussions ment Goals more achievable.
of energy transitions. All low-carbon scenarios foresee a
step-change in efficiency, so that essential services such Put together, these arguments make a compelling
as mobility, thermal comfort, or mechanical power can be case for the circular economy not just as one option to
achieved with much lower levels of energy input. Across consider in the quest to meet climate targets, but as an
2°C scenarios, energy efficiency improvements account invaluable part of the transformation we need for a pro-
for some 40% of the emissions reductions. Reducing total sperous and sustainable future.
energy demand also makes it easier to grow low-carbon
energy sources to fully meet global needs, and to keep In the section that follows, we explore the potential for
costs manageable. CO2 emission reductions. Then, in Section 3, we examine
the economics of a more circular economy. We focus both
The same happens with industry in a more circular analyses on the EU, an advanced economy with ambitious
economy. The strategies we explore in this report effec- climate targets and a strong record in pioneering low-car-
tively increase the ‘materials efficiency’ of the economy, bon technologies. The EU could spearhead many of these
reducing the amount of new materials required. This nar- strategies and reap the benefits while showing the rest of
rows the emissions gap that needs to be closed through the world how they, too, can reduce industrial emissions
supply-side measures, buying time and reducing the need without compromising on economic development.
16
Many strategies would also create new jobs close
to where materials are used, and some would
make key services, such as transportation, more
accessible and affordable.
17
1.2 A more circular economy
can cut 2050 emissions from
heavy industry by 56%
To quantify the potential of a more circular economy net emissions in the EU around mid-century or before
to reduce industrial GHG emissions, we conducted an – leaving no or little room for residual emissions from
in-depth study of steel, plastics, aluminium and cement industry. Yet existing analyses do not articulate how this
in the European Union, as well as two key supply chains would be achieved for heavy industry. For example, the
– passenger cars and buildings. As we describe in detail European Commission’s 2011 ‘Roadmap 2050’ fore-
below, we found that a more circular economy could re- saw emissions reductions from all of industry of 259 Mt
duce emissions from the production of these four mate- CO2, while leaving 170 Mt CO2 in place, even with ass-
rials by more than half: set against emissions of 530 Mt umptions of very large volumes of CCS13. Sector road-
CO2 per year, the circular economy abatement potential maps from steel, cement and chemical industry associ-
we identify is almost 296 Mt CO2 per year by 2050 ations left 300 Mt CO2 in place, even after abatement of
(Exhibit 1.5). 150 Mt CO2 through CCS14. The European Commission
is expected to produce a new mid-century roadmap in
Mobilising this potential would make an indispensable 2019. Our analysis suggests that demand-side strate-
contribution to the EU’s climate commitment. Meeting gies should have a central place in net-zero scenarios
targets under the Paris Agreement requires near-zero for European industry.
18
Exhibit 1.5
Circular opportunities can cut 2050 emissions from steel,
plastics, aluminium and cement by 56%
EU EMISSIONS REDUCTIONS POTENTIAL FROM A MORE CIRCULAR ECONOMY, 2050

Mt OF CARBON DIOXIDE PER YEAR
530 178
-56 %
56
62
234


19
Mt OF CARBON DIOXIDE PER YEAR, EU
LOWER VEHICLE WEIGHT
More varied car sizes with shared fleet
Advanced materials more profitable
HIGHER PROFITABILITY OF
CIRCULAR BUSINESS MODELS
Without deep transformation,

emissions would amount to 530 Mt CO 2 in 2050
END-OF-LIFE VALUE
Higher EOL value (modularity,
valuable materials etc.)
In order to gauge the potential impactEOL
of flows
a more circular in fleet60-70% of steel, cement and aluminium use, and some
more predictable
owned system
economy on industrial emissions, we looked more closely 30% of plastics. Conversely, steel, cement, aluminium and
at each of the four materials and the two value chains. Ex- plastics account for 85% of the materials CO2 footprint of
hibit 1.6 provides an overview. Cars andLOWER
buildings account buildings and passenger cars.
COST OF TRANSPORT
New design and high utilisation reduce
total cost
Exhibit 1.6
Competitiveness with other modes of
transport increases
Study scope: four materials and two value chains that together
account for most of industrial CO 2 emissions
FOUR KEY MATERIALS CATEGORIES

Used across economy in In advanced economies, Key uses include Used for construction
construction, packaging is a major use, buildings, automotive, of buildings and infra-
66%
transportation, followed by construction electrical machinery structure
industrial machinery, and automotive and packaging Production is primarily
and consumer products 100 kg/capita is consum- The EU imports 40% local due to high avail-
40% of demand is ed annually in Europe, of its aluminum, some- ability of raw materials
served by secondary of which secondary times from locations and high cost of transport- OF INDUSTRIAL CO2 EMISSIONS
production in the EU, plastics only represent with very high CO2 ation
but the industry still 10% of demand intensity of production CO2 emissions are more
releases some 230 CO2 130 Mt CO2 are re- In total, the CO2 foot- than 110 Mt per year,
Mt per year leased annually from print of EU demand is of which 55% are due to
European production around 80 Mt annually the process chemistry
rather than energy
TWO KEY VALUE CHAINS AND PRODUCT CATEGORIES
CARS/MOBILITY
~50%
Automotive sector uses 20% of steel, 10% of plastics and 20% of aluminium
CO2 emissions from materials in passenger cars sold in the EU are around 50 Mt per year
Embodied emissions in materials are becoming a larger share of total CO2 footprint
PASSENGER CARS AND BUILDINGS

JOINTLY ACCOUNT FOR ALMOST
HALF OF EMISSIONS FROM STEEL,
BUILDINGS/SHELTER PLASTICS, ALUMINIUM AND CEMENT
Buildings account for 33% of steel, 20% of plastics, 25% of aluminium and 65% of cement
The CO2 footprint of building materials in the EU is 250 Mt per year
Embodied carbon is 10-20% of EU buildings’ CO2 footprint today, but already 50% in
countries with low-carbon energy
SOURCE : SEE SUBSEQUENT CHAPTERS OF THIS REPORT.
20
Steel, cement, aluminium and plastics account
for 85% of the materials CO2 footprint of
buildings and passenger cars.
21 .
Total emissions from the production of the four mate- tional emissions in 2050. Accounting for those emissions
rials for EU consumption amounts to 564 Mt CO2.15 The as well as modest consumption growth, annual emis-
baseline scenario sees a varied development of demand: sions from plastics are as high as 233 Mt CO2. Clearly,
steel demand falls slightly, as the total steel stock sa- the incineration of fossil-based plastics cannot continue
turates, whereas cement use stays constant, and plastics in a low-carbon economy.
and aluminium use grows somewhat. The current pro-
duction processes and practices continue to be used, In the steel sector, on the other hand, emissions decli-
so that plastics production is fossil-fuel based, cement is ne even in the baseline scenario. This is due to several
based on clinker production from limestone, and prima- factors: reduction in demand for steel as the EU steel stock
ry steel continues to use basic oxygen furnaces based stabilises, a gradual and substantial shift toward more se-
on coal. However, trends towards lower emissions also condary and less primary production, and the decarbonisa-
continue: the efficiency of the processes improves by 10- tion of electricity inputs to secondary steel production. The
20% through the rapid adoption of best available produc- EU steel sector thus is already on a path to a more circular
tion techniques; materials use becomes more circular economy that would significantly reduce emissions reduc-
even in the baseline scenario; and the energy system is tions – but as we discuss below, there is potential to go
assumed to be decarbonising, so that emissions from EU much further than is possible with current practice.
electricity fall to near-zero levels, in line with the 2011 Eu-
ropean Commission Roadmap 2050. Nonetheless, there For aluminium, the combination of a higher share
is only a small net decrease in industrial emissions by scrap-based production and the decarbonisation of electri-
2050, to 530 Mt CO2 (Exhibit 1.7). city production more than offset the growth in emissions
that would otherwise occur as the volume of aluminium
The largest net growth in emissions in our baseline increases. For cement, the baseline scenario shows a very
scenario occurs in plastics. This is only partly because small decline in emissions, as improvements in process
consumption increases, but more because plastics con- efficiency slightly outstrip growth in demand.
tains substantial embedded carbon in the material itself,
which is released as CO2 when plastics are incinerated. The bottom line is that despite increased recycling, espe-
Today, when plastics are burned instead of other fossil cially of steel, and the transformation of EU energy supply,
fuels, the net increase in emissions is relatively small. emissions from materials production in 2050 would be lar-
That would change dramatically in a decarbonised en- gely similar to today. Achieving the much-larger reductions
ergy system, where the alternative would be zero-carbon needed for a low-carbon economy will thus require stra-
energy sources. Thus, a continuation of the current shift tegies to reduce demand, even as new technologies are
towards burning plastics would result in substantial addi- deployed to reduce industrial process emissions as well.
22
Exhibit 1.7
Without demand-side measures, EU emissions from
materials production barely decline by 2050

Mt OF CARBON DIOXIDE PER YEAR, EU
564
STEEL 530
• Steel use remains at close to 160 Mt / year as EU stock saturates,
• Scrap-based production increases from 40% to 65%
• The CO2 intensity of production falls by more than 50% as a result 104
234
PLASTICS
• Consumption increases from 49 Mt today to 62 Mt 2050
• Emissions increase chiefly because of embedded emissions:
the carbon in the plastic itself. In a low-carbon energy system,
burning plastics has high net emissions 233
ALUMINIUM
132
• Aluminium use grows from 12 to 16 Mt per year,
but stabilises at this level
• Clean electricity means lower CO2 intensity for both
EU production and imported metal
• The net result is almost stationary emissions attributable
83 to EU consumption 80
CEMENT
• Consumption stays similar today, reaching 184 Mt per year
114 113
• Emissions falls by 10 percent through production process
improvements
TODAY 2050
SOURCE : MATERIAL ECONOMICS MODELLING AS DESCRIBED IN TEXT.
STEEL: PRIMARY VS. RECYCLED

23 PLASTICS: PRIMARY VS. RECYCLED
PRIMARY PRIMARY
t CO2 / t STEEL t CO2 / t PLASTICS
RECYCLED RECYCLED
Three circular strategies work

together to cut EU industrial emissions
The demand-side potential includes a range of strate- • More material-efficient products jointly make up op-
gies. We include under the umbrella of ‘circular economy’ portunities with 56 Mt CO2 of abatement potential. These
any opportunity to provide the same economic service opportunities have in common that they reduce the total
with less primary material. In this section, we explore the materials input required to produce a given product or
potential associated with each of three categories of me- structure, through lightweighting, reduced waste, high-
asures, which are further summarised in Exhibit 1.8. strength materials, and other strategies.
• Materials recirculation opportunities provide more • New circular business models could reduce emissions
than half of the abatement potential, at 178 Mt CO2 per year, by by 62 Mt CO2. The main lever here is a large increase in
increasing the share of materials needs that are met by using the materials productivity of both mobility and buildings: in-
recycled rather than primary materials. This means increasing creasing the benefits derived from each building or vehicle
both the volume and the quality of secondary materials. through shared use and measures to prolong their lifetime.
Exhibit 1.8
Three circular economy strategies make better use of

materials and products to reduce GHG emissions
1 2 3
MATERIAL RECIRCULATION PRODUCT MATERIAL EFFICIENCY CIRCULAR BUSINESS MODELS
GHG MATERIALS PRODUCT

GHG
EMISSIONS
= USEFUL
SERVICE
X X X
MATERIALS PRODUCT USEFUL SERVICE
Underlaying benefit,
More high-value recycling  Less material input Fewer products required to
such as passenger
less primary material required for each car, achieve the same benefits or
or freight kilometers
production lower emissions building, etc (e.g. service (e.g number cars
of transportation, or
per tonne of material tonnes of material per produced for a given amount
effective available
car) of transportation)
building area
High value recycling, Improved production Higher utilisation and

through: process intensive use of products
Increased collection Less production waste Sharing of products
rates Avoid over-specification Product as service
Design for disassembly
and improved Reuse of components Longer lifetime of
materials separation products
Less contamination Designing products with Design for durability
and downgrading of less materials and disassembly
materials High-strength materials Long lasting materials
New design principles Improved maintenance
Variation in size Remanufacturing
24
Materials recirculation opportunities
provide more than half of the abatement
potential.
25
1. MATERIALS RECIRCULATION STRATEGIES COULD CURB EMISSIONS BY 178 Mt CO 2 PER YEAR
From a circular economy perspective, Europe is a recycled steel can cut emissions by 90% if using largely
treasure trove of recyclable materials – those that have decarbonised electricity.
already been discarded, as well as millions of tonnes
• Plastics: The largest potential to improve circularity
more that will become available as buildings, vehicles
is in plastics, where recycling rates today are low (recyc-
and products are taken out of service in the future. As
led volumes are just 10% of plastics in the market), and
shown in Exhibit 1.9, producing secondary materials
CO2 gains would be substantial. A detailed assessment
through recycling results in far lower emissions per tonne
of plastics categories and uses identifies opportunities
than producing primary materials. Thus, every piece of
to greatly increase recycling levels, especially for the
scrap holds the promise of CO2 savings. Across mate-
biggest five plastic types that make up 70% of demand.
rials, realising the potential requires creating robust sys-
More than half of plastics volumes could be recycled
tems for collection, avoid contamination by additives or
mechanically, and a further 11% through chemical re-
mixing of different qualities of materials, and producing
cycling approaches.
materials of sufficient quality to serve as genuine substi-
tutes for the corresponding primary material. However, • Aluminium: It is less certain to what extent the stock
the scope for increasing the reuse of different materials of aluminium will stabilise, but the amount of post-consu-
varies significantly: mer scrap available will nonetheless increase. By 2050
potentially, post-consumer scrap generated in Europe
• Steel: Current EU steel production is more than 60%
could amount to as much as 75% of the production re-
based on primary production, i.e. produced from iron ore.
quired to meet European demand. The opportunity is to
However, a detailed analysis of steel stock evolution and
enable the continued use of this scrap in a wide range
scrap flows suggests that, in decades to come, the EU
of applications, so that it can replace a greater share of
will approach the point where the need to maintain a
primary metals production.
near-constant stock of steel can be served to a large
extent by recirculating steel that has already been pro- • Cement: Cement cannot be recycled in the conven-
duced. Doing so will require reducing losses of steel, tional sense, though structural elements can be reused.
changing how steel scrap is handled and traded, and In addition, as we explain below, approaches are under
avoiding contamination of the steel stock with copper, development to recover unreacted cement from concrete,
as we discuss below. The CO2 prize is substantial, as which can then be recycled into new concrete production.
26
• Consumption stays similar today, reaching 184 Mt per year
114 113
• Emissions falls by 10 percent through production process
improvements
TODAY 2050
Exhibit 1.9
Materials recycling cuts CO 2 emissions

from materials significantly
STEEL: PRIMARY VS. RECYCLED PRIMARY PLASTICS: PRIMARY VS. RECYCLED PRIMARY
t CO2 / t STEEL t CO2 / t PLASTICS
RECYCLED RECYCLED
2.4
2.3
2.2
1.9
14 %
0.4 0.4
0.3
0.1
TODAY 2050 TODAY 2050
ALUMINIUM: PRIMARY VS. RECYCLED PRIMARY CEMENT: PRIMARY VS. RECYCLED

PRIMARY
t CO2 / t ALUMINIUM t CO2 / t CEMENT
RECYCLED RECYCLED
13.5
0.7
0.6
9.7
14 %
0.3
0.1
0.3 0.2
TODAY 2050 TODAY 2050
SOURCE : MATERIAL ECONOMICS ANALYSIS AS DESCRIBED IN SUBSEQUENT CHAPTERS OF THIS REPORT.
27
2. MORE MATERIAL-EFFICIENT PRODUCTS CAN Fully achieving the potential gains in this category
would require deep changes in how we deliver services
CUT EMISSIONS BY 56 Mt CO 2 PER YEAR and goods, which in turn would drive major changes
in how vehicles and buildings are initially built and ma-
The amount of materials required also can be reduced sub- intained. But the rewards would be substantial. As we
stantially through a range of approaches. One is to reduce elaborate below, there are major opportunities to boost
the amount of materials that are discarded as scrap or waste productivity and achieve significant co-benefits for the en-
in the manufacturing or construction process. For example, vironment, public health, and quality of life.
some 15% of buildings materials are wasted in construction,
a share that can be reduced substantially through best prac- In practice, these broad strategies translate into a
tice. Another strategy is to use more advanced materials and wide range of measures: from changes to product design
construction techniques. High-strength steel, for instance, has and materials choice, to improved technology, larger sca-
the potential to cut materials use by 30–40% in a range of le in secondary materials industries, and even extensive
applications, from heavy machinery to buildings and furniture. changes in the value chains and organisation of mobility
It also is possible to save materials by reducing over-specifica- and buildings.
tion; by one estimate, as much as 50% of steel used in buil-
dings is in excess to what is strictly required to meet structural
needs. Finally, lightweighting of products can also be achieved
by tailoring individual products better to specific uses; for ex- * * *
ample, while cars built for individual ownership typically need
to be able to carry at least four passengers and some bagga-
ge, cars built for a shared fleet could vary in size, with many The benefits of circular approaches are strengthe-
just big enough for a one- or two-passenger trip in the city. ned by the fact that many of these measures work well
together. For example, reducing the materials intensity
of buildings and vehicles reduces the total steel stock
3. NEW CIRCULAR BUSINESS MODELS IN needed, which secondary steel production needs to
MOBILITY AND BUILDINGS CAN REDUCE grow less to meet the demand. Likewise, new business
models that boost the value realised from each product
EMISSIONS BY 62 Mt CO 2 PER YEAR can drastically improve the economics of measures to
make products more materials-efficient. Cumulatively,
The third pillar of the abatement potential is opportuni- these opportunities can result in a step change in re-
ties that increase the utilisation and prolong the lifetime of source efficiency.
materials-intensive assets in the economy. In both cases,
the same initial materials input can then provide a much Tracing one material in one value chain makes clear
larger amount of use benefits, whether passenger travel, the extent the opportunity. For example, circular strate-
or effective occupied square metres. The opportunity is gies could jointly cut the amount of primary steel requi-
most significant for personal vehicles; as we describe red to serve mobility needs by 70% (Exhibit 1.10). Put
below, a system of fleet-managed, shared vehicles could differently, the productivity of materials use has incre-
require just 25% of the materials inputs of today’s system ased so that each tonne of steel supports more than
of individually owned cars. three times as much transportation.
28
Exhibit 1.10
Circular measures can reduce the primary steel required

to support mobility by 70 percent
PRIMARY STEEL USED FOR MOBILITY SERVICES

TONNE PRIMARY STEEL PER MILLION PASSENGER KILOMETRE
MATERIALS RECIRCULATION PRODUCT MATERIALS EFFICIENCY CIRCULAR BUSINESS MODELS
0.29 II
0.35
III -70 %
1.84 0.65
0.56
BASELINE CIRCULAR
SCENARIO
A circular scenario results in mobility services being provided with much less primary steel used per unit, through:
I MATERIALS RECIRCULATION II MORE MATERIAL III IMPROVED PRODUCTIVITY OF USE

already is a major part of steelmaking, with -EFFICIENT PRODUCTS has particularly large potential for vehicles. Today’s passenger cars use
secondary (recycled) steel making up around can reduce steel requirements by only 2% of their capacity, because they are parked most of the time and
one-third of current global production. Still, some 30%. The first step is to often carry only one or two people when used. Replacing a large share of
there is much potential to increase recycling by reduce process losses; up to half of personal vehicles with a system of fleet-managed, shared vehicles would
improving collection rates, avoiding steel now ends up as process scrap mean that cars could not only be used more intensively, but such use
contamination with copper, reducing losses rather than in the final product. would make a range of strategies feasible that extend their lifetimes. These
during remelting, and avoiding downgrading. A second step is to make cars in a include more durable design and higher-value materials, a greater share
Together, these measures could reduce the much wider range of sizes, and to of intrinsically more durable electric vehicles, predictive and fleet-man-
need for global primary steel production by use more advanced materials that aged maintenance, and modular design for reuse and remanufacturing.
another 14% by 2050, and as much as 80% may cost more, but pay off when The combination of sharing and longer lifetimes could almost halve the
in a more mature economy such as the EU. vehicles are used more intensively. amount of materials needed per passenger-kilometre travelled.
NOTE : ANALYSIS IS BASED ON GLOBAL STEEL PRODUCTION.
29
Demand- and supply-side measures are both

required for deep decarbonisation of industry
The emissions reductions described above are net year, and associated emissions by 296 Mt CO2. If emis-
reductions, accounting for the inputs required to imple- sions are to be reduced by an illustrative 90%, that would
ment relevant measures, including transportation and still leave about 209 Mt CO2 of emissions to abate. Supp-
electricity. The numbers highlight another advantage of ly-side measures are crucial to closing that gap.
these approaches: they can benefit from synergies with
the decarbonisation of transport and electricity systems. Sharply reducing emissions from materials production
The overall resource requirements also are modest when would require fundamental changes to industrial proces-
compared with the requirements for supply-side abate- ses. Examples include replacing coal with hydrogen as
ment – for example, the energy requirements are much an input to primary steel production, the synthesis of
lower per tonne material produced. The intrinsic higher plastics from non-fossil feedstock (e.g. bioplastics), and
resource efficiency of a more circular economy therefore novel cements. While there are promising options un-
pays off not only in reduced demand for primary raw ma- der exploration, they are at low technological readiness,
terials, but also in terms of inputs for low-carbon energy, and far from commercial use. Developing and deploy-
putting less strain on other aspects of the overall transi- ing them is a major undertaking, with uncertainty both
tion to a low-carbon economy. about which options will prove successful, and how long
they will take to scale up. Additional CO2 reductions may
Yet nothing in this report should be taken to mean that be possible through further energy efficiency improve-
circular economy approaches should replace supply-side ments, wider use of renewable energy and electrification
measures. Both are needed if Europe is to decarbonise its as well.
industry. Exhibit 1.11 shows the potential for demand-
side emissions reductions combined with the potential on The other main option is to capture CO2 from existing
the supply side. Total 2050 baseline demand across steel, processes. This is being trialled at some industrial si-
plastics, aluminium and cement is over 400 Mt per year, of tes, but is still at demonstration phase. After more than
which almost 300 Mt are primary materials. As noted ear- a decade of active development, CCS still faces large
lier, even assuming the use of the best available technologi- challenges, including a lack of viable storage sites for
es, energy efficiency improvements, the use of low-carbon CO2, and the cost and logistics of fitting a large number
energy, and circularity consistent with current trends, total of dispersed industrial sites with CCS technology.
emissions from the production of these materials would be
530 Mt CO2 per year, barely changed from today. Overall, circular economy opportunities thus substanti-
ally reduce the scale of supply-side abatement required,
The demand-side measures explored in this report potentially providing the majority of the emissions reduc-
would reduce primary materials needs by 173 Mt per tions required for EU 2050 targets.
30
-EFFICIENT PRODUCTS
secondary (recycled) steel making up around can reduce steel requirements by only 2% of their capacity, because they are parked most of the time and
one-third of current global production. Still, some 30%. The first step is to often carry only one or two people when used. Replacing a large share of
there is much potential to increase recycling by reduce process losses; up to half of personal vehicles with a system of fleet-managed, shared vehicles would
improving collection rates, avoiding steel now ends up as process scrap mean that cars could not only be used more intensively, but such use
contamination with copper, reducing losses rather than in the final product. would make a range of strategies feasible that extend their lifetimes. These
during remelting, and avoiding downgrading. A second step is to make cars in a include more durable design and higher-value materials, a greater share
Together, these measures could reduce the much wider range of sizes, and to of intrinsically more durable electric vehicles, predictive and fleet-man-
need for global primary steel production by use more advanced materials that aged maintenance, and modular design for reuse and remanufacturing.
another 14% by 2050, and as much as 80% may cost more, but pay off when The combination of sharing and longer lifetimes could almost halve the
in a more mature economy such as the EU. vehicles are used more intensively. amount of materials needed per passenger-kilometre travelled.
Exhibit 1.11
The role of demand- and supply-side measures in decarbonising

emissions from materials production
EMISSIONS INTENSITY
t CO2 / t MATERIAL
2,0 REMAINING EMISSIONS AFTER TOTAL EMISSIONS

DEMAND-SIDE ABATEMENT IN THE EU 2050
234 Mt 530 Mt
CO2 CO2
1,5
ILLUSTRATIVE SUPPLY-SIDE
REDUCTION OF 209 Mt CO2
THROUGH:
CIRCULARITY REDUCES PRIMARY
• EFFICIENCY AND MATERIALS DEMAND, RESULTING
RENEWABLE ENERGY IN 296 Mt CO2 ABATEMENT
1,0 • ELECTRIFICATION
AND PROCESS
BREAKTHROUGHS
• CARBON CAPTURE
AND STORAGE
0,5
25 Mt
CO2
REMAINING EMISSIONS AFTER
DEMAND – AND SUPPLY-SIDE
ABATEMENT
PRIMARY
0,0 MATERIALS
Mt
40 80 120 160 200 240 280 320 MATERIAL
31
Globally, a more circular economy

could avoid 3.6 Gt CO 2 per year by 2050
The EU is well positioned to pioneer many circular eco- Altogether, we estimate the global abatement potential
nomy principles, especially those focused on materials from circular measures at 3.6 billion tonnes of CO2 per year
recirculation. It helps to have large stocks of materials by 2050, or nearly 45% of baseline emissions (increasing
already available for recycling. Although there is potential to 60% later in the century). This development course is
for recycling also in developing countries, it is smaller when one much more consistent with climate objectives; in terms
much of production goes towards building up the total stock. of cumulative emissions to 2100, a more circular economy
could cut 333 out of the 918 billion tonnes of CO2 in the
Strategies focused on materials efficiency and produc- baseline scenario (Exhibit 1.12). Rapidly adopting current
tivity of use, on the other hand, could benefit developing low-carbon technologies could cut another 178 billion ton-
countries at least as much as the EU. Instead of replicating nes, so that just over 400 billion tonnes remain. To meet
the inefficiencies of the mobility and buildings systems of climate objectives, emissions would need to fall still further
today’s OECD economies, these countries can ‘leapfrog’ – through additional circular economy measures beyond
to more efficient technologies and business models. Ma- those investigated in this study, materials substitution, and
king construction more materials-efficient, for example, is cru- the development and gradual adoption of even lower-car-
cial for countries that are rapidly expanding their built area. bon production processes than those available today.
32
Exhibit 1.12
A more circular economy could cut global emissions

by more than 300 billion tonnes to 2100
CUMULATIVE EMISSIONS FROM MATERIALS PRODUCTION, 2015 – 2100

Gt CO2
STEEL ALUMINIUM
1000 PLASTICS (PRODUCTION) CEMENT

918 333
PLASTICS (EMBEDDED)
900
800
700
178
600 To meet climate targets:
Additional circularity
Production process breakthroughs
500 Material substitution
408
400
300
200
100
0
BASELINE DEMAND-SIDE CURRENT REMAINING FURTHER EMISSIONS
EMISSIONS MEASURES LOW-CARBON EMISSIONS REDUCTIONS REQUIRED FOR
(CIRCULARITY) PROCESSES CLIMATE TARGETS
COST OF PASSENGER TRANSPORT 33 CO-BENEFITS FROM A SHIFT TOWARDS SHARED CARS

EUR PER 1000 PASSENGER KILOMETRES EUR PER 1000 PASSENGER KILOMETRES
1.3 MANY CIRCULAR
ECONOMY STRATEGIES ARE
ECONOMICALLY ATTRACTIVE
In Section 2, we showed the substantial potential for Some care is required in interpreting a cost curve of
circular economy strategies to reduce industrial emis- this type. While a negative cost shows that a measure
sions. The obvious next question is, are they cost-effective? can ‘pay for itself’ when viewed from a societal perspecti-
In this section, we quantify the cost of different circular ve, that does not mean that it will be implemented without
economy options; Exhibit 1.13 summarises our findings policy action. If measures that have net benefits remain, it
in a cost curve. One axis shows the abatement potential is because there are barriers or market failures that pre-
available, divided into groups of measures that add up to vent their implementation. In such situations, the market
296 Mt CO2.16 The other axis shows the estimated cost will not act by itself, even if prices are favourable; policy
of each measure, expressed in EUR per tonne of CO2 interventions will be needed to overcome the barriers.
avoided.
Some negative cost measures may also require sys-
It is striking how economically attractive many measu- temic shifts in a value chain or sector. Car-sharing is a
res appear to be, with low and even negative costs. case in point; shared fleets of cars could lower costs sig-
Many measures are economically attractive, if barriers nificantly relative to individual ownership, but this would
to their implementation can be overcome; many others require a large-scale shift towards a different innovation
have a cost of no more than 50 EUR/t CO2, less than focus and overall system for how to serve mobility. Rea-
many other ways to reduce emissions (including many lising the potential CO2 and economic benefits therefore
supply-side opportunities for industrial emissions). We requires large-scale and systemic shifts, not just incre-
caution that this cost curve is indicative, with many un- mental change. A similar argument applies to high-value
certainties, and must be followed up with deeper ana- plastics recycling: negative costs are not available today,
lysis to improve the estimates. Nonetheless, the results but depend on a combination of technical progress,
show a clear potential for circular economy strategies to change in the way that plastic is used in products and
make very cost-effective contributions to a low-carbon handled at end-of-life, and large scale in recycling acti-
economy. Moreover, ex-ante analyses of this kind often vities.
fail to capture many of the technological developments
that will become available. In subsequent chapters of To further understand the economic significance of dif-
the report, we provide more details about the numbers ferent measures, it helps to distinguish the costs and be-
underlying our analysis. nefits that arise in each category, see following sections.
34
Exhibit 1.13
A cost curve for circular economy measures

to reduce EU CO 2 emissions
ABATEMENT POTENTIAL AND COST FOR CIRCULAR ECONOMY LEVERS, 2050

Mt CO2 PER YEAR; EUR PER t CO2 ABATED
ABATEMENT COST
EUR / t CO2
Cars - Sharing
Cars - Prolong lifetime
Cars - Remanufacturing
Buildings - Floor space sharing
Other - Sharing and lifetime
100 Plastics - Higher quality recycling
Buildings - Reduced waste in construction
Plastics - Reuse
Steel - Reduce copper
50 Aluminium - increase collection
50 100 EMISSIONS
0
150 200 250 300 SAVINGS
Mt CO2 / YEAR
Cars - Lightweighting
-50 Buildings - Materials efficiency
Cement - Cement recycling
Steel - Reduce fabrication losses
-100 Plastics - Chemical recycling
Buildings - Reuse
Other - Materials efficiency
Plastics - Increased recycling at current quality
Aluminium - avoid downgrading
Steel - Increase collection
HOW TO READ THE CO2 ABATEMENT COST CURVE

The CO2 abatement
PLASTIC COST CURVE cost 2050
curvePOTENTIAL
summarises the opportunities to reduce CO2 emissions through circular eco-
nomy
EUR PERopportunities.
ABATED tThe CO2width of each2050
, EUROPE, bar represents the CO2 potential of that opportunity to reduce CO2
emissions in 2050, relative to a baseline scenario. It accounts for overlaps and interactions that arise when
all the opportunities are pursued simultaneously.
ABATEMENT COST
The
EUR height
/ t CO2 of each bar represents the average cost of avoiding 1 tonne of CO2 emissions by 2050 through
that opportunity. In several cases, the cost is a weighted average across several sub-opportunities. All costs are
Agriculture
in 2015 Reuse The cost of abatement is calculated from a societal perspective (excluding taxes and sub-
real Euros.
B&C Reuse
sidies), but does not account for reduced externalities or other benefits associated with the opportunities. The
opportunities are Other plasticsleft
ordered Hightoquality
rightincreased
with therecycling
lowest-cost abatement opportunities to the right and highest
to the left. Both volume B&C andHigh
costquality
estimates arerecycling
increased of course uncertain in such a long timeframe
B&C Low quality increased recycling
Automotive High quality increased recycling
EEE High quality increased recycling
100 Other plastics Reuse
100
Agriculture High quality increased recycling
35
Packaging High quality increased recycling
Automotive Reuse 55 57
50 43
35
1. MATERIALS RECIRCULATION CAN BUILD

ON HIGH-VALUE PRODUCTION OF RECYCLED MATERIALS
The economic attractiveness of materials recircula- Second, technology continues to improve, making
tion depends in large measure on the capacity to retain circular economy measures more viable – through eve-
the material value over the use cycle. Substantial values rything from advanced sensors, to automation, to infor-
are created when materials are initially produced. If the- mation technology and mobile apps. The costs of techno-
se can be retained and recovered, the CO2 benefits of logy also continue to decline, to the point that the value
avoided primary production can be coupled to the indu- of the materials to be recovered would more than offset
strial opportunity of preserving the intrinsic value of the the cost of implementing the necessary tools.
secondary materials. This is why already, 83% of steel
and 77% of aluminium is recycled. The CO2 and other Plastics recycling offers a particularly instructive ex-
resource benefits in turn depends on the ability to produ- ample of the dynamic. Aging, additives, contamination,
ce high-quality secondary materials that are capable of mixing of different plastic types and other factors conspi-
substituting for primary materials. re to hold back this market. As a result, recycling of-
ten results in low-value secondary materials produced
For further recycling to be economically attractive, it at relatively high cost. In CO2 terms, abatement costs
is necessary to either reduce the cost of collection and are around 50–100 EUR per tonne CO2. Even though
secondary materials production, or to increase the value plastics recycling rates as high as 30% are reported in
of the material produced. A key finding of our analysis is the EU, the actual volume of secondary materials in the
that the scope for high-quality secondary materials pro- market is only 10%. As illustrated in Exhibit 1.14, a range
duction is far from exhausted, for two main reasons: of measures to improve the collection and processing
of plastics for recycling could transform the economics.
First, there are numerous barriers and market failures Once we have a system that retains substantial value,
that now hinder what would otherwise be cost-effective recycling plastics can become profitable, and take off on
materials recovery and secondary materials production. a larger scale. Our analysis suggests that this is feasible
One major problem is that current processes for product for the top five plastic types.
design and dismantling, from cars to buildings and packa-
ging, lead to significant materials degradation. Even minor
modifications could capture large additional values.
36
VOLUME
tonnes
TODAY 55% COLLECTION CIRCULAR SYSTEM
90% value loss >80% of value still lost, 50% of value retained
“30% collection” Large process losses in recycling Higher recycling yields increase
but just 16% Uphill struggle to increase of volume
recycling volume further – at a net cost High value retention provides
Rebound effects as downgraded revenue for investment in self-
plastics cannot replace virgin sustaining systems and scaling
plastics in many uses Recycling can replace virgin
production
Exhibit 1.14
Plastics recycling can be made economically

viable through a range of measures
REDUCED RECYCLING COSTS HIGHER REVENUES
EUR / TONNE TREATED PLASTICS EUR / TONNE TREATED PLASTICS
924 926
-16%
COLLECTION 186
774
+71%
119
SORTING 191
147 PACKAGING EXAMPLE
540
LOGISTICS 15
10
337
RECYCLING 439
162
INVESTMENT 93 REVENUES
2017 2050 2017 2050
Cleaner flows from materials choices and Increased yields give larger revenues per tonne
product design optimised for recycling treated plastic (better technology, cleaner inflows)
Increased scale reduces unit cost Higher quality enables pricing closer to virgin
Specialisation and regional integration of markets materials (additional sorting, improved technology,
Technological improvement boosts efficiency but cleaner flows and optimised products)
requires higher investment Raised cost of virgin materials improves
Reduced risk from regulations and market willingness to pay for secondary products
uncertainty Better functioning markets reduce current
commercial risk to buyers
DEEP TRANSFORMATION THROUGH COORDINATED AND SUSTAINED ACTION ACROSS THE VALUE CHAIN
SOURCE : SEE CHAPTER 3.
EMISSIONS SAVINGS
Mt CO2 / YEAR 37
80
2. MAJOR VALUE CHAINS COULD REDUCE THE AMOUNT OF MATERIALS REQUIRED FOR PRODUCTS
Measures aimed at using materials more efficiently • Increased automation can reduce the trade-off
cover a wide gamut and can vary significantly in the tra- between materials efficiency and labour cost. This often
de-offs between reduced materials use and other costs. is the limiting factor today; for example, a major reason
What all these have in common is that they reduce ma- for over-specification of steel in buildings is the additio-
terials use by changing production processes or product nal labour cost of using more custom-made components.
design – encompassing measures such as reduced pro- With automation of more of the construction process, this
cess scrap in manufacturing, reduced materials waste trade-off could be reduced.
in construction, more materials-efficient production tech- • Increased utilisation and lifetime – as part of sha-
niques such as customised structural elements in buil- ring models of mobility and buildings use – can justified
dings, and the use of high-strength materials. the use of lighter but more costly materials, by spreading
their cost over a much more extensive period of use.
The financial benefit is primarily the avoided cost of
materials. The costs of the measures, on the other hand, • Reducing the need for spare capacity – a real
can include factors such as larger inventories, smaller possibility with new business models – would address
scale of operation, higher labour cost, or slower produc- a major driver of materials use today. Personal vehicles
tion. Interviews with market actors suggest that the ba- are a good example. Although most people rarely use
lance between these two depends strongly on specific every seat in their car, few would want to own a car that
circumstances, and we therefore have been cautious with can only carry one or two people. However, in a system
cost estimates in the above abatement cost curve. On of shared fleets of cars, such vehicles would be more
the other hand, several companies indicated in interviews economically viable. A fleet could have some larger cars,
that they have found low-hanging fruit that enabled sub- and others designed for one- or two-person trips. In that
stantial materials savings without large costs. context, ‘lightweighting’ could be achieved at scale at low
or even negative costs simply by making smaller cars.
For all the caveats, there are four broad trends that
support a move to greater cost-effectiveness in improving None of this means that increased materials efficiency
materials efficiency: will necessarily happen by itself. The materials footprint
of products and structures often is low on the agenda
• Technical advances can drastically lower the cost of today. But it shows that important CO2 gains could be
reducing waste in production. A prominent example is available without incurring very high costs.
‘additive’ manufacturing methods such as 3D printing,
which can almost eliminate production scrap. Another is
advances in building information management systems,
which mark and track construction materials much more
closely, and have shown significant potential for reducing
waste of materials in the construction phase.
38
Steel use in buildings is almost 50% higher
than what is strictly required to meet design
specifications.
39
3. NEW CIRCULAR BUSINESS MODELS COULD YIELD MAJOR PRODUCTIVITY GAINS AND CO-BENEFITS
Some of the most economically attractive options are kilometres – with much lower costs as a result. Similar
to be found in circularity strategies that focus on using approaches could be taken to use buildings (and indi-
products more efficiently. Several of these measures vidual spaces within them) more efficiently, and to ex-
would reduce CO2 emissions at a significant net profit. tend lifetimes (of whole buildings or parts) so that future
This is because they involve making large systemic im- construction activity could be reduced.
provements to boost productivity in the value chains for
mobility and for buildings. As we discuss below, the real issue with such system
shifts is not whether they would be more productive and
Today, large transaction costs undermine the econo- lower-cost – they undoubtedly would, as shown in Exhibit
mic efficiency with which we use key assets. The average 1.15. Moreover, they could have substantial co-bene-
European car is used at 2% of its capacity, and even fits, from lower pollution to reduced traffic congestion.
during business hours, the average European office spa- However, they would constitute a major shift in how mo-
ce is only used at about 40% of capacity. In contrast, bility and buildings use are organised. What the aba-
a system of shared vehicles, designed to be optimised tement cost analysis shows is that these shifts would
for intensive use and with much longer lifetimes, would make a major contribution towards a low-carbon transi-
spread the cost of cars over a much greater number of tion, while also offering economic advantages.
40
300
200
100
0
BASELINE DEMAND-SIDE CURRENTExhibit 1.15
REMAINING FURTHER EMISSIONS
EMISSIONS MEASURES LOW-CARBON EMISSIONS REDUCTIONS REQUIRED FOR
A systemic shift towards circularity is a major productivity

(CIRCULARITY) PROCESSES CLIMATE TARGETS
opportunity – both for individual users and society
COST OF PASSENGER TRANSPORT CO-BENEFITS FROM A SHIFT TOWARDS SHARED CARS

EUR PER 1000 PASSENGER KILOMETRES EUR PER 1000 PASSENGER KILOMETRES
264
155
-74 %
-77 %
14 % 14 %
68
35
CURRENT CIRCULAR CURRENT CIRCULAR

SCENARIO SCENARIO
COST FOR MATERIALS, COST OF SOCIETY, INCLUDING INFRASTRUCTURE,

MANUFACTURING AND DISTRIBUTION PARKING AND LAND
COST OF DRIVING, INCLUDING FUEL, COST FOR CONGESTION

INSURANCE, PARKING ETC.
COST FOR MAINTENANCE OTHER EXTERNALITIES, INCLUDING ACCIDENTS,

NOISE AND AIR POLLUTION
Shared car fleets and higher utilisation per car lower cost per kilometre Fewer cars due to higher occupancy per car result in improved
Larger share of secondary materials and components reduces cost air quality and reduce noise level and congestion
for materials Autonomous cars are safer, reduce need for signs, lanes
Shared cars that are to a larger extent electric, autonomous and of and other infrastructure and can optimise trafic flows
smaller average size reduce cost of materials and cost of driving More electric vehicles reduce noise level and have positive impact
Longer car lifetime reduces costs per kilometre, while more durable on air quality, especially when shitfing towards renewable energy
materials, autonomous cars etc. also have negative impact on cost savings sources
41
1.4 SEIZING THE OPPORTUNITY
WILL REQUIRE LEADERSHIP
AND CONCERTED ACTION
A key take-away from our analysis is that the poten- what the concrete mitigation and industrial opportunity
tial of circular economy measures can only be realised is in each materials category and value chain, summari-
if policy-makers and businesses actively pursue these sed in Exhibit 1.16. We then turn to a discussion of the
opportunities. In this section, we lay out a broad imple- feasibility of different measures, barriers that need to be
mentation agenda. We start by describing in more detail overcome, and key policy interventions needed.
42
Exhibit 1.16
Abatement opportunities across materials and key measures
ABATEMENT OPPORTUNITY KEY MEASURES AND ENABLERS CO2 ABATEMENT
As the EU steel stock saturates, there Enable high-quality secondary production

is potential to meet demand to a large ex- through more advanced scrap markets
tent through secondary steel production. In Avoiding contamination of steel by copper
combination with low-carbon electricity, this Increase collection of post 41 Mt
STEEL makes possible deep cuts to emissions consumer scrap
Reduce fabrication scrap
Plastics recycling is very low, despite Product design that facilitates recycling
high technical potential. Secondary plas- Large-scale and specialised recycling
tics production driven by intrinsic materials operations and regional integration of markets 117 Mt
value could see more than half of total plas- Technology development for sorting,
PLASTICS tics recycled automation, and chemical recycling
Aluminium recycling can be increased Reduce collection losses

further by reducing losses of aluminium, Increase alloy separation in scrap
process scrap, and by avoiding the downg- recycling to avoid downgrading and
26 Mt
rading of aluminium through the mixing of thereby increase the usefulness of
ALUMINIUM different qualities of metal secondary aluminium
Reduce scrap forming during production
Cement recycling is at an early stage, Increased development of smart

but both re-grinding and re-use of building crushers and increased use of recovered
structural segments are now being trialled concrete in construction 25 Mt
CEMENT Development of local markets for
re-use of structural segments
Materials in cars have the potential to be Increased lifetime and more durable
more efficiently used when cars are shared materials will be keys to the economic
and more intensively used logic of the sharing model
19 Mt
PASSENGER CARS Vehicle size customized for the number
of people riding saves materials as cars
can be smaller
Material use for buildings can decrease Material savings during construction
by 30% as they are used more efficiently by reduction of waste
Through engineering for light weighting
can less material be used
Development of local markets for 55 Mt
BUILDINGS re-use of building components
Increased sharing of space to reduce
total floor space
Similar CO2 abatement opportunities for Light weighting to reduce materials

other product groups, such as rest of trans- per product
OTHERS portation, machinery etc. Development of local markets for 13 Mt
re-use of building components
Prolonged lifetime
Leasing models to increase utilisation
43
The Circular
– a Powerful
EconomyForce
– a Powerful
for Climate
Force
Mitigation / The
for Climate climate/ Executive
Mitigation potential of
summary
a circular economy
Understanding the
opportunities and
barriers across
materials and
value chains
STEEL
The CO2 abatement potential for steel use arises from Building such a circular system would require three
the possibility that future EU steel needs could be met to main actions. First, current losses of steel need to be
a large extent by recycling existing steel. Current practice reduced by higher collection of process scrap and scrap
is not set up for this possibility, but instead requires a from end-of-life products. Second, secondary steel pro-
substantial share of primary production – both to replace duction needs to improve to match the quality of primary
steel that is lost at various points in the use cycle, and to steel. Today, secondary steel is disproportionately used
compensate for downgrading of steel quality. By addres- for relatively basic construction steels, while for more de-
sing these issues, the EU could transition to a much manding uses, primary metal is typically used. If secon-
more circular steel system by 2050, replacing 30 Mt of dary production is to serve also more demanding product
today’s 92 Mt of primary steel production by secondary groups, a more developed market is required, matching
production, and reducing emissions by 57 Mt CO2 per scrap inputs to secondary steelmaking with the needs
year by mid-century. and tolerances of high-quality steel production. This is
doable, as demonstrated by several producers who rely
Achieving this would require a major reconfiguration of on a combination of good control of scrap supplies and
EU steel production. The current industry reflects a long advanced metallurgy to make some of the highest-quality
legacy of building up a substantial steel stock – i.e., the steels in the world – entirely from scrap. It also would
total amount of steel in use in the economy at any one create new sources of value and business opportunity in
time. This has required constant additions of new, pri- a more advanced scrap market.
mary metal. Primary production now accounts for nearly
60% of production, and the large majority of the sector’s Third, it is crucial to address the problem of copper
CO2 emissions of around 230 Mt per year. pollution. Copper often enters steel scrap at the point of
recycling, as products containing both steel and copper
These parameters may change in decades ahead. The are dismantled. When cars are scrapped, for example,
growth in the total stock is slowing and may well saturate it is common for the steel to have more than 0.4% of
in the next decades. At this point, demand is concentra- copper content even after basic processing. This redu-
ted on producing steel to replace annual stock turnover ces the quality and potential uses of the secondary steel.
of 2–3%. A detailed analysis of available steel stocks and Even levels of 0.2–0.3%, which are seen in several EU
flows suggests that the volume of scrap available will countries, lower the value of the steel. Moreover, once
approach total steel requirements, likely by the 2030s. mixed in, copper cannot be separated from the steel with
This is a major shift, with the prospect that, for the first any commercially viable technique. To keep the quality of
time, an industrial economy could meet its steel requi- the steel stock, processes must be improved to dismant-
rements largely by recirculating the stock it has already le products more carefully at end of life, to sort better,
built up. Given the much lower CO2 emissions from se- and to separate high-copper scrap from purer varieties.
condary steel production, this also could drastically redu- Design improvements can also make it easier to avoid
ce CO2 emissions from steel production. cross-contamination of materials during recycling.
44
PLASTICS
Most plastics can be recycled and used multiple times, product groups. Better design can make it easier to se-
but as noted above, actual volumes of secondary plas- parate different types of plastic, for example, and make
tics production amount to just 10% of total demand in used plastics easier to clean. This represents a major
the EU. A detailed assessment of plastic types and use change from today’s practice, where plastics recycling
categories suggests that 56% of plastic volumes could has sprung mostly from a waste handling logic, with little
be mechanically recycled or reused, with the recovered or no adaptation ‘upstream’. In tandem, secondary plas-
material value paying for much of the cost. Another 11% tics production must transition from today’s fragmented
could be recouped through chemical recycling techni- and small-scale activity to large-scale operations that can
ques (such as pyrolysis and depolymerisation). Together, reap substantial benefits of scale and enable specialisa-
these measures would reduce emissions from 233 to tion. Finally, major investments are needed to accelerate
144 Mt CO2 per year, compared with producing new the development of technologies to mark different plas-
plastics and incinerating them at end-of-life. tic types, automate sorting and processing, and recycle
plastics chemically.
Indeed, plastics offer the greatest untapped potential
identified in our analysis. It arises both because of the Although these changes can seem daunting given
low starting point for recycling, and because the contri- today’s low starting point, much can be gained by first
bution of plastics to CO2 emissions will grow drastically focusing on three key product categories – packaging,
over time. Right now, the main alternative to recycling is automotive and buildings – and on the five plastic types
incineration of plastics for energy recovery – essentially, that jointly represent more than 70% of plastics use. Over
using them as a fossil fuel, and in the process releasing time, successful approaches can be rolled out more bro-
as much CO2 as was created when the plastics were adly to fully realise the potential for cost savings and CO2
first produced. Today the net emissions from incinerating reductions.
plastics are rarely noted and arguably modest, as another
fossil fuel would likely be used instead. By mid-century, Moreover, achieving this would have significant econo-
however, heat and electricity production in a low-carbon mic value: our analysis suggests that concurrent impro-
EU would need to be largely emissions free. At this point, vements in product design and materials choice; increa-
using plastics for energy production becomes a major sed scale in collection and recycling, and improvement
source of fossil CO2 emissions. Recycling helps avoid in underlying technologies jointly can substantially im-
that problem. prove the economics of recycling. Much of it could be
driven by underlying value of the material, or else have
To achieve recycling of 56% of plastics volumes (rather a low cost of abatement compared to many other me-
than the 10% we see today) will require change throug- asures to reduce CO2. In tandem, a secondary plastics
hout the value chain. Above all, the potential for high-qu- industry of this scale would have revenues of some 30
ality recycling must be built into the design of the main billion per year.
45
ALUMINIUM
The resource gains from recycling are particularly large are often mixed together and used to produce cast alu-
in the case of aluminium, as remelting existing metal re- minium. However, the recycled material has limited uses;
quires just 5% of the energy of new production. Moreover, most aluminium products need to be made from metal
the EU imports about one-third of its aluminium, much of with much more tightly controlled alloy content. Over the
it produced with coal-fired electricity and carrying a CO2 next decades, a continuation of downgrading of alumi-
footprint as high as 18 tonnes CO2 per tonne of alumini- nium therefore threatens to undermine recycling. This
um – almost nine times more than one tonne of primary is even more so as cast aluminium is used primarily in
steel.17 Recycling aluminium can reduce CO2 emissions components of internal combustion engine drivetrains of
by as much as 98% relative to this. cars. As global markets shift more and more to electric
vehicles, demand for cast aluminium could fall, even as
Reducing the need for primary aluminium, particularly more and more aluminium is alloyed to the point where it
imports, could thus yield significant energy savings and has no other uses than to make cast products.
CO2 reductions. We estimate that improvement on current
practice could avoid 3-5 Mt per year of primary alumini- In decades to come, continued aluminium recycling
um production that otherwise would be required to serve will thus require that additional flows of aluminium are
European demand by 2050, or 40-60% of the total. The kept in ‘closed loops’, so that metal can be recycled and
CO2 avoided could amount to 30-50 Mt CO2 per year. used for the same purpose repeatedly, similar to current
practice for beverage cans. As with steel, a continued
Two changes to current practice will be necessary for circular system for aluminium would require product de-
this. A first step is to reduce losses, which now amount sign that enables separation of individual qualities, more
to almost 30%. This requires higher collection and better developed dismantling of products at end-of-life, advan-
end-of-life treatment for a range of products, alongside ced sorting technology, and additional deposit schemes.
design of those products so that aluminium can be re-
covered. Overall, aluminium therefore is yet another example
of how a future circular system will require change of
It also will be necessary to improve recycling practi- current practice, with large CO2 benefits as a result. This
ce, as current methods result in irreversible downgrading also creates new opportunities for value creation, in pre-
that in time will become an obstacle to recycling. Spe- serving and monetising the inherent value in secondary
cifically, today’s aluminium recycling is an ‘open loop’: materials.
after first use, a variety of specialised aluminium alloys
46
PASSENGER CARS
As noted above, passenger cars offer a major opportu- ring of vehicles also makes possible a much closer match
nity to provide the same service with much lower materi- of vehicle size to the needs of individual trips, thus redu-
als requirements. We find that in a system where profes- cing the average size of vehicles substantially.
sionally managed, shared vehicle fleets meet two-thirds
of travel demand, materials requirements could fall by as These factors combined can reduce materials require-
much as 75%, reducing annual CO2 emissions associa- ments dramatically. The average car would be smaller, far
ted with materials production by 43 Mt by 2050. more durable, and better maintained. The initial design and
materials choices would be optimised for much more inten-
Today’s cars are optimised for the use pattern associ- sive use. Combined with an electric drivetrain, the effective
ated with ownership by individual households. The result lifetime therefore could more than double. Lightweighting
is overcapacity of individual vehicles (five-seat cars used techniques that use advanced materials would be far more
mostly for one-passenger trips), and vehicles that are economic than when applied to a personal car. The same is
stationary 95% of the time. Vehicle design therefore also true of using automation to reduce accident risk, and app-
is optimised for this structure of use and ownership. lying more advanced manufacturing methods such as 3D
printing to reduce materials losses at the production stage.
A car system built around professionally managed fle-
ets of shared cars would change many of the underlying Although we describe this primarily in terms of materi-
incentives. Sharing enables much more intensive use of als requirements, the main motivation for such a system
each vehicle. Once the use of these fleets achieves suffi- is the much wider productivity opportunity that it repre-
cient scale, there will be enormous incentives for changes sents. As noted above, we calculate that the cost per pas-
to the design of vehicles and for innovation. Higher uti- senger-kilometre could be as much as 77% lower. Major
lisation justifies much more investment in upfront costs, externalities also could be reduced by three-quarters,
from the higher cost of electric-vehicle drivetrains, to more including major costs from factors such as traffic cong-
advanced automation technology, or higher-performance estion, air pollution and collisions. Although the pace of
materials. Professionally managed fleets in turn also ena- change will depend on many factors, not least travellers’
ble much greater control over vehicle maintenance, parts expectations and norms about car ownerships, the analy-
inventory, reuse of components, and remanufacture. Sha- sis makes plain that the incentives are very strong.
47
BUILDINGS AND CEMENT

The EU buildings sectors uses as much as 1.6 billion ton- Key opportunities to reduce emissions include:20
nes (Gt) of materials per year, a number that is likely to grow • Design for longevity and disassembly: Much of the
as the total stock is slowly expanding, and a wave of post-war abatement potential lies beyond 2050 but requires action
buildings will require substantial renovation or replacement today to avoid saddling future generations with unneces-
in decades ahead. The CO2 footprint is significant, and in sary construction needs.
the EU countries that have gone furthest in improving energy
efficiency and decarbonising heat, the construction phase • Floor space sharing: Sharing principles include com-
now accounts for as much as half of the lifetime CO2 foot- munal space-sharing in residential settings, co-working
print of a building (with the use-phase the remaining half).18 offices, etc.
By 2050, we estimate that just cement, steel, aluminium and • Reuse of existing buildings: adaptive reuse (i.e.
plastic used for construction would result in emissions of reuse of the building structure for a new purpose, e.g.
230 Mt CO2 if they were made with today’s production pro- office building becoming residential building), relocation
cesses. Demand-side measures could reduce this by 80 Mt of buildings (reuse of structure on a new site) and reuse
CO2 per year, or 30%. of building components can save up to 70% of materials
that otherwise would go to waste.
The abatement potential springs from addressing sig-
• Materials efficiency: reducing materials needed
nificant underlying inefficiencies in the construction value
per building through new materials choices, new de-
chain. The building sector is characterised by multiple
sign of building components, less over-design, and new
stakeholders with diverse interests and contract structu-
construction techniques.
res from design, through to construction and use. These
result in deeply split incentives; for example, a common • Reduced waste in construction: waste during
situation is that the client pays for materials, whereas the construction can be reduced by 5–10% of total materi-
builder can only improve profit by reducing labour costs als through increased standardization, improved planning
– skewing incentives strongly against strategies to redu- and appropriate storage and transportation.
ce materials use. The rate of innovation and knowledge • Cement recycling and reuse: there is some potential
diffusion also is slow, as a result of high industry fragmen- both to reprocess concrete to recover some unreacted
tation, high workforce turnover, a project-based business, cement, and to directly reuse structural elements. Scaling
and low investment and R&D in a cyclical and volatile bu- these would depend on highly developed, local markets.
siness. One marker of the inefficiencies is stagnant labour
productivity, with no growth over the past 15 years (even In many cases, these strategies are underused today
as manufacturing productivity increased by a third). not because they are inherently expensive, but because a
range of barriers and split incentives prevent their deploy-
Unsurprisingly, these features also translate to signi- ment. This resembles the situation for energy efficiency in
ficant structural overuse and waste of materials. For ex- buildings, where the solution often has been to mandate
ample, an estimated 50% of steel used in buildings is in minimum energy efficiency standards to capture potential
excess of what is needed to achieve the desired structu- that is cost-effective. In other cases, the business case
ral properties.19 In some segments, utilisation of building does not stack up today, but could improve with new
space is very low; for example, some 60% of European technology. The clearest examples are improved materi-
office space is unoccupied even during working hours. als efficiency and reduced waste in construction, both of
Between 10–15% of delivered materials are not employ- which are limited today by the concern to reduce labour
ed in actual construction process but wasted. Buildings cost. Digitalisation of materials management and auto-
often are demolished wholesale, even when structural mation of construction could significantly reduce the cost
elements (accounting for 85% of the GHG footprint of of techniques to reduce materials use.
materials) have long remaining technical lifetime.
48
Other value chains

Passenger cars and buildings jointly account for sectors is likely to be lower than the 75% reduction
just under half of emissions from cement, steel, plas- in the case of passenger cars, but a conservative
tics and aluminium. However, many of the underly- analysis suggests that some 13 Mt CO 2 of additional
ing principles also apply to other products, including abatement would be available in the EU through a
heavy-goods vehicles, machinery in agriculture and range of circular economy strategies across additio-
industry, packaging, some infrastructure, and several nal transport, consumer goods, and infrastructure and
consumer goods categories. The potential in these construction value chains.
49
A new policy agenda to realise the opportunities

The opportunities explored above, cumulatively, would As with other climate policy challenges, a concerted
significantly reduce CO2 emissions from industry. Mor- effort is needed to understand potentials, identify and
eover, as shown in Section 3, many of the options are test new approaches, recognise and avoid unintended
economically attractive, powered by the inherently higher consequences and undue costs, and coordinate efforts.
productivity of shared mobility systems and the savings The agenda may seem daunting at first, but in many ca-
from reduced waste and avoided materials costs. Yet ses, policy-makers can build on frameworks already in
it is also clear that multiple barriers stand in the way. place.
Some, such as the problem of copper contamination of
steel, will require sector-specific measures and speciali- Our analysis did not investigate the merits of specific
sed technologies; others, such as high-value plastics re- policy instruments, so we do not offer specific recom-
cycling, require much more alignment of incentives along mendations. However, we sketch below the elements of
value chains, from the initial design of a product to its this policy agenda at a high level, recognising that much
end-of-life treatment. more investigation is required to elaborate concrete poli-
cy proposals. Four broad approaches are needed:
Energy efficiency again offers a helpful example. As
the barriers that hinder progress have been recognised, 1. Set the direction and recognise the potential of the
policy-makers have adopted a wide range of interventions circular economy as a major contributor to climate tar-
to capture cost-effective potential – from aggregate targets.
gets, through to quota systems, financing mechanisms,
subsidies, and detailed product-level standards and la- 2. Create enablers: from publicly funded research, to
belling schemes. Similarly, multi-faceted approaches will infrastructure, to regulatory frameworks.
be needed to promote circularity.
3. Level the playing field, to remove incentives to waste
Circular economy policy will require action across waste, and improve the financial case for investment in circular
transportation, and the built environment, as well as infra- economy measures.
structure, industrial and innovation policy. Some will cover
familiar ground – such as effective waste management – 4. Take government action across a range of areas,
but some will push policy-makers into unfamiliar realms and especially where government already is a major actor
require learning and ingenuity. Along with EU-level policies, shaping outcomes.
these measures will require national implementation as well
as attention by cities and regional authorities.
50
Exhibit 1.17
Elements of a policy agenda: examples of

potential measures to investigate
1. SET THE DIRECTION 2. CREATE ENABLERS 3. LEVEL THE PLAYING FIELD 4. TAKE GOVERNMENT ACTION
Targets for high value Support for innovation Carbon pricing Public procurement
recycling and technology deve-
Extended producer
lopment Waste regulation systems
responsibility
Improved transparency
and statistics for waste Standards for Quotas or other support
and recycling secondary materials for demand
MATERIALS
RECIRCULATION Waste regulations and Improved end-of-life
landfill bans handling (e.g. shred-
ding, demolition)
Regulation of long-term
destructive practices Product design (e.g.
(e.g. copper, additives) Ecodesign directive)
Targets for efficient Information systems Waste charges Stimulate re-use and
materials use and re-use and platforms (e.g. recycling markets
in key sectors BIM) Support for design for
disassembly Require high materials
Materials passports and Labelling schemes for efficiency (re-use, design
PRODUCT documentation materials efficiency in for disassembly, etc.) in
MATERIALS construction etc. public procurement
EFFICIENCY
Fund innovation and
technology development
Improved evidence base Create supportive city Pricing or regulation of City plans for longevity
and conviction about regulations (e.g. par- congestion, air pollution and adaptability of
benefits of shared mobi- king for shared cars) and other externalities buildings
lity and buildings
Revise barriers in exis- Include efficient mate- Integrate car sharing
CIRCULAR Endorse shared mobility ting regulations (e.g., rials use or labelling in with public transport
BUSINESS systems as target vision insurance) building standards systems
MODELS Adapt data regulations

to support in new
business models
51
1. SET THE DIRECTION 2. CREATE ENABLERS
The potential of circular economy measures to contri- The circular economy will be built mainly by the private
bute to climate targets is far from recognised in cur- sector: the companies that produce major materials, the ma-
rent climate or industrial policy. Climate action plans nufacturers, construction firms and consumer goods sellers
at international, national, or city level make no or little who put the materials to use, and waste management and
reference to this agenda. Thus, as a first step, there recycling firms. The crucial role of the public sector in this
is a need to focus the attention of policy-makers, civil context is to facilitate and encourage private action.
society, researchers and business leaders. Policy can
help by clarifying the intention to incorporate circula- One key way to do this is to stimulate innovation. Technolo-
rity among other aspects of EU climate and industrial gical advances are making circularity measures more viable
policy. than ever before. Opportunities to capture value abound: from
platforms for car-sharing, to autonomous vehicles, to sensor
The EU’s ‘20-20-20’ 2020 targets for emissions reduc- and sorting technology for scrap metals, tracking of materials
tions, renewable energy and energy efficiency provide inflows to buildings, chemical marking of plastics types for easy
inspiration in this regard. Although some criticise the sorting, automation of construction processes, automated disas-
mixing of many targets, the long-term commitment has sembly of complex products, chemical recycling technologies,
also been a significant catalyst in changing expectations methods to remove copper from steel, etc. EU companies that
about the direction of the energy system. At the time of seize these opportunities could position themselves as future
the targets’ introduction in 2009, there was considera- global leaders. Private actors may not yet be ready to make ma-
ble doubt and disagreement about whether renewable jor investments in these innovations, however. The public sector
energy would be a major factor in the energy system can step up the pace by funding R&D and early deployment.
over the next decade. The targets agreed by EU Mem-
ber States coordinated efforts to ensure that it would Regulations may also need to be adapted or introdu-
be, and made a major contribution towards channelling ced to enable circular economy opportunities. As ever,
the investment, supply-chain industrialisation, and inn- regulatory approaches must be weighed carefully, to ba-
ovation that, in turn, helped drive down costs. Today, lance different objectives. Areas of waste regulations to
many assessments indicate that a largely renewable consider include how to handle current incentives for inci-
energy-based electricity system is not just feasible, but neration of waste; the impact of additives and toxicity on
potentially available at a cost similar to one run on fossil plastics recycling; ‘end-of-waste’ policy that can result in
fuels. A strong political signal was part of what made obstacles to the trade of secondary materials; and how to
that possible. ensure clarity about ownership of and access to end-of-life
materials. Standards can offer another area of developme-
The circular economy needs a similar articulation of nt, and especially to encourage European or international
ambitions. From a CO2 perspective, the key aim is to standards for secondary materials.
develop business models and industry structures that
enable us to maintain our modern economy and living Other regulatory areas to investigate include measu-
standards with much lower levels of primary materials res to increase transparency and reporting. For example,
production. That, in turn, requires a) secondary materi- would it be warranted to introduce reporting and track-
als production, b) improved materials efficiency in major ing of materials efficiency metrics or the materials con-
product groups, and c) increased lifetime and sharing of tent of buildings, much like energy performance or safety
materials-intensive capital assets in the economy. Set- standards are today? Or would the impact be limited, and
ting targets for these would be a good starting point for costs large? Several reports on the circular economy have
the EU. also proposed that the EU expand its Ecodesign directive
to consider materials more systematically, whereas others
In parallel, policy-makers can help ensure that they have considered this too intrusive an approach.
are moving in the right direction by kick-starting a ma-
jor knowledge effort on the circular economy. Hund- Policy also has strong influence on many of the enablers
reds of times more research has been done on energy for new business models based on sharing vehicles or buil-
efficiency than on the efficiency of materials use. We dings. These often are data-intensive, and sound approaches
need new knowledge to ensure that long-term policy is to data protection and cybersecurity can be a major factor
based on a much fuller understanding of the potential, in their long-term viability. Likewise, infrastructure and local
the barriers, and the economics of circular economy policies (from parking rates, to congestion charges and city
measures. planning) strongly influence mobility choices, as can public
regulations that influences insurance, liability laws, etc.
52
nisms. Although these different options show the range of
3. LEVEL THE PLAYING FIELD available possible instruments, significant additional investi-
gation would be required to evaluate first whether such sti-
To a great extent, market conditions today are skewed mulation of demand is warranted, and if so which options are
in favour of waste and against materials efficiency and reu- best suited for different markets and specific situations.
se. To improve the business case for more circular business
models, the financial incentives will need to change. Policy intervention may also be required to limit ‘re-
bound’ effects that can arise from increased efficiency. Re-
The fact that so many negative-cost opportunities have bound arises when the savings from increased efficiency
yet to be seized suggests that carbon prices and related lead to increased consumption. This is well known in energy
tools will not be enough. Many of the relevant materials are efficiency policy, but has been explored less for materials.
internationally traded commodities, making it difficult for the An example would be the possibility that much lower cost
EU to unilaterally introduce carbon prices at the levels re- of transportation in a shared car system leads to increased
quired for deep cuts to emissions – whether on the supply travel and reduced public transit use – already a challenge
or the demand side. But even high carbon prices may not in some cities. Another risk is that the availability of inexpen-
suffice, because many of the barriers are non-financial. Exis- sive recycled materials (particularly if they are of low quality)
ting regulations also steer key sectors such as buildings and will lead to increased consumption. Policy interventions can
waste management in ways that may encourage or discoura- reduce rebound effects, even if they cannot be fully avoided.
ge circularity. Thus, while a carbon price can help, policy-ma-
kers will need to do much more to unleash the full potential.
4. TAKE GOVERNMENT ACTION
A major issue is how best to provide incentives to manu-
facturers to account for the impact of materials and design Finally, government can also be a principal actor to enable
choices on component and materials values at the product’s the transition to more circular outcomes. The public sector is
end of life. To date, the main approach has been ‘extended both a major provider and user of mobility services, owner or
producer responsibility’, but in practice its implementation pro- operator of much of the built environment, and buyer of a range
vides little incentive to modify product design. The externality of materials-intensive products. That means governments are
therefore remains largely in place, and it should be a priority well positioned to push the market towards a more circular
to consider creative approaches for how it can be addressed. economy through their own investments and purchases.
This report has not evaluated which way would be the most Government also already is a major shaper of outcomes
promising way forward, and therefore does not provide spe- for transport systems and the built environment, particularly
cific recommendations. To give a flavour of the options that through decisions taken at the city level. One priority is to en-
have been proposed, one possibility is that new marking and sure public transport systems are integrated with shared car
sensor technology could enable systems that create more systems. City planning also influences the nature of buildings.
individual product responsibility incentives. If recycling com- As the potential for reduced materials use shows, issues such
panies can distinguish products from different manufacturers, as materials efficiency and flexibility in the building stock are
they could then also in principle pay different amounts, depen- strong contenders to be considered alongside more traditional
ding on how adapted the products are for reuse or materials issues such as the density of occupation, social issues, etc.
recovery. Another idea that has been floated is to introduce
a system of charges or subsidies that mimics differentiated Public procurement is another area open for conside-
producer charges, depending on the costs that different de- ration. Starting to incorporate recycled content and ma-
sign choices impose at end of life. If workable, these solutions terials efficiency into purchasing decisions could help
could avoid some of the risks with more direct regulation of boost nascent markets. As with other proposals discus-
products, but they also carry their own complexities. sed here, further evaluation is required.
A second approach is to create demand for recycled * * *

materials or materials-efficient products. The ‘pledging’ initia-
tive in the Commission’s Plastics Strategy, where companies A common concern about the transition to a more circular
commit to using recycled plastics, is one attempt to create a economy is that it requires ‘systemic’ change, which is seen
basis for investment in additional recycling capacity to crea- as difficult if not infeasible. It is true – realising the full po-
te high-quality secondary plastics. Research has identified tential described in this report would require transformative
a quota system for recycled content as a possible option, change. But what is most exciting about this opportunity is
similar to quotas for renewable energy and energy efficien- that smaller, manageable policy packages, combined with
cy. Another proposal is to support nascent markets directly an ambitious vision for the future and strong political will,
(e.g. reused structural building segments, recycled cement, could unleash a wave of innovation and new investment that
recycled plastics), through subsidy schemes or other mecha- pays off for generations.
53
The Circular Economy – a Powerful Force for Climate Mitigation / Steel
2. steel
towards a circular steel system
Steel offers particular promise for a major industrialised economy could meet
circular economy. Already today, one-third its needs for a fundamental material almost
of the steel we use comes from recycled entirely through recycling. CO2 emissions
metal, and the analysis presented here from steelmaking would drop by 80% from
shows that within five decades, seconda- current levels if low-carbon electricity is
ry steel could meet nearly half the world’s used. Thus, the EU would approach a CO2-
steel needs. In fact, even with rapid global free steel sector in Europe – a global first.
economic growth, future growth in steel de-
mand could be met largely by reusing steel, Realising this opportunity will require
with primary steel production held steady. significant changes to minimise losses in
The climate benefits would be significant: volume and quality from one use cycle to
that level of recycling would cut the average the next. The contamination of steel with
CO2 emissions per tonne of steel produced copper could pose particular challenges.
by 60%, reducing overall emissions from EU secondary steel production would also
steelmaking by nearly 4 billion tonnes of have to be restructured, while continued pri-
CO2 per year by the end of the century. mary production would be more focussed
on exports. However, the industry already
For the EU, the opportunity is even more faces major challenges: global over-capa-
immediate. Europe already has a large city, flagging profitability, high CO2 emis-
stock of steel, near the saturation point. Our sions, and the threat of tariffs. A circular ste-
analysis shows that if downgrading of steel economy offers a promising path forward
el is avoided, secondary steel production by boosting productivity and making the EU
could meet as much as 85% of the EU’s a pioneer and leader in the technologies of
steel needs by 2050. That would be a re- the future.
markable achievement: for the first time, a
54 54
Available scrap could cover
85% of EU steel
needs by 2050.
55
2.1 UNDERSTANDING FUTURE STEEL NEEDS

Steel is used across all areas of a modern, industrial with stocks ranging from about 5 tonnes per capita in
economy. About half the steel produced today is used in China, to barely 1 tonne in much of Africa and India.3
construction and infrastructure; another 16–17% each in
transportation and machinery and electrical equipment, We estimated future demand for steel by quantify-
and the remaining 15% in a variety of metal products and ing the production levels needed to enable the entire
appliances.1 With few exceptions, the benefits from steel world to gradually transition to per-capita steel stocks
derive from the total amount of steel available, and the similar to those in OECD countries, following similar
services this steel stock enables: shelter/buildings, infra- patterns of use (see Exhibit 2.1 for an explanation
structure, mobility, urban development, industrial produc- of our methodology).4 Steel production would have
tion, and more. Steel flows, such as annual production, to rise steadily, to 2.5 times the current level. CO2
have little benefit except by enabling societies to develop emissions levels increase less, in part because pro-
and maintain this stock. duction technology improves but mostly because
there is a shift towards more recycled steel even in
The analysis presented in this chapter focuses on the a baseline scenario. These are not best understood
EU, but to understand the developments in one region, as forecasts, but as an illustration of what would be
one needs to also take a global view. There is significant required if the current pattern of steel use in matu-
international trade in raw materials, steel products and re economies were adopted globally. As we discuss
scrap, and steel firms operate across borders. Global elsewhere in this report, this may not be necessary:
growth in steelmaking today is driven mainly by deve- steel needs could fall sharply if we could make do
loping and emerging economies, as they build up their with less steel to provide the same mobility, shelter,
stock of steel: the total amount of steel available in the or other services. Also, as we elaborate below, it is
economy. In mature economies such as the U.S. and EU, possible to reduce losses and increase the share of
this stands at about 11–14 tonnes per person2, and steel recycled steel beyond such a ‘business as usual’ de-
demand is already driven largely by the need to repla- velopment. On both counts, there is significant poten-
ce older products or structures. In much of the world, tial to reduce the need for primary steel production,
however, there is still a large unmet demand for steel, and therefore also CO2 emissions.
56
Exhibit 2.1
Baseline projections for steel demand,

emissions, and production
3 500
PRODUCTS
3 000
2 500
MACHINERY
2 000 TRANSPORTATION
DEMAND BY PRODUCT GROUP
Mt STEEL PER YEAR, 1 500 CONSTRUCTION
2015-2100, GLOBAL 1 000
500
0
2020 2030 2040 2050 2060 2070 2080 2090 2100
4 500
4 000
3 500
3 000
CO2 EMISSIONS 2 500
Mt CO2 PER YEAR, 2 000

2015-2100, GLOBAL 1 500
1 000
500
0
2020 2030 2040 2050 2060 2070 2080 2090 2100
TOTAL STEEL DEMAND

4 000
3 500 MCKINSEY (2018), MEDIUM SCENARIO

3 000 ALLWOOD ET AL. (2012)
XYLIA ET AL. (2017)
PRODUCTION PER ROUTE 2 500
PAULIUK ET AL. (2013)
Mt STEEL PER YEAR, 2 000 GLOBAL CALCULATOR, LOW
VAN RUIJVEN & VAN UUREN (2016)
2015-2100, GLOBAL 1 500
IEA-ETP (2017)
1 000
500
0
2020 2030 2040 2050 2060 2070 2080 2090 2100
SOURCE : PRODUCTION PROJECTIONS FROM MULTIPLE SOURCES. 10
57
E T T VÄ R D E B E S TÄ N D I G T S V E N S K T M AT E R I A L S Y S T E M / P L A S T
STEEL: METHODOLOGY AND MODELLING APPROACH

The modelling approach used in this study follows that developed and described in Pauliuk et al.
(2013).5 Steel demand is modelled in four end-use sectors and 11 world regions. Demand in each
sector and region is modelled with a stock-based approach, with population forecasts based on UN
Population Prospects 2017,6 and per-capita saturation is largely completed by 2100.7 Historical steel
stocks are based on Pauliuk, Wang and Müller (2013),8 while future stock turnover is modelled with
different lifetimes of products in each end-use sector following a normal distribution. Production is
calibrated to recent volume and production route data for key world regions (World Steel Association,
2017; Bureau of International Recycling, 2017).9 The steps of the steel supply chain and use cycle
are directly modelled, including losses in production, new scrap formation, the remelting process,
and collection of post-consumer scrap. The model also tracks inmixing of copper in the steel stock,
and the limitations this places on use. It further represents degrees of international trade in scrap, and
the dilution of steel scrap with primary steel. Note the difference between demand, which denotes the
requirements for steel in final products, and production, which is the amount of crude steel required
to service this demand. CO2 emissions estimates include direct emissions as well as indirect emis-
sions from electricity, and are based on the gradual adoption of best available techniques for current
production processes by 2050. The high-level results of a baseline scenario are shown below, with
further elaboration throughout this chapter.
58
1 000
500
0
The Circular Economy – a Powerful Force2020
for Climate
2030 Mitigation
2040 Steel 2070
2050 / 2060 2080 2090 2100
TOTAL STEEL DEMAND

4 000
A closer look at the EU highlights the importance3 500 EU demand would fall at first, then stabilise M atCKINSEY
repla-(2018), MEDIUM S
of saturation (Exhibit 2.2). The EU steel stock already3 000 cement levels of around 150 million tonnes per year.
ALLWOOD ET AL. (2012)
approaches 12 tonnes per person, and the population Production by the EU steel industry could ofXYLIA course
ET AL. (2017)
PRODUCTION
level is flattening and expected PER ROUTE
to decrease. Once a
2 500
PAULIUK ET AL. (2013)
still increase, but would have to be driven by increasing
GLOBAL CALCULATOR, LOW
saturation point is reached, EUMtsteel
STEEL PER YEAR,
demand would be2 000 exports. As we discuss below, the stabilisation of de-
VAN RUIJVEN & VAN UUREN
2015-2100,
driven almost entirely by the need GLOBAL
to replace products1 500 mand even as stocks accumulate would enableIEA-ETP Europe(2017)
and structures as they reach the end of their life, about1 000 to rely much more heavily on recycling to meet its own
2–3% of the total stock each year. In this scenario, 500 steel needs.
0
2020 2030 2040 2050 2060 2070 2080 2090 2100
Exhibit 2.2
EU steel demand can stabilise as the stock saturates
EUROPEAN STEEL PRODUCTION

Mt STEEL PER YEAR, 2015-2050
220
HISTORICAL DEMAND
200
PROJECTED DEMAND
180
160
140
120
100
80
60
40
20
0
2000 2010 2020 2030 2040 2050
CO2 INTENSITY OF STEEL PRODUCTION

t CO2/t STEEL
BASIC OXYGEN FURNACE (BOF)

BASIC OXYGEN 2.3 The large majority of current steel production uses coal to reduce
FURNACE (BOF) NOTE : THE FIGURE DEPICTS THE PRODUCTION OF CRUDE STEEL REQUIREDiron ore and EU
TO SERVICE produce
DEMANDsteel in integrated steelworks.
SOURCE : PROJECTED DEMAND MATERIAL ECONOMICS MODELLING AS DESCRIBED IN TEXT; HISTORICAL PRODUCTION BASED ON WORLD STEEL ASSOCIATION (2017) . 11
INCREASED PROCESS EFFICIENCY

BOF, WITH BEST
AVAILABLE TECHNOLOGY
1.9 An estimated 15% process efficiency improvement is possible within
the current BOF process.
59
2.2 CO 2 EMISSIONS FROM STEEL PRODUCTION AND USE

The need for much more steel in the future presents a di- In contrast, clean power does not help much to reduce
lemma for climate objectives, as the production of steel CO2 emissions from primary steelmaking. Most of the CO2
results in large emissions of CO2. Most steel production from primary steel production results from the chemical pro-
in Europe and globally is primary production, conver- cess in which iron ore is reduced, not from the energy used
ting iron ore to steel through the basic oxygen furnace for heat. Unlike in power production or heating, it therefore is
(BOF) route. This emits, on average, just over 2 tonnes not possible to use renewable energy directly. Established
of CO2 per tonne of steel produced.12 Doubling primary options can reduce emissions by around half. For deeper
steel production would thus double a major CO2 source. emissions cuts, it would be necessary to use methods of
steel production that are still under development (Exhibit 2.3).
However, as the global steel stock grows, so does
the opportunity to use remelted end-of-life steel as an To quantify the climate impact of continued heavy reli-
alternative to primary production. Secondary steel is ance on primary steelmaking and plausible scenarios for
made in electric arc furnaces (EAF), with one-fifth the technological improvements, we constructed two scena-
CO2 emissions of primary steelmaking even when much rios for how to meet projected demand growth:
of the electricity comes from fossil fuels. If power from • Baseline scenario: This sees continued reliance on
renewable sources is used in EAF, secondary steel can the BOF route for primary steel production, but with best
be decarbonised almost entirely. A more circular steel available technology rapidly adopted. Steel recycling
flow therefore both is inherently far less emissions-in- continues with the same collection rates, loss levels and
tensive, and easier to decarbonise through the use of practices as today.
renewable energy; it is one of many examples in this re-
port of how greater circularity moves the emissions from • Low-carbon process: For illustration, we also show
‘hard to abate’ sources and towards ones with much what the gradual adoption of currently proven abatement
more established ways to cut emissions. technologies in steel production would entail, so that the CO2
intensity of primary steel production falls by half by 2050.14
60
220
HISTORICAL DEMAND
200
PROJECTED DEMAND
180
160
140
120
100
Exhibit 2.3
80
Steel recycling has significantly lower emissions than

60
other available options for steel production

40
20
0
2000 2010 2020 2030 2040 2050

t CO2/t STEEL

FURNACE (BOF) iron ore and produce steel in integrated steelworks.
INCREASED PROCESS EFFICIENCY

BOF, WITH BEST
AVAILABLE TECHNOLOGY
1.9 An estimated 15% process efficiency improvement is possible within
the current BOF process.
NNE BIO-BASED INPUTS

BOF, WITH BIO FUELS 1.1 Bio-based fuels can substitute for some of the coal input, with
emissions reductions of around 50%.
DIRECT REDUCED IRON (DRI)

DIRECT REDUCED IRON
(DRI) 1.1 This route uses natural gas to reduce iron ore, which almost
halves emissions. DRI accounts for 5% of current world production.
CARBON CAPTURE AND STORAGE (CCS)

BOF + CARBON CAPTURE
AND STORAGE (CCS)
0.9 Capturing the CO2 from the blast furnace of an integrated steel plant
can reduce overall emissions by 60%.
ELECTRIC ARC FURNACE (EAF)

ELECTRIC ARC FURNACE 0.4 The main route for secondary steel uses electricity to melt steel scrap and /
or direct reduced iron, with only small onsite emissions.
(EAF)
ZERO-CARBON ELECTRICITY (EAF):

EAF + ZERO CARBON ELECTRICITY 0.1 By using zero-carbon electricity, nearly all the emissions from
steelmaking can be eliminated.
CO2 EMISSIONS, STEEL PRODUCTION CUMULATIVE CO2 EMISSIONS, STEEL PRODUCTION

Gt CO2 PER YEAR, 2015-2100, GLOBAL Gt CO2 2015-2100, GLOBAL
7.0
6.0 298
5.0
4.0 SOURCE : MATERIAL ECONOMICS ANALYSIS OF MULTIPLE SOURCES. 13 198
3.0
2.0
61
1.0
t CO2/t STEEL
Steel
The Circular Economy – a Powerful Force for Climate Mitigation /
ThereFURNACE
are good(BOF) and bad news in this scenario (Exhibit billion iron
tonnes
ore and(Gt)
produceCOsteel by 2100. Yet as noted in Chap-
2 in integrated steelworks.
2.4). The good news is that even current practices would ter 1, to stay within a carbon budget consistent with the
enable a much higher share of recycled steel in the futu- commitments of the Paris Agreement, emissions from all
re, as the amount of available scrap grows over time (see materials production
INCREASED PROCESS couldEFFICIENCY
not exceed 300 Gt CO2. Esta-
next BOF
section). EAF production could thus grow from1.9
, WITH BEST 26% blishedAnsupply-side measures
estimated 15% process efficiencyin the low-carbon
improvement scena-
is possible within
AVAILABLE
today, TECHNOLOGY
to 36% by 2050. This alone would reduce the CO2 the current
rio help, but even BOF process.
if those technologies were fully rolled
intensity of steel by one third compared with the current out by 2050, as in our ‘low-carbon production’ scenario,
situation. We are thus already set to reap some of the cumulative emissions would still reach 198 Gt CO2.
climate benefits of increased circularity of steel use. BIO-BASED INPUTS
BOF, WITH BIO FUELS 1.1 The main
Bio-basedconclusion from
fuels can substitute thisof the
for some analysis
coal input,iswiththat de-
The bad news is that this leaves much of the potential emissions
mand-side reductions of are
measures around 50%. to meeting climate ob-
crucial
for demand-side improvements on the table, through a jectives. Achieving deep cuts on the supply side alone
combination of steel losses and downgrading. CO2 emis- would require extraordinarily rapid, global implementa-
sions would therefore be higher than they need to be. tion ofDIRECT REDUCED
processes for IRON (DRI)
steelmaking that are still unproven
AsDIRECT REDUCED IRON
we detail
(DRI) 1.1 practices
in subsequent sections, improving at scale.
This This is not
route uses to gas
natural sayto that
reducedeveloping
iron ore, whichand almostdeploying
halves emissions. DRI accounts for 5% of current world production.
could raise the share of scrap-based steelmaking to al- these technologies should not be a priority: they are ne-
most half by 2050, further reducing emissions by several cessary to handle the primary production that will still be
hundred million tonnes of CO2 per year. necessary. But demand-side action is urgently needed
CARBON
as well, both CAPTURE
becauseAND STORAGE
of the emissions (CCS) reductions they
BOF +other
The piece of bad news is0.9
CARBON CAPTURE that emissions re- Capturing
provide, and becausethe CO 2
from the blast furnace
they reduce the of an challenge
integrated steelon
plant
the
AND STORAGE (CCS) can reduce overall emissions by 60%.
main far higher than required to meet climate objecti- supply-side significantly by reducing the volume of pri-
ves. Cumulative emissions from steelmaking reach 298 mary steel required.

(EAF)
Exhibit 2.4
A baseline scenario sees largeZERO-CARBON

reductions ELECTRICITY in(EAF):
the CO 2
intensity 0.1 of production, but large
steelmakingemissions remain
EAF + ZERO CARBON ELECTRICITY By using zero-carbon electricity, nearly all the emissions from
can be eliminated.
CO2 EMISSIONS, STEEL PRODUCTION CUMULATIVE CO2 EMISSIONS, STEEL PRODUCTION

Gt CO2 PER YEAR, 2015-2100, GLOBAL Gt CO2 2015-2100, GLOBAL
7.0
6.0 298
5.0
4.0 198
3.0
2.0
1.0
0.0
2020 2040 2060 2080 2100 BASELINE SCENARIO BASELINE
+ LOW CARBON
CURRENT EMISSIONS INTENSITY PROCESS
BASELINE SCENARIO
BASELINE + LOW-CARBON PROCESS
GLOBAL STEEL PRODUCTION AND AVAILABLE STEEL SCRAP 62

Even in a growing world, recycling of
steel could meet most of the growth in
future global demand.
63
(EAF)
ZERO-CARBON ELECTRICITY (EAF):
2.3 REALISING0.1THE POTENTIAL FOR

EAF + ZERO CARBON ELECTRICITY By using zero-carbon electricity, nearly all the emissions from
steelmaking can be eliminated.
CIRCULARITY IN THE STEEL SECTOR

CO EMISSIONS, STEEL PRODUCTION
2 CUMULATIVE CO2 EMISSIONS, STEEL PRODUCTION
The steel 2 PER YEAR,
Gt COsector has 2015-2100, GLOBAL
a head start in the circular eco- de, with striking 2015-2100,
Gt CO2 results: In GLOBAL
principle, the scrap avai-
7.0
nomy, especially in mature markets such as Europe – but lable around the world would be sufficient to meet
as steel stocks saturate, there is significant potential to most of future growth in steel demand with secondary
6.0 298
scale up recycling and minimise demand for virgin steel. steel (Exhibit 2.5). Primary production is still required,
Already, EU steel demand is driven mainly by the need to but only for net additions to the steel stock in rapidly
5.0
replace products and structures at the end of their useful growing countries. For economic and environmental re-
life. As4.0the total steel stock saturates, the resulting scrap asons, it makes sense to seize this opportunity
198 – and
can be re-melted to provide an ever larger share of the that means making it a priority to adopt the practices re-
flow of3.0steel required. quired to enable future recycling: from collection at end-
of-life, to keeping steel stocks pure from contaminants
We modelled
2.0 the amount of scrap available worldwi- that could otherwise limit the uses of recycled steel.
1.0
0.0
+ LOW CARBON
CURRENT EMISSIONS INTENSITY
BASELINE SCENARIO
Exhibit 2.5 PROCESS
Even in a growing world, recycling of steel could

meet most of the growth in future demand

GLOBAL STEEL PRODUCTION AND AVAILABLE STEEL SCRAP
4 000
STEEL PRODUCTION
3 000
2 000
AVAILABLE STEEL SCRAP*
1 000
0
2020 2030 2040 2050 2060 2070 2080 2090 2100
STEEL PRODUCTION TO MEET EU DEMAND AND AVAILABLE STEEL SCRAP

180 64
160 STEEL PRODUCTION
0.0
+ LOW CARBON
CURRENT EMISSIONS INTENSITY
The Circular Economy – a Powerful Force for Climate Mitigation / Steel PROCESS
BASELINE SCENARIO
GLOBAL STEEL PRODUCTION AND AVAILABLE STEEL SCRAP

The results
Mt STEEL are
PER still2015-2100
YEAR, more striking for the EU (Exhi- the full benefits will require significant changes in how
bit 2.6). Europe already produces around 100 Mt of steel flows are handled. There are three key issues, which
scrap every year. In 2016, EAF plants in the EU pro- we discuss next:
4 000
duced some 65 Mt of steel, while 18 Mt of scrap was STEEL PRODUCTION
exported, and the remainder used in integrated steel • Reducing the losses of steel throughout the use cycle;
plants.15 Looking ahead, the amount of available scrap • Preventing downcycling and enabling the
will increase, and we estimated that by mid-century, the production of high-quality secondary steel; and
3 000
amount of available scrap covers 80% of EU steel re-
quirements. • Reducing the contamination of steel – especially
by copper.
Both globally and in Europe, then, the future of steel
2 000
could be much more circular than it is. However, reaping AVAILABLE STEEL SCRAP*
1 000
Exhibit 2.6
0
2020 2030 Scrap
2040 availability
2050 2060 in Europe
2070 2080 rapidly
2090 2100
approaches the level of demand

STEEL PRODUCTION TO MEET EU DEMAND AND AVAILABLE STEEL SCRAP
180

140
120
100
AVAILABLE STEEL SCRAP
80
60
40
20
0
2020 2030 2040 2050
NOTE : EXCLUDING SCRAP EXPORTED OR USED IN INTEGRATED STEELMAKING.

LOSSES IN STEEL PRODUCTION
Mt STEEL 2015
65
NOT COLLECTED POST CONSUMER SCRAP
REDUCING LOSSES OF STEEL

Steel has a high recycling rate: around 83% of end- • Low scrap collection rates: While overall collection
of-life steel is collected for recycling16, and in some rates are relatively high, they are low for some product groups
countries, this reaches 90%.17 Still, overall losses are lar- and regions – as low as 50% for some consumer products18,
ge (Exhibit 2.7). We calculate that almost 150 Mt steel and for production scrap (new scrap), just 70%.19
are lost each year – an amount close to the EU’s annual • New scrap: For certain product groups, such as au-
production. The cost of producing new steel to replace tomobiles, only 50% of the steel used actually makes it
these losses is on the order of 70 billion EUR, while the into products. The remainder becomes scrap at the for-
CO2 emissions resulting from replacing them are more ming or fabrication stage. Overall, some 25–30% of steel
than 320 Mt CO2 per year. becomes new scrap.20 This means that more steel needs
to be produced and circulating at any given point. It also
Losses in the steel system occur in four main categories: leads to additional losses, as noted above.
• Obsolete stock and inaccessible scrap: Some • Remelting losses: Remelting of steel is highly optimi-
steel structures are simply abandoned and not scrap- sed to maximise yields, but losses nonetheless reach an
ped. Examples are submarine or underground pipes, or estimated 4–5%.21 One factor that increases losses is the
abandoned buildings or structures. In some categories, need to remove substances (mainly trace or alloying me-
the share of end-of-life steel that is not collected can tals) that would otherwise affect steel purity and quality
reach 10%. (slagging). Slagging losses could be reduced with input
materials that better matched the desired output.
66
180

140
120
100
AVAILABLE STEEL SCRAP
80
60
40
Exhibit 2.7
20 Almost 150 million tonnes of steel
0
2020 2030
are lost2040annually 2050
LOSSES IN STEEL PRODUCTION

Mt STEEL 2015

146 On average, 4-5% of steel is lost in the remelting
process as slag when impurities and alloying
elements are removed
22
NEW SCRAP LOST
Large losses at fabrication and forming
means up to 27% of steel does not reach
products
70-90% of resulting scrap is collected
48
OLD SCRAP LOST

15% of end-of-life products are not collected for recycling
– rising to 50% or more in some product categories
69
OBSOLETE STOCK
1-10% of end-of-life structures are
inaccessible due to corrosion or permanent
loss (e.g. underground structures)
6
MILLION TONNES
OF STEEL LOST
MAXIMUM COPPER CONTENT BY STEEL PRODUCT CATEGORY

Wt% COPPER IN STEEL
67
REBAR 0.40
146 On average, 4-5% of steel is lost in the remelting
process as slag when impurities and alloying
elements are removed
22
DOWNCYCLING TONEW
CONSTRUCTION
SCRAP LOST STEEL
Large losses at fabrication and forming
Steel comes in many different
means up to 27% ofgrades
steel does and qualities,
not reach show below, there are limits to downcycling as a strategy
products
and many products require precise control over the purity in a world with higher shares of secondary steel. Eventu-
70-90% of resulting scrap is collected
of the inputs and other elements of the production pro- ally, high shares can only be achieved if secondary steel
48
cess. This can be a challenge for secondary steelmaking. production can also produce other, high-quality products,
Scrap with very different content often is mixed together meeting the needs of all sectors of the economy, inclu-
in today’s supply chains, making it harder to precisely con- ding more demanding applications.
trol the inputs to production. This results in downcycling of
steel.22 For example, by one estimate, only around 8% of There is no fundamental reason that this could not be
the steel originally used in cars maintains sufficient quality achieved. There are several secondary steelmakers that
to be used for the same purpose again.23 already manage to produce very high-quality and even speci-
OLD SCRAP LOST
alty steels from scrap today. This requires a tightly controlled
15% of end-of-life products are not collected for recycling
Today, this is handled by using secondary steel prima- supply of scrap – keeping different qualities apart, matching
– rising to 50% or more in some product categories
rily for relatively basic construction steels (rebar, but also input content to product requirements, and removing unde-
structural elements). These can tolerate both a less pre- 69 elements – as well as advanced metallurgy. Replica-
sired
cise composition and a higher share of ‘tramp elements’ ting this across the EU and global economy would require a
– i.e., substances that are not desired. However, as we much more sophisticated handling of end-of-life steel.
OBSOLETE STOCK
1-10% of end-of-life structures are
inaccessible due to corrosion or permanent
loss (e.g. underground structures) Exhibit 2.8
6
Copper levels in steelMILLION
scrap TONNES already exceed
OF STEEL LOST
tolerances for many important products
Wt% COPPER IN STEEL
REBAR 0.40
STRUCTURAL 0.12
COMMERCIAL 0.10
GLOBAL PRODUCT MIX 0.08
FINE WIRE 0.07
DRAWING 0.06
CURRENT LEVEL OECD SCRAP
(0.2-0.25%)
DEEP DRAWING 0.04
SOURCE : MATERIAL ECONOMICS ANALYSIS; DAEHN ET AL. (2017) . 28
CO2 EMISSIONS FROM GLOBAL STEEL PRODUCTION 68

Mt CO2 PER YEAR, 2015-2100
COPPER CONTAMINATION – A POTENTIAL MAJOR CONSTRAINT ON FUTURE SECONDARY STEEL

A particularly important aspect of downgrading is the Today, the copper problem is managed in part
contamination of steel with copper. Even low levels of through dilution, either with scrap containing lower le-
copper drastically affect the quality of steel. At a share of vels of copper, or with fresh primary steel. Downcycling
0.15% or less, steel becomes unusable for important pro- also helps manage the problem: copper contamination
duct groups (Exhibit 2.8).24 The copper content of OECD is a major reason why secondary steel tends to be
scrap is around 0.2–0.25%, a level that exceeds the reused for construction steels, which can tolerate higher
quirements for many product categories except rebar.25 levels.
Moreover, unlike many other elements, copper cannot
be removed from the steel.26 In effect, the addition of Nonetheless, this creates major challenges as
copper results in the permanent and long-term downgra- copper continues to accumulate, and as the share of
ding of the steel stock. The addition of copper and other scrap-based production increases. As we show below,
elements also increases the energy required to produce continuing today’s practice would put a major obstacle
secondary steel, in some cases by as much as 40%.27 in the way of a more circular steel economy, with signifi-
cant increases in CO2 emissions as a result.
Copper is added to steel primarily at the point of re-
cycling, as products containing both steel and copper
are dismantled.29 Examples include components such as
electrical motors or cables with magnetic parts, or entire
products with high joint iron and copper content. End-of-
life vehicles are a major source, as cars can have a cop-
per content of 0.5% or more and often are shredded after
at most basic dismantling.30 Even to make rebar, fresh
primary steel or less contaminated scrap must therefore
be added. With more advanced handling of cars, such
as automated or manual dismantling, the copper mixing
can be reduced.
69
6
MILLION TONNES
OF STEEL LOST
2.4 A CIRCULAR SCENARIO FOR STEEL IN 2050

Wt% COPPER IN STEEL
To explore the benefits of a more circular steel sys- As explored elsewhere in this report, it is also pos-
tem, we constructed a scenario for the steel sector in EXHIBIT 2.9
sible to implement circular economy principles at the
which all three issues discussed above are addressed: product level – through longer lifetimes, more intensive
REBAR 0.40
use, light-weighting, and other strategies. In a ‘materials
1. Reduce the losses of steel;
and product circularity’ scenario that captures some of
2. Enable the use of secondary steel across a wider the opportunities in transport and buildings, the emis-
STRUCTURAL
range of product groups globally; and 0.12 sions reductions are still larger: as much as 52% of the
3. Reduce and eventually eliminate the problem of cop- emissions in 2100 are eliminated, and emissions fall in
per COMMERCIAL
contamination. absolute terms compared with current levels.
0.10
Globally, we find that this ‘materials circularity’ scena- The materials circularity scenario also sees a diffe-
rio cuts CO emissions significantly (Exhibit 2.9): by 590 rent profile of steel production vs. the baseline. Instead
GLOBAL PRODUCT2 MIX 0.08 of rising to 2.2 Gt per year, primary production is never
Mt CO2 per year (20%) by 2050, and as much as 1.1
Gt CO2 per year (29%) towards the end of the century. higher than 1.6 Gt. Secondary steel takes a majority role
By anyFINE
measure, by mid-century, and nearly all growth in steel demand
WIRE demand-side measures 0.07 for steel are a
major contributor to the objective of making materials can thus be met through recycling.
use consistent with climate objectives.
DRAWING 0.06
DEEP DRAWING 0.04 2.9
Exhibit (0.2-0.25%)
A circular scenario with reduced losses and technology

improvements has large potential to reduce emissions
CO2 EMISSIONS FROM GLOBAL STEEL PRODUCTION EXHIBIT 2.10
BASELINE SCENARIO
5 000
MATERIALS CIRCULARITY
4 500
MATERIALS AND PRODUCT
4 000 CIRCULARITY
-29%
3 500
3 000
-20%
-52%
2 500
500
1500
1 000
500
0
2020 2030 2040 2050 2060 2070 2080 2090 2100
EXHIBIT 2.10
70
GLOBAL STEEL PRODUCTION BY ROUTE

(0.2-0.25%)
DEEP DRAWING 0.04
CO2 EMISSIONS FROM GLOBAL STEEL PRODUCTION

This scenario implies a major reconfiguration of glo- Another major reconfiguration is in regional pat-
bal steel production, from an overwhelming emphasis terns of steel production and trade. The above scenario
on primary production, and towards a circular steel sys- depends on reducing the contamination of steel with
BASELINE SCENARIO
5 000 tem. In fact, if the potential for circularity is realised, the- copper. With current practice, this would require signifi-
re might be no need to add further primary steelmaking cant scale-up of tradeMATERIALS
betweenCIRCULARITY
regions. For dilution to
4 500
capacity beyond the 2020s, except to replace obsolete be feasible, steel scrap with high
MATERIALS ANDcopper
PRODUCTcontent would
4 000 plants (there already is very large overcapacity in steel- need to be mixed with CIRCULARITY
new primary steel produced in
making).31 This is a major break with the past. Even in countries-29%
with fast-growing demand. However, this is not
3 500 the last decades, two-thirds of new global demand has a viable approach as the problem grows, as perfectly
been met by building new capacity for primary steelma- mixing all available scrap and primary steel globally to
3 000
king (notably, as demand has -20%grown at unprecedented keep average copper content down is unrealistic. At a
2 500 pace in China). 32
By contrast, in a circular scenario, minimum, it -52%
is at odds with historical patterns for steel,
the analysis suggest that all the blast furnaces required where most industrialised regions have aimed to meet
500 globally could be built by the 2020s. most of their steel needs via domestic production.
1500
1 000
500
0
2020 2030 2040 2050 2060 2070 Exhibit
2080 2.10
2090 2100
A circular scenario has large impact on primary

steel production globally
GLOBAL STEEL PRODUCTION BY ROUTE

4 000
3 500
SECONDARY PRODUCTION
3 000 Circular scenario
2 500
PRIMARY PRODUCTION
2000 Basline scenario
1500
47%
PRIMARY PRODUCTION
500
0
2020 2030 2040 2050 2060 2070 2080 2090 2100
EUROPEAN STEEL PRODUCTION BY ROUTE

71
1 000
500
0 The Circular Economy – a Powerful Force for Climate Mitigation / Steel

2020 2030 2040 2050 2060 2070 2080 2090 2100
EU CO 2 EMISSIONS AND PRODUCTION

Turning to the EU, even more is at stake by improving groups. Instead, some of the high-copper scrap would
circularity (Exhibit 2.11). As noted, the region is transition need to be exported, and ultimately diluted with pri-
to a situation where available scrap could meet a much mary steel in other markets. Meanwhile, primary pro-
higher share of demand. The scenario below shows the duction of around 50 Mt would be required to dilute
production profile resulting if the EU were to use all remaining scrap in the EU.
domestically available GLOBAL STEEL
scrap PRODUCTION
to meet its ownBYdemand.
ROUTE
In principle, primaryMtsteel
STEELproduction
PER YEAR, 2015-2100
could then decli- This development would also have a major impact on
ne steeply, falling to just 15% of current levels by the emissions within the EU. In a reference case, emissions
2050s. However, this would require that the circularity remain at 104 Mt CO2 per year in 2050, as primary ste-
4 000
levers be successfully implemented, so that nearly all el production continues to serve 30% of EU demand. In
available scrap is utilised,
3 500 high-quality steel production a circular scenario, by contrast, emissions could fall by
from scrap is feasible, and copper contamination is re- half, to 47 Mt per year.33 SECONDARY PRODUCTION
3 000
solved. If current practices were maintained instead, the Circular scenario
EU could not use its own scrap to meet demand, as it This depicts the potential to meet demand using
2 500
would not hold the quality required for major product scrap-based steelmaking. The actual production
PRIMARY profile
PRODUCTION
2000 Basline scenario
1500
47%
PRIMARY PRODUCTION
500 Exhibit 2.11
Improved circularity
0
2020 2030 2040could
2050 significantly
2060 2070 2080 reduce
2090 2100 the need for
primary steel production to serve EU demand
EUROPEAN STEEL PRODUCTION BY ROUTE

180
160
140
120
100 SECONDARY PRODUCTION

Circular scenario
80
60 PRIMARY PRODUCTION
Circular scenario without
40
managed copper levels
20
PRIMARY PRODUCTION
0 Circular scenario
2020 2030 2040 2050
72
of the future EU steel industry is a much more
complex question. For a start, it would of cour-
se be possible to produce steel for export, in
which case additional primary production would
be required. As another option, it would be pos-
sible to export ever larger amounts of scrap (al-
ready, some 18% of scrap is exported, as noted
above), and continue primary steel production to
meet some share of domestic demand. Either way,
however, stabilisation of the EU steel stock will likely
bring significant structural change to the industry,
and a much smaller future role for blast furnaces.
Nonetheless, the analysis shows that a nearly

fully circular steel economy is possible within the
EU, with large resulting reductions in CO2. This
alone is a remarkable prospect. To achieve that,
a number of enablers need to be put in place, as
we discuss below.
A CIRCULAR SCENARIO CAN SUPPORT

SUPPLY-SIDE ABATEMENT
For clarity, the above numbers compare emis-
sions reductions when primary steel is still lar-
gely produced through the BOF route. However,
a more circular scenario can also support efforts
to decarbonise primary steel production.
Reducing emissions from primary steel produc-

tion faces significant challenges. First, as noted
in section 2.2, though there are established tech-
nologies that could halve emissions, they are a
long way from being deployed at scale, while truly
deep cuts in emissions would depend on rapid
deployment of breakthrough process innovations
that are still at the research and development sta-
ge. This creates a significant and uncertain inn-
ovation challenge – and it would also require ex-
isting steelmaking facilities to be replaced before
the end of their useful life, at a high capital cost.
In that context, demand-side measures are not

only valuable on their own right, but also to buy
time and take the pressure off – much like ener-
gy efficiency does on low-carbon energy supply.
Circularity is thus an integral part of any deep
mitigation scenario. As noted, in the long run, en-
suring effective recycling globally could avoid the
need to produce – and decarbonise – nearly 600
Mt of primary steel production per year, rising to
1,000 Mt with products circularity.
73
2.5 HOW TO GET THERE – MEASURES

TO PROMOTE CIRCULARITY IN STEEL
The different circularity strategies are closely interrela- First, systems to collect end-of-use products need to
ted and would need to be pursued simultaneously. Cop- be set up more widely and expand their capacity and
per contamination becomes a much bigger concern as accuracy as volumes of steel scrap grow. New initiatives
secondary steel is used for a wider range of purposes, and incentives will be needed to encourage consumers
including the most demanding uses that require very high to recycle their products and to ensure that iron and steel
quality. Similarly, unless copper is addressed, measures products are separated when buildings are demolished.
to reduce losses or prevent downcycling are limited in
how much they can displace primary production. As no- The forming of new scrap could be reduced with clo-
ted earlier, without solving the copper problem, the world ser-to-shape semi-finished products and with designs
as a whole would need to produce additional, “excess” that consider the production process and aim to mini-
primary steel just to dilute the contaminated steel stock. mise new scrap. New technologies that leave much less
By 2050, primary steel demand for that purpose alone production scrap, such as 3-D printing and powder me-
could amount to as much as 100–150 Mt per year. In tallurgy, could be adopted far more widely.34 In addition,
fact, unless the copper problem is solved, other circula- the quality of the collected new scrap could be improved
rity measures such as increased collection and reduced through increased sorting. The advantage of new scrap
losses rapidly run into a limit. over post-consumer scrap is known alloy content and
clean streams of carbon steel. Efforts should be made to
MEASURES TO REDUCE LOSSES keep alloys separated to be able to utilise the properties
of the scrap after remelting.
To arrive at the circular scenario, as little steel as pos-
sible can be lost in production and collection. To minimise Reducing remelting losses will be challenging as more
losses, measures need to be taken in three main areas. scrap enters the system. One way of reducing losses is
to limit the mixing of alloys prior to melting to minimise
impurities, as the process for removing undesired ele-
ments also leads to higher losses of iron. Better sorting
can also reduce energy requirements of the EAF process
by as much as 30%.35 With a better alloy sorting sys-
tem, alloying elements would not have to be removed,
and instead their properties could be valuable for a new
batch of steel being produced.
74
AVOIDING DOWNGRADING THROUGH A MORE • Product design for reduced contamination: The
design of products can also improve the sorting process.
SOPHISTICATED MARKET FOR STEEL SCRAP Design principles for recycling and for disassembly could
facilitate the removal of copper components by make
Alloy-to-alloy sorting is another key component to av-
them easier to see and to access and remove. Material
oid downgrading. If the contents of steel used for secon-
substitution is sometimes an option, such as replacing
dary production are known, steel scrap can be mixed
copper cables and wires with optic fibre or aluminium
much more closely to correspond to end products. The
equivalents.
technologies required to achieve this are under rapid
development. For example, laser-induced breakdown
• Metallurgy to increase copper tolerance: Produc-
spectroscopy technologies are improving and falling in
tion processes can be designed to be more tolerant to
cost fast, making it possible to quickly determine the con-
copper by avoiding the temperature interval where cop-
tent of alloys. When the content is known, a much more
per causes problems.36 Although not in itself a long-term
differentiated marketplace can be created so that steel
solution, this mitigates the problem.
manufacturers can specify and obtain the scrap required
also for exacting products.
• Separation of copper from steel: There currently
is no commercially viable method for removing copper
FOUR WAYS OUT OF COPPER CONTAMINATION from steel once it has been added. Some assessments
Copper’s effect on steel has been known for a long time, have been pessimistic that this will ever be viable.37 No-
but the problem has so far been relatively easy to handle, netheless, some research is ongoing into methods such
because secondary steel demand was limited. Looking as sulphide slagging, vacuum distillation and the use of
ahead, four key strategies need to be implemented: O2/CL2 gas.38
• Improved separation at end of life: The first step What will it take for these measures to take root? Ar-
is to avoid adding high-copper scrap to otherwise clean guably, current markets are poorly equipped to really ac-
flows, as is often done today to dispose of flows such count for the long-term impact of current practices on the
as copper-alloyed steel or some vehicle scrap. Beyond long-term quality of the global steel stock. It may there-
this, it will be necessary to increase the separation of fore be necessary to consider regulation as the route to
copper and steel in the recycling process. This already address copper pollution before it becomes a significant
happens to some extent, but practices vary widely, and problem for future steel recycling.
the extent of sorting fluctuates with the copper price,
since removing copper can be costly, manual work.
To avoid the cost of manual labour, more technologies
for automated sorting are being developed. More clo-
sed-loop recycling would also be necessary to keep
some scrap flows very pure and enable the use of scrap
in especially copper-sensitive applications.
75
The Circular Economy – a Powerful Force for Climate Mitigation / Plastics
3. plastics
finding a place within a low-carbon economy
Plastics are light, versatile, and cheap 60% of plastics demand by 2050, cutting
materials with a wide range of uses, and in CO2 emissions by half. To make this eco-
the EU we now use 100 kg of plastics per nomically attractive, the key is to focus on
person and year. Yet the use of plastics also systems that enable high quality recycling
leads to emissions of carbon dioxide (CO2): and preserve the value of plastics upon
both in production and because fossil car- recycling. This requires a large-scale and
bon is a major building block of all plastics. regionally integrated secondary plastics in-
On current course, emissions would nearly dustry, technology development, and above
double by 2050 in a baseline scenario with all changes to product design: our use of
today’s recycling rates, to more than 200 plastics must have recycling in mind from
million tonnes of CO2 per year – far excee- first use.
ding levels compatible with the EU’s com-
mitments under the Paris Agreement. Given This amounts to an ambitious transfor-
their ubiquitous role in our society, finding mation. Achieving circular plastics use will
a way to make plastics compatible with a require integration along value chains, and
low-carbon economy is an urgent agenda. concerted effort to overcome a range of bar-
riers. Yet much can be achieved by focus-
Recycling can offer a large part of the sing on what matters most: the five largest
answer to this. Most plastics are recycla- plastics types, and their uses in packaging,
ble, and recycling saves 90% of the CO2 automotive, and construction applications.
emissions arising from new production. In By starting now, we could supply more than
a detailed assessment of plastics types, half our plastics needs through recycling by
flows, and uses we find that a combina- 2050.
tion of re-use and recycling could provide
76 76
More than half of plastics
needs could be supplied through
recycling by 2050.
77
3.1 UNDERSTANDING FUTURE PLASTICS USE

Plastics are versatile, durable and low-cost materials total global oil use in a 2 °C scenario.1 By the end of the
that are used widely across modern economies. More century, plastics consumption would need to rise further,
than 30 types of plastics are in common use, with dif- to 1345 million tonnes per year – more than four times
ferent properties and applications in numerous sectors. today’s levels.1 By any measure, plastics therefore are a
Since the 1950s, global plastics production has risen major determinant of future fossil fuel use – and hence
dramatically, reaching more than 320 million tonnes in also of carbon dioxide (CO2) emissions, as we discuss
2015 and now growing by about 10 million tonnes per below.
year. In the EU, we now use 100 kg of plastics per person
per year. In Europe, a mature market, plastics demand actually
declined with the economic downturn, and although it
Global growth is set to continue. Plastics consumption has since rebounded, with a growth of about 1% per year,
increases as economies develop and begin using them production is no higher than in 2000. Still, the EU uses
in a range of products and consumer goods. Our ana- 49 million tonnes of plastics per year – about 100 kg per
lysis suggests that by 2050, global plastics production person. With a continued shift to lighter materials and
would double to over 800 million tonnes per year (Exhibit packaging, increased use of plastics in automotive appli-
3.1). If made with today’s fossil feedstock, global plas- cations, and increased demand for insulation to improve
tics production would then require 900 Mt of oil per year, buildings’ energy efficiency, we project that EU plastics
more than the European Union uses today, and 23% of demand will grow to 62 million tonnes per year by 2050.
Exhibit 3.1
Plastics demand will increase significantly
PLASTICS DEMAND BY REGION
Mt PER YEAR, 2015-2100
1400
MIDDLE EAST & AFRICA
1200 OECD ASIA
CHINA
1000 DEVELOPED ASIA
LATIN AMERICA
800 EUROPE
NORTH AMERICA
600
400
200
0
2020 2030 2040 2050 2060 2070 2080 2090 2100
78
3.2 CO 2 EMISSIONS FROM PLASTICS PRODUCTION AND USE

PLASTICS DEMAND BY REGION
Mt PER YEAR, 2015-2100
Finding a way to sustainable use of plastics has many plastics that are not recycled are increasingly incinerated
1400
dimensions. Plastics make up a large and growing share for energy generation, which leads to the immediate relea-
of solid waste around the world.2 Marine plastic pollu- se of all embedded carbon as CO2. MIDDLE EAST & AFRICA
1200
tion is a large and growing problem.3 Waste also decays OECD ASIA
into microplastics that pollute the entire food chain.4 In At the global level, current scenarios andCHINA roadmaps for a
addition,
1000 growing plastics use will be a major contri- 2°C economy tend to overlook plastics. Their productionASIA
DEVELOPED ma-
butor to CO2 emissions. kes up just 2% of CO2 emissions from industry and energy,
LATIN AMERICA
800 and embedded emissions often are not attributed to materi-
EUROPE
On average, each tonne of plastics produced results in als in emissions inventories, but just classified as emissions
NORTH AMERICA
2.5 tonnes
600 of CO2 emissions from the production process
from waste. However, if plastics demand continues to grow
alone.5 In addition, carbon embedded in the material cor- as projected, and a larger share of landfilling is replaced with
responds to another 2.7 tonnes of CO2. When this carbon incineration, cumulative CO2 emissions associated with plas-
400 depends on how plastics are treated at end
is released tics could grow very large. We calculate that the combined
of life. In landfills (or if plastics are released into the en- emissions from plastics production and embedded carbon
200 that process occurs slowly, as plastics take a
vironment), would be as much as 287 billion tonnes by 2100 (Exhibit
very long time to break down. However, in the EU, where 3.2). This corresponds to more than a third of the whole car-
0
policy-makers aim to eliminate the landfilling of plastics, bon budget for a 2°C economy. Addressing CO2 emissions
2020 2030 2040 2050 2060 from plastics 2080
2070 is crucial for a successful
2090 2100 low-carbon transition.
Exhibit 3.2
Plastics alone risk exceeding the carbon budget

quota available for industrial emissions
CO2 EMISSIONS, PLASTICS PRODUCTION CUMULATIVE CO2 EMISSIONS, PLASTICS PRODUCTION
Gt CO2 PER YEAR, 2015-2100, GLOBAL Gt CO2, 2015-2100, GLOBAL
287
5
226
0
2010 2020 2030 2040 2050 2060 2070 2080 2090 2100 BASELINE SCENARIO BASELINE + LOW
CARBON ENERGY
CURRENT BASELINE BASELINE + LOW
SCENARIO CARBON PROCESS
SOURCE : MATERIAL ECONOMICS MODELLING.
79 PRODUCTION
EMISSIONS INTENSITY OF PLASTICS
TONNES CO2 EMISSIONS PER TONNE PLASTICS PRODUCED
In the context of Europe’s climate targets, plastics must gy efficiency, and by adding external, low-carbon energy.
be addressed to reach commitments under the Paris However, this leaves embedded emissions in place, and
Agreement: today’s emissions of 132 Mt CO2 per year even if 57% of production emissions were eliminated, the
otherwise risk growing to 233 Mt, an increase of 76% (Ex- total emissions are 3.7 t CO2 per tonne plastics. Mechani-
hibit 3.3). Today’s emissions are dominated by production cal recycling – the cleaning, re-melting, and upgrading of
emissions, and end-of-life treatment adds little to this (ar- used plastics materials – produces less than 20% of the
guably, the net emissions from incinerating plastics are CO2 emissions associated with making new plastics – even
low when the alternative would be to burn other fossil fuels if the transportation, heat and electricity used are not yet
with a similar CO2 intensity). However, this picture changes decarbonised.7 However, when recycling results in downg-
by 2050. First, demand grows, adding 34 Mt CO2. More rading of plastics quality, the net emissions are higher, as
than that, the emissions from end-of-life treatment increase it may not replace new primary plastics on a one-to-one
by 90 Mt CO2, even if today’s recycling rates are maintai- basis. Even so, total emissions of 1.4 t CO2 per tonne plastic
ned. This is both because the current trend of increased are far lower than those from new production. Moreover, the
incineration would continue, and because, as the energy emissions from mechanical recycling can be all but elimi-
sector is decarbonised, instead of replacing fossil fuels, nated – provided the transportation and energy inputs used
plastics burnt to produce energy would replace cleaner are low-carbon, and high-quality plastics are produced, ca-
alternatives, increasing emissions. Even with a 15% re- pable of displacing new plastics use on a one-to-one basis.
duction in the emissions from plastics production, CO2
emissions from plastics increase to 233 Mt CO2 per year, For plastics that cannot be mechanically recycled, the-
an amount exceeding even the total emissions implied by re are a variety of chemical recycling approaches, whe-
the Paris Agreement commitments.6 reby plastics are broken down into smaller constituents
that can be reassembled to new materials, chemicals, or
There are a variety of strategies that can reduce CO2 fuels. Chemical recycling involves some loss of carbon
emissions from plastics (Exhibit 3.4). Emissions from as CO2, but the CO2 savings can be as much as 75% with
production can be reduced both by improving ener- technologies now under development.
Exhibit 3.3
Net CO 2 emissions from plastics in the EU

could grow by 76% by 2050
EMISSIONS FROM PLASTICS WITHIN THE EU
Mt CO2 PER YEAR 24
90
+76%
34
233
132
2017 DEMAND GROWTH INCREASED EMISSIONS IMPROVED PRODUCTION 2050

AT END OF LIFE
80
WITH CURRENT DEVELOPMENT, CO2 EMISSIONS FROM RECYCLING OFFERS ONE OF THE FEW WAYS TO
3
0
2010 2020 2030 2040 2050 2060 2070 Exhibit
2080 3.4
2090 2100 BASELINE SCENARIO BASELINE + LOW
CARBON ENERGY
Plastics
CURRENT recycling
BASELINE hasBASELINE
significantly
+ LOW lower CO 2 emissions
than other options for plastics production
SCENARIO CARBON PROCESS
EMISSIONS INTENSITY OF PLASTICS PRODUCTION

TONNES CO2 EMISSIONS PER TONNE PLASTICS PRODUCED
PRODUCTION EMBEDDED
Primary plastics production leads to large emissions

PRIMARY PLASTICS, 2017 5.1
as well as embedded carbon in the material
Improving production efficiency reduces production

PRIMARY PLASTICS, 2050 4.8
emissions, but does not address embedded emissions
PRIMARY PLASTICS WITH

3.7 Using renewable energy inputs can cut production
RENEWABLE ENERGY emissions but does not address embedded emissions
MECHANICAL RECYCLING, The recycling process has emissions of ~0.4 tCO2 / t plastics
LOW QUALITY 2017 1.4 Low-quality recycling may not lead to full replacement
of primary plastics
MECHANICAL RECYCLING, High-quality recycling in a largely decarbonised

HIGH QUALITY 2050 0.1
energy system can remove most emissions
Chemical recycling results in some CO2 emissions,

CHEMICAL RECYCLING 1.0 but eliminates embedded emissions from new
fossil feedstock
SOURCE : MATERIAL ECONOMICS ANALYSIS BASED ON MULTIPLE SOURCES. 8
81
ALUMINIUM 66
800
BORTTAGEN CEMENT 254
3.3 REALISING THE POTENTIAL FOR

CIRCULARITY
307
IN THE PLASTICS SECTOR
STEEL
300
The good news is that there is significant untapped (40%), buildings and construction (20%), automobiles
potential to recirculate plastics. More than 30 types of (9%) and electronics (6%). In addition, five big plastics ty-
plastics are in common use, with different properties and pes account for more than 70% of use, and all can be re-
applications 288 sectors. However, much of the
PLASTICS in numerous cycled mechanically.9 That means that although the plas-
volume of plastics can be addressed by focussing on tics industry is very complex, a circularity strategy that
the largest flows (Exhibit 3.5). Four main value chains started with just those four value chains and five plastics
account for three-quarters of EU plastics use: packaging
BAT ASSUMING CURRENT types could address
2° CARBON 60% of total plastics
BUDGET FOR volumes.
CARBON BUDGET
EU LEVEL CONSUMPTION INDUSTRY AND ENERGY AVAILABLE FOR
GLOBALLY MATERIALS
Exhibit 3.5
Five plastic types in four value chains represent 60% of demand
SHARE OF EUROPEAN PLASTICS DEMAND (49MT)
%, 2015
60% OF PLASTICS DEMAND
PACKAGING BUILDING & AUTOMOTIVE ELECTRONICS OTHERS

CONSTRUCTION
PE
(POLYETHYLENE)
PP
(POLYPROPYLENE)
PS
(POLYSTYRENE)
PVC
(POLYVINYL
CHLORIDE)
PET
(POLYETHYLENE
TEREPHTHALATE)
OTHER
PLASTICS
40% 20% 9% 6% 26%
SOURCE : PLASTICS EUROPE (2018B) . 10
PLASTIC VOLUMES IN EUROPE, 2015

Mt PER YEAR
82
Export in products
Stock build-up
(POLYVINYL
CHLORIDE)
PET
(POLYETHYLENE
TEREPHTHALATE)
OTHER
Exhibit 3.6
PLASTICS
Current recycled volumes are ~10% of demand – far lower
than
40%
the ~30% cited
20%
in official
9%
statistics
6% 26%
PLASTIC VOLUMES IN EUROPE, 2015

Mt PER YEAR
Export in products
Stock build-up
Misclassified waste
Etc
19
58 4 -90%
49
10
30 26 7
4
9 5
PLASTICS NET DEMAND USE BALANCE END OF LIFE UNSORTED COLLECTED LANDFILL ENERGY COLLECTED PROCESS RECYCLING
PRODUCTION EXPORT AND STOCK PLASTICS PLASTICS IN PLASTIC RECOVERY FOR LOSSES IN
BUILD-UP HOUSHOLD WASTE RECYCLING RECYCLING
WASTE
SOURCE : PLASTICS EUROPE (2018B) , DELOITTE AND PLASTICS RECYCLERS EUROPE (2015) . 14
Even though most plastics thus are recyclable, actual From a technical standpoint, three main factors inhibit the
effective plastics recycling rates in the EU are very low. It production of high-quality secondary plastics:
is often said that more than 30% of plastics are recycled • Mixed and contaminated flows. High-quality re-
in Europe11, but that is an FROMincomplete figure (Exhibit 3.6). cyclingTO requires that plastics are separated into streams
It refers to volumes collected for recycling, rather than of single plastics types, separated from other matter. This
actual recycled volumes, and collected volumes as a is hindered by today’s product design and collection sys-
1. share of identified plastics waste VIRGIN PLASTICS carry cost for
VIRGIN PLASTICS do notrather
carry than total plastics tems, which
consumed. Totalcost reported plastics waste in 2015 was embodied
of externalities 30 carbon (othermix or fuse different types of plastics, add
RAW MATERIAL other materials such as paper or metal, require costly
million tonnes, but as noted above, actual demand was externalities likely via regulation)
sorting, and may make separation impossible.
49 million tonnes. One reason is misclassification; for
instance, in Sweden, a study found that only 50% of plas- • Additives. Plastics often contain additives, from co-
tics waste, identified DESIGN does
PRODUCTthrough a not
detailed mapping, was MATERIALSlourants,
AND DESIGNtochoices
stabilisers,
to to flame retardants. These are dif-
2. ficult to trace or to remove, and can contaminate plastics
reported as waste bearincosts for downstream
official statistics. Another is that
12 make reuse and recycling the
PRODUCT
externalities
some of the plastics usedand arecosts
added to the stock (e.g.,intended destination at end of life
or make them unsafe or impossible to use in new pro-
in buildings or long-lived products). In addition, because ducts.
plastics streams for recycling are highly contaminated • Contamination. Plastics may also be contaminated
and mixed, onlyPOLICY
some AND60 percent
SYSTEMS focusof
on the volume sent
FOCUS ONby ENABLING raw
the substances they held, leading to stains or smells
3.
for recycling is actually reprocessed
collection into new materials.materials
volumes – similar 13
flows for secondary
that lower quality, or adding harmful chemicals that pre-
COLLECTION to otherplastics
‘waste’ flows
Actual EU secondary production as a share materials
of production
vent recycling entirely. Some contaminants, such as med-
demand is therefore closer to 10% than 30%. ical waste, require plastics to be incinerated.
END-OF-LIFE TREATMENT and PRODUCTS DESIGNED for disassembly

4. dismantling without focus on and dismantled to retain secondary
END-OF-LIFE retaining material value
83 value
material
The fundamental reasons are not just technical, for secondary materials production. For example, the sh-
however, but arise from a combination of policy, market, redding of cars results in plastics that are too mixed and
and industry features throughout the whole value chain of contaminated to recycle (and in some cases, even to inci-
plastics use (Exhibit 3.7). We start from a position where nerate safely). Similarly, when buildings are demolished,
plastics recycling is an extension of end-of-life and waste plastics often are not recycled (with the exception of PVC
policy, and need to move towards one where recycling is recycling in some EU countries).
the intended destination for plastics used in products.15
5. Recycling – create investment certainty and higher
1. Raw materials production – get the prices right. yields: As noted above, more than 40% of the plastics
EU CO2 prices for plastics production are significantly collected for recycling in the EU are never actually tur-
lower than either estimates of the social cost of carbon, ned into secondary materials. The materials may not be
or the implicit CO2 prices in other climate policy (such easily recyclable, or they may be degraded or contami-
as subsidies for renewable energy). There are many re- nated, making them unusable – or only usable for very
asons for this, including the disadvantages of imposing low-value secondary materials. As important is to enable
costs on EU industry without corresponding measures in investment in the capacity for high-quality recycling. This
other geographies. The result is that primary plastics are is held back by uncertain ownership of, and access to,
cheaper than they would be if the negative environmental end-of-use plastics in many EU countries.17 Large scale
consequences of their use were accounted for. For il- is crucial to cost efficiency in recycling. Recycling pro-
lustration, with a CO2 price of 50 EUR per tonne CO2, the cesses are often locally managed or inconsistent across
full CO2 cost (included embedded emissions) would be different localities, making it difficult to achieve sufficient
20% of the price of primary plastics, a price increase that scale. Waste regulations sometimes inhibit trade in used
would significantly improve the economics of recycling. plastics across borders, preventing the emergence of a
regionally integrated market.
2. Product design – address externalities for recycling:
Many plastic items are designed in ways that make re- 6. Secondary materials market – create demand pull
cycling difficult or impossible. Different plastics may be and high quality: Due to the problems noted above, as
used and fused together; plastics may be dyed (black well as a lack of quality standards for secondary plas-
plastics are difficult to recycle, and colours reduce the qu- tics, most secondary plastics are perceived to be of low
ality and commercial value of recycled plastics); there may quality, with limited uses. They therefore often trade at a
also be additives that cannot be easily removed, as well discount, with average prices as low as 50% of corres-
as adhesives. All these factors reduce the value that can ponding primary plastics in some categories – a major
be recovered, but producers do not bear any of the cost. source of value destruction. Recycled plastics thus end
This is a major, unrecognised market failure, as there is no up in lower-value applications such as flower pots, traffic
realistic mechanism for a secondary materials industry to cones or garbage bags. The combination of high cost
coordinate with those upstream in the value chain to indu- and low quality also has resulted in limited demand for
ce the changes that would retain more material value.16 secondary plastics. There is a chicken-and-egg dynamic
at work, where a fragmented and small-scale recycling
3. Collection – target the right quantities. Plastics mana- industry cannot produce the consistent quality and vo-
gement policies today target collection rates, not actual lumes required for large-scale use, even as lack of de-
production of high-quality secondary materials. The resul- mand holds back the investment that would enable such
ting collection systems are far from optimised for the re- production in the first place.
cycling process. The result is additional costs of recycling,
with low outputs and low revenues for the recycling industry. The key lesson from this diagnosis is that barriers are
Future targets should instead focus on maximising the out- found throughout the entire value chain. The remedy the-
put of value retention – which requires a combination of refore must also address all of these, not just focus on
high collection rates, minimal loss of materials, and high-qu- more ambitious targets for collection. Plastics recycling
ality outputs that are real substitutes for primary plastics. needs to become a major secondary materials industry,
rather than an extension of waste handling systems. This
4. End-of-life treatment – prevent downgrading and also suggests the agenda for improvement: consciously
loss of materials. The dismantling of products at their designing products for recycling and thus enabling a lar-
end of life often takes little account of the implications ge-scale secondary materials industry.
84
19
58 4
49
10
30 26 7
4
9
FÄRGPALETT PLASTICS NET DEMAND USE BALANCE END OF LIFE UNSORTED COLLECTED LANDFILL ENERGY COLLECTED PROCESS
PRODUCTION EXPORT AND STOCK PLASTICS PLASTICS IN PLASTIC RECOVERY FOR LOSSES IN
Exhibit 3.7
BUILD-UP HOUSHOLD WASTE RECYCLING RECYCLING
WASTE
Improving plastics recycling will require transformation

across the entire value chain
FROM TO
HIBIT 3.7
1. VIRGIN PLASTICS carry cost for

VIRGIN PLASTICS do not carry
cost of externalities embodied carbon (other
RAW MATERIAL
externalities likely via regulation)
Ludde gör denna

direkt i InDesign?
PRODUCT DESIGN does not MATERIALS AND DESIGN choices to
2.
bear costs for downstream make reuse and recycling the
PRODUCT
externalities and costs intended destination at end of life
POLICY AND SYSTEMS focus on FOCUS ON ENABLING raw

3.
collection volumes – similar materials flows for secondary
COLLECTION to other ‘waste’ flows materials production
END-OF-LIFE TREATMENT and PRODUCTS DESIGNED for disassembly

4. dismantling without focus on and dismantled to retain secondary
END-OF-LIFE retaining material value material value
LOW-QUALITY inputs and LARGE-SCALE INDUSTRY with high-

5.
investment uncertainty results quality outputs with high retained
SECONDARY
in low yields and small-scale, value and the ability to replace
MATERIAL
fragmented industry primary materials one-to-one
PRODUCTION
6. RELIABLE PRODUCTS and demand-side

INCENTIVES focused on supply
MARKET FOR incentives create market certainty
into recycling process
RECYCLED and stimulate investment in capacity
HIBIT 3.8 MATERIAL
PLASTICS PRODUCTION EMISSIONS

Mt CO2, EUROPE, 2050
116
85
Addition
233 Substitut
VIRGIN PLASTICS carry cost for
VIRGIN PLASTICS do not carry The Circular Economy – a Powerful Force for Climate Mitigation / Plastics
cost of externalities embodied carbon (other
RIAL
externalities likely via regulation)
3.4 A CIRCULAR SCENARIO FOR EU PLASTICS USE IN 2050
To estimate
PRODUCT DESIGNthe does
CO2not emission reduction potential
MATERIALSfromAND DESIGN Increase
• choices to the collection rate for mechanical recycling
a more circular plastics
bear costs for downstream system, we constructed
makea scenario
reuse and to 73%.
recycling The
the rate varies by value chain and plastics type,
T but is underpinned by collection up to 85% of the five lar-
for 2050 that captures
externalities and costs the potential for intended destination at end of life
plastics recycling
much more fully. As noted above, in the baseline scenario, gest plastic types. Collection rate remains as low as 30%
production of plastics for EU use and associated end-of-life for other, smaller-scale plastics, which although often very
treatment would produce 233 Mt of CO2 emissions in 2050. valuable, often also entails small volumes that make the
AND SYSTEMS focus on FOCUS ON ENABLING raw
This POLICY
includes emissions from production and disposal, and economics of recycling difficult. The non-collected portion
collection volumes – similar materials
assumes all plastics are incinerated, with no carbon capture flows for secondary
of 27% also includes a large share of thermosets, which
ON to other ‘waste’ flows materials
and storage (Exhibit 3.8). In a circular scenario, emissions are productioncan be chemically but not mechanically recycled.
reduced by nearly 50% by making reuse and recycling the • Significant yield improvements in sorting and re-
standard for end-of-life plastics. To reach climate objectives, cycling to 76%: The share of plastics that are collected
these emissions
END-OF-LIFE and to be reduced further,
would need
TREATMENT PRODUCTS DESIGNED for disassembly
through for recycling but not turned into secondary materials falls
dismantling
additional without
recycling, and dismantled
focus on with other materials,
substitution use of re- to retain
from secondary
more than 40% today to 24%. Combined with a 73%
IFE retaining
newable material
energy value
in production, material
replacement of fossil value
feedstock collection rate, this means that the output from re-use
with bio-based alternatives or chemicals synthesised from recycling is 56% of total end-of-life plastics volumes.
(non-fossil) CO2, and other production process innovation.
LOW-QUALITY inputs and
• Chemical recycling of 25% of remaining flows. This
LARGE-SCALE INDUSTRY with high-
investment
Exhibit 3.9 showsuncertainty results componentsquality
the different outputs withfocuses
of a circular on the remaining 44% of plastics that are difficult
high retained
RY
in lowfor
scenario yields and small-scale,
EU plastics, all of them contributingvalue and the ability to replace through mechanical recycling. Thus, 11% of
to significant to handle
AL total end-of-life plastics are chemically recycled.
fragmented
reuse industry
of end-of-life primary
plastics and substitution of primary materials one-to-one
plastics
ION
with secondary materials. Key features of this scenario include:
RELIABLE PRODUCTS and demand-side

INCENTIVES focused on supply
OR incentives create market certainty
into recycling process
ED and stimulate investment in capacity
AL
Exhibit 3.8
Implementing a circular plastics scenario

can reduce 2050 emissions by 50%
PLASTICS PRODUCTION EMISSIONS
Mt CO2, EUROPE, 2050
116
Additional recycling
233 Substitution with other materials
Renewable energy in production
Biobased or CO2 feedstock
Process innovation
117
2050 WITH 100% CIRCULAR ABATEMENT REMAINING ADDITIONAL REMAINING

ENERGY RECOVERY POTENTIAL 2050 EMISSIONS POTENTIAL ABATEMENT 2050 EMISSIONS
TREATMENT OF END OF USE PLASTICS, 2017 AND 2050 86

MILLION TONNES; % OF PLASTIC DEMAND
Renewable energy in production
Biobased or CO2 feedstock
Process innovation
117
2050 WITH 100%

ENERGY RECOVERY
CIRCULAR ABATEMENT
POTENTIAL
Exhibit 3.9
REMAINING
2050 EMISSIONS
ADDITIONAL
POTENTIAL ABATEMENT
REMAINING
2050 EMISSIONS
An ambitious circular scenario for plastics in 2050
TREATMENT OF END OF USE PLASTICS, 2017 AND 2050

MILLION TONNES; % OF PLASTIC DEMAND
TOTAL
49.0 61.6
11%
2%
WHAT IT WILL TAKE TO GET THERE
NOT COLLECTED
/ STOCK 47% 21% 85% of 5 most common plastics collected for
recycling with significant yield increases (90% yield)
Shift from landfill to energy recovery for non-recycled
Reduced stock build-up and increased collection for
15%
recycling
Reuse of:
8% of packaging, especially industrial
LANDFILL 20%
3% of B&C plastics, esp. PVC carpets, pipes etc.
4% of large automotive parts
46%
ENERGY RECOVERY 15%
YIELD LOSSES 8%
RECYCLED
10%
REUSE 0% 5%
2017 2050
PACKAGING BUILDING & CONSTRUCTION AUTOMOTIVE

19.5 24.5 9.8 12.3 4.5 5.7
5%
NOT COLLECTED 11%
21% 2%
18%
/ STOCK
16% 2%
2% 47%
LANDFILL 25% 15% 17% 32%
76%
12%
ENERGY RECOVERY 22% 18%

53% 30%
47%
YIELD LOSSES 13%
8% 5% 32%
RECYCLED 9% 11%
19%
3%
REUSE 0% 10% 0% 5% 4% 0% 6% 4%
2017 2050 2017 2050 2017 2050
SOURCE : DELOITTE AND PLASTICS RECYCLERS EUROPE (2015) , PLASTICS EUROPE (2018B) . 18
REDUCED RECYCLING COSTS HIGHER REVENUES PACKAGING EXAMPLE

87
924 926
Overall, this is a highly ambitious scenario. Realising the A deepdive into the economics of recycling gives an
circular scenario described above will require transforming the indication of how costs and revenues would have to shift.
system to address barriers across the value chain. This may Today, costs often exceed the available revenuesVOLUME from
seem daunting, but the prize would be worth it: in the circular often low-quality secondary plastics output. Exhibittonnes
3.10
scenario, the reuse
TODAYand recycling 55% of secondary
COLLECTION plastics cre- shows a case
CIRCULAR SYSTEM study example for plastic packaging. Our
ates revenues of90%27 value
billionlossEUR per year. Plastics could then
>80% of value still lost, analysis
50% of value retained that costs can be reduced by 16% by
suggests
move a good way “30%towards
collection” where Large
steelprocess
or aluminium are to-
losses in recycling creating cleaner
Higher recycling flows from materials choices and pro-
yields increase
day, with recycling
butdriven
just 16%by the value of struggle
Uphill secondary materials.
to increase of duct
volumedesign optimised for recycling; increasing the scale
recycling volume further – at a net cost High value retention provides
Rebound effects as downgraded revenue for investment in self-
plastics cannot replace virgin sustaining systems and scaling
plastics in many uses Recycling can replace virgin
Exhibit 3.10 production
Recycling is financially challenging today

but can be made economically attractive
REDUCED RECYCLING COSTS HIGHER REVENUES
924 926
-16%
COLLECTION 186
774
+71%
119
SORTING 191
147 PACKAGING EXAMPLE
540
LOGISTICS 15
10
337
RECYCLING 439
162
INVESTMENT 93 REVENUES
2017 2050 2017 2050
Cleaner flows from materials choices and Increased yields give larger revenues per tonne
product design optimised for recycling treated plastic (better technology, cleaner inflows)
Increased scale reduces unit cost Higher quality enables pricing closer to virgin
Specialisation and regional integration of markets materials (additional sorting, improved technology,
Technological improvement boosts efficiency but cleaner flows and optimised products)
requires higher investment Raised cost of virgin materials improves
Reduced risk from regulations and market willingness to pay for secondary products
uncertainty Better functioning markets reduce current
commercial risk to buyers
DEEP TRANSFORMATION THROUGH COORDINATED AND SUSTAINED ACTION ACROSS THE VALUE CHAIN
88
Other plastics Low quality increased recycling (21)
50Force for Climate Mitigation /

The Circular Economy – a Powerful Plastics
B&C Low quality increased recycling (-66)
of recycling; creating more specialised but also regional- Achieving this transformation holds the promise of an
ly integrated recycling; adopting new technologies; and industry driven by the intrinsic value of plastics.19 From a
40
reducing risks due to regulations and market uncertainty. climate perspective, adopting those measures could also
At thePackaging
same High time, revenues
quality can be
increased recycling increased by as much
(-96) make the difference between plastics recycling being a
as 71%, through increased yields, which give higher re- costly abatement option, or an economically attractive
venues per tonne of treated plastic (better technology, one. Today, recycling is expensive seen from a pure CO2
cleaner
Agricultureinflows); higherrecycling
High quality increased quality,
(-115)which enables pricing 30 perspective (although it can have other benefits). With the
closer to virgin materials (an
EEE High quality increased recycling (-138) average of 70% of virgin transformation described above, this changes by 2050.
prices in our scenario); higher costs for virgin materials, Exhibit 3.11 shows a particularly striking example, with
B&C High
which quality increased
improve buyers’ recycling (-138)
willingness to pay for recycled ma- today’s high abatement cost for packaging. By 2050, the
terials; and better-functioning markets that reduce current net CO2 abatement per tonne of recycled plastic goes up
20
commercial
Automotive risk to recycling
High quality increased buyers, thus incentivising the use of
(-149) (for the reasons described in section 3.2), while the net
secondary plastics and investments in recycling capacity cost of recycling pivots to a net benefit. The abatement
and innovation. cost is therefore negative, meaning that the transformed
Other plastics High quality increased recycling (-170)
recycling system can not only cut CO2, but also create a
Alongside all of these, strong technology development 10 net financial gain.
provides aPackaging
significant
Reuse boost.
(-179) Recycling is strongly synergis-
tic with digitalisation, which is rapidly reducing the cost of These cost estimates need careful interpretation. First,
EEE Reuse (-192)
marking ofB&C different materials and products, use of sensor
Reuse (-202) the negative cost does not mean that large profits will accrue
technology, and
Automotive automation
Reuse (-203) of dismantling and sorting – to recycling industries. Rather, the benefit will be distributed
all of which will be fundamental
-200 -150 to -100
achieve-50more cost-ef-
0 50 across
100 the
150 value
200 chain,
250 including
300 through changed prices
fective plastics recycling. This requires increased capex, for end-of-life plastics and for secondary plastics. Second,
but this is more than offset by the higher revenues ABATEMENT COST
availa- achieving this cost depends on the full transformation descri-
ble from higher-quality outputs. EUR/t CO 2
bed above. It is not something one actor can create on its own.
Exhibit 3.11
Reducing abatement cost requires higher net abatement,

improved yields and higher quality
ABATEMENT COST (PACKAGING EXAMPLE)
EUR PER TON ABATED CO2
257
43
-68
-101
2017 COST INCREASED ABATEMENT PER REDUCED COST OF RECYCLING, INCREASED REVENUES FROM
TONNE OF PLASTIC RECYCLED ESPECIALLY THROUGH HIGHER-QUALITY SECONDARY
INCREASED YIELDS MATERIALS
SOURCE : MATERIAL ECONOMICS ANALYSIS AS DESCRIBED IN TEXT.
89
Performing the above analysis for all segments and
for different qualities of plastics flows, it is possible to cre-
ate a cost curve for CO2 reductions through plastics re-
cycling (Exhibit 3.12). The horizontal axis shows the total
abatement potential, which sums to the 116 Mt CO2 per
year, described above. The costs are estimated similarly
to that described for the packaging example described
above. There are of course many uncertainties in project
technology developments, the feasible quality of secon-
dary plastics, and other factors leading to 2050. Also,
although we have attempted to account for the increased
cost of adapting products for recycling, the actual cost
will not be known until incentives to do so are created.
As with other areas where incentives have been lacking
to date, there is likely to be significant low-hanging fru-
it, but there will also be instances where adapting for
recyclability would create significant compromises with
function, or impose costs. For all these reasons, these
costs are exploratory and indicative. Nonetheless, the
analysis shows that plastics recycling can not only be
a substantial abatement opportunity in terms of volume,
but also has the potential to be economically attractive
compared to much else that needs to happen to put the
EU on a low-carbon pathway.
90
50 100 EMISSIONS
0
150 200 250 300 SAVINGS
Mt CO2 / YEAR
Cars - Lightweighting
-50 Buildings - Materials efficiency
Cement - Cement recycling
Exhibit 3.12 Steel - Reduce fabrication losses
-100 Plastics - Chemical recycling
With changes throughout the value chain, recycling could be Buildings - Reuse
a highly cost-effective form of Plastics abatement

CO 2- Increased by 2050
Other - Materials efficiency
recycling at current quality
Aluminium - avoid downgrading
Steel - Increase collection
PLASTIC COST CURVE 2050 POTENTIAL

EUR PER ABATED t CO2, EUROPE, 2050
ABATEMENT COST
EUR / t CO2
Agriculture Reuse
B&C Reuse
Other plastics High quality increased recycling
B&C High quality increased recycling
B&C Low quality increased recycling
Automotive High quality increased recycling
EEE High quality increased recycling
100 Other plastics Reuse
100
Agriculture High quality increased recycling
Packaging High quality increased recycling
Automotive Reuse 55 57
50 43
35
23
10 20 30 40 50 EMISSIONS
0
-3 60 70 80 90 100 110 120 SAVINGS
Mt CO2 / YEAR
-26 EEE Low quality
-50 increased recycling
-53 -49
-59 Automotive Low
-70 -62 quality increased recycling
-100 -93 Agriculture Low
-100 -100
quality increased recycling
Packaging Resuse
Chemical recycling
Packaging Low quality increased recycling
EEE Reuse
Other plastics Low quality increased recycling
NOTE : BUILDINGS AND CONSTRUCTION (B&C) , ELECTRICAL AND ELECTRONIC EQUIPMENT (EEE) .
91
COMPLEMENTING STRATEGIES TO FURTHER REDUCE EMISSIONS

Even ambitious recycling as described above would leave To further reduce emissions, a combination of additio-
half of emissions in place. These arise from a variety of sour- nal strategies will be necessary:
ces, which also suggest ways to reduce them (Exhibit 3.13). • Substitution with other materials. The scenario here
is based on continued growth in plastics use. However,
Even with ambitious mechanical and chemical re- there are other options, including substituting to other
cycling, a small share of plastics will be difficult to collect materials that either can be more easily recycled, or ones
and treat. There are also substantial yield losses from that are based on bio-feedstock.
plastics that are collected for recycling, but which do
not become secondary materials. In both cases, primary • Use of renewable energy in production. As noted
plastics must be produced to replace the material lost. above, adding external renewable energy to the plastics
In addition, the recycling processes also result in some production process can reduce emissions to some de-
CO2. Both losses and direct CO2 emissions are particu- gree. With further process innovation, it is possible that
larly large for chemical recycling, and an innovation pri- even more of the production emissions could be abated.
ority therefore must be to find more effective processes As noted, however, this does not address the fossil car-
with a higher yield, including by introducing external, re- bon in the feedstock.
newable energy to generate the heat necessary. • Non-fossil feedstock. Replacing fossil fuels with
other feedstocks could address a large share of both
There are also emissions from the production of pri- production and embedded emissions. Bio-based feed-
mary plastics required to add to the total stock of plas- stocks have been mooted as one alternative, although
tics in the economy. How much will be required is very the volumes of biomass required would be large.
uncertain – it is possible that the total stock will saturate, Another option would be to use either CO2 or car-
removing much of these emissions. Finally, we have al- bon-monoxide as feedstock.20 However, these would
lowed for some rebound effects: an increase in plastics have to come from non-fossil sources if total emissions
consumption due to the increased supply of some relative- are to be brought to net-zero. Using such routes also
ly lower-quality recycled materials. This can be reduced by would have large resource requirements, especially for
improving quality further, or if the amount of lower-quality renewable electricity.
plastics could be absorbed by existing demand pools.
92
PRODUCT DISPOSAL
Reward practices conducive to
END-OF-LIFE high value recycling
TREATMENT
STANDARDISATION
Quality standards for recyclates and
guarantees for recycled content SECONDARY
MATERIAL
Exhibit 3.13
PRODUCTION
PUBLIC PROCUREMENT
To create demand pull in nascent
Deeper transformation and innovation is required

markets
to reduce the remaining

REGULATION OF ADDITIVES
Limiting toxic substances (current) and use
MARKET FOR
RECYCLED
65% of emissions
DEMAND PULL
Quotas to create demand
of additives inimical to recycling (future) MATERIAL pull for nascent markets
REMAINING EMISSIONS 2050

Mt CO2
REBOUND
117 Recycled plastics of lower quality do not replace
the same volume of virgin plastics and may lead
to an increased demand of total plastics
11
PRODUCTION FOR STOCK BUILD
Around 10% of demand results in net stock
build-up, requiring additional primary production
22
REUSE AND RECYCLING EMISSIONS
117 80% from yield losses in chemical recycling, the
remainder from mechanical recycling
14
YIELD LOSSES
Yield losses in pre-treatment, recycling and
reuse process sent to energy recovery
Process improvements as well as product design and
choice of material important to maximize yields 42
AGED, DEGRADED, CONTAMINATED AND

NOT COLLECTED “COMMON PLASTICS”
Relatively small given option of chemical recycling
Remaining share is aged, contaminated, or not
captured by collection systems
THERMOSETS AND SMALL VOLUME
SPECIALTY PLASTICS 16
Challenging to achieve scale in small specialty
plastics streams
12
93
3.5 HOW TO GET THERE – MEASURES

TO PROMOTE PLASTICS RECYCLING
As noted above, current policies are focused mainly there is significant latent demand from companies that
on waste management, with increasing pressure on re- increasingly seek to use secondary materials, but with
cyclers to maximise collection volumes. The recent EU undeveloped markets, this cannot support the necessa-
Plastics Strategy sought to complement this, including ry investment in recycling infrastructure. The ‘pledging’
to introduce measures in other parts in the value chain. mechanism in the Plastics Strategy sought to address
this. If policy-makers discover that more far-reaching in-
We propose that the bedrock of a future sustainable tervention is warranted, options to evaluate could include
plastics system – and secondary plastics industry – must quotas for recycled plastics in selected product categories,
be to enable high value of secondary plastics. Exhibit financial instruments linked to the use of secondary ma-
3.14 illustrates examples of policy areas that could sup- terials, and market creation through public procurement.
port or accelerate a transition. The full change described
above will take time – possibly measured in decades – so Fourth, waste collection systems and practices for hand-
it is important to start on this path as soon as possible. ling end-of-life products need to be revisited to account
for their impact on high-value recycling. Today’s collec-
This report has not evaluated individual policy propo- tion systems are fragmented, with patchy coverage of
sals, and is in no position to make specific proposals for plastics streams and end-use segments. Much of infra-
policies. However, a number of high-level principles can structure is optimised to meet policy targets that are not
guide policy-makers as they approach this area. necessarily conducive to actual production of high-qua-
lity secondary materials. Another area to investigate are
First, changes must occur simultaneously across the the practices for dismantling end-of-life products, notably
entire value chain: primary production, product design, the demolition of buildings and shredding and mixing of
recycling infrastructure and business models, the market materials from end-of-life vehicles.
for secondary plastic, and legacy waste handling regu-
lations and systems. One route to this could be further Fifth, waste regulations may also need updating. Lar-
industry integration, perhaps of primary and secondary ge-scale use of secondary plastics requires that markets
materials production. However, it is also possible that a operate at a scale and to the standards of global value
range of barriers – rooted in the economics of incom- chains for raw materials. By contrast, current regulations
plete contracts and transaction costs – will continue to often treat recycling as an off-shoot of local, small-scale
prevent market actors from achieving this outcome. If so, waste handling. Key issues to consider include restric-
policy will have an important coordinating role. tions on trade across borders, lack of clarity about ow-
nership and access to plastics flows, and the resulting
Second, a starting point for policy must be to analyse investment uncertainty. As with other policy areas, chang-
and correct for current externalities. One of these is the es would need to be evaluated for their impact on other
lack of a price on emitting CO2, to reflect either the so- objectives, notably that of ensuring safe waste handling.
cial cost of carbon, or the prices that will eventually be
required across the economy to reach climate targets. Finally, governments need to kick-start innovation. Re-
More broadly, low taxes on resources and pollution relati- cycling is undergoing a quiet technology revolution. Tech-
ve to those on the inputs to recycling (notably, labour and nology is already improving fast, with recent plants sig-
transport fuels) reduce incentives for recycling. A second nificantly more capable than ones available only a few
is to recognise and address the heavy externalities that years ago. Digitisation is a major force to harness, with
product design choices impose on recycling. This is a applications across marking, remote sensors, real-term
new agenda, and there is an urgent need to evaluate tracking, and automation. In parallel, there is a need for
the costs and benefits of different options to address it materials development, notably to find long-term solu-
– whether the gradual introduction of product regulation, tions for plastics that are hard to recycle, such as engine-
voluntary agreements, standards, industry design proto- ering plastics and thermosets. Policy can support these
cols, financial incentives, or in other ways. agendas both through fundamental research, develop-
ment, and demonstration, and by supporting new tech-
Third, policy-makers need to consider the nature of bar- nology deployment – analogous to the approach taken
riers to demand for secondary plastics. Engagements to renewable energy technologies or other long-term key
with a range of companies for this study suggest that solutions to climate mitigation.
94
INCREASED YIELDS MATERIALS
Exhibit 3.14
Current policies cover only parts of the range of

barriers to high-value plastics recycling
EXAMPLES OF POLICIES – ALL WOULD NEED CAREFUL EVALUATION
CURRENT POLICY FOCUS NO / LITTLE POLICY FOCUS TODAY PARTIAL POLICY COVER TODAY
FEES
Correct for externalities for primary
production of plastics
RAW MATERIAL
PRODUCT DESIGN
Regulation and fees to correct
PRODUCER RESPONSIBILITY for externalities on secondary
Differentiation to reflect downstream materials
impacts on recyclability PRODUCT
COLLECTION TARGETS
Are the main denominator of
LANDFILL / ENERGY current policies
RECOVERY TAX / FEE
To boost value of recycling
COLLECTION
PRODUCT DISPOSAL
Reward practices conducive to
END-OF-LIFE high value recycling
TREATMENT
STANDARDISATION
Quality standards for recyclates and
guarantees for recycled content SECONDARY
MATERIAL
PUBLIC PROCUREMENT
PRODUCTION To create demand pull in nascent
markets
REGULATION OF ADDITIVES MARKET FOR DEMAND PULL

Limiting toxic substances (current) and use RECYCLED Quotas to create demand
of additives inimical to recycling (future) MATERIAL pull for nascent markets
REMAINING EMISSIONS 2050

Mt CO2
REBOUND
117 Recycled plastics of lower quality do not replace
the same volume of virgin plastics and may lead
95 to an increased demand of total plastics
11
PRODUCTION FOR STOCK BUILD
The Circular Economy – a Powerful Force for Climate Mitigation / Aluminium
4. ALUMINIUM
realising the potential for recycling
Aluminium is a heavily used material for share of aluminium demand. If this were to
which the potential for circularity is growing continue, there would eventually be ‘excess
fast. Even with rapid demand growth, the scrap’, even as demand for new aluminium
potential for aluminium recycling is increa- kept growing. A shift to electric vehicles will
sing: By 2050, the share of reused mate- exacerbate this problem by further redu-
rial could, in principle, double – even as cing demand for the cast aluminium com-
total aluminium use grows strongly. In the ponents that absorb most of the recycled
EU, there will be enough end-of-life alumi- aluminium today.
nium to serve more than half of demand by
mid-century. Moreover, the resource gains A more circular scenario for aluminium
from circularity are particularly large with requires reducing losses when products
aluminium, as remelting existing metal re- are first manufactured, and better separa-
quires just 5% of the energy of new produc- ting and sorting end-of-life aluminium. This
tion, sharply reducing CO2 emissions. would have significant climate benefits. Glo-
bally, CO2 emissions could be reduced by
Capturing this opportunity, however, re- 300 Mt per year in 2050, and in Europe, by
quires significant changes. A first step is close to 30 Mt. The latter represents 36%
to reduce losses of aluminium in each use of emissions from meeting EU aluminium
cycle, which now amount to 25–30%. This demand in a baseline scenario.
requires better end-of-life treatment and col-
lection for a range of products, as well as Overall, aluminium therefore is yet
product designs that facilitate the recovery another example of large CO2 benefits from
of aluminium. End-of-life aluminium will also enabling a more circular system. This also
need to be handled differently. Today much opens up new opportunities for value crea-
of it is ‘downcycled’ into a mix of different tion, in preserving and monetising the inhe-
alloys that can only be used for a small rent value in secondary materials.
96
Globally, CO2 emissions could
be reduced by 300 Mt per year
in 2050.
97
4.1 UNDERSTANDING FUTURE ALUMINIUM NEEDS

Aluminium is a relatively young, but versatile materi- is a global commodity. Raw materials, aluminium and
al with multiple and ever-increasing uses in packaging, buil- scrap metal are all traded internationally, and the EU im-
dings, automotive and other sectors. No other metal except ports a significant share of the aluminium it uses. Thus,
for steel is more widely used around the world. Produc- any effort to develop a more circular aluminium economy
tion of aluminium from ore (‘primary aluminium’) now needs to take global trade into account.
stands at just under 60 million tonnes per year, and it
continues to grow. Like other metals use, the benefits from aluminium de-
pend on the available stock: the total amount of metal
In recent years, growth in aluminium production and embodied in structures and products such as cars, buil-
use has been dominated by developing countries. China dings or machinery. Demand for aluminium derives from
in particular has been a major driver of aluminium de- the need to build up and maintain this stock – which is still
mand, accounting for more than 70% of the increase in growing in all world regions.1 To meet future needs, the
production in the past decade. As a result, 55% of prima- global stock per capita would need to increase by 350%
ry aluminium production is now in China. (see Exhibit 4.1 for an explanation of our methodology).
To keep up, aluminium production would have to increase
The analysis we present in this chapter focuses on the rapidly. By 2050, our scenario shows global aluminium
EU, but the global context is crucial because aluminium production would more than double from current levels.
Exhibit 4.1
Global demand for aluminium projected if by 2100, aluminium stocks

worldwide matched those in OECD countries today
GLOBAL ALUMINIUM PRODUCTION
Mt ALUMINIUM PER YEAR, 2016 – 2100
300
200
100
0
2020 2030 2040 2050 2060 2070 2080 2090 2100
98
The modelling of future demand follows the methodology described in Liu, Bangs,
and Müller (2013).2 Demand is modelled for six end-use sectors (consumer durables, construction,
electrical equipment, machinery, transport and other) and five world regions (Europe, North America,
Japan, China and Rest of World). Demand in each sector and region is modelled with a stock-based
approach, with population forecasts based on UN Population Prospects 2017.3 Per-capita saturation
at around 425-475 kg per person, depending on the region, follows projected GDP develoments
and is completed in all regions by 2100. In addition, the scenario includes a bottom-up model of
the evolution of vehicle production, including the shares of electrical and internal combustion engine
vehicles, and the assumed aluminium content in these categories. Historical aluminium stocks are
based on the data described in Bertram et al. (2017)4, while the split of stocks between end-use
sectors is based on Liu and Müller (2013).5 Future stock turnover is modelled with different lifetimes
of products in each end-use sector following a normal distribution. The steps of the aluminium supply
chain and use cycle are directly modelled, including losses in production, new scrap formation, and
collection of post-consumer scrap. Losses and other parameters are calibrated so production and
shares of products match those in data published by World Aluminium (2018a)6, and also based on
various estimates in the literature. The model also tracks separate flows of casting alloys vs. wrought
aluminium. It further represents degrees of international trade in scrap, and the dilution of casting
aluminium with other scrap or with primary metal. CO2 emissions estimates include emissions from
alumina refining, aluminium smelting, and casting, but do not include other sources sometimes
included in full life-cycle assessments (such as the emissions from the production of anodes, the
transportation or raw materials or products, or the mining of bauxite). Both direct emissions from fuel
use and indirect emissions from electricity production are included. In keeping with the approach
used elsewhere in this study, the baseline scenario includes the gradual decarbonisation of the en-
ergy system, as described elsewhere in the text, and by 2100, both alumina refining and electricity
production are therefore assumed to be fully CO2-free. The resulting global production of aluminium
is shown on the previous page, with further results elaborated throughout the chapter.
SOURCE : MATERIAL ECONOMICS ANALYSIS

Exhibit 4.2
EU aluminium demand grows only slowly to 2050
ALUMINIUM PRODUCTION REQUIRED TO MEET EU DEMAND

Mt ALUMINIUM PER YEAR, 2016 – 2050
15
10
0
2020 2030 2040 2050
NOTE : THE FIGURE DEPICTS THE PRODUCTION OF ALUMINIUM REQUIRED TO SERVE EU DEMAND, NOT PRODUCTION IN THE EU.
100
Within the EU, future demand growth will depend Second, the level of demand depends on when the
strongly on when saturation is reached, and at what level. stock will saturate. As noted, the above profile achieves
Although aluminium stocks vary significantly by country, an 80% increase in the stock by 2050. With still hig-
the current average is 250 kg per person. Our scenario her levels, demand would have to be higher. However,
assumes that aluminium stocks continue to grow until there also are strong reasons why continued growth
they reach 450 kg per person by mid-century, an 80% in the aluminium stock beyond these levels might not
increase. Thereafter, demand would be driven not by a be necessary. As the analysis in Chapter 5 sets out,
further increase in the total stock of aluminium, but by the demand for materials from the automotive sector could
need to replace products at end of their life. be as much as 75% lower in a future shared mobility
system than in today’s system of individually owned
However, this does not mean that production would need cars. Chapter 6 shows that materials use in buildings
to grow much. Today, the production required to serve EU de- also could be reduced significantly. These two sectors
mand amounts to just over 13 million tonnes per year of com- account for half of aluminium demand. Moreover, as
bined primary and secondary (i.e. recycled) aluminium, but this we discuss in Section 4.4, this does not in fact signi-
already includes continued additions to the stock.7 Achieving ficantly affect our findings; achieving a more circular
the 80% increase in stock by 2050 would require only a slight use of aluminium remains important even with higher
increase, to 14 million tonnes per year (Exhibit 4.2).8 growth.
Several things are notable about this projection. First, Finally, it is important to distinguish where alumi-
in keeping with our focus on the demand side, it covers nium is used from where it is produced. The alumini-
only the aluminium production required to meet demand um used for manufacturing in the EU has a variety of
for aluminium within the EU. Europe also manufactures sources; a third is imported. In that regard, aluminium
and exports a wide range of products containing alu- differs from the other materials investigated in this re-
minium – for example, it exports cars containing some port, which the EU produces more or less in the same
800,000 tonnes of aluminium each year.9 In principle, Eu- amounts as it uses domestically. In addition, the EU
ropean aluminium production could therefore grow faster imports half of the alumina used (alumina is the in-
than demand for aluminium in the EU, driven by exports termedia product in producing aluminium metal from
of either aluminium or aluminium-containing products. bauxite ore).
101
10
4.2 CO2 EMISSIONS FROM ALUMINIUM PRODUCTION AND USE

5
Primary production of aluminium requires very large amounts Unlike with steel production, however, emissions from
of energy. The smelting of aluminium uses 13–15 MWh per aluminium production can be reduced dramatically by using
tonne of metal produced. Multiply that by the almost 60 million low-carbon power. With a decarbonised electricity supply,
tonnes of aluminium now produced around the world, and the emissions from primary aluminium production are reduced
electricity usage amounts to as much as 5% of global electri- to just over 3 tonnes CO2 per tonne. Indeed, a majority of
city production – roughly
0 what is used for household lighting the aluminium produced in Europe is relatively low-carbon,
around world.10 When using fossil2020 2030
fuels for electricity produc- 2040hydro- and nuclear power.
because it is made with 205012
tion, large amounts of CO2 emissions result. In addition, gre-
enhouse gas emissions arise from the production of alumina, Yet as dramatically as low-carbon electricity can lower
and there are further emissions associated with the smelting emissions from aluminium production, recycling aluminium
process and with casting the metal into ingots. still has even lower emissions, because remelting requires
just 5% of the energy of primary production. This results in
Adding these up, emissions from aluminium production can emissions savings of up to 98% (Exhibit 4.3).
vary significantly. The key determinant is whether the electri-
city is low-carbon. Globally, 60% of aluminium today is pro- Accounting for both direct emissions and indirect emis-
duced using coal-fired electricity,11 which results in emissions sions, aluminium production emits some 800 Mt CO2
per tonne of aluminium produced of as much as 18 tonnes per hear (Exhibit 4.4).14 If global electricity production was
CO2. For comparison, this is more than 7 times the emissions decarbonised to meet climate targets, CO2 emissions from
associated with producing one tonne of primary steel. aluminium production would increase far less dramatically
Exhibit 4.3
Recycled aluminium has far lower emissions

than new primary production
CO2 INTENSITY OF PRIMARY ALUMINIUM
TONNES CO2 PER TONNE ALUMINIUM
17
12
10
3
3
0.3
COAL-BASED WORLD AVERAGE GAS-BASED PRODUCTION BASED RECYCLED

PRODUCTION PRODUCTION ON CO2 -FREE POWER ALUMINIUM
SOURCE : MATERIAL ECONOMICS CALCULATIONS BASED ON MULTIPLE SOURCES. 13
102
COAL-BASED WORLD AVERAGE GAS-BASED PRODUCTION BASED RECYCLED
PRODUCTION PRODUCTION ON CO2 -FREE POWER ALUMINIUM
Exhibit 4.4
Low-carbon electricity can dramatically cut emissions from

aluminium but cumulative emissions to 2100 nonetheless are large
ANNUAL EMISSIONS FROM GLOBAL ALUMINIUM PRODUCTION CUMULATIVE EMISSIONS FROM GLOBAL
Mt CO2 PER YEAR, 2015 – 2100 ALUMINIUM PRODUCTION, Gt CO2, 2015 – 2100
113
1600 100
1400
80
1200
79
1000
60
800
600 40
14 %
400
20
200
0 0
2020 2030 2040 2050 2060 2070 2080 2090 2100 BASELINE REFERENCE SCENARIO
SCENARIO BASED ON SLOWER
ENERGY TRANSITION
BASELINE SCENARIO REFERENCE SCENARIO BASED ON
SLOWER ENERGY TRANSITION
SOURCE : INTERNATIONAL ENERGY AGENCY (2017); MATERIAL ECONOMICS MODELLING AS DESCRIBED IN TEXT. 16
than demand for the metal, and starting in mid-century, ry aluminium required to meet economic needs. To put
they would actually decline.15 In a more typical ‘reference’ this in perspective, if primary aluminium production grows
scenario, however, where electricity production is decarbo- as rapidly as our analysis suggests, fully decarbonising
nised much more slowly than the goals of the Paris Agre- aluminium smelting through low-carbon power would re-
ement would imply, emissions from aluminium production quire an additional 1,335 TWh per year of zero-carbon
would gradually rise to 1.5 billion tonnes of CO2 per year. electricity in 2050 – almost as much electricity as India
Cumulative emissions (which are what matters most for uses today. Demand-side measures are therefore crucial
the climate) would be about 80–110 Gt CO2, or 27–37% to reduce the claim that aluminium makes on clean ener-
of the 300 Gt carbon budget available for all of industry in gy, and thus reduce the cost and increase the feasibility
a low-carbon scenario. Note that the ‘reference’ scenario of decarbonising the industry.
is shown for illustration of the impact of a slow energy tran-
sition on emissions only; it is not used for any evaluation of The good news is that there is significant potential to
the benefits of greater circularity described in this report. improve on current practices, to enable a larger share of
production through recycling and thus reduce the need
As this shows, even if emissions from aluminium were for future primary aluminium production even in a world
sharply reduced by changing production technologies17, with strongly growing demand.
there is much value in also reducing the amount of prima-
103
4.3 REALISING THE POTENTIAL FOR

CIRCULARITY IN THE ALUMINIUM SECTOR
The core of the circular economy opportunity for alumini- REDUCING LOSSES OF ALUMINIUM
um is to make the best use possible of metal that has already
been produced. This requires minimising the losses of alumi- For many consumers, aluminium is a prime example
nium, and ensuring that secondary production (i.e., recycling) of effective recycling. For decades now, many countries
can meet as much of aluminium needs as possible. have run successful aluminium recycling programmes, and
even if many other items are discarded, beverage cans are
The potential for more circular aluminium will grow in the carefully collected and reused. This system for recycling
future (Exhibit 4.5). Today, aluminium recycled from end- beverage cans is particularly effective, and in general, alu-
of-life products covers just 20% of total global demand. minium scrap collection rates are relatively high. Globally,
By 2050, however, this share will almost double, even as 77% of aluminium from end-of-life products is collected for
overall demand for aluminium more than doubles. Half the recycling, though the rate varies significantly across regions
growth in demand could thus be met by recycling. In Eu- and product groups.18 Still, this means 23% of end-of-life
rope, secondary production could be still more important, aluminium goes to waste.
as the total metal available for recycling would grow from
27% today, to 55% in 2050. In addition, large amounts of scrap aluminium are created
during the casting, fabrication or manufacturing processes
Both globally and in Europe, then, recycling could be – around 35% of the material ends up not in products, but
a major part of supplying future aluminium needs – with as ‘new scrap’.19 While almost all this new scrap is collected,
major savings of energy and CO2 emissions. However, some 2% is not returned for remelting or refining.20 In ad-
there is a long way from available scrap to production dition, for the 98% that is reused, some 2–3% is lost during
of high-quality, secondary aluminium that can be used processing, remelting and refining.21 All of this contributes
across applications. There are two key issues: to higher overall losses of aluminium.
• Reducing the losses of aluminium throughout the
Altogether, in each use cycle, 25–30% of aluminium is
use cycle; and
lost. To counter this, the single most important measure is to
• Preventing downcycling and enabling the produc- increase collection rates of post-consumer scrap, although
tion of high-quality secondary aluminium. reducing the amount of new scrap created is also helpful.
104
Exhibit 4.5
Post-consumer scrap could meet a large share of

future growth in global aluminium demand
AVAILABLE ALUMINIUM POST-CONSUMER SCRAP COMPARED TO TOTAL PRODUCTION
GLOBAL EU
Mt ALUMINIUM PER YEAR, 2016 – 2100 Mt ALUMINIUM PER YEAR, 2016 – 2050
300
14
250
12
200 10
8
150
100
50
2
0 0
2020 2030 2040 2050 2060 2070 2080 2090 2100 2020 2030 2040 2050
TOTAL PRODUCTION POST-CONSUMER SCRAP
NOTE : ‘AVAILABLE SCRAP’ REFERS TO THE AMOUNT OF POST-CONSUMER SCRAP THAT WOULD BE COLLECTED WITH CURRENT PRACTICES.
SOURCE : MATERIAL ECONOMICS ANALYSIS AS DESCRIBED IN TEXT
105
EMISSIONS INTENSITY
DOWNCYCLING OF ALUMINIUM TO CAST PRODUCTS

Although aluminium can be remelted indefinitely, there • There are many dozens of different alloys in major use, but
are three factors that greatly complicate secondary alu- a key divide is that wrought alloys contain much smaller quanti-
minium production: ties of alloying elements than do casting alloys.22 Wrought alloys
• Aluminium is rarely used in its pure form, but typi- with lower alloying content can generally be recycled into casting
cally alloyed with a range of other elements to attain alloys, but this is a one-way street, and wrought products cannot
the desired properties. Those properties often depend be created from aluminium that has once been a casting alloy.
on keeping alloying composition within a tight range, so • Finally, a major difference from steel is that most al-
even small deviations can undermine the quality for a loying elements cannot realistically be removed once they
specific purpose (Exhibit 4.6). have been added.23
Exhibit 4.6
Aluminium needs to be combined with alloying

elements to precise specifications
PURE ALUMINIUM BEVERAGE CAN AIRPLANE ENGINE BLOCK
• LOW DENSITY LID: 2 % CU 17 % SI

• HIGH CONDUCTIVITY 2,6 % MG 3 % MG 5 % CU
• CORROSION RESISTANT 0,25 % CR 6 % ZN 0,5 % MG
• SOFT 1 % ZN
CONTAINER:
1,2 % MN
1 % MG
ALUMINIUM FOIL WINDOW FRAME BICYCLE FRAME CAR WHEEL
> 99 % ALUMINUM 0,4 % SI 0,6 % SI 7 % SI

0,05 % CU 0,25 % CU 0,4 % MG
0,8 % MG 1,2 % MG
0,2 % CR
SOURCE : ADAPTED FROM MÜLLER (2017) . 24
106
Today, remelters and refiners handle these challeng- from a range of sectors into cast aluminium auto parts.
es by downcycling from wrought aluminium to less-pu- The main exception is used beverage cans, which are
re casting alloys. Only around 20% of end-of-life scrap recycled in a ‘closed loop’, where the same metal is
is turned into wrought aluminium, even though wrought used repeatedly for the same purpose. (Exhibit 4.7).25
products account for two-thirds of all aluminium in use. In addition to downcycling, both refining and remelting
The main uses for casting alloys, in turn, are in the involve diluting (‘sweetening’) more highly alloyed alu-
automotive sector. Overall, a large share of aluminium minium with up to 25% of either primary aluminium, or
recycling today involves turning end-of-life aluminium low-alloyed scrap.26
Exhibit 4.7
DRIVE TRAIN
The current aluminium cycle depends on downcycling
to cast aluminium used in the automotive sector
LOW ALLOY
PACKAGING
LOW ALLOY
TECHNOLOGY/OTHER
LOW ALLOY
BUILDINGS
CAST ALUMINIUM
LOW ALLOY
TRANSPORT
SOURCE : ADAPTED FROM MODARESI AND MÜLLER (2012) . 27
107
Downcycling to cast aluminium will no

longer be a viable strategy as the automotive
industry shifts production to electric vehicles.
Up until now, downcycling makes economic sen- demand for casting alloys, even with extensive trade
se. Continuous growth in the demand for casting al- across regions. Unless post-consumer scrap can be
loys of aluminium has been able to absorb available used for wrought products, more primary aluminium
post-consumer scrap. Europe and North America have will be required. Much of the potential for circularity
already reached the point where there is ‘excess scrap’ would then be wasted.
– i.e., scrap that exceeds domestic requirements for
cast aluminium, but which is too mixed or highly al- One other factor makes this issue even more pressing.
loyed to be used for other categories. However, this Today, around two-thirds of the aluminium content of
has been handled by exporting the scrap to other re- vehicles is cast aluminium.30 However, a large share of
gions.28 Downcycling has thus helped fulfil demand for that material is used in drive trains, motor blocks, heat
cast aluminium while minimising the cost of separation exchangers, and other components linked to the use
of products and sorting of scrap by alloy category. The of internal combustion engines. Yet it is increasingly
main loss is one of value, as less-pure secondary alu- clear that electric vehicles will make up a fast-growing
minium trades at a discount to purer forms.29 It is not share of the market in the coming decades.31 As the
likely that it has resulted in any additional production automotive industry shifts production to electric vehic-
of primary aluminium. les, the cast aluminium content per car could fall by
half.32 Demand for cast aluminium therefore looks set
However, this is not sustainable in the long term. to slow, even as post-consumer scrap of aluminium
As noted above, the amount of post-consumer scrap grows. In that context, downcycling will no longer be a
is growing globally. At some point, it will exceed the viable strategy. New approaches are needed.
108
Exhibit 4.8
A higher share of electric vehicles could significantly

reduce the demand for cast aluminium alloys
CAST ALUMINIUM COMPRISES THE LARGEST VOLUME OF ALUMINIUM CONTAINED IN TODAY’S PASSENGER CARS
BREAKDOWN BASED ON CONTAINED ALUMINIUM
CAST ALUMINIUM
LOW ALLOY
34% 14 %
66 %
ELECTRIC VEHICLES CHALLENGE THE USEFULNESS OF CAST ALUMINIUM

SHARE (%) CAST ALUMINIUM OF TOTAL CONTAINED ALUMINIUM
DRIVE TRAIN 12 %
OTHER ENGINE PARTS 10 %
Half of the total amount
HEAT EXCHANGER 9% of cast aluminium is in
components that are not
HEAD 8 % 14 % used in an electric vehicle
ENGINE BLOCK 7%
PISTON 0.4 %
FRAME PARTS 6%
BREAKS 3%
STEERING LINKAGE 3%
SUSPENSION ARM 2%
OTHER PARTS 6%
SOURCE : MATERIAL ECONOMCIS ANALYSIS OF MULTIPLE SOURCES. 33
109
Source: Excess capacity in the global steel industry: the current situation and ways forward, OECD (2015)
4.4 A CIRCULAR SCENARIO FOR ALUMINIUM

Reducing losses and avoiding the need for downcycling more than half of primary production. However, we must
could make a major difference to the sustainability of alumini- keep stressing that significant changes on current practice
um. Exhibit 4.9 contrasts the effect of a baseline scenario, with are needed to achieve this.35
only marginal improvement on current practice, with a circular
scenario with much higher degree of separation and sorting of The CO2 implications are similarly stark (Exhibit 4.10).
scrap.34 In a baseline scenario, emissions increase to 1,300 Mt CO2
per year, but in a circular scenario are kept to 1000 Mt CO2,
The results are striking: the circular scenario requires or 24% lower. The baseline scenario includes the switch to
far less primary aluminium to serve the same demand. By low-carbon energy described above. However, this transition
2050, the difference is 30 Mt per year, half of current prima- takes time. Increased circularity reduces the need for primary
ry production volume. Later in the century, the gains are still metals production while it is still highly emissions-intensive.
greater: instead of rising to 174 Mt per year, primary pro- To give a sense of the scale of resource savings: if the circu-
duction can peak at 116 Mt per year, and then decline even lar opportunity were not captured, but same CO2 savings had
as total aluminium demand continues to grow. This shows to be achieved by decarbonising the energy supply, the alu-
the great potential of aluminium as a circular material, and minium sector would require as much low-carbon electricity
the gains of ensuring that the aluminium stock is kept as as was produced by all wind and solar power in the EU in
pure as possible. By reducing losses and enabling high-qu- 2016. It is clear that recycling can be a major contributor to
ality recycling, secondary production could then replace aligning aluminium use with climate objectives.
Exhibit 4.9
A circular scenario can reduce the need for primary

aluminium production by 30 M t by 2100
ALUMINIUM PRODUCTION TO MEET GLOBAL DEMAND, BY ROUTE, 2016-2100
Mt ALUMINIUM PER YEAR PRIMARY PRODUCTION BASELINE SCENARIO
SECONDARY PRODUCTION CIRCULAR SCENARIO
300 PRIMARY PRODUCTION CIRCULAR SCENARIO
250
200
150
100
50
2020 2030 2040 2050 2060 2070 2080 2090 2100
110
Exhibit 4.10
CO 2 emissions are 300 M t lower globally

in a circular scenario
GLOBAL CO2 EMISSIONS FROM ALUMINIUM PRODUCTION BY ROUTE

Mt CO2 PER YEAR, 2016 – 2100
1400
1200
-24 %
1000
800
600
400 BASELINE
SCENARIO
200 CIRCULAR
SCENARIO
0
2020 2030 2040 2050 2060 2070 2080 2090 2100
SOURCE : MATERIAL ECONOMICS MODELLING AS DESCRIBED IN TEXT, SEE ALSO ENDNOTE. 36
111
ALUMINIUM PRODUCTION TO MEET EU DEMAND, BY ROUTE, 2016-2050
EU CO 2 EMISSIONS AND PRODUCTION for example, Norsk Hydro now markets certified primary alu-
minium with life-cycle emissions of no more than 4 tonnes
Turning to the EU, the gains from greater circularity are CO2 per tonne of metal.39
even more pronounced (Exhibit 4.11). As discussed in Se-
ction 4.1, the availability of post-consumer scrap is set to The gradual decarbonisation of the aluminium supply
increase significantly in the EU, creating the potential for a in the EU can go hand in hand with efforts to improve the
much more circular system. With current practice, however, circularity of European aluminium use. By both decarbonising
this potential would not be captured. More scrap would supply, and reducing the need for (imported) primary metal
become available, but the EU would need to use primary to serve demand, the EU can act on two complementary
aluminium instead – the same amount in 2050 as today. fronts to bring aluminium in line with low-carbon objectives.
Exporting excess cast aluminium scrap would not solve
the problem, as other regions also would have an excess.
In contrast, by improving circularity, and making sure that
SENSITIVITY ANALYSIS – THE FINDINGS ARE
more wrought aluminium demand can be met from availa- ROBUST TO A RANGE OF ASSUMPTIONS
ble post-consumer scrap, the need for primary aluminium
could be cut by 38%. Casting alloys of aluminium, and automotive aluminium
in particular, are the main potential future bottleneck to a
Given the much higher energy-intensity of primary pro- more circular aluminium system, and therefore the focus of
duction, it is clear that reducing the need for primary alu- this analysis. Wrought aluminium is treated as one category,
minium also makes a big contribution to reducing the CO2 which is a simplification, as there are limitations on which
emissions arising from European aluminium use. After the wrought alloys can be turned into other wrought alloys.
closure of several smelters, Europe now imports half of its
primary aluminium. Any reduction in demand for primary On the other hand, the circular scenario is not a theore-
aluminium in the EU thus avoids whatever CO2 emissions tical ‘100% circular’ benchmark, but is intended as an am-
are associated with production of those imports. In our sce- bitious, yet realistic, representation. Instead of assuming full
nario, this corresponds to almost 10 tonnes CO2 per tonne circularity for aluminium, it involves some continued dilution
of primary aluminium, even in 2050.37 A circular scenario and downcycling.40 Of course, if these limitations could be
for aluminium in the EU alone would thus reduce global overcome, the CO2 reductions from a more circular alumi-
emissions by 29 million tonnes of CO2. nium system would be even bigger than indicated above.
This raises an important difference between aluminium The analysis underlying the results depends on many
and other materials investigated in this study: most emis- factors that are uncertain in a 2050 perspective (let alone in
sions savings from circularity are likely to occur not within scenarios to 2100). However, the overall findings are robust
the EU, but in the countries from which aluminium is im- to variation in many key assumptions. In particular, they do
ported. Nonetheless, they are real, net savings to global not rely on a relatively flat demand profile. If the EU stock
emissions: a more circular system in Europe reduces the were to grow much more to 2050 (as suggested in some
global need for primary aluminium, after accounting for all industry forecasts), production might need to reach 20 Mt of
the possibilities of trade and adaptation of use of alumi- aluminium per year, and primary metal would have a much
nium in other markets. It therefore constitutes an unusual greater share. However, despite the fast-growing stock,
opportunity for Europe to adjust not just its own territori- the gains from circularity would in fact be still greater, with
al emissions, but also the emissions associated with its abatement potential increasing from 29 Mt CO2 per year
consumption. in 2050, to 44 Mt CO2 per year. Likewise, assuming a less
aggressive adoption of electric vehicles, or a larger pool
There is an additional opportunity for European industry of future cast aluminium demand more generally, reduces
to produce and market CO2-free aluminium. As noted ear- the abatement potential by only around 15%. Overall, the
lier, overall, European production uses mainly low-carbon findings therefore seem robust to a range of assumptions.
energy. Life-cycle emissions from EU aluminium are about
7 tonnes CO2 per tonne of aluminium, versus a global av- The main question instead is what actions need to be
erage of 18 tonnes CO2, and 20 tonnes in China.38 Some taken to achieve greater circularity, and whether policy sup-
companies are already seizing this business opportunity; port is needed to get there – the topic of the next section.
112
SCENARIO
0
2020 2030 2040 2050 2060 2070 2080 2090 2100
Exhibit 4.11
minimum primary aluminium production to serve EU demand

is 38% lower in a circular scenario by 2050
ALUMINIUM PRODUCTION TO MEET EU DEMAND, BY ROUTE, 2016-2050

Mt ALUMINIUM PER YEAR
14
12
10
6 -38 %
2020 2030 2040 2050
MINIMUM PRIMARY PRODUCTION BASELINE SCENARIO
POTENTIAL SECONDARY PRODUCTION CIRCULAR SCENARIO
MINIMUM PRIMARY CIRCULAR SCENARIO
SOURCE : MATERIAL ECONOMICS ANALYSIS AS DESCRIBED IN TEXT.
113
4.5 HOW TO GET THERE – MEASURES TO PROMOTE

THE CIRCULARITY OF ALUMINIUM
What would it take to transition from today’s practices to MEASURES TO PHASE OUT DOWNCYCLING
a more circular scenario? The key topics are the reduction
of losses, and addressing downcycling. The key issue for downcycling is to improve the hand-
ling of post-consumer scrap. If, as in the circular scenario
MEASURES TO REDUCE LOSSES described above, two-thirds of end-of-life aluminium is to be
reusable as wrought aluminium, significant changes will be
Most of the losses of aluminium occur as products required on current practice.
reach their end of life, so this is where to focus attention.
Collection systems is one aspect: while collection from buil- The good news is that there will be strong market in-
dings and cars is already very high, the rates are much centives for this, especially as the amount of excess scrap
lower for consumer products that are often handled through grows. Note that ‘excess’ secondary metal does not mean
public waste management systems. One option could be to that aluminium would be discarded, but that its price would
introduce additional deposit systems, similar to those used fall. Today’s price difference between secondary and prima-
for beverage cans today. Where waste is publicly handled, ry aluminium is relatively small (on the order of 10-15%, de-
it would help to separate metals ahead of other treatment. pending on circumstances), but this could grow significantly
More generally, attention throughout the chain would help if the supply-demand balance shifted. There would then be a
reduce losses. For example, a greater share could be re- strong underlying financial gain from taking action to avoid it,
covered if initial product design ensured easy separation of and today’s remelters would have strong incentives to deve-
aluminium components upon dismantling. lop and invest in additional separation and sorting.
Reducing process scrap would also reduce losses of Although much could be achieved by greater sorting of
aluminium. Collection rates are already high, so the main existing scrap flows, this does not mean that the aluminium
promise is likely in reducing formation of new scrap in the industry should be the only one to act. The cost of achieving
first place. There are large differences today: some car ma- purer scrap flows could be much smaller if not achieved just
nufacturers, for example, generate more than 40% scrap, post-hoc, but instead simultaneously addressed by actors
even as best practice is closer to 25%. New production throughout the value chain: manufacturers using aluminium,
techniques, such as additive manufacturing, could also lead public deposit systems, collection systems of end-of-life me-
to lower scrap volumes. tal, practices for dismantling products, scrap markets, and
users of recycled metal. The more integrated the aluminium
industry, the easier it will be to get there, but in any scenario it
will also be necessary to engage manufacturing companies
and others.
114
This is an area that needs much more investigation, but • New processes for dismantling of products: To-
the broad areas of action are as follows: day’s dismantling is often rudimentary, whether it involves
demolition of buildings or shredding of vehicles. The result
• Improved product design: As noted throughout this is that different categories of aluminium are mixed, and
report, there are currently few incentives for product design other metals enter aluminium flows. Given the importance
to account for the impact on recycling. With aluminium this of the automotive sector for aluminium use, a long-term pri-
already has some negative effects (chiefly in the form of los- ority should be to create effective automated disassembly
ses), but with the need for much more detailed separation systems to separate auto parts before shredding. Even with
and sorting of scrap, it becomes more urgent. Among op- today’s technology, however, it is possible to carry out more
tions to discuss, the EU framework for producer responsibility component separation of cars and thereby reduce the mix-
could address the feasibility of achieving pure scrap flows. ing of different aluminium categories.41
• Materials specification: Aluminium continues to de- • Increased separation and sorting: A large share
velop, with new alloys invented to serve new uses, and to of the effort towards purer scrap flows must be with the
improve on previous versions. This has many benefits, and process of sorting scrap flows. The aluminium industry is
doubtless will continue, making it possible to use aluminium already investing significantly in this, and new possibilities
in new applications. At the same time, the proliferation of dif- arise with cheaper sensor and automation technologies.
ferent alloys also makes it much more difficult to use recyc- Further reducing the cost of sorting must be a priority, and
led content. Thus, there needs to be a countervailing move- could be a significant competitive advantage in a future in-
ment to avoid needless over-specification. One step would dustry with a much greater share of post-consumer scrap.
be for the industry to shift away from the current practice of • A more sophisticated market for scrap: In parallel,
specifying the precise composition of alloys, to specifying there will be a need for a more developed market for scrap,
and buying aluminium based on function instead – leaving with additional specification of categories. One possibility
it up to the supplier to meet this in the best way possible. is to move towards much more real-time information about
Recycled aluminium could then meet more of the demand. different scrap flows available in a geography, perhaps
It also is worth considering whether regulations and particu- tracked through distributed ledgers or similar information
lar practices prevent such developments. Examples include technology.
the intellectual property invested in some alloys, or industry
standards. • Production process development: Finally, it is pos-
sible that methods could be developed to improve the
• Reuse: Downcycling can be avoided entirely by reusing commercial viability of removing impurities from aluminium.
components rather than remelting metal, a topic returned to Many methods for refining aluminium are already in use and
later in this report. could be developed further, including electrolysis, electro-
• Closed-loop recycling systems: Reusing metal for slag refining, fractional solidification, fluxing refining, etc.
a very similar purpose is the shortest route to handling the
issue of secondary aluminium alloys. Today the only major Taken together, the above is an ambitious list for indu-
flow of this kind is used beverage cans, but it would be pos- strial development. Vertical integration, coordination across
sible to do it for additional product categories as well. For the supply chain, and new technology are among the key
consumer products, publicly run systems are the most likely themes. Policy may well have a role to play as well, not least
to succeed. Deposit schemes would be one way to support to coordinate the important action required in the areas of
this. However, much more analysis is required to identify the product design, public collection systems, and end-of-life
pros and cons of such initiatives. dismantling. All are areas that needs further investigation.
115
The Circular Economy – a Powerful Force for Climate Mitigation / Mobility
5. mobility
the promise of a shared car system
Personal mobility offers a major opportu- hand, would enable much more intensive
nity for a more circular economy. Transpor- use of each vehicle. Once the use of these
tation is of undoubted economic and social fleets achieves sufficient scale, there will be
importance, accounting for 15% of house- enormous incentives for changes to the de-
hold disposable income. It is a major fea- sign of vehicles and for innovation. Higher
ture of everyday life and shapes the cities utilisation justifies much more investment
and communities we live in. In this chapter, in upfront costs, from the higher cost of
we show how circular economy principles electric-vehicle drivetrains, to more advan-
could enable the same travel to occur with ced automation technology, or higher-per-
much lower materials requirements – and formance materials.
at a lower cost.
Professionally managed fleets also en-
We focus on passenger cars, which pro- able much greater control over vehicle ma-
vide 83% of travel today. In a scenario whe- intenance, parts inventory, reuse of com-
re professionally managed, shared vehicle ponents, and remanufacture. In addition, the
fleets account for two thirds of travel, mate- cars used can be matched much more close-
rials requirements could fall by as much as ly to the needs of individual trips, thus redu-
75%, reducing annual CO2 emissions from cing the average size of vehicles substantially.
materials production by 43 Mt by 2050.
The average car would be smaller, requi-
The reason why the potential savings are ring far less material, and be far more dura-
so large is that our use of cars today is ex- ble and better maintained. The initial design
tremely inefficient. The vast majority of cars and materials choices would be optimised
are owned and operated by a single hou- for much more intensive use, and the ef-
sehold, and this results is overcapacity: fi- fective lifetime could more than double.
ve-seat cars used mostly for one-passenger Lightweighting techniques that use advan-
trips, and vehicles that are stationary 92% ced materials would be far more economic
of the time. Vehicle design is also optimised than when applied to a personal car. The
for this structure of use and ownership. same is true of using automation to redu-
ce accidents, and applying more advanced
A system built around professionally ma- manufacturing methods to reduce materials
naged fleets of shared cars, on the other losses at the production stage.
116 116
A shared car system
reduces material requirements
for passenger cars by 75%.
117
Exhibit 5.1
A shared mobility scenario is a highly attractive

vision for passenger cars
CO2 IMPACT OF MATERIALS TOTAL COST OF OWNERSHIP EXTERNALITIES AND COST TO SOCIETY
Mt CO2 PER YEAR, EUROPE EUR PER 1000 PKM EUR PER 1000 PKM
60 155 264
STEEL
-70%
-77% -74%
ALUMINIUM
18
68
35
PLASTICS
OTHER
BASELINE CIRCULAR CURRENT CIRCULAR CURRENT CIRCULAR

SCENARIO, 2050 SCENARIO, 2050 SCENARIO, 2050 SCENARIO, 2050
PKM = Passenger-kilometre
LIFECYCLE CO2 EMISSIONS FROM INTERNAL COMBUSTION ENGINE VEHICLES (ICEVs) VERSUS ELECTRIC VEHICLES (EVs)
g CO2 PER CAR-KM NOTE : PKM DENOTES PASSENGER-KILOMETRES.
216 167 37
Although the focus of this report is on materials use 9% congestion, air pollution and
from factors such as traffic
and associated CO2 emissions, the main motivation for collisions. Although the pace of change will depend
such a system is the much wider productivity opportunity on many factors, not least travellers’ expectations and
that it represents. We calculate that the cost per pas- 58%
norms about car ownership, it is clear that the benefits
senger-kilometre could be as much as 77% lower than of achieving a mobility system based on shared cars
that of individually 77% Major externalities could
USE PHASEowned cars. would be very large – contributing to economic produc-
be reduced by three-quarters, including major costs tivity as well as climate goals.
91%
42%
PRODUCTION
23%
AND MATERIALS
118
ICEVs EVs, CURRENT ENERGY EVs, RENEWABLE
MIX (EUROPE) ENERGY
5.1 UNDERSTANDING FUTURE MOBILITY NEEDS

In its 2011 White Paper on transportation, the Europe- In addition, today’s transport system has large impacts
an Commission wrote: “Transport is fundamental to our on public health, well-being, and the environment. Road
economy and society. Mobility is vital for the internal mar- traffic leads to local air pollution, traffic congestion, noise
ket and for the quality of life of citizens as they enjoy their and accidents, and transport fuels are a major source of
freedom to travel.”1 CO2 emissions. Car travel also depend on large public
expenditure on infrastructure, and it claims valuable land
For personal travel, mobility is overwhelmingly a story area. As we discuss below, these costs can in fact ex-
of passenger cars. Despite the undoubted importance of ceed even the substantial direct costs of transport by
public transport, cars account for 83% of land transporta- passenger car.
tion in the EU.2 Passenger transport grew by 5% between
2004 and 2014,3 and the number of cars is growing fas- These issues set the frame for the main focus of this
ter than the population.4 study, which is the materials use associated with car
transport.9 The average car contains 1.4 tonnes of mate-
Cars are also major capital assets, both for households rials, dominated by steel, aluminium and plastics.10 Still
and in the wider economy. The price of a new car is on more materials are used in manufacturing, as it is not
the same level as average annual household disposa- uncommon that 40% of metals used are lost as scrap
ble income.5 There are more than 250 million passenger on the factory floor.11 As a result, the automotive sector
cars in the EU with a replacement value of more than uses 25% of aluminium, 12% of steel, and 9% of plastics
5,700 billion EUR6 – twice as much as total investment in consumed in the EU.12 Car manufacturing has also been
the entire EU28 economy in a year.7 Adding the running a major force behind materials development, from new
costs of fuel, maintenance, insurance and more, total ex- applications of aluminium, to high-strength steel, to rein-
penditure on driving takes up some 13% of household forced plastics and carbon fibre. At end of life, vehicles
disposable income in the EU.8 are also a major factor in materials recycling.
119
The circular economy opportunity
for transportation is enabling the same
distance and convenience of transport,
but requiring much less input of
materials.
120
ALUMINIUM
18
68
35
PLASTICS
OTHER
BASELINE CIRCULAR CURRENT CIRCULAR CURRENT CIRCULAR

SCENARIO, 2050 SCENARIO, 2050 Exhibit 5.2SCENARIO, 2050 SCENARIO, 2050
Emissions from materials and production will be

the main source of CO 2 from future cars
g CO2 PER CAR-KM
216 167 37
9%
58%
USE PHASE 77%
91%
42%
PRODUCTION
23%
AND MATERIALS
ICEVs EVs, CURRENT ENERGY EVs, RENEWABLE

MIX (EUROPE) ENERGY
Note: Emissions calculated per car kilometre for the large vehicle segment
with an average weight of 1528 kg over a lifetime of 180,000 km.
NOTE : THE DATA SHOW LIFECYCLE GREENHOUSE GAS EMISSIONS FROM PRODUCTION AND THE USE -PHASE
FOR A LARGE PASSENGER CAR WITH A LIFETIME DRIVING RANGE OF 180,000 KM.
PRODUCT MATERIALS EFFICIENCY CIRCULAR
SOURCE BUSINESS
: ELLINGSEN ET AL. MODELS
(2016) 13 MATERIALS INTENSITY OF TRANSPORT
TONNES MATERIALS PER CAR NUMBER OF CARS REQUIRED TONNES MATERIALS TO ENABLE TRAVEL
As detailed
1. REDUCE WEIGHT in the PER preceding
VEHICLE chapters of this report,
3. SHARING Given the economic importance MATERIALS of cars,
INPUTSand their lar-
REQUIRED
the production
Lower averageof these
vehicle size materials
in a shared leads to large CO Large emis- ge share of materials use, TO
2 increase in utilisation in a
it matters
SUPPORT greatly
MOBILITYhow cars
sions. car Tosystem,
date,better
analyses of CO
matching vehicles to 2 emissions fromshared-car
vehicles system are used. As discussed below, Largeour current
reduction in totalmodel
amount ofofmaterials
car
actual trip needs
have focused on the use phase rather than on emissions Higher occupancy per vehicle with and use is highly
ownership thatmaterials intensive:
must be produced large
for useful travel
Advanced design and materials choices mobility as a service
from materials production and manufacturing. This is un- volumes of steel, aluminium,(materials per passenger kilometres)
plastics and other materi-
derstandable, as emissions from fuels account for 70% of als are required to achieve relatively little transportation
total2.lifecycle
However,
REUSE AND
aswaste
Reduced
REMANUFACTURING
emissions
fuelin vehicles
emissions
X 4. LONGER
of a typical car today (Exhibit
are reduced and eventually
production
LIFESPANS
5.2).
tion isincentives
More intensive use increases
=
benefit. The circular economy opportunity for transporta-
to improve this, enabling the same distance and
reach Modular
zero, the design and replacement of
emissions from materials become for durable design
the re- convenience of transport, but requiring much less input
components with limited lifetime Electric drivetrains with intrinsically
maining obstacle
Increased re-use oftodurable
CO2components
-free transportation. In anbetter durability of materials.
electric
vehicleatusing zero-carbon energy, almost all the remaining
end-of-life Proactive maintenance of fleet-managed
emissions come
Controls from vehicle
over inventory and flows production
in a and materials
vehiclesuse.
fleet-managed system of shared cars Modular design to enable replacement
of components with shorter lifespans
121
5.2 THE CIRCULAR ECONOMY OPPORTUNITY IN MOBILITY

g CO2 PER CAR-KM
In previous chapters, the focus was on the potential In the case of passenger cars, the relevant ‘service’ is
216 167 37
of materials recycling to replace new production of ma- mobility, best measured in ‘passenger-kilometres’ (Exhibit
terials such as steel, plastics and aluminium. However, 5.3). Materials efficiency9%for cars centres primarily on two
the opportunities for a circular economy also encom- opportunities: reducing the average weight of each car, and
pass many strategies that focus on the products them- reducing the input of materials by reducing scrap in manu-
selves. By reducing the amount of material required for facturing and increasing reuse and remanufacturing. Two
58%
each product, materials efficiency can increase. Then, additional opportunities are central to providing more use-
by ensuringUSEthat 77% provides more useful
each product
PHASE ful travel with each vehicle: increased sharing to increase
services, the resource claims can be reduced even the utilisation and occupancy of each vehicle, and longer
further. lifespans to make it possible
91% for each car to serve a greater
number of passenger-kilometres once produced.
42%
PRODUCTION
23%
AND MATERIALS
ICEVs
Exhibit 5.3
EVs, CURRENT ENERGY EVs, RENEWABLE
MIX (EUROPE) ENERGY
Four circular economy strategies can sharply reduce
Note: Emissions calculated per car kilometre for the large vehicle segment
the materials requirements ofofmobility
with an average weight 1528 kg over a lifetime of 180,000 km.
PRODUCT MATERIALS EFFICIENCY CIRCULAR BUSINESS MODELS MATERIALS INTENSITY OF TRANSPORT

TONNES MATERIALS PER CAR NUMBER OF CARS REQUIRED TONNES MATERIALS TO ENABLE TRAVEL
1. REDUCE WEIGHT PER VEHICLE 3. SHARING MATERIALS INPUTS REQUIRED

Lower average vehicle size in a shared Large increase in utilisation in a TO SUPPORT MOBILITY
car system, better matching vehicles to shared-car system Large reduction in total amount of materials
actual trip needs Higher occupancy per vehicle with that must be produced for useful travel
Advanced design and materials choices mobility as a service (materials per passenger kilometres)
2. REUSE AND REMANUFACTURING

Reduced waste in vehicles production
X 4. LONGER LIFESPANS
More intensive use increases incentives
=
Modular design and replacement of for durable design
Increased re-use of durable components better durability
at end-of-life Proactive maintenance of fleet-managed
Controls over inventory and flows in a vehicles
1 3 5
TRAVEL DEMAND CAR OWNERSHIP CAR WEIGHT
TRILLION PASSENGER-KM BY CAR PER YEAR 122EU28
CARS PER CAPITA, TONNES PER CAR
4.7 0.57 1.39

4.5 0.53 1.34
Today’s cars have an effective
utilisation of just 2%. Increasing
this is at the core of the circular
opportunity.
123
THE CURRENT STATE OF CIRCULARITY IN MOBILITY: POTENTIAL WAITING TO BE UNLEASHED
Our current use of cars is highly wasteful. For all the Why is the picture so bleak? Unsurprisingly, there are
economic importance of passenger cars, they are used powerful reasons for the current situation, rooted in the
very little. Cars are parked for 92% of daylight hours14, model of individual car ownership. Individually owned cars
and when in use, only 1.5 of the typical five seats are must be able to serve each household’s peak travel needs.
occupied.15 The overall utilisation rate thus is 2%16 – They serve relatively low annual travel needs, are maintai-
an extremely low number compared to other assets of ned in households’ leisure time, and are used only when
similar economic importance. Utilisation matters: the individuals in each household need to travel.
current situation amounts to locking in a large store of
value and embodied CO2 but getting only a low level of The case of vehicle lifetimes provides perhaps the clea-
benefit in return. rest example of why this model of ownership and operation
limits circularity. Cars are driven some 200,000 km befo-
Summarising the current trends, there is at best slow re being scrapped20, but as each car is driven a relatively
movement towards reducing cars’ claims on resources short distance, it takes many years to get to that point.21 The
(Exhibit 5.4). As noted above, total travel demand is average lifetime of all cars in use is more than 10 years,
increasing in the EU, and the number of cars is growing and vehicles are on average 14 years old when scrapped.22
faster than the population, leading to lower rather than Any improvement to lifetime is therefore felt only long after
higher utilisation for each vehicle. Although there have purchase, too far off to justify a higher price, and also of
been major efforts to reduce vehicle weight to lower fuel limited benefit to users (as cars 15–20 years old would be
consumption and to comply with emissions regulations, out of date, lagging behind in technology, safety and design).
the total weight of vehicles is rising rather than falling.17 Even incremental change is difficult, mainly a by-product of
As noted, manufacturing remains wasteful, with 40% or improving reliability. A step-change in lifetime, on the other
more of metals lost as manufacturing scrap. There is hand, is all but precluded by the current ownership model:
also no discernible increase in the average occupancy cars designed and maintained to last for, say, 500,000 km
of individual vehicles. Car lifetime is the main area whe- would be 35 years old when finally scrapped. There would
re there is some improvement, as cars last for around be little or no consumer demand for such vehicles. The main
20% more kilometres of travel today than they did 10 lesson is that today’s more limited car lifetime is no absolute,
years ago.18 technical limitation, but follows from the low rate of utilisation.
124
1. REDUCE WEIGHT PER VEHICLE 3. SHARING MATERIALS INPUTS REQUIRED
Lower average vehicle size in a shared Large increase in utilisation in a TO SUPPORT MOBILITY
car system, better matching vehicles to shared-car system Large reduction in total amount of materials
actual trip needs Higher occupancy per vehicle with that must be produced for useful travel
Advanced design and materials choices mobility as a service (materials per passenger kilometres)
2. REUSE AND REMANUFACTURING

Reduced waste in vehicles production
X 4. LONGER LIFESPANS
More intensive use increases incentives
=
Modular design and replacement of for durable design
Increased re-use of durable components
at end-of-life
Exhibit 5.4
better durability
Proactive maintenance of fleet-managed
Current trends see little improvement in

Controls over inventory and flows in a vehicles
the materials intensity of EU mobility
1 3 5
TRAVEL DEMAND CAR OWNERSHIP CAR WEIGHT
TRILLION PASSENGER-KM BY CAR PER YEAR CARS PER CAPITA, EU28 TONNES PER CAR
4.7 0.57 1.39

4.5 0.53 1.34
2005 2015 2005 2015 2005 2015
2 4 6
OCCUPANCY PER CAR CAR LIFETIME SCRAP IN CAR MANUFACTURING
CAR OCCUPANCY (PASSENGER PER CAR), UK CAR KM PER CAR DURING LIFETIME % SCRAPPED MATERIALS
1.57 1.56
170
140
~40%
2005 2015 2006 2013
DOWNGRADING AND COPPER MIXING OF STEEL PREVENTION OF PLASTICS RECYCLING
SOURCE : MATERIAL ECONOMICS ANALYSIS OF A RANGE OF SOURCES. 19
ONLY 8% OF STEEL RECYCLED FROM VEHICLES HIGH-VALUE RECYCLING IS NOT FEASIBLE FOR
CAN BE USED AS MATERIALS FOR NEW CARS A LARGE SHARE OF VEHICLE PLASTIC VOLUMES
Shredding of vehicles mixes copper with steel which is
125 Current practices leaves plastics in a mixed fraction
a serious long term contaminant of the steel stock that often is landfilled / incinerated
Alloys are not separated, leading to lost values of alloy Material substitution for lightweighting leads to fibre
metals, waste of critical materials, and downcycling of reinforced plastics that contaminates other plastics flows
The current ownership model also creates a disconnect ed wrought aluminium products are mixed with highly alloy-
between car manufacturing and the subsequent treatment of ed cast aluminium components when cars are dismantled,
2005 2015 2005 2015 2005 2015
cars at end of life. Recycling of end-of-life vehicles focuses and other, undesired metals are introduced to the aluminium
primarily on avoiding the release of hazardous substances scrap.24 In the case of plastics, recycling tends to be limited
2
and recovering spare parts. Materials recovery is far less of
a concern. Although regulation now requires that most mate-
4 6
to a few, large pieces, but overall recovery rates are low.25
OCCUPANCY
rials PER the
be recycled, CARdegrada-
CAR process results in significant LIFETIME For these reasons, theSCRAP IN CARvalue
scrapping MANUFACTURING
of a car is close to
CAR OCCUPANCY (PASSENGER PER CAR), UK CAR KM PER
tion of quality and loss of materials value (Exhibit 5.5). The CAR DURING LIFETIME % SCRAPPED MATERIALS
zero, even though just the major raw materials that go into a car
steel in vehicles is often highly specialised, including highly have a value of some 2,000–3,000 EUR. In many countries,
1.57 1.56
alloyed varieties, but shredding results in scrap that is often cars are returned for recycling only after a delay,27 and many
so contaminated and mixed that it is usable only for basic vehicles170
are not fully accounted for.28 Just as with extending life-
construction steels. By one estimate, only 8% of the steel140 re- time, however, there is little incentive for the manufacturer to en-
~40%
cycled from vehicles is of a quality that it could be used again able higher-quality recycling of materials as long as that value
for its original purpose. In the case of aluminium, low-alloy-
23
accrues only 15 years in the future, and then to another party.
Exhibit 5.5
2005 2015 2006 2013
Cars are at the centre of key challenges
to a more circular materials system
DOWNGRADING AND COPPER MIXING OF STEEL PREVENTION OF PLASTICS RECYCLING
ONLY 8% OF STEEL RECYCLED FROM VEHICLES HIGH-VALUE RECYCLING IS NOT FEASIBLE FOR
CAN BE USED AS MATERIALS FOR NEW CARS A LARGE SHARE OF VEHICLE PLASTIC VOLUMES
Shredding of vehicles mixes copper with steel which is Current practices leaves plastics in a mixed fraction
a serious long term contaminant of the steel stock that often is landfilled / incinerated
Alloys are not separated, leading to lost values of alloy Material substitution for lightweighting leads to fibre
metals, waste of critical materials, and downcycling of reinforced plastics that contaminates other plastics flows
steel and that is difficult to recycle
DOWNGRADING OF ALUMINIUM LIMITED RECYCLING OF RARE CRITICAL METALS
MIXING OF ALUMINIUM ALLOYS RESULTS IN END-OF-LIFE VEHICLES LEADS TO LARGE LOSSES

DOWNGRADING OF WROUGHT ALUMINIUM OF CRITICAL METALS
AUTO PARTS Rare critical metals can make up 1% of total vehicle
materials
Cars make up >40% of cast aluminium demand, a key
Only eight out of 25 scarce metals are recycled, with
‘sink’ for aluminium recycling
the remainder lost to carrier metals, construction and
Aluminium used in cars is downgraded when mixed
backfilling materials and landfills
with cast aluminium, precluding other uses
SOURCE : MATERIAL ECONOMICS ANALYSIS BASED ON MULTIPLE SOURCES. 26
126
The scrapping value of a car is close
to zero, even though just the major
raw materials that go into a car have
a value of some 2,000–3,000 EUR.
127
DOWNGRADING OF ALUMINIUM LIMITED RECYCLING OF RARE CRITICAL METALS
MIXING OF ALUMINIUM ALLOYS RESULTS IN END-OF-LIFE VEHICLES LEADS TO LARGE LOSSES

DOWNGRADING OF WROUGHT ALUMINIUM OF CRITICAL METALS
AUTO PARTS Rare critical metals can make up 1% of total vehicle
Cars make up >40% of cast aluminium demand, a key
Exhibit 5.6 materials
Only eight out of 25 scarce metals are recycled, with
A circular scenario see major changes in car ownership,
‘sink’ for aluminium recycling
the remainder lost to carrier metals, construction and
Aluminium used in cars is downgraded when mixed
backfilling materials and landfills
design, and materials use
with cast aluminium, precluding other uses
FROM: AN OWNED CAR ECONOMY TO: A SHARED CAR ECONOMY
A FUNDAMENTAL SHIFT IN Effective utilisation of a typical European Cars shared and operated in fleets increase
OWNERSHIP AND OPERATING car is 2%, reflecting the needs of individual both utilisation and occupancy per car
MODEL car owners High utilisation dramatically changes
Slow adoption of capital intensive incentives for durability, maintenance,
technologies (such as autonomous drive and re-use, as well as a range of more
electric drivetrains) due to high investment capital-intensive solutions
cost and low utilisation
CARS DESIGNED AND LIFETIME OF ~12 YEARS OR 230,000 KM, LIFETIME INCREASED TO 700,000 KM, MATCHING
MANAGED FOR RUN-TIME MATCHING THE NEEDS OF INDIVIDUAL OWNERS THE NEEDS OF INTENSIVELY USED VEHICLES
Low utilisation places limits on up-front Electric drivetrains with intrinsically

investment and on benefits of longer lifetime longer lifetime
Investments in more durable materials
Limited maintenance, and disconnect (more attractive for highly utilised cars)
between end-of-life vehicles and first owners Proactive maintenance and modular
design enable re-manufacturing
MATERIALS IMPACT LOW UTILISATION, HIGH WEIGHT, AND LIMITED RANGE OF SIZES, ADVANCED MATERIALS, AND
MINIMISED LIFETIME RESULTS IN HIGH MATERIALS INTENSITY AUTONOMOUS DRIVING REDUCE FOOTPRINT
Cars weigh 1.4 tonnes, as each vehicle Smaller average size adapted to needs of
needs carry for five seats and have high each trip
safety standard Fleet management and more expensive
materials incentivise re-use of parts and
EOL recovery
MATERIALS PER CAR NUMBER OF CARS MATERIALS INTENSITY OF TRANSPORT

TONNES NEW MATERIALS CARS PER MILLION PKM TONNES MATERIALS PER MILLION PKM
128
1.2 2.6 3.2
5.3 A CIRCULAR SCENARIO FOR MOBILITY

CAN CUT CO 2 EMISSIONS BY 70%
The above diagnostic may sound like a litany, but in than offset by benefits over just a few years of intensive
fact it points the way to a major productivity opportunity. use (instead of paying off over more than 14 years, as
In a 2050 perspective, not just an incremental shift is is the case today). The switch to electric vehicles is an
possible, but there is an opportunity to reorganise mobili- integral part of longer technical lifespans, as the 20-or-so
ty from an owned car economy to one dominated by sha- moving components of an electric motor are intrinsically
red cars (Exhibit 5.6). This comes with major advantages much longer-lasting than the 2,000 moving parts of an in-
not just in terms of CO2 emissions and materials use, but ternal combustion engine.29 Improved maintenance also
in drastic cost reductions, and it helps address major helps; in a fleet-managed system this could resemble
negative effects on public health and the environment. that of aircraft today, with detailed and predictive mainte-
The core of this scenario is a major shift in how ve- nance schedules for each component.
hicles are owned and operated. Instead of individually • Increased reuse and remanufacturing: The trans-
owned vehicles that are mostly stationary, 64% of cars action costs and inventory problems that limit reuse and
would be operated as shared vehicles, with much higher remanufacturing would also be much easier to overco-
utilisation (the total time the vehicle is in use), and higher me in a fleet-managed system. Longer lifetime need not
average occupancy (the number of passengers per trip). depend on all parts of a car living longer, but can be
Shared vehicles would operate as a service to end-users, enabled by a modular design that allows for components
who would be able to call on a car when needed without with shorter lifespans (whether for technical, aesthetic
having to own individual vehicles. Ownership and ma- or other reasons) to be easily replaced, and intrinsically
nagement instead would lie with service providers, who long-lived components to be reused even when vehicles
would own and operate fleets of cars. Self-driving vehic- are scrapped.
les would be integral to this business model, making it • Smaller and lighter vehicles: It would also be pos-
possible to call on vehicles when needed without large sible to vary car size so it matches the needs of each trip.
costs. This in turn can unlock a number of step-change Instead of each vehicle dimensioned for its maximum ca-
improvements in the productivity of transportation: pacity (typically, five passengers), cars could range in
• Faster adoption of electric and automated vehicles: size from single- to multi-person vehicles, matched to the
Higher utilisation would fundamentally shift the incentives needs of each trip. The cost of more advanced materials
for how vehicles are designed and operated. The same required for lighter weight, such as high-strength steel,
upfront investment would be paid back earlier, over a advanced plastics, carbon fibre or aluminium, would also
much larger number of kilometres driven. This makes be much easier to motivate with higher rates of utilisation.
electric vehicles more economic, as the higher upfront • Better end-of-life management: Finally, with more
cost can be rapidly recovered. It also enables more ra- expensive materials, more scope for remanufacturing,
pid adoption of connected and self-driving vehicles, both and more predictable end-of-life flows, it would be much
because the upfront cost is easier to bear with higher more profitable to design vehicles for full recovery of ma-
utilisation, and because its value is much higher in a sha- terials at end of life. Manufacturers and operators of sha-
red-car business model. red-car fleets could directly link up (or vertically integrate)
• Increased lifespan: A shared model also creates in- to monetise the benefits. Vehicles could be designed for
centives for cars with much longer lifetimes. Cars might much more automated disassembly, which also is inte-
not be older, counted in years, but would cover the same grated with the increased reuse and remanufacturing
distance in a much shorter period of time. The additional described above.
cost of achieving a longer lifetime would thus be more
129
ANATOMY OF A SHARED, FLEET-MANAGED CAR IN 2050

To see the extent of improvement possible, it is useful This is an astounding productivity improvement – but is it
FROM: ANcar
to consider the ‘materials intensity’ of current OWNEDtravel. CAR ECONOMY
realistic? We would argue TO:that
A SHARED CAR ECONOMY
it is. Already today, the early in-
Today, to achieve one million passenger-kilometres of tra- dications are that shared cars can replace 4–10 private cars
vel, some 3,200 kg of materials must be supplied. This in Europe and 9–13 private cars in North America.30 In terms
may not sound like much,
A FUNDAMENTAL SHIFT INbut given the 4.7 trillion
Effective utilisationpas-of a typical
of lifetime,
EuropeanLondon ‘black Carscabs’
shared already have ainlifetime
and operated of up
fleets increase
OWNERSHIP AND OPERATING car is 2%,
senger-kilometres served by cars in the EU each year, it reflecting the needs of individual both
31 utilisation and occupancy per car
to 1,600,000 kilometres, about 9 times greater than that of
adds up to more than 20 million tonnescar
MODEL perowners
year of steel, ordinary passenger cars. HighThe key is adramatically
utilisation custom-made changes
vehicle,
aluminium, plastics and other materialsSlow adoptionjust
annually, of capital
to intensive
designed for durability incentives
and easyformaintenance
durability, maintenance,
during intensi-
technologies (such as autonomous drive and re-use, as well as a range of more
keep the car fleet going. electric drivetrains) due to high veinvestment
use. With improved technology
capital-intensiveand greater use of electric
solutions
cost and low utilisation drivetrains by 2050, the 700,000 kilometres assumed here
These changed incentives of a shared-car system does not seem far-fetched. In terms of weight reduction, the
could transform the key parameters in this equation. The 600 kg average size need not mean that all vehicles need to
average materials requirements per car could be cut in be small, let alone unsafe. Rather, it reflects a smaller aver-
CARS DESIGNED AND LIFETIME OF ~12 YEARS OR 230,000 KM, LIFETIME INCREASED TO 700,000 KM, MATCHING
half, from 1,200 kg to 600 kg per vehicle. Still more im- age size required whenTHEvehicles are matched
NEEDS OF INTENSIVELY to the typical
USED VEHICLES
MANAGED FOR RUN-TIME MATCHING THE NEEDS OF INDIVIDUAL OWNERS
portant, each vehicle could service much more travel. two-person trip rather than to the maximum five-person
Electric drivetrains with intrinsically
trip.
Low utilisation places limits on up-front
Today, serving one million passenger-kilometres requi- The average
investment and on benefits of longer lifetime would include vehicles
longer lifetime that are both smaller and
res the equivalent of 2.6 cars. With lifetime increased to
Limited maintenance, and disconnect
larger. The lower accident rates with
Investments in autonomous driving also
more durable materials
450,000 kilometres, occupancy increased toend-of-life
1.93, and enable (more attractive for highly
lighter vehicles, as does the greater viabilityutilised cars)
of more
between vehicles and first owners Proactive maintenance and modular
11% of components remanufactured, this falls to just 0.6. expensive and advanced materials for a vehicle
design enable re-manufacturing with higher
Putting these effects together, materials requirements fall utilisation. Likewise, the higher average occupancy would in-
to just one-eighth: from 3,200 kg to just 400 kg of mate- clude solitary trips, but some share of travel would take place
rials per million passenger-kilometres
MATERIALS IMPACT provided. in vehicles with more passengers than today.
RANGE OF SIZES, ADVANCED MATERIALS, AND
LOW UTILISATION, HIGH WEIGHT, AND LIMITED
Exhibit 5.7
safety standard Fleet management and more expensive
Circular strategies jointly reduce the materials materials incentivise re-use of parts and
EOL recovery
intensity of transport by 88%
MATERIALS PER CAR NUMBER OF CARS MATERIALS INTENSITY OF TRANSPORT

1.2 2.6 3.2
-88%
X =
0.6
0.6
0.4
OWNED CAR SHARED CAR OWNED CAR SHARED CAR OWNED CAR SHARED CAR
NOTE : PKM DENOTES PASSENGER-KILOMETRES
2050 FLEET AVERAGE IN A BASELINE VERSUS CIRCULAR SCENARIO

130
PRODUCT MATERIALS EFFICIENCY CIRCULAR BUSINESS MODELS
TONNES MATERIALS PER CAR NUMBER OF CARS
With two-thirds of travel in a shared-car
system, materials requirements can fall by
70%, cutting CO2 emissions by 40 million
tonnes per year.
131
safety standard
The Circular / Mobility
Economy – a Powerful Force for Climate MitigationFleet management and more expensive
materials incentivise re-use of parts and
EOL recovery
A CIRCULAR SCENARIO CUTS CO 2 EMISSIONS BY 70%
The above describes the characteristics of an indivi- kg to 900 kg. Average lifetime increases, from 280,000 to
dual vehicle. To understand the impact on total materials 450,000 kilometres, while average occupancy goes up by
MATERIALS
demand PERandCAR NUMBER
CO2 emissions from materials, OF neces-
it is CARS MATERIALS
18%, to 1.91 passengers INTENSITY
per car. OF TRANSPORT
sary to construct a scenario for the mobility system as a
whole. Combining these, the effect on CO2 from vehicle materials
falls by as much as 70%, or almost 40 million tonnes of CO2
1.2
We construct a circular scenario to 2050 to 2.6 3.2
explore this, per year (Exhibit 5.9). Reuse/remanufacturing, lightweigh-
with two-thirds of travel served by shared and fleet-ma- ting, and sharing all contribute some part of this, but more
naged vehicles as described above, and the remainder than half occurs because of longer lifetimes. -88%
This is thus a
by individually owned cars much like those in use today. prime example of how demand-side strategies can address
X =
In aggregate terms, the outcome is an average of a sha- CO2 emissions, and in an important sector. It is also an im-
red and an individually owned system (Exhibit 5.8). The portant element of the saturation of metal stocks discussed
0.6 per vehicle falls to 89% with rein Chapters 2 and 4. In fact, with a shared mobility system,
share of new materials
manufacturing, while the average weight falls from 1,300 stocks could fall rather than increase or saturate.
0.6
0.4
Exhibit 5.8
OWNED CAR SHARED CAR OWNED CAR SHARED CAR OWNED CAR SHARED CAR
A circular scenario reduces materials

requirements in multiple ways
2050 FLEET AVERAGE IN A BASELINE VERSUS CIRCULAR SCENARIO
PRODUCT MATERIALS EFFICIENCY CIRCULAR BUSINESS MODELS

TONNES MATERIALS PER CAR NUMBER OF CARS
REUSE AND REMANUFACTURING LIGHTWEIGHTING PROLONGED LIFETIME OCCUPANCY

% OF INPUT MATERIALS TONNES PER CAR, FLEET AVERAGE 1000 CAR KM PER CAR PASSENGERS PER CAR
1.3 450 1.91

5% 11%
1.63
0.9
280
BASELINE CIRCULAR BASELINE CIRCULAR BASELINE CIRCULAR BASELINE CIRCULAR

SCENARIO SCENARIO SCENARIO SCENARIO SCENARIO SCENARIO SCENARIO SCENARIO
REMANUFACTURED COMPONENTS
NEW MATERIALS
CARBON DIOXIDE EMISSIONS FROM MATERIALS USED IN EU PASSENGER CARS, 2050

Mt CO2, PER YEAR
132
11 60 4
9
% OF INPUT MATERIALS TONNES PER CAR, FLEET AVERAGE 1000 CAR KM PER CAR PASSENGERS PER CAR
1.3 450 1.91

5% 11%
1.63
0.9
280
Exhibit 5.9
2050 CO 2 emissions from materials fall

BASELINE
SCENARIO
CIRCULAR
SCENARIO by 70% in
BASELINE
SCENARIO a circular
CIRCULAR
SCENARIO scenario
BASELINE
SCENARIO
CIRCULAR
SCENARIO
BASELINE
SCENARIO
CIRCULAR
SCENARIO
REMANUFACTURED COMPONENTS
NEW MATERIALS
CARBON DIOXIDE EMISSIONS FROM MATERIALS USED IN EU PASSENGER CARS, 2050

Mt CO2, PER YEAR
11 60 4
9
49
21
-70%
18
CURRENT VOLUME BASELINE REUSE AND LIGHT- LONGER SHARING CIRCULAR

INCREASE SCENARIO 2050 REMANUFACTURING WEIGHTING LIFETIME SCENARIO 2050
COST REDUCTION POTENTIAL

EUR PER 1000 PASSENGER KILOMETRES
155
MATERIAL INPUT KEY DRIVERS OF LOWER COST OF OWNERSHIP
SHARING
Higher utilisation per car together with reduced cost
through autonomous car fleets lower costs per kilometre
PARTS +
CAR MANUFACTURING
PROLONGED LIFETIME
Reduced total cost per kilometre despite increased cost
for input materials
-77%
DISTRIBUTION AND M&S LIGHTWEIGHTING
133 Smaller average cars in shared fleets reduce materials
input and total cost, despite increased cost for lightweight
materials
FUEL/ENERGY
5.4 A SHARED-MOBILITY SYSTEM IS HIGHLY COST-EFFECTIVE

Although the shared-mobility system described abo- If travel becomes both cheaper and more convenient,
ve would be much more resource-efficient, the main mo- one likely response is to boost total distance travelled –
tivating factor would be its much greater economic pro- a so-called ‘rebound’ effect. The above analysis allows
ductivity. By using both inputs and products much more for this, assuming a 20% rebound effect. The rebound
intensively, the costs would be lower, and technologies could very well grow larger still if left unmanaged, but
that cut running costs (such as electric drivetrains) would policy-makers have a number of options to contain it,
be more profitable. whether through charges or regulation. As important as
the total volume of travel is the integration of shared and
This can be explored by looking at the costs of indivi- self-driving cars with other transport modes, and especi-
dual ownership vs. the shared car model (Exhibit 5.10). ally public transport. A major task for future city planners
Even when allowing for increased costs associated with will be to ensure integrated transport solutions that serve
factors such as more durable design, valuable materials, overall social objectives.
higher capital expenditure for electric vehicles, and more
advanced automation for self-driving vehicles, costs fall A shift towards a shared mobility system could also help
substantially. The extra costs are simply overpowered by mitigate many of the negative effects of today’s transport
the greatly increased utilisation, smaller size of the aver- system (Exhibit 5.11). These are very significant today.
age vehicle, and productivity gains from reuse and re- Congestion costs up to 2% of GDP in major EU cities.34
manufacturing. Comparing the circular scenario with one Every year, 26,000 people die in traffic accidents in the
dominated by individually owned vehicles, the net result EU, and another 1.5 million are injured.35 Road transport
is a reduction in the average cost per passenger-kilome- also accounts for 39% of nitrous oxide and 11% of parti-
tre of as much as 77%. This may seem like an astounding culate matter (PM2.5) air pollution in the EU, which are at
claim, but other analyses have reached similar findings.32 the root of 475,000 premature deaths each year.36 Noise
This gives a clear example of how new circular business pollution is another major factor, with costs only very par-
models can combine improvements in economic produc- tially accounted for today. Added to these ‘externalities’,
tivity with greater resource efficiency. a number of costs of transport are socialised, such as
when public funds pay for infrastructure, valuable land
The reduced cost of travel would also come with gre- is given up for roads, and parking is subsidised. Valuing
ater convenience, as passengers would not need to own these effects is notoriously tricky, and the exact amounts
their own vehicles, including the risk of value depreciation can be controversial. Nonetheless, standard estimates
and hassle of repairs and maintenance. Self-driving cars from the literature suggest externalities and other public
can also free up driving time for other purposes, reducing costs add up to more even than the total cost paid by
the trade-off between travel and either work or leisure. car owners.
134
21
-70%
9
Exhibit 5.10
18
Externalities and costs to society can be cut
significantly in a circular scenario
CURRENT VOLUME BASELINE REUSE AND LIGHT- LONGER SHARING CIRCULAR

INCREASE SCENARIO 2050 REMANUFACTURING WEIGHTING LIFETIME SCENARIO 2050
COST REDUCTION POTENTIAL

155
MATERIAL INPUT KEY DRIVERS OF LOWER COST OF OWNERSHIP
SHARING
Higher utilisation per car together with reduced cost
through autonomous car fleets lower costs per kilometre
PARTS +
CAR MANUFACTURING
PROLONGED LIFETIME
Reduced total cost per kilometre despite increased cost
for input materials
-77%
DISTRIBUTION AND M&S LIGHTWEIGHTING
Smaller average cars in shared fleets reduce materials
materials
FUEL/ENERGY
REUSE AND REMANUFACTURING
Larger share reduces input materials and average cost of
materials, modular design makes maintenance cheaper
MAINTENANCE
35
CAPITAL COST,
INSURANCE,
PARKING
CURRENT CIRCULAR
SCENARIO 2050
EXTERNALITIES AND PUBLIC COSTS OF CAR TRANSPORTATION

264
SOURCE : MATERIAL ECONOMICS ANALYSIS. 33
REDUCED EXTERNALITIES AND PUBLIC COST
REDUCED NUMBER OF CARS

135
Due to higher occupancy and increased utilisation
result in improved air quality, reduced noise level, less
WAITING TIME
congestion and less need for inner-city parking
Autonomous vehicles offer the prospect

of reducing accidents sharply, as more
than 90% of accidents are attributable
to human error.
A shared mobility system would help reduce many of Like with rebound, this outcome depends on carefully ma-
these costs. The total number of cars required to serve naging the transition to a shared system. For example, early
a given travel need would be lower, because of higher experiments with car sharing have resulted in more conges-
occupancy and sequential sharing over the course of tion rather than less in some cities, by taking passengers out
a day. Autonomous vehicles offer the prospect of redu- of public transport systems and into shared cars.40 Individual
cing accidents sharply, as more than 90% of accidents car-sharing services have also been controversial at times,
are attributable to human error.37 Self-driving and con- mostly for reasons unrelated to the potential benefits we des-
nected cars can also optimise traffic flows, drive more cribe above (e.g. tax status, employment rights, impact on ex-
densely than is possible with human drivers, and do isting taxi services, etc.). However, these are particular to indi-
not need parking space in city centres – thus reducing vidual situations and represent at most incremental change;
congestion as well as land use requirements for roads there is little reason to think that these are intrinsic to shared
and parking.38 By speeding up the adoption of electric mobility. Nonetheless, they serve as a reminder that large
vehicles, a shared car system would also contribute to transitions in fundamental systems such as personal mobility
reduced air and noise pollution. A systematic analysis often unfold in bumpy ways. Policy-makers and business le-
of different factors suggests as much as 70% of exter- aders alike need to be alert to problems that may arise. For
nalities and public costs could be eliminated (Exhibit all that, the intrinsic higher productivity and other benefits of
5.11). such a system are so large that it is well worth aiming for.
136
Smaller average cars in shared fleets reduce materials
materials
FUEL/ENERGY
REUSE AND REMANUFACTURING
Larger share reduces input materials and average cost of
materials, modular design makes maintenance cheaper
MAINTENANCE
Exhibit
35 5.11
Externalities and cost of society can be significantly

CAPITAL COST,
reduced in a circular scenario by 2050
INSURANCE,
PARKING
CURRENT CIRCULAR
SCENARIO 2050
EXTERNALITIES AND PUBLIC COSTS OF CAR TRANSPORTATION

264
REDUCED EXTERNALITIES AND PUBLIC COST
REDUCED NUMBER OF CARS

Due to higher occupancy and increased utilisation
result in improved air quality, reduced noise level, less
WAITING TIME
congestion and less need for inner-city parking
AUTONOMOUS AND CONNECTED CARS

Are safer, reduce need for signs, lanes and other
-74% infrastructure and can optimise traffic flows, which
reduces congestion and land use for roads
ELECTRIC VEHICLES
ACCIDENTS Reduce noise level and have positive impact on air
quality, especially when shifting towards renewable
AIR POLLUTION energy sources
NOISE
INFRASTRUCTURE
68
PARKING COSTS
LAND
CURRENT CIRCULAR
SCENARIO 2050
SOURCE : MATERIAL ECONOMICS ANALYSIS. 39
MAJOR SHIFT OF INNOVATION FOCUS

137
FROM MAXIMIZING UPFRONT SALES VOLUME AND PRICE,
TO MAKING THE CAR A WELL-FUNCTIONING EFFECTIVE PART OF URBAN MOBILITY
HOW TO GET THERE – A POSITIVE INNOVATION SPIRAL

Although car-sharing is growing strongly, it has only Nonetheless, there are some areas for policy-makers to
taken the first tentative steps in the EU.41 Even if it is an pay attention to. First, the more that air pollution, conges-
end-state worth aiming for, is there a credible path from tion and other negative effects of today’s mobility system
today’s system to the one described above? are discouraged, the more force there will be behind a
move to a shared system with lower such impacts. Se-
The key to unlocking this is a set of self-reinforcing dy- cond, policy-makers can ensure that various other areas
namics – a positive innovation spiral (Exhibit 5.12). The of regulation are adapted so they account for shared mo-
shared-car system requires vehicles that are designed bility. One example could be data policy, to safeguard
and operated to different specifications, with joint inno- end-user privacy in what will be a highly data-intensive
vation in materials, design, manufacturing, and business business model. Another is to promote the emergence
models. Once started, these changes become self-re- of effective insurance solutions for shared cars, and yet
inforcing in multiple dimensions. Higher utilisation ena- another to clarify and strike the right balance for safety
bles earlier adoption of costlier automation technologies. and liability regulation for self-driving vehicles. Third, it is
These in turn enable business models based on sha- important not to let the early disappointments and contro-
ring, in turn boosting utilisation further. Once a business versies surrounding some of today’s car-sharing muddy
model for mobility-as-a-service takes root, it creates the the waters; the potential benefits of shared mobility are
precondition for fleet-managed vehicles, enabling better extensive and real, whatever the precise performance of
maintenance, component reuse, remanufacturing, and early experiments.
more control over end-of-life flows. Longer lifetime in turn
becomes an enabler of the intensive use underpinning Finally, city authorities will have an important task to
a sharing business model, which reinforces the higher incorporate shared mobility with other important city ini-
utilisation that helps motivate the higher upfront costs of tiatives. Given the last shift ahead, they can make clear
more durable vehicle design. their support for the transition – such as the city of Hel-
sinki’s stated ambition to enable a fully functional sys-
In multiple ways, then, the overall shift is away from max- tem of ‘mobility on demand’ to provide a service equi-
imising upfront sales volumes and prices, and towards ma- valent to that of private cars, by 2025. As noted, it will
king the passenger car a well-functioning and effective part be important to integrate shared mobility solutions with
of urban mobility. Technology development and innovations public transport. City planning will also need to adapt as
within the shared car segment are likely to have positive patterns of mobility change in response to the different
spillovers for other vehicles, leading to even larger producti- circumstances of shared car system.
vity gains. The agenda for industry is relatively clear.
The other major shift required is in expectations and * * *

attitudes. Today, car ownership is bound up with many
things besides travel: cars have symbolic value, signal- In its 2011 transport White Paper, the European Com-
ling a range of attributes from status to freedom. Partly mission wrote: “Since the first big oil crisis 40 years ago
for that reason, even the circular scenario sees one-third – despite technical progress, potential for cost-effective
of travel continue in individually owned cars. At the same energy efficiency improvements and policy efforts – the
time, norms shift over time, and the greater convenience transport system has not fundamentally changed” – no-
and lower cost of shared and autonomous mobility are ting the challenges this creates for greenhouse gas emis-
powerful forces in a perspective to 2050. sions, cost, oil import dependence, congestion, pollution
and more. The possibility of a shared mobility system
The final area of change is policy and regulations. now offers the prospect of such fundamental change –
Though much is at stake for public policy objectives, poli- with significant gains for resource efficiency, lower costs,
cy may in fact have a relatively limited role to play. Private public health, and the environment. A more circular eco-
and public incentives are, on the whole, powerfully aligned. nomy can be a major force for a better transport system.
138
Exhibit 5.12
Sharing can catalyse a positive spiral of new innovation

opportunities for passenger cars
MAJOR SHIFT OF INNOVATION FOCUS
FROM: Maximizing upfront sales volume and price

TO: Making the car a wellfunctioning effective part of urban mobility
VEHICLES RE-DESIGNED AND

HIGHER UTILISATION PER CAR MANAGED TO MAXIMISE RUN TIME
SHARING BUSINESS MODELS LONGER VEHICLE LIFETIMES

New business models to suit new High returns to durability
customer groups Longer-lived electric vehicles profitable
Integration with public transport system Professionally maintained fleet
DIGITISATION AND AUTOMATION MODULARITY AND REUSE

Self-driving cars enables new service Design for quick repair and upgrade-
models ability
Data-intensive optimisation of traffic Reuse built into vehicle design
LOWER VEHICLE WEIGHT

More varied car sizes with shared fleet
Advanced materials more profitable
HIGHER PROFITABILITY OF
CIRCULAR BUSINESS MODELS
END-OF-LIFE VALUE
Higher EOL value (modularity,
valuable materials etc.)
EOL flows more predictable in fleet
owned system
LOWER COST OF TRANSPORT

New design and high utilisation reduce
total cost
Competitiveness with other modes of
transport increases
139
The Circular Economy – a Powerful Force for Climate Mitigation / Buildings and cement
6. Buildings
and cement
more value from less materials
The EU uses as much as 1.6 billion ton- ponents, on site or within local markets.
nes (Gt) of materials for buildings per year, Improved design of building components,
and that amount is likely to grow as building new construction techniques, and use of
stock continues to slowly expand, and as high-strength steel and concrete could re-
large numbers of post-war buildings require duce the materials needed for new buil-
substantial renovation or replacement in the dings. Increased standardization, impro-
decades ahead. The CO2 footprint of the- ved planning, and appropriate storage and
se materials is significant, and in the EU transportation could reduce waste. New bu-
countries that have gone furthest in impro- siness models for buildings in which spa-
ving energy efficiency and decarbonising ces are shared, e.g. co-working spaces,
heat, construction now accounts for as could reduce the total built area needed.
much as half of the lifetime CO2 footprint of Finally, there is potential to recycle cement
a building (and the use phase for the rest).1 from demolition, by reprocessing concrete
to recover some unreacted clinker.
How we use construction materials
therefore matters greatly for future clima- Cost-effective strategies may be under-
te targets. By 2050, just the cement, steel, used because of diverging interests: the
aluminium and plastic used for construction party making decisions is not the one who
will result in emissions of 230 Mt CO2 in a would benefit (similar to how landlords have
baseline scenario where they are made with little incentive to invest in energy efficien-
today’s production processes. Demand-side cy if their tenants pay the energy bills).
measures could reduce this by more than In other cases, the business case is not
half, or 123 Mt CO2, by the second half of this strong enough yet, but could improve with
century. Of this, 80 Mt CO2 per year would be new technology. Increased digitalisation
available by 2050, making a major contribu- of the construction process will be a key
tion to EU mid-century climate targets. factor for the adoption of circular opportuni-
ties, through the use of building information
Key measures that could reduce de- modelling, and the gradual automation of
mand for building materials include design more of the construction process. Both of
changes to increase buildings’ longevity these can significantly reduce the cost of
and adaptability; disassembly at the end techniques to reduce materials use.
of life; and reuse of intact structural com-
140
Increased digitalisation of the construction
process will be a key factor for the adoption
of circular opportunities, through the use of
building information modelling system and
increased automation.
141
6.1 UNDERSTANDING FUTURE BUILDING NEEDS

We spend most of our time in buildings. EU residents targets, the EU will also have to retrofit buildings to improve
use an average of 48 m2 of floor area – for housing, of- their energy efficiency at two or three times today’s pace.5
fices, education, lodgings and other purposes; housing
costs alone consume 22% of disposable income.2 The ef- Even though the turnover rate for buildings is slow,
fort that goes into creating and maintaining this building construction is a major user of materials. Buildings ac-
stock also is a major economic sector, accounting for count for two-thirds of cement use, more than a third of
8.6% of the EU’s GDP and employing 18 million people.3 steel, a quarter of aluminium, and almost 20% of plastics.7
Adding aggregates and other materials such as bricks,
The building stock changes only slowly, so the decisions gypsum, lime and copper, buildings alone use some 1.6
made today have very long-term consequences. Each year, billion tonnes of materials per year, as noted below. Pro-
1% of EU buildings are renovated, and new buildings are ducing these materials, in turn, results in about 250 mil-
added that make up 1–1.5% of the total stock.4 The EU’s lion tonnes (Mt) of CO2 emissions annually (Exhibit 6.2).
building stock is aging – 45% of buildings are over half a Cement, steel, aluminium and plastics account for almost
century old, and a large share of housing was built just after 80% of those emissions. Construction and demolition (in-
World War II – so in the coming years, significant renovations cluding infrastructure) also generate 25–30% of total was-
and upgrades will be needed (Exhibit 6.1). To meet climate te volumes,8 far more than any other sector.
Exhibit 6.1
The building stock evolves slowly, with 1% renovation

and 1–1.5% new construction per year
AGE DISTRIBUTION OF BUILDING STOCK, 2013
% OF TOTAL BUILDING STOCK, EUROPE
AFTER 2010 3%
2000-2010 12%
T 6.1
1990-1999 10%
1980-1989 13%
New buildings represent

approximately 1.0 - 1.5%
1970-1979 17% of total stock
mmet
int Renovation represents
approximately~1%
of stock per year
1945-1969 24%
BEFORE 1945 21%
SOURCE : ANALYSIS BASED ON MULTIPLE SOURCES 6
142
MATERIALS COMPOSITION IN BUILDINGS CO2 FOOTPRINT OF 250 Mt PER YEAR CONSTRUCTION AND DEMOLITION
1980-1989 13%
New buildings represent

approximately 1.0 - 1.5%
1970-1979 17% of total stock
Renovation represents
approximately~1%
of stock per year
1945-1969 24%
Exhibit 6.2
EU building21%materials have a CO 2 footprint

BEFORE 1945
of 250 million tonnes per year
MATERIALS COMPOSITION IN BUILDINGS CO2 FOOTPRINT OF 250 Mt PER YEAR CONSTRUCTION AND DEMOLITION WASTE
% OF MASS, AVERAGE BUILDING1 Mt CO2 PER YEAR; % Mt PER YEAR; %
KS OTHER 3
~250 Mt ~2500 Mt
4%
ALUMINIUM
0.5% 7%
STEEL 2% OTHER 3 22%
PLASTICS
0.2%
OTHER ~1600
CONCRETE 42% (65%)
CEMENT 30%
ALUMINIUM 12%
PLASTICS 11%
AGGREGATES2 45%
C&D WASTE
~900
(35%)
STEEL 25%
LIFECYCLE CO2 EMISSIONS FROM BUILDINGS

% OF LIFETIME CO2 EMISSIONS DURING LIFETIME
NOTE : 1 DATA FOR MATERIALS COMPOSITION OF BUILDINGS INCLUDE RESIDENTIAL, PUBLIC AND COMMERCIAL BUILDINGS, BUT NOT INDUSTRIAL,
2 EXCLUDING AGGREGATES USED IN CONCRETE, 3 INCLUDING WOOD, STONE, GLASS, AND CLAY
SOURCES : ECORYS AND COPENHAGEN RESOURCE INSTITUTE (2014) , EUROSTAT (2017) .
143
USE PHASE
INCLUDING ENERGY
The Circular Economy –ALUMINIUM
12%
a Powerful Force for Climate Mitigation / Buildings and cement
PLASTICS 11%
AGGREGATES2 45%
Exhibit 6.3
~900
Materials will be a25%major future source(35%)
C&D WASTE
STEEL
of CO 2 emissions from buildings
LIFECYCLE CO2 EMISSIONS FROM BUILDINGS

% OF LIFETIME CO2 EMISSIONS DURING LIFETIME
USE PHASE
INCLUDING ENERGY
50%
8-15%
MATERIALS AND
CONSTRUCTION
WITH CURRENT ENERGY SYSTEM WITH LOW-CARBON ENERGY

(UK AND BELGIUM EXAMPLES) (SWEDEN EXAMPLE)
GLOBAL CO2 EMISSIONS FROM CEMENT PRODUCTION CUMULATIVE CO2 EMISSIONS, 2015-2100
Gt CO2 PER YEAR Gt CO2
SOURCE : IVA (2014) , ROSSI, MARIQUE AND REITER (2012) , CUÉLLAR-FRANCA AND AZAPAGIC (2012) 13
4 254
To date, most discussions of the CO2 impact of buil- emissions, the CO2 ‘embodied’ in materials will become
3
dings have focused on the use phase, especially on ever more important. In countries where electricity
184 and
energy consumption. This is understandable: buildings heating systems are already low-carbon, building mate-
account for 40% of final energy demand in the EU, and rials already account for half of the total CO2 footprint of
36%2 of CO2 emissions. Still, around 15% of total life-
10
new buildings.11 Other EU countries will follow a similar
cycle CO2 emissions from EU buildings today are attri- route as their energy systems decarbonise. Thus, it is
butable to materials and construction (Exhibit 6.3). As high time that discussions about sustainable buildings
energy efficiency and low-carbon energy reduce energy pay attention to building materials.12
1
0
2020 2030 2040 2050 2060 2070 2080 2090 2100 BASELINE SCENARIO BASELINE SCENARIO
144 + LOW CARBON ENERGY
BASELINE SCENARIO
It is high time that discussions
about sustainable buildings pay
attention to building materials.
145
EMISSIONS FROM CEMENT PRODUCTION

Cement accounts for a large share of the embodied large chemical process emissions would remain. Even
emissions in buildings, and is a major source of global fully decarbonised energy leaves in place 2 Gt per year
CO2 emissions. The chemical reaction involved in the of CO2 emissions from cement production, claiming 184
production of clinker (the main ingredient of cement) re- Gt of the global CO2 budget to 2100.
sults in some 1.4 billion tonnes of CO2 per year global-
ly.14 Another 1 Gt CO2 results from the large amounts of In the face of such daunting numbers, many low-car-
energy required to produce cement. The total of 2.4 Gt bon scenarios assume carbon capture and storage as
CO2 corresponds to 7% total global CO2 emissions from the main way to address future cement emissions.18 Oth-
energy and industry.15 ers see great potential in materials substitution, by using
existing alternatives to clinker,19 or switching to polymer
Looking ahead, global cement production is set to grow cements or other innovative options.20
further. Many developing countries still have stocks of 5
tonnes per person or less, less than a quarter of many Cement production is a significant source of emissions
developed countries.16 To explore the emissions implica- in the EU as well: 114 Mt CO2 per year today. As noted
tions of global cement use for CO2 emissions, we model- above, construction activity and demand are likely to in-
led a scenario in which demand rises to 7 billion tonnes crease somewhat. Even with incremental process impro-
per year by 2100.17 Even using the best available techni- vements, emissions in 2050 therefore would be similar to
ques for production, total emissions would rise to almost today’s, at 113 Mt CO2 per year. Given this challenge, any
4 Gt CO2 per year, with the cumulative emissions of more opportunity to reduce the demand for cement – without
than 250 Gt by 2100 (Exhibit 6.4). A rapid switch to compromising on well-being – could make an important
low-carbon energy in cement production would help, but contribution to EU climate targets.
146
USE PHASE
INCLUDING ENERGY
50%
Exhibit 6.4
Emissions from global cement production risk claiming most of the

carbon budget available for all materials production
8-15%
MATERIALS AND
CONSTRUCTION

(UK AND BELGIUM EXAMPLES) (SWEDEN EXAMPLE)
Gt CO2 PER YEAR Gt CO2
4 254
3 184
0
2020 2030 2040 2050 2060 2070 2080 2090 2100 BASELINE SCENARIO BASELINE SCENARIO
+ LOW CARBON ENERGY
BASELINE SCENARIO
BASELINE SCENARIO + LOW CARBON ENERGY
MATERIALS RECIRCULATION BUILDING MATERIALS EFFICIENCY CIRCULAR BUSINESS MODELS

TONNES CO2 PER TONNE MATERIALS TONNES MATERIALS PER m2 USEFUL SERVICE FROM EACH m2
CEMENT RECYCLING AND VALUE REDUCED WASTE IN CONSTRUCTION PROLONG LIFETIME

CAPTURE AT END OF LIFE Reduced loss of materials during Adapt and renovate of buildings to avoid
Improved waste management, better construction through improved materials demolition and new build
separation of materials and increased management Improved maintenance to extend lifespan
recycling of key components
Less downcycling of concrete through MATERIALS EFFICIENCY OF Design for flexible use and to enable deep
adoption of new technologies to recover renovation
BUILDINGS
unused cement and aggregates
”Design for disassembly” as a key principle Less material per building through less
SHARING TO REDUCE FLOOR SPACE
over-specification, improved design,
and high-strength materials
REQUIREMENTS
Reduce total area required through
REUSE OF BUILDING COMPONENTS mechanisms such as peer-to-peer sharing,
CONSTRUCTION

(UK AND – a Powerful
BELGIUM Force for Climate Mitigation
EXAMPLES) / Buildings
(SWEDEN EXAMPLE)and cement
6.2 THE CIRCULAR ECONOMY OPPORTUNITY IN BUILDINGS

GtThe PER YEAR economy opportunities in buildings are in
CO2 circular materials; Gt
andCOreusing
2 buildings and building components.
three main categories (Exhibit 6.5): • Circular business models: Increasing the useful ser-
• Materials recirculation: Recycling materials from vice from materials, by extending the lifetime of buildings,
4end-of-life buildings, and designing and dismantling and/or by increasing
254 the utilisation of floor space through
buildings so that high-value recycling is possible (the re- sharing and other mechanisms.
circulation of steel, plastics, aluminium was discussed in Together, these opportunities could significantly redu-
Chapters 2–4, so we focus on cement below). ce the amount of construction materials needed
3 184 in the EU.
• Building materials efficiency: Reducing the amount of Before discussing them in detail, however, it is useful to
materials that are required for a given floor area. The main consider some of the reasons why materials use is so inef-
opportunities are to reduce waste during the construction
2
ficient today. After all, building materials constitute a major
process; reduce the amount of material in each building by part of the total cost of construction, so there should al-
avoiding over-specification and using user higher-strength ready be strong economic incentives to optimise their use.
0
2020 2030 2040 2050 2060 2070 2080 2090Exhibit
2100 6.5 BASELINE SCENARIO BASELINE SCENARIO
+ LOW CARBON ENERGY
A broad palette of circular economy opportunities

BASELINE SCENARIO
can improve materials use in buildings
BASELINE SCENARIO + LOW CARBON ENERGY
MATERIALS RECIRCULATION BUILDING MATERIALS EFFICIENCY CIRCULAR BUSINESS MODELS

TONNES CO2 PER TONNE MATERIALS TONNES MATERIALS PER m2 USEFUL SERVICE FROM EACH m2
CEMENT RECYCLING AND VALUE REDUCED WASTE IN CONSTRUCTION PROLONG LIFETIME

CAPTURE AT END OF LIFE Reduced loss of materials during Adapt and renovate of buildings to avoid
Improved waste management, better construction through improved materials demolition and new build
separation of materials and increased management Improved maintenance to extend lifespan
recycling of key components
Less downcycling of concrete through MATERIALS EFFICIENCY OF Design for flexible use and to enable deep
adoption of new technologies to recover renovation
BUILDINGS
unused cement and aggregates
”Design for disassembly” as a key principle Less material per building through less
SHARING TO REDUCE FLOOR SPACE
REQUIREMENTS
office sharing, and more communal space
Reuse of structural components at end
of life and during renovations through
new local markets
LABOUR PRODUCTIVITY IN CONSTRUCTION VERSUS OVERALL ECONOMY

INDEXED: 100 = 2001, EU28
148
115
”Design for disassembly” as a key principle SHARING TO REDUCE FLOOR SPACE
REQUIREMENTS
office sharing, and more communal space
Reuse of structural components at end
of life and during renovations through
new local markets
Exhibit 6.6
The construction sector faces a major

productivity challenge

INDEXED: 100 = 2001, EU28
115
OVERALL ECONOMY
110
105
100
CONSTRUCTION
95
90
85
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
SOURCE : EUROPEAN COMMISSION (2015) 22

CO2 EMISSIONS FROM MATERIALS USED IN BUILDINGS
Mt CO2 PER YEAR STEEL ALUMINIUM
PLASTICS CEMENT
However, the231 construction 13 industry faces significant For these reasons, and because of the social and econo-
challenges.21 First of all, it is fragmented, comprising nu- mic importance of buildings, regulators often intervene heavily
10
merous,
208 often small companies. It is cyclical and volatile,
20 in the sector. The long chain of actors and time lag between in-
and EU construction companies are still recovering from itial construction and eventual occupation (let alone demolition)
the sharp downturn that came with the financial crisis. mean -80 (-34%)
24 that commercial contracts can rarely cover all aspects of
The workforce is relatively unstable, and profit margins public importance. Regulations and building standards cover -123 (-53%)
and capitalisation are lower than in many other sectors. 13 safety, energy performance, aesthetics, per-
many issues – fire
The stakeholders and value chain are fragmented, with mitting, construction products, structural safety, etc. However,
complex contracting across multiple parties that often materials efficiency is not yet included in such regulations.
results in inefficiencies and poorly aligned incentives.
Construction has also been slow to adopt new technolo- Another important theme is that many barriers could mi-
gies or methods; there has been no analogue of manu- tigated by increased digitalisation. High transaction costs,
facturing’s ‘lean’ improvements. For all these reasons, risk, and labour costs are three of the main factors in the
unlike in any other sector of comparable economic current industry structure that lead to the inefficient use of
importance, productivity improvements in construction materials, and digitalisation-related strategies could help
have stalled (Exhibit 6.6). address them, as discussed below.
149
MATERIALS EFFICIENCY IN THE CONSTRUCTION PROCESS

The main reasons for overuse of materials are was- steel can reduce materials needs and CO2 footprint by
te during construction, over-specification in the building, 15–20% in some building types.26 There also are benefits
and discarding instead of reusing structural elements be- from ultra-high strength concrete in some applications.27
fore their end of life.
Reuse of components: Many of the components of
Waste of materials during construction: Industry in- buildings are still fully functional even when the building
siders indicate that as much as 15% of buildings materi- reaches its end of life, creating an opportunity for reuse,
als are wasted in the construction process, ranging from especially of concrete and steel components. There are
10% in the best cases to 20% or more in the worst. already some examples of successful reuse. A new resi-
dential area of Copenhagen is being built out of structural
There are many reasons for this waste. Contracts are elements from abandoned houses28, while a German re-
often structured in such a way that those making deci- sidential area near Berlin similarly reused precast concre-
sions about materials use do not bear the cost of any te elements, reducing the cost of construction by 30%.29
waste – but they do bear the cost of delays, so they order Although these examples demonstrate the viability of reu-
extra materials as a buffer. Due to lack of standardisation se, they are rare exceptions, and the amount of concrete
and local markets, however, surpluses often cannot be being reused is still tiny relative to new concrete demand.
sold on or returned. Many construction companies do not There also are many examples of successful reuse of ste-
even have firm data on the amount of materials they was- el30, but overall, only 5% of end-of-life steel from buildings
te, showing how little attention the issue has received. is in fact reused, a small share of the total potential.
Over-specification of materials inputs: Up to 85% of A range of barriers stand in the way.32 The most im-
the materials in a building are in the structural elements, mediate obstacles are availability and coordination: even
with steel and concrete making up the large majority.23 when reusable parts are in fact available, there often are
There is a widespread tendency for these elements to be no local markets capable of matching connecting those
over-specified, using more materials than are needed for dismantling buildings with those who could use the parts,
a sound and safe structure. This is particularly true with or of storing the parts until someone can use them. In-
steel: detailed case studies show its use could be cut by consistent or unpredictable supplies of materials could
half without compromising on design standards.24 This is result in costs or delays for other parts of a building pro-
an astounding number, with immediate significance for ject, and the quality may be difficult to trace or certify. As
total steel use; as noted above, construction accounts important, the practice is simply unfamiliar, at odds with
for more than half of total steel demand. It also shows established routines. Industry experts also say there is
clearly how little attention there is in today’s construction little demand from clients to test them.
process to efficient materials use.25 This also suggests
that greater use of digital tools and automated assembly To overcome these barriers, it would be necessary
could do much to address this issue. both to ‘design for deconstruction’, and to enable much
closer tracking of materials in- and outflows to the buil-
High-strength materials offer another route to reduce ding stock.
the total use of materials. For example, high-strength
150
Steel use for buildings could be
cut by half without compromising
on design standards.
151
MAKING MORE OF EACH BUILDING: INCREASED LIFETIME AND UTILISATION

Increasing the building lifetime is one of the single most Above all, a longer life for buildings will require a diffe-
effective actions that can be taken to reduce the need for rent mindset, where buildings are intended and expected
construction materials. As noted above, around 85% of the to last, and to be reconfigured multiple times over long
CO2 footprint of materials is associated with structural ele- time periods. This may require public intervention, as the
ments. Yet when buildings are demolished, it is typically not societal benefits of ‘build to last’ principles are felt only a
because their structures are no longer usable in any technical long time in the future.
sense – most are perfectly sound.33 Instead, old buildings are
razed for social or economic reasons: because architectural The need for materials in buildings also can be redu-
tastes have changed, new types of housing or commercial ced by making more of the building stock that is already
space are needed; the neighbourhood has evolved; or the in place. Sharing offers one option. For example, much
cost of renovating the building is higher than demolition and European office space is occupied only for 40% of the
new construction. There is also significant variation in how time35, and there is a trend both towards more compact
long buildings stand: some low-quality developments are torn and shared offices. Services such as Airbnb also offer
down only some decades after construction, whereas at the the possibility of using spare or idle residential space
other extreme many European cities take pride in high-quality, more effectively.
beautiful buildings that may have stood for centuries. Despite
what is actually possible, the technical lifetime of new buil- In the residential sector, the opportunity is different.
dings is often assumed to be just 50 years.34 One aspect is a mismatch of space and occupancy.
Many European countries have long had a discussion
This suggests a major agenda to extend the life of buil- about tax systems discouraging downsizing from large
dings. The key is for buildings to be adaptable to the chan- to small houses. Architects also have highlighted the
ging requirements of their users: whether they change in opportunity to make better use of communal space
function (e.g. from commercial to residential), or in appea- in new residential developments – perhaps chiefly to
rance and design. This requires high quality and an emp- create a more socially attractive place to live, but also
hasis on design that is modular and enables deep renova- making it possible for each individual dwelling to be
tion. In addition, improved maintenance can extend lifetimes smaller without compromising on residents’ require-
by preventing the deterioration of facades and other parts, ments.
which in turn become the trigger for wholesale demolition.
152
IMPROVED END-OF-LIFE
DISASSEMBLY FOR HIGH-VALUE
REUSE AND RECYCLING
The building stock in Europe consti-
tutes an enormous repository of highly
specialised materials: a range of diffe-
rent metals, specific alloys or composi-
tes, plastics with a wide range of diffe-
rent properties, insulation materials, and
much more. The prospect of ‘buildings
as a materials bank’ expresses the no-
tion that, properly managed, end-of-life
buildings could become a major source
of materials for future needs.
The starting point is very different,

however. Standard practice when buil-
dings reach their end of life is demolition.
Although official statistics state that 47%
of EU construction and demolition waste
is reused or recycled, this stretches any
meaningful definition of recycling: most-
ly, materials with high embodied energy
and CO2 are turned into bulk aggrega-
tes or backfill, sometimes at no net CO2
benefit.36 Concrete may be crushed to
recover reinforcing steel, and then either
landfilled or used as aggregates. Steel
elements are typically remelted as scrap,
or lost. As noted in Chapter 3, some
plastics from buildings are recycled, but
far from the full potential. Except for me-
tals and some plastics recycling, current
practice is at best very heavy downcy-
cling, combined with no small share of
landfilling.
To change this, several things must

happen. One is to properly track the
content of buildings through ‘materials
passports’ and logbooks, so that the po-
tential for materials recovery is known at
end of life, and to track changes during
a building’s lifetime. Another is to design Properly managed, buildings
buildings for deconstruction, as noted abo-
ve, and otherwise for separation into clean could be a major materials
streams of high-value materials. The demo-
lition process would need to change, as to- bank for future use..
day’s speed-focused methods often dama-
ge or mix the materials to the point where
they can only be recovered as aggregates.
153
LOWER CO 2 INTENSITY OF BUILDING MATERIALS AND CEMENT RECYCLING

The remaining key strategy to reduce CO2 emissions lands has found in pilots that it is possible not only to
from buildings is to use materials with lower CO2 emis- recover unreacted cement, but also to retrieve aggrega-
sions per tonne. One demand-side option is to substitute tes from end-of-life concrete with improved quality relative
from high-carbon to low-carbon materials (see Box 1 for to new aggregates.38 In turn, the recovered aggregates
two prominent examples). The other is to recycle materi- with improved strength require up to 15% less cement
als (here we focus on cement, but see Chapters 2–4 for than new aggregates. Overall, the recovered cement can
discussions of recycling steel, plastics and aluminium). replace up to 80% of new cement in construction, saving
almost half of the CO2.39
Cement cannot be recycled in the same sense as metals
or plastics, as the process of making cement clinker is irrever- These technologies have yet to enter into commerci-
sible. Nevertheless, there are opportunities to recover unused al use at scale. In addition to technology development,
cement from concrete at end of life. As much as 30–40% of it is necessary to set up the regulatory framework and
the clinker often remains unused (or unhydrated)37 and can adapt standards to accommodate recycled materials. Gi-
therefore in principle be used again to replace new cement. ven transport costs, markets for recovered cement would
mostly be local, or even integrate the demolition of exis-
Technologies to capture these opportunities are under ting structures directly into the production of raw materi-
development. For example, SmartCrusher in the Nether- als for replacement buildings.
154
BOX 1: MATERIALS SUBSTITUTION IN BUILDINGS
Clinker substitutes in Portland cement clinker-based cements: There is ongoing research to in-
vestigate options for reducing the CO2 footprint of cement. One track looks to entirely novel cements,
either based on polymers, or in part synthesised from CO2. More near-term efforts focus on substitu-
ting clinker with other raw materials, since the process emissions of cement production occur during
the limestone-to-clinker process. The most common clinker alternatives are granulated blast-furnace
slags and coal ash. Together, these have an estimated potential to replace up to 15–25% of clinker.
However, neither steel blast furnace production nor coal-fired power production are expected to
grow much (indeed, phasing out coal power is crucial to climate objectives), whereas global cement
production is expected to increase significantly. Another promising option is to use calcinated clays,
with the potential advantage of relatively low cost and high availability. This is still under develoip-
ment, but some investigations point to the possibility of replacing as much as 50% of clinker without
compromising on cement quality.40
Wood-based construction: Wood-based construction components can have markedly lower CO2
intensity than steel and concrete, and even constitute a carbon sink if managed well. Timber can
have a strength similar to that of reinforced concrete (hardwood is slightly stronger and softwood is
slightly weaker), but it is less stable and less able to handle compression, and it poses higher fire
risks. Nonetheless, innovation is extending the range of potential applications for wood in buildings.
For instance, studies of the UK construction sector have shown that novel off-site, modular timber
frame systems can save up to 50% of embodied carbon and 35% of embodied energy compared
with traditional residential building methods and materials.41 Cross-laminated timber (CLT) is another
promising substitute for concrete, particularly suited to multi-storey buildings. CLT advantages of
traditional timber include improved load bearing and shear capacity, and greater potential for prefa-
brication. A common concern is the shorter lifecycle of timber compared with long-lasting concrete.
However, buildings are rarely demolished due to degradation of the main structure. There are also
potential treatments to extend service life, such as coating, impregnation, chemical/mechanical mo-
dification, and design details that limit the exposure to wetting and direct sunlight.42 There is a need
to analyse the realistic potential to use wood at scale in EU construction – an important gap to fill in
considering more sustainable buildings.
NOTES : PKM DENOTES PASSENGER-KILOMETRES

SOURCE : MATERIAL ECONOMICS ANALYSIS AS DESCRIBED IN TEXT
155
6.3 A 100CIRCULAR SCENARIO FOR BUILDINGS CAN CUT CO 2

INDEXED: = 2001, EU28
EMISSIONS BY MORE THAN HALF

115
OVERALL ECONOMY
The above discussion lays out a rich set of options to from materials for buildings are likely to grow from 208 Mt
reduce
110
the demand for materials in construction without CO2 today to just over 230 Mt CO2 per year by 2050. The
compromising on the benefits derived from buildings. To circular opportunities jointly could cut emissions by 80 Mt
explore the impact of those options on CO2 emissions, CO2 by 2050, and a further 43 Mt CO2 later in the century,
105 a circular scenario with the gradual adoption of a
we built as the effect of longer building lifetimes takes hold.43 The
wide-ranging palette of circular opportunities in buildings largest reductions are from materials efficiency strategies
by 2050.
100 The analysis shows that CO2 emissions from ma- (less waste, less over-specification, use of high-strength
terials in buildings could be reduced by as much as half materials) and reuse. The total reduction of 123 Mt CO2
CONSTRUCTION
(Exhibit 6.7). Without further action, the total emissions thus cuts emissions by more than half.
95
90
Exhibit 6.7
85
2001 2002
A2003circular
2004 2005
scenario
2006 2007
reduces
2008 2009
CO
2010
2
emissions
2011 2012 2013
from building materials by 53%
CO2 EMISSIONS FROM MATERIALS USED IN BUILDINGS

Mt CO2 PER YEAR STEEL ALUMINIUM
PLASTICS CEMENT
231 13
10
208 20
24 -80 (-34%)
-123 (-53%)
13
CURRENT REFERENCE CEMENT WASTE REUSE OF MATERIALS SHARING CIRCULAR PROLONGED CIRCULAR
SCENARIO RECYCLING REDUCTION BUILDING EFFICIENCY SCENARIO LIFETIME 1 SCENARIO
2050 COMPONENTS 2050 (BEYOND 2050)
156
RECIRCULATION OF MATERIALS BUILDING MATERIALS EFFICIENCY CIRCULAR BUSINESS MODELS

CURRENT REFERENCE CEMENT WASTE REUSE Exhibit 6.8 SHARING
OF MATERIALS CIRCULAR PROLONGED CIRCULAR
SCENARIO RECYCLING REDUCTION BUILDING EFFICIENCY SCENARIO LIFETIME 1 SCENARIO
A circular scenario sees improvement across

2050 COMPONENTS 2050 (BEYOND 2050)
all aspects of building materials use
RECIRCULATION OF MATERIALS BUILDING MATERIALS EFFICIENCY CIRCULAR BUSINESS MODELS

t CO2 PER t MATERIALS TONNES MATERIALS PER M2 USEFUL SERVICE FROM EACH m2
CEMENTS RECYCLING REDUCED WASTE, REUSE OF BUILDING COMPONENTS, SHARING PROLOGNED

t CO2/t CEMENT MATERIALS EFFICIENCY m2 PER PERSON 1 LIFETIME
t NEW MATERIALS PER M2 YEARS
0.62 2.45 0.10 52 91

0.21 49
0.22
0.48 1.92
65
BASELINE CIRCULAR BASELINE REDUCED REUSE OF MATERIALS CIRCULAR BASELINE CIRCULAR BASELINE CIRCULAR
SCENARIO SCENARIO SCENARIO WASTE BUILDING EFFICIENCY SCENARIO SCENARIO SCENARIO SCENARIO SCENARIO
COMPONENTS
NOTE: 1 ACROSS ALL BULIDING TYPES EXCEPT INDUSTRIAL BUILDINGS
Design Construction Use Phase EoL Treatment
To achieve this, a range of changes are necessary • Circular business models enable both higher uti-
From: conflicting objectives, lack of transparency and accountability
(Exhibit 6.8.): lisation and longer lifetime. Sharing reduces floor area
• Involvement only at early • Incentive to reduce • requirements
Frequent shift inbyownership
5%, summed • No documentation
across of cate-
all building
• Cement recycling
stage, limited becomes construction
accountability widespread, cutting
costs the
through with limited transfer of materials content
average CO2-intensity
for outcomes (e.g. EoL)of cement production by 23%,
low-cost materials from
and fast gories.
knowledgeTheoraverage lifetime• ofRapid
value capture buildings increases
demolition and by
•0.62
Long term assets and construction
to 0.48 tonnes CO2 per tonne cement. 40%,
of through
investment a combination of adaptive
resulting buildings
degrading of and
frequent change of owners • Constructor optimises labour • the
Ownership
‘build tomodel
last’with risk of
principles andcomponets
improved andmaintenance
materials
• creates
The risk for higher efficiency
materials up-front cost, few incentives
strategies reduce to the lock-in effects and ineffective
described
investment reduce waste or materials use of spaceabove. Combining this, the benefit of each
amount of new building materialsuse that are required, from tonne of building material increases: for each year of
an average of 2.45 tonnes of materials for each squ- service from a building, the materials input is one-third
are metre of building, to 1.92 tonnes. To achieve this, lower in the circular scenario – although as noted, the
waste during construction is reduced to 5%, while steel benefit of longer lifetimes takes a long time to show up
and cement use between
To: collaboration falls by actors,
20–30%, as a resultand
transparency of long
reduced
term objectives
in actual materials demand.
over-specification and use of higher-quality materials.
•15%BIMofenables information
structural • New construction
building components techniques
are reused. • ’Real estate as a service’ • Design for disassembly
transfer and enables return on and digitalisation improve enabled by digitalisation enables value capture at EoL
investments in circular productivity and reduce over- • Size adapted to needs • BIM facilitates EoL treatment
principles (design for specification • Digitalisation enables and ensure quality and
disassembly, durable and • Lower trade-off with labour ongoing optimisation, e.g. of usability of recovered
flexible designs) cost results in improved 157 office space materials and components
• Buildings designed with view materials usage and quality
to future ’material banks’
0.48 1.92
65
Exhibit 6.9
New business models

BASELINE CIRCULAR BASELINE
and digitalisation
REDUCED REUSE OF
enable
MATERIALS CIRCULAR
improvement
BASELINE CIRCULAR BASELINE CIRCULAR
SCENARIO SCENARIO
across the entire
SCENARIO WASTE
building value chain
BUILDING
COMPONENTS
EFFICIENCY SCENARIO SCENARIO SCENARIO SCENARIO SCENARIO
FROM: TO:
CONFLICTING OBJECTIVES, LACK OF TRANSPARENCY COLLABORATION BETWEEN ACTORS, TRANSPARENCY
AND ACCOUNTABILITY AND LONG TERM OBJECTIVES
Involvement only at early stage, limited accountability BIM enables information transfer and enables return on
for outcomes (e.g. EOL) investments in circular principles (design for disassembly,
Long term assets and frequent change of owners creates durable and flexible designs)
risk for higher up-front investment DESIGN Buildings designed with view to future ’material banks’
Incentive to reduce construction costs through low-cost New construction techniques and digitalisation improve
materials and fast construction productivity and reduce over-specification
Constructor optimises labour cost, few incentives to Lower trade-off with labour cost results in improved
reduce waste or materials use CONSTRUCTION materials usage and quality
Frequent shift in ownership with limited transfer of ’Real estate as a service’ enabled by digitalisation
knowledge or value capture of investment Size adapted to needs
Ownership model with risk of lock-in effects and Digitalisation enables ongoing optimisation,
ineffective use of space USE PHASE e.g. of office space
No documentation of materials content Design for disassembly enables value capture at EOL
Rapid demolition and resulting degrading of BIM facilitates EOL treatment and ensure quality and
components and materials EOL usability of recovered materials and components
TREATMENT
BUILDING INFORMATION MANAGEMENT NEW DIGITAL TECHNOLOGIES IN CONSTRUCTION

The circular scenario illustrated here is intended to be alignment of incentives, risk-sharing, and an emphasis
ambitious, but still mostly incremental. There are signifi- on ‘real estate as a service’. Some of these may require
cant changes in parts of the construction supply chain, much more extensive vertical integration, with construc-
but nothing that amounts to a revolution in practices (Ex- tion companies more involved in the use phase of buil-
hibit 6.9). A common
BUILDING INFORMATION theme is to betterAND
DRONES align incentives dings. AND
AUTOMATION New practices also INmust be able to 3D
ROBOTICS bridge the
across MODELLING
different phases
(BIM) of design, construction, occu-PREFABRICATION
3D SCANNERS time gap betweenCONSTRUCTION
initial decision and specification,
PRINTING and
pancy and end-of-life treatment of buildings. New busi- the ultimate effect on materials use and recovery.
Platform will
ness models for integrated
be required inCansome
be usedcases,
for construction Improves quality and
with better Enables complex tasks with Enables construction of
design, modelling, plann- site inspections precision human monitoring complex shapes
ing and collaboration Reduces labour costs and time Reduces labour cost, Improves quality and Reduces construction time
Reduces cost and saves Improves safety for workers delivery time and waste precision and cost of customised
construction time in construction Reduces labour cost and components
Facilitates re-use and delivery time and improves Reduces waste in construct-
recycling of EOL materials safety ion due to better precision
158
DIGITALISATION OF CONSTRUCTION IS A KEY ENABLER OF CIRCULAR OPPORTUNITIES

BASELINE CIRCULAR BASELINE REDUCED REUSE OF MATERIALS CIRCULAR BASELINE CIRCULAR BASELINE CIRCULAR
ManySCENARIO
of these changes are SCENARIO
SCENARIO made possible
WASTE by digitali-
BUILDING BIM can be
EFFICIENCY complemented
SCENARIO SCENARIO bySCENARIO
drones and 3D scanners
SCENARIO SCENARIO
sation, which could help overcome many of theCOMPONENTS
barriers to provide real-time data on the construction process.
to improved materials use in today’s system. Above all, a
range of data-driven strategies enable coordination that New digital construction technologies: The other ma-
makes economically viable arrangements that otherwi- jor leg of digitalisation is the automation of the construc-
se are possible only in coordinated, whole supply chain tion process. Some of this consists of moving construction
approaches.
FROM: Some of the key aspects of this change are off-site, through
TO: pre-fabrication – in turn enabled by more
set out in Exhibit
CONFLICTING 6.10. LACK OF TRANSPARENCY
OBJECTIVES, advancedCOLLABORATION
and customised inventory
BETWEEN ACTORS, andTRANSPARENCY
manufacturing
AND ACCOUNTABILITY processes. AND
AsLONG TERM OBJECTIVES
important, the building process itself can
Data-driven strategies during construction and ope- be automated. One major construction company summa-
ration of buildings:
Involvement Ubiquitous
only at early stage, use of much more ad- rised its vision
limited accountability BIMfor 2050
enables by saying:
information “The
transfer and construction
enables return onsite
vanced Building
for outcomes Information Modelling (BIM) is a cen- of 2050 willinvestments
(e.g. EOL) be human-free.
in circularRobots
principleswill work
(design in teams to
for disassembly,
Long term
trepiece ofassets
theand frequentscenario.
circular change of owners
BIM creates
has the potential build complex durable and flexibleusing
structures designs)dynamic new materials.
DESIGN
notriskjust
for higher up-front investment
to store detailed 3D object data, but also to en- Elements of Buildings the build designed with view to future45’material banks’
will self-assemble.” Whether or not
able advanced control over scheduling, cost, and opera- this vision materialises, the first forays into construction
tions and tomaintenance
Incentive reduce construction during the use
costs through phase. This can robotics areNew
low-cost already proving
construction capable
techniques of performing
and digitalisation some
improve
address
materialstheandtransaction costs and lack of information that tasks faster productivity
fast construction than unaided human
and reduce labour,46 and it is clear
over-specification
Constructor
hold back aoptimises
numberlabour cost, few opportunities
of circular incentives to Lower with
in buildings: that the trade-off trade-off with labour
labour cost results
cost that now inunderlies
improved much
reduce waste or materials use CONSTRUCTION materials usage and quality
cutting buildings materials waste by tightly managing the materials overuse, over-specification, and risk aversion
flow of construction materials; creating shared platforms can be mitigated to a very large extent through automation.
or Frequent
exchanges shift infor the flow
ownership with of end-of-life
limited transfer ofcomponents for Add to this ’Real construction techniques
estate as a service’ enabled bysuch as 3D printing
digitalisation
knowledge
reuse and or value capture
materials forofrecycling;
investment enabling more advan- and self-assembling. Size adapted 3D to needs
printing already is in advanced
cedOwnership model with
inventories andrisk of lock-in effects
advanced and
construction techniques development, Digitalisation enables ongoing
with the Chinese optimisation,
company WinSun 3D using
ineffective use of space USE PHASE e.g. of office space
that are required to reduce over-specification; and ser- the technology at large scale. Finally, automation also 47
ving as the repository for information required for buil- has significant promise for disassembly of buildings in ways
dings as materials
No documentation of materials notably materials passports. that avoid the
banks,content destructive
Design practices
for disassembly enablesof today’s
value capturedemolition.
at EOL
Rapid demolition and resulting degrading of BIM facilitates EOL treatment and ensure quality and
components and materials EOL usability of recovered materials and components
TREATMENT
Exhibit 6.10
Digitalisation is a key enabler of circular

opportunities in buildings
BUILDING INFORMATION MANAGEMENT NEW DIGITAL TECHNOLOGIES IN CONSTRUCTION
BUILDING INFORMATION DRONES AND AUTOMATION AND ROBOTICS IN 3D

MODELLING (BIM) 3D SCANNERS PREFABRICATION CONSTRUCTION PRINTING
Platform for integrated Can be used for construction Improves quality and Enables complex tasks with Enables construction of
design, modelling, plann- site inspections precision human monitoring complex shapes
ing and collaboration Reduces labour costs and time Reduces labour cost, Improves quality and Reduces construction time
Reduces cost and saves Improves safety for workers delivery time and waste precision and cost of customised
construction time in construction Reduces labour cost and components
Facilitates re-use and delivery time and improves Reduces waste in construct-
recycling of EOL materials safety ion due to better precision
159
6.4 HOW TO GET THERE – ACTIONS TO PROMOTE

A MORE CIRCULAR BUILDINGS SECTOR
Regulators intervene heavily in the buildings sector, business models. This could include exchange platforms
but to date have paid little attention to the efficient use of for building materials, or standards for reused compo-
materials. Strategies used to promote energy efficiency nents.
may be helpful here. For example, it might be possible
to introduce labelling schemes similar to those used for Finally, the public sector has major roles both as a
energy performance for aspects related to materials uses customer of building and infrastructure projects, and as a
– such as the amount of reused content, measures of principal actor through its influence on city planning. The-
the degree of over-specification, or adherence to design se are key areas to start for the transformation required to
principles conducive to high-value disassembly. give buildings longer lifetimes, through greater durability and
adaptability. A bold first step would be to articulate a vision
Regulations also have a major role in the end-of-life tre- and roadmap at the city level towards much more long-lived
atment of buildings. To date the focus has been prima- structures within a continuously developing urban landscape.
rily on the safe disposal of construction and demolition
waste. To realise the circular opportunities described For industry, many of the practices described in this stu-
here, it would be necessary to enable a range of other dy would require changes to business practices – but of-
practices. One example is the introduction of materials ten more in the nature of integration and extension of cur-
passports, which exist in detail in some EU countries rent practice than any radical new departure. As European
but not in others. construction seeks a path towards greater productivity, a step
change in digitalisation, and more sustainable construction,
A third area where regulation can play a role is in pro- circular economy opportunities are a central component. The
moting the emergence of coordination and integration pioneers of these business models and practices could well
mechanisms that currently are lacking to enable circular find the innovations a source of competitive advantage.
160
As European construction seeks a path
towards greater productivity, a step change
in digitalisation, and more sustainable
construction, circular economy opportunities
are a central component.
. 161
Endnotes
Chapter 1.
1
This includes both direct emissions and the emissions from the production of electricity used by industry (International Energy Agency, 2017). Direct emissions are
closer to 30% (IPCC, 2014).
International Energy Agency (2017). Energy Technology Perspectives 2017: Catalysing Energy Technology Transformations. International Energy Agency, Paris.
Intergovernmental Panel on Climate Change (2014). Climate Change 2014: Mitigation of Climate Change. Working Group III Contribution to the IPCC 5th
Assessment Report: Technical Summary.
2
Pauliuk, S., Wang, T. and Müller, D. B. (2013). Steel all over the world: Estimating in-use stocks of iron for 200 countries. Resources, Conservation and Recycling,
71. 22–30. DOI:10.1016/j.resconrec.2012.11.008.
3
The remaining 25% includes some additional materials (such as lime, other non-ferrous metals, ceramics, and glass), but also manufacturing and chemicals produc-
tion.
4
See subsequent chapters for discussion and comparison with other scenarios. In overview, the steel scenario assumes somewhat higher growth than other long-range
scenarios, which tend to assume that steel stocks will only saturate much later, into the 22nd century. For cement, the scenario is in the middle of ranges for other
scenarios, but somewhat higher than what is used in many ’integrated assessment models’ to study low-carbon economic pathways. For aluminium, views diverge
widely, but the saturation assumption is again close to the middle of other assumptions in the literature. Finally, much less long-range work has been done on plastics,
but the assumption results in similar growth of plastics production as in industry and consultant forecasts for the next decades.
5 European Commission (2015). The Paris Protocol – A blueprint for tackling global climate change beyond 2020. Climate Action - European Commission. https://
ec.europa.eu/clima/policies/international/negotiations/paris_en.
6 This is complicated by the fact that most scenarios assume substantial amounts of ‘negative emissions’ that can offset some of the remaining emissions. However, the
250–300 Gt budget share refers to actual emissions from industry, even accounting for the fact that large amounts of negative emissions are assumed to be available.
Specifically, around 400 Gt CO2 is available for all of industry, but at least 100 Gt must be devoted to other industrial activity – from manufacturing to chemicals,
pulp & paper, and other uses.
7 Intergovernmental Panel on Climate Change, IPCC (2014). IAMC AR5 Scenario Database. https://secure.iiasa.ac.at/web-apps/ene/AR5DB/.
8 Specifically, this assumes zero emissions from all thermal energy in cement production is zero-carbon as well as the electricity inputs to aluminium smelting steel
electric arc furnaces. For plastics, the addition of external zero-carbon heat can reduce production emissions by around half. Finally, for primary steel production,
zero-carbon energy does little, but the scenario posits that production is converted to the lowest-emitting current process proven at scale, which is the direct reduction
of iron ore using natural gas.
9 The carbon budget for materials production is based on the average of available scenarios in the IPCC AR5 database (Intergovernmental Panel on Climate Change,
2014), adjusted for the needs of other industrial activity. The overall carbon budget is based on Mercator Research Institute on Global Commons and Climate
Change, ‘Remaining Carbon Budget’ (2018), which gives a remaining budget of 705 billion tonnes of CO2 (range of 335-885) for a 2 °C scenario. Adjusting for CO2
emissions data for recent years from Global Carbon Project (2017), ‘Global Carbon Budget 2017’, this corresponds to a budget around 800 billion tonnes for the
period 2015-2100. CO2 emissions from materials production are based on the gradual adoption of current ‘best available technique’ as described in subsequent chap-
ters of the report, but no other decarbonisation. The low-carbon production scenario sees the gradual elimination of energy emissions from cement, half of emissions
from steel (e.g. through natural-gas direct reduction or the use of biomass feedstock), decarbonisation of electricity as an input to aluminium production, and use of
renewable energy to produce plastics, but not a switch to bio-feedstock.
Intergovernmental Panel on Climate Change, IPCC (2014). IAMC AR5 Scenario Database.
Mercator Research Institute on Global Commons and Climate Change (2018). Remaining carbon budget. https://www.mcc-berlin.net/en/research/co2-budget.html.
Global Carbon Project (2017). Global Carbon Budget 2017. 60.
10
Plastics emissions include emissions from production (representing 50% of emissions in the best available technique (BAT) scenario and 36% in the zero carbon
scenario), incineration and landfill.
11
International Energy Agency (2016). 20 Years of Carbon Capture and Storage - Accelerating Future Deployment.
12
In China, the cement and steel industry jointly account for more than half of the particulate matter pollution (Lei et al., 2011; Wang et al. 2016).
Lei, Y., Zhang, Q., Nielsen, C. and He, K. (2011). An inventory of primary air pollutants and CO2 emissions from cement production in China, 1990–2020. Atmo-
spheric Environment, 45(1). 147–54. DOI:10.1016/j.atmosenv.2010.09.034.
Wang, K., Tian, H., Hua, S., Zhu, C., Gao, J., Xue, Y., Hao, J., Wang, Y. and Zhou, J. (2016). A comprehensive emission inventory of multiple air pollutants
from iron and steel industry in China: Temporal trends and spatial variation characteristics. Science of The Total Environment, 559. 7–14. DOI:10.1016/j.scito-
tenv.2016.03.125.
13
European Commission (2011). Energy Roadmap 2050 Impact Assessment and Scenario Analysis. SEC(2011) 1565 final. https://ec.europa.eu/energy/sites/ener/
files/documents/roadmap2050_ia_20120430_en_0.pdf.
14
The European Steel Association (2013). A Steel Roadmap for a Low Carbon Europe 2050.
162
The European Cement Association (2013). The Role of Cement in the 2050 Low Carbon Economy. http://lowcarboneconomy.cembureau.eu/.
CEFIC (2013). European Chemistry for Growth: Unlocking a Competitive, Low Carbon and Energy Efficient Future. http://www.cefic.org/Industry-support/.
15
This includes the emissions from the generation of electricity used by these sectors.
16
The calculation accounts for a number of ways in which different circular economy opportunities interact. For example, a higher share of steel recycling reduces
the CO2 intensity of steel production, which in turn reduces the savings that are availble from reduced steel use in products. We account for this in the cost curve by
showing the potential available for each measure (each ‘bar’ in the cost curve) available when all other measures are also implemented. If calculated in isolation, most
measures would have larger potential.
17
The CO2 intensity of aluminium varies strongly with the source of electricity used. Coal-based production in China and some other parts of Asia results in a CO2
intensity of 18 t CO2 per tonne aluminium, whereas gas-based electricity in the Middle East is closer to 7-8 tCO2 / tonne aluminium. Aluminium produced from cle-
an electricity (such as Icelandic geothermal or Norwegian hydropower) can reach CO2 emissions as low as 2 t CO2 per tonne aluminium. Unlike the other materials
discussed here, aluminium therefore has potential to address the large majority of the CO2 emissions through the use of renewable energy (Hydro, 2017).
Hydro (2017). Capital Markets Day 2017. https://www.hydro.com/en/investor-relations/presentations/capital-markets-days/.
18
IVA, S. B. (2014). Klimatpåverkan från byggprocessen. http://www.iva.se/publicerat/.
19
Moynihan, M. C. and Allwood, J. M. (2014). Utilization of structural steel in buildings. Proceedings. Mathematical, Physical, and Engineering Sciences / The
Royal Society, 470(2168). DOI:10.1098/rspa.2014.0170.
20
Another major opportunity is to substitute construction materials, a topic not covered in this report. The use of cross-laminated timber has been proposed as an op-
tion to reduce the use of cement and steel in the construction of buildings, potentially with a large negative impact of CO2 both from avoided emissions in materials
production, and from an effective CO2 sink within the wood (Kuilen et al., 2011)
Kuilen, J. W. G. V. D., Ceccotti, A., Xia, Z. and He, M. (2011). Very Tall Wooden Buildings with Cross Laminated Timber. Procedia Engineering, 14. 1621–28.
DOI:10.1016/j.proeng.2011.07.204.
Chapter 2.
1
World Steel Association (2014). Steel Solutions in the Green Economy. https://www.worldsteel.org/en/dam/jcr:a3776cb0-7b75-4de7-810a-dcab78122464/PD-
F%253A+Affordable+social+housing.pdf.
2
Pauliuk, S., Wang, T. and Müller, D. B. (2013). Steel all over the world: Estimating in-use stocks of iron for 200 countries. Resources, Conservation and Recycling,
71. 22–30. DOI:10.1016/j.resconrec.2012.11.008.
3
Pauliuk et al. (2013). Steel all over the world.
Graedel, T. E. (2010). Metal Stocks in Society: Scientific Synthesis. UNEP, Nairobi.
Pauliuk, S., Wang, T. and Müller, D. B. (2012). Moving Toward the Circular Economy: The Role of Stocks in the Chinese Steel Cycle. Environmental Science &
Technology, 46(1). 148–54. DOI:10.1021/es201904c.
4
The methodology used here builds on the excellent and foundational work described in Pauliuk, Milford, et al. (2013), Milford et al. (2013), and Daehn et al.
(2017), which has been crucial to developing the estimates and insights presented here. The scenarios and assumptions we use differ in some respects, especially in
relating future steel demand more closely to recent projected GDP developments. However, the implementation of a stock-driven model of future demand as well as
the foundational data are the same.
Pauliuk, S., Milford, R. L., Müller, D. B. and Allwood, J. M. (2013). The Steel Scrap Age. Environmental Science & Technology, 47(7). 3448–54. DOI:10.1021/
es303149z.
Milford, R. L., Pauliuk, S., Allwood, J. M. and Müller, D. B. (2013). The Roles of Energy and Material Efficiency in Meeting Steel Industry CO2 Targets. Environ-
mental Science & Technology, 47(7). 3455–62. DOI:10.1021/es3031424.
Daehn, K. E., Cabrera Serrenho, A. and Allwood, J. M. (2017). How Will Copper Contamination Constrain Future Global Steel Recycling? Environmental Science
& Technology, 51(11). 6599–6606. DOI:10.1021/acs.est.7b00997.
5
Pauliuk et al. (2013). The Steel Scrap Age.
6
United Nations, Department of Economic and Social Affairs, Population Division (2017). World Population Prospects: The 2017 Revision. ST/ESA/SER.A/399.
https://esa.un.org/unpd/wpp/.
7
Specifically, saturation takes place once regions reach a GDP (purchasing power parity) of 40,000 USD per capita, as per the forecasts used in the Shared Socioeco-
nomic Pathway 2 ‘Middle of the Road’ (Fricko et al. 2017). This is significantly more accelerated than in the original article by Pauliuk, Milford, et al. (2013), and
accounts for some of the differences in the results.
163
Fricko, O., Havlik, P., Rogelj, J., Klimont, Z., Gusti, M., et al. (2017). The marker quantification of the Shared Socioeconomic Pathway 2: A middle-of-the-road
scenario for the 21st century. Global Environmental Change, 42. 251–67. DOI:10.1016/j.gloenvcha.2016.06.004.
8
9
World Steel Association (2017). Steel Statistical Yearbook 2017. Brussels/Beijing. https://www.worldsteel.org/en/dam/jcr:3e275c73-6f11-4e7f-a5d8-23d9bc5c508f/
Steel+Statistical+Yearbook+2017.pdf.
Bureau of International Recycling (2017). World steel recycling in figures 2012-2013, Steel Scrap - a Raw Material for Steelmaking.
Total steel demand; Material Economics projection, builds on Pauliuk (2013) but with other assumptions for saturation.
10
McKinsey (2018). Decarbonizing Steel. Workshop Document.
Global Calculator (2015). Global Calculator Technical Documentation: Manufacturing Sector. http://uncached-site.globalcalculator.org/sites/default/files/Manu-
facturing%20technical%20documentation%20part%202.pdf.
van Ruijven, B. J., van Vuuren, D. P., Boskaljon, W., Neelis, M. L., Saygin, D. and Patel, M. K. (2016). Long-term model-based projections of energy use and CO2
emissions from the global steel and cement industries. Resources, Conservation and Recycling, 112. 15–36. DOI:10.1016/j.resconrec.2016.04.016.
Xylia, M., Silveira, S., Duerinck, J. and Meinke-Hubeny, F. (2017). Weighing regional scrap availability in global pathways for steel production processes. Energy
Efficiency, . 1–25. DOI:10.1007/s12053-017-9583-7.
Allwood, J. M., Cullen, J. M., Carruth, M. A., Cooper, D. R., McBrien, M., Milford, R. L., Moynihan, M. C. and Patel, A. C. (2012). Sustainable Materials: With
Both Eyes Open. UIT Cambridge Ltd.
11
World Steel Association (2017). Steel Statistical Yearbook 2017.
12
In estimates presented here we include direct emissions at steel plants, as well as indirect emissions from the production of the electricity used. We do not attempt
to account for full life-cycle emissions (which would include items such as the mining of ore and transportation of raw materials). The emissions factors used build on
the detailed discussion in Milford et al. (2013).
Milford et al. (2013). The Roles of Energy.
13
Increased process efficiency: The BOF route is already highly optimised and efficient. Current estimates suggest that the amount of additional efficiency potential
through use of the best available technologies is on the order of 15% (Milford et al. 2013). Direct reduced iron: DRI cuts emissions by almost 50%, with estimates
of emissions intensity ranging between 1.1-1.3 tCO2 per tonne steel. (International Energy Agency 2017; Milford et al. 2013; Birat and Maizières-lès-Metz 2010).
Bio-based inputs: Bio-derived raw materials are used in some countries, such as Brazil, where local conditions are favourable. Emissions reductions can be around
50% (Mathieson et al. 2011). Carbon capture and storage: capturing the CO2 from the blast furnace of an integrated steel plant can reduce overall emissions by 60%
(Rootzén 2015; Birat and Maizières-lès-Metz 2010). To date, only one such facility is operational, but the technology features heavily in long-range roadmaps for the
reduction of CO2 emissions from steelmaking. Fundamental process innovation: One proposed route is to use pure hydrogen instead of coal or natural gas to produce
DRI. This is now under active research, but has not been achieved at scale. Electrolysis offers a much more speculative, long-term option (Milford et al. 2013) If these
can be achieved, they offer the advantage that, like EAF production, they could use low-carbon electricity as the key energy input.
Milford et al. (2013). The Roles of Energy.
International Energy Agency (2017). Energy Technology Perspectives 2017.
Birat, J. P. and Maizières-lès-Metz, D. (2010). Global Technology Roadmap for CCS in Industry - Steel Sectoral Report. UNIDO and ArcelorMittal. https://hub.
globalccsinstitute.com/sites/default/files/publications/15671/global-technology-roadmap-ccs-industry-steel-sectoral-report.pdf.
Mathieson, J., Rogers, H., Somerville, M., Ridgeway, P. and Jahanshahi, S. (2011). Use of biomass in the iron and steel industry - An Australian perspective. 1st
International Conference on Energy Efficiency and CO2 Reduction in the Steel Industry (EECR Steel 2011) - incorporated in METEC InSteelCon 2011 10p. ,
Dusseldorf, Germany. The Steel Institute VDEh.
Rootzén, J. (2015). Pathways to Deep Decarbonisation of Carbon-Intensive Industry in the European Union: Techno-Economic Assessments of Key Technologies
and Measures. Doktorsavhandlingar vid Chalmers Tekniska Högskola. Chalmers University of Technology, Göteborg.
14
This would require a combination of replacement of BOF-based capacity with DRI, or the retrofitting of CCS, or the large-scale use of bioenergy. In addition, in
this scenario, we also assume that CO2 emissions from electricity production are eliminated by 2050, reducing the emissions intensity of EAF production.
15
Willeke, R. (2017). EU-28 Steel scrap statistics. https://www.euric-aisbl.eu/facts-figures/statistics/download/231/63/32.
164
16
World Steel Association (2009). The Three Rs of Sustainable Steel. Brussels/Beijing.
17
Jernkontoret Research (2013). The Steel Eco-Cycle - Environmental Research Programme. Closing the Loop in the Manufacture and Use of Steel in Society. Scien-
tific Report 2004-2012. http://www.jernkontoret.se/en/publications/steel-research/open-reports-serie-d/d-853/.
18
World Steel Association (2009). The Three Rs.
19
Allwood et al. (2012). Sustainable Materials.
20
Cullen, J. M., Allwood, J. M. and Bambach, M. D. (2012). Mapping the Global Flow of Steel: From Steelmaking to End-Use Goods. Environmental Science &
Technology, 46(24). 13048–55. DOI:10.1021/es302433p.
21
Price, D. (2009). Yield improvement in the steel industry. Ironmaking & Steelmaking; London, 36(7). 482–86.
22
Reck, B. K. and Graedel, T. E. (2012). Challenges in Metal Recycling. Science, 337(6095). 690–95. DOI:10.1126/science.1217501.
23
Ohno, H., Matsubae, K., Nakajima, K., Nakamura, S. and Nagasaka, T. (2014). Unintentional Flow of Alloying Elements in Steel during Recycling of End-of-Life
Vehicles. Journal of Industrial Ecology, 18(2). 242–53. DOI:10.1111/jiec.12095.
24
Daehn et al. (2017). How Will Copper Contamination Constrain.
25
Björkman, B. and Samuelsson, C. (2014). Chapter 6 - Recycling of Steel. In Handbook of Recycling. Elsevier, Boston. 65–83. DOI:10.1016/B978-0-12-396459-
5.00006-4.
26
We focus on copper here, as the most significant constraint on steel recycling. However, there also are other tramp elements, such as tin and nickel, that can cause
analogous problems. The solutions in many cases must be similar to those discussed in relation to copper. See e.g. Ohno et al. (2014).
27
Haupt, M., Vadenbo, C., Zeltner, C. and Hellweg, S. (2017). Influence of Input‐Scrap Quality on the Environmental Impact of Secondary Steel Production. Jour-
nal of Industrial Ecology, 21(2). 391–401. DOI:10.1111/jiec.12439.
28
Global product mix: Material Economics modelling. Copper limits: Daehn et al. (2017).
Daehn et al. (2017). How Will Copper Contamination Constrain – Supplementary Information.
29
In addition, some steel alloys contain copper for corrosion resistance.
30
Nakamura, S., Kondo, Y., Matsubae, K., Nakajima, K., Tasaki, T. and Nagasaka, T. (2012). Quality- and Dilution Losses in the Recycling of Ferrous Materials
from End-of-Life Passenger Cars: Input-Output Analysis under Explicit Consideration of Scrap Quality. Environmental Science & Technology, 46(17). 9266–73.
DOI:10.1021/es3013529.
31
OECD (2015). Excess Capacity in the Global Steel Industry and the Implications of New Investment Projects. OECD Science, Technology and Industry Policy
Papers, (18). DOI:10.1787/5js65x46nxhj-en.
32
World Steel Association (2017). Steel Statistical Yearbook 2017.
33
Global emissions are a different matter. If the EU were instead to export large volumes of scrap, emissions within the EU would increase, but emissions elsewhere
could fall, as less primary production would be required there.
34
Azevedo, J. M. C., CabreraSerrenho, A. and Allwood, J. M. (2018). Energy and material efficiency of steel powder metallurgy. Powder Technology, 328. 329–36.
DOI:10.1016/j.powtec.2018.01.009.
35
Haupt et al. (2017). Influence of Input‐Scrap Quality.
36
Direct strip casting is one method: the steel is cast in a protected atmosphere, which combines short oxidation time with rapid cooling and direct rolling. This poses
a suitable alternative for products that do not contain high levels of copper during their useful life.
37
Nakajima, K., Takeda, O., Miki, T., Matsubae, K. and Nagasaka, T. (2011). Thermodynamic Analysis for the Controllability of Elements in the Recycling Process
of Metals. Environmental Science & Technology, 45(11). 4929–36. DOI:10.1021/es104231n.
38
Daehn et al. (2017). How Will Copper Contamination Constrain.
165
Chapter 3.
1
Each tonne of plastics uses c. 1.1 tonnes of oil (Levi and Cullen, 2018), which means 903 million tonnes of oil would be needed for plastics production in 2050.
2050 oil consumption in the International Energy Agency’s (2017) 2 °C scenario is just over 3900 million tonnes. In addition, producing this volume of plastics
would require large volumes of natural gas.
Levi, P. G. and Cullen, J. M. (2018). Mapping Global Flows of Chemicals: From Fossil Fuel Feedstocks to Chemical Products. Environmental Science & Technology,
52(4). 1725–34. DOI:10.1021/acs.est.7b04573.
2
Geyer, R., Jambeck, J. R. and Law, K. L. (2017). Production, use, and fate of all plastics ever made. Science Advances, 3(7). e1700782. DOI:10.1126/sci-
adv.1700782.
3 Haward, M. (2018). Plastic pollution of the world’s seas and oceans as a contemporary challenge in ocean governance. Nature Communications, 9(1). 667.
DOI:10.1038/s41467-018-03104-3.
4
However, microplastics and plastics in the ocean are a broader problem that also depend on plastic products in fishing industries, microplastics added to cosmetics
and consumer products, plastic fibres in textiles and wear on tires (Boucher and Friot, 2017).
Boucher, J. and Friot, D. (2017). Primary Microplastics in the Oceans: A Global Evaluation of Sources. IUCN International Union for Conservation of Nature.
DOI:10.2305/IUCN.CH.2017.01.en.
5
Emissions differ by plastics type, from 1.6 tonnes CO2 per tonne plastic for polypropylene, 1.8 for polyethylene, 1.99 for PVC, 2.15 for PET, 3.3 for polystyrene,
and 4.8 for a weighted average of the remaining mix of plastics in use (Deloitte and Plastics Recyclers Europe, 2015; Plastics Europe, 2018a). These numbers include
the indirect emissions from electricity, as well as the production of precursors such as chlorine.
Deloitte and Plastics Recyclers Europe (2015). Increased EU Plastics Recycling Targets: Environmental, Economic and Social Impact Assessment. http://www.plas-
ticsrecyclers.eu/sites/default/files/BIO_Deloitte_PRE_Plastics%20Recycling%20Impact_Assesment_Final%20Report.pdf.
Plastics Europe (2018a). Eco-profiles. http://www.plasticseurope.org/plastics-sustainability-14017/eco-profiles.aspx.
6
This on the assumption that a net-zero 2050 position would involve some residual emissions, offset by ‘negative emissions’ from a variety of sources, such as CO2
sinks.
7
Deloitte and Plastics Recyclers Europe (2015). Increased EU Plastics Recycling Targets.
8
Plastics Europe (2018a). Eco-profiles.
Zhu, J.-B., Watson, E. M., Tang, J. and Chen, E. Y.-X. (2018). A synthetic polymer system with repeatable chemical recyclability. Science, 360(6387). 398–403.
DOI:10.1126/science.aar5498.
9
PE (29%); PP (19%); PVC (10%); PS (7%); PET (7%).
10
Plastics Europe (2018b). Plastics - the Facts 2016. Plastics Europe. http://www.plasticseurope.org/Document/plastics---the-facts-2016-15787.aspx?FolID=2.
11
European Commission (2018). A European Strategy for Plastics in a Circular Economy. http://ec.europa.eu/environment/circular-economy/pdf/plastics-strategy.
pdf.
12
Material Economics (2018). Ett Värdebeständigt Svenskt Materialsystem. http://resource-sip.se/content/uploads/2018/01/vardebestmtrlsystemrapport180118.pdf.
13
14
Plastics Europe (2018b). Plastics - the Facts 2016.
15
Ellen MacArthur Foundation (2016). The New Plastics Economy: Rethinking the future of plastics & Catalysing action. https://www.ellenmacarthurfoundation.
org/publications/the-new-plastics-economy-rethinking-the-future-of-plastics-catalysing-action.
16
Ellen MacArthur Foundation (2016). The New Plastics Economy.
17
Also, Europe has long sent almost half of it to China as a lower-cost alternative, which has discouraged investment in European recycling capacity. This is now
changing dramatically, as China has recently restricted plastic waste imports, setting higher quality and purity requirements.
18
Plastics Europe (2018b). Plastics - the Facts 2016.
19
Other assessments have come to a similar conclusion, notably Ellen MacArthur Foundation (2017).
Ellen MacArthur Foundation (2017). The New Plastics Economy - Catalysing Action. Ellen MacArthur Foundation. https://www.ellenmacarthurfoundation.org/
assets/downloads/New-Plastics-Economy_Catalysing-Action_13-1-17.pdf.
20
Bazzanella, A. M. and Ausfelder, F. (2017). Low Carbon Energy and Feedstock for the European Chemical Industry. DECHEMA Gesellschaft für Chemische
Technik und Biotechnologie e.V. https://dechema.de/dechema_media/Technology_study_Low_carbon_energy_and_feedstock_for_the_European_chemical_indu-
stry-p-20002750.pdf.
166
Chapter 4.
1
Aluminium use patterns are more complex and show more variation than for example steel. For example, the United States has an aluminium stock of about 600 kg
per person, while that in European countries varies between 200–500 kg per person (Liu and Müller, 2013). The stock data used for the analysis here is aggregated
for regions as given in (World Aluminium, 2018a) and described in (Bertram et al., 2017). These data show that stocks have grown steadily in countries at all levels of
development, although recently, its growth has slowed in Japan and North America. Expert interviews also indicate that stocks are saturating in some sectors also in
Europe, notably in buildings.
Liu, G. and Müller, D. B. (2013). Centennial Evolution of Aluminum In-Use Stocks on Our Aluminized Planet. Environmental Science & Technology, 47(9).
4882–88. DOI:10.1021/es305108p.
World Aluminium (2018a). Mass Flow Statistics. http://www.world-aluminium.org/statistics/massflow/.
Bertram, M., Ramkumar, S., Rechberger, H., Rombach, G., Bayliss, C., Martchek, K. J., Müller, D. B. and Liu, G. (2017). A regionally-linked, dynamic material
flow modelling tool for rolled, extruded and cast aluminium products. Resources, Conservation and Recycling, (125). 48–69. DOI:10.1016/j.resconrec.2017.05.014.
2
Liu, G., Bangs, C. E. and Müller, D. B. (2013). Stock dynamics and emission pathways of the global aluminium cycle. Nature Climate Change, 3(4). 338–42.
DOI:10.1038/nclimate1698.
3
United Nations, Department of Economic and Social Affairs, Population Division (2017). World Population Prospects: The 2017 Revision. ST/ESA/SER.A/399.
https://esa.un.org/unpd/wpp/.
4
Bertram et al. (2017). A regionally-linked.
5
Liu and Müller (2013). Centennial Evolution.
6
World Aluminium (2018a). Mass Flow Statistics.
7
Note that this refers to European demand, not European production. Published data by World Aluminium (2018a) gives total input to fabrication as more than 16
million tonnes in 2016. However, this appears to be an anomalous number, with a much lower number of 13.8 million tonnes stated for 2015, and 13.4 for 2017.
8
The shape of future demand is very uncertain, of course. There is a strong logic behind the notion that stocks in the EU will saturate, but it is much less clear at what
level, and when. For illustration, if stock build-up were to accelerate instead, and reach 600 kg per person by 2050, some 20 million tonnes per year of production
would be needed to meet EU demand. However, as discuss in Section 4.4, the saturation level is not critical to the main findings of our analysis. The higher the
overall demand, the more important it is to pursue the circularity measures presented in this chapter.
9
Europe exported 5.6 million cars in in 2017 (ACEA - European Automobile Manufacturers’ Association, 2017), with an average aluminium content of around 140
kg per car.
ACEA - European Automobile Manufacturers’ Association (2017). Economic and Market Report Q4 2017. http://www.acea.be/statistics/article/economic-and-mar-
ket-report-key-takeaways-about-the-eu-auto-industry-for-17.
10
Based on data from International Energy Agency (2017).
11
World Aluminium (2018d). Primary Aluminium Smelting Power Consumption. http://www.world-aluminium.org/statistics/primary-aluminium-smel-
ting-power-consumption/.
According to World Aluminium (2018d), 88% of European production uses non-fossil sources of electricity.
12
World Aluminium (2018d). Primary Aluminium.
13
Emissions are calculated from data in World Aluminium Association (2018a, 2018b, 2018c, 2018d) and the International Energy Agency (2017). Emissions from
anodes are estimated based on expert interviews.
World Aluminium (2018b). Metallurgical Alumina Refining Fuel Consumption. http://www.world-aluminium.org/statistics/metallurgical-alumina-refining-fu-

el-consumption/.
World Aluminium (2018c). Perfluorocarbon (PFC) Emissions. http://www.world-aluminium.org/statistics/perfluorocarbon-pfc-emissions/.

14
The International Energy Agency (2017) estimates that direct emissions from aluminium production were 261 Mt CO2 in 2014 and electricity consumption of 926
TWh, with associated emissions of 542 Mt CO2, which gives total emissions of 803 Mt CO2. For comparison, World Aluminium (2018d), which gives electricity
consumption as 844 TWh for 2016 for smelting. The two sources give a range of emissions (including perflourocarbons as well as CO2) of 12.5-15 t CO2 per tonne
aluminium.
167
15
Strictly speaking this is greenhouse gases measured in ‘carbon dioxide equivalents’ rather than just carbon dioxide. This distinction arises because a small but not
insignificant share of emissions come from perflourocarbons, a very potent greenhouse gases that is produced in small quantities during aluminium smelting. In the
following, the shorthand ‘CO2’ is used for the sake of brevity and familiarity.
16
The baseline scenario sees a gradual decarbonisation of all energy use in alumina refining and aluminium smelting, complete by 2100. The ‘reference’ scenario
instead follows the decarbonisation in the International Energy Agency’s (2017) ‘Reference Technology Scenario’. Note that the ‘reference’ scenario is only shown to
illustrate the impact of a slow energy transition on emissions; it is not used for any evaluation of the benefits of greater circularity described in this report. Specifically,
the reference scenario sees the electricity input to the 70% share of aluminium production now powered by fossil fuels following the trajectory of Asian electricity
generation in the International Energy Agency’s (2017) ‘Reference Technology Scenario’. This translates into a 60% reduction in emissions intensity of the current
fossil share. After 2060, the emissions intensity is kept constant.
17
There are also ongoing efforts to reduce other emissions. Improving the smelting production process could reduce energy demand somewhat, but there are technical
limits; even a 10–15% reduction would be a major achievement. Perflourocarbon (PFC) emissions have already fallen dramatically already but could be reduced
further (World Aluminium, 2018b). Emissions from alumina refining could in principle be addressed by using zero-carbon energy inputs, although this is not done
today. Fully eliminating direct emissions from smelting, on the other hand, would require a process breakthrough. One potential route for this was announced by
Elysis Technologies in 2018 (Alcoa, 2018).
World Aluminium (2018c). Perfluorocarbon (PFC) Emissions.
Alcoa (2018). Alcoa and Rio Tinto Announce World’s First Carbon-Free Aluminum Smelting Process. http://news.alcoa.com/press-release/alcoa-and-rio-tinto-anno-
unce-worlds-first-carbon-free-aluminum-smelting-process
18
The figure of 77% is based on the rates reported in Bertram et al. (2017). Ciacci et al. (2013) gives a similar number for Italy, while European Aluminium (2017)
states that more than 90% of end-of-life scrap from buildings and automotive sectors is collected in Europe. However, all estimates are highly uncertain, and some
estimates differ. For example, Ciacci et al. (2013) find lower rates in the United States.
Bertram et al. (2017). A regionally-linked, dynamic material flow modelling tool.
Ciacci, L., Chen, W., Passarini, F., Eckelman, M., Vassura, I. and Morselli, L. (2013). Historical evolution of anthropogenic aluminum stocks and flows in Italy.
Resources, Conservation and Recycling, 72. 1–8. DOI:10.1016/j.resconrec.2012.12.004.
European Aluminium (2017). 2015 Sustainability Highlights. https://www.european-aluminium.eu/media/1836/20170323-sustainability-performance-report.pdf.
19
Cullen, J. M. and Allwood, J. M. (2013). Mapping the Global Flow of Aluminum: From Liquid Aluminum to End-Use Goods. Environmental Science & Techno-
logy, 47(7). 3057–64. DOI:10.1021/es304256s.
20
Liu et al. (2013) gives a collection rate of 98%.
Liu et al. (2013). Stock dynamics.
21
Bertram et al. (2017) gives rates used in this study.
22
Cullen and Allwood. (2013). Mapping the Global Flow of Aluminum.
23
Nakajima, K., Takeda, O., Miki, T., Matsubae, K. and Nagasaka, T. (2011). Thermodynamic Analysis for the Controllability of Elements in the Recycling Process
of Metals. Environmental Science & Technology, 45(11). 4929–36. DOI:10.1021/es104231n.(Nakajima et al. 2011)
24
Müller, D. B. (2017). The global aluminum cycle: challenges and solution options.
Modaresi, R. and Müller, D. B. (2012). The Role of Automobiles for the Future of Aluminum Recycling. Environmental Science & Technology, 46(16). 8587–94.
25
DOI:10.1021/es300648w.
Løvik, A. N., Modaresi, R. and Müller, D. B. (2014). Long-Term Strategies for Increased Recycling of Automotive Aluminum and Its Alloying Elements. Environme-
ntal Science & Technology, 48(8). 4257–65. DOI:10.1021/es405604g.
Cullen and Allwood (2013). Mapping the Global Flow of Aluminum.

26
Modaresi and Müller (2012). The Role of Automobiles
27
Modaresi and Müller (2012). The Role of Automobiles
28
29
Koffler, C. and Florin, J. (2013). Tackling the Downcycling Issue—A Revised Approach to Value-Corrected Substitution in Life Cycle Assessment of Aluminum
(VCS 2.0). Sustainability, 5(11). 4546–60. DOI:10.3390/su5114546.
Material Economics (2018). Ett Värdebeständigt Svenskt Materialsystem. http://resource-sip.se/content/uploads/2018/01/vardebestmtrlsystemrapport180118.pdf.
Ducker Worldwide (2016). Aluminium Content in Cars. https://european-aluminium.eu/media/1721/european-aluminium-ducker-study-summary-report_sept.

30
pdf.
31
International Energy Agency (2017) projections for electric vehicles.
168
32
This is based on expert interviews, and similar to the assumption in (Modaresi and Müller 2012). CRU Group (2018), an industry consultant, points out that some
of the components of electric vehicles, such as battery casings, could use either wrought or cast components, but nonetheless assume an even larger decline in the
share of secondary castings per vehicle than is assumed here.
Modaresi and Müller (2012). The Role of Automobiles.
CRU Group (2018). Electric vehicles to transform aluminium demand. https://www.crugroup.com/knowledge-and-insights/insights/2018/electric-vehicles-to-trans-

form-aluminium-demand/.
33
Material Economics analysis building on Ducker Worldwide (2016), European Aluminium (2011), Magma (2018), and Keshavaram (1999), and Ducker Worldwi-
de (2011).
Ducker Worldwide (2016). Aluminium Content in Cars.
European Aluminium (2011). The aluminium automotive manual. https://www.european-aluminium.eu/resource-hub/aluminium-automotive-manual/.

Magma (2018). Manufacturing State of the Art Aluminum Cylinder Head Castings. https://www.magmasoft.com/en/company/references/reference/manufacturing-
state-of-the-art-aluminum-cylinder-head-castings/.
Keshavaram, B. N. (1999). Aluminum Alloys for Automotive Disc Brake Calipers. 1999-01–0346. SAE International, Warrendale, PA. DOI:10.4271/1999-01-0346.
Ducker Worldwide (2011). Ducker Worldwide Presents at CRU’s 16th World Aluminum Conference. Ducker. http://www.ducker.com/news-insights/duck-
er-worldwide-presents-crus-16th-world-aluminum-conference-full-presentation.
34
The scenarios differ in the degree of sorting and separation and losses. Sorting and separation: Both scenarios see a gradual change from current practice to 2050. In
the baseline scenario, only one-third of wrought end-of-life scrap is sorted separately so that it can be recycled into wrought scrap again. In the circular scenario, this
increases to two-thirds by 2050. In both scenarios, all new scrap is sorted separately into wrought and cast alloys. Losses: Losses are lower in the circular scenario, with
a gradual halving of the losses that take place today, chiefly by increasing the collection rate for end-of-life scrap, and to a lesser extent by reducing the formation of
new scrap and associated remelting losses. Trade: In both scenarios, 70% of any ‘excess scrap’ can be exported, as long as there are other world regions that still have a
shortage of post-consumer scrap to serve demand for cast alloys. Trade is not 100% perfect; as recent developments in both China and the United States have shown,
countries are keen to have some self-sufficiency in aluminium production. Electric vehicles: demand for cast aluminium is modelled bottom up for vehicles, and a fix-
ed share of other demand is assumed to be cast aluminium in other applications. The penetration of electric vehicles follows that in Bloomberg New Energy Finance
(2017) ‘New Economic Outlook’ scenario. The aluminium content vehicles is based on (Ducker Worldwide 2016) and the cast aluminium content of electric vs. ICE
vehicles on expert interviews, with numbers similar to those in Modaresi and Müller (2012). Dilution: the dilution rate is 20% in all scenarios. This is less than the
rate reported in Cullen and Allwood (2013) but on the same level as the baseline assumption in Modaresi and Müller (2012)
Bloomberg New Energy Finance (2017). New Energy Outlook 2017. p.201. https://about.bnef.com/new-energy-outlook/.
Ducker Worldwide (2016). Aluminium Content in Cars.
Modaresi and Müller. (2012). The Role of Automobiles.
Cullen and Allwood. (2013). Mapping the Global Flow of Aluminum.
35
These findings may seem surprising, but they are in fact in line with other estimates. For example, Modaresi and Müller (2012) find that there could be 6 Mt excess
scrap in 2030, rising to up to 18 Mt in 2050. Our scenario estimates 7 Mt by 2030, and 23 Mt by 2050.
Modaresi and Müller. (2012). The Role of Automobiles.
36
Scenarios assume different degrees of losses of aluminium and sorting of post-consumer scrap to avoid downcycling. The baseline scenario keeps current practi-
ce, whereas the circular scenario sees a significant increase in the feasibility of using post-consumer scrap for wrought aluminium by 2050. In both scenarios, the
CO2-intensity of aluminium production falls over time, as described above.
37
There is a more nuanced point here about how to evaluate the emissions savings from reducing primary aluminium use in the EU. First, in an international
commodity market, increases or reductions of global demand over the course of decades serves to reduce supply at the margin: less demand in the EU leads to less
primary capacity overall. A more circular system in the EU therefore avoids the emissions associated with global marginal supply. This is the case regardless of what
exact sources the might EU import from in 2050, which is anyway impossible to predict. If anything, global marginal supply is likely to be more CO2-intensive than
the average emissions factor used here. Second, it is less relevant over time what share of current EU aluminium smelting is powered by legacy low-carbon sources
of electricity production. By 2050, if low-carbon electricity production were not used for primary aluminium, the same low-carbon resource could serve some other
electricity demand. In a 2050 perspective, reduced demand for primary aluminium production therefore has the same effect as any energy efficiency measure, helping
accelerate the transition to a low-carbon energy system, and to contain the amount of low-carbon power required to achieve full decarbonisation. For both these rea-
sons, the CO2 reductions resulting from reduced primary aluminium demand in Europe are calculated using the world average CO2 intensity of production. As noted
above, in the baseline scenario this average follows a pattern of gradual decarbonisation where current fossil production reduces its CO2 intensity of electricity used by
half to 2050, and all electricity is fully CO2-free by 2100. (An alternative interpretation would be to take a weighted average of EU domestic emissions intensity and
global, marginal supply; although this has less of a basis in economic logic, it happens this gives a very similar result).
38
European Aluminium (2017). 2015 Sustainability Highlights.
39
Norsk Hydro (2017). Low-carbon aluminium: Hydro 4.0 and Hydro 75R. https://www.hydro.com/en/products/low-carbon-aluminium-hydro-4.0-and-hydro-75r/.
40
Specifically, the circular scenario assumes that 20% primary metal is still required to dilute and ‘sweeten’ recycled aluminium, and one-third of wrought aluminium
scrap remains mixed to the point where it can only be used for casting alloys.
41
Løvik et al. (2014). Long-Term Strategies.
169
Chapter 5.
1
European Commission (2011). Roadmap to a Single European Transport Area – Towards a Competitive and Resource Efficient Transport System. 0144 final. Euro-
pean Commission. http://eur-lex.europa.eu/legal-content/EN/ALL/?uri=CELEX:52011DC0144.
2
This share is nearly constant over time, varying between 83.0- 83.7% in the period 2004-2014 (Eurostat Statistics Explained, 2017).
Eurostat Statistics Explained (2017). Passenger transport statistics. http://ec.europa.eu/eurostat/statistics-explained/index.php/Passenger_transport_statistics#Relati-
ve_growth_of_passenger_transport_and_the_economy.
3
Eurostat Statistics Explained (2017). Passenger transport statistics.
4
Vehicle ownership increased from 530 vehicles per 1000 people in 2005 to 570 in 2015 (ACEA - European Automobile Manufacturers’ Association,2017b).
ACEA - European Automobile Manufacturers’ Association (2017b). Report: Vehicles in Use. http://www.acea.be/statistics/article/Report-Vehicles-in-Use.
5
Average value per car is approximately 22 thousand EUR (ACEA - European Automobile Manufacturers’ Association, 2017a) which can be compared with adjusted
gross disposable income of households per capita that was 21,903 EUR in 2016 (Eurostat, 2018a).
ACEA - European Automobile Manufacturers’ Association (2017a). Economic and Market Report Q4 2017. http://www.acea.be/statistics/article/economic-and-mar-
ket-report-key-takeaways-about-the-eu-auto-industry-for-17.
Eurostat (2018a). Adjusted gross disposable income of households per capita in PPS. http://ec.europa.eu/eurostat/tgm/table.do?tab=table&language=en&pcode=-
tec00113.
6
252 million cars in the EU, with an approximate replacement value of 22 thousand EUR each, based on number of passenger cars exported from the EU and total
value of passenger car exports (ACEA - European Automobile Manufacturers’ Association, 2017a).
ACEA - European Automobile Manufacturers’ Association (2017a). Economic and Market Report Q4 2017.
7
Eurostat (2018b). Gross fixed capital formation (investments). http://ec.europa.eu/eurostat/tgm/table.do?tab=table&init=1&language=en&pcode=tec00011&plu-
gin=1.
8
Eurostat (2016). Mean consumption expenditure by detailed COICOP level (in PPS). http://appsso.eurostat.ec.europa.eu/nui/submitViewTableAction.do.
9
This study focuses on the materials implications of a shared mobility system. For a perspective on the wider economic implications within a more circular economy,
see Ellen MacArthur Foundation (2015), which also has provided inspiration and insight for the analysis presented here.
Ellen MacArthur Foundation (2015). Growth Within: A Circular Economy Vision for a Competitive Europe. Ellen MacArthur Foundation together with Stiftungs-
fonds für Umweltökonomie und Nachhaltigkeit (SUN) and McKinsey Center for Business and Environment. https://www.ellenmacarthurfoundation.org/assets/
downloads/publications/EllenMacArthurFoundation_Growth-Within_July15.pdf.
10
ACEA - European Automobile Manufacturers’ Association (2017c). Vehicles in Use - Europe 2017. http://www.acea.be/statistics/article/vehicles-in-use-euro-
pe-2017.
11
See Horton and Allwood (2017). However, performance varies widely, and expert interviews indicate that the best carmakers achieve on the order of 25% scrap,
showing the significant potential for reduction.
Horton, P. M. and Allwood, J. M. (2017). Yield improvement opportunities for manufacturing automotive sheet metal components. Journal of Materials Processing
Technology, 249. 78–88. DOI:10.1016/j.jmatprotec.2017.05.037.
Plastics Europe (2018). Plastics - the Facts 2016. Plastics Europe. http://www.plasticseurope.org/Document/plastics---the-facts-2016-15787.aspx?FolID=2.
12
13
The bars show the lifecycle greenhouse gas emissions from materials and use phase per car, as calculated in the article by Ellingsen et al. (2016). The article defines
the use phase as a total driving range of 180,000 km. The emissions are displayed per type of vehicle comprising internal combustion engine vehicles, electric vehicles
powered by the current European electricity mix and electric vehicles in a prospective green energy scenario. The article looks at four car segments: mini, medium, lar-
ge and luxury car, out of which we have used at the large car segment with an average weight of 1528 kg per car. All numbers have been recalculated from ton CO2-eq
per car over the lifetime of 180,000 km to g CO2-eq. per car km.
Ellingsen Ager-Wick, L., Singh, B. and Hammer Strømman, A. (2016). The size and range effect: lifecycle greenhouse gas emissions of electric vehicles. http://iopsci-
ence.iop.org/article/10.1088/1748-9326/11/5/054010.
Pasaoglu, G., Fiorello, D., Martino, A., Scarcella, G., Alemanno, A., et al. (2012). Driving and Parking Patterns of European Car Drivers: A Mobility Survey.
14
Publications Office, Luxembourg.
15
European Environment Agency (2010). Occupancy rates of passenger vehicles. https://www.eea.europa.eu/data-and-maps/indicators/occupancy-rates-of-pas-
senger-vehicles/occupancy-rates-of-passenger-vehicles. Indicator Assessment.
UK Department for Transport (2017). National Travel Survey: England 2016.
16
For clarity: 1.5/5 multiplied by 8% gives a utilisation rate of 2.4 percent.
17
ICCT - The International Council on Clean Transportation (2018). European Vehicle Market Statistics 2017/2018. 63.
18
Dun, C., Horton, G. and Kollamthodi, S. (2015). Improvements to the definition of lifetime mileage of light duty vehicles.(1). 68.
170
19
Dun et al. (2015) Improvements to the definition of lifetime mileage.
European Environment Agency (2017b). Passenger transport demand. https://www.eea.europa.eu/data-and-maps/indicators/passenger-transport-demand-version-2/

assessment-9. Indicator Assessment.
UK Department for Transport (2017). National Travel Survey
ICCT - The International Council on Clean Transportation (2016). European Vehicle Market Statistics: Pocketbook 2016/2017. Berlin.
ACEA - European Automobile Manufacturers’ Association (2018). Vehicles Per Capita, by Country. http://www.acea.be/statistics/tag/category/vehicles-per-capi-
ta-by-country.
Horton and Allwood (2017) Yield improvement opportunities.
20
Annual driving range per car is on average 12,000 km, based on average lifetime of 14 years per car an170,000 km driven per car (Dun et al., 2015).
21
22
23
Modaresi, R. and Müller, D. B. (2012). The Role of Automobiles for the Future of Aluminum Recycling. Environmental Science & Technology, 46(16). 8587–94.
24
DOI:10.1021/es300648w.
Løvik, A. N., Modaresi, R. and Müller, D. B. (2014). Long-Term Strategies for Increased Recycling of Automotive Aluminum and Its Alloying Elements. Environme-
ntal Science & Technology, 48(8). 4257–65. DOI:10.1021/es405604g.
25
See discussion in Chapter 3.
26
Sources include Daehn et al. (2017); Luben et al. (2003); Andersson et al. (2017); and expert interviews.
Daehn, K. E., Cabrera Serrenho, A. and Allwood, J. M. (2017). How Will Copper Contamination Constrain Future Global Steel Recycling? Environmental Science
& Technology, 51(11). 6599–6606. DOI:10.1021/acs.est.7b00997.
Luben, S., Volkova, E. and Janke, D. (2003). Copper and Tin in Steel Scrap. http://www.rmz-mg.com/contents.htm#Volume%2050,%20No.%204%20(Decem-
ber%202003).
Andersson, M., Ljunggren Söderman, M. and Sandén, B. A. (2017). Are scarce metals in cars functionally recycled? Waste Management, 60. 407–16. DOI:10.1016/j.
wasman.2016.06.031.
For example, in Sweden 10% of the stock consists of vehicles that have reached the end of service, but are not being returned for recycling (Copenhagen Econo-
27
mics, 2017), and a similar number is reported for Germany (Pauliuk et al., 2017).
Copenhagen Economics (2017). Den Samhällsekonomiska Kostnaden Av Skrotbilar.
Pauliuk, S., Kondo, Y., Nakamura, S. and Nakajima, K. (2017). Regional distribution and losses of end-of-life steel throughout multiple product life cycles—In-
sights from the global multiregional MaTrace model. Resources, Conservation and Recycling, 116. 84–93. DOI:10.1016/j.resconrec.2016.09.029. (Supplementary
materials).
In 2013, 33% of deregistered ELVs could not be accounted for, probably largely because of illegal exports (Lorz, 2017).
28
Lorz, B. (2017). The End-of-life vehicles (ELV) Directive and its implementation.
29
Seba, T. (2014). Clean Disruption of Energy and Transportation: How Silicon Valley Will Make Oil, Nuclear, Natural Gas, Coal, Electric Utilities and Conventio-
nal.
30
Dr Scott Le Vine, Dr Alireza Zolfaghari and Professor John Polak (2014). Car sharing: evolution, challenges and opportunities. 20.
31
Hook, L. (2018). Race to build a million-mile car becomes a reality. Financial Times, 13 January. https://www.ft.com/content/1255be4c-f680-11e7-88f7-
5465a6ce1a00.
32
Wadud, Z., MacKenzie, D. and Leiby, P. (2016). Help or hindrance? The travel, energy and carbon impacts of highly automated vehicles. Transportation Research
Part A: Policy and Practice, 86. 1–18. DOI:10.1016/j.tra.2015.12.001.
33
This draws on analysis from Ellen MacArthur Foundation (2015).
Ellen MacArthur Foundation (2015). Growth Within.
34
Centre for Economics and Business Research (2014). The Future Economic and Environmental Costs of Gridlock in 2030. https://www.ibtta.org/sites/default/files/
documents/MAF/Costs-of-Congestion-INRIX-Cebr-Report%20(3).pdf. Victoria Transport Policy Institute (2018).
Transportation Cost and Benefit Analysis II – Congestion Costs. http://www.vtpi.org/tca/tca0505.pdf.
35
European Commission Directorate General for Mobility and Transport (2018). Road safety evolution in EU. https://ec.europa.eu/transport/road_safety/sites/roads-
afety/files/pdf/statistics/historical_evol.pdf.
36
European Environment Agency (2017a). Air Quality in Europe 2017.
171
37
Eurofot (2013). Reducing the number of traffic accidents in the EU. European Commission Community Research and Development Information Service. https://
cordis.europa.eu/news/rcn/35406_en.html.
NHTSA (2008). National Motor Vehicle Crash Causation Survey: Report to Congress. US Department of Transportation, Springfield.
One detailed study found that reducing individually owned cars to 40% ‘essentially makes congestion disappear’ (The International Transport Forum – OECD,
38
2016).
The International Transport Forum - OECD (2016). Shared Mobility - Innovation for Livable Cities. International Transport Forum. https://www.itf-oecd.org/sites/
default/files/docs/shared-mobility-liveable-cities.pdf.
39
This draws on analysis from Ellen MacArthur Foundation (2015). Key sources include: Centre for Economics and Business Research (2014); European Environme-
nt Agency (2017a).
Ellen MacArthur Foundation (2015). Growth Within.
Centre for Economics and Business Research (2014). The Future Economic and Environmental Costs of Gridlock in 2030.
European Environment Agency (2017a). Air Quality in Europe 2017.
40
Gehrke, S. R., Felix, A. and Reardon, T. (2018). Fare Choices Survey of Ride-Hailing Passengers in Metro Boston. https://www.mapc.org/farechoices/.
Clewlow, R. and Kulieke, S. (2017). Disruptive Transportation: The Adoption, Utilization, and Impacts of Ride-Hailing in the United States. UCD-ITS-RR-17-07.
Institute of Transportation Studies, University of California, Davis.
Membership in car sharing schemes increased from 0.35 million in 2006 to 2.2 million in 2014, and the number of shared vehicles increased from 8 thousand to
41
40 thousand during the same period (Dr Scott Le Vine et al., 2014).
Dr Scott Le Vine et al. (2014). Car sharing.
Chapter 6.
1
IVA, S. B. (2014). Klimatpåverkan från byggprocessen. http://www.iva.se/publicerat/
2
Eurostat (2018). Share of housing costs in disposable household income, by type of household and income group - EU-SILC survey. 18 May. http://appsso.eurostat.
ec.europa.eu/nui/show.do?dataset=ilc_mded01&lang=en.
3
Building Radar (2016). Construction Industry in Europe. https://buildingradar.com/construction-blog/construction-industry-europe/.
4
Wiedenhofer, D., Steinberger, J. K., Eisenmenger, N. and Haas, W. (2015). Maintenance and Expansion: Modeling Material Stocks and Flows for Residential Buil-
dings and Transportation Networks in the EU25. Journal of Industrial Ecology, 19(4). 538–51. DOI:10.1111/jiec.12216.
Artola, I., Rademaekers, K., Williams, R. and Yearwood, J. (2016). Boosting Building Renovation: What Potential and Value for Europe? http://www.europarl.euro-
pa.eu/RegData/etudes/STUD/2016/587326/IPOL_STU(2016)587326_EN.pdf.
Energy Efficiency Financial Institution Group (2015). Energy Efficiency – the First Fuel for the EU Economy. http://bpie.eu/wp-content/uploads/2015/10/Final_Re-
port_EEFIG_v_9.1_24022015_clean_FINAL_sent.pdf.
5
Sandberg, N. H., Sartori, I., Heidrich, O., Dawson, R., Dascalaki, E., et al. (2016). Dynamic building stock modelling: Application to 11 European countries to
support the energy efficiency and retrofit ambitions of the EU. Energy and Buildings, 132. 26–38. DOI:10.1016/j.enbuild.2016.05.100.
Buildings Performance Institute Europe (BPIE) (2011). Europe’s Buildings under the Microscope. http://bpie.eu/wp-content/uploads/2015/10/HR_EU_B_un-
der_microscope_study.pdf.
6
Buildings Performance Institute Europe (BPIE) (2017). State of the Building Stock Briefing. http://bpie.eu/wp-content/uploads/2017/12/State-of-the-building-
stock-briefing_Dic6.pdf.
Artola et al. (2016). Boosting Building Renovation.
Wiedenhofer et al. (2015). Maintenance and Expansion.
Energy Efficiency Financial Institution Group (2015). Energy Efficiency
7
Plastics Europe (2018). Plastics - the Facts 2016. Plastics Europe. http://www.plasticseurope.org/Document/plastics---the-facts-2016-15787.aspx?FolID=2.
Moynihan, M. C. and Allwood, J. M. (2012). The flow of steel into the construction sector. Resources, Conservation and Recycling, 68. 88–95. DOI:10.1016/j.
resconrec.2012.08.009.
8
European Commission (2016). Construction and demolition waste. http://ec.europa.eu/environment/waste/construction_demolition.htm.
9
Ecorys and Copenhagen Resource Institute (2014). Resource Efficiency in the Building Sector. Rotterdam.
172
Eurostat (2017). Waste statistics - Statistics Explained. http://ec.europa.eu/eurostat/statistics-explained/index.php/Waste_statistics.
European Commission (2018). The Energy Performance of Buildings Directive. https://ec.europa.eu/info/sites/info/files/epbd_factsheet_20180503_dc_v03e_final.

10
pdf.
11
IVA (2014). Klimatpåverkan från byggprocessen.
For example, a recent 300-page book on ‘The Transition to Sustainable Buildings’ from the International Energy Agency makes no mention of materials use or the
12
CO2 that results from the production and use of building materials, but focusses instead 100% on energy efficiency and low-carbon energy.( International Energy
Agency (2013). Transition to Sustainable Buildings).
International Energy Agency, ed. (2013). Transition to Sustainable Buildings: Strategies and Opportunities to 2050. IEA Publ, Paris.
13
IVA (2014). Klimatpåverkan från byggprocessen.
Rossi, B., Marique, A.-F. and Reiter, S. (2012). Life-cycle assessment of residential buildings in three different European locations, case study. Building and Environ-
ment, 51. 402–7. DOI:10.1016/j.buildenv.2011.11.002.
Cuéllar-Franca, R. M. and Azapagic, A. (2012). Environmental impacts of the UK residential sector: Life cycle assessment of houses. Building and Environment,
(54). 86–99. DOI:10.1016/j.buildenv.2012.02.005.
14
Andrew (2018) gives an estimate of 1.5 Gt CO2 in 2014, which fits with the International Energy Agency’s estimate that 63% of the total 2.23 Gt of CO2 were
process emissions.
Andrew, R. M. (2018). Global CO2 emissions from cement production. Earth System Science Data, 10(1). 195–217. DOI:https://doi.org/10.5194/essd-10-195-
2018.
15
This is based on International Energy Agency (2017) and Andrew (2018), which give estimates for 2014. However, Andrew (2018) notes that there has been little
growth since. The number includes indirect emissions from electricity use, which is 12% of total energy used.
Andrew (2018). Global CO2 emissions from cement production.
16
Cao, Z., Shen, L., Løvik, A. N., Müller, D. B. and Liu, G. (2017). Elaborating the History of Our Cementing Societies: An in-Use Stock Perspective. Environmen-
tal Science & Technology, 51(19). 11468–75. DOI:10.1021/acs.est.7b03077.
17
Forecasts used by the international climate modelling community cluster strongly around 4 billion tonnes per year in 2050 and no higher than 5 billion tonnes
by 2100 (Edelenbosch et al. 2017), although a somewhat higher forecast is presented in (van Ruijven et al. 2016). However, this belies the uncertainty about future
cement use. The total amount of cement ‘stock’ in use varies widely even between developed countries, from 15 t per person in the United States, to 20 in France, and
more than 30 in Italy (Cao et al. 2017). Existing forecasts also underestimate the cement intensity of recent economic growth, and especially China’s development,
and thus understate even current production by as much as 500 Mt per year. There clearly is much more uncertainty than the span of commonly used projections
imply. One exception is the Global Calculator project, which explored scenarios up to more than 8 billion tonnes in 2050 (Global
Calculator 2015). The forecast used in this study is for 6 Gt per year in 2050, rising to 7 Gt by 2100.
Edelenbosch, O. Y., Kermeli, K., Crijns-Graus, W., Worrell, E., Bibas, R., et al. (2017). Comparing projections of industrial energy demand and greenhouse gas
emissions in long-term energy models. Energy, 122. 701–10. DOI:10.1016/j.energy.2017.01.017.
van Ruijven, B. J., van Vuuren, D. P., Boskaljon, W., Neelis, M. L., Saygin, D. and Patel, M. K. (2016). Long-term model-based projections of energy use and CO2
emissions from the global steel and cement industries. Resources, Conservation and Recycling, 112. 15–36. DOI:10.1016/j.resconrec.2016.04.016.
Cao et al. (2017) Elaborating the History of Our Cementing Societies
Global Calculator (2015). Global Calculator Technical Documentation: Manufacturing Sector. http://uncached-site.globalcalculator.org/sites/default/files/Manu-
facturing%20technical%20documentation%20part%202.pdf.
18
International Energy Agency and The World Business Council for Sustainable Development (2009). Cement Technology Roadmap 2009: Carbon Emissions
Reductions up to 2050. OECD/IEA and The World Business Council for Sustainable Development. https://www.iea.org/publications/freepublications/publication/
Cement.pdf.
International Energy Agency and The World Business Council for Sustainable Development (2018). Technology Roadmap - Low-Carbon Transition in the Cement
Industry. http://www.iea.org/publications/freepublications/publication/TechnologyRoadmapLowCarbonTransitionintheCementIndustry.pdf.
19
Scrivener, K. L., John, V. M. and Gartner, E. M. (2016). Eco-Effcient Cements: Potential, Economically Viable Solutions for a Low-CO2, Cement-Based Materials
Industry. United Nations Environment Programme.
20
Gartner, E. and Hirao, H. (2015). A review of alternative approaches to the reduction of CO2 emissions associated with the manufacture of the binder phase in
concrete. Cement and Concrete Research, 78. 126–42. DOI:10.1016/j.cemconres.2015.04.012.
World Economic Forum (2016). Shaping the Future of Construction. http://www3.weforum.org/docs/WEF_Shaping_the_Future_of_Construction_full_report__.

21
pdf.
22
European Commission (2015). Report on Single Market Integration and Competitiveness in the EU and Its Member States. http://publications.europa.eu/resource/
cellar/93a2b2cc-7d5a-11e5-b8b7-01aa75ed71a1.0001.03/DOC_2.
23
Wiedenhofer et al. (2015). Maintenance and Expansion.
173
Moynihan, M. C. and Allwood, J. M. (2014). Utilization of structural steel in buildings. Proceedings. Mathematical, Physical, and Engineering Sciences / The
24
Royal Society, 470(2168). DOI:10.1098/rspa.2014.0170.
Moynihan, M. C. (2014). Material Efficiency in Construction. UIT Cambridge.
25
In the case of steel, key reason for the over-use is that it is easier to use a smaller number of different section designs, as this cuts labour costs both during design and
construction (Moynihan 2014). There is much less evidence on the degree of over-use of concrete, the other main structural element, but industry experts suggested
that significant savings could be made if it were made a focus of the construction process.
Moynihan (2014). Material Efficiency in Construction.
26
Jernkontoret Research (2013). The Steel Eco-Cycle - Environmental Research Programme. Closing the Loop in the Manufacture and Use of Steel in Society. Scien-
tific Report 2004-2012. http://www.jernkontoret.se/en/publications/steel-research/open-reports-serie-d/d-853/.
27
Larsen, I. L., Aasbakken, I. G., O’Born, R., Vertes, K. and Thorstensen, R. T. (2017). Determining the Environmental Benefits of Ultra High Performance Concre-
te as a Bridge Construction Material. IOP Conference Series: Materials Science and Engineering, 245(5). 052096. DOI:10.1088/1757-899X/245/5/052096.
28
Auken, I. (2018b). This Danish city is rebuilding itself out of recycled rubble. World Economic Forum. https://www.weforum.org/agenda/2018/03/copenha-
gen-denmark-rebuilding-recycled-rubble/.
Salama, W. (2017). Design of concrete buildings for disassembly: An explorative review. International Journal of Sustainable Built Environment, 6(2). 617–35.
29
DOI:10.1016/j.ijsbe.2017.03.005.
Gorgolewski, M. (2008). Designing with reused building components: some challenges. Building Research & Information, 36(2). 175–88.
30
DOI:10.1080/09613210701559499.
31
Steelconstruction.info (n.d.). The recycling and reuse survey. www.steelconstruction.info/The_recycling_and_reuse_survey.
It is striking how much more has been written by researchers than by companies on the opportunities for re-use. Examples include Iacovidou and Purnell (2016);
32
Dunant et al. (2017); Dunant et al. (2018); Ness et al. (2015); Densley Tingley et al. (2017).
Iacovidou, E. and Purnell, P. (2016). Mining the physical infrastructure: Opportunities, barriers and interventions in promoting structural components reuse. Science
of The Total Environment, 557–558. 791–807. DOI:10.1016/j.scitotenv.2016.03.098.
Dunant, C. F., Drewniok, M. P., Sansom, M., Corbey, S., Allwood, J. M. and Cullen, J. M. (2017). Real and perceived barriers to steel reuse across the UK construc-
tion value chain. Resources, Conservation and Recycling, 126. 118–31. DOI:10.1016/j.resconrec.2017.07.036.
Dunant, C. F., Drewniok, M. P., Sansom, M., Corbey, S., Cullen, J. M. and Allwood, J. M. (2018). Options to make steel reuse profitable: An analysis of cost and
risk distribution across the UK construction value chain. Journal of Cleaner Production, 183. 102–11. DOI:10.1016/j.jclepro.2018.02.141.
Ness, D., Swift, J., Ranasinghe, D. C., Xing, K. and Soebarto, V. (2015). Smart steel: new paradigms for the reuse of steel enabled by digital tracking and modelling.
Journal of Cleaner Production, 98. 292–303. DOI:10.1016/j.jclepro.2014.08.055.
Densley Tingley, D., Cooper, S. and Cullen, J. (2017). Understanding and overcoming the barriers to structural steel reuse, a UK perspective. Journal of Cleaner
Production, 148. 642–52. DOI:10.1016/j.jclepro.2017.02.006.
33
A survey in Finland found that, in a country with one of the highest demolition rates in the EU, only 9% of demolition took place because the buildings had
reached the end of their technical life (Hradil et al. 2014). Interviews confirm that this is common practice across the EU: most buildings reach their end-of-life not
because they are no longer structurally sound, but for other reasons.
Hradil, P., Talja, A., Wahlström, M., Huuhka, S., Lahdensivu, J. and Jussa Pikkuvirta (2014). Re-Use of Structural Elements; Environmentally Efficient Recovery of
Building Components. VTT Technical Research Centre of Finland. DOI:10.13140/2.1.1771.9363.
34
Méquignon, M. and Ait Haddou, H. (2014). Lifetime Environmental Impact of Buildings. SpringerBriefs in Applied Sciences and Technology. Springer Internatio-
nal Publishing, Cham. DOI:10.1007/978-3-319-06641-7.
Ellen McArthur Foundation, SUN and McKinsey Center for Business and Environment (2015). Growth Within: a Circular Economy Vision for a Competitive
35
Europe. https://www.ellenmacarthurfoundation.org/assets/downloads/publications/EllenMacArthurFoundation_Growth-Within_July15.pdf.
Huuhka, S., Kaasalainen, T., Hakanen, J. H. and Lahdensivu, J. (2015). Reusing concrete panels from buildings for building: Potential in Finnish 1970s mass
36
housing. Resources, Conservation and Recycling, 101. 105–21. DOI:10.1016/j.resconrec.2015.05.017.
37
Nusselder, S., Maqbool, A. S., Deen, R., Blake, G., Bouwens, J. and Fauzi, R. T. (2015). Closed Loop Economy: Case of Concrete in the Netherlands. TU Delft.
https://www.slimbreker.nl/downloads/IPG-concrete-final-report(1).pdf.
38
According to SmartCrusher , this is possible by using an specially developed grinder that does not break through the sand and gravel but only crushes the cement
stone, making it possible to separate unhydrated cement, hydrated cement, sand and gravel from each other (Nusselder et al. 2015).
Nusselder et al. (2015) Closed Loop Economy
39
44% is saved one a one-to-one basis (Ecofys et al. 2009), and in combination with 15% less cement input the CO2 savings can be up to 52% (Nusselder et al.
2015).
Ecofys, Fraunhofer Institute for Systems and Innovation Research and Öko-Institut (2009). Methodology for the Free Allocation of Emission Allowances in the EU
ETS Post 2012. https://ec.europa.eu/clima/sites/clima/files/ets/allowances/docs/bm_study-lime_en.pdf.
174
Nusselder et al. (2015). Closed Loop Economy
40
Scrivener et al. (2016). Eco-Effcient Cements.
41
Ramage, M. H., Burridge, H., Busse-Wicher, M., Fereday, G., Reynolds, T., et al. (2017). The wood from the trees: The use of timber in construction. Renewable
and Sustainable Energy Reviews, 68. 333–59. DOI:10.1016/j.rser.2016.09.107.
42
See Ramage et al. (2017). and Kuilen et al. (2011).
Ramage et al. (2017). The wood from the trees.
Kuilen, J. W. G. V. D., Ceccotti, A., Xia, Z. and He, M. (2011). Very Tall Wooden Buildings with Cross Laminated Timber. Procedia Engineering, 14. 1621–28.
DOI:10.1016/j.proeng.2011.07.204.
43
These are emissions from the production step for each materials category, including direct CO2 emissions as well as CO2 from the use of electricity. The calculation
accounts for the overlaps of jointly pursuing several strategies at the same time.
44
World Economic Forum (2016). Shaping the Future of Construction.
45
Balfour Beatty (2017). Innovation 2050: A Digital Future for the Infrastructure Industry. https://www.balfourbeatty.com/media/244600/balfour-beatty-innova-
tion-2050.pdf.
46
Wilkinson, C. (2018). Bot the builder: the robot that will replace bricklayers. Financial Times, 23 February. https://www.ft.com/content/db2b5d64-10e7-11e8-
a765-993b2440bd73.
47
Future of Construction (2016c). Winsun. https://futureofconstruction.org/case/winsun/.
175
a Powerful Force for Climate Mitigation
Transformative innovation for prosperous
and low-carbon industry
This report investigates how a more circular economy can

contribute to cutting CO2 emissions. It explores a broad range of
opportunities for the four largest materials in terms of emissions
(steel, plastics, aluminium, and cement) and two large use seg-
ments for these materials (passenger cars and buildings).
The key conclusion is that a more circular economy can make

deep cuts to emissions from heavy industry: in an ambitious sce-
nario, as much as 296 million tons CO2 per year in the EU by
2050, out of 530 Mt in total – and some 3.6 billion tonnes per year
globally. Making better use of the materials that already exist in
the economy thus can take EU industry halfway towards net-zero
emissions. Moreover, doing so often is economically attractive. In-
itiatives for a more circular economy therefore deserve a central
place in EU climate and industrial policy.
Material Economics Sverige AB 176

Grev Turegatan 30, 114 38 Stockholm, Sweden materialeconomics.com

Material Economics Circular Economy

Uploaded by

Copyright:

Available Formats

Material Economics Circular Economy

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Material Economics Circular Economy

Uploaded by

Copyright:

Available Formats

The Circular Economy

a Powerful Force for Climate Mitigation

THE CIRCULAR ECONOMY – A POWERFUL FORCE FOR CLIMATE MITIGATION (2018)

Per-Anders Enkvist Per Klevnäs

EU EMISSIONS REDUCTIONS POTENTIAL FROM A MORE CIRCULAR ECONOMY, 2050

STEEL PLASTICS ALUMINIUM CEMENT

2050 BASELINE MATERIALS PRODUCT MATERIALS CIRCULAR 2050 CIRCULAR

MATERIALS PRODUCTION EMISSIONS, 2015 AND 2050

Global materials demand is set to increase 2-

With current patterns of materials use, global demand

Growth in materials use risks exhausting

the carbon budget and imperil climate objectives

Existing scenarios imply that at most 300 G t of the remaining

CUMULATIVE INDUSTRY CO2 EMISSIONS IN 2ºC SCENARIOS, 2015 – 2100

Materials production alone risks exceeding the total

CO2 EMISSIONS AND CARBON BUDGET

CARBON BUDGET TO 2100 CO2 EMISSIONS FROM MATERIALS PRODUCTION

2°C CARBON BUDGET CARBON BUDGET MATERIALS MATERIALS

Low-carbon energy is not enough: even with 100% low-carbon

CUMULATIVE EMISSIONS, 2015 – 2100

Best available energy efficiency fully

Emissions intensity reduced by half through full

Half of production emissions eliminated. (e.g.,

38 Zero-carbon electricity used for electrolysis and

All energy-related emissions (fuel, electricity,

Plastic emissions include emissions from production as well as embedded emissions

Source: Material Economics analysis as described in subsequent chapters of this report.

A MORE CIRCULAR ECONOMY IS INDISPENSABLE FOR CLIMATE TARGETS

EU EMISSIONS REDUCTIONS POTENTIAL FROM A MORE CIRCULAR ECONOMY, 2050

STEEL PLASTICS ALUMINIUM CEMENT

2050 BASELINE MATERIALS PRODUCT MATERIALS CIRCULAR 2050 CIRCULAR

MATERIALS PRODUCTION EMISSIONS, 2015 AND 2050

Without deep transformation,

FOUR KEY MATERIALS CATEGORIES

STEEL PLASTICS ALUMINIUM CEMENT

TWO KEY VALUE CHAINS AND PRODUCT CATEGORIES

PASSENGER CARS AND BUILDINGS

materials production barely decline by 2050

MATERIALS PRODUCTION EMISSIONS, 2015 AND 2050

SOURCE : MATERIAL ECONOMICS MODELLING AS DESCRIBED IN TEXT.

STEEL: PRIMARY VS. RECYCLED

Three circular strategies work

Three circular economy strategies make better use of

GHG MATERIALS PRODUCT

High value recycling, Improved production Higher utilisation and

1. MATERIALS RECIRCULATION STRATEGIES COULD CURB EMISSIONS BY 178 Mt CO 2 PER YEAR

Materials recycling cuts CO 2 emissions

TODAY 2050 TODAY 2050

ALUMINIUM: PRIMARY VS. RECYCLED PRIMARY CEMENT: PRIMARY VS. RECYCLED

TODAY 2050 TODAY 2050

SOURCE : MATERIAL ECONOMICS ANALYSIS AS DESCRIBED IN SUBSEQUENT CHAPTERS OF THIS REPORT.

Circular measures can reduce the primary steel required

PRIMARY STEEL USED FOR MOBILITY SERVICES

MATERIALS RECIRCULATION PRODUCT MATERIALS EFFICIENCY CIRCULAR BUSINESS MODELS

I MATERIALS RECIRCULATION II MORE MATERIAL III IMPROVED PRODUCTIVITY OF USE

NOTE : ANALYSIS IS BASED ON GLOBAL STEEL PRODUCTION.

Demand- and supply-side measures are both

The role of demand- and supply-side measures in decarbonising

2,0 REMAINING EMISSIONS AFTER TOTAL EMISSIONS

Globally, a more circular economy