Sign In

Create an account

Create a free IEA account to download our reports or subcribe to a paid service.

Join for freeJoin for free


Not on track
Shutterstock 697273432

About this report

The direct CO2 intensity of aluminium production remained relatively flat during the past couple of years. In the Net Zero Emissions by 2050 Scenario, however, emissions intensity declines 3% annually to 2030. Getting on track with this scenario will require improved end-of-life scrap collection and sorting to enable greater production from scrap, as well as further development of new technologies to reduce emissions from primary production.

Governments can stimulate action by better coordinating aluminium scrap collection and sorting, funding RD&D and adopting mandatory CO2 emissions reduction policies.

Direct CO2 intensity of aluminium production in the Net Zero Scenario, 2018-2030

Tracking progress

Primary aluminium production is highly energy-intensive, with electricity making up a large share of the energy consumed. Getting on track with the Net Zero by 2050 Scenario will require efforts on multiple fronts. 

Material efficiency strategies can help maximise the collection of post-consumer scrap to enable greater secondary production and reduce the total amount of metal used while delivering the same services. 

Meanwhile, R&D is needed on innovative alternative production methods that reduce primary production process and combustion emissions, and more energy-efficient equipment and operations would be beneficial.  

Given the considerable amount of electricity consumed in the aluminium subsector, decarbonising its power sources would help reduce indirect emissions and is thus a key complement to reducing direct aluminium emissions. 

The global energy intensity of overall aluminium production fell by 1.2% in 2019, similar to average annual reductions over the past several years. This includes both primary production from bauxite ore and secondary production from scrap. Primary production is approximately ten times more energy-intensive than secondary production.  

Primary aluminium production involves two key steps: alumina refining, to refine bauxite ore into alumina (aluminium oxide), and aluminium smelting, to convert alumina to pure aluminium. 

Average annual drops in alumina refining energy intensity for 2019 and 2020 were similar to those during 2010-2018 (an average 3.0% per year). Meanwhile, the global average energy intensity of aluminium smelting has trended slightly upwards in the past three years (an average 0.3% per year), in contrast with declines during 2010-2017 (an average -0.6% per year). 

Falling global energy intensity of alumina refining and primary aluminium smelting since 2000 has resulted largely from developments in China. Capacity expanded significantly over the past two decades, such that China accounted for more than half of global primary production in 2020. This enabled robust energy intensity declines and has put China’s average aluminium smelting energy intensity at the best-available-technology performance level since 2014. With China’s potential for energy intensity improvements essentially fully exploited, global average reductions have become smaller in the past few years. 

Energy intensity of primary aluminium smelting by region, 2000-2020


Energy intensity of alumina refining by region, 2000-2020


 The proportion of aluminium produced from recycled metal (secondary production) also affects the overall average energy intensity of aluminium production. In 2019, 34% of aluminium produced came from new and old scrap, of which 58% was old (new scrap refers to scrap created during product manufacturing, while old scrap refers to end-of-life scrap; internal scrap produced in aluminium fabrication facilities is not included here). The share of secondary production has remained relatively constant at 31-33% since 2000, with 34% in 2019 the highest share during this period. 

Scrap-based production tends to cost less than primary production, so the key constraint is scrap availability. In 2019, collection rates for aluminium were over 95% for new scrap and just over 70% for old. While these rates are high, there is potential to improve old-scrap collection. Still, aluminium remains locked within products until their lifetime ends, so even with better collection rates there is an upper limit on potential for recycled production. 

For alignment with the Net Zero Emissions by 2050 Scenario, it will be important to continue reducing the energy intensities of primary and secondary aluminium production, and to expand secondary production by improving old-scrap collection and sorting, as well as reducing losses within the recycling system. The global energy intensity of aluminium production overall (primary and secondary combined) needs to fall at least at 0.8% annually to 2030.  

An increasing share of secondary production will be the primary catalyst of energy intensity improvements. The combined share of aluminium produced from recycled new and old scrap needs to reach nearly 40% (at least 70% of this from old scrap) by 2030 to attain the Net Zero Emissions by 2050 pathway. Achieving this share will require better scrap collection and sorting, particularly for old scrap, since stronger material efficiency efforts under the scenario will reduce the availability of new scrap.

If the aluminium emissions boundary is enlarged to include indirect emissions from power generated for use in aluminium production, those power emissions would currently account for just under 70% of total (direct plus indirect) global aluminium emissions.  

A considerable portion of these emissions are within the control of the aluminium industry, given that about 55% of power consumed by the industry globally is self-generated rather than purchased from the grid. The share of self-generation is particularly high in Asia (~65% in China and over 95% in the rest of Asia), and moderate in North and South America (~45%). Meanwhile, most power for aluminium production is purchased in Europe, Africa and Oceania. 

The average world power mix supplying the aluminium industry differs considerably from the average total global power mix. Hydropower is currently used for about 25% of global aluminium production – even though it accounts for only 15% of the total power mix – but this share has fallen since 2010, when 40% of aluminium production was fuelled by hydropower. The shift is largely due to expanding aluminium production in China powered by coal-based electricity, where coal supplies close to 90% of production. Meanwhile, in Europe, North America and South America, hydro still supplies 75% or more of production.  

In the Net Zero Emissions by 2050 Scenario, the emissions intensity of the total power mix declines roughly 75% from today’s level by 2030. The aluminium industry should aim to reduce the intensity of its power supply by at least this much, including by reducing reliance on coal-generated power. 

Global aluminium industry power mix compared with the global total power mix, 2020


Global aluminium industry power mix compared with the global total power mix, 2010


Global aluminium production remained flat in 2019 and 2020, in contrast with average annual 5.5% growth registered in 2010-2018. Despite this temporary slowdown, production is expected to continue expanding, driven by population and GDP growth. Clean energy transitions will also impact aluminium demand, with potential for upward pressure from technology shifts that require greater use of aluminium, e.g. for lightweight vehicles and solar energy.1

Adopting material efficiency measures can help curb demand growth, however. Examples include reducing scrap generation during fabrication and manufacturing, reusing old scrap, and designing products with recycling in mind. Although demand growth in the Net Zero Emissions by 2050 Scenario slows to an average annual rate of 0.8%, this still represents nearly 10% growth in total demand from the current level. Measures will therefore be needed to limit rises in energy and emissions intensities. 

Global aluminium production in the Net Zero Scenario, 2010-2030


Although energy efficiency and scrap-based production are important for alignment with the Net Zero Emissions by 2050 Scenario, on their own they cannot decarbonise the subsector. Transformational change is required, particularly to deal with process emissions from primary production, and the groundwork for breakthrough technologies needs to be laid before 2030.  

Some solid progress has been made on this front in recent years. Currently, primary aluminium smelting relies on carbon anodes, which produce CO2 as they degrade. Innovation efforts are under way to develop inert anodes, which are made from alternative materials, do not degrade, and produce pure oxygen rather than CO2. Two key initiatives have made considerable progress in the past couple of years: 

  • In 2018, Alcoa and Rio Tinto announced they have developed an inert anode technology and have formed a joint venture called Elysis to further develop the technology. Construction of its first commercial-scale prototype cells at a smelter in Quebec, Canada, began in June 2021. They are aiming to complete demonstration by 2024, with commercialisation to follow. 
  • RUSAL's Krasnoyarsk plant in Russia has produced primary aluminium using inert anode technology at industrial scale (1 tonne of aluminium per day per cell). Test deliveries of a pilot batch of aluminium commenced in the spring of 2021, and the company aims for mass-scale production by 2023. 

Applying CCS could be another option for low-emission aluminium smelting, although it is still at the concept stage because the very low concentration of CO2 in exhaust gases presents a major challenge. In France, initial investigations are under way to determine a capture technology suitable for aluminium smelters. 

Another area of innovation involves adapting aluminium production to provide flexibility to the power grid, given that aluminium smelters are major electricity consumers. This will become increasingly important as the share of variable renewable power rises.  

In May 2019, TRIMET began the first successful industrial-scale operation of the EnPot demand-response technology, consisting of 120 pots at its plant at Essen, Germany, following a smaller 12-pot trial that began five years earlier. A 10-pot trial is also being installed at Hamburg. This “virtual battery” concept relies on adjustable heat exchangers that can maintain the energy balance in each electrolytic cell irrespective of shifting power inputs. 

Furthermore, it will be important to develop alternative ways to produce heat for alumina refining, which currently relies primarily on fossil fuels. The use of 30% biomass has been successfully tested in Australia, while a consortium – also in Australia – is working to obtain 50% of energy from concentrated solar power. 

Expanding secondary production through better scrap collection and sorting will be important to raise energy efficiency and decarbonise the aluminium subsector. 

Stakeholders should work to increase scrap collection and recovery by improving recycling channels and sorting methods, and by better connecting participants along supply chains. Focusing on end uses that currently have low collection rates will be important. 

The aluminium industry, aluminium product manufacturers and waste collectors can work together to ensure that manufacturing and end-of-life scrap is channelled back to aluminium producers. Engineers should consider reusability and recyclability in product design, and governments can assist by setting requirements and coordinating channels for end-of-life material reuse and recycling. 

Participants all along the value chain (aluminium producers, engineers, construction companies and product manufacturers) can also adopt material efficiency strategies to reduce overall aluminium demand. 

Furthermore, ensuring efficient equipment operations and maintenance will help guarantee optimal energy performance. This can be reinforced by implementing energy management systems. 

Reducing emissions from primary production is important, as scrap availability will put an upper limit on the potential for secondary production. 

The aluminium industry should prioritise RD&D of alternative production methods that reduce process emissions from primary production, such as the use of inert anodes. R&D will be needed within the next decade to enable widespread deployment post-2030. 

Governments and financial investors need to increase support for RD&D, particularly to advance the large-scale demonstration and deployment of technologies that have already shown promise. 

Public-private partnerships can help, as can green public procurement and contracts for difference that generate early demand and enable producers to gain experience and bring down costs. Government coordination of stakeholder efforts can also direct focus to priority areas and avoid overlap. 

Given the high electricity requirements of aluminium production, efforts to decarbonise the grid will be necessary to reduce the subsector’s indirect emissions. 

The aluminium subsector can in turn assist with grid decarbonisation by providing flexibility services that would help integrate a higher portion of variable renewables. Electricity producers can help by offering electricity pricing incentives to aluminium producers using demand management systems.

Policymakers can promote CO2 emissions reduction efforts by adopting mandatory reduction policies, such as a gradually rising carbon price or tradeable industry performance standards that require the average CO2 intensity for production of each key material to decline across the economy and that permit regulated entities to trade compliance credits. 

Adopting these policies at lower stringencies in the short term (i.e. within the next three to five years) will provide an early market signal, enabling industries to prepare and adapt as stringency increases over time. It can also help reduce the costs of low-carbon production methods, softening the impact on aluminium prices in the long term. 

Complementary measures may be useful in the short to medium term, such as differentiated market requirements (i.e. a government-mandated minimum proportion of low-emission aluminium in targeted products). 

Ideally, these policies would be applied globally at similar strengths. Since aluminium is highly traded, measures will be needed to help ensure a level global playing field if the strength of policy efforts differs from one region to another. Possibilities include adopting border carbon adjustments or the free allocation of allowances for emissions below a targeted benchmark in an emissions trading system.  

Governments can extend the reach of their efforts by partaking in multilateral forums to facilitate low-carbon technology transfer and to encourage other countries to also adopt mandatory CO2 emissions policies. 

Improving the collection, transparency and accessibility of energy performance and CO2 emissions statistics on the aluminium subsector would facilitate research, regulatory and monitoring efforts (including, for example, multi-country performance benchmarking assessments).  

Better data on recycled production levels, the energy intensity of recycled production, and scrap availability are particularly needed. Industry participation and government coordination will both be important to improve data collection and reporting. 

Notes and references
  1. Solar energy systems use aluminium for various components, including for mounting and framing solar PV panels and for reflectors in concentrating solar power systems.

  2. IAI (International Aluminium Institute) (2021), Current IAI Statistics,