Phase-in and phase-out policies in the global steel transition

ABSTRACT

To reach the goals of the Paris Agreement, global emissions should be reduced to net zero by mid-century. The steel sector is an emission-intensive industrial subsector where low-carbon production routes are emerging, and recent studies have shown that rapid decarbonization of the sector is technically possible. However, several barriers block the sector-wide diffusion of low-carbon steelmaking. Inertia and barriers to exit inhibit the closure of emission-intensive plants, thus driving overcapacity and trade conflicts which in turn risk undermining the global steel transition. Drawing on the industrial transitions literature, we find that phase-out policy has a key role to play in overcoming inertia and barriers to exit, increasing the pace of exits in the steel sector, and enabling market space for low-emission steelmaking. Still, reviewing policy mixes in the top four steelmaking jurisdictions, we observe that these are primarily oriented towards phasing-in low-emission capacity rather than phasing-out emission-intensive capacity. In an analysis of low-emission steel projects in the LeadIT Green Steel Tracker, we find that almost half of these projects are financially backed by government, revealing that support for phase-ins is sparking a renaissance for subsidies in the steel sector. At the same time, we find that green steel projects, in aggregate, are increasing total steelmaking capacity. To minimize overcapacity and trade conflicts in the steel transition, policymakers should develop new corresponding phase-out policies that support and increase the pace of closures, to enable a rapid sector-wide diffusion of low-emission steelmaking.

Key policy insights

• While rapid emission reductions are technically possible in the steel sector, frictions such as overcapacity and trade conflicts risk impeding the transition.

• The steel sector’s inertia and barriers to exit are exacerbating these frictions, and targeted phase-out policies are needed to overcome these blockages.

• Steel decarbonization policies in top steelmaking jurisdictions are oriented towards phasing-in green steel, not on phasing-out emission-intensive steelmaking.

• Green steel projects are increasing total steelmaking capacity, and to a high degree are financially supported by governments.

KEYWORDS:

• Decarbonization
• transition
• steel
• phase-out
• policy

1. Introduction

The Paris Agreement implies that global greenhouse gas emissions should peak as soon as possible and reach net zero in 2050. Reaching zero emissions requires phasing out the use of unabated fossil fuels across all sectors of society. Industry directly contributed 24% of total emissions and 34% including indirect emissions from power and heat in 2019, second only to the energy supply sector (IPCC, 2022). Among industrial sectors, steel is a major emitter, contributing 7% of total energy system emissions in 2019 (International Energy Agency, 2019).

A transition to steelmaking with very low CO₂ emissions (less than 95% compared to current practices) is emerging, supported by various policies such as direct subsidies, export credits and loans, green procurement, and in some regions (notably the EU) a high-cost penalty for emitting CO₂ from traditional steelmaking. At the time of writing, the Green Steel Tracker (Vogl et al., 2023) presents 81 projects in 24 countries that target green steel production, with the potential to significantly reduce CO₂-emissions from primary steelmaking. This development started in the EU but is quickly spreading to other parts of the world, not least because the main green steel option today – replacing coal with green hydrogen from renewables – is increasingly competitive in countries with access to low-cost renewable electricity, including Australia, Brazil and Oman.

Several scenarios developed over the past years illustrate pathways on how to decarbonize the global steel production (see e.g. Agora Energiwende & Wuppertal Institute, 2020; Agora Industry & Wuppertal Institute, 2023; Bataille et al., 2021; Delasalle et al., 2022; International Energy Agency, 2022; Yu et al., 2021). These studies have in common that they assume a rational and frictionless transition where older emission-intensive capacities are phased out at the time of the next relining, and concurrently with the introduction of new green capacities. Steel capacities are thus assumed to develop according to the overall market demand. However, new investments in green steel production capacity will only result in absolute global CO₂ reductions if they ‘crowd-out’ or replace conventional high-emitting steel production capacities. Furthermore, most future scenarios suggest that cost-effective steel decarbonization requires maximized secondary steelmaking, and a limit to overall steel consumption (Bashmakov et al., 2022). Hence, high-emitting capacity needs to be phased out faster than the introduction of green primary capacities. However, due to high barriers to exit (Rimini et al., 2020) and inertia (Grubb, Chapuis, & Duong, 1995; Wesseling et al., 2017), the sector is slow to adjust capacity downwards.

‘Deliberate decline’ is an emerging field for studying how transitions can be enabled by overcoming lock-ins to existing systems by intentionally introducing policies for destabilizing or phasing-out unsustainable substances, technologies or systems (Kivimaa & Kern, 2016; Rogge & Reichardt, 2016; Rosenbloom & Rinscheid, 2020; Trencher et al., 2023). Phase-outs are often introduced gradually over a given timeframe, to minimize social disturbance and frictions (Trencher et al., 2022). Destabilization (or phase-outs) represents the ‘flipside’ of the innovation process central to transitions and has been called ‘exnovation’ due to its effects on the innovation process (Arne Heyen et al., 2017; David, 2017). It also has a wider transformative effect on market demand in the relevant sectors as ‘creating space in future markets, destabilization policies would perhaps best be seen as “transformative” demand pull instruments to be added to the policy mix to accelerate transition processes’ (Rogge & Johnstone, 2017). Policies for deliberate decline and phase-outs have been studied for the coal sector, internal combustion engines and several substances such as mercury and lead (Trencher et al., 2023), but so far not for the steel sector (or any other heavy industry), and not as an integral part of a policy package for a global sectoral transition.

The aim of this article is to analyse the need for a deliberate decline approach to global steel decarbonization policy and to discuss the how the gaps in the emerging policy landscape may drive frictions that impede the transition. This article thus contributes to the literature on steel decarbonization, first by showing how a lack of a coordinated approach to phase-ins and phase-outs may spark transition frictions – i.e. political and economic frictions related to the growth in capacity and competition for green steel market shares – that risk impeding the transition. See Burgess et al. (2024) for a discussion of frictions at a national level. We thereby contribute to the transitions literature and steel phase-out literature, such as Vogl (2023); Vogl et al. (2021), by making a global phase-out policy study on a heavy industry sector.

The remainder of the paper is structured as follows. The next section outlines the methodology of the paper. The third section discusses the emission intensity of steel and strategies for decarbonization. In section four, the underlying theory of inertia and barriers to exit in the steel sector is introduced. The fifth section presents the empirical analysis. Thereafter, the findings are discussed before a final conclusion of the paper.

2. Methodology

First, we review and analyse the literature on steel decarbonization, identify the main elements and discuss their interrelatedness. Second, we review the literature on inertia and barriers to exit in the steel sector, which we find both in more recent scholarly work on industrial decarbonization and in the steel industry literature, i.e. grey literature and academic literature on the characteristics of the steel industry. Insights are then discussed using the theoretical lens of phase-outs and transitions. The empirical work consists of two parts, where we analyse both projects and policies.

The project analysis is based on the LeadIT Green Steel Tracker dataset. This dataset includes investment announcements for near-zero emission primary production of iron and steel and includes a methodological protocol for the selection of projects (see Vogl et al., 2023). Each project announcement in the Tracker is further analysed through press releases, announcements and news articles regarding whether they replace or adapt an existing plant or imply the construction of a new plant. Thereafter, each project is reviewed again for whether it has been supported by government funding. This is done by analysing whether subsidies, credit guarantees, investment support, investment from public banks, or OPEX support has been announced in relation to the project.

For the study of policies, we conduct a document analysis compiling information from scholarly publications, reports from international organizations such as the IEA and the OECD, and private consultancies, as well as official government publications aiming to review the major relevant policies affecting steel decarbonization. To triangulate the results from the document analysis, and as well to guide further desk research, eight semi-structured, open-ended interviews were conducted with steel experts across the four major steelmaking jurisdictions China, the European Union, the United States, and India. The policy study is delimited to these top four jurisdictions due to their large share in global steelmaking. We are aware that the project and policy landscape for global steel decarbonization is emerging and rapidly changing; however, the aim is to provide a snapshot to guide policy insights.

3. Global steel production and the making of green steel

Today, steel is produced through two main routes: the iron ore-based primary route which constitutes around 75% of world production, and the scrap-based secondary route constituting the remaining 25% of world production (Bashmakov et al., 2022). In the dominating primary route, iron ore is reduced into iron in a blast furnace by applying carbon monoxide made from coking coal, and the iron is thereafter converted into steel in a basic oxygen furnace. This route is highly emission-intensive, producing around 2.3 tonnes of CO₂ per tonne of crude steel globally (WorldSteel, 2023), albeit with significant regional variations in emission intensity. The secondary route uses steel scrap as an input that is smelted in an electric arc furnace (EAF). This is less emissions-intensive at 0.7 tonnes of CO₂ per tonne of steel (WorldSteel, 2023), and emissions are mainly restricted to the electricity powering the electric arc furnace, making the secondary route easier to fully decarbonize. However, the problem with the secondary steelmaking route is producing steel at sufficiently high quality, as tramp elements embodied in the scrap lower the purity and thus the quality of the steel (Watari et al., 2020). There is also a limitation regarding access to scrap, which is especially a problem for growing economies with low levels of steel in stock and increasing steel demand (Watari, Hata, et al., 2023). The remaining 5% of steel is produced using a direct reduction furnace, which reduces the iron using fossil gas into direct reduced iron (DRI), also called ‘sponge iron’, which is then melted in an electric arc furnace or in some cases through a basic oxygen furnace (BOF). This route produces about 1.7 tonnes of CO₂ per tonne of steel, including emissions from purchased electricity (WorldSteel, 2023).

Since conventional primary steel is produced with technology that is difficult or impossible to adapt into a zero-emission plant (for example, a blast furnace cannot be fuelled solely by hydrogen), firms will have to retire physical assets and replace them with new ones (Grubb, Mercure, Salas, Lange, & Sognnaes, 2018). This has to be done rapidly, as iron and steelmaking capacity that uses unmitigated fossil fuels should be phased out by mid-century to decarbonize in line with the Paris Agreement (Bataille et al., 2021). However, the global fleet of blast furnaces is young, with many years left to reach the typical lifetime of a blast furnace of 40 years (International Energy Agency, 2019). At the same time, new conventional emission-intensive capacity is continuing to be added, which also will have to be phased out by mid-century; see Figure 1.

Figure 1. Blast furnace capacity (MT) announced and under construction in the top 10 countries (Swalec & Grigsby-Schulte, 2023).

Considering the remaining lifetime of the global fleet of ironmaking assets, they will not be able to run to the full extent of the typical lifetime with multiple relinings, as estimated by the IEA. Understood differently, if the exit rate of assets remains at the historic level, the number of highly emitting steelmaking assets phased out by 2050 will not provide the emissions reductions needed from the steel sector (Rimini et al., 2020). However, Vogl et al. (2021) argue that the point of relining – with a historical median relining cycle of 19, 16 and 10.5 years for the first, second and third cycles, respectively – provides a window of opportunity for phasing-out conventional steelmaking assets, as capital expenditures for relining lie at one-third to one-half of costs for constructing a new blast furnace.

3.1. Elements of steel decarbonization

No single strategy alone can decarbonize steelmaking, and several elements are needed in parallel: increased material and energy efficiency, reduced steel demand, and low-emission breakthrough technologies, many of which use renewables (Pathak et al., 2022; Vogl et al., 2018). These elements affect both how much steel is produced and demanded, and how and where steel will be produced in a decarbonized world.

3.1.1. Lower steel use

Material efficiency in use, reuse, and recycling is a key measure to lower steel emissions. In the International Energy Agency’s Sustainable Development Scenario, where temperature increase is limited to ‘well-below 2 degrees’, steel demand only grows by 10% from 2019 to 2050 (International Energy Agency, 2019). In the Net-Zero Scenario – reaching the target of net-zero emissions by 2050 – steel demand stagnates until 2050 (International Energy Agency, 2023). This is because while key low-carbon technologies require steel, this growth is more than offset by a larger increase in material efficiency and lower steel demand, as demand falls in advanced economies, and emerging economies mature and require less infrastructure investment and construction. Therefore, while steel demand for energy infrastructure increases, demand for transport and buildings fall until 2030 and 2050. According to the IPCC, steel demand can be reduced by up to 40% compared to a business-as-usual scenario through material efficiency (Pathak et al., 2022), and lowering demand will be key for meeting climate targets (Creutzig et al., 2022). Gast and Allwood (2023) argue that a lack of biomass, renewable hydrogen and CCS will limit the supply of low-emission steel, and supply will be 300–930 mt lower than demand in 2050, necessitating significant restrictions of global steel demand to meet global climate ambitions.

3.1.2. Technological change

Technological change is a second element of steel decarbonization, where new production routes with very low, or zero emissions replace conventional primary steelmaking. Producing steel with low emissions requires increasing the share of secondary steelmaking and decarbonizing its power production while using new technologies for decarbonizing primary steel. The two main paths for very low-emission steelmaking (here defined as <95% of emissions from the traditional BF-BOF route) are hydrogen-based steelmaking and coal-based steel with carbon capture and storage (CCS) and a partial replacement of coal with biomass where possible (Vogl et al., 2021). Hydrogen can be used in a direct reduction process replacing fossil gas for later smelting in an electric arc furnace (Vogl et al., 2018). The DRI can be blended with scrap to ensure sufficient quality of the product while increasing the rate of recycling. The CCS route is also an alternative, but since CCS cannot capture 100% of emissions, at least some of the fossil fuels supplied as PCI¹ have to be replaced with biomass (Mandova et al., 2019). The incomplete capture rate and the lack of industrial scale CCS-steel projects, however, indicates that CCS may have a limited role in steel decarbonization (Agora Industry et al., 2021). There are other technologies that can contribute to the decarbonization of steel, such as Molten Oxide Electrolysis, HIsarna with CCS, or Siderwin, but none are yet at a sufficiently high level of technological readiness (International Energy Agency, 2023).

3.2. Geographical aspects of a steel transition towards renewables

The current decarbonization trajectory, shifting from coal-based steel to steel based on renewable electricity (H-DR (hydrogen direct reduction) + EAFs), will also change the comparative advantages of steel manufacturing countries. This is specific to H-DR, as the cost of electricity for green hydrogen reaches 25–33% of the total costs at electricity prices between €40/MWh and €60/MWh (Vogl et al., 2018). Such changing geographic comparative advantages will have an effect internally on locations within regions (Schneider, 2022) but could, foremost, have effects on the organization of the global value chain. Samadi et al. (2023) call this phenomenon ‘renewable pull’ and note that the two most susceptible sectors are steel and petrochemical feedstock production, see also Devlin et al. (2023) and Gielen et al. (2020). The renewable pull effect will increase the pressure on older steel regions in e.g. the EU or the US who have less advantageous renewable electricity potentials compared to the MENA region, South Africa, Chile, or Australia, to reduce capacities and leave room for emerging steel producing countries (Trollip et al., 2022). Where these effects have been modelled, the assumption is either that existing regions decarbonize their own existing assets (Bataille et al., 2021; International Energy Agency, 2019; Yu et al., 2021) or that the advantages of relocation of iron and/or steel is fully exploited (Bataille et al., 2021; International Energy Agency, 2019; Yu et al., 2021). Apart from renewables, scrap is another resource where availability exacerbates new geographic tensions for a decarbonized steel sector (Agora Industry & Wuppertal Institute, 2023; Åhman et al., 2023; Su & Assous, 2022; Watari, Giurco, et al., 2023; Xylia et al., 2017). Developed countries are restricting scrap exports, undermining this low-cost, low-emission steelmaking route in developing economies (Li et al., 2022; Tarasenko, 2022). Regardless, changing comparative advantages and a geographical relocation push increase the frictions in this transition as they relate directly to the tensions between trade, climate policy and domestic industrial policy (Åhman et al., 2017).

The three elements of steel decarbonization described above are interrelated, as demand and conditions for production improve or decline in various locations. In aggregate, stagnating global demand combined with a higher share of secondary steelmaking will reduce the need for primary steel. Furthermore, the emergence of new green steelmaking countries will require an even quicker reduction of primary steelmaking in existing steel countries. Thus, primary steelmaking capacity will need to decline rapidly and only be partially replaced by green steel investments at existing sites.

4. Inertia and barriers to exit in the steel sector

There is significant inertia in the steel sector, locking in the continued use of fossil fuels. This section describes three specific phenomena, high up-front investments, lumpiness, and societal embeddedness within the steel sector that cause this inertia – also known as barriers to exit – thereby inhibiting a smooth and efficient phase-out of old technologies.

One key source of inertia in the steel sector originates in steel plants requiring high up-front investments that have to be recouped over long time horizons (Rimini et al., 2020; Wesseling et al., 2017). Investments in new plants are made with the intent to use the plant for several decades, and infrastructure like ports and railroads transporting iron ore, coal and gas, and a workforce that is skilled and large enough to man a sizable steel plant is built up over several years. Thus, the cost of changing production systems in the steel sector is high, as it takes a long time to renew the physical capital stock (Deily, 1988). Altering input factors and supply chains is costly and difficult. Due to these factors, innovation is usually more targeted at incremental improvements over radical innovation, and brownfield investments are usually favoured over greenfield (Wesseling et al., 2017).

High up-front investments may induce firms facing diminishing demand to lower prices for steel or capacity utilization, increasing the time it takes to recoup the investment since this is a better option than to close operations and recoup nothing. As Rimini et al. point out, sunk costs also create significant risks associated with entering the sector. This hinders competition, as more efficient firms avoid entering the sector at all, which ‘discourage(s) exit indirectly by shielding unproductive incumbents from competition, enabling them to remain in the market’ (Rimini et al., 2020). Therefore, the steel sector is highly path-dependent and slow to adjust to lower demand or adopt new technology. It is thus subject to carbon lock-in (see Unruh, 2000).

As new primary steel plants are large and rare, capacity is usually increased or decreased in large increments in contrast to small or modular assets, as in for example the wind or solar sectors. The steel sector is energy-intensive, with increasing returns to scale through more effective utilization of energy and infrastructure with the size of a plant. Over time, furnaces used for primary steelmaking have grown in size to increase efficiency and lower production costs (Nielsen, 2017; Gold, 1974). New green iron plants also see significant benefits of increasing size (Nykvist & Olsson, 2020). Smoothly phasing in and out new capacity is difficult, as a new plant requires large infrastructure investments, workforce, and large orders of inputs. Thresholds are therefore high for firms to commit to decarbonization.

Societal embeddedness – here defined as strong links with local communities and governments (Grubert & Hastings-Simon, 2022; Turnheim, 2022; Unruh, 2000) – is a third and related phenomenon that characterizes the steel sector (Rimini et al., 2020). A plant that is unprofitable should, in a well-functioning market, cease operations, and its resources should be used elsewhere for more productive purposes. However, governments have historically supported domestic production of steel and capacity expansions for the purpose of economic development, job creation, and security of supply to drive growth. This blocks the exit of steel capacity, as it is supported by governments around the world, no matter its profitability (Rimini et al., 2020). Since steel plants are large production units, they are usually significant sources of employment and income in the communities where they are based. Due to their reliance on the plant, local communities, unions, and workers may also hinder plant closures, to the extent possible, and try to apply political pressure to keep unprofitable steel plants alive (Moore, 1996). There may also be various legal requirements on firms to pay for environmental clean-up after a plant has closed and related infrastructure is no longer used. Such clean-up costs also limit firms’ willingness to cease production, and instead incentivise them to keep production going, despite low profitability (Rimini et al., 2020).

While ‘barriers to exit’ is a concept belonging to the steel industry literature, it provides two types of lock-in mechanisms relevant for transitions. The first is economic lock-in mechanisms arising from the capital intensity and specificity of physical assets in the steel industry (a blast furnace cannot be dismantled and sold to another industry producing other goods, thus limiting its salvage value (Deily, 1988)); the second is political blockages to closure of plants. These are sector-specific sources of inertia and path-dependency in the steel industry. Due to these phenomena, the process of ‘creative destruction’ will not work within reasonable timeframes for decarbonizing the steel industry in line with the Paris Agreement. Rather, deliberately lowering barriers to exit and closing emission-intensive plants would have three benign effects on the steel transition. First, by incentivising innovation and development of low-carbon steelmaking though ‘exnovation’ (Arne Heyen et al., 2017; David, 2017). Second, by supporting market demand for low-carbon steel and thus supporting the long-term profitability of low-carbon steel beyond early demonstration projects (Rogge & Johnstone, 2017; Rogge & Reichardt, 2016). And third, by minimizing social and political harm and disturbance by providing clear timelines and enabling a transition away from emission-intensive production (Grubert & Hastings-Simon, 2022; Trencher et al., 2022).

5. Projects and policies

In this section, we first review projects and their effect on capacity and the degree of policy support. We then review policies in the top four steelmaking jurisdictions. Finally, we discuss potential designs and examples of phase-out policies.

5.1. Steel decarbonization projects

At the time of this analysis (September 2023), the Green Steel Tracker included 81 projects in 24 countries (including 10 countries in the European Union), announced between 2015 and 2023. Projects included in the tracker are R&D partnerships, pilot plants, demonstration plants, and full-scale plants. R&D partnerships are excluded from our analysis as these are not assumed to have a direct effect on capacity. Each project announcement was reviewed for its relationship to existing capacity. Projects where it was not possible to distinguish an effect on steelmaking capacity (17 projects) – such as hydrogen production for steelmaking or carbon storage – were excluded from the analysis. Another three projects, where it was not possible to identify whether they replace, adapt, or construct a new plant, were also excluded from the compilation. In total, 36 green steel projects are thus included.

The results are shown in Figure 2. The largest share of green steel projects (52%) replaces existing capacity, and a small share (6%) adapt existing capacity. However, as much as 42% of green steel projects are new plants that will add to existing primary capacity, even though demand is set to stagnate, and secondary steelmaking is due to increase.

Figure 2. Green steel projects’ contribution to net capacity. Authors’ calculation.

There is a low share of plant adaptations. This is likely because the physical capital associated with conventional steel is difficult to fully decarbonize, and that interest in Carbon Capture and Storage (CCS) – which could be used in capturing emissions from existing assets – is weak compared to other technologies such as hydrogen steelmaking. Thus, while the majority of commercial-scale green steel projects do replace or adapt existing steel plants, green steel projects are in aggregate contributing to total steelmaking capacity. For example, while the Hybrit project in Sweden is associated with the planned closure of blast furnace capacity, the H2 Green Steel project in Northern Sweden will add another 5 mt of capacity to Sweden’s current total of 6 mtpa.

Of the 36 projects included in the analysis, financial backing from the government could be identified for 16 of them. In 20 cases, government financing is possible, but could not be established. Identified policy support includes direct capex subsidies, loans from public development banks, and green credit guarantees.

Since almost half of announced green steel projects have financial backing from the government, two conclusions can be made. First, it is clear that such financial support is important in the development of green steel projects. Without such backing, the development of green steel would have been slower. Secondly, government subsidies for steel are seeing a return with decarbonization policy. Even though much diplomatic effort has been made to create international frameworks that restrict subsidies and state aid as they risk distorting the steel market, they are seeing a renaissance driven by the urgency to decarbonize.

5.2. Steel decarbonization policy

The last few years have seen rapid growth in policies for decarbonizing steel. The following key trends were identified in the respective jurisdictions: (1) an expansion of subsidies for steel decarbonization and development on green trade barriers in the US and the EU; (2) improved but inadequate carbon pricing in the EU associated with the phase-out of free emission allowances; (3) failed capacity management in China; and (4) a large, planned steel expansion in India with heavy reliance on coal. Below, we describe and discuss each in turn.

The EU Green Deal has led to significant changes in how the EU regulates funding and subsidies for sectors such as steel. The EU changed the mandate of the EIB to turn it into ‘the EU climate bank’ in line with the Green Deal of 2019 (European Investment Bank, 2019). Since then, the EIB has decided to lend €750 million to H2 Green Steel in Northern Sweden, building a 2.5 mtpa-capacity H-DR-EAF integrated plant (European Investment Bank, 2021). The EU also makes exemptions from its own state aid rules for climate, environmental protection, and energy (European Commission, 2021b), which has enabled governments to provide state aid for green steel projects such as ArcelorMittal’s projects in Hamburg, Germany (Reuters, 2021) and Gijon, Spain (NNR, 2021). The EU Innovation Fund has also supported both the HYBRIT (HYBRIT, 2022) and the H2 Green Steel projects (European Commission, 2023) in Northern Sweden.

In late 2022, the US Congress passed a number of spending bills such as the Inflation Reduction Act (IRA) and the CHIPS and Science Act, which enables significant amounts of subsidies for decarbonization projects, including for steel. Included in the CHIPS and Science Act is the Steel Upgrading Partnerships and Emission Reduction provision (SUPER), which is a programme for research, development, demonstration, and commercial applications for low-emission steel manufacturing (Bennet, 2022). A second relevant programme under the IRA is the Industrial Demonstrations Program, providing $6.3 bn for decarbonization of ‘hard to abate’ sectors including iron and steel (Department of Energy, 2022). The IRA also provides subsidies of up to $3 per kg for green hydrogen, which will significantly reduce the operating costs of hydrogen-based steelmaking in the US.

The EU Carbon Border Adjustment Mechanism (CBAM) will set carbon tariffs on the import of several goods such as steel, based on their emissions (Åhman et al., 2022; Grubb et al., 2022; Hermwille et al., 2022). The EU is in talks with the US on what has been called a Global Agreement on Sustainable Steel and Aluminium (GASSA) (European Commission, 2021a), which could impose penalties on imports of emission-intensive steel and protect domestic industries from global overcapacity. The agreement is currently only under discussion and no details are finalized.

The introduction of the CBAM in the EU would be combined with a phase-out of free emission allowances for emission-intensive industries, such as steel, to price embodied carbon associated with steel purchased in the EU, no matter whether it is imported or produced in the union. However, the phase-out will be gradual until 2034 (European Parliament, 2023) and in the initial phase, the fine for missed reporting is believed to be low (Hancock, 2023).

China is struggling with overcapacity and structurally falling demand in the long term and is therefore mainly focused on capacity management, limiting the growth of steel capacity, and ensuring that new capacity additions are met with a corresponding capacity reduction. China is thus the only country in our study that has specific limitations to steelmaking capacity, and that aims to close plants. In 2016, China announced plans to phase-out 100–150 mt of capacity until 2020, and in 2019 there was a ban on net capacity additions. In the Guidance on Promoting High-Quality Development of the Iron and Steel Industry, the government states a plan to phase-out obsolete capacity (Li et al., 2022).

The country’s limit on capacity and introduction of ‘capacity swaps’ is significant. The capacity swap mandates that there is a reduction in capacity relative to any capacity addition. The latest version of the capacity swap mandate requires a reduction ratio of 1.5:1 in environmentally sensitive regions, and 1.25:1 in the rest (OECD, 2021). However, the OECD states that firms and provinces are using loopholes in the regulations to avoid the mandatory reductions, rendering the capacity swap system ineffective. The government is also softening the capacity swap ratio to 1:1 for environmentally friendly steelmaking facilities, such as EAFs or hydrogen-based steelmaking. The Chinese steel policy focus until 2030 is on reducing overcapacity and shifting towards more secondary steelmaking via EAFs, as more scrap is becoming available domestically or imported. GHG-emissions from steel production should peak by 2030, and policies for achieving this are in place (Li, Andersson, Nilsson, & Åhman, 2023). However, China still has almost 150 mtpa of BF capacity under development (Swalec & Grigsby-Schulte, 2023). Targets and policies up to 2060, when China is to be carbon neutral, are still vague with no specific support schemes or market interventions announced. However, in the Chinese political economy, it is the state-owned enterprises (SOEs) (HBIS, Angang and others) that are expected to lead the way and help the government in achieving its goals (Moretz, 2018). Several of these big steel companies are exploring options for deep decarbonization, including H-DR, but also BF/BOF with CCUS. The technology menu is still open, with a focus on CCUS, but attention to H-DR has recently increased.

India is the second world producer, with a current production capacity of 118 mt and, as a fast-developing country, has ambitions and targets to expand drastically over the years to over 300 mt by 2030 and 500 mt by 2050 (Hall et al., 2020). This growth builds on an anticipated increase in demand as the country is developing, and as part of industrial policy to supply this steel mainly by domestic production. The Indian steel industry is complex, with some of the largest global actors (ArcelorMittal, JSW, etc.) but also many minor local producers that use coal-gasification in smaller DRI units (Hall et al., 2020). For reaching long-term decarbonization targets, the IEA (2019) roadmap for India suggested that the country turns to BF-CCS, followed by an expansion of H-DR later, whereas Hall et al. (2020) focus more directly on moving towards H-DR. India has recently acknowledged its renewable potential and launched a ‘National Green Hydrogen Mission’ (Government of India, 2023) to increase green hydrogen production to at least 5 million metric tonnes (MMT) by 2030. The development of CCS in India has also been slow due to inadequate research on technologies and sites (Gupta & Paul, 2019). The main challenge for India’s decarbonization is the persistent emission-intensive development in the steel sector, including the 150 mt unmitigated blast furnace capacity currently under development (Swalec & Grigsby-Schulte, 2023).

5.3. Possible phase-out policies

The above review identifies what has been characterized as an ‘innovation bias’ in current steel decarbonization policy mixes (Arne Heyen et al., 2017) following from a lack of deliberate termination of unsustainable practices. However, existing literature and historical examples illustrate how such policies could be designed. Rogge and Reichardt (2016) define policy instruments as ‘the concrete tools to achieve overarching objectives’ and group them into economic instruments, regulations, and information. Economic instruments create the economic incentives for change, while regulations are laws allowing or banning certain practices, and information relates to training, rating, and labelling programmes and public information and cooperative research programmes (David, 2017; Rogge & Reichardt, 2016).

In the economic instruments category, we have the EU ETS, which prices emissions and creates economic disincentives for fossil-based steel capacity. However, jurisdictions that choose to avoid a carbon-pricing approach may opt for more direct economic instruments. Examples can be found in the steel sectors’ history as well as other sectors. In the 1990s, overcapacity was a major challenge for the sector, and based on the 1994 association agreement between the EU and Romania, Romania is allowed to exceptionally grant state aid to its steel industry, provided that it is part of a rationalization and reduction of steel capacity (Europe Agreement between the EEC and Romania, 1994). Problems remained, and in 2002 the secretary-general of the OECD sent a letter to the World Bank encouraging it to ‘play a supportive role in addressing the issue of inefficient excess capacity and the social cost of closing facilities’ (Johnston, 2002). Similarly, state aid and international financial organizations can be used to cover the financial and social costs of closures of emission-intensive plants on a systematic, international scale. The EU-Egypt partnership, where the EU will help finance the shut-down of 5 GW gas-based power generation capacity and replace it with 10 GW renewable capacity, and the Just Energy Transition Partnership (JETP) with Indonesia, are examples of a bilateral approach to the same effect (European Commission, 2022; JETP Secretariat, 2023).

An example of a regulation could be simple bans on coal-based production or on the emissions-reducing the air quality near major population centres, requiring the closure of plants. However, such measures have been found to be poorly enforced in the steel industry when there is a risk of closure (Deily & Gray, 1991). Rather, such measures may, similarly to the automotive industry, develop when aligned with industrial policy goals of diversification and upgrading (Meckling & Nahm, 2019). Finally, information instruments could be labour force development and research related to economic diversification (Greco, 2022), similar to measures taken for the oil and gas industry (Krawchenko & Gordon, 2022). However, attempts to diversify risk yielding temporary results if not combined with a consistent policy to phase-out the fossil-based activity (Mäkitie et al., 2019).

6. Discussion

While previous studies show that a rapid decarbonization of the steel sector is technically possible, inertia and barriers to exit inhibit the closure of emission-intensive steel, yielding frictions such as overcapacity and trade conflicts that may slow the transition. The literature on industrial transitions suggests that phase-out policies are important in destabilizing existing techno-economic regimes, setting a clear directionality for sectoral decarbonization, and enabling demand for new low-emission technologies. Such policies are important but lacking in the steel sector. As argued by Vogl et al. (2021), retiring conventional plants does not have to wait until low-emission steel has become fully commercialized. Rather, retiring conventional plants may be necessary to enable diffusion of low-emission steelmaking. Thus, such policies may be particularly important for the global steel sector, as low-carbon technologies are emerging, but structural overcapacity and trade conflicts risk locking-in steel sector emissions, undermining long-term demand for low-emission steel, and impeding market access for low-emission steel through trade barriers.

We find that the ongoing government-supported green capacity expansion is unlikely to crowd out emission-intensive capacities, due to inertia and barriers to exit. This snapshot suggests a lack of support for phase-outs of conventional plants, although it is possible that such measures might be announced after the time of analysis (September 2023). The lack of phase-out policies risks exacerbating frictions, impeding the global steel transition. As global steel demand stagnates, driven by a fall in demand in China and material efficiency strategies, there is a considerable risk that steel overcapacity and related trade frictions are exacerbated if emission-intensive plants are not phased out at a considerably higher pace than is currently the case. In addition, the global pool of scrap will increase as the Chinese steel stock reaches end-of-life. This will put further downward pressure on demand for primary steel, enhancing the need for phase-outs of emission-intensive primary steelmaking.

The renewables-pull effect provides an opportunity for low-income countries to develop their own steel industries but is also likely to drive frictions between ‘old’ and ‘new’ steelmakers, where old steel manufacturing countries with established value chains will want to retain their industries, while new entrants will want to leverage access to renewable energy to expand their production. Such a dynamic risks becoming another source of frictions that may impede the transition through trade barriers or government support for steelmaking in what will become uncompetitive areas in a low-emission global steel sector.

The phase-out of emission-intensive steel is likely to be politically fraught, but no less important. Fairness and climate justice will have to be central to future discussions of where, how, and when emission-intensive assets are to be phased out. While we have not focused on such elements in this article, there are obvious challenges relating to the principle of common-but-differentiated responsibilities for developed and developing economies. Developed economies have a larger pool of accumulated scrap than developing economies, which will need more steel per capita to build the necessary infrastructure. Large numbers of emission-intensive facilities have recently been built, are under construction, or have been announced in developing economies, and will have to close early to realize net-zero emissions by mid-century. Most green steel projects are in wealthy countries, while demand will expand in poorer countries. Addressing such challenges in a political process will be difficult albeit important to ensure that the entire global steel sector is decarbonizing on necessary timelines to realize the climate ambitions signed up to by the global community under the Paris Agreement.

Historically, the combination of stagnating demand and new entrants to the global steel sector has driven frictions that have been resolved in a political process of restructuring, negotiations, and coordination (see for example D'Costa, 1994; Moore, 1996, 1998). The analysis in this article suggests that the global steel transition will similarly require a political process that supports a parallel phase-in of low-emission steelmaking and phase-out of emission-intensive steelmaking. International forums such as the OECD Steel Committee and the G7 Climate Club could facilitate such coordination.

7. Conclusion

While the growing literature on steel decarbonization using model-based projections shows that the steel sector could become a ‘fast-to-abate’ sector (Agora Industry & Wuppertal Institute, 2023), we argue that such a rapid transition can be impeded by frictions that inhibit the long-term steel transition.

Identified frictions are overcapacity and trade conflicts which undermine long-term demand for low-emission steelmaking, limit market access for low-emission steel, and slow the exit rate of emission-intensive plants. Inertia and barriers to exit in the steel sector will drive overcapacity and trade conflicts by blocking the closure of plants. Therefore, phase-out policies may have a particularly important role in the decarbonization of the steel sector in ensuring both emission reductions and market space for low-emission steelmaking.

In this paper we have analysed green steel projects, finding that 42% of projects are new plants, and establishing that 16 out of 36 have financial support from government. Decarbonization is therefore sparking a renaissance for subsidies in the steel sector, risking increasing overcapacity and trade conflicts unless combined with corresponding phase-out policies. Based on our analysis of steel decarbonization policy mixes in the top four steelmaking jurisdictions, there is a lack of policies that lower barriers to exit for emission-intensive plants. Rather, policy mixes in the EU and US are oriented towards the phase-in of low-emission steelmaking, and in India around increasing capacity. In China, there is an attempt to limit steelmaking capacity growth, although it has been unsuccessful in stopping capacity additions, let alone reducing capacity in line with the expected fall in domestic demand.

A variety of potential phase-out policies are conceivable; historical examples from the steel industry include state aid for capacity reductions and international finance for closures. Examples from other sectors include policies for economic diversity in relevant regions, bans or bilateral agreements on financing closures, and economic diversification. The introduction of such phase-out policies would enable a faster expansion of low-carbon steelmaking.

Meanwhile, new comparative advantages arising from a transition to renewable energy and scrap as key inputs in steelmaking provide opportunities for new regions and emerging markets to develop steelmaking. However, without coordination for leveraging such potentials, new comparative advantages risk exacerbating trade conflicts between ‘old’ steelmakers wanting to retain their market shares and new entrants, which may slow the steel transition by increasing government support and protection of emission-intensive steel production.

Considering the inertia and barriers to the exit of the steel industry, the many different pathways to decarbonization, and the heterogeneity of countries’ respective steel sectors, it is likely that the steel sector will be neither hard-to-abate nor fast-to-abate, but fractious-to-abate. A focus on – and coordination of – phase-outs of emission-intensive assets and phase-in of low-emission steelmaking may help ease such frictions and help deliver a fast and deep steel transition.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the European Union’s NextGenerationEU, via the Swedish Energy Authority under The Leap for Industry.

Notes

1 PCI = Pressurized Coal Injection; done in the lower tiers of the blast furnace.