Made to measure: tailoring steel standards to policy goals

Traditional steelmaking remains heavily dependent on carbon-intensive inputs, including iron ore, coal and natural gas. Achieving climate neutrality in the sector therefore requires a rapid shift towards production routes that reduce fossil-fuel use, make greater use of renewable energy and integrate circularity throughout the steel value chain.

Renewable electricity will be central to this transformation. Hydrogen-based direct reduction of iron (DRI) in particular requires large volumes of electricity to produce renewable hydrogen, making access to abundant and competitively priced clean power a decisive factor in determining where new near-zero iron- and steelmaking capacity is likely to emerge. Renewable electricity costs and availability differ widely across regions (see Exhibits 1-6), creating uneven starting conditions for steel producers and project developers.

Regions with strong wind and solar resources have a structural advantage (see Exhibit 2). High-quality renewable resources allow more intensive use of generation assets and can lower hydrogen production costs. Yet resource availability alone does not determine competitiveness. Constructing renewable generation capacity, electrolysers and associated industrial infrastructure requires significant upfront investment. Macroeconomic stability and financial conditions therefore play an important role by influencing the cost of capital and, ultimately, project viability (see Exhibit 3).

Access to firm, low-carbon electricity is another important factor. Although electrolyser technologies and system-integration approaches are evolving, hydrogen production and large-scale industrial facilities may still benefit from a more stable electricity supply. Locations with access to low-carbon and well-developed electricity grids could therefore gain an advantage by attracting the first wave of near-zero steelmaking projects (see Exhibits 1-6). Where grid infrastructure is limited, developers may need additional flexibility and storage, including batteries, hydrogen storage or other balancing solutions; this can raise system costs, particularly during extended periods of low renewable output.

Location and system design are therefore closely linked. Developers must assess renewable resource quality, grid access, infrastructure, financing conditions, logistics, proximity to downstream industries and value chains, and process requirements as part of a single investment strategy. Electricity procurement (see Exhibit 5) must deliver sufficient volumes at the lowest feasible cost while meeting the required carbon-intensity thresholds and complying with the applicable production standards and regulatory requirements.

No single technology or resource endowment will deliver steel decarbonisation. The transition will require renewable generation, electricity grids, hydrogen infrastructure, storage and industrial assets to be developed in a coordinated manner. Countries and regions able to combine these elements with supportive policy frameworks, risk-sharing instruments and access to finance will be best positioned to attract early investment and build competitive steel value chains with near-zero emissions (see Exhibit 6).

These factors provide the basis on which we assess how standard design choices affect the uptake and cost of cleaner technologies across countries.

The key findings on emission thresholds are that when emissions are calculated within our widest scope (including steel production, energy inputs, and upstream value chain processes): 

• BF-BOF cannot meet any of the four stringency levels (Levels 1–4).
• NG-DRI-EAF can reach Level 2 at best. 
• Level 3 and Level 4 (near-zero steel) can be achieved only through H₂-DRI-EAF, alternative clean primary steel technologies, or scrap-based EAF.

The sliding scale therefore differentiates technologies effectively according to their emission performance. More importantly, Level 3 represents the minimum threshold needed to incentivise near-zero technologies when emissions are assessed across the full value chain. This leads us to our next question about scope coverage.

Narrowing the scope changes the picture (see Exhibit 11). In the scenario that is referred to here as narrower coverage, namely Scope 1–3 light:

• BF-BOF can meet Level 1 at most, as emissions from upstream material inputs are not included here. 
• NG-DRI-EAF can reach Level 3 under this narrower scope. 
• Only hydrogen-based steelmaking and scrap-based EAF can meet Level 4. 

This means that, under a narrower emission scope, Level 4 is needed to maintain a strong incentive for near-zero-emission technologies.

Broad value chain coverage should remain the long-term objective, but it may need to be introduced progressively. In the short term, where low-carbon inputs or supplier-level data are unavailable, applying default emission values to upstream inputs could prevent even hydrogen-based steelmaking from meeting near-zero thresholds. As data availability and value-chain coordination improve, policies can progressively incorporate emissions from these upstream inputs. 

Our analysis also shows that production costs are dictated to a greater extent by increasing stringency than by expanding scope. Moving between levels (1-4) requires technology shifts, whereas expanding coverage from Scope 1 to Scope 3 has relatively little impact on production costs. This conclusion derives mainly from our analysis of electricity costs. The analysis finds that the least-cost electricity supply to an H2-DRI-EAF plant is typically one that is renewable-dominated and very low-emission. Consequently, for most locations, bringing electricity emissions within scope (moving from Scope 1 to Scope 2) does not necessitate a change in the electricity mix required for a plant to meet the Level 4 emission threshold. The reality could be different if non-modelled factors were to increase the local costs of clean electricity. Our findings also reflect assumptions about the relatively low costs of eliminating emissions from material inputs such as pellets and lime.  

Costs nevertheless vary across countries (Exhibits 11 and 12), with electricity costs, grid carbon intensity and renewable availability shaping the economics of early investment. The optimal configuration generally combines dedicated renewables with around 20–30 percent grid electricity, avoiding the higher energy storage costs of relying entirely on dedicated renewables. Brazil has the lowest-cost renewable-powered steel production in our analysis, while Sweden benefits from relatively cheap electricity (and has expensive natural gas). By contrast, South Africa’s carbon-intensive grid limits its advantage despite the country’s strong renewable potential. These cost dynamics are likely to evolve as carbon prices rise and clean technologies scale up.

In any case, near-zero-emission steel costs more under full value chain coverage (Scope 1–3) because upstream materials also need to be decarbonised. Where low-carbon inputs are already available – such as DRI pellets from Sweden and Canada or alloys produced with low-carbon electricity –, actual upstream emissions can be lower than default values. 

The key implications for policy are as follows:

• Stringency and scope are complementary levers: higher stringency is critical for driving the deployment of near-zero technologies, while broader value chain coverage increases incentives to reduce emissions across the wider value chain. Where narrower scope coverage is used in the short term, Level 4 stringency is needed to maintain a strong incentive for near-zero-emission technologies.
• A sequencing approach may be appropriate: apply high stringency (Level 4) initially to drive investment in near-zero technologies, while progressively broadening value chain coverage as supplier-level data and coordination improve. Broad value chain coverage should remain the long-term objective, its implementation being expanded as the necessary data and coordination become available and this being done in alignment with countries’ economy-wide net-zero emission goals.

In the near term, the potential for high scrap shares of input to narrow the cost gap between H₂-DRI and fossil-fuel-based routes entails an important advantage. For early movers in Germany, for example increasing scrap use from 20 to 70 percent can reduce the carbon abatement cost of H₂-DRI-EAF relative to BF-BOF by more than half, from around 190 to 70 euros (EUR) per tCO₂e (see Exhibit 13). This is not a trivial finding because steelmakers have only relatively recently gained confidence in the potential to make high-grade steel using high scrap shares of input. This implies that a carbon price of EUR 70/tCO₂e, which may be considered foreseeable within current European Union policy frameworks, could offset the higher cost of H2-DRI as the ore-based metallic production route. 

In Brazil, increasing the scrap share from 20 to 70 percent reduces H₂-DRI’s abatement cost from around EUR 110 to EUR 40/tCO₂e (see Exhibit 14), while in South Korea it falls from around EUR 300 to EUR 110/tCO₂e (see Exhibit 15). In scrap-rich regions with quality controls already in place and highly mature collection and sorting mechanisms, this could provide commercial viability at a much faster pace than H₂-DRI with a lower scrap share. 

However, increasing the volume of near-zero-emission ore-based steel production is a critical challenge for the transition because scrap-based steel couldn’t fully meet global steel demand even if it were able to meet all quality requirements. Agora Industry’s ‘Achieving climate neutral steel by 2050’ report modelled a scenario where scrap is able to make up around 50 percent of global production by mid-century. 

Without explicit rules, scrap-based EAF could outcompete green H₂-DRI, potentially discouraging investment in low-carbon primary production. The key implications of our scrap-content analysis are as follows:

• H2-DRI steel plants may use high shares of scrap inputs to narrow the cost gap to conventional steel. 
• at the same time, limits on scrap availability mean that an important policy objective is to ensure the commercial viability of near-zero iron production.

This is why, when policies refer to or make use of standards, the relevant design choices may need to be tailored to the specific policy objectives. The next section illustrates this by showing how the findings can inform the development of a scrap-sensitive green steel premium calculator. 

The analysis has shown how different choices on thresholds, scrap share and scope coverage translate into different costs depending on the base assumptions in the eleven different countries of focus in this study. Our interpretation of these results leads us to conclude that the design choices do not need to be one-off, finalised decisions. Instead, different policies can operate on different timelines. For instance, a public procurement policy could tighten over time, as more near-zero emission steel becomes available. A single, static design choice would find it hard to serve all of these purposes equally well.

This is where tailoring of standards becomes essential. i.e., how standard-related choices are applied differently across policy contexts. For example, subsidies intended to support the deployment of near-zero-emission steel technologies would need to operate on the basis of a different emission threshold than a regulation intended to prevent investment in the highest-emitting production processes. 

The tailoring of standard-related choices to support specific policy objectives should not be done in a way that undermines governments’ strategic objectives. In particular, the calibration of stringent thresholds, scope, and scrap share to ensure the effectiveness of each policy instrument should be consistent with governments’ long-term goal of a carbon-neutral steel sector.

Take the example of clean technology deployment policies. Policies in this category are those that support the deployment of clean technologies for steel production. They include

• subsidies for green iron or clean steel production, 
• clean steel production mandates, 
• clean steel tariff reductions or exemptions.

These policies are designed to enable technologies to be deployed that are consistent with the end goal of a near-zero-emission steel sector. This can include the first deployment of such technologies within a given market and the increasing diffusion of those technologies. Early deployment is important in the process of technology evolution: as investment increases, technologies are improved, and costs can be reduced. These policies may also aim to support the development and diffusion of other elements of the clean steel system, such as supply chains and infrastructure. Standard-related choices for these policies involve decisions of scope, stringency, and scrap share. 

Policies in this category will create the strongest incentives for deploying clean steel technologies if only these near-zero-emission technologies are eligible for support; this means also excluding any technologies that are inconsistent with the near-zero-emission end-goal of the transition. The analysis shows that this can be achieved by using either Level 4 stringency (with any scope), or Level 3 stringency with Scope 1, 2 and 3. Early in the transition, before supply chain emission data becomes widely available and practices for decarbonising steps such as pelletisation and lime production are not well established, achieving Level 3 with Scope 3+ may involve an inappropriate level of difficulty, meaning it may be more practical to use Level 4 stringency with Scope 1 or 2.

In addition, targeted support policies may need to be carefully designed to take the use of scrap steel into account. Generally, the analysis suggests that increasing the share of scrap input reduces the cost of meeting Level 2, Level 3 or Level 4 stringency on the sliding scale, regardless of which scope is chosen, all the way to 100 percent scrap input. Our technoeconomic results clearly show that ore-based metals will be more expensive to decarbonise than higher shares of scrap. This needs to be responded to with policies that reflect the investments needed for building out these near-zero capacities. Keeping ore-based and scrap-based steel routes in the same standard is important as EAF producers are increasingly entering high-grade steel markets – meaning the route of production is no longer such an important differentiator. 

In the same way that increasing the scrap share of input reduces the carbon price required for H2-DRI-EAF steel to reach cost parity with BF-BOF steel, it also reduces the subsidy required to achieve the same effect. This is illustrated by the Green Steel Premium Calculator’ (see below, Exhibit 16). Subsidy policies will need to be designed to reflect this difference so that clean steel produced with any share of scrap input receives the appropriate level of support.  

In contrast, a public procurement policy creating a lead market for near-zero-emission steel would need to balance the stringency necessary to drive investments in near-zero steel facilities with the availability of steel produced with fully low-carbon inputs such as green hydrogen rather than fossil gas. This could mean either using a stringent threshold (Level 4) with Scope 1- 3 light initially applied to only a small share of public procurement, with this share expanding over time, or using a less stringent threshold (Level 3) and then tightening it once it is clear that enough green steel will be produced for procurement officers to achieve their institutional goals. Another Agora publication has argued that enough near-zero steel will be available by 2035 to stipulate that the EU automotive industry must procure only near-zero steel from 2035 (Agora Industry, 2026).

The calculator includes the input costs for near-zero steel in the eleven different countries. From a techno-economic perspective, it is clear that the higher the steel product’s ore-based component is, the higher the cost difference between the coal-based and the near-zero-based production route will be. The costs decrease with the amount of scrap added, meaning that near-zero steel with a high scrap share has a lower cost difference than with a low scrap share. 

The hope is that this calculator can inform policymakers about the cost gap that needs to be bridged through policies. These policies would de-risk investments to the extent needed for companies to be confident that any investments they make will receive policy support.