Beyond the GWP Number: Rethinking how we identify ‘low emissions’ steel

With every heat wave, the need for ‘low emissions’ steel intensifies. And yet awareness of the complexity of measuring and comparing emissions and GHG performance across the steel industry is also growing. The growing use of Environmental Product Declarations (EPDs), especially in the construction sector, signals a welcome shift towards greater transparency around emissions and other environmental impacts. But they also come with risks when used in isolation.

In the recently published report “From Data to Decisions: Comparable GHG Emissions Data for Steel Procurement”, ResponsibleSteel explores to what extent product-level Global Warming Potential (GWP) values reported in EPDs can be reliably linked to the underlying GHG emissions performance of the steelmaking site. Supported by a data collection and conversion process with four steelmaker members, the research suggests the answer is: not yet.

The research examined how GWP values for steel products are calculated in EPDs and whether these values, together with information on Secondary Materials, could be used to determine a steelmaking site's Decarbonisation Progress Level (DPL). This analysis is framed against the Steel Standards Principles' (SSP) Criteria for Transparency in GHG Reporting which is a benchmark already recognised across the sector to improve transparency, understanding and comparability of GHG metrics.

The findings reveal significant challenges in translating between product-level EPD data and site-level measures of decarbonisation.

GWP values for the same steel product could vary substantially even when calculated using the same Product Category Rule. EPDs also do not consistently disclose the use of Secondary Materials. To bridge these gaps, the research team collected more than 30 additional quantitative and qualitative data points directly from participating steelmakers. Even with this additional information, significant variation remained between cases, making the approach unsuitable for reliable conversion.

Three factors were particularly important: system boundaries, co-product allocation and background datasets. Methodological choices within EPDs can have a greater influence on the reported number than genuine differences between steel producers. Using EPD GWP thresholds alone to define 'low emissions' steel could therefore produce misleading procurement outcomes that can stall progress rather than driving meaningful emissions reductions.

This does not mean EPDs should be abandoned. They remain valuable and well-established tools for communicating environmental information, including declarations beyond the GWP value. Instead, the research points to three practical improvements: greater transparency about what each GWP value represents; tighter application of existing standards to reduce methodological variation; and a consistent approach to reporting scrap, distinguishing clearly between scrap, Secondary Material and recycled content.

Ultimately, the product-level GWP value and site-level decarbonisation metric answer different questions. The most credible approach is to use them together: EPDs can provide product-level environmental information, while complementary metrics such as ResponsibleSteel's DPLs can provide evidence of genuine decarbonisation performance.

The SSP Transparency Criteria reinforce this direction bycalling for disclosure of information such as the reporting boundary, steel production route, primary data share, scrap inputs and key accounting choices. Greater transparency of this kind would make it easier for buyers to understand what a reported GWP actually represents and how it can, and cannot, be compared with other emissions metrics.

The direction of travel is already clear. Initiatives including the Industrial Deep Decarbonisation Initiative (IDDI), Catena-X,CEN's draft steel-specific standard, ISO harmonisation work and the SSP are all pointing towards greater transparency and methodological consistency within EPDs.  

If steel procurement is to genuinely drive decarbonisation, we need to move beyond a single GWP number. EPD programme operators, steelmakers, manufacturers, specifiers, standards bodies, policymakers and civil society should work together to make product-level emissions data more transparent, comparable and fit for purpose.

The goal is not to replace EPDs, but to make the information they provide more meaningful—and to ensure it is used alongside the right complementary metrics to identify genuinely lower-emissions steel.

What does this mean for steel procurement?

Procurement is one of the most powerful levers for change. But to genuinely drive decarbonisation, buyers need the right data, used in the right way.

We put forward the following recommendations to producers and users of the EPD system:

1. Make EPDs more transparent. Policies, labels and programmes using EPD GWP values for steel should recognise only EPD programmes that meet the SSP Criteria for Transparency in GHG Reporting and disclose key information, including crude steel emissions intensity and scrap share.

2. Look beyond GWP alone. ‘Low emissions’ steel specifications should consider both GHG emissions intensity and scrap share, using a defined, scrap-variable framework such as ResponsibleSteel’s DPLs. EPDs should therefore disclose these metrics alongside GWP, enabling specifiers and procurers to assess emissions performance more meaningfully.

3. Harmonise how steel EPDs are calculated. EPD programme operators, Programme Category Rule (PCR) developers and steel-sector stakeholders should work together to reduce methodological inconsistencies in calculating steel product GWP values, improving comparability while building on the EPD systems already in use.

‍

Read the full report where we dive deeper into the methodology, findings and recommendations for applying DPLs and EPDs together in decision making.

Sign up for the upcoming webinar

This study builds on the findings of the recent policy paper, titled “Global Steel Decarbonisation: From complexity to coherence”.