Executive summary

India is the world’s second-largest steel producer, accounting for 7.4 per cent of global output in 2023 (Verma et al. 2024). The steel industry contributes approximately 2 per cent to India’s GDP (IBEF 2025) and supports nearly 2.8 million jobs. Yet, India’s steelmaking remains among the most emissions-intensive globally, emitting 2.54 tonnes of carbon dioxide per tonne of crude steel (tCO₂/tcs) compared to the global average of 1.91 tCO₂/tcs. Among the key factors driving this gap is the predominance of coal-based direct reduced iron (C-DRI)—a steel-making route where India is the world’s largest producer. C-DRI contributes nearly 30 per cent of the country’s crude-steel output and emits between 2.7 and 3.1 tCO₂/tcs (Verma et al. 2024). Crude steel capacity in India increased at a rate of 10 per cent between 2003–04 and 2023–24, reaching 179.5 million tonnes (Mt) (JPC 2023–24). The National Steel Policy (MoS 2017) projects capacity to rise to 300 Mt by 2030–31. Expansion pathways differ; integrated steel plants are adding capacity mainly through the blast furnace route, while the sponge iron sector is expanding via coal-based rotary kilns. According to the Sponge Iron Manufacturers Association (SIMA), an additional 27 Mt of C-DRI capacity is planned to be added by 2030, on top of the current 48.2 Mtpa, potentially locking in high-carbon capacity for decades without targeted intervention.

While integrated steel plants generally possess the financial and technical capacity to pursue decarbonisation, the C-DRI sector is dominated by small, standalone units with limited resources. These units are typically clustered in resource-rich regions such as Chhattisgarh, Odisha, etc. Decarbonising C-DRI can reduce costs for smaller firms, increase competitiveness, and unlock new economic activity. With the steel sector accounting for nearly 12 per cent of India’s total emissions (Verma et al. 2024), achieving the national targets of 45 per cent emissions-intensity reduction by 2030 and net-zero by 2070 will require the targeted transformation of C-DRI clusters.

Despite variations in production mix, input use, and energy sources, C-DRI units within each cluster exhibit similar operational characteristics, enabling the deployment of shared infrastructure and common decarbonisation solutions. To harness this opportunity, we developed an evaluation framework supported by a survey of prominent DRI clusters to assess and compare clusters in terms of cost and readiness for a low-carbon transition. The framework evaluates emissions profiles, infrastructure gaps, and mitigation options. Insights from the framework inform cluster-specific strategies and crosscutting recommendations with phased timelines, offering a roadmap to accelerate low-carbon transition in one of India’s most emissionsintensive industrial segments and align it with national climate and development goals.

Evaluation framework for cluster-level decarbonisation in the steel industry

The evaluation framework assesses India’s C-DRI clusters along two key dimensions— annualised decarbonisation cost and readiness (Figure ES1). It serves as a decision-support tool to compare decarbonisation potential, cost-effectiveness, and implementation feasibility across clusters. Clusters are plotted on a cost–readiness matrix, with the x-axis representing annualised decarbonisation cost (USD/tCO₂ mitigated) and the y-axis representing readiness. The cost dimension captures the expenditure per tonne of CO₂ reduced, while readiness reflects each cluster’s preparedness to decarbonise through levers such as energy efficiency (EE), renewable energy (RE), and ecosystem readiness, which together act as key enablers.

Figure ES1. The evaluation framework considers the decarbonisation potential and readiness of a cluster

Based on these dimensions, clusters are grouped into four categories (Figure ES2). From an implementation perspective, clusters with low annualised decarbonisation costs and high readiness (dark green) are better suited for immediate decarbonisation and new capacity addition, whereas clusters that have high decarbonisation costs and low readiness (red) are not prepared for transition and will need significant efforts and policy support to decarbonise.

Figure ES2. Categories of clusters explained based on annualised decarbonisation cost and readiness levels

Cluster-level surveys, findings, and evaluation results

To operationalise the framework, we surveyed DRI units across three major clusters— Sundargarh and Jharsuguda (Odisha) and Bellary (Karnataka)—covering about 29 per cent of installed capacity and 20 per cent of the number of units in these clusters. Insights from earlier work in Raipur and Raigarh (Chhattisgarh) (Nitturu et al. 2024) were incorporated to complete the analysis across five clusters, which together account for over half of India’s C-DRI production. The surveys captured plant-level data on technology use, fuel mix, and operational barriers. The key findings from the survey are discussed in Chapter 3 of the report.

Based on the survey data, we developed an emissions inventory for the five clusters (detailed in Chapter 4). Our results indicate that C-DRI units primarily drive emissions within steel clusters (ranging between 78–86 per cent). Larger clusters, such as Raipur (7.75 Mt) and Raigarh (10.66 Mt), record higher overall emissions owing to their greater DRI capacity of 3.4 and 3.46 Mt, respectively, as well as higher-capacity induction furnace (IF) (4.01 Mt) and electric arc furnace (EAF) (6.28 Mt) plants. Clusters with significant domestic fuel use, such as Sundargarh with 88 per cent, exhibit disproportionately higher emissions. In the next step, we use the emissions inventory and the penetration levels of various decarbonisation levers across clusters to evaluate pathways for achieving net-zero emissions. We followed the approach outlined in our earlier research (Elango et al. 2023), assuming that these measures can be implemented immediately under current conditions. The financial requirements associated with each decarbonisation lever are represented through a marginal abatement cost (MAC) curve (Figure ES3).

Figure ES3. Marginal abatement cost curves (MACs) for all the decarbonisation levers in the five clusters

Generally, mitigation costs vary across clusters depending on the price of incumbent fuels, energy sources, and decarbonisation measures. Across all clusters, WHR consistently delivers cost savings (–153 to –22 USD/tCO2 ) by displacing more expensive power sources. Scrap use also has a negative abatement cost (–4 to –7 USD/tCO2 ) but contributes only marginally (0.76 Mt across 5 clusters) to overall emission reduction. In contrast, pelletisation, firm and dispatchable renewable energy (FDRE), alternative fuels, and CCUS have positive MAC values. Although pelletisation incurs lower abatement costs, its impact on emission reduction remains marginal. While FDRE can significantly mitigate emissions (2.52 Mt across 5 clusters), the cost of mitigation varies across clusters, depending on local cost of electricity and open access (OA) charges. The use of imported coal (an alternative fuel) is particularly expensive and is relevant mainly to clusters in Odisha that rely heavily on domestic coal. Among the available decarbonisation levers, CCUS dominates both in terms of cost and the share of emissions abated across each cluster.

Figure ES4. Cluster-wise relative readiness scores for all eight readiness levers

The readiness score (Figure ES4) aggregates scores across eight sub-levers—EE, ME, RE, CCUS potential, finance, policy support, alternative fuels, and ecosystem readiness—to produce a composite readiness index for every cluster. The detailed computation methodology and weighting structure are presented in Chapter 5 of the report. Overall, EE readiness is high across clusters, given the widespread presence of ESCOs and basic efficiency measures, though project-level engagement remains limited. ME readiness is strongest in Bellary and Raipur, which have sufficient pellet capacity, while Raigarh and the Odisha clusters lag behind. RE readiness is highest in Chhattisgarh and Karnataka, supported by operational green-OA rules and low tariffs, whereas Odisha’s draft rules constrain progress.

For CCUS, favourable storage geology in Chhattisgarh and Odisha improves readiness. Financing conditions are better in Chhattisgarh and Odisha due to the presence of SIDBI branches, and a larger share of non-standalone units. Ecosystem readiness remains weak across clusters, reflecting limited awareness and low PAT coverage. Alternative fuels readiness is relatively stronger in Odisha, which has a state hydrogen policy, while Chhattisgarh faces high delivered NG prices. Policy support remains uniform nationwide, with central initiatives such as taxonomy in place but no state-level GPP frameworks yet operational.

Integrating cost and readiness as annualised decarbonisation cost and readiness score (methodology described in Chapter 5) yields cluster plots. Figure ES5 presents three cluster plots that illustrate annualised decarbonisation cost (x-axis) and readiness (y-axis) across three scenarios: (i) with all decarbonisation measures, (ii) excluding CCUS, and (iii) excluding both CCUS and alternative fuels. The volume of emissions mitigated is represented by the size of the bubble.

When all decarbonisation levers are applied, all clusters exhibit high costs and—except for Bellary—high readiness. Bellary records a lower annualised decarbonisation cost because its existing WHR capacity fully meets its power demand, and it already uses 100 per cent imported coal. Consequently, there is limited scope for additional RE or alternative fuel interventions, which would increase the cost of decarbonisation. As CCUS dominates across all clusters, both in terms of its mitigation share and its cost. This heavily skews the potential and readiness scores of all clusters. For example, in Bellary, 78 per cent of total mitigation depends on CCUS. However, because Bellary lacks CCS potential in the form of saline aquifers or basalt rocks, its readiness scores are significantly lower. By contrast, all other clusters possess CCS potential in the form of saline aquifers.

Figure ES5. All clusters have higher decarbonisation costs due to carbon capture utilisation and storage (CCUS)-related costs

The second cluster plot excludes CCUS to better isolate lower-cost, near-term decarbonisation opportunities. Without CCUS, the annualised decarbonisation costs of all clusters decrease substantially, with the threshold for a high-decarbonisation cluster dropping from USD 30 per tonne CO2 per year to USD 10. The emissions mitigated also decline significantly, indicated by the reduced bubble size.

• In this scenario, Raipur and Bellary emerge as low-decarbonisation cost clusters, while Raigarh’s costs also approach low-cost levels.
• Bellary transitions from low readiness to high readiness. Its annualised decarbonisation cost turns negative, as it has no requirement for FDRE or imported fuels and substantial WHR potential, resulting in net cost savings.
• The MAC of FDRE is low in Raipur, owing to higher grid dependence and lower OA charges, which make FDRE deployment more economical, resulting in lower overall costs.
• Sundargarh becomes the highest cost cluster, as it must displace 88 per cent of domestic coal, and transitioning to imported coal remains expensive (as shown in Figure ES3).

Since using imported coal as a decarbonisation measure significantly increases costs, we developed another cluster plot that excludes alternative fuels in addition to CCUS. The clusters’ annualised cost scores are now primarily shaped by the remaining four levers: WHR, scrap utilisation, FDRE, and pelletisation. In terms of readiness, the 80 per cent weight assigned to decarbonisation levers is allocated to EE, ME, and RE in proportion to the emissions mitigated by each measure. Emission reductions further decline in clusters that need alternative fuels (coal substitution with imported coal) to decarbonise.

• In this scenario, the Sundargarh cluster shifts from a high-cost to a low-cost cluster, as alternative fuels are excluded from the assessment.
• Jharsuguda remains a high-cost cluster owing to higher FDRE requirements than Sundargarh and a low WHR potential.
• The Raigarh cluster incurs higher costs because it depends on FDRE, whereas WHR only partially meets its power requirements.
• All clusters exhibit high readiness, ranging from 0.50 in Jharsuguda to 0.87 in Bellary. Odisha faces lower readiness scores due to higher OA charges, and its Green Energy Open Access Rules (GEOAR) are still in draft form. This is particularly relevant to clusters requiring FDRE. In addition, Odisha’s clusters have lower pelletisation capacity than those of other states, further reducing their readiness.

Regardless of changes in the annualised decarbonisation cost and readiness levels arising from the inclusion or exclusion of specific decarbonisation levers, the evaluation framework remains scalable. It allows for new clusters and improved data to be incorporated over time. With additional surveys and stakeholder inputs, the framework can evolve into a robust national planning framework for decarbonising India’s secondary steel sector.

Policy recommendations and conclusions

Cross-cutting recommendations

Given our findings on the limited near-term feasibility and costs of CCUS and expensive alternative fuels through imported fuel use, this section concentrates on interventions that can be implemented immediately and suggests piloting CCUS.

Ecosystem enablers (short-term)

• The Udyam platform should be leveraged to disseminate information on potential support and policy mechanisms introduced by the Ministry of Steel (MoS) for small and medium enterprises (SMEs) in the steel sector. Further, SIMA and state-level associations, supported by the MoS, BEE, and other relevant institutes (such as SIDBI and think tanks), should enhance policy awareness among SMEs through SIMA-led conferences, cluster-level workshops, and the establishment of instant messaging channels designed to disseminate knowledge tailored explicitly to DRI units.
• The BEE should implement disclosure and reporting of emissions data by DRI units by formulating standardised monitoring and verification frameworks supported by digital tools, including AI-based and other platforms. These datasets should be made publicly accessible and updated annually.
• The MoS, in collaboration with various ecosystem players, should develop a programme to identify two or three sustainability champions within state sponge iron associations. These champions would work with the ESCOs to implement decarbonisation initiatives and develop model plants within the cluster. They should also coordinate SME participation, promote engagement, and recognise the top three SMEs with rewards and incentives.

EE (short-term)

• The MoS should assess the feasibility of pooled procurement models to aggregate demand for WHR and to benefit from economies of scale. State sponge iron associations should coordinate cluster-level procurement to achieve economies of scale and reduce procurement costs.

ME (medium-term)

• The BEE should mandate the inclusion of standalone pellet plants in National Carbon Market (NCM) frameworks, including the Carbon Credit Trading Scheme (CCTS) and future emissions trading schemes. Emissions abated from pellet use should be fully credited downstream to C-DRI producers.

RE (short to medium-term)

• State energy departments and discoms seeking coal power capacity should assess opportunities to reallocate the power from captive coal power plants through power purchase agreements (PPAs) or by selling power in secondary markets, thus freeing up capacity for RE integration.
• State electricity regulatory commissions (SERCs) should rationalise cross-subsidy surcharges (CSSs) and additional surcharges for RE-based OA procurement. They should reduce CSSs as per Karnataka’s tariff order (KERCb 2024) and adopt models implemented in Tamil Nadu and Gujarat to assess additional surcharges more accurately.
• State industry associations should aggregate RE demand at the cluster level. States can establish or utilise their renewable energy implementing agencies (REIAs). Additionally, the respective state renewable development agencies should introduce single-window clearance systems to streamline approvals, land access, and grid connectivity for RE developers.
• State energy departments, in collaboration with SERCs, should explore the feasibility of behind-the-meter RE, peer-to-peer energy trading, and group captive procurement models to increase renewable adoption and bypass high OA and wheeling charges.

Demonstration and pilots (short to medium-term)

• The MoS should establish a model low-carbon DRI plant using the natural-gas route and rotary kilns equipped with all decarbonisation levers at the cluster level. The plant would demonstrate integrated decarbonisation strategies and serve as a replicable learning hub for SMEs.
• The MoS should also explore the feasibility of developing a pilot group-captive vertical-shaft furnace for gas-based DRI production. The Steel Research and Technology Mission of India (SRTMI) can play a crucial role in guiding the research and development (R&D) activities and disseminating information about the pilot study across all clusters.

Cluster-specific recommendations (Cluster-wise recommendations are depicted in Chapter 6)

Clusters such as Bellary and Sundargarh, with high WHR potential, should focus on maximising WHR utilisation and enabling local sale of surplus power to nearby induction furnaces. Clusters with limited access to pellets (Odisha and Raigarh) require policy and infrastructure support to expand pelletisation capacity and lower transport costs through slurry pipelines. Chhattisgarh should reduce the value-added tax on natural gas, which currently raises delivered prices by nearly 30 per cent compared to Karnataka. States with lower OA charges, such as Chhattisgarh, should leverage these advantages to expand RE supply across clusters. Karnataka should notify green hydrogen policies and reduce OA charges to enable early adoption and offtake aggregation in industrial clusters. Strengthening institutional mechanisms, including the revival of defunct associations such as OSIMA in Odisha, will be crucial for coordinating decarbonisation efforts in the state.