Reducing Blast Furnace Dependence on Coke: A Review of Low-Carbon Strategies

Abstract

Reducing coke consumption is one of the most effective approaches for lowering CO₂ emissions in blast furnace (BF) ironmaking. Coke plays both chemical and physical roles in the BF, acting as a reductant, providing heat, and supporting the burden to maintain gas permeability. While part of its chemical function can be replaced by alternative carbon-based fuels and reducing agents, its physical load-bearing role remains difficult to substitute. This review summarizes recent studies on the partial replacement of coke in BFs from the perspective of low-carbon ironmaking. Rather than treating different coke-saving technologies as separate methods, this review discusses coke reduction according to the specific functions of coke being replaced. Alternative carbon sources, H₂-rich reductants, CO₂-to-CO recycling, load-bearing packed-bed materials, and burden structure optimization are considered. This perspective helps explain why alternative fuels, H₂-rich gases, and recycled CO can reduce the chemical demand for coke, while deeper coke reduction remains limited by the difficulty of replacing coke’s structural role. Overall, meaningful coke reduction is more likely to rely on the combined use of several approaches while maintaining process stability and engineering feasibility. This function-based understanding may help guide the development of more practical lower-carbon BF ironmaking strategies.

Introduction

Over the past three decades, global energy-related CO₂ emissions have risen sharply, increasing from 22.6 billion tonnes in 1996 to more than 38 billion tonnes in 2025, largely driven by energy combustion and industrial processes.¹ The steel industry contributes approximately 8–9% of global CO₂ emissions, with the blast furnace–basic oxygen furnace (BF-BOF) route remaining the dominant production pathway.^2–4 The BF-BOF process generates around 1.8 t CO₂ per ton of steel, with over 80% of the total CO₂ emissions arising from the BF ironmaking stage.^5–8 The reduction of BF carbon emission is therefore urgent for combating global warming challenges.^9,10

Coke plays an essential role in conventional BF operation, serving three primary functions. First, it serves as the primary reductant and heat source (chemical function). Coke reacts with iron oxides to form Fe (either or both direct and indirect) and the combustion of coke through the tuyere region creates a high-temperature environment, providing heat for the reduction reaction in the BF.¹¹ Meanwhile, carbon and CO₂ also react to produce CO, which reduces iron oxides.^12,13 In recent decades, technologies have been developed/under development to replace part of coke’s chemical function by using auxiliary fuels, including pulverized coal, natural gas, H₂-rich gases, and biomass fuels.^14–18 More recently, CO₂ capture, conversion and utilization strategies, such as CO₂ electrolysis and CO₂ splitting using perovskite-based oxygen carriers,^19–23 have been proposed to regenerate CO from the captured CO₂ and re-inject it into the BF. Under the premise that CO₂ can be efficiently converted into CO and recycled to participate in indirect reduction reactions in the BF, the chemical function demand for coke has a great potential to be substantially reduced. Second, coke serves as a carburizer. The dissolution of carbon into the hot iron metal is essential for stable production, typically resulting in pig iron containing 4–4.5 wt.% C.^24,25 Third, coke serves critical physical and structural support. It maintains gas permeability through its porous bed and acts as a stabilizer in the cohesive zone where softening and melting of iron-bearing materials occur.^26,27 Even as gaseous reducing agents (CO and H₂) increasingly replace coke’s chemical duties, the BF still requires a robust, high-temperature-resistant internal structure to ensure mechanical stability and permeability.^28,29 These physical functions remain the most difficult to replace. Thus, the way forward is not simply to address the challenges of the coke elimination but also to explore potential materials and operational strategies to partially replace the load-bearing function of coke. Such materials should not melt, incur excessive undesired reactions, or generate unwanted by-products under BF conditions. This issue represents an important long-term research challenge for deeper coke reduction in BF ironmaking.

This paper aims to systematically review recent efforts toward the partial replacement of coke in BFs for low-carbon ironmaking. The purpose of this review is not to present the chemical and structural functions of coke as a new concept, since these roles are already well recognized in BF ironmaking. Rather, these functions are used as an analytical framework for critically assessing the real substitution potential of different coke-reduction strategies. Many existing studies discuss auxiliary fuel injection, hydrogen enrichment, biomass use, carbon recycling, and burden optimization as separate approaches. However, these routes target different functions of coke and therefore have different limitations. For example, alternative carbon sources and H₂-rich gases can reduce the chemical demand for coke, but they do not directly replace the load-bearing and permeability-maintaining function of coke in the cohesive zone and lower furnace. Conversely, potential load-bearing materials may address the structural role of coke, but their feasibility is still highly uncertain because of slag attack, abrasion, cost, recyclability, and compatibility with dynamic BF operation. Therefore, this review evaluates coke-reduction strategies by considering not only their carbon-saving potential but also the coke function targeted, industrial constraints, and possible influence on furnace stability. This distinction does not imply that the chemical and structural roles of coke can be physically separated in real BF operation. In the cohesive zone and lower furnace, these roles are strongly coupled: solution-loss reactions consume carbon and weaken coke strength and permeability, while coke layers control gas flow, liquid drainage, heat transfer, and carburization. This framework helps explain why moderate coke reduction can be achieved through several mature technologies, whereas deeper coke reduction remains limited by the difficulty of replacing the structural role of coke. Based on this assessment, the review identifies practical near-term routes and longer-term research needs for lower-carbon BF ironmaking.

This article is structured in the following manner. The "Introduction" section above introduces the background to low-carbon BF ironmaking and the challenge of reducing coke dependence while maintaining furnace stability. The "Technical Routes for Reducing Coke Dependence of BFs" section summarizes the main strategies for partial replacement of coke functions, focusing on alternative carbon sources, H₂-rich reduction, and potential load-bearing materials. The "Burden, Furnace Structure, and Operational Optimization as Indirect Coke-Saving Routes" section discusses burden, furnace structure, and operational optimization as indirect coke-saving routes. The "CO₂-to-CO Conversion as an Enabling Route for Low-Coke BF Operation" section introduces CO₂-to-CO conversion technologies as an enabling approach for carbon recycling within BF system. The "Overall Comparison of Coke-Reduction Strategies" section provides an overall comparison of coke-reduction strategies. Finally, key conclusions and future perspectives are presented in the "Conclusions and Perspective" section.

Technical Routes for Reducing Coke Dependence of BFs

Since coke performs both chemical and physical roles, most studies focus on partially replacing specific functions rather than eliminating coke entirely. Existing approaches can be generally divided into three categories (see Fig. 1): (1) alternative carbon sources to provide part of the carbon input, (2) H₂-rich reduction atmospheres, where H₂-bearing gases directly decrease the demand for carbon-based reduction, and (3) load-bearing packed-bed materials that may help maintain burden structure and permeability. Among these routes, the first two mainly target the chemical demand for coke, whereas the third attempts to address its much harder-to-replace structural role.

Fig. 1.

Main approaches for reducing coke consumption and structural dependence in BF ironmaking.

Alternative Carbon Sources

Low-Rank Coal and Semi-coke

Low-rank coals and semi-coke powders usually exhibit high oxygen content and well-developed porosity, which lead to significantly higher CO₂ reactivity after carbonization compared with conventional BF coke.^30–34 Considerable efforts have been made to upgrade these materials prior to briquetting. For example, Mollah et al.³⁵ showed that hydrothermal dewatering combined with acid washing reduced oxygen functionality and enhanced structural ordering of brown coal. After hot briquetting (150–230°C) and carbonization at 1000°C, the resulting products showed strength close to that of BF coke, and exhibited lower surface area and reactivity compared to untreated coal-derived chars, which was attributed to the development of a more graphitic structure. In a later study, the same group found that alkali treatment followed by hot briquetting produced low-surface-area carbon materials with improved mechanical integrity.³⁶ However, even after these treatments, the CO₂ reactivity remained higher and the graphitic ordering remained lower than that of standard BF coke. This suggests that such briquette cokes may be suitable for partial replacement, but are still difficult to use as a full substitute for coke in BFs.³⁷

In addition to these pretreatment methods, semi-coke fines from low-rank coal pyrolysis have also been consolidated using carbonated consolidation processes. Guo et al.³⁸ reported that optimization of the briquetting pressure to 93.63 MPa, together with Ca(OH)₂-assisted carbonation, produced coke with good compressive strength, acceptable heating value, and reduced SO₂/NO_x emission. Therefore, it is clear that pretreatment methods directly affect pore development, surface area, and CO₂ reactivity, which are critical factors for stable BF operation. However, even under such modification, low-rank coal-based-formed coke still cannot match the hot strength of traditional coke around 1200–1400°C. The main reason lies in its insufficient plastic phase formation and less optimized meso-/macropore structure during carbonization.^35,36 Overall, these studies show that low-rank coal and semi-coke are more suitable as partial carbon substitutes than as direct replacements for the high-temperature structural function of coke. Their main value lies in reducing fossil carbon input, but their weaker hot strength remains a key limitation for further substitution.

Biomass-Based Materials

Biomass-based carbon materials, including wood chips, straw, and other agricultural and forestry wastes, have been explored as renewable substitutes for fossil-based coke in BF operation to reduce net CO₂ emissions.^39,40 The study by Seo et al.⁴¹ showed that biomass–coal mixtures could produce bio-coke with calorific values exceeding 7000 kcal/kg and ignition temperatures in the range of 400–600°C, while maintaining the basic fuel requirements. However, Florentino-Madiedo et al.⁴² reported that biomass addition often lowers coal blend fluidity and weakens thermoplastic behavior, unless suitable binder systems or pre-treatment methods are applied. Thus, pretreatment plays a critical role in improving structural performance. For example, Wibawa et al.⁴³ found that binderless hot briquetting of torrefied woody biomass followed by carbonization at 1000°C produced formed coke with indirect tensile strengths ranging from 8 MPa to 32 MPa, depending strongly on torrefaction temperature and particle size. An optimal torrefaction temperature at around 275°C was found to enhance interparticle bonding and densification during carbonization, suggesting that biomass pretreatment is essential for achieving acceptable mechanical properties.⁴³ More recently, biomass upgrading routes combining hydrothermal carbonization and pyrolysis have also been proposed to produce higher-quality BF injection fuels, with reduced alkali metal content, ash content, and specific surface area, as well as improved carbon ordering and gasification behavior.⁴⁴

At the industrial scale, biomass-containing briquettes have demonstrated technical feasibility. Mousa et al.⁴⁵ reported in BF trials that charging 55% bio-briquettes (~ 64 kg/tHM) reduced carbon consumption by ~ 10 kg/tHM and lowered total CO₂ emissions by 33–40 kg/tHM, without significant deterioration in shaft permeability or gas utilization. The process model conducted by Orre et al.⁴⁶ further indicated that combining the biomass injection with the top charging strategy can reduce CO₂ emissions by up to 30–35%, while maintaining stable operation of the furnace. Recent feed-flexible BF concepts further suggest that tailored biochar may be used in different forms, including tuyere injectant, biocoke blend component, and limited top-charged lump biochar, although each route remains constrained by material quality, availability, and furnace stability.⁴⁷

Nevertheless, biomass-based carbon materials generally exhibit higher intrinsic reactivity and a lower degree of graphitization than conventional BF coke. As a result, their resistance to high-temperature degradation in the cohesive zone remains limited, restricting their role primarily to partial substitution rather than full structural replacement.⁴⁸ Therefore, the main significance of biomass-based materials is not that they can replicate coke performance, but that they can lower the net carbon footprint of BF ironmaking when introduced in controlled proportions.

Carbon-Bearing Wastes

Carbon-bearing waste materials from both steel plants and other waste sources provide an additional route for reducing fossil coke consumption. Residues generated within steel plants, including coke breeze, BF dust and sludge, etc., contain various amounts of carbon.^3,49–52 Cold-bonded briquetting is widely adopted to convert these materials into mechanically stable agglomerates.³⁴

Li et al.⁵³ studied reduction kinetics of cold‐bonded briquettes prepared from return fines of sinter with carbon monoxide and coke, and observed that briquettes containing coke often exhibit a higher reduction velocity index and a lower apparent activation energy (32.4 kJ mol⁻¹) compared with coke-free briquettes (38–57 kJ mol⁻¹), leading to more complete phase transformation and improved metallurgical performance. For briquettes prepared from BF dust or sludge, self-reduction can start at around 500°C and is nearly complete by 1200°C. During heating, the decomposition of hydrates and carbonates generates gaseous intermediates that promote carbon gasification and accelerate the stepwise transformation of iron oxides.⁵⁴ In addition, Fe-bearing process wastes can also be recycled through RHF-based composite pellets to produce DRI for BF use. Chung et al.⁵² reported that pellets prepared from FINEX process wastes and anthracite achieved over 84% oxygen removal at 1300°C, indicating the potential of converting integrated steel plant wastes into recyclable BF feedstock, although problems such as incomplete reduction at lower temperatures, pellet bursting, and impurity-related operational problems still need to be addressed. Coke breeze has also been improved through pitch-bonded briquetting followed by carbonization, producing formed coke with compressive strengths above 30 MPa and enabling partial substitution of metallurgical coke in BF operation.^55,56 Diez et al.⁵⁷ reported that briquetted coal-tar sludge and oily residues could be added to coking blends at up to 10 wt.% in coke ovens, and when binder type and particle size were optimized, coke reactivity and post-gasification strength remained largely unchanged, allowing coke quality to meet BF requirements. Waste plastics can also serve as supplementary carbon sources, mainly through tuyere injection. Their high hydrogen content supports combustion and indirect reduction, but the low mechanical strength, low softening temperature, and poor gas permeability limit their application in large proportions.^58,59 Recent flash-pyrolysis and burnout studies of plastic- and paper-derived refuse fuels further indicate that such alternative reductants must be evaluated under raceway-relevant heating rates and short residence times before large-scale BF injection.⁶⁰

These studies show that carbon-bearing wastes can provide carbon and help reduce landfill disposal. However, recycling carbon-bearing wastes requires the careful control of impurity accumulation, thermal decomposition behavior, and possible effects on slag chemistry, so as to avoid negative effects on furnace permeability and slag quality.⁶¹ Compared with low-rank coal and biomass, the key advantage of carbon-bearing wastes is circular resource utilization. However, their larger compositional variation and impurity burden make process control more critical. In practice, their role is again mainly chemical substitution rather than replacement of the structural function of coke.

For practical BF operation, the use of carbon-bearing wastes should also be assessed from the viewpoint of long-term impurity accumulation. Elements such as alkali metals, Zn, Pb, S, P, and Cl, as well as ash-forming oxides, may circulate or accumulate in the furnace and influence burden degradation, gas permeability, slag chemistry, viscosity, melting behavior, and slag–metal separation.^51,62 Therefore, the suitability of such wastes should be evaluated not only by their carbon content and calorific value, but also by impurity mass balance, high-temperature softening/melting behavior, slag–property changes, and long-term pilot- or plant-scale monitoring.

H₂-Rich Reduction Atmospheres

Partial replacement by H₂ in the reduction process can weaken the solution loss reaction and reduce coke degradation in the lower part of the BF. In this way, the structural demand for coke can also be lowered to some extent by changing the furnace atmosphere.^63–65 This route does not replace coke physically, but reduces the rate at which coke is consumed and degraded, thereby indirectly preserving its supporting role.

Reducing Coke Solution Loss in the BF

Coke degradation in the BF is largely governed by the following carbon solution loss reactions:^66–68

$$\text{C}\left(\text{s}\right)+{\text{CO}}_2\left(\text{g}\right)=\text{2CO}\left(\text{g}\right)$$ 1

$$\text{C}\left(\text{s}\right)+{\text{H}}_2\text{O}\left(\text{g}\right)=\text{CO}\left(\text{g}\right)+{\text{H}}_2\left(\text{g}\right)$$ 2

The extent and location of these reactions strongly affect coke porosity development and strength loss.^69,70 A recent review further emphasized that H₂O generated under H₂-rich BF conditions can become an important gasifying agent and accelerate coke structural degradation, making H₂O control critical for maintaining coke integrity.⁷¹ Shin and Jung⁶⁹ observed that, compared with CO₂, involving H₂O (g) in the gasification process, the reaction was more likely to occur near the surface of the coke, which led to a more significant decrease in the strength of the coke. Therefore, controlling the generation of CO₂ and H₂O (g) in the high-temperature region is crucial for maintaining coke integrity and its supporting role in the furnace. Lan et al.⁷² found from thermodynamic calculations and experimental results that increasing H₂ content in the reducing gas could promote the reduction of iron oxides to lower temperature regions and reduce the generation of CO₂ in the high-temperature zone. When φ(CO)/φ(H₂) decreases, the main carbon gasification temperature zone becomes narrow and moves downward in the BF, effectively weakening coke solution loss in the furnace bosh and cohesive regions.⁷² This trend is consistent with coupled coke gasification and sinter reduction experiments in CO-N₂-(25%)H₂ atmospheres, which showed that the enrichment of H₂ changed the evolution of the carbon structure of coke and the formation process of pores. With less CO₂ produced and a modified reaction pathway, H₂-rich conditions can weaken coke structural degradation and help retain more post-reaction strength under moderate H₂ levels.^73,74 Thus, H₂-rich operation indirectly stabilizes the coke structural role by relocating and weakening solution loss reactions rather than strengthening coke directly. The key finding here is that H₂ can lower coke degradation not because coke becomes intrinsically stronger, but because the reaction environment becomes less destructive to coke in the most critical furnace zones.

Indirect Reduction with H₂

Hydrogen reduction of iron oxides proceeds via:⁷⁵

$$\text{FeO}\left(\text{s}\right)+{\text{H}}_2\left(\text{g}\right)=\text{Fe}\left(\text{s}\right)+{\text{H}}_2\text{O}\left(\text{g}\right)$$ 3

Compared with CO, H₂ has higher diffusivity and generally faster reaction kinetics, enhancing indirect reduction rates in the shaft region of the BF.^76–78 Numerical simulations by Mauret et al.¹⁵ showed that intensive H₂ injection, via coke oven gas, natural gas, or pure H₂, can increase the fraction of indirect reduction by H₂ at the expense of CO-based reduction, thereby lowering the carbon demand per mole of oxygen removed, significantly reducing the coke rate, and maintaining stable furnace operation within acceptable process limits. Okosun et al.⁷⁹ carried out CFD analysis and reported that up to ~ 35 kg H₂/tHM injection can reduce coke consumption by approximately 60 kg/tHM under controlled flame temperature conditions. Zhao et al.⁸⁰ further found that shaft injection of reformed coke oven gas could enhance near-wall reduction degree, narrow the cohesive zone, and decrease coke rate through improved H₂ utilization efficiency. Similarly, a multi-fluid BF modeling was performed by Yu and Shen,⁸¹ which indicated that H₂ enrichment could promote earlier reduction in the upper shaft, facilitate the cohesive zone upward, and increase gas utilization efficiency.

However, H₂-based reduction also presents several challenges. H₂ reduction may suffer from diffusion limitations in the later stages of iron oxide reduction, which can hinder complete metallization and influence overall process efficiency.⁶⁴ Therefore, although H₂ can promote indirect reduction and lower coke consumption, these kinetic limitations must be considered when evaluating the H₂-enhanced ironmaking process. Overall, the benefit of H₂-rich operation lies in increasing indirect reduction efficiency and lowering carbon demand, but the extent of substitution is still constrained by kinetics, thermal balance, and modeling uncertainties related to gas utilization, temperature distribution, and operational limits.⁸²

Behavior of Water–Gas Shift Reaction

A critical aspect of H₂-rich operation is the water–gas shift reaction:⁸³

$$\text{CO}\left(\text{g}\right)+{\text{H}}_2\text{O}\left(\text{g}\right)={\text{CO}}_2\left(\text{g}\right)+{\text{H}}_2\left(\text{g}\right)$$ 4

H₂ enrichment increases H₂O generation through H₂-based indirect reduction (i.e., reaction 3), which in turn promotes the above water–gas shift reaction. As a result, the balance among CO, CO₂, H₂, and H₂O is changed, directly influencing both CO and H₂ utilization efficiencies.¹⁵ Kemppainen et al.⁸⁴ carried out experimental studies on pellet beds and found that the water–gas shift reaction becomes both thermodynamically feasible and kinetically significant at 300–500°C, particularly in the presence of magnetite, which exhibits a catalytic effect. Nogami et al.⁸⁵ also investigated the effect of this water–gas shift reaction by numerical simulation, further demonstrating that this reaction plays a decisive role in the formation of the gas composition field, and therefore should be considered in H₂-intensive BF models. Thus, proper control of the water–gas shift reaction enables improved CO efficiency and enhances overall gas utilization. However, the increased H₂O content under H₂-rich conditions may also accelerate coke gasification through the C-H₂O reaction, leading to coke solution loss, pore development, strength degradation, and possible deterioration of burden permeability. The extent of this effect depends on the local H₂O partial pressure, temperature, residence time, coke reactivity, and gas-flow distribution. Although the direct effect of H₂O on oxide furnace materials is usually less significant than slag/metal corrosion, the associated change in local oxygen potential should still be considered in H₂-rich BF operation.

H₂ injection is also constrained by its impact on the thermal state of the lower furnace.^86–88 Recent modeling of the raceway adiabatic flame temperature indicates that H₂-rich gas injection must be coupled with suitable oxygen enrichment, blast-temperature control, and injection-rate optimization to maintain an acceptable tuyere-raceway temperature.⁸⁹ This is because H₂-related reactions are strongly endothermic, and excessive H₂ addition can lower the flame temperature and may destabilize the lower furnace.⁷⁹ Recent numerical simulations further show that H₂ and pulverized coal co-injection, as well as the interaction between H₂ injection and oxygen enrichment, strongly affect PCI combustion, raceway gas composition, and temperature distribution.^90,91 Moreover, the type and injection amount of H₂-rich gas, together with natural gas and pulverized coal co-injection strategies, determine the feasible operation window for reducing carbon emissions while maintaining raceway stability and lower-furnace heat balance.^92,93 In addition, numerical simulations of biomass- and coke oven gas-derived H₂-rich fuels co-combusted with pulverized coal further suggest that alternative fuel injection should be evaluated together with combustion efficiency, temperature distribution, and lower-furnace thermal stability.⁹⁴ Similar thermodynamic limitations have also been reported for CH₄ injection, where excessive endothermic reaction reduces the heat supply to the lower zones.⁹⁵ In this case, top gas recycling oxygen BF systems are proposed which enable H₂-rich operation by recycling CO-rich streams, thereby reducing overall carbon demand, fuel rate, and CO₂ emissions while maintaining thermal balance.^96,97 More recently, Zhang et al.⁹⁸ investigated the effect of H₂-rich gas composition on energy consumption and raceway state in an oxygen BF, further showing that the composition and injection strategy of H₂-rich gas must be carefully controlled to balance carbon reduction, energy consumption, and raceway stability. Furthermore, Babich et al.⁹⁹ clarified that the optimization of H₂ injection distribution at the shaft and tuyere could effectively reduce the coke ratio, while maintaining the stability of raceway temperature and cohesive zone, thereby achieving a higher H₂ utilization rate. Therefore, the practical value of H₂-rich BF operation depends not only on reduction efficiency, but also on whether gas composition, heat balance, and lower-furnace stability can be simultaneously maintained.

Industrial Demonstrations and Practical Limits of H₂-Rich BF Operation

In addition to laboratory studies and numerical simulations, several industrial or near-industrial demonstrations have provided useful evidence for the practical implementation of H₂-rich BF operation. For example, H₂ injection was tested in an operating BF at the Duisburg site of Thyssenkrupp Steel,^100,101 where H₂ was injected through tuyeres as a partial replacement for pulverized coal. These tests provided important operational information on lance position, flow and pressure conditions, and the interaction between H₂ injection and high-temperature BF operation. In Japan, the COURSE50 project has further investigated H₂ injection into BFs using coke oven gas, room-temperature H₂, and heated H₂ in experimental BF systems.¹⁰²

These demonstrations confirm that H₂-rich BF operation is technically feasible and can reduce the carbon demand associated with iron oxide reduction. However, they also show that H₂ injection should be regarded as a partial decarbonization route rather than a complete replacement for coke. The achievable benefit depends strongly on H₂ availability and cost, tuyere and shaft injection strategy, oxygen enrichment, raceway adiabatic flame temperature, gas utilization efficiency, and the control of H₂O-induced coke gasification. Moreover, H₂ injection mainly replaces part of the chemical function of coke and does not directly replace its load-bearing and permeability-maintaining role in the cohesive zone and lower furnace. Therefore, industrial H₂ BF demonstrations are valuable as bridge technologies for lowering coke consumption, but deeper coke reduction still requires simultaneous control of furnace thermal balance, burden permeability, and coke structural degradation.^15,101

Load-Bearing Packed-Bed Materials

Instead of replacing coke with reactive carbonaceous materials or using an H₂-rich atmosphere, another approach is to introduce load-bearing packed-bed materials into the high-temperature zone of the BF, which can reduce the reliance on coke for mechanical support and maintain the permeability and structural stability under BF conditions.¹⁰³ Potential materials must satisfy some critical requirements: (1) high-temperature mechanical strength under loading, (2) chemical stability with CaO-SiO₂-Al₂O₃-MgO-FeO slags and molten iron, and (3) a sufficiently stable pore structure to maintain gas permeability. However, most existing studies focus on refractory or ceramic material performance in static corrosion environments rather than in typical dynamic slag-metal systems relevant to real BF operation. Therefore, the materials reviewed in following subsections should be regarded mainly as exploratory candidates that may provide guidance for future investigation. This route is conceptually important because it directly targets the structural role of coke, but it is also the least mature among all available approaches. The present discussion should therefore not be interpreted as evidence that ceramic or hollow packed-bed materials are ready for industrial BF application. Instead, these materials are reviewed to define the performance requirements and failure modes that must be addressed before any structural coke substitute can become realistic.

Oxide-Based Ceramic Systems

Porous oxide ceramics with high melting temperatures have been considered as potential load-bearing materials. Among them, MgO-Al₂O₃ materials are attractive due to their high melting temperatures and spinel-forming capability. Recent machine-learning-assisted optimization of multi-element spinel refractory materials also suggests that composition design can be used to improve mechanical properties, providing useful guidance for future development of spinel-based load-bearing ceramic candidates.¹⁰⁴ Salomão et al.¹⁰⁵ investigated porous alumina–spinel ceramics for high temperature applications and found that MgAl₂O₄-containing porous ceramics could retain a stable porous structure even after high-temperature treatment above 1300°C, which suggests the possibility of maintaining structural integrity at BF-related temperatures. However, their interaction with BF slags still requires careful consideration. Wang et al.¹⁰⁶ combined thermodynamic calculations with experimental validation and showed that Al₂O₃-rich refractory materials could react with CaO-containing slags to form calcium aluminates such as CaAl₁₂O₁₉ and CaAl₄O₇. Interface investigations by Park et al.¹⁰⁷ further revealed the formation of reaction layers including CaAl₁₂O₁₉ and Mg(Mn)Al₂O₄ spinels in CaO-CaF₂-SiO₂-Al₂O₃-MgO-MnO slags. While these reaction layers may sometimes act as diffusion barriers under static conditions, their stability under dynamic BF-relevant environments remains uncertain because slag flow, metal dripping, and thermal cycling may disrupt or dissolve the calcium aluminate layer. Therefore, Al₂O₃-MgO systems should be regarded as semi-inert materials with conditional interfacial protection rather than fully stable load-bearing substitutes.

ZrO₂-based materials are also frequently investigated due to their extremely high melting temperature and potential for forming protective calcium zirconate layers. In high-basicity slag systems, ZrO₂ can react with CaO to form CaZrO₃ and related high-melting compounds that limit further erosion and reaction of the slag.¹⁰⁶ However, Kumar et al.¹⁰⁸ reported that, under oxidizing and MnO-rich slag conditions in ZrO₂-C refractories, the consumption of phase-stabilizing Y₂O₃ from ZrO₂ granules may result in their fragmentation and subsequent diffusion in slag-rich regions, eventually leading to weakening of their structural strength. In comparison, MgO-ZrO₂ composite systems showed improved resistance in some metallurgical slags due to the formation of high-melting interfacial phases of ZrSiO₄ and MgO-ZrO₂-based stable solid solutions, which can reduce slag wettability and impede the penetration path.^109,110 This suggests that composite oxide designs may provide better semi-inert performance than simple oxide systems in some cases, although their long-term behaviors under dynamic BF conditions still require further evaluation. In short, current research in oxide-based ceramic systems demonstrate that high-temperature stability alone is not sufficient, and the decisive issue is whether the material can retain both structure and chemical compatibility under continuous slag attack and abrasion.

Hollow Spheres and Engineered Spherical Materials

From a structural perspective, hollow ceramic spheres are among the material forms closest to porous load-bearing materials, where the particle size, the pore structure, and the permeability can be controlled through synthetic parameters. Qi et al.¹¹¹ investigated hollow sphere ceramics and showed that, as the sintering temperature increased from 1400°C to 1600°C, the samples shrunk increasingly and the porosity decreased from 59% to 42%, which led to an increase in the strength of the alumina foams from 6.9 MPa (at 1400°C) to 100.0 MPa (at 1550°C). They further found that the mechanical strength of these materials was highly controlled by the contact area between the hollow spheres, which could be enlarged by increasing the sintering temperature, reducing the hollow sphere size or introducing slurry infiltration.

Hollow alumina sphere castables have also been explored to construct lightweight aggregate systems with low thermal conductivity and reasonable mechanical strength. Xu et al.¹¹² studied hollow alumina sphere castables containing in situ-formed Y₃Al₅O₁₂, and found that interfacial strengthening at the sphere–matrix interface improved the strength of individual particles, increased cold crushing strength and high-temperature modulus of rupture, and further reduced thermal conductivity. These studies suggest that interfacial engineering and multiphase design can help maintain mechanical performance under higher temperatures. The key insight is that structural design at the particle and interface levels can substantially improve strength; however, evidence under real BF conditions with slag and hot metal remains lacking.

Industrial Concepts from Cupola Furnace Systems

In addition to laboratory studies, several patents have proposed the use of engineered spherical materials to replace the structural role of coke in cupola furnaces rather than BFs. Ceramic balls prepared from Al₂O₃-SiO₂-C-binder systems have been suggested as packed-bed materials to support the burden while maintaining gas permeability at high temperature.¹¹³ Some designs also proposed to employ high density carbon balls in modified cupola furnaces, where the high-carbon balls replace traditional water-cooled grates and provide structural support for the burden.¹¹⁴ Simultaneously, they also performed a critical carburizing function by enriching the iron droplets with carbon to achieve the desired final chemistry. Although such concepts are mainly proposed for cupola-type systems rather than conventional BFs, they still provide an alternative approach for using engineered spherical materials to reduce the dependence on traditional coke.

Critical Remarks on Load-Bearing Packed-Bed Materials

Although oxide ceramics, hollow spheres, and engineered spherical materials provide useful concepts for reducing structural coke dependence, their transfer into real BF operation faces major physical, chemical, and economic challenges. In the lower part of the furnace, molten slag and metal droplets, together with severe abrasion and thermal cycling, are present simultaneously. Under such conditions, pores or interparticle voids may act as pathways for slag infiltration, leading to blockage, weight gain, or structural disassembly. In addition to these physical challenges, chemical stability remains a further concern. Al₂O₃-based materials can react with CaO-rich slag, while ZrO₂ may form protective calcium zirconate layers but can still degrade under BF conditions. Moreover, the behavior of materials is very different under dynamic abrasion and slag infiltration in real FeO-rich furnace environments.¹¹⁵

Beyond the technical feasibility, cost and large-scale production also present major barriers. The porous aggregates such as ZrO₂ or spinel are much more expensive than coke and would need to be supplied continuously at large scale in the BF. Issues related to slag separation, recycling, and coexistence with the existing BF conditions also remain unresolved. Therefore, these materials are better regarded as exploratory options and a long-term research direction for structural substitution, rather than immediately practical low-carbon solutions for industrial BF ironmaking.

Burden, Furnace Structure, and Operational Optimization as Indirect Coke-Saving Routes

An industrially feasible approach to reduce the coke content is not to replace the coke structural role directly, but to lower the mechanical and permeability demands through burden and furnace structure optimization. In addition to burden optimization, conventional operational measures also play important roles in reducing coke ratio in industrial BF practice. Increasing the hot-blast temperature can provide additional sensible heat to the lower furnace and reduce the amount of coke required for heat supply.^116,117 Oxygen-enriched blast increases combustion intensity in the tuyere-raceway region, promotes pulverized coal burnout, and allows a higher auxiliary fuel injection rate.¹¹⁸ High-pressure operation can improve gas–solid contact and gas utilization efficiency, helping to stabilize the furnace and lower the coke rate, although its practical benefit depends on furnace condition and operational constraints.¹¹⁹ However, these measures mainly reduce coke consumption through thermal and gas-flow optimization, rather than directly replacing the load-bearing and permeability-maintaining function of coke. Therefore, they should be considered as complementary industrial measures for coke saving, while deeper coke reduction still requires simultaneous consideration of furnace stability, cohesive-zone permeability, and coke structural function.

The softening-melting behavior of burden materials plays a key role in determining cohesive zone thickness and gas permeability. Recent studies under H₂-rich or hydrogen-enriched BF conditions further show that primary slag formation, carburization of metallic iron, and Al₂O₃-related slag characteristics can significantly affect cohesive-zone behavior and burden permeability.^120,121 Chiu et al.¹²² showed under simulated BF conditions that grain coarsening of Fe and FeO and liquid phase evolution could strongly influence dripping behavior and permeability. Mixed-burden experiments further showed that interaction between acid and basic components above 1150°C could modify liquid formation and interfacial reactions, thereby affecting cohesive zone structure.^123–125 In addition, optimization of MgO and Al₂O₃ contents in sinter or high-alumina burdens can adjust softening–melting intervals, slag characteristics, and permeability of mixed ferrous burdens.^121,126 In the upper shaft region, Barustan et al.¹²⁷ analyzed factors affecting the burden surface and porosity distribution, and showed that H₂-containing gases might enhance gas diffusion into particles and promote the hematite–magnetite transformation. Depending on burden type, this can induce internal cracking, alter degradation behavior, and change gas permeability. These effects also influence fines generation and therefore affect the level of structural support and permeability control required from coke.

From a structural perspective, burden distribution control is another critical concern. Simulation studies carried out by Wei et al.¹²⁸ and Roeplal et al.¹²⁹ showed that the material charging approach could strongly affect the radial distribution of burden materials, which relates to bed porosity and permeability. These changes redistribute the ascending gas flow and consequently influence the shape and position of the cohesive zone. Layered cohesive zone modeling by Dong et al.¹³⁰ further confirmed that proper burden layering could form low-permeability ore layers and highly permeable coke layers, causing gas to flow preferentially through the coke layers and liquids to pass through these permeable paths in the lower furnace.

Overall, improving burden quality, optimizing softening–melting behavior, and adjusting charging practice can help maintain acceptable gas permeability and liquid flow even at lower coke rates. Under these conditions, more coke mainly acts as a permeability stabilizer and carburizer, rather than serving as the primary structural role, providing a practical route for low-carbon BF operation. This route is particularly important because it does not require a new substitute material for coke. Instead, it reduces the extent to which the furnace relies on coke by making the burden-furnace system itself easier to operate at lower coke rates. In that sense, burden and furnace optimization is likely to be one of the most practical near-term approaches for reducing coke dependence.

CO₂-to-CO Conversion as an Enabling Route for Low-Coke BF Operation

Recent studies suggest that part of CO required in a BF can be provided by self-generated CO₂ rather than produced from coke, such as thermochemical CO₂ splitting using BCNF-type perovskite materials and electrochemical CO₂ conversion.^19–23 BCNF-based systems can convert CO₂ into CO through redox cycles at around 700–800°C, providing a potential approach to recycle CO₂ within the ironmaking process.^22,23 At the same time, CO₂ electrolysis technologies have also been proposed to transform captured BF CO₂ into CO-rich gas streams that can be re-injected into the furnace, which may significantly reduce fresh coke consumption while lowering overall CO₂ emissions.^19,21 In addition, CO₂ utilization through co-injection with H₂ into BF tuyeres has been evaluated as another possible route for carbon recycling and coke-rate reduction.¹³¹ These approaches suggest that CO used for iron oxide reduction does not necessarily have to come only from coke combustion but can also be regenerated from captured CO₂.

However, even if part of the reducing gas is supplied from recycled CO or H₂-rich gas streams, the BF still requires a solid carbonaceous material to maintain its internal structure. Therefore, most current low-carbon BF concepts focus on reducing rather than completely eliminating coke. This means that CO₂-to-CO conversion should be understood as an enabling technology for lowering the chemical demand for coke, not as a standalone solution to the coke problem. Therefore, in actual industrial settings, the process toward lower coke consumption is usually achieved through a combination of several approaches. One option is the partial replacement of conventional coke with the above-mentioned carbonaceous materials, which can decrease the demand for fossil carbon while maintaining acceptable burden permeability. Another is the injection of H₂-rich or CO-rich gases to increase the indirect reduction ratio and reduce the amount of carbon required for reduction reactions. In addition, optimization of burden composition and the cohesive zone behavior is also necessary to maintain stable gas flow and permeability when the coke rate is lowered.

In short, technologies such as BCNF-based CO₂ splitting or CO₂ electrolysis can be viewed as enabling approaches for lowering CO₂ emissions and generating CO. By converting unavoidable CO₂ emissions into effective CO, these technologies provide an additional source of reducing gas and help close the carbon loop within the BF system. When combined with H₂ enrichment, recycled gas injection, and improved burden design, coke can gradually shift from being the main reductant to primarily serving as a structural support material and carburizer in the furnace. Thus, the overall strategy for reducing coke demand in a low-carbon BF can be considered as a comprehensive optimization of three aspects: potential materials for partial replacement of coke function, burden and furnace structure optimization, and CO₂-to-CO conversion and recycling, as illustrated in Fig. 2. This suggests that meaningful coke reduction will likely require a combination of measures rather than relying on any single technology alone. Overall, these strategies do not reduce coke dependence in the same way. Some mainly lower the chemical demand for coke, while others help to reduce the structural dependence on coke. As shown in Fig. 3, the more mature options are generally those that reduce coke use indirectly. In contrast, direct replacement of coke’s structural role is still at an early stage.

Fig. 2.

Overall strategies to reduce coke demand in low-carbon BF.

Fig. 3.

Relative maturity and principal route of different coke-reduction strategies.

Overall Comparison of Coke-Reduction Strategies

Functional Comparison of Coke-Reduction Strategies

Table I summarizes the main coke-reduction routes from the viewpoint of coke function, carbon-saving role, and key industrial limitations, while their relative maturity and functional positioning are further illustrated in Fig. 3. This comparison shows that different strategies should not be judged only by their theoretical carbon-saving potential. Their practical value depends strongly on which coke function they target and whether furnace stability can be maintained.

Table I.

Critical comparison of representative coke-reduction routes according to coke function, carbon-saving role, and industrial limitations

Strategy	Main coke function targeted	Carbon-saving role	Key industrial limitation
Low-rank coal/semi-coke	Chemical demand reduction	Partial fossil carbon substitution	High reactivity; weak hot strength
Biomass/biochar/bio-coke	Chemical demand reduction	Lower net fossil CO2 input	Supply limitation, quality variation, weak structural role
Carbon-bearing wastes/by-products	Chemical demand reduction	Waste carbon recovery	Impurity accumulation, slag chemistry, permeability
Auxiliary fuel injection (especially PCI)	Chemical demand reduction	Reduces coke as fuel and reductant	Combustion efficiency, raceway stability, thermal balance
H2-rich gas injection	Chemical demand reduction	Lowers carbon-based reduction demand	H2 cost, thermal balance, H2O-induced coke degradation
Hot-blast temperature, oxygen enrichment, and high-pressure operation	Chemical demand reduction	Indirect coke saving	Limited by furnace stability and operating window
Top-gas recycling/oxygen blast furnace	Chemical demand reduction	High CO2-reduction potential	CO2 separation, O2 demand, energy penalty, system integration
Burden and furnace structure optimization	Indirect reduction of structural demand	Maintains permeability at lower coke rate	Ore quality, softening-melting control
CO2-to-CO conversion	Chemical demand reduction	Reuses captured CO2 as reducing gas	Conversion efficiency, energy demand, reactor integration
Load-bearing packed-bed materials	Direct structural support	Potential structural coke replacement	Slag attack, abrasion, cost, recyclability, large-scale supply

The comparison shows that current coke-reduction strategies should not be evaluated on the same basis. Alternative carbon sources, biomass-derived materials, and carbon-bearing wastes mainly reduce fossil carbon input, but their high reactivity, weaker hot strength, or impurity burden limits their ability to replace the structural function of coke. H₂-rich reduction and recycled CO injection can reduce the chemical demand for carbon and improve gas utilization, but they still require a stable coke bed to maintain permeability, liquid drainage, and cohesive-zone stability. Industrial operational measures, such as hot-blast temperature, oxygen enrichment, high-pressure operation, and burden optimization, are more mature and can reduce coke rate indirectly, but their effects are limited by the thermal and permeability window of the BF. In contrast, load-bearing packed-bed materials directly target the most difficult structural role of coke, but this route remains exploratory because such materials must withstand simultaneous slag attack, metal dripping, abrasion, thermal cycling, and large-scale economic constraints. Therefore, meaningful coke reduction in BF ironmaking is more likely to be achieved by combining mature chemical-substitution and operational-optimization routes, while treating structural coke replacement as a longer-term research challenge.

Techno-Economic and Implementation Considerations

Besides technical feasibility, the practical deployment of coke-reduction strategies must also be assessed from a techno-economic perspective. Mature routes such as PCI optimization, hot-blast temperature control, oxygen enrichment, high-pressure operation, and burden distribution optimization are more attractive for near-term application because they can be integrated into existing BF systems with limited reconstruction. In contrast, H₂-rich operation, top-gas recycling/oxygen BF systems, and CO₂-to-CO conversion require additional infrastructure for hydrogen supply, oxygen production, CO₂ separation, compression, reheating, and re-injection, which may introduce extra energy supply and costs. Therefore, near-term coke reduction is more likely to rely on mature operational optimization and partial chemical substitution, while deeper coke reduction will require coordinated development of low-carbon energy supply, gas recycling, carbon management, and new materials capable of maintaining furnace permeability.

Conclusion and Perspective

This review clarifies coke reduction not simply as a fuel-substitution problem, but as a functional substitution problem. From this perspective, most existing low-carbon strategies mainly reduce the chemical demand for coke, whereas the physical requirement for coke as a load-bearing and permeability-maintaining material remains the key bottleneck for deeper coke reduction. This review has examined several strategies currently being explored for low-carbon BF ironmaking, including alternative carbonaceous materials, H₂-rich reducing gases, recycled CO derived from CO₂ conversion, load-bearing packed-bed materials, and burden structure optimization. Alternative carbon sources, H₂-rich gases, and recycled CO can contribute to lowering fossil carbon input and improving gas utilization. However, they do not fully replace the structural role of coke in the lower furnace and cohesive zone. Candidate load-bearing materials have been proposed to address this issue, but their practical application still faces major challenges in chemical compatibility, slag reactivity, mechanical degradation, cost, and large-scale supply.

Main Conclusions

From this review, several main conclusions can be drawn:

• Alternative carbonaceous materials, including low-rank coal and semi-coke, and carbon-bearing wastes, are primarily effective as partial substitutes for the chemical function of coke, but generally cannot provide equivalent high-temperature structural performance;

• H₂-rich operation can reduce coke demand indirectly by promoting H₂-based indirect reduction and weakening coke solution loss, although its application remains constrained by gas-reaction complexity, reduction kinetics, and lower-furnace thermal balance;

• Load-bearing packed-bed materials are attractive because they aim to replace the structural role of coke directly, but they are still at an exploratory stage owing to unresolved issues of slag attack, abrasion resistance, cost, and large-scale application;

• Burden and furnace structure optimization offers a particularly practical route for lowering coke dependence because it reduces the structural burden placed on coke without requiring a fully new substitute material;

• CO₂-to-CO conversion technologies should be regarded as supportive approaches for carbon recycling and gas re-utilization, rather than complete solutions for coke replacement.

Perspectives

From an industrial perspective, reducing coke consumption is more likely to depend on the combination of several approaches than on any single technology. In this context, partial replacement of coke functions, burden and furnace optimization, and CO₂-to-CO conversion can work together in a complementary way. Under such conditions, the role of coke may gradually shift from being the main reductant and structural support to mainly maintaining permeability, burden support, and process stability. However, the practical value of any low-coke strategy ultimately depends on whether carbon reduction can be achieved without affecting furnace productivity, thermal efficiency, and operational stability.

Conflict of interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.