1. Introduction

In 2011 the European Commission presented “A Roadmap for Moving to a Competitive Low Carbon Economy in 2050” outlining the milestones, among them also the 83–87% CO₂ reduction of the industry sector (1, 2). Primary steel production via the blast furnace/basic oxygen furnace (BF/BOF) route or so-called integrated steel plant, is a predominant and well-established process, contributing 70.8% of the world’s 1807 million tonnes of crude steel production in 2018 (3). The reduction process of the iron ore to crude steel is linked to CO₂ emissions, resulting in 2016 for a total of 7% (160 million tonnes of CO_2eq) of EU-28’s greenhouse gas emissions (4). The energy efficiency potential of a modern integrated steel plant has already been exploited to a great extent through conventional process optimisations. It is, therefore, necessary to transfer steel production to climate-friendly processes through new and innovative approaches. Hydrogen-based direct reduction processes and electrolytic reduction methods are alternatives for the reduction of iron ore, but they require on the one hand huge amounts of renewable energy, for instance for green hydrogen production in water electrolysis, and cause on the other hand significant investment demand as the existing production infrastructure has to be replaced (5). Carbon capture and utilisation (CCU) processes are a second option which are near-term actionable, as they can be added to the existing infrastructure without a significant change in the steel production itself (6). The first step of a CCU process chain is the energy intensive separation of CO₂ from diluted exhaust or process gases. If these gases also contain carbon monoxide, as is the case in the steel industry (Table I), carbon monoxide is avoided of utilisation since the carbon capture processes selectively separate CO₂ (8). The separated CO₂ is then either biologically or catalytically converted to usable products (6).

The process gases in a steel plant, BFG, BOFG and COG, contain high shares of carbon monoxide and CO₂ (Table I). These low-calorific gases are currently utilised in an integrated steel plant as an energy carrier, i.e. in heating processes, and as fuel in the power plant. In Figure 1 the energy flows in an integrated steel plant are depicted. The main part of the BFG (white letter A in Figure 1) is directed to the enrichment process where it is mixed with BOFG (white letter B in Figure 1). The main share of the enriched gas fuels the power plant. COG (white letter C in Figure 1) is mainly used in internal processes and in the power plant as well. The power plant covers almost the total electricity demand of the steel mill. NG (white letter D in Figure 1) is used for heating in downstream processes, like in the hot strip mill, and in the power plant as well. The quantity of the process gases BFG, BOFG and COG covers up to 40% (9) of the steel plant’s energy demand, where the remaining part is provided by electrical energy and fossil fuels, like NG.

The Sankey diagram of Figure 2 indicates the composition of the different byproduct gases and their internal use. It is obvious that a withdrawal of these low-calorific gases has to be compensated by the supply of other energy carriers, either synthetic natural gas (SNG) or external electricity, since the main part of the gases are used in the power plant downstream of the enrichment process. In the enrichment process, the byproduct gases are mixed and buffered in gasometers.

However, due to their high carbon monoxide and CO₂ concentrations, the process gases may be conceived for further use as a carbon source for catalytic conversions. As summarised in the review from Frey et al. (10) the alternative utilisation of steel plant process gases for the production of ammonia, methanol or recovery of its derivatives has already been under examination since the early 1950s. The latest review from Uribe-Soto et al. (11) outlined three alternatives: (a) thermal use of the process gases (state-of-the-art); (b) recovery of the valuable compounds (hydrogen, methane and carbon monoxide) via different separation technologies; and (c) thermochemical synthesis to high-added value products (methanol, dimethyl ether, urea), with the focus on the latter. The consideration of their utilisation as a potential carbon source and coupling it with renewable energy within the context of power-to-X technology, has gained attention in the last years especially in Europe, resulting in various theoretical studies as well as research projects. The largest German steel producer, thyssenkrupp, is leading the “Carbon2Chem” project (12), where different scenarios for the synthesis of methanol (13, 14), ammonia or urea (15), as well as higher alcohols and polymers (16) from BFG, COG and BOFG are being investigated. In an ongoing research project the possibilities of converting BOFG and BFG into methanol and methane are explored under dynamic conditions (17, 18).

All studies referenced above utilise pure CO₂ which is separated in a first step from the process gases, and is subsequently converted with hydrogen in a catalytic synthesis. In this study, the catalytic conversion of BFG and BOFG to methane is investigated without a prior separation of CO₂. The process gases of the steel plant are only pre-cleaned upstream of the catalytic conversion by dust removal (i.e. by a venturi scrubber and a bag house filter) and separation of the sulfur compounds sulfur dioxide, carbonyl sulfide, mercaptans (i.e. in a series of two adsorbers in order to remove the catalyst poisons), see Figure 3. Consequently, the catalytic conversion is carried out with significant shares of nitrogen in the feed gas. Therefore, the following advantages arise:

The aim of the present study is the assessment of different process chains for the direct utilisation of BFG and BOFG in catalytic methanation without a prior CO₂ separation. Since the conversion of CO₂ and carbon monoxide to methane requires renewable hydrogen, the hydrogen supply is ensured by a water electrolysis powered by renewable electricity (P2G plant) as well as by an additional biomass gasification plant (Figure 3). Therefore, a variety of possible implementation scenarios arise, and the following fundamental research questions have to be answered:

A withdrawal of process gases, particularly of the comparatively high calorific COG, would result in a shortage of internal energy supply in the integrated steel plant (Figure 1) which has to be substituted by NG or electric energy sourced externally. Therefore, in order to avoid significant changes of the existing steel production infrastructure, in this study COG was not considered, and BFG as well as BOFG are solely used as a carbon source for a potential utilisation process (Question 1(a)). Furthermore, it has been deliberately decided that the product gas from the methanation, nitrogen diluted SNG, substitute fossil NG currently used in the integrated steel plant, mainly for heating processes. Alternatively, it is used as reducing agent in the blast furnace, for example as substitute for pulverised coal injection (PCI). An injection into the NG grid is not possible since the required specifications are not met (Question 3). The technoeconomic and ecological questions (Questions 1(b), 2 and 4) have been treated by Rosenfeld et al. (19). Supporting laboratory experiments for biomass gasification have been published by Müller et al. (20). The focus of this study is on Questions 5 and 1(b).

2. Integration Scenarios

Figure 3 provides an overview of the possible integration of a P2G plant (Figure 3(b)) as well as a dual fluidised biomass gasification plant (Figure 3(c)) into the integrated steel plant (Figure 3(a) in dashed lines). By integrating a P2G plant, renewable energy is used for the production of hydrogen by water electrolysis and subsequently for the catalytic methanation of the process gases BFG and BOFG. The combination with a dual fluidised biomass gasification (20–22) provides an additional biogenic hydrogen source. The biogenic CO₂ is vented to the atmosphere. Alternatively, it could be stored in a carbon capture and storage (CCS) process resulting in negative CO₂ emissions (bioenergy with carbon capture and storage (BECCS)) which is not further considered here (23). In addition, the oxygen from the water electrolysis has the potential for utilisation in steel production as well as in the biomass gasification process. The produced nitrogen diluted SNG is directly utilised in the steel plant as a substitute for NG in various processes, and PCI in the blast furnace.

To explore the integration potential, a number of different scenarios has been defined and three of them, supported by the experimental results at a laboratory catalytic methanation plant, will be presented in detail in the present work. The three chosen scenarios provide a good overview of the order of magnitude of the required renewable energy, as well as the resulting CO₂ reduction potential.

The scenarios with different integration variations were based on Austria’s biggest steel production sites. The integration of renewable energy by a P2G plant and a biomass gasification plant has been analysed by three extreme value scenarios and three constrained scenarios. The results are reported in (19). The three extreme value scenarios described a maximum utilisation of the process gases, either individually or in combination. The required hydrogen for the methanation was balanced, half from water electrolysis and half from biomass gasification. The constrained scenarios are realistic in the medium term. They are limited by the maximum plant size of the biomass gasification plant (100 MW_th), based on the current biomass fuel availability and already installed gasification capacity in Europe (21). The main cost influencing factor throughout all six scenarios is the energy supply cost, both for electricity and for biomass (19).

The aim of the aforementioned scenarios was the minimisation of, or complete substitution of, the integrated steel plant’s demand for fossil fuels like NG and PCI. The steel plant process gases (BFG and BOFG) were used as carbon source for the methanation. The three scenarios which are the basis for the considerations in this paper are:

For these three scenarios, the required amount of the renewable electricity, biomass as well as the withdrawal amount of the process gasses (BFG or BOFG) has been determined. The main evaluation criteria for all scenarios were set by the CO₂ reduction potential.

3. Fluidised Bed Biomass Gasification

Dual fluidised bed gasification systems consist of two reactors, the gasification reactor (650°C) and the combustion reactor (900°C). In contrast to the conventional systems, the presented system uses the sorption enhanced reforming (SER) process. It allows selective transport of CO₂ between the gasification reactor and the combustion reactor, by the use of calcium oxide as bed material, resulting in a product gas with a high hydrogen (up to 75 vol%) and low CO₂ concentration. The hydrogen rich product gas of the biomass gasification substitute green hydrogen from the electrolysis, and thus reduces the demand of renewable electric power (22, 24). Additionally, when pure oxygen is used instead of air for the combustion (oxySER), an almost pure CO₂ stream can be obtained as an exhaust (flue) gas, suitable as biogenic CO₂ source (Table II). The data given in Table II are based on the gasification of wood chips. A thermal gasification power of 100 MW_th consumes 50,400 kg h^–1 wood chips from Austria as fuel, and produces 28,800 Nm³ h^–1 product gas with the composition according to Table II (21).

4. Experimental Tests

The experimental tests were performed at a laboratory test plant, which consists of three fixed-bed reactors (R1–R3) connected in series with the purpose of achieving a multi-stage fixed-bed methanation. A detailed description of the test plant can be found in Kirchbacher et al. (25) and Medved (26). The conversion of CO₂ and carbon monoxide was investigated for synthetic gas compositions of BFG and BOFG under different flow rates, variation of hydrogen surplus and presence of nitrogen, with the focus on achieving a complete CO_x conversion.

A commercial bulk catalyst with 20 wt% nickel load was used. The operating pressure was set to 4 bar, which coincided with the steel producer’s gas supply system. The reactor load was limited to gas hourly space velocity (GHSV) of 4000 h^–1 (GHSV = V̇_feedgas/V_catalyst) for the synthetic BFG and BOFG gas composition and added hydrogen. The temperature in the reactor was determined by multi-thermocouples (Figure 4).

Seven measuring points in each reactor, five in the catalyst bed and one below and above the catalyst zone, gave an understanding of the axial temperature profile in the catalyst bed. The methanation gas composition for BFG and BOFG for the stoichiometric ratio with hydrogen according to Equations (i) and (ii) is listed in Table III.

(i)

(ii)

5. Results and Discussion

The experimental results obtained from the methanation of BFG and BOFG were used as support for the further analysis of the three selected scenarios. In the following, the experimental results and the scenarios are presented separately in the subsections.

5.1 Methanation

The methanation of process gases is a combination of CO₂ and carbon monoxide conversion according to Equations (i) and (ii).

A suitable parameter for the description of the stoichiometry is the ratio r_{H 2} of molar hydrogen flow and molar flows of CO and CO₂, respectively, in the feed gas given in the Equation (iii):

(iii)

r_{H 2} equals 1 for stoichiometric mixtures, r_{H 2} <1 for sub- and r_{H 2} >1 for over-stoichiometric mixtures, respectively.

Achieved CO_x conversion rates for each reactor (R1–R3), with variation of hydrogen surplus (r_{H 2} = 1; 1.02; 1.04; 1.05) with and without nitrogen for a synthetic BFG and BOFG feed gas compositions, can be seen in Figures 5 and 6. On the right y-axis, the mean reactor temperature represents the average of the measured catalyst bed temperatures, as well as the calculated heating values of the product gas in each reactor. For experiments without nitrogen in the feed gas, marked with -N₂, the H₂:CO_x ratio and the amount of the reactive gas in the experimental series remained the same, meaning that the GHSV was reduced to 3260–3280 h^–1 for BFG and 3680–3690 h^–1 for BOFG due to the absent inert gas flow.

Fig. 5

Methanation of BFG with and without nitrogen and hydrogen-surplus variation

Fig. 6

Methanation of BOFG with and without nitrogen and hydrogen-surplus variation

5.1.1 Methanation of Blast Furnace Gas

Complete CO_x conversions are achieved downstream of the third reactor with an over-stoichiometric ratio of 1.05, both with and without the nitrogen in the feed gas (Figure 5). The mean reactor temperatures are approximately 50°C lower with nitrogen present, which is a result of the additional heat capacity of the inert gas. Despite these lower temperatures, approximately 4% better conversions are reached in R1 for all ratios in the absence of nitrogen. The withdrawal of nitrogen from the feed gas resulted in lower GHSV, consequently prolonging the residence time in the reactor leading to slightly better CO_x conversions. The temperature decrease in R2 and R3 was expected, since the majority of the reactive gas converted in R1, resulting in lower release of the exothermic reaction heat. Therefore, nitrogen in the feed gas only has a significant influence on the heating value of the product gas. In the case of the product gas (R3) with nitrogen the heating values vary from 19.4–19.8 MJ m^–3 (r_{H 2} = 1.05–1), whereas without nitrogen the values almost double (36.0–37.9 MJ m^–3). Although with the higher hydrogen surplus better conversions are achieved, the unconverted hydrogen decreases the heating value of the product gas, due to its lower volumetric heating value compared to methane.

5.1.2 Methanation of Basic Oxygen Furnace Gas

As for BFG, similar test series were conducted for the methanation of BOFG. As shown in Figure 6, on account of lower nitrogen share in the feed gas (8.2%), no noticeable effect on the temperature and consequently conversion can be recognised. Furthermore, a complete CO_x conversion at 4% hydrogen surplus is achieved with or without nitrogen in the feed gas. When comparing the mean reactor temperatures and CO_x conversion in R1 of BOFG with the BFG test series, temperatures are 50–100°C higher and conversions 5–10% lower, respectively. This can be attributed to the higher carbon monoxide share in BOFG, resulting in higher reaction heat release. Therefore, the conversion in the first reactor (R1) is clearly thermodynamically limited. Since its lower share in BOFG compared to BFG, the influence of nitrogen on the heating value of the product gas is lower, and values vary from 27.2–28.8 MJ m^–3 (r_H2 = 1.05–1) with nitrogen and between 34.9–37.7 MJ m^–3 (r_H2 = 1.05–1) without nitrogen.

The product gas composition downstream of the methanation is given in Table IV. A complete conversion is achieved at 5% hydrogen surplus for BFG and at 4% hydrogen surplus for BOFG, where the unconverted hydrogen is a result of its over-stoichiometric addition.

Table IV

Product Gas Composition for the Methanation of BFG and BOFG

	Product gas molar fraction
	CH4	CO2	CO	N2	H2
BFG (rH2 = 1.05)	0.446	0	0	0.434	0.120
BOFG (rH2 = 1.04)	0.679	0	0	0.215	0.106

6. Conclusions

In this study, three different scenarios for the implementation of a P2G plant and a biomass gasification in an integrated steel plant have been investigated. The aim was the quantification of the CO₂ emission reduction potential of steel production, avoiding significant modifications in the existing steel plant infrastructure. Furthermore, a carbon capture step shall be avoided as well, in order to improve the efficiency of the CCU process chain, resulting in a catalytic conversion of BFG and BOFG to methane in the presence of nitrogen.

Basic evaluation of the three chosen scenarios confirmed the possibility of CO₂ emission reductions between 0.81 and 4.6 million tonnes CO_2eq per year without considerable interference with existing steel production. The required plant sizes and the necessary fuel demand (renewable power and biomass, respectively) substantially exceed the current realistic possibilities of a P2G plant (electrolyser power 784–2877 MW_el) as well of a biomass gasification (100–3162 MW_th). Even for the scenarios realistic in the medium term, the amount of required renewable electricity beyond 700 MW_el cannot be provided in the foreseeable future. This underscores the need for new technologies for the production of CO₂-free hydrogen.

Experimental tests have shown that the methanation of BFG and BOFG is technically possible without separating the inert gas nitrogen, and thus saving the energy intensive carbon capture unit with the additional benefit of carbon monoxide utilisation contained in the process gases from the steel production. A complete conversion of CO_x was achieved with a 4–5% hydrogen surplus for both process gases, BFG and BOFG, with and without nitrogen, with three-stage methanation. The lower heating value enrichment of the unrefined BFG (up to 19.8 MJ m^–3) and BOFG (up to 28.8 MJ m^–3) via methanation without the necessity of nitrogen removal as lean product gas showed a utilisation potential in the integrated steel plant as a substitute for NG and PCI.

The first evaluation presented here provides a good overview on the order of magnitude of required renewable energy and biomass for the transition of the integrated steel plant towards renewable gas supply by adding a CCU process chain. Additionally, particularly for the utilisation of the product gases of catalytic methanation within the steel plant, intricate CO₂ separation is not required as has been shown by the experimental investigations.