Quantitative Evaluation of CO₂ Emission Reduction of Active Carbon Recycling Energy System for Ironmaking by Modeling with Aspen Plus

Abstract

The use of the Active Carbon Recycling Energy System in ironmaking (iACRES) has been proposed for reducing CO₂ emissions. To evaluate the performance of iACRES quantitatively, a process flow diagram of a blast furnace model with iACRES was developed using Aspen Plus, a chemical process simulator. The CO₂ emission reduction and exergy analysis was predicted by using the mass and energy balance obtained from the simulation results. iACRES used a solid oxide electrolysis cell (SOEC) with CO₂ capture and separation (CCS), an SOEC without CCS, and a reverse water-gas shift reactor as the CO₂ reduction reactor powered by a high-temperature gas-cooled reactor. iACRES could provide a CO₂ emission reduction of 3–11% by recycling carbon monoxide and hydrogen, whereas the effective exergy ratio decreased in all cases.

1. Introduction

Steel manufacture accounts for 42%¹⁾ of the CO₂ emissions from Japan’s industrial sector. Additionally, secure access to coal is a major problem because Japan is the world’s second-largest coal importer and has very limited domestic coal resources.²⁾ To reduce CO₂ emissions and carbon usage by the steel industry, the introduction of the Active Carbon Recycling Energy System (ACRES) to the ironmaking process (iACRES) has been proposed. A schematic of iACRES is shown in Fig. 1. The core concept of iACRES is based on the reduction of CO₂ to CO by electrochemical or thermochemical processes and recycling the CO to the blast furnace. The exergy used for reducing CO₂ must be generated by a low-carbon primary energy source, such as renewable or nuclear energy to decrease CO₂ emissions by iACRES. High-temperature gas-cooled reactors (HTGRs), a type of nuclear reactor, are one of the most suitable power sources for iACRES, which can generate electric power, high temperatures (~950°C), hydrogen (H₂), and oxygen (O₂) without CO₂ emissions. In the iACRES process, CO₂ is separated from blast furnace gas (BFG) by a conventional amine process and reduced to CO. The CO is recycled to the blast furnace and used for the reduction of iron oxides into pig iron. The trial introduction of ACRES^3,4) and unit operations for CO₂ reduction, such as those using solid oxide electrolysis cells (SOECs)⁵⁾ and reverse water-gas shift reactors (RWGSs), have already been reported. However, quantitative evaluation and discussion of iACRES has not yet been performed. In this work, a blast furnace using iACRES (Fig. 1) was simulated by Aspen Plus, a commercial chemical process simulator (Aspen Technology, Inc.), to evaluate the saving in the coke input and resulting reduction of CO₂ emissions and exergy balance quantitatively. The exergy source is an HTGR and the evaluation boundary is encircled by dashed lines.

Fig. 1.

Schematic illustration and evaluation boundary of iACRES.

2. Blast Furnace Model

In a typical blast furnace, solid materials (e.g., coke, iron ore sinter) are introduced from the top of the furnace and come into contact with up-flowing gases (e.g., hot air, O₂, steam). The process flow diagram (PFD) of the blast furnace simulated by Aspen Plus is shown in Fig. 2. Input streams and output streams are encircled by solid and dashed lines, respectively. The details of the streams are listed in Tables 1(a) and 1(b). Modeling the heat transfer and chemical reactions occurring in a blast furnace as counter-current gas-solid interactions in the PFD accurately reflects actual blast furnace operation. From top to bottom, the blast-furnace shaft is composed of low-, middle-, and high-temperature Gibbs reactors. Additionally, two stoichiometric reactors are placed at the solid outlet of the high-temperature Gibbs reactor to simulate other reactions in the actual blast furnace. One of these reactors directly reduces silicon dioxide (SiO₂) and manganese oxide (MnO) with carbon, and the other simulates the partial oxidation of pulverized coal in the raceways. The adiabatic combustion temperature of the raceway was specified at 2200°C.

Fig. 2.

PFD of blast furnace on Aspen Plus.

Table 1. (a) Stream ID (input). (b) Stream ID (output).

(a)

Stream ID	Stream Composition
I1-IR-OR	Sintered ore
I2-OT-OR	Other ore
I3-AUX-M	Auxiliary material
I4-COKE	Coke
I5-PC	Powdered coal
I6A-AIR	Hot air
I6B-E-O2	Oxygen
I6C-MOI	Moisture

(b)

Stream ID	Stream Composition
O1-H-MET	Hot metal
O2-SLAG	Slag
O3-BFG	BFG
O4-DUST	Dust

The validity of this PFD was verified with reference data based on actual operational data.⁶⁾ The input conditions for the reference data is listed in Table 2. Tables 3(a) and 3(b) show the simulation results for the output material and heat streams, respectively. Compared with the reference data, the error in the output material and heat streams was within 5%.

Table 2. Input material for the calculation.

Material	Weight or Volume [kg/t of hot metal (THM)or N m3/THM]
Sintered ore	1254
Other ore	339
Auxiliary material	15
Coke	385
Powdered coal	112
Hot air	1045
Oxygen	25
Moisture	27

Table 3. (a) Comparison of the simulation result with material balance reference data. (b) Comparison of simulation results with enthalpy balance reference values.

(a)

Stream	Chemical species	This work [kg/THM]	Reference [kg/THM]
Hot metal	C	45.04	47.00
	Fe	949.29	945.60
	Si	5.30	5.30
	Mn	2.50	2.50
Sum		1002.14	1000.40
Slag	SiO2	93.31	97.72
	Al2O3	37.84	35.88
	CaO	126.74	116.72
	MgO	18.72	14.64
	MnO	1.38	1.81
Sum		279.99	266.77
BFG	H2	6.17	6.50
	N2	1032.20	1004.57
	CO	433.78	452.85
	CO2	708.99	674.33
	H2O	32.94	33.05
	CH4	1.06	0.00
Sum		2215.14	2171.30

(b)

Name	Sum [MJ/THM]		Heat and Pressure [MJ/THM]		Chemical [MJ/THM]
	This work	Reference	This work	Reference	This work	Reference
Hot metal	9620	9820	940	1118	8680	8702
Slag	898	890	439	437	459	453
BFG	5355	5277	–32	0	5387	5277
Electricity	113	112	–	–	–	–
Heat loss	900	744	900	744	–	–
Other loss	307	315	307	315	–	–
Sum	15549	15727	909	1118	14527	14497

This strong agreement between the simulation results and reference data confirmed the validity of the model. The scale of blast furnace used in this simulation is shown Table 4; the size of the associated iACRES equipment was scaled to match the blast furnace.

Table 4. Assumed scale of the blast furnace.

Scale indexes	Value
Volume of blast furnace	5000 m3
Iron trapping ratio	2.1 t/(dam3)
Product of hot metal	3833 kt/y

3. iACRES Model

3.1. CO₂ Separation

In iACRES, CO is circulated to the blast furnace as a substitute for additional carbon input. The source of CO is CO₂ in BFG. Because BFG has an enormous flow rate and contains large quantities of N₂ and CO, CO₂ reduction requires a large scale and a large amount of additional energy consumption. However, the CO contained in BFG can be used directly as a gaseous reductant. In this work, we considered that 30 to 40% of BFG is separated at the top of blast furnace and is recirculated to the blast furnace via ACRES either directly or through separating CO₂, as shown in Fig. 3. The conventional mono-ethanol amine process is used for CO₂ capture and separation (CCS). A simple component splitter block was used for the CCS model in the ASPEN Plus and an additional heat of 4.0 GJ/t-CO₂ at 140°C is included in the total heat input. The CO₂ recovery rate was assumed to be 90%.

Fig. 3.

Schematic of BFG separation and CCS model.

SOECs can be operated without CCS in the model, whereas N₂ and CO from BFG remain in the recirculated gas stream and can disturb the electrolyzation of CO₂. However, An RWGS without CCS was not evaluated because circulating CO and H₂O strongly disrupt the RWGS reaction.

3.2. SOEC Model

Figure 4 shows a schematic of the SOEC system. The CO₂ introduced into the cathode is reduced to CO and O^2– by electrolysis, and the O^2– conducted through the electrolyte is oxidized to O₂. In this simulation, the operation temperature of the SOEC is assumed to remain constant at 800°C and the conversion rate of CO₂ is assumed to be 60% or 70%. The theoretical electrolysis voltage (E_t at 800°C) was calculated from the Gibbs free energy change (ΔG = 189 kJ/mol) in the electrolysis reaction of CO₂ → CO + 0.5O₂. Assuming adiabatic conditions, the electrolysis of 1 mol of CO₂ requires enthalpy of 282 kJ (800°C); part of the required enthalpy must be supplied as electric energy, equivalent to ΔG = 189 kJ, and the rest of the enthalpy can be supplied as heat of TΔS = 93 kJ. However, the applied voltage, E, is usually higher than the theoretical electrolysis voltage, E_t, due to overvoltage (Fig. 5). Current density, i, for electrolysis is given by the amount of CO₂ to be electrolyzed (mol/s) according to Faraday’s law and current leakage. The applied voltage, E, and the overvoltage, η, are determined from the experimental i-V performance curve obtained by Ishihara and co-workers⁷⁾ as shown in Fig. 5. Figure 6 shows the calculation method for the required electric energy and heat input. Electric energy, E·i, is consumed during the electrolysis of CO₂ at the rate of CO₂ electrolyzation in an SOEC. Part of the applied electric power, equivalent to η·i, is converted to heat and makes up part of the required heat input; therefore, the heat input required from external sources decreases. However, the exergy of electric power is destroyed by converting it to heat at a low temperature. Figure 7 shows the breakdown of the total heat input, TΔS, versus current density. In this calculation, the scale of one SOEC was set by assuming a 10 kW solid oxide fuel cell.⁸⁾ The total heat input TΔS increases in negative in proportion to current density in accordance with Faraday’s law. Concurrently, the internal heat generation caused by overvoltage, η, also increases, and the remainder has a peak value at i = 0.21 A/cm². To minimize the exergy loss, the operating current density, i, was set at 0.21 A/cm² to maximize the direct heat input. In this calculation, the current density was set by setting number of SOEC modules.

Fig. 4.

Schematic of SOEC and electrochemical reactions.

Fig. 5.

Experimental i-V performance of SOEC.⁸⁾

Fig. 6.

Schematic illustration of required enthalpy change, and input of electric power and heat in SOEC process.

Fig. 7.

Heat flow calculation of current density.

3.3. RWGS Model

The catalytic reduction of CO₂ by H₂ in the RWGS is expressed in Eq. (1), and the unit operation was modeled by an equilibrium reactor in Aspen plus.   

$$\mathrm{C}{\mathrm{O}}_2+{\mathrm{H}}_2\to \mathrm{C}\mathrm{O}+{\mathrm{H}}_2\mathrm{O},\hspace{1em}\mathrm{\Delta }H{°}_{298}=+41.2\mathrm{k}\mathrm{J}/\mathrm{m}\mathrm{o}\mathrm{l}$$(1)

H₂ was assumed to be produced in iodine-sulfur (IS) process, a thermochemical water splitting process. Exergy source of the H₂ production was HTGRs, hydrogen production thermal efficiency was 37.0% as a result of a process flow simulation by Aspen Plus. The relationship between reaction temperature and conversion of CO₂ at molar ratios of H₂/CO₂ of 1, 2, 2.55, and 3 is shown in Fig. 8. Considering the maximum hydrogen production from the HTGR, the maximum H₂/CO₂ ratio is 2.55. Higher H₂ concentrations and temperatures produce higher yields of CO. However, higher temperatures may induce catalyst degradation. H₂O is also produced in Eq. (1) and this can increase the mole fraction of H₂O in circulated gas. Excess H₂O affects the coal gasification in the blast furnace, as shown in Eq. (2), which may cause a temperature drop due to an endothermic reaction. To prevent the temperature drop, the reduced gas was cooled to 40°C to condense the steam before circulating to the blast furnace. Considering the appropriate temperature for the Ni-based catalyst used in a typical RWGS,⁹⁾ the operating temperature of the RWGS was set at 640°C, so as to induce 60% CO₂ conversion, and a H₂/CO₂ ratio of 2.55.   

$$\mathrm{C}+{\mathrm{H}}_2\mathrm{O}\to \mathrm{C}\mathrm{O}+{\mathrm{H}}_2,\hspace{1em}\mathrm{\Delta }H{°}_{298}=+132\mathrm{k}\mathrm{J}/\mathrm{m}\mathrm{o}\mathrm{l}$$(2)

Fig. 8.

Equilibrium CO₂ conversion vs temperature at various H₂/CO₂ ratios.

3.4. Operating Conditions

Applying iACRES to the blast furnace model changes the operating conditions from those of the original system in several ways. In particular, the carbon content in hot metal, heat flux ratio, and the temperature of the raceway were important operating conditions. The carbon content in hot metal affects the melting point. The heat flux ratio and raceway temperature affect the temperature distribution in the blast furnace, which may result in incomplete iron ore reduction. The ratio of CO and CO₂ in BFG is changed by the recirculated gas and must be calculated by using the non-isothermal rate-based model. In this simulation, the operating temperature of the low-temperature block was adjusted by applying the top gas composition of the blast furnace calculated by other one-dimensional simulation code to the gas output stream of the block. Figure 9 shows the CO/(CO+CO₂) ratio of the BFG outlet, where z is calculated by one-dimensional simulation code. z is approximated by   

$$z=(0.0056x-0.8747)y+47.4$$(3)

where x and y are the CO/(CO+CO₂) ratio and CO+CO₂ molar flow of the circulation gas, respectively.

Fig. 9.

Relationship between CO+CO₂ input and the CO/(CO+CO₂) ratio of BFG.

Table 5 summarizes the operating conditions and corresponding adjusted variables. In this simulation, all the exergy inputs, such as the electric power and heat for SOEC or hydrogen and heat for RWGS are supplied by 600 MW_th HTGRs. The number of reactors and generation efficiencies are summarized in Table 6.

Table 5. Operating conditions and adjusted variables.

Conditions	Value	Adjusted variables
Carbon content [kg/THM]	45.04	Coke input
Heat flux ratio	0.75	Flow rate of enriched O2 and　hot air
Temperature of raceway [°C]	2196	Temperature of circulated gas

Table 6. Summary of efficiencies and exergy output of HTGR.

	SOEC	RWGS
Reactor output	600 MWth
Number of reactors	0.5	2.0
H2 production efficiency ηH2 [%]		37.0
Power generation efficiency ηelec [%]	48.1	47.0
H2 exergy [MJ/THM]		2905
Electricity exergy [MJ/THM]	1073	22
Heat exergy [MJ/THM]	72.5	99.4

4. Results and Discussion

4.1. CO₂ Emission Reduction

Figures 10(a)–10(d) depicts carbon flows for the evaluation boundary shown in Fig. 1, for the original system, iACRES using SOEC with CCS, iACRES using SOEC without CCS, and iACRES using RWGS with CCS, respectively The mass flows of carbon in a year (kt/y) for each flow and block calculated based on a 3833 kt/y hot-metal production are shown in Table 4. Table 7 compares the CO₂ emission reduction of iACRES against the original blast furnace system. The base case used the following operating conditions: 30% BFG circulation and 60% CO₂ conversion by SOEC/RWGS. The effects of these operating conditions on CO₂ emission are discussed in Table 8.

Fig. 10.

(a) Carbon flow (original system) (unit: kt/y). (b) Carbon flow (iACRES using SOEC with CCS) (unit: kt/y). (c) Carbon flow (iACRES using SOEC without CCS) (unit: kt/y). (d) Carbon flow (iACRES using RWGS with CCS) (unit: kt/y).

Table 7. Comparison of results of CO₂ emission reduction with ACRES.

		CO2 emission reduction [%]
		SOEC	RWGS
CCS	w/	3.43	8.34
	w/o	11.4

Table 8. Effect of CO₂ conversion and BFG circulation on CO₂ emission reduction with SOEC with iACRES.

		BFG circulation [%]
		30	40
CO2 conversion [%]	60	3.43	4.99
	70	4.11

When iACRES is used, the circulated CO is used to reduce iron oxides, which decreases the coke input and thus CO₂ emissions from the boundary shown in Fig. 1. The original system shown in Fig. 10(a) inputs a 1288 kt/y carbon flow as coke into the blast furnace in comparison with 1232, 1102, and 1155 kt/y shown in Figs. 10(b)–10(d), respectively. The CO₂ emission reduction rate of RWGS was higher than that of SOEC because the H₂, which is unused for reducing CO₂ in the RWGS, was recirculated into the blast furnace and used to reduce iron oxides. In the carbon flow in Fig. 10(d), the circulated gas and the BFG flow into the stove are less than those in Fig. 10(b). However, the hydrogen flow, which can act both as a reductant in the blast furnace and a heat source in the stove, is hidden in this carbon flow. The CO₂ reduction rate of SOEC without CCS was higher than that of SOEC with CCS, because both CO and the circulated H₂ were used as a reductant for iron oxides. The circulated gas shown in Fig. 10(c) inputs a 538 kt/y carbon flow, which is mainly contained as carbon monoxide, compared with 244 kt/y shown in Fig. 10(b). However, because the nitrogen in the hot gas is decreased to adjust the heat flux ratio, the heat duty and resulting CO₂ emission from the stove is almost comparable in both cases.

Table 8 shows the effects of the conversion and BFG circulation on the CO₂ emission reduction by using SOEC as the CO₂ reduction reactor. The CO₂ emission decreased with an increase in both BFG circulation and CO₂ conversion because the mole fractions of gaseous reducing agents, such as CO and H₂, increased. Meanwhile, an increase in BFG circulation boosted the CO₂ emission reduction, whereas the increase in the ratio of CO₂ conversion also increased the emission reduction by around 1.2 times.

4.2. Exergy

Figure 11 shows the total exergy input and total effective exergy output for each iACRES setup compared with the original system. Figure 12 shows the effective exergy ratio in each iACRES setup, which is defined as the ratio of the total effective exergy output to the total exergy input. In the iACRES setups using an SOEC and RWGS with CCS, the exergy inputs increased, whereas the exergy outputs were very similar. This indicated that the additional exergy input was more than the reduction in the carbon input arising from CCS. For iACRES using an SOEC and without CCS, the exergy input decreased, in contrast to the system with CCS. A combination of input and output effects resulted in a decrease in effective exergy ratio compared with the original system. In particular, iACRES without CCS had a lower effective exergy ratio than iACRES with CCS because the unused BFG contained a large amount of CO₂ and H₂O, which have no chemical exergy.

Fig. 11.

Comparison of exergy input and effective exergy output in various iACRES with those of the original system.

Fig. 12.

Comparison of effective exergy ratio in various iACRES with the original system.

For iACRES using an RWGS, exergy inputs increased significantly because of the excess hydrogen. The production efficiency of H₂ was lower than the generation efficiency, as shown in Table 6, which means that a greater exergy input is required in an RWGS than that in an SOEC. However, the exergy outputs were almost the same. This resulted in a lower effective exergy ratio in iACRES using an RWGS than that using an SOEC because BFG contained a large amount of CO₂ and H₂O, which have a low chemical exergy.

4.3. Blast Furnace Temperature

Because the temperature drop in the reactor block blast furnace has the potential to make operation impossible, the temperature change in the blast furnace as a result of using iACRES was evaluated. As shown in Fig. 2, the blast furnace model consists of three pairs of adiabatic Gibbs reactors and flash blocks. The high-temperature block is used as an index for temperature change because the highest temperature block is affected significantly by changes in the input gas conditions. Table 9 shows the calculated temperature of the high-temperature block in each case. Table 10 shows the sensitivity analysis of the temperature in the SOEC with CCS model. In the original system, the high-temperature block is at 1160°C and the temperatures in all cases dropped by up to 98°C.

Table 9. Calculated temperature of HIGH-TEMP block (°C).

		Reduction
		SOEC	RWGS
CCS	w/	1097	1115
	w/o	1082

Table 10. Calculated temperature of HIGH-TEMP block in SOEC with CCS model (°C).

		Circulation (%)
		30	40
Conversion rate (%)	60	1097	1062
	70	1122

Using iACRES changes temperature in the blast furnace because of the carbon solution reaction and the reduction of iron oxides by hydrogen.

The unreduced CO₂ circulated into the blast furnace can promote the carbon solution reaction in Eq. (4).   

$$\mathrm{C}+\mathrm{C}{\mathrm{O}}_2\to 2\mathrm{C}\mathrm{O}\hspace{1em}\mathrm{\Delta }H{°}_{298}=+97.5\mathrm{k}\mathrm{J}/\mathrm{m}\mathrm{o}\mathrm{l}$$(4)

The temperature drop in the blast furnace increased with the increase in BFG circulation and thus in CO₂ flow rate. For the same reason, the high conversion rate suppresses the temperature drop in Table 10.

The hydrogen circulated to the blast furnace reduces the iron oxides, as shown in Eq. (5). However, it also suppresses the direct reduction of iron oxide by carbon as shown in Eq. (6), which has a much larger endothermic enthalpy change than that of the hydrogen reduction. Therefore, the temperature drop in the RWGS is less than that in the SOEC.   

$$\mathrm{F}\mathrm{e}\mathrm{O}+{\mathrm{H}}_2\to \mathrm{F}\mathrm{e}+{\mathrm{H}}_2\mathrm{O}\hspace{1em}\mathrm{\Delta }H{°}_{298}=+27.2\mathrm{k}\mathrm{J}/\mathrm{m}\mathrm{o}\mathrm{l}$$(5)

  

$$\mathrm{F}\mathrm{e}\mathrm{O}+\mathrm{C}\to \mathrm{F}\mathrm{e}+\mathrm{C}\mathrm{O}\hspace{1em}\mathrm{\Delta }H{°}_{298}=+158\mathrm{k}\mathrm{J}/\mathrm{m}\mathrm{o}\mathrm{l}$$(6)

This simulation model contained only equilibrium reactors and thus cannot give sufficient data to discuss the operational feasibility of the modified blast furnace. Detailed analysis using rate-based blast furnace models is required in future work.

To mitigate the negative effect of the temperature drop, additional countermeasures are required. For example, increasing the hot air temperature is a quick way to compensate for the temperature drop, although an additional heat source with an exergy higher than that at 1160°C is necessary. The combination of an HTGR and coal firing is one possible heat source. CO₂ separation from the reduced gas before recirculation is another possible countermeasure.

5. Conclusions

A PFD of iACRES including the blast furnace model was developed and its effect was quantitatively evaluated in terms of CO₂ emission reduction resulting from the saving of coke input and exergy analysis. The following results were obtained.

(1) Compared with the original blast furnace, iACRES can reduce CO₂ emission by 3–11%, and decrease the effective exergy ratio by 1–7%.

(2) The CO₂ separation (CCS) from BFG before the CO₂ reduction reactor contributes to suppressing the decrease in effective exergy ratio. However, because BFG contains H₂ and CO, which function as reductants, iACRES without CCS can also reduce the coke input.

(3) In iACRES without CCS or with the CO₂ reduction reactor operating at a low CO₂ conversion, the circulated CO₂ causes a temperature drop in the blast furnace of up to 100°C. Detailed analysis using rate-based blast furnace models is necessary to assess whether the system is operationally feasible.

References

• 1)  Greenhouse Gas Inventory Office of Japan: The GHGs Emission Data of Japan, NIES, Ibaraki, Japan, (2013).
• 2)  Ministry of Economy Trade and Industry (METI): The White Paper for Energy in 2013 FY, METI, Tokyo, (2013).
• 3)   Y.  Kato: ISIJ Int., 50 (2010), 181.
• 4)   Y.  Kato,  T.  Obara,  I.  Yamanaka,  S.  Mori,  A. L.  Dipu,  J.  Ryu,  Y.  Ujisawa and  M.  Suzuki: Prog. Nucl. Energ., 53 (2011), 1017.
• 5)   T.  Ishihara: J. Ceram. Soc. Jpn., 122 (2014), 122.
• 6)  Research and Study of Future Technology for Reducing CO₂ Emission in Iron and Steel Manufacturing Process (Tekkou prosesu ni okeru jisedai CO₂ sakugen gijutu tyousa kentou), ISIJ, Tokyo, (2007).
• 7)   S.  Wang,  A.  Inoishi,  J.  Hong,  Y.  Ju,  H.  Hagiwara,  S.  Ida and  T.  Ishihara: J. Mater. Chem. A, 40 (2013), 12455.
• 8)  Report of Developing Solid Oxide Fuel Cell in the 18 to 19 Year of Heisei Ara. (Kotaisankabustu gata nenryo denchi sisutemu gijutu kaihatu heisei 18 nendo ~ heisei 19 nendo houkokusyo), NEDO, Kanagawa, Japan, (2009).
• 9)   L.  Wang,  S.  Zhang and  Y.  Liu: J. Rare Earth., 26 (2008), 66.