Introduction

Despite the dominance of the blast furnace ironmaking process, increasing attention is being paid to the development of new technologies with lower energy consumption and CO2 emissions. It should also require much less capital investment than the blast furnace/coke oven combination. At the University of Utah, an alternative technology to meet these demands, called the flash ironmaking technology (FIT), has been under development [1–7]. This novel technology eliminates the highly problematic cokemaking and pelletization/sintering steps by directly utilizing iron ore concentrate, which is currently produced in large quantities in North America and elsewhere. In this article, recent advances on this work are discussed.

The FIT is an innovative process that uses iron ore concentrate directly without further treatment. The fineness of the concentrate particles allows a very rapid reaction rate, thus requiring residence times measured in seconds instead of the minutes and hours it takes to reduce pellets and even iron ore fines.

Other processes for the gaseous reduction of iron oxide can be grouped into two broad types: (1) shaft furnaces (Midrex [8] and Energiron [9]), and (2) fluidized-bed reactors (FINMET and the earlier FIOR [10], CIRCORED Process [11], and SPIREX [12]). Shaft furnaces require pellets or sinter, which are of 10–12-mm size and require minutes or even hours to be fully reduced. The fluid-bed processes use “iron ore fines,” i.e., particles in the range of +0.1 mm to −10 mm. These processes are slow because the larger pellets and particles react slowly, and the process cannot be operated at high temperatures, under which they suffer the problems of sticking and fusion of particles. Processes that can replace the blast furnace must be sufficiently intensive to meet the large production rates required for economic competitiveness. The FIT reduces iron ore concentrates of <100-μm sizes, which are smaller than fines by one or two orders of magnitude. Historically, fluidized-bed processes using iron ore fines have been less than successful. Compared with concentrate particles, even ore ‘fines’ used in a fluidized-bed process take much longer time to reduce than the time available in a typical flash reduction process.

The FIT removes many of the limitations associated with the other processes discussed above. Specifically, (1) it uses iron oxide concentrates directly without the need for pelletization or sintering; (2) coke ovens are not required; (3) high temperatures can be used because there will be little particle sticking or fusion problems (and thus, the process can be operated at high temperatures); (4) it is possible to produce either solid or molten iron; and (5) the raw materials can be fed continuously. The process also has the possibility of direct steelmaking in a single unit, as depicted in Fig. 1.

Fig. 1
figure 1

A schematic diagram of a possible direct steelmaking process based on the Flash Ironmaking Technology (FIT) [61]

Full size image

In this paper, we will describe major progresses made so far on the Flash Ironmaking research, especially the kinetics of the reduction process, the chemistry of the slag, and the economic and environmental aspects of the process.

Reduction Kinetics

Previous research on the gaseous reduction of iron oxide has typically involved samples larger than concentrate particles, such as ore fines and pellets as well as temperature ranges substantially lower than in the new flash ironmaking process. Ray and Kundu [13] reduced hematite in CO–N2 mixtures in the temperature range 1073–1273 K (800–1000 °C) and determined an activation energy of 90 kJ/mol. Edstrom [14] studied the reduction of natural hematite crystals by hydrogen below 1273 K (1000 °C) and found that the hematite crystal swelled at first, but shrunk during the later stages when reduced by hydrogen. Hematite particles of 105–140-µm sizes were reduced by hydrogen in a silica-tube fluidized-bed reactor by Srinivasan and Sheasby [15] in the temperature range of 923–1173 K (650–900 °C). They found that a pore-free magnetite layer was formed adjacent to the hematite phase, but the magnetite layer developed porosity with further reduction. Piotrowski et al. [16] reduced hematite particles of 91-µm average size to wustite by carbon monoxide at 973–1173 K (700–900 °C). The nucleation-and-growth kinetic model was applied to the initial stages of the reduction process, which gradually shifted to diffusion control as reduction proceeded. They determined activation energy to be 58 kJ/mol. Sturn et al. [17] found that the phase boundary reaction was the most likely rate-controlling step for the reduction of hematite particle of 177-µm diameter by hydrogen in the temperature range 673–803 K (400–530 °C). Fruehan et al. [18] investigated the reduction of hematite particles of 180–250-µm size by hydrogen in the temperature range 873–1273 K (600–1000 °C) and found that a shell of dense iron formed around the wustite grains that significantly slowed the reduction rate, limited by the diffusion of oxygen atoms through solid iron when the grains were larger than 2 µm. Recently, Monazam et al. [19] studied the reduction kinetics of hematite (average size of 200 µm) to wustite with hydrogen at temperatures of 973–1223 K (700–950 °C). They reported that the reduction process can be divided into two stages: an induction period, and a nucleation-and-growth period. The induction period became shorter and the entire reduction process could be considered as nucleation/growth period at higher temperatures of 1123–1223 K (850–950 °C) with an apparent activation energy of 64 kJ/mol.

There has been previous work on the reduction of fine iron ore particles; but most of it has been performed at temperatures lower than 1273 K (1000 °C), because the existing gaseous reduction processes operate at lower temperatures than in the new flash ironmaking process, which is expected to be operated at a temperature above 1500 K.

There had been little work on the gaseous reduction of iron oxide concentrates under the conditions of the new flash ironmaking technology [1, 20, 21]. In conjunction with the development of the new process, Sohn and coworkers [1, 20, 21] have investigated the gaseous reduction of magnetite concentrate particles. Their work included the measurement of the intrinsic kinetics of the hydrogen reduction of magnetite concentrate particles in the temperature range of 1423–1673 K (1150–1400 °C) [1]. Most importantly, they have established that magnetite concentrate can be reduced to a degree higher than 90 pct within several seconds, which presents sufficiently rapid kinetics for a flash reduction process.

As an integral part of the development of this novel process, Sohn and coworkers [1, 22, 23] have determined detailed kinetics of the gaseous reduction of magnetite concentrate particles, aimed at generating a database to be used for the design of a flash ironmaking reactor; before their work, there had been little research reported on this subject under the conditions of the FIT.

A number of kinetic equations were tested on the obtained experimental data to determine the best-fitting equation: the power law, contraction models, kinetics-order models, and the nucleation-and-growth model. From a mechanistic point of view, it is not uncommon in the gaseous reduction of very fine metal oxide particles to follow the nucleation-and-growth process. Supported by the goodness of the fit as well as the SEM micrographs shown in Fig. 2, the nucleation-and-growth model best represented the kinetics of hematite concentrate reduction by hydrogen. In the nucleation-and-growth kinetics, hydrogen is adsorbed on active sites on the surface and then reacts with oxygen and produces Fe nuclei that grow with time, as shown in Fig. 3. For small particles, the period of the formation and the growth of nuclei occupy essentially the entire conversion range. The reduction of hematite to iron in the temperature range of this work proceeds through the formation of Fe3O4 and FeO with changes in reactivity. However, for small particles, it is extremely difficult to follow the irregular evolution of morphology involving Fe3O4 and FeO formation and the overlapping regions of different reduction extents. Furthermore, the reduction of Fe2O3 to FeO accounts for only the first 33 % of conversion, whereas the main interest in this work and the new ironmaking technology is the attainment of much higher degrees of reduction. Thus, the following global nucleation-and-growth rate equation for the overall reduction process is appropriate:

$$ [ - Ln(1 - X)]^{1/n} = [k \cdot f_{p} (p{\text{H}}_{2} ,p{\text{H}}_{2} {\text{O}})] \cdot t = k_{\text{app}} \cdot t $$
(1)

where X is the fractional reduction degree; t is the residence time; n is the Avrami parameter [24–26]; k is the rate constant; pH2 and pH2O are the partial pressures of hydrogen and water vapor, respectively; fp(pH2, pH2O) is the concentration driving force; and k app is the apparent rate constant. Sohn et al. [1, 20, 22, 23] measured the kinetics of the reduction of magnetite or hematite concentrate particles by hydrogen or carbon monoxide in the temperature range of 1423–1673 K (1150–1400 °C). They obtained the following rate expression for the hydrogen reduction of magnetite concentrate particle in the temperature range of 1423–1623 K (1150–1350 °C) given by:

$$ - Ln(1 - X) = 1.23 \times 10^{7} \times e^{{\frac{ - 196,000}{\text{RT}}}} \times \left( {p{\text{H}}_{2} - \frac{{p{\text{H}}_{2} {\text{O}}}}{K}} \right)_{\text{lm}} \times t $$
(2)

where R is 8.314 J mol−1 K−1, p is in atm, t is in seconds, and K is the equilibrium constant for the reaction: Fe0.947O + H2 = 0.947 Fe + H2O. It was shown previously [20] that a logarithmic mean (lm) between the input and the off-gas values appropriately represents the average hydrogen driving force for first–order dependence with respect to pH2, and an arithmetic mean (am) is the appropriate average driving force for a half-order reaction.

Fig. 2
figure 2

Coexistence of various iron oxide phases along with iron as evidenced by a SEM micrographs of reduced sample particles (X = 0.14); and b XRD of samples with different reduction degrees [23]

Full size image
Fig. 3
figure 3

SEM micrographs of hematite samples with different reduction degrees in the temperature range of 1473–1623 K (1200–1350 °C) with pH2 ranging from 0.1 to 0.4 atm [23]

Full size image

Although the FIT is being developed for magnetite concentrate, it is expected to be equally applicable to hematite concentrate. Thus, Sohn and coworkers [23] studied the hydrogen reduction kinetics of hematite concentrate particles of average size 21 µm in the temperature range of 1423–1623 K (1150–1400 °C). The results clearly indicated that the hematite concentrate can also be reduced to greater than 90 pct degree by hydrogen in the several seconds of residence time typically available in a flash reactor. In their work [23], results at various temperatures and partial pressures of hydrogen showed that the nucleation-and-growth rate equation with an Avrami parameter of 1 well describes the kinetics of hematite reduction. The reduction rate has a 1st-order dependence on the partial pressure of hydrogen. The activation energy of hydrogen reduction of hematite concentrate is 214 kJ/mol [23]. The following rate equation was obtained, which satisfactorily represent the reduction kinetics of hematite particles:

$$ - Ln(1 - X) = 4.41 \times 10^{7} \times e^{{\frac{ - 214,000}{\text{RT}}}} \times \left( {p{\text{H}}_{2} - \frac{{p{\text{H}}_{2} {\text{O}}}}{K}} \right)_{lm} \times t $$
(3)

where R is 8.314 J/mol K, T is in K, p is in atm, and t is in seconds.

Compared with the activation energy for the hydrogen reduction of magnetite concentrate of 196 kJ/mol in the temperature range 1423–1673 K (1150–1400 °C), the reduction of hematite concentrate has a similar activation energy of 214 kJ/mol in the similar temperature range.

Hematite reduction kinetics by CO + H2 mixtures was also investigated to represent the case of using natural gas or coal gas as the reductant/fuel in the flash ironmaking process [27]. The kinetics in this case is complicated due to the simultaneous reduction by the two reductants. It was found, however, that within a few seconds of residence time, a reduction degree of over 90 pct was achieved using CO + H2 at temperatures above 1573 K (1300 °C), as shown in Fig. 4. It is also clear that adding CO to the H2 boosts the reduction kinetics as compared with the reduction by a single gas H2 or CO. Increasing CO partial pressure from 0.05 to 0.2 atm while holding hydrogen partial pressure at 0.2 atm did not affect the reduction kinetics of hematite concentrate, as Fig. 4 shows. The measurement of the kinetics of the flash reduction is continuing in order to determine the rate equation for the reduction of magnetite and hematite concentrate with mixtures of CO and H2 to be used in designing a flash reactor that will run on natural gas.

Fig. 4
figure 4

Reduction degree comparisons between single reductant gas (H2) and CO + H2 mixtures at 1573 K (1300 °C) (adapted from [27])

Full size image

Meanwhile, Sohn et al. [28, 29] studied the kinetics of the re-oxidation of iron particles produced by the FIT in various oxidizing gas mixtures. As the gas–particle mixture cools down in the lower part of the flash reactor, the re-oxidation of iron could take place because of the decreasing equilibrium constant and the high reactivity of the freshly reduced fine iron particles. The last stage of hydrogen reduction of iron oxide, i.e., the reduction of wustite, is significantly limited by equilibrium. The effects of temperature [823–973 K (550–700 °C)] and H2O partial pressure (34–86 kPa) on the re-oxidation rate were examined. The nucleation-and-growth model was shown to best describe the re-oxidation kinetics. Pressure dependence was of first order with respect to water vapor, and the activation energy was 146 kJ/mol. A complete rate equation that adequately represents the experimental data was developed [29]. They also studied the oxidation in pure CO2 gas in the temperature range of 873–1073 K (600–800 °C). Their findings indicated that within the several seconds of residence time typically available in a flash reduction process the re-oxidation degree of iron particles in water vapor should be <0.24 % in the temperature range of 823–973 K (550–700 °C) and the results implied that the oxidation will be negligible in the flash reduction process where CO2 from partial combustion of natural gas with oxygen accounts for less than 10 % in the gas mixture. In the collector where product iron particles may be kept for up to 1 h at around 673 K (400 °C), the expected re-oxidation degree would be around 0.02 % [28–30]. This indicates that re-oxidation within the reactor due to the shift in equilibrium as temperature decreases toward the bottom of the reactor is of no concern.

The pyrophoricity of direct reduced iron when it comes into contact with air is of concern during storage or transportation. Thus, oxidation of flash-reduced iron in O2–N2 gas mixtures was investigated to determine the effects of temperature [673–873 K (400–600 °C)] and O2 partial pressure (3.4–18 kPa). The rate data were analyzed based on the nucleation-and-growth model, which resulted a pressure dependence of first order with respect to oxygen and an activation energy of 14.4 kJ/mol [28]. Most importantly, it was concluded that flash-reduced iron is much less vulnerable to oxidation than conventional direct reduced iron particles. This is because the flash ironmaking process uses higher reduction temperatures, leading to a lower specific surface area of the product iron. This can be clearly seen from the micrographs presented in Fig. 5, which shows iron oxide reduced at a lower temperature at which conventional DRI is produced (Fig. 5a) compared with iron produced at a flash ironmaking temperature (Fig. 5b). This work determined that at temperatures lower than 573 K (300 °C), the oxidation of flash-reduced iron is not of great concern [28]. In addition, Sohn and coworkers have developed a relationship between the fractional oxygen content (F) and time [29]:

$$ \left[ {Ln\left( {\frac{{1 - F_{0} }}{1 - F}} \right)} \right]^{1/1.26} = 941.5 \cdot \exp \left( { - \frac{146000}{\text{RT}}} \right) \cdot \left( {p_{{{\text{H}}_{2} {\text{O}}}} - p_{{{\text{H}}_{2} }} /K} \right) \cdot t $$
(4)

where F 0 is the initial fractional oxygen content; K is the equilibrium constant of oxidation of iron to wustite [or magnetite less than 833 K (560 °C)] by H2O(g); T is in K; p is the pressure in kPa, and t is the time in seconds. Also, they developed a rate equation for the re-oxidation of flash-reduced iron particles in O2–N2 gas mixture as follows [28]:

$$ Ln\left( {\frac{{1 - F_{0} }}{1 - F}} \right) = 0.006\;\exp \left( { - \frac{14,400}{\text{RT}}} \right) \cdot p_{{{\text{O}}_{2} }} \cdot t $$
(5)

where F is the fractional oxygen content; R is the gas constant in J/mol K; \( p_{{{\text{O}}_{2} }} \) is the partial pressure of oxygen in kPa; and t is the time in seconds.

Fig. 5
figure 5

Comparison of microstructures of different iron particles: a H2-reduced iron at 1073 K (800 °C); b flash-reduced iron produced at 1623 K (1350 °C) [28]

Full size image

Slag Chemistry

Slag is one of the key components that determine the quality of iron produced during a smelting process. The capacity of the slag to absorb elements such as sulfur and phosphorus is mainly controlled by its composition together with oxygen potential and temperature. For instance, high basicity of the slag increases its sulfide capacity. Further, the slag chemistry has a strong impact on the activity of FeO in the slag, which impacts the yield at the end of ironmaking and steelmaking. It is important to understand how the slag chemistry and structure affect slag–metal reaction equilibria as well as slag properties such as viscosity, density, and conductivity. Ironmaking slags contain silica and other complex-forming components, and the structure of the silicate has a significant effect on the structure and behavior of the slag. In this laboratory, the effects of CO/CO2, H2/H2O, and CO/CO2/H2/H2O gas atmospheres on the chemistry and structure of ironmaking slag were determined. Little work on slag chemistry under H2–H2O atmosphere had previously been reported in the literature. The results of this work are important in the development of an ecofriendly, novel, flash ironmaking process with the potential for steelmaking in a single continuous process [1, 6, 7, 22, 23]. The solubility of water in various slags has been studied by many researchers [31–41]. In this work, the effects of the water vapor on the equilibrium distribution of elements such as sulfur [42, 43], phosphorus [7, 42], and manganese [44] in addition to the effects on the activities of iron and magnesium oxides were investigated [6, 45, 46].

In slag, silicate anions are composed of silicon cations surrounded by four oxygen anions forming tetrahedral units. These tetrahedra are joined together in chains or rings by bridging oxygen (BO), as shown in Fig. 6. Cations are classified into two categories with respect to their impact on the silicate chains or rings (silicate polymer): the first type is the network breaker where the cations tend to break the BO bond to form nonbridging oxygen (NBO): O− and free oxygen, O2−. This process is called depolymerization of silicate melt, which is usually expressed by the ratio of nonbridging oxygen atoms to the number of tetrahedrally coordinated atoms, or simply denoted as NBO/T ratio [47]. It is worth noting that the rates and equilibria in slag–metal reactions as well as the physical properties of slag are strongly dependent on the NBO/T ratio [48, 49]. Examples of the network breaker cations are Ca2+, Mg2+, and Fe2+. In addition to this ratio, Si anions are frequently classified by the term structon or Q-notation (Qn where n = 0, 1, 2, 3, and 4) which was defined as single atom (or ion or molecule) surrounded by others in a specified manner [50, 51]. Figure 6 shows examples of silicate ions with their Q-notation and NBO/T ratios in addition to other distinguishing features. A Si polyanion may have more than one isomer with different Q values, as will be shown subsequently.

Fig. 6
figure 6

Models of silicate melt structural entities with their Q-notation, NBO/Si ratios, and structural units in addition to showing the bridging oxygen (BO) and nonbridging oxygen (NBO), respectively [6, 57]

Full size image

The second type of cations is the network former such as Al3+ and P5+ which form AlO4 5− and PO4 3− units. To balance the electrical charge, cations such as Ca2+, Mg2+, or Fe2+ are required. The cations involved in charge-balancing duties do not participate in network breaking. Thus, Al2O3 increases the polymerization of the silicate melt [49]. Like Al2O3, P2O5 increases polymerization since it forms phosphate complexes that have greater affinity for cations, which are mostly of the network modifier type than silicate ions. This affinity toward cations accounts for the ability of phosphate to increase silica polymerization since phosphate consumes the silicate network modifiers and decreases the attack of these cations on the silicate polymers. It is noted that P2O5 is less likely to form P–O–Si bonds compared with the aforementioned phosphate complexes that are based on the PO4 3− units [49].

The silicate melts comprise various 3-dimensionally interconnected units such as SiO2, Si2O5 2−, Si2O6 4−, Si2O7 2−, and SiO4 4− coexisting in the melt [49]. The nature of these units is affected by the nature of the cation in the silicate melt. The SiO2 and SiO4 4− units are stabilized by small cations of high valences, e.g., Mg2+ > Ca2+ > Na+. In the presence of these basic metal oxides, silicate polymers are broken according to the following pattern: Si6O18 12− → Si4O12 8− → Si3O9 6− → Si2O7 6− → SiO −44 . This explains the absence of isolated SiO2 molecules in industrial slags. Instead, SiO2-based slags are known to consist of polymers of tetrahedral units (SiO −44 ) [48, 52, 53].

Mohassab and Sohn [7, 43–46, 54, 55] have applied a number of analytical techniques to determine the structures and properties of the slags formed under the conditions of flash ironmaking. Several features of such instrumental analysis have direct correlations with important properties of slag such as its affinity to impurities. These techniques were applied to slags of interest for the Flash Ironmaking Technology (FIT), especially in terms of the effects of water vapor expected to be present in high contents in the new process [6, 55–58]. Table 1 summarizes the analysis results. The ratio of the more polymerized anions to the less polymerized anions, (Q2 + Q3)/(Q0 + Q1), in the H2O-containing atmosphere is at least twice that in the CO/CO2 atmosphere, in agreement with the FTIR results discussed elsewhere [6, 54]. In addition, the XPS and the NBO/T suggest that the studied slag is more depolymerized under a CO/CO2 atmosphere and more polymerized under the H2/H2O atmosphere [57]. It is, therefore, notable that all the results consistently lead to the conclusion that the CO/CO2 slag has the lowest polymerization degree of all the three, whereas the two other slags have a similar degree of polymerization. This suggests that water affects the slag more significantly than CO2 when they coexist in the gas mixture [6, 57]. Water vapor was found to be advantageous for lowering FeO in slag [46], and manganese [44]. Moreover they found that water vapor in the atmosphere of the new process would keep lining wear low based on the low MgO solubility in slag under its expected operating conditions [45]. The research on slag also found that water vapor decreased sulfur [43] and phosphorus [45] in produced molten iron.

Table 1 Comparison of a few structure analysis results [57]
Full size table

Economic and Environmental Aspects

A commercial scale plant flow sheet for the Flash Ironmaking Technology was constructed and its operation was simulated through detailed material and energy balance calculations [2, 4, 59, 60, 62]. Based on the process simulation, the optimum operating condition with low cost and high energy efficiency was selected. Further, an economic feasibility analysis was performed based on the process simulation results. The capital and operating costs for a commercial plant were estimated, and the net present value (NPV) after a 15 year operation was calculated. For details of the process simulation and economic analysis, the reader is referred to earlier articles from this laboratory [2, 4, 59, 60, 62].

Two potential process flow sheet configurations, the “1-step process” and “2-step process,” shown in Fig. 7, were developed using hydrogen or natural gas. The main difference between the two is in the number of reactors; the reduction of magnetite to iron [reactions (6) and (7)] takes place in one reactor in the 1-step process, and it takes place in two reactors in the 2-step process. In case of hydrogen, the two steps are:

Fig. 7
figure 7

Schematic diagram of iron production using hydrogen: a 1-step process and b 2-step process [59]

Full size image
$$ {\text{Fe}}_{3} {\text{O}}_{4} + {\text{H}}_{2} = 3{\text{FeO}} + {\text{H}}_{2} {\text{O}}({\text{g}}) $$
(6)
$$ {\text{FeO}} + {\text{H}}_{2} = {\text{Fe}} + {\text{H}}_{2} {\text{O}}({\text{g}}) $$
(7)

In the 1-step process, all materials required for producing iron are fed into the reactor, and the discharged off-gas is directly sent to heat recovery and cleaning processes for the removal of water and dusts prior to recycling the hydrogen.

The 2-step process has an iron ore pre-reduction reactor where wustite is produced. Since the reduction of wustite by hydrogen is significantly limited by equilibrium in the presence of water vapor, the off-gas from the iron producing reactor that contains a significant amount of hydrogen is used for pre-reduction of magnetite to wustite. Also, as the iron producing reactor requires a higher temperature than the pre-reduction reactor, a large amount of energy is provided in this reactor. The sensible heat contained in the off-gas from this reactor is sufficient for pre-reduction [59].

The compositions of iron ore concentrate and flux including major gangue components are given in Fig. 8. The composition of the slag after reduction reaction is also shown in Fig. 8. The slag basicity was set to be 1.2, which would provide appropriate slag viscosity to be separated from hot product iron [59]. Detailed flow sheets can be found elsewhere [2, 4, 59, 60, 62].

Fig. 8
figure 8

Mass and composition of input and output materials around the iron producing reactor (and the pre-reduction reactor) in the pilot scale 1-step and 2-step ironmaking processes using hydrogen [59]

Full size image

The hydrogen requirement was decreased by preheating the hydrogen to as high a temperature as possible by transferring the off-gas sensible heat, or decreasing the reactor operating temperature. Overall, the 2-step process had lower hydrogen requirement than the 1-step process. If the 1-step process would be applied, it would be critical for this process to utilize as much sensible heat of the reactor off-gas as possible for preheating hydrogen. The comparison between the 1-step and 2-step processes operated at 1500 °C and the excess driving force of 0.5 and the blast furnace process showed that both processes required less energy than the blast furnace process. However, this comparison did not include the energy required to produce purchased hydrogen [59].

An economic feasibility analysis based on “discounted cash flow analysis” was performed on the commercial scale flash ironmaking processes using natural gas or hydrogen producing one million tons of hot product iron a year. The net present value (NPV) was calculated at the discount rate of 10 % using estimated capital and operating costs for the project with 2-year construction and 15-year plant operation periods.

Process flow sheets for the commercial scale reformerless 1-step and 2-step ironmaking processes in which natural gas was directly used were constructed and simulated. The process flow sheets were constructed with hydrogen recycle, natural gas heating, a water gas shift (WGS) step, a waste heat boiler and a pressure swing adsorption step. Simulations were performed with the reactor operating at 1500 °C and an excess driving force of 0.5. The reactor feed gas preheating temperature was set to be 900 °C in the 1-step process, and 650 °C in the 2-step process. The total required energy for either reformerless process was lower than that for the blast furnace process [2, 4, 59, 60]. The results of the economic analysis indicated that the proposed “flash ironmaking process” will be economical even today if the process operates with in situ partial combustion of natural gas [62]. The total required energy for both reformerless processes was smaller than that for the blast furnace process. NPV estimation for the reformerless processes indicated that these processes would be more economic than the process using purchased hydrogen and the process combined with the steam–methane-reforming (SMR) process as will be discussed below, especially when the 1-step process was applied [2, 4, 59, 60, 62].

Process flow sheets for the commercial scale 1-step process combined with a built-in SMR step were constructed and simulated. The 1-step ironmaking process was constructed and simulated where (1) hydrogen is first produced and (2) syngas from the SMR unit is directly fed to the ironmaking unit. Simulations were performed with the reactor operating temperature of 1500 °C, the excess driving force of 0.5, and the reactor feed gas preheating temperature of 900 °C. The results showed that it would require less fresh natural gas when hydrogen was first produced and used in the iron producing reactor, which made it more economical by overcoming the additional cost of increased process steps and facilities. Also, fresh natural gas for fuel in the reformer was largely decreased by connecting the tail gas from the pressure swing adsorption (PSA) step in the ironmaking unit to the reformer burner. Compared to the blast furnace process, the total energy required in the 1-step process combined with the SMR process was 8 % larger when hydrogen was produced and 36 % larger when syngas was produced. NPV estimation for the 1-step ironmaking process with built-in hydrogen production was estimated to be $103 million (in 2010 dollars and conditions), which benefited from the lower operating cost. Natural gas price has a large effect on NPV. CO2 emissions credit did not affect NPV as much as the process using purchased hydrogen did [59].

In the process using purchased hydrogen, estimated NPV was negative $475 million (in terms of 2010 dollars and conditions) in the 1-step process, and negative $530 million (in 2010 dollars and conditions) in the 2-step process, respectively. However, the 1-step process could be economic depending on the availability of less-expensive hydrogen, CO2 emissions credit trading, and the price of iron [59].

These results indicate that the proposed “flash ironmaking process” will be economical even today if the process operates with in situ partial combustion of natural gas. Major results are summarized in Figs. 9 and 10. Figure 9 shows potential energy saving, lower carbon footprint, and economic feasibility (except when purchased H2 is used) relative to the BF. However, application of a $50 per ton of CO2 credit should make the use of hydrogen economically feasible at the 2010 price of hydrogen, as Fig. 10 indicates [4, 60, 62]. The values of CO2 emissions for the hydrogen-based flash ironmaking processes assume only those from limestone calcination and none from the ironmaking step. It is recognized that hydrogen production may involve CO2 generation, but since there are many possible methods of its production with varying degrees of CO2 emissions, including the reforming of natural gas or biomass or the electrolysis of water using nuclear or solar energy, the carbon footprint of hydrogen production was not considered in this analysis.

Fig. 9
figure 9

Energy consumption (GJ/ton of iron) and NPV (million dollars, relative to 2010 dollars and conditions using $2.5/kg for hydrogen, and $6/million Btu for natural gas) applying $0 CO2 credit for a 1 Mt/year plant of different configurations of the Flash Ironmaking Technology compared with average blast furnace. (H1S: H2-based 1-step process, H2S: H2-based 2-step process, SMRH1S: 1-step process with hydrogen production from SMR (Steam-Methane Reforming), SMRS1S: 1-step process with syngas production from SMR, R1S: Reformerless1-step process, R2S: Reformerless 2-step process) [4, 60, 62]

Full size image
Fig. 10
figure 10

The energy consumption (GJ/ton of iron), NPV (million dollars, irelative to 2010 dollars and conditions using $2.5/kg for hydrogen, and $6/million Btu for natural gas) applying $50 CO2 credit, and CO2 emissions (tons/ton of iron) for a 1 Mt/year plant of different configurations of the Flash Ironmaking Technology compared with average blast furnace [4, 60, 62]

Full size image

This transformative technology has significant energy saving and reduced CO2 emissions compared with the Blast Furnace process. It has been proved that iron particles of more than 95 % metallization can be obtained by reduction with hydrogen or a mixture of carbon monoxide and hydrogen in 2–7 s at temperatures of 1573 K (1300 °C) or above. The product of Flash Ironmaking Technology (FIT), which is expected to operate at temperatures higher than 1273 K (1000 °C), is significantly less reactive toward oxygen compared with DRI produced by the current technologies. Also, the molten iron produced by the FIT will contain less impurities and the relevant slag would be less corrosive to furnace lining.

Conclusions

The Flash Ironmaking Technology (FIT) was conceived as a novel ironmaking process at the University of Utah. This technology is based on fine particle reduction using a reductant gas such as hydrogen, natural gas, coal gas, or a combination thereof. It is the first flash-type ironmaking process converting iron ore concentrates directly to metallic iron in-flight, which would be suitable for an industrial-scale operation. This process will produce iron without requiring pelletization or sintering of iron ore and also avoids the need for coke. In addition, this process concept takes full advantage of the fine particle size of the concentrate with a large surface area, which permits rapid reduction by a gas. Another potential benefit of this process is the possibility of steelmaking in a single, continuous process. This transformative technology is expected to allow significant energy saving and reduced CO2 emissions compared with the Blast Furnace process.