The production of high value pig iron nuggets from steelmaking by-products – A thermodynamic evaluation

ABSTRACT

Zinc contaminated steelmaking by-products such as blast furnace (BF) dust and basic oxygen steelmaking (BOS) dust present a significant recycling challenge at integrated steelmaking plants. Rotary Hearth Furnaces (RHFs) provide an attractive route for recovery of Fe and Zn from these materials through carbothermal reduction of the oxides in the material to yield Direct Reduced Iron (DRI) and crude zinc oxide. The next generation of RHF processes such as ITmk3 and e-nugget produce pig iron nuggets from iron ore concentrates and coal, as iron and gangue separate in-situ without an additional melting unit. A computational study of the metal-slag system using FACTSAGE 7.3 suggests that production of pig iron nuggets of good quality (93.75 wt.% Fe, 4.3 wt.% C, 0.116 wt.% S, 0.66 wt.% Mn) can be produced from BF dust and BOS dust in a ratio of 37:63 with addition of 5.4 wt.% SiO₂ and 1.51 wt% MgO at 1450 °C. These computational results are in good agreement with experimental studies on similar material and, as such, suggest a practically feasible process.

Introduction

Iron rich by-product dusts and sludges are an unavoidable consequence of ironmaking and steelmaking activity. The recycling of these fine dusts is usually simple, reincorporating the material into the process through agglomeration in a sinter plant (Umadevi et al., 2013) or via cold bonded briquetting/pelletization (Andersson et al., 2019); however, dusts recovered from the off-gas dedusting systems in the blast furnace (BF) and basic oxygen steelmaking (BOS) plants at integrated works are often contaminated with zinc. The deleterious effect of zinc on the blast furnace is well known (Besta et al., 2013; Narita et al., 1981) and even relatively low concentrations of zinc (100–120 g Zn per ton of hot metal) in feedstock material can have negative effects on process stability and thus the economics of production. The removal of zinc from steelmaking materials via hydrometallurgical means (Cantarino et al., 2012; Jha et al., 2001; Oustadakis et al., 2010; Shawabkeh, 2010; Trung et al., 2011; Zhang et al., 2017) and pyrometallurgical means (Abdel-Latif, 2002; Assis, 1998; Brocchi and Moura, 2008; Mombelli et al., 2016a; Stewart and Barron, 2020) is well documented but these materials are often still stockpiled and underutilized by steel producers. Landfilling of these materials is not a long term economically and environmentally sustainable solution. The steel industry and the governments of countries the industry operates within are actively looking to avoid the negative environmental consequences of huge stockpiles of solid waste on defunct steel manufacturing sites, that must be remediated at huge expense before the land may be reutilized, as has been the case in many high profile plant closures (ENDS Report, 1992). Pyrometallurgical separation of zinc from these dusts in an on-site rotary hearth furnace (RHF) is becoming more commonplace at newer integrated works. These facilities utilize carbon in coal or in the by-products themselves to reduce, volatilize and capture the zinc from the materials in the form of a crude zinc oxide.

The major advantage of a carbothermal reduction route such as the RHF is the iron product is partially reduced and is known as direct reduced iron (DRI). High quality DRI can be >90% metallized (McClelland and Metius, 2003) and charging DRI back to the BF can reduce the coke rate of the furnace, which carries an economic and environmental benefit (Chatterjee, 2012). Unfortunately, DRI produced from recycled materials is often limited in quality. High gangue content due to the nature of the input materials is inevitable and carries penalties such as increased slag volumes, but another key value driver of DRI is sulfur content. As DRI is substantially metallized, it can also be used to displace scrap within a BOS vessel or an electric arc furnace (EAF). Scrap steel is substantially more expensive per unit of iron than virgin ore (Yellishetty et al., 2011) and DRI that is of high enough quality to be recycled to the BOS vessel rather than the BF will have substantially higher value in use (VIU) to the steelmaker.

The BOS process is sensitive to sulfur addition in the scrap charge, because despite the presence of a basic slag, the high oxygen activity in the converter makes the dissolution of sulfur into the hot metal (HM) thermodynamically favorable (Schrama et al., 2017). Many high-performance steels require extremely low levels of residual sulfur within the HM, with the production costs sharpening steeply as ultralow sulfur concentrations (<0.001 wt.%) become necessary (Pehlke and Fuwa, 1985). As such, low sulfur scrap substitutes are expensive and thus low residual sulfur levels within DRI are desirable for the DRI producer.

Next generation RHF processes such as ITmk3 go one step further than the purely solid-state reduction of iron oxides. Through careful manipulation of slag chemistry, the iron in the ITmk3 process is sufficiently carburized to liquify and coalesce into pig iron nuggets (Fig. 1), completely separate from an immiscible, liquid slag phase (Kikuchi et al., 2010; Kobe Steel ltd, 2010). Similar nugget making processes include E-Iron, which uses an adapted linear tunnel furnace to produce iron nuggets from non-zinc bearing mill wastes (Simmons, 2015) and Hi-QIP which uses an un-agglomerated feed on a bed of carbonaceous reductant (Ishiwata et al., 2009). The production of pig iron nuggets from zinc bearing mill wastes has been reported previously (Birol, 2019; Wang et al., 2010; Zhao et al., 2012), implying the feasibility of the production of nuggets based on experimental laboratory conditions in inert gas atmospheres.

Sulfur control in DRI production is mostly achieved by raw material input. DRI produced using coal as a reductant is too high in sulfur (0.44 wt%) to be utilized in a BOS/EAF melt shop (Seetharaman, 2014) and thus is usually utilized via a blast furnace, which is less sensitive to sulfur. However, the addition of a melting step and a molten slag, such as in ITmk3, offers an opportunity to sequester sulfur from the metallic iron within a distinct slag phase. This is not dissimilar to the mechanism by which sulfur is removed within the blast furnace (Schrama et al., 2017). Another potential avenue to add value to iron nugget production is by the simultaneous reduction of manganese oxides often found in BOS dust in small, but not insignificant amounts, into the pig iron nugget product (Longbottom et al., 2016; Mombelli et al., 2016a). Manganese is a common alloying element in the production of steel and is often added on a supplementary basis with the scrap charge to a BOS vessel or EAF in the form of ferromanganese alloys. This is a cost to the steel producer, therefore simultaneously reducing Mn into the hot metal produced during pig iron nugget making would be economically desirable. Srivastava reported the reduction of Mn_xO_y to the liquid iron phase during nugget making reactions using a manganiferous ore with minimal losses of Mn to the slag phase (18% at 1450 °C for 40 min) (Srivastava, 2014). However, this was performed using a high manganese iron ore rather than with steelmaking by-products and used graphite and polyethylene as a reducing agent.

Herein we propose that an exploration of the thermodynamics of the complicated slag-iron system present in nugget production, the ability for the nugget making process to desulfurize the produced pig iron and reduce manganese into the hot metal at a range of temperatures, slag chemistries and reductant levels can be determined. Three conditions need to be met to produce pig iron nuggets in a manner similar to the ITmk3 process or E-Iron;

• a low liquidus temperature slag, to allow for molten iron droplets to coalesce,

• sufficient carbon to reduce iron and zinc oxides to their elemental form, and

• sufficient carbon to further carburize the metallic iron such that the melting point is low enough to form molten pig iron at the reaction temperature.

Using BF dust and BOS dust from Tata Steel Port Talbot in the United Kingdom as a case study, the effect of various blends and temperatures is explored with respect to the thermodynamics of the chemical system. The study is limited in its scope to thermodynamics, without exploring the complicated reaction kinetics of the multiphase system and a comparison is made to experimental studies from literature performed on similar material. By comparing results of a thermodynamic study such as this to the experimental observations made in the literature to determine whether the relatively short reaction times in pig iron nugget making (<60 min) are sufficient to allow the slag-metal system to reach equilibrium or whether kinetic factors play a substantial role in determining Fe yield and Mn/S partition between the hot metal and the slag.