News

Pre-treatment of Hot Metal

• satyendra
• February 3, 2024
• 7 Comments
• BF, BOF converter, Dephosphorization, Desiliconization, desulphurization, Hot metal, Kanbara reactor, LD-NRP, LD-ORP, MURC, Pre-treatment, SMP, SRP, ZSP,

Pre-treatment of Hot Metal

Since the 1980s, the quality of raw materials and fuels used in ironmaking and steelmaking has increasingly deteriorated around the world. This trend has led to a deteriorating quality of hot metal in terms of higher contents of impurities. Also, during this period, there has been growing demand for the production of clean steel with low and ultra-low contents of impurity elements. The decreasing availability of higher quality raw materials and fuels with lower contents of unwanted and harmful elements and the increasing pressure to lower production cost by purchasing cheaper raw materials have led the iron and steel industry in a challenging condition.

The western steel plants have responded to the above challenge by (i) optimization of the facilities, (ii) improvement in the degree of automation, (iii) introduction of bottom blowing of inert gas in the BOF (basic oxygen furnace) converters, and (iv) improvement of the productivity of the production units to very high levels. However, these efforts have a limit and for the achievement of further progress needed alternative concepts. On the other hand, Japanese steelmakers opted for a different very radical path and focused on intensive hot metal pre-treatment of hot metal taking advantage of the benefits of the low treatment temperature. In this way, they established a different but competitive way of steelmaking. The concepts and the technologies of hot metal pre-treatment and their application to the conventional steelmaking route provide an efficient way of producing clean steel in a cost-effective manner.

The introduction of BOF converter technology in Japan has occurred at a time of limited availability of high-quality scrap, and, as a result, the desire has been to minimize the use of this expensive resource. Steel production has been focused on the use of controlled, prepared raw materials. The technologies developed for the efficient removal of silicon and phosphorus from the hot metal, both fundamentally endothermic when carried out using the customary oxide reagents, provided an economic benefit by consuming chemical energy otherwise available for melting scrap in the BOF converter.

Increasingly stringent quality requirements have increased the demand for steels with very low levels of impurities such as phosphorus (P), sulphur (S), hydrogen (H2), nitrogen (N2), and oxygen (O2), and of non-metallic inclusions such as manganese sulphide (MnS), silica (SiO2), and alumina (Al2O3). Such high purity cannot be achieved by blowing the BOF converter for decarburization since its refining capability is limited. Hot metal produced in the blast furnace (BF) is conventionally transferred either to a hot metal ladle or to a torpedo ladle car and is then transported to the steel melting shop for charging of the hot metal into the BOF converter.

The oxygen blowing process in which hot metal is decarburized and converted to steel is carried out mainly in the BOF converter. However, a method for dividing the refining capability and allocating the divided function to processes before and after the BOF converter has been put into practical use. The processes in which impurities are removed from the hot metal are called hot metal pre-treatment, whereas the processes in which the molten steel tapped from the BOF converter is subjected to further refining and degassing are called secondary refining. At present, an integrated process of smelting in the blast furnace, hot metal pretreatment, decarburizing in the BOF, and secondary refining has become the standard manufacturing process for high grade steels.

Hot metal pre-treatment – Process steps which are used to remove impurities from the hot metal can be summarized as hot metal pre-treatment. These process steps are inserted between the blast furnace and the BOF converter. Main objective is the adjustment and control of the hot metal composition by the removal of unwanted impurities like sulphur (desulphurization), silicon (desiliconization), and phosphorous (dephosphorization) along with their undesired inclusions (oxides, borides, nitrides, carbides, and chlorides). This is a pre-condition to produce clean steel. Less than 100 ppm total impurities are standard requirement for quality steelmakers today.

Pre-treatment of hot metal is the adjustment of the composition and temperature of blast furnace produced hot metal for optimal operation of the BOF converter process. As such, it is one of the interdependent chain of processes which constitute modern steelmaking. Fig 1 shows the progress of these processes. When taken to the extreme case, the BOF converter process function is reduced to scrap melting and carbon reduction subsequent to the prior removal of silicon, phosphorus, and sulphur in preparatory steps under thermo-dynamically favourable conditions. An important benefit of removing phosphorus and sulphur from the hot metal prior to the BOF converter process is the ability to produce steels with phosphorus and sulphur contents lower than otherwise achievable without severe penalty to the BOF converter process. Silicon removal is beneficial to the converter to reduce the chemical attack on the basic refractory lining and to allow the use of only minimal quantities of slag-making fluxes, hence maximizing process yield.

Fig 1 Progress of interdependent chain of processes which constitute modern steelmaking

Desulphurization of hot metal is a standard practice carried out in nearly all steel plants around the world while desiliconization and dephosphorization have not been in the main focus of the Western steel plants. This has not been different in Japan, where the steel plants have investigated and developed hot metal pre-treatment technologies. A variety of hot metal pre-treatment processes and technologies have been developed, applied, and reported. The processes which are the most common ones are (i) slag minimum process (SMP), (ii) simple refining process (SRP), (iii) zero slag process (ZSP), (iv) LD-optimized refining process (LD-ORP), (v) LD-new refining process (LD-NRP), and (vi) multi-refining converter process (MURC).

All the above processes use different vessels for carrying out the metallurgical work of desiliconization, dephosphorization, and desulphurization. The main metallurgical vessels are (i) BF runner, (ii) torpedo car, (iii) open top hot medal ladle (transfer ladle or charging ladle), and BOF converter. All kind of combinations using different vessels and logistics have been tried and operated during the years.

One of the early pioneers has been Nippon steel. It has developed and introduced the slag minimum process (SMP) in 1983 as shown in Fig 2a. The main idea of this process is to separate the desiliconization reaction from the dephosphorization and the decarburization reactions and to separate the process slag. In industrial scale, this process consisted of hot metal desiliconization in the ladle using a KR (Kanbara reactor) impeller system. The BOF converter is then utilized for dephosphorization and decarbonization. Advantages compared to the conventional process using the BOF converter for removal of sulphur, silicon, phosphorous, and carbon have been found in a reduced consumption of fluxes (mainly lime), a reduced slag quantity, and an increased metallic yield. It is reported that the slag quantity of this process for a hot metal silicon content of 0.15 % is half compared to the conventional process. Further investigations have led consequently to technologies using the injection of suitable reagents into (oversized) torpedo cars. This operation procedure offered considerable benefits in terms of ‘less slag’ production and slag recycling etc. but it has also some disadvantages because of the excessive handling needs and the huge temperature losses.

Fig 2 Flowsheets of slag minimum process and slag refining process

Nippon steel has installed and operated the torpedo car based ‘simple refining process’ (SMP) for hot metal dephosphorization in its Kimitsu steel plant and Yawata steel plant as shown in Fig 2b. Another Japanese steel producer, Kawasaki steel plant has installed and operated similar processes in its Mizushima steel plant and Chiba steel plant, also using the torpedo car as reactor for dephosphorization. The plants applied the intensive pre-treatment operations in dedicated hot metal pre-treatment shops located between blast furnace and steel melting shop. Chiba steel plant firstly used gaseous oxygen blowing during dephosphorization in order to reduce the tremendous temperature losses. The process flow of hot metal pre-treatment process at the Chiba steel plant is shown in Fig 3a.

Fig 3 Hot metal pre-treatment process

The large slag generation because of blowing oxygen in the torpedo car during dephosphorization and a highly restricted scrap usage ratio, led Nagoya steel plant of Nippon steel to modify its BOF converter into a hot metal pre-treatment converter for desiliconization, dephosphorization, and desulphurization. After tapping and separation of the slag, the hot metal is then charged to another converter, where decarbonization is done. Fig 3b shows the flow diagram of this process, which has been named ‘LD – ORP’ process.

The first hot metal pre-treatment shop using torpedo cars for desiliconization and open top ladles for dephosphorization and desulphurization has been installed and operated at the Oita steel plant of Nippon steel. Fig 4a shows this process which is completely based on the deep injection technology. The use of a turn-table ladle turret with four treatment stations for sampling, two number deslagging, and injection (dephosphorization and desulphurization) is another highlight of hot metal pre-treatment process at the Oita steel plant.

Fig 4 Ladle based hot metal pre-treatment and zero slag process

NKK’s Fukuyama works (now known as JFE) operated a hot metal pre-treatment process where the silicon content of hot metal from the blast furnace is already lowered to 0.2 %. The hot metal is then sent to the desiliconization station, where it becomes ultra-low silicon hot metal with a silicon content of less than 0.1 %. At the desiliconization shop, oxygen gas is used along with sintered iron ore (iron oxide) as reagents for desiliconization. The reaction vessel is an open top ladle, and the hot metal is vigorously stirred by injecting lime through an immersed lance. This method provides a highly efficient and stable supply of ultra-low silicon hot metal which considerably improved the efficiency of lime for dephosphorization. As a result, the slag generation throughout the entire steelmaking process can be lowered to a minimum quantity. Hence, NKK named this completely ladle-based process as ‘zero slag process’ (ZSP) which is shown in Fig 4b.

A further development has been the ‘multi-refining converter process’ (MURC) which allows desiliconization / dephosphorization and decarbonization treatment in one vessel without tapping and recharging. The desiliconization / dephosphorization is carried with a low basicity slag, high (% total Fe) and a hot metal without previous desiliconization treatment. After dephosphorization the phosphorus rich slag is tapped and sent for dumping. The slag of the following decarbonization step is totally recycled as it remains as a hot feed-stock in the converter for the next desiliconization / dephosphorization treatment as shown in Fig 5a.

Fig 5 Multi refining converter process and possible flow of hot metal pre-treatment within BF-BOF route

There are a number of processes in the secondary metallurgy in order to ‘fix’ deficiencies of the produced steel and for ensuring highest cleanness. Some of them are necessary from the metallurgical point of view but some treatments which are applied to liquid steel at high cost can be beneficially transferred to hot metal at lower cost. Further, beside the ability to produce clean steel with low and ultra-low contents of sulphur and phosphorous, there are additional benefits of hot metal pre-treatment.

Benefits of hot metal desulphurization with regards to the blast furnace operation are (i) the blast furnace can be released from some metallurgical work of desulphurization which increases productivity of blast furnace, (ii) coke, coal, and flux charges to the blast furnace can be reduced and the hot metal yield can be increased, (iii) the blast furnace alkali balance can be improved because of operating at a lower basicity and a cost saving ‘lean slag’ production becomes possible.

Benefits of hot metal desiliconization and dephosphorization with regards to BOF converter operation are (i) reduced refractory attack by the less acid slag, (ii) double slag technique, frequently applied in case of high silicon hot metal, can be avoided and this also shortens the tap-to-tap time and increases productivity, (iii) the slag generated during oxygen steelmaking using dephosphorized hot metal has a low phosphorus content which allows utilization of this slag in the sinter plant as part of the feed-stock for the blast furnace (the recycling of BOF slag in the blast furnace without previous hot metal dephosphorization is limited because of an enrichment of the phosphorus content in the hot metal), (iv) lower contents of silicon, phosphorous, and sulphur in the hot metal allow operating the BOF converter for more or less decarburization only with minimum slag (less than 1/3rd of the standard quantity) and hence minimizing lime consumption, besides the obvious saving, the so reduced slag quantity leads to a smoother blowing behaviour, less slopping, and an easier control of the end-point targets, and (v) BOF converter operation with minimum slag allows (partially) substitution of ferro-manganese added at tap by much cheaper manganese ore (because of the small slag quantity the major portion of the manganese from the manganese ore is dissolved in the steel instead of being oxidized to the slag) and this operation allows substantial cost savings.

Anyhow, the rather smart hot metal pre-treatment technologies did not have a major break-through outside of Japan and some installations in Taiwan and South Korea. Presently, the conditions have changed and there is an increasing need for steel plants to investigate opportunities to benefit from applying hot metal pre-treatment steps into their existing conventional BF-BOF production route in order to release their aggregates from some metallurgical work.

Hot metal pre-treatment includes the desiliconization, dephosphorization, and desulphurization of hot metal. The silicon (Si) in the hot metal is oxidized in the BOF converter, where it reacts with added lime (CaO) and iron oxide (FeO) to form a CaO-FeO-SiO2 slag. If the silicon content of the hot metal is low, this reaction is shortened in the BOF converter, the production efficiency is improved, and the volume of slag generated is small, hence, decarburizing with a high iron yield is possible. Desiliconization is hence conducted as a pre-treatment process by adding iron oxides such as mill scale and sintered ore fines to hot metal in the runners in the cast house of the BF or in the hot metal ladle or the torpedo car.

Dephosphorization is normally carried out by injecting a dephosphorizing agent containing lime, iron oxide, and fluorspar etc. into the hot metal in the open top hot metal ladle or a torpedo car together with a gas. This promotes the transfer of the phosphorus in the hot metal to the slag phase, which is then discharged. Dephosphorization is normally carried out after desiliconization, since the dephosphorization reaction proceeds more quickly at lower silicon contents.

Although hot metal is desulphurized to some extent by the dephosphorization treatment, extra low sulphur steels need further desulphurization, which is performed by separate injection of desulphurizing agents such as CaO (lime), Na2CO3 (sodium carbonate), CaC2 (calcium carbide), and Mg (magnesium) into the hot metal. Such treatments can be made more effective by identifying and improving the elementary rate steps which control the dephosphorization and desulphurization processes. A good understanding of thermodynamics and transport phenomena is indispensable in achieving these objectives.

By 1983, a large number of pre-treatment facilities have been in use. Initially, these pre-treatment processes have been performed by adding iron ores or sinter to the hot metal during its flow in the blast furnace runner. Further, improvements and control over chemical results have been achieved by the addition through a sub-surface injection of the reagents in dedicated vessels, such as oversized torpedo cars. This brought on the use of a variety of chemical reagents, including soda ash (sodium carbonate), which also provides for considerable removal of sulphur.

When using iron oxides for desiliconization, it is necessary to separate, i.e., remove, the process slag before the hot metal is desulphurized as this operation needs low oxygen potential for efficient performance. It is important to recognize that phosphorous removal occurs only in hot metal containing less than 0.15 % Si, additionally, phosphorus held in the slag can be subject to reduction, i.e., reversal, into the hot metal if it is present during desulphurization. An interesting technical development is the combination of dephosphorization and desulphurization in a single vessel whereby phosphorous is reacted with the oxidizing reagents as they rise in the liquid and sulphur is removed by the top slag in the vessel as shown in Fig 6a.

Fig 6 Pre-treatment of hot metal

Desiliconization and dephosphorization are accompanied by losses of carbon from the hot metal and evolution of carbon di-oxide (CO2) from carbonate reagents. Hence, control strategies such as addition of coke breeze or equipment accommodations are to be made in the reaction vessel and gas capture systems to contain foaming and flame evolution. In the recent timeframe, environmental considerations over disposal of sodium-containing slags have forced the use of limestone-based reagents, frequently mixed with iron ore or sinter fines and delivered with oxygen, the latter used to diminish the thermal penalty from the pre-treatment process. Oxygen consumption in these process steps is shown in Fig 6b.

Hot metal desiliconization (De-Si) – The silicon in blast furnace hot metal comes mainly from the ash of the blast furnace coke and the blast furnace operation. Silicon has a higher oxygen affinity than phosphorous and carbon. It is hence the first element to be oxidized during oxygen blowing in the BOF converter. The competitive reactions of phosphorous and in particular carbon starts with decreasing silicon content in the hot metal.

High silicon content in hot metal of more than 0.7 % to 1 % leads to problematic oxygen blowing because of (i) increased consumption of lime for maintaining a suitable basicity, (ii) increased iron losses in the slag, (iii) decreased capacity of BOF converter because of the increased slag quantity, (iv) increased oxygen blowing time and hence decreased productivity (sometimes even a double slag practice is necessary), and (v) increased danger of slopping (instable blowing process). Hence, control of the silicon content below, say 0.7 %, is desirable and one of the benefits of hot metal desiliconization. Another reason is that a silicon content of less than 0.15 % is necessary before effective dephosphorization can start.

The process of hot metal desiliconization needs oxygen which can be supplied as gaseous oxygen or in form of iron oxides (solid oxygen). Hot metal desiliconization with solid oxygen is endothermic and drops the temperature by 10 deg C per 0.1 % of removed silicon (using mill-scale as source of solid oxygen). The slag skimming also accounts for large temperature losses in case of high initial silicon content in the hot metal. Desiliconization by means of gaseous oxygen is more effective compared to the addition of a desiliconization agent. Assuming initial silicon content of 0.7 %, around 1.4 normal cubic meter (N cum) of О2 per ton per 0.1 % of removed silicon is needed. Fig 7a compares the removal of silicon as a function of time for adding solid oxygen (here mill-scale), oxygen injection, and oxygen top blowing. Blowing of gaseous oxygen for desiliconization is exothermic and increases the temperature by around 27 deg C per 0.1 % of removed silicon. The large quantity of generated slag makes it difficult to control this process in ladle.

Fig 7 Desiliconization behaviour of hot metal

Fig 7b shows the temperature variation in case of solid oxygen addition (here mill-scale) and injection of gaseous oxygen as a function of removed silicon. High mixing power of the desiliconization reagents, metal, and slag is needed for efficient desiliconization. This can be achieved by several methods, e.g., natural flow, gas stirring by injection, and mechanical stirring with an immersed impeller. In general, the desiliconization efficiency of a process is to be high enough to lower the silicon content to match with the needs of proper oxygen blowing in the BOF converter or to lower the content to allow efficient dephosphorization. The methods which are applied in industrial scale for desiliconization of hot metal are described below.

The first is top addition of desiliconization reagents into the blast furnace runner with or without slag separation by means of a skimmer blade. The natural flow of the hot metal is utilized to generate mixing power. The second is the injection of desiliconization reagents into torpedo car with or without simultaneous oxygen blowing. The third is the injection of desiliconization reagents into open top ladle with or without simultaneous oxygen blowing, and the fourth is the top addition of desiliconization reagents into open top ladle using a KR (Kanbara reactor) impeller stirring system for providing mechanical stirring.

Method 1 is cheap and simple since less equipment is needed but efficiency and predictability are poor. An addition of 10 kilogram (kg) to 20 kg sinter dust fines per ton of hot metal achieves 30 % removed silicon. Injecting the desiliconization reagents into the blast furnace runner increases the desiliconization rate to 40 %. Nakasuga along with his colleagues introduced a process using mechanical stirring with an impeller immersed in the hot metal runner. They report improved desiliconization rates up to 55 % compared with the injection method. All these methods have fluctuations in the achieved silicon content, caused by varying temperature and varying flow velocity of the hot metal. The desiliconization rates achieved with these methods are normally sufficient to control the hot metal silicon content below say 0.7 % making it suitable for oxygen blowing in the BOF converter. But desiliconization down to 0.15 % which is the pre-condition to proceed with dephosphorization needs procedures to be hollowed as per methods 2 to 4.

Method 2 saves process time as the reactions can take place ‘on the road’. Because of the shape of the torpedo car, it is difficult to remove the high quantities of slag completely. As a result, the capacity of the torpedo car is decreasing with every treatment as accretions accumulate inside and at the mouth of the torpedo car. Finally, the cast house layout and its facilities are frequently limiting the application of this procedure.

Method 3 and method 4 allow precise control of the final silicon content and adjustment of the temperature to certain extend. Furthermore, these methods frequently allow usage of existing facilities already being used for desulphurization. Mill scale, magnetite fines, hematite fines, or sinter dust are used as sources of solid oxygen in the desiliconization reagent with similar desiliconization efficiencies. Lime and silica are added to flux the slag, to adjust the basicity, and to prevent excessive foaming.

Hot metal desulphurization (De-S) – Agricola has written four- and one-half centuries ago that ‘…sulphur is frequently found in metallic ores, and, generally speaking, is more harmful to the metals, except gold, than other things. It is most harmful of all to iron…’. From ancient times, through puddling furnaces and into blast furnaces, the control and elimination of sulphur has been a major task for the steelmaker. The cost of sulphur is every high. In its simplest form, a modern coal to steel flow sheet involves separating more than 99 % of the sulphur dug out of the ground at the coal pits. For controlling the 11 kg/ton of sulphur contributed by the coal and other feed-stocks, a steel plant spends a substantial quantity of money in addition to capital charges for equipment and exclusive of processes necessary to be followed during secondary steelmaking for sulphide shape control in the steel product.

Sulphur in hot metal is mainly coming from fuels such as coke, coal, and oil. With the exception of some very specific steel grades, sulphur is undesirable and has to be removed. Although the blast furnace is a very efficient facility for desulphurization and able to remove typically around 85 % of the total sulphur input, 15 % are remaining in the produced hot metal. This is still very high for satisfying the requirements of the steel grades. Hence, additional external desulphurization outside of the blast furnace is needed.

In the oxygen driven operation of BOF converter, the heavily oxidizing environment of the metal and slag and the inability to achieve the equilibrium sulphur partition ratio between slag and metal limit the sulphur removal capability of the process. Hence, to bring the sulphur content of the steel to within the range manageable by the far more costly steel desulphurization, the lower cost hot metal treatment technologies have been developed for removing sulphur prior to the oxygen steelmaking step.

Initially these technologies have been used to help the steelmaker, but, in time, it has been recognized that considerable cost savings and production increases in ironmaking result if sulphur limits formerly imposed on the blast furnace operation are lifted. In several steel plants, the hot metal leaves the blast furnace containing 0.04 % sulphur to 0.07 % sulphur, while the oxygen converters are charged with hot metal containing as little as 0.01 % sulphur to 0.001 % sulphur, to conform to limits on steel composition set by caster operations and final product quality requirements.

The importance of sulphur management and the large costs involved have led to worldwide efforts to develop and implement an range of different desulphurization technologies. The different reagent and delivery systems in use are the result of local economic and environmental factors and the preferences of technical and operating management at the individual plant sites.

The desulphurization during the production of steel is carried out by several methods namely (i) injection of desulphurization reagents into the torpedo car, (ii) injection of desulphurization reagents into the open top ladle, (iii) addition of desulphurization reagents into open top ladles using the KR impeller stirring system, (iv) desulphurization in the BOF converter, and (v) desulphurization in the steel teeming ladle.

As a comparison, the estimated cost in for the removal of I kilogram of sulphur is around 30 units in the blast furnace, around 11 units from the hot metal in the ladle, around 180 units in the BOF converter, and around 65 units from the steel in the steel teeming ladle. This cost comparison clearly shows that desulphurization of hot metal is beneficial compared to a higher effort needed to improve the desulphurization in the blast furnace. The BOF converter is the most expensive vessel to perform desulphurization and steel desulphurization is only to be applied if necessary.

Hot metal desulphurization by injection technology is carried out by deep injection of powdery reagents like lime, calcium carbide, magnesium, soda ash, or mixtures thereof into ladles (torpedo ladle, open top ladle, or BOF charging ladle). Refractory lined lance is immersed deep into the hot metal while the reagents are injected with high-speed using mainly nitrogen as transport gas. Several pneumatic injection technologies like mono-injection, co-injection, or multi-injection are used depending on the selected reagents and operational requirements.

Efficient desulphurization needs intensive mixing of reagents, metal, and slag. This can be achieved with a gas / solid mixture deeply injected into a ladle ensuring a minimum bath level of 1.5 meters in order to maximize the residence time of the particles in the bath before they reach the surface. Hence, injection desulphurization in open top ladle has a higher efficiency compared to torpedo car having an unfavourable shape. The consumption of reagents in the torpedo car is around 15 % to 25 % higher compared to open top ladle. Another possibility for achieving high mixing power is mechanical stirring with an impeller. This technique is used by KR (Kanbara reactor) system.

A Kanbara reactor is a hot metal pre-treatment facility which can remove sulphur in the hot metal to lower levels and at a cheaper cost than conventional facilities. The Kanbara reactor is a mechanical mixer-type hot metal desulphurization facility. It is characterized by its acceleration of the desulphurization reaction through the submersion and rotation of an impeller in hot metal, and subsequent mechanical mixing of the hot metal and desulphurizing reagent. The KR process efficiently advances the desulphurization reaction by means of high-speed (around 120 revolutions per minute, rpm) rotation of the impeller. As a result, it enables the reduction of sulphur concentration in the hot metal to low levels of within several dozen ppm (parts per million) using only inexpensive lime as a desulphurizing reagent, without the need for expensive magnesium. Desulphurizing reagent costs can be considerably reduced by replacing injection-type desulphurization facilities which use magnesium and calcium carbide with KR process. On comparing the injection method using magnesium and calcium carbide to the KR method using lime, the desulphurizing cost can be reduced by around 45 %.

Proper removal of the desulphurization process slag is necessary before the hot metal is charged into the BOF converter. The main chemical reactions for sulphur removal from hot metal, the range of process permutations, the specifics of reagent delivery systems, the importance of reaction vessel selection and of slag management issues bear on a well-functioning system.

The variety of process permutations adopted worldwide depend on one or a combination of several reactions namely (i) Na2CO3(s) + S + C = Na2S(l) + CO2(g) + CO(g) (equation 1), (ii) Mg(s) + S  = MgS(s) (equation 2), (iii) CaC2 + S = CaS(s) + 2C (equation 3), (iv) CaO + S + C = CaS(s) + CO(g) (equation 4), (v) Mg + CaO + S = CaS(s) + MgO(s) (equation 5), (vi) CaO + 2Al + S + 3O = (CaO.Al2O3) (S) (equation 6), and (vii) (CaO.Al2O3)(s) + S = (CaO.Al2O3) (S) (equation 7).

Initially, majority of the plants relied on reaction at equation 1, i.e., the addition of soda ash (Na2CO3) at the blast furnace cast house or at the steel plant while filling hot metal ladle (open top or charging ladle) or torpedo car. This approach has been abandoned as process control and environmental management has been very difficult. In several plants, mechanical stirrers (KR) have been introduced into the blast furnace runners, and later for use in open top ladle and torpedo car.

The next step has been dependent on reaction at equation 2 with the use of Magcoke (a product made by filling the pores of coke with magnesium and submerging this material into the hot metal in a sequence of multiple dips). Results have been reproducible, but, achievement of sulphur contents of less than 0.02 % have been costly in reagents and process time. Capture of the copious magnesium fumes has been nearly impossible without total building evacuation.

The chemical behaviour of magnesium in hot metal has been the subject of extensive study. It is important to realize that the solubility product of magnesium and sulphur is strongly dependent on temperature as well as the silicon and carbon content of the iron. This results in improved sulphur removal, or the reduced need for magnesium for colder iron. The practical effect is to lessen the cost penalty of having to load relatively cold hot metal with magnesium for attainment of sulphur levels lower than 0.002 % sulphur. An interesting technical side effect of the increase in solubility of magnesium in hot metal at low sulphur levels, e.g., near 10 ppm sulphur, is the observation which some magnesium appears to be oxidized from the iron as soon as the raker blade clears the surface of slag. The effect is noticeable as a light white fume even after emptying the hot metal ladle into the BOF converter. Sampling has shown the plume material to consist mostly of MgO.

A critical step in process development has been adoption of sub-surface pneumatic injection of calcium carbide powder (reaction at equation 3) and of pulverized lime (reaction at equation 4) or combinations, i.e., mixtures, of pulverized magnesium and lime (reaction at equation 5). Since calcium carbide is inert, it is difficult to distribute it throughout the liquid. For improving on this, one of two reagents is added (around 15 % to 20 %) to create surface and stirring, limestone, which cools the liquid or diamid lime, which is less endothermic. The latter version, known as CaD, has been developed by SKW in Germany.

An important improvement came in the development of co-injection technology which consists of the controlled mixing in the transport line of reagents supplied separately. This technique, now in universal use, allows for a wide array of reagent combinations and permits independent adjustment of the rates of the delivery of the reagents during the process. This is very useful for magnesium-based systems wherein splash and fuming during lance insertion and removal can be kept to a minimum by starting and stopping magnesium reagent flow with the lance tip at the deepest immersion in the hot metal. Another advantage is that the rate of delivery of the magnesium can be reduced at low sulphur levels when low magnesium solubility limits its dissolution in the iron.

Co-injection affords a cost benefit by allowing the user to purchase the individual components from the least costly material supplier rather than being solely reliant on a supplier for a proprietary mixture. A further benefit over blended reagent mixtures is elimination of segregation (i.e., separation) of individual reagent components with differing size or density while the mixture is in transit or in storage. Although earlier carbide-based systems have been favoured for a number of years, several shops also have adopted co-injection for lime and magnesium.

In another variant, carbide and magnesium are used in combination by co-injection. For a fixed magnesium feed rate (splash limit) carbide + magnesium is faster than lime + magnesium. In some plants, magnesium granules, coated with passivating layers of salts, have been in use with delivery by subsurface injection. However, the use of this reagent has environmental concerns. Further, because of the environmental constraints on the disposal of slags containing residual quantities of calcium carbide, reagent systems based on vigorous stirring of lime and soda-based reagents (KR process) have been used successfully. The benefit of intense hot metal reagent mixing is demonstrated in one plant where the spent desulphurization slag from a shop equipped for conventional injection treatment is used as the main reagent in a shop using the KR process. Recently, environmental concerns on disposal of soda slags have brought on adoption of lime plus magnesium systems.

A variant to the use of reactive agents like calcium carbide and magnesium is the use of lime powder either preceded by addition of aluminum to the hot metal, or lime delivered with organic stirring agents such as natural gas and / or solid hydrocarbons. The latter, used as ground solid hydrocarbons, has been adopted for improving mixing even for magnesium-based reagents. With the use of aluminum, CaO / Al2O3 globules form which have a large solubility for sulphur. Recently, in some plants, desulphurization by injection of pre-fluxed 2CaO.Al2O3 has been introduced with some commercial success (reaction at equation 7). Although these reagents have lower unit cost than carbide or magnesium, there is a limitation shared with lime systems, i.e., the higher mass of reagent needed increases the time needed for treatment and for the follow-on raking step.

Two other methods for delivery of desulphurization reagents into hot metal transfer ladles are worthy of note. One approach, paralleling the use of cored wires for steel ladle treatment, is to feed magnesium cored wire at high rates to reach the ladle bottom for release of the reagent at maximum depth. Magnesium in this form is far costlier than as an injectable solid, although the delivery system is simpler to operate and maintain. Another approach, with somewhat larger implementation, known by the commercial name ISID, consists of feeding the reagents through a rotatable bayonet system installed low in the wall of the open top ladle or torpedo car. Maintenance concerns and cost have limited broad implementation of this technology.

Hot metal dephosphorization (De-P) – Phosphorous in the iron ore fed to the blast furnace is converted into the hot metal by a rate of higher than 90 %. The only possibility to limit the phosphorus content of the hot metal is to use a low phosphorus containing iron burden. It is hence clear, that higher phosphorus content in hot metal needs higher efforts in steelmaking operations. The BOF converter is a highly efficient facility to perform dephosphorization especially when equipped with strong bottom stirring. However, several limitations call for alternatives like hot metal desiliconization and dephosphorization. Applying hot metal dephosphorization reduces the needed slag quantity during the oxygen blowing in BOF converter, and ensures achieving consistently ultra-low phosphorous contents at end of the blow. In this sense, hot metal pre-treatment is of big help in reducing the metallurgical work load of the oxygen blowing process.

Basically, hot metal dephosphorization makes use of the same reagents as desiliconization. Oxygen in solid and / or gaseous form is needed. Iron oxides like mill scale, sinter dust fines, iron ore fines (e.g., magnetite or hematite fines), etc. are used as sources of solid oxygen. Blowing oxygen onto or into desiliconized hot metal initiates strong decarburization simultaneously. Because of the {CO} gas generation, such a process is impossible to be managed safely in an open top ladle. Hence, hot metal dephosphorization is done either in ladles with the sole use of solid oxygen or in a BOPF converter which provides enough space to handle the foamy slag.

Similar to desiliconization, also dephosphorization of hot metal needs high mixing power. A simple addition of the reagent on the surface of the metal results in unacceptable long treatment time and tremendous temperature losses. Application of gas stirring through bottom bricks in the ladle improves the situation only slightly. Only deep injection of the reagents or mechanical mixing with impellers provides sufficient mixing power.

The benefits of hot metal dephosphorization can be (i) dephosphorization is more effective at lower hot metal temperatures than at steel temperatures, (ii) a lean slag BOF converter blowing process for decarburization can be established, (iii) slag from a BOF process using dephosphorized hot metal contains no free lime (CaO) and allows direct utilization as construction material e.g., for road layers etc., and (iv) slag from a BOF process using dephosphorized hot metal can also be utilized for hot metal desiliconization.

Integration of hot metal pretreatment process steps into the conventional BF-BOF steelmaking route – In order to evaluate the technical feasibility and the economic benefits, an optimum sequence has to be defined. Tab 1 summarizes favourable conditions for the individual hot metal pretreatment steps.

Considering this and the typical logistics of the conventional BF-BOF steelmaking route the process flow shown at Fig 5b is realized with little modifications. Desiliconization is executed in the blast furnace runner with separation of the desiliconization slag during pouring into a torpedo car by means of a skimmer blade. This is followed by based on injection of dephosphorization agents into the torpedo car. After reladling into the BOF charging ladle, dephosphorization slag is skimmed off. Then desulphurization is carried out by means of injection technology or a KR system. After the desulphurization slag has been removed, the hot metal is charged into the BOF converter and the decarburization process is executed by means oxygen blowing.

It is known that during desiliconization with a proper desiliconization reagent also a certain quantity of sulphur is removed. This advantage can be utilized when desiliconized is carried out before desulphurization. Although this sequence is a ‘natural’ sequence taking advantage of the ‘low cost’ desiliconization in the BF runner as the first step, another sequence is frequently more advantageous from thermodynamic point of view. Desulphurization as well as desiliconization, both favour high temperature. In opposite, dephosphorization favours low temperature. Considering this, the sequence shown in Fig 8 is preferable.

Fig 8 Alternative sequence of process step within the conventional BF-BOF step

Summarizing it can be said that sequence has to be selected for each installation individually depending on available facilities particular, existing vessels for metallurgical targets, and temperature need to be considered. Several desulphurization stations for open top ladles are in operation worldwide based on either injection technology or mechanical stirring technology (KR systems). Very frequently these existing stations can be easily extended to hot metal pre-treatment providing desiliconization and dephosphorization additionally to desulphurization when needed and beneficial.

The major reasons why hot metal desiliconization and dephosphorization in open top ladles have not a real push through in the iron and steel industry are (i) limited capacity of treatment facilities since the existing treatment facilities especially for desulphurization are normally sized to meet the needs of the converter cycle and additional treatments in these facilities lead to an increased load and hence the converter cycle can no longer, (ii) large temperature losses since every treatment step causes a temperature drop of the hot metal such as desulphurization injection, slag skimming, desiliconization injection, slag skimming, dephosphorization injection, and again slag skimming. In total, these accounts to more than 100 deg C temperature drop which frequently hinders execution of all steps because of a too low final hot metal temperature for the following BOF converter process. In particular, the last dephosphorization step frequently is not executed because of this reason.

Oxidation of carbon in a ladle process is especially critical because of the generation of {CO} gas and the excessive foaming of the slag. Hence, the oxidation of carbon during desiliconization and dephosphorization has to be limited and controlled to the best possible extend.

Kuttner has designed a process for desiliconization and dephosphorization of hot metals which overcomes the drawback of large temperature losses using chemical heating with gaseous oxygen. The generated heat is taken from the exothermic reaction of silicon with gaseous oxygen in contrast to the endothermic reaction of silicon with iron oxides. In order to control the increase of temperature during the removal of silicon to the desired extend, a balanced supply of gaseous oxygen and injected solid oxygen from iron oxides deep into the ladle is used. As the silicon has to be removed anyway for dephosphorization, this effect can be utilized to compensate the temperature losses to some extent.

This process needs the injection of a mixture of lime (CaO), and iron oxide. Some additives are used to control slag composition and excessive slag foaming. The needed reagents are injected as a premix in mono-injection mode or as single components using co-injection or multi-injection systems. At an incoming silicon content of 0.5 % and 1,325 deg C, the consumption of desiliconization reagent for achieving 0.15 % [Si] is 20 kg/t.

Phosphorous removal needs a very high mixing power density to achieve a low phosphorous content. A dephosphorization rate of 70 % (e.g., reduction of phosphorus from 800 ppm to 250 ppm) is achieved by injecting around 22 kg/ton of a dephosphorization agent. impossible to achieve sufficiently high mixing power with top addition and lance stirring or bottom plug stirring, injection of solids or mechanical stirring by an impeller are the preferred methods for the dephosphorization of the hot metal.

Production of clean steel needs removal of all undesired impurities especially sulphur and phosphorous. The removal of sulphur from hot metal is already being used by several steel plants but dephosphorization is traditionally carried out in the BOF converter by the majority of the steel plants. Since, this process is frequently already optimized to its limits, additional incoming phosphorous contents can only be handled by paying a penalty in terms of increased production cost and reduced productivity. Purchasing premium iron ores with low phosphorous contents is frequently the only way to avoid these issues.

On the other hand, premium low phosphorous iron ores are considerably more expensive than ores with higher phosphorous contents. Hence, economic benefits can be gained by introducing pre-treatment steps before charging the hot metal into the BOF converter. With respect to desiliconization and dephosphorization, small investments have frequently very short payback periods. By blowing pre-treated hot metal in the BOF converter, ultra-low phosphorous steel grades with contents of less than 50 ppm can be easily achieved.

Another benefit arises from a lean slag BOF converter operation which allows partial substitution of ferro-manganese with several times cheaper manganese ore. Also, the generated BOF slag can be utilized as desiliconizing agent or blast furnace feedstock. Pre-treatment of hot metal means basically shifting of anyhow necessary process steps from the steelmaking phase (BOF converter or secondary metallurgy) to the thermodynamically more favourable hot metal stage.

Since several already existing desulphurization shops based on the injection technology or mechanical stirring (KR) are suitable to be used also for desiliconization and dephosphorization, extension of desulphurization to triple-D (desiliconization, dephosphorization and desulphurization) is frequently just a short step.

The main drawback of the ladle-based hot metal pre-treatment technologies is the large temperature drop of the hot metal with the negative effects on the downstream operations. The concept of the Kuttner utilizes the exothermal oxidation of silicon with gaseous oxygen in order to generate heat needed to compensate the temperature losses at least partially. This process can be adopted into the production sequence of nearly all plant concepts and layouts and hence helps to create capabilities of processing cheaper raw materials.

Transport systems – Delivery and use of pulverized desulphurization reagents, e.g. magnesium, lime, calcium carbide, involve distinct technological needs, principally the avoidance of contact with air. Magnesium powder, produced by atomization or grinding, are to be transported in sealed, air-tight containers of limited capacity (20 tons each). Hence, to provide the capability to move this material in bulk, the industry developed a 90 % Mg – 10 % lime product which is flowable and can be delivered and stored in bulk trailers. Calcium carbide is also to be kept from the moisture in air and is transported and stored in bulk, sealed, pneumatic system equipped trailers. Salt-coated magnesium is relatively impervious to moisture and normally is stored in bulk trailers as well.

An important adjunct to facilitate the use of pulverized materials such as lime and calcium carbide has been the development of a technology for improvement of the flowability of these materials by application of silicone oil-based flow aids during pulverization. Powders prepared in this manner can be delivered by dense phase injection techniques, which minimize the quantities of reagent and iron droplets carried out of the liquid by the transport gas (e.g., for lime, transport line loading of 2 ton/cum of gas). This allows delivery of injection reagents at rates of 50 kg/minute through an 18 mm transport line.

Slag management – As in all metallurgical processes, management of the slag produced during hot metal desulphurization is critical to success. After conclusion of treatment, the slag normally is removed with a raking device, which typically is an articulated arm and paddle assembly. The raking process needs some time which can become a production penalty in some operations. Process yield suffers as some hot metal is lost from the ladle with each stroke of the paddle.

In some shops, the iron transfer ladle has a retention dam across the mouth with hole(s) for the metal to pour out. The slag is retained in the ladle as the hot metal is charged into the furnace. While effective at separating the slag and minimizing yield loss, this can slow the rate of charging the vessel and, hence, extend BOF converter heat cycle. Additionally, the slag retained in the ladle is to be dumped after each use, since this is done by reversing direction (to keep the pour holes open), this step may create difficulties in some shops.

During raking the post-treatment slag takes with it a considerable metallic content (around 40 % by weight), this can represent a yield loss of nearly 1 %. Methods to minimize this loss include the use of dense phase injection (to minimize the volume of gas for delivery of the injected powders), and the addition of a fluxing agent, 5 % fluor spar (CaF2) or Na2CO3 to the desulphurizing reagents, to produce a less viscous liquid slag for release of the iron globules. A further help is to provide a small quantity of gas bubbling during the raking process and this promotes flotation of the slag towards the lip of the ladle and thereby reduces the number of strokes needed for slag removal.

Typical figures for slag removal are in the range of 15 kg/ton to 25 kg/ton hot metal for several shops. Two viscosity related factors combine to make the quantity of slag raked from transfer ladles dependent on hot metal temperature. These are colder hot metal appears to ‘hold’ more entrained blast furnace slag and the retention of iron droplets in slags increases as temperature drops. Disposal of the spent slag normally is by mixing it into the blast furnace slag management system despite the remaining unused sulphur holding capacity. In some plants special controls are in place to cope with the effect of remaining unreacted reagents such as carbide or soda ash. This is not an issue with magnesium or lime.

Lance systems – The commonly used injection lance designs have nozzles which are directed vertically or exit the side of the lance at different angles. Typical life experience is 80 treatments or up to 1,200 minutes for the hockey stick design, which prevents the reactive gases from attacking the refractory coating. A lower figure, 70 treatments, is typical for the simpler lazy L lances. Lances typically are constructed of lengths of square steel tubing which contain the transport pipe(s). These assemblies are then cast within a refractory mould and cured in temperature controlled drying ovens. The cross-sectional shape can be square or round with an area of 40 square centimeters to 50 square centimeters. The refractory typically is high alumina.

A recent development directed at increasing the injection rate of magnesium containing combinations, without causing otherwise intolerable violence at the surface of the metal in the open top ladle or torpedo car, is to use two transport lines contained within a single lance each fed by a separate injection materials source. The outlet nozzles are positioned to deliver the reagent at 30 degrees from the vertical on opposite sides of the lance. It is also possible to use two separate lances immersed into the metal at the same time. These approaches have reduced injection time by 50 % without increasing iron losses to the ladle slag or reduction in reagent efficiencies.

Process control – Modern facilities are normally fully automated to control the operation from the start of treatment to finish. Powder initiation and lance insertion are automatically controlled along with lance withdrawal after the pre-determined quantity of reagent has been injected. Reagent rates can be adjusted during treatment by controlling injection tank pressure or transport line throat diameter.

In addition to the automatic functioning, the controlling computer provides for process data capture and storage. Real-time graphics provide information on process efficiency and dispenser performance to aim standards. Typical, for a well-controlled operation is to achieve within 0.002 % sulphur of aim at end points ordered less than 0.003 % sulphur.