THE USE OF TONNAGE OXYGEN IN STEELMAKING

BY E. KLEIN, B.Sc.(Eng.), M.S., AND K. O. R. GEBHARD, Dr. Eng., (Visitors) - The Certificated Engineer March 1965.

ABSTRACT

The role of oxygen in steelmaking. Application of tonnage oxygen in steelmaking. Surface blowing and penetrating blowing. The L-D process, its development, and characteristic features. Carbon exploitation in the L-D process. The utilization of L-D waste gases. The Kaldo process, its principles and advantages, and problems. Higher scrap consumption than in the L-D process but lining life still not satisfactory. The principles of the Graef Rotor process. Experience with Iscor's Graef Rotor plants and its development to the Iscor Rotor process. Advantages and disadvantages of the Iscor Rotor process. Low carbon exploitation. The Tandem furnace process a logical development of the Iscor Rotor process. Carbon exploitation in two different units.

The use of oxygen in open-hearth furnaces. Roof lancing. The Ajax furnace. The problematic use of oxygen in a hearth furnace. The impact of oxygen steelmaking on the scrap market. The survival of the open hearth shop. Economic aspects. The factors influencing the economy of oxygen utilization.

1. INTRODUCTION

The electrical and mechanical engineer outside the steel industry has little opportunity of coming into contact with steel-making processes. By the nature and responsibilities of his work, he has even less chance of getting acquainted with the background of many of the new processes which have come into being by the use of tonnage oxygen. This paper gives a survey of modern trends in this field. While most engineers have read references to these new processes in the popular and technical press, we have attempted to give as comprehensive a picture as possible of this field.

 2. THE ROLE OF OXYGEN IN STEELMAKING

The term oxygen steelmaking has become common and it is meant to describe the use of oxygen of high purity in steelmaking. It should not be overlooked that even in the days before the production of oxygen of high purity in air-separation plants became an economic proposition, oxygen was used in steel-making. As a matter of fact, unless 100 percent scrap is used, steel cannot be made without the use of some form of oxygen. Broadly speaking, pig iron or hot metal is converted to steel by reducing its carbon content and other oxidizable impurities, like silicon, manganese, phosphorus, and sulphur by oxidation and by simultaneously increasing the bath temperature from approximately 1 250°C to 1 600°C.

The bulk of the steel made in the world is produced in open-hearth furnaces. The source of oxygen in this process is from iron oxides, i.e. ores and the oxygen content of the combustion gases above the bath. It requires an external source of heat to liberate the oxygen in the ore and to raise the temperature of the bath. One can well imagine how startled Sir Henry Bessemer's contemporaries were when at a meeting in Cheltenham in 1856 he announced that he had made steel without fuel. His ingenious invention was based on the recognition that the carbon, silicon, manganese, phosphorus and sulphur contained in hot metal are fuels and that, when these fuels are burnt with air, sufficient heat is generated to raise the temperature of the molten iron to the required tapping temperature of the steel. Figure 1 shows the heat balance of the basic Bessemer process. As is known, in this process, iron with a high phosphorus content is refined. A Considerable amount of heat is carried away by the waste gases which contain large quantities of nitrogen and carbon monoxide. The heat lost by radiation and from other causes amounts to approximately 4 percent and is not shown on the heat balance. The amount of scrap which can be charged is small and depends on the analysis of the hot metal.

Sir Henry Bessemer was aware of these short-comings of his process and in 1856 took out a patent on the use of pure oxygen instead of air or his pneumatic process. It was his misfortune that he was ahead of his time by approximately 80 years. It was only in 1932 that, because of another development, namely the air separation by the Linde-Frankie process, the production of tonnage oxygen became an economic proposition and thus the era of oxygen steelmaking commenced.

Figure 2 shows the world production of steel by various processes. The greatest amount of steel is still produced by the open-hearth process, but oxygen steelmaking is growing rapidly. Oxygen steelmaking will continue to increase over the years to come and we will discuss later certain problems that the wide application of oxygen steelmaking has created and which are still unsolved. The impact that oxygen steelmaking is having on the open hearth operations will also be considered.

Figure 3 shows a survey of the steelmaking processes in which tonnage oxygen is used.

The pneumatic processes can be divided into two groups which are characterized by the method in which the oxygen is applied. These are the surface blowing processes and the penetrating blowing processes. This grouping is based on the basic principles of the processes. In practice, a certain amount of overlapping might occur. In their original concepts, the LDAC process, its modification designed to refine iron with a high phosphorus content, the LDAC process, and the Kaldo process, are surface blowing processes. Of these, the L-D process has had the biggest impact on steelmaking. The Graef rotor process and the Iscor rotor process, which was developed from it, and also the Tandem furnace process, rely on the penetration of the oxygen jet. There is as yet no Tandem furnace plant in operation, but the first Tandem furnace plant should be commissioned in a few months' time.

In order to understand the sub-division into the surface and penetrating blowing processes, one has to study the history of the development of the L-D process.

 3. THE L-D PROCESS

Valerian Schwarz took out a patent in 1939 on oxygen steelmaking in which the oxygen was to be injected into the hot metal as a jet at a high velocity so that it penetrated into the bath like a solid body. Schwarz's ideas were taken up by Durrer, Professor of Metallurgy at the Technical University at Berlin. The first trials were made in Berlin and continued after the war at van Roll'sche Eisenwerke at Gerlafingen, Switzerland. The Vereinigte Ooestereichische Hiittenwerke at Linz in Austria (VOEST) then became interested and it was at Linz that the first successful heats were made using oxygen of high purity as a refining agent.

This process, known as the 'Linz-Dii en Verfahren,' or L-D process, has gained a well-deserved world reputation. In a VOEST publication 'Three Years L-D Steel'(1), the development of oxygen blowing from "1902 to 1950 is shown (see Fig. 4). None of the various methods that were tried to introduce oxygen as a solid body into a bath of molten metal succeeded. It was only after the idea of thoroughly mixing oxygen and metal was discarded that a final breakthrough was achieved. It was found that hot metal could be refined at a fast rate provided the oxygen did not penetrate the bath but impinged only on the surface. The pressure at which the oxygen jet strikes the surface should not exceed 0.75 Kg/cm2 (2). H. Hautmann states that only about 3 percent of the total surface of the bath is directly exposed to the oxygen jet(1) and that at this ignition spot temperatures of 2 000°C to 2 500°C are generated. The results obtained under these circumstances, i.e. the limitation of the contact between oxygen and bath to a small surface area, were unexpected.

From a small beginning in l S-ton and 30-ton vessels, L-D vessels of ever-increasing capacity have been built. Vessels of 300-ton capacity are now in operation. The blowing times are in the order of 30 to 40 minutes. Including the auxiliary time, for charging, sampling, and tapping, etc., tap to tap times of under one hour are achieved regularly, resulting in production rates of 300 tons per operating hour. Normally at least two vessels are installed, one being on the line, whilst the other one is being relined. The initial difficulties in respect of lining wear have been overcome and lining life is long enough so that breaking out, relining, and heating up of the spare vessel is completed while the other vessel is still blowing. The availability of the vessel is high and an L-D plant consisting of two 300-ton vessels can produce up to 2 500 000 ton/annum.

Figure 5 shows the cross-section of an L-O vessel having a nominal capacity of 250 tons. The bath level is indicated. As can be seen, the ratio between total capacity and the bath volume is rather high and a figure of 35 feet^3/ton capacity is quite common. In order to facilitate relining, modern vessels are equipped with a removable bottom. A recent development uses a transport car, which shifts a worn-out vessel to a relining stand and the newly relined vessel to the blowing stand. Thus, for the one vessel on the line operation less fixed equipment is required. There is one blowing stand only, which is nearly always in operation. Up to now, it was common practice to have two blowing stands, one of which was in operation at a time, and a worn-out vessel was relined in situ in the second stand. Thus part of the fixed equipment was in use for only half the available time The large space above the bath is required because of some special features of the L-D process. At first glance, the high rate of refining obtained is difficult to explain because the oxygen impinges on 3 percent of the total surface area only and does not penetrate appreciably into the bath.

In a paper published in 1960, E. Piockinger(3) and co-workers studied the behaviour of an oxygen jet in the air. Fig. 6 shows how a jet aspirates air. At a distance of 800 mm from a 10 mm diameter nozzle, its average oxygen content has decreased to 32 percent. In other words, under the test conditions, the jet has aspirated approximately six times more air than its original volume. At a blowing rate of, say, 6 000 feet^3 36 000 feet^3 of air would be aspirated. An oxygen jet in an L-D vessel will act similarly and Plockinger suggests that a flow pattern as shown in Fig. 7 will develop in an L-D vessel. He points out that the aspirating action of the jet will have the following results:

• Due to the carbon monoxide which evolves constantly from the bath, the atmosphere above the bath consists mainly of carbon monoxide. The oxygen jet aspirates part of this carbon monoxide, burning it to carbon dioxide. As can be calculated, the combustion of carbon monoxide, which has a temperature of approximately 1 500°C, with pure oxygen generates flame temperatures of approximately 2 800°C. This temperature could be even higher if the carbon dioxide formed was not dissociated partially at these temperatures. Depending on physical Conditions, such as lance distance, jet velocity, etc., a mixture of oxygen, carbon monoxide and carbon dioxides of high temperature impinges on the bath surface. The extremely high ignition spot temperatures, which have been mentioned previously, can thus be explained without assuming that it is generated only by the combustion of iron.

It is of interest that refining by a hot mixture of oxygen, carbon monoxide, and carbon dioxide is the subject matter of a U.K. patent.(4)

How is it possible that, in view of these extremely high temperatures, the good life of the lining is obtained in L-D vessels? We think that a statement made by H. Hautrnann(1) and the specification of the VOEST U.S.A. patent(2) on the L-D process, offer an explanation. This patent specification states that during the first minutes of the blowing time, an emulsion of slag, iron oxide, and iron particles is formed. This emulsion, which according to reports on L-D vessel operation, sometimes reaches the mouth of the vessel, rises on the walls of the vessel. H. Hautmann states that there is a temperature difference of approximately 1 000 degrees C between the slag near the ignition spot, and at the walls of the vessel. The abovementioned emulsion thus shields the walls from the high temperatures prevailing at the ignition spot. It is observed that one of the conditions for obtaining good lining life is a central and vertical position of the jet. This supports the theory outlined above.

It was mentioned that a 30U-ton L-D vessel produces at a rate of 300 tons per hour from tap to tap, i.e. over 7 000 tons per day. These high production rates call or special care in the planning of the material handling facilities. Special scrap charging facilities have been developed. An interesting design is the Calderon scrap charging machine (see Fig. 8). This machine fulfills two tasks. Assorted scrap is charged into a chute of special design. When this chute is tilted the scrap pieces, which are positioned at random, align themselves whilst sliding down and are discharged to the feeding chute parallel to each other.

It takes a few minutes only to charge the required 20 percent to 30 percent scrap in modern L-D plants using either Calderon machines or other suitable facilities. Fluxes are normally charged from hoppers positioned above the mouth of the vessel.

Figure 9 shows the cross-section through an L- D plant. The tall building which houses the lance-holding device, the flux bunkers, and the waste gas flues, including the waste heat boilers, which are common in most plants, is typical. The waste gases of the L-D process contain 80 percent to 90 percent of carbon monoxide. The theory of Plockinger that part of the carbon monoxide evolved from the bath is burnt by the oxygen jet to carbon dioxide can be reconciled with a high carbon monoxide content in the waste gas. One molecule of oxygen can burn 2 molecules of carbon monoxide to two molecules of carbon dioxide according to:

O2 + 2CO = 2CO2

The reaction of 2 molecules of carbon dioxide with the carbon in the bath yields, however, 4 molecules of carbon monoxide according to:

2C02 + 2C = 4CO

Therefore, twice the amount of carbon monoxide is generated than can be burnt by the oxygen supplied. Half the volume of the carbon monoxide evolved from the bath has to leave the vessel as waste gas. This hot waste gas, rich in carbon monoxide, represents a valuable source of heat. As is known, the combustion of 1 lb of carbon to carbon monoxide generates 4 000 Btu and to carbon dioxide gen rates 14 100 Btu, i.e. the combustion of carbon to carbon monoxide releases approximately one-third only of the potential heat. A heat balance on the L-D process published by Rinesch(5) shows (see Fig. 10) that nearly 40 percent of the total heat involved leaves the vessel in the form of hot gas, rich in carbon monoxide. It is quite common to recover part of this heat in waste heat boilers. The combustion of the carbon monoxide in the waste gases with air increases considerably the final volume of the waste gas. At a blowing rate of, say, 6 000 feet^3/min the gas cleaning plant has to handle approximately 150 000 feet^3/min. The cleaning of the gas is therefore expensive. It is for this reason and because steelmakers have a certain aversion to operating a plant of which they are not sure, whether it be a boiler house or a steelmaking plant, those other solutions were suggested and adopted. One consists of a device that prevents the infiltration of air between the mouth of the vessel and the waste gas duct. Thus after-combustion of carbon monoxide to carbon dioxide in the waste gas duct is suppressed. In Japan, a plant is in operation in which waste gas containing carbon monoxide is recovered and is stored in a gasometer for further use.(6) Approximately 50 percent of the potential heat can be recovered by this cold method as compared with 75 percent in the hot direct recovery in waste heat boilers.7 The recovered and stored gas contains about 74 percent carbon monoxide and it has a dust content of .22 grains per feet^3. Cleaning is therefore required when the recovered gas is drawn from the gasometer before it can be put to further use.

4. THE KALDO PROCESS

It was mentioned that the scrap consumption in the L-O process is in the order of 20 percent to 30 percent. The amount of scrap which can be charged depends on the metalloid content of the hot metal, especially in regard to silicon and phosphorus. The L-O process is, therefore, able to consume the scrap which arises in the processing of the ingots produced in an L-O plant. As is known, the amount of scrap accumulating in old established industrial countries is so high that in open-hearth furnaces, which are still the main source of ingot production, 50 percent, and more scrap is charged. Even now, the steadily widening application of the L-O process has had an adverse effect on the overall consumption of scrap in steel production. Several suggestions have been made to increase consumption in the L-O process. Rinesch(5) proposed either to add fuel during the blowing period or to preheat the scrap. The addition of fuel during the blowing period is, however, of doubtful value. The fuel utilization would below. As mentioned, the waste gas of the L-O process consists mainly of carbon monoxide. Thus, the carbon contained in the fuel could burn to carbon monoxide only and its hydrocarbons would be mainly cracked because water is not stable in an atmosphere containing carbon monoxide at 1 500°C to 1 600°C. K. Stone(8) therefore rejects this proposal and also advocates preheating of the scrap. If the scrap is preheated in the vessel before blowing commences, the rate of production would decrease drastically and one of the attractive features of the L-O process, namely the high rate of production, would suffer. Stone states that 'preheating of scrap in the L-O process represents some unsolved problems, especially with equipment.' Some years previously B. Kalling(9) in Sweden and R. Graef(10) in Germany, realized almost simultaneously but independently that the consumption of scrap in an oxygen steelmaking process could be increased if the heat from the combustion of carbon to carbon dioxide could be utilized fully within the vessel. They suggested burning the carbon monoxide evolved to carbon dioxide above the bath. The waste gas of such a process would then consist mainly of carbon dioxide.

The process developed by Kalling has become known as the Kaldo process and Graef's process is known as the Graef-Rotor process In the Kaldo process oxygen is blown at an angle on to the surface of a bath of hot metal, which is contained in a converter type vessel. As can be seen in Fig. 11 the vessel normally works in an inclined position but it can be tilted into various positions for blowing, charging, adding of fluxes, etc. It can also be rotated. The oxygen lance is inserted through the mouth of the vessel and the waste gases are withdrawn through the same opening. That portion of the oxygen not directly absorbed by the bath is deflected and burns the carbon monoxide evolved to carbon dioxide above the bath. The refining action is further enhanced by the rotation of the vessel. The vessel is built to rotate at up to 30 revs/min and thus slag and steel are brought into good contact

with each other. The temperature of the slag and steel and the rate of refining can be controlled by the angle of blowing and the rate of blowing.

At present eight Kaldo plants are in operation gild two more are planned. To enable continuous operation, the plants have either two blowing stands, which are used alternately or one blowing stand with interchangeable vessels. These plants represent a remarkable achievement in mechanical design. It is not an easy matter to design a plant in which a vessel weighing approximately 600 tons must rotate at 30 revs/mill, and at the same time be able to be tilted and which must be able to be transported from the blowing to the relining stand. Judging by the reports published on the operation of Kaldo plants, no serious trouble, have occurred either in respect of the metallurgical aspects or in the mechanical design. Up to 45 percent of scrap is charged regularly and iron with a high phosphorus content is refined to the steel of good quality. The yield of iron is high. However, the position in respect of the con sump ion of the refractory lining does not appear to be satisfactory. In a paper, published in the Journal at the Iron and Steel Institute, Kalling(11) stated that a contributing factor to the rapid wear of the lining could be the high temperatures generated by the after-combustion of carbon monoxide above the bath. In a more recent publication(12) hope is expressed that as more experience is gained in operation and with better refractories available, the life of the lining will improve.

5. THE GRAEF ROTOR PROCESS

Both Kalling and Graef published their first reports ill 1957. The aim of both processes is similar and in both processes, the carbon monoxide evolved from the bath is burnt to carbon dioxide above the bath.

The processes differ, however, not only in the equipment used but also in the manner in which the oxygen is introduced. The failure of the tests which preceded the L-D process had brought refining using a deep penetrating jet of oxygen into disfavour. Thorough intermixing of bath and oxygen was considered to be detrimental. Graef took the bold step of discarding these opinions. The principles of his process are depicted in Fig. 12. Primary oxygen is introduced from a submerged lance parallel to the bath surface and secondary oxygen is blown above the bath. Thus the carbon monoxide generated by the action of the primary oxygen is burnt to carbon dioxide by the secondary oxygen. By changing the ratio of primary to secondary oxygen and by adjusting the purity of the secondary oxygen from between 21 percent to 95 percent, the heat generated above the bath and the temperatures of the bath and slag were able to be controlled. The temperature of the lining above the bath could be expected to be high, but it was thought that the submersion of the hot portion of the lining into the bath during the rotation would not only cool the lining but also transfer heat to the bath.

Iscor operates two Rotor plants, one at the Vanderbijlpark Works and one at the Pretoria Works. The Vanderbijlpark plant was commissioned in November 1959, and the Pretoria plant in July 1960. It must be realized that in most oxygen blowing processes the final temperature of the steel is controlled by the addition of coolant in the form of scrap or ore.

The Graef Rotor process was still being developed when Iscor decided to install two plants. It might be of interest to know why this decision was taken.

Firstly, the possibility of utilizing fully the heat of combustion of the carbon contained in the hot metal was appealing. It opened the prospect of charging such a high percentage of ore that approximately 7 percent of the ingots produced would have been made directly from this source. This would have been an economic proposition despite a higher consumption of oxygen. In view of the shortage of scrap in the Republic, high consumption of ore was thought to be a necessity. Secondly, at that time the Rotor process was the only oxygen process that could handle heats of 100 tons tapping weight and this was the minimum size of heat required to tie in with the capacity of the soaking pit. Thirdly, it was thought that the cleaning of the waste gases issuing from the Rotor process would be easier because of the larger particle size than was the elimination of the brown fumes from the L-D process.

We were aware that the life of the lining was an unsolved problem and that considerable research would have to be carried out in this field. Before operational aspects of the Rotor process are discussed, a short description of the plant is given.

Figure l3 shows a cross-section through a Rotor plant. The vessel rests on a cradle which is mounted on a rotating table. The vessel can rotate in the cradle and simultaneously can be tilted for tapping and for rolling out slag. It can be slewed into any position required for charging, blowing, tapping, and sampling. These positions can be pre-selected. All the controls are housed in a pulpit adjacent to the vessel. Primary and secondary oxygen lances are mounted on a lance car and by slewing the vessel through 180 degrees either end of the vessel can be the blowing end of the waste gas end. Ore and fluxes are fed by a charging machine operated from the floor and molten metal is fed via a launder. No facilities were provided for the charging of scrap. On each side of the blowing stand is a relining stand (see Fig. 14). There are two vessels for one blowing stand. A vessel change is carried out as follows:

The worn-out vessel is tilted into the vertical position in the cradle and rests on the horns of the cradle. By means of hydraulic jacks, which are mounted on the transfer car, the vessel is transferred to this car which then moves to the relining stand on to which the vessel is lowered. The transfer car can now pass underneath the cradle and is available to pick up the

relined and preheated vessel from the other relining stand. This vessel is transferred to the cradle and brought into the horizontal position. The vessel is raised to the operating temperature with the aid of a powerful tar-oxygen burner which also serves as a fettling burner. A vessel can be changed over in 7 to 9 hours.

Each plant was laid out for an annual production of 300 000 tons and is served by a 100 ton per day oxygen plant. There are quite a number of gears, pinion, motors, drives, controls and interlocks in the Rotor plant, to the delight of the engineers but to the dismay of the operator. So far, however, the plant, intricate as it is has not given any serious trouble.

It is beyond the scope of this paper to give a detailed report on our operational experience. This has been done in previous publications.(13) Our experiences were sometimes quite exciting and not always encouraging, but we succeeded in producing the scheduled 25 000 ton/month after a few months of operating. As expected, the wear of the lining was the greatest problem. The metallurgy of the process did not cause undue difficulties. We had to revert to fettling in order to keep one vessel in the blowing stand long enough to enable breaking out, relining, and preheating of the spare vessel to be completed. An effective fettling, the procedure was developed which ensured continuous operation. We realized. of course, that valuable production time was lost by fettling. The consumption of refractories and oxygen increased above our original concept. Our aim was to obtain a life of 201 to 300 heats from the lining without fettling. It soon became evident that there was little prospect of reaching this target as long as we operated according to the original concept of the Graef Rotor process. The after-combust-on of the carbon monoxide caused the lining to be attacked severely. The cooling of the lining by contact with the colder bath during rotation was so limited that the life of the lining was improved. We had to give up the idea that the heat of combustion of the carbon in the hot metal could be fully exploited.

6. THE ISCOR ROTOR PROCESS

The modifications in operational procedure introduced over the years have resulted in such drastic changes in the Graef Rotor process that, for the sake of clarity, the term Iscor rotor process might be used for our present operational practice. This practice is characterized by the following features:

No secondary oxygen is used normally, i.e. the after-combustion of the carbon monoxide is suppressed. Primary oxygen is injected at supersonic speed from a lance, which is positioned at an acute angle near the surface of the bath. The rate of blowing has been increased from approximately 1 200 feet^3/min to approximately 5 000 feet^3/min. As a coolant ore has been replaced by scrap. Due to the lack of proper facilities for charging scrap, only 5 percent of scrap can be charged at present. Further, we have replaced the original lance by one of a new design. This lance reciprocates and reaches to about the middle of the vessel. This has resulted in a better life of the refractory material. Fig. 15 shows how the middle portion of the lining was hardly consumed when the original lance was used. With the new lance wear of the lining is more uniform. A life of 250 to 350 heats is obtained regularly from the lining without fettling. It should be mentioned that the quality of the refractory material supplied has also improved considerably. Magnesite of high purity and especially the tar-impregnated bricks have given encouraging results and the best performance to date is 464 heats.

In September 1964, the Pretoria rotor plant produced 53,600 tons, which is 113 percent more than scheduled. Only 100 tons of oxygen is available per day. With this limited amount of oxygen available a production rate of 1 780 tons per day was possible because the oxygen consumption is below 1 400 cubic feet per ingot ton. Approximately 350 feet^3/ton are supplied from air drawn in at the blowing end. We are confident that the production rate will be increased considerably when the additions and modifications now in hand, come into operation. These additions are:- A further oxygen plant, faster scrap charging facilities, and the strengthening of the cranes so that a tapping weight of 150 tons can be handled.

We have established that under our conditions, it is more economical to lose the heat in the waste gas containing excessive carbon monoxide than to try to recover this heat within the vessel by burning the carbon monoxide. However, we have not lost sight of the fact that we are dissipating valuable heat in the waste gas line by injecting water into it. The heat loss is equal to the heat input to a 200-ton open-hearth furnace. This is the main reason why we have decided to shelve plans for the construction of another Rotor plant and to embark on the design and construction of a Tandem furnace.

7. THE TANDEM FURNACE

The idea of utilizing the waste heat of our Rotor process by recovering the waste gases, as is being done successfully in Japan, did not appeal to us. We felt that this potential source of heat should be used, if possible, directly to preheat scrap. Calculations, backed by tests in the Rotor plant, have shown that input of 40 percent to 45 percent of scrap should be possible provided the scrap could be preheated to approximately 900°C.(14) These figures agree well with the calculations of Stone.(8) The amount of carbon monoxide issuing from the oxygen refining of a bath containing 55 percent to 60 percent of hot metal would suffice to preheat 40 percent to 45 percent of scrap to 900°C. The problem was how to do it. Installing two vessels end to end, i.e. to replace the present waste gas line by a second vessel did not seem an attractive proposition. Such a plant would be complicated mechanically and its operation would be intricate, mainly because the only access to the rotating vessels would be from the ends.

It will be remembered that the original Rotor process had to be carried out in a rotating vessel in order to transfer heat via the lining to the bath. This did not work as contemplated. In the Iscor Rotor process, the after-combustion of carbon monoxide above the bath is suppressed. The result is that the temperature of the lining above the bath is lower than the temperature of the steel at tapping. The need to operate in a rotating vessel falls away if such a temperature distribution prevails.

The concept of a tandem furnace is then possible. We have designed a tandem furnace. Fig. 16 is a photograph of a model of the tandem furnace of 120 tons tapping weight, which should be commissioned at the Vanderbijlpark Works towards the middle of 1965.

The two furnaces are joined by a throat. Each will function alternately as a refining furnace and as a preheating furnace, the reversal taking place when the charge in the refining furnace has been completed. The oxygen will be introduced into the refining furnace using a lance similar in design to the lance which is in use in our Rotor plants. The method of blowing will be the same, i.e. the after-combustion of carbon monoxide to carbon dioxide above the bath will be suppressed. It can be expected that the metallurgy of the refining furnace will not differ from the metal-lurgy in our Rotor plants. The carbon monoxide-rich waste gas is passed through the throat into the preheating furnace. Slag from the previous charge will be retained in this furnace and oxygen will be injected through two lances which will be positioned in the preheating furnace adjacent to the throat and the hot waste gas thus burnt to carbon monoxide. The most favourable position of these after-combustion lances was determined by experiments in models. Fig. 17 shows how these tests were carried out. The waste gas is simulated by the water of a certain acidity and the oxygen is stimulated by potassium permanganate. The concentrations of the two liquids are so adjusted that the combustion is completed to 60 percent at the point where the red colour of the permanganate disappears. The flame will reach temperatures of approximately 2 800°C and it will be luminous due to the dust content of the gas. The effect of the high flame temperature on the lining should not be too severe, however, because firstly, the roof and walls are positioned at a considerable distance from the flame, and secondly, a large proportion of the heat liberated will be absorbed by the cold scrap and fluxes. The large temperature difference between flame and scrap should favour the efficient transfer of heat. Hot metal will be added to the preheated scrap when the heat in the refining furnace nears completion. The direction of the flow of gas is then reversed so that the preheating furnace now becomes the refining furnace and vice versa.

It is intended to charge scrap to the preheating furnace while blowing proceeds in the refining furnace. Sampling, temperature measurement, the flushing off of slag, and the addition of fluxes should also be possible during the blowing period. Thus we expect that the interval between alternative blowing periods will be short. The ratio of blowing time to auxiliary time should be high.

The gas cleaning plant consists of a single-stage venturi washer and is relatively small because the waste gas is not burdened with a large amount of nitrogen. The design of the gas cleaning plant was based on tests carried out in a pilot gas cleaning plant which was installed at the Rotor plant.

This short description of the anticipated operation of the Tandem furnace sounds simple and we may seem over-confident about its success. We have confidence in this venture, but we do not hesitate to admit that we do not know all the answers. Unforeseen difficulties might temporarily block our way to final success.

Provided we succeed in developing our prototype Tandem furnace process to a technically and economically sound proposition, one could still ask if there is a need for a new oxygen steelmaking process. The L-D process and the Kaldo process have been established and, as far as Iscor is concerned, the Rotor process is successful. Do they not meet the demands of the iron and steel industry fully?

We think that there is room for another process that combines a high scrap and/or ore consumption with the simplicity of equipment and high production rates. Further, a Tandem furnace could be housed in a building similar to an open hearth shop, and would not take more space than that occupied by a 150 to 200-ton open-hearth furnace but would replace the production of several furnaces.

 8. THE USE OF OXYGEN IN OPEN HEARTH FURNACES

The dominating position which the open-hearth process has occupied as the main producer of steel for many years is being strongly challenged by the unpredicted success and rapid growth of oxygen steelmaking. Long before pneumatic oxygen steelmaking was invented, oxygen was used in open-hearth furnaces. The rate of carbon removal was increased towards the end of heat by injecting oxygen into the bath through consumable lances, which were inserted through the doors. The rate at which oxygen could be supplied by this method was limited. New methods were developed. They have to lead to roof lancing, on the one hand, and the Ajax furnace on the other hand. Further, oxygen is being used to enrich the air for combustion, sometimes in combination with roof lancing. The main field for the use of oxygen in open-hearth furnaces can be covered, we think, by a short description of roof lancing and of the Ajax process.

In quite a number of open-hearth furnaces, oxygen is introduced through lances, which are inserted through the roof. Fig. 18 shows a typical roof lance arrangement. Normally a multi-hole lance is used.

The speed of refining and the rate of production can be increased considerably by roof lancing.

An Iscor's Vanderbijlpark Works roof lancing was tried. We found that over 60 tons per hour could be produced from charge to tap if simultaneously the amount of scrap charged was decreased to 25 percent. We found further, that the attack on the lining was severe. Fettling, and hot and cold repairs required so much time that the production rate per calendar hour decreased to 32 tons. This rate of production has been achieved without roof lancing and with a low input of scrap. Wear on the roof is less severe in bigger furnaces in which the roof is positioned higher above the level of the bath than in smaller furnaces. It is known that a 500-ton open-hearth furnace at the Steel Company of Canada produced regularly in the order of 100 tons per hour tap to tap.(15) The consumption of scrap in this furnace is in the order of 45 percent and approximately 1 100 feet^3 of oxygen is used per ingot ton. The fuel consumption is approximately 2.0 x 106 Btu/ingot ton. This is less than half the fuel used without oxygen but the consumption of oxygen is high when compared with the 1 400 feet^3/ton used in Iscor's Rotor plants.

Oxygen refining in an open hearth furnace was approached in quite a different way by Jackson.(16) Fig. 19 shows the Ajax furnace. Oxygen is introduced through an inclined lance. The furnace is operated similarly to an open hearth furnace, i.e. the furnace is fitted with regenerators and fuels are used. The 200-ton furnace produces 28.5 tons per hour tap to tap from a 100 percent hot metal charge. The oxygen consumption is 1 450 feet^3/ton, which is practically the same as in the Iscor Rotor plants. Some 0.95 x 10^6 Btu are used per ton. The phosphorus content of the hot metal is 1.1 percent and it is possible that this phosphorus content justifies the operation of his furnace.

Looking broadly at the various methods of oxygen application in open-hearth furnaces and at the results obtained, one gains the impression that, in an effort to survive, open-hearth furnace operators have been forced to operate their furnaces in a manner for which the furnaces were not designed. In the old open-hearth furnaces, for example, the distance between roof and level of the bath is determined by the necessity to keep the flames down onto the charge and in order to benefit from the radiation from the roof. Experience has shown that the life of the roof improves considerably if the roof is raised in furnaces using oxygen for refining. It is not inconceivable that the consumption of fuel might increase in furnaces which have a higher roof. It remains to be seen whether a compromise in design can be found which will satisfy the somewhat divergent demands on the furnace which are-to function as a hearth furnace and, simultaneously, to function as an oxygen refining vessel.

One should not only view the attempts to boost open-hearth production with the aid of oxygen as a matter of prestige. It should be borne in mind that an enormous capital is invested partly in the open hearth furnaces but mainly in the buildings, cranes charging equipment, and ancillary apparatus required.

Serious disruption of the scrap market can arise if the op n hearth process is replaced by the L-D process. The demand for scrap would then decrease and the price of it might become so low that its collection might not be profitable. As a solution, one could think of the use of 100 percent scrap in big arc furnaces or in hot blast cupola furnaces for the conversion of scrap to hot metal. This would require new installations and still make open-hearth plants obsolete.

Naturally, our thoughts on this problem move along different lines. Should we succeed in developing the Tandem furnace into a technically and economically Sound oxygen steelmaking unit, then we think this process should fill the present gap. As mentioned, the Tandem furnace could be housed in many existing open-hearth shops, and buildings, cranes and charging equipment would not become obsolete.

9. ECONOMIC ASPECTS

This survey has shown that tonnage oxygen can be used in many different ways in steelmaking. Which is the most economic method of using oxygen in steelmaking? It is not possible to give a clear-cut answer to this question. There are too many factors to be considered and the importance of such factors varies from country to country. For example, one can compare the capital investment cost per ton per year of plants to be built on a new site. Processes with a high hourly rate of production will show up more favourably than, say, the modified open-hearth processes.

Such a comparison does not take into account the losses, which will occur if existing plants have to be demolished to make way for the new, or the savings that could be made if part of an existing plant is to be used.

In the pneumatic oxygen steelmaking processes, the consumption of oxygen per ingot ton does not differ significantly. It should be borne in mind that the cost of oxygen is, in fact, a fixed cost, because the oxygen plants cannot be shut down when little or no oxygen is required. Continuity in operation in oxygen steelmaking processes is of greater importance than in, say, the Bessemer process.

If fuel and oxygen are used, the calculation becomes even more involved and the relative cost of the two items and the influence of their consumption on the rates of production has to be studied.

Concerning man-hours per ingot ton, high rates of production are, of course, an advantage. The consumption of the refractory lining influences the cost of production directly and indirectly. A high consumption rate of the refractory can mean prolonged downtimes and, as we have experienced in our Rotor operation, discontinuity in production, which is expensive because the oxygen plant produces at a constant rate.

The greatest influence on the cost of producing steel is the input cost. This amounts to 70 percent to 80 percent of the total cost. The input cost depends not only on the mix, i.e. the ratio of hot metal to scrap, but also on the absolute cost of these two items, and their relative cost. Kalling(12) states that the cost of conversion in the Kaldo process is higher than in the basic Bessemer, i.e. the Thomas, process. Nevertheless, ingots can be produced more cheaply by the Kaldo process than by the Thomas process. The reason is that the Kaldo process consumes more scrap and scrap is cheaper than Thomas iron. A change in the ratio of the cost of iron to the cost of scrap has a great influence on the economy of a steelmaking process. The flexibility of the good old open heart process whereby the amount of scrap that can be charged can be varied has enabled the open hearth operator to take advantage of fluctuations in price. 

10. CONCLUSION

For a long period the pneumatic steelmaking processes, in which air was used as the refining agent, have played an important role in the mass production of steel. They suffered from two drawbacks. The nitrogen burden of the air lowered the heat economy of the processes and only small amounts of scrap could be consumed. Further, the high nitrogen content of the steel produced to a certain extent limited their use.

Despite the high production rates and low production costs of the pneumatic steelmaking process, they were replaced to an ever-increasing extent by the open hearth and electric arc furnace processes.

When high purity oxygen became available in large quantities, oxygen could be substituted as a refining medium. It was, however, not possible just to blow high purity oxygen instead of air when using the existing plant and equipment. New methods of bringing oxygen into contact with the hot metal had to be developed. New equipment was required.

New melting facilities, which were robust as well as delicate, were developed with the able assistance of the electrical and mechanical engineer. The oxygen processes, in use now, produce high-quality steel at rates hitherto unheard of. The development in this field has not yet come to an end. As in the past, further progress will depend, to a large extent, on the co-operation between the electrical and mechanical engineer and the metallurgist.

List of References

1. Three Years L-D Steel; VOEST 1956 p.30.
2. The U.S.A. Patent 2741 555.
3. E. PLOCKINGER ET AL. Stahl und Eisen 80 (1960) p.1477/86.
4. U.K. Patent 944479.
5. R. RINESCH-journal of Metals, July 1962, p.497-501.
6. M. YUKAWA AND K. OKANIWA. Iron & Steel Engineering. Dec. 1962, p.141-147.
7. A. PEARSON AND L. SHORE-Steel Times, 8th May 1964, p. 608-612.
8. K. STONE, Iron & Steel Engineer, June 1963, p.67-78.
9. B. KALLING, Stahl und Eisen 77 (1957) p. l308-13J5. (1964) p. 181-190.
10. R. GRAEF, Stahl unl Eisen 77 (1957) p, 1.
11. B. KALLING & F. JOHANSSON, Journal Iron & Steel Institute, 1959, Aug., p. 330-334.
12. F. JOHANSSON AND B. KALLING. Stahl and Eisen 8-1
13. Congres International Sur les Acieres a l'oxygene, Le Touquet 23-25 September 1963. Collective report presented by DR. C. M. KROGER, p, 62-73.
14. F. BARTU. Steel Times, 15th March 1964, p.646-651.
15. A. K. MOORE, Steel Times, June 26, 1964, p. 858-865.
16. A. JACKSON. Oxygen Steelmaking for Steelmakers, 1964. Published by George Newnes, London, W.C.2.

DISCUSSION

Professor D. B. Ie Roux (Visitor): The paper is particularly valuable because it deals first hand with the experiences of men who were at the forefront of an interesting and important development: The use of tonnage oxygen in steelmaking. Moreover, it is the exciting story of the successful application of the scientific method in overcoming almost innumerable problems. We who are interested in steelmaking but who cannot act in the frontline of the adventure, are taken step by step through the development of the process and it is an enriching experience.

In reading through the paper one is struck by two aspects. In the first place, there is the boldness of the men who pioneered this steelmaking method and their willingness to accept the challenge knowing beforehand the magnitude of the problems that would be encountered. Secondly, there is much evidence that the steps taken in this development were based on theoretical considerations and the metallurgical behaviour was to a large extent predicted by basic theory.

I am sure we are now satisfied that there is a real gap in the steelmaking processes into which the Tandem process will fit and all eyes will be on this operation that is to start at Vanderbijlpark in a few months' time.

In order to throw a little more light on the operation, I ask the authors to give us some details on the following points: How are the following operations going to carried out without interfering with the blow:

(a) Charging the pre-heating furnace with scrap?

(b) Tapping slag from the previous blow and charging raw material for new slag?

Also of considerable interest would be some details on the design of the coupling between the two furnaces.

Dr. A. G. Raper (Visitor): I have a few comments to make and some questions to ask about the various processes.

The greatest advance in oxygen steelmaking has been achieved by the L-D or basic oxygen furnace process. Tap to tap times of 60 minutes or even less are now being achieved on large-capacity vessels and it seems that the use of multi-nozzle lances has allowed the higher blowing rates to be achieved whilst giving an improvement in lining life and the reduction in the build-up of slag which for a long time proved troublesome with large scale L-D furnaces blown With a single nozzle lance.

The authors mention the possibility of interchangeable furnaces thereby halving the hood and blowing equipment since a vessel requiring repair would be removed from the blowing station. However, It has been our view that due to the maintenance requirements on hoods and the higher cost of the interchangeable furnaces, it is probably preferable to pay for the extra blowing station.

Mention is made of the problem of the high heat loss in the exhaust gas from the L-D process and the alternative solution thereto. Waste heat boilers are in operation, but in view of the high cost, large variations in steam output and the maintenance requirements, these have become somewhat less popular in recent years because to achieve a steady demand either a large proportion of auxiliary firing is necessary or large accumulator sets are required. Perhaps the authors would care to comment on this.

The unburnt gas systems seem to have had mixed success. In France, considerable difficulty was encountered during the commissioning stages. However, 1 believe that in Japan the equipment is working satisfactorily. I think for any type of exhaust gas system he first requisite should be high availability.

The authors mention the possibility of adding fuel via a lance to pre-heat the scrap and thus raise the scrap melting potential of the process. They tend to feel that this is of doubtful value since the fuel utilized will below. The table shows that to increase the scrap melting capacity from 17 ? percent to 35 percent, approximately 50 percent additional oxygen is required, and also due to the period required to transfer the heat, the steel production rate is reduced to about 57 percent of the original value.

As far as the Kaldo process is concerned, in Great Britain, three Kaldo plants have recently been commissioned. The mechanical operation of the Kaldo has not been troublesome and the metallurgical results obtained are good. However, in each plant, considerable difficulty has been experienced inlining wear. In all three plants, lining lives with dolomite were initially between 20 and 30 heats and although improvement has been obtained to about 60 heats, this has been achieved by the use of abnormally thick linings. The cost of refractory material per ton of steel is still high. It does appear that until cheaper high-grade refractory materials become available, the application of the Kaldo process will be limited.

As far as the Graef-Rotor process is concerned, the authors state that there is some similarity to the Kaldo process. However, I wonder whether they feel that the mixing between metal and slag in the rotor process is as effective as in the Kaldo process, due to its somewhat lower speed, and if, therefore, they feel that the ability of the pressed to handle high phosphorus metal is limited. Probably, however, for most applications, a more basic disadvantage to the process is that by virtue of its shape, scrap charging is a rather slow operation.

It is interesting to hear how, by reducing secondary combustion of carbon monoxide, long lining lives have been obtained on the Iscor Rotor process. I would like to ask the authors whether this has resulted in any change in the characteristics of the fumes evolved from the furnace and if the gas cleaning problems are simplified.

As far as the Tandem furnace is concerned, of course, we are anxiously awaiting results. However, would like to ask the authors whether they feel there is a control problem in the process which might cause delays. Depending upon the nature of scrap in the furnace and the combustion conditions in the scrap heating furnace, varying levels of pre-heats may be achieved during the pre-heat period, which will result in an out of balance heat during the normal blowing stage of the furnace, following furnace reversal. If such variations did occur, additional time may well be required to adjust the steel composition in the blowing section of the furnace.

Finally, I think one aspect which is clear from the paper is that the scrap balance of a traditional steelworks is disturbed by the introduction of oxygen steelmaking processes, although alternative steelmaking methods can help. However, this is not the only change that results, since there is also a radical change in the energy balance of the works, using, for example, the L-D process as opposed to the open-hearth process and based upon blast furnace practice. There is a larger surplus of energy from the L-D works in the form of coke oven and blast furnace gas, although, of course, instantaneous demands may reduce this surplus considerably. Under these conditions, the recovery of further energy for example in waste heat boilers may not be worthwhile.

Mr. C. E. Mavrocordatos (Visitor): In 1958, just a year after the publication of Dr. Graef's paper on the rotor process, Mr. Klein delivered a paper to the S.A. Institute of Mechanical Engineers, in which he reviewed the modern trends in iron and steel making. In this paper, not only did he describe the pneumatic processes already developed and established, but also one which was not yet fully developed, to wit, the Graef Rotor. He informed his audience that two such plants had been ordered and were in process of being erected at Iscor. The people who already knew of this bold and daring development by Iscor stood aghast and apprehensive. One of the greatest prob-lems to the Graef Rotor was the problem of life and indeed the very nature of the lining. One could not visualize the behaviour of the lining which was supposed to be in direct contact with extremely hot gases and which was to be a heat-bank to transfer this heat to the bath itself, particularly under the highly erosive conditions prevailing.

The South African technical experts and, I am sure, even those outside this country, followed the development with fear and awe. Their feelings soon changed to applause, approval, and admiration because in the subsequent years a great story of success emerged from the nebulous brown fumes of the rotors. Within five years, Professor Kruger and Mr. Klein not only succeeded in overcoming the apparently insurmountable difficulties but they also gave the world a better rotor, one which surpassed every expectation in its performance and life.

In comparing the design of the rotor as it was in 1958 with what it is today, one can see the following main changes:

1. The original idea was to use only ore as coolant. Today 5 percent scrap is used as a coolant and this figure will increase as scrap charging facilities are installed.
2. The oxygen lance is not stationary under the bath, but it is reciprocating just above the surface.
3. Secondary oxygen is not delivered at the top of the furnace, but ordinary air is sucked in to burn part of the carbon monoxide to carbon dioxide.

Mr. Klein was defeated in that he had to use scrap after all, and could not utilize all the heat which could be liberated during the process. But Mr. Klein does not take setbacks lightly. If this design of furnace or that process could not do what Mr. Klein wanted it to do, why, he just invents a new design and develops a new process. Hence the tandem furnace was designed and is going into operation shortly.

The experts who followed the struggle for the development of the Iscor rotor process with bated breath, are confident that such an able team of experts cannot but succeed this time.

Mr. Klein's paper covers all the pneumatic processes used to-day including the Iscor rotor process. In so doing, he wrote a full and useful paper but he detracted from his main theme. He should have written instead of a paper on 'The romance of developing a new steelmaking process.' May I suggest that in three to four years' time he gives us such a paper with the title 'Two steelmaking processes; the Iscor rotor process and the Iscor tandem process.'

Having started my contribution with well-deserved admiration, praise, and eulogy for a most wonderful achievement, I now ask Mr. Klein a few questions.

The thermodynamics of gas-slag-metal reactions have been worked for conditions that one could consider to approach equilibrium. To wit, in an atmosphere of comparative calmness, quiescence of the charge, and over a period of six hours in the open hearth bath. Now we have totally different conditions to encounter. Temperatures of over 2 000°C are obtained, the whole charge is in continuous turbulence and the reaction time have come down to about one hour.

I would like to know how the accepted equilibria fit in with this vastly changed picture.

What about scrap? I understand that the one aspect which captured his imagination, when, in the late 1950s, he studied the development of the Graef Rotor, was the possibility of complete exclusion of scrap from the charge. To him, this was an important aspect in view of the world shortage of scrap which has now also reached South Africa. Yet, in the new development of the tandem process, 45 percent of the charge will be scrap. Where is this scrap coming from? Is it only from hot tops from ingots due to increased production?

The authors appear to have decided that, if the open-hearth process is not already obsolete, at least it is on its last legs. In my opinion, this is rather a pessimistic view and the open-hearth process will survive alongside the new methods of steelmaking in the same way as it survived the advent of the electric furnaces. With the introduction of the Ajax system and the raising of the roof, the development of basic roofs, and improved instrumentation, its great flexibility concerning the consumption of scrap and its ability to produce any grade of steel, will ensure prolonged life for the basic open-hearth process of steelmaking.

The authors did not give us any information on the possibility of further developments at Iscor in the field of direct ore reduction. In 1958 he hinted that he was interested in the H-iron process and that he was awaiting developments on the BISRA Cyclosteel method. Can Mr. Klein say what the attitude of Iscor is towards developments in this field?

This paper is of tremendous economic significance to this country. From its title, it appeared to be a review of the pneumatic methods of steelmaking. We were, however, expecting to hear about the developments at Iscor. We heard the outline, we were referred to a paper which gives details, but I think this is the time to emphasize the socio-economic implications of these developments.

Put simply, it means that this country is not only taking its place among the self-sufficient, highly industrialized countries of the world but that also visionary, enlightened and imaginative management are making possible the development of new processes in this most important field of heavy industry. It means that South Africa does not draw parasitically from the technical achievements of other countries but, instead, is able to contribute to the advancement of science and technology.

Mr. P. Whyte (Visitor): The paper has given an interesting review of the history and development of the main methods of steelmaking in which tonnage oxygen is used, and of the reasons why the Rotor process and subsequently the Tandem furnace process have been chosen by Iscor in preference to the more fashionable L-D process.

As the authors point out, the excellent result obtained with the L-D method in which oxygen is blown on to a small portion of the bath surface is rather surprising, as one would expect that intimate mixing f oxygen and metal would be necessary for efficient oxygen usage. That this fact still troubles the minds of some researchers is shown by the development of the Graef Rotor process, where oxygen is injected below the metal surface. A more recent and a different approach to the problem is seen in a paper by Churcher and Lemlin(1), in which is described work n the development of a water-cooled oxygen probe that is installed in the side near the base of a converter vessel.

Tested in a converter of 150 lb capacity, the water-cooled probe was found to operate reliably and could offer an attractive means of blowing oxygen in converter vessels. It is felt by Churcher and Lemlin that the use of submerged oxygen probes may result in less splashing, greater oxygen efficiency, and lower fume losses. The oxygen probe would also be much smaller in size and hence cheaper than the lances used in the L-D process. No handling gear would be required.

Regarding the question of penetrating and non-penetrating blowing, it seems apparent, from the literature, that for early dephosphorisation at a relatively high carbon level, the lance must be positioned high above the metal surface, thereby inducing a highly oxidizing slag. In the Rotor furnace, it is claimed that dephosphorisation is performed by an emulsion of metal and slag being thrown, by the submerged primary oxygen lance, into the secondary oxygen stream.

One wonders, therefore, what effect the Iscor modifications to the Rotor process have had on the progress of dephosphorisation.

It would also be of interest to know how the nitrogen content of the steel has varied between the Graef Rotor and the Iscor Rotor processes. In the L-D process, oxygen of the order of 99.5 percent purity is said to be necessary, whereas, in the Rotor process, it is claimed that 95 percent pure oxygen can be used without deleterious effect on the nitrogen content. Could the authors give some indication of their experience in this respect?

It has been explained that the Tandem furnace is a logical development from the Iscor Rotor process. It is known that, with the Rotor process, it is necessary to blow oxygen alternately from opposite ends of the vessel, so as to even outlining wear. This is done by the Rotor vessel about a vertical axis at intervals. Since this movement will not be possible with the Tandem furnace, could the authors explain what steps will be taken to avoid excessive lining wear at the blowing end?

It appears that the Tandem furnace will be suited to the treatment of a high-scrap charge. Since scrap is known to be in short supply in South Africa, it is presumed that the extra scrap for the Tandem furnace will be diverted from the open hearth furnaces to the Tandem furnace. We, therefore, have a rather unusual position where the open hearth will be used to treat a high hot metal charge and an oxygen blowing process to treat a high scrap charge.

Finally, the installation of the Tandem furnace at Iscor represents a bold and interesting step, and the development of this new process will undoubtedly be watched with great interest by all those concerned with oxygen steelmaking.

Reference: 1. Churcher and Lemlin-Steel and Coal. 9th August 1963.

AUTHORS' REPLY TO DISCUSSION

Reply by Dr. K. O. R. Gebhard:

In reply to Professor D. P. le Roux:

We are providing rapid charging methods for solid materials. It is not our intention to interrupt blowing during charging. Furthermore, as mentioned; the furnaces are tilting so that the removal of slag will present no problem.

As the installation at Vanderbijlpark is a prototype Tandem furnace, we would rather not comment on any details of the design.

In reply to Dr. A. C. Raper:

We have no practical experience in the L-D process. The mention of the inter-changeable vessels was a development which I thought would interest an audience of mechanical engineers.

With regard to the problem of high heat loss, it should be borne in mind that the carbon monoxide evolves from the carbon in the hot metal and this carbon is delivered free of charge, so it is possible to operate economically without recovering any of the heat value of this gas. We are in agreement with Dr. Raper's remarks on waste heat boilers.

It is our experience that dephosphorization and the speed of rotation bear no relation to each other. The Rotors were designed to rotate up to a maximum of 5 rev/min. At present we rotate at a very slow speed. As you have seen from one of the diagrams, the diameter of the vessel is small in relation its length and we have found by experiments with models that if it is rotated fast then the difference in chemical analysis between the blowing end and the exhaust end is even greater than by slow rotation, as any longitudinal current which is set up is destroyed by a cross-current set up by the rotation. We have carried out tests with a synthetic iron-containing over 2 percent of phosphorus. Blowing this iron by the Iscor-practice, i.e. with no secondary oxygen, we have had I no difficulty in dephosphorization. The slag, however, at the beginning of the blow is in such a condition that every open-hearth operator would be astounded if he saw it. We have definitely found that there is no need for inter-mixing of this slag with the steel as is done in the Kaldo process.

It was mentioned by Mr. Klein in a recent paper that we still have to work out the theory. In reply to the question of dust during a rotor blow, we did not notice any difficulties due to the variations ill our operation. With our present blowing practice, it is characteristic that the dust is black, which indicates that it consists mainly of magnetite particles and not haematite particles. The brown fumes are observed mainly during the short time when the lance is introduced or withdrawn or on the occasion when surface blowing has to be resorted to in order to gain temperature.

In the contemplated Tandem operation, the question of balance has also exercised our minds. It must be borne in mind, however, that this will be a new process and the object of the prototype installation is to enable u to answer these and similar questions.

In reply to Mr. C. E. Maorocordatos:

The questions asked by Mr. Mavrocordatos I shall leave to Mr. Klein, especially those which refer to the paper he gave some years ago to the Institute of the Mechanical Engineers.

I would like to add, however, that even with the vastly changed picture, the accepted equilibria are attained.

In reply to Mr. P. Whyte:

The question of the efficiency of the Iscor Rotor process as a dephosphorizer has been answered. In regard to nitrogen, we started using oxygen of 95 percent purity and we found that the nitrogen content of the steel wire in the order of .005 and .006 which, while good enough for more steels, is not satisfactory for deep drawing qualities and sundry other sheet products. We have recently operated one of the Rotor plants with 99.5 percent purity oxygen and the nitrogen content under these conditions is .002 to .003 so, as in the L-D process, the purity of the oxygen has a decided influence on the nitrogen content of the steel.

In regard to the question of how lining wear can be minimized in the Tandem furnace since there IS no rotation and also because we cannot blow from both ends I failed to mention that with our present Rotor blowing process which will also be used in the Tandem furnace, the slag temperature at tapping is lower than the temperature of the steel. We have measured temperature differences between 40 and 70 degrees C, i.e. we develop more heat inside the bath rather than above the bath. The original concept of the Rotor was to generate more heat above the bath than 10 the bath. Having reversed this concept, we see no need for rotation. Furthermore, the Tandem furnace i5 so designed that the roof and walls are more generously spaced than in the present Rotor. We feel that lining wear in the Tandem furnace should not cause us undue difficulties.

Regarding the use of scrap, we will have to balance our open hearth and Tandem furnace inputs. While the figure of 45 percent input of scrap to the Tandem furnace has been mentioned, it must be pointed out that this is a maximum figure and might not necessarily be used in our practice. Against common belief, we found that our Vanderbijlpark open-hearth furnace gave their maximum outputs with a 25 percent input of scrap.

Reply by Mr. E. Klein:

Mr. Mavrocordatos has taken several friendly tilts at some of my remarks in this and other papers. It might interest him and others that somewhat surprisingly we are building a new 400-ton open-hearth furnace. We always maintained and always will maintain that the open hearth furnace is the most versatile of all steelmaking units. You can use low scrap, high scrap, or all scrap in the charge and you can make a large range of steels, but an all pig iron charge should not be used. You can even make a large range of steel. High-speed tool steels and stainless steels have not yet been made on the op n hearth furnace. We made during the last war 14 percent manganese steel in a 150-ton open-hearth furnace. I would not advocate this as a normal practice.

The object of building this 400-ton furnace will be to enable us to discard some of our older open-hearth furnaces if the Tandem furnace is a success. We are toying with the idea of 200 ton Tandem furnace in the future.

Mr. Mavrocordatos seems to think that our intention to use 100 percent ore in the Rotor as a coolant might have been optimistic. The main objective of building any steelmaking unit is to make steel and if you find that some of your earlier concepts are wrong, well, find some other concept that works.

I might add that Mr. Whyte said that there is a shortage of scrap in the Republic and that to find 4 percent of scrap for the Tandem furnace might cause difficulties. This has been answered by Dr. Gebhard. We hope, however, that the Tandem furnace is not going to be used only in South Africa. It might with advantage be used in countries where there is a surplus of scrap.

Mr. Mavrocordatos also made remarks regarding my thoughts on direct reduction. It is quite correct that in the earlier paper I did mention direct reduction and H-iron. The direct reduction must not be ignored, but I think that I did say, and if I did not, I ought to have said it, that as long as one has good metallurgical coke or a good blast furnace fuel and good ore the orthodox blast furnace is the most economical unit in an integrated steel plant, and I venture to say that our conception of the blast furnace as we knew it five years ago and we will know it five years hence, will be totally different. Furnaces of 28 feet in diameter, using all modern techniques, are making 3 000 tons of iron a day and this may even be exceeded but I am not going to venture any more bold statements.

In regard to the Tandem furnace, as Dr. Gebhard has mentioned, we no doubt will have our headaches, disappointments, and other little difficulties. Further work has been done overseas which, unfortunately, I am not at liberty to mention. All I can say is that it tends to give us more optimism than we have ever had.