US3551140A

US3551140A - Controlled refining of pig iron

Info

Publication number: US3551140A
Authority: US
Inventors: Paul Emile Nilles; Etienne Marie Denis
Original assignee: Centre de Recherches Metallurgiques CRM ASBL
Current assignee: Centre de Recherches Metallurgiques CRM ASBL
Priority date: 1966-10-11
Filing date: 1967-10-09
Publication date: 1970-12-29
Anticipated expiration: 1987-12-29
Also published as: LU52150A1; NL142453B; DE1583228B1; GB1153938A; FR1548479A; NL6713782A; BE704338A; DE1583228C2

Description

Dec. 29, 1-970 P. E. NlLLES ErAL 3,551,140

CONTROLLED REPINING 0F PIG IRON Filed Oct. 9, 1967 //VV[/\/70/ .5 PAUL EMILE NILLES ET.AL

United States Patent 52,1 0 Int. Cl. C21c 5/32 US. Cl. 7560 7 Claims ABSTRACT OF THE DISCLOSURE A process for the dynamic control of the pig iron refining operation carried out in a solid base converter, characterised in that, preferably relying on theoretical bases, the starting materials necessary for carrying out a predetermined refining of a given pig iron are determined, that during the course of the refining operation the value of the surface of penetration of the jet or jets of refining oxygen into the molten bath is regulated in such a Way that the amount of carbon removed from the bath per unit of time follows a predetermined time graph, and that while respecting this latter condition, the conditions of blowing are regulated in such a way that the amount of oxygen absorbed per unit of time by the various constituents of the slag also follows a predetermined time graph.

The invention relates to a method for the dynamic control of the pig iron refining operation conducted in a solid bottom converter.

The general method of conducting a refining operation can be thought of in two complementary ways; on the one hand in a static way which starts from a precalculated theoretical amount of the elements necessary for entering into reaction or which should enter into reaction in the course of the refining operation being conducted, and on the other hand, in a dynamic way, by following as closely as possible the progress of the refining and guiding it as frequently as it appears to be desirable.

Speaking more specifically, the static phase of the refining operation can be set out as follows:

One has available at the beginning:

(a) Pig iron whose composition and temperature are known.

(b) Lime, the analysis and temperature of which are known.

(c) Cooling additives such as scrap, ores and so on, of

known analysis and temperature.

(d) Oxygen of known purity.

One desires to obtain at the end of the refining process:

(a) A given weight of steel.

(b) A predetermined temperature for the said steel.

(0) A predetermined, content of carbon, sulphur and phosphorous in the steel.

From these indications it is possible to determine what is to be put in the furnace i.e. the amounts of pig iron, lime and scrap to be introduced at the beginning into the converter as well as the total amount of oxygen to be blown in.

The dynamic phase of the refining operation is the controlling of the steps of the operation to obtain at the end of the refining process a desired steel, having desired properties, while maintaining the various physical and chemico-physical reactions within the limits of a range which is considered to be correct.

These two phases which are particularly important for obtaining at the end of the refining process the desired product, in the manner desired, have already separately been the object of numerous research projects.

As far as the static phase is concerned, the major part of the methods worked out for effecting the determinations of what is to be put in the furnace relies upon certain estimations which are empirically fixed, or on data obtained from the practical experience of operators. This state of affairs obviously makes such methods very uncertain, diminishes the accuracy and allows too great a discrepancy of results.

Among the methods commonly used for carrying out the control of a refining operation in the course of blowing, is one which consists in following by means of a suitable device, a graph revealing the evolution during the course of the refining of a certain parameter, for instance, the temperature of the converter fumes. This graph is then compared with a predetermined one which is considered as being representative of the ideal evolution of this parameter during the course of the refining operation. The comparison of these two graphs makes it possible to draw certain indications regarding the state of progress of the refining and to make if necessary, any correction which would appear to be necessary. All these graphs, considered as ideal, are determined by statistical and experimental observations, and in general they take into account factors pertaining to a given steel works and to a given converter.

This method has some disadvantages; in the first place it does not allow to intervene to a completely for the connection existing between the blowing conditions and the reactions which take place in the converter, and furthermore it is based on ideal graphs which are representative only of a portion of the reactions taking place in the converter. It follows from this that the corrective measures indicated by this method are often incomplete and sometimes inadequate.

The method indicated above has up to the present practically always been used separately and in actual fact has not been able to provide effective dynamic control of the refining operation. Up to the present no method of control has been proposed based on the simultaneous utilization of the two phases i.e. static and dynamic, leading to a completely effective control of the refining process.

The present invention has as its object a method for the dynamic control of the refining operation of pig iron making it possible to improve upon the incomplete and imprecise nature of the methods known theretofore. This method makes possible the control and regulation of the refining operation during its entire course.

In the description which follows, as well as in the appended claims, the expression surface of penetration has been used which should be understood as the surface at the locality of the point or points of the jet or jets of refining gas for each of which the dynamic pressure is equilibrated by the static pressure of the molten metal bath and the slag. A method for determining such a surface is given below.

The method forming the subject of the present invention is essentially characterized in that in accordance with a known method, and based preferably on theoretical bases, one determines the charge necessary to put into the furnace to carry out the required refining of a given pig iron, and in that, during the course of this refining operation, one automatically regulates the surfaces of penetration of the jet or jets of refining oxygen in such a way that the amount of carbon removed from the bath per unit of time follows a time-graph which is determined by prior calculation, and in that while maintaining the observance of this first condition, one also automatically reg- 3 ulates the blowing conditions in such a way that the amount of oxygen required per unit of time by the various constituents of the slag, also follows a time-graph determined by prior calculation.

It has in actual fact been found that during the entire refining operation and in well determined operational conditions, the surface of penetration is a function of the amount of carbon escaping from the bath per unit of time and that the amount of O absorbed per unit of time by the various constituents of the slag, as well as the surface of penetration, is a function of the blowing conditions. It suflices accordingly, within the limits fixed by the operational conditions, to trace a curve or relationship which gives as a function of value of this surface of penetration the amount of carbon escaping from the bath as well as the curve or relationship which compares the blowing conditions to the amount of O The principle of the present method consists in following these two graphs. Among the operational conditions which influence the regulating characteristics of the surface of penetration there should be enumerated the capacity of the converter, its geometry, and the nature of the pig iron put into the furnace.

In other words, the regulating process of the invention can be described, once the furnace charge has been determined and the steps of the refining process has been fixed, in the following way:

(a) One knowns the graphs representing the ideal evolution as a function of time of the rate of departure of carbon from the bath and of the rate of absorption of oxygen by the slag.

(b) One determines the relationship which relates the surface of penetration S to the rate of departure of carbon from the bath.

(c) One automatically calculates the value of this surface S as well as the actual rate of departure of the carbon. By verifying whether this actual rate of departure of the carbon corresponds to that indicated by the ideal graph, one can make a check to see whether the refining is progressing in the desired way.

(d) If this is not the case, S is modified in accordance with the relationship determined by step b above, until the value of the actual rate of departure of the carbon is equal to that indicated by the ideal graph.

(e) One acts similarly as far as the graph of absorption of O by the slag is concerned. These two aspects of the regulation are examined in greater detail below.

(A) For measuring the surface of penetration of the jet or jets of refining oxygen as a function of the amount of C removed from the bath at a determined moment, one can for example proceed as follows:

(I) The theoretical calculation of the maximum depth of penetration of a gas in a liquid, normally when its upper surface is taken to be free and at rest, has been adapted to the case where by means of a nozzle, oxidising gas is injected, i.e. oxygen is injected into a bath of molten metal contained in a converter and covered with a layer of slag. One has accordingly obtained the following formula:

-In this equation:

4 l =Length of supersonic core, in metres; it is given by the equation:

d Diameter of the throat of the nozzle in metres.

s =Weight of the slag floating on the bath in kg.

a Numerical coefficient making it possible to transform in equivalent of ferrostatic height the thickness of the layer of slag floating on the bath. This thickness is obtained by dividing the weight s of slag by the product of its density through the section of the converter at the level of the slag. The coefficient a appears as being the ratio density slag/ density metal.

H =Height above the bath of the lower end of the lance,

in metres.

p=Maximum penetration of the jet into the metallic bath on the axis of the jet. This depth of peneration represents the distance existing between the surface of the bath before injection and the lowest point of the surface of penetration (in metres).

A, B, C represents values which are a function of the parameters such as the temperature of the jet, the density of the oxygen, the density of the liquid and the distribution of rates in the jet of gas. These values A, B, C are known since the relations connecting the said values to the parameters are themselves well known. (II) The locality of the points of the jet for which the penetration p is lower than p is determined as a function of the radius r, representing the distance from the point considered to the axis of the jet.

Use is made of the Equation 1 given above, in which:

p is replaced by p. A is replaced by A, A having the value A.exp[ED(r/r') with an d.

(III) The combination of the Equations 1 and 2 accordingly makes it possible to know p as a function of r. It suflices to vary p from zero to p to known point by point the value of r. Supposing that the surface of penetration is one of revolution about the axis of the jet, one will obtain its value by the equation S=f 2.nr(p').dp" since r is known as a function of p. The integration can be carried out by any known method.

The calculation which has been described above is, as Was stated above, drawn up within well defined conditions. Of course the protection claimed is not restricted to this method of calculation but also covers any type of calculation of the surface of penetration whatever may be the form, inclination, or simple or multiple character of the jet.

(IV) The relation existing between the surface of penetration of the jet in the bath and the amount of carbon AC/At leaving the bath during the unit of time can be obtained in accordance with the invention on the basis of the considerations given below.

It has been observed that for numerous refining operations which have been carried out the amount of carbon removed from the bath during the unit of time is substantially proportional to the value of the surface of penetration of the jet in the bath of metal (coeflicient K substantially constant) since (a) The depth of penetration of the jet is between and 75% of the height of molten metal located in the converter, the metal being taken to be at rest;

(b) The content in carbon of the molten bath is between 100% and of the initial carbon content of the pig iron.

Starting out from these premises one can easily trace a graph or a network of graphs, giving S as a function of AC/At for different depths of penetration. The relation between AC/At and the speed of decarbonization being well known, one can accordingly determine the effect of a modification of S on the rate of decarbonisation.

If one considers what takes place at the end of the refining operation, i.e. when the content in C of the bath is lower than 15 of the initial carbon content of the pig iron, one must take into account the fact that the decarbonisation is very advanced and that the same surface of penetration of the jet will have more reduced influence on the disseminated carbon remaining in the bath. For taking into account this phenomenon, one affects the coeflicient of proportionality K of a factor K of reduction decreasing K in a substantial way linearly between the moment when the carbon content of the bath is 15 of the initial carbon content of the pig iron (at this moment K'=1) and that when this content cor responds to the final content allowed for (at this moment K'=0).

When the refining is carried out by means of a plurality of jets of oxidising gas operating simultaneously one puts successively into the Equation 1 or a similar equation, for the penetration of the various partial jets and one makes a total of the surfaces obtained by applying as many times as there are partial jets the Equation 2. The coefiicient of proportionality K which one uses then represents substantially the mean of these corresponding to each partial surface.

The relation of the surface of penetration S to the removed rate of C from the bath thus being known it is now possible to follow the ideal known graph, giving the removal rate of C from the bath as a function of time. In actual fact it suffices to measure the actual AC/At, at a given moment, and to compare the result of the measurement with the corresponding value of the ideal graph. One accordingly knows whether this graph is followed or not and one modifies AC/At by means of S to approach the ideal graph. The way in which S should be modified is known by virtue of the equation given above. This equation gives directly the effect of, for example, the rate of flow of 0 of its pressure and of the height of the lance, on the value of S.

(B) Once the surface of penetration has been regulated it is necessary to adjust the blowing conditions in such a way as to absorb a given amount of oxygen in the constituents of the slag, at a given moment.

For this purpose it will be recalled that:

(a) The blowing conditions which one can influence are the following:

(1) Height of the lower end of the lance above the bath.

(2) Rate of oxygen passing through the lance.

(3) Pressure of refining oxygen (possibly charged with lime).

(4) Geometric characteristics of the blowing nozzle.

(b) One knows as a function of time the total amount of oxygen supplied to the bath; all that has to be done is to measure it.

(c) Knowing the relation existing between the surface of penetration and the amount of carbon escaping from the bath, one can, i.e. by means of the hypothesis that the ratio of the concentratioins of the CO and C0,, of the gases issuing from the converter is approximately constant, determine the amount of oxygen in relation to the carbon. This amount can be controlled by measuring the rates of the conversion gases in the hood and by analyzing them.

(d) One obtains from the difference, the total amount of O absorbed in the slag. It sufiices to compare this amount with that supplied by the predetermined ideal graph to know in which direction it may be necessary to modify this amount of oxygen.

For modifying the amount of oxygen absorbed by the slag action is taken in the sense desired on one or more of the above-mentioned conditions. It should neverthe less be remembered that the determined relationship between S and the amount of C removed from the bath per unit of time should be followed. The complexity of the control operations, the measuring and calculating operations to be carried out for putting in operation the process of the invention in actual fact requires the permanent use of suitable electronic computers.

In accordance with an advantageous variation of the process of the invention one carries out continuously the various regulations necessary for keeping on or bringing back to the ideal graphs the actual graphs representative at any moment of the surface of penetration, the amount of C removed from the bath and the amount of 0 absorbed by the slag.

For checking whether the rate of C removal from the bath is deviating from predetermined ideal conditions, apparatus is arranged by which it is possible in particular to take the following standard measurements:

The temperature of the converter fumes.

The energy emitted by the flame of the converter.

The sound emitted by the converter during the course of refining.

The content in oxygen of the converter fumes.

The content in CO and CO of the converter fumes.

The amount of vapour produced in the hood.

If these measurements indicate that the evolution of the removal of C is not in accordance with the ideal evolution (for instance in accordance with the indications of the equations given above), one modifies the surface of penetration in the sense desired to obtain the desired conditions.

In order to respect the second condition imposed, i.e. the amount if 0 absorbed by the slag at a particular time, use is also made of the above-mentioned apparatus and, in accordance with their indications, the supply of O to the molten bath is modified in the sense desired. By means of at least one of the other blowing conditions, the influence which the modifications of the rate of 0 could have had on the value of the surface of penetration is compensated.

The present method is able on the one hand to simultaneously regulate several values, each is a characteristic of an aspect of the state of progress of the refining process, which values are dependent on each other, and on the other hand the corrections which appear necessary are introduced in a permanent and automatic way, which rapidly reduces any deviation between theactual graph and the ideal graph.

The use of basic theories for calculating what is to be put into the furnace also makes possible a number of advantages, among which one can mention;

An increase in the uniformity of the weight of steel and a decrease in the ingot wastage.

A decrease in the content in P in rephosphorisation and in the dispersion of the content in P. Decrease in the final temperature gradient of the steel.

Decrease in the amount of lime per metric ton.

These advantages can be obtained by means of the process of calculating what is to be put into the furnace comprising the following steps in sequence:

(A) From known chemico-physical data of the nature of the pig iron and of the ternary diagram CaO-FeO-P O One determines the composition of the slag to be produced, this being saturated with lime. The main constituents of the slag (CaOSiO P O Fe O MnO-MgO) are calculated by means of a system of six equations with six unknowns. These six equations translate into mathematical language the following six conditions:

(1) The main constituents of the slag, i.e.

account for a determined percentage (nearly 100%) of the total weight of slag. It is known for a given steel-works.

(2) One adopts for MgO a suitable mean value which can be estimated for the steel works concerned.

(3) It is assumed that the content in MnO depends on that of the iron.

(4) The value of the ratio SiO /P O 0f the slag supplies a fourth ratio taking into account that the total amount of P 0 is equal to that of slagged P 0 Whereas the total SiO is composed of slagged SiO and SiO of the lime.

(5) It is assumed that in the ternary diagram used (CaO-P O FeO used in the case of phosphorus pig iron and C210, FeO, SiO for hematite pig iron), the line of saturation of the slag in lime is a straight line.

What this means in actual fact is setting up a simple relation between the basicity of the slag and its content in Fe, since use is made of a slag saturated with lime. This relation can take the following Well known form:

In the case where hematite pig iron is being treated, use is made of the ternary diagram CaO-FeO-SiO and always for a slag saturated with CaO, the above equation has the form CaO/SiO =A.Fe+B.

(6) A content in P of the steel corresponds at the temperature envisaged for the steel to each composition of slag calculated as a function of the Fe content of the slag.

(B) One deduces from these 6 equations and a balance of CaO the weight of the slag per metric ton of pig iron and the lime used per metric ton.

C) A balance oxygen supplies the value of the volume of O necessary for the refining operation.

(D) The amount of pig iron to be used as well as the cooling additions are determined by salving a system of two equations, one representing the Fe balance and the other the thermal balance well known per se.

The method of calculating the amounts used as given above can be utilized not only as a basis for determinations relating to the refining operations carried out in accordance with the customary Thomas process but also in accordance with variations in the processes of top blowing. A few simple modifications will simply be made in accordance with the case concerned.

(1) Thus, for instance, in the LD-AC process in accordance with which one may or may not retain the slag the refining or of the preceeding phase, one must take into account the amount of lime, Fe and of O reintroduced into the converter.

(2) If one is concerned with refining carried out in accordance with a LD process where it is particularly difficult to achieve a low sulphur content, e.g. 0.02, one makes use of the value found for the ultilization of lime per metric ton necessary for obtaining the content in phosphorus aimed at. This value, found from the diagram CaOFeOSiO makes it possible to determine the basic ity index of the final slag envisaged. In accordance with a known equation, this basicity index corresponds to a partition coefiicient for sulphur between the slag and the metal. An equation setting up the sulphur balance makes it possible to calculate the content in sulphur which one should look for in the steel. If this content is considered as excessive, one decreases the phosphorus content to be obtained until the sulphur content reaches the minimum value envisaged.

It should be pointed out here that once the amounts are determined, for instance in accordance with the method given above, and consequently once the evolution of the refining process has been determined, one

8 knows as a function of time the graph representative of the ideal evolution of the rate of decarbonisation or that of the rate of removal of the carbon from the bath.

As the latter is directly related to the surface of penetration of the jet of refining gas in the molten metal bath, for the refining to follow the ideal course, it is necessary for this surface to be varied in such a way that the rate of removal of C follows its pre-determined ideal graph.

It should be mentioned here that although the regulations according to the invention in the above examples have been set up as a function of the momentary rates of removal of C from the bath and of the momentary rate of absorption of 0 by the slag, it is not going beyond the scope of the invention to set up these regulators as a function of other parameters which are functions of these rates (for instance as a function of the total amount of C which has been removed from the bath at a given moment, of the total amount of O absorbed by the slag at a given moment).

In order to increase the effectiveness of the process described in the present invention provision is made for doubling or multiplying all the control and regulating installations provided by the said process.

EXAMPLE OF REFINING OPERATION CARRIED OUT IN ACCORDANCE WITH THE INVENTION, BUT IN NO RESTRICTIVE SENSE II.--'First phase One introduces 28.8 metric tons of scrap and 0.61 metric tons of scale;

'One blows 5800 m. of 0 for 16 /2 minutes at an average rate of 350 m. /rnin.;

One incorporates 5.23 metric tons of lime in the blowing 0 from the 6th to the 12th minute.

At the end of this first phase, the analysis of the metal was C=O.54%, P=0. 168%, Mn=0.36%, and the slag had 51% CaO, 8.8% SiO 19% P 0 9% Fe.

IIL-Second phase Duration 8 minutes;

One introduces 8.5 metric tons of scrap, 2.5 metric tons of rolling mill scale;

One blows 2060 m. of 0 One incorporates 7.57 metric tons of lime in the oxygen in the course of the first five minutes of blowing.

At the end of this second phase, the analysis of the steel is as follows: C=0.047%, P:0.015%, Mn=0.08% with a temperature of 1596 C. The slag comprises: 46.8% CaO, 5% SiO' 9.5% P 0 IV.-'Characteristics of the diagrams of rates of decarbonization and rate of flow of oxygen which the operator must follow to keep in conformity with the process of the invention (see attached drawing) (1) First phase (a) Before the injection of lime (from t- /2 min. to t=6 min.)

9 AC/At varies from 180 to 320 kg./min.

A varies from 260 to 130 imfi/min.

(b) During the injection of lime (from t:6 min. to

t=l2 min.)

AC/ At remains constant at 230 kg./ min.

A0 remains constant at 180 mfi/min.

(c) After the injection of lime (from t=12 min. to

r=16v2 min.)

AC/At varies from 250 kg. to 160 kg./min.

A0 varies from 150 to 801m. /min.

(2) DesIagging.-From t: 16 /2 to t: 18 min.

(3) Second phase.From t: 18 min. to t=26 min.

(a) During the injection of lime (from t=18 min. to

t:23 min.)

AC/At remains constant at 120 kg./min.

A0 remains constant at 180 m. min.

(b) After the injection of lime (from t=23 min. to

t=26 min.)

AC/At varies from 180 to 50 kg./'min.

A0 varies from 250 to 400 m. min.

In accordance with the method of the invention the graph AC/At is simply related to the depth of penetration of the O jet in the molten metal bath. One predetermines a graph AC/At which one attempts to follow by varying the pressure of the oxygen and the height of the lance. One also predetermines a graph A0 which one attempts to follow by modifying the rate of flow of 0 but while still following the first predetermined graph.

The attached figure shows as abscissae the duration of the refining in minutes, and as ordinates there are plotted (graph 1) the rate of decarbonisation in kg./min. and (graph II) the rate of 0 in mfi/min. The two flat levels correspond to the injection of lime in each of the two phases; the interruption of these two graphs corresponds to the deslagging.

Although the invention has been described and illustrated in detail, it is to be understood that this does not delimit the invention. The spirit and scope of this invention is limited only by the language of the appended claims.

We claim:

1. A method of controlling the refining of pig iron in a solid bottomconverter comprising the steps of:

(A) determining the quantitative charge of pig iron, lime and cooling additives to be introduced into the converter as well as the total quantity of O to be blown therein, based on the composition and temperature of the charge materials and the purity of the 0 as well as on the desired weight, temperature and amount of P, S and C in the finished steel;

(B) introducing at least a part of the so determined charge into the converter, forming a molten metal bath, and blowing 0 through lance means onto the upper surface of the bath;

(C) continuously measuring the surface of penetration of at least one jet of 0 on the molten bath surface, the measuring being determined from the flow and pressure of the 0 through the lance means, the height the exit end of the lance means is from the bath, the shape of the lance means, length of the supersonic core, and the weight of slag floating on the bath;

(D) continuously determining the actual rate of carbon removal from the bath, which rate is substantially proportional to the surface of penetration;

(E) continuously comparing the actual rate of carbon removal to a predetermined theoretical rate which ideally would obtain a physicochemical equilibria between the slag and the elements to be slagged in the molten metal bath by the blown 0 (F) automatically adjusting the surface of penetration by altering at least one of the parameters selected from the group consisting of the flow of 0 the pressure of 0 and/or the height of the lance means above the bath, until the actual rate of carbon removal coincides with the predetermined theoretical rate;

(G) continuously determining the actual quantity of O absorbed per unit of time by the constituents of the slag;

(H) continuously comparing the actual quantity of O absorbed per unit of time to a predetermined theoretical quantity absorbed per unit of time which would ideally be required; and

(I) automatically adjusting the actual quantity of O absorbed per unit of time by altering at least one of the parameters set forth in step F until the actual quantity of O coincides with the predetermined theoretical quantity per unit of time.

2. A method as claimed in claim 1 wherein the adjusting steps of F and I are sequential and the surface of penetration is further adjusted by means of at least one of the parameters set forth in step F to compensate for the influence the adjusting step I might have had on the adjusted surface of penetration as determined in step F.

3. A method as claimed in claim 2 wherein the various sequential steps are continuously carried out.

4. A method as claimed in claim 1 wherein the main components of the slag are determined on the basis of the following conditions, the slag being considered as saturated in lime:

(A) that CaO, SiO P 0 Fe O *MnO, and MgO account for a substantial percentage of the total weight of the slag, the particular percentage being unique to each steel works;

(B) that a suitable mean value of MgO is estimated;

(C) that it is assumed that the MnO content depends upon the Fe content;

(D) that the value of the ratio SiO /P O of the slag provides a fourth ratio, taking into account that the total amount of P 0 is equal to that of slagged P 0 whereas the total SiO is composed of slagged S10 and the SiO of the lime;

(E) that it is assumed that in the ternary diagrams used, the line of saturation of the slag in lime is a straight line'thereby permitting a simple ratio between the basicity of the slag and its Fe content to be established; and

(F) that the P content of the steel at a predetermined steel temperature corresponds to each composition of slag calculated as a function of the Fe content of the slag;

based on the above conditions and a balance of CaO the weight of slag per metric ton of pig iron and the gross weight of lime per metric ton are calculated; the volume of required 0 necessary for refining is determined by calculating an oxygen balance; and the amount of pig iron to be used as Well as the cooling additives are calculated by solving a system of two equations with two unknowns, one representing the Fe balance and the other the thermal balance.

5. A method as claimed in claim 1 wherein for depths of penetration of the jet of 0 between 10% and of the height of the molten metal bath and for carbon content of the bath between and 15% of the initial carbon content of the pig iron, the adjusting of the surface of penetration is determined by a relation in which the surface of penetration of the jet is related to the amount of carbon removed from the bath per unit of time by a coefficient of proportionality K which is substantially constant.

6. A method as claimed in claim 1 wherein for depths of penetration of the jet of 0 between 10% and 75% of the height of the molten metal bath and for carbon content of the bath which is less than or equal to 15% of the initial carbon content of the pig iron, the adjustment of the surface of penetration is determined by a relation in which the surface of penetration of the jet is related to the amount of carbon removed from the bath per unit of time by a coefiicient of proportionality K decreasing substantially linearly from 1 to zero between the moment when the carbon content of the bath is 15% of the initial carbon content of the pig iron and when the carbon content of the bath corresponds to the final desired carbon content.

12' of O jets are simultaneously utilized and the continuous measurement of the surface of penetration is the sum of the surfaces of penetration of each jet.

References Cited UNITED STATES PATENTS L. DEWAYNE RUTLEDGE, Primary Examiner 7. A method as claimed in claim 1 wherein a plurality 15 WHITE, Assistant EXaIniIlflr