US3847593A - Process for refining metals, in particular liquid pig iron, in oxygen converters with continuous control of the operative procedure - Google Patents

Abstract

A process for refining metals, in particular liquid pig iron into steel, in stationary or rotating reactors, by top-blowing oxidizing gases, particuarly oxygen, onto the liquid metal or slag bath, by means of control and/or continuous adjustment, made by an electronic computer, of the flow rate of the refining gas blown with a lance, and/or, - in the alternative - of the distance between the delivery head of the said lance and the metallic bath surface, wherein, on the basis of determined characteristic process indexes, an instantaneous metallurgical parameter m for the functioning of the refining process is established and for each type of operation there are prescribed a curve of optimal variation of the said metallurgical parameter m as a function of the blowing time, and a proper allowance band, within which the actual instantaneous values of the said metallurgical parameter can oscillate around the before mentioned optimal curve, these data being stored into the memory of the before mentioned electronic computer; during a given type of refining and during the whole blowing period, the said computer continuously and automatically performing the calculation of the actual instantaneous value of the said metallurgical parameter m, determining the actual variation curve with the time thereof, said curve being continuously and automatically compared with the prescribed optimal curve and the instantaneous actual value of m being continuously brought back to the corresponding optimal value or, at least, being maintained in the interval of the said allowance band.

Description

United States Patent 1191 Ramacciotti et al.

[ Nov. 12, 1974 [75] Inventors: Aldo Ramacciotti, Rome; Giancarlo Eminian, Taranto, both of Italy [73] Assignee: Centro Sperimentale Metallurgico S.P.A., Rome, Italy [22] Filed: June 29, 1972 [21] Appl. No.: 267,694

[30] Foreign Application Priority Data July 13, 1971 Italy 31604/71 [52] US. Cl. 75/60, 75/59 [51] Int. Cl. C21c 5/30 [58] Field of Search 75/60, 59

[56] References Cited UNITED STATES PATENTS 3,720,404 3/1973 Carlson 75/60 3,594,155 7/1971 Ramachandran. 75/60 3,598,386 8/1971 Murphy 75/60 3,700,429 10/1972 Ramachandran... 75/60 3,723,099 3/1973 Marukawa 75/60 3,533,778 10/1970 Nilles 1 75/60 3,485,619 12/1969 Maatsch.... 75/60 3,372,023 3/1968 Krainer 1 75/60 3,719,469 3/1973 Roessing 75/60 3,475,599 10/1969 Schwartzenberg 75/60 Primary Examiner-C. Lovell Assistant Examiner-Peter D. Rosenburg Attorney, Agent, or Firm-Young & Thompson [571 ABSTRACT A process for refining metals, in particular liquid pig iron into steel, in stationary or rotating reactors, by top-blowing oxidizing gases, particuarly oxygen, onto the liquid metal or slag bath, by means of control and- /or continuous adjustment, made by an electronic computer, of the flow rate of the refining gas blown with a lance, and/or,in the alternative-of the distance between the delivery head of the said lance and the metallic bath surface, wherein, on the basis of determinedpharacteristic process indexes, an instan t aneous metallurgical parameter m for the functioning oi the refining process is established and for each type of operation there are prescribed a curve of optimal variation of the said metallurgical parameter m as a fu nction o f the blowing time, w allowance band, within which the actual instantaneous values of the said metallurgical parameter can oscillate around the before mentioned optimal curve, these data being w iin the memory of t e e qremsatie e199; tronic computeri du r irig a given type of refining and during the whole blowing period, the said computer continuously and automatically performing the calculation of the actual instantaneous value of the said ngtallurgical parameter nb determining the actual v ar: iatiodcur ve with the time thereof, 5211512696 being continuously and automatically compared with the prescribed optimal curve and the instantaneous actual value of m being continuously brought back to the corresponding optimal value or, at least, being maintained in the interval of the said allowance band.

1 Claim, 4 Drawing Figures PROCESS FOR REFINING METALS, IN PARTICULAR LIQUID PIG IRON, IN OXYGEN CONVERTERS WITH CONTINUOUS CONTROL OF THE OPERATIVE PROCEDURE The present invention covers a process for refining metals, in particular liquid pig iron, in oxygen converters with continuous control of the operative procedure. More particularly the invention relates to a liquid pig iron refining process, where an instantaneous determination of the refining pattern is performed as well as a continuous bringing back to a pattern preestablished as a standard model to be attained for the type of heat, by monitoring the oxygen instantaneous balance, for the purpose of keeping all the heats with temperature percentage of the final C and iron yield, within the allowance limits preestablished for each type of heat.

In the refining processes of pig iron in top-blown converters, the amount of oxygen blown under pressure and the distance between the lance and the metal bath, assume great importance as operative control variables for the progress of the processes themselves. These values, in fact, are those which determine the composition and the temperature of the resulting gas jet which impinges on the bath and the depth of its penetration into the slag layer and the liquid metal, by influencing, in this way, the pattern of the metallurgical refining process.

As a rule, both the distance lance-metal bath and the flow of the oxygen jet, are preestablished before the beginning of the refining, on the basis of considerations suggested by the experience and modified, during the refining itself, on the basis of the variation of some values, such as for example the intensity of some frequencies of the noise of the converter, or on account of the subjective impressions of the operating staff. This kind of operation which requires a great deal of attention and of skill from the staff, turns into an actual uncertainty as far as the obtainable results are concerned; in fact, the variation of the operational values considered, affects in various ways the trend of the refining, depending on the conditions existing in the converter, such as the temperature and composition of the bath and of the slag, the amount of slag, etc. It is clear then, that in a situation like this, (where all the values are variable, and vary in a different way, depending on the general conditions existing in the converter), modification of some parameters without any possibility of an objective control, does not insure the required reliability in carrying out the refining process.

In order to obviate these drawbacks, refining processes in which an objective control of the operational procedure is effected, have been proposed to this purpose. Substantially, mathematical models allowing the determination of the optimal course of the process have been proposed.

A model proposed by the Italian Pat. No. 766,770 is based on the following equation:

at the stack. By integration of the equation, the amount of oxygen to be blown, c, in order to reach the required final value of carbon in the bath, is calculated a at a certain instant approaching the end point. At the same time, by means of another empirical model, not amenable to a simple formula. which expresses the var iation of the bath temperature as a function of the amount of blown oxygen, the amount of oxygen 0 still to be blown, in order to reach the final required temperature, is calculated. If 0, 0 everything is proceeding normally; if instead O 0, one must intervene either adding coolants or modifying the distance between the lance and the bath.

The operational procedure proposed in the patent under consideration, presents a number of limitations, among which the most significant are the following:

the analysis of the flue gases at the stack supplies data affected by errors and it is not possible to determine, with sufficient precision, the various values and especially the constant k.

in the case where 0 is different from 0 the proper corrections of the process conditions are performed, but in so doing the value of the constant k, calculated before the corrections, is also modified; therefore, in order to make the model correspond to the actual process conditions, it would be necessary to recalculate the new k value and consequently the new 0 and O values, but in pratice theie is no sufficient time to perform these calculations and the further corrections of the process conditions.

the fundamental parameter on which the control and the specific decarburization velocity is based; in other words, the model takes practically into account the decarburization reaction, while the oxygen supplied through the lance, besides reacting with the carbon of the bath, reacts also with the other present slag-forming elements, as well as with the carbon monoxide.

Other techniques have been proposed, in order to make it possible monitoring the refining pattern, but for some of them, as in the technique mentioned above as an example, an intervention is planned by the end of the refining, aiming at determining the amount of oxygen still to be blown, in order to attain the required conditions. Obviously these techniques cannot be reliable, insofar as they do not work during most of the refining process and they are put into operation only when a short time is still left for influencing with a certain effectiveness.

Furthermore, the control of the refining pattern is performed by means of an analysis of the fumes at some location of the fume collecting and depuration system, far away from the converter and generally downstream of the waste heat boilers, and from these data a determination of the fumes analysis at the converter mouth is effected. Thesemethods, however, are made questionable both by the large amounts of nitrogen present in the analyzed gases, coming from the air drawn into the hood and by the uncertainty inherent to the determination of the extent of the combustion reactions of CO into CO which occur outside of the converter.

The object of the present invention is a refining process for metals, particularly for molten pig iron, in stationary or rotating oxygen converters, which allows to obviate the afore mentioned limitations, by continuously monitoring the refining process.

The process, which is the object of the present invention, is characterized by an uninterrupted control'technique of the refining, in which recourse is made to a dynamic model based on the instantaneous balance of the oxygen blown through the lance, calculated by means of the analysis of the discharge gases at the hood and at the stack, taking into account that the said oxygen reacts not only with the carbon of the bath, to form carbon monoxide, but also with other elements present in the bath, to form slag, and also, in the region above the bath, with the carbon monoxide obtained in the initial reaction, to form carbon dioxide.

In order words, the following three reactions are taken into consideration:

a. /20 C C c. 0 Fe, Si, Mn. .=Slag which represent respectively the decarburization reaction of pig iron, the combustion reaction of carbon monoxide and the slag forming reaction.

Each of these three reactions, considered separately, yield essential information for the process control. Reaction (a) indicates how the decarburization is proceeding; reaction (b) allows to know how the thermal yield of the conversion itself is varying in the course of the refining; reaction (c) gives a measure of the amount of slag which is forming and this is of a great interest as far as the regularity of the refining and the control of the metallic yield are concerned.

Actually, as it will be seen subsequently, only reactions (a) and (b) are used directly in the oxygen balance instantaneous calculation, since the amount of oxygen forming the slag is obtained by subtracting the amounts which react according to (a) and (b) from the total amount of blown oxygen. However, the partial balance according to (c) is important, in that it explains why, in some periods of the blow, more oxygen is used for reactions (a) and (b) than the amount blown (which means that a portion of the slag previously formed, is decomposing, bringing back into the liquid metal bath, the corresponding portions of iron, silicon, manganese, etc. mainly iron which previously had formed the slag).

According to the present invention, a metallurgical reference parameter is determined, as a function of the reactions occurring inside the converter during the refining, and of the amount of oxygen blown. The said reference parameter is defined by the following instantaneous experimental relationship:

where A instantaneous flow rate of the oxygen employed for the combustion of C into CO, in Nm /min.

B instantaneous flow rate of the oxygen used for the combustion, inside the converter, of CO to CO in Nm /min.

Q instantaneous flow of the total oxygen blown through the lance in Nm lmin K a constant characteristic of the type of refining.

tionship cannot be used to establish in advance a blowing pattern.

It must be observed that even though the distance of the lance from the bath is a control variable'very important for the development of the refining reactions, it does not appear explicitly in the formula which defines parameter m; that is due to the fact, experimentally observed, that the refining process reacts with a considerable inertia to the height variations of the distance between the lance and the bath, while the response to flow rate variations is much prompter.

In order to utilize the metallurgical parameter as above defined, according to the present invention optimal curves are experimentally established, indicating the variations of the said parameter m as a function of the blowing time, each one corresponding to one of the various types of refining which are planned for execution, then for each curve, a range or variability band is established where the actual value' of m can oscillate without any negative effect on the satisfactory progress of the heat. These data are stored in the memory of an electronic computer.

For an actual refining, the type of the charge and the characteristics required for the steel at the end point are preestablished. On. the basis of these data the com puter chooses the relevant type of optimal curve and, during the blowing, receives the data which allow it to calculate A and B and thence the ratio K (A B)/Q m. Then it compares the value of this ratio at eac instant of the blowing to the optimal value preestablished for this instant, and if necessary, modifies the flow rate Q in such a way that the actual value of m be the nearest possible to the optimal value, or at least, be comprised within the preestablished variability range.

In this way a considerable progress is insured over the known art in this technological field, where the chosen metallurgical parameters did not allow, as previously mentioned, to obtain reliable results, or where separated consideration and evaluation of different variables even though collectively combined in the herein mentioned parameter m did necessarily impose an arduous correlation of several data.

The methods of sampling and analyzing the gases and thepalculation model for A, B and margde s cribgd hereunder.

The refining control is performed by means of a continuous series of samplings and analyses of the gases issuing from the converter, for the purpose of ascertainmg:

the composition of the gases (CO +CO2) at the converter mouth, before their combustion with air. the composition of the gases at the stack of the steel ShOp. ()2 the measure of the pressure P, of the pressure drop AP, of the temperature T, of the flue gases of the stack for the calculation of the flow rate of the gases in the dry state in Nm /sec.

According to the present invention, the analysis of the gases at the converter mouth is performed by an indirect measurement, by sampling of the gas with a sampler within the hood, at adistance from the converter mouth ranging from 0.3 to 2.5 meters. A long series of tests has shown that, in the hood, at the above mentioned distances from the converter mouth, only the outer zone is strongly concerned with the variations of composition due to air entry, while in the central zone these variations are minimal, thus causing errors which are within the required precision limits. The sampler, which can be tilted from 0 to 90 in respect to the horizon, is thus introduced into the hood, in such way that its suction extremity is located approximately in the center of the hood itself, at a distance of 0 to 0.5 meters from the axis of the hood itself, and at a distance from the converter mouth ranging from 0.3 to 2.5 meters.

According to the present invention, it is also possible to perform the sampling of the gases at a distance from the converter mouth, lesser than 0.3 meters or even inside the converter itself, by means of a sampler directly incorporated in the blowing lance or attached to it, or separated from it and introduced into the converter mouth; however the preselected technique (that is the sampling of the gases in the hood) is to be preferred as being the most practical, since no further lifting apparatus for the sampler is necessary, and the chance that a spray of slag or liquid metal plugging the sampler is less probable; besides, the composition differences between the gases sampled inside the converter and those sampled with the suggested method are not too important.

The gas sampled by the sampler is first depurated of dust and then conveyed to a continuous analyzer, of a conventional type, for CO and CO2 analysis and then to another continuous analyzer, also of a conventional type, for the analysis of oxygen. The nitrogen percentage possibly present is calculated as a complement to one hundred of the total of the C0, C02 and 02 per centages. The instantaneous values of the thus determined percentages, are recorded and simultaneously fed to an electronic computer.

In the stack, the gases are practically cold and saturated with water vapor, because of the countercurrent water cooling and are already depurated of dust; they are sampled by a sampler and conveyed directly into an analytical system, similar to that installed in the hood, and the analysis data thus obtained are also recorded and simultaneously fed to an electronic computer.

In the stack, by means of a venturimeter-and a series of thermoresistances, measurements of the differential pressure AP, and of temperature T, are also performed. These data too are recorded and fed into the computer.

As soon as the blowing is initiated, the computer be gins to read, with a proper frequency, the values corresponding to the single measurements mentioned above. In real time too the computer elaborates the input data calculates the instantaneous value of m and compares it with the value preestablished for this particular type of refining at the corresponding instant.

The calculation procedure is the following where the data a e indicated with:

AP the differential pressure at the stack in kg/m P absolute pressure at the stack in kg/m T temperature of gases at the stack in C Q the oxygen .flow rate from the lance in Nm-"/min.

% COZF 1.532 7472+ 0.532 2.532 P 53.2)/(% co (TB- .2. Water vapor pressure in the gases at the stack, in

kg/m

P exp (23.43097 5260.704/7C 273) where 7C corresponds to the temperature of the gases in the stack.

3. Specific weight of the gases saturated in water vapor at the stack, in kg/m 4. Flow rate of the dry gases at the stack, in Nm lmin Q 1 LF! P- P /10330 273/T+ 273 where K is the venturimeter constant in the stack. 5. Oxygen used for the combustion of C into CO, in Nm /min 6. Oxygen used for the combustion in the converter of CO into CO in Nm lmin B CO /l00 X A 7. Parameter m m K A B/QO,

Then, the computer compares continuously, during the blowing, the actual instantaneous value of m with that indicated by the preestablished curve, and gradually, by trials, prescribes the corrections to be possibly made to the distance between the lance and the bath and/or, preferably, to the flow rate of the oxygen delivered by said lance, in order to bring the instantaneous value of m back to the optimal value.

The present invention will be hereunder described in greater detail with reference to the embodiments supplied, in a non limitative way, in the following examples, and with reference to the drawings where:

FIG. 1 shows an optimal curve 1 of the parameter m, as a function of the blowing time, with the relevant allowance band or strip preestablished for a certain type of charge and for a certain requested result and delimited by the dashed lines 2 and 3.

FIG. 2 shows the actual curve 4 of m, related to the heat indicated in Example I, superimposed over the curve of FIG. 1.

FIG. 3 shows the actual curve 5 of m, relative to the curve of the heat indicated in Example 2 and superimposed over the curve of FIG. 1;

FIG. 4 shows the actual curve 6 of m related to the heat indicated in the Example 3, superimposed over the curve of FIG. 1.

Some examples of practical embodiment of the pro-' cess, according to the invention, are now following.

EXAMPLE 1 Characteristic data liquid pig iron 245.4 tons steel scrap 88.6 tons lime l5.6 tons and the total volume of O to be blown 15,600 Nm After the charge of liquid and solid pig iron and of steel scrap has been effected, the blowing is initiated with the following blowing parameters values flow rate 700 Nm /min lance distance from the bath 1.800 m Since the beginning of the blow the computer performs the calculation of m, with a frequency of 3.84 sec, but the automatic control of the blowing parameters initiates only at the 4th minute of blowing.

During the first two minutes 13.6 tons oflime and 0.5 tons of ore are loaded into the converter.

At the 4th minute the automatic control initiates and consequently the lance is lowered down to 1.50 meters.

During the remainder of the blowing period, the lance position is kept constant, while the oxygen flow rates are permitted to vary within an interval ranging from 600 to 800 Nm /min.

At the 17th minute of the blowing, 2 tons oflime and 600 kg of fluorite are loaded.

At the 19th minute the oxygen flow rate is 800 Nm /min, the lance distance from the bath is still 1.50 meters. The automatic control is interrupted at about 2 minutes before the end point.

As can be seen in FIG. 2, the actual parameter m remained within the optimal interval. The results at the end point were the following:

Bath temperature l605C 71 carbon in the bath 0.054

"/2 Mn do. 0.23

% P00 in the slag l7.2 Steel weight at the end point 328 tons Iron yield 95.7 7:

The iron yield, at the allowance limits, is due to a sparking period during the first 3 minutes of blowing.

EXAMPLE 2 Liquid pig iron T l300C; '/r Analysis C 4.52

Si 0.53, Mn 0.88; S 0.039. P 0062 Solid pig iron 7.0 tons Iron ore 4.6 tons Conditions required at the end point:

I600 i 10C Iron yield; )5 "i slag hasiuity index 4.0

The on-line computer on the basis of the previous data has indicated the following compositions of the charge:

Liquid cast iron 272.6 tons scrap 57.9 tons lime 14.0 tons and the oxygen total volume to be blown 15,000 Nm After the solid and liquid pig iron and steel scrap have been loaded, the blowing is initiated with the following values of the blowing parameters:

0 flow rate 650 Nm"/min Lance distance from the bath 1.80 m

During the first 3 minutes of blowing all the lime, the ore and 200 kg of fluorite are loaded.

From the 4th minute, the automatic control is initiated. The course of the parameter m during the refining is given in FIG. 3.

A slight slopping occurred from 6 to 6'40 of the blowing.

The final data after 20 of blowing are:

T of the bath 1602C 71 S 0.020 Steel weight 3l2.3 tons lron yield 96.2 71

72 FeO in the slag 20% For the purpose of stressing the advantages connected with the method proposed by the present invention, a series of heats was performed in which the computer recorded the course of parameter m during the blowing. but did not order the corrections whereas the control of the refining was effected by hand following the subjective impressions of the staff responsible for the heat. An example of heat performed in this way is given hereunder:

EXAMPLE 3 Characteristic data Liquid pig iron T= l,270C; Analysis:

C =4.5l Si =0.78; Mn =0.8 S 0.032; P 0.063 Iron in pigs 12.3 tons iron ore 2.5 tons Required conditions at the end point:

Bath temperature l600C l0C 71 C 0.060 t 0.020 Iron yield a 71 Slag basicity index 3.5

The on-line computer, on the basis of the previous data, indicated the charge composition as follows:

Liquid pig iron 270.2 tons Scrap 53.0 tons Lime 17.8 tons and the oxygen total volume to be blown 15,700 Nm After the solid and liquid pig iron and-steel scrap have been loaded, the blow is initiated with the following blowing parameters:

;- flow rate 700 Nm- /min Lance distance from the bath 1.8 m

The lime, the ore and 200 kg of fluorite were loaded during the first 3.

At the 4th minute the lance distance from the bath I was brought to 1.50 meters. The oxygen flow rate was maintained constant until the 17th minute, and thereafter was brought to 800 Nm /min.

The curve showing the course of the parameter m during the blowing is given in FIG. 4.

The final data of the metal bath are:

"/r Mn 0.28

Steel weight 298.6 tons Iron yield 92.2 71

The present invention has been described with particular reference to some specific and preferred embodiments thereof, but it is intended that modifications or variations may in practice be introduced therein, without departing from the scope of the present invention as defined by the appended claims.

Having thus described the present invention, what is claimed is:

1. In a process for refining liquid pig iron into steel in a liquid bath, comprising blowing a gas containing oxygen onto the bath, and analyzing the gas above the bath; the improvement comprising the steps of predetermining a desired graphical curve of m versus time on the basis of at least the actual initial and desired final physical characteristics of the bath, in which A being the instantaneous flow rate of the oxygen in said analyzed gas employed for the combustion of C into CO in Nm /min., B being the instantaneous flow rate of the oxygen in said analyzed gas used for the combustion of CO to CO in Nm /min., and Q being the instantaneous flow rate of the total oxygen in said blown gas in Nm /min., k being a constant; determining at a plurality of time intervals the actual instantaneous values of m on the basis of the physical characteristics of gas emitted from the bath; and varying the flow rate of said blown gas so as to decrease the difference between each said actual value of m and the predetermined value of m at the same said time interval on said graphical curve until physical characteristics close to said desired final physical characteristics of the bath are achieved.

Claims (1)

1. IN A PROCESS FOR REFINING LIQUID PIG IRON INTO STEEL IN A LIQUID BATH, COMPRISING BLOWING A GAS CONTAINING OXYGEN ONTO THE BATH, AND ANALYZING THE GAS ABOVE THE BATH; THE IMPROVEMENT COMPRISING THE STEPS OF PREDETERMINING A DESIRED GRAPHICAL CURVE OF M VERSUS TIME ON THE BASIS OF AT LEAST THE ACTUAL INITIAL AND DESIRED FINAL PHYSICAL CHARACTERISTICS OF THE BATH, I WHICH

Images