DE2234204B2 - Process for refining metals - Google Patents

Description

4- B

is determined, where k is a constant characteristic of the type of hairdressing.

2. The method according to claim I. characterized in that that gas samples are taken from both the flue gas chimney and the hood or converter.

3. The method according to claim 2, characterized in that that the gas samples of the hood at a distance of 0.3 to 2.5 m from the converger mouth can be removed.

4. The method according to claim 3, characterized in that that the gas samples of the hood at a distance of 0 to 0.5 m from the height axis can be removed.

55

The invention relates to a method for refining metals, in particular liquid pig iron

Steel, in converters or the like. By inflating lydating gases with the help of a computer (»0 Subsequent measurement and control of the flow rate of the gases and / or the distance of the first bowl from the surface of the melt, based on certain characteristic processing conditions a metallurgical parameter that changes over time for the process of l'riliens determined and nuciicbcn in the memory of the computer, which is during the fresh process automatically the one from exhaust gas measurements The actual value of the metallurgical parameter is calculated, compared with the optimal target value and, from this, the Control derives.

It depends on the implementation of deletion procedures. that constantly, i.e. / u at every point in time, The 1-year and the state of hairdressing determined and the process is uninterrupted and the level of organization is normal Program for the melt, in particular the Temperaluruerl. is returned. by the Sauer - ;.> ffbcdarf is monitored with the Aim to keep all the heat in the melt in temperature of the final C content and the steel components in the door the batch provided Hold area / u.

When refining pig iron by blowing oxygen onto the surface in the designated area Converters play the natives under pressure of inflated oxygen and the distance of the lance from the surface of the melt plays a decisive role as control parameters for the process.

Both the distance between the lance and the molten metal and the size of the oxygen flow are frequently used before the start of hairdressing due to Irfahrunt'swericn preselected and modified during the freshness process. e.g. due to the frequency band of the blowing noises of the converter or the subjective pressure of the operating personnel.

This type of control of the process flow. the attention and experience of the operating staff makes high demands is unreliable in terms of the results achieved. In indeed affects a change in said procedural conditions which are dependent on the state present in the crucible. /. B. the temperature and composition of the melt and the slag, their proportion, etc. the process of hairdressing in vetschicdcner Respect.

Of course there is in a situation like this (in which all values are changeable and dependent of the overall state in the converter can change in different directions) changing some Parameters without a possibility of objective control do not guarantee the correctness of the Procedure for freshening.

To avoid these disadvantages, processes for freshening have already been proposed an objective control of the process sequence is carried out. It is for that purpose even proposed mathematical models and programs to determine the optimal process flow been.

The model proposed in Italian patent 7 66 770 lies in the equation

in which R indicates the percentage of burnt carbon per cing-blown Nm ¹ O ₂ as a function of the percentage carbon content c of the melt. .1 and k are arithmetically determined constants based on exhaust gas measurements.

_{By integrating the equation, the amount of oxygen O 1} required to achieve the final C content of the melt is calculated assuming a certain point in time at the end of the line. At the same time, through another empirical model that is not through a simple one

tneihematische equation expressing läl.U and the change of Schmelzenlemper.itur depending ton the amount of oxygen blown representing the amount 0 of the Hi submission 'the required Fndtemperatur still to be blown oxygen ^s is calculated. When O ₁ - O is. the procedure runs normally: however, if O ₁ differs from O '■'. you either have to provide cooling or the distance between the animal lance. the melt change.

The one referred to in this patent. Vci- (tihren however, there are defects. among which the following are the most important:

The analysis of the flue gases from the chimney does not give correct values and it is impossible. the different quantities, especially the constant λ. with sufficient accuracy / u determine.

At different O ₁ and O _(the Korreklurwerte will probably zo for Verfahrctishedingun- gen calculated changed with these as well as the value determined before the corrections of the constant L. I'm the model to the existing at the moment the process conditions adapt, must therefore first the constant />: -, and then again O ₁ and O, can be calculated. In practical operation, however, the time is not sufficient to carry out this calculation and to make further corrections
As a basic parameter, according to which the control and the specific decarburization speed are based, the model is practically based on the combustion of carbon, while the oxygen blown out by actually blowing out the oxygen in addition to burning deep carbon also with the elements in the slag and reacts with the CO.

With other techniques / to monitor the firing process. as in the previous example for ^ ₀ the desired conditions against Inde of Frisehens fcrreichung a Finschrcilen for determining ties planned still to be blown oxygen.

It is evident that these techniques are not reliable because they are true! the duration of the ₄ ς t-'rischens are not effective and are widely applied. When there is little time left for effective inflow.

In addition, the process is checked by analyzing smoke gases in the process Flue gas collection and rejection system far from the crucible and generally towards the flue gas / heat exchanger. and from these measured values one then deduces the condition at the crucible outlet |) The value of this method is questionable in view of the grumpy Stieksloffantcil in the analyzed l'ias. that is sucked in with the air, and as a result animal I am certain about those outside the crucible before you jichendc combustion of CO and CO-.

The same applies to the well-known ancestor animal ₍ , o initially mentioned type (Dl-OS 14 33 4431. in which also from animal composition animal already with additional air mixed smoke gases must be inferred on the conditions within the converter. The accuracy of the control allows accordingly (, _ <; ιίι to be desired.

The task is based on the Procedure tier eimianiis sienannien kind to this effect / 11 that a control of the fresh process becomes possible with increased accuracy.

To solve this problem, the method according to Tier Hrfmdung is characterized by the fact that the target word tics mechanical parameters οι as a time-dependent curve with tolerance band from the: instantaneous flow rate 1 of the only carbon monoxide reacting SauciVioHes. which is a measure for the degree of Aufkohhmgsgrad the melt, and the current flow rate Ii of the carbon dioxide reacting oxygen with the formation of carbon dioxide, which is a measure of the thermal efficiency of the I mscl / ung, as well as momentarily with the first / e supplied oxygen flow rate O ₁₁ according to the equation

,, 1 ■ Ii

m! ■:

is determined, where /. cmc for the type of I riseben-indicative Is constant.

L- is a dynamic regulation, taking into account that the cuige hl a sen e oxygen only add flour with the carbon animal melt Kohicnmonowd. but also with other themes in tier melt reacts to form slag and 'one above tier melt / e the first reaction from tier he / eu;! le carbon dioxide burns to carbon dioxide.

In the method according to Tier FiTmdiirig one obtains direct information about which processes within of the converter run and what amounts of oxygen on the one hand for the carbonization and on the other hand IVu consumed the oxidation of carbon monoxide graze. The stenciling is therefore the respective one actual condition within ties convoy icis. compared based on the specified optimal setpoint. Control takes place accordingly direct and is no longer dependent on falsifications, such as those caused by subsequent oxidation among animal Directly of additional air are unavoidable.

Expressed in formulas, in principle! 'Are eradicating Reaction schemes to consider:

; i | I 2 (), i C = CO.

bl I 2O ₁ + CO = - CO ₁ .

O O, -! - "Fe. Si. Mn..." -. R slag.

with which the decarburization of deep pig iron, the burning ties coal moiiowds and the formation of Slag are described.

Each of the three reactions provides essential information for supervising the process, namely there

as the progress of decarburization.

b) Information about the change in the thermal Efficiency during the duration of the hairdressing.

this is and is a measure of the visual lacquer vtin great interest in assessing animal Regular wedge of hairdressing and control of the SlahlausboMie.

In fact, only the reactions according to tl equations are al tint) b | for the calculation of the currently required amounts of oxygen are used, there is the sludge forming the sludge

by subtracting the after the reactions after a) and It) required amount of oxygen from the total the amount of blown sauce is determined and. I lotzdeni the determination of c) is important because it provides information there is why in certain phases of hairdressing the need for oxygen for the reactions a) and h | is greater than the inflated amount (which means that a part of the visual varnish reduces! and and the corresponding quantities VDiI liiseii in the molten metal. Silicon. Manganese etc .. specifically iron that previously formed the slag, returns).

The previously mentioned values Λ and B are, in addition to their dependency on () ₍ >, during the duration of hairdressing also variable over time depending on other parameters, such as the distance of the lance from the surface of the melt, the size of the The area of impact of the gas jet on the molten metal, the temperature of the molten metal, the amount of slag, etc. in a manner that cannot be calculated, with the result that the experimental relationship previously used cannot be used for a successful, predeterminable freshness program.

It must be noted that. although the distance between the lance and the melt is a very influential control variable. this does not appear explicitly in the equation for the parameter m . This is due to the empirically found fact that the process of hairdressing reacts with remarkable inertia to changes in the height of the lance from the surface of the melt, while the reaction to changes in the flow rate is much more immediate.

For the application of the above-defined metallurgical parameter according to the invention, optimal curves have been determined experimentally which show the change in this parameter m as a function of the time of hairdressing and each of which corresponds to one of the various types and programs of hairdressing . In addition, a tolerance band is defined for each curve, within which the measured value of m can fluctuate without adverse consequences for the satisfactory execution of the program. These data are in the memory of the computer.

For a specific application of hairdressing, the type of charge and the values that are characteristic and required for the steel at the end of hairdressing are preselected. On the basis of this data, the computer selects the optimum curve available and receives the measured values during the blowing which enable the calculation of A and B and then the ratio k {A + B) ZQo, ⁼ '". It compares this value during of hairdressing with the optimal value given for the moment and changes, if necessary, the flow rate Q _{Oj in} such a way that the current value of w comes as close as possible to the optimal value or at least lies within the tolerance band.

In this way, a considerable advance over the prior art is achieved, in which the chosen metallurgical parameters as stated above do not make it possible to achieve satisfactory results, or separate consideration and evaluation of various variable quantities even if these are combined and to the parameter mentioned in summarized are necessary

often lead to a difficult interrelationship between different data.

The method of taking and analyzing the gas proheii and the framing scheme for Λ. Β s and / 11 will be described below: The control iles Frisehens is performed using continuous sampling and analyzing of samples of the discharged gases from the converter with the aim of determining or measuring

the composition of the gases (CO and CO ₂ I at the converter mouth before they are burned with air.

the composition of the gases in the flue gas chimney (CO, O ₂ and N ₂ ).

of the pressure csp and the pressure drop \ p. the temperature 7 "of the exhaust gases in the flue gas chimney for the calculation of the flow rate of the same in the dry state in Nnr ^v see.

According to the invention, the gases at the converter mouth are analyzed by indirect measurement, namely by taking samples by means of a sampling line in the hood at a distance between 0.3 and 2.5 m from the converter mouth.

A longer series of tests has shown that in the hood at this distance from the converter mouth only the composition of the outer zone is changed by adding air. while in the core area these changes are very, slightly

and cause errors that are within the accuracy limits lie. The sampler. the one opposite the horizontal between 0 and 90 can be tilted, is inserted into the hood, that its proboscis is approximately in the middle of the hood, between 0 and 0.5 m \ on their axis removed and 0.3 to 2.5 m away from the converter mouth.

According to the method according to the invention, it is also possible to take the gas samples in a small

ncrcn distance than 0.3 m from the converter mouth or even from inside the converter itself by means of an existing on the lance, attached to it or separately next to it in the converter mouth Introduced sample taker to carry out. However, the one described first is preferable Technique of taking gas samples from the hood, since then there is no special device for lifting the Sampler is required and its clogging by the molten metal or slag drops un-

is likely. Incidentally, it should be mentioned. that differences in composition between the gases taken from inside the converter and the gases taken from the hood are practically insignificant.
The Ga taken with the sampler:

is first freed of dust, then by a conventional, continuously working analyzer for the determination of CO and CO ₁ , and finally by a second, also continuous analyzer for the determination of O ₁

Any proportion of N ₂ that may be present results in the calculation difference compared to the proportion of CO. CO ₂ and O ₂ related to "00". The currently determined percentages are recorded and fed into a computer at the same time

The gases are practical in the flue gas chimney cold, saturated with water vapor because of the countercurrent water cooling and already vo Dust cleaned. You are going through a rehearsal here

taken enlnelimer and directly from the Aualvsatoi-Sysiem supplied, similar to the one built into the hood. Also these won here Analysis data are recorded and immediately / quickly sent to the computer 1.

In Rauehgaskamin a number \ be on the pressure drop thermocouples I /) and the temperature / measured by means of a venturi tube uni \. These data are also recorded and fed into a computer.

As soon as the sound begins, the computer begins to read the measured values obtained from the measurements mentioned at a suitable frequency. With the reading of the incoming data, the computer calculates the respective existing value of in and compares it with the value of this moment and this special value of the level in the window.

In the following invoice, the cm / a 1 \ iten mean:

'«CO: content of CO in the 1 la exercise in percent by volume. "« ('() ,: Gehail an CO, in i] cv hood in percent by volume.
¹ H θ ",: content of O, in the hood in percent by volume.

"ο CO ₂ : content of CO in the flue gas chimney in percent by volume.

"·> O ,: Gehall to O, in the rough gas chimney in percent by volume.

I / ;: pressure drop in the flue gas chimney in kg no. p: absolute pressure in the rough gas chimney ii

kg irr.

7: Temperature in the rough gas chimney in C.
Qa: flow rate of the O ₂ from the lance it Nm ³ min.

The computer then takes the following calculation before:

I. CO ₂ -AnieU at the Konververtermündiing in percent by volume:

1.532 _{% rf} r t 0.532., Cö + 2.532, ^ - 53.2
% ΓΌ +% UJ,

2 Water vapor pressure in the gases in the flue gas chimney in kg no:

(23.43

4309.00

where T C corresponds to the gas temperature in the rough gas chimney.

3. Specific weight of the gases saturated with water vapor in the flue gas chimney in kg'm ³ :

0.N04

ι on ρ

1.429.Y =

1,250 x "

P 273

10330 7T273

-I. Flow rate of dry gases in the smoke paskamin in just 'mm:

O ^ K

. j IP / '- ■ Pum 273

10330 '/ ■ -'- 273'

where K. is the N'enturircihr constant in the smoke gas chimney.

> ■ 1-iir the combustion of ("." In CO bcnotigier Oxygen m only min:

■ 1 - "0.5 ■ O

% CO ₂ 100

'"> for the combustion of CO to CO _: oxygen required in the converter in Nm ¹ min:

Parameter no.

»Ι - K

Then \ ergleicht the computer during the Frisehens continuously the instantaneous value ^v "n m with the target value of the predetermined curve and writes stepwise by experiments the corrections ^v ° r., The 7th example, by changing the distance of the lance ^v" of the surface n of the Melt and / or preferably the flow rate of the oxygen supplied through the lance are to be carried out to the instantaneous value of. ■ ?; to bring it back to its optimal value.

The invention is described below with reference to specific exemplary embodiments and the schematic Drawing explained in more detail. In the drawing shows

1 shows an optimal curve 1 for the parameter m as a function of the time of freshening with a predetermined Tolcranz band delimited by the dashed lines 2 and 3 for a certain type of batch and a certain desired result.
Fig. 2 ais curve 4 the actual course of m.

<-o plotted against the curve according to FIG. 1, for the batch from Example 1.

3, as curve 5, shows the actual course of m. Plotted over the curve according to FIG. 1 for the batch of example 2, and

F i g. 4, as curve 6, the actual course of nu plotted against the curve according to FIG. 1 for the batch of example 3.

Some examples of the practical application of the method according to the invention are given below described:

Example I.
Characteristic data

Liquid pig iron 7 = 1420 ^r C: percentage analysis: C = 4.62: Si = 0.69; Mn = 0.68: S = 00 "> 0 P = 0.060.
Ingots: 2Ί.3 t. Iron ore = 0.51.

Required end dates:

T = 1600 C ± 10 C.
% C = 0.060 ά 0.020.
Yield ^ 95%.
Basicity of the slag

= 3.5.

The computer as a process computer has due to the output data and the preselected program the following composition is calculated for the batch:

Liquid pig iron 245.4 t

Steel scrap 88.61

Surcharges 15.6 l

Total amount of O ₂ 15 60ONm 'to be blown

On Grünt, the computer as the process computer indicated earlier Dalen as the composition of the batch

Liquid pig iron 272.6 t

SchroU 57.91

Limestone 14.0 t

Total oxygen requirement 15,000 Nm

After introducing the liquid and solid pig iron; and put scrap steel in the crucible, it became fresher started with the following values of the parameters

Amount of flow of oxygen 650 Nnv ¹ min Distance of the lance from the

Melt 1.8 m

After the batch of liquid and solid pig iron and steel scrap has been introduced, the process begins blowing with the following parameters

Flow rate O ₂ 7 (X) Nm ³ min

I distance from the
Melt 18 (X) m

From the beginning of the freshness, the computer calculates the parameter m at a frequency of 3.84 seconds, but the automatic control of the freshness values does not begin until the fourth minute of blowing.

During the first 2 minutes, 13.6 t of lime and 0.5 t of ore are fed into the crucible. After the start of the automatic control in the 4th minute, the lance is lowered to 1.5 m. During the further refining, the height of the lance will remain constant, while the flow rate of the oxygen is changed between 600 and 800 Nm ³ min. In the 17th minute of freshening, 2 tons of lime and 600 kg of fluorspar are added. In the 19th minute the flow rate of oxygen S (K) Nm ³ min and the distance between the lance and the melt is still 1.5 m. The automatic control is switched off 2 minutes before the end. As one can see from Fi g. 2, the parameter »! within the tolerance band. The dates at the end of the procedure were:

Melt temperature 1605 C

Percentage of C in the melt. 0.054
Proc. Manganese content in the

Melt..? 0.23

Proc. Sulfur content in the

Melt 0.016

Percentage of FeO in the slag 17.2

Melted steel 328 t

Iron yield 95.7%

The iron yield at the tolerance limit is due to sparkling in the first 3 minutes of freshening traced back.

Example 2

Liquid pig iron T = 1300 C: percentage analysis: C ^ = 4.52. Si = 0.53. Mn = 0.88. S = 0.039. P = 0.062.

Solid pig iron 7.01.

Iron ore 4.6 t.

Desired dates at the end of freshening

Melt temperature 1600 ± 10 C

Carbon content 0.060 ± 0.020

Ice cream yield _95%

Basicity of the slag 4.0

is during the first 3 minutes of freshening the entire amount of lime, the ore and 200 kJ fluorspar are added to the melt.

Automatic control was carried out from the 4th minute. The course of the parameter η during the freshening is shown in FIG. 3 shown.

A slight spitting occurred from 6 'to 6'4 (V after the start of freshening. Fresh after 20 minutes The following data was available at the time:

Melt temperature 1602 C

Percentage C content 0.062

Proc. Manual switch 0.24

Percent sulfur content 0.020

Melted steel quantity 312.3 t

Iron yield 96.2%

Proc. FeO-Gchall in the slag 20

For the purpose of testing the advantages possible with the method according to the invention, a series of batches were run with the computer, in which case the computer kept the course during the freshness; Parameters m recorded without requiring the corrections zi, while the control of the freshening in front of the hand took place according to subjective impressions by the staff. The result of such a manual stcucrunt is shown below:

Example 3

Liquid pig iron T = 1270 CC = 4 51 "Si = 0 78 Mn = 0.88: S = 0.032: P = 0.063. Iron bar 12.3 t.
Iron ore 2.5 t.

Desired dates at the end of freshening

Melt temperature 1600 ± 10 C

Carbon content in% 0.060 ± 0.020

Iron yield> 95%

Basicity of slag 375

The computer, as a process computer, calculated the composition of the dei on the basis of earlier data Batch:

Liquid pig iron 270.2 t

Scrap 53Ot

Limestone Υ. '.'. '.'. '.', J 18t

Total oxygen requirement 15 700 Nm

After the solid and liquid pig iron and steel scrap had been placed in the crucible, the Fresh process with a flow rate of O, before 700Nm-min and a distance of the lance "uei the melt of 1.8 m started.

The limestone, ore and 200 kg fluorspar were watched for the first 3 minutes.

/ f 22

U Ii

In the 4th minute the lance was on a leash Absland \ on! .5Om brought. The amount of oxygen was his / ui 17. Minnie kept constant and then brought to SOO \ ητ 'mm. The curve with the course iles parameter m during the Iriscliens is in Fig.4 shown.

The dates / round of the proceedings were.

Melt temperature 1625 (

Per/. C-Cichall 0.0X4

ψπν /. Manganese hall 0.2N

ψτο /. Sulfur content 0.24

| * ro /. Fisenoxide content in the

Slag l ^l )%

Trschniol / ene steel quantity 2'iS.6 t

tiscnausbcute 92.2% ' ^s

204

Auller the advantages of the constant temperature and composition of the melt and an iron output Of more than 45%, the method according to the invention brings some advantages that from the point of view of precise control of the process are important how /. B. a limited amount of Fiscnoxydstaut in the exhaust gases (red smoke I. an almost completely consumed slom of the ranch gases csentuell with the known Changes during the whole process, shortest possible blowing time for each Type of batch. Slag guidance according to its composition and quantity at any given moment the requirements of the procedure. Obtaining a slag in the best-looking condition that reduces the risk overflow is limited to a minimum.

1 sheet of drawings

Claims (1)

Palent claims: 1. Process / for refining metals, in particular from liquid pig iron to steel, in converters or djil. by inflating oxidizing gases with the help of a computer measuring and controlling the flow rate of the gases and / or the distance of the lance head from the surface of the melt. whereby, on the basis of certain characteristic process conditions, a metallurgical parameter that changes over time for the hairdressing process is specified and entered into the memory of the computer, which automatically calculates the actual value of the metallurgical parameter determined during the freshening process, including exhaust gas measurements compares the optimal target value and derives the control from it, thereby showing that the target value of the metallurgical parameter 111 is a time-dependent curve with a tolerance band from the instantaneous flow rate 1 of the acidic substance which reacts with the carbon of the melt to form carbon monoxide, which is a measure of the degree of Aufkohlimgsgrad the melt, and the current flow rate B of the oxygen reacting with the carbon monoxide to form carbon dioxide, which is a measure of the thermal efficiency of the conversion, and the sow currently being stirred with the lance material flow rate O ₍₎ , according to the equation