JP4140939B2 - Converter blowing method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a converter blowing method, and more particularly to a method for determining the amount of lime input necessary for appropriately controlling the end point molten steel P (phosphorus) in the converter blowing.
[0002]
[Prior art]
Conventionally, in converter blowing, it has been an important subject to appropriately control the molten steel temperature, molten steel carbon (C) concentration, and molten steel phosphorus (P) concentration at the end of blowing. Among these, as for the method of controlling the molten steel temperature and the molten steel C concentration, various static controls and dynamic controls have been proposed and put into practical use.
As for phosphorus (P), as disclosed in JP-B-58-36644, a method of controlling P by adjusting the amount of coolant introduced, JP-B-58-39202, JP-A-4-41611. As disclosed in Japanese Laid-Open Patent Publication No. 4-187709, a method for controlling phosphorus (P) by performing trajectory correction after sub-lance intermediate measurement has been proposed. In addition to the above-mentioned methods, an effective method is proposed for a method of appropriately controlling the end point molten steel phosphorus (P) by optimizing the total amount of lime (the amount of lime input) input at the initial stage of blowing. The lime input is determined using a fixed standard, or the lime input is determined and input by modifying the standard each time depending on the experience and intuition of the site. It's real.
[0003]
[Problems to be solved by the invention]
Since the conventional converter blowing method is configured as described above, the following problems exist. That is, when the method of determining the amount of lime input based on a fixed lime standard among such conventional methods is employed, the following problems exist.
That is, the factors that determine the amount of lime input are: hot metal silicon concentration, hot metal phosphorous concentration, end point molten steel target temperature, end point phosphorus concentration target value, end point carbon concentration target value, main raw material blending ratio, furnace body erosion status, bottom blowing As the mouth conditions and lance nozzle conditions vary, not all of these factors can be ascertained, and it is impossible to create an optimal standard for all combinations, so the lime input was not optimal. .
In addition, there is a method of changing the lime input amount based on experience and intuition based on the lime standard, but even if this method is adopted, there are individual differences in the lime input amount, and the lime input amount is also optimal. did not become.
Therefore, in the case of excessive lime input, the cost increases. In the case of excessive lime input, the end point of molten steel carbon (C) is blown down in order to prevent the phosphorus (P) component value from being deviated or the phosphorus (P) component from being deviated. The subject that the cost raise accompanying the raise of slag FeO density | concentration had generate | occur | produced.
[0004]
The present invention has been made to solve the above-described problems, and in particular, according to changes in the blowing condition, by increasing the accuracy of the optimum amount of lime input, it is possible to prevent P detachment, It aims at providing the converter blowing method which reduced the unit.
[0005]
[Means for Solving the Problems]
In the converter blowing, the converter blowing method according to the present invention is a function of calculating lime input amount or calculating procedure of lime input amount TCaO = F (HMSi, HMP, TE, PE, CE, x1,..., Xm , Δa)
However,
TCaO: Conversion lime (CaO total amount) basic unit per converter charged pig iron,
HMSi: Hot metal silicon concentration,
HMP: hot metal phosphorus concentration,
TE: Converter end point molten steel temperature,
PE: Converter end point molten steel phosphorus concentration or post process molten steel phosphorus concentration,
CE: Converter end point molten steel carbon concentration,
x1, ..., xm: other factors,
Δa: A learning term that is updated for each charge using the blow smelting charge record,
, The end point molten steel temperature target value, the end point molten steel phosphorus concentration target value or the post-process molten steel phosphorus concentration target value, the end point molten steel carbon concentration target value are respectively substituted into the TE, PE, and CE, and the HMSi, HMP, x1,. , Xm is substituted with actual or planned values of each factor, lime input is determined and blown,
After each blowing, the value of Δa is updated using the actual values of TCaO, HMSi, HMP, x1,..., Xm and the estimated values of TE, PE, and CE.
The estimated values of TE, PE, and CE are
In the dynamic control model formula for TE, PE, and CE including the oxygen amount ΔO ₂ after the intermediate measurement, the coolant amount ΔSORE after the intermediate measurement, the carbon content CS of the molten steel at the intermediate measurement, and the intermediate temperature TS of the molten steel, , TS is substituted for the measured value in each sublance intermediate measurement, and ΔO ₂ , ΔSORE is substituted with the actual values from the respective sublance intermediate measurement to the converter end point.
[0006]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, preferred embodiments of a converter blowing method according to the present invention will be described with reference to the drawings.
First, as a result of analyzing the on-site operation data, the present inventor found that the optimum converter operation related to the determination of lime input amount is “within the range where excessive dephosphorization does not occur, until the end point molten steel carbon ( C) ascending "operation. Therefore, in consideration of the latest changes in blowing conditions, the end-point molten steel (C) target that has risen to a level that does not cause a problem in the subsequent process and the appropriate end-point molten steel phosphorus target that does not cause excessive dephosphorization are realized. It became clear that a procedure for determining the amount of lime input should be given.
Although this is the procedure for determining the amount of lime input, a fixed standard such as the conventional method cannot cope with changes in various factors and other changes in the blowing condition. The formula for dephosphorization was determined, and this learning term was added, and this was updated by using actual data every time the charge was completed. A method for learning about such a control equation is generally widely used in an end point temperature control model of static control, an end point molten steel (C) control model, and various model equations of dynamic control. However, regarding the determination of lime input, there is no recognition that the operation that changes the amount of lime input according to the change in the unexplained blowing condition that cannot be expressed numerically is the optimum operation, and static control. In, the dephosphorization formula is not included in the control model, and the learning term is not added to the lime input standard. The inventor of the present invention has completed the present invention, focusing on the fact that whatever the lime input standard is, it can be directly used as a dephosphorization (P) control formula.
Therefore, in converter blowing, lime input calculation formula or function expressing lime input calculation procedure TCaO = F (HMSi, HMP, TE, PE, CE, x1,..., Xm, Δa)
However,
TCaO: Conversion lime (CaO total amount) basic unit per converter charged pig iron,
HMSi: Hot metal silicon concentration,
HMP: hot metal phosphorus concentration,
TE: Converter end point molten steel temperature,
PE: Converter end point molten steel phosphorus concentration or post process molten steel phosphorus concentration,
CE: Converter end point molten steel carbon concentration,
x1, ..., xm: other factors,
Δa: A learning term that is updated for each charge using the blow smelting charge record,
, The end point molten steel temperature target value, the end point molten steel phosphorus concentration target value or the post-process molten steel phosphorus concentration target value, and the end point molten steel carbon concentration target value are substituted for the TE, PE, and CE, respectively, and the HMSi, HMP, x1,. , Xm are substituted with actual values or planned values of the respective factors, and the amount of lime is determined and blown, and after each blowing, the TCaO, HMSi, HMP, x1,..., Xm The value of Δa was updated using the actual value of TE and the estimated values of TE, PE, and CE.
Here, as estimated values of TE, PE, and CE,
In the dynamic control model formula for TE, PE, and CE including oxygen amount ΔO ₂ after intermediate measurement, coolant amount ΔSORE after intermediate measurement, carbon content CS of molten steel at intermediate measurement, and intermediate temperature TS of molten steel, CS , TS is substituted with the measured value in each sublance intermediate measurement, and the values obtained by substituting the actual values from after each sublance intermediate measurement to the converter end point into ΔO ₂ and ΔSORE are employed.
Further, in this case, the update of the value of Δa is preferably performed by correcting the TCaO to a constant HMSi in order to prevent a significant fluctuation effect of the molten silicon concentration.
As described above, the optimum converter operation is an operation of “increasing the end point molten steel (C) within a range not causing excessive dephosphorization until no problem occurs in the post-process”, thereby achieving an end point molten steel phosphorus concentration target. Is a numerical value that does not cause the phosphorus (P) to be out of specification, and the end point molten steel carbon concentration target value may be the end point molten steel (C) that has risen to a level that does not cause a problem in the subsequent process.
In the converter blowing method according to the present invention, as described above, an appropriate amount of lime is determined using a lime amount calculation formula including a learning term, so that the control accuracy of the phosphorus (P) concentration is improved. It is possible to increase the molten steel phosphorus (P) concentration within a range that does not cause the phosphorus (P) component to be out of specification. Thereby, a required lime amount can be reduced and a total cost can be minimized.
The present applicant has previously proposed an invention related to the present invention in Japanese Patent Application No. 9-279146. The content was a method using actual values instead of estimated values as TE, PE, and CE at the time of learning (update of Δa) of the method of the present invention. However, it is more effective to use the method of the present invention. Give the reason.
In the conventional dynamic control for controlling the end point C and the end point temperature, the following can be said empirically. For example, when blowing is performed under the same conditions and the same amount of lime, when CE (end point C) = 0.07%, the result is stopped at CE = 0.07%, and CE = 0. When aiming at 10% and finally stopping at CE = 0.07%, the level of PE (end point P or subsequent process P) (expected value, average value) is clearly different. The latter has a higher PE level. However, when CE = 0.07% is eventually blown off at CE = 0.07%, and CE = 0.07% is blown off at CE = 0.10%. However, if the target is CE = 0.07%, oxygen corresponding to that is used, and even if CE remains high, the slag is sufficiently oxidized and has sufficient dephosphorization ability. It is done. The actual CE information is not useful in estimating the phosphorus concentration after the intermediate measurement, but rather the amount of oxygen and coolant used since the sublance intermediate measurement time. There is. This can also be said from the fact that the end point P estimation accuracy hardly changes even if the end point C analysis value is used in the analysis of the end point P estimation formula.
That is, under the same blowing conditions and the same amount of lime input, the actual PE is an estimated CE (basically estimated CE = target CE for automatic blow-off in conventional control) rather than the actual CE. Has a strong correlation. Therefore, also in the update (learning) of the learning term Δa of the formula for calculating the input lime amount, as CE, the accuracy is better when the estimated value is used than the actual value.
The same can be said for TE. Originally, if TE (end point temperature) is high, phosphorus distribution will deteriorate and PE (P) will also increase, but if a high actual temperature is obtained for a low target temperature, it will oxidize a large amount of iron. The PE level is almost the same as when it was blown off at a low temperature as intended. For this reason, also in the update (learning) of the learning term Δa of the formula for calculating the converted lime input amount, the accuracy is better when the estimated TE value is used than the actual TE value.
Regarding PE, CE and TE cannot be said, but the following can be said for any of PE, CE and TE.
The static control model formula is usually a formula that controls the end point of converter blowing. However, after the sublance intermediate measurement, the static control is not used and the control is performed by the dynamic control type. Considering this, the static control substantially controls the sublance intermediate measurement appropriately. Therefore, when learning is performed, it is desirable to perform the learning by subtracting an unpredictable error after the intermediate measurement. In other words, only the error up to the intermediate measurement is learned. That is, the learning term Δa may be updated using an estimated value using information at the time of intermediate measurement. Therefore, learning is more stable when learning is performed using the estimated values of TE, PE, and CE at the time of intermediate measurement.
For the reasons described above, in the present invention, unlike Japanese Patent Application No. 9-279146, TE, PE, and CE values used for learning are not actual values but estimated values using dynamic control equations. To do. Thereby, as shown in FIG.1 and FIG.2, a precision improvement is obtained even compared with Japanese Patent Application No. 9-279146 previously, and the variation in the amount of lime and P is reduced.
[0007]
【Example】
Below, one Example of the converter operation by this invention is described. First, the concept of the lime input standard will be described. The lime standard is obtained by simultaneously solving the following dephosphorization equilibrium formula, heat balance formula, and mass balance formula. Specific calculation formulas are as follows.
[0008]
(1) Formula: Dephosphorization equilibrium type log ((P ₂ O ₅ ) / PE ² /(T.Fe) ⁵ )
= A1 · log ((CaO) + a2 · (MgO) + a3 · (T.Fe))
+ A4 / TE + a5
[0009]
(2) Formula: Heat balance formula TE = b1 · CaO + b2 · CaCO ₃ + b3 · MgCO ₃
+ B4.SV. (T.Fe) + b5.HMSi + b6.SORE
+ G (x1, ..., xn) + b7
[0010]
(3) Formula: Relational expression of (T.Fe) in slag (T.Fe) = c1 · CE + c2 · (SiO ₂ ) + c3 / TE + c4
[0011]
(4) Formula: P material balance formula 10 · (HMP · HMR + CMP · CMR + SCP · SCR)
= 0.429 · (P ₂ O ₅ ) · SV + d1 / [P]
0.429 is a stoichiometric coefficient of content in P ₂ O ₅ .
[0012]
(5) Formula: Si material balance formula 0.467 · (SiO ₂ ) · SV = 10 · (HMSi · HMR
+ CMSi · CMR) + 0.023 · SORE
0.467 stoichiometric coefficients of content in SiO ₂ Si, 0.023 is the content of the Si Sore.
[0013]
(6) Formula: CaO material balance formula (CaO) · SV / 100 = TCaO + 0.1 · SORE
0.1 is the CaO content in SORE.
[0014]
(7) Formula: MgO material balance formula (MgO) · SV / 100 = 0.15 · MgCO ₃ + e1
0.15 is MgO content in raw dolomite
Formula (8): Slag amount calculation formula SV · (100-1.29 (T.Fe)) / 100
= 2.14 · (HMSi · HMR + CMSi · CMR) + TCaO
+ 0.15 · MgCO ₃ + 0.15 · SORE + e1 + f1
[0016]
(9) Formula: Equivalent lime amount calculation formula TCaO = 0.95 · CaO + 0.55 · CaCO ₃ + 0.34 · MgCO ₃
0.95, 0.55, and 0.34 are the ratios of pure lime in each raw material.
here,
(CaO): CaO concentration in slag (%) Calculation formula (SiO ₂ ): SiO ₂ concentration in slag (%) Calculation formula (MgO): MgO concentration in slag (%) Calculation formula (T.Fe): in slag T.A. Fe concentration (%) Calculation formula (P ₂ O ₅ ): P ₂ O ₅ concentration in slag (%) Calculation formula PE: Converter end-point molten steel phosphorus concentration (%) Target value, estimated value TE: Converter end-point molten steel temperature (° C) Target value, estimated value CE: Converter end-point molten steel carbon concentration (%) Target value, estimated value HMSi: Molten silicon concentration (%) Actual value HMP: Molten phosphorus concentration (%) Actual value CMSi: Cold silicon concentration (%) Actual value CMP: Cold phosphorus concentration (%) Actual value SCP: Steel scrap phosphorus concentration (%) Actual value HMR: Hot metal ratio (%) Actual value CMR: Cold iron ratio (%) Actual value SCR: Steel scrap rate (%) actual CaO: burnt lime consumption rate (Kg / T-main raw material) formula CaCO _3: lime Ishihara units (Kg / T-main raw material) fixed amount MgCO _3: raw dolomite Hara (Kg / T-main raw material ) Calculation formula SORE: Sinter unit consumption (Kg / T-main raw material) Calculation Formula TCaO: Conversion lime basic unit (Kg / T-main raw material) Calculation formula SV: Slag amount basic unit (Kg / T-main raw material) Calculation formula G (•): Real value function
In the construction of the relational expressions of the above-described calculations (1) to (9), the coefficients in the relational expressions were determined mainly by multiple regression using the operation data of the 185ton converter used by the applicant. .
In the relational expressions of the calculations (1) to (9), HMSi, HMP, CMSi, CMP, SCP, HMR, CMR, and SCR are substituted with actual values, PE, TE, and CE are substituted with respective target values, and CaCO ₃ , If MgCO ₃ is a fixed amount, there are nine unknowns (T.Fe), (SiO ₂ ), (CaO), (MgO), (P ₂ O ₅ ), CaO, TCaO, SORE, and SV. Since there are nine, by solving this, the amount of main lime necessary for setting the end point temperature, end point carbon concentration, and phosphorus concentration before the start of RH treatment to the target values is obtained. The above calculation is very complicated and not suitable for on-site calculation. For simplicity, the following approximate expression was created and used together with the solution obtained by the above method.
(Approximate expression that can be used when HMSi ≧ 0.2%)
[0019]
Since the amount of lime input varies in a curved manner with respect to the molten silicon concentration, the approximate expression is a power function with respect to HMSi.
TCaO = g1 * (HMSi) ^g2 * [g3-g4 * (F '(CE / 10) -F' (g5)) + g6 * (TE-g7) + g8 * (HMP-g9) + g10 * (g11-PE) + X1 +... + Xm + Δa / g12]
(10)
here,
Δa: A learning term that is updated for each charge using the blowing charge record,
F ′: a real value function described later.
It is assumed that the input amount of limestone = 20 kg / ton-main raw material and the input amount of raw dolomite = 10 kg / ton-main raw material.
By substituting the end point target temperature, end point target carbon, phosphorus concentration before starting RH treatment, and actual values of hot metal silicon concentration and hot metal phosphorus concentration, which are preset for each steel type, into the above formula The amount could be calculated, and the phosphorus concentration in the end-point molten steel could be controlled to an appropriate range with high accuracy.
In addition, in order to obtain | require the estimated value of TE, PE, and CE used for learning, the estimated value was calculated | required using the dynamic model formula disclosed by Unexamined-Japanese-Patent No. 4-187709.
[0020]
That is,
Temperature model formula (1a)
TE = 0.81TS + 13.6ΔO ₂ −2.58 · ΔSORE / WCH
+ 0.27HMR-77.8 / F ′ (100 · CS) + 247.0 + learning term
Oxygen model formula (2a)
ΔO ₂ / WCH = F ′ (100 · CE) −F ′ (100 · CS)
-0.12 · ΔSORE / WCH + 0.017 · TCaO / WCH
−0.012 · SORE / WCH−0.005 · CaCO ₃ / WCH
−0.22 · CaF ₂ /WCH+0.35+learning term where
C 'was divided as follows to obtain F' (C).
When 0 <C <5 (10 ⁻² %) F ′ (C) = − 0.928C + 12.93
When 5 <C <25 (10 ⁻² %) F ′ (C) = 0.73 * 1n (C) −0.13C + 23.7 / C + 3.1
(However, In is a natural logarithm)
When 25 <C (10 ⁻² %) F ′ (C) = − 0.11C + 5.7
[0022]
Phosphorus model formula (3a)
PE = (0.10HMSi + 0.093HMP-1.13 · O ₂ / PiG
-0.39 · ΔO ₂ /WCH-0.14·SORE/WCH
-0.59 · ΔSORE / WCH-0.36HMR
-0.37CaF ₂ /WCH-0.019TCaO/PiG
+0.15 (100 ・ CS) −0.12CMR + 0.12TS
-87.1) / 1000 + learning term, where
TE: Molten steel end point molten steel temperature (° C)
TS: Intermediate temperature of molten steel (° C)
O ₂ : oxygen amount until intermediate measurement (Nm ³ )
ΔO ₂ : oxygen amount after intermediate measurement (Nm ³ )
SORE: Coolant amount (Kg)
ΔSORE: Amount of coolant after intermediate measurement (Kg)
CE: Molten steel carbon content (%)
CS: Carbon content of molten steel during intermediate measurement (%)
WCH: Main raw material (tons),
HMR: Hot metal ratio (%)
CaCO ₃ : Limestone (Kg),
HMSi: Hot metal silicon (10 ^-2 %)
PE: Phosphor end point phosphorus (10 ⁻³ %)
PiG: Pig iron
CMR: refrigeration rate (%),
HMP: Hot metal phosphorus (10 ^-3 %)
CaF ₂ : Fluorite (Kg)
Note that sintered ore was used as the coolant.
Substituting actual values for all variables other than TE, PE, and CE in equations (1a), (2a), and (3a), and solving for TE, PE, and CE is an estimated value of TE, PE, and CE. It was.
However, for CE, which is the end-point carbon content of the molten steel, in order to simplify the calculation, the equation (2a) is solved for F ′ (100 · CE) to be an estimated value.
The learning term Δa is updated by first calculating the calculated lime intensity values by substituting the actual values of HMSi, HMP, x1,..., Xm and the estimated values of TE, PE, and CE into the equation (10), respectively. , TCaO (cal) and actually put the converted lime intensity, hand corrected to a certain HMSi using the following equation TCaO (R).
Correction TCaO (cal) = TCaO (cal) · g12 / (g1 · (HMSi) ^g2 ) (11)
Correction TCaO (R) = TCaO (R) .g12 / (g1. (HMSi) ^g2 ) (12)
The learning term Δa in the equation (10) is learned and updated from the difference between the corrected TCaO (cal) and the corrected TCaO (R) obtained from the above equation.
[0023]
【The invention's effect】
Since the converter blowing method according to the present invention is configured as described above, the end point phosphorus control accuracy is improved, and the average value of the phosphorus component can be increased without increasing the off-specification ratio of the phosphorus component. It has become possible.
Further, by updating the hand correcting the learning term Δa calculated values and actual values of TCaO a constant HMSI, it is possible to perform learning HMSI is correspondingly vary.
Furthermore, for the P at the time of pouring into the RH degassing apparatus having the end point temperature and the phosphorus concentration in the post-process molten steel, the estimated value based on the dynamic model formula is used instead of the actual value, and the F ( By using the CE) value, it becomes possible to ignore the variation in the end point that cannot be avoided in the dynamic control, and the variation in the calculated value of the lime input can be reduced.
[Brief description of the drawings]
FIG. 1 is a characteristic diagram showing a comparison of phosphorus (P) detachment rates between a conventional method and a method of the present invention.
FIG. 2 is a characteristic diagram showing a comparison of phosphorus (P) concentration distribution between the conventional method and the method of the present invention.
[Explanation of symbols]
TE Converter end-point molten steel temperature PE Converter end-point molten steel phosphorus concentration or post-process molten steel phosphorus concentration CE Converter end-point molten steel carbon concentration x1, ..., xm Other factors Δa Learning term updated for each charge

Claims (1)

In converter blowing, lime input calculation formula or lime input calculation procedure TCaO = F (HMSi, HMP, TE, PE, CE, x1, ..., xm, Δa)
However,
TCaO: Conversion lime (CaO total amount) basic unit per converter charged pig iron,
HMSi: Hot metal silicon concentration,
HMP: hot metal phosphorus concentration,
TE: Converter end point molten steel temperature,
PE: Converter end point molten steel phosphorus concentration or post process molten steel phosphorus concentration,
CE: Converter end point molten steel carbon concentration,
x1, ..., xm: other factors,
Δa: A learning term that is updated for each charge using the blow smelting charge record,
, The end point molten steel temperature target value, the end point molten steel phosphorus concentration target value or the post-process molten steel phosphorus concentration target value, the end point molten steel carbon concentration target value are respectively substituted into the TE, PE, and CE, and the HMSi, HMP, x1,. , Xm is substituted with actual or planned values of each factor, lime input is determined and blown,
After each blowing, the value of Δa is updated using the actual values of TCaO, HMSi, HMP, x1,..., Xm and the estimated values of TE, PE, and CE.
The estimated values of TE, PE, and CE are
In the dynamic control model formula for TE, PE, and CE including the oxygen amount ΔO ₂ after the intermediate measurement, the coolant amount ΔSORE after the intermediate measurement, the carbon content CS of the molten steel at the intermediate measurement, and the intermediate temperature TS of the molten steel, , TS is substituted with the measured value in each sublance intermediate measurement, and is obtained by substituting the actual value from after each sublance intermediate measurement to the end point of the converter into ΔO ₂ , ΔSORE. .

Images