Method and apparatus to determine and control the carbon content of steel in a BOF vessel

[0002] This invention is directed to a method and apparatus for controlling or determining the carbon content of a heat in a BOF vessel, and more particularly, to a method for determining the in-blow carbon content and the First Turn Down Carbon (FTDC), in low carbon steel BOF heats containing 0.06% or less carbon.

[0003] Users of flat rolled steel product demand low carbon grade steel because of its good formability properties. For example, in the automotive industry, such low carbon steel permits auto manufacturers to stamp and form complex automobile shapes without encountering steel spring-back after the foraging operations. This makes it necessary for steelmakers to accurately manage and control the carbon content of their BOF heats to produce a product having the proper metallurgical requirements.

[0004] In the BOF steehnaking steelmaking process carbon saturated liquid iron is poured into the vessel along with various amounts of steel scrap. tligh and high velocity oxygen is blown into the BOF vessel at the surface of the molten steel bath where it reacts with the carbon to form CO and CO₂. This reaction removes excess carbon in the steel bath and produces a finished product having the desired carbon content.

[0005] There are many BOF process control methods available to preset day steelmakers. These controls range from sophisticated predictive models that are managed through the use of computers in combination with sensor instruments such as gas analyzers, thermocouples, load cells, etc.

[0006] In the past, various attempts have been made to control the carbon content in a vessel using flame drop measurements. One such past attempt is shown in U.S. Pat. No. 3,652,262 granted to Denis. This patent discloses using a sensor to detect infrared radiation emitted from a BOF vessel. The signal from the infrared sensor is processed to generate a curve representing a function of radiation intensity against time. In his patent, Denis compared his radiation curve with a decarburization curve generated by using readings taken from a first gas pickup used to measure the concentration of CO₂and CO in the off-gas of a BOF vessel, and from a second gas pickup used to measure the total gas output. He then compared the two curves and concluded that his time/radiation curve was useful in providing an instantaneous carbon reading during BOF steelmaking operations. However, if the two different graphs are compared, it can be seen that wide variations in predicted carbon levels occur between the off-gas curve and the radiation curve. Therefore, although Denis has provided some improvement in providing an instantaneous reading of carbon content during a heat, his patent shows a wide margin of error in his predicted carbon levels based upon his flame drop readings.

[0007] Additionally, in a study found in chapter fifteen entitled “BOF Control”, of an Iron & Steel Society publication “BOF STEELMAKING” dated 1977, J. H. Cox, et al. teach that flame intensity is a function of the carbon in the bath. However, the authors also teach that carbon predictions, based upon flame intensity measurements, are not satisfactory for the more stringent present day needs.

[0008] Such beliefs have become widespread throughout the steelmaking industry. They have led steelmakers to use control strategies based on statistical, predictive-adaptive control models, or highly sophisticated control systems based on a continuous or periodic measurement of variables such as carbon, temperature, etc. (J. H. Cox, et al. “BOF STEELMAKING”). One such measurement process is based upon mass/temperature calculations to determine the carbon content of a BOF heat. It is well known that such mass/temperature calculations contain a margin of error, and they often lead to either overblowing or underblowing the BOF heats.

[0009] In instances where a heat is overblown various undesirable chemical reactions take place within the vessel. For example, in an overblown heat, the oxygen consumes an excessive amount of carbon and a steel product having an undesirable low carbon level is produced. The excess oxygen also reacts with the molten iron to form iron oxides. This reduces the iron yield of a heat. Overblowing a heat will also overheat the steelmaking vessel, cause premature wear on its protective refractory lining, and reduce the service life of the vessel.

[0010] In those instances where a heat is underblown, the heat may have to be reblown to further reduce the carbon level. This increases production time and cost, and causes excess refractory wear. The excess refractory wear is due to the iron oxides that are formed in the slag during the reblow. Iron oxides in the slag make the slag more corrosive to the refractory lining.

[0011] Another problem encountered with BOF control systems is dealing with the hostile environment adjacent the hot BOF vessel. The radiant heat emitted from a BOF vessel during the steelmaking process overheats sensitive electronic equipment located near the vessel and causes system failures. Dust and fume released from the vessel also settles on equipment located throughout the steelmaking shop, including the various sensor devices used to control the steelmaking process. The dust and fume causes control equipment sensors to become fouled and dirty, and results in poor readings and inaccurate metallurgical analysis.

[0012] It is therefore an object of this invention to provide a method for determining the carbon content of a BOF heat.

[0013] It is still a further object of this invention to accurately control or determine the carbon content of a BOF heat within a margin of error of about 0.004% C.

[0014] It is still a further object of this invention to accurately control or determine the carbon content of a BOF heat by measuring a difference in visible light intensity emitted from the BOF vessel in relation to the amount of oxygen blown into the vessel.

[0015] It is still a further object of this invention to provide apparatus for controlling and/or determining carbon content of a BOF heat, the apparatus being resistant to high temperatures encountered in the steelmaking process.

[0016] And finally, it is still a further object of this invention to provide apparatus, for controlling and/or determining the carbon content of a BOF heat, that includes a self cleaning means to prevent an accumulation of fume or dust on its sensor devices.

[0017] We have discovered that the foregoing objects can be attained by measuring a drop in visible light intensity emitted from a BOF vessel with a light sensor housed within a temperature regulated case having a sighting window including air wipe means to shield the light sensor from steelmaking dust and fume. The drop in visible light intensity is measured from a point of maximum light intensity emitted from the BOF vessel in relation to the amount of oxygen blown into the BOF vessel from such point of maximum light intensity, to the end of the oxygen blow. The light intensity and oxygen readings are used to compute continuous, real time in-blow % carbon levels until an aim carbon content is reached for the heat.

[0018] FIG. 1 is a schematic view showing the preferred system used to carry out the steps of the method to determine carbon content of a BOF heat.

[0019] FIG. 2 is an elevation view of the preferred embodiment of the light sensor shown in the preferred system of FIG. 1.

[0020] FIG. 3 is an enlarged view showing a portion of the light sensor of FIG. 2.

[0021] FIG. 4 is a graph showing a BOF heat having a 0.053% FTDC.

[0022] FIG. 5 is a graph showing a BOF heat having a 0.045% FTDC.

[0023] FIG. 6 is a graph showing a BOF heat having a 0.028% FTDC.

[0024] FIG. 7 is a flow chart of one program that may be used to determine in-blow carbon content and FTDC based upon flame drop measurements and oxygen blown.

[0025] Referring to the drawings, FIG. 1 shows the preferred embodiment of the sensor system 10 for carrying out the steps of the method to determine in-blow % carbon content and FTDC levels of BOF heats. Throughout this specification, “in-blow carbon content” shall mean realtime carbon levels determined at any point during an oxygen blow, and “FTDC” shall mean First Turn Down Carbon determined at the end of the first oxygen blow into a BOF.

[0026] The system comprises a sensor device 1, an oxygen source 2, and a Programmable Logic Controller 3, (PLC). The sensor device 1 comprises a case or housing 5 to hold various electronic components of the sensor device including a light meter 4 having means to amplify light intensity signals to a level suitable for use in the PLC, a power supply 6, and a cold air supply 7 to prevent the hot steelmaking environment from overheating the sensor device. The sensor device further includes a sighting window 8 that extends through a wall of the case 5 to expose the light meter 4 to visible light emitted from the mouth of the BOF vessel. An air wipe 9 emanates from the open sighting window 8 to prevent dust and fume from entering the interior space of the case. The air wipe 9 encircles the light sensitive portion of the light meter 4 and shields it from the steelmaking dust and fume.

[0027] The sensor device is designed to generate light intensity signals when it is exposed to light emitted from the mouth of the BOF vessel, and the signals are sent to the PLC for processing. Likewise, the oxygen source 2 includes means to generate signals that correspond to the amount of oxygen blown during the heat, and these signals are also sent to the PLC for processing. The light intensity signals and the oxygen blown signals provide data to continuously calculate in-blow % carbon content of the heat during the oxygen blow. The PLC receives and processes the light and oxygen signals to provide a continuous in-blow % carbon content reading in real time based upon the drop in visible light intensity in relation to the amount of oxygen blown, hereinafter referred to as “the flame drop method.” In-blow % carbon content predictions based upon the flame drop method are found to contain an error of prediction of about 0.004% carbon.

[0028] Referring now to FIGS. 2 and 3 of the drawings, the preferred embodiment of the sensor device 1 is shown comprising a case 5 having a hinged lid 11 to provide an opening 12 for access to the interior space 13 of case 5. A gasket or seal 14 extends along the periphery of the opening 12, and the gasket cooperates with hinged lid 11 to prevent dust or fume from entering the interior space 13 when the lid 11 is in its closed position. A cold air supply 7, shown as a vortex tube in the preferred embodiment, extends through a wall of the case to inject cold air 15A into the interior space 13. The cold air cools the interior space 13. A thermometer 16 is also attached to a wall of case 5. The thermometer includes a temperature probe 17 that extends into interior space 13 to provide an inside space temperature reading on the thermometer. The inside temperature reading is used to control the interior temperature of the case by regulating the cold air supply 7.

[0029] In the preferred embodiment, the cold air supply is shown comprising a vortex tube cooler. However any known air conditioning device, having means to regulate temperature, may be used to supply cold air to the interior space 13. The vortex tube 7 comprises an air inlet 18 through which pressurized air 15 is introduced into a vortex chamber 19 comprising a cold air vent 20 that injects cold air 15A into the interior space 13, and an exhaust tube 21 for venting hot air exhaust 15B to the atmosphere. The vortex tube further includes a valve mechanism having a valve screw 42 that extends through a wall of the exhaust tube 21 to regulate the temperature of the cold air 15A being injected into the interior space 13. The temperature of the cold air is either increased or decreased by turning the valve screw 42 in either a clockwise or counter clockwise direction. This decreases or increases the flow of the hot air exhaust and in turn either raises or lowers the temperature of the cold air 15A.

[0030] The light meter 4 is housed within the interior space 13 of case 5. In the preferred embodiment, a model P401025 Davis Instruments light meter is used. However, any suitable light meter capable of measuring the visible light intensity emitted from the mouth of a BOF vessel may be used with the sensor apparatus 1. The Davis light meter has a selenium photo-voltaic cell 22 separate from its amplifying circuit 23. This light meter arrangement makes it possible to independently mount the light sensitive cell 22 and the amplifying circuit 23 within case 5. The light sensitive cell 22 is attached to a slidable adjustment 24, and the amplifying circuit 23 to a fixed non-conductive mounting board 25 that is attached to a wall of the case. An electrical wire 26 connects the photo-voltaic cell 22 to the amplifying circuit 23, and the amplifying circuit increases the level of the light intensity signals received from the flow of electrons that respond to visible light falling on photo-voltaic cell 22. A power source (not shown) supplies power to the light meter 4 through line 27, and the amplified electrical signals from the light meter are transmitted through line 28 to the PLC shown in FIG. 1. The power supply may be either internal as shown by reference number 6 in FIG. 1, or external as shown by the outside power line 27 in FIG. 2.

[0031] Referring to the enlarged FIG. 3, the slidable adjustment 24 includes a mounting plate 29, a clamp arrangement 33 attached to a first end of the mounting plate, and a screw mechanism 32 attached to a second end of the mounting plate opposite the clamp arrangement 33. The mounting plate 29 also includes elongated slots 30. A Pin or fastener 31, having one end fixed to the mounting board 25, extends through each slot 30 to slidably attach the mounting plate 29 to the non-conductive mounting board 25.

[0032] The screw mechanism includes a first threaded boss 34 attached to the non-conductive mounting board 25, and a second threaded boss 35 attached to the mounting plate 29. A threaded shaft 36, having an adjustment knob 37 located at one end thereof extends through the threads of the first and second boss. The threaded shaft 36 provides a force to move the mounting plate 29 within the limits of engagement between the elongated slots 30 and pins 31 when shaft 36 is rotated in either a clockwise or counter clockwise direction.

[0033] The clamp arrangement 33 is located opposite the screw mechanism 32 and includes a back plate 38 attached to mounting plate 29 and positioned in a direction parallel to the sighting window 8. Clamp 33 also includes a transparent front plate 39 positioned adjacent the sighting window 8 and attached to the back plate by fasteners 40. The photo-voltaic cell 22 is positioned between plates 38 and 39, and the fasteners are tightened to clamp cell 22 between the two plates and hold it in place, adjacent and parallel to the sighting window 8.

[0034] The cold air 15A, injected into the interior space 13 from the cold air vent 20, provides a cold air flow throughout the interior of case 5. The cold air How travels in a direction toward the open sighting window 8 and cools the electronic components housed within the case. The cold air encircles the photo-voltaic cell 22 positioned adjacent the open sighting window 8, and it exits through the window to provide an air wipe 9. The air wipe 9 prevents dirt, dust and fume from entering the interior space 13 through the sighting window 8 and thereby maintains the transparent plate 39, and the light sensitive surface of the photo-voltaic cell 22, in a clean condition. It should be understood, however, that although the air wipe is described as encircling the photo cell 22, the apparatus is not limited to a photo cell having a disk shape configuration. The shape of the photo cell is not important to the scope of this invention, and it may comprise any suitable shape such as a square, rectangle, etc., and still be considered encircled by the air wipe 9.

[0035] If it is observed that the photo cell is becoming contaminated by steelmaking dust and fume, the screw mechanism 32 may be rotated to move the photo cell in a direction toward the sighting window. This will reduce the air gap 41 and increase the velocity of the air wipe 9. The screw mechanism is rotated until the air wipe velocity is increased to a level that prevents dust and fume from contaminating the light sensitive surface of the photo-voltaic cell 22. Similarly, if it is discovered that less air wipe velocity is needed to maintain the light sensor in a clean condition, the screw mechanism may be rotated in the opposite direction to increase the air gap 41 and reduce the velocity of the air wipe.

[0036] Visible observations of the flame emitted from the mouth of steelmaking vessels have long been used to estimate the FTDC of steelmaking heats. Such control methods were commonly used with the Bessemer converters, and in some instances, such flame observations have been used in BOF steelmaking operations. However, it appears that the flame drop control methods of the past have failed to recognize the correlation between a drop in light intensity measured from a point of maximum light intensity emitted from the BOF vessel, and the amount of oxygen blown into the BOF vessel from the point of maximum light intensity to the end of the oxygen blow.

[0037] It is well known that carbon dissolved in the liquid iron reacts with the oxygen that is blown into the BOF vessel. The reaction forms CO and CO₂in the off-gas of the steelmaking vessel. In the early part of an oxygen blow the incoming oxygen generates far more carbon monoxide than carbon dioxide, and therefore, CO makes tip up a large part of the vessel off-gas. The CO in the off-gas post-combusts at the mouth of the vessel and creates a flame. Near the end of the blow, when the molten bath has been converted to steel and the bath contains much lower carbon levels, the carbon-oxygen reaction generates much less CO in the off-gas. At this point, because there is little CO in the off-gas, the post-combustion flame at the mouth of the vessel decreases to a somewhat constant low level light intensity.

[0038] In an effort to correlate the drop in flame intensity with the amount of oxygen blown and with the FTDC of BOF heats, light intensities (LI) for more than 300 BOF heats were measured as a function of SCF of oxygen blown. It was found that the FTDC for these observed heats correlated with the DO₂, DLI/ILI and DO₂kink variables listed below.

[0039] The FTDC variables were then correlated through statistical regression analyses to develop an FTDC equation that could be used to determine the in-blow % carbon content for any low carbon heat. This was done by first defining the DO₂, DLI/ILI and DO₂kink variables as follows to simplify the calculations:

[0040] X1=DO₂/100,000,

[0041] X2=DLI/ILI, and

[0042] X3=DO₂kink/10,000.

[0043] The laboratory analysis for each FTDC of the analyzed heats was identified with its measured variables X1, X2 and X3 as shown by the example in the following Table A.

[0044] It should be pointed out that although the above TABLE A lists actual values determined during the numerous heat observations, the values shown in TABLE A are not necessarily listed in their actual heat number order. Also, as indicated above, the X2 variable is a ratio between the two LI quantities, (DLI/ILI) and is a measurement of the degree of flame drop in a BOF. The LI values used in this work are expressed in arbitrary units. Because X2 is only a ratio between two LI quantities, the units in which LI is measured have no effect in characterizing the intensity of the flame. As long as the units of LI measurement are consistent, any unit of measurement, either arbitrary or absolute, may be used to measure the LI of the flame emitted from the BOF vessel.

[0045] Knowing the actual measured values for X1, X2 and X3 variables, and also knowing the actual FTDC for each heat, determined by laboratory analysis, the following equation was developed and the values of “a” through “k” were calculated for each heat. $\begin{array}{cc}\mathrm{FTDC}=a+\mathrm{bX1}+{\mathrm{cX1}}^2+{\mathrm{dX1}}^3+\mathrm{eX2}+{\mathrm{fX2}}^2+{\mathrm{gX2}}^3+\mathrm{hX1X2}+{\mathrm{iX1X2}}^2+{\mathrm{jX1}}^2\mathrm{X2}+k\sqrt{\mathrm{X3}}& \mathrm{Eqaution}1\end{array}$

[0046] As heretofore stated, the FTDC variables were correlated through statistical regression analyses to develop an FTDC equation to determine the in-blow % carbon content for any low carbon heat. This was done by first defining the X1, X2 and X3 variables as disclosed above. The values for FTDC, X1, X1², X1³, X2, X2², X2³, X1X2, X1X2², X1²X2 and the squareroot of X3 for the heats 1 through 300 were then listed in separate columns on a Microsoft Excel Worksheet. The linear regression program provided with MicroSoft Excel was run with FTDC being the dependent variable and the values of X1 through the squareroot of X3 being the independent variables. The program output specifies the values of the coefficients “a” through “k”. For example, the value of “a” is the same as the “Intercept”, (the value of “a” is the same as FTDC when X1, X2 and X3 are equal to zero). The value for “b” is the same as the coefficient for the variable X1, “c” is the same as the coefficient for X1², and so on, through to the last value showing “k” the same as the coefficient for the squareroot of X3. An example of a completed FTDC equation showing actual calculated “a” through “k” values for the observed heats is shown below. $\begin{array}{cc}\mathrm{FTDC}=0.09993125+0.03013298\left(\mathrm{X1}\right)-0.0587246\left({\mathrm{X1}}^2\right)-0.0266337\left({\mathrm{X1}}^3\right)-0.0879685\left(\mathrm{X2}\right)-0.0666153\left({\mathrm{X2}}^2\right)+0.12504982\left({\mathrm{X2}}^3\right)+0.10246922\left(\mathrm{X1X2}\right)-0.2269549\left({\mathrm{X1X2}}^2\right)+0.14953375\left({\mathrm{X1}}^2\mathrm{X2}\right)-0.0003159\left(\mathrm{Squareroot}\mathrm{of}\mathrm{X3}\right)& \mathrm{Equation}2\end{array}$

[0047] The calculated values for the coefficients “a” through “k” vary from BOF to BOF. In general, their values are dependent upon shop conditions and the type of BOF vessel observed. Some vessels emit more visible light than others due to such conditions as skull build up at the mouth of the vessel and the physical features of the vessel design. The “a” through “k” values may also depend upon the location of the light meter with respect to the BOF. Therefore, it can be seen that it is necessary to determine the values of “a” through “k” for each BOF vessel before % carbon is determined using the flame-drop method.

[0048] The calculated coefficient values “a” through “k” are entered into a program that is written based upon the program diagram shown in FIG. 7. Code for one such possible program is attached herewith as Appendix A. It should be understood, however, that the attached code in Appendix A is only one example of many such codes that can be written following the program diagram shown in the drawings. Based upon the calculated values for “a” through “k”, and also based upon real time measurements of the variables X1, X2 and X3, the program shown in FIG. 7 calculates real time % carbon using the FTDC equation.

[0049] It is important to realize that the form of the FTDC equation is not important. The equation is only a means to carry out the steps of the invention for determining % carbon based upon the flame-drop method. A polynomial equation is one way to describe the variation of FTDC with variables X1, X2 and X3. However, other forms of equations involving logarithmic, exponential, higher order polynomial terms or a combination of any of these terms may be used for carbon prediction. Neural Network programs may also be used to predict carbon for this purpose.

[0050] After the FTDC equation and program shown in FIG. 7 were developed, additional BOF heats were monitored using the flame drop method to determine their respective FTDC levels. These heats were monitored using the sensor system 10 shown in FIGS. 1-3. The PLC was programmed according to the flow diagram shown in FIG. 7, and the program analyzed the LI signals received from the light sensor 1 and the O₂signals received from the oxygen source 2 to determine the X1, X2, and X3 values for the additional heats. Using the FTDC equation, the program then predicted the in-blow % carbon content based upon the previously calculated “a” through “k” values and the realtime X1, X2 and X3 values determined from the continuous LI, DO₂and DO₂kink measurements. Each predicted FTDC for the additional heats was later contrasted with an actual FTDC compositions determined in a laboratory through chemical analyses. It was found that the predicted values were very accurate. The mean absolute error for the predicted in-blow % carbon content values, based on the flame drop method, was found to be about 0.004% carbon, and the standard deviation of error was determined to be about 0.006% carbon. It was determined that the predictions were best suited for BOF heats having a carbon content range of about 0.05% carbon or lower because above the 0.05% carbon range accuracy of the FTDC predictions decrease.

[0051] Referring to the BOF heat examples shown in the graphs of FIGS. 4 and 5, it can be seen that the light intensity emitted from a BOF increases steadily for approximately the first 80% of the in-blow before it reaches a point of maximum light intensity. The increase in light intensity tip up to maximum LI is shown as (ILI). Realtime LI and O₂signals are sent continuously to the PLC from the start of the blow until the end of the blow to plot a complete LI curve and to record total oxygen consumption. The time averaged LI signals are sent once every minute for the first 80% of the blow and once every 4 seconds during the last 20% of the blow. When the LI curve drops from its point of maximum light intensity, the PLC begins to continuously process the realtime LI and O₂signals to determine in-blow carbon content. Again, the in-blow % carbon content is determined by running a computer program that calculates continuous realtime % carbon based upon an FTDC equation similar to the one disclosed. As the predicted % carbon falls to within a range of approximately 0.05% C or below, a continuous display showing the predicted % carbon is given to the operators until the end of the blow is reached. In the two curves plotted in FIGS. 4 and 5, the realtime O₂and LI signals are relayed to the PLC for processing until the computer readout indicates that the % carbon in the steel bath has been reduced to the desired aim carbon level for the end product. The oxygen blow is then stopped, and the steel is poured into a ladle for further processing such as continuous casting into finished products.

[0052] The third BOF heat example shown in FIB. 6, depicts a curve that continues past a “kink” in the LI measurements. The kink in the LI curve is an indication of a smaller post combustion flame, and a lower visible light level at the mouth of the BOF vessel. Flame reduction and lower light intensity are a result of the oxygen blow reducing the excess carbon in the steel bath to a level where suddenly a somewhat constant flame, having a low light intensity level, is observed at the mouth of the BOF vessel because of the small amount of CO in the off-gas. Consequently, the steady low level of the light intensity gives LI measurements that generate the kink in the LI curve shown in FIG. 6. Under such steady low light level conditions realtime DO₂kink values are determined continuously by the PLC along with the realtime DO₂and LI values, and the computer continuously calculates in-blow % carbon content based upon DLI/ILI, DO₂and DO₂kink as shown in the example below. When the program readout indicates that the % carbon in the steel bath is equal to the desired aim carbon content, the oxygen blow is discontinued and the steel is poured for further processing into finished products.

[0053] As seen in the three example curves plotted using the flame drop method, after the maximum LI point has been reached, the values for X1, X2 and X3 can be calculated continuously as the heat is being blown, and the in-blow % carbon content of the heat can be determined in realtime until the aim carbon is predicted by the PLC output. For example, in FIG. 4 a point of maximum LI is reached after about 500,000 SCF of O₂is blown into the vessel. At the point of maximum LI, the light intensity drops from a high of approximately 730, and the oxygen blow continues until a metal composition of 0.053% carbon is reached. As the LI falls from its maximum level of 730, the PLC receives continuous realtime LI and O₂signals from the sensor device 1 and the oxygen supply 2, and displays continuous in-blow % carbon content for the heat starting at about 0.05% carbon and below until the end of the oxygen blow is shut down. In this example the 0.053% carbon level is reached at a light intensity of about 480 corresponding to a DO₂of approximately 45,000 SCF.

[0054] In FIG. 5, an LI curve for a low carbon BOF heat is shown having an FTDC of 0.045% carbon. A point of maximum LI is reached at about 780 at just under 500,000 SCF of oxygen, and the 0.045% carbon level is predicted at a light intensity of about 400 corresponding to a DO₂of approximately 40,000 SCF. Once again, during the oxygen blow the LI and O₂signals are continuously sent to the PLC, and the program calculates the X1, X2 and X3 values and continuously predicts the in-blow % carbon content.

[0055] Referring to FIG. 6, a BOF heat is shown having an FTDC of 0.028% carbon. In this example the oxygen is blown past the point of DO₂kink shown in the LI curve. As heretofore stated, the “kink” occurs at the stage of the steelmaking process where the excess carbon has been reduced by the oxygen to such a low carbon level that the small amount of CO in the off-gas suddenly produces a somewhat constant low light level flame as described above. Consequently, the dark post combustion flame at the mouth of the BOF vessel produces X2 variables that produce the “kink” in the LI curve.

[0056] Control of the oxygen blow is more critical through the DO₂kink portion of the heat. The PLC readout must be watched to prevent overblowing the heat and consuming excessive amounts of carbon. The heat shown in FIG. 6 reaches its point of maximum LI at about 710 at 360,000 SCF of oxygen. The LI intensity then drops until it reaches the “kink” at an LI measurement of about 280. The oxygen blow continues past the kink until the PLC indicates that a predicted steel composition of 0.028% carbon has been reached. In this example the FTDC of 0.028% is reached when DLI/ILI=1.05, DO₂=155,000 SCF, and DO₂kink=25,000 SCF.

[0057] While this invention has been described as having a preferred design and method for predicting in-blow % carbon content and -FTDC in a BOF heat, it is understood that it is capable of further modifications, uses, and/or adaptations following in general the principle of the invention and including such departures from the present disclosure as come within known or customary practice in the art to which the invention pertains, and as may be applied to the essential features set forth herein, that fall within the scope of the appended claims.