Converter Mouth Research Articles

Dynamic prediction of flue gas information driven by converter mouth flame spectral information

Dynamic prediction of flue gas information driven by converter mouth flame spectral information

Research on prediction model of converter temperature and carbon content based on spectral feature extraction

The flame of converter mouth can well reflect the change of temperature and composition of molten steel in the furnace. The flame characteristics of converter mouth collected by device can well predict the smelting process of converter. Based on the flame spectrum data set of converter mouth, this paper uses the BEADS algorithm and rough set attribute reduction algorithm optimized by genetic algorithm to extract the features of 2048-dimensional wavelength data. Through the model, eight indexes that contribute greatly to temperature and carbon content are selected, which are f-507, f-520, f-839, f-1073, f-1371, f-1528, f-1727 and f-1826. The MIC coefficients of the eight indicators with temperature and carbon content are calculated, and the MIC coefficients of the variables is small, and the selected indicators are representative. There was a significant correlation between temperature and C content. In BP neural network of temperature prediction model, it is found that the prediction accuracy of the training set is 0.99, the prediction accuracy of the test set is 0.99, the prediction accuracy of the verification set is 0.99, and the prediction accuracy of the whole set is 0.99. Through statistics, it is found that the hit rate of the temperature model in the range of ± 5 K is 88.7%, and the hit rate in the range of ± 10 K is 98.4%. and the RMSE parameter analysis shows that the average prediction error is 3.85 K. In BP neural network of carbon content prediction model, it is found that the prediction accuracy of the training set is 0.99, the prediction accuracy of the test set is 0.99, the prediction accuracy of the verification set is 0.98, and the prediction accuracy of the whole set is 0.99. Through statistics, it is found that the hit rate of the carbon contents model in the range of ± 0.05% is 94.0%, and the hit rate in the range of ± 0.10% is 98.3%, and the RMSE parameter analysis shows that the average prediction error is 0.021%. Finally, the universality of the model is verified by MIV algorithm.

Industrial cyber-physical system driven intelligent prediction model for converter end carbon content in steelmaking plants

Industry 4.0 provides an opportunity for realizing converter intelligent steelmaking. Based on a cyber-physical system (CPS) framework for steelmaking plants, we developed a non-contact intelligent prediction model for determining molten steel carbon content. We achieved this by leveraging big data in-depth learning and optical information from converter mouth flame characteristics to realize accurate control of end-point carbon content in the process of intelligent steelmaking. The spectral information of a converter port flame was gathered using a USB2000+ spectrometer, and its main characteristics were extracted via factor analysis. The sampling frequency of the converter flue gas analysis mass spectrometer was set to obtain the carbon content of continuous molten steel during the late stages of steelmaking consistent with the spectral information collection frequency. A large sample dataset was constructed to continuously predict the carbon content of the molten steel based on the flame spectrum information of the furnace orifice. Furthermore, the parameters of a support vector regression algorithm were optimized, and the quantitative relationships among the sample data were intelligently mined to obtain a dynamic prediction model of the carbon content of 150 furnace samples. To improve the universality of the model, the initial steelmaking conditions were applied as the optimization criteria of the carbon content prediction model generated by matching different furnace samples. A support vector machine algorithm was applied to construct a model optimization classifier to obtain a universal prediction model. The results indicate that in the late stages of smelting, the spectral information of the flame in the furnace exhibited a high correlation with the carbon content of the molten steel. After testing, the prediction accuracy of this model was more than 90%. With the help of big data deep learning and CPS framework, we can build robust and accurate steelmaking intelligent control systems.

Slopping index for LD converters based on sound and image data fusion by fuzzy Kalman filter

ABSTRACT The LD (Linz Donawitz) steelmaking process is the most used in the steel industry due to its high-volume capacity and low cost per ton of steel produced. However, the basic oxygen steelmaking process in LD converters is subjected to potential steel charge overflows, often called ‘slopping’. Besides yield losses, slopping events can damage the environment and expose employees to danger. More than ever, steelmaking plants need to avoid this type of event to keep producing as environmental impacts are no more tolerable by society. Steelmaking plants already use different methods to monitor and detect slopping events, but they are often limited and unreliable. Therefore, this paper proposes a multi-sensor data fusion process to generate a reliable slopping index to warn operators of potential slopping events and detect the triggered ones. The work is based on sound and image data (67 heats with 27 slopping events) collected on previous trials at a 350-ton LD converter. The Kalman filter was applied as a data fusion agent of two indexes, one resulted from computer vision analysis of the LD converter mouth (image data), the other resulted from digital signal analysis of sound captured on the converter’s hood (sound data). Fuzzy sets were applied for adaptative tuning of the Kalman filter to improve the data fusion process. Besides the increase of alarm accuracy and heat classification, the data fusion index worked better on different scenarios and produced a more reliable indicator for a slopping prevention system.

Research on terminal control model of intelligent mining of flame spectral information of converter mouth in late smelting stage

ABSTRACT A USB4000 spectrometer was installed in the converter mouth to collect spectral information of the flame in theprocess of steelmaking in real time, Gaussian fitting algorithm and wavelet analysis algorithm are used to extract the stableand unstable eigenvalues of flame spectral information. The spectral information characteristic value and the accumulatedoxygen consumption information index data in the smelting process were input as samples. Combined with the corresponding static model, the corresponding carbon content and temperature values were calculated as sample outputs,and to establish a one-to-one correspondence sample set. A continuous intelligent prediction model of carbon content andtemperature in the later stage of steelmaking was established by using the backpropagation neural network algorithm, witha model forecast accuracy in the experimental stage of more than 95% and above 89% in practical steelmaking, andprovides new research ideas to control the endpoint of steelmaking.

Using shell-type tuyeres at Pierce–Smith horizontal converters of the Nadezhda Metallurgical Plant

Since 2015 the processing capacity reconfiguration at the Polar Branch of MMC Norilsk Nickel (hereinafter PB) sets new goals for conventional pyrometallurgical processes of smelting and converting. The design flowsheet of Kolesnikov Nadezhda Metallurgical Abstract: Plant (hereinafter NMP) provided for «cross-converting» when copper matte was first processed in one converter to produce blister copper followed by nickel matte processing to yield copper-nickel converter matte bypassing the discharge of dry coagulated slag. This flowsheet allowed for converter heat balance optimization, decreasing the formation of refractory reverts and significant extension of the converter campaign. PB Nickel Plant shutdown resulted in copper processing elimination at NMP and switching the converters to the conventional nickel converting flowsheet. In turn, it gave rise to the need for solutions to extend converter campaign while maintaining the possibility to process large amounts of nickel slag from the second converting stage at the PB Copper Plant. For this purpose the series of lab experiments were carried out to develop the technology and design documentation for the system to supply oxygen-enriched air (up to 45 %) to horizontal converters using shell-type tuyeres. In addition, literature data were analyzed on this topic along with the experience of smelters in this area. Process design calculations were done. The efforts were taken in cooperation with the PB engineering personnel and Laboratory of Pyrometallurgy of LLC «Gipronickel Institute». The use of reduced diameter shell-type tuyeres to inject the oxygen-air mixture was found to decrease the converter blowing and off-gas volumes. The decline in off-gas quantity leads to reduced heat load on the converter mouth and flue duct system, as well as to lowered converter dust entrainment. The use of oxygen-enriched blowing implies the higher smelt heating rate. Excess heat compensation requires timely charging of cold reverts and flux. In emergencies (if cold reverts are not available) the oxygen content of the blowing has to be reduced until switching over to air blowing. The series of the above efforts will offer a possibility to use the shell-type tuyeres keeping the converter off-gas temperature at the current level. Thus continuous monitoring and efficient control will ensure the off-gas temperature and volume at the inlets of gas cooling and cleaning systems not exceeding the limiting values. The introduction of the reduced diameter shell-type tuyeres for air-oxygen mixture injection does not require any upgrade of the existing gas cooling and cleaning systems. Moreover, switching to these tuyeres will reduce gas load on the flue duct system and heat load on the water-cooled dust cap, lower dust entrainment and non-recoverable dust losses after the gas cleaning system.

Industrial IoT for Intelligent Steelmaking With Converter Mouth Flame Spectrum Information Processed by Deep Learning

In this article, based on the fact that the converter mouth flame is the comprehensive external appearance of the physical and chemical reactions in the converter during the steelmaking process, the continuous spectrum information of the converter mouth flame was obtained by USB4000 spectrometer. In the framework of the Internet of Things, a bidirectional recursive multiscale convolution depth neural network algorithm can take into account the characteristics of frequency domain structure and time domain dynamic sequence. It is applied to the deep-learning of the converter mouth flame spectrum information. The dynamic prediction model of carbon content and temperature value in molten steel at the later stage of steelmaking is constructed. The static control system and dynamic prediction model of automatic steelmaking are intelligently fused to realize one-key steelmaking control. The results show that the average hit rate of carbon content, temperature, and carbon temperature at the end of steelmaking is 94.78%, 98.41%, and 93.43%, which makes the end-point control of steelmaking more stable and the blowing rate less than 1%.

Analysis of Converter Combustion Flame Spectrum Big Data Sets Based on HHT

The characteristics of the converter combustion flame are one of the key factors in the process control and end-point control of steelmaking. In a big data era, it is significant to carry out high-speed and effective processing on frame spectrum data. By installing data acquisition devices at the converter mouth and separating the spectrum according to the wave length, high-dimensional converter flame spectrum big data sets are achieved. The data of each converter is preprocessed after information fusion. By applying the SM software, the correspondence with the carbon content is obtained. Selecting the relative data of the two peak ratios and the single-peak absolute data as a one-dimensional signal, due to the obvious nonlinear and nonstationary characteristics, using HHT to do empirical mode decomposition and Hilbert spectrum analysis, the variation characteristics after 70% of the converter steelmaking process are obtained. From data acquisition, data preprocessing to data analysis and results, it provides a new perspective and method for the study of similar problems.

The research on intelligent extraction of furnace mouth flame characteristics based on DNN

Deep neural networks are a focus of artificial intelligence and big data analysis in recent years. The monitor of the converter mouth is essential to the quality of the steel material production while the requirement of the steel material production is increasingly higher in China. The end-point control of converter blowing is the ultimate regulation of the carbon content and temperature. The severity of carbon-oxygen reaction and the temperature of molten steel can be reflected by the converter mouth flame. Operators judge the end of the steel by watching the converter mouth flame, the converter mouth spark and the time of oxygen supply. So, it is very important to offer a quantitative analysis to converter mouth flame characteristics. We quote the deep neural network into the intelligent extraction of the flame characteristics of the furnace mouth and construct a flame color recognition algorithm based on the deepness letter neural network. This paper belongs to the data science problem in the intelligent research of steel production. By observing the converter flame during the steel flame changes, this paper records the data of light intensity and end-point carbon content of each steel making furnace. When this paper then uses the temperature of flame emission spectrum to deduce and the absorption of the molten steel to judge the contents of the carbon during the converter steel blew process, it is more feasible and accurate than watching by operators. At the same time, by using deep learning algorithm, this paper makes the control process get automatic learning ability and achieve intelligent production so that we can provide a basis for solving the problem of predicting the end-point carbon content in molten steel during the blowing process.

DEVELOPMENT OF AUTOMATIC SYSTEM OF GAS-DYNAMIC CUT-OFF OF SLAG FOR CONVERTER WITH ROTATING VESSEL SHELL

The steelmaking unit for the processing of poor-quality charge and metal-containing wastes with a high content of harmful impurities must satisfy to some requirements which are considered at the article. Also the expediency of using the oxygen converter with rotating vessel shell as a prototype is justified. To increase the efficiency of such converter, it is proposed to supply it with a system of gas-dynamic slag cut-off at automatic mode during the steel tapping into the casting ladle. The proposed system is structurally a twin lever mechanism ensuring (during the metal overflow into the ladle through the edge of converter's mouth) holding at the constant distance from it the gas distribution chamber with slotted nozzles, from which flat gas jets run under excess pressure, directed at a given angle of attack to the surface of slag melt covering the teemed steel. Due to the dynamic impact of these jets on the layer of liquid slag, it is pushed away from the edge of the mouth of the inclined vessel shell towards its bottom, i.е. it is retained in the bath of melting unit. Synchronous movement of the structural elements of the lever mechanism with the end part of converter’s mouth is achieved by force interaction of the stops rigidly fixed on the supporting ring of the melting unit with swivel arms kinetically connected with the gas distribution chamber. At the same time, activation of the gas-dynamic system of slag cut-off and its shutdown occur automatically by a mechanical connection. After the finishing of steel tapping the slag is poured into the bowl, turning the vessel shell in the opposite direction from the steel casting ladle. The results of experimental verification of the operation of the proposed converter slag gas-dynamic cut-off system performed on its model confirmed correctness of the accepted technical solutions. The practical application of this development will significantly reduce the material losses associated with the ingress of a large amount of aggressive converter slag into the casting ladle (deoxidizing and ligature burning, intensive wear of the refractory lining and the possibility of metal rephosphorization).

Applying flame spectral analysis and multi-class classification algorithm on the BOS endpoint carbon content prediction

Aiming at an accurate endpoint control of the basic oxygen steelmaking (BOS) with two blowing stages, a new prediction method based on the flame spectral analysis and multi-class classification is proposed for those small-sized and middle-sized converters. Based on the fact that the experienced operators mainly control the blowing process by the flame variation at the converter mouth, a converter radiation information acquisition system is adopted to analyze the end period of the blowing from the perspective of different flame spectral characteristics. Besides, different blowing processes are divided into four sub-classes corresponding to different endpoint carbon content, and support vector machine is used to construct the multi-class classifier. The proposed model is validated by real industrial dataset, and results show that the mentioned method can play a good auxiliary role in the operators’ endpoint prediction.

基于多光谱测温法的转炉炉口火焰温度预测研究

基于多光谱测温法的转炉炉口火焰温度预测研究

Numerical Investigation of Combined Top and Lateral Blowing in a Peirce-Smith Converter

Abstract Typical current operation of lateral-blown Peirce-Smith converters (PSCs) has the common phenomenon of splashing and slopping due to air injection. The splashing and wave motion in these converters cause metal losses and potential production lost time due to intermittent cleaning of the converter mouth and thus reduced process throughput. Understanding of the effect of combined top and lateral blowing could possibly lead to alternative technology advancement for increased process efficiency. In this study, computational fluid dynamics (CFD) simulations of conventional common practice (lateral blowing) and combined (top and lateral blowing) in a PSC were carried out, and results of flow variables (bath velocity, turbulence kinetic energy, etc.) were compared. The two-dimensional (2-D) and three-dimensional (3-D) simulations of the three-phase system (air–matte–slag) were executed utilizing a commercial CFD numerical software code, ANSYS FLUENT 14.0. These simulations were performed employing the volume of fluid and realizable turbulence models to account for multiphase and turbulent nature of the flow, respectively. Upon completion of the simulations, the results of the models were analysed and compared by means of density contour plots, velocity vector plots, turbulent kinetic energy vector plots, average turbulent kinetic energy, turbulent intensity contour plots and average matte bulk velocity. It was found that both blowing configuration and slag layer thickness have significant effects on mixing propagation, wave formation and splashing in the PSC as the results showed wave formation and splashing significantly being reduced by employing combined top- and lateral-blowing configurations.

Applying Flame Spectrum on SVC-RVM Modeling for BOF Endpoint Prediction

A new non-contact method for predicting the basic oxygen furnace(BOF) end point carbon content is proposed in this study. A model applying the flame spectrum of the converter vessel mouth is constructed to carry out the prediction. This model consists two parts, viz. a classifier based on support vector classification to classify the whole period of one BOF heat into two main phases, and a relevance vector machine working at the posterior phase to predict the carbon content. Compared with current non-contact methods of end point carbon content prediction, the proposed method can make better use of the information of the flame of the converter mouth. Simulations on industrial data show that this method yields good results on the classification as well as end point carbon content prediction.

Detection of Slopping in Basic Oxygen Steelmaking using a Camera Viewing the Converter Mouth

In basic oxygen steelmaking, foam overflow or slopping causes operational and environmental problems. In most steel plants slopping is neither detected nor recorded. In this paper a slop detection algorithm is presented, which is based on images taken by a camera viewing the converter mouth. The proposed algorithm has a sensitivity and specificity of 0.74 and 0.94 respectively, is relatively simple and can easily be used in on-line applications. When using this algorithm, slopping is quickly detected and can be halted for the majority of slopping batches.

Detection of slopping in basic oxygen steelmaking using a CMOS camera aimed at the converter mouth

The paper introduces a slop detection algorithm, based on images taken by a CMOS camera aimed at the converter mouth. The proposed algorithm, with an accuracy and selectivity of 74% and 94% respectively, is relatively simple and can therefore easily be used in on-line applications.

銅製錬転炉における造銅期の反応解析

Oxygen probes were applied to a PS converter in copper-making operation, and the oxygen pressures of white metal and blister copper in the converter were measured through the converter's mouth and through a tuyere, respectively. Based on the obtained results, the following reaction mechanism has been deduced.(1) The copper-making stage of a PS converter may be classified into three stages, i.e. the early substage where the tuyeres are completely in white metal, the intermediate substage where both phases of white metal and blister copper coexist in the converter and the tuyeres are in the blister copper, and the final substage where the white metal has disappeared from the converter.(2) In the early stage, direct oxidation of white metal by oxygen gas contained in the reaction air progresses in the vicinity of the tuyeres.Cu2S(l) + O2(g) = 2Cu(l) + SO2(g) …………… (1)(3) In the intermediate substage, the following three reactions proceed near the tuyeres, at the interface of blister copper and white metal phases, and in the bulk of the blister copper, respectively.O2(g) = 2O(l) …………… (2)Cu2S(l) = 2Cu(l) + S(l) …………… (3)S(l) + 2O(l) = SO2(g) …………… (4)Here the most important factor is that reaction(4) progresses under the condition of PSO2=1atm. It is not dependent on how high the SO2 pressure of the waste gas from the converter is. Because the condense phases, i.e. the white metal and blister copper phases, are not in equilibrium with the gas phase.(4) In the final substage, reactions(2) and (4) only proceed because the white metal is no longer in the converter. Some of the blister copper can be excessively oxidized to the extent that Cu2O(s) is formed according to equation(5). However the final degree of oxidation of the whole blister copper does not reach this level.2Cu(l) + 1 / 2 O2(g) = Cu2O(s) …………… (5)

Splashing and Spitting Behaviour in the Combined Blown Steelmaking Converter.

The amount of iron droplets ejected in the BOF affects on metallic yield, refactory wear and the progress of decarburisation. The purpose of this study was to investigate the effect of lance height, lance nozzle angle, lance position, top gas flow rate, bottom blowing and foamy slag on splashing and spitting. A new cold model method for investigating the effects of above-mentioned parameters on the location and quantity of liquid splashed on the walls of the model was utilised. According to the model tests, reduction of the nozzle angle increased the total amount of splashing and spitting considerably. Consequently, reduced productivity due to an increase in metal losses, skulling of the cone and converter mouth and further increased time for skull removal is expected. Introduction of bottom blowing increased splashing significantly on lower parts of the vessel. Lance position has an effect on total amount of splashing when bottom blowing is used. The presence of even minor foam layer on water surface reduced the amount of total splashing significantly.

銅製錬転炉反応解析への平衡計算の応用

Changes of the matte grade and oxygen pressure during the slag making stage in a PS copper converter were calculated by equilibrium calculations. In the calculation a slag making stage was simulated by many calculation steps in series, each of the steps being assumed in an equilibrium state. Results of the calculation were compared with experimental values measured in pilot and commercial converters.1. Calculated and measured matte grade changes were in good agreement with each other during the slag making stage. In particular, both of the changes showed that in the beginning of the stage there was a period of time for which the matte grade did not increase appreciably in spite of blowing. According to the present calculation, this was attributed to an increase of oxygen disolved in the matte.2. Calculated oxygen pressures at the beginning and the end of the stage agreed relatively well with those measured for the matte before blowing and for the converter slag after the stage, respectively. However oxygen pressures measured through the converter mouth during the stage were rather higher than those predicted by the calculation. Some of the measurements gave higher oxygen pressures than those measured even at the end of the stage.3. To explain the above mentioned phenomena by equilibrium calculation, a two-zone converter model has been developed. In the model the matte holding zone in a converter consists of two zones, i. e. a magnetite formation zone and a magnetite reduction zone, which are assumed in a different equilibrium state. By setting a proper ratio of the two zones, differences in oxygen pressure between the white metal and the converter slag at the end of the slag making stage were explained to some extent.

Mechanisms of dust generation in a stainless steelmaking converter

The four possible mechanisms of dust formation in converter steelmaking are evaporation of metal, ejection of slag, ejection of metal and entrainment of solids. In order to describe the relation between process operation and dust formation and take appropriate measures to minimize the quantity of dust, more detailed information is needed on the amounts contributed by the different dust generation mechanisms. For their determination, dust samples were taken from the off‐gas during refining of five austenitic heats in a modified 80‐t AOD converter. On the basis of dust, metal and slag analyses and process and product gas flow rates, the contributions of the different dust generation mechanisms were quantitatively determined for blowing step I (pure oxygen blowing) and reduction step R (pure argon blowing). The predominant mechanism is ejection of metal and slag droplets by bursting of bubbles at the bath surface. Dust formation via the vapour phase in AOD steelmaking is of less importance than previously expected, the relative amount is only about 3 %. During the addition of fluxes through the converter mouth, the entrainment of solids plays a role, too.