高炉喷吹氢气的预测数学模型及工业验证

Prediction-oriented mathematical model and industrial validation of hydrogen-injected blast furnace

高炉炼铁系统能耗和排放占据钢铁全流程总能耗和总排放的70%以上,具有很大的节能减排潜力。实践表明,高炉风口喷吹氢气不仅可以实现低碳炼铁并且能够取得较好的经济效益。基于物质平衡、能量平衡,以及机器学习建模迭代确定的间接还原度方程建立了“高炉喷吹氢气的预测数学模型”。应用该数学模型研究了氢气喷吹后高炉冶炼指标的变化规律,包括对高炉燃料比、理论燃烧温度、直接还原度、鼓风量、炉腹煤气体积、碳素消耗等的影响。基于晋南炼铁厂高炉风口喷吹氢气实际工业生产数据,验证了模型的合理性和可靠性,燃料比和煤气利用率相对误差可以控制在3%的范围内。以鼓风含氧率和氢气喷吹量为主要考察因素,分别研究并预测单因素变化和两因素协同变化时的高炉冶炼行为规律。研究发现,单因素改变时,仅提高鼓风含氧率,高炉燃料比、理论燃烧温度升高,直接还原度、鼓风量和炉腹煤气体积降低;仅提高氢气喷吹量,高炉燃料比、理论燃烧温度、直接还原度均降低,鼓风量降低速度减缓,炉腹煤气体积先降低后略有升高的趋势。两因素协同调整时,每提高10 m^(3)/t氢气喷吹量,同时提高0.43%鼓风含氧率,能使理论燃烧温度稳定为(2142±2)℃,同时降低碳素消耗。通过将传统高炉数学模型与机器学习优化算法相结合,可以节约富氢高炉工业试验成本,优化工业试验方案,为工业试验的稳定运行和合理预测提供理论指导。

The energy consumption and emissions of the blast furnace ironmaking account for over 70%of the total energy consumption and emissions of the entire steelmaking process.Consequently,it has a great potential to reduce energy consumption and emissions of the blast furnace process.It is shown that hydrogen injection in blast furnace tuyere have promising benefits in low-carbon ironmaking.A prediction mathematical model of blast furnace with hydrogen-injected at tuyeres is established based on material balance,heat balance and the machine-learning dynamic model of indirect reduction degree.The mathematical model was applied to study the impact of hydrogen injection on the blast furnace indices such as fuel ratio,theoretical combustion temperature,direct reduction degree,blast volume,bosh gas volume,carbon consumption,and so on.The rationality and reliability of the model were validated using the actual industrial data of hydrogen-injected blast furnace of Jinnan Ironmaking Plant,and the relative errors of fuel ratio and gas utilization rate are controlled within 3%.Taking the oxygen content of the blast furnace and the amount of hydrogen injection as the main factors,the campaign behavior of blast furnace with single factor variation and two factors co-variation were studied and predicted.When just the oxygen content of the blast furnace is increased,the fuel ratio and theoretical combustion temperature of the blast furnace increase,while the direct reduction degree,blast volume and the bosh gas volume fall.Only when the hydrogen injection rate was increased did the fuel ratio,theoretical combustion temperature,and direct reduction degree of the blast furnace decrease,the blast volume decreases and the speed slows down,and the bosh gas volume declined first and then marginally increased.When the two factors are adjusted cooperatively,the theoretical combustion temperature can be controlled within(2142±2)℃ by increasing the hydrogen injection by 10 m^(3)/t while simultaneously increasing the blast oxygen content by 0.43%.Finally,the carbon consumption could be reduced.By combining the traditional blast furnace mathematical model with machine-learning optimization algorithm,the industrial tests of hydrogen-enriched blast furnace could be cost-saving and optimized,and the theoretical guidance for the stable operation and prediction is realized.