Silicon Control in Blast Furnace
Silicon Control in Blast Furnace Hot Metal is one of the most critical problems in today’s Iron Making. With deterioration in quality of iron ore day by day, silica load in the burden material has a steady increase. The same story is with the Coke ash, which passes good amount of silicon in the hot metal. Every 0.1% reduction in silicon in the hot metal gives rise to 1.5 % increase in hot metal production. In the competitive scenario of today’s Iron Making hence it is necessary to reduce every bit of Silicon in hot metal to boost up furnace production and productivity. Experience of Blast Furnace Operation finds 0.5 to 0.7 % . Blast Furnace # 1 in Bhilai Steel Silicon in hot metal is the optimum for furnace productivity and furnace health. Plant, India had a typical problem of producing high silicon metal since July 2005. Other same capacity furnaces viz. Furnace 2 & 3 of the same plant did not show such fluctuation in silicon. Although the wide variation in input variables were known but the exact reason was difficult to identify due to complexities in Blast Furnace Process. Hence a Six Sigma Black Belt Project had been taken up to control the silicon level of BF # 1 hot metal and maintain it at an optimum level below 0.7 %. Pareto Analysis and Graphical Summary of Historical Data helped to find out a baseline of 64.88 % casts having silicon in hot metal more than 0.7% and an entitlement of 33% casts on daily average data base. A target of reduction of percentage casts having silicon more than 0.7% from 64.88% to 40 % had been taken up.
Through DMAIC process initially 53 inputs were identified in Detailed Map, which was scaled down to 5 critical inputs after C&E Matrix and FMEA. Regression Analysis was done on real time data and best-fitted line were plotted to find out the effects of those critical inputs on hot metal silicon. Finally a Regression Equation was found and trial values approximately near to the calculated value of the
Through DMAIC process initially 53 inputs were identified in Detailed Map, which was scaled down to 5 critical inputs after C&E Matrix and FMEA. Regression Analysis was done on real time data and best-fitted line were plotted to find out the effects of those critical inputs on hot metal silicon. Finally a Regression Equation was found and trial values approximately near to the calculated value of the