Inter-influencing factors in affecting BF coke rate – statistical approach

A blast furnace (BF) is a multivariate system which is subjected to a large number of inter-influencing variables affecting its performance: productivity and coke rate. It is important to isolate the inter-influence before drawing any conclusions. It is essential to support the parameters’ effect (on the responses) through theoretical consideration. . The explanatory variables affecting the coke rate have been statistically analysed in case of typical BF at Bhilai Steel Plant, SAIL and discussed in the present paper. The results (sixteen years’ monthly average data points) show that some variables are not orthogonal which have been explained in terms of (a) natural cause and effect relationship amongst the variables and (b) apparent inter-relationship amongst the explanatory variables and masking the effect. The inter-influences of the variables have been isolated and their effect on coke rate has been discussed which are substantiated by the theoretical consideration. The study shows that the variable, burden rate was the most significant one followed by temperature of hot blast. A change in burden rate has been mainly reflected by a change in weight of raw limestone in the burden.. The adverse effect of the variables, ash content of coke was inter-influenced by the variable, limestone consumption. However, analysis shows the variable, ash content in coke affected the coke rate, though marginally when the inter-influence was isolated. The range studied against this variable during the period was quite high: 12 to 554 kg/ THM which indicates that lump ore constituted high proportion in the burden with nearly zero sinter consumption during some period. The empirical equation, established through the analysis of more recent data shows that an increase in ash content of coke (range 21.5- 26.6%) increased the carbon rate and the correlation coefficient was similar to those observed previously. The study indicates that an equation developed need to be modified when the value of an independent variables deviates from the range of data considered in the analysis