Study On The Analysis Of Coking Coal Properties And The Prediction Of Coke Quality In Laisteel

Though coal reserves are abundant in our country, high-quality coking coals are very poor such as coking coal and fat coal. Coking enterprises need to do some changes based on their own condition, such as adjusting the composition of coal blending reasonably and reducing the cost for the blending coal, which has important practical value for coal blending and coking production. In recent years,for the stable production and improving management level of coking enterprises, the research of coke quality prediction model has become more and more important. Although people have established all kinds of the coke prediction model, the prediction model do not have general applicability due to the different coking technology, the different coking coals and so on.In this paper, we are in a research object of Laisteel's coking plant.First of all, we have chosen the 10 coking coals as the main analysis object of coal species,by measuring and analyzing the various indicators of coking coals and we can make some conclusions:When the average maximum reflectivity of coking coal is about1.35,the maximum thickness of Y is about 17 mm in the colloid layer and the bond index G value is about 89,the quality of coke produced by coking coal is better than others; its M25 is about 89%, M10 is less than 6%, CRI is about 16% and CSR is around 76%.The ash content of coke is more correlation to the quality of coke than others,its correlation coefficient is above 0.7.When we studied the additive property of Laisteel's coal properties,we found: The blending coal have good additive property with coal ash, sulfur and volatile between the single coal, but the additive property of the G value of the coal blending index is without ideal.We have analyzed the industrial properties of 34 common coal in details, and then we picked out the reasonable parameters effected on coal quality, which was based on the coking data in 1~9 months historical data of coking plant 4 coke ovens in 2015. And by the coking data in 1~5 months of 2016 to test the model of the coke's quality prediction model were inspected on Laisteel's 4 coke ovens. The correlation coefficient of the coke prediction model is in a reasonable range, which can meet the coke oven production demand;This paper use the Vdaf, G and Ad to predict the cold strength of coke(M10 and M40) and to predict the thermal properties of coke( CRI and CSR).And it is feasible in the situation that the coking ovens are stable.