Flotation of carbon values from blast furnace flue dust using statistical design
The blast furnace flue dust constitutes a substantial quantity of the wastes generated in the integrated steel plants. It is generally stored inside the plant premises causing environmental and space problems. Chemical analysis of a typical BF flue dust sample indicates that it contains high values of iron and carbon accompanied by harmful elements like Na, K, Zn, Pb, etc. depending on the chemistry of raw material used in the blast furnace operation. These elements are considered a major concern for sustaining stable operation of a blast furnace. However, this material can be utilized if both carbon and iron are recovered economically through a cost-effective technology.
A representative sample of blast furnace flue dust containing ~30.34% carbon, 4.6% volatile matter (VM), and 65.03% ash was collected from Tata Steel, India. Flotation studies were carried out to recover carbon values from this blast furnace flue dust. The variables chosen for the experiments and their levels are given in the table. The experiments were designed according to a 24 full factorial design method, and regression equations for recovery of carbon and product grade were developed from the data. The equations are as follows:
YR = 65.94 + 2.05X1 – 0.81X2 + 1.42X3 + 10.2X4 – 1.56X1X2 + 0.98X1X3 + 0.94X1X4 – 0.60X2X3 + 3.23X2X4 + 2.91X3X4 + 2.73X1X2X3 – 2.03X1X3X4 – 1.01X1X2X3X4 (1)
YG = 61.81 – 1.30X1 – 2.30X4 + 0.87X1X2 + 1.33X1X3 – 1.04X1X4 – 0.92X2X3 + 0.37X2X4 – 0.56X3X4 – 0.62X1X2X3 + 0.66X1X2X4 + 0.77X1X3X4 – 1.02X2X3X4 – 0.76X1X2X3X4 (2)
YR and YG = the recovery of carbon and grade of carbon, respectively, and
X1, X2, X3, X4 = dimensionless coded factors for dispersant, pH, frother, and collector, respectively.
From the statistical analysis, it was found that all the main effects on recovery are significant at 95% confidence level. The effect of dispersant, collector, and frother were positive, whereas that of pH is negative. The collector has the strongest effect on the recovery followed by dispersant, frother, and pH. In the case of grade, the main effect of dispersant, collector, and all the interaction effects are significant at 95% confidence level. The main effect of both the dispersant and collector are negative and the effect of the collector is stronger than that of dispersant. Experiments were performed to optimize recovery and grade using the method of steepest ascent. The recovery of carbon was optimum at 6.887 kg/t sodium hexa-
metaphosphate, 5.94 kg/t diesel oil, 1.35 kg/t MIBC, and pH 8. The recovery of carbon was 90.0% and the product contains 59.5% carbon, 5.6% volatile matter, and 33.7% ash. The maximum carbon grade of 66.4% was obtained at 2.492 kg/t sodium hexametaphosphate and 2.25 kg/t diesel oil, 0.9 kg/t MIBC, and pH 8. The recovery was 50.5% and the composition of the concentrate was 66.4% carbon, 5% volatile matter, and 27.3% ash.