Iron ore sinter is produced from fine-grained ore in order to provide a direct charge to the blast furnace. During the sinter production process fine sinter is produced that is not acceptable as feedstock for the blast furnace. This fine material is screened off and returned to the sinter plant to be recycled. The production of these recirculating fines therefore results in loss of revenue for the sinter producing facility. The aim of this study is to compare the composition and phase chemistry of the fine and normal sinter to determine if chemical or operational changes can be made to reduce the amount of fine sinter formed. This will be achieved through the analyses of real industrial sinter samples, synthetically produced lab-based samples and sinter test pot samples. Iron ore sinter was collected from the Vanderbijl Park sinter plant in South Africa. The samples were prepared for X-ray diffraction (XRD) and X-ray fluorescence (XRF) analysis to test the reproducibility of the equipment used as well as the reproducibility of the sample preparation method. The results obtained showed that the methods applied produced accurate results and the preparation method was then applied on all samples that were analyzed. The XRD analyses show that the sinter contains spinel (mainly magnetite with variable Mg), hematite, dicalcium silicate (C2S) and silico ferrite of calcium and aluminum (SFCA) and that there are distinct differences between the fine and normal sinter. Fine sinter contains more hematite and less SFCA than normal sinter. The presence of the SFCA is considered to be essential for the production of strong sinter. The XRF analyses show that there are no distinctive differences in the chemistry of fine and normal sinter. The samples were then analyzed with an electron microprobe. It was found that the compositions of some of the SFCA phases present in the samples do not correspond to those described in the literature. Optical microscopy combined with point counting was conducted on the fine and normal sinter in order to determine differences and to compare the point counting data to the XRD results. The point counting results showed that the hematite present in the fine sinter is largely relict or unreacted hematite. Sinter pot test samples were analyzed with XRD and XRF. It was found that the pot test samples exhibited similar trends as the samples taken at the sinter plant. This shows that it is not only plant parameters such as sample transport that result in the formation of fines, but that carbon addition, flame temperature and reaction time may also play a role in the formation of fine sinter material. It has been concluded that the production of fine sinter is a direct function of the amount of hematite present in the sinter. The proposed hypothesis for this phenomenon involves the incomplete reaction of the sinter material during processing. Suggestions to decrease the amount of fine sinter formed include: uniform heat distribution during ignition, pO2 alteration by reductant addition, lower ignition temperature, regulating the cooling regime and decreasing the grain size of lime.
Dissertation (MSc(Applied Science))--University of Pretoria, 2008.