Development of coke properties during the descent in the blast furnace

Sammanfattning: The efficiency in use of reducing agents in blast furnace (BF) ironmaking has been significantly improved over the years. At most blast furnaces, auxiliary fuels are injected to replace some of the coke. To further reduce carbon consumption, prevention of losses and modification of raw materials or process conditions are required. In this study coke properties, their development during the descent in the BF under different process conditions, contribution to off-gas dust as well as modifications of coke aiming for reduced energy consumption in the BF have been investigated. The changes in chemical composition of coke ash, mineral phase developments and coke participation in the endothermic gasification reaction with CO2 and coke graphitisation were studied. Also, in order to reduce the losses of material in the off-gas, a characterisation of the off-gas dust and its connection to BF conditions and coke property development in the BF has been made. Coke properties have been studied in the laboratory and in samples taken out from different parts of the LKAB Experimental BF (EBF®) during varied operational set-ups. Properties of coke from the BF high-temperature region were studied via tuyere core-drilling of samples from the EBF and an industrial-scale BF.The results showed that the coarser dry flue dust is mechanically formed. Coke fines originating from the upper shaft dominates the coarser fractions. Gasification in the shaft has a negligible effect on the high quality coke used in the EBF. The finer dust fraction, sludge, consists mainly of chemically formed spherical particles in the order of <1 μm. Gaseous compounds formed in the BF high-temperature area precipitate from the ascending gas as the temperature decreases. Flow conditions in the top of the BF and the fluidisation properties of fine particles determine out-flow of off-gas dust. Low off-gas temperatures, and thus lower off-gas velocities, are favourable for low flue dust amounts expelled from the blast furnace. The strength and reactivity of coke at high temperature, measured in the standardised test, was compared with coke from basket samples charged into the EBF. Coke reactivity in the EBF is considerably lower for all coke types studied compared to coke studied in the CRI/CSR test. Due to higher gasification in the laboratory test the ash content is higher in that test. However, gaseous components in the EBF, such as recirculated alkali, contribute to changed ash composition in BF coke. Laboratory studies using fixed bed reactor (FBR) and a thermo gravimetric analyser (TGA) on the same material, original as well as Fe- and Ca-activated coke samples, identified the effects on reaction behaviour and were the basis for selection of the potentially best method for activating nut coke to achieve higher reactivity with CO2 when charged into iron-bearing layers. All types of Fe- and Ca-containing activation agents used increase the apparent reaction rates for coke with low reactivity and contents of catalytic components in the ash. Activation with a solution of Fe(NO3)3 has the strongest effect on the reactivity, followed by slurry of Ca(OH)2 and iron oxides (Fe2O3 and Fe3O4) in descending order. Use of a non-specific weight loss in TGA as a basis for apparent reaction rate demands measurements in a temperature interval free from other reactions involving the activation agents. The reaction rate is reduced when CO is present in the reaction gas. Aiming for a method for determination of thermal history in BF samples the correlation between temperature and coke graphitisation was studied using three different data processing methods for removing the influence from overlapping peaks during X-ray diffraction measurements. Structural changes in coke during heat treatment are accompanied by chemical transformations in ash compounds and changed the pattern of disturbing peaks. Key ash phases were SiO2 and mullite. At high temperatures (~1500°C) the SiO2 peaks are reduced in magnitude and finally disappear to be replaced with SiC peaks. The graphitisation was correlated with temperature and time in argon atmosphere and the impact from hot metal and slag was investigated. The graphitisation degree in coke samples from the EBF raceway and hearth was estimated and the evolution of structural order was found to be suitable in order to estimate the thermal history of coke in blast furnaces.Gaseous compounds formed in the BF high temperature area are precipitated from the ascending gas as the temperature is decreased. Flow conditions in the top of the BF and the fluidisation properties of fine particles determine out-flow of off-gas dust. Low off-gas temperatures, and thus lower off-gas velocities, are favourable for low flue dust amounts expelled from the blast furnace. The strength and reactivity of coke at high temperature, measured in the CSR/CRI test was compared with coke from charged basket samples into the EBF. Coke reactivity in the EBF was considerably lower for all coke types studied compared to in the CRI/CSR test. Due to higher gasification in the laboratory test the ash content was higher in the test. However, gaseous components in the EBF such as recirculated alkali contribute to changed ash composition in BF coke. Laboratory studies using fixed bed reactor (FBR) and thermal gravimetric analysis (TGA) on the same material of original as well as Fe- and Ca-activated coke samples stated the effects on reaction behaviour and were the basis for selection of the potentially best method for activating nut coke to achieve higher reactivity with CO2 when charged into iron bearing layers. All types of Fe and Ca containing activation agents used increased the apparent reaction rates for coke with low reactivity and contents of catalytic components in the ash. Activation with a solution of Fe(NO3)3 had the strongest effect on the reactivity, followed by slurry of Ca(OH)2 and iron oxides (Fe2O3 and Fe3O4) in descending order. Use of a non-specific weight loss in TGA as basis for apparent reaction rate demands measurements in a temperature interval free from other reactions involving the activation agents. The reaction rate was reduced when CO was present in the reaction gas. Aiming for a method for determination of thermal history in BF samples the correlation between temperature coke graphitisation degrees was studied using three different data processing methods for remove the influence overlapping peaks during X-ray diffraction measurements. Structural changes in coke during heat treatment were accompanied by chemical transformations in ash compounds and changed the pattern of disturbing peaks. Key ash phases were SiO2 and mullite. At high temperatures ~1500°C the SiO2 peak are reduced in magnitude and finally disappear to be replaced with SiC peaks. The graphitisation could be correlated with the temperature and time in argon atmosphere as well and the impact from hot metal and slag. Coke samples from the EBF raceway and hearth were also measured and estimated and the evolution of structural order was found to be suitable in order to estimate the thermal history of coke in blast furnaces.