Minimisation of Chromium Leaching from Low-Alloy Electric Arc Furnace Slag by Mineral Modifications

Sammanfattning: Iron and steel are produced continuously as a vital part of modern life. The largest by-product of metal production is slag. Slag is an essential part of the production of steel since, in the furnace, the molten slag removes impurities from the metal. When the slag which consists of oxides, solidifies minerals form. Slag can then be used in different applications: fertiliser, water purification, cement, concreate and as building material. By using slag, natural resources can be saved. In the electric arc furnace, EAF, scrap is melted to produce steel. The scrap contains chromium which partly distributes to the slag. Unfortunately, excessive leaching of chromium restricts slag usage. Chromium leaching occurs when chromium-containing minerals in slag dissolve. In low-alloy electric arc furnace slag there are three chromium containing minerals: spinel (Mg,Fe)Cr2O4, magnesiowüstite (Mg,Mn,Fe)O and brownmillerite Ca2(Al,Fe)2O5. Of these minerals the spinel has already been determined to be stable. The aim of this thesis is to minimise the chromium leaching of low-alloy EAF slag by modifying the minerals and/or mineralogy in the slag so that the chromium-containing minerals do not dissolve. In addition, it was discovered that ageing of low-alloy electric arc furnace slag may increase the chromium leaching. Autoclave treatment of remelted slag resulted in chromium leaching at the same magnitude as before remelting. Autoclave treatment was performed before the leaching test to determine the chromium leaching of aged slag.Instead of indirectly examining the chromium leaching from magnesiowüstite and brownmillerite by comparing chromium leaching from slag samples, the dissolution properties of the minerals were investigated individually. Dissolution of different compositions of magnesiowüstite and brownmillerite were studied at pH 7 and pH 10 for 40 hours. The conclusion was that increased iron content decreased the dissolution rate of both magnesiowüstite and brownmillerite. Magnesiowüstite (Mg,Fe)O did not dissolve at pH 10 when it contained 60 wt% FeO and did neither dissolve at pH 7 or pH 10 at higher FeO content. According to thermodynamic calculations, the FeO content in the magnesiowüstite can be affected and this was investigated. Laboratory and full-scale experiments showed that the composition of magnesiowüstite in slag could not be correlated to chromium leaching. The conclusion was that magnesiowüstite was not the main mineral leaching chromium.Brownmillerite composition is difficult to control in slag and had a significant impact on the dissolution rate; therefore, brownmillerite in slag should be avoided to minimise leaching of chromium. From thermodynamic calculations two options to avoid brownmillerite formation were identified: decreasing the basicity (CaO/SiO2 ratio) or increasing the cooling rate. Both methods were tested in laboratory scale using low-alloy electric arc furnace slag. When the basicity was decreased by SiO2 addition, the chromium leaching of the slag decreased. The chromium leaching was correlated to the brownmillerite content, the chromium leaching decreased with the decreasing basicity until brownmillerite was no longer be detected by XRD analysis. After that the chromium leaching did not decrease significantly. This indicates a correlation between basicity, brownmillerite and chromium leaching. When cooling of slag was investigated the slower cooling rates increased the chromium leaching and fast cooling rates decreased the chromium leaching. The results were verified by full-scale experiments. Si-sand addition was used to decrease the basicity to 2.2; according to thermodynamic calculations this would prevent brownmillerite formation. The targeted basicity of 2.2 was difficult to achieve, as the basicity varied between batches since the scrap composition and other parameters changed. Batches from other experiments were included for more data. It was seen that the chromium leaching of batches below basicity 2.2 did not reach the same magnitude as the batches above basicity 2.2. The full-scale trials with different cooling methods confirmed that the faster-cooled water-sprayed slag had lower chromium leaching than slag that had cooled by its own.Basicity between 2.0 and 2.2 minimises chromium leaching but decreasing the basicity to specific values in full-scale is challenging. A decrease in basicity is the most efficient method of decreasing chromium leaching, but faster cooling also decreases the chromium leaching.