Statistical analysis of the spray roasting operation for the production of high quality Fe2O3 from steel pickling liquors

Statistical tools were used to analyse the spray roasting operation so as to identify the main operating variables for the production of commercial value iron oxide (Fe2O3) by pyrohydrolysis from steel pickling liquors. The study was carried out using a 24 factorial design of experiments and the effects of the following variables were investigated based on large scale experimental data: temperature, liquor feed flow rate, pressure and feed air/COG (coke oven gas) ratio. The quality of Fe2O3 was evaluated according to the contents of Cl, Fe, Si, Mn and H2O, as well as pH and granulometry. For the conditions investigated, the major influencing variables were temperature and feed flow rate; in fact, commercial value Fe2O3 was obtained solely at higher evaluated levels of temperature and flow rate. The contents of Fe, Si and Mn in the Fe2O3 were affected by the content of respective species in the feed liquor, which in turn depends on the type of steel treated in the pickling line. Simple statistical models were proposed to describe some Fe2O3 characteristics. Before this study, only 10% of the Fe2O3 produced at USIMINAS was commercialized, due mainly to the high content of chloride (approximately 0.45%, w/w). Today, all oxide produced meets quality requirements.

Research highlights▶ A large-scale pyrohydrolysis process was evaluated statistically. ▶ Temperature and feed flow rate were found the major influencing factors. ▶ Simple statistical models valid at industrial scale were proposed. ▶ Production efficiency was improved and environmental pollution reduced considerably. ▶ Commercial value Fe2O3 powder from the effluent of steel pickling line was obtained.