Iron chloride solutions are a waste product in steel pickling plants. A technique to recover the spent solutions is a spray roasting process, where the spent solution is sprayed into a hot reaction atmosphere and solid iron oxide particles are formed. The particle history in spray roasting reactors has an important influence on the efficiency of the recovery process and on the quality of the by-product Fe2O3. The iron oxide underlays strong quality demands for further processing. The particle quality is influenced by the plant design and operation parameters. To investigate the influence of those properties on the iron oxide produced in the spray roasting process, a model for CFD simulations has been developed. It describes the particle formation and chemical reaction of the iron chloride solution in the spray roasting reactor in a simplified way suitable for CFD simulations. Simulations of two different industrial reactor configurations show the capability of the model to predict the influence of geometric variations on the composition of the resulting iron oxide. (C) 2013 Elsevier B.V. All rights reserved.