Cement industry is highly energy and emissions intensive. Thermal energy is majorly used in the kiln and calciner systems, accounting for over 90% of total energy use in cement plant, and virtually all of the fuel use. Therefore, the precalciner is one of the key equipment in cement manufacturing process impacting the energy efficiency of the pyroprocessing unit. The purpose of this paper is to demonstrate the potential of statistical modeling to support the optimization of precalciners. According to the current knowledge of the authors, this is the first investigation using statistical models to optimize combustion system operating conditions by analyzing historical and experimental design data. Multiple linear regressions were obtained for each of the two approaches with statistically comparable variances. It should be noted that the optimum obtained with the experimental design is better because the corresponding model describes a larger range of operating variables. On the basis of the experiment-based model, 60% of the total unburned organic carbon could be reduced. (C) 2018 Elsevier Ltd. All rights reserved.