Cement in-line calciner kiln modeling for heat optimization using a design of computer experiments

Abstract

A model of a dry process cement kiln was implemented in Aspen Plus V9 and validated with a 2000 metric ton per day in-line calciner cement kiln data. A simpli ed model of the process was obtained using a 20-run Plackett-Burman design of experiments on Aspen Plus simulations with 19 input variables related to false air (in-leakage air), oxygen concentration, calciner temperature, cyclones e ciency and clinker's cooler bed height. Linear metamodels for the speci c heat consumption (SHC) and other response variables were obtained. The metamodel indicated that (i) false air in cyclones 4 and 5, (ii) calciner's control temperature, (iii) oxygen concentration at the calciner exit and (iv) cooler's clinker bed height were the most signi cant input variables a ecting SHC. The SHC obtained from simulation based on the optimized values from the metamodel resulted in a reduction of ca. 29 kcal/kg of clinker. A sensitivity analysis indicated that the two most impacting variables were the oxygen concentration at the calciner's exit and the cooler's clinker bed depth. The SHC metamodel is a powerful tool for capturing the complexity of the process simulations on a simple and easy to use model.