Quantification of Impact of Uncertainty on Emissions in a Cement Manufacturing Plant: Surrogate Modeling-Based Approach

Abstract

A stable and efficient cement manufacturing process is essential to minimizing raw material and utility consumption while maximizing productivity. However, process uncertainty arising from feed composition and the process conditions such as flow rate, temperature, etc., put a challenge to realizing stable and efficient operation of a cement manufacturing plant. In this study, a quantitative analysis based on polynomial chaos expansion (PCE) is performed to evaluate the collective impact of input uncertainty on emissions. To achieve this, models based on least-squares booting and Artificial Neural Networks (ANNs) were developed to forecast CO₂, O₂, CO, and NO_x. Then, the relatively more accurate model, the ANN model, was used as a surrogate within the PCE for uncertainty quantification. The predictions were based on varying input variables having ±10% uncertainty in the feed flow rate, kiln air flow rate, tertiary air flow rate, and coal flow rate. The proposed framework accurately quantifies the impact of uncertainty on the emissions of a cement manufacturing plant.