Author(s)

C. Lutz, J. Matthes, P. Waibel, U. Precht, U. Schweike, G. Beuchle, K. Garbev, P. Stemmermann, H. B. Keller

Abstract

The production of cement is one of the most energy and CO₂ intensive industries. At the Karlsruhe Institute of Technology a new “green” process for the production of cementitious materials was developed with a high potential for saving energy and CO₂-emissions. In a two-stage process a new hydraulic binder (Celitement) comparable to Portland cement is produced from limestone and sand. Until now, the product was only analysed with complex analytical methods. In the future it is necessary to have a robust measurement system, which is able to determine the product quality online during its manufacture. In this paper, the development of a method for the prediction of auxiliary quantities for characterization of the product quality based on an online near-infrared-spectroscopy (NIR) is presented. Initial methods with data pre-processing and partial-least-squares-regression for systematic development of near-infrared calibration models and their application to new data sets are presented. In addition, a new method for data pre-processing by adapting the target vector to the working point in comparison to the calculation of calibration models with absolute values is shown.

Keywords

Celitement, cementitious material, NIR-spectroscopy, calibration model, partial-least-squares-regression