The effect of statistical analytical measurement variations on the plant control parameters and production costs in cement manufacturing – a case study

Abstract

Raw materials used in cement manufacturing normally have varying chemical compositions and require regular analyses for plant control purposes. This is achieved by using several analytical instruments, such as XRF and ICP. The values obtained for the major elements Ca, Si, Fe and Al, are used to calculate the plant control parameters Lime Saturation Factor (LSF), Silica Ratio (SR) and Alumina Modulus (AM). These plant control parameters are used to regulate the mixing and blending of various raw meal components and to operate the plant optimally. Any errors and large fluctuations in these plant parameters not only influence the quality of the cement produced, but also have a major effect on the cost of production of cement clinker through their influence on the energy consumption and residence time in the kiln. This paper looks at the role that statistical variances in the analytical measurements of the major elements Ca, Si, Fe and Al can have on the ultimate LSF, SR and AM values calculated from these measurements. The influence of too high and too low values of the LSF, SR and AM on clinker quality and energy consumption is discussed, and acceptable variances in these three parameters, based on plant experiences, are established. The effect of variances in the LSF, SR and AM parameters on the production costs is then analysed, and it is shown that variations of as large as 30% and as little as 5% can potentially occur. The LSF calculation incorporates most chemical elements and therefore is prone to the largest number of variations due to statistical variances in the analytical determinations of the chemical elements. Despite all these variations in LSF values they actually produced the smallest influence on the production cost of the clinker. It is therefore concluded that the LSF value is the most practical parameter for plant control purposes.