Skip to main navigation Skip to search Skip to main content

Simulation aided flow sheet optimization of a cement grinding circuit by considering the quality measurements

Research output: Contribution to journalArticlepeer-review

14 Citations (Scopus)

Abstract

The paper focusses on improving the energy utilization of a cement grinding circuit by changing the flow sheet of the process. The circuit was comprised of ball mill, static classifier and dynamic classifier and in the current flow sheet the fine product of static classifier is sent to feed stream of dynamic classifier. However, it was experimentally proved that sending static fine to the final product silo had several benefits. In order to quantify the benefits, 250 kg of static fine stream was collected then mixed with the final product and series of tests i.e., compressive strength, particle size, chemical assays and water demand, were undertaken. The results showed that mixing static fine with final product increased the 28-day strength by 2.3%, 4.3% and 4.8% for CEM I, CEM II and CEM IV type cements respectively. Improved strength results were attributed to having wider size distributed product (reduced slope value) that decreased the water demand of the cement. Such an increase in strength enabled to coarsen the final product at the same quality figures. Therefore, in the next phase of the study, the grinding circuit was simulated as the static fine stream would be sent to final product silo that indicated the production rates of the circuit could be increased by 9.7% for CEM II, 7.4% for both CEM I and CEM IV types.

Original languageEnglish
Pages (from-to)1242-1251
Number of pages10
JournalPowder Technology
Volume301
DOIs
Publication statusPublished - 1 Nov 2016

Keywords

  • Cement
  • Dry grinding
  • Fine grinding
  • Modelling
  • Optimization
  • Simulation

Fingerprint

Dive into the research topics of 'Simulation aided flow sheet optimization of a cement grinding circuit by considering the quality measurements'. Together they form a unique fingerprint.

Cite this