A 4-component model for settling slurry pipe flow has been previously described by Wilson et al (2006) and Sellgren et al (2014) to predict pipeline friction loss (hydraulic pressure gradient) over a range of slurry compositions: from fine to coarse particle size, narrow to broad particle size distribution, and low to high solids concentration. The method applies a weighted average of established pipeline models for various settling slurry flow regimes, according to the volume fraction of solids falling within the applicable size range for each model. Further development of the model was undertaken by Visintainer et al (2017a, 2017b), based on a comprehensive set of laboratory tests in 203 mm (8 inch) and 103 mm (4 inch) pipelines, and it was also adapted to the modelling of slurry pump performance derates. However, this work was all performed with solids having a specific gravity near 2.65, as is typical for many mineral processing and dredging applications.
The goal of the present work is to test the applicability of these models for settling slurries having a higher solids specific gravity, as may often be seen in the mineral processing of iron ore deposits. To that end, a test program was carried out in a 103 mm (4 inch) pipe loop using various compositions of an iron ore product having a solids specific gravity of 4.75 and containing both coarse and fine solids. By screening and flushing operations, a range of particle size distributions were created having different proportions of the coarse and fine elements. In all, 19 tests were performed with d50 particle sizes ranging from 50 μm to 3.2 mm and delivered solids concentrations from 10 per cent to 43 per cent by volume. Particle size distributions varied from very narrow to very broad, with d85/d50 ratios ranging from 1.75 to 26. Pipeline pressure gradient and pump performance data were collected and used to test the applicability of the previously developed 4- component models and to propose improvements to the models for the handling of high-density solids.