(83a) Measurement and Prediction of Process Relevant Segregation for Multi-Component Bulk Materials

Bulk materials segregate due to a variety of mechanisms. Fines may sift down through a matrix of coarser particles, depositing below the process charge point. Angle of repose differences may cause components within the mixture to separate as they flow down a pile. Fines may also be carried by air currents, depositing where these air currents decrease. Large particles may impact on semi-fluid layers, penetrating the top layer and causing top-to bottom segregation. In order to mitigate segregation within existing processes, it is critical to determine the types of segregation that a prescribed material may be sensitive to within the process. The magnitude of the potential segregation, along with the characteristic segregation pattern, must be accurately known when the bulk material is exposed to feed conditions similar to process conditions. It is critical that the segregation potential, or intensity of segregation problems, be measured at conditions relevant to the process for accurate scale up. Almost any mixture can segregate due to any mechanism if the proper stimulus is induced in the material. The real question to be answered is: will the mixture in question segregate when exposed to a prescribed stimulus similar to plant conditions? Segregation tests must be done with process scale-up in mind. Under these test conditions, the segregation pattern and segregation magnitude, coupled with the computed flow profile in process equipment, determines if the process will promote separation of particles or mitigate segregation. The following paper presents a new testing methodology to identify the potential process relevant segregation mechanisms and magnitude. This new methodology uses a controlled feed method coupled with spectral measurement of segregated material to estimate the spatial concentration of chemically different materials or different particle sizes in the mixture. It also outlines a methodology to predict specific segregation leaving handling systems using easily measured flow properties. Thus, prediction of segregation profiles expected from process equipment is possible by using results from simple segregation potential and standard flow property tests.