Modelling granular flow in caving mines: large scale physical modelling and full scale experiments

Abstract

Over the last 20-30 years, the trend toward the mining of deeper, lower cost, higher tonnage deposits, has made caving mining methods increasingly attractive to the global mining industry. During this period the scale of these caving operations has increased significantly.Understanding and planning for granular flow is acknowledged as key to the safe and efficient operation of caving mines, however a detailed review of flow research revealed that little fundamental research had been done to justify these increases in mining scale. At present, flow can not be reliably simulated, as developing numerical models have no reliable data against which they can be calibrated and validated, and existing flow paradigms have been developed from physical models which do not satisfy the requirements of similitude.Extensive experimental programmes were therefore designed and implemented to gather reliable experimental data on granular flow for modem, large scale caving environments. This data has been used to develop new understanding of the mechanisms controlling flow, to show that the scale increases undertaken by the industry have outmoded previous design assumptions, and to develop new design criteria for modem caving environments.For block caving mines, the largest and most complex physical model yet built for the modelling of flow in caving mines was designed and constructed. Using this model, an experimental programme was carried out to remove confusion about the existence of a particle size effect. This effect was identified, and quantified for the first time on a model in which flow is not significantly affected by discrepancies in similitude.Hypotheses explaining the mechanisms driving this particle size effect were developed, and existing theories on flow geometry assessed in light of new experimental evidence. This assessment showed that industry standard design guidelines describing the development of flow geometry significantly underestimated the width of draw zones, probably due to their origins in small scale physical models.For sublevel caving mines, an extensive full scale experimental programme was carried out, yielding arguably the most comprehensive set of flow measurements in existence for a modem SLC geometry. The results of this programme allowed the accurate measurement of draw in large scale sublevel caving rings. Reliable calculations of primary and secondary recovery for these rings were made, and the variation in these calculations quantified. Experimental evidence was obtained that the source of early dilution entry was initially from above the ore side of the ring, rather than the waste side of the ring, as had been hypothesised previously. Interaction of draw envelopes was shown not to have a significant effect in a modern SLC layout.Depth of draw was also effectively measured for the first time for a burden greater than 2.5m. It was found that shallow draw is common, and that frequently, the full depth of the fired ring is not recovered. Hypotheses explaining the mechanisms behind this draw behaviour were discussed, and the importance of secondary recovery in boosting overall recovery was shown.Combined with these experimental results was a unique set of visual observations of fired rings (believed to be the first of its kind), and large scale physical modelling results, which allowed the full scale experimental results, and the hypotheses on the mechanisms driving them to be confirmed visually. Based on the results of this experimental programme, potential improvements to standard modem SLC layouts were discussed.The possibility of the observed behaviours being endemic to the SLC method (due to the expansion in mining scale with little effective accompanying research) was proposed, and the need for further full scale experiments, in both block caving and SLC environments established, to independently confirm the new findings in this Thesis.The experimental programmes carried out for this Thesis have indicated that existing industry standard theories on flow do not reliably model flow for modern caving geometries. New paradigms and standards are required for the modelling of flow for modern, large scale layouts, and this Thesis presents new standards for the modelling of flow in both block caving and sublevel caving environments.