An investigation into the drainage characteristics and behaviour of hydraulically placed mine backfill and permeable minefill barricades

Hydraulic fills are quite popular as backfilling materials for underground voids created in the process of mining. For ease of transport through pipes, they are placed in the form of slurry and allowed to settle freely under self weight. Stopes can be approximated as rectangular prisms, and may extend as high as 200 metres. The horizontal access drives, used for transporting the ore, are blocked by porous barricade bricks before backfilling. Failures of barricades, and subsequent in-rush of wet hydraulic fill into the mines, have claimed several lives and contributed to severe economic loss world-wide. The objective of this research is to carry out a thorough experimental study of the hydraulic fills and barricade bricks, with particular emphasis on load-deformation and drainage characteristics. Two separate numerical models were developed in FLAC and FLAC3D, to simulate the backfilling process and to monitor the pore water pressure developments, fill and water heights and discharge rates. So far, the findings of the research were disseminated through four journal papers, a book chapter and seven papers in refereed international conferences. Overall, this study will improve the current state-of-the-art in hydraulic filling of mine stopes significantly. More than 25 different hydraulic fills, from five different mines, were studied. All Australian hydraulic fills fall within a narrow band of grain size distribution and are classified as silty sands or sandy silts. Their specific gravity values range from 2.8 to 4.5. Constant head and falling head tests were carried out on reconstituted samples, produced from the hydraulic fill slurry, in a process that replicates the sedimentation process in the mine. The hydraulic fills settled to a porosity of 37%-48%, void ratio of 0.58-0.93, dry density (g/cm3) of 0.58 times the specific gravity and relative density of 50%-80%. These values are in very good agreement with those measured in situ in Australia and U.S. The permeability values of the hydraulic fills, as determined in the laboratory, ranged from 10 mm/hr to 30 mm/hr, significantly less than the 100 mm/hr preferred for use in design by the mining industry. In situ, hydraulic fills with these permeabilities have performed satisfactorily with adequate drainage in the mine stope, implying that the 100 mm/hr limit may be excessively conservative when mine efficiency is considered. Full barricade bricks, cylindrical samples cored from the bricks, and specially cast samples were tested for uniaxial strength, Young's modulus, failure strain and permeability. Uniaxial compression tests were performed on more than 50 cores in an attempt to carry out an extensive statistical analysis. Beta distributions were fitted to describe the strength, stiffness and failure strain. It is shown that wetting the bricks reduces the strength by about 25%. A unique permeameter was developed to simulate one-dimensional flow through the bricks and to measure the permeability. This was the first ever attempt to measure the permeability of barricade bricks, and it was shown that barricade bricks are 2-3 orders of magnitude more permeable than the hydraulic fill, thus justifying the assumption in the numerical models that the fill-barricade boundary is free draining. A 2-dimensional numerical model was developed in FLAC that compared very well with Isaacs and Carter model, while having better features. The model simulates hydraulic filling process in the mines, and monitors the pore pressure developments throughout the stope, fill height and water height at all times. This model was extended to three dimensions using FLAC3D.