The UNSW coal pillar design formulae were developed in the mid 1990s using the same maximum likelihood statistical approach as used by Salamon and Munro to develop their famous formula for South African conditions in the late 1960s. However, the UNSW formulae have been extended to account for squat and rectangular pillars. The formulae find extensive application in Australia and remain the only design approach which assigns a probability of stability to the design outcome. This paper reviews some of the considerations arising out of the application of pillar design formulae in general, and the UNSW pillar design methodology in particular. Issues evaluated include: the effects of water pressure on pillar load and stability in flooded workings, confidence limits associated with the strength of rectangular shape pillars; pitfalls in modified pillar design formulations that are based on relationships between pillar width to height ratio and pillar safety factor; the time dependent strength of coal pillars; and the impact of the assumed statistical pillar failure distribution on the computed probability of stability.
The history of mining is blemished with failures of pillar systems. The Australian coal mining industry has not been immune from such events, with at least 18 failures occurring over the last 25 years. Six events occurred suddenly and without warning in active working sections, fortuitously on holidays or idle shifts. Another eight were creep events, three of which caused the loss of main development roadways in active mines. A number of fatalities in pillar extraction operations during this period can also be indirectly attributed to a poor pillar design knowledge base. Since the early 1990s, there has been a research focus in Australia on coal pillar design. One outcome has been the UNSW Pillar Design Methodology, which now finds extensive application. In all likelihood, this design procedure would have prevented all 18 known pillar failure events and a number of the pillar extraction mishaps. However, it has to be applied correctly and judiciously. This is not always the case,
resulting in two collapses since 2001 in bord and pillar workings that were less than 2 years old. There is no one design procedure for all circumstances and it is important that end users do not misapply pillar design methodologies or push the limits of a pillar design procedure too far.