Pillar design – a science or an art?

Pillar design in a hard-rock mining environment typically takes the form of either empirical, analytical or, in recent years, numerical analyses (using a failure criterion to determine the strength and then obtain a factor of safety, or strength factor). In many instances, particularly in mining studies, the pillar-forming material properties are averaged and the pillar dimensions determined using these values as input. Figure 1 shows examples of some methods used to evaluate pillar stability.

Figure 1

Figure 1 Examples of empirical pillar design methods

Heasley (2000) identified that the overlooked part of a safety factor determination is the pillar load. With the increased use of modelling codes to determine the safety factor, very little attention is paid to how the pillar will perform while being loaded. This has been evident, in particular, where weak layers are present either within the pillar-forming material or in close proximity to the pillar foundations. The effect of these layers is often only recognized once the stability of the pillar has been compromised, and it is too late to minimize the effect of the instability (Figure 2).

Figure 2

Figure 2 Weak layer in a predominantly kimberlite pillar in North America

It has been observed on a number of occasions that it is not necessarily only the thickness of the weak layer that affects the pillar integrity under load, but also the composition of the layer. There have been five recorded occurrences in southern Africa over the past decade of large-scale instability, three of which resulted in mine closure and the other two requiring the re-establishment of alternate access into the mine. In all these cases, an empirical design solution was implemented, which neglected the effect that the presence of a shear layer as thin as 10 mm would have on a pillar under load (Figure 3).

Figure 3

Figure 3 Three southern African mines experiencing pillar stability issues due to a weak layer

Design methodologies have been proposed to take the effect of joints into account for the pillar strength determination. This effect takes the form of downgrading the uniaxial compressive strength of a rock type to estimate the rock mass strength of the pillar-forming material. However, the presence of one or more geologic structures with weak infill material is often overlooked when using published rock mass classification or characterization methods.

Unfortunately, in the absence of well-preserved geologic drill cores or clear, scaled core photographs, the designer has to rely on the geotechnical core logs. The quality of data recorded is determined by the experience of the person/s who conducted the core logging. From experience, the presence of a shear zone or sheared infill in joints, is overlooked. In some cases the geologic log was not consulted during the geotechnical logging process, with the result that the occurrence of core loss, which may be the result of the washing out of sheared or weak material, was overlooked.

While empirical methodologies exist to estimate pillar strength, pillar load is often determined using a numerical modelling code (unless it is a regular bord/room and pillar layout that lends itself to tributary area load estimation). The safety factor is then determined by the ratio of pillar strength to pillar load.

The art behind pillar design lies in the visualization of the most likely loading scenario and interpreting how the pillar may respond to the load. The ability to visualize the behaviour, and identify the shortfall of the design methodology followed, can only be gained through experience in the field.

visualization pillar design

Figure 4

Reference

Heasley, K.A., The forgotten denominator, pillar loading, National Institute of Occupational Safety and Health, Pittsburg, Pa., USA, 2000