The objective of installing ground support is to mitigate the risk of injuries and costs associated with rehabilitating excavation damage and associated production downtime. It follows that the greater the likelihood of stress damage or rockfalls, due to the inherent rock mass conditions, the more work is required from the support system. In addition, the exposure of personnel in these excavations and the quantity of production serviced by the excavation will significantly influence the support requirements.
While support systems have evolved over many years to meet these demands and experienced rock engineering practitioners generally apply appropriate engineering judgment, most design methods are based on a simple factor of safety and do not account for proper risk evaluations. In nature, rock mass characteristics are variable and considerable effort is required to understand the variability, particularly when the geology is complex. Stress fields are often unknown or are simply inferred from one or two stress measurements. Model bias and other uncertainties also need to be considered.
SRK, in conjunction with the Australian Centre for Geomechanics, has developed a risk-based approach to support design, which takes the variability of rock mass conditions and uncertainty into consideration. The approach includes methods of data collection and determining the variability of inputs into the subsequent analyses. Two potential failure modes are considered, namely joint-controlled rockfalls and stress damage.
The probability that rockfalls of different sizes occur in a given length of tunnel is estimated using a statistical block stability method. A simple discrete fracture network is used to generate blocks, which are analysed using limit equilibrium methods in a Monte-Carlo simulation. The results are presented as a frequency distribution of potential rockfalls for a range of rockfall dimensions.
Stress damage can be analysed using elastic and elasto-plastic numerical methods and excessive stress damage can be represented by the depth of failure or deformation exceeding a prescribed serviceability criterion. Monte-Carlo simulations are not practical if the analyses are complex and time consuming. Probabilistic methods, such as the Response Surface Method or the well-known Point Estimate Method, are then used to estimate the probability of excessive stress damage. A probability distribution of the length of the tunnel that exceeds this criterion is then determined by applying a binomial distribution.
The potential financial losses due to excavation damage are determined by estimating the rehabilitation time and production downtime based on the mining layout and site specific considerations. These are presented on a typical risk matrix for decision making.
The risk of injury is determined by assessing exposure of personnel and spatial coincidence. In this way, both the economic and safety risks are evaluated. Better ground support systems can be introduced to mitigate the risks, if required.
