Implementing Spatial Variance and Investigating the Effect on the Stability of a Mine Entry Roof in a Longwall Coal Mine

Roof falls are known as a potential source for injuries and fatalities in underground coal mine. The roof rock is heterogeneous and therefore they are normally spatially correlated. The spatial correlation varies from far away to close distance. In the past, research used uniform scaling approaches to predict the roof failure, which ignored the inherent spatial variance. This has often produced unrealistic behavior in the rock mass that ranged from under design to over design of structures. In this paper, we have shown the effect of including spatial variance in the property on the stability of gate road entry in a longwall mine. The realistic field database is created using Extreme Value stochastic model. Two scaler parameters from both horizontal and vertical directions are added to control the spatial correlation. A few cutting sequences are considered to observe the influence of the spatial variance on roof behavior. The stochastic model showed that the variability of cohesion and friction fields, especially the weak area, causes the failure to propagate in the immediate roof, which is different from the deterministic model. Additionally, comparing the principal stress magnitude, the traditionally deterministic model was more conservative than the spatial correlated random model. The model results showed that the spatial variance is a critical factor in roof stability analysis.