Using Simul8 to model underground hard rock mining operations

CIM Bulletin, Vol. 98, No. 1090, 2005

N. Vayenas and G. Yuriy

The file is a zipped PDF document.The combination of a predictive reliability assessment model and discrete-event simulation can be an effective methodology to study the impact of equipment failures on the capacity of production systems in mines. In this study, a predictive reliability assessment model using genetic algorithms is combined with a discrete-event simulation model to analyze mine equipment systems.Discrete-event simulation studies the evolution of a system as variables change instantaneously at separate points in time. Computer simulation allows rapid manipulation of the major parameters of a system without the need for real-life experimentation. The Simul8 simulation software, by Visual Thinking, is used in this research. Simul8 is different from many other simulation packages because its model-building approach is not based on laborious programming, but rather on the ability to draw a scenario on screen and simply filling in numerical data where necessary.The simulation model developed for the purpose of this research depicts a typical two-level, sub-level stoping method section of an underground mine in the Sudbury region in Ontario. While some simulation packages include sophisticated 3D visualization of machinery operations, Simul8 offers a two-dimensional representation of equipment movement in the road network. The animation is simple and effective, portraying the behaviour of the simulated entities over time.In the research work, the reliability assessment model based on genetic algorithms provides input in the form of times between failures (TBFs) to predict load-haul-dump (LHD) failures in the simulation model, affecting equipment utilization and, most importantly, overall production throughput. Genetic algorithms represent selection or search algorithms that are based on the principals of natural selection. They combine a structured, though somewhat randomized, mode of information exchange with the basic rule of survival of the fittest gene (or gene pool). The reason for selecting genetic algorithms is the fact that the reliability of mining equipment changes over time due to its dependence upon several covariates/factors (e.g. the operating environment, and number and quality of repairs). These factors create a combined and complex impact on the reliability function. This impact encapsulates and inherits, to some degree, the individual characteristics of the factors as they evolve over time. By using genetic algorithms, it is attempted to capture the impact of the factors on the reliability function of a piece of equipment by mimicking the process of heredity and natural selection.Through a case study, it is intended to assess the impact of LHD breakdowns on the production throughput. A comparative analysis to assess the impact of LHD failures on production was performed in two parts: first, by taking into consideration LHD failures, and subsequently without considering LHD failures. The parameters of interest under study are tonnage throughput, equipment utilization, and LHD availability. As expected, the average tonnage is significantly greater when not taking into consideration the LHD mechanical failures. Under the assumed conditions for this study, the difference is approximately 29% greater production when the LHD is operating with 100% availability. It is indicated that when the LHD failures are included, the equipment utilization for the other types of machinery would then be lower due to longer periods of time waiting for the LHD to become available. To avoid low equipment utilization, multi-drifting, multiple stoping operations may be considered as an alternative to increase the rate of equipment utilization.This paper demonstrates how Simul8 can be applied to model hard rock mining operations and assesses the impact of equipment failures on the development cycle in an underground mine. The applied methodology considers the merging of a reliability assessment modelling technique, based on genetic algorithms, with a discrete event simulation model. By isolating LHD mechanical failures as a point of research, it is shown that there is vast potential for further analysis through case studies involving failures of other equipment. This study also shows that low-cost, two-dimensional simulation tools, such as Simul8, can provide an effective analysis of mine production systems.
Keywords: Simulation, Underground, Hard rock, Load haul dump (LHD), Mean time between failures (MTBF)