Key challenges integrating caving geomechanics simulation into mine planning processes start with the considerable effort to build realistic models and to map production schedules and cave back geometries into the simulation. Currently, calibrating parameters for the complex failure mechanisms that define the interface (cave back, possible airgap, muck pile) between solid and flow domain can be extremely time consuming. This also often requires a high level of expertise in modelling. This work investigates the numerical efficiency of automated mesh and model building strategies and advantages of using high-performance computing on regular fine grids for non-linear finite element simulation. This allows direct mapping of cellular automaton results, and in return, predictions of rock mass failure without loss of accuracy at a higher frequency, maximising the use of information available from calibrated flow models for production scheduling. An important goal for such models must be the ability to simulate cave growth in complex geological settings and replicate realistic behaviour for relevant benchmark problems that reflect industry experience in block caving. These automated processes will not just accelerate the cave modelling processes and reduce manual processing time but also allow use of the full simulation cycle in case studies, sensitivity analysis and optimisation in an environment of uncertainty and constant changes to the available data. Integrated simulation and optimisation tools significantly improve understanding of realistic geomechanics behaviour driven by the inherent characteristics of the rock mass and structural geological setting, by the extraction strategy and by other engineering decisions (interaction with underground infrastructure). This greater level of understanding reflects in key performance indicators related to safety, revenue maximisation (strategy on how best to exploit the mineral resource), and operation excellence (productivity).

Palabora Mining Company commenced mining of the Lift 1 block cave in 2001. Lift 1 is now nearing completion, and this paper will discuss the challenges around the reconciliation and forecasting of grades during the life of Lift 1. Particular emphasis will be placed on the impact of the open pit failure which took place in 2004 and its impact on the grades later on in Lift 1. The paper will look at the different caving mechanisms which have contributed to, and impacted on, the recovered metal from the block cave. It will also look at how the cave management was impacted by the coarse fragmentation of the cave, which was a first worldwide at the time, and the interaction between practical mining constraints and idealised caving strategies. In the latter stages of the mine, some particularly surprising grade behaviour was experienced, and plausible explanations for this behaviour will be presented relating to the open pit failure material and cave back shapes. Concluding remarks will discuss how the experiences learned from Palabora Lift 1 can potentially benefit Palabora Lift 2 and other future caving operations.

This paper presents an overview of a new application specifically developed for the planning and scheduling of sublevel caving (SLC) projects and operations. Conventional mine planning software tools are not well suited to the evaluation of this type of deposit due to their inability to model the dilution behaviour. The overall purpose of this application is to enable a user to schedule a deposit effectively and efficiently for the feasibility of mining using the SLC method. Much of the logic in the module is similar in concept to that used in Gemcom's PCBC software (Diering, 2000) for block cave operations. Tools for layout generation of tunnels and rings provide a basis for the subsequent generation of production schedules. Ore recovery and dilution modelling are included, but rely strongly on reasonable calibration against historic mining or other models such as REBOP (Itasca Consulting Group, 2000) (REBOP is a code for rapid simulation of the flow of fragmented rock in cave mining operations based on PFC, Particle Flow). Some examples are provided showing either small test cases of a few rings or larger problems of several thousand rings. https://papers.acg.uwa.edu.au/p/1002_15_Villa/ A new mine planning tool for sublevel caving mines D. Villa and T. Diering 238

Material mixing and its impact on the block cave operations is one of the main aspects for evaluating and operating any caving project. Initial modelling with limited geotechnical data and calibration using the actual data are the two main phases to build the flow model. Challenges on the way are inevitable; how to predict the flow in the first place, and how to calibrate even when such data is available. In order to achieve realistic results, the right tools are a necessity to be able to mimic the actual cave. Vertical Mixing (VM), Template Mixing (TM), and CA3D tools in GEOVIA PCBC have been extensively used for more than two decades by many users across the globe; very useful tools but slow and limited on graphics. To overcome these challenges, Marker Mixing (MM) was introduced in 2021; a powerful material flow simulator with unique graphical capabilities and much higher speed compared to other existing tools. The stand-alone feature in MMIX provides the option to run a mixing model without even running a productions schedule. Implementing cave back and other geotechnical and operational constraints are easier and visually assessable, increasing the transparency while making it possible to improve the caving model in a more efficient approach. Markers simulate the material movements from their original location to extraction (from drawpoints) within the cave. Simulation is done step by step with vertical and horizontal movements, toppling, riling, erosion and frozen concepts all captured and visualised in the new MMIX tool. The theory and experience behind MMIX can help define the initial flow model, then in the next step, actual production data can be used to calibrate the parameters. In addition, information gathered from Beacons (smart markers) installed in the cave zone can be visualised, interpreted, and implemented using the MMIX play back tool. This feature can significantly improve the calibration process for achieving more precise models by mimicking the actual flow in the cave. The residual block model can also be generated using MMIX much faster and with better visual checks compared to CA3D. This paper introduces Marker Mixing tool, its features, and capabilities, with comparisons to existing Vertical Mixing, Template Mixing, and CA3D tools in PCBC.

This paper discusses a new software tool developed specifically for use in a sublevel caving project/mine. The tool includes a number of new algorithms or technologies which when combined provide a very efficient way to evaluate a complete sublevel layout in a matter of seconds. The core technologies imbedded in the tool are simulation of material mixing for dilution forecasting using Template Mixing algorithm, optimisation of the draw factors or extraction percentages per ring, and rapid sequencing/scheduling of the resulting ore tonnages. Examples are presented showing how the rapid analysis can allow sensitivity studies on tunnel spacing, mining rate and economics, face shapes and draw or extraction strategies. A real-world example from the De Beers Venetia Underground project is presented. Concluding remarks will discuss how this new tool fits in with the existing industry tools for open pit, open stope, block cave and sublevel cave operations optimisation.