Metal mining is beginning a significant transition from large-scale open pit to large-scale underground operations as pittable reserves become exhausted. It is interesting to note that pit depths for recent and forecast transition projects are typically more than 800m. This has significant implications to transitioning, including higher stress levels, greater development required to establish underground production (i.e., longer time to first ore, significant value implications due to time cost of capital), the presence of the pit which can act as a funnel for water and the major issue of a production gap, as the switch is made to underground. Therefore careful decisions have to be made to ensure a smooth transition process.
Transitioning to underground bulk mining raises several questions:
- Method selection: generally bulk methods will consist of SLC/block or panel caves, though SLOS may be used as an interim method to reduce any potential production gap.
- Geotechnical hazards and their management.
- Minimum thickness of the surface crown pillar required to allow simultaneous surface and underground operations.
- How long may open pit and underground mining be carried on simultaneously.
- Early failure of the surface crown pillar could jeopardise continuing open pit and underground operations.
- Optimum height of the ore column to be caved.
- Will the cave propagate upwards through the full block height.
- The nature and extent of subsidence, its’ timing, and its’ effect on surface and underground infrastructure
- If caving propagation is arrested and a void is created, sudden failure of the cave back could cause an air blast, with significant, and possibly fatal, consequences (as occurred at Northparkes).
- Surface subsidence could influence open pit operations and surface and underground infrastructure
- The open pit will act as a catch basin for rainfall, increasing the risk of water inflows or mud rushes into the underground
- Caving in deep, relatively massive rock masses could induce seismicity and strain bursts
Mining Methods
Notwithstanding the above questions, there is substantial precedent for underground mining at rates of 30,000 to 50,000tpd, see next figure (note: this does not include the recent experience at “super” caves such as Cadia and Grasberg). Broadly, entry methods such as Drift and Fill typically fall in the range 100 to 8000 tpd, SLOS methods fit in the range 1,000 to 20,000tpd (though there are some exceptions). Over about 15,000tpd the method of choice tends to be SLC, Block or Panel Caving.
The rock mass characteristics will have a significant impact on mining method, layouts, sequences, production performance and costs.
Typically, only the bulk methods are considered for transitioning to underground due to production rates that can be achieved and the enhanced safety that arises due to non-entry mining. The principal bulk mining methods showing relative OPEX are illustrated the next figure.
Block/Panel Caving
The estimate of caveability is sensitive to inputs on rock mass characteristics, intact rock strength being a particularly important factor. The range of strengths, the juxtaposition of weaker and stronger rocks and the presence and distribution of faulting will influence caveability and overall production performance. The sensitivity of HR (hydraulic radius is defined as the ratio of the cave footprint area divided by the cave plan view perimeter) for sustained caving to rock strength and rock mass quality is critical when the margins between the required HR and available HR are small. Another factor that can strongly impact caveability is the stress regime. Good Orebody Knowledge is critical before a decision on the production level and sequence is made.
The industry trend is for deeper, higher lift caves, driven by the need to maintain margins to compensate for falling grades, deeper mining, and more onerous support requirements.
A high lift cave can be financially attractive due to the elimination of a production level(s) and the associated infrastructure. However, more time is required for development and resource may be lost due orebody geometry. Thus, there is a trade-off between recovery (revenue), lift height and footprint CAPEX and OPEX. The elevation of the production footprint is based on assessment of value and production risk
Orebody Knowledge
Orebody Knowledge (OK) is typically not given enough attention in transition investigations as it is often assumed that information from the pit is broadly adequate. Thus, there is often limited geological and geotechnical information for the areas proposed for underground mining. There must be a focus on Orebody Knowledge for any Transition and Caving project.
- Comprehensive investigations will be required to establish the distribution of rock quality and strength within the orebody and the country rock. It is important to determine the rock mass strength (a function of rock quality and intact strength). An increase in rock quality/strength with depth (as is common in other orebodies) will impact the decision in locating production level elevation if block caving is adopted.
- Drilling will be required to better determine country rock and orebody characteristics, including orebody geometry, contact characteristics, detailed information on the distribution of rock strength and other characteristics (triaxial and UCS testing, etc.). Point load testing is also a simple inexpensive technique to assist in domaining of the rock mass.
- A comprehensive understanding of geological and structural/stress setting is a necessity. This will require the appropriate geology and structural studies to be undertaken.
- Similarly, hydrogeologic studies will be required to determine groundwater characteristics and water handling strategies.
- Most new caving operations use early development or specific “Exploration” drives to obtain additional Orebody Knowledge to reduce uncertainty (de-risk). Examples included Argyle, Palabora, Resolution, Grasberg, Oyu Tolgoi.
The next figure provides target levels of data confidence for each study stage. Orebody knowledge is required for the entire orebody and the surrounding country rock that may be impacted by underground operations.
Rock Mass Characterization
Rock mass characterization methods are essential to block cave design. Empirical approaches that are typically referenced for block cave applications include the Mining Rock Mass Rating (MRMR), Barton’s Tunneling Quality Index (Q) and the Mathews stability number, N.
The MRMR classification method was introduced for cave mining applications and has been related to estimations for caveability, subsidence angles, failure zone, fragmentation, undercutting sequence and support design (Laubscher, 2011). The method relies on a variation of the Rock Mass Rating (RMR) value, as defined by Bieniawski, and adjusts it to account for in situ and induced stresses, join orientations, stress changes and the effects of blasting, water and weathering (Hoek, 2007). A limitation of empirical rock mass characterization methods, such as MRMR, is the sensitivity of their application to the judgement of the practitioner. Additionally, the relevance of MRMR for cave design is limited to the range of applications that are included in the empirical dataset. Future caving projects are pushing the mining method to new extremes in terms of its application, thereby prompting industry to supplement empirical methods with modelling tools.
Discrete Fracture Network (DFN) methods are one example of the modelling tools that are used for rock mass characterization. DFN uses statistical distributions to characterize each discontinuity set within a structural domain; variables included in the characterization include orientation, persistence and spatial location of discontinuities (Elmo et al., 2010). A major outcome of DFN modeling is the spatial distribution of fracture intensity, referred to as P32 and expressed in units of m2/m3 (fracture area/unit volume). DFN-based methods are considered to be particularly advantageous as they rely on quantifiable fracture data that is collected from field analysis of the rock mass. Since these fracture properties are preserved during the modeling process, the heterogeneity of the fracture system is better defined, resulting in an appropriate method to describing local scale problems (Rogers et al., 2010).