Held in Canada’s globally important nickel centre of Sudbury, Ontario, this year’s MassMin event has grown again since the last get together in Lulea, Sweden in 2008; reflecting the ongoing development of major new block cave projects by Rio Tinto, Newcrest, Codelco, Freeport and others. In the opening session, attended by IM, Allan Moss, Rio Tinto’s General Manager Underground Technology Centre, had some interesting points to make about the long term success of today’s planned “supercaves”, which will involve a step change in approach and scale.

The industry is looking at tonnages of up to 160,000 t/d, in the case of the Grasberg Block Cave (GBC) in which Rio Tinto is a joint venture partner with Freeport; and at lift heights increasing from 100-200 m up to 500 m or even 800 m. These tonnages, even compared to the highest tonnages mined in caves today such as DOZ at 80,000 t/d, are enormous and will involve the management of thousands of drawpoints. By 2022, about 40-50% of Rio Tinto’s production will be underground, a huge change for a company known more for open pit iron ore, coal and copper mining. Oyu Tolgoi is set to handle 90,000 t/d and Resolution 110,000 t/d, in addition to GBC. As a company, Rio Tinto estimates it will need 10 new production shafts, 12,000 drawpoints and over 500 mobile units; a massive undertaking and investment.

To efficiently mine caves on this scale, Moss argues that a new approach to measurement and calibration of cave models is needed. A better understanding of the caves is crucial to avoid unwanted surprises such as clay migrating to drawpoints or major surface collapses related to caving, as well as unexpected behaviour in the block in terms of fragmentation. But planning in caving on this scale and timeframe is difficult: there are less than 15 active block caves in the world, with a lot of anecdotal knowledge rather than real data. Time is a big factor with these new supercaves. Support/installation can be as much as 50% of development time, and 30% of construction time. For example, the GBC access tunnel work started 2004, but the mine not going into production till 2022 – as such staff that started the project will likely have retired by the time it is producing with the resultant loss of experience. If the industry is going to achieve the holy grail of an “ore factory” through supercaves, it has got to achieve a better reliability of grades coming out of the cave, and caving models should be better able to adapt to ongoing measurements, not just relying on a reactive approach. Block cave predictive plans and designs often do not agree with what is seen underground – as an example rilling (lateral movement) has been seen to be much more common than thought. The industry needs to constantly measure and refine process based on measurements. As Moss states, “what you don’t measure, you can’t manage.” The mine has to collect data underground that gives models and modellers the data they need to calibrate their models and make progress from the lessons learned. And it is important to remember that the people who design the mines are not the same people that operate them.

It is a paradigm shift from the world of open pit mining. Blocks of ore in a cave move and fragment, whereas in a pit they stay put. The block of ore also may not follow the most direct route to drawpoints. In a pit it is a straight fixed geometry exercise, but in a cave it is anything but. It is crucial to accurately describe the rock mass in the cave but this is easier said than done when you are having to characterise huge areas of rock, where different sets of joints and veinlet structures all have an effect on caving behaviour. Proper measurement will help to reduce dilution and increase recoery while reducing drawpoint damage. Moss concluded: Performance forecasts have largely been based on sets of rules and empiricisms established through trial and error at previous operations….substantial effort is required to obtain the data that describes cave performance before modelling techniques can be further advanced. This entails a return to quality measurements of key performance data such as cave back propagation, fragmentation size, material movement in the draw column, draw control and drawpoint availability. Such data must be fully integrated with operational information, and to be effective as a proactive cave management tool, it must be available in real time and well communicated to all the stakeholders.”