Diavik goes underground

It's as rare and unique as the gems it produces - a diamond mine under a lake, and now going underground.

Situated in a remote part of Canada's Northwest Territories, only 200 kilometres from the Arctic Circle, the Diavik diamond mine (Rio Tinto: 60 per cent) started production in January 2003. In 2006, total output amounted to nearly ten million carats.

Diavik is notable for the high quality of its sparkling white diamonds and the fact that its diamond bearing kimberlite pipes lie under the icy waters of a lake. This geological inconvenience meant that before Rio Tinto could begin mining, it first had to construct large water tight dykes out into Lac de Gras and pump out the water, capturing and releasing the fish as it did so. Only then could open pit mining get under way in the lake bed, safely behind the barriers holding back the water.

Today, open pit operations have been completed at the A154 North pipe while surface mining continues at A154 South. The dyke wall at the adjacent A418 pipe has recently been completed in preparation for open pit mining to begin there. A fourth kimberlite pipe (A21) is now being studied to evaluate its potential for both open pit and underground mining. But the dyke walls impose a limit on the ultimate size of the pits: as the depth of each pit increases, so too does its width. Extending the dykes further into the lake would not be an economic proposition, so if diamond production is to continue at Diavik in the long term, the only solution will be to convert from open pit to underground mining.

To this end, Rio Tinto has over the past two years been conducting extensive feasibility studies of various underground mining options. These studies are due to finish in mid-2007. Two exploratory declines have been excavated. The first is allowing an investigation of underground mining at the A154 (North and South) and A418 pipes.

The aim is to develop a comprehensive resource model and gain a good understanding of the ground conditions and thereby establish what ground support will be needed for the spiral tunnel that will circle downwards round the carrot shaped kimberlite pipes. The second decline is being used to withdraw bulk samples of the A21 pipe, located some three kilometres to the south of the other three pipes, to assess its potential both for open pit and eventually underground mining.

Some US$200 million has so far been committed to the project. This covers the cost of the mining and other equipment brought in via the so-called "ice road" from Yellowknife (open only during a brief winter weather window when the lake ice can support heavy vehicles) and also for the contractors to carry out the development. A decision will be taken towards the middle of 2007 on whether to invest a further US$200 million to complete all the work necessary to convert Diavik to underground mining and whether to build a dyke to allow open pit mining of the A21 pipe.

One of the underground options currently being evaluated is to use roadheaders, a mining machine. Costing US$5 million each, these bulldozer type vehicles have a large revolving drum mounted on an arm at the front. Embedded in the drum are tungsten carbide teeth which cut into the diamond bearing kimberlite. Roadheaders are used quite commonly in the coal business and are well suited to the soft rock conditions found, for example, in the A418 pipe. Tests are under way to determine whether using them allows faster development than the more conventional drill and blast method.

If underground mining is given the full go ahead in mid 2007, underground production will not start in earnest until early 2009. Diavik's production profile will then become progressively more complicated. There will, in due course, be three open pits - though not all will be mined at the same time - and four underground mining areas operating. The last open pit operation is scheduled to be phased out in 2016.

All this will demand careful planning and the meticulous coordination of operations. Furthermore, each ore source has its own characteristics, so a good understanding of the geotechnical parameters of the ore pipes is important, as is feeding the right blend to the processing plant.

Ore from the A154 South pipe, for example, is relatively hard and needs to be blended with softer ore from the other pipes in order to allow the processing plant to operate at its planned capacity of 2.3 million tonnes a year.