Editorial

One of the more pleasant decisions in the mineral industry must be whether to sink a shaft or develop a decline to extract a newly found ore deposit. Both are mining methods that might be considered when the ore is too deep for open pit methods, and the balance between shafts and declines has changed substantially over the last century.

In the late 1990s, I was engaged in informal discussions with colleagues as to whether a shaft or decline was the better option for a recently discovered greenstone gold system that was known to a depth of a few 100  m and appeared to continue farther. I left that discussion accepting that it was an engineering decision and involved a relatively simple calculation that compared the cost of the two options for different parameters that included gold price and discount rates. I recognise now that passing the shaft decision over to engineers and financiers was a major abrogation of duty and that we should have been offering substantial geological input into that decision. A quarter of a century later that mine is now exploring at 3 km depth with on-going success.

The shifting fortunes of different mining methods are well illustrated by the largest goldfield in the greenstone belts of the Yilgarn Craton of Western Australia. At the start of the 1980s, there were 100 shafts within the Golden Mile at Kalgoorlie and most reached depths of a few hundred metres to extract gold from multiple auriferous shear zones. These shafts were a legacy of developments decades previously. By 2000, those shafts had disappeared and been replaced by one very large open pit. By the standards of some of the World's other great goldfields, the 800  m maximum depth in the Golden Mile at Kalgoorlie in 1980 was hardly extreme, and it was up to 1 km shallower than other great gold deposits such as Macassa in Canada, Kolar in India, Morro Velho in Brazil and Ashanti in Ghana. Kalgoorlie was discovered in 1893 and its development pathway of shafts followed by the large open pit of today reflects timing, grade, geometry, the economy, gold price, availability of skilled labour and many other factors hardly considered at the time of discovery.

The nature of gold occurrences globally is that thousands have been discovered by early prospectors and after a little digging most came to very little. However, a small number of these have become substantial deposits as they continued to great depth. Newer gold discoveries of the 1980s have been developed in a quite different way to the Kalgoorlie of 1893 with fast and cheap access via an open pit followed by a decline. The latter allows access by driving a tunnel off the bottom of the open pit once that pit had reached its maximum economic depth. With an expected mine life of 5–10 years, the option of an open pit followed by a decline was sensible for the time. As some of those declines have gone deeper than expected and new reserves have been discovered through brownfields exploration below and ahead of mining the obvious choice of a decline is being questioned.

Some of the gold discoveries of the 1980s have already progressed through the phases of open pit and then decline to be well over 1 km deep and some are over 1.5 km. With a gradient of a decline being 1-in-7, it can require 10 km of driving from the surface to the working areas, and this travel is done along a small tunnel at slow speed and giving way to larger ore trucks ascending and descending at even slower speed. The journey down the decline for the geologist going to inspect and advise on some ore face can be 1 hour and the return journey will be no faster. The two-way trip for the ore trucks bringing material to the surface is even slower. Left to the current mining method of a decline, the probability arises that mines become less economic or even uneconomic around 2 km depth. At this time, management and shareholders might wish that their 2-km-deep mine already had been using a shaft that could easily be extended a farther many hundred metres in depth. However, with a marginally economic decline reaching to 2 km depth, will there be any corporate appetite or funds to start a shaft at this stage? Should the shaft option have been commenced much earlier and if so at what depth? Do modern methods such as raise boring mean that the relative cost of a shaft and of a decline from 1 to 2 km depth has changed over time so that they might be similar except that the shaft positions the mine much better for further advancement? Or should all ideas of a shaft be bypassed and instead there be some acceptance that the high-grade ore deeper than 2 km is sterilised given currently available mining options?

It is in this context of the decline travel time and costs continually increasing that companies using a decline might look with considerable envy at the large shaft systems such as found on the Witwatersrand gold mines in South Africa. In these goldfields, there are single-drop shafts descending at 60 km h⁻¹ and reaching 2 km depth in minutes. From the 1960s, it was an understood procedure that once these Witwatersrand mines needed to go beyond 2 km depth an internal shaft would be built nearby that started at 2 km depth and ended at over 3 km depth. Today at the Moab Khotsong mine near Klerksdorp in South Africa, there is a single lift to 3 km depth which takes less than 5 minutes. No one is saying that these very deep shaft systems are easy or cheap to build but with many years of experience the industry was able to advance major shaft systems by 10  m vertically per day in exceptional cases.

There is more to be learned from the Witwatersrand mines, their 2–4 km deep shafts and their geology. Many of these shafts were built from the 1950s to 1970s following the discoveries of the Carletonville (1930s), Welkom (1940s) and Evander (1951) goldfields. Some of those shafts were sunk to 2 km depth based upon a small number of drill holes. The decision to proceed with a shaft was not because the gold grades were extraordinary, though some were, but primarily because there was supreme confidence that there was geological continuity between holes by matching the planar mineralised unconformity surfaces, i.e. Carbon Leader Reef, Basal Reef, Kimberley Reef and many more. Some of these shafts were exceptionally successful, whereas others were regrettable in hindsight and absorbed into the major operations surrounding them. This period of confidence and optimism on the Witwatersrand goldfields changed from the 1980s in the midst of an on-going 60-year period during which no new goldfield was discovered. Confidence in geological models and predictions waned, there was reduced production from 1000 tpa to 150 tpa of gold which virtually no one predicted, and the building of new shafts almost ceased.

Fast forward to today and we are looking at a real example of a multi-million-ounce goldfield in Australia that is mining high-grade ore by decline methods at 700  m depth. If anything, the ore is improving with depth. What is the appropriate development method to mine deeper than the present 700  m depth? Financial and engineering models that use the projected price of steel, digging, labour and electrical work to name a few items coupled with an informed discount rate and gold price estimate, should play a role here. Also, the availability of shaft sinking expertise must be considered. However, the decision of shaft or decline is not to be left solely to engineers and financial planners as nobody wants a shaft that has no ore to haul, and neither do future generations want to see a mining operation cease at 2 km depth and their jobs disappear because it has become too slow and expensive to drive vehicles up and down its decline.

One of the most important factors in all this decision making is the geometry of the orebody. If geologists are prepared only to commit that the high-grade orebody today might stretch to 750  m depth, then there is little reason to discuss a shaft option, and continuing the decline is optimal. Alternatively, if the geological indicators suggest that this could be one of the world's great orebodies with substantial depth extension then the discussion of the shaft option becomes a reality. A decision needs to be made, and it will need to be made without exact information about the tonnes and grade of gold nor about the price of steel or labour.

Geological decisions about mining methods will utilise the descriptive knowledge of our tens of thousands of mines, and even more important is being able to recognise which orebodies are likely to continue to great depth because of high gold grade and multiple ore-hosting structures. In deciding about mining methods and whether a shaft is viable, engineering costs do play a part and so do financial models. That the shaft or decline question is important is not difficult to show but providing helpful answers around this topic is not as easy. When is the right time to be thinking of a shaft remembering that in the nineteenth century the answer was almost as soon as digging started? Are there ways to modularise shaft building to reduce the initial cost outlay but to allow up-scaling if ore continues? Is the shaft versus decline discussion too narrow as there may be alternative options such as hydraulic hoisting of crushed ore? And then there are tensions between shorter term versus longer term timeframes such as how does a company rationalise the financial comfort of a decline that will easily suffice for the term of a Board and a CEO against the interests of communities and governments who might have a longer timeframe and do not want to see one vertical kilometre of ore beneath the bottom of a decline left behind? Boards may have made some commitment to shareholders for rapid progress from exploration to production; this is a difficult time for all but a few companies to be promoting the importance of optimal development over two to three decades.

Whether it is a greenstone gold deposit like Kalgoorlie, unconformity-related system like the Witwatersrand, or our quartz-vein type gold deposit in a slate belt, many of the important decisions affecting the future of mining operations need expert geological input. We might not be able to provide exact grade and Reserves for 1 km ahead even with drilling but the better this can be done the more robust the mine planning decisions become for the benefit of next year and following decades.