There have been a number of interesting developments in the field of computer modeling of ore deposits that involve complicated applied mathematics and the application of other magic wands to determine the spatial orientation of deposits and to make an assessment of the variability of the mineral content. When you think of it, it is the variability that often embarrasses mining companies and so it is important to know something about such things. 

For instance, let’s take a hypothetical case of an interesting gold play in, let’s say… Busang. If this deposit is drilled on 1 kilometer centers and the type and grade of gold mineralization is consistent in each drill hole, then we would be justified in becoming very excited and quoting reserve figures of tens of millions of ounces, right? If you have understood anything about this website you would agree that summary execution would be too good for purveyors of such wild extrapolation. And you would feel justified in taking part in the firing squad if you lost your life savings on such a project. But if you invested money in such a project after reading the section on Reserves you should also feel embarrassed because you would know that there has to be a feasibility study to substantiate the use of the word “reserve” and nothing can be considered consistent over 1,000 meters. Variability of mineralization is a huge issue - especially in the case of gold deposits. When grades are quoted in terms of “grams per tonne” they are being referenced in terms of parts per million. So a gold grade of 9 grams per tonne is really saying there are 9 parts gold per million parts total material. These are very little bits indeed and in many mines, drill holes on 25 foot centers is not enough to understand the variability of the deposit. So be very careful with mineralization variability as it can be quite tricky. It is also important to know that many deposits can be very consistent over large areas but highly variable over short distances. This is the case in bulk deposits like oil sands and lateritic nickel, for example. In laterite deposits, nickel grade can change by 50 percent over centimeters but be very consistent, on average, over several tens of meters. But it takes a lot of drilling to substantiate that this is the case - there are no shortcuts.

This discussion begs the question of what are the practical implications of resource variability? First of all, variability in mineral occurrence is a bit like variability in the boss’ mood. The variability is not the problem. It is the accurate prediction of variability that is the issue. If you can accurately predict the boss’ mood when you arrive at work each day you will make the necessary adjustments to your routine to maximize the relationship, be promoted and one day fire the SOB. Likewise in a mine if the variability of the mineralization is understood the mine plan and processing strategies can be adjusted appropriately and the maximum amount of good bits extracted from the rock. This is a good thing because it is the good bits that make all the money. There are different strategies for dealing with variability in the mineralization and these strategies should be discussed in the feasibility study. If they are not then you have a bunch of zinger questions for your brother-in-law. 

In some cases it is best to build a big stockpile of ore and blend it to get a consistent feed type going to the plant. In other cases it is possible to make quick changes in the plant to react to changes in the ore feed type. As a general rule, the larger the mine (higher production rate) the greater the tendency to blend the feed and the smaller the mine (lower production rate) the greater the tendency to “campaign” feed specific mineral types. It all depends on the size of blending stockpiles required and the complexity of the process plant. Stockpiles, which are ugly and unromantic, can be hugely expensive and are to be avoided if possible. For example, many of the large iron ore mines in Brazil have blending stockpiles that contain up to two weeks of process plant feed material. That means the mine works for two weeks without getting paid for all the manpower, diesel fuel, electricity and Windex used during those two weeks. 

So in making your mining investment decisions learn something about the ore grade variability. And a good tool used to understand ore grade variability is called geostatistics.