A commonly overlooked factor in the resource estimation and eventual mine planning phase during feasibility studies is rock density. Density is most critical in determining the tonnage and contained metal of a resource and therefore translates directly to the financial feasibility of a deposit. In common scientific terms density is defined as mass per unit volume and expressed as g/cc, g/cm3 or t/m3. Tonnage and contained metal are thus calculated by filling a resource model’s volume with density values, interpolated from each sample taken from drill core. Each cubic meter of the resource model will then have an assigned mass and the grade associated with it will give contained metal. From this we see that density estimation is as important as grade estimation. Assumptions regarding density can be as detrimental to a resource model as assumptions regarding grade…
A well known explorer in North America very recently published a feasibility study conducted for them by a independent consultant. The conclusion of the study was that there is a “medium to high” confidence in the feasibility of the project with a envisioned post-tax IRR of near on 30%. What was interesting about this particular study was the resource estimation, and particularly the way the company had gone about estimating the densities for their ore zone. They claim in their report that the data bank of analytical drill core samples is over 10,000. A couple of years ago they sent a batch of a couple of hundred ore samples for pycnometer density determination. From this they formulated a simple regression which was used to assign densities to the rest of the deposit. The feasibility study further reported that confidence in the resource estimation is medium to high, with a low risk associated to it. Can you imagine producing a resource estimation, particularly deposit tons and ounces, having only analysed a small number of samples for there grade and using a regression model to predict the rest? If density is equally as important as grade, why is this acceptable in the minerals industry?
Another unforeseen implication of a lack of high quality density data in a resource model is the effects on mine planning, particularly in the area of dilution. If the density of the ore zone as well as that of the hanging wall and foot wall lithologies are not equally well understood, a mine has no way of quantifying the effects of dilution. Dilution during mining is a given. The amount of dilution can be minimised by good mine planning and disciplined mining standards, but to count it out of the picture is dangerous. Every percentage dilution directly results in a equal percentage loss in contained metal. Depending on the thickness of the reef/ ore zone being mined, even an over-break of a couple of decimetres can have a large percentage influence. Imagine, for example, an underground thin tabular reef of 1 -2 meter thickness commonly being over-broken by 20 – 30 cm.
Best practice is to analyse density for each sample of ore, hanging and foot wall. Please share your thoughts and current practise around this simple but fundamentally important aspect of resource estimation and feasibility evaluation in the comments below. I would love to hear what norms are practiced in other countries, companies and projects.
If you are interested in further reading on this subject, a study was recently conducted on the effects of the different density determination methods (namely the hydrostatic submersion method vs. the Grabner Minidens gas pycnometer method) on resource estimation and mine planning on South African platinum mines.