Geostatistics for the Mining Industry

- Exploration/Pre-feasibility Stage...
when you need to understand the mineralization to evaluate the potential of the orebody and delimit the inferred ore resource.
Is your drilling grid based on tradition or mineralization?
The basic tool of geostatistics, the variogram, provides you with a 3D image of the spatial behavior of the mineralization. It allows you to identify the directions or maximum grade continuity (anisotropy), or to better understand the short scale grade variability (nugget effect) or the different phases of mineralization (nested structures). As such it can be used as an important aid in the geological understanding of the orebody without ever replacing good geology. The unique nature of an orebody is characterized by the specific form of the variogram from which we can provide you with the optimal drilling strategy for your budget that best respects the mineralization.
Don’t you need a correct resource evaluation?
You need to know the in-situ resource to properly evaluate the potential of an orebody either in its entirety or for medium to long term planning purposes. Overestimating the resource can lead to disastrous financial implications for any mining project while underestimating it means that you will not maximize the project’s potential. Geostatistics and kriging have been proven historically to provide an unbiased resource evaluation because they take the spatial behavior of the mineralization and the stope size into account. The other commonly used estimation methods (based on polygons or inverse distances) do not. Moreover geostatistics has the additional advantage of being able to take the correlation between several variables (via the cokriging technique) into account in the resource evaluation.
- Feasibility Stage...
when you need to know if the orebody is technically and economically viable to re-classify the ore resource as ore reserves.
Have you chosen optimally sized mining blocks?
The choice of the mining block (SMU) size is determinant in the overall performance of a mine. The smaller the block the more selectively you can mine and hence the higher the recovered grade. On the other hand using a larger mining block will lead to a higher tonnage that may be required for the plant or to prolong the life of the mine. Using geostatistical change of support models we can show you how sensitive the reserves are to the mining block size and hence let you choose the block size that best respects the specific requirements of your mining project.
Project Management - Do you know the risks involved?
One reserve evaluation is essential for mine planning but not enough to quantify the risk associated with your project. Risk can only be correctly assessed from a series of possible images of your orebody obtained from advanced geostatistical conditional simulation techniques. We can provide you with such a series of possible images from which you can calculate the whole range of possible reserves and the probability associated with those values. This uncertainty associated with the real unknown reserves must then be incorporated your financial risk analysis of the project. Doesn’t knowing the odds increase your chances of making the right management decision ?
Is your mine plan based on the best reserves evaluation?
From widely spaced drillholes you need an accurate unbiased assessment of those small mining blocks (SMU’s) above the economic cut-off. But no estimation technique allows you to realistically estimate individual mining blocks from widely spaced exploration drillholes. Moreover, then applying a cut-off to such estimated mining blocks leads to biased grade-tonnage estimates, with the corresponding financial repercussions when making investment decisions. Only non-linear geostatistical techniques can provide you with an evaluation from which you can confidently plan production or design pits that maximize the cash flow during the financially critical early years. We have the expertise to choose and apply the most suitable non linear geostatistical technique for your particular orebody, be it vein type, tabular or massive mineralization.
- During Production...
when you know that good grade reconciliation and plant recovery can make or break a mining project.
Isn’t good grade control vital to your project’s success?
For short term mine planning you need to know the grades of individual mining blocks or stopes to process the ore properly and maximize the potential of your project. Only the kriging (or cokriging) technique takes the spatial behavior of the mineralization and the stope size into account when estimating stope grades from all the available data: production and exploration drillhole samples. In this way the geostatistical grade evaluation that we can provide you with means that you actually recover what is predicted. Isn’t that what you want ?
Do you want optimal production sampling ?
You need a production sampling pattern that provides the most information about the blast block or stope. Geostatistical simulation techniques allow us to test different sampling patterns by assuming that one simulation represents the ’ground truth reality’. For each sampling pattern we firstly estimate the different stopes and then compare these values with the ’real’ stope grades. Thus we can provide you with the most suitable one for you sampling budget, that is the pattern that leads to the best correlation between estimated and ’real’ stope grades.
Are you concerned by grade variability at the plant ?
If plant recovery depends on a constant feed grade then you need to understand how the grade varies throughout the orebody. Only conditional simulation techniques allow you to do this because, as opposed to estimation techniques, geostatistical simulations reproduce the observed variability of the mineralization. We will provide you with simulated images of your orebody that allow you to define the blending procedure for an optimal plant recovery or a stocking and retrieving procedure that guarantees that your product meets the control criteria imposed by your clients.
![[Geovariances]](/layout/index_r1_c3.jpg)



![[Geostatistics]](/layout/geostat-efficient.gif)