Page 258 - Introduction to Mineral Exploration

10: EVALUATION TECHNIQUES 241

10.4.4 Grade calculation methods Area of influence can also be used as part of
the weighting function. For example, to calcu-
Conventional methods late the weighted average grade of a stratiform
deposit which is to be assessed by polygons of
Often borehole chips or cores, or channels, different sizes, the area of each polygon is used
are sampled on a significantly smaller interval as the weighting function. Thus:
than a bench height in an open pit or a stope
width in an underground mine. It is the geo- ×
=
logist’s responsibility to calculate a composite G ∑ A G
grade or quality value over a predetermined ∑ A
interval from the samples taken from the cores
or channels. In an open pit the predetermined where A = area of influence of each polygon.
interval is often a bench height. In some in-
stances where selective mining is to take place, Computational methods
such as in a coal deposit, the sample compos-
ites are calculated for coal and waste separately In large ore bodies, or deposits which have a
(see Chapter 13 for a more detailed description large amount of data, such as Trinity Silver
of sampling and weighted average quality cal- Mine (see Chapter 16), computers are used
culations). In an underground vein mine, the to manipulate the data. Commercial soft-
composite usually represents the minimum ware packages are available (e.g. MineSight,
mining width. Sometimes a thickness × grade Datamine, see section 9.3) which can per-
accumulation is used to calculate a minimum form all the conventional methods described
mining width (e.g. section 14.6). above. Computers offer speed and repeatabil-
The general formula given above for cal- ity, as well as the chance to vary parameters
culating the weighted average thickness value to undertake sensitivity analyses. In addition
n
for the 1/D block model, also applies when computers can use exactly the same procedures
calculating weighted average grade or qual- as described for geostatistical mineral resource
ity values. The weighting function changes estimation methods. As well as using thick-
though, and three variables, length, specific ness to calculate volume and (with SG) tonn-
gravity (SG), and area are used. For example age, grade or quality can be estimated along
when compositing several core samples for a with the estimation error.
particular bench in a proposed open pit, length
of sample is used. This is particularly import-
ant when the geologist sampling the core has 10.4.5 Software simulation
done so using geological criteria to choose Geostatistical tools may be used for spatial
sample lengths and samples are of different data analysis, to model the heterogeneity of
lengths (e.g. Fig. 10.12). Thus:
mineral deposits, to assess the uncertainty
×
∑ L G (risk) in the estimation process, and to use this
=
G information in the decision process (Deutsch
∑ L 2002). The fundamental problem exploration
geologists have is to create a model from
where G = weighted average grade of each limited sample data. Methods to mitigate
borehole on that bench, G = grade of each core this, such as manual contouring or kriging, are
sample, L = length of sample. described above. Once a project appears to
When samples are to be composited from, have economic potential, engineers become
say, core, and the rocks are of significantly involved in the evaluation process. They may
different densities, then the weighting factor wish to predict grade changes as the deposit is
should have bulk density (BD) included. “mined” in a manner that mimics the proposed
Thus: mining method. This can be achieved using
conditional simulation.
×
×
∑ BD G Normally, sample points will support three
L
=
G
×
∑ BD L estimated (kriged) points. Thus, on a 250 m