Geoscience Reference
In-Depth Information
local measure of uncertainty, which often leads to unreason-
able expectations in the resource model. Current practice for
resource classification includes different methods that have
conceptual similarities. Some common ones are:
• Using the number of drill holes and samples near
each block is geometric in nature and easy to explain,
although it frequently tends to be simplistic in its imple-
mentation.
• The kriging variance provides an index of data configura-
tion (Chap. 8), that is, a measure of how well each block
in the model is informed at the time of estimation.
• Using different search radii to estimate blocks in a step-
wise process, while keeping track of when the blocks
get an estimated value. The more information is used to
obtain an estimate, the more certain it will be.
• Deciding according to geologic criteria what drill hole
grid spacing is required for the resource to belong to a cat-
egory (measured, indicated, or inferred), and then search-
ing throughout the deposit for that nominal grid spacing,
thus classifying the different areas of the deposit.
Purely geometric criteria could be supplemented with con-
ventional statistical criteria, that is, defining the expected
grade and a corresponding range of possible grades around
it. For example, measured resources may be defined as those
predicted to be known ± 15 %, 90 % of the time for a volume
equivalent to 3 months production. The model (numerical or
subjective) used to come up with such a statement is most
important to the effectiveness of the classification scheme.
There are shortcomings and pitfalls in the practice of re-
source classification. Many of these can be resolved with
a defendable model of uncertainty based on geostatistical
simulation. Inevitably, the process of classifying resources
depends on the circumstances and conditions of the mining
project being assessed in addition to purely geologic con-
ditions and technical issues. Nevertheless, in all cases, the
classification must be defendable by the professional that
signs off on the resource model.
determine the economic consequences of uncertainty. This
can be further refined by applying existing mine plans to
the simulation models, such that, for a specific mine plan,
an evaluation of the impact of new drilling on recovered re-
serves can be made.
In practice, this type of analysis is based on production
volumes, such as metal sold in a month. If the parameters
that describe metallurgical plant performance are known,
then the uncertainty of the tonnages and grades fed to the
mill can be directly linked to the risk of not achieving the
expected production plan.
The typical question asked by the project development
manager is “how many drill holes do I need?” The answer
to this question requires a definition of the objectives of the
new drilling in terms of uncertainty. Then, the applicable
optimality criteria can be developed and the value of new
drilling can be assessed. This could be expressed in dollar
values, in terms of uncertainty and risk reduction, or in terms
of reduction of cash flow and net present value (NPV) risk.
1.3.10
Medium- and Short-term Models
Medium- and short-term models are auxiliary models used
to improve the local estimation of the long-term resources
model. These are reserve models that are used in an operat-
ing mine for production purposes. Medium- and short-term
models are used to improve the estimation of relatively small
volumes of the deposit. This is useful because mine opera-
tions plan on smaller, shorter-term volumes. The definition
of what is long-, medium-, and short-term varies from one
operation to another; however, common use of the terms
suggest that long-term refers to production periods of a year
or longer, while medium-term refers to three to 6 months
production, and short-term implying 1 month production or
less. The periods chosen will be related to the budget and
forecast cycles of the operation.
At most medium to large mining operations there is a
yearly budget that updates the material movement and corre-
sponding expected cash flows of the original long-term mine
plan. It provides a cash flow prediction for the following year.
Additionally, this budget is itself updated by a short-term
forecast, usually done on a semi-annual, quarterly, or monthly
basis, depending on the characteristics of the operation.
The update of the existing long-term model is accom-
plished by incorporating infill drilling and production
information. Since this work is to be performed within a
production environment, the procedures and methods used
in updating the resource model are constrained by time and
human resources. The definition of the most appropriate and
practical methodology to update the geological and grade
models can become a significant challenge.
1.3.9
Optimal Drill Hole Spacing
Drill hole spacing should be optimal for a given cost-benefit
analysis, which is dependent on the project development
stage. New drill holes must reduce the uncertainty of the
resources to a tolerable, pre-defined level, as required for
project advancement.
A cost-benefit analysis of potential new drill holes re-
quires assessing the benefit of decreasing the uncertainty
of the resource model by a given amount. This amounts to
quantifying the value of new information. If the consequenc-
es of errors in the resource estimates can be defined and
quantified, then it is feasible to use simulated realizations to
