Geoscience Reference
In-Depth Information
depend on the study objectives, the available data and the pro-
fessional time available to complete the study.
Resource estimates should be complemented with a mea-
sure of uncertainty. All numerical models have multiple
significant sources of uncertainty including the data, the
geologic interpretation, and the grade modeling. A statement
quantifying the uncertainty in the predicted variables is re-
quired for good and best practices.
These four main subjects are covered in 14 chapters. Each
chapter concludes with an exercise that summarizes the key
points and helps interested readers test their understanding
of the material presented. No solutions to the exercises are
provided.
Perceived limitations and risk areas should be documented.
The process of model validation and reconciliation is itera-
tive. The calibration of a recoverable resource model against
production, if available, is particularly important to ensure
future predictions are as accurate as possible. Proper and de-
tailed documentation is required for each step. An audit trail
must be created during the entire resource estimation process
to allow a third party to review the modeling work. Transpar-
ency and the ability to allow for peer-reviews are essential
components of the work.
1.3
Critical Aspects
The estimation of resources and reserves requires detailed
consideration of a number of critical issues. Like a chain,
they are linked such that the quality of the overall resource
estimate will be equal to the quality of the weakest link; any
one of them failing will result in an unacceptable resource
estimate. Resource estimators must deal with these issues on
a daily basis.
The quality of the mineral resource estimate depends
firstly on the available data and the geological complexity of
the deposit; however, the resource estimate is also strongly
dependent on the overall technical skills and experience of
the mine staff, how the problems encountered are solved, the
level of attention to detail at every stage, the open disclosure
of basic assumptions along with their justifications, and the
quality of the documentation for each step.
The emphasis on documenting every aspect of the work is
stressed throughout this topic because it is the final and, pos-
sibly, the most important link in the chain. Justification and
documentation of every important decision serves as quality
control of the work, because it forces detailed internal re-
views. In addition, it also facilitates third-party reviews and
audits, which are a common requirement in industry. Some
basic issues to be dealt with in resource estimation are brief-
ly discussed next.
1.2
Scope of Resource Modeling
The collection, gathering, and initial analysis of data are the
first steps in mineral resource modeling. Sufficient qual-
ity controls and safeguards are required to achieve an ad-
equate degree of confidence in the data. The overall process
of Quality Assurance and Quality Control (QA/QC) should
encompass field practices, sampling, assaying, and data
management. This is necessary to ensure confidence in the
resource model.
The data are subset within different geological domains.
These domains may be based on a variety of geological con-
trols such as structure, mineralogy, alteration and lithology.
Categorical variable models are constructed to subdivide the
data and focus analysis in different regions of the subsurface.
Domains are commonly assigned to a gridded block model.
The block model must have sufficient resolution to represent
the geological variations and provide the required resolution
for engineering design. Of course, the number of blocks must
not be too large. At the time of writing this topic, it is common
to use 1 to 30 million blocks. Larger models are possible, but
they require more computer resources and managing multiple
realizations of many variables becomes time consuming.
Statistical analyses of the available data are required be-
fore decisions can be made about geological domains. Min-
eralization controls interact to control the spatial distribution
of grades. Compositing the original data values is common
practice. This is done partly to homogenize the support of
the data used in estimation, but also to reduce the variability
of the dataset. Further statistical analyses are performed to
understand and visualize the data distributions and to define
the most appropriate form of estimation.
After defining the block model geometry and geological
domains, it is necessary to assign grades. The choice of an
estimation method and the formulation of plans for grade in-
terpolation are described in later chapters. Special consider-
ations required for simulation are also discussed.
Each step in mineral resource estimation requires as-
sumptions and decisions that should be explicitly stated.
1.3.1
Data Assembly and Data Quality
The quality of the resource estimate is directly dependent on
the quality of the data gathering and handling procedures.
Many different technical issues affect the overall quality of
the data. Some important ones are mentioned here.
The concept of data quality is used in a pragmatic way.
The concept is that data (samples) from a certain volume
will be collected and used to predict tonnages and grades of
the elements of interest. Decisions are made based on geo-
logical knowledge and statistical analyses applied in con-
junction with other technical information. Therefore, the
numerical basis for the analyses has to be of good quality to
provide for sound decision-making. This is particularly im-
