Geoscience Reference
In-Depth Information
Contamination of the sample can be in the form of dust
falling on the sample, material left over in the sampling
circuit, abrasion on the sampling equipment, or corrosion.
These errors can either increase or decrease the critical con-
tent for the component of interest and once the sample has
been contaminated there is nothing that can be done remove
the contamination.
Losses may occur as fines in dust, material left in the sam-
pling circuit, or smearing of the critical component. Usually
the component being sampled is a minor portion of the over-
all material and it is significantly different in composition
from the gangue material. This means that losing a specific
fraction of the sample, the fine or coarse fraction, will result
in the sample no longer being representative.
Changes in the chemical composition of the sample must
be avoided. Some potential changes are oxidation, or fixa-
tion of water or carbon dioxide. Alterations in the chemical
nature of the material will impact the assay results and the
expected recovery of the material as it is processed.
One time human errors have the potential to be signifi-
cant and it is very hard to determine the source of the error.
Dropping the samples, mixing of samples, improper label-
ling, poor maintenance of the equipment, contamination, etc,
are mistakes that may introduce error. Care and attention to
detail as well as following the sampling protocol will ensure
that errors of this type are minimal.
Fraud and sabotage are the intentional alteration of the
samples used to increase, or decrease, the value of a sample
for personal or corporate gain. Inflation of the mineral con-
tent in the deposit to increase share value has occurred in the
past and as a result the requirements for sampling quality
and documentation of the sampling undertaken have become
more stringent.
There are no universally accepted procedures for QA/QC,
although certain basic steps are always recommended. An
outline of a recommended set of procedures, derived mostly
for gold sampling, will be presented. Good general referenc-
es are Long ( 1999 ) and Roden and Smith ( 2001 ).
5.4.1
General Principles
The main objective of the QA/QC program is to minimize
errors introduced due to sampling, sample preparation, and
sample assaying procedures. The QA/QC program is a con-
tinuous process providing information necessary to correct
defects in the shortest amount of time possible.
Accuracy and precision are two terms used to evaluate
the quality of the information provided by analytical labo-
ratories. Accuracy is a measure of the degree of agreement
of the assayed sample value to the true unknown value of
that sample. An indication of accuracy can only be obtained
through re-assaying samples of known values such as stan-
dards or reference materials.
Precision is a measure of the reproducibility of the sample
value, which can be estimated by re-assaying the same sam-
ple a number of times. Precision and accuracy are different
concepts. A laboratory could have any combination of good
or bad precision and accuracy.
Figure 5.7 illustrates the concepts of accuracy and preci-
sion using the common analogy of the shooter's bullseye.
The left image shows a precisely inaccurate set of three
shots; the center figure shows an accurate but imprecise se-
ries, while the right image shows the case where the shooter
has been both accurate and precise.
All QA/QC check samples sent for analysis to the labora-
tories should be blind, meaning that the laboratory should not
be able to differentiate a check sample from a regular sub-
mission. The internal checks that analytical laboratories often
implement are performed with the technicians being aware of
the fact that they are assaying duplicate samples. These in-
ternal checks, often reported by the laboratories as measures
of their sampling precision and accuracy, should never be
considered as part of a formal QA/QC program. This applies
both to company-owned and external laboratories.
The minimum control unit should be the batch of samples
sent originally to the laboratory. The batch concept derives
from the fact that gold fire assays are done by oven batches.
Typically a set of 40 samples is introduced into the oven. It
is a useful concept that has been extended to other types of
assaying in the context of QA/QC.
Any check sample that fails implies that the complete
batch to which the control sample was incorporated should
be re-assayed. This applies to drill hole samples, but not
necessarily for production samples, since there is no time to
5.4
Sampling Quality Assurance and Quality
Control
The process of mineral resource estimation requires a strict
program of quality assurance and quality control (QA/QC)
to provide confidence about the precision and accuracy of
the drill hole data used for estimation. The QA/QC program
implemented may have somewhat different characteristics if
implemented for mine operations (production sampling), but
it will have the same general objective.
A QA/QC program of an appropriate standard is re-
quired by most international resource reporting standards.
Published resource estimates should be accompanied by a
description and statement of the data quality. This is also a
basic item in any third-party review of resource models, and
may have a significant impact on the overall perception of
the model quality.
 
Search WWH ::




Custom Search