Page 177 - Introduction to Mineral Exploration

160 C.J. MOON

and it is not desirable to operate two number- 65°C so that volatile elements such as mercury
ing systems, so some arrangement should be are not lost. Drying is generally followed by
made with the laboratories to randomize the gentle disaggregation and sieving to obtain
samples within an analytical batch. The results the desired size fraction. Care should be taken
can then easily be re-sorted by computer. to avoid the use of metallic materials and to
When laboratory results become available avoid carryover from highly mineralized to
the data should be plotted batch by batch to background samples. Preparation of rocks and
examine the within- and between-batch effects. vegetation is usually carried out in the labor-
The effects observable within a batch are pre- atory and care should be taken in the selec-
cision and systematic instrumental drift. Pre- tion of crushing materials. For example, in a
cision should be monitored by calculating the rock geochemical program a company search-
average and difference between duplicates on a ing for volcanic-associated massive sulfides
chart such as is detailed by Thompson (1982). found manganese anomalies associated with
The precision for most applications should be a hard amphibolite. They were encouraged by
less than about 20%, although older methods this and took it as a sign of exhalative activity.
may reach 50%. If precision is worse than anti- Unfortunately further work showed that the
cipated then the batch should be re-analyzed. manganese highs were related to pieces of man-
Systematic drift can be monitored from the ganese steel breaking off the jaw crusher and
reference materials within a batch and the contaminating the amphibolite samples. Other
reagent blanks. Between-batch effects represent less systematic variation can be caused by
the deviation from the expected value of the carryover from high grade samples, for example
reference material and should be monitored by not cleaning small grains of mineralized vein
control charts. Any batch with results from ref- material (e.g. 100,000 ppb Au) will cause signi-
erence materials outside the mean ±2 standard ficant anomalies when mixed with background
deviations should be re-analyzed; commercial (1 ppb Au) rock. Contamination can be elimin-
laboratories normally do this at no further ated by cleaning crushing equipment thor-
charge. Long-term monitoring of drift can show oughly between samples and by checking this
some interesting effects. Coope (1991) was able by analyzing materials such as silica sand.
to demonstrate the disruption caused by the The Bre-X scandal, in which alluvial gold
moving of laboratories in a study of the gold ref- was added to drill pulps before sending the
erence materials used by Newmont Gold Inc. samples for analysis (see section 5.4), empha-
sizes the importance of recording the methods
of sample preparation and controlling access to
8.2.2 Sample collection and preparation
the samples.
Samples should be collected in nonmetallic
containers to avoid contamination. Kraft paper
bags are best suited for sampling soils and 8.2.3 Analytical methods
stream sediments because the bags retain their Most analysis is aimed at the determination of
strength if the samples are wet and the samples the elemental concentrations in a sample and
can be oven dried without removing them from usually of trace metals. At present it is impos-
their bags. Thick gauge plastic or cloth are pre- sible to analyze all elements simultaneously
ferred for rock samples. All samples should be at the required levels, so some compromises
clearly labeled by pens containing nonmetallic have to be made (Fig. 8.3). In exploration for
ink. base metals it is usual to analyze for the ele-
Most sample preparation is carried out in the ments sought, e.g. copper in the case of a copper
field, particularly when it involves the collec- deposit, and as many useful elements as pos-
tion of soils and stream sediments. The aim of sible at a limited extra cost. With modern tech-
the sample preparation is to reduce the bulk of niques it is often possible to get 20–30 extra
the samples and prepare them for shipment. elements, including some that provide little
Soils and stream sediments are generally dried extra information but a lot of extra data for in-
either in the sun, in low temperature ovens, or terpretation. The major methods are as shown
freeze dried; the temperature should be below in Table 8.3 but for detail on the methods used