Page 233 - Introduction to Mineral Exploration

216 M.K.G. WHATELEY & B. SCOTT

TABLE 10.5 Errors in assays on a prepared sample of familiar and the following remarks mainly
2.74 grams per tonne gold. This is an example where apply to work at remote prospects.
the population parameters are known since the As much background information as possible
material was artificially prepared. The results of should be collected from previous publications
the four different sets of five identical samples and reports. A preliminary description of the
(A to E), each analyzed separately at four different exercise serves as a source of stimulation that
laboratories (1 to 4), are given below. (After Springett
1983b.) will help to isolate points on which further
thought is needed. The tasks to be completed
− − − − −1
Laboratory (g t gold) have to be briefly but adequately described.
Logistical factors are important. How much
1 2 3 4 time is available for the exercise? Can the site
be revisited or is this the only opportunity for
Sample sampling? What budget restraints are there? Is
A 5.72 2.74 4.39 2.81 assistance necessary and if so of what type –
B 4.18 2.40 4.01 2.78 surveyors, drivers, cooks, laborers, etc.? Decide
C 5.66 2.06 3.81 2.67 on the objectives and how it is proposed to
D 4.42 3.43 4.01 2.67
E 3.26 2.06 3.81 2.71 achieve them; are they cost-effective?
What is the purpose of the sampling? Is
Mean 4.65 2.54 4.01 2.73 it purely qualitative, where it is necessary
S 2 0.87 0.26 0.04 0.003 to prove the presence or absence of certain
S 0.93 0.51 0.21 0.06
minerals or chemical elements, or is a stat-
1. The range of values at the 95% confidence level (which is istically valid estimate of the grade of (say) a
the mean ±2S) for laboratory 1 is 2.79–6.51 g t . This range base metal prospect required? Is the study pre-
−1
does not contain the population mean (2.74 g t ) and the liminary (i.e. a return visit is possible) or final
−1
variance is high. For laboratory 2 the range is 1.52 to (i.e. no chance of return)? Certainly, before
3.56 g t which does contain the population mean, but the
−1
−1
variance is high. At laboratory 3 the range is 3.59–4.43 g t . carrying out extensive systematic sampling at
This does not contain the population mean, although the various locations, an orientation survey should
variance is acceptable. Laboratory 4 results have a range of be completed to ascertain the main character-
2.61–2.85 g t , which contains the population mean and istics of the prospective target population
−1
has a low variance. – a preliminary mean, standard deviation, geo-
2. Only the results from laboratory 4 are acceptable.
chemical threshold, and probability of error
(see also section 8.3). Finally, check that the
sampling procedure is sound and adequate
and that the projected sample size (weight
samples (e.g. every, say, tenth sample is and number) is large enough for the purpose
repeated) and by introducing artificially pre- but not too large. Carefully assess the cost-
pared reference materials (e.g. every 30th or effectiveness of the proposed exercise.
40th sample). Some laboratories are found to
perform relatively poorly and a rigorous pro-
gram of interlaboratory checking by the use Safety
of blanks, duplicates and prepared materials is Safety aspects are paramount. Before visiting a
important (Reimann 1989). new site or abandoned mines, samplers should
undertake a risk assessment (Grayson 2001).
This will consist of hazard identification and
10.1.7 Sample acquisition
risk assessment, followed by developing pro-
cedures to manage the risk and control the
Before sampling
hazards. Well-designed risk management pro-
Pre-sampling procedures consist of gaining as grams will also bring health aspects into
much familiarity as possible with the geology consideration. Effective assessment of hazards
and possible problems before visiting the loca- required good analytical tools and judgment.
tion to be tested. Obviously if the sampling is at Ask for experienced help when starting out
an operating mine these problems are already in the field. A formalized safety code will be