|
Titel |
Sampling and analysis of chemical element concentration distribution in rock units and orebodies |
VerfasserIn |
F. P. Agterberg |
Medientyp |
Artikel
|
Sprache |
Englisch
|
ISSN |
1023-5809
|
Digitales Dokument |
URL |
Erschienen |
In: Nonlinear Processes in Geophysics ; 19, no. 1 ; Nr. 19, no. 1 (2012-01-05), S.23-44 |
Datensatznummer |
250014162
|
Publikation (Nr.) |
copernicus.org/npg-19-23-2012.pdf |
|
|
|
Zusammenfassung |
Existing sampling techniques applied within known orebodies,
such as sampling along mining drifts, yield element concentration values for
larger blocks of ore if they are extended into their surroundings. The
resulting average concentration values have relatively small "extension
variance". These techniques can be used for multifractal modeling as well
as ore reserve estimation approaches. Geometric probability theory can aid
in local spatial covariance modeling. It provides information about increase
of variability of element concentration over short distances exceeding
microscopic scale. In general, the local clustering of ore crystals results
in small-scale variability known as the "nugget effect". Parameters to
characterize spatial covariance estimated from ore samples subjected to
chemical analysis for ore reserve estimation may not be valid at local scale
because of the nugget effect. The novel method of local singularity mapping
applied within orebodies provides new insights into the nature of the nugget
effect. Within the Pulacayo orebody, Bolivia, local singularity for zinc is
linearly related with logarithmically transformed concentration value. If
there is a nugget effect, moving averages resulting from covariance models
or estimated by other methods that have a smoothing effect, such as kriging,
can be improved by incorporating local singularities indicating local
element enrichment or depletion. Although there have been many successful
applications of the multifractal binomial/p model, its application within the
Pulacayo orebody results in inconsistencies, indicating some shortcomings of
this relatively simple approach. Local singularity analysis and universal
multifractal modeling are two promising new approaches to improve upon
results obtained by commonly used geostatistical techniques and use of the
binomial/p model. All methods in this paper are illustrated using a single
example (118 Pulacayo zinc values), and several techniques are applied to
other orebody datasets (Whalesback copper deposit, Witwatersrand goldfields
and Black Cargo titanium deposit). Additionally, it is discussed that nugget
effects exist in a binary series of alternating mostly gneiss and metabasite
previously derived from KTB borehole velocity and lithology logs, and within
a series of 2796 copper concentration values from this same drill-hole. |
|
|
Teil von |
|
|
|
|
|
|