The semivariogram is a geostatistical tool to show several geological features which are important in mineral deposit evaluation, such as the continuity, size and shape and it is used for many mining purposes.
Stationary conditions must be satisfied if the semivariogram calculated in one part of the mineralization is to be used in another part, or if the model developed for the entire mineralization is to be accepted for local block evaluation.
Since,the ore recovery varies with sample size, it is important to know the variance (semivariogram) of samples of any size. The practical problem which occur may be formulated in a reverse way; the size is the unknown, then one may make dogmatic statement like if : the size of sample is unknown the recovery cannot be predicted or if the recovery of samples is unknown, no variogram can be computed. Academically these statement are true. In practice one should be much more conciliatory.
The structural tool that is used to recognize the spatial distribution of the chosen variables was the semivariogram.
Experimental semivariogram for Fe%, Mn% and Cl% components of Ghorabi iron ore deposit are constructed from drill holes data.
At first, let us see in the case of good continuity (case 1): the theoretical dispersion variance of point samples can be derived from Krige’s formula as follows:
γ(l) = F(ω) – F(l/a)
where F(ω) is a special function equal to the sill variance of the semivariogram, i.e., F(ω) = C, where C = 10 (%)² in our case and F(l/a).
In the iron Fraction (Fe%), where the semi-variogram has no nugget effect, the length of sample is not effective i.e., any length of sample is accepted.
