190
Figure 5. Uncertainty (defined as the P90–P10) off tonnage mass with cut-off 46%.
tatively states that the uncertainty relative to the mineral resources is inversely proportional
to the sampling density (Figs. 4, 5).
Considering the Al2O3 cut-off grade as 46% and adopting the difference between the P10
and P90 as the measure of uncertainty at 80% confidence limit. We can obtain the measure of
uncertainty associated with the different data spacings (i) 25 × 25 × 0.5; (ii) 100 × 100 × 0.5; (iii)
200 × 200 × 0.5; (iv) 400 × 400 × 0.5; (v) 800 × 800 × 0.5, and (vi) 1200 × 1200 × 0.5. Respectively,
the obtained uncertainty was 0.20%, 0.75%, 1.63%, 2.63%, 4.50% and 6.50% for the tonnage
above cut-off (Fig. 5) at 80% confidence limit. These results show an almost linear increase in
uncertainty as the data spacing increases. The same way, the uncertainty of grade average curve
with cut-off above 46% relative to those grids, is 0.008%, 0.033%, 0.051%, 0.111%, 0.234% and
0.247% (Fig. 6). As checked, a characteristic of the ore is a low variation in its average grade.
We also evaluated the uncertainty for the selective mining unity (SMU) scale. The uncertainty was measured using the probability to be around 15% of the mean. The mean is the
Figure 6. Uncertainty (defined as the P90–10) of average grade with cut-off 46%.
Figure 7. Boxplot—probability of the SMU to be around the average by 15%.
~
c
·n;
1::
.. ... c
:::l
UNCERTAINTY TONNAGE -CUT-OFF 46%
/
6.50%
/
4 .50%
.. 0.20%
~ 2.63%
__,.. 1.63%
,..... o.75%
25x25
l OOxlOO 200x200 4 00x400 800x800 1200x1200
Grids
UNCERTAINTY AVAREGE GRADE- CUT- OFF 46%
/ ' ·=-"' "
. . /0.111
-
0.033
........ 0.051
.. 0.008
25x25
lOOxl OO 200x200 400x4 00 800x800 1200x1200
Grids
25
100
200
400
800 1200
Grids size (m)