244
5 CONCLUSIONS
In the mining industry data can be obtained from different sampling methods, with different support size or where the assays were determined from different technical analytics. In
order to assess the different data sets in the same estimation process it is crucial to perform a
statistical and geostatistical analysis, at global and local scale. This will determine if the different data sets can be merge directly or not. This paper proposes a methodology used when
the different data sets cannot be merged directly. The KVEM combined with TB conditional
simulation are two solutions to estimate resources, associating a measurement error to the
unreliable or inaccurate data set. The errors are estimated by cokriging from the same element of each involved data set. TB conditional simulations allows going a step further to
perform a risk analysis and to calculate the probability that the mean grade of a group of
Figure  11. Overlays of Mean grade versus cutoff curves for Corvo orebody for all 100 conditional
simulations.
Figure 12. Overlays of total tonnage versus and cutoff curves for Corvo orebody for all 100 conditional simulations.
11
11
1 0
1 0
w
9
9
:;::
"(j
ID
Ill
~
' "
"'
[.!)
~
8
8
GJ
"
Ill
~
w
0.
:;:
ID
7
7
6
6
Corvo- Cu 100 simus
Corvo - Cu Si mu 7
corvo- c u Simu 11
Corvo-Cu Simu 1 5
4
5
6
7
8
9
corvo-cu Simu 28
Cutoff
Corvo-Cu Simu 46
2
3
8 0
70
I
: .:. : .1.,
I
_: :_:1_:.
7 0
6 0
I
I
. '. • • I •
6 0
w
:.1.: : ... 1.:
01
I
I
Ill
50
~
o.l:.·
1~
0
I
I
E-<
4 0
>-3
0
50
rt"
~
>4 0
>-3
0
.....
Ill
3 0
_,_,
"'
30
"'
~
0
lQ
E-<
ID
2 0
20
Corvo-Cu 1 00 sirnus
Corvo-Cu Simu 7
corvo- c u Si mu 11
Corvo-Cu Simu 1 5
c orvo-cu Simu 28
10
0
9
. . . I . .
" "•" 'l'"•"
3
4
5
6
7
8
Cutoff
Corvo- Cu Simu 46