Case Studies - Mineral Resource Estimation - page 267

cant, from the sulfide-rich secondary enrichment blanket
below. The TOS surface can be defined in absolute terms,
i.e., there are no sulfide mineralization above it; or it can
be defined in relative terms, i.e., there are very few sulfide
minerals above it, generally either mixed with oxide miner-
als or with minerals characteristic of the leached zone. This
last definition is the one that was applied at Escondida Norte.
For the 2002 Escondida Norte Resource Model, all min-
eralization units above the TOS surface were modeled using
an indicator Kriging technique. These included:
• Leach cap;
• Oxide mineralization;
• Mixed mineralization, where both oxide and sulfide min-
erals are observed in the core; and,
• Partial leach mineralization, where both sulfide minerals
and evidence of leaching of Cu minerals is observed.
These units are usually no more than a few tens of meters
across in diameter with very erratic, structurally controlled
limits. Even on a 50 m drill hole spacing the accurate inter-
pretation and modeling of these units is difficult. Therefore,
an indicator approach can be used to estimate the likelihood
that each block contains a particular mineralization unit. The
main phases of the work are:
1. Database confirmation : a comparison was made between
the logged information available in the database (from
geologic mapping) and the definition of oxide, mixed,
and sulfide mineralization resulting from the chemical
analyses of Total and Soluble Copper (TCu and SCu).
This comparison is intended to ensure the consistency
between geologic mapping and sample chemical assays.
In the case on Escondida Norte, the agreement was good,
such that the logged mineralization units were partially
used to define mineralization units.
2. Definition of estimation units : the second stage involves
the definition of the estimation domains (Chap. 4). This
process defines the areas that are considered homogenous
for the purposes of estimating oxide and sulfide indicators.
3. EDA and variogram models : all necessary statistics and
variography was performed for each of the indicators de-
fined (Chaps. 2 and 6).
4. Block model definition : a block model covering the neces-
sary volume with an appropriate block size was chosen.
5. Indicator Kriging and final mineralization unit assignment :
the indicators defined on the available drill hole data were
used to krige the probability of each mineralization unit.
6. Model checks and validations : significant checking and
validation was performed before accepting the proposed
probabilistic mineralization model.
The comparison of logged oxidation units from visual obser-
vations and chemical SCu/TCu ratios was performed using
original samples. For each sampled interval in the drill holes
there will be a logged mineralization unit and a correspond-
ing SCu/TCu ratio. The mineralization units were re-defined
combining the assayed values and the logged information

Table 14.14 Definition of mineralization units
Condition CHMIN

Code

Mineralization
unit

LOGMIN = “LEACH” & TCu < 0.2 % 1 Leach

LOGMIN ≠ “LEACH” & TCu < 0.1 % 1 Leach

SCu ≥ 0.2 % & SCu/TCu > 0.5 2 Oxide

CHMIN > 1 & SCu/TCu ≤ 0.15 4 Partial leach

CHMIN ≠ 1,2, or 4, & SCu/TCu > 0.0 5 Mixed

oxide-sulfide

according to the criteria shown in Table 14.14 . In this table,
“CHMIN” signifies the chemically defined mineralization
unit while “LOGMIN” is the logged mineralization unit.
Table 14.14 shows, for example, that oxide mineralization
is defined as a soluble Cu grade greater or equal than 0.2 %,
and the ratio of SCu to TCu is greater than 50 %. The limits
on the SCu/TCu ratio are derived from definitions used by
the processing plant.
The basic statistics of the leach, partial leach, and oxide
material are similar, except that the ratio of the chemically
defined oxides has a sharp boundary at 50 %, while there are
samples geologically described as oxides with a ratio of less
than 50 %. The largest difference between the two is in the
mixture of oxide and sulfide minerals. There are significantly
more mixed samples chemically defined than there are logged.
This is explained by the geologist's natural tendency to clas-
sify as mixed those samples with approximately the same
number of oxide and sulfide particles. If one or the other is
clearly prevalent, then the geologist will probably classify the
unit as oxides or sulfides, according to the majority observed.
Given that the distributions of both SCu/TCu ratios can
be considered similar, the decision was made to use the
logged database, relying on the geological definition of the
mineralization units.
The next step was the definition of the estimation do-
mains, a process similar to what was described in Chap. 4.
The estimation domains result from a combination of lithol-
ogy and alteration units, as well as structural zones within
the deposit. Figure 14.38 shows the structural zones defined
in the Escondida Norte deposit. Table 14.15 shows the defi-
nitions in terms of oxide and sulfide indicators. The com-
bined indicators define the unit in question: for leach, both
indicators have to be 0, i.e., there are no sulfide or oxide
minerals; while for mixed, both types of minerals need to be
present. There are a total of 5 oxides and 3 sulfide Domains.
After defining the estimation domains, the exploratory
data analysis and variography was completed for each indi-
cator and for each domain. This defined the estimation strat-
egy and the indicator kriging estimation plans for each unit.
Indicator kriging was completed using the oxide and sul-
fide indicators defined in the database, i.e., a weighted linear
combination of zeros and ones that result in estimated values
between 0 and 1. These interpolated values can be interpret-
ed either as the probability of each estimated block of having