Resource Model Validations and Reconciliations - Mineral Resource Estimation - page 200

Comparisons based on longer periods (quarterly, semi-an-
nually, or even yearly) may be necessary and the only avail-
able option if the quality of the information is relatively poor,
or if the operation operates on a very small scale. Estimating
average values of larger volumes of data (periods) is easier
than for smaller periods. Also, longer periods may be neces-
sary if the operation handles a significant number of stock-
piles, since accurately measuring tonnage and grades drawn
from and put into stockpiles is difficult. The suggested meth-
od does not include internal reconciliation of the processing
facilities (mill balances), which should be an integral part of
the reconciliation process, but outside the scope of this topic.
The basic information to be compiled includes:
• Tonnage, grades, and metal content of the long-term
resource model for the period. This implies obtaining the
mine advance positions for the period, and to superim-
pose them on the block model.
• Similarly, tonnages, grades, and metal content should be
obtained from short-term models for the same periods, if
they exist.
• Tonnage, grades, and metal content should be obtained
from the daily production model (grade control model).
This information should be gathered daily, but compiled
into the proper reconciliation period (monthly). In open
pit mines, the grade control model for the same periods,
with all its potential problems and pitfalls, including
sample quality and sometimes inadequate modeling tech-
niques, represents the best possible “in-situ” information
available, given the typical density of information. Some-
times this is also true for underground mines, but more
commonly, the only reliable information is the short-term
model from which the final stope designs are made.
• Tonnage, grades, and metal content as reported by the
mine should also be compiled. The grade usually cor-
responds to the grade assigned to the extracted panel or
stope by the grade control model. This may include some
downgrading to consider operational dilution and ore
loss, as well as blast movement. Reported tonnage may
be from truck weights (preferably avoiding truck factors),
or by direct topographic measurement of the extracted
volumes. Sometimes the only tonnage available is the one
reported by the grade control model.
• Tonnage, grades, and metal content informed as head
grades and tons. This should be based on direct sampling ,
as opposed to back-calculated from tailings grades and
adjusted recoveries. Back-calculated head tonnage and
grades should not be used for model optimization.
There may be stockpiles to be considered in between the
mine and the mill. Also, material within the metallurgical
stream itself should be accounted for. However, some or all
of these stockpiles may not be relevant to the reconciliation
program if they are completely “turned over” or replaced
within the reporting period. In that case, they can simply be
ignored for production reconciliation.

With the information described above, several dimen-
sionless factors can be calculated:
1. F 1 factor, defined for tonnage, grade, and metal content
(F 1t , F 1g , F 1m ). It is calculated from the corresponding ton-
nage, grade, and metal of the long-term and short-term
models as:

= Short-term
Long-term

F 1

2. F 2 tonnage, grade, and metal content factors (F 2t , F 2g ,
F 2g ). These compare the grade control (production) model
versus the short-term model (if it exists), and are calcu-
lated as:

= Grade-control
Short-term

F 2

3. F 3 factors (F 3t , F 3g , F 3m ) may be defined based on the ton-
nages, grades, and contained metal of the monthly mine
report versus the grade control model. Sometimes mine
reports for tonnage and grades are simply taken from the
grade control model, and are informed as material sent
to the mill. In other instances the mine reports the grade
provided by the grade control model (in open pits gener-
ally there is no other option, although for underground
operations may be based on additional sampling), but the
reported tonnage is based on truck weights, counts, or
volumetric measurements of the advances. If applicable,
the F 3 factors are calculated as:

= Mine-reported
Grade-control

F 3

4. F 4 factors (F 4t , F 4l , F 4f ) based on tonnages, grades, and
metal content of the “received-at-mill” material versus
the mine reported material. The F 4 factors may be calcu-
lated as:

= Received-at-mill
Mine-reported

F 4

Not all these factors need to be defined, as for example
when no short-term model exists. Note that, as defined, a
factor greater than 1.0 implies underestimation, while a fac-
tor smaller than 1.0 implies overestimation. From these fac-
tors, several performance measures can be readily obtained.
For example, to quantify, the performance of the long-term
model in terms of tonnage and grade of ore delivered to the
mill, the F LT M factor is obtained as:

= Received-at-mill
Long-term

F LTM = F 1 × F 2 × F 3 × F 4