6
tests conducted on a large number of samples. These tests
are instrumental in identifying variability patterns across
geological domains. The Geopyörä breakage test (Bueno
et al, 2021), in particular, offers a cost-effective solution
by leveraging small assay samples (e.g., 1m of half/quar-
ter core) already collected during exploration. Its ability to
process a high volume of samples enables a more detailed
initial characterization of orebody heterogeneity, which
serves as the foundation for subsequent, more detailed test-
ing campaigns.
Drawing on the framework proposed by Koch and Link
(1971), the iterative sampling strategy emphasizes progres-
sively increasing the number of samples tested until the
coefficient of variation (CV) stabilizes within each domain.
The CV, calculated as the ratio of the standard deviation
to the mean, provides a robust measure of parameter vari-
ability and ensures that the sampling effort is aligned with
the complexity of the orebody. Stabilization of the CV
indicates that sufficient data has been collected to reli-
ably characterize domain-specific variability, minimizing
the risks associated with under-sampling while controlling
project costs. By incorporating statistical and geostatistical
tools, this approach ensures that sampling density evolves
dynamically based on the observed variability, particularly
in high-risk zones.
Finally, this adaptive sampling design allows for risk-
weighted resource allocation, prioritizing domains with
higher variability or operational significance. The integra-
tion of machine learning algorithms further enhances the
efficiency of this framework by using early test results to
optimize sampling grids and reduce redundancy. This itera-
tive and adaptive methodology strengthens the reliability of
throughput forecasts and ensures cost-effectiveness, align-
ing comminution test intensity with the financial and tech-
nical risk profiles of mining projects.
CONCLUSION AND
RECOMMENDATIONS
This study highlights the critical importance of optimising
comminution sampling and testing practices, particularly
for projects employing AG and SAG mills. Current prac-
tices prioritize geochemical assays at the expense of com-
minution testing, which can create substantial uncertainty
in throughput forecasting and project economics. This is
Figure 7. NPV to throughput variability relationship
tests conducted on a large number of samples. These tests
are instrumental in identifying variability patterns across
geological domains. The Geopyörä breakage test (Bueno
et al, 2021), in particular, offers a cost-effective solution
by leveraging small assay samples (e.g., 1m of half/quar-
ter core) already collected during exploration. Its ability to
process a high volume of samples enables a more detailed
initial characterization of orebody heterogeneity, which
serves as the foundation for subsequent, more detailed test-
ing campaigns.
Drawing on the framework proposed by Koch and Link
(1971), the iterative sampling strategy emphasizes progres-
sively increasing the number of samples tested until the
coefficient of variation (CV) stabilizes within each domain.
The CV, calculated as the ratio of the standard deviation
to the mean, provides a robust measure of parameter vari-
ability and ensures that the sampling effort is aligned with
the complexity of the orebody. Stabilization of the CV
indicates that sufficient data has been collected to reli-
ably characterize domain-specific variability, minimizing
the risks associated with under-sampling while controlling
project costs. By incorporating statistical and geostatistical
tools, this approach ensures that sampling density evolves
dynamically based on the observed variability, particularly
in high-risk zones.
Finally, this adaptive sampling design allows for risk-
weighted resource allocation, prioritizing domains with
higher variability or operational significance. The integra-
tion of machine learning algorithms further enhances the
efficiency of this framework by using early test results to
optimize sampling grids and reduce redundancy. This itera-
tive and adaptive methodology strengthens the reliability of
throughput forecasts and ensures cost-effectiveness, align-
ing comminution test intensity with the financial and tech-
nical risk profiles of mining projects.
CONCLUSION AND
RECOMMENDATIONS
This study highlights the critical importance of optimising
comminution sampling and testing practices, particularly
for projects employing AG and SAG mills. Current prac-
tices prioritize geochemical assays at the expense of com-
minution testing, which can create substantial uncertainty
in throughput forecasting and project economics. This is
Figure 7. NPV to throughput variability relationship