Mining Goes Digital – Mueller et al. (Eds)
© 2019 Taylor & Francis Group, London, ISBN 978-0-367-33604-2
309
A data science model on production level pillar stability
at El Teniente mine
R.J. Quevedo & R.F. Quezada
Resident Engineers at El Teniente Mine, Universidad de Santiago, Chile
R.A. Zepeda & S.A. Balboa
Mining Engineers, Department of Geomechanics, CODELCO División El Teniente, Chile
J.P. Vargas & S.A. Pérez
Mining Engineering Department, Universidad de Santiago de Chile, Chile
ABSTRACT: Collapses have been a major geomechanical issue at El Teniente Mine, due
to its difficult treatment and the known fact that actually there are neither methodologies
nor tools to estimate their possible occurrences or successfully holding their progression. The
term “collapse” is understood as a physical process consisting in major large-scale deformations whose later manifestation is the total closure of affected excavations and/or drifts. The
objective of this paper is to illustrate the development and results of a pillar vulnerability
index, built through a machine learning classification model, in order to estimate collapserisky areas in production sectors and future projects that are to be put in operation in the
coming years, given the 34 years of history of primary ore extraction at El Teniente mine.
1 INTRODUCTION
El Teniente Mine is an underground operation owned by CODELCO, which is located in the
first elevations of Los Andes in the central zone of Chile, about 70 km SSE from the capital
city, Santiago.
El Teniente mine is a typical porphyry copper-molybdenum deposit, composed mainly by
andesite, diorite and hydrothermal breccias of the Miocene eras as the main lithologies.
The main structural feature of the orebody is a stockwork of multi-directional veins and
veinlets. The veins are principally cemented with anhydrite, quartz and sulphides.
Geometrically, the deposit is located around a chimney composed of subvolcanic breccias,
known as ‘Braden Pipe’, which postdates the copper-moly mineralization. The hydrothermal mineralization is distributed around this pipe over a variable radial extension of 400 to
800 m, with mineralogical associations of variable strength.
The mineralization shows two distinct forms: the secondary ore is located near the surface
and the primary mineralization is located at a greater depth.
The primary ore can be described as a high cohesion and impermeable rockmass. The
stock-work veins, containing the original mineralogy, are sealed. According to the geomechanical characterization, the primary ore is harder than the secondary. It’s caving results
in large fragmentation. Primary rockmass could exhibit brittle, often violent failure mechanisms under a high-stress condition.
2 RESEARCH CONTEXT
Since the beginning of the mining operation in 1982, a series of geomechanical instabilities
known as ‘collapses’ have occurred. A collapse is defined mainly as a progressive deformation
© 2019 Taylor & Francis Group, London, ISBN 978-0-367-33604-2
309
A data science model on production level pillar stability
at El Teniente mine
R.J. Quevedo & R.F. Quezada
Resident Engineers at El Teniente Mine, Universidad de Santiago, Chile
R.A. Zepeda & S.A. Balboa
Mining Engineers, Department of Geomechanics, CODELCO División El Teniente, Chile
J.P. Vargas & S.A. Pérez
Mining Engineering Department, Universidad de Santiago de Chile, Chile
ABSTRACT: Collapses have been a major geomechanical issue at El Teniente Mine, due
to its difficult treatment and the known fact that actually there are neither methodologies
nor tools to estimate their possible occurrences or successfully holding their progression. The
term “collapse” is understood as a physical process consisting in major large-scale deformations whose later manifestation is the total closure of affected excavations and/or drifts. The
objective of this paper is to illustrate the development and results of a pillar vulnerability
index, built through a machine learning classification model, in order to estimate collapserisky areas in production sectors and future projects that are to be put in operation in the
coming years, given the 34 years of history of primary ore extraction at El Teniente mine.
1 INTRODUCTION
El Teniente Mine is an underground operation owned by CODELCO, which is located in the
first elevations of Los Andes in the central zone of Chile, about 70 km SSE from the capital
city, Santiago.
El Teniente mine is a typical porphyry copper-molybdenum deposit, composed mainly by
andesite, diorite and hydrothermal breccias of the Miocene eras as the main lithologies.
The main structural feature of the orebody is a stockwork of multi-directional veins and
veinlets. The veins are principally cemented with anhydrite, quartz and sulphides.
Geometrically, the deposit is located around a chimney composed of subvolcanic breccias,
known as ‘Braden Pipe’, which postdates the copper-moly mineralization. The hydrothermal mineralization is distributed around this pipe over a variable radial extension of 400 to
800 m, with mineralogical associations of variable strength.
The mineralization shows two distinct forms: the secondary ore is located near the surface
and the primary mineralization is located at a greater depth.
The primary ore can be described as a high cohesion and impermeable rockmass. The
stock-work veins, containing the original mineralogy, are sealed. According to the geomechanical characterization, the primary ore is harder than the secondary. It’s caving results
in large fragmentation. Primary rockmass could exhibit brittle, often violent failure mechanisms under a high-stress condition.
2 RESEARCH CONTEXT
Since the beginning of the mining operation in 1982, a series of geomechanical instabilities
known as ‘collapses’ have occurred. A collapse is defined mainly as a progressive deformation