1045
Forecasting the Recovery Potential of a Tailings Storage Facility
with Particle-Based Separation Models
Lucas Pereira, Silum Ghebreyesus, Max Frenzel, Duong H. Hoang, Martin Rudolph,
Gerald van den Boogaart, Jens Gutzmer, Raimon Tolosana-Delgado
Helmholtz-Zentrum Dresden-Rossendorf, Helmholtz Institute Freiberg for Resource Technology, Freiberg, Germany
ABSTRACT: The reprocessing of tailings storage facilities (TSFs) is highly relevant given the potential to
increase metal production and reduce environmental issues. Yet, in these TSFs, minerals are often contained
within fine and non-liberated particles, which not only pose a challenge to their efficient concentration but
also to forecasting the outcome of such activities. In this case, particle-based separation models, which account
for particle complexity in their recovery estimation, are fundamental. This study covers the construction of a
recovery model for a TSF from Germany, in which five flotation tests are required to reach a high accuracy in
recovery predictions. Differences in flotation behavior in the different areas of the TSF seem to be hindering the
quality of grade predictions.
INTRODUCTION
Tailings storage facilities (TSFs) may host considerable
amounts of raw materials while also being potential envi-
ronmental threats. Reprocessing of TSFs thus became a
major interest in the last years to minimize environmen-
tal issues while recovering marketable metals (Araya et
al., 2020 Kinnunen and Kaksonen, 2019). Often, these
metals are carried by fine and non-liberated mineral par-
ticles (Blannin et al., 2022 Pereira et al., 2019)—the rea-
son for being lost to the tailings of the original operations
in the first place. Thus, TSF reprocessing activities are not
trivial and require not only new processing strategies but
also better techniques for understanding and forecasting
operations.
The field of automated mineralogy has largely con-
tributed to improving the recovery of complex ores (Gu et
al., 2014 Lotter, 2011) by quantifying in high resolution
the microstructural properties of individual particles in a
sample—fundamental information to better understand
particle separation processes. Such particle datasets are the
cornerstone for a series of methodologies to enable a multi-
dimensional understanding of particle separation processes
(Hannula et al., 2018 Lamberg and Vianna, 2007 Pereira
et al., 2021 Schach et al., 2019). The latest of these meth-
ods, the particle-based separation models (PSMs) intro-
duced by Pereira et al. (2021), enables the use of all relevant
particle data collected with automated mineralogy, result-
ing in the recovery quantification of all individual particles
present in a sample, based on their size, shape, and com-
position, both in terms of liberation and association. This
method has been successfully applied to a series of complex
ores (Pereira et al., 2022, 2021), but not yet to tailings.
This study aims to combine PSMs and geostatistical
tools into a methodology to estimate the recovery poten-
tial of TSFs with only a few processing experiments. As
a case study, we evaluate the potential of minimizing the
environmental impact of a historical TSF in Germany with
bulk sulfide flotation, to inactivate the remaining tailings
material.
Forecasting the Recovery Potential of a Tailings Storage Facility
with Particle-Based Separation Models
Lucas Pereira, Silum Ghebreyesus, Max Frenzel, Duong H. Hoang, Martin Rudolph,
Gerald van den Boogaart, Jens Gutzmer, Raimon Tolosana-Delgado
Helmholtz-Zentrum Dresden-Rossendorf, Helmholtz Institute Freiberg for Resource Technology, Freiberg, Germany
ABSTRACT: The reprocessing of tailings storage facilities (TSFs) is highly relevant given the potential to
increase metal production and reduce environmental issues. Yet, in these TSFs, minerals are often contained
within fine and non-liberated particles, which not only pose a challenge to their efficient concentration but
also to forecasting the outcome of such activities. In this case, particle-based separation models, which account
for particle complexity in their recovery estimation, are fundamental. This study covers the construction of a
recovery model for a TSF from Germany, in which five flotation tests are required to reach a high accuracy in
recovery predictions. Differences in flotation behavior in the different areas of the TSF seem to be hindering the
quality of grade predictions.
INTRODUCTION
Tailings storage facilities (TSFs) may host considerable
amounts of raw materials while also being potential envi-
ronmental threats. Reprocessing of TSFs thus became a
major interest in the last years to minimize environmen-
tal issues while recovering marketable metals (Araya et
al., 2020 Kinnunen and Kaksonen, 2019). Often, these
metals are carried by fine and non-liberated mineral par-
ticles (Blannin et al., 2022 Pereira et al., 2019)—the rea-
son for being lost to the tailings of the original operations
in the first place. Thus, TSF reprocessing activities are not
trivial and require not only new processing strategies but
also better techniques for understanding and forecasting
operations.
The field of automated mineralogy has largely con-
tributed to improving the recovery of complex ores (Gu et
al., 2014 Lotter, 2011) by quantifying in high resolution
the microstructural properties of individual particles in a
sample—fundamental information to better understand
particle separation processes. Such particle datasets are the
cornerstone for a series of methodologies to enable a multi-
dimensional understanding of particle separation processes
(Hannula et al., 2018 Lamberg and Vianna, 2007 Pereira
et al., 2021 Schach et al., 2019). The latest of these meth-
ods, the particle-based separation models (PSMs) intro-
duced by Pereira et al. (2021), enables the use of all relevant
particle data collected with automated mineralogy, result-
ing in the recovery quantification of all individual particles
present in a sample, based on their size, shape, and com-
position, both in terms of liberation and association. This
method has been successfully applied to a series of complex
ores (Pereira et al., 2022, 2021), but not yet to tailings.
This study aims to combine PSMs and geostatistical
tools into a methodology to estimate the recovery poten-
tial of TSFs with only a few processing experiments. As
a case study, we evaluate the potential of minimizing the
environmental impact of a historical TSF in Germany with
bulk sulfide flotation, to inactivate the remaining tailings
material.