3744
Optimising SAG Mill Operations: Insights from Real-Time
Acoustic Monitoring and Analysis
Kwaku B. Owusu
Future Industries Institute, UniSA STEM, University of South Australia, Mawson Lakes Campus, Australia
William Skinner, Richmond K. Asamoah
Future Industries Institute, UniSA STEM, University of South Australia, Mawson Lakes Campus, Australia
ARC Centre of Excellence for Enabling Eco-Efficient Beneficiation of Minerals,
University of South Australia, Mawson Lakes Campus, Australia
AISRF for Advanced Recovery of the Battery Materials and REE from Ores and Wastes,
University of South Australia, Mawson Lakes Campus, Australia
ABSTRACT: This paper presents an advanced investigation on optimising SAG mill operations using
acoustic sensor for ore characteristics and mill state monitoring. Specifically, a series of systematically designed
experiments, involving a laboratory-scale SAG mill equipped with an acoustic sensor, for unravelling ore
charge compositions (ore, steel ball, and water), and ore heterogeneity (feed hardness and size distribution)
was undertaken. The study observed that intensified mill-generated noise serves as an indicative parameter of
multifaceted operational events, notably involving steel balls-ore-lifters/liners interactions. Also, different ore
hardness and size distributions showed unique acoustic signatures through simple regression and extended
classification models, relevant for mill optimisation.
Keywords: Semi-autogenous mill, ore heterogeneity, acoustic response, real-time monitoring, operational
efficiency
INTRODUCTION
The role of comminution (crushing and milling), which
encompasses mainly particle size reduction and liberation,
has been underscored as highly energy intensive, empha-
sising the need for continuous improvements and opti-
misation strategies (Ballantyne et al., 2012 Owusu et al.,
2021h). Typically, in milling, the autogenous/semi-autog-
enous (AG/SAG) mills possess more challenges as primary
mills compared with secondary mills (e.g., ball mills). One
major concern of the AG/SAG mill revolves around the
intricate real-time in-mill monitoring, given the non-trans-
parency and harsh operating environment of the mill.
Owing to the complexities of online in-mill monitor-
ing, various techniques are employed, encompassing the
traditional approach (e.g., input-output technique, the
expertise of mill operators), simulations and modelling,
as well as sensor applications (e.g., vibration, acoustic)
(Hodouin, 2011 Hodouin et al., 2001 Hosseini et al.,
2011 Owusu et al., 2020a Owusu et al., 2021f Owusu
et al., 2021h Pax, 2001 Pax, 2011a Pax, 2016 Pax and
Thornton, 2019).
Optimising SAG Mill Operations: Insights from Real-Time
Acoustic Monitoring and Analysis
Kwaku B. Owusu
Future Industries Institute, UniSA STEM, University of South Australia, Mawson Lakes Campus, Australia
William Skinner, Richmond K. Asamoah
Future Industries Institute, UniSA STEM, University of South Australia, Mawson Lakes Campus, Australia
ARC Centre of Excellence for Enabling Eco-Efficient Beneficiation of Minerals,
University of South Australia, Mawson Lakes Campus, Australia
AISRF for Advanced Recovery of the Battery Materials and REE from Ores and Wastes,
University of South Australia, Mawson Lakes Campus, Australia
ABSTRACT: This paper presents an advanced investigation on optimising SAG mill operations using
acoustic sensor for ore characteristics and mill state monitoring. Specifically, a series of systematically designed
experiments, involving a laboratory-scale SAG mill equipped with an acoustic sensor, for unravelling ore
charge compositions (ore, steel ball, and water), and ore heterogeneity (feed hardness and size distribution)
was undertaken. The study observed that intensified mill-generated noise serves as an indicative parameter of
multifaceted operational events, notably involving steel balls-ore-lifters/liners interactions. Also, different ore
hardness and size distributions showed unique acoustic signatures through simple regression and extended
classification models, relevant for mill optimisation.
Keywords: Semi-autogenous mill, ore heterogeneity, acoustic response, real-time monitoring, operational
efficiency
INTRODUCTION
The role of comminution (crushing and milling), which
encompasses mainly particle size reduction and liberation,
has been underscored as highly energy intensive, empha-
sising the need for continuous improvements and opti-
misation strategies (Ballantyne et al., 2012 Owusu et al.,
2021h). Typically, in milling, the autogenous/semi-autog-
enous (AG/SAG) mills possess more challenges as primary
mills compared with secondary mills (e.g., ball mills). One
major concern of the AG/SAG mill revolves around the
intricate real-time in-mill monitoring, given the non-trans-
parency and harsh operating environment of the mill.
Owing to the complexities of online in-mill monitor-
ing, various techniques are employed, encompassing the
traditional approach (e.g., input-output technique, the
expertise of mill operators), simulations and modelling,
as well as sensor applications (e.g., vibration, acoustic)
(Hodouin, 2011 Hodouin et al., 2001 Hosseini et al.,
2011 Owusu et al., 2020a Owusu et al., 2021f Owusu
et al., 2021h Pax, 2001 Pax, 2011a Pax, 2016 Pax and
Thornton, 2019).