Progress and Challenges in Electrical Comminution by High-Voltage Pulses

Abstract

The recent progress in electrical comminution using high-voltage pulses and the technical challenges in order to bring this technology to the mining industry are outlined. Pre-weakening ore particles and preferential liberation of minerals at coarse sizes are the two major research outcomes that may have potential benefits for the industry. A particle pre-weakening characterization method by single-particle/single-pulse test has been developed. The emerging challenges for the mining industry to realize the benefits of this novel comminution technology include scale-up for industrial application, hybrid circuit design, maximization of pulse-induced cracks, and study of the downstream processing effects.

Full text

Progress and Challenges in Electrical
Comminution by High-Voltage Pulses
The recent progress in electrical comminution using high-voltage pulses and the
technical challenges in order to bring this technology to the mining industry are
outlined. Pre-weakening ore particles and preferential liberation of minerals at
coarse sizes are the two major research outcomes that may have potential benefits
for the industry. A particle pre-weakening characterization method by single-
particle/single-pulse test has been developed. The emerging challenges for the
mining industry to realize the benefits of this novel comminution technology
include scale-up for industrial application, hybrid circuit design, maximization of
pulse-induced cracks, and study of the downstream processing effects.
Keywords: Electrical comminution, Energy, High-voltage pulses, Mineral liberation
Received: September 27, 2013; revised: October 28, 2013; accepted: November 21, 2013
DOI: 10.1002/ceat.201300660
1 Introduction
In 2009–2010, the mining industry in Australia used 509 PJ
energy, 8.6% of the total energy consumed in Australia [1].
This corresponds to 136 million tons of CO
2
emission. Com-
minution, including crushing and grinding, consumes about
one third of the energy used by the mining industry. This is
equivalent to the output from five 1400-MW coal-fired power
stations, all energy being necessary to crush and grind rocks.
Improved methods for mineral comminution are continually
being sought, in order to achieve the size reduction and miner-
al liberation required for better recovery of valuable minerals
at lower energy consumption and lower emissions.
High-voltage pulses to break rocks have attracted the atten-
tion of researchers for the past half century [2–12]. Previous
research in the mineral industry mainly focused on mineral
liberation which consumed significantly high amounts of ener-
gy. In the past five years, the Julius Kruttschnitt Mineral
Research Centre (JKMRC), in collaboration with SELFRAG
AG, has made a significant effort towards developing an elec-
trical comminution technique for the mineral industry. The
major outcomes and emerging challenges to overcome for the
mineral industry to realize the benefits are reported.
2 Electrical Comminution
The high-voltage (HV) pulse power technology consists of an
HV power supply, an HV pulse generator, and the process area
(Fig. 1 a, [13]). Materials are immersed in a liquid in the pro-
cess area. Dielectric liquids, like water, have a high dielectric
strength when the voltage rising time is kept below 500 ns
(Fig. 1b). Consequently, the water acts as a special electrical in-
sulator to prevent electrical discharge occurring outside the
rocks [14, 15].
A number of technical names related to HV pulse technolo-
gy appear in the literature. Tab. 1 lists their characteristic fea-
tures based on the method of transferring energy and the HV
rising time. The term electrical comminution is used to cover
all those applying HV pulses.
Note that the voltage rising time in electrohydraulic disinte-
gration is larger than 500 ns. Compared with other pulse tech-
niques using shorter voltage rising times, the energy efficiency
of electrohydraulic disintegration may be lower, and the results
of rock pre-weakening and mineral liberation may be different.
At present, two brands of HV pulse equipment appear on the
market: selFrag manufactured by SELFRAG AG based in Swit-
zerland, and Spark by CNT-MC based in Canada. The JKMRC
has installed a selFrag Lab machine. The selFrag Lab is
designed to treat samples in the 1-kg range per batch. The
selectable process parameters are: number of pulses, discharge
voltage (91–200 kV) which controls the energy per pulse, elec-
trode gap (10–40 mm), and frequency of discharge (1–5 Hz).
Chem. Eng. Technol. 2014,37, No. 5, 1–6 © 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.cet-journal.com
Fengnian Shi
Emmy Manlapig
Weiran Zuo
The University of Queensland,
Sustainable Minerals Institute,
Julius Kruttschnitt Mineral
Research Centre, Brisbane,
Australia.
–
Correspondence: Dr. Fengnian Shi (f.shi@uq.edu.au), The University of
Queensland, Sustainable Minerals Institute, Julius Kruttschnitt Mineral
Research Centre, 40 Isles Road, Brisbane, 4068, Australia.
Table 1. Features of high-voltage pulse technology.
Technical name Electrode
contact
Voltage rising
time [ns]
Electrohydraulic disintegration Water > 500
Electrical disintegration Rock < 500
Electrodynamic disintegration Water < 500
Electrical pulse disaggregation Water < 500
Review 1

To avoid overbreaking the target minerals, a process vessel with
interchangeable sieve bottoms is used. Sieve apertures from 4
to 0.3 mm are typically applied.
3 Recent Development
More than ten ore samples including copper, gold, lead/zinc,
platinum ores, and industrial minerals from various mine sites
around the world were treated by HV pulses equipment and
by conventional comminution facilities for comparison. In
total, over 5 t of ore samples in the size range of 10–45 mm
have been processed by the JKMC. Data cumulated from the
extensive experimental work were used to develop applications
of electrical comminution for the mineral industry and to gain
knowledge on the factors affecting the efficiency of electrical
comminution.
3.1 Pre-Weakening Ore Particles
A new application of HV pulse power has been developed
which applies the specific energies of 1–3 kWh t
–1
to pre-weak-
en ore particles. Evidence of cracks and microcracks measured
by X-ray tomography (Fig. 2) and mercury porosimetry sup-
ported the principle of HV pulse power-induced damage on
rocks, resulting in a reduction of particle strength and, conse-
quently, a reduction in the energy consumption in the down-
stream process [16].
A residual breakage resistance indicator, A*b, of the pulses’
breakage product and mechanical breakage product was mea-
sured by a rotary breakage tester (JKRBT [17]). The typical
A*b values in the JKMRC database which consists of more
than 2000 standard breakage testing data for ore particles, are
between 20 to 300, A*b values less than 40 indicating very high
resistance to breakage and larger than 100 indicating less resis-
tance to breakage. Changes in the A*b values from 9 % to 52 %
(using 700 g per batch test) were observed between the two
products fragmented with similar specific energy levels. Up to
24 % reduction in the Bond ball mill work index in the selFrag
product was observed. Large variations of the pre-weakening
effect were found with these ore samples.
3.2 Preferential Liberation
Investigation of preferential liberation by HV pulses was con-
ducted on three ore samples (one copper ore, one gold ore,
and one platinum group metal (PGM) ore), in comparison
with the mechanical breakage products with identical specific
energy levels [18]. Mineral liberation analysis was performed
using a mineral liberation analyzer (MLA). Close to 400 MLA
datasets were collected for statistical analysis of the compara-
tive results.
www.cet-journal.com © 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eng. Technol. 2014,37, No. 5, 1–6
t
~500 ns
Water
a) b)
Figure 1. The high voltage pulse technology [2], a) three major components; b) breakdown strength in relation to voltage rising time.
2000μm
Figure 2. X-Ray tomography image of a gold-copper ore particle
treated with one electrical pulse, dark lines showing cracks/
microcracks.
2Review

Distribution of the liberated minerals demon-
strates that in the electrical comminution product
a large percentage of the liberated minerals appear
in size fractions coarser than 53 lm, while in the
mechanical comminution product the liberated
minerals are accumulated in fine and very fine size
fractions (Fig. 3). Therefore, there may be potential
benefits in recovering the coarse liberated minerals
in the electrical comminution product, prior to
further grinding.
3.3 Factors Affecting Electrical
Comminution
A number of factors associated with HV pulse
operations were investigated through experimental
work and numerical simulations of electrical field
distribution using the software COULOMB 3D. It
was found that particles having the following dis-
tinguishing features may enhance the electrical
comminution performance: coarse grained miner-
als of interest, large feed particle size, conductive
minerals embedded within gangues, angular parti-
cle shape, and large difference in the electrical
properties between the valuable mineral-hosting
and gangue phases [19].
Numerical simulation also indicates that high electrical field
intensity was created around the boundaries of two mineral
phases with large differences in their permittivity and conduc-
tivity. Fig. 4 demonstrates that the electrical field intensity in
silicate is 4 kV mm
–1
under 100 kV loading. The electrical field
intensity in the boundaries between chalcopyrite and silicate
rises to 5.4 kV mm
–1
. Similar intensity occurs along the bound-
aries between pyrite and silicate. The larger electrical filed
intensity will cause higher tensile stress, leading to breakage
taking place along boundaries between minerals of different
permittivity. This elucidates the mechanism of preferential
liberation in electrical comminution.
3.4 Characterization of Pre-Weakening Effect
Recent research found that HV pulse discharges develop in an
array around the electrode axis, and particles further from the
electrode axis do not receive the same level of pulse energy as
the ones directly under the electrode [20]. The pre-weakening
effect determined through the standard 700-g batch tests is an
averaged effect of treated and mistreated particles.
A single particle subjected to a single pulse discharge meth-
od has been developed which effectively eliminates the influ-
ence of machine-inefficiency factor. Tab. 2 compares the pre-
weakening results by the 700-g batch tests and the single-parti-
cle tests on the same gold/copper ore sample. The A*b values
are compared at a nominal size of 30 mm.
Chem. Eng. Technol. 2014,37, No. 5, 1–6 © 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.cet-journal.com
Figure 3. Distribution of the > 95 % liberated chalcopyrite deportment in the products of a copper ore sample, comminuted mechanically
and electrically using 9 kWh t
–1
and 22 kWh t
–1
respectively, with the error bars indicating the 95 % confidence interval.
Applied Voltage 100 kV
X-plane
Chalcopyrite Pyrite
5.4 kV/mm5.4 kV/mm
Water
0 kV/mm
Silicate
(4 kV/mm)
Figure 4. Electrical field distribution of chalcopyrite and pyrite spheres em-
bedded in the K-feldspar immersed in water, with an electrical potential differ-
ence of 100 kV.
Review 3

Tab. 2 indicates very different pre-weakening results deter-
mined by the two methods. Some of the ore samples tested pre-
viously, showing a poor response to electrical comminution,
may not be due to their poor amenability, but due to the ma-
chine-inefficiency factor associated with the testing method.
Fig. 5 displays the calculated A*b values in relation to parti-
cle size for pulse-treated and untreated SAG (semi-autogenous
grinding) mill feed of a gold-copper ore. For both pulse-treat-
ed and untreated samples it is obvious that the larger particles
are weaker (higher A*b values) than the smaller particles (both
showing an increased trend). This is known as particle size ef-
fect on breakage. The size effect is more pronounced in the
HV pulse-treated fragments than in the untreated material.
This confirms that the larger fragments would achieve a better
pre-weakening effect than the smaller fragments when sub-
jected to HV pulses. The characterization result suggests that
using HV pulses to pre-weaken AG/SAG mill feed may achieve
more significant benefits in terms of energy saving or increased
throughput than pre-weakening the ball mill feed.
4 Challenges and Current Research
A number of challenges have emerged in the application of
electrical comminution technology. These need to be ad-
dressed in the current research, in order to realize the full ben-
efits of such technology for the mineral industry.
4.1 Scale-up for Industrial Application
The majority of the JKMRC research outcomes were achieved
based on laboratory-scale selFrag systems. Scaling up to treat
large particle sizes with high throughput in a continuous
operation mode is a challenge, but not a problem. The pilot-
scale unit of a pre-weakening station with 1 t h
–1
capacity is
available on the market [21], the engineering design of 10 t h
–1
pilot pre-weakening station is complete, and 100–1000 t h
–1
multiple modules can be developed.
One of the real challenges is to maximize the probability of
every particle presented in the feed receiving the energy in each
pulse discharge. The single-particle characterization test has
demonstrated significant improvement in the pre-weakening
result. In a large-scale continuous operation system, how to
maximize the efficiency in delivering the pulse energy to all
feed particles holds the key to achieving better electrical com-
minution results for the mineral industry.
4.2 Hybrid Circuit Design
The electrical comminution unit needs to be integrated in the
traditional mechanical comminution circuit to achieve opti-
mal technical and economic benefits. To design a hybrid com-
minution circuit, factors considered include the selection of
the SAG mill feed size fractions to be treated by HV pulses, the
treatment of SAG mill pebbles, the transfer size between SAG
mill and ball mill, and the energy distributions among HV
pulse pre-weakening, SAG milling and ball milling.
Currently, the JKMRC is developing an electrical comminu-
tion model to describe the breakage probability and pre-
weakening index (PWI). It is expected the electrical commi-
nution model be integrated in the existing JKSimMet, the
popular comminution software used in the mineral industry,
to design the hybrid comminution circuit and its optimization,
based on ore breakage characteristics.
4.3 Cracks/Microcracks Generation
The X-ray tomography technique gives evidence to support
that cracks/microcracks induced by HV pulses are the main
reason for the ore particle pre-weakening effect. Recent
research found that the pulse-induced cracks/microcracks may
be exhausted in the first stage mechanical breakage, e.g., SAG
milling, leading to the secondary grinding (ball mill) feed ore
retaining the original hardness. The challenge is how to
increase the microcracks’ density in a particle induced by HV
pulses.
A research program utilizing numerical simulation tools,
e.g., DEM and FEM, to understand cracks/microcracks genera-
tion mechanisms from fundamental approach will be under-
taken. The recently installed and commissioned Versa 500
www.cet-journal.com © 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eng. Technol. 2014,37, No. 5, 1–6
Table 2. Comparison of the pre-weakening results; PWI – pre-
weakening index.
Method 700-g batch Single particle
Particle size [mm] 26.5–45 26.5–45
Particle mass [kg] 21 5.0
selFrag Ecs [kWh t
–1
] 4.6 1.6
A*b (untreated) 31 31
A*b (treated) 52 84
Change in A*b [%] 68 171
PWI [% change in A*b
per kWh t
–1
]15 107
Figure 5. Comparison of A*b values in relation to fragment size
between the high voltage pulse treated and the untreated SAG
mill feed of a copper ore sample.
4Review

X-ray tomography device at the JKMRC, in addition to the
existing Sky scanner, will provide useful facilities to validate
and guide the numerical simulations.
In the previous experiment, large variations in the pre-weak-
ening results were observed for various ore samples. The devel-
opment of a single-particle/single-pulse test has largely re-
moved the machine inefficiency factor in the laboratory
characterization process, yet the variations still exist. For
example, one copper ore sample achieved 81 % of particles
fragmented with one pulse treatment, but another copper ore
sample only had 40 % particles fragmented under the identical
selFrag operational conditions and for the same feed particle
size. Further research in this area is undertaken at the JKMRC
through fundamental study to investigate how ore properties
and selFrag operational conditions affect particle pre-weaken-
ing and liberation results, upon which a model for electrical
comminution will be developed.
4.4 Downstream Effects
Previous research has reported the pre-weakening effect and
the preferential liberation effect associated with HV pulses.
However, whether or not the pre-weakening effect on ore par-
ticles can result in decreased comminution energy or increased
mill throughput, and whether or not the better liberation can
lead to better recovery of valuable minerals in the downstream
processes are yet to be confirmed in the real mineral processing
operation.
Another challenge is the adverse effects of HV pulses on sur-
face chemistry or mineral phase of particles which, if exhibited,
may affect grade or recovery of valuable minerals in the down-
stream processes.
Another research program is currently being undertaken at
the JKMRC to evaluate the effects of HV pulses on down-
stream processing and how to overcome the adverse effects, if
any, in order to maximize the benefits of electrical comminu-
tion in the mineral industry.
5 Conclusions
The recent progress on electrical comminution using HV
pulses to improve grinding circuit energy efficiency is re-
viewed. To bring this technology to the mineral industry, chal-
lenges on facilities scaling-up, circuit design, cracks/micro-
cracks generation, and downstream effects are discussed.
Acknowledgment
The authors would like to acknowledge financial support from
the Australian Research Council Linkage Scheme (AMSRI –
LP0667828) and the project sponsors. Support from Anglo
American, Newcrest, Newmont, Rio Tinto, Xstrata, and Vale
mining companies is gratefully acknowledged. The collabora-
tive research partnership with SELFRAG AG is enjoyable and
appreciated.
The authors have declared no conflict of interest.
References
[1] DRET (Australian Government Department of Resources,
Energy and Tourism), Energy in Australia 2012.
[2] L. A. Yutkin, Electrohydraulic Effect, U.S. Dept. of Commerce,
Office of Technical Services, Document #62-15184, MCL
1207/1-2, 1955.
[3] K. W. Curley-Macauly, Chem. Process Eng. 1968,48, 87–92.
[4] U. Sparks, D. L. Mills, R. Warren, T. Holstein, A. A. Maradu-
din, L. J. Sham, E. Loh, D. King, Phys. Rev. B 1981,26, 3579.
[5] K. J. Touryan, W. M. Money, C. T. Aimone, J. W. Benze, Proc.
of 7th IEEE Pulsed Power Conf., Monterey, CA 1989, 69–72.
[6] U. Andres, Int. J. Miner. Process. 1977,4, 33–38.
[7] Anonymous, World Min. Equipment 1986,10, 4–19.
[8] U. Andres, Miner. Process. Extr. Metall. Rev. 1995,14 (2),
87–110.
[9] U. Andres, I. Timoshkin, J. Jirestig, H. Stallknecht, Powder
Technol. 2001,114, 40–50.
[10] R. Lastra, L. J. Cabri, P. Weiblen, in Proc. XXII Int. Mineral
Processing Congress (Eds: L. Lorenzen et al.), Cape Town
2003,1, pp. 251–260.
[11] L. J. Cabri, N. S. Rudashevsky, V. N. Rudashevsky, V. Y. Gor-
kovetz, Miner. Eng. 2008,21, 463–470.
[12] U. Andres, Int. J. Miner. Process. 2010,97, 31–38.
[13] F. Shi, A. Weh, E. Manlapig, E. Wang, Proc. of XXVI Int.
Mineral Processing Congress, Indian Institute of Mineral En-
gineers, New Delhi, 2012, Paper 627.
[14] H. Bluhm, W. Frey, H. Giese, P. Hoppé, C. Schultheiß,
R. Sträßner, IEEE Transact. Dielectr. Electr. Insul. 2000,7,
625–636.
[15] U. Andres, I. Timoshkin, M. Soloviev, Trans. Inst. Min.
Metall., Sect. C 2001,110, 149–157.
[16] E. Wang, F. Shi, E. Manlapig, Miner. Eng. 2011,24, 455–462.
[17] F. Shi, T. Kojovic, S. Larbi-Bram, E. Manlapig, Miner. Eng.
2009,22, 602–612.
[18] E. Wang, F. Shi, E. Manlapig, Miner. Eng. 2012,27–28,
28–36.
[19] E. Wang, F. Shi, E. Manlapig, Miner. Eng. 2012,34, 48–54.
[20] F. Shi, W. Zuo, E. Manlapig, Miner. Eng. 2013,50–51, 69–76.
[21] selFrag website: http://www.selfrag.com/; last accessed Sep-
tember 05, 2013.
Chem. Eng. Technol. 2014,37, No. 5, 1–6 © 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.cet-journal.com
Review 5

Review: Comminution consumes about
one third of the energy used by the
mining industry. Therefore, improved
methods for mineral comminution are
continually being sought. A particle pre-
weakening characterization method by
single-particle/single-pulse test has
been developed. Challenges in terms
of facilities scaling-up, circuit design,
generation of cracks, and downstream
effects are discussed.
Progress and Challenges in Electrical
Comminution by High-Voltage Pulses
F. Shi*, E. Manlapig, W. Zuo
Chem. Eng. Technol. 2014,37 (5),
XXX … XXX
DOI: 10.1002/ceat.201300660
Applied Voltage 100 kV
X-plane
Chalcopyrite Pyrite
5.4 kV/mm5.4 kV/mm
Water
0 kV/mm
Silicate
(4 kV/mm)
www.cet-journal.com © 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eng. Technol. 2014,37, No. 5, 1–6
6Review