Deep Mining 2014 — A.A. Editor and B. Editor (eds)
© 2014 Australian Centre for Geomechanics, Perth, ISBN 978-0-98709xx-x-x
Portable Analytical Instruments for Deep Mining
DC Arne CSA Global, Canada
A Piotrowski Northern ANI Solutions, Canada
Abstract
Modern mining requires critical geochemical and mineralogical information for mine planning and
scheduling in a timely and cost-effective manner. Non-destructive testing methods using portable analytical
instrumentation can provide this information in some situations to allow rapid decision-making
synchronised with underground mine schedules. Portable X-ray fluorescence (pXRF) is a well-established
technology that has been widely adopted throughout the resource industry to provide rapid geochemical
information for a wide range of commodity and associated elements. Portable X-ray diffraction (pXRD) is a
relatively recent innovation developed from a standard laboratory technique (XRD) that can be used to
identify a wide variety of minerals. Portable hyperspectral analysis (pHSA) in the visible to near infrared
(VNIR) and shortwave infrared (SWIR) spectrum is also a well-established technology that has been used
widely in mineral exploration to assist in the identification of hydrous, carbonate and oxide minerals but
which has seen less utilization within the mining industry. These technologies can be used independently or
integrated together to provide grade control data, identify both favourable and deleterious minerals
associated with mineralization, assist in targeting and provide near-real time data on drilling material to
allow timely decisions to be made on drill hole shutdowns.
1
Introduction
Modern mining operations require information on commodity grades in a timely and cost-effective manner
so that planning of development can proceed with minimal disruption to the mine schedule. This
imperative takes on added importance in underground operations and is compounded further at increasing
depth. The consequences of inactive headings pending assay results for face samples or drilling material
extend from the working face at depth, through material transport schedules to crushing and mill
operations. The need for the rapid turnaround of assay results in mining operations has traditionally been
addressed through the installation of an on-site laboratory or the use of visual estimates from the mine
geologists undertaking the face sampling.
An alternative to these approaches for some commodity elements is provided by the use of portable X-ray
fluorescence (pXRF). The technology has been widely adopted in exploration where it provides rapid, semiquantitative geochemical information to allow near-real time assessment of exploration targets. Its use has
resulted in the compression of field programs from several campaigns separated by the receipt and
interpretation of laboratory results, to single field campaigns in which decisions are made “on-the-fly”
based on pXRF data (eg. Arne et al. 2014). Portable XRF has also been used for grade control purposes for
some commodities (eg. Houlahan et al. 2003) where careful sample preparation and calibration protocols
can produce quantitative data that are “fit-for-purpose”.
X-ray diffraction (XRD) has been a keystone mineralogical technique that is both well-established and
widely applied by mineralogists. However, it has until recently been a technique restricted to the laboratory
given the need for a goniometer to vary the positions of the X-ray tube and/or detector with respect to the
sample in order to change the angles of incident and reflected radiation. Recent innovations by NASA with
further modifications by Olympus have produced a portable XRD device (Olympus Terra Portable XRD
System) that uses ultrasonic vibration to agitate a small amount of sample within a chamber and nonmechanical goniometers to provide quantitative XRD data similar in quality to laboratory XRD results (eg.
Uvarova et al. 2014). Quantitative XRD data can be used to identify minerals associated with hydrothermal
Event 201x, City, Country | 1
Portable Analytical Instrumentation for Deep Mining
D.C. Arne and A. Piotrowski
alteration (eg. Bierlein et al. 2000), and thus help near-mine exploration geologists vector towards
mineralization, as well as identify deleterious minerals such as serpentine that may have a negative impact
on mineral processing.
Hyperspectral analysis has been widely used in the mineral exploration industry through the use of a broad
spectrum of reflected electromagnetic radiation from visible to far infrared generated by satellites, such as
the advanced spaceborne thermal emission and reflection radiometer (ASTER), and other airborne
technologies. Portable handheld devices have been in common use for several decades, with the portable
infrared mineral analyser (PIMA) seeing widespread use in the short wave infrared (SWIR) part of the
spectrum between 1300 and 2500 nm (Figure 1). It has subsequently been superseded by other detectors
covering the visible to near infrared range (VNIR) from 350 to 1300 nm, both in portable and core scanning
devices. Applications have mainly been directed towards mineral exploration with a focus on hydrothermal
mineral deposits (eg. Arne et al. 2008). There are very few published case studies of its application to
mining or grade control, although there has certainly been research undertaken with these objectives in
mind.
Figure 1 Visible and infrared electromagnetic radiation (to be re-drafted in final version)
2
Portable X-ray Fluorescence
There are numerous potential applications for pXRF in deep underground mining operations. These range
from direct grade control to analysis of drill material in order to determine whether underground drill holes
should be shut down or continued. It first must be established whether the commodity elements of interest
are detected at the necessary levels before proceeding with pXRF. In the case of gold, the lower level of
detection (LOD) may not always provide the required accuracy at the levels necessary for grade control.
Cut-off grades are often sub-1 ppm in open pits, but the increasing costs of deep mining means that cut-off
grades may begin to fall within the realm of detection for the most recent pXRF instrumentation. Whether
the necessary precision can be obtained at the cut-off grade requires orientation testing to establish.
Alternatively, in the case of gold, it may be possible to use an associated element such as arsenic or
antimony as a surrogate provided a strong positive correlation can be established and the associated
element occurs at levels readily detectable by pXRF. An example is provided in Figure 2 from the VG Zone
discovery hole on the QV property of Comstock Metals Ltd in the Yukon Territory. The mineralized zone is
enriched in antimony, arsenic and a variety of other associated (i.e. pathfinder) elements at concentrations
higher than observed for gold. These elements then become useful surrogates for gold in the use of pXRF to
determine on site whether the drill hole has likely encountered gold mineralization, especially when
integrated with other information.
2 |Deep Mining 2014, Sudbury, Canada
Proceedings Section/Chapter
Figure 2 Comparison of gold, antimony, sodium/aluminium molar ratio, and wavelength
position of the main white mica AlOH infrared absorption peak around 2200 nm from
the discovery hole for the VG Zone discovery on the QV property of Comstock Metals
Ltd.
The LODs attainable for pXRF instrumentation continue to fall and are approaching levels comparable with
some laboratory instrumentation for some elements. Table 1 presents example LODs for a range of
elements calculated for data obtained for an Olympus Delta Premium using a Ta anode. The data were
obtained from contaminated soil samples in 3-beam soil mode using a total count time of 120 seconds. The
LODs were calculated using a value of 3 standard deviations above the mean of the background signal for
each particular element. This is normally calculated by the instrument for each individual sample as each
may have a slightly different matrix. A separate LOD calculation must be done externally from the device
for a sample population in order to provide a coherent LOD for each particular element in a data set.
The latest Delta Premium instruments use an accelerated count rate detector that captures much more of
the secondary radiation emitted by the sample compared to previous versions and allows the detection of
some elements at lower levels than previously possible. Many of these values are far lower than necessary
for many grade control situations, although it should be borne in mind that the levels of quantification
(LOQ) are approximately an order of magnitude higher than these values. They are presented here to
demonstrate the current LODs that can be obtained with modern instrumentation and to illustrate the
potential use of elements that may be associated with gold mineralization. These LODs apply only to the
data set for which they were calculated and are not indicative of the LODs that might be obtained from
other material. Each application requires specific orientation and calibration work to establish what sort of
LODs might be obtained during routine operations.
Event 201x, City, Country | 3
Portable Analytical Instrumentation for Deep Mining
Table 1
D.C. Arne and A. Piotrowski
Example lower detection limits calculated for an Olympus Delta Premium in a soil
matrix
Element
Detection Limit
Sulphur
102 ppm
Chlorine
32 ppm
Potassium
33 ppm
Calcium
69 ppm
Titanium
9 ppm
Chromium
1 ppm
Manganese
1.3 ppm
Nickel
2.6 ppm
Copper
1.2 ppm
Zinc
1.0 ppm
Arsenic
0.4 ppm
Molybdenum
0.4 ppm
Cadmium
2.8 ppm
Lead
0.7 ppm
Mercury
0.5 ppm
As with any method of grade control, the method of sampling is likely to have the greatest impact on data
accuracy. This is likely to be most extreme where mineralization is unevenly distributed within the mining
face and where exact placement of the device will strongly influence the results. Houlahan et al. (2003)
provided case studies on the use of pXRF from open pit Cu and Ni mines, as well as an underground Cu
mine, and demonstrated that data accuracy improved dramatically in samples that have been pulverized
prior to analysis. Therefore consideration must be given to some degree of sample preparation in order to
provide sufficient data precision for grade control purposes.
The other requirement for grade control usage is calibration of the instrument. This can be done using
certified reference materials (CRMs) if matrix-appropriate material is available, but is best done using
project specific samples covering the expected range of grades to be encountered. Emphasis should be
placed in providing calibration control around the anticipated cut-off grade. Care must also be taken to
ensure that laboratory analysis of the samples used for calibration provide total results for the elements of
interest. Calibrations based on linear regressions can be entered directly onto each individual device and
each device should be calibrated individually to allow for between-instrument variations. Post-processing
of the data may be required where the calibration regression is non-linear but can easily be built into a
database in order to automate the corrections.
4 |Deep Mining 2014, Sudbury, Canada
Proceedings Section/Chapter
Semi-quantitative pXRF data can be used to validate the locations of drill holes inside hydrothermal
alteration haloes based on the presence of low levels of commodity elements or, in the case of gold,
associated elements that may be more readily detectable using pXRF (Figure 2). Many of the issues raised
with respect to sampling also apply to analysis of drill core by pXRF and similar conclusions can be drawn.
For example, Somarin et al. (2012) used a modified angle grinder to obtain pulp samples (i.e. fillet samples)
down the length of drill core mineralized with base metals and silver and obtained results comparable to
those from laboratory pulp samples. Other strategies can be assessed in order to obtain data that are “fitfor-purpose” but these require a pilot study to validate. However, once an effective sampling strategy is
developed, and the accuracy and precision of the pXRF device established, the metholology can be used to
determine whether the drill hole is within mineralization or perhaps within an associated alteration halo
(eg. Arne et al., 2008).
3
Portable X-ray Diffraction
X-ray diffraction (XRD) is a standard method of identifying mineral species and extensive libraries of XRD
patterns exist for most minerals and their variants. The advent of software packages that can extract
quantitative estimates of mineral percentages from XRD data allows the information to be used both for
vectoring purposes in mineral exploration (Figure 3) as well as for geometallurgy. An example illustrating
the potential use of pXRD data is presented by the following case study.
Hydrothermal alteration associated with central Victorian, sediment-hosted gold mineralization has
been well described in recent years (eg. Bierlein et al. 2000; Arne et al. 2008). Alteration consists of
sulphidic (pyrite, arsenopyrite, pyrrhotite), phyllic (sericite replacement of detrital albite) and ferroan
carbonate (ankerite, siderite, ferroan calcite) haloes, some of which is cryptic in that it is not readily
apparent in fresh diamond drill core. A number of orientation studies have described the geochemical
and mineralogical extent of these haloes away from mineralized structures, and have demonstrated
the significant increase in the size of the deposit footprint where cryptic alteration is recognized.
Quantitative XRD analysis can be used to track the presence of ferroan carbonate minerals in metasedimentary rocks as mineralized structures are approached. The amount of chlorite in the metasedimentary rocks varies inversely with ferroan carbonate, either reflecting the replacement of
metamorphic chlorite in the host rock or a change to an outer, more distal alteration halo. The key
marker is the “crossover” in the proportions of chlorite to ferroan carbonate determined by
quantitative XRD that can be tracked effectively using the Fe carbonate/chlorite ratio in the samples.
Coincidently, this change occurs about 60 m away from the mineralized structures at both Ballarat East
and Fosterville, which are otherwise distinctly different styles of gold mineralization (quartz vein hosted vs.
disseminated). One geochemical marker that also reflects the inferred outer limit of the cryptic
alteration zone is the arsenic (As) content of the samples, particularly that of disseminated pyrite
crystals within the rock. A combined approach using both field portable XRD and XRF would effectively
identify this alteration.
Event 201x, City, Country | 5
Portable Analytical Instrumentation for Deep Mining
D.C. Arne and A. Piotrowski
Figure 3 Variations in ferroan carbonate, chlorite, arsenic and gold with proximity to gold
mineralization at Ballarat East and Fosterville deposits (data from Beirlein et al. 2000)
4
Portable Hyperspectral Analysis
Portable hyperspectral analysis (pHSA) also has the potential to provide mineralogical information in deep
mining situations. Application of the technique is restricted to certain hydrous, carbonate and oxide
minerals within the spectral range it is usually applied to (350 to 2500 nm). Of particular interest for many
hydrothermal mineral deposits is the position of the main AlOH absorption peak for white micas around
2200 nm (Figure 2). The peak position appears to be a reliable indicator of mica composition, which in turn
is related to the hydrothermal system in which the mica developed. The following case study demonstrates
the use of hyperspectral analysis on drill core samples after crushing at the laboratory prior to geochemical
analysis. However, the method could equally be applied in the field using a portable device directly on drill
core.
An example of this approach is provided by integrated hyperspectral and geochemical analysis of diamond
drill core from the recent VG Zone discovery on the QV property of Comstock Metals Ltd in the White Gold
District of the Yukon Territory of Canada (Figure 2). Selected major and trace elements from drill core
6 |Deep Mining 2014, Sudbury, Canada
Proceedings Section/Chapter
samples reflect mineralogical changes in the felsic orthogneiss host rock during hydrothermal alteration
associated with gold mineralization. The core samples were analysed using a near-total 4-acid digestion in
order to obtain data that will detect changes in silicate mineralogy during hydrothermal alteration, as well
as information on important pathfinder elements known to be enriched (or depleted) at the same time that
gold was introduced into the rocks. A split of coarse (<2 mm) material from most core samples was also
analysed using VNIR and SWIR on a HyLoggerTM operated by Bureau Veritas Minerals Pty Ltd in Australia.
The HyLoggerTM uses a similar ASD detector as that found in the portable range of analysers developed by
PANalytical (eg. TerraSpec) and so provides similar data to that generated by pHSA.
Rather than interpret the hyperspectral data simply in terms of mineralogy, spectral parameters that
characterize changes in mineral composition within hydrothermally altered rocks have been plotted in
order to demonstrate the correlation of the hyperspectral response with the lithogeochemical data. The
antimony data previously discussed extend the observed down hole width of the hydrothermal alteration
halo by an additional 80 % compared to gold data alone. The hyperspectral data increase the down hole
width of the halo by as much as 50 %, based on the shift in the peak position for the AlOH absorption peak
for white mica around 2200 nm. This zone corresponds to the presence of visible sericite alteration in the
drill core as well as the near total loss of sodium in the rock due to the breakdown of albite during
hydrothermal alteration.
While this case study is relatively straight forward to interpret (ie. intercept of interest encompasses a
single uniform lithology), complications introduced by lithological variations and structural complexities
require the integrated use of multiple data types in order to produce a rigorous interpretation. The
combination of a decrease in the main AlOH absorption peak for white mica with elevated antimony
appears to provide a reliable indicator for the presence of gold mineralization. Both responses can be
detected on site using portable instrumentation and decisions to continue or shutdown drilling could be
made based on this basis.
5
Conclusions
Portable analytical instrumentation, including XRF, XRD and HAS, can provide near-real time analytical data
for mining operations. Applications include grade control, as well as drill targeting and evaluation of drill
materials. Some of the techniques are well established (eg. pXRF and pHSA) but recent innovations have
allowed for new field portable applications (i.e. pXRD). Other portable analytical instrumentation has
recently been developed and this range of technologies will provide both the exploration and mine
geologist with a new suite of tools to further shorten the time between data acquisition and decisionmaking.
Acknowledgement
The authors would like to thank Chris Butz for providing the pXRF data used to generate the LODs for the
Delta Premium pXRF. Matt Bodnar is thanked for preparing the illustration of the pXRD data presented
here. Bureau Veritas Minerals Pty Ltd provided the hyperspectral data for the VG Zone case study and
Rasool Mohammad of Comstock Metals is thanked for allowing us to present it.
References
Arne, D C, House, E & Lisitsin, . 2008 Lithogeochemical haloes surrounding central Victorian gold deposits: Part 1 – Primary
alteration systems. Geoscience Victoria Gold Undercover Report 4.
Arne, D C, Mackie, R A & Jones, S A 2014, The use of property-scale portable x-ray fluorescence data in gold exploration:
advantages and limitations, Geochemistry: Exploration, Environment, Analysis, http://dx.doi.org/10.1144/geochem2013233
Bierlein, F P, Arne, D C, McKnight, S, Lu, J, Reeves, S., Besanko, J, Marek, J & Cooke, D 2000, Wallrock petrology and geochemistry in
alteration haloes associated with mesothermal gold mineralisation, central Victoria, Australia. Economic Geology, vol. 95,
pp. 283-312.
Event 201x, City, Country | 7
Portable Analytical Instrumentation for Deep Mining
D.C. Arne and A. Piotrowski
Houlahan, T, Ramsay, S & Povey, D, 2003, Use of field portable x-ray fluorescence spectrum for grade control – a presentation of
case studies, in Proceedings of the 5th International Mining Conference, pp. 377–385 (The Australasian Institute of Mining
and Metallurgy: London).
Somarin, A K, Lopez, R, Herrera, M & Guiza-Gonzalez, S 2012, Application of the Thermo Scientific portable XRF analyzer in
geochemical exploration: An example from the Francisco I. Madero Zn-Pb-Cu-(Ag) deposit, Zacatecas, Mexico, Proceedings
of the Geological Association of Canada/Mineralogical Association of Canada Joint Annual Meeting, St. John’s,
Newfoundland, May, 2012.
Uvarova, Y A, Cleverley, J S, Baensch, A & Verrall, M 2014, Coupled XRF and XRD analyses for rapid and low-cost characterization of
geological materials in the mineral exploration and mining industry. Explore no. 162, pp. 1-14.
8 |Deep Mining 2014, Sudbury, Canada