Journal of Geochemical Exploration 137 (2014) 11–28

Contents lists available at ScienceDirect

Journal of Geochemical Exploration

journal homepage: www.elsevier.com/locate/jgeoexp

Identification of acid rock drainage sources through mesotextural classification at abandoned mines of Croydon, Australia: Implications for the rehabilitation of waste rock repositories Anita K. Parbhakar-Fox a,⁎, Mansour Edraki b, Kathleen Hardie c, Oskar Kadletz d, Tania Hall d a Co-operative Research Centre for Optimising Resource Extraction (CRC ORE) Ltd, School of Earth Sciences, University of Tasmania, Private Bag 79, Hobart, Tasmania 7001, Australia b Centre for Mined Land Rehabilitation (CMLR), Sustainable Minerals Institute, The University of Queensland, St. Lucia, QLD 4072, Australia c Xstrata Coal, Queensland, Level 26 111 Eagle St., Brisbane, QLD 4000, Australia d Department of Natural Resources and Mines, Level 16, 61 Mary Street PO Box 15216, City East, QLD 4001, Australia article info abstract

Article history: Developing effective strategies to manage acid rock drainage (ARD) from historic and abandoned mine sites is a Received 16 January 2013 significant rehabilitation challenge. In Australia, there are more than 50,000 recorded abandoned mine sites, Accepted 31 October 2013 many of which have associated ARD and water quality issues. Traditional rehabilitation strategies focus on Available online 9 November 2013 utilising a blanket approach to management. However, if sources of ARD were instead thoroughly characterised, cost-effective management strategies based on mineralogy could be formulated, potentially enhancing site reha- Keywords: bilitation and ensuring longer-term success. Acid drainage Static tests A mesotextural method was developed to domain waste rocks into groups based on their mineralogical, textural Mineralogy and chemical similarities, using routine geological tools and field-based analytical instrumentation. This was Texture tested at the abandoned mining operations at Croydon, North Queensland, from which uncapped sulphidic Sulphide oxidation waste rock piles were sampled. Surface water and sediment samples collected from creeks up to 10 km down- stream of the site showed elevated concentrations of As, Cd, Cu, Ni, Pb, S and Zn relative to local background levels, indicating the necessity for effective rehabilitation strategies to be implemented at these sites. Ten mesotextural waste rock groups (A to J) were identified in the piles across both mine sites and comprise of hydrothermally altered rhyolites, and massive sulphides. Three major sulphide-bearing groups were identified (G, H and J). Mineralogical and geochemical data indicated that group J (quartz–pyrite) was acid forming, with pyrite containing significant concentrations of As, Pb, Zn and Cu. Pyrite was in early weathering stages with some hydrous ferric oxides observed on grain rims and fractures. Group H (arsenopyrite–quartz–pyrite) was also acid forming; with scorodite extensively precipitated in fractures and rims, likely retarding arsenopyrite ox- idation. Significant quantities of Zn and Cd were leached from Group G (quartz–sphalerite–galena) in first flush experiments, and were also measured downstream of the Glencoe site (at which the majority of group G material was identified). Microtextural analyses showed galena had partial weathering to anglesite, suggesting a potential Pb source. High concentrations of Fe and Cd (8.5 wt.% and 0.19 wt.% respectively) were measured in sphalerite, which likely encouraged oxidation, and subsequent release of Zn. Considering the diversity of the sulphide mineralogy and the associated weathering pathways, a rehabilitation strategy which focuses on segregating waste on the basis of mesotextural classification should be considered. Crown Copyright © 2013 Published by Elsevier B.V. All rights reserved.

1. Introduction 1999). Once metals enter streams, complex pH and redox dependant processes (including transformation, speciation and complexation) in- Effective management of acid rock drainage (ARD) is a significant fluence the transport and fate of metals and determine their concentra- rehabilitation challenge for abandoned mine sites. At these sites, the tions in both surface and subsurface environments (Caruso and Bishop, exposure of sulphides to water, air and microorganisms, leads to oxida- 2009). Subsequently, aquatic and terrestrial ecosystems downstream of tion and ARD generation (Egiebor and Oni, 2007; Evangelou and Zhang, mine works are at risk of significant environmental degradation (David, 1995). Under these acidic conditions, liberation of dissolved compo- 2003; Gray, 1997; Hudson-Edwards and Edwards, 2005; Luís et al., nents including heavy metals (e.g., Cd, Co, Cu, Hg, Ni, Pb and Zn) and 2009). metalloids (e.g., As, Sb) is promoted (Ashley et al., 2004; Plumlee, In Australia, there are over 50,000 registered abandoned mines which range from isolated minor surface works, to large and complex ⁎ Corresponding author. sites (Franco et al., 2010; Unger et al., 2012). Features of these sites E-mail address: [email protected] (A.K. Parbhakar-Fox). can include waste rock piles, tailings storage facilities, mineral

0375-6742/$ – see front matter. Crown Copyright © 2013 Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gexplo.2013.10.017 12 A.K. Parbhakar-Fox et al. / Journal of Geochemical Exploration 137 (2014) 11–28 processing wastes, and remains of mining infrastructure. Abandoned 142”15’ Mine Locations waste rock piles are significant sources of ARD (cf. Ashley et al., 2004; Carron River Aykol et al., 2003; Harris et al., 2003; Lottermoser et al., 2005; 1. Federation Marescotti et al., 2007; Mudd, 2005; Smuda et al., 2007; Tarras- 2. La Perouse 3. Glencoe

Tabletop Creek Wahlberg and Nguyen, 2008). Current mining practices dictate that Deadhorse Creek waste rock piles are engineered based on geochemical classifications, with waste rock classes or types defined by acid forming/neutralising characteristics (e.g., Andrina et al., 2006; Brown et al., 2009; Hutchison and Brett, 2006; Smith et al., 2009, Tran et al., 2003). However, at aban- doned mine sites, waste rock piles were not constructed in this manner 3 (Ashley et al., 2004; Harris et al., 2003; Hudson-Edwards and Edwards, 1 2 2005; Lottermoser et al., 1999), with costs of remediating associated ARD estimated at AUD$100,000 or more per hectare (Harries, 1997). Current rehabilitation strategies are responsive in their nature (i.e., only 18’10 implemented if acid rock drainage occurs). Consequently, a ‘blanket approach’ to management is adopted whereby techniques such as lime CROYDON dosing and waste rock capping are implemented, but have mixed success (e.g., Edraki et al., 2009, 2012; Gasparon et al., 2007; Gore et al., 2007; Mudd and Patterson, 2010). Alternatively, undertaking detailed and effec- tive predictive characterisation on an individual site basis may allow for the breakage of source–pathway–receptor chains (Vik et al., 2001), and 0 15km improve rehabilitation long-term. Coral Sea

The objective of this study was to develop a systematic approach to Croydon Volcanic Group Gulf of characterising waste rock pile materials and identifying ARD sources at Carpentaria Cairns abandoned mine sites. Therefore, a mesotextural classification method Esmerelda Granite based on mineralogical and textural differences observed in hand- Croydon Townsville Recent Sediments specimen samples to definewasterockgroupsisproposed.Themethod Mount Isa was tested at the abandoned Croydon gold mines, north Queensland, Rivers/Creeks Australia, from which ARD (pH b 4) is emanating as measured through Roads QUEENSLAND a local geochemical study of sediments and surface waters. Following mesotextural grouping, samples were subjected to ARD predictive tests according to the geochemistry–mineralogy-texture (GMT) approach fi proposed in Parbhakar-Fox et al. (2011). This contribution demonstrates Fig. 1. Simpli ed geology of the Croydon area showing locations of abandoned gold mine sites (after Bain et al., 1998). that through adopting a systematic mesotextural classification scheme, ARD sources are readily identified and can be prioritised for remediation as part of an effective long-term rehabilitation plan. 2.3. Geology and mineralisation

The geology of the Croydon district is dominated by the 2. Croydon mining area Mesoproterozoic rhyolitic Croydon Volcanic Group (CVG) and Esmerelda Supersuite (Fig. 1). The Croydon lode gold deposits are hosted by the 2.1. Mining history CVG, which is overlain by the Gilbert River Formation. The lodes consist of major quartz, potassium feldspar, muscovite, plagioclase, minor illite, The Croydon gold mining district is located approximately 15 km kaolinite, sulphides (pyrite, arsenopyrite, sphalerite, galena), and traces northeast of the Croydon Township and 400 km northeast of Mt. Isa, of pyrrhotite and chalcopyrite (Van Eck and Child, 1990). north Queensland (Fig. 1). Small-scale historic mining of reef gold was The CVG has been subjected to varying degrees of hydrothermal al- undertaken in the 1880 to 1890s, and modern open pit mines targeted teration, with evidence of silicification, kaolinitisation and sericitisation 2.84 Mt of ore (3.4 g/t Au) from 1981 to 1991 at two main sites: observed in wall rock adjacent to the quartz veins (Van Eck and Child, Federation/La Perouse and Glencoe (Van Eck and Child, 1990). The 1990). In terms of acid forming potential, the host rocks to mineralisation mine workings and waste rock piles have remained undisturbed since have little potential for buffering acid produced from sulphide oxidation, 1991. Currently, the Department of Natural Resources and Mines are as carbonates are notably absent. Effective silicate neutralising minerals in management of this site, with estimated rehabilitation liabilities of (e.g., biotite, chlorite and serpentinite) as defined by Bowell et al. AUD $1.8 million for the waste rock piles alone (DME, 2008). (2000) and Jambor et al. (2002) are also absent.

2.2. Physiography and climate 2.4. Site description

Theregionhasatropicalsavannahtypeclimatewithanaverage The Federation/La Perouse site consists of two pits (Federation: annual rainfall of 750 mm, much of which falls between December to 320 m × 160 m × 35 m; and La Perouse: 270 m × 180 m × 40 m), March. The average annual temperature is 33.8 °C, with maximum tem- two waste rock piles (Federation/La Perouse pile: 1.5 million m3 and peratures experienced during November to January (Fig. S1; Bureau of 35,000 m3), one stockpile (25,000 m3), heap leach pads (55,000 m3), a Meteorology, 2013). Tabletop Creek and Deadhorse Creek drain the catch dam (170 m × 65 m), a seepage collection pond (100 m × 30 m) Federation/La Perouse and Glencoe sites respectively (Fig. 1). Deadhorse and relict mining infrastructure including a crusher platform. The waste Creek is a tributary of Tabletop Creek, with the confluence approximately rock piles comprise materials ranging from boulder (N0.5 m diameter) 10 km from the mining operations. Tabletop Creek is in turn a tributary of through to coarse sand crushings (0.2–1 cm) and abundant fines the Carron River, which flows into the Gulf of Carpentaria. Much of the (b0.2 cm). The entire waste rock piles comprise approximately 70% Croydon district is used for grazing, including the immediate mine area flow-banded rhyolite, 20% red-stained rhyolites and tuffs, and 10% (DME, 2008). quartz–sulphide vein material (DME, 2008). Most of this material displays A.K. Parbhakar-Fox et al. / Journal of Geochemical Exploration 137 (2014) 11–28 13 dark to medium brownish–red Fe-staining, with metallic bluish black 28 N Mn-stains also observed. 27 07 Federation pit captures runoff and seepage from the main Federation/ 26 La Perouse waste rock pile. A catch dam was constructed below Federa- 29 tion pit for the purpose of containing seasonal overflow from the pit 06 lake. A seepage pond was constructed below this to contain and pump back seepage to the catch dam. However, during the wet season, water 25 overflows from the catch dam and seepage pond into Tabletop Creek. 05 Consequently, the catch dam is acidic with an average pH of 2.9 (DME, Tabletop Deadhorse 2008). There is also seepage from the waste rock pile directly into Federation and Tabletop Creeks during the wet season. Operations Creek Creek 04 were smaller at Glencoe with one open pit (330 m × 60 m × 25 m) 3 and a waste rock pile (483,000 m ). Seepage from this waste rock pile 24 30 03 enters Deadhorse Creek. Field observations indicate that galena and 23 sphalerite dominate the sulphide mineralogy of this pile. 22 Acid rock drainage has been established immediately downstream 21 31 (b2 km) of the mine workings, with elevated concentrations of Cd 02 (max. ~80 μg/L) and Zn (max. ~8000 μg/L) relative to the local baseline measured within 10 km of the operations. Remedial works were under- Glencoe 32 01 taken in November 2007 to improve the water quality in Federation pit, 38 through addition of 140,000 t of lime (CaO) to raise pH. Lime was also Federation/ 33 37 sprayed on the pit walls and deposited on the surface of the Federa- La Perouse 35 00 tion/La Perouse waste rock pile. Whilst initially pH values rose (pH 11 34 36 39 to 12), within two months pH values had declined to pH 3 to 4 (DME, 9 11 8 2008). Further remediation works (2009 to 2011) were conducted to 10 7 reduce the volume of contaminated seepage waters entering Tabletop 12 6 99 4 3 Creek, with the construction of the seepage pond described. Although 5 these works have reduced the volumes of seepage entering Tabletop 1 Creek, seepage water quality has not improved. Therefore, additional re- 2km 98 habilitation efforts are required and should instead focus on the identi- 35 36 37 38 2 40 fication and management of ARD sources rather than the treatment of ARD waters. Fig. 2. Plan view of stream sediment and water sample locations (with numbers given) both upstream and downstream of the Federation/La Perouse and Glencoe mine opera- 3. Materials and methods tions, Croydon.

3.1. Sampling and sample preparation comparison. For water and sediment analyses, sampling equipment and HDPE sample bottles were cleaned prior to sampling by soaking

Field work was conducted in May 2008. Hand-specimen sized them in HNO3 (trace metal grade) and rinsing in deionised (DI) water. (c.2kg)wasterocksamples(n=53)wereselectedtoprovidearange of lithologies from four different locations across the waste rock piles. 3.2. Mesotextural classification Samples were sawn, with one piece kept for textural studies, and the other jaw crushed to b5 cm (University of Tasmania (UTAS), Hobart, Previously, only three lithological groups were identified in these Australia). A split was taken, and the remaining material ground in a waste rock piles (flow-banded rhyolite, red-stained rhyolites and tuffs, ring mill to b125 μm for mineralogical and geochemical characterisation. and quartz–sulphides; DME, 2008). However, when considering the Surface water samples (n = 29) were collected directly from Table- styles of mineralisation and alteration, it is likely that additional groups top and Deadhorse Creeks, and at their confluence (Fig. 2). These samples exist. Therefore, a mesotextural classification method was developed as were collected around the district to allow for comparison of water qual- a means of identifying the major waste rock lithologies, and measuring ity upstream and downstream of the mine operations. Only duplicate their acid forming characteristics. Polished slices were prepared from samples were obtained due to the limited amount of surface water avail- each waste rock sample to facilitate the identification of primary and al- able at most locations. Additionally, samples were obtained from the sur- teration minerals and textures, which was performed using a handlens faces of Federation (35 m depth), La Perouse (40 m depth) and Glencoe and a binocular microscope. Lithologies were described, with particular (25 m depth) pit lakes. Water samples were collected for analysis of attention given to the texture (e.g., porphyritic, flow-banded), and esti- major cations and anions (unfiltered), and trace metals and metalloids mating the modal mineralogy. As the groundmass of rhyolite samples

(0.45 μm filtered) with samples preserved using 10% HNO3 at pH b 1. was fine-grained, a portable short-wave infrared (SW-IR) mineral Electrical conductivity (EC) and pH were measured in the field. Values analyser (Terraspec) was used for mineral identification. In this analysis, of pH were measured using a TPS WP-81 meter, which was calibrated three 30 mm2 spots across each polished slab were analysed, with the to pH 4 and 7 prior to each measurement. The EC was also measured spectra interpreted using The Spectral Geologist™ software. Pressed using this instrument which was calibrated at the start of each sampling powder pellets were also prepared from each sample and analysed day using a 0.01 M KCl solution. using a field-portable X-ray fluorescence instrument (InnovX X50). Stream sediment samples (n = 34) were also collected upstream Based on the mineralogical and textural differences observed in hand- and downstream of the Federation/La Perouse site, and represented specimen, samples were categorised into mesotextural groups. Ten background and ‘mine-impacted’ materials. Samples were taken at, mesotextural groups were identified (A to J), with one representative and downstream of Glencoe only. Duplicate sediment samples were col- sample from each group shown in Fig. 3. lected from the middle of streams at a depth of 0 to 10 cm. Samples One polished slice from each mesotextural group was evaluated by were dry sieved (using a stainless steel sieve) to b63 μm, with both the acid rock drainage index (ARDI), whereby textural parameters the whole and the fine fraction (b63 μm) analysed for geochemical known to influence acid formation were examined. As this is a site- 14 A.K. Parbhakar-Fox et al. / Journal of Geochemical Exploration 137 (2014) 11–28

A B Fe-ox staining qtz veins bearing msc altered disseminated py ksp phenocrysts mm-scale qtz veins qtz mscmsc msc graphite clots phenocrysts subhedral qtz mscmsc msc phenocrysts fiamme 1 cm msc 1 cm C DD 1 cm intensely msc <1cm qtz veins weathered Fe-ox rind bearing mm-scale disseminated py

msc mscmsc msc mscmsc qtz veinlets mscmsc msc msc altered ksp 1 cm qtz phenocrysts qtz phenocrystsqtz phenocrysts phenocrysts E msc altered ksp F F weathered qtz phenocrysts vein qtz msc phenocrysts msc ksp msc qtz msc phenocrysts mscmsc phenocrysts

ksp 1 cm 1 cm 1 cm H G gl qtzqtz

lithic fragments splspl massive asp

qtzqtz lithic 1 cm fragments I J 1 cm

chlchl

mscmsc pypy

weathered rind qtz qtz qtzqtz phenocrystsphenocrysts 1 cm

Fig. 3. Representative mesotextures (A to J) of the ten main lithologies observed at the Federation/La Perouse and Glencoe waste rock piles (scale bar = 1 cm). Stars indicate areas analysed by short-wave infrared, with the identified mineral phase given in italic. NB. Mineralogical descriptions of mesotextural groups are given in Table 1. Abbreviations: asp, arsenopyrite; chl, chlorite; Fe-ox, iron-oxide; gl, galena; ksp, potassium feldspar; msc, muscovite; py, pyrite; qtz, quartz; spl, sphalerite. A.K. Parbhakar-Fox et al. / Journal of Geochemical Exploration 137 (2014) 11–28 15 specific index, samples were screened prior to evaluation to define the each tested for every five samples. Both reference materials are com- ranking criteria. The ARDI evaluates sulphide content; sulphide monly used in ARD studies (e.g., Goodall, 2008; Paktunc, 2001). The alteration; sulphide morphology; neutralising mineral content; and sul- relative standard deviation calculated between the standard measure- phide mineral associations following the methodology described in ments was b5%. Parbhakar-Fox et al. (2011).Thefirst three parameters (A to C) are First flush experiments were performed on 2 kg samples (b10 mm ranked from 0 to 10; and the latter two (D and E) −5 to 10. Scores and b4 mm) from mesotextural groups E (flow-banded rhyolite con- were averaged across each sample to calculate an overall ARDI value. taining disseminated sulphide), G (quartz−sphalerite−galena), H To refine the ARDI value, evaluations were performed on a polished (arsenopyrite−quartz−pyrite) and J (quartz−pyrite). Samples were thin section made from the slab using a petrographic microscope. loaded into a Buchner funnel with water added until the surface of the Values obtained from both hand-specimen and thin section analyses material was saturated. The pH and EC of leachates were measured were averaged to obtain a final score. Samples scoring from 41 to 50 after 24 h (using the same instruments as with paste pH testing, and fol- were classified as extremely acid forming (EAF); 31 to 40 as acid lowing the same calibration methods). Selected trace elements (e.g., Ag, forming (AF); 21 to 30 as potentially acid forming (PAF); 11 to 20 are Al, As, Ba, Be, Cd, Co, Cr, Cu, Fe, Mn, Ni, Pb, Sb, Se, Zn) were determined non-acid forming (NAF); 1–10 are NAF or have a potential neutralising using ICP-MS techniques (Agilent 4500 ICP-MS; UTAS, Australia). capacity (PNC); and −10 to 0 have an acid neutralising capacity (ANC). During ICP-MS analysis, three internal standards and a multi-element These values are recommended for use alongside static geochemical calibration standard (MISA29; Choice Analytical) were used before, data to enhance waste classification (Parbhakar-Fox et al., 2011). and at the end of each sample run, in addition to two blank samples. The relative standard deviation for these data was b5%. 3.3. Chemical and mineralogical analyses

The bulk elemental composition (major: Al2O3,CaO,Fe2O3,K2O, MgO, 3.4. Textural analyses MnO, Na2O, P2O5,PbO,SiO2,TiO2;trace:Ag,As,Bi,Cd,Cu,Ni,Pb,S,Sb,Zn, Zr) of all waste rock samples was assessed by X-ray fluorescence (XRF; Samples from mesotextural groups C, E, G, H and J were selected Philips PW1480 X-ray Spectrometer, UTAS, Australia). In-house stan- for microtextural analysis (FEI Quanta 600 environmental scanning dards (TASGRAN, TASBAS, TASMONZ, TASDIOR, and a blank) were electron microscope (ESEM); Central Science Laboratory (CSL), analysed during this run, in addition to standard reference materials UTAS, Australia). Relationships between primary sulphides and sec- (i.e., BCR-2, BHVO-2, RGM-1, W-2, WS-E, AC-E, GSP-2). These standards ondary minerals (e.g., scorodite, anglesite, rhomboclase which were were run at the start of the analysis, and at the end with the relative identified by XRD; Table 1), were examined in this analysis. Addi- standard deviation calculated as b1.5%. One sample from each tionally, inhomogeneties, which may influence trace element distri- mesotextural group was analysed for their mineralogical composition bution (e.g. compositional zoning or mineral inclusions), were by quantitative XRD (Siemens D501 diffractometer, University of Ballarat, observed. Additionally, one sample from each of these groups (i.e., C, Australia). E, G, H and J) were subjected to textural mapping to examine sulphide One inch polished laser mounts (n = 15) were prepared from mineral associations (FEI Quanta 600 mineral liberation analyser scan- major sulphide bearing mesotextural groups (G, H and J), and electron ning electron microscope (MLA-SEM), CSL, UTAS, Australia). One inch probe microanalysis (EPMA) performed (Cameca SX100 electron mi- polished tiles (3 cm × 3 cm) were prepared and analysed using the ex- croprobe; Central Science Laboratory (CSL), UTAS, Australia). Natural tended back scattered electron (XBSE) technique as described by and synthetic materials were used as standards and included Fandrich et al. (2007). Data were processed in MLA Image Viewer and sphalerite-ast modified (S, Zn), marcasite (Fe), Cd–metal-UTAS4 (Cd), in-house Texture Viewer software to produce classified images for cuprite–ast modified (Cu) and astimex. Spot analyses were performed each sample based on a site-specific mineral library. operating with a 20 keV accelerating voltage, 15 nA beam current and a 2 μm beam diameter to measure concentrations of minor ele- ments (e.g., As, Cd, Co, Ni) for comparison with Laser-Ablation ICP- 3.5. Sediment analyses MS (LA-ICPMS) data. In the case of sphalerite, Zn measurements were used as an internal standard. Both spot and mapping analyses Selected sediments were analysed for their mineral composition by were performed on these samples using LA-ICPMS (Agilent HP4500 XRD powder diffraction (Bruker D8 Advance X-Ray diffractometer; Quadripole ICPMS; UTAS, Australia). Calibration was performed UQ, Brisbane, Australia). Both the whole sediment and the fine- using an in-house standard (STDGL2b-2) comprising powdered grained (b63 μm) sediment fraction were partially digested in hot sulphides doped with certified element solutions and fused to a lithium aqua regia. The resulting extractants and the water samples collected borate glass disc (Danyushevsky et al., 2011). These analyses were per- from around the Croydon district were analysed for selected trace ele- formed to quantify trace elements in sulphides and observe their spatial ments (e.g., Ag, Al, As, Ba, Be, Cd, Co, Cr, Cu, Fe, Mn, Ni, Pb, Sb, Se, Zn) distribution. using ICP-MS techniques (UQ, Brisbane, Australia). During ICP-MS anal- Static acid base accounting (ABA) tests were performed on all waste ysis, two reference standards were used, a multi-element calibration rock samples (at both UTAS and University of Queensland (UQ), Bris- standard-2A (Agilent Technologies) and an arsenic reference standard bane, Australia), and included paste pH, Sobek, modified Sobek and a (Eawag aquatic research, Swiss Federal Institute of Aquatic Science range of standard and advanced net acid generation (NAG) tests, follow- and Technology). Both were analysed before and at the end of each ing procedures given in White et al. (1999) and the AMIRA P387A Hand- sample run in addition to two blank samples. The relative standard de- book (Smart et al., 2002). During paste pH testing, solutions were viation for these data was less than 10%. measured in triplicate (per sample) using a Eutech Instruments 510 Determination of the solid speciation of selected metals, was per- pH meter. The pH probe was calibrated to pH 4 and 7 using standard formed through a six-step sequential extraction analysis (Centre for buffer solutions (Merck Ltd.) after each sample measurement. Sample Mined Land Rehabilitation (CMLR), UQ, Australia) on Fe-rich stream blanks (deionised water) were tested before and at the end of each sediments (n = 6) collected at the Federation/La Perouse site only. sample batch. The EC was measured using a TPS WP-81 meter, with The analytical procedure of Dold (2003) was followed, which differenti- the probe calibrated prior to use with a 0.01 M KCl solution. During ated between water soluble, ion exchangeable, Fe3+ oxyhydroxide, Sobek and modified Sobek testing, KZL-1 (sericitic schist) and Fe3+ oxide, organic/sulphide and residual fractions. An additional step

NBM-1 (altered feldspar porphyry) standards (obtained from CANMET, was added to include Mn oxides (20 mL 0.1 M NH2OH−HCl; pH 2; Natural Resources, Ottawa, Canada) were used, with one sample of shake 2 h). 16 A.K. Parbhakar-Fox et al. / Journal of Geochemical Exploration 137 (2014) 11–28

Table 1 4. Results Mineralogy and Croydon waste rock samples as measured by quantitative X-ray diffrac- tometry (QXRD). One sample is shown per mesotextural group (A to J). 4.1. Mineralogy and textural groups Mesotextural Lithological Gangue Primary Secondary group description minerals sulphides sulphates and The mineralogy of each mesotextural group is summarised in Table 1. (hydro)oxides From the ten mesotextural groups, four were porphyritic rhyolites A Porphyritic pale- Major Minor Minor (groups A, C, F and I), three were flow-banded rhyolites (groups B, D mid grey rhyolite Quartz Pyrite Szmolnokite and E) and the remainder quartz–sulphides (groups G, H and J; with sub-cm mus- Muscovite Chalcopyrite Anglesite covite pheno- Kaolinite Galena Scorodite Table 1). The porphyritic rhyolite groups differed in their component min- crysts with mm- Minor eral proportions, presence/absence of quartz veinlets, phenocryst and al- scale quartz vein- Fluorite teration type. For example, in group C textural evaluations identified trace lets. sulphides (b1 wt.%) in both the groundmass and quartz veinlets. This B Flow-banded blue- Major Minor Major consisted of euhedral–subhedral quartz-associated disseminated grey rhyolite with Quartz Arsenopyrite Goethite b sub-rounded Muscovite Pyrite Minor ( 1 mm) pyrite and arsenopyrite, but similar sulphides were not quartz phenocrysts Kaolinite Galena Anglesite observed in other porphyritic rhyolite groups. Additionally, whilst with graphite clots Szmolnokite flow-banded rhyolite groups B and D are similar in appearance; group present. Gypsum B had been more silicified, and group D weathered, so were classified Rhomboclase Scorodite differently. C Porphyritic light- Major Major Minor Mesotextural groups G, H and J differed significantly in terms of their mid grey rhyolite- K-feldspar Pyrite Anglesite sulphide mineralogy and texture (Fig. 4). Group G was dominated by tuff with quartz Quartz Minor Gypsum sphalerite and galena (Fig. 4A), with minor pyrite also identified. In gen- phenocrysts. Sub Muscovite Arsenopyrite Szmolnokite eral, larger (N500 μm) sphalerite grains appeared more weathered than cm-scale quartz Minor Chalcopyrite Rhomboclase veinlets containing Fluorite Galena smaller grains, with secondary minerals pervasively developed within mm-scale pyrite. grains. Galena alteration was observed, with fine-grained anglesite D Flow-banded dark Major Minor Minor identified as the alteration product (Fig. 4B). Sphalerite containing galena grey rhyolite with Quartz Pyrite Szmolnokite inclusions appeared strongly weathered (Fig. 4C). This is likely the result mm-scale quartz K-feldspar Galena Anglesite veins & mm-scale Muscovite Rhomboclase of galvanic interactions between these sulphides, with sphalerite prefer- graphite clots. Kaolinite entially weathering due to its lower rest potential (−0.24 V) relative to Minor galena (0.28 V; Kwong et al., 2003; Lottermoser, 2010). Fluorite Mesotextural group H displayed a massive arsenopyrite–quartz– Chlorite pyrite texture with scorodite extensively precipitated at the inter- EFlow-banded Major Minor Minor beige-grey rhyo- Quartz Pyrite Anglesite face of these minerals, and within fractures (Fig. 4D). Euhedral pyrite lite containing K-feldspar Chalcopyrite Szmolnokite grains appeared relatively unweathered when encapsulated in scorodite. mm-disseminated Muscovite Galena Scorodite However, when intergrown with arsenopyrite, pyrite had weathered to a pyrite. Minor Rhomboclase greater degree. Galena micro-inclusions were a common feature within Fluorite b μ Chlorite pyrite (Fig. 4E). Smaller ( 200 m) quartz-associated arsenopyrite grains FPorphyriticmid-Major Minor Minor appear unfractured and unweathered. Scorodite layers within massive grey-pink Quartz Galena Szmolnokite arsenopyrite had a relatively uniform thickness (c. 20 μm), but was occa- rhyolite-tuff with K-feldspar Pyrite Anglesite sionally observed as spherules (Fig. 4F). Weathering of scorodite to amor- mm-quartz phe- Muscovite Rhomboclase phous ferric arsenate phases rich in As, Cu, Fe and Pb was also recognised nocrysts and cm- Minor scale quartz vein- Fluorite (cf. Murciego et al., 2011). ing. Chlorite Pyrite was observed as both grains and very fi0ne (b10 μm) veinlets G Massive quartz Major Major Major in group J (Fig. 4G). Smaller, euhedral grains containing galena micro- with sub cm-scale Quartz Sphalerite Szmolnokite inclusions (Fig. 4H) were less weathered than larger euhedral– sphalerite and Muscovite Galena Anglesite galena inter- Minor Minor Minor subhedral pyrite grains which were highly fractured (Fig. 4I). Secondary growths and Albite Chalcopyrite Gypsum products of pyrite oxidation (i.e., coatings) were not frequently Minor mm-scale K-feldspar Scorodite observed. pyrite. Kaolinite Rhomboclase H Massive quartz Major Major Minor 4.2. Waste rock chemistry with cm-scale Quartz Arsenopyrite Scorodite arsenopyrite- Apatite Pyrite Szmolnokite pyrite inter- Hematite 4.2.1. Major and trace elements per group

growths and mm- Anglesite All groups were dominated by SiO2 (55 to 87 wt.%) except in group scale disseminated Rhomboclase H, where Fe2O3 dominated (36 wt.%) with values measured by XRF. The galena. feldspar–biotite model ((2Ca + Na + K) / Zr vs. Al / Zr) of Downing IPorphyriticblue– Major Minor Minor grey rhyolite, Quartz Pyrite Szmolnokite and Madeisky (1997) was constructed to assess, based on whole rock silicified. Microcline Galena Anglesite chemistry, the buffering potential of each group (Fig. S2). This model al- Albite Chalcopyrite Rhomboclase lows for an assessment of alteration based on the stoichiometrically de- Muscovite fined co-variation of the sum of the alkalis with Al in anorthite Chlorite Minor (CaAl2Si2O8), albite (NaAlSi3O8), orthoclase (KAlSi3O8) and biotite Fluorite [K(Fe,Mg)3AlSiO10(OH)2] used. Unaltered rocks plot on the model line Kaolinite of slope 1, and altered rocks plot either above or below the model line, J Massive quartz Major Major Major depending on whether they gained or lost alkalis during alteration pro- containing cm- Quartz Pyrite Rhomboclase cesses (Downing and Madeisky, 1997). Group I was the least altered, scale pyrite with Minor Minor Szmolnokite cm-scale Fe-oxide Fluorite Galena Anglesite plotting above the unaltered feldspar line. All other groups were altered, weathering rind. Muscovite Minor plotting below the model line indicating no significant buffering poten- Magnetite Gypsum tial. The model suggests that groups G, H and J have undergone extreme A.K. Parbhakar-Fox et al. / Journal of Geochemical Exploration 137 (2014) 11–28 17

AA DB G py py

gl sc

Fig.6g qtz sp asp msc qtz

5050mm mm qtz 5050mm mm 50 mm

sphalerite pyrite galena anglesite arsenopyrite pyrite galena scorodite galena muscovite potassium feldspar

quartz muscovite kaolinite unknown quartz muscovite unknown invalid pyrite quartz secondary sulphate C CE BF AH BI B qtz C E F gl H I gl qtz gl asp sphericule py gl sc qtz py gl spl sc qtz gl ang ribbons 20 µm 50 µm gl py 50 µm py 100 µm qtz 100 µm 500 µm

Fig. 4. Classified mineral map of MLA tile (3 cm × 3 cm) and BSE images from mesotextural groups G, H and J: (A) Classified XBSE mineral map image of mesotextural group G material; (B) BSE image of altered reaction interface between galena and anglesite; (C) oxidised sphalerite grains (skeletal grain outlined) intergrown with galena; (D) classified XBSE mineral map image of mesotextural group H material; (E) massive arsenopyrite and pyrite; (F) scorodite microtextures (ribbons, spherules and masses) identified in pyrite; (G) classified XBSE mineral map image of mesotextural group J material; (H) unweathered pyrite with galena micro-inclusions; (I) highly fractured pyrite with galena micro-inclusions. Abbreviations: ang, anglesite; asp, arsenopyrite; gl, galena, msc, muscovite; py, pyrite; qtz, quartz, sc, scorodite; spl, sphalerite.

1,000,000 1000 100,000 A B 10,000 100 1000 100 10 As (ppm) Cd (ppm) 10 1 1 024681012141618 024681012141618 S (wt. %) S (wt. %)

100,000 1000 D C 10,000 100 1000

100 10 Pb (ppm) Cu (ppm) 10

1 1 0 2 4 6 8 1012141618 024681012141618 S (wt. %) S (wt. %)

100,000 E 10,000

1000 A B C D E 100 Zn (ppm) 10 F G H I J

1 024681012141618 S (wt. %)

Fig. 5. Concentrations (ppm; measured by XRF) of As, Cu, Cd, Pb and Zn shown against S (wt.%) for Croydon waste rock materials grouped by mesotextural characteristics (A to J). 18 A.K. Parbhakar-Fox et al. / Journal of Geochemical Exploration 137 (2014) 11–28 acid leaching as several samples plotted along the X-axis (cf. Downing Bethke, 1987; Cook et al., 2009). Lead was present in veins (Fig. 6E) as and Madeisky, 1997). Field evidence for this manifested as vuggy galena intergrowths. A slight decrease in Zn concentration at the grain quartz–sulphide textures observed in waste rock materials belonging boundary was observed (Fig. 6G), implying the dissolution of sphalerite to these particular sulphide-dominated groups. to form a Zn-deficient layer (Cook et al., 2009; Weisener et al., 2004). Concentrations of As, Cd, Cu, Pb and Zn (measured by XRF and given Galena present in group G (subjected to LA-ICPMS spot analysis; in ppm) were plotted against sulphur (wt.%) per mesotextural group n = 61; data not shown) was identified as relatively enriched in Bi (Fig. 5A–E) to demonstrate their concentration ranges within these (maximum: 1508 ppm; average: 454 ppm; sd: 225 ppm) and Sb (max- groups. Average concentrations of Bi, Ni and Sb were b100 ppm for all imum: 1026 ppm; average: 867 ppm; sd: 185 ppm), with similar ele- groups and are therefore not shown. Mesotextural groups A to F and I ment signatures reported in Diehl et al. (2008). Element mapping contained low concentrations of As, Cd, Cu, Pb, S and Zn. In relative indicated no trace element zonation, and is therefore consistent with terms, Cd, Pb and Zn measured high in group G (Fig. 5B, D and E), As observations made by Bethke and Barton (1971) for lead sulphides as- in groups H and J (Fig. 5A), and Cu in groups G, H and J (Fig. 5C). One sociated with igneous activity, or heated after formation (i.e., orogenic group J sample contained high Pb (~35,000 ppm), but relatively low S areas). (Fig. 5D) for the group, suggesting that the Pb is possibly hosted by a Electron microprobe spot analyses of both larger arsenopyrite grains secondary hydrous ferric oxide (HFO) phase in this group. Zinc concen- in mesotextural group H (n = 29) reported concentrations of Cd, Cu, trations were relatively low in groups H and J (Fig. 5E). The greatest Co, Ni, Pb, Sb and Zn below detection limit. However, LA-ICPMS anal- quantities of S were measured in groups G, H and J, confirming the yses (spot: n = 11, mapping: n = 3) showed that in these grains Sb presence of sulphides as they are the major sulphur bearing mineral (maximum: 200 ppm; average: 140 ppm; sd: 31 ppm) is pervasively group identified by XRD (Table 1). Based on this, the abundance of distributed, with minor Co (maximum: 27 ppm; average 10 ppm; sd: several environmentally significant elements (including As, Cd, Cu, Pb 5 ppm) and Ni (maximum: 5 ppm; average 1 ppm; sd 2 ppm) demon- and Zn) was investigated further in sulphides from mesotextural groups strating a banded distribution (Fig. 7). Concentrations of Cd, Pb and Zn G, H and J. This was performed to allow for accurate determination of the are relatively high in grain fractures and rims compared to arsenopyrite element contents and distribution in sulphides in order to identify con- and quartz, potentially indicating adsorption of these elements to sec- trols on sulphide oxidation. ondary scorodite (Fig. 7). Element distribution maps and spot analyses (n = 7 and n = 57 re- 4.2.2. Sulphide mineral chemistry spectively; analysed by LA-ICPMS) were performed on pyrite grains in Sphalerite in group G is iron rich, with EPMA spot analyses (n = 58) mesotextural group J, with a representative example shown in Fig. 8. measuring average contents of 8.4 wt.% Fe (sd: 1.28 wt.%) and The BSE images (Fig. 8A and B) and Fe-distribution map (Fig. 8F) show 0.19 wt.% Cd (sd: 0.05 wt.%). The bulk chemical composition was calcu- the grain outline, and the absence of fractures. Arsenic was relatively lated as (Zn0.85,Fe0.15)S. These grains are likely to be relatively suscepti- enriched in areas of the grain (Fig. 8C; maximum: 19,080 ppm; average: ble to weathering (compared to trace element poor sphalerite; cf. 3800 ppm; sd: 5770 ppm), and showed an antithetic distribution to the Weisener et al., 2004; Stanton et al., 2006). Distributions of Cd and Fe metals. Pyrite (particularly towards the core) was rich in Cu (Fig. 8E; max- were homogeneous, with both pervasively distributed across the grain imum: 8290 ppm; average: 430 ppm; sd: 1139 ppm), Pb (Fig. 8G; maxi- as shown by LA-ICPMS mapping, with an example shown in Fig. 6. mum: 3164 ppm; average: 150 ppm; sd: 446 ppm) and Zn (Fig. 8H; This implies that Fe and Cd are in solid solution, as is typical for these el- maximum: 4720 ppm; average: 820 ppm; sd: 1499 ppm). Galena ements (Cook et al., 2009). Sub-5 μm blebs of chalcopyrite (Fig. 6C) micro-inclusions as shown in BSE images (Fig. 4H), were again observed characteristic of chalcopyrite disease were recognised (cf. Barton and with localised highs, and coincident distribution of Ag and Bi

cps cps cps 1e6 1e5 1e6 A B C 1e5 D 1e4 1e4 1e5 spl 1e3 1e3 1e4 gl qtz 1e2 1e2 10 10 1e3 1000 µm Cu Fe Cd 1 1

cps cps cps 1e9 1e7 E 1e8 1e7 F G 1e6 1e6 1e5 1e5 1e5 1e4 1e4 1e3 1e2 1e3 10 1e4 Pb S Zn 1e2 1

Fig. 6. Relative element distribution (cps; analysed by LA-ICPMS) in a sphalerite grain (with galena intergrowths) from Croydon waste rock mesotextural group G: (A) Back scattered electron image of the original grain; (B) Cd; (C) Cu; (D) Fe; (E) Pb; (F) S; (G) Zn. Abbreviations: gl, galena; qtz, quartz; spl, sphalerite. A.K. Parbhakar-Fox et al. / Journal of Geochemical Exploration 137 (2014) 11–28 19

ABqtz qtz C asp asp asp qtz 50 µm 200 µm 1.5 cm

cps cps cps 1e5 1000 1000 1e4 Cu 100 1e3 100

10 1e2 10

1 10 1 cps cps cps

1e5 1e5 1e6 1e4 1e5 Co 1e4 1e3 1e4 1e3 1e2 1e3 10 1e2 100 1 cps cps cps 1e5 1e5 1e4 1e4 1e4 1e3 Ni 1e3 1e3 1e2 1e2 1e2 10 10 10 1 1 1 cps cps cps

1e5 1e5 1e6 1e4 1e5 1e4 Sb 1e3 1e4 1e3 1e2 1e3 10 1e2 1e2

cps cps cps 1e7 1e5 1e6 1e5 1e5 Zn 1e4 1e4 1e4 103 1e3 1e3 1e2

Fig. 7. Back scattered electron image (A), reflected light image (B) and photograph (C) of three arsenopyrite grains from Croydon waste rock mesotextural group H, with relative element distribution maps for Cu, Co, Ni, Sb and Zn shown (cps; analysed by LA-ICPMS). Abbreviations: asp; arsenopyrite; qtz, quartz.

(b10 ppm). These Zn–Pb–Cu zones were rimmed by a fine band of (PAF). These three groups also returned relatively high average values

Co (~10 ppm). for total-sulphur (STotal) of 1.21 wt.%, 14.10 wt.% and 10.5 wt.% respec- tively. The remainder of groups returned STotal values below b0.25 wt.%. 4.2.3. Static geochemical tests Accordingly, groups G, H and J returned relatively high MPA values

A summary of static geochemical results are shown in Table 2.Aver- (approximate range 37 to 431 kg H2SO4/t; Table 2). Sobek ANC age paste pH values were NpH 5.5 for all groups except G (pH 4.5); H values were low for all groups (approximately −4to5kgH2SO4/t; (pH 5.09) and J (pH 3.76), classifying these as potentially acid forming Table 2), reflecting the absence of carbonates. Additionally, silicate 20 A.K. Parbhakar-Fox et al. / Journal of Geochemical Exploration 137 (2014) 11–28

A B Py Py Qtz Qtz LA-ICPMS pits 2000 µm

cps cps 1e5 C 1e5 D 1e4 1e3 1e4 100 10 1 As Co cps cps 1e6 1e6 E 1e5 F 1e5 1e4 1e4 1e3 1e3 100 100 Cu Fe cps cps 1e5 1e6 G H 1e4 1e4 1e3 1e2 1 100 Pb Zn

Fig. 8. Relative element distribution (cps; analysed by LA-ICPMS) in a pyrite grain from Croydon waste rock mesotextural group J: (A) Back scattered electron image of grain; (B) back scattered electron image of mapped area; (C) As; (D) Co; (E) Cu; (F) Fe; (G) Pb; (H) Zn. minerals which may contribute to ANC (e.g., serpentinite, chlorite, 4.2.4. First-flush chemistry olivine; Jambor et al., 2002, 2007)werenotidentified in these samples. Using the same cut-off criterion for paste pH (whereby leachates

Average NAPP values N20 kg H2SO4/t were calculated for groups G, H measuring NpH 5.5 are considered PAF) all samples from groups E, G, and J only, identifying these as PAF (Skousen et al., 2002), as confirmed H and J were classified as PAF (Fig. 9A; range: pH 3.0 (group J, in the NAPP against NAG plot shown in Fig. S3. Several samples from b4 mm) to 4.7 (group E, b10 mm fraction)). This indicates the potency group C were identified as PAF (Fig. S3). However, ARDI assessments of pyrite to generate acid in the absence of neutralising minerals for this considered this group overall as NAF (Table 2). Several samples from type of mine waste even when pyrite is present in very low concentra- group E plotted in the PAF field (Fig. S3), with the ARDI value for this tions i.e., group E where b0.5 wt.% was measured by QXRD. No signifi- group supporting this classification. cant differences between pH values from the finer (b4mm)and

Table 2 Static test geochemical data for materials representative of Croydon waste rock mesotextural groups A to J, with average (avg) and standard deviation (sd) values shown (where appropriate) for each group. (* = kg H2SO4/t). Abbreviations: ANC, acid neutralising capacity; (M)NAG, (multi-addition) net acid generation; ARD, acid rock drainage.

Group Paste pH Total-S (%) Maximum potential acidity* Sobek ANC* Net acid producing potential* (M)NAG* NAG pH ARD index (/50) No. of samples

A avg: 6.91 avg: 0.14 avg: 4.14 avg: 4.81 avg: −0.67 0avg:5.2082 sd: 0.07 sd:0.19 sd:5.85 sd:0.64 sd:6.49 sd:0.19 B avg: 7.51 avg: 0.02 avg: 0.74 avg: 1.17 avg: 12.57 0avg:5.3284 sd: 0.86 sd: 0.02 sd: 0.85 sd: 2.73 sd:27.51 sd:0.25 C avg: 6.12 avg: 0.25 avg: 7.55 avg: 0.49 avg:7.06 avg: 3.44 avg: 3.54 16 10 sd: 0.82 sd: 0.23 sd: 6.89 sd: 3.2 sd: 8.44 sd: 0.49 sd: 0.96 D avg:7.65 avg: 0 avg: 0 avg: 4.01 avg: −4.01 0avg:4.8982 sd: 0.18 sd: 0 sd: 0 sd: 2.19 sd: 2.19 sd: 0.10 E avg: 6.37 avg: 0.13 avg: 3.84 avg: 2.01 avg: 1.84 avg: 1.33 avg: 4.53 21 5 sd: 1.70 sd: 0.17 sd: 5.26 sd: 3.48 sd: 8.52 sd: 0.26 sd: 1.39 F avg: 7.38 avg: 0.02 avg: 0.76 avg: 3.63 avg:-2.87 avg: 0.07 avg: 5.24 813 sd: 0.70 sd: 0.05 sd: 1.52 sd: 1.54 sd:1.91 sd:0.22 sd: 0.59 Gavg:4.5avg:1.21 avg: 36.88 avg: −0.44 avg: 37.31 avg: 13.79 avg: 2.69 22 3 sd: 1.12 sd: 0.96 sd: 29.32 sd: 2.46 sd: 29.85 sd: 8.49 sd: 0.67 H avg: 5.09 avg: 14.1 avg: 431.47 avg: 1.84 avg: 429.97 avg: 487.8 avg: 1.46 41 2 sd:0.06 sd: 0.28 sd: 8.67 sd: 2.60 sd: 5.60 sd: 3.21 sd: 0.21 I avg: 7.39 avg: 0.01 avg: 0.36 avg: 4.43 avg: −4.07 avg: 6.55 avg: 5.73 92 sd: 0.82 sd: 0.01 sd: 0.50 sd: 0.01 sd: 0.50 sd: 0.92 sd: 1.03 J avg: 3.63 avg: 10.5 avg: 321.47 avg: −3.76 avg: 261.66 avg: 144.54 avg: 1.73 43 10 sd: 1.05 sd: 4.58 sd: 140.20 sd: 4.42 sd: 158.02 sd: 75.61 sd: 0.09 A.K. Parbhakar-Fox et al. / Journal of Geochemical Exploration 137 (2014) 11–28 21

pH 4,000 5 A B Al 4 3,000 3 2,000 pH 2 mg/L 1,000 1 0 0 EGHJ1J2J3 EGHJ1J2J3

15,000 140 D C 120 As Cd 10,000 100 80

mg/L 60 mg/L 5,000 40 20 0 0 EGHJ1J2J3 EGHJ1J2J3

1,500 120,000 E 100,000 F Cu Fe 1,000 80,000 60,000 mg/L mg/L 500 40,000 20,000 0 0 EGHJ1J2J3 EGHJ1J2J3

8,000 30,000 G 25,000 H 6,000 Pb Zn 20,000 15,000 4,000 mg/L

mg/L 10,000 2,000 5,000 0 0 EGHJ1J2J3 EGHJ1J2J3 Sample Sample

Size fraction < 4 mm < 10 mm

Fig. 9. First flush chemistry (mg/L; measured by ICP-MS) of leachates derived from material representative of Croydon waste-rock mesotextural groups E, G, H and J: (A) pH; (B) Al; (C) As; (D) Cd; (E) Cu; (F) Fe; (G) Pb and (H) Zn. Values for both the b10 mm and b4 mm size fractions are shown. coarser (b10 mm) fractions were measured (Fig. 9A). However, a greater For group J, the b4 mm fraction consistently measured highest reactivity was expected from the coarser (b10 mm) fraction as a function values for each element, with the exception of Pb in sample 3 of surface area (cf. Stomberg and Banwart, 1999), and this was observed (Fig. 9G). Of the group J samples, J1 contained the highest As (b4mm: in kinetic column leach experiments performed on these materials ~210 mg/L; Fig. 9C),followedbygroupJ3(b4 mm: ~130 mg/L; (Parbhakar-Fox et al., 2013). The lowest elemental concentrations (mea- Fig. 9C) and sample J2 the lowest (b10 mm: ~2 mg/L; Fig. 9C). These sured by ICP-MS) were from Group E, with the lowest quantities of Al, As, values increase with pyrite content, inferring As release during oxida- Cd, Cu, Fe, Pb and Zn measured (Fig. 9). Such a first-flush leachate signa- tion, and its consequent redistribution to soluble secondary minerals. ture was anticipated for this group when considering the low pyrite abun- The presence of significant sulphide sources of Cd (i.e., sphalerite) in dance, and its textural form (i.e., euhedral pyrite encapsulated in a this group was not inferred by this data (as confirmed by QXRD; quartz–muscovite groundmass). The highest values for Al and Zn were Table 1). However, Cd concentrations ~120 mg/L were measured from measured from the b4 mm fraction of group G (Fig. 9B; 3626 mg/L and the b4 mm fraction of sample J3 (Fig. 9D). This suggests that secondary 9 g; 26, 260 mg/L respectively). High concentrations of Zn in first-flush minerals present in this group are Cd-bearing. Irwin et al. (1997) and leachate was anticipated when considering the presence of sphalerite in Tauson et al. (2004) reported of Cd-bearing anglesite, therefore this is this material (Table 1). In group H the highest As concentrations were considered here a potential Cd source in material from this group. measured (~14,000 mg/L; Fig. 9C) corresponding to the high arsenopy- High concentrations of Cu were measured in leachate from samples J1 rite and scorodite contents (Table 1). Generally, relatively similar element (b4mm:~2900mg/L;Fig. 9E) and J3 (b4mm:~1500mg/L;Fig. 9E), concentrations were measured between the two size fractions, however and likely relate to the presence of Cu in pyrite (Fig. 8E), furthermore, slightly higher As, Cd, Pb and Zn were measured from the b4mmfraction minor quantities of chalcopyrite were identified in SEM and MLA stud- (Fig. 9). High concentrations of Pb were measured from both size frac- ies. The highest concentrations of Fe (range: ~133,780 to 997,800 mg/L; tions, and were likely sourced from Pb micro-inclusions identified in Fig. 9F) were measured in this group, specifically from the b4mm pyrite intergrown with arsenopyrite (Fig. 4D). fraction, and likely represents the dissolution of soluble secondary 22 A.K. Parbhakar-Fox et al. / Journal of Geochemical Exploration 137 (2014) 11–28

Fe-sulphates, which were identified by QXRD (Table 1). Measurements value) and the ISQG-High concentration. The trigger value is a threshold of Pb are linked to weathering of galena to anglesite with both mea- concentration, and below this the frequency of biological effects is ex- sured the highest in sample J3 (Fig. 9G). Zinc is most likely sourced pected to be very low (Simpson et al., 2005). The ISQG-High concentra- from pyrite and its alteration products rhomboclase and anglesite tion is intended to represent a concentration, above which adverse which were detected in minor quantities (Fig. 9H; Table 1; cf. Buckby biological effects are expected to occur more frequently (Simpson et al., et al., 2003; Giere et al., 2003; Nordstrom, 2004). 2005). The upstream (local baseline), on site, and downstream concentra- 4.3. Stream sediments tions of selected metals and metalloids in both the total and fine (b63 μm) fractions are shown (for both mine sites) in Fig. 10 (measured Element contents of stream sediment samples (measured by ICP-MS) by ICP-MS). Generally, background (upstream of mine operations) sam- collected in the Croydon area were compared against Australian and ples contain low concentrations of As (b20 mg/kg), Cd (b1.5 mg/kg), New Zealand interim sediment quality guidelines (ISQGs). These em- Cu (b65 mg/kg), Pb (b50 mg/kg) and Zn (b200 mg/kg). These values pirical guidelines are derived from the North American effects database are within the ANZECC (2000) ISQG-High values. Several background (ANZECC/ARMCANZ, 2000a; in Simpson et al., 2005). The guidelines samples with elevated As, Pb and Sb relative to ISQG-Low values contain two concentrations, the ISQG-Low concentration (or trigger (Fig. 10A, D and E) were measured and likely indicate the presence of

As Location number Cd Location number 35 7 6 8 35 14 16 17 18 19 20 31 21 23 24 25 27 28 29 30 32 34 33 39 36 2 9 7 6 12 22 3 4 5 8 14 16 17 18 19 20 31 21 23 24 25 27 28 29 30 32 34 33 39 36 2 9 12 22 3 4 5 10 1 37 10 1 A 37 B Federation 1,800 Tabletop Creek Deadhorse Creek 12 Tabletop Creek Federation Deadhorse Creek La Perouse La Perouse 1,600

ISQG-High Glencoe Glencoe 10 1,400 Confluence Confluence 1,200 8 1,000 6 800 600 4 400 Concentration (mg/kg) Concentration (mg/kg) 2 ISQG-Low 200 ISQG-High 0 0 ISQG-Low 3.3 3.4 3.6 3.9 4.3 7.8 8.3 4.3 2.8 1.9 1.7 1.3 1.1 0.9 0.7 3.3 3.4 3.6 3.9 4.3 7.8 8.3 4.3 2.8 1.9 1.7 1.3 1.1 0.9 0.7 -4.4 -3.7 -3.6 -2.8 -2.6 -1.8 -1.7 -1.6 -1.3 -1.2 -1.1 -4.4 -3.7 -3.6 -2.8 -2.6 -1.8 -1.7 -1.6 -1.3 -1.2 -1.1 10.1 10.0 10.1 10.0 Distance upstream (km) Distance downstream (km) Distance upstream (km) Distance downstream (km)

Cu Location number Pb Location number 35 35 7 6 7 6 8 8 14 16 17 18 19 20 31 21 23 24 25 27 28 29 30 32 34 33 39 36 14 16 17 18 19 20 31 21 23 24 25 27 28 29 30 32 34 33 39 36 2 2 9 9 12 22 12 22 3 4 5 3 4 5 10 10 1 1 C 37 D 37 Federation Federation ISQG-High Tabletop Creek Deadhorse Creek 250 Tabletop Creek Deadhorse Creek La Perouse La Perouse 2,000 Glencoe 200 Glencoe Confluence

Confluence 1,500 150 1,000 100 ISQG-Low 500

Concentration (mg/kg) 50 Concentration (mg/kg) ISQG-High ISQG-Low 0 0 3.3 3.4 3.6 3.9 4.3 7.8 8.3 4.3 2.8 1.9 1.7 1.3 1.1 0.9 0.7 3.3 3.4 3.6 3.9 4.3 7.8 8.3 4.3 2.8 1.9 1.7 1.3 1.1 0.9 0.7 -4.4 -3.7 -3.6 -2.8 -2.6 -1.8 -1.7 -1.6 -1.3 -1.2 -1.1 -4.4 -3.7 -3.6 -2.8 -2.6 -1.8 -1.7 -1.6 -1.3 -1.2 -1.1 10.1 10.0 10.1 10.0 Distance upstream (km) Distance downstream (km) Distance upstream (km) Distance downstream (km)

Sb Location number Zn Location number 35 35 7 6 7 6 8 8 2 2 14 16 17 18 19 20 31 21 23 24 25 27 28 29 30 32 34 33 39 36 14 16 17 18 19 20 31 21 23 24 25 27 28 29 30 32 34 33 39 36 9 9 3 4 5 3 4 5 12 22 12 22 1 1 10 10 E 37 F 37 Tabletop Creek Deadhorse Creek 700 Tabletop Creek Federation Deadhorse Creek 25 Federation La Perouse ISQG-High 600 La Perouse Glencoe Glencoe 20 Confluence 500 Confluence ISQG-High 15 400 300 10 200 ISQG-Low

Concentration (mg/kg) 5 ISQG-Low Concentration (mg/kg) 100 0 0 3.3 3.4 3.6 3.9 4.3 7.8 3.3 3.4 3.6 3.9 4.3 7.8 8.3 4.3 2.8 1.9 1.7 1.3 1.1 0.9 0.7 8.3 4.3 2.8 1.9 1.7 1.3 1.1 0.9 0.7 -4.4 -3.7 -3.6 -2.8 -2.6 -1.8 -1.7 -1.6 -1.3 -1.2 -1.1 -4.4 -3.7 -3.6 -2.8 -2.6 -1.8 -1.7 -1.6 -1.3 -1.2 -1.1 10.1 10.0 10.1 10.0 Distance upstream (km) Distance downstream (km) Distance upstream (km) Distance downstream (km)

Total fraction < 63 µm fraction ISQG -High ISQG -Low

Fig. 10. Trace element content (mg/kg; measured by ICP-MS) in stream sediments (total and b63 μm fractions shown) around the Croydon district. ANZECC (2000) ISQG-High and ISQG-Low values are shown for comparison: (A) As; (B) Cd; (C) Cu; (D) Pb; (E) Sb; (F) Zn. NB. Sample location numbers are shown at the top of each graph, and correlate to locations shown on Fig. 2. A.K. Parbhakar-Fox et al. / Journal of Geochemical Exploration 137 (2014) 11–28 23 sulphide bearing lodes and mineralised rocks, causing localised element 4.4. Surface waters enrichment. At the Federation/La Perouse site, elevated concentrations of As The pH values measured across the Croydon district were compared (total and 63 μm fractions), Cu (b63 μm), Cd (b63 μm), Pb (b63 μm) against concentrations of As, Cd, Cu, Pb and Zn (Fig. 12). In Tabletop and Sb (b 63 μm) relative to ISQG-Low values were measured. Only As Creek upstream of the Federation/La Perouse site the pH range was 6.4 was elevated relative to the ISQG-High value at the Glencoe site. The to 8.6, with low concentrations of Al (b67 μg/L), As (b6.1 μg/L), Cd spatial distribution of sediment-associated metals downstream of both (b2 μg/L), Cu (b3 μg/L), Pb (b1.1 μg/L), S (0.42 mg/L) and Zn (b9 μg/L) sites did not display a simple distance–metal concentration decay pat- measured. Concentrations of Fe were elevated upstream of the mine op- tern. Downstream of Federation/La Perouse in Tabletop Creek, relative erations (Fig. 12E). At the Federation/La Perouse site, pH values decreased to the ISQG-High value, As remained generally in the b63 μm fraction to 3.7, with elevated concentrations of Al, As, Cd, Cu, Ni, Pb, S and Zn rel- to approximately 8 km (location 25; Fig. 10A). Cadmium was generally ative to local background/upstream value measured (Fig. 12). Down- below the ISQG-Low value from approximately 3 km downstream of stream, pH values generally increased with a maximum of pH 7.6 the Federation/La Perouse site (Fig. 10B) until the confluence (at ap- recorded. Generally, concentrations of Cd, Cu, Pb and Zn decreased proximately 10 km; location 27). Maximum concentrations of Cd downstream. (b63 μm) and Zn (b63 μm) exceeding ISQG-High values were mea- At Glencoe, pH values were between 4.3 and 6.3 (Fig. 12), and ele- sured in Deadhorse Creek approximately 0.9 km downstream of the vated concentrations of As (20.4 μg/L), Cd (53 μg/L), Pb (14 μg/L) and Glencoesite(location35;Fig. 10B and F). These concentrations fluctuated Zn (7715 μg/L) relative to ANZECC (2000) drinking water guidelines to the confluence, with occasional highs (e.g., location 30) again potential- (DWG) values were measured. In Deadhorse Creek adjacent to the site ly indicating the presence of a mineralised sulphide body in proximity to (location 36; Fig. 2), pH 3.95 was measured. However, by c. 1.7 km the sampled location. Copper concentrations fluctuated downstream of (location 33; Fig. 2), pH ~11 was measured. High dissolved Fe the Glencoe site to the confluence (Fig. 10C). Lead concentrations also (2125 μg/L) was also measured at this location. At the confluence of Ta- fluctuated for approximately 8 km downstream of the Glencoe site, bletop and Deadhorse Creeks, a near-neutral pH 6.7 was measured, after which it was measured below the ISQG-High value (Fig. 10D). At with only Cu (13 μg/L), Ni (12 μg/L), S (18 mg/L) and Zn (137 μg/L) el- the confluence of Tabletop and Deadhorse Creek concentrations for As, evated relative to background/upstream concentrations (Fig. 12), but Cu, Cd, Pb and Zn were measured below ISQG-High values. Monitoring are not elevated in comparison to ANZECC (2000) DWG values. (since 1998) of Tabletop Creek below the confluence (c. 17 and 30 km Element concentrations and pH values for Federation, La Perouse downstream of the Federation/La Perouse site) reported concentrations and Glencoe pit lakes are given in Table 3. Federation pit was the most below ISQG-Low values for As, Cu, Cd, Pb and Zn (DME, 2008). acidic (pH 3.9) whilst La Perouse and Glencoe were only mildly acidic Sequential extraction results for six samples collected at the with measured values of pH 6.1 and 6.3, respectively. These values are Federation/La Perouse site are presented in Fig. 11 as cumulative per- similar to previous monitoring data collected by the DEEDI since 1998 centages (element concentrations measured by ICP-MS). Arsenic is (DME, 2008). The Federation Pit lake contained the highest concentra- mainly the immobile residual fraction (45%) and the less mobile Fe tions of Cd (83.6 μg/L), Cu (989.6 μg/L), Ni (65.6 μg/L), Pb (71.3 μg/L) (III) oxide fraction (42%), with values below detection limit measured and Zn (1918 μg/L), and the highest As concentration was measured for the water soluble and exchangeable fractions. Cadmium belongs at Glencoe (20.4 μg/L). Elevated concentrations relative to ANZECC mainly to the water soluble (43%) and exchangeable (32%) fractions. (2000) DWG values were detected for As (Glencoe) and Pb (Federa- Copper is dominantly associated with the Fe (III) oxyhydroxide (36%) tion). Overall, La Perouse and Glencoe pit lakes have better water qual- and Fe (III) oxide (32%) fractions. Maximum concentrations of Ni, Pb ity than Federation. Since the construction of the catch dam, water and Sb were measured from the residual fractions (83, 66 and 100% quality in Federation pit has deteriorated as in the dry season, water respectively). Zinc showed a diverse behaviour and was divided into quality in the catch dam worsens due to evapo-concentration of solutes, the following fractions: 22% water soluble, 20% Fe (III) oxide, 14% ex- and in the wet season the water bodies become linked. changeable, 11% Mn oxide, and 11% Fe (III) hydroxide. In general terms, a higher proportion of Cd, Zn and Cu are associated with the readily mobile fractions of sediments with the following general order 5. Discussion of mobility: observed Cd N Zn N Cu N As N Pb N Ni N Sb. Similar trends of element mobility in mining affected areas were reported by Bird et al. 5.1. Acid forming groups and metal/metalloid sources (2003) and Teixeira et al. (2003). Detailed microtextural and chemical studies of the main sulphide- As Cd Cu Ni Pb Sb Zn 100 bearing groups revealed diversity of minerals and, in conjunction with geochemical data helped to identify the current ARD and metal/metal- loid sources. 80 Paste pH experiments indicate that material representative of group G (quartz–sphalerite–galena) is weakly acid forming, with the ARDI 60 classifying them as PAF (Table 2). This group is a significant source of Cd, as indicated by first-flush experiments (Fig. 9). Specifically Cd is fi b μ 40 sourced from ner-grained sphalerite ( 200 m), which is undergoing oxidation at a greater rate than coarser-grained sphalerite (N200 μm) as a function of surface area, higher Fe-content and the presence of galena 20 inclusions (cf. Lottermoser, 2010; Moncur et al., 2009; Stanton, 2005;

cumulative percentage (%) percentage cumulative Weisener et al., 2004). Zinc is also primarily sourced from sphalerite. 0 Neither Zn nor Cd is retained surficially on sphalerite, indicating that Water soluble Exchangeable Mn oxide on oxidation, metal deficient surface layers have formed (cf. Buckley Fe (III) hydroxide Fe (III) oxide Organic/secondary sulphide et al., 1989). Chalcopyrite was not detected as a major sulphide mineral Residual in any mesotextural group. However, it was observed in sphalerite as micro-inclusions. The presence of these micro-inclusions causes sphaler- Fig. 11. Cumulative percentage of As, Cd, Cu, Ni, Pb, Sb and Zn (mg/kg; measured by ICP-MS) in different steps of sequential extraction performed on sediment samples collected at the ite lattice destabilisation, and enhancing oxidation (cf. Lottermoser, 2010, Federation/La Perouse mine site, Croydon. Urbano et al., 2007). 24 A.K. Parbhakar-Fox et al. / Journal of Geochemical Exploration 137 (2014) 11–28

25,000 25 A Al B As 20,000 20

15,000 15

mg/L 10,000 10

5000 5

0 0 1234567891011121314 1234567891011121314 pH pH

90 600 80 C Cd D Cu 500 70 60 400 50 300 40 µg/L µg/L µg/L 30 200 20 100 10 0 0 1234567891011121314 1234567891011121314 pH pH

3000 90 Fe 80 Ni 2500 E F 70 2000 60 50 1500 µg/L µg/L 40 1000 30 20 500 10 0 0 1 2 3 4 5 6 7 8 9 1011121314 1234567891011121314 pH pH

70 140 60 G Pb 120 H S 50 100 40 80 µg/L µg/L 30 60 20 40 10 20 0 0 1234567891011121314 1234567891011121314 pH pH

9,000 8,000 I Zn Tabletop Creek (Upstream) 7,000 6,000 Federation/La Perouse 5,000 Glencoe

µg/L 4,000 3,000 Tabletop Creek (Downstream) 2,000 Deadhorse Creek (Downstream) 1,000 0 Confluence 1234567891011121314 pH

Fig. 12. Trace element concentration (μg/L; measured by ICP-MS) versus pH in surface water samples collected from around the Croydon district: (A) Al; (B) As; (C) Cd; (D) Cu; (E) Fe; (F) Ni; (G) Pb; (H) S; (I) Zn.

Galena in group G is rich in Ag, Bi and Sb increasing its potential for oxidation in accordance with Liu et al. (2008). Diehl et al. (2008) stated weathering and Pb release, with anglesite developed as a secondary that anglesite does not function as a protective barrier against fluid infil- product (cf. Diehl et al., 2003; Diehl et al., 2007; Savage et al., 2000). tration because it is porous and fine-grained. However, Moncur et al. Finer-grained (b200 μm) galena appeared more weathered than (2009) stated that anglesite rims slow oxidation progress. Observations coarser-grains, indicating that grain size poses a significant control on from this study support Diehl et al. (2008). A.K. Parbhakar-Fox et al. / Journal of Geochemical Exploration 137 (2014) 11–28 25

Table 3 contents (i.e., b1 wt.%), only a small contribution to net-ARD is likely Element concentrations (μg/L; analysed by ICP-MS) and pH values measured in Federa- from material representative of these groups. tion, La Perouse and Glencoe pit water samples (obtained from the surface). Drinking water guidelines (DWG) published by ANZECC (2000) are shown for comparison. 5.2. Metal and arsenic dispersion at Croydon Federation La Perouse Glencoe DWG

pH 3.9 6.1 6.3 6.5 to 8.5 At Croydon, surface water downstream of both mine sites is classified Al 17,680 25 11 – as acid-low metal to neutral-low metal by the general Ficklin diagram Ag 1 1 1 100 given in Plumlee (1999). As pH values increased above 4.5 downstream As 4 2 20 7 Cd 84 BDL 12of both the Federation/La Perouse and Glencoe sites, concentrations of Co 2 BDL BDL – Al, As, Co, Cd, Ni, Pb, S and Zn declined. However, concentrations of Fe Cr 4 6 6 50 were measured higher in background and downstream samples. Around Cu 990 2 1 2000 pH 4.5 to 5 As and Pb have likely sorbed onto Fe-bearing precipitates, Fe 44 105 56 – with sequential extraction results indicating concentration of these Mn 609 53 77 – Ni 66 3 4 – elements to this fraction (cf. Ashley et al., 2004; Hudson-Edwards Pb 71 BDL BDL 20 and Edwards, 2005; Lottermoser, 2010). Sediment loads of Cu, Cd Rb 14 7 10 – and Zn declined downstream, with Cu sorbed onto Fe-bearing precipi- Se 11 BDL 110tates around neutral pH (Hudson-Edwards and Edwards, 2005). Sr 44 6 3 – Ti 743 10 20 – Cadmium and Zn concentrations correlate with Mn indicating co- Tl BDL BDL BDL – precipitation of these elements with manganese oxides. At the conflu- U39 BDL BDL 10 ence approximately 10 km downstream of both sites, neutral pH was V2 3 3 – measured, with metal (Al, Cd, Cu, Pb and Zn) concentrations similar to Zn 1918 17 43 3000 upstream values and below guideline values recommended by govern- ment authorities (e.g., ANZECC, 2000), indicating effective attenuation. However, at Croydon seasonal fluxes of metals may occur as a function Material representative of mesotextural group H (arsenopyrite– of low flow conditions (low pH) or heavy rainfall events (high pH) caus- quartz–pyrite) was classified as extremely acid forming by geochemical ing desorption of elements (cf. Ashley et al., 2004; Harris et al., 2003; tests and the ARDI. Highly fractured massive arsenopyrite dominated, and Nordstrom, 2009). is in early stages of weathering, with scorodite dominating the secondary Generally water quality is similar in both creeks. Elevated Cd and Zn mineralogy. Scorodite behaves as a protective weathering barrier under concentrations measured in Deadhorse Creek provide an indication that acidic conditions as its dissolution is slow (10−9 to 10−10 mol m2 s−1; material representative of mesotextural group G dominates in the Harvey et al., 2006). Scorodite precipitates within arsenopyrite fractures Glencoe waste rock pile. Consequently, there is a potential risk to cattle and on grain boundaries as confined laminated layers of uniform thick- from ARD seepage at the site. However, these results suggest that Cd ness parallel to grain boundaries. Where scorodite growth is unconfined concentrations are relatively low from 1 km downstream of Glencoe. (e.g., in proximity to pyrite), a greater diversity of microtextures is ob- Instead, as the Croydon mine sites are currently designated grazing served, including spherules (DeSisto et al., 2011; Murciego et al., 2011) land, attention should be given to the presence of plant Calotrope and ‘ribbons’. Paste pH results indicate that to an extent, scorodite is (Calotropis procera) observed around the Croydon mines, as it is toxic retarding acid formation, as Craw et al. (2003) summarised that even if consumed (cf. Lottermoser, 2011). at submicron-scale, this mineral offers protection. Mesotextural group J samples (quartz–pyrite) were consistently 5.3. Implications for site rehabilitation classified the most acid forming group at Croydon by all geochemical methods and the ARDI. In general, larger pyrite grains were more Historic and abandoned sulphidic mine sites require rehabilitation. oxidised as a result of extensive fracturing. Group J samples had the However, extensive ARD processes, limited capital and excessive costs highest total concentration of metals (Cd, Cu, Ni, Pb and Zn). Arsenic associated with planning and remediation works mean that comprehen- in pyrite decreases resistance to oxidation (cf. Blanchard et al., 2007; sive site rehabilitation is rarely achievable. At these sites, rehabilitation is Plumlee, 1999), here this has caused accelerated weathering of these driven primarily by impacts on receiving environments i.e. soil and cores, with weathering further enhanced by the presence of galena water quality (e.g., Alvarez and Ridolfi, 1999; Mudd and Patterson, micro-inclusions (b10 μm diameter) straining the lattice (cf. Jambor, 2010). 1994; Kwong, 1995; Lottermoser, 2010; Plumlee, 1999). HFO have pre- If water quality data alone is used to assess the requirement for reha- cipitated in pyrite fractures, and have adsorbed elements released on bilitation at Croydon, the site would not be prioritised under Queensland's pyrite oxidation, particularly As (cf. Foster et al., 1998). In addition, Abandoned Mine Lands Program (AMLP). Our observations indicate that szomolnokite and rhomboclase have precipitated and also represent a sulphidic mine wastes are in early weathering stages by the general transient store of trace elements (Buckby et al., 2003; Lottermoser, mine waste paragenesis proposed by Moncur et al. (2009). However, 2010). These minerals are highly soluble under the pH range (3.0 to first-flush data indicated that there would be significant discharge of 4.7) measured in paste pH tests (cf. Harris et al., 2003), and therefore, metals and As after a heavy rainfall event. Thus, both Croydon sites stud- a likely source of As, Cd, Cu, Pb and Zn. ied are considered as long-term contamination sources posing significant Observations made in minor sulphide bearing groups i.e., A environmental risk to the downstream environment. (muscovite-altered porphyritic rhyolite with minor disseminated pyrite Previous rehabilitation strategies at the site focused on raising pH in in quartz veins), C (muscovite-altered porphyritic rhyolite with dissemi- pit lakes and managing ARD seepage downstream. However, as shown nated pyrite in quartz veinlets) and E (muscovite-altered porphyritic in this study, this strategy can only have a temporary effect and a rhyolite with disseminated pyrite in the groundmass) show that more sustainable long-term strategy should focus on the relocation of groundmass-associated sulphides are more susceptible to weathering ARD sources into newly constructed impoundments, containing waste than quartz-associated sulphides. Smuda et al. (2007) made similar ob- rock pile material domained into the ten mesotextural groups identi- servations at the Excelsior waste rock dump, Cerro de Pasco, Peru. These fied. Whilst upfront financial costs of adopting such a scheme may be authors concluded that fine-grained disseminated pyrite in a volcanic considered high, a long-term cost saving is anticipated when consider- groundmass oxidised much faster than massive pyrite from the ore ing that global liability costs associated with current and future ARD re- body due to the high porosity. Considering the small total sulphide mediation is estimated at US $100 billion (Tremblay and Hogan, 2001 in 26 A.K. Parbhakar-Fox et al. / Journal of Geochemical Exploration 137 (2014) 11–28

Hudson-Edwards et al., 2011), indicating that a proactive rather than re- Additionally, waste rock piles are treated as mineralogically homoge- sponsive approach to ARD management is required. nous entities. Consequently, rehabilitation schemes can have limited Material representative of mesotextural group H is the dominant success, with significant associated financial costs. This study demon- source of As, with its concentration in dissolved waters controlled in- strated that through improving the mineralogical and geochemical un- part by scorodite formed in waste rock material. However, evidence of derstanding of ARD sources in waste rock piles, better management scorodite weathering was observed, indicating a likely temporal in- schemes can be developed, rather than utilising reactive strategies. crease in dissolved As in waste pile leachates. Material representative Materials collected from waste rock piles at the abandoned Croydon of mesotextural group G contains significant quantities of Cd (and Zn) mine operations (north Queensland, Australia), were used to develop which is homogenously distributed in sphalerite grains and therefore and test a mesotextural grouping methodology. This method accurately identified as a long-term source of these elements. Whilst relatively domains material based on chemical, mineralogical and textural low metal and As concentrations were measured from mesotextural similarities by using geological logging techniques, pXRF and SW-IR group J in kinetic trials (Parbhakar-Fox et al., 2013), it is identified as a analyses. Mineralogical variability is better accounted for than if using significant contamination source due to its high pyrite contents (i.e., lithology or alteration classes, with samples domained based on ARD generating low pH conditions). Considering these diverse chemical and forming characteristics prior to ARD testing. Mesotextural grouping mineralogical characteristics, consideration should be given to handling of Croydon waste materials identified ten major groups (A to J), and managing mesotextural group G, H and J materials separately. Co- which comprised of hydrothermally altered rhyolites and massive/ disposal of these materials has potential to increase the environmental semi-massive sulphides. risk posed, through geochemical interactions. Such waste segregation Three major acid forming mesotextural groups were identi- practices are described in the GARD guide (2010). For example, sphalerite fied: J (quartz–pyrite), H (arsenopyrite–quartz–pyrite), and G leaching is enhanced at low pH (2 to 4; Stanton et al., 2008), and first- (quartz–sphalerite–galena). Group J contained moderate-high concen- flush pH measurements from group H were below 4. Therefore, co- trations of As, Pb and Zn sourced from pyrite, and was consistently the disposal may enhance sphalerite oxidation, releasing increased amounts most acid forming. Some secondary HFO had developed. However, of dissolved Cd and Zn. high concentrations of elements Pb, Zn and Cu were measured in As an alternative to impermeable covers, neutralisation strategies first flush experiments, suggesting dissolution of secondary sul- could be implemented in the impoundments containing material repre- phates (e.g., rhomboclase, szomolnokite). Group H contained major sentative of mesotextural group G and J (and E, which can be co- arsenopyrite; however scorodite had extensively precipitated in frac- disposed with J) waste in order to raise pH and reduce heavy metal tures and rims, thus oxidation is controlled by its solubility. Group G and metalloid mobility (cf. Ashley et al., 2004). Alternatives to lime contains high amounts of Zn, Cd and Pb sourced from sphalerite and ga- should be sought as neutralising materials given the limited success lena. Galena had undergone partial weathering to anglesite with Pb lib- previously experienced (DME, 2008). As scorodite has extensively pre- erated, as observed in first flush experiments. Elevated Zn and Cd also cipitated in mesotextural group H, the pH in the repository containing it measured in these experiments was a consequence of the absence of should be maintained in a range where it is relatively insoluble. Bluteau armouring secondary minerals on sphalerite, and as sphalerite was and Demopoulos (2007) reported low solubility rates (0.35 mg/L−1)of Fe-rich with abundant micro-inclusions, continued oxidation is As at pH 5, with its dissolution from pH 5 to 9 similarly low. However, expected. Krause and Ettel (1988) reported that ~90 mg/L−1 As was leached Elevated Cd, Pb and Zn relative to ANZECC (2000) ISQG were from scorodite at pH 5. Drahota and Filippi (2009) proposed that rea- measured within 10 km downstream of the site operations, indicating sons for these differences may be linked to the degree of scorodite that these sources of contaminants are impacting the local environ- crystallinity used, with two types of scorodite texture observed in ment. This study indicates that a rehabilitation strategy involving the here (spherules, and acicular grains). Therefore, prior to recommending segregation of sulphidic material into repositories designated for each the geochemical conditions at which this material representative of this mesotextural groups (i.e., G, H and J) should be considered as an option mesotextural group should be maintained at, further work is required to for future waste management at this site. Such a strategy will limit in- elucidate the crystallinity and stability of scorodite in this mesotextural teractions between these materials, thus limiting liberation of environ- group. mentally significant elements. Furthermore, different pH conditions can The mesotextural classification method developed in this study is be maintained in each pile to encourage formation of secondary min- simple and cost-effective to perform at abandoned mine operations be- erals and ensure their stability. cause it allows for the rapid assessment of large quantities of waste ma- Further studies should focus on a better understanding of the terials. Such a site-specific methodology should be designed primarily long-term geochemical behaviour of waste rock, identifying the loca- by geologists (in conjunction with site managers), as an understanding tion and flow rates of seepage points, measuring volumes of water in of the waste rock mineralogy is fundamental for scheme to be effective. sub-catchments and understanding the characteristics of shallow Based on the collection of real data (i.e., pXRF, SW-IR) and improved groundwater. mineralogical and textural assessments of waste materials, the selection Supplementary data to this article can be found online at http://dx. of the most appropriate samples for standard geochemical tests and in- doi.org/10.1016/j.gexplo.2013.10.017. depth mineralogical evaluations is permitted. This allows for full ARD characterisation i.e., identifying sulphide oxidation controls and under- standing of ARD evolution on a mesotextural group basis. This method- Acknowledgements ology has been adapted for use at several operational Australian mine sites, where such mesotextural evaluations were performed to support The authors acknowledge the Queensland Department of Natural deposit-wide domaining of ARD characteristics, and has allowed for the Resources and Mines for allowing access to the Croydon sites for sample development of improved waste management schemes. collection. Thanks are extended to Dr. Karsten Goemann (CSL, UTAS), Katie McGoldrick (CODES, UTAS), Dr. Nathan Fox (CODES, UTAS), 6. Conclusions Sarah Gilbert (CODES, UTAS) and Ian Little (CODES, UTAS) for analytical assistance. Additional thanks are extended to Professors Bernd Rehabilitation strategies at historic and abandoned metalliferous Lottermoser, Dee Bradshaw, Dave Craw and Bernard Dold for reviewing mine sites are only undertaken when ARD is generated. Consequently, the Ph.D thesis from which this manuscript is synthesised, and two a ‘blanket approach’ to rehabilitation is typically adopted whereby tech- anonymous reviewers, whose comments have greatly improved the niques such as lime dosing and waste rock capping are performed. quality of this manuscript. A.K. Parbhakar-Fox et al. / Journal of Geochemical Exploration 137 (2014) 11–28 27

References Drahota, P., Filippi, M., 2009. Secondary arsenic minerals in the environment: a review. Environ. Int. 35, 1243–1255. Alvarez, S.M., Ridolfi, B.A., 1999. Moon Creek reclamation project creative solution for an Edraki, M., Baumgartl, T., Fletcher, A., Mulligan, D., Fegan, W., Munawar, A., 2009. Hydro- historic mine site. SME Annual Meeting March 1999, Denver Colorado, pp. 1–7. geochemical evolution of an uncapped gold tailings storage. In: Wiertz, J., Moran, C. Andrina, J., Wilson, G.W., Miller, S., Neale, A., 2006. Performance of the acid rock drainage (Eds.), Proceedings of the 1st International Seminar on Environmental Issues in the mitigation waste rock trial dump at Grasberg mine. 7th International Conference on Mining Industry. Santiago, Chile. Acid Rock Drainage (ICARD), pp. 30–44. Edraki, M., Forsyth, B., Baumgartl, T., Bradshaw, D., 2012. Geochemistry of tailings and ANZECC, 2000. Australian guidelines for water quality monitoring and reporting. National seepage from three tailings storage facilities in Australia — uncapped, capped, and ac- Water Quality Management Strategy Paper No 7, Australian and New Zealand Envi- tive tailings. Proceedings of the Life-of-Mine Conference, Brisbane, Australia, ronment and Conservation Council & Agriculture and Resource Management Council pp. 269–277. of Australia and New Zealand, Canberra. Egiebor, N.O., Oni, B., 2007. Acid rock drainage formation and treatment: a review. Asia Ashley, P.M., Lottermoser, B.G., Collins, A.J., Grant, C.D., 2004. Environmental geochemistry Pac. J. Chem. Eng. 2, 47–62. of the derelict Webbs Consols mine, New South Wales, Australia. Environ. Geol. 46, Evangelou, V.P., Zhang, Y.L., 1995. A review: pyrite oxidation mechanisms and acid mine 591–604. drainage prevention. Crit. Rev. Environ. Sci. Technol. 25, 141–199. Aykol, A., Budakoglu, M., Kumral, M., Gultekin, A.H., Turhan, M., Esenli, V., Yavuz, F., Fandrich, R., Gu, Y., Burrows, D., Moeller, K., 2007. Modern SEM-based mineral liberation Orgun, Y., 2003. Heavy metal pollution and acid drainage from the abandoned analysis. Int. J. Miner. Process. 84, 310–320. Balya Pb–Zn sulfide Mine, NW Anatolia, Turkey. Environ. Geol. 45, 198–208. Foster, A.L., Brown Jr., G.E., Parks, G.A., Tingle, T.N., 1998. Quantitative speciation of arsenic Bain, J.H.C., Withnall, I.W., Black, L.P., Etminan, H., Golding, S.D., Sun, S.S., 1998. Towards in mine tailings using X-ray absorption spectroscopy. Am. Mineral. 89, 553–568. an understanding of the age and origin of mesothermal gold mineralisation in the Franco,S.,Kadletz,O.,Stevens,R.,Nguyen,X.,Reid,M.,Scougall,J.,McLean,R., Etheridge Goldfield, Georgetown region, north Queensland. Aust. J. Earth Sci. 45, Stutsel,M.,Long,T.,2010.Strategic framework for managing abandoned mines 247–263. in the minerals industry. Ministerial Council on Mineral and Petroleum Barton Jr., P.A., Bethke, P.M., 1987. Chalcopyrite disease in sphalerite: pathology and Resources (MCMPR), p. 44. epidemiology. Am. Mineral. 72, 451–467. Gasparon, M., Smedley, A., Jong, T., Costagliola, P., Benvenuti, M., 2007. Acid mine drainage Bethke, P.M., Barton Jr., P.A., 1971. Distribution of some minor elements between at Mount Morgan, Queensland (Australia): experimental simulation and geochemical coexisting sulfide minerals. Econ. Geol. 66, 140–163. modeling of buffering reactions. In: Cidu, R., Frau, F. (Eds.), IMWA Symposium: Water Bird, G., Brewer, P.A., Mackin, M.G., Balteanu, D., Driga, B., Serban, M., Zaharia, S., 2003. The in Mining Environments, pp. 433–436. solid state partitioning of contaminant metals and As in river channel sediments of Giere, R., Sidenko, N.V., Lazareva, E.V., 2003. The role of secondary minerals in controlling the the mining affected Tisa drainage basin, northwestern Romania and eastern migration of arsenic and metals from high-sulfide wastes (Berikul gold mine, Siberia). Hungary. Appl. Geochem. 18, 1583–1595. Appl. Geochem. 18, 1347–1359. Blanchard, M., Alfredsson, M., Brodholt, J., Wright, K., Catlow, C.R.A., 2007. Arsenic incor- Goodall, W., 2008. Automated mineralogy in the prediction of acid rock drainage: acces- poration into FeS2 pyrite and its influence on dissolution: a DFT study. Geochim. sible mineralogy using QEMSCAN. Proceedings of the 2008 Society for Mining, Metal- Cosmochim. Acta 71, 624–630. lurgy and Exploration (SME) Annual Meeting and Exhibit, Salt Lake City, Utah, United Bluteau, M.C., Demopoulos, G.P., 2007. The incogruent dissolution of scorodite: solubility, States. kinetics and mechanism. Hydrometallurgy 87, 163–177. Gore, D.B., Preston, N.J., Kirstie, A.F., 2007. Post-rehabilitation environmental hazard of Cu, Bowell, R.J., Rees, S.B., Parshley, J.V., 2000. Geochemical predictions of metal leaching and Zn, As and Pb at the derelict Conrad Mine, eastern Australia. Environ. Pollut. 148, acid generation: geologic controls and baseline assessment. Geology and Ore 491–500. Deposits: The great Basin and Beyond Proceedings, 2, pp. 799–823. Gray, N.F., 1997. Environmental impact and remediation of acid mine drainage: a Brown, D., Salzsauler, K., Verburg, R., Sifuentes, R., Aranda, C., 2009. Geochemical charac- management problem. Environ. Geol. 30, 62–71. terization of waste rock using field kinetic testing at the Antamina mine. 8th Interna- Harries, J.R., 1997. Acid Mine Drainage in Australia: Its Extent and Potential Future Liabil- tional Conference on Acid Rock Drainage (ICARD) and Securing the Future: Mining, ity. Supervising Scientist, Canberra (94 pp.). Metals & the Environment in a Sustainable Society, pp. 311–321. Harris, D.L., Lottermoser, B.G., Duchesne, J., 2003. Ephemeral acid mine drainage at the Buckby,T.,Black,S.,Coleman,M.L.,Hodson,M.E.,2003.Fe-sulfate-rich evaporative Montalbion silver mine, north Queensland. Aust. J. Earth Sci. 50, 797–809. mineral precipitates from the Rio Tinto, southwest Spain. Mineral. Mag. 67, 263–278. Harvey, M.C., Schreiber, M.E., Rimstidt, J.D., Griffith, M.M., 2006. Scorodite dissolution Buckley, A.N., Woods, R., Wouterlood, H.J., 1989. An XPS investigation of the surface of kinetics: implications for arsenic release. Environ. Sci. Technol. 40, 6709–6714. natural sphalerites under flotation-related conditions. Int. J. Miner. Process. 26, Hudson-Edwards, K.A., Edwards, S.J., 2005. Mineralogical controls on storage of As, Cu, Pb 29–49. and Zn at the abandoned Mathiatis massive sulphide mine, Cyprus. Mineral. Mag. 69, Bureau of Meteorology, 2013. www.bom.com.au. 695–706. Caruso, B.S., Bishop, M., 2009. Seasonal and spatial variation of metal loads from natural Hudson-Edwards, K.A., Jamieson, H.E., Lottermoser, B.G., 2011. Mine wastes: past, present flows in the Upper Tenmile Creek Watershed, Montana. Mine Water Environ. 28, and future. Elements 7, 375–380. 166–181. Hutchison, B.J., Brett, D., 2006. Savage river mine — practical remediation works. 7th In- Cook, N.J., Ciobanu, C.L., Pring, A., Skinner, W., Danyushevsky, L., Shimizu, M., Saini- ternational Conference on Acid Rock Drainage (ICARD), pp. 810–819. Eidukat, B., Melcher, F., 2009. Trace and minor elements in sphalerite: a LA-ICP-MS Irwin, R.J., Van Mouwerik, M., Stevens, L., Seese, M.S., Basham, W., 1997. Environ- study. Geochim. Cosmochim. Acta 73, 4761–4791. mental Contaminants Encyclopedia Cadmium Entry. National Parks Service, Colarado, Craw, D., Falconer, D., Youngson, J.H., 2003. Environmental arsenopyrite stability and US 88. dissolution: theory, experiment and field observations. Chem. Geol. 199, 71–82. Jambor, J.L., 1994. Mineralogy of sulphide rich tailings and their oxidation products. In: Danyushevsky, L., Robinson, P., Gilbert, S., Norman, M., Large, R., McGoldrick, P., Shelley, Blowes, D.W., Jambor, J.L. (Eds.), The Environmental Geochemistry of Sulphide M., 2011. Routine quantitative multi-element analysis of sulphide minerals by laser Mine Wastes. Mineralogical Association of Canada, Short Course Series, 22, ablation ICP-MS: standard development and consideration of matrix effects. pp. 59–102. Geochem. Explor. Environ. Anal. 11, 51–60. Jambor, J.L., Dutrizac, J.E., Groat, L., Raudsepp, M., 2002. Static tests of neutralization David, C.P.C., 2003. Establishing the impact of acid mine drainage through metal bioaccu- potentials of silicate and aluminosilicate minerals. Environ. Geol. 43, 1–17. mulation and taxa richness of benthic insects in a tropical Asian stream (the Jambor, J.L., Dutrizac, J.E., Raudsepp, M., 2007. Measured and computed neutralization Philippines). Environ. Toxicol. Chem. 22, 2952–2959. potentials from static tests of diverse rock types. Environ. Geol. 52, 1019–1031. DeSisto, S.L., Jamieson, H.E., Parsons, M.B., 2011. Influence of hardpan layers on arsenic Krause, E., Ettel, V.A., 1988. Solubility and stability of scorodite: new data and further mobility in historical gold mine tailings. Appl. Geochem. 26, 2004–2018. discussion. Am. Mineral. 73, 850–854. Diehl, S.F., Smith, K.S., Desborough, G.A., Goldhaber, M.B., Fey, D.L., 2003. Trace-metal Kwong, J.Y.T., 1995. Thoughts on ways to improve acid drainage and metal leaching pre- sources and their release from mine wastes: examples from humidity cell tests diction for metal mines. U.S. Geol. Surv. Water Resour. Rep. Invest. 95–4227. of hard-rock mine waste and from Warrior Basin coal. National Meeting of the Kwong, Y.T., Swerhone, G.W., Lawrence, J.R., 2003. Galvanic sulphide oxidation as a American Society of Mining and Reclamation Symposium, Billings, MT, metal-leaching mechanism and its environmental implications. Geochem. pp. 232–253. Explor. Environ. Anal. 3, 337–343. Diehl, S.F., Koenig, A.E., Hageman, P.L., Smith, K.S., Fey, D.L., Lowers, H.A., 2007. From the Liu, Y.G., Zhou, M., Zeng, G.M., Wang, X., Fan, T., Xu, W.H., 2008. Bioleaching of heavy micro to the macroscale: a textural and chemical perspective of characterising waste- metals from mine tailings by indigenous sulfur-oxidizing bacteria: effects of substrate rock material. Proceedings of the 2007 Society for Mining, Metallurgy and Exploration concentration. Bioresour. Technol. 99, 4124–4129. (SME) Annual Meeting and Exhibit, and the 109th National Western Mining Conference, Lottermoser, B.G., 2010. Mine Wastes: Characterization, Treatment and Environmental Denver, Colorado. Society for Mining, Metallurgy and Exploration Inc., United States, Impacts, third ed. Springer-Verlag, Berlin Heidelberg. Preprint 07-021, pp. 1–16. Lottermoser, B.G., 2011. Colonisation of the rehabilitated Mary Kathleen uranium mine Diehl, S.F., Hageman, P.L., Smith, K.S., 2008. What is weathering in mine waste? Mineral- site (Australia) by Calotropis procera: toxicity risk to grazing animals. J. Geochem. ogic evidence for sources of metals in leachates, Chapter A. In: Verplanck, P.L. (Ed.), Explor. 111, 39–46. Understanding Contaminants Associated With Mineral Deposits: U.S. Geological Sur- Lottermoser, B.G., Ashley, P.M., Lawie, D.C., 1999. Environmental geochemistry of the Gulf vey Circular, 1328, pp. 4–7. Creek copper mine area, north-eastern New South Wales, Australia. Environ. Geol. 39, DME (Queensland Department of Mines and Environment), 2008. Report on Value Engi- 61–74. neering Study, Remediation of Abandoned Federation and Glencoe Mines, Croydon. Lottermoser, B.G., Ashley, P.M., Costelloe, M.T., 2005. Contaminant dispersion at the reha- Dold, B., 2003. Speciation of the most soluble phases in a sequential extraction procedure bilitated Mary Kathleen uranium mine, Australia. Environ. Geol. 48, 748–761. adapted for geochemical studies of copper sulphide mine waste. J. Geochem. Explor. Luís, A.T., Teixeira, P., Almeida, S.F.P., Ector, L., Matos, J.X., Ferreira da Silva, E.A., 2009. Im- 80, 55–68. pact of Acid Mine Drainage (AMD) on water quality, stream sediments and periphytic Downing, B.W., Madeisky, H.E., 1997. Lithogeochemical methods for acid rock drainage diatom communities in the surrounding streams of Aljustrel Mining Area (Portugal). studies and prediction. Explor. Min. Geol. 6, 367–379. Water Air Soil Pollut. 200, 147–167. 28 A.K. Parbhakar-Fox et al. / Journal of Geochemical Exploration 137 (2014) 11–28

Marescotti, P., Carbone, C., De Capitani, L., Greco, G., Lucchetti, G., Servida, D., 2007. Mineral- Stanton, M.R., Taylor, C.D., Gemery-Hill, P.A., Shanks III, W.C., 2006. Laboratory studies of ogical and geochemical characterisation of open-air tailing and waste-rock dumps from sphalerite decomposition: applications to the weathering of mine wastes and potential the Libiola Fe–Cu sulphide mine (Eastern Liguria, Italy). Environ. Geol. 53, 1613–1626. effects on water quality. Paper presented at the 7th International Conference on Acid Moncur, M.C., Jambor, J.L., Ptacek, C.J., Blowes, D.W., 2009. Mine drainage from the Rock Drainage (ICARD), March 26–30, 2006, St. Louis MO. R.I. Barnhise. Published by weathering of sulfide minerals and magnetite. Appl. Geochem. 24, 2362–2373. the American Society of Mining and Reclamation (ASMR), 3134 Montavesta Road, Mudd, G.M., 2005. An environmental history of uranium mining in Australia: a scientificre- Lexington, KY, p. 40502. view. Proceedings of the Australian Uranium Conference, Fremantle, WA, pp. 1–19. Stanton, R., Gemery-Hill, P.A., Shanks, W.C., Taylor, C.D., 2008. Removal of zinc and trace Mudd, G.M., Patterson, J., 2010. Continuing pollution from the Rum Jungle U–Cu Project: a metal release from dissolving sphalerite at pH 2.0 to 4.0, Appl. Geochem. 23, critical evaluation of environmental monitoring and rehabilitation. Environ. Pollut. 136–147. 158, 1252–1260. Stomberg, B., Banwart, S., 1999. Weathering kinetics of waste rock from the Aitik copper Murciego, A., Álvarez-Ayuso, E., Pellitero, E., Rodríguez, M.A., García-Sánchez, A., Tamayo, A., mine, Sweden: scale dependent rate factors and pH controls in large column experi- Rubio, J., Rubio, F., Rubin, J., 2011. Study of arsenopyrite weathering products in mine ments. J. Contam. Hydrol. 39, 59–89. wastes from abandoned tungsten and tin exploitations. J. Hazard. Mater. 186, 590–601. Tarras-Wahlberg, N.H., Nguyen, T.L., 2008. Environmental regulatory failure and metal con- Nordstrom, D.K., 2004. Modeling low-temperature geochemical processes. In: Drever, J.I. tamination at the Giap Lai Pyrite mine, Northern Vietnam. J. Environ. Manag. 86, (Ed.), Surface and Ground Water, Weathering, and Soils. Treatise of Geochemistry: 712–720. Elsevier Pergamon, Amsterdam, pp. 37–72. Tauson, V.L., Babkin, D.N., Parkhomenko, Y., Menshikov, V.I., 2004. On the mechanism of Nordstrom, D.K., 2009. Acid rock drainage and climate change. J. Geochem. Explor. 100, trace-element uptake during the hydrothermal growth of sulphide mineral crystals. 97–104. Crystallogr. Rep. 49, 149–157. Paktunc, A.D., 2001. MODAN a computer program for estimating mineral quantities based Texeira, E.C., Rodrigues, M.L.K., Alves, M.F.C., Barbosa, J.R., 2003. Study of the geochemical on bulk composition: Windows version. Comput. Geosci. 21, 883–886. distribution of heavy metals in sediments in areas impacted by coal mining. In: Locat, Parbhakar-Fox, A.K., Edraki, M., Walters, S., Bradshaw, D., 2011. Development of a textural J., Galvez-Cloutier, R., Chaney, R., Demars, K. (Eds.), Contaminated Sediments: Charac- index for the prediction of acid rock drainage. Miner. Eng. 24, 1277–1287. terisation, Evaluation, Mitigation, Restoration and Management Strategy Perfor- Parbhakar-Fox, A., Bradshaw, D., Lottermoser, B., 2013. Evaluating waste rock mineralogy mance. Proceedings of the Second International Symposium on Contaminated and microtexture during kinetic testing for improved acid rock drainage prediction. Sediments (Quebec City, Canada, May 26–28, 2003). ASTM International, West Miner. Eng. 52, 111–124. Conshohocken, PA, pp. 72–86. Plumlee, G.S., 1999. The environmental geology of mineral deposits. In: Plumlee, G.S., Tran, A.B., Miller, S., Williams, D.J., Fines, P., Wilson, G.W., 2003. Geochemical and miner- Lodgson, M.J. (Eds.), The Environmental Geochemistry of Mineral Deposits Part A: alogical characterisation of two contrasting waste rock dumps — the INAP waste rock Processes, Techniques and Health Issues, Reviews in Economic Geology, vol. 6B. Society dump characterisation project. 6th International Conference on Acid Rock Drainage of Economic Geologists, United States, Littleton, CO, pp. 71–116. (ICARD), pp. 939–948. Savage, K.S., Tingle, T.N., O'Day, P.A., Waychunas, G.A., Bird, D.K., 2000. Arsenic speciation Tremblay, G.A., Hogan, C.M., 2001. Mine Environment Neutral Drainage (MEND) Manual in pyrite and secondary weathering phases, Mother Lode Gold District, Tuolumne 5.4.2d: Prevention and Control, Canada Centre for Mineral and Energy Technology, County, California. Appl. Geochem. 15, 1219–1244. Natural Resources Canada, Ottawa (352 pp.). Simpson, S.L., Batley, G.E., Chaiton, G.E., Stauber, A.A., King, C.K., Chapman, J.C., Hyne, R.V., Unger, C., Lechner, A., Glenn, V., Edraki, M., Mulligan, D.R., 2012. Mapping and prioritising Gale, S.A., Roach, A.C., Maher, W.A., 2005. Handbook for Sediment Quality Assessment rehabilitation of abandoned mines in Australia. Life-of-Mine Conference 2012, Brisbane, (CSIRO, Bangor, NSW). 126. Australia, pp. 1–14. Skousen, J., Simmons, J., Ziemkiewicz, P., 2002. Acid–base accounting to predict post- Urbano, G., Melendez, A.M., Reyes, V.E., Veloz, M.A., Gonzalez, I., 2007. Galvanic interactions mining drainage quality on surface mines. J. Environ. Qual. 31, 2034–2044. between galena–sphaleriteandtheirreactivity.Int.J.Miner.Process.82,148–155. Smart, R., Skinner, W.M., Levay, G., Gerson, A.R., Thomas, J.E., Sobieraj, H., Schumann, R., Van Eck, M., Child, R., 1990. Croydon gold deposits. In: Hughes, F.E. (Ed.), Geology and Weisener, C.G., Weber, P.A., Miller, S.D., Stewart, W.A., 2002. ARD Test Handbook: Mineral Deposits of Australia and Papua New Guinea, the Australian Institute of Mining Project P387A Prediction and Kinetic Control of Acid Mine Drainage. AMIRA, Interna- and Metallurgy, Monograph. , 14, pp. 979–982. tional Ltd, Ian Wark Research Institute, Melbourne, Australia. Vik, E.A., Bardos, P., Brogan, J., Edwards, D., Gondi, F., Henrysson, T., Jensen, B.K., Smith, L.J., Neuner, M., Gupton, M., Moore, M., Bailey, B.L., Blowes, D.W., Smith, L., Sego, Jorge, C., Mariotti, C., Nathanail, P., Papassiopi, N., 2001. Towards a framework D.C., 2009. Diavik waste rock project: from the laboratory to the Canadian arctic. for selecting remediation technologies for contaminated sites. Land Contam. Reclam. 8th International Conference on Acid Rock Drainage (ICARD). 9, 119–128. Smuda, J., Dold, B., Friese, K., Morgenstern, P., Glaesser, W., 2007. Mineralogical and geo- Weisener, C.G., Smart, R.St., Gerson, A., 2004. A comparison of the kinetics and mechanism chemical study of element mobility at the sulphide-rich Excelsior waste rock dump of acid leaching of sphalerite containing low and high concentrations of iron. Int. from the polymetallic Zn–gPb–(A –Bi–Cu) deposit, Cerro de Pasco, Peru. J. Geochem. J. Miner. Process. 74, 239–249. Explor. 92, 97–110. White, W.W., Lapakko, K.A., Cox, R.L., 1999. Static test methods most commonly used to Stanton, M.R., 2005. Baseline laboratory studies of sphalerite (ZnS) dissolution: effects on predict acid mine drainage: practical guidelines for use and interpretation. In: aqueous metal concentrations and solubilization rates. In: Barnhisel, R.I. (Ed.), Pro- Plumlee,G.S.,Lodgson,M.J.(Eds.),TheEnvironmental Geochemistry of Mineral Deposits ceedings 22nd National Conference. American Association of Mining Reclamation, Part A: Processes, Techniques, and Health Issues, Reviews of Economic Geology, vol. 6A, Breckenridge, CO, pp. 1155–1165. pp. 325–338.