Case Studies Chapter 4
Geochemical Characterisation in Australia
Geochemical Characterisation to Quantify ARD Risk And Facilitate Management at a Gold Mine in Australia
An ARD Geochemical Characterisation case study prepared by Earth Systems Pty. Ltd.
A waste material characterisation study was carried out during the pre-feasibility stage of an Australian gold mining project. This case study documents how appropriate geochemical characterisation, from sample selection, static and kinetic geochemical testwork, through to integration with a mine block model, forms the basis for quantifying ARD risk and for ARD management planning.
While rock and tailings were the main waste materials to be generated from mining activities, this case study focuses on waste rock only. The approach, however, is equally applicable to tailings, pit wallrock, ore/concentrate stockpiles and heap leach pads.
Sample selection and collection
Selection of sufficient and representative waste rock samples for geochemical analysis was a critical part of the characterisation study. Available data on geology (map and cross section), geochemistry (assay), mineralogy and the pit shell were used to identify seven distinct waste rock lithologies. A lithology is defined as a primary rock type that has been altered by a mineralising overprint (protolith plus alteration features). Therefore a single rock type may be represented by multiple lithologies. These may be further sub-divided into degrees of weathering.
An initial 151 waste rock samples were selected from the seven lithologies for static geochemical testwork. These comprised several 1 m interval samples of each lithology, chosen from drill holes across the vertical and lateral extent of the pit shell. If a single lithology contained fresh, partially oxidised or fully oxidised sections, samples were also collected to represent each of these materials. Total S (sulfur) and Total C (carbon) assay data were already available for the entire deposit, and this assisted sample selection by enabling a broad range of S and C values to be chosen. The number of samples collected for each lithology (and weathering sub-category) was proportional to the relative proportion of that lithology within the pit shell that was identified as waste rock.
Static geochemical characterisation and ARD risk classification
The following static geochemical parameters were determined for all waste rock samples:
- Total Sulfur;
- Maximum Potential Acidity (MPA, kg H2SO4 / tonne);
- Acid Neutralising Capacity (ANC, kg H2SO4 / tonne);
- Net Acid Producing Potential (NAPP = MPA-ANC, kg H2SO4 / tonne);
- pH after oxidation (NAG pH, pH units);
- Net Acid Generation at pH 4.5 (NAG4.5, kg H2SO4 / tonne);
- Net Acid Generation at pH 7.0 (NAG7.0, kg H2SO4 / tonne);
- Total Carbon (wt% C) and Total Organic Carbon (TOC, wt% C) (Total C-TOC = Carbonate Carbon wt.%).
Based on these results, samples were classified using AMDact v.2.5 (by Earth Systems) according to their ARD risk into the following categories:
- High potential for acid generation – Category 1 (AG1).
- Moderate/high potential for acid generation – Category 2 (AG2).
- Moderate potential for acid generation – Category 3 (AG3).
- Low potential for acid generation – Category 4 (AG4).
- Unlikely to be acid generating (UAG).
- Likely to be acid consuming (LAC).
- Inconsistent data (ID).
Average static geochemical results and corresponding ARD risk classifications are presented in Table 1 and Figure 1. The majority of waste rock samples were characterised by positive NAPP values and NAG pH below 4.5, corresponding to various acid generating categories. Several samples with NAG pH values above 4.5, despite positive NAPP values were considered unlikely to be acid generating (UAG). Negative NAPP values generally corresponded to NAG pH values greater than 4.5, indicating that they were likely to be acid consuming (LAC). A relatively close relationship between NAPP and NAG7.0 values for most lithologies indicates that pyrite is the dominant sulfur species.
Samples generally contained relatively low concentrations of Total C and essentially no TOC, and hence Total C values were effectively equivalent to Carbonate C and therefore consistent with the generally low ANC values. Likewise, there was generally a good correlation between measured NAPP values (Table 1) and NAPP values calculated from existing Total Sulfur and Total Carbon values for the same samples. Based on this correlation, NAPP values could be calculated for each 1m interval within the pit shell using the S and C assay data. This calculation was incorporated into the mine block model to create an environmental geochemistry layer. The ARD risk classification system (above) was then applied to the mine block model, based on the calculated NAPP layer, permitting estimation of the annual production of waste rock within each ARD risk category (Figure 2).
Kinetic geochemical testwork and estimation of annual acidity generation rates
Oxygen consumption testwork was conducted on bulk samples of all seven lithologies to clarify pyrite oxidation rates (POR) for the purpose of estimating acidity generation rates from waste rock materials. This approach provides more rapid results than other kinetic testwork methods (eg. column leach, humidity cells) and also has the ability to assess PORs as a function of key controls such as moisture content, oxygen concentration and particle size.
The measured oxygen consumption rate for each bulk sample was assumed to be proportional to the mass of pyrite in the sample and was converted into a POR using the stoichiometric relationship in the complete reaction for pyrite oxidation by oxygen (ie. 3.75 moles of oxygen are consumed to fully oxidise 1 mole of pyrite). This assumes that all sulfur present is in the form of reactive pyrite and that the pyrite oxidation reaction is driven to completion. Oxygen dilution associated with carbon dioxide generation from carbonate dissolution (independently measured) was also taken into account in oxygen consumption calculations.
Despite the broad range in sulfur contents of the bulk samples, the sulfide-normalised PORs for all lithologies were similar and averaged 0.2 wt% FeS2 / year. These POR units (wt% FeS2 / year) mean that 0.2 wt% of all pyrite exposed to atmospheric oxygen will be oxidised to form sulfuric acid (H2SO4) per year. In this form, PORs can be used to produce annual (or monthly) acidity generation rates for any unsaturated sulfidic material, as long as the total mass of the material and its sulfide content are known (see box below).
For example, in Year 1 of operations, it was estimated that 3,668 kt of unsaturated waste rock (category AG4) would be produced, with an average 0.9 wt% FeS2. For a POR of 0.2 wt% FeS2 per year, this material can be expected to generate approximately 108 tonnes H2SO4 per year without any ARD management intervention (Table 2). This is equivalent to 29 kg H2SO4 per tonne of waste rock per year. Using this approach, the annual acidity generation loads produced during Years 1 to 5 of mining operations were quantified for all lithologies (see Table 2).
Table 2: Estimated annual acidity generation loads from unsaturated waste rock for Years 1 to 5 in the absence of ARD management strategies. The LAC (likely to be acid consuming) category was disregarded to provide a conservative assessment.
Development of waste rock designs
Once all of the mine waste rock materials were characterised, classified, scheduled for extraction and assessed for POR and annual acidity loads, management strategies were then formulated to lower the ARD risk, with a better understanding of limitations and resources.
A key conclusion from the data provided above is that the waste rock will require management to minimise or prevent offsite acidity discharges. Slightly more than 50% of the waste rock extracted will be acidity generating. The quantity of acid consuming material is very small and will only be produced in significant quantities in the first 2 years.
Based on material types, acidity generation rates and scheduling limitations, the following key principles of waste rock dump design were developed:
- Avoid end dumping construction techniques.
- Construct thin lifts from AG1-AG4 materials, starting from the base of the dump, optimise moisture addition and ensure maximum compaction to minimise air-entry.
- Encapsulate AG1-AG4 lifts with minimum 15 m thick UAF layers around AG1-AG4 cells to reduce oxygen ingress to diffusion control mechanisms.
- Ensure optimum moisture addition to UAF materials and maximum compaction to minimise air-entry.
- Permit thinly layered dump to grow upwards in thin AG1-AG4 lifts with protective UAF margins.
- Cap AG1-AG4 cells with UAF materials at various stages of dump construction when sufficient material is available, and continue with moisture optimisation and compaction of capping layers during construction.
- Selectively mine and stockpile materials that are LAC (in first 2 years) and utilise as natural alkalinity producing cover materials during the final stages of dump construction to optimise sulfide passivation and thereby progressively lower acidity generation.
- Install a “store and release” cover over the alkalinity producing cover materials in order to lower net percolation into the AG1-AG4 materials.
In combination, all of these approaches are expected to dramatically lower sulfide oxidation rates and acidity flux rates to the stage where natural dilution or possibly passive treatment can deal with the residual acidity release.