Overview of Acid Mine Drainage Prediction Methods

The potential of mineral wastes to form acid, mobilise pollutants and affect natural resources (land, soil, water, ecosystems, biodiversity and others) is often evaluated by field measurements, various laboratory tests and predictive modelling approaches. These techniques constitute what is sometimes known as the "Wheel Approach" for predicting the future drainage chemistry and potential environmental impacts as summarised in Figure 2.1 (Morin and Hutt, 1998; Maest et al., 2005; Elaw, 2010).

Price (2009) proposed three main steps to follow in order to identify and predict AMD formation, namely (1) determination of the environmental baseline conditions at the mine site (water quality objectives, environmental values, etc.), (2) measurement of the existing drainage chemistry and (3) determination of the potential future drainage chemistry. The types of analyses and tests that are conducted under each of these prediction steps have been

FIGURE 2.1

The Wheel approach to predict acid mine drainage formation. (From Morin and Hutt, 1998.)

reviewed in many publications (Sobek et al., 1978; Coastech Research, Inc., 1989; Skousen et al., 1990; USEPA, 1994; Lawrence and Wang, 1996; White et al., 1999; Morin and Hutt, 1997; 1998; 2001; Lapakko, 2002; Rae et al., 2007; Price, 2009; Lottermoser, 2015; Jones et al., 2016).

The Wheel approach given in Figure 2.1 is the accepted standard practice in AMD prediction and metal leaching studies whereby several geochemical static and kinetic tests are carried out and the results compared (Morin and Hutt, 1998). The most common static tests are acid-base accounting (ABA), mineralogy studies, total metal content and whole-rock analysis, retention (and reaction product) tests and net acid generation (NAG) tests. The main kinetic tests are laboratory-based kinetic tests (humidity cells, leach columns and the international kinetic database [IKD]) and field-based kinetic tests (large-scale bins or cribs, i.e., test pads or piles), mine wall stations and routine site monitoring (Morin and Hutt, 2001; 1997; 1998). The IKD contains pre-test characterisation and overall test results that help to compare kinetic data with that from other mine sites (Morin and Hutt, 2001). According to Morin and Hutt (2001) and Price (2009), there is no single test that can provide a basis for a reliable AMD prediction alone, but rather prediction data should be derived from various tests and sources.

2.3.1 Water Quality Survey

The best method to assess the present situation of the water quality and identify AMD presence is by measuring the chemistry of existing drainage through field observations, measurements and monitoring of surface seepage and groundwater quality (i.e., field water quality survey) discharged from all types of mineral wastes such as overburden material, ore stockpiles, heap and dump leach residue materials, tailings and waste rocks impoundments and bedrocks (walls of open pit mines, underground mine workings, etc.) (Rae et al., 2007; Price, 2009). Apart from water quality survey, the laboratory prediction tests on all mine waste materials are necessary to confirm AMD formation (ZCCM-IH, 2005). The poor effluent quality in the form of AMD may contain sulphuric acid, toxic metal ions and sulphates that present a risk to water resources, vegetation, aquatic life and human livelihoods.

One of the simplest and most useful field parameters to measure is effluent pH, which is an indicator of free acid. The pH value also gives an indication of whether (or not) sulphide oxidation has exhausted the acid neutralising capacity of the mineral wastes. Electrical conductivity (EC) is another useful parameter that indicates leachable soluble salts (salinity) (Jones et al., 2016). The typical chemical characteristics of AMD may be low pH (1.5-4.0), high concentration of metal ions (Fe2+/3+, Cu2+, Pb2+, Zn2+, Al3', Mn2+, Cd2+, As2+/3+), high sulphate concentration (500-10,000 mg/L) and low turbidity (total suspended solids). Some of the key visual indicators of AMD presence are red coloured water, orange-brown precipitates (of iron oxides), death of fish and corrosion of steel/concrete structures (Price, 2009).

The pH, EC, oxidation/reduction (redox) potential and total dissolved solids (TDS) of AMD samples are easily measured by portable pH meters. The liquid samples are also commonly analysed for the concentration levels of metal ions and sulphates by techniques such as atomic absorption spectrophotometry (AAS) and inductively coupled plasma atomic emission spectroscopy (ICP-AES). The measured values may be compared with water and effluent discharge standards to determine whether the drainage meets the water quality objectives. All these measurements are carried out so as to characterise the status of any exposed sulphidic material.

2.3.2 Geochemical Static Tests

Many types of static tests have been developed to measure the geochemical properties of mineral samples at mine sites in order to predict the potential future drainage chemistry. They mainly aim to quantify the maximum capacities of mine wastes to either produce acid or consume acid by measuring the theoretical balance between acid producing and neutralising components of a waste material. The static tests are simple, low-cost and rapid laboratory test procedures that are conducted in a matter of hours or days, but at one point in time to evaluate the net acid generation potential (AP) of samples. These tests form the basis of predicting AMD formation (Morin and Hutt, 2001; Price, 2009; Lottermoser, 2015). According to Price (2009), static tests data is useful for identifying materials with little AMD potential and can also help to develop criteria to classify and segregate mineral wastes for separate disposal and mitigation procedures.

2.3.2.1 Acid-Base Accounting Method

The commonest static tests fall under the ABA methods and its various modifications. The characteristic tests in the ABA procedure include the determination of sulphur species (sulphide, sulphate and total sulphur), AP, acid neutralisation potential (NP) and the net acid NP. The other key analytical tests that must accompany ABA methods are the mineralogy, elemental composition, soluble components, paste pH and the related direct NAG test procedure (Sobek et al., 1978; USEPA, 1994; White et al., 1999; Morin and Hutt, 2001; Rae et al., 2007; Price, 2009; Lottermoser, 2015; Jones et al., 2016). A combined data interpretation from static tests, kinetic tests, drainage chemistry and baseline water quality is necessary to fully understand and predict the potential of AMD formation in the future. The objectives, procedures and limitations of various static tests are discussed in the sections that follow.

The earliest approaches in the development of the ABA method in the 1960s focussed on assessments of the potential of coal mine wastes (i.e., overburden and mine spoil) for revegetation based on rock types, acidity and alkalinity. This included an assessment of the need for lime and suitability for plant species. By the 1970s, the AP of rocks from coal seams was reported and a system to balance acid and alkaline producing potential of rocks was developed. Later, the important role of acid neutralising minerals was recognised and quantified, and this became known as the NP (West Virginia University, 1971; Grube et al., 1973; Smith et al., 1974; Smith et al., 1976; Skousen et al., 1990; Kania, 1998; Perry, 1998). Based on these early publications, Sobek et al. (1978) formally presented the detailed laboratory procedures for performing the ABA method, which is now termed the Sobek method and is frequently cited and considered as the source document and basis of AMD prediction methods.

The ABA method, also known as the EPA 600 ABA, is the best known and most widely used method for predicting the potential future drainage chemistry from mine wastes (Ferguson and Erickson, 1988; Morin and Hutt, 1997; 2001). The theoretical principle in the ABA method is that AMD formation in the future can be predicted by a quantitative determination of the total amount of acidity and alkalinity a particular mine waste material can potentially produce. The ABA method consists of geochemical analyses and calculations to evaluate the potential for mine wastes to produce net acidic drainage if exposed to air and water. It is, thus, a quantitative estimate of the balance between the acid generated from the oxidation of sulphide minerals and the acid consumption by the carbonate minerals. The capacity of a mineral sample to generate net acidity is based on measurement of sulphur species, the calculation of AP and the determination of the acid NP.

The difference between AP and NP values indicates the net acid production potential (NAPP) of the waste material (Morin and Hutt, 1997; 2001; Kania, 1998; Lottermoser, 2015). One of the inherent limitations of the ABA methods as discussed by Lottermoser (2015) is that they do not predict the time when acid generation would occur. In addition, the tests do not consider the role of microbial activity in catalysing acid formation. Furthermore, the ABA method provides no information about effluent quality and the individual mineral reaction rates for acid generation and neuralisation. The large differences between laboratory tests and actual mine site conditions have also been criticised (Morin and Hutt, 2001; Lottermoser, 2015; Parbhakar-Fox and Lottermoser, 2015). Thus, kinetic tests have been developed to compliment static tests data by calculating the rates of acid formation in order to fully assess the potential of future water quality. The details of the main laboratory procedures in the ABA method and its modifications are discussed in the sections that follow.

2.3.2.2 Sulphur Analysis and Acid Generation Potential Calculation

The acid generation potential or simply AP, also referred to as the maximum potential acidity (MPA), is the total acid that mine wastes can produce. Thus, AP is determined by calculating the theoretical amount of acid that can be produced if the total amount of sulphur in the mineral sample is oxidised to sulphuric acid (Lawrence and Wang, 1996). It is used to estimate the maximum acid production potential based on the concentration of sulphur and/ or its forms (i.e., total sulphur (S), sulphide (S2~ / S2~), sulphate (SO2 ) and organic sulphur). Thus, sulphur is the primary source of acid through oxidation (Perry, 1998; Price, 2009). In the original Sobek method, the AP is stoichiometrically calculated using the percent (%) total sulphur (S) content that occurs only as pyrite (FeS2) as given in Equation 2.1. The AP is commonly expressed in units of kilogram CaC03 equivalent per tonne of the sample (kg/t) (or kg H2S04 per tonne). Alternatively, a conversion factor of

30.6 is used in Equation 2.1 to estimate MPA values in kg H2S04/t (Weber et al., 2005).

Thus, AP (kg CaC03/t) = % total sulphur x 31.25 The conversion factor (31.25) is derived from the stoichiometric consideration of the standard pyrite oxidation reaction (2.2) and the acid neutralisation reaction (2.3) for calcite (Morin and Hutt, 1997; 2001; Price, 2009). It is assumed, in Equation 2.2, that all sulphur occurs only as a sulphide (S2~) and also that the sulphide occurs only in pyrite (FeS2) with oxygen and water being the only oxidants. In addition, it is assumed that pyrite is completely oxidised to sulphate (S04 ) and ferric hydroxide, and H+ ions are neutralised by CaCO, (Morin and Hutt, 1997; 2001).

Overall reaction is given in Equation 2.4 as follows:

Therefore, from Equation 2.4, one mole of oxidised pyrite requires two moles of calcite to neutralise the acid formed. By mass basis, 1 g of sulphur present requires 3.125 g of calcite. Using the units of "parts per thousand" (ppt) of mine waste, for each 10 ppt sulphur (i.e., 1% sulphur) present, 31.25 ppt calcite is required to neutralise the acid (Morin and Hutt, 2001; Perry, 1998). Similarly, from Equation 2.2, for every one (1) mole of pyrite that is oxidised two moles of sulphuric acid are produced. Therefore, the MPA value of a mine waste material containing 1% S (as pyrite) is 30.6 kg of H2S04 per tonne (Smart et al., 2002; Weber et al., 2005). Hence, the MPA value may be calculated from Equation 2.1 using a conversion factor of 30.6. Although different conversion factors for sulphur to AP exist if pyrite is oxidised by other oxidants, for example, 15.63 using Fe3+ and 125.0 using Mn, the standard and more accurate conversion factor is 31.25 (Morin and Hutt, 2001).

However, the use of total sulphur (mainly a sum of sulphate and sulphide species) to estimate AP in mine wastes is considered as a conservative approach because not all forms of sulphur produce acid (Morin and Hutt, 2001). For example, the sulphate form of sulphur or simply sulphate-sulphur in gypsum, anhydrite and barite minerals does not generate acid, but sulphate in jarosite, alunite and melanterite is acid generating. The sulphide form of sulphur or sulphide-sulphur that generate acid occurs mainly in iron sulphide minerals (pyrite, pyrrhotite, arsenopyrite) and chalcopyrite, but chalcocite and covellite yield less acid than pyrite while sphalerite and galena may not yield acid at all (refer to Chapter 3 as well). Furthermore, organic sulphur (in coal mines) does not produce net acidity (Price, 2009; Jones et al., 2016). Thus, the use of total sulphur may overestimate the AP in a sample. For these reasons, the modified ABA (or Sobek) method uses sulphide-sulphur rather than total sulphur in Equation 2.1 to estimate AP (Coastech Research, Inc., 1989).

The accurate determination of sulphur species is a crucial step in the prediction of AP. A variety of analytical methods exist to determine the concentration of sulphur species in mine wastes, and these include volatilisation (roasting/pyrolysis), wet chemical extraction, mineralogical analysis and solid phase elemental analysis (Price, 2009). The standard analytical procedure for total sulphur analysis involves volatilisation of the sample by roasting (or pyrolysis) at 1500-1700°C in a Leco high temperature induction furnace followed by sulphur dioxide analysis in the gas phase. In the American Society of Testing and Materials (ASTM) method (E-1915-97) as presented in ASTM (2000) and reviewed by Lapakko (2002), the mineral sample is ignited at 1500-1700°C to convert all sulphur species to sulphur dioxide gas which is then analysed by absorption spectrometric techniques. To measure the sulphide-sulphur content, the sample is heated at a relatively lower temperature of 550°C in a muffle furnace. Here, it is assumed that only sulphides are converted to sulphur dioxide and not sulphate-sulphur. The sulphate content is the difference between the total sulphur in the original sample material, measured at 1500-1700°C, and total sulphur (also measured at 1500-1700°C) in the residue material that is left after pyrolysis that is carried out at lower temperature of 550°C as given in the ASTM method E-1915-99 (ASTM, 2000; Lapakko, 2002; Price, 2009).

The response to pyrolysis treatment differs for different minerals. The sulphide in iron sulphides (pyrite (FeS2), marcasite (FeS2), arsenopyrite (FeAsS) and pyrrhotite (Fe0.X)S)) may be volatilised completely. However, the loss of sulphide from copper sulphides (bornite (Cu5FeS4), chalcopy- rite (CuFeS2)), pentlandite ((Fe,Ni)qSs), galena (PbS) and sphalerite (ZnS) is reported to be minor, which then underestimates the total sulphide content. In addition, hydrated sulphate minerals (gypsum (CaS04-2H20), jarosite (KFe3(S04)2(0H)„)) may decompose partially, resulting in overestimation of the sulphide species (Bucknam, 1999; Li et al., 2007). Therefore, mineralogical analysis must always be taken into account when selecting a method for determining the sulphur species (Lapakko, 2002). The importance of mineralogy in AMD prediction is discussed further in Subsection 2.3.2.7.

The estimation of sulphur species can also be carried out by various wet chemical extraction procedures where mine wastes are digested either by water to measure soluble sulphates (FeS04-7H20, MgS04-7H20 and CaS04-2H20) (Li et al., 2007), or by hydrochloric acid to remove sulphate- sulphur (Sobek et al., 1978; Tuttle et al., 2003; Ahern et al., 2004; Li et al., 2007), or by sodium carbonate to remove less soluble sulphate minerals (Bucknam,

  • 1999) and by nitric acid to measure sulphide-sulphur (Sobek et al., 1978). The detailed laboratory procedures can be found in various relevant references. For example, in the EPA-600 method, a mine waste sample is milled and leached with 40% hydrochloric acid to determine sulphate-sulphur (acid soluble). The leach residue material is then analysed for residue total-sulphur by the high temperature pyrolysis technique in a Leco induction furnace. The sulphur lost or dissolved by leaching, that is, acid soluble sulphate-sulphur, is taken as the difference between total sulphur in the sample before and after extraction. The sulphate-sulphur is then subtracted from the total-sulphur to determine the sulphide-sulphur content, which is used to calculate the AP of the material. This method is not suitable if acid insoluble sulphate minerals are present since AP may be underestimated (Price, 2009).
  • 2.3.2.3 Acid Neutralisation Potential

The acid NP or acid neutralisation capacity (ANC) method measures the capacity of mine waste materials (tailings, waste rocks) to neutralise the acid that is produced. The alternative to the ANC test is the costly acid buffering characteristic curve (ABCC) test that also measures the ANC of waste materials. The main sources of acid neutralisation are carbonate minerals (calcite (CaC03), dolomite (Ca,Mg(C03)2), magnesite (MgC03)) and some reactive silicate minerals. The most common analytical methods used to measure NP are based on the Sobek method and its modified forms that use either bulk acid neutralisation or carbonate acid neutralisation techniques (Lawrence and Wang, 1996; Price, 2009). In the Sobek or US EPA-600 method (Sobek et al„ 1978), NP is determined experimentally by adding hydrochloric acid to a finely ground sample and digesting under boiling conditions. The volume and strength of hydrochloric acid added is determined by the fizz test. The strength of the fizz resulting from the addition of drops of acid to the sample is an indicator of the amount of reactive carbonate minerals present. For example, for "no fizz", add 20 mL 0.1 N; "slight fizz", add 40 mL 0.1 N; "moderate fizz", add 40 mL 0.5 N and for a "strong fizz", 80 mL 0.5 N of hydrochloric acid is added (Price, 2009). After the sample has been cooled, the residue acid in the slurry is determined by titration with sodium hydroxide to pH 7.0 so that the amount of acid consumed by the waste material can be calculated (in kg CaC03/t). The NP value is then compared with AP value to calculate the NAPP of a sample. Although the basic laboratory procedure is the same, the NP method based on Sobek et al (1978) has been modified by Coastech Research, Inc. (1989) and Lawrence and Wang (1996) mainly to provide longer acid digestion time for samples. Furthermore, the pH end point during titration is different (pH 8.3 for the modified NP method). All these methods aim to ensure that sufficient acid is added to dissolve all carbonates and reactive silicates (Price, 2009).

According to Price (2009), the Sobek method is fast (3-4 h) and widely used to determine NP, while the modified NP method by Lawrence and Wang

(1996) require over 25 hours which may be costly, time consuming and practically difficult to implement. The errors in these methods may come from misinterpretation of fizz test results, leading to incorrect determination of the amount of acid to be added, and any strong acid employed may dissolve minerals that do not contribute to acid neutralisation.

Other less common bulk NP methods include the British Columbia (BC) Research Initial test (Duncan and Bruynesteyn, 1979) and the Lapakko test (1994) in which NP is determined by the amount of acid required to attain pH values of 3.5 and 6.0, respectively. It must be noted that different methods can produce different NP values for the same sample due to differences in particle size, digestion conditions (acid used, pH, temperature and duration) and endpoint pH of titration (Lapakko, 1992; Lapakko, 1994).

Another technique that can be used to estimate the acid NP is called the carbonate neutralisation potential (CO,-NP). This is the ANC that a waste material possesses if all the carbonates reacted like calcite. The carbonate concentration is calculated from total carbon (i.e., carbon present as carbonate, organic carbon and graphite), inorganic carbon or carbon dioxide as shown in Equation 2.5:

where C is percentage of total carbon.

If all carbon occurs as carbonate, the total carbon is measured by Leco furnace facilities. For high concentration of non-carbonate carbon (coal, organic matter, graphite), the acid soluble analysis of carbonate-carbon is used whereby carbonate minerals are converted to carbon dioxide by hydrochloric acid (Price, 2009).

2.3.2.4 Net Acid Production Potential

The NAPP, as used in Australia and Asian countries, is calculated as the difference between the MPA and ANC values (Equation 2.6a) (Price, 2009; Jones et al., 2016). The NAPP value gives an acid-base account of the mine waste materials, which is then interpreted to predict AMD formation. When the NAPP value is negative (i.e., MPA < ANC), then the mine waste has sufficient neutralisation capacity to prevent acid formation. However, the material may generate acid if the NAPP value is positive (i.e., MPA > ANC). Thus, NAPP data may then be used to classify mine waste materials as either potentially acid forming (PAF) or non-acid forming (NAF). Figure 2.2 shows that the risk of acidic drainage is high if NAPP values are located in the risk domain. However, a related term that is commonly used in North America is the net neutralisation potential (NNP) which is the difference between NP and AP values (Equation 2.6b). Similar to the interpretation of NAPP, a mine waste may generate acid if NNP value is found to be negative (i.e., NP < AP) and may neutralise any acid formed if NNP value is positive (i.e., NP > AP) (Coastech Research, Inc., 1989; USEPA, 1994; White et al., 1999; Price, 2009).

Another ABA classification criterion is based on the net potential ratio (NPR), that is, ANC/MPA or NP/AP ratio that is used to assess the acid producing

FIGURE 2.2

Acid-base accounting data interpretation plot. (From Jones et al., 2016.)

or acid consuming (ACM) potential of mine wastes (Morin and Hutt, 1994; Price, 2009; Jones et al., 2016). Lawrence and Wang (1996) stated that the use of NPR to interpret ABA test data is favoured because it clarifies the relative amounts of acid producing and ACM phases. A range of NPR values has been recommended in the literature to differentiate the potentially acid producing materials from ACM ones. Jones et al. (2016) cited NPR values of 1.5 to 3, but recommended safe values of >2 to ensure that the material will remain near-neutral in pH and that it will not result in AMD formation. According to the universal criteria given in Table 2.1 by Morin and Hutt

  • (1997), an NPR < 1 or NNP < 0 t CaC03/1000 t indicates that the material has insufficient neutralisation capacity and, therefore, should eventually result in net acidic outcome. However, the material may remain near-neutral or alkaline indefinitely if NPR > 1 or NNP > 0 t CaC03/1000 t.
  • 2.3.2.5 Paste pH

The analytical procedure for paste pH as presented by Sobek et al (1978) provides a simple method to assess whether a material will be acidic, neutral or alkaline at the time of analysis. It involves pulverising a sample (to dp < 100 pm) and mixing it with distilled water at -pH 5.3 to form a slurry from which paste pH is measured. As presented in Table 2.1, a paste pH of less than 5.0 indicates net acidity, pH values between 5.0 and 10.0 are regarded as near neutral and values above pH 10.0 indicate net alkalinity at the time of the test, but the future drainage pH is non-predictable in all cases.

TABLE 2.1

Universal ABA Criteria for Assessing and/or Predicting Drainage pH from Mine Wastes

Criteria

Prediction / Current Condition

Paste pH

I

Paste pH <5.0

Currently acidic; future non-predictable

II

5.0 < /Paste pH < 10.0

Currently near neutral; future non-predictable

III

Paste pH > 10.0

Currently alkaline; future non-predictable

NPR or NNP*

A

NPR < 1.0 or NNP < 0.0 t CaC

Eventually acidic

В

1.0 < NPR < 2.0 or 0 < NNP < 20

Uncertain future

C

NPR > 2.0 or NNP > +20 t CaC03/1000 t

Indefinitely near neutral or alkaline

Source: Morin and Hutt, 2001.

  • •' Tine static tests are based on sulphide-sulphur.
  • 2.3.2.6 Net Acid Generation

According to Price (2009), the NAG test utilises hydrogen peroxide (H,02) to rapidly oxidise sulphides in order to assess the capacity of mine waste materials to neutralise acid formed by sulphide oxidation, for example, pyrite (Equation 2.7). The NAG test is based on the original hydrogen peroxide method by Sobek et al. (1978). The acid formed may subsequently solubilise carbonate (or neutralising) minerals to give a net effect between acid forming and neutralising reactions which can then be measured directly in the NAG test. This implies that the test directly measures the net quantity of acid formed by a mine waste sample. The test is simple, rapid and relatively cheap, which helps to identify PAF waste materials at mine sites. The NAG test is often used in association with the NAPP values to classify the acidgenerating potential of waste materials (Miller et al., 1994; Lapakko, 2002; Smart et al., 2002).

Several types of the NAG test have been developed to accommodate the wide geochemical variability of mine waste materials. These include the single addition NAG test (low S^"), sequential NAG test (high Si-), partial ABA (for total S, NP and paste pH), kinetic NAG test and the ABCC (Smart et al., 2002; Price, 2009). The theory, detailed experimental procedures and limitations of these tests can be found in laboratory manuals by Smart et al. (2002) and Price (2009). The NAG test procedures commonly used are the single addition NAG test and the sequential NAG test. The sequential NAG test is applicable if mine waste samples have high total sulphur or sulphide sulphur in order to measure the total acid-generating capacity, and also for samples with high ANC.

The NAG test is done by mixing hydrogen peroxide with finely milled mine waste samples and heating the mixture until effervescence stops followed by cooling and recording the pH (or NAG pH). The mixture is filtered and titrated with sodium hydroxide (NaOH) to pH 4.5 and 7.0. The first pH value of 4.5 accounts for acidity due to Fe, A1 and most of the H+ ion, and the pH between 4.5 and 7 indicates soluble metals (e.g., Cu and Zn). A pH after the sulphide oxidation reaction (NAG pH) of less than 4.5 indicates that the mine waste sample is PAF or potentially acid generating (PAG). However, if NAG pH is greater than 4.5, the sample may be considered as non-acid forming (NAF) or non-PAG. The amount of acid produced is determined by titration and expressed in the same units as NAPP (kg H2S04/tonne) (Smart et al., 2002) as given by Equation 2.8. The classification criteria for mine waste material based on NAPP and NAG test data are shown in Table 2.2.

where V = amount of NaOH (mL), M = concentration of NaOH (moles/L), W = weight of rock sample (g).

The data from the NAG test is useful to assess the risks of acid formation, sulphide reactivity and the quantity of acid that may form. This information helps to develop plans to implement mine waste management strategies to prevent AMD formation (Miller et al., 1994). However, one of the limitations of the NAG test is that hydrogen peroxide may decompose before it reacts with all sulphides and thus underestimate the total acid-generating

TABLE 2.2

Geochemical Classification Criteria

Primary Geochemical Material Type

NAPP (kg H2S04/tonne)

NAG pH

Potentially acid forming (PAF)

>10

<4.5

Potentially acid forming - low capacity (PAF-LC)

0 to №

<4.5

Non-acid forming (NAF)

Negative

>4.5

Acid consuming (ACM)

Less than -100

>4.5

Positive

>4.5

Uncertain15

Negative

<4.5

Positive

<4.5

Sources: Smart et al., 2002; Jones et al., 2016. a Site-specific, 5-20 kg H2S04/ tonne. b Further testing required.

potential. The use of pH 4.5 criteria also can imply that Non-PAG classification might include weakly acidic mine waste samples that can potentially solubilise toxic trace elements. In addition, the rapid oxidation of sulphides may not simulate expected field conditions (Price, 2009).

2.3.2.7 Mineralogical and Elemental Analyses

The prediction of AMD formation also depends on mineralogy as shown in the "Wheel approach" by Morin and Hutt (1998). Thus, mineralogical characterisation of mine waste materials is essential to understand the actual mineral phases that drive AMD formation. A mineralogical study should reveal mineral type, composition, abundance, particle sizes, liberation and grain size distribution of minerals. A combination of characterisation methods are often used such as optical mineralogy, petrographic, X-ray diffraction (XRD), scanning electron microscopes (SEM), micro-probes, quantitative evaluation of materials by scanning electron microscopy (QEMSCAN) and bulk elemental analyses (such as X-ray fluorescence (XRF), ICP-AES). The chemical analysis of elements is accomplished by whole rock or near-total solid phase analysis by hot acid digestion method and leachate analysis by XRF or ICP-AES techniques (Price, 2009). The aim of the characterisation is to identify the sulphide and carbonate minerals that are responsible for acid formation and neutralisation, respectively. However, silicates (e.g., anor- thite, olivine and chlorite) also have long-term acid neutralising properties but may be ineffective due to slow reaction kinetics. The knowledge of mineralogy not only helps to indicate the mineral phases that may contribute to acid formation and NP in static tests, but also to establish the likely source of harmful elements in the leachates. The characterisation may also provide information about the concentration of elements of interest relative to background rocks/soils. The common acid-forming sulphides are pyrite (FeS2), pyrrhotite (FeS), marcasite (FeS2), chalcopyrite (CuFeS2) and arsenopyrite (FeAsS). However, not all sulphides are acid-generating (see Chapter 3) during oxidation, for example, covellite, sphalerite, galena, chalcocite and born- ite may not form acid when oxidised, but may release metals when exposed to acidic water (Morin and Hutt, 2001; Smart et al., 2002; Rae et al., 2007; Price, 2009).

2.3.3 Geochemical Kinetic Tests

According to Coastech Research, Inc. (1991), the kinetic prediction tests are follow-up geochemical test procedures that may be carried out on mine wastes if the results from static tests are either uncertain or predict potential for AMD formation. Therefore, the aim of kinetic tests is to understand whether, and/or when, AMD is likely to occur by predicting the long-term oxidation (or dissolution) rates of mine wastes (in months to years) (Coastech Research, Inc., 1991; Parbhakar-Fox and Lottermoser, 2015; Dold, 2017). The specific objectives (Coastech Research, Inc., 1991; Lapakko, 2003; Price, 2009; Lottermoser, 2015) may be to:

  • • Determine the rates of acid generation and neutralisation,
  • • Determine whether AMD is likely to occur and its time frame,
  • • Estimate the rate of metal leaching from mine wastes,
  • • Establish the chemical composition and leachate quality (drainage chemistry),
  • • Identify the main chemical weathering reactions, and
  • • Confirm results from static tests, hence, AMD generation potential.

A number of kinetic tests have been developed to complement static tests data in the prediction of AMD formation. These kinetic tests are categorised as either laboratory-based (e.g., humidity cells, leach columns, shake flasks, biokinetic tests, soxhlet extraction, BC research confirmation tests and the use of the international kinetic database) or field-based (e.g., large-scale field test pads (i.e., bins or cribs or piles), mine wall stations and routine site monitoring) (Coastech Research, Inc., 1991; Morin and Hutt, 2001). The basic concepts, principles and inherent limitations of the most popular kinetic tests with wide applications, i.e., humidity cell test, leach columns and biokinetic tests, are briefly discussed in the sections that follow.

2.3.3.1 Humidity Cell

Morin and Hutt (1997) reported that the humidity cell was the kinetic test of choice. This is now the recommended and most popular kinetic test. As discussed by Lapakko (2003), the original humidity cell testing method based on Sobek et al. (1978) has been revised by many researchers, with the notable method being the ASTM5744-96 (revised 2007-2018). The humidity cell test simulates the geochemical weathering of mine wastes (e.g., rocks or tailings) in order to estimate mainly the rate of acid generation and the quality of the leachate. The test results may indicate whether the sample will generate acidic, neutral or alkaline drainage and types of dissolved species (e.g., metals and sulphates) and rates of their release under controlled conditions. The test also helps to address any uncertainty in data from static tests.

The principle of the humidity cell test is that a suitable sample size of the waste material is crushed to a particular particle size and loaded into a column of suitable dimensions. The loaded column is then subjected to alternating cycles of dry air (3 days) and humid (or moist) air (3 days). Thereafter, water is percolated through the column to soak, rinse or leach the material (1 day) which is later discharged as a leachate. The quality or chemistry of the leachate is analysed for metal ions, sulphate, pH, conductivity, redox potential, acidity and alkalinity (Sobek et al., 1978; Lawrence et al., 1989; Morin and Hutt, 1997; 2001; Lapakko, 2003; Coastech Research, Inc., 2008, Lottermoser, 2010).

The ASTM D8187-18 method (ASTM, 2018) provides a detailed discussion on the interpretation of results from the humidity cell test. Basically, the results are analysed to assess the potential of the mineral wastes in forming AMD based on leachate composition, for example, a pH of 3-5 indicates that the sample is PAF. In addition, the acidity and alkalinity measurements refer to the balance between acid and alkaline generating minerals and a high redox potential (>500 mV) indicates a high sulphide oxidation rate by ferric ions. Plots of pH, acidity/alkalinity, metal ions with time help to establish the variation in leachate chemistry and deduce evidence for the potential to form AMD (Coastech Research, Inc., 2008).

The humidity cell test is simple and reliable. It predicts the leachate quality expected from waste materials, and thus, it is useful in helping to assess the initial plans that can prevent AMD formation. However, the humidity cell test is costly and requires a lot of time to be completed or to generate meaningful data (Coastech Research, Inc., 2008). Furthermore, the equilibrium weathering conditions that are expected in actual mine waste dumps might not be reproduced in the humidity cell test due to relatively limited reaction time that is employed under laboratory conditions (Parbhakar-Fox and Lottermoser, 2015).

2.3.3.2 Leach Columns

The leach column tests are designed to simulate leaching conditions in mine waste materials under partial or full saturation conditions and/or deprived of oxygen. These tests are often carried out on scale that is larger than that of humidity cell test (Lawrence and Day, 1997; MEND, 2000). The most commonly used column type is the free draining leach column test that utilises the Buchner funnel or drums depending on the size of the sample. The columns are loaded with crushed mine waste material and exposed to cycles of wetting and drying (with heat lamps) to oxidise the material and flush the reaction products. The loaded samples are wetted by addition of water (weekly) to the surface of the column and leachates collected at the column base (monthly). The testing period depends on the characteristics of the waste material and the objectives of the project. The leachate quality is analysed for metal solubility, oxidation kinetics, sulphide reactivity and the leaching characteristics of mine waste materials (Smart et al., 2002). However, columns can be operated such that they are sub-aerial type (well- drained waste materials) or sub-aqueous leach columns (waste materials are flooded) (Price, 2009).

Apart from determining the rates of acid formation, sulphide oxidation and depletion of NP, a column test may reliably predict the drainage quality and dissolved metals expected from mine wastes (Coastech Research, Inc.,

  • 2008). The column test set-up may be modified to accommodate different grain sizes and sample mass of waste materials and the frequency of leachate collection can also be varied (Parbhakar-Fox and Lottermoser, 2015). It is claimed by some studies that experimental conditions in the column test closely simulate actual field conditions than the humidity cell test (Bradham and Caruccio, 1990; Maest et al., 2005), but this view was refuted by Morin and Hutt (1997), who noted that laboratory conditions may not reproduce field conditions and that columns rarely attain steady-state conditions to give reliable estimates of oxidation rates. Furthermore, solution channelling in columns can be a challenge (MEND, 2000), and there is no provision to integrate mineralogical assessment during the test (Parbhakar-Fox and Lottermoser, 2015).
  • 2.3.3.3 Biokinetic Tests

The effect of iron and sulphur oxidising bacteria (e.g., Acidithiobacillusferrooxi- dans) on the rate of acid formation by mine waste materials has been studied by the shake flask test (USEPA, 1994), the biological acid producing potential (BAPP) test (Parbhakar-Fox and Lottermoser, 2015) which is based on the BC Research confirmation test (USEPA, 1994) and the relatively recent biokinetic test (Hesketh et al., 2010). In the BC Research confirmation test, a mine waste sample is milled to form an acidic slurry in shake flasks, which are then inoculated with Acidithiobacillus ferrooxidans and incubated for the bacterial adaptation followed by frequent monitoring of pH and other water quality parameters. A final pH value that is below 3.5 indicates microbial activity and the potential for the sample to form acid. For the pH above 3.5, the sample is considered NAF. One major criticism in this test is the use of a single bacterial species (Hesketh et al., 2010; Parbhakar-Fox and Lottermoser, 2015). The shake flask test is similar to the BC Research confirmation test except that consideration of the effect of bacteria on the kinetics of acid formation is optional.

The biokinetic test, as presented by the work of Hesketh et al (2010), is an improvement to the existing shake flask tests. The objective of this test is to determine the potential and likelihood of acid formation by mine wastes in the presence of bacteria and to determine the rates of acid formation and neutralisation reactions. In the test, the mine waste materials are milled to form an acidic slurry at pH 2 through sulphuric acid addition in flasks and inoculated with more than one bacterial species. The bacteria is a mixed culture of Acidithiobacillus ferrooxidans, Leptospirillum ferriphilum, Acidithiobacillus cal- dus and Sulfobacillus benefaciens that simulate typical microbial environments found at actual mine sites with AMD. The biokinetic test data (e.g., final pH, acid consumption) are able to validate the NAG pH, ANC and waste classification criteria under the static ABA methods. However, the timing is different between ACM and acid-generating reactions. The ANC of the waste material is depleted rapidly and controlled by the more reactive carbonate phases. The sulphides oxidise and form more acid over long term than that predicted by static tests. Thus, the biokinetic shake flask test is seen to be relatively simple, fast and of low-cost. However, there is still no practical use of this test to predict AMD formation (Parbhakar-Fox and Lottermoser, 2015).

 
Source
< Prev   CONTENTS   Source   Next >