Reactive transport modelling of porewater geochemistry and sulfur isotope fractionation in organic carbon amended mine tailings

Highlights

• Integrated simulation of sulfide oxidation, reduction and $\mathrm{\delta }$³⁴S–SO₄ fractionation.

• Sulfate reduction and metal-sulfide precipitation promote S, Fe, Mn and Zn removal.

• Simulated$\mathrm{\delta }$³⁴S–SO₄ fractionation demonstrates microbial control on SO₄ reduction.

• Simulation of observed geochemistry quantitatively supports conceptual model.

• Carbon amendments evaluated for in situ treatment of tailings porewater.

Abstract

Field experiments previously conducted to assess organic carbon (OC) amendments for in situ biological treatment of tailings porewater at the Greens Creek Mine (Alaska, USA) showed sulfate reduction, metal-sulfide precipitation, and decreased fluxes of sulfate and dissolved metals. Here, we develop two reactive transport models using the reactive transport code MIN3P to simulate hydrogeochemical processes and $\mathrm{\delta }$³⁴S–SO₄ isotope fractionation over four years in test cells containing unamended (control) and amended (5 vol % OC) tailings. These models successfully simulate observed data including pH, SO₄, $\mathrm{\delta }$³⁴S–SO₄, Ca, Fe, K, Mg, Mn, Si, and Zn. The models also indicate that dissolution of carbonate and, to a lesser extent, aluminosilicate minerals neutralize acidic porewater generated from sulfide mineral oxidation and sulfate reduction reactions. Application of a constant kinetic fractionation factor of 0.9820 to simulate measured $\mathrm{\delta }$³⁴S–SO₄ trends confirms that sulfate removal principally results from microbially-mediated sulfate reduction in conjunction with OC oxidation and subsequent metal-sulfide precipitation. Gypsum precipitation/dissolution and thiosulfate disproportionation have negligible effects on modelled porewater $\mathrm{\delta }$³⁴S–SO₄ signatures. Our simulations are consistent with the previous findings that metal-sulfide precipitation controls Fe and Zn attenuation in amended tailings and that coprecipitation reactions contribute to metal removal. Overall, these simulations demonstrate that coupled reactive transport modelling incorporating stable isotope fractionation can improve the understanding of hydrogeochemical and biogeochemical controls within in situ treatment systems, further illustrating the benefits and limitations of this technique for improving water quality of mine drainage.

Introduction

The oxidation of sulfide minerals in mine wastes can pose considerable environmental risks to regional water quality and the health of surrounding ecosystems if effluent is left untreated (Bussière, 2009; Blowes et al., 2014; Nordstrom et al., 2015). Sufficient quantities of carbonate minerals such as calcite [CaCO₃] and dolomite [CaMg(CO₃)₂], can neutralize acid generated by sulfide-mineral oxidation and maintain circumneutral pH porewater and drainage (Al et al., 2000; Blowes et al., 2014). Neutral mine drainage (NMD) from sulfide- and carbonate-rich mine wastes typically exhibits pH between 6 and 9, and contains elevated concentrations of SO₄, Ca, Mg, metals and metalloids (e.g., As, Cd, Cr, Co, Cu, Fe, Mn, Mo, Ni, Se, Sb, Zn) (Lindsay et al., 2015; Nordstrom et al., 2015; Vriens et al., 2019). Enhanced mobility of these sulfide-mineral oxidation products under circumneutral pH conditions creates greater potential for contamination of downgradient surface and groundwater resources if mitigation or management efforts are not utilized effectively (Blowes and Jambor, 1990; Vriens et al., 2020).

Although preventing sulfide-mineral oxidation in subaerial mine waste deposits is challenging, various passive remediation strategies have been developed to minimize these reactions and mitigate fluxes of associated reaction products (Johnson and Hallberg, 2005; Blowes et al., 2014). Organic carbon (OC) has been used in bioreactors (Neculita et al., 2007), permeable reactive barriers (PRB) (Blowes et al., 2000), and constructed wetlands (Webb et al., 1998) for remediation of mine-impacted drainage. These remediation strategies promote microbially mediated dissimilatory sulfate reduction (DSR), which follows the overall reaction:

$$2{\text{CH}}_2\text{O}+{\text{SO}}_4^{2-}\stackrel{SRB}{\to }2{\text{HCO}}_3^{-}+{\text{H}}_2\text{S}$$

In reaction (1), sulfate reducing bacteria (SRB) catalyze sulfate reduction coupled with the oxidation of simple OC compounds (CH₂O). This anaerobic respiration process generates HCO₃⁻ and H₂S, and can promote precipitation of secondary metal (Me²⁺) sulfide minerals:

$${\text{Me}}^{2+}+{\text{H}}_2\text{S}\to {\text{MeS}}_{\left(\text{s}\right)}+2{\text{H}}^{+}$$

The principle was first demonstrated by Tuttle et al. (1969) for treatment of acidic mine waters. Hulshof et al. (2003, 2006) expanded this approach by demonstrating that OC amendments can reduce fluxes and, therefore, discharge of sulfide-mineral oxidation products in saturated sulfide mine tailings.

Detailed investigations by Lindsay et al. (2009b, 2009c, 2011a, 2011b, 2011c) have further evaluated this technology using in situ OC amendments for limiting porewater concentrations of toxic elements in sulfide- and carbonate-rich tailings from the Greens Creek Mine (Alaska, USA). As part of this research, Lindsay et al. (2009b, 2009c, 2011a, 2011c) conducted a four-year field experiment to assess sulfate and metal removal in variably saturated test cells containing unamended tailings and tailings amended with different OC sources. Lindsay et al. (2011b) also performed column experiments under fully saturated conditions to examine the impact of tailings OC amendments on sulfate and metal removal rates.

Field experiment results showed that OC amended tailings removed >45% of the dissolved S mass from porewaters and decreased the mobility of Cu, Fe, Mn, Ni, Pb, Tl, and Zn, which was suggested to occur by (co)precipitation of secondary minerals and adsorption onto metal-oxy-hydroxide and OC surfaces. Observations from scanning electron microprobe (SEM) micrographs revealed secondary Fe–S and Zn–S precipitates associated with OC solids. A primary component of the field and laboratory experiments also involved assessing the microbiological control in DSR reactions, which included quantifying $\mathrm{\delta }$³⁴S–SO₄ and $\mathrm{\delta }$¹³C-DIC enrichment factors as a way of indicating the isotopic signature and overall influence of SRB. The measured $\mathrm{\delta }$³⁴S–SO₄ and $\mathrm{\delta }$¹³C-DIC enrichment illustrated the significance of SRB in the amended tailings, consistent with the overall decrease in sulfate concentrations in comparison to the unamended tailings.

Factors controlling $\mathrm{\delta }$³⁴S–SO₄ enrichment in the field experiments were suggested to include thiosulfate disproportionation and sulfide production as a function of increased SRB populations, whereas possible gypsum dissolution could lead to $\mathrm{\delta }$³⁴S–SO₄ depletion. Other possible geochemical processes that could influence the observed $\mathrm{\delta }$³⁴S–SO₄ isotope ratio include the oxidation of authigenic metal-sulfide minerals and/or dissolution of other secondary minerals, in addition to physical processes including dispersion and mixing. However, the extent to which these mechanisms influence the measured $\mathrm{\delta }$³⁴S–SO₄ isotope signatures is unclear. Developing a quantitative understanding of the relative effects of the hydrogeochemical and microbiological mechanisms affecting sulfate reduction and metal removal is a critical step in assessing the strengths and limitations of this treatment method.

Multicomponent reactive transport models are powerful tools for simulating dynamic hydrogeochemical systems in porous media, and have been applied to investigations that involve minimizing sulfide-mineral oxidation in mine-wastes through dry cover systems (Molson et al., 2008; Pabst et al., 2017; Raymond et al., 2020); mitigating sulfate and metal(loid) mobility from mine drainage through microbially-mediated DSR in OC layers/PRB systems (Amos et al., 2004; Mayer et al., 2006), and understanding the influences of DSR on sulfur isotope fractionation in OC-based mine drainage treatment (Gibson et al., 2011; Druhan et al., 2014). Collectively these studies demonstrate that by applying quantitative approaches through modelling, a more rigorous assessment of original conceptual models can be obtained. For example, Amos et al. (2004) was able to assess the contribution of secondary carbonate and sulfide minerals to iron attenuation in organic carbon treatment. Likewise, Druhan et al. (2014) assessed $\mathrm{\delta }$³⁴S signatures of dissolved and solid-phase sulfur species by applying a constant kinetic fractionation factor for microbially-mediated sulfate reduction reactions.

MIN3P (Mayer et al., 2002) is a mechanistic multicomponent reactive transport code that can be applied to variably saturated media. The code is capable of integrating solute and gas transport processes with a coupled reaction network that includes aqueous complexation, hydrolysis, isotope fractionation, redox reactions, ion exchange, as well as thermodynamically and kinetically controlled mineral dissolution-precipitation (Mayer et al., 2002, 2012; Jurjovec et al., 2004; Brookfield et al., 2006; Gibson et al., 2011; Wilson et al., 2018).

In this study, we use MIN3P to simulate geochemical reactions and transport processes within test cell experiments at Greens Creek Mine containing sulfide- and carbonate-rich tailings in the presence and absence of organic carbon amendments. Our principal objective is to provide a quantitative assessment of the conceptual models developed by Lindsay et al. (2009b, 2009c, 2011a, 2011b, 2011c) and to establish representative hydrogeochemical models of sulfide-mineral oxidation and sulfate reduction processes. These models were subsequently used to investigate the isotopic changes in $\mathrm{\delta }$³⁴S–SO₄ over time and to better understand the influence of biogeochemical reactions on DSR and metal-sulfide precipitation processes at circumneutral pH. The inclusion of $\mathrm{\delta }$³⁴S–SO₄ within the models is also used to illustrate how simulation of microbially-controlled mechanisms can be achived at the pilot-scale level, and the utility of isotopes when evaluating the effectiveness of passive strategies for treatment of mine drainage.