Bioremediation of Acid Mine Drainage Using Grass

Abstract

Large-scale mining of South Africa’s gold in the Witwatersrand, and coal in the Mpumalanga areas has contributed substantially to economic wealth and energy provision for the South African economy. Unfortunately, these operations have also resulted in considerable damage to surrounding environments. One of the largest sources of this large-scale mining-induced environmental damage in South Africa is acid mine drainage (AMD). AMD is formed when sulfidic rock is exposed to oxygen and water, and results in highly mineralised, acid water, which can infiltrate and contaminate surrounding groundwater, land, and surface watercourses (McCarthy, 2011a; Simate and Ndlovu, 2014). Where AMD volumes surpass the natural dilution effects, plants, animals, humans, and economic activities such as agriculture suffer negative impacts. Since South Africa is a water-stressed country, remediating AMD is critical to protect limited water resources. Active chemical treatment options for AMD remediation are generally implemented owing to their ease of control; however, these treatment options carry high capital and operating costs and produce large volumes of hazardous sludge waste. A more sustainable and cost-effective alternative to chemical treatment is biological treatment via the dissimilatory reduction (DSR) remediation process. DSR uses sulfate reducing bacteria (SRB) to reduce sulfates (SO42-) to hydrogen sulfide (H2S). The alkalinity produced from the reaction increases the pH and precipitations metals present in the system as metal sulfides (Ruehl and Hiibel, 2020; Burman, Harding and Sheridan, 2021). The DSR process requires supplementation with an organic carbon source to drive the remediation process (Van Rooyen, Van Staden and Du Preez, 2021). Grass can be used as a lignocellulosic substrate to provide the organic carbon and electron donor source to drive the DSR process. Grass is a renewable, secondary waste resource that is readily available as natural growth, garden waste, and roadside vegetation, and thus does not compete with primary agricultural output for food production, a particularly important consideration for countries which experience food insecurity (Magowo, Rumbold and Sheridan, 2015). However, the complex organic structure of the grass needs to be broken down to release the monomeric sugars in a form amenable to SRB before it can be used in the DSR process. In this work, the acidity of AMD is used as a weak acid source to disrupt the grass lignocellulosic structure. This disrupted lignocellulosic structure would render the cellulosic component in the grass amenable to subsequent enzymatic hydrolysis and fermentation for biofuel production. Using the grass residue from AMD bioremediation for bioethanol production provides a means to derive value from the remediation process which can help to offset the remediation costs. This study aimed to contribute to existing knowledge for the development of a more sustainable process for the biological remediation of AMD. The study also sought to assess the effects of varying AMD bioremediation process conditions on subsequent biofuel production. This research commenced with a characterisation of AMD from two AMD receiving dams in the Mpumalanga coalfields region in South Africa, and an evaluation of the seasonal variations on the physical and chemical parameters present in the AMD. Spatial and temporal trends presented changes in physio-chemical concentrations, which provides valuable information to inform remediation process development, and to identify possible limitations in prospective remediation processes. Data showed the presence of several metal(oid)s that were in excess of legislative limits. The removal of elevated concentrations of manganese (Mn) and zinc (Zn) are difficult by adsorption and are thus not easily supported by passive treatment systems. High concentrations of aluminium are difficult to remove via precipitation mechanisms owing to the amphoteric nature of aluminium hydroxides and would thus be less suited to active chemical treatment processes. This work demonstrated a strong correlation between electrical conductivity (EC) and total dissolved solids (TDS). Other major AMD parameters, such as iron (Fe) and sulfate (SO42-), showed similar temporal trends to EC and TDS. The research continued with a statistical evaluation of all the data from the two Mpumalanga AMD sites. This confirmed the strong EC/TDS correlation noted in the previous work and identified further EC correlated parameters which included total acidity (TA), dissolved iron (Fe), and dissolved sulfur (S). EC was thus identified as a useful correlative measurement for the prediction of TA, Fe and S concentrations in these acidic AMD waters. These empirical correlations were then leveraged to derive regression equations which were used to calculate EC values corresponding to legislative water quality limits and thus provide a novel rapid method for testing compliance. In a similar manner, EC can also be used as a control measure in remediation processes, particularly in the event of an uncontrolled AMD breach, where real-time measurements could trigger a fast response to minimise potential environmental harm. In the next part of the work, locally sourced grass types that were co-located with AMD sites in the Mpumalanga coalfields region, South Africa were identified, classified, and their physical and chemical properties determined. Experimental test work was conducted to assess the amenability of these grass types to increase the pH of AMD, thereby indicating their suitability for use in AMD bioremediation. Grass types, Chrysopogon zizanioides and Hyparrhenia hirta achieved the greatest increase in pH after single and multiple grass additions. Renewing the organic substrate improved the performance of the more poorly performing grass types. This was supported by an improvement in the anticipated positive correlation between the hydrogen ion concentration (H+) and oxidation-reduction potential (ORP). Decreased inorganic cation concentrations observed in the grass samples after contact with AMD indicated the likelihood of an ion exchange reaction between the H+ ions in the AMD and inorganic cations associated with the anions present in the grass. In the next component of the study, naturally occurring sources of SRB cultures were experimentally evaluated to determine their suitability to support SRB growth and activity in AMD remediation. The performance of the selected SRB culture sources was assessed at three different pH environments. Raw sewage (RS) collected from the Bushkoppies wastewater treatment facility near Soweto, and mud inoculum sampled from the reed beds (RM) surrounding a Mpumalanga AMD source, achieved the greatest conversion of sulfate to sulfide. The reduction of sulfate to sulfide is indicative of the DSR process which was shown to improve at neutral pH. Lastly, experimental work was conducted to evaluate the effect of hydraulic retention time (HRT) on AMD bioremediation using grass, and subsequent bioethanol production. To produce bioethanol from grass, the lignocellulosic structure of the grass must first be broken down to render the grass amenable to enzymatic hydrolysis. The premise for this work was that If grass were used for AMD remediation, the acidity from the AMD would be sufficient to disrupt the lignocellulosic structure of the grass. The opened grass structure will facilitate enzyme access to the remaining sugars in the grass in a subsequent simultaneous hydrolysis and fermentation (SSF) process to produce bioethanol products. This would provide a dual use for the grass, whereby it is first used for AMD remediation, and subsequently the waste grass from this process is used for bioethanol production, thereby creating an opportunity to derive value from the overall process. Various investigations into grass contacted with AMD (to achieve optimal AMD remediation, and to achieve the greatest conversion to bioethanol after SSF were performed. The AMD remediation and grass pre-treatment experiments were conducted with and without SRB addition. Data showed that the synthetically prepared weak acid AMD solution was sufficient to disrupt the grass lignocellulosic structure at ambient temperature. Extending the retention times to eleven and thirteen days improved the pH increase and sulfide production, particularly in filtered AMD solutions after grass had been removed. SRB amendment improved pH increase, and sulfide production was increased by orders of magnitude more than in non-amended samples. An untreated grass sample that had not had prior exposure to acid hydrolysis by AMD, produced marginally higher ethanol concentrations during SSF than the non-amended pre-treated day eleven sample, but less than the SRB-amended day eleven HRT sample. As such, SRB amendment and longer grass-AMD contact times, favoured improved AMD remediation as well as bioethanol production. Testwork was successfully carried out to address all objectives. The outcomes from each study bring new knowledge to improve the sustainability and potentially improve the cost-effectiveness of AMD bioremediation techniques.