Effectiveness of sulfate-reducing passive bioreactors for treating highly contaminated acid mine drainage: I. Effect of hydraulic retention time

Abstract

Sulfate-reducing passive bioreactors have proved to be an effective technology for the treatment of acid mine drainage (AMD) contaminated waters over relatively short periods of time (1–5 a). However, long-term efficiency can be limited by several factors including problems related to the hydraulic properties of the reactive mixture. In this study, the effect of two hydraulic retention times (HRTs) of 7.3 d and 10 d on the performance of passive bioreactors was evaluated over an 11-month period for the treatment of a highly contaminated AMD. Evolution of the porosity and hydraulic conductivity of the reactive mixture was also evaluated during the 15-month operation of two bioreactors. Results indicated that bioreactors were effective at both HRTs for increasing the pH and alkalinity of contaminated water and for SO₄ and metal removal (60–82% for Fe and up to 99.9% for Cd, Ni and Zn). Although the quality of treated effluent was significantly improved with the 10 d HRT compared to the 7.3 d HRT, results showed that the higher HRT reduced the porosity and the permeability of the reactive mixture which might lead to hydraulic related problems and, eventually, to limited efficiency in long-term operation compared to a shorter HRT. The choice of HRT for a passive bioreactor must therefore consider both the desired quality of treated effluent and the potential for deterioration of hydraulic properties in the reactive mixture.

Introduction

Acid mine drainage (AMD) contaminated waters have low pH and high SO₄ and metal concentrations which need to be treated before being discharged into the environment (Blowes et al., 2003). Sulfate-reducing passive bioreactors are preferred to traditional water treatment plant technologies because they allow higher metal removal at low pH and generate more stable sludge with lower operation costs and minimal energy consumption (Zaluski et al., 2003). Over the last 12 a, several pilot- and full-scale passive bioreactors containing mixtures of natural organic materials have been installed at several former mine sites (e.g. Ferris–Haggarty Mine, Wyoming, USA; Cadillac Molybdenite, Northern Québec, Canada; Silver Cycle Adit and National Tunnel Adit, Colorado, USA,) (Guesk, 2005, Kuyucak et al., 2006, Figueroa et al., 2007).

Generally, passive bioreactors operated over relatively short periods of time (up to 5 a) meet their treatment objectives in terms of increasing the pH and alkalinity, and for SO₄ and metal removal (Dvorak et al., 1992, Reisinger et al., 2000, Zaluski et al., 2003, Kuyucak et al., 2006, Figueroa et al., 2007). However, in long-term operation, their efficiency has been limited by substrate composition (Beaulieu et al., 2000, Chang et al., 2000, Cocos et al., 2002, Waybrant et al., 2002, Zagury et al., 2006, Figueroa et al., 2007, Neculita et al., 2007), hydraulic retention time (HRT) (Younger et al., 2002), as well as by AMD toxicity and variations in flow (Reisman et al., 2003, Kuyucak et al., 2006). In fact, in a field-bioreactor, after the selection of the appropriate substrate, optimization of HRT is the most important design objective, and yet the most difficult to achieve (Younger et al., 2002). Recent research suggests that the nominal HRT of field-bioreactors should be at least 40 h, while a HRT of 4 or more days is required for efficient treatment of highly contaminated AMD (Younger et al., 2002).

Finally, the long-term efficiency of passive bioreactors is also limited by problems related to the hydraulic properties of the reactive mixture such as clogging, compaction, segregation, and development of preferential flow paths (Younger et al., 2002). Suspended solids from AMD, (oxy)hydroxide, carbonate and sulphide minerals formed by metal precipitation, and biomass and metabolic products generated by bacterial activity can induce changes in substrate properties (Rockhold et al., 2002). These solids potentially decrease the porosity and permeability, affect the longevity and the performance of bioreactors and, ultimately, can result in their failure (Neculita et al., 2007). Saturated hydraulic conductivity (k_sat) of the substrate materials is therefore one of the important parameters to be determined (Bolis et al., 1992). Usually, values of k_sat are substrate-related which typically vary, for most substrates, from 10⁻⁴ cm/s (compost) to between 10⁻² and 10⁻³ cm/s (sawdust) (URS, 2003). Moreover, experience has shown that k_sat needs to be evaluated before installing the bioreactor because laboratory tests can more easily be carried out. For field-bioreactors, however, hydraulic parameters are very difficult to evaluate because the installation of the large number of piezometers to monitor the system may change the hydraulic behavior of the system itself (Younger et al., 2002).

In addition, the volume of substrate needed to achieve a given nominal HRT is related to its hydraulic conductivity and porosity by Darcy’s law (assuming a saturated medium), which can be expressed as follows:

$$v=k_{\text{sat}}xi\text{,}\text{with}{k}_{\text{sat}}=f(n)$$

where v = seepage velocity (cm/s), k_sat = saturated hydraulic conductivity (cm/s), i = hydraulic gradient, and n = porosity of the porous medium.

The porosity of the substrates is therefore another important hydraulic parameter which must be estimated prior to final selection of mixtures and to design appropriate dimensions for the treatment system (Younger et al., 2002). Usually, field-tested porosities are in the range 0.15–0.35, while laboratory-based values are in the range 0.35–0.63 (Younger et al., 2002, Amos and Younger, 2003).

Several studies have evaluated the influence of biofilm (biomass growth and metabolic products generated by bacterial activity) and related changes in the physical and hydraulic properties of porous media (Taylor and Jaffé, 1990, Taylor et al., 1990, Rockhold et al., 2002, Anello et al., 2005, Polo et al., 2006). However, they used very simple matrices (e.g. sand), liquid organic C sources (e.g. lactate, methanol) and sometimes only aerobic media. Moreover, some studies found that there appears to be a limit beyond which no further permeability reductions can occur (Taylor and Jaffé, 1990). Expressed as the ratio of the hydraulic conductivity k_sat to the initial hydraulic conductivity (k_{sat 0}), this limit has been reported to be on the order of 10⁻⁴ (Taylor and Jaffé, 1990). Limited results are, however, available with respect to the evolution of the hydraulic properties during the long-term operation of laboratory and field-scale passive bioreactors filled with mixtures constituted from complex organic C sources and mineral waste materials (Bolis et al., 1992, Reisinger et al., 2000).

The main objective of the present work was to study the evolution of effluent quality, as well as the evolution of hydraulic parameters during the operation of column bioreactors treating a highly contaminated AMD, at two different HRTs. Porosity and hydraulic conductivity were evaluated before, during, and after the testing period.