The microbiology of redox processes : development of a redox model

The work described in this report forms part of the BioTran project, which was initiated to examine the effects of microbes on transport processes, especially in the context of contaminant properties of host rocks. An understanding of these microbial processes is also relevant to other areas such as bioremediation of contaminated land, borehole and reservoir ‘bioclogging’ and microbially enhanced oil recovery. More broadly, these processes impinge on aquifer recharge, pathogen survival, and ultimately on groundwater protection. To date, the project has comprehensively reviewed the available literature and developed methodologies for experimental studies to provide information and data for existing transport models (Bateman et al., 2006; Coombs et al., 2008; West et al., 2008). This report develops this work further by examining the influences of microbial activity on redox processes in the subsurface. Along with pH, Eh – or the extent to which a groundwater is oxidising or reducing – defines the thermodynamic stability of minerals and dissolved species and hence the solubility and transport of elements and the extent of their interactions with surfaces. Redox gradients or fronts are locations where mineral dissolution/precipitation rates may be high and may also be key locations for colloid formation or destabilisation. Chemotrophic organisations can utilise the energy of inorganic redox reactions for their life processes and hence particularly at fronts, high activity levels may be found. This report discusses the significance of microbial catalysis in redox reactions in both nearsurface and deep geological environments. In near-surface groundwaters, particularly in contaminated aquifers where microbial population densities may be high, the changes in bulk water chemistry as a result of biologically driven redox reactions can be significant. Here the micro-organisms utilise the energy from catalysed exoenergetic redox reactions, effectively coupling inorganic reactions to the production of the energy transport molecules (predominantly ATP) that drive their life processes. In such systems, profiles along the direction of flow show characteristic stepwise development of increasingly reducing conditions as the more reactive oxidants are consumed, this usually results in initial consumption of dissolved oxygen followed by nitrate, Mn(IV), Fe(III), sulphate and, finally, carbonate. For contaminated waters with high loadings of organic carbon, this serves as the predominant reductant. In contrast, most deep geological environments, especially those being considered for applications such as waste disposal or carbon dioxide storage, are characterised by very low energy fluxes. Unlike the case of contaminated aquifers, the consequences of microbially mediated redox reactions in these deep environments may be subtle and build up slowly over very long periods of time. An example is discussed which provides evidence for direct involvement of microorganisms in redox reactions involving oxidation of compacted rock; the rock matrix serving as a substrate for microbial communities that are – albeit very slowly – living off the energy provided by the reactions that they catalyse. The fundamental redox concept necessary for deep geological environments assumes that the sequence of microbial population groups is similar to that defined in shallow environments. The key difference in a deep system when compared to the near surface, is that the reductants are present in the rock and thus, rather than zones that extend over hundreds of metres or kilometres in aquifers that are defined by the kinetics of redox reactions, zones may be on the scale of mm or cm within oxidation rims of mineral grains or rock matrix and be defined by the kinetics of slow diffusive transport of dissolved oxidants. However, in order to extend the scoping calculations as illustrated in this report, it will be necessary to develop a more rigorous solute diffusion/reaction model that will individually characterise the reactions in the different redox zones. The future model would have three coupled components: 1. A solute transport module, which initially will focus on 1D diffusion from a constant or stepwise variable source but, ideally, should be expandable to 2D with the option to include advective transport 2. A redox profile generation module that identifies the most favourable redox reactions in different zones as a function of solute concentrations and available solid phases. 3. An integrated chemical / microbial reaction module that couples redox reactions with other relevant reactions (e.g. inorganic pH buffering) to quantify changes in solution and solid phases (including biomass) in particular zones. This module also tracks net changes in the porosity of the rock. In order to support applications that can have significant environmental impacts, the model used would have to be rigorously tested. For example, the siting of a repository for radioactive waste is critical and if a deep redox front is evidence of past intrusion of oxidising water, such an observation may be sufficient to disqualify a potential site. If, on the other hand, such a redox front results from slow buffering reactions by the rock matrix, this can be an indication of a potentially very powerful far-field geosphere barrier. The model must be capable of making predictions that allow these two cases to be distinguished in an unambiguous manner.