Model‐based analysis of the role of biological, hydrological and geochemical factors affecting uranium bioremediation

Abstract

Uranium contamination is a serious concern at several sites motivating the development of novel treatment strategies such as the Geobacter-mediated reductive immobilization of uranium. However, this bioremediation strategy has not yet been optimized for the sustained uranium removal. While several reactive-transport models have been developed to represent Geobacter-mediated bioremediation of uranium, these models often lack the detailed quantitative description of the microbial process (e.g., biomass build-up in both groundwater and sediments, electron transport system, etc.) and the interaction between biogeochemical and hydrological process. In this study, a novel multi-scale model was developed by integrating our recent model on electron capacitance of Geobacter (Zhao et al., 2010) with a comprehensive simulator of coupled fluid flow, hydrologic transport, heat transfer, and biogeochemical reactions. This mechanistic reactive-transport model accurately reproduces the experimental data for the bioremediation of uranium with acetate amendment. We subsequently performed global sensitivity analysis with the reactive-transport model in order to identify the main sources of prediction uncertainty caused by synergistic effects of biological, geochemical, and hydrological processes. The proposed approach successfully captured significant contributing factors across time and space, thereby improving the structure and parameterization of the comprehensive reactive-transport model. The global sensitivity analysis also provides a potentially useful tool to evaluate uranium bioremediation strategy. The simulations suggest that under difficult environments (e.g., highly contaminated with U(VI) at a high migration rate of solutes), the efficiency of uranium removal can be improved by adding Geobacter species to the contaminated site (bioaugmentation) in conjunction with the addition of electron donor (biostimulation). The simulations also highlight the interactive effect of initial cell concentration and flow rate on U(VI) reduction.

Variance-based global sensitivity analysis for multiple scenarios and models with implementation using sparse grid collocation

“…Various methods have been developed to reduce the computational cost. A common practice is to first conduct a Morris analysis to screen out unimportant parameters so that global sensitivity analysis is only conducted for important parameters (e.g., Chu-Agor et al 2011, Zhao et al, 2011. The quasi-random sampling method developed by Saltelli et al (2010) is the most popular MC method because the number of model executions needed for Sobol" sensitivity analysis is dramatically reduced in comparison with conventional MC methods.…”

The design of long‐term effective uranium bioremediation strategy using a community metabolic model

“…The genome‐scale model of G. sulfurreducens has been previously published and experimentally verified (Mahadevan et al, 2006; Segura et al, 2008; Zhuang et al, 2011), and the same acetate kinetic parameters used in the Geobacter and Rhodofera x competition study (Zhuang et al, 2011) were used here. However, we have expanded the model to include a uranium reduction pathway based on the electron capacitance model (Zhao et al, 2010, 2011) as well as multiple Fe(III) uptake pathways to reflect the existence of multiple Fe(III) types in the subsurface. Two separate genome‐scale Geobacter models were used; the attached and planktonic Geobacter are treated as two separate organisms, and we assumed that only planktonic Geobacter cells are able to reduce uranium, and only attached Geobacter cells are able to reduce Fe(III) (Fig.…”

“…The mechanisms of uranium reduction are not yet clearly understood. One hypothesis suggested that electrons can be temporarily stored in reduced cytochromes and subsequently transferred to uranium (Zhao et al, 2010, 2011), although recent evidence suggested that electrons can be directly transferred to uranium via pili (Cologgi et al, 2011). In addition, recently it has been reported that Fe(II) can reduce uranium under specific conditions (Du et al, 2011).…”

“…To incorporate this lumped reaction into the community model, we expanded the genome‐scale metabolic model of Geobacter by including a hypothetical reaction for uranium reduction. The uptake of uranium is constrained by Equation (5), which reflects the uranium reduction kinetics predicted by the electron capacitance model (Zhao et al, 2010, 2011). …”

“…The attachment and detachment kinetics of Geobacter have been previously modeled based on the experimental observation that Geobacter primarily existed in planktonic form during Uranium reduction (Zhao et al, 2010, 2011). Recent experimental evidence showed that during the iron reduction phase, Geobacter primarily existed in planktonic form, whereas during the sulfate reduction phase, Geobacter primarily existed in attached form (Shabir et al, personal communication).…”