Predicting Contaminant Mobility at the Plume Scale
Predictions of plume evolution, migration, and remediation efficacy at the plume scale often fail because of the great simplifications that are typically made in the representation of heterogeneity and the coupled hydrological and biogeochemical processes taking place within the subsurface. Although DOE is steward of many subsurface contaminant plumes, there are few scientific efforts that focus on predicting contaminant mobility at the plume scale as is needed to guide environmental stewardship. Some examples of flow and transport problems relevant to the plume scale include:
- The characterization and prediction of plume distribution, key biogeochemical processes, hydrogeological heterogeneity, and key mechanisms of plume evolution over time and as a function of groundwater composition (e.g., pH, Eh, etc.) with sufficient fidelity over large spatial regions. This obstacle is particularly difficult to tackle at plume edges, where large compositional gradients often occur.
- Understanding scale-dependent processes, such as the relationship between key biogeochemical processes (which typically operate at pore to grain scale) and plume migration (which is mostly governed by gradients over large scales).
- Developing mechanistic reactive transport capabilities that include kinetic models for biodegradation, mineral precipitation and dissolution, and sorption behavior, as needed to predict plume behavior over longer time frames and with more accuracy than is capable with current models, which are mostly empirical.
- Quantifying sensitivity to model parameters over long time frames and identifying parameters that could be used to reduce overall model uncertaintly.
Recognizing the wide range of scale-dependent processes at play, the complexities inherent in biogeochemical reaction networks, and the potential impact of heterogeneity and biogeochemical-hydrological feedbacks on flow and transport, there is a significant need to develop methods to guide the characterization and prediction of contaminant mobility accurately, tractably, and at field-relevant scales. To fill this need the Plume Challenge explores the predictive understanding of contaminant mobility at the plume scale using the concept of geochemical gradients and reactive facies. Biogeochemical (e.g. pH, contaminant concentration) gradients develop at the leading and trailing edges of plumes as a result of variations in source terms and the coupling of transport and reactive processes that can occur preferentially as a function of the initial hydrogeochemical heterogeneity of the system (e.g., Denham, 2008). We also explore the concept of using a reactive facies framework to bridge hydrogeochemical measurements at multiple scales. The reactive facies construct takes advantage of the (nonrandom) spatial distribution of geological units and the linkage between hydrophysical and geochemical properties that often exists in nature, with the goal of providing sufficient information for reactive transport models that provide more accurate predictions.
This research Challenge is being performed at the Savannah River F-Area, where we focus primarily on U and I (iodine) mobility. There is a wealth of hydrogeological and geochemical datasets available for this site, including data from over 100 groundwater-monitoring wells that are sampled regularly (some since the 1980s). Contaminant plume in groundwater beneath the F-Area of the Savannah River Site were created by the disposal and leakage of low-level acidic radioactive solutions to seepage basins. The plumes extend from the basins approximately 600 m downgradient to a stream and contain a large number of contaminants; from a risk perspective, the most important contaminants include uranium isotopes, Sr-90, I-129, Tc-99, tritium, and nitrate.
This challenge explores the impact of a migrating pH gradient and the power of the novel "reactive facies" concept as an organizing principle to integrate laboratory and field information about properties and mechanisms in order to make reliable and computationally tractable predictions at the plume scale. The challenge includes an evaluation of conceptual model complexity and sensitivity in simulations of contaminant mobility over stewardship frames. Research in this challenge will be performed through testing the following objectives:
- Objective 1 - Understand key geochemical and hydrological processes dictating contaminant mobility at the Savannah River F-Area.
- Objective 2 - Investigate whether subsurface sediments at this site can be grouped into several sediment types that each display common physical or geochemical characteristics affecting reactive transport (i.e., reactive facies).
- Objective 3 - Investigate whether identified reactive facies can be linked to depositional processes (e.g., lithogic facies) and whether geophysical methods can be used to spatially distribute reactive facies and associated transport parameters over plume scales.
- Objective 4 - Build reactive transport models incorporating the identified (and mapped) reactive facies and conduct simulations to predict contaminant mobility at various scales (e.g., local- and plume-scales).
- Objective 5 - Develop a methodology and criteria to assess the optimal level of model detail needed for reliable predictions of contaminant mobility at the plume scale over stewardship time frames.
The research will be carried out through integrated laboratory geochemical, synchrotron, isotopic, geophysical, and reactive transport modeling tasks. In collaboration with EM-supported SRNL scientists (led by Miles Denham), the resulting reactive transport model will be used to assess the natural attenuation capacity of the acidic, Uranium and Strontium plume at the F-Area of the Savannah River Site.
Research in this challenge is expected to lead to several outcomes, such as documentation of:
- The ability to characterize sorption/precipitation behavior as a function of sediment classes;
- Development of a biochemical reaction rate model for iodine transformations;
- The ability to exploit physical-chemical linkages and geophysical methods in estimating the spatial distribution of reactive facies;
- The utility of interpreting isotopic signatures in terms of plume history and mineral interactions.
- The utility of the reactive facies concept to provide a sufficiently complex model parameterization in a tractable manner at the plume scale;
- The predictive capability of reactive transport models to capture U and I-129 mobility at the local scale through a comparison of predicted versus push-pull field measurements;
- The level of model and heterogeneity complexity needed to accurately predict contaminant mobility at F-Area.
Research in this challenge will be carried out by the following LBNL multidisciplinary research team with expertise in experimental geochemistry, reactive transport modeling, isotope geochemistry, synchrotron science, and subsurface characterization using disparate, multi-scale datasets.
- John Christensen
- Susan Hubbard
- Mike Kowalsky
- Jim Davis
- Sumit Mukhopadhyay
- Wenming Dong
- Doug Sassen
- Nic Spycher
- Carl Steefel
- William Stringfellow
- Jiamin Wan
- Miles Denham (SRNL)
- Jack Istok (OSU)
- John Seaman
- Dan Kaplan (IIT)
- Jim Hunt (UCB)
- Arthur Wiedmer (UCB)
- Peter Santschi (Texas A & M University)
This LBNL SFA Challenge research will be performed in collaboration with a parallel and significant effort led by Miles Denham (SRNL) and funded by the EM-32 Soil and Groundwater Program, entitled “Attenuation-based Remedies for Metal and Radionuclide Contaminated Groundwater.” The LBNL scientific team will closely collaborate with and leverage on this EM-funded effort, which involves scientists from SRNL as well as other labs and universities, site personnel, and regulators, who are striving to identify steps that can be taken to advance the use of current scientific understanding for guiding MNA.
We are also collaborating with the EM-32 supported Advanced Subsurface Computing for Environmental Management (ASCEM) team, who are developing and testing an open-source, modular, high performance computing environment for the purpose of guiding EM cleanup and closure activities.