Climate forcing controls on carbon terrestrial fluxes during shale weathering.

Climate influences near-surface biogeochemical processes and thereby determines the partitioning of carbon dioxide (CO2) in shale, and yet the controls on carbon (C) weathering fluxes remain poorly constrained. Using a dataset that characterizes biogeochemical responses to climate forcing in shale regolith, we implement a numerical model that describes the effects of water infiltration events, gas exchange, and temperature fluctuations on soil respiration and mineral weathering at a seasonal timescale. Our modeling approach allows us to quantitatively disentangle the controls of transient climate forcing and biogeochemical mechanisms on C partitioning. We find that ~3% of soil CO2 (1.02 mol C/m2/y) is exported to the subsurface during large infiltration events. Here, net atmospheric CO2 drawdown primarily occurs during spring snowmelt, governs the aqueous C exports (61%), and exceeds the CO2 flux generated by pyrite and petrogenic organic matter oxidation (~0.2 mol C/m2/y). We show that shale CO2 consumption results from the temporal coupling between soil microbial respiration and carbonate weathering. This coupling is driven by the impacts of hydrologic fluctuations on fresh organic matter availability and CO2 transport to the weathering front. Diffusion-limited transport of gases under transient hydrological conditions exerts an important control on CO2(g) egress patterns and thus must be considered when inferring soil CO2 drawdown from the gas phase composition. Our findings emphasize the importance of seasonal climate forcing in shaping the net contribution of shale weathering to terrestrial C fluxes and suggest that warmer conditions could reduce the potential for shale weathering to act as a CO2 sink.

Dissolved Carbonate and pH Control the Dissolution of Uranyl Phosphate Minerals in Flow-Through Porous Media.

Uranyl phosphate minerals represent an important secondary source of uranium release at contaminated sites. In flow-through column experiments with background porewater (BPW) of typical freshwater aquifer composition (pH 7.0, ∼0.2 mM total carbonate (TC)), dissolution of K-ankoleite (KUO2PO4·3H2O), Na-autunite (NaUO2PO4·3H2O), and Ca-autunite (Ca(UO2)2(PO4)2·6H2O) was controlled by mineral solubility at steady-state U release. Effluent concentrations indicated exchange with BPW cations, and postreaction characterization showed alteration of the initial mineral composition, changes in structure (decreased crystallinity, increased disorder, and distortion of U-P mineral sheets) and possible neoformation of phases of similar structure. Increasing the BPW pH and TC to 8.1-8.2 and 2.2-3.7 mM, respectively, resulted in mineral undersaturation and produced ca. 2 orders-of-magnitude higher U and P release without reaching steady state. Minerals incorporated less BPW cations into their structures compared to low carbonate BPW experiments but showed structural disorder and distortion. Faster dissolution rates were attributed to the formation of binary and ternary uranyl carbonate complexes that accelerate the rate-determining step of uranyl detachment from the uranyl-phosphate layered structure. Calculated dissolution rates (log Rs between -8.95 and -10.32 mol m-2 s-1), accounting for reaction and transport in porous media, were similar to dissolution rates of other classes of uranyl minerals. In undersaturated solutions, dissolution rates for uranyl phosphate, oxyhydroxide, and silicate minerals can be predicted within 1-2 orders-of-magnitude from pH ∼5-10 on the basis of pH/carbonate concentration.

Exascale applications: skin in the game.

As noted in Wikipedia, skin in the game refers to having 'incurred risk by being involved in achieving a goal', where 'skin is a synecdoche for the person involved, and game is the metaphor for actions on the field of play under discussion'. For exascale applications under development in the US Department of Energy Exascale Computing Project, nothing could be more apt, with the skin being exascale applications and the game being delivering comprehensive science-based computational applications that effectively exploit exascale high-performance computing technologies to provide breakthrough modelling and simulation and data science solutions. These solutions will yield high-confidence insights and answers to the most critical problems and challenges for the USA in scientific discovery, national security, energy assurance, economic competitiveness and advanced healthcare. This article is part of a discussion meeting issue 'Numerical algorithms for high-performance computational science'.

Incorporating Nanoscale Effects into a Continuum-Scale Reactive Transport Model for CO2-Deteriorated Cement.

Wellbore cement deterioration is critical for wellbore integrity and the safety of CO2 storage in geologic formations. Our previous experimental work highlighted the importance of the portlandite (CH)-depleted zone and the surface dissolution zone in the CO2-attacked cement. In this study, we simulated numerically the evolution of the CH-depleted zone and the dissolution of the cement surfaces utilizing a reduced-dimension (1D) reactive transport model. The approach shows that three nanoscale effects are important and had to be incorporated in a continuum-scale model to capture experimental observations: First, it was necessary to account for the fact that secondary CaCO3 precipitation does not fill the pore space completely, with the result that acidic brine continues to diffuse through the carbonated zone to form a CH-depleted zone. Second, secondary precipitation in brine begins via nucleation kinetics, and this could not be described with previous models using growth kinetics alone. Third, our results suggest that the CaCO3 precipitates in the confined pore space are more soluble than those formed in brine. This study provides a new platform for a reduced dimension model for CO2 attack on cement that captures the important nanoscale mechanisms influencing macroscale phenomena in subsurface environments.

Alteration and Erosion of Rock Matrix Bordering a Carbonate-Rich Shale Fracture.

A novel reactive transport model has been developed to examine the processes that affect fracture evolution in a carbonate-rich shale. An in situ synchrotron X-ray microtomography experiment, flowing CO2 saturated water through a single fracture mini-core of Niobrara Shale provided the experimental observations for the development and testing of the model. The phenomena observed included the development of a porous altered layer, flow channeling, and increasingly limited calcite dissolution. The experimental observations cannot be explained by models that consider only mineral dissolution and development of an altered layer. The difference between the fracture volume change recorded by the microtomography images and what would be expected from mineral dissolution alone suggest that there is erosion of the altered layer as it develops. The numerical model includes this additional mechanism, with the erosion rate based on the thickness of the altered layer, and successfully captures the evolution of the geochemical reactions and morphology of the fracture. The findings imply that the abundance (with a threshold of approximately 35%) and reactivity of the rapidly reacting mineral control the development and erodibility of the altered layer on the fracture surfaces, and therefore fracture opening.

Reoxidation of Chromium(III) Products Formed under Different Biogeochemical Regimes.

Hexavalent chromium, Cr(VI), is a widespread and toxic groundwater contaminant. Reductive immobilization to Cr(III) is a treatment option, but its success depends on the long-term potential for reduced chromium precipitates to remain immobilized under oxidizing conditions. In this unique long-term study, aquifer sediments subjected to reductive Cr(VI) immobilization under different biogeochemical regimes were tested for their susceptibility to reoxidation. After reductive treatment for 1 year, sediments were exposed to oxygenated conditions for another 2 years in flow-through, laboratory columns. Under oxidizing conditions, immobilized chromium reduced under predominantly denitrifying conditions was mobilized at low concentrations (≪1 µM Cr(VI); ∼ 3% of Cr(III) deposited) that declined over time. A conceptual model of a limited pool of more soluble Cr(III), and a larger pool of relatively insoluble Cr(III), is proposed. In contrast, almost no chromium was mobilized from columns reduced under predominantly fermentative conditions, and where reducing conditions persisted for several months after introduction of oxidizing conditions, presumably due to the presence of a reservoir of reduced species generated during reductive treatment. The results from this 3-year study demonstrate that biogeochemical conditions present during reductive treatment, and the potential for buildup of reducing species, will impact the long-term sustainability of the remediation effort.

Water Table Dynamics and Biogeochemical Cycling in a Shallow, Variably-Saturated Floodplain.

Three-dimensional variably saturated flow and multicomponent biogeochemical reactive transport modeling, based on published and newly generated data, is used to better understand the interplay of hydrology, geochemistry, and biology controlling the cycling of carbon, nitrogen, oxygen, iron, sulfur, and uranium in a shallow floodplain. In this system, aerobic respiration generally maintains anoxic groundwater below an oxic vadose zone until seasonal snowmelt-driven water table peaking transports dissolved oxygen (DO) and nitrate from the vadose zone into the alluvial aquifer. The response to this perturbation is localized due to distinct physico-biogeochemical environments and relatively long time scales for transport through the floodplain aquifer and vadose zone. Naturally reduced zones (NRZs) containing sediments higher in organic matter, iron sulfides, and non-crystalline U(IV) rapidly consume DO and nitrate to maintain anoxic conditions, yielding Fe(II) from FeS oxidative dissolution, nitrite from denitrification, and U(VI) from nitrite-promoted U(IV) oxidation. Redox cycling is a key factor for sustaining the observed aquifer behaviors despite continuous oxygen influx and the annual hydrologically induced oxidation event. Depth-dependent activity of fermenters, aerobes, nitrate reducers, sulfate reducers, and chemolithoautotrophs (e.g., oxidizing Fe(II), S compounds, and ammonium) is linked to the presence of DO, which has higher concentrations near the water table.

A 2.5D Reactive Transport Model for Fracture Alteration Simulation.

Understanding fracture alteration resulting from geochemical reactions is critical in predicting fluid migration in the subsurface and is relevant to multiple environmental challenges. Here, we present a novel 2.5D continuum reactive transport model that captures and predicts the spatial pattern of fracture aperture change and the development of an altered layer in the near-fracture region. The model considers permeability heterogeneity in the fracture plane and updates fracture apertures and flow fields based on local reactions. It tracks the reaction front of each mineral phase and calculates the thickness of the altered layer. Given this treatment, the model is able to account for the diffusion limitation on reaction rates associated with the altered layer. The model results are in good agreement with an experimental study in which a CO2-acidified brine was injected into a fracture in the Duperow Dolomite, causing dissolution of calcite and dolomite that result in the formation of a preferential flow channel and an altered layer. With an effective diffusion coefficient consistent with the experimentally observed porosity of the altered layer, the model captures the progressive decrease in the dissolution rate of the fast-reacting mineral in the altered layer.

Pore-scale controls on calcite dissolution rates from flow-through laboratory and numerical experiments.

A combination of experimental, imaging, and modeling techniques were applied to investigate the pore-scale transport and surface reaction controls on calcite dissolution under elevated pCO2 conditions. The laboratory experiment consisted of the injection of a solution at 4 bar pCO2 into a capillary tube packed with crushed calcite. A high resolution pore-scale numerical model was used to simulate the experiment based on a computational domain consisting of reactive calcite, pore space, and the capillary wall constructed from volumetric X-ray microtomography images. Simulated pore-scale effluent concentrations were higher than those measured by a factor of 1.8, with the largest component of the discrepancy related to uncertainties in the reaction rate model and its parameters. However, part of the discrepancy was apparently due to mass transport limitations to reactive surfaces, which were most pronounced near the inlet where larger diffusive boundary layers formed around grains and in slow-flowing pore spaces that exchanged mass by diffusion with fast flow paths. Although minor, the difference between pore- and continuum-scale results due to transport controls was discernible with the highly accurate methods employed and is expected to be more significant where heterogeneity is greater, as in natural subsurface materials.

Divergent aquifer biogeochemical systems converge on similar and unexpected Cr(VI) reduction products.

In this study of reductive chromium immobilization, we found that flow-through columns constructed with homogenized aquifer sediment and continuously infused with lactate, chromate, and various native electron acceptors diverged to have very different Cr(VI)-reducing biogeochemical regimes characterized by either denitrifying or fermentative conditions (as indicated by effluent chemical data, 16S rRNA pyrotag data, and metatranscriptome data). Despite the two dramatically different biogeochemical environments that evolved in the columns, these regimes created similar Cr(III)-Fe(III) hydroxide precipitates as the predominant Cr(VI) reduction product, as characterized by micro-X-ray fluorescence and micro-X-ray absorption near-edge structure analysis. We discuss two conflicting scenarios of microbially mediated formation of Cr(III)-Fe(III) precipitates, each of which is both supported and contradicted by different lines of evidence: (1) enzymatic reduction of Cr(VI) to Cr(III) followed by coprecipitation of Cr(III) and Fe(III) and (2) both regimes generated at least small amounts of Fe(II), which abiotically reduced Cr(VI) to form a Cr-Fe precipitate. Evidence of zones with different levels of Cr(VI) reduction suggest that local heterogeneity may have confounded interpretation of processes based on bulk measurements. This study indicates that the bulk redox status and biogeochemical regime, as categorized by the dominant electron-accepting process, do not necessarily control the final product of Cr(VI) reduction.

Influence of salinity gradients on the diffusion of water and ionic species in dual porosity clay samples.

Most of the available data on diffusion in natural clayey rocks consider tracer diffusion in the absence of a salinity gradient despite the fact that such gradients are frequently found in natural and engineered subsurface environments. To assess the role of such gradients on the diffusion properties of clayey materials, through-diffusion experiments were carried out in the presence and absence of a salinity gradient using salt-diffusion and radioisotope tracer techniques. The experiments were carried out with vermiculite samples that contained equal proportions of interparticle and interlayer porosities so as to assess also the role played by the two types of porosities on the diffusion of water and ions. Data were interpreted using both a classical Fickian diffusion model and with a reactive transport code, CrunchClay that can handle multi-porosity diffusion processes in the presence of charged surfaces. By combining experimental and simulated data, we demonstrated that (i) the flux of water diffusing through vermiculite interlayer porosity was minor compared to that diffusing through the interparticle porosity, and (ii) a model considering at least three types of porous volumes (interlayer, interparticle diffuse layer, and bulk interparticle) was necessary to reproduce consistently the variations of neutral and charged species diffusion as a function of salinity gradient conditions.

Upscaling calcite growth rates from the mesoscale to the macroscale.

Quantitative prediction of mineral reaction rates in the subsurface remains a daunting task partly because a key parameter for macroscopic models, the reactive site density, is poorly constrained. Here we report atomic force microscopy (AFM) measurements on the {1014} calcite surface of monomolecular step densities, treated as equivalent to the reactive site density, as a function of aqueous calcium-to-carbonate ratio and saturation index. Data for the obtuse step orientation are combined with existing step velocity measurements to generate a model that predicts overall macroscopic calcite growth rates. The model is quantitatively consistent with several published macroscopic rates under a range of alkaline solution conditions, particularly for two of the most comprehensive data sets, without the need for additional fit parameters. The model reproduces peak growth rates, and its functional form is simple enough to be incorporated into reactive transport or other macroscopic models designed for predictions in porous media. However, it currently cannot model equilibrium or pH effects and it may overestimate rates at high aqueous calcium-to-carbonate ratios. The discrepancies in rates at high calcium-to-carbonate ratios may be due to differences in pretreatment, such as exposing the seed material to SI ≥ 1.0 to generate/develop growth hillocks, or other factors.

Resolving experimental biases in the interpretation of diffusion experiments with a user-friendly numerical reactive transport approach.

The reactive transport code CrunchClay was used to derive effective diffusion coefficients (De), clay porosities (ε), and adsorption distribution coefficients (KD) from through-diffusion data while considering accurately the influence of unavoidable experimental biases on the estimation of these diffusion parameters. These effects include the presence of filters holding the solid sample in place, the variations in concentration gradients across the diffusion cell due to sampling events, the impact of tubing/dead volumes on the estimation of diffusive fluxes and sample porosity, and the effects of O-ring-filter setups on the delivery of solutions to the clay packing. Doing so, the direct modeling of the measurements of (radio)tracer concentrations in reservoirs is more accurate than that of data converted directly into diffusive fluxes. While the above-mentioned effects have already been described individually in the literature, a consistent modeling approach addressing all these issues at the same time has never been described nor made easily available to the community. A graphical user interface, CrunchEase, was created, which supports the user by automating the creation of input files, the running of simulations, and the extraction and comparison of data and simulation results. While a classical model considering an effective diffusion coefficient, a porosity and a solid/solution distribution coefficient (De-ε-KD) may be implemented in any reactive transport code, the development of CrunchEase makes it easy to apply by experimentalists without a background in reactive transport modeling. CrunchEase makes it also possible to transition more easily from a De-ε-KD modeling approach to a state-of-the-art process-based understanding modeling approach using the full capabilities of CrunchClay, which include surface complexation modeling and a multi-porosity description of the clay packing with charged diffuse layers.

Understanding the hydrological response of a headwater-dominated catchment by analysis of distributed surface-subsurface interactions.

We computationally explore the relationship between surface-subsurface exchange and hydrological response in a headwater-dominated high elevation, mountainous catchment in East River Watershed, Colorado, USA. In order to isolate the effect of surface-subsurface exchange on the hydrological response, we compare three model variations that differ only in soil permeability. Traditional methods of hydrograph analysis that have been developed for headwater catchments may fail to properly characterize catchments, where catchment response is tightly coupled to headwater inflow. Analyzing the spatially distributed hydrological response of such catchments gives additional information on the catchment functioning. Thus, we compute hydrographs, hydrological indices, and spatio-temporal distributions of hydrological variables. The indices and distributions are then linked to the hydrograph at the outlet of the catchment. Our results show that changes in the surface-subsurface exchange fluxes trigger different flow regimes, connectivity dynamics, and runoff generation mechanisms inside the catchment, and hence, affect the distributed hydrological response. Further, changes in surface-subsurface exchange rates lead to a nonlinear change in the degree of connectivity-quantified through the number of disconnected clusters of ponding water-in the catchment. Although the runoff formation in the catchment changes significantly, these changes do not significantly alter the aggregated streamflow hydrograph. This hints at a crucial gap in our ability to infer catchment function from aggregated signatures. We show that while these changes in distributed hydrological response may not always be observable through aggregated hydrological signatures, they can be quantified through the use of indices of connectivity.

Timing the onset of sulfate reduction over multiple subsurface acetate amendments by measurement and modeling of sulfur isotope fractionation.

Stable isotope fractionations of sulfur are reported for three consecutive years of acetate-enabled uranium bioremediation at the US Department of Energy's Rifle Integrated Field Research Challenge (IFRC) site. The data show a previously undocumented decrease in the time between acetate addition and the onset of sulfate reducing conditions over subsequent amendments, from 20 days in the 2007 experiment to 4 days in the 2009 experiment. Increased sulfide concentrations were observed at the same time as δ(34)S of sulfate enrichment in the first year, but in subsequent years elevated sulfide was detected up to 15 days after increased δ(34)S of sulfate. A biogeochemical reactive transport model is developed which explicitly incorporates the stable isotopes of sulfur to simulate fractionation during the 2007 and 2008 amendments. A model based on an initially low, uniformly distributed population of sulfate reducing bacteria that grow and become spatially variable with time reproduces measured trends in solute concentration and δ(34)S, capturing the change in onset of sulfate reduction in subsequent years. Our results demonstrate a previously unrecognized hysteretic effect in the spatial distribution of biomass growth during stimulated subsurface bioremediation.

Physicochemical heterogeneity controls on uranium bioreduction rates at the field scale.

It has been demonstrated in laboratory systems that U(VI) can be reduced to immobile U(IV) by bacteria in natural environments. The ultimate efficacy of bioreduction at the field scale, however, is often challenging to quantify and depends on site characteristics. In this work, uranium bioreduction rates at the field scale are quantified, for the first time, using an integrated approach. The approach combines field data, inverse and forward hydrological and reactive transport modeling, and quantification of reduction rates at different spatial scales. The approach is used to explore the impact of local scale (tens of centimeters) parameters and processes on field scale (tens of meters) system responses to biostimulation treatments and the controls of physicochemical heterogeneity on bioreduction rates. Using the biostimulation experiments at the Department of Energy Old Rifle site, our results show that the spatial distribution of hydraulic conductivity and solid phase mineral (Fe(III)) play a critical role in determining the field-scale bioreduction rates. Due to the dependence on Fe-reducing bacteria, field-scale U(VI) bioreduction rates were found to be largely controlled by the abundance of Fe(III) minerals at the vicinity of the injection wells and by the presence of preferential flow paths connecting injection wells to down gradient Fe(III) abundant areas.

Strontium and cesium release mechanisms during unsaturated flow through waste-weathered Hanford sediments.

Leaching behavior of Sr and Cs in the vadose zone of Hanford site (Washington) was studied with laboratory-weathered sediments mimicking realistic conditions beneath the leaking radioactive waste storage tanks. Unsaturated column leaching experiments were conducted using background Hanford pore water focused on first 200 pore volumes. The weathered sediments were prepared by 6 months reaction with a synthetic Hanford tank waste leachate containing Sr and Cs (10(-5) and 10(-3) molal representative of LO- and HI-sediment, respectively) as surrogates for (90)Sr and (137)Cs. The mineral composition of the weathered sediments showed that zeolite (chabazite-type) and feldspathoid (sodalite-type) were the major byproducts but different contents depending on the weathering conditions. Reactive transport modeling indicated that Cs leaching was controlled by ion-exchange, while Sr release was affected primarily by dissolution of the secondary minerals. The later release of K, Al, and Si from the HI-column indicated the additional dissolution of a more crystalline mineral (cancrinite-type). A two-site ion-exchange model successfully simulated the Cs release from the LO-column. However, a three-site ion-exchange model was needed for the HI-column. The study implied that the weathering conditions greatly impact the speciation of the secondary minerals and leaching behavior of sequestrated Sr and Cs.

Modeling geogenic and atmospheric nitrogen through the East River Watershed, Colorado Rocky Mountains.

There is a growing understanding of the role that bedrock weathering can play as a source of nitrogen (N) to soils, groundwater and river systems. The significance is particularly apparent in mountainous environments where weathering fluxes can be large. However, our understanding of the relative contributions of rock-derived, or geogenic, N to the total N supply of mountainous watersheds remains poorly understood. In this study, we develop the High-Altitude Nitrogen Suite of Models (HAN-SoMo), a watershed-scale ensemble of process-based models to quantify the relative sources, transformations, and sinks of geogenic and atmospheric N through a mountain watershed. Our study is based in the East River Watershed (ERW) in the Upper Colorado River Basin. The East River is a near-pristine headwater watershed underlain primarily by an N-rich Mancos Shale bedrock, enabling the timing and magnitude of geogenic and atmospheric contributions to watershed scale dissolved N-exports to be quantified. Several calibration scenarios were developed to explore equifinality using >1600 N concentration measurements from streams, groundwater, and vadose zone samples collected over the course of four years across the watershed. When accounting for recycling of N through plant litter turnover, rock weathering accounts for approximately 12% of the annual dissolved N sources to the watershed in the most probable calibration scenario (0-31% in other scenarios), and 21% (0-44% in other scenarios) when considering only "new" N sources (i.e. geogenic and atmospheric). On an annual scale, instream dissolved N elimination, plant turnover (including cattle grazing) and atmospheric deposition are the most important controls on N cycling.

Modeling acid-gas generation from boiling chloride brines.

BACKGROUND: This study investigates the generation of HCl and other acid gases from boiling calcium chloride dominated waters at atmospheric pressure, primarily using numerical modeling. The main focus of this investigation relates to the long-term geologic disposal of nuclear waste at Yucca Mountain, Nevada, where pore waters around waste-emplacement tunnels are expected to undergo boiling and evaporative concentration as a result of the heat released by spent nuclear fuel. Processes that are modeled include boiling of highly concentrated solutions, gas transport, and gas condensation accompanied by the dissociation of acid gases, causing low-pH condensate. RESULTS: Simple calculations are first carried out to evaluate condensate pH as a function of HCl gas fugacity and condensed water fraction for a vapor equilibrated with saturated calcium chloride brine at 50-150 degrees C and 1 bar. The distillation of a calcium-chloride-dominated brine is then simulated with a reactive transport model using a brine composition representative of partially evaporated calcium-rich pore waters at Yucca Mountain. Results show a significant increase in boiling temperature from evaporative concentration, as well as low pH in condensates, particularly for dynamic systems where partial condensation takes place, which result in enrichment of HCl in condensates. These results are in qualitative agreement with experimental data from other studies. CONCLUSION: The combination of reactive transport with multicomponent brine chemistry to study evaporation, boiling, and the potential for acid gas generation at the proposed Yucca Mountain repository is seen as an improvement relative to previously applied simpler batch evaporation models. This approach allows the evaluation of thermal, hydrological, and chemical (THC) processes in a coupled manner, and modeling of settings much more relevant to actual field conditions than the distillation experiment considered. The actual and modeled distillation experiments do not represent expected conditions in an emplacement drift, but nevertheless illustrate the potential for acid-gas generation at moderate temperatures (<150 degrees C).

Cesium migration in Hanford sediment: a multisite cation exchange model based on laboratory transport experiments.

Cs+ transport experiments carried out in columns packed with uncontaminated Hanford formation sediment from the SX tank farm provide strong support for the use of a multisite, multicomponent cation exchange model to describe Cs+ migration in the Hanford vadose zone. The experimental results indicate a strong dependence of the effective Cs+ Kd on the concentrations of other cations, including Na+ that is present at high to extremely high concentrations in fluids leaking from the Hanford SX tanks. A strong dependence of the Cs+ Kd on the aqueous Cs+ concentration is also apparent, with retardation of Cs+ increasing from a value of 41 at a Cs+ concentration of 10(-4) M in the feed solution to as much as 282 at a Cs+ concentration of 5x10(-7) M, all in a background of 1 M NaNO3. The total cation exchange capacity (CEC) of the Hanford sediment was determined using 22Na isotopic equilibrium exchange in a flow-through column experiment. The value for the CEC of 120 microeq/g determined with this method is compatible with a value of 121.9 microeq/g determined by multi-cation elution. While two distinct exchange sites were proposed by Zachara et al. [Geochim. Cosmochim. Acta 66 (2002) 193] based on binary batch exchange experiments, a third site is proposed in this study to improve the fit of the Cs+-Na+ and Cs+-Ca+ exchange data and to capture self-sharpened Cs+ breakthrough curves at low concentrations of Cs+. Two of the proposed exchange sites represent frayed edge sites (FES) on weathered micas and constitute 0.02% and 0.22% of the total CEC. Both of the FES show a very strong selectivity for Cs+ over Na+ (K(Na-Cs)=10(7.22) and 10(4.93), respectively). The third site, accounting for over 99% of the total CEC, is associated with planar sites on expansible clays and shows a smaller Na+-Cs+ selectivity coefficient of 10(1.99). Parameters derived from a fit of binary batch experiments alone tend to under predict Cs+ retardation in the column experiments. The transport experiments indicate 72-90% of the Cs+ sorbed in experiments targeting exchange on FES was desorbed over a 10- and 24-day period, respectively. At high Cs+ concentrations, where sorption is controlled primarily by exchange on planar sites, 95% of the Cs+ desorption was desorbed. Most of the difficulty in desorbing Cs+ from FES is a result of the extremely high selectivity of these sites for Cs+, although truly irreversible sorption as high as 23% was suggested in one experiment. The conclusion that Cs+ exchange is largely reversible in a thermodynamic sense is supported by the ability to match Cs+ desorption curves almost quantitatively with an equilibrium reactive transport simulation. The model for Cs+ retardation developed here qualitatively explains the behavior of Cs+ in the Hanford vadose zone underneath a variety of leaking tanks with differing salt concentrations. The high selectivity of FES for Cs+ implies that future desorption and migration is very unlikely to occur under natural recharge conditions.