PyLEnM: A Machine Learning Framework for Long-Term Groundwater Contamination Monitoring Strategies.

In this study, we have developed a comprehensive machine learning (ML) framework for long-term groundwater contamination monitoring as the Python package PyLEnM (Python for Long-term Environmental Monitoring). PyLEnM aims to establish the seamless data-to-ML pipeline with various utility functions, such as quality assurance and quality control (QA/QC), coincident/colocated data identification, the automated ingestion and processing of publicly available spatial data layers, and novel data summarization/visualization. The key ML innovations include (1) time series/multianalyte clustering to find the well groups that have similar groundwater dynamics and to inform spatial interpolation and well optimization, (2) the automated model selection and parameter tuning, comparing multiple regression models for spatial interpolation, (3) the proxy-based spatial interpolation method by including spatial data layers or in situ measurable variables as predictors for contaminant concentrations and groundwater levels, and (4) the new well optimization algorithm to identify the most effective subset of wells for maintaining the spatial interpolation ability for long-term monitoring. We demonstrate our methodology using the monitoring data at the Savannah River Site F-Area. Through this open-source PyLEnM package, we aim to improve the transparency of data analytics at contaminated sites, empowering concerned citizens as well as improving public relations.

Improving Long-term Monitoring of Contaminated Groundwater at Sites where Attenuation-based Remedies are Deployed.

This study presents an effective approach to tackle the challenge of long-term monitoring of contaminated groundwater sites where remediation leaves residual contamination in the subsurface. Traditional long-term monitoring of contaminated groundwater sites focuses on measuring contaminant concentrations and is applicable to sites where contaminant mass is removed or degraded to a level below the regulatory standard. The traditional approach is less effective at sites where risk from metals or radionuclides continues to exist in the subsurface after remedial goals are achieved. We propose a long-term monitoring strategy for this type of waste site that focuses on measuring the hydrological and geochemical parameters that control attenuation or remobilization of contaminants while de-emphasizing contaminant-concentration measurements. We demonstrate how this approach would be more effective than traditional long-term monitoring, using a site in South Carolina, USA, where groundwater is contaminated by several radionuclides. A comprehensive enhanced attenuation remedy has been implemented at the site to minimize discharge of contamination to surface water. The immobilization of contaminants occurs in three locations by manipulation of hydrological and geochemical parameters, as well as by natural attenuation processes. Deployment of our proposed long-term monitoring strategy will combine subsurface and surface measurements using spectroscopic tools, geophysical tools, and sensors to monitor the parameters controlling contaminant attenuation. The advantage of this approach is that it will detect the potential for contaminant remobilization from engineered and natural attenuation zones, allowing potential adverse changes to be mitigated before contaminant attenuation is reversed.

Climate change impact on residual contaminants under sustainable remediation.

This study investigates the potential impact of climate change on residual contaminants in vadose zones and groundwater. We assume that the effect of climate changes can be represented by perturbations in the natural recharge through the aquifer system. We perform numerical modeling of unsaturated/saturated flow and transport and consider different performance metrics: contaminant concentrations at observation wells and contaminant export at the site's boundary. We evaluate the effect of increasing and decreasing recharge as well as the impact of potential failure of surface capping structures employed to immobilize vadose zone contaminants. Our approach is demonstrated in a real case study by simulating transport of non-reactive radioactive tritium at the U.S. Department of Energy's Savannah River Site. Results show that recharge changes significantly affect well concentrations: after an initial slight dilution we identify a significant concentration increase at different observation wells some years after the recharge increase and/or the cap failure, as a consequence of contaminants' mobilization. This effect is generally emphasized and occurs earlier as the recharge increases. Under decreased aquifers' recharge the concentration could slightly increase for some years, due to a decrease of dilution, depending on the magnitude of the negative recharge shift. We identify trigger levels of recharge above which the concentration/export breakthrough curves and the time of exceedance of the Maximum Contaminant Level for tritium are remarkably affected. Moreover, we observe that the contaminant export at the control plane, identified as the risk pathway to the downgradient population, may only be minimally affected by shifts in the natural recharge regime, except for some extreme cases. We conclude that more frequent sampling and in-situ monitoring near the source zone should be adopted to better explain concentrations' anomalies under changing climatic conditions. Moreover, the maintenance of the cap is critical not only to sequester residual contaminants in the vadose zone, but also to reduce the uncertainty associated with future precipitation changes. Finally, realistic flow and transport simulations achieved through proper calibration processes, rather than conservative modeling, should be adopted to identify non-trivial trade-offs which enable better allocation of resources towards reducing uncertainty in decision making.

In Situ Monitoring of Groundwater Contamination Using the Kalman Filter.

This study presents a Kalman filter-based framework to establish a real-time in situ monitoring system for groundwater contamination based on in situ measurable water quality variables, such as specific conductance (SC) and pH. First, this framework uses principal component analysis (PCA) to identify correlations between the contaminant concentrations of interest and in situ measurable variables. It then applies the Kalman filter to estimate contaminant concentrations continuously and in real-time by coupling data-driven concentration-decay models with the previously identified data correlations. We demonstrate our approach with historical groundwater data from the Savannah River Site F-Area: We use SC and pH data to estimate tritium and uranium concentrations over time. Results show that the developed method can estimate these contaminant concentrations based on in situ measurable variables. The estimates remain reliable with less frequent or no direct measurements of the contaminant concentrations, while capturing the dynamics of short- and long-term contaminant concentration changes. In addition, we show that data mining, such as PCA, is useful to understand correlations in groundwater data and to design long-term monitoring systems. The developed in situ monitoring methodology is expected to improve long-term groundwater monitoring by continuously confirming the contaminant plume's stability and by providing an early warning system for unexpected changes in the plume's migration.

Unrefined humic substances as a potential low-cost amendment for the management of acidic groundwater contamination.

The present study explores a novel application of Huma-K, a commercially available, unrefined humic substance, as a promising low-cost source of organic matter for in situ remediation of contaminated acidic groundwater plumes. This can be achieved by creating a humic-rich coating on the surface of minerals which can enhance the sorption of contaminants from groundwater. Huma-K was characterized by means of scanning electron microscopy equipped with energy dispersive spectroscopy, Fourier-transform infrared analysis, and potentiometric titrations. Batch experiments were performed to investigate the sorption-desorption behavior of Huma-K and to evaluate what conditions (pH, contact time, and initial Huma-K concentration) affect these processes upon injection into aquifer sediments. As evidenced by potentiometric titrations, Huma-K possesses functional groups that have an acidic nature, with pK values in the range of 4-6 (carboxylic) and 9-10 (phenolic). Sorption, homogeneous precipitation, and surface-induced precipitation seem to be favored in the presence of sediment at pH 4, where there is less deprotonation of acidic functional groups. As the pH is increased, functional groups become negatively charged, leading to electrostatic repulsion and dissolution of Huma-K from sediment. Kinetic experiments indicate that Huma-K sorption is a slow-rate process, most likely governed by film diffusion. The enhanced sorption of Huma-K in acidic conditions suggests that it may be used to create a subsurface treatment zone in acidic aquifers for the sequestration of contaminants such as uranium. The treatment zone will persist as long as the pH does not increase sufficiently to cause soil-bound Huma-K to be released, remobilizing aqueous contaminants.

Identifying key controls on the behavior of an acidic-U(VI) plume in the Savannah River Site using reactive transport modeling.

Acidic low-level waste radioactive waste solutions were discharged to three unlined seepage basins at the F-Area of the Department of Energy (DOE) Savannah River Site (SRS), South Carolina, USA, from 1955 through 1989. Despite many years of active remediation, the groundwater remains acidic and contaminated with significant levels of U(VI) and other radionuclides. Monitored Natural Attenuation (MNA) is a desired closure strategy for the site, based on the premise that regional flow of clean background groundwater will eventually neutralize the groundwater acidity, immobilizing U(VI) through adsorption. An in situ treatment system is currently in place to accelerate this in the downgradient portion of the plume and similar measures could be taken upgradient if necessary. Understanding the long-term pH and U(VI) adsorption behavior at the site is critical to assess feasibility of MNA along with the in-situ remediation treatments. This paper presents a reactive transport (RT) model and uncertainty quantification (UQ) analyses to explore key controls on the U(VI)-plume evolution and long-term mobility at this site. Two-dimensional numerical RT simulations are run including the saturated and unsaturated (vadose) zones, U(VI) and H(+) adsorption (surface complexation) onto sediments, dissolution and precipitation of Al and Fe minerals, and key hydrodynamic processes are considered. UQ techniques are applied using a new open-source tool that is part of the developing ASCEM reactive transport modeling and analysis framework to: (1) identify the complex physical and geochemical processes that control the U(VI) plume migration in the pH range where the plume is highly mobile, (2) evaluate those physical and geochemical parameters that are most controlling, and (3) predict the future plume evolution constrained by historical, chemical and hydrological data. The RT simulation results show a good agreement with the observed historical pH and concentrations of U(VI), nitrates and Al concentrations at multiple locations. Mineral dissolution and precipitation combined with adsorption reactions on goethite and kaolinite (the main minerals present with quartz) could buffer pH at the site for long periods of time. UQ analysis using the Morris one-at-a-time (OAT) method indicates that the model/parameter is most sensitive to the pH of the waste solution, discharge rates, and the reactive surface area available for adsorption. However, as a key finding, UQ analysis also indicates that this model (and parameters) sensitivity evolves in space and time, and its understanding could be crucial to assess the temporal efficiency of a remediation strategy in contaminated sites. Results also indicate that residual U(VI) and H(+) adsorbed in the vadose zone, as well as aquifer permeability, could have a significant impact on the acidic plume long-term mobility.

Persistent source influences on the trailing edge of a groundwater plume, and natural attenuation timeframes: the F-Area Savannah River Site.

At the Savannah River Site's F-Area, wastewaters containing radionuclides were disposed into seepage basins for decades. After closure and capping in 1991, the U.S. Department of Energy (DOE) has being monitoring and remediating the groundwater plume. Despite numerous studies of the plume, its persistence for over 20 years has not been well understood. To better understand the plume dynamics, a limited number of deep boreholes were drilled to determine the current plume characteristics. A mixing model was developed to predict plume tritium and nitrate concentrations. We found that the plume trailing edges have emerged for some contaminants, and that contaminant recharge from the basin's vadose zone is still important. The model's estimated time-dependent basin drainage rates combined with dilution from natural recharge successfully predicted plume tritium and nitrate concentrations. This new understanding of source zone influences can help guide science-based remediation, and improve predictions of the natural attenuation timeframes.

Evaluation of a radioiodine plume increasing in concentration at the Savannah River Site.

Field and laboratory studies were carried out to understand the cause for steady increases in (129)I concentrations emanating from radiological basins located on the Savannah River Site, South Carolina. The basins were closed in 1988 by adding limestone and slag and then capping with a low permeability engineered cover. Groundwater (129)I concentrations in a well near the basins in 1993 were 200 pCi L(-1) and are presently between 400 and 1000 pCi L(-1). Iodine speciation in the plume contained wide ranges of iodide, iodate, and organo-iodine concentrations. First-order calculations based on a basin sediment desorption study indicate that the modest increase of 0.7 pH units detected in the study site groundwater over the last 17 years since closure of the basins may be sufficient to produce the observed increased groundwater (129)I concentrations near the basins. Groundwater monitoring of the plume at the basins has shown that the migration of many of the high risk radionuclides originally present at this complex site has been attenuated. However, (129)I continues to leave the source at a rate that may have been exacerbated by the initial remediation efforts. This study underscores the importance of identifying the appropriate in situ stabilization technologies for all source contaminants, especially if their geochemical behaviors differ.

Characterization of DNAPL from the U.S. DOE Savannah River Site.

The composition of chlorinated hydrocarbon DNAPLs (dense non-aqueous phase liquids) from field sites can be substantially different than the material originally purchased for use as a solvent. Waste management practices at the U.S. Department of Energy's (DOE) Savannah River Site (SRS) included co-disposal of a wide range of organic and inorganic wastes. In 1991, a clear, orange-colored DNAPL was found in two wells near the SRS M-area settling basin. Waste effluent from the fuel and target fabrication facilities that were discharged to this settling basin included acids, caustics, metals and chlorinated solvents. The characterization of the SRS DNAPL suggests that numerous constituents partitioned into the DNAPL during its use as a solvent, co-disposal and ultimate migration through the subsurface. Trace constituents in the DNAPL include metals, from processing operations or co-disposal practices and subsurface minerals, high molecular weight hydrocarbons and alkyl esters, and acids. This complex mixture results in DNAPL-water interfacial properties that are substantially different than would be expected from a simple mixture of PCE and TCE. Under conditions when there is a high DNAPL to water volume ratio, a semi-rigid film accumulates on water droplets suspended in the DNAPL. It is concluded that the array of precipitated metal species comprising this film contributes to the interfacial tension that is over an order of magnitude lower than expected for a "clean" PCE/TCE mixture.

Influence of sources on plutonium mobility and oxidation state transformations in vadose zone sediments.

Well-defined solid sources of Pu(III) (PuCl3), Pu(IV) (Pu (NO3)4 and Pu (C2O4)2), and Pu(VI) (Pu02(NO3)2) were placed in lysimeters containing vadose zone sediments and exposed to natural weather conditions for 2 or 11 years. The objective of this study was to measure the release rate of Pu and the changes in the Pu oxidation states from these Pu sources with the intent to develop a reactive transport model source-term. Pu(III) and Pu(IV) sources had identical Pu concentration depth profiles and similar Pu release rates. Source release data indicate that PuIV(C2O4)2 was the least mobile, whereas Pu(VI)O2(NO3)2 was the most mobile. Synchrotron X-ray fluorescence (SXRF) revealed that Pu was very unevenly distributed on the sediment and Mn concentrations were too low (630 mg kg(-1)) and perhaps of the wrong mineralogy to influence Pu distribution. The high stability of sorbed Pu(IV) is proposed to be due to the formation of a stable hydrolyzed Pu(IV) surface species. Plutonium X-ray absorption near-edge spectroscopy (XANES) analysis conducted on sediment recovered at the end of the studyfrom the Pu(IV)(NO3)4- and Pu(III)(III)Cl3-amended lysimeters contained essentially identical Pu distributions: approximately 37% Pu(III), 67% Pu(IV), 0% Pu(V), and 0% Pu(VI). These results were similar to those using a wet chemistry Pu oxidation state assay, except the latter method did not detect any Pu(III) present on the sediment but instead indicated that 93-98% of the Pu existed as Pu(IV). This discrepancy was likely attributable to incomplete extraction of sediment Pu(III) by the wet chemistry method. Although Pu has been known to exist in the +3 oxidation state under microbially induced reducing conditions for decades, to our knowledge, this is the first observation of steady-state Pu(III) in association with natural sediments. On the basis of thermodynamic considerations, Pu(III) has a wide potential distribution, especially in acidic environments, and as such may warrant further investigation.