2g-3: Legacy site assessment and management

Remediation planning, design and construction is underway for the South River and South Fork Shenandoah River in Virginia to address legacy mercury. Mercury was used as a catalyst in manufacturing from 1929 to 1950 at a former DuPont facility in Waynesboro, VA. In the 1970s, a fish consumption advisory was administered, and it remains in effect today. Following investigation, an engineering remedial alternatives analysis and public input, the state selected a Monitored Natural Recovery plan and initiated a monitoring program in 1984. Because mercury levels in tissues of some species of fish were not declining as predicted, the South River Science Team was established in 2001 to understand why fish tissue levels are not decreasing and to evaluate options to address this challenge.

A conceptual model describing the fate and transport of mercury in the South River watershed and possible actions to reduce mercury in fish tissue was developed with input from the South River Science Team. The conceptual model indicates that erosion of bank soils containing legacy mercury is the principle source of loading to the aquatic system. However, due to the size of the river and the complexity of mercury cycling, uncertainties in the conceptual model and the effectiveness of possible remedial actions remain. Moreover, because remedial actions are being constructed on public and private properties, community stakeholders and landowners have questioned the net benefit of the remedial efforts. Remedial designs have been revised to incorporate stakeholder preferences. In order to maximize the likelihood of remedy effectiveness, minimize short-term and long-term risks and address other potential stakeholder concerns, an adaptive management approach to remediation is being implemented.

Adaptive management requires a structured process of prediction, action, monitoring, feedback and adjustment. Based on pre-remedy and post-remedy monitoring, we will show how the initial phase of remediation has achieved the best short-term outcome while reducing technical and social uncertainty via system monitoring and stakeholder feedback.

Mercury was used as a catalyst in rayon production at a former DuPont facility in Waynesboro, VA from 1929 to 1950. In the 1970s, a fish consumption advisory was established and remains in effect. Fish mercury concentrations have not declined as expected and range from approximately 0.2 to >4 µg/g in 250-300 mm Smallmouth Bass along a 40 km river section downstream of the former rayon facility. DuPont and the Virginia Department of Environmental Quality established the South River Science Team in 2001 to understand why and to evaluate the feasibility of remedial options. Contaminated river banks continue to supply mercury to the river. Total mercury concentrations in surface sediments (0-3 cm) are 10-20 µg/g in some areas. Methylmercury concentrations in surface sediments are above 150 ng/g at some locations. Mechanistic modeling is being carried out to help understand reasons for the slow natural recovery and evaluate the potential effectiveness of remediation options. A mechanistic model of mercury cycling and bioaccumulation in aquatic systems (Dynamic Mercury Cycling Model, D-MCM) was used to simulate 44 km of the contaminated river downstream of the former DuPont facility. The modeled portion of the river was divided into 47 segments ranging in length from 0.5 to 1.1 km. Inorganic Hg(II), elemental mercury and methylmercury were simulated in the water column, sediments and food web. Hydrodynamics, bank erosion, bank leaching, particle transport and key processes in the aquatic mercury cycle were included. Simulations of existing conditions spanned 2006-2014. Model simulations reasonably captured key spatial trends in the contaminated system. Observed and modeled concentrations of total mercury and methylmercury in water, sediments and biota, reached peak levels 5-25 km downstream of the original point source. Ongoing erosion of mercury-contaminated solids from river banks was an important source of continued elevated mercury levels in the simulated system. Predicted and observed concentrations of total mercury and methylmercury in water, sediments, and biota will be presented. Factors predicted to be responsible for spatial and temporal trends at the site will be discussed, as well as the potential benefits of selected remediation options, e.g. stabilizing selected reaches of river banks.

Mechanistic computer models that simulate the methylation of mercury in anoxic sediments are an underutilized tool to support remediation measures. Computer models have the advantage of investigating responses of a system to various perturbations, such as different management strategies or the effect of climate change. Application of a mechanism-based model for remediation studies requires building the model using thermodynamic and kinetic constraints and testing its sensitivity to variation of model parameters. A biogeochemical reaction-transport model using the PhreeqC program was modified from prior studies to improve the model description of Hg methylation. Data sets from two mercury-impacted water bodies in California were used to establish a range of values for model input and verify the accuracy of the model response. Rate equations that describe Hg methylation resulting from sulfate or iron reduction were implemented in different ways to compare model response to factors such as sulfate concentration, organic matter, and inorganic mercury (Hg(II)) concentration. The model can take into account the adsorption of Hg(II) to sediments, and dissolution and precipitation of nanosize mercury sulfide to determine Hg(II) concentration. Another parameter that is important but difficult to implement accurately is the model treatment of dissolved organic matter since it can both reduce the bioavailability of Hg(II) through complexation that leads to particles that are too big to pass cell membranes, and it can increase methylation by stimulating bacterial growth. Moreover, the model had to be adapted to the different conditions of the two study sites, since they vary from each other in that one is a managed wetland while the other one is a reservoir impacted by legacy mercury. We compared the model simulations of the two sites as examples demonstrating how to build mechanistic models that can be used to plan remediation measures. For all processes described in the model, a balance between accuracy and practicality had to be found to make the model applicable.

Elevated concentrations of mercury (Hg) distributed in a range of environmental compartments, including rivers, sediments, lakes, and oceans, are a world-wide concern. A number of remediation methods are available for removing or stabilizing Hg at contaminated sites, including natural attenuation, sediment dredging, soil washing, phytoremediation, and soil and sediment amendments. Here, we investigate the effectiveness of switchgrass biochars (pyrolyzed at 300 and 600°C) as amendments to stabilize Hg under anaerobic conditions. Switchgrass is widely available in North America and can be used to produce biochar at low cost. Microcosm experiments were conducted to evaluate the control of total Hg and methylmercury (MeHg) by mixing Hg-contaminated sediment, biochar and river water under reducing conditions over 600 days. The results indicate aqueous concentrations of total Hg and MeHg were greatly reduced in the presence of biochars over most of the experimental period, with the exception of a spike in MeHg concentration observed at ~440 days in the 600°C biochar system. The aqueous concentrations of Hg were observed to vary over time. Initially, moderate concentrations of Hg were leached from the sediment. With the onset of Fe(III)-reducing conditions, large increases in Hg concentrations were observed, likely due to the release of Hg during reductive dissolution of Fe(III)oxides. As sulfate-reducing conditions were established, aqueous concentrations of Hg declined, likely through the formation of Hg-sulfide phases. The first peak of MeHg concentrations was observed during the onset of Fe(III)- and sulfate-reducing conditions. At late times, methanogenic conditions were established and a second peak of MeHg was observed. Micro-X-ray fluorescence mapping and confocal micro-X-ray fluorescence imaging showed the co-occurrence of Hg with S, Fe, Cu and other elements within or on the surface of biochar particles. Hg LIII edge X-ray absorption near edge structure (XANES) and extended X-ray absorption fine structure (EXAFS) analyses indicated that the Hg was present within the biochar particles as Hg-sulfide phases. This study suggests that Hg is bound as stable phases within biochar particles, potentially limiting its rate of transfer back to the aqueous phase and availability for methylation.

Mercury polluted contaminated sites are very serious problems in many countries round the globe. History of industrial area in Neratovice (company Spolana, Czech Republic) started at 1898. Spolana used amalgam electrolysis to produce chlorine since 1948. In sixties, the production of chlorinated pesticides started including components for Agent Orange. In the production of these types of chemicals, a number of employees fell seriously ill and production led to the extensive contamination by dioxins and other persistent organic pollutants.

In 2001, the activities concerning to the decontamination and decommission production building operations contaminated with dioxins and mercury. In August 2002, during the catastrophic floods affecting the lower basin of the Vltava and Elbe Spolana site was inundated by overflowing Labe.

Waste from old amalgam electrolysis Spolana were weighed incinerator SPOVO Ostrava during the years 2010 - 2012. For disposal of these waste the combustion part consists of a rotary kiln and secondary afterburner chamber where the wastes are removed at 1 100-1 200 C, was used. Noncombustible leaving the rest in the form of slag. Providing the necessary combustion temperature during start, operation, and shutdown is achieved by additional gas burners. Secondary afterburner chamber with a retention time of about 2 s to ensure total destruction of highly stable hazardous substances (eg. PCBs, CFCs). Cleaning flue gas - a set of several technics allowing to remove pollutants from flue gases - two-stage wet scrubbing, dioxin filter, DeNOx catalyst.

Wastes were, according to the load mercury content, divided into two groups, namely waste and bulk piece with the different content of mercury.

Overall, therefore, in the material contained an estimated 3 963 kg Hg, 6.5 g of 2,3,7,8-TCDD and about 30 kg of PCB. SPOVO technology without issue warrants listed POPs disposal, if we start from the above it can be assumed that up to around 3 600 kg Hg would be left in the form of vapor into the separator device, which must be captured.

On the basis of the operating parameters of a hazardous waste incinerator SPOVO can assume that the device is completely satisfactory for the decomposition of the material present in the reporting of persistent organic pollutants. Likewise, it is reasonable to assume that the mercury present will largely transferred to the gas phase and its disposal sufficient decides the effectiveness of cleaning of the incinerator. Following these procedures can ensure appropriate disposal of such waste.

The remediation of mercury (Hg) contaminated terrestrial sites is a challenging area because of the elements mobility and tendency to form toxic compounds which can bio-accumulate. Nature may have provided a remedy. Certain bacteria have the ability to tolerate and disburse Hg from their surroundings using intracellular reduction of Hg2+ cations to elemental Hg (Hg0), which is volatile and disbursed passively in gaseous elemental form (GEM). Enhanced Hg emissions can be readily captured with existing technology. This biological redox reaction is controlled by a mercuric reductase enzyme that is coded for by the merA gene, and using merA carrying bacteria followed by Hg capture and storage may assist in reducing the biogeochemical cycling of this toxic element. Delivery for bio-augmenting sites is challenging considering transport issues for microbial cultures in large volumes, the heterogeneous nature of the terrain, and locations of Hg contaminated sites.

To overcome this delivery issue, this research used a biopolymer to immobilize Pseudomonas veronii, on a solid bulk substrate (zeolite). This bio-functionalized zeolite was dried and stored for several months, then applied to heavily Hg contaminated material. GEM flux from the treated and control material was measured directly using a dual setup of Tekran 2537 ambient mercury analysers using dynamic flux chambers. Oxidised mercury species were also measured using a dual pump set up with CEM filters that were later desorbed and total oxidized Hg content measured using a Tekran 2600 analyser. Two watering regimes were used on both inoculated and control soil trays, 15 and 50 percentv/v, and no nutrients were applied. Bio-functional zeolite was added in a ratio of 50%v/v or 15%w/w. Gaseous elemental mercury flux was measured for a minimum of twelve days post inoculation and three days prior.

Results show a dramatic and distinct spike in GEM emissions after inoculation, with higher emissions under both watering regimes compared to water only and background emissions. A corresponding reduction in gaseous oxidised Hg emissions was noted, as would be expected given the enhanced redox reactions going on. This evidence strongly suggests P.veronii were able to enhance GEM emissions to the point where capture and extraction become viable. Further work is required to optimize the watering and nutrient regime to establish a colony sufficiently self- sustaining such that remediation management requirements are limited, and to establish kinetics parameters.

Mercury is present in the Liquid Waste System at the Savannah River Site (SRS), a US Department of Energy facility located near Aiken, South Carolina. This system contains legacy radioactive liquid waste that was generated over several decades of nuclear material production in support of the US nuclear stockpile. Historically inorganic mercury nitrate was used as a catalyst to dissolve aluminum assemblies following irradiation. Approximately 60,000 kg of the mercury is present within the approximately 37 million gallons of radioactive waste that is present within this system. The mercury has the potential to impact waste disposal requirements of the final stabilized material.

The liquid waste system uses evaporators to vaporize water from the liquid solution to minimize volume (storage) requirements. Analysis of the feedstock to the evaporators has revealed that the total mercury concentration is 234 mg/L. Mercury present in the soluble fraction (186 mg/L) is distributed between the elemental state (3.46 mg/L), inorganic ionic (50.5 mg/L), and organic (134 mg/L) species. During evaporator operations elemental mercury accumulates and condenses as an immiscible liquid phase that is drained by gravity from the evaporator system. In 2015 over 33-liters of elemental mercury was removed from the liquid waste system during evaporator operations. The removal process for elemental mercury is based upon chemical thermodynamics that dictate the partitioning of elemental mercury between the liquid and vapor phase. This provides one of the most effective purge points for the removal of mercury from the system.

In this investigation chemical reduction and reductive demercuration processes are evaluated as mechanisms to increase the removal of mercury from the evaporator unit operation. The objective is to increase the removal of elemental mercury by converting a fraction of either the ionic or the organic mercury that is present in the system. Non-radioactive waste simulants were used in laboratory scale investigations to determine the effectiveness of various industrial reducing agents in converting the various mercury species that are present in this complex-alkaline solution. The investigation included a selected reductant that is known to facilitate reductive demercuration, an organic mercury removal component that is often coupled with oxymercuration processes used in organic synthesis. Full-scale implementation of the processes within the evaporator unit operation could significantly increase the volume of mercury removed from an existing purge point in the liquid waste system.

Monomethylmercury and other organomercury species which are present in the Savannah River Site (SRS) Liquid Waste System (LWS) decrease the effectiveness of mercury removal /control systems and increase the leachability of mercury from final wasteforms. An overarching objective of ongoing SRS LWS mercury activities is removal of mercury from the LWS at a rate that would reduce overall mercury mass in the system over time thus making operations sustainable as the entire radioactive waste inventory is processed into final stabilized forms (primarily high level waste glass canisters and a cementitious saltstone). An emerging strategy to help meet this objective is deployment of technology to alter mercury speciation within the LWS to control mercury behavior. Advanced oxidation processes, such as photoreactions and ozone reactions, have the potential to break down methylmercury into ionic and elemental mercury forms. In this speciation-control example, conversion of organomercury to inorganic ionic mercury or elemental mercury would measurably improve saltstone performance and would provide a purge point for mercury from the LWS an existing high-quality environmentally-protective wasteform.

As a scoping step, the mechanisms and rates for the conversion of methylmercury to inorganic mercury species by advanced oxidation processes were evaluated/quantified. Two technology classes were studied, photooxidation with UVC wavelength light and oxidation by ozone. All phases of the research emphasized conditions that are applicable to the complex chemistry of alkaline liquid waste solutions/simulants. The work was performed in several phases: 1) preparation of methylmercury simulants/solutions, 2) determination of the optical properties of the simulants/solutions, 3) development of rapid mercury speciation methods, 4) photoreactor tests, and 5) ozone tests. The research confirmed that advanced oxidation techniques, particularly photoreactions, are capable of converting monomethylmercury into inorganic mercury species such as inorganic ionic mercury and elemental mercury (along with other more recalcitrant mercury species). The rate of reactions was relatively slow, particularly in the complex redox chemistry of alkaline LWS simulants which contain a number of metals and high levels of nitrate/nitrite. The data suggest that developing effective oxidation processes will be a challenge and that such processes are likely to result in significant collateral impacts on the chemistry and flowsheet for the LWS.