Loading...
HomeMy WebLinkAboutNC0004961_11. RBSS_CAP Part 2 Appx H_FINAL_20160212 Appendix H Monitored Natural Attenuation Technical Memorandum This page intentionally left blank Geochemical, LLC PO Box 1468 Socorro, NM 87801 USA (575) 838-0505 www. geochemical. com Technical Memorandum To Mark Filardi / HDR Engineering, Inc. From Gregory P. Miller Date October 14, 2015 Subject Riverbend Steam Station Corrective Action Plan Monitored Natural Attenuation – Tier I and II This technical memorandum presents the Tier I and Tier II evaluation of the attenuation of compounds dissolved in groundwater at the Riverbend Steam Station (RBSS), Gaston County, North Carolina. The memorandum consists of three sections: Background on Monitored Natural Attenuation (MNA) as a remedial alternative for coal combustion products; the Findings of the Tier I MNA Demonstration activity; and the Findings of the Tier II Demonstration. Investigative and data evaluation efforts are in progress as of the date of this memorandum. The findings herein represent the understanding as of this date and are subject to revision as new information becomes available. BACKGROUND Introduction to Monitored Natural Attenuation for Groundwater MNA is a strategy and a set of procedures used to demonstrate that physiochemical and/or biological processes in an aquifer will reduce concentrations of undesirable substances to levels below regulatory concern. It has been broadly applied to releases of petroleum hydrocarbons in many hydrogeological environments. There has been less application of MNA to the remediation of inorganic or radiologic substances in groundwater than to organic compounds. While MNA of organic contaminants is a readily accepted remedy at the state and federal level, inorganic attenuation is more complicated and there is limited implementation experience in industry, science, and government. Metals, radionuclides, and/or other inorganic compounds are found in all aquifers. The mechanisms that regulate their release from solids and movement through aquifers are for the most part the same processes that control movement of inorganics in aquifers impacted by Coal Combustion Products (CCP) leachate. These processes attenuate the concentration of inorganics in groundwater by depositing inorganics on aquifer solids. Unlike organic compounds that break down from hydrolysis or bacterial action, a reduction in concentration of dissolved inorganic compounds requires water to be removed from the aquifer, or the compounds immobilized by conversion to, or adsorption onto, solids. When demonstrated that the mechanism and permanence of natural processes will result in attenuation of undesirable compounds to acceptable levels, then remediation can be conducted by verifying that the remediation proceeds as predicted. MNA relies on natural processes to remove contaminants of concern from groundwater resulting in lower cost, and less disturbance of ground and infrastructure. Active remediation of inorganic compounds in the subsurface may rely on processes that are similar to or the same as natural attenuation mechanisms. Active remediation often seeks to improve or hasten these natural processes through chemical or biological augmentation of the aquifer. There are not clear distinctions between the late phases of conventional groundwater M. Filardi RSS MNA Tier I & II October 14, 2015 Page 2 of 24 remediation and MNA, but such distinctions may be unnecessary. Relying on limited or low- energy active remediation as a preliminary step, then achieving final remedial goals through MNA is sometimes called “enhanced MNA”, or “enhanced attenuation.” Active remediation methods for petroleum products often leave residual dissolved contamination that is degraded slowly, requiring monitoring under regulations allowing MNA. Enhanced MNA is a common strategy in hydrocarbon remediation and the principle is adaptable to inorganic MNA. Authorizing Statutes MNA is a federally recognized remedial technology that can meet Resource Conservation and Recovery Act (RCRA) Corrective Action requirements and could be included when remedial alternatives for legacy CCP management sites are evaluated. Implementation of MNA for inorganics at RCRA facilities is clearly and favorably proposed by the United States Environmental Protection Agency (USEPA) in the 1999 Office of Solid Waste and Emergency Response (OSWER) Directive 9200. 4-7P. In that directive MNA is defined as: “the reliance on natural attenuation processes (within the context of a carefully controlled and monitored site cleanup approach) to achieve site-specific remediation objectives within a time frame that is reasonable compared to that offered by other more active methods. The ‘natural attenuation processes’ that are at work in such a remediation approach include a variety of physical, chemical, or biological processes that, under favorable conditions, act without human intervention to reduce the mass, toxicity, mobility, volume, or concentration of contaminants in soil or groundwater. These in-situ processes include biodegradation; dispersion; dilution; sorption; volatilization; radioactive decay; and chemical or biological stabilization, transformation, or destruction of contaminants.” Site characterization and assessment of MNA is different for organic and inorganic strategies. The focus in organic MNA is on a working understanding of the transformation rates and daughter products when organic chemicals break down in aquifers. Complex organic compounds are reduced to water, carbon dioxide, methane, and salts by biotic and abiotic processes in organic MNA. The science and engineering looks to the dissolved phase concentrations derived from groundwater samples. From those analytical results and aquifer dimensions and flow rates, the mass and flux of organic contaminants is evaluated over time. Meeting a single objective, mass reduction of organic compounds, is all that is needed to validate the organic MNA remedial approach. For inorganic contaminants, the scientific and engineering investigations must consider both the dissolved and solid phase. This is because inorganic compounds are not destroyed (with the exception of some radioactive decay chains), but become comingled with the aquifer solids. Inorganic compounds that are removed from the dissolved phase are transferred to the aquifer solids in inorganic MNA processes. The undesirable inorganic compound is not destroyed; its location is changed, from dissolved and mobile to immobile via incorporation into/onto a solid. Because inorganics are left in place, there is stakeholder concern about the stability and longevity of the remedial action. It is partly this uncertainty regarding inorganic MNA that drives the need for more in-depth study. While much of the chemistry is abiotic, biotic reactions can be very important in some inorganic MNA pathways. Inorganic MNA has two focus points: 1. Demonstrate that the attenuation process is reducing contaminant mass in groundwater; and, 2. Demonstrate that the remediation will meet long-term stability criteria without intervention. M. Filardi RSS MNA Tier I & II October 14, 2015 Page 3 of 24 Principles of Inorganic Attenuation The 1999 OSWER Directive provides limited policy and implementation guidance for inorganic MNA. USEPA indicates that: the site-specific mechanisms of attenuation of inorganic contaminants should be known; and, the stability of the remedial action should be evaluated under potential changes in conditions. The OSWER Directive is fairly specific about the classes of attenuation processes that may be proven effective, stating: “MNA may, under certain conditions (e.g., through sorption or oxidation-reduction reactions), effectively reduce the dissolved concentrations and/or toxic forms of inorganic contaminants in groundwater and soil. Both metals and non-metals (including radionuclides) may be attenuated by sorption reactions such as precipitation, adsorption on the surfaces of soil minerals, absorption into the matrix of soil minerals, or partitioning into organic matter.” “Oxidation-reduction (redox) reactions can transform the valence states of some inorganic contaminants to less soluble and thus less mobile forms (e.g., hexavalent uranium to tetravalent uranium) and/or to less toxic forms (e.g., hexavalent chromium to trivalent chromium). Sorption and redox reactions are the dominant mechanisms responsible for the reduction of mobility, toxicity, or bioavailability of inorganic contaminants.” “It is necessary to know what specific mechanism (type of sorption or redox reaction) is responsible for the attenuation of inorganics so that the stability of the mechanism can be evaluated. For example, precipitation reactions and absorption into a soil’s solid structure (e.g., cesium into specific clay minerals) are generally stable, whereas surface adsorption (e.g., uranium on iron-oxide minerals) and organic partitioning (complexation reactions) are more reversible.” “Complexation of metals or radionuclides with carrier (chelating) agents (e.g., trivalent chromium with EDTA) may increase their concentrations in water and thus enhance their mobility. Changes in a contaminant’s concentration, pH, redox potential, and chemical speciation may reduce a contaminant’s stability at a site and release it into the environment. Determining the existence, and demonstrating the irreversibility, of these mechanisms is important to show that a MNA remedy is sufficiently protective.” “Inorganic contaminants persist in the subsurface because, except for radioactive decay, they are not degraded by the other natural attenuation processes. Often, however, they may exist in forms that have low mobility, toxicity, or bioavailability such that they pose a relatively low level of risk. Therefore, natural attenuation of inorganic contaminants is most applicable to sites where immobilization or radioactive decay is demonstrated to be in effect and the process/mechanism is irreversible.” There have been numerous advances in the understanding of the environmental geochemistry of inorganic compounds and metals since publication of the Directive in 1999. The inorganic MNA topics discussed in the Directive are still valid, and form a first step to evaluate inorganic MNA. Over the period of 2006-2010 USEPA has released much more detailed guidance on implementation of inorganic MNA, and the Interstate Technical and Regulatory Council (ITRC) has published on practical and regulatory aspects of implementing MNA inorganic at the State regulatory level. These guidance documents are summarized in a following section. M. Filardi RSS MNA Tier I & II October 14, 2015 Page 4 of 24 Governing Equations The concentration of a dissolved compound in flowing groundwater changes over time due to movement, mixing, and chemical reactions. The concentration of constituents of interest (COIs) in RBSS groundwater is governed by these same processes. At RBSS, groundwater movement and mixing probably changes COI concentrations as much or more than chemical reactions do. Following source control, COI concentrations in RBSS groundwater will decrease due to the combination of these processes, resulting in COI concentration attenuation. The change in COI concentration with respect to time and place can be described by the advection-dispersion equation (ADE). The ADE is used to determine the concentration of COI in groundwater during remedial scenarios, including MNA. In differential form, including reversible and irreversible sorption to solids, the ADE is: Ct S x Cvx CDt C b µδ δ θ ρ −−∂ ∂−∂ ∂=∂ ∂ 2 2 where: C = concentration of solute in water t = time x = distance from source D = Dispersivity v = average groundwater velocity ρb = solid bulk density ϴ = solid porosity S = concentration of solute on solid µ = rate of solute irreversible sorption The first and second terms in the ADE describe the change in COI concentration due to mixing and diffusion in granular aquifers (D - Dispersivity), and the change in concentration from groundwater flow moving the solute away from the point of interest (v - average groundwater velocity), respectively. The third term in the ADE, related to concentration of solutes on solids and reversible sorption processes, has alternate formulations that are very useful in solute fate and transport modeling. The relationship between the concentrations of reversibly adsorbed solutes on solids and the dissolved concentration in groundwater can be stated as a proportionality – Kd, where: 𝐾𝐾𝑑𝑑=𝑆𝑆𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 where: SB is for the concentration of COI on solids, and CB is for the concentration of COI in the water, when COI chemistry is at equilibrium. Kd is a proportionality of concentration between the solids and the liquids in an aquifer. It commonly has units of volume over mass. In order to use Kd in the ADE formulation it is often expressed as Retardation – R, where: 𝑅𝑅=1 +𝜌𝜌𝜃𝜃𝐾𝐾𝑑𝑑 M. Filardi RSS MNA Tier I & II October 14, 2015 Page 5 of 24 In the Retardation equation we see the reappearance of bulk density and porosity from the ADE with the surface concentration term contained in the proportionality Kd. The effect of R and D on solute transport time and concentration can be seen from the figure below for model transport of retarded and non-retarded solutes with equal Dispersivity. Note that in the retarded case ‘a’ that velocity is divided by Retardation; an R value of 1 indicates no retardation. Retardation and Dispersivity are conservative with respect to solute mass. If a COI has a high Kd, it will have a high R. A high R puts higher contaminant mass on the solids in proportion to the amount dissolved in water. When the water concentrations decrease to background over time, the solids will release COIs until that mass is exhausted to background concentrations. If R is low, that return to background concentrations on solids will happen relatively quickly. The final term in the ADE is related to irreversible sorption processes. To describe the attenuation of COI at RBSS the inputs to the ADE are satisfied as follows: ADE Variable Data Source C Concentration measured from groundwater samples. t Time is determined by the nature of the problem to be investigated, e.g., the time for COI in groundwater to fall below the 2L Standard. x Distance is determined by the nature of the problem to be investigated. D Dispersivity is estimated from published values or determined from solute concentrations over time using numerical models. v Average groundwater velocity varies in space and time and is supplied by a calibrated groundwater flow model. ρb Bulk density measured from samples. ϴ Porosity measured from samples. S Concentration measured from solid samples. Calculated Kd values from laboratory batch and column testing. µ Irreversible component of S. Not determined at this time. Kd Calculations and Interpretation There are many different mathematical models to describe sorption, Kd, Langmuir Isotherms, and Freundlich Isotherms are common methods (USEPA 1999). All of the models have advantages and limitations. If you know the solution concentration, and can model or measure the solid concentration, you can calculate Kd. If confidence in the model is high, the solution concentration can be assumed and surface concentrations of COI calculated directly (Goldberg et al. 2000; Dzombak and Morel 1990). 0 1 C/ C o Increasing Distance → D and R D Only a b tvxb=  + = d a K tvx θ ρ1 M. Filardi RSS MNA Tier I & II October 14, 2015 Page 6 of 24 One limitation to the Kd approach is that it does not describe the effect of variable pH conditions. The question of pH is important on many levels. We know that COI will take on an electrical charge based on pH, that the surfaces of minerals that adsorb COI are variably charged-again dependent on pH, and that the propensity of COI to approach a surface depends on all of these charges. To handle these complexities surface complexation theory was developed (Dzombak and Morel 1990). Surface complexation theory states that we can describe these sorption reactions the same way as we describe chemical reactions, as mass-action equations that have equilibrium constants. K’s are numerical constants for chemical mass-action equations. They express the equilibrium relationships between chemicals. Setting an equilibrium chemical reaction with A and B as reactants, C and D as products, where the variable is concentration: DCBA+=+, the mass-action equation takes the form: BA DCK=. K is constant at constant temperature and pressure. Three constants are needed to model the surface complexation approach, one for the COI adsorption to the surface and two to describe the charging of the surface. The surface charging constants K- and K+ are determined for pH- dependent protonation-deprotonation reactions at the oxide surface. The protonation reactions with the surface, SOH, are described by the two step reversible process below. The mass-action relationship for COI (KCOI) can be expressed in a generic form as: +−+⇔+HCOISHCOIHSOHyxyx [][] [][]() ()RT F COIHSOH HCOISH COI yx yxK 0expψ−+−=, where square brackets indicate concentrations (mol/L), F is the Faraday constant (C/mol), Ψ is the surface potential (V), R is the molar gas constant (J/mol K), T is the absolute temperature (K), and the exponential terms represent solid-phase activity coefficients that correct for charges on the surface complexes. Each KCOI mass-action expression must be constructed from a balanced equation using the generic form above as an example. Mass-action constants can be determined from adsorption experiments (Kd determinations) using the code FITEQL (Herbelin and Westall 1999) and several surface complexation sub-models are available in PHREEQC. Irreversible Sorption of COIs Irreversible sorption of COIs has been observed in/on a range of geomedia for over 100 years. Here the irreversible sorption for COIs is defined as: COIs that are retained in the structure of [] [][] () ()RT F HSOH SOH sK 02exp ψ + +=+ [][] [] () ()RT F SOH HSO sK 0exp ψ− − +−= ++⇔+2SOHHSOH SOHHSO⇔++− M. Filardi RSS MNA Tier I & II October 14, 2015 Page 7 of 24 minerals and mineral-like inorganic and organic compounds on a time scale significantly longer than reversible surface adsorption reactions. In most cases, reversible surface sorption reactions are instantaneous as compared to groundwater flow rates. Irreversible sorption processes for COIs are common in the natural environment. COIs will be taken up to form part of the mineral lattice of clay minerals, hydrous metal oxides, carbonates, and composted organic material, as examples. Clay minerals and hydrous metal oxides are common at the site. Release of COI from these materials is slow. The chemical equilibrium between solutions and solids is a balance between the rate of forward reactions and the rate of back reactions. Surface complexation adsorption processes are reversible equilibrium reactions. Irreversible sorption is a process where the forward reaction is so much faster than the back reaction that the reverse reaction is insignificant on the time scale of interest. The remobilization mechanism is too slow to be significant. Irreversible sorption is a COI sink; COI is removed from the flow system. First-order decay is one way to represent irreversible sorption in a numerical model. First-order decay rates are also used to calculate the half-life of a compound. Inverse Modeling of Hydrologic Parameters Determination of most of the physical properties that affect solute transport in granular aquifers can be accomplished using laboratory or field tracer tests. Dispersivity, retardation, and effective porosity can be quantified this way. These are important parameters for modeling of solute transport in groundwater. Laboratory and field tracer tests are controlled experiments often involving forward modeling of expected conditions to constrain the experimental design. Determining the appropriate concentration of tracer to be used and how long to apply it to have a successful experiment (source term) would be an example of forward modeling. Inverse modeling is fitting of aquifer parameters to observed concentration/distance/time data. Laboratory and field breakthrough curves (BTCs) generally trace smooth curves due to frequent sampling and the ability to rerun experiments. Data from groundwater monitoring programs can be used for determination of aquifer parameters in the same manner as laboratory tests, if the data provide a fairly complete record of changes in concentration of a dissolved solute. Ideally, monitoring would be frequent and the rising limb and falling tail of the solute BTC would be captured with accuracy and precision. Forward modeling of field conditions is possible if the source term is known or assumed. BTC data from groundwater monitoring generally has an imperfect understanding of the concentration (C) and duration (t) of the contaminant source over time – the source term. The distance between the source (z) and the monitored well may also be hard to quantify. A public-domain program for numerical evaluation of BTCs is available. CXTFIT, Version 2.1 (Toride et al. 1999), is useful for determining values for average linear velocity (v), dispersivity (D), retardation (R), and decay (µ), which is used to account for irreversible adsorption. Independent data for v are available from the calibrated flow model, tests of hydraulic conductivity, and measurement of gradients. Similarly, R has been determined independently from laboratory measurements of bulk density (ρb), effective porosity (n), and partition coefficients (Kd). Laboratory or field determined values for dispersivity (D) or decay (µ) are not available at this time. M. Filardi RSS MNA Tier I & II October 14, 2015 Page 8 of 24 Coal Combustion Product Leachate Characteristics Inorganic MNA relies on being able to observe, quantify, and in some cases modify subsurface chemistry. There are major differences in the chemical characteristics of leachate across the major groups of CCP (fly ash, bottom ash, pollution control residue), and complexities introduced by co-disposal of different CCPs. The composition of leachate from CCP is highly variable. Controls on leachate chemistry are numerous and are strongly influenced by site- specific factors. Leachate chemistry is influenced to variable and unequal degrees by: • The type of coal; • The basin, sub basin, mine, or seam it is produced from; • Variability of coal composition within the seam; • Coal cleaning and pretreatment processes; • Combustion conditions; • Pollution control operations; • CCP handling; • CCP management strategies – impoundments, landfills or combined methods; and, • Climate and environmental factors - such as hydrology, geology, and weather. The preceding list of factors, that ultimately control the success of inorganic MNA, is imperfect and not meant to be exhaustive. There is a large knowledge base on the geochemistry of CCP and CCP leachate. That information is very helpful to constrain remedial strategies for legacy CCP sites, but does not replace the detailed site-specific assessment of ash and aquifer geochemistry required for inorganic MNA. The Electric Power Research Institute (EPRI) has conducted extensive research on the occurrence, concentration, and mobility of CCP leachate components. The selection criteria the used to identify constituents of concern (COCs) included prevalence, mobility, and risk as determined by comparison to water quality standards. EPRI’s assessment identifies five leachate constituents that are probably present in porewater at all legacy CCP management sites; arsenic, boron, chromium, selenium, and sulfate. Each of these substances has triggered at least one remedial action at a legacy CCP site. While these constituents are the primary focus of concern due to their prevalence and mobility, several other inorganic constituents may be of concern at legacy CCP management sites. From EPRI’s 2006 effort the following constituents are noted: “The leachate data indicate that concentrations of antimony, arsenic, cadmium, chromium, selenium, and thallium were higher than health-based MCLs in at least 10 percent of the samples. In addition, the 90th percentile concentrations of boron, lithium, manganese, molybdenum, sodium, sulfate, and vanadium were higher than alternative drinking water criteria. These constituents are more likely to trigger a remedial action in the event of a leachate release than constituents that typically have leachate concentrations lower than drinking water standards.” The EPRI-listed constituents above are primary candidates for MNA because of their occurrence, mobility, and propensity to exceed drinking water standards at legacy CCP sites. Not all of these constituents are amenable to remediation through MNA processes because of their chemical properties. Table 1 compares the leachate chemistry as identified by the previously listed EPA and EPRI sources, with the compounds suitable for inorganic MNA as found in the EPA, EPRI, and ITRC guidance documents. M. Filardi RSS MNA Tier I & II October 14, 2015 Page 9 of 24 The EPA sources cited identify the following 13 COCs based on the potential for human health and/or ecological impacts using a screening risk assessment: aluminum, arsenic, antimony, barium, boron, cadmium, cobalt, chromium, lead, mercury, molybdenum, selenium, and thallium. One of the findings of the 2009 EPA report was nine of the 13 COCs listed above were found in CCP leachate at levels of concern for groundwater protection, as indicated in Table 1. Published Guidance on Inorganic Monitored Natural Attenuation United States Environmental Protection Agency Guidance The USEPA is the primary source of guidance on MNA. The USEPA has published their inorganic MNA guidance in three volumes. Volume 1 is a procedural and technical guide for implementing MNA; Volume 2 is a detailed exploration of issues associated with inorganic MNA for nine compounds; and Volume 3 contains discussion of MNA issues unique to radionuclides and a detailed exploration of inorganic MNA for 10 radionuclides. Volumes 1 and 2 were published in 2006 and Volume 3 in 2010. The USEPA guidance documents cover inorganic MNA at a high level of scientific detail. The USEPA guidance, while not specific to CCP, is a technical resource on evaluation of the feasibility of inorganic MNA for compounds as listed in Table 1. Volume 1 introduces USEPA’s “Tiered Analysis Approach to Site Characterization”; the tiers being four demonstrations or findings to be accomplished when implementing inorganic MNA. Volume 1 provides specific technical guidance on implementing inorganic MNA using the four- tiered approach. Volume 2 expands on the technical guidance to include specific strategies and chemical mechanisms for arsenic, cadmium, chromium, copper, lead, nickel, nitrate, perchlorate, and selenium inorganic MNA. Volume 3 presents the evaluation of MNA for radionuclides. Table 2 presents a simplification of the four tiered approach found in the EPA guidance. CCP inorganic MNA is not considered by USEPA to be a viable remedial alternative for groundwater unless the leachate additions to groundwater have been controlled. Developing a site-specific understanding of the mechanism by which inorganic MNA may attenuate (Tier II) and the capacity of the aquifer to do so (Tier III) is a complex and iterative task. Volume 1 provides the technical basis for inorganic MNA evaluations to accomplish Tier II and III goals. Monitoring of inorganic MNA is necessarily more detailed than for many active remediation strategies and is discussed in Volume 1. Interstate Technology Research Council Guidance In late 2010, ITRC published a decision framework for inorganic MNA. The document contains a summary of the three-volume EPA guidance, expanded to decision framework for determining MNA feasibility. Case studies are included and compared to the EPA four tiered approach. The MNA implementation at the state level was surveyed and reported. The ITRC document summarizes the topic of inorganic MNA, presents an introduction to the science, and presents a strategy for training and implementation of inorganic MNA under state laws and regulations. This guidance is not specific to CCP. As with the USEPA guidance, it is a technical resource on evaluation of the feasibility of inorganic MNA for compounds in Table 1. M. Filardi RSS MNA Tier I & II October 14, 2015 Page 10 of 24 Table 1 Compounds in CCP leachate as compared to MNA guidance EPA CCP EPRI CCP EPA MNA ITRC MNA EPRI MNA Aluminum (Al) X XX Antimony (Sb) XX XX Arsenic (As) XXX XXX X X X Barium (Ba) XXX X Beryllium (Be) X X Boron (B) XX XXX X Cadmium (Cd) XX XX X X Calcium (Ca) X X Chromium (Cr) XXX XXX X X X Cobalt (Co) X X X Copper (Cu) X X X Fluorine (F) X X Iron (Fe) X X Lithium (Li) XX Lead (Pb) X X X X Magnesium (Mg) X Manganese (Mn) X XX Mercury (Hg) X X Molybdenum (Mo) XX XX Nickel (Ni) X X X X Potassium X X Selenium (Se) XXX XXX X Silicon (Si) X Silver (Ag) X X Sodium (Na) X X Strontium (Sr) X X Thallium (Th) XX XX Vanadium (V) X XX Zinc (Zn) X X X pH X X Sulfate (SO4) X XXX X Dissolved Solids X XX Radionuclides X X EPA CCP: X=mentioned, XX= of concern, XXX=exceeds toxicity criteria EPRI CCP: X= evaluated, XX= of concern, XXX=of concern and common Table 2 Summary of US EPA four-tiered feasibility evaluation for inorganic MNA Tier I Source Control Is the contaminant mass in the plume decreasing? Tier II Attenuation Mechanism Is the chemical mechanism well understood? Tier III Attenuation Capacity Is the capacity and permanence of the mechanism sufficient? Tier IV Monitoring & Contingency How will monitoring be conducted? What actions will be taken if monitoring indicates attenuation is lacking? M. Filardi RSS MNA Tier I & II October 14, 2015 Page 11 of 24 TIER I DEMONSTRATION FOR THE RIVERBEND STEAM STATION MNA is a candidate remedial technology for COIs exceeding 2L Standards in groundwater at the RBSS. Attenuation processes reduce the concentration or toxicity of compounds in groundwater primarily by chemical interaction with aquifer solids. When attenuation processes result in a reduction of contaminant concentrations in a timeframe similar to other remedial technologies, attenuation can be an acceptable remedial alternative (USEPA 2007). This section documents that the first milestone in the Four-Tiered MNA process has been achieved; attenuation of certain COIs has been observed. This successful Tier I demonstration is based on the methods contained in EPA guidance (USEPA 2007). The MNA demonstration for certain COIs will now be advanced to Tier II/III. Data were collected on the distribution of COIs in groundwater, porewater, and soil in three dimensions to evaluate if attenuation was occurring. A strong positive correlation between COI concentration in water and solid pairs indicates attenuation (USEPA 2007) and is the first step (Tier I) in evaluation of MNA as a remedial technology. EPA guidance defines Tier I (USEPA, 2007, page 6) as: “Demonstration that the ground-water plume is not expanding and that sorption of the contaminant onto aquifer solids is occurring where immobilization is the predominant attenuation process.” Determination of plume stability or non-expansion generally follows source control. Quantitative analysis of plume stability assuming source control was conducted as part of the numerical modeling effort supporting the CAP. Fate and transport modeling used the site-specific partitioning coefficients developed as part of a laboratory testing program to determine site- specific attenuation capacities (Kd) for COIs at RBSS. Tier I objectives were accomplished by chemical evaluation of collocated aquifer solids and groundwater (i.e., solid/water pairs) as described in the USEPA guidance (USEPA 2007, page 6): “Determination of contaminant sorption onto aquifer solids could be supported through the collection of aquifer cores coincident with the locations of ground-water data collection and analysis of contaminant concentrations on the retrieved aquifer solids.” MNA requires additional testing and data interpretation beyond Tier I to allow reasonable comparison to other remedial alternatives with respect to efficacy, cost, stakeholder acceptance, and time to meet remedial goals. This will be accomplished for COIs by completing Tiers II and III of MNA Demonstration. The groundwater COI’s for RBSS are antimony, arsenic, boron, chromium, cobalt, iron, lead, manganese, sulfate, thallium and TDS. Antimony, cobalt, iron, manganese, and vanadium occur naturally in regional groundwater. Sulfate and TDS are generally not attenuated by reactions with solids, but are reduced in concentration by diffusion, mechanical mixing, or dilution. Here we consider water-solid pairs and site-specific Kd data as evidence for attenuation for arsenic, boron, chromium, selenium, and thallium. Water and solid chemistry data from the comprehensive site assessment (CSA) were used to conduct the Tier I analysis. A single extraction method was used on solids, USEPA SW-846 M. Filardi RSS MNA Tier I & II October 14, 2015 Page 12 of 24 Method 3051A. When a water sample could be matched with more than one solid sample in the screened section of the well, multiple solid-water pairs were created. Non-detection (U values, below detection limit) were not used in the comparison or were substitute values used. J values (presence indicated, below quantification limit.) were used as reported. As reflected in Table 3 a total of 28 solid-water pairs were used for the RBSS Tier I Demonstration. Tier I Analysis Tier I analysis uses two lines of evidence for attenuation. Solid-water pair comparison of COI concentrations are performed, a mutually rising relationship indicating attenuation (USEPA 2007); and, laboratory determination of the solid-water partitioning coefficient or Kd value (USEPA 1999) is used as measure of the propensity of COIs to adsorb to site-specific solids and be attenuated. If a solid-water pair comparison chart is not presented, there was insufficient data for Tier I analysis for that element. Laboratory determination of Kd was performed by the University of North Carolina Charlotte (UNCC) under the supervision of Dr. William Langley on 14 site-specific samples of soil, or partially weathered rock (saprolite) from the transition zone (Table 4). Solid samples were batch equilibrated to measure the adsorption of COIs at varying concentrations. COIs tested were arsenic, boron, cadmium, chromium, molybdenum, selenium, thallium, vanadium. Antimony, cobalt; iron and manganese were not tested. These multiple data points for each COI and sample are evaluated to determine if the observed data can be fit to an adsorption isotherm. If fitting to an isotherm is supported, a Kd is calculated; blanks in Table 4 indicate that a reasonable isotherm was not observed in that test. Batch tests are conducted in duplicate to evaluate error. Table 4 also presents calculated median batch-test Kd values for each COI averaged across all tests, the minimum and maximum values, and the ratio of the maximum to the minimum value as an indication of the variance in the results. Iron and manganese Kd determinations were difficult due to the high concentrations of poorly crystalline iron and manganese oxides and oxyhydroxides present in the Piedmont soils and saprolite. The variability in Kd measurement across the 14 samples is dependent on the COI. Kd values vary from thousands of ml/g to single digits. The variation in the ratios of max/min Kd is not correlated to the calculated medians of Kd. We can use the max/min ratio as a subjective indicator of the potential for COI to have a Kd that is variable across geomedia. The relative strength in binding based on the batch Kd data is: Arsenic > Vanadium > Chromium > Selenium Molybdenum > Cadmium > Thallium > Boron. The potential for a variable Kd across the site is in the order: Chromium > Boron > Thallium > Cadmium > Vanadium > Molybdenum > Arsenic > Selenium. The most variable COIs are all oxy-anions whose sorption is variable with pH conditions. Antimony Antimony attenuation using solid-water pairs was not observed. There were less than three data pairs, preventing analysis at this time. The calculations of a median Kd of 120 ml/g for antimony indicated that it should be moderately to strongly attenuated at the site. The Max/Min ratio of 81. 5 suggests that some variability in Kd across the site is to be expected. Natural attenuation of antimony should be observed and the COI should be carried through to Tier II. Arsenic Arsenic attenuation using solid-water pairs was not observed. There were less than three data pairs, preventing analysis at this time. Kd determinations for arsenic support attenuation with a calculated median Kd of 904 ml/g. Variability in observed Kd should be high with a Max/Min M. Filardi RSS MNA Tier I & II October 14, 2015 Page 13 of 24 ration of 474. Laboratory attenuation of arsenic has been observed, and the COI should be carried through to Tier II. Boron Figure 1 depicts the solid-water pair analysis for boron. Attenuation is not observed in this plot. Calculated site-specific boron Kd values are low (1.7 – 3.1 ml/g) but are consistent with Kd values observed for arsenic uptake in soils (Goldberg 1999). Boron attenuation is often controlled by clay mineralogy. The large amounts of illite clay observed, boron’s favored sorption by illite, and observations of partitioning and attenuation in controlled laboratory conditions suggest that boron attenuation is taking place, albeit in smaller amounts than other COIs. The attenuation of boron has been observed and it should be carried through to a Tier II evaluation. Figure 1 Chromium The solid-water pair concentrations of chromium from Table 3 are plotted in Figure 3. Attenuation is observed in this plot. Calculated site-specific chromium Kd values are high (median of 648 ml/g). It was difficult to fit isotherms for many samples. The attenuation of chromium at Riverbend has been observed and chromium should be carried through to a Tier II evaluation. 0 500 1000 1500 2000 0 20 40 60 Bo r o n , d i s s o l v e d ( µg / L ) Boron (mg/kg) M. Filardi RSS MNA Tier I & II October 14, 2015 Page 14 of 24 Figure 2 Selenium Selenium attenuation using solid-water pairs was not observed. There were less than three data pairs, preventing analysis at this time. The calculations of a median Kd of 573 ml/g for selenium (batch) indicated that it should be moderately to strongly attenuated at the site. The Max/Min ratio of 1457 suggests that high variation in Kd across the site is to be expected. Natural attenuation of selenium should be observed and the COI should be carried through to Tier II on the basis of the high Kd observations in batch tests. Thallium Thallium attenuation using solid-water pairs was not observed. There were less than three data pairs, preventing analysis at this time. The calculations of a median Kd of 118 ml/g for thallium indicated that it should be moderately attenuated at the site. The Max/Min ratio of 20 suggests that little variation in Kd across the site is to be expected. Natural attenuation of thallium should be observed and the COI should be carried through to Tier II on the basis of the moderate Kd observations in batch and column tests. Tier I Findings The groundwater COI’s for RBSS are antimony, arsenic, boron, chromium, cobalt, iron, lead, manganese, sulfate, thallium and TDS. Tier I analysis indicates that arsenic, boron, chromium, selenium and thallium should be carried through to Tier II determinations of mechanism. TIER II DEMONSTRATION FOR THE RIVERBEND STEAM STATION MNA is a candidate remedial technology for the groundwater COIs originating on RBSS property. A strong positive correlation between COI concentrations in solid-water pairs indicates attenuation. Correlation was demonstrated as documented in the Tier I demonstration. This section covers Tier II demonstration of mechanism and rate for COI attenuation. Attenuation processes reduce the concentration or toxicity of compounds in groundwater primarily by advection, dispersion, and chemical interaction with aquifer solids. When attenuation processes result in a reduction of contaminant concentrations in a timeframe similar to other remedial technologies, attenuation can be an acceptable remedial alternative (USEPA 2007). In completion, the Tier I, II, and III MNA demonstrations indicate that MNA is operable on a timescale comparable to active remedial technologies (e.g., pump and treat) that have been R² = 0.6601 0 0.5 1 1.5 2 2.5 3 3.5 0 50 100 150 200 Ch r o m i u m , d i s s o l v e d ( µg / L ) Chromium (mg/kg) M. Filardi RSS MNA Tier I & II October 14, 2015 Page 15 of 24 screened for the RBSS groundwater portion of the CAP. MNA is being considered in the CAP for RBSS groundwater as a remedial technology. Data were collected on the distribution of COI in groundwater and geomedia in three dimensions to evaluate if attenuation was occurring. The first use of the data was to evaluate the premise of COI attenuation (Tier I) using a subset of results from the approved data collection program. The groundwater COIs for RBSS are antimony, arsenic, boron, chromium, cobalt, iron, lead, manganese, sulfate, thallium and TDS. Tier I analysis indicates that arsenic, boron, chromium, selenium, and thallium should be carried through to Tier II determinations of mechanism. The process was intended to be flexible based on findings (the observational approach) or in response to agency comments or new scientific opportunities that may arise. USEPA guidance defines Tier II (USEPA 2007, page 6) as; “Determination of the mechanism and rate of the attenuation process.” and, continues on page 7 with further clarification on rate: “The objective under Tier II analysis would be to eliminate sites where further analysis shows that attenuation rates are insufficient for attaining cleanup objectives established for the site within a timeframe that is reasonable compared to other remedial alternatives.” MNA tasks require an observational approach and the reliance on predictive models increases over the duration of the project (Tier I to III). The MNA demonstration process results in increasing levels of confidence in the reliability of MNA as a remedial solution. There are two technical paths that contribute to the same site conceptual model. There is a technical path primarily related to calculations and modeling of groundwater flow, contaminant transport, and how fate and transport of the COIs is affected by remedial alternatives, attenuation, or no action. Supporting those calculations are geochemical activities primarily related to technical evaluation of COI attenuation capacity, permanence, and remedial effectiveness over time. The technical activities proposed for the CSA relied on our pre-Tier I best understanding of COI attenuation processes in RBSS groundwater and were subject to change based on observation, study objectives, or Tier goal attainment by alternate methods. Site Conceptual Model for Attenuation Lithologic Controls on Water Quality The U. S. Geological Survey (USGS) and North Carolina Department of Environment and Natural Resources recently completed regional studies (USGS 2009; USGS 2013) of the groundwater chemical quality in the Blue Ridge and Piedmont Provinces. RBSS is located in the Felsic Intrusive Geozone of USGS 2009, a major province occupying about 20. 6% of the surface area of the Blue Ridge and Piedmont Provinces in North Carolina. The Felsic Intrusive in the vicinity of the RBSS is composed of metamorphosed quartz diorite and tonalite. Bedrock groundwater in the Felsic Intrusive Geozone varies from calcium-sodium bicarbonate to a calcium-magnesium bicarbonate type with the calcium-magnesium composition dominant. A notable consideration from the two recent USGS papers is that is it possible to roughly rank the potential for natural background occurrence of heavy metals (Sb, As, Ba, Co, Cr, V, and Zn as examples) by rock type as Mafic Crystalline > Felsic Crystalline > Siliciclastic > Carbonate > Recent Alluvium. Mafic material (igneous-origin rock that contains larger percentages of Fe and Mn silicates than quartz or feldspars) has a link to background levels of heavy metals at Riverbend. Mafic minerals have the same connection. The Riverbend surface geologic mapping indicates a granodiorite (5-25% mafic minerals) to tonalite composition (10-40% mafic minerals). M. Filardi RSS MNA Tier I & II October 14, 2015 Page 16 of 24 Surface geologic mapping indicates the near proximity of mafic rock types (gabbro) that are also expected to influence subsurface trace metal occurrence. Mineralogical Controls on Attenuation COIs are adsorbed to organic matter, oxide minerals, and clay minerals. These are the predominant sources and sinks for COIs in groundwater, soil, or sediment. All of these solids are present in subsurface materials at RBSS. COIs adsorbed on these solids equilibrates with solution COI concentrations. Some COIs adsorbed to clays migrates from clay edges to clay inter-layer spaces. The strength of partitioning of COIs from groundwater to solids ranges from loosely bound by ion exchange reactions, to irreversibly bound in clay structures or chemical precipitates. By example, through the method of selective or partial extractions (Tessier et al. 1979) mineral phases and sediment fractions associated with higher concentrations of COIs through attenuation can be identified. By mineral or size separation and chemical extraction the solids that contribute to attenuation can be classified and segregated for degree of effect. The degree of COI partitioning (Kd and µ) over a range of concentrations can be determined by testing co- located solids and groundwater, or the examination of COI concentration over time in groundwater at a fixed location for the purpose of solving the ADE. This and other information is used to calculate the amount of COI removal, and time to achievement of remedial objectives using a MNA remedial approach. Role of Site Lithology Typically geologic materials are sampled offsite and away from potential contamination to establish baseline values for comparison to those analyzed from study sites. In lieu of detailed offsite sampling, published values for average crustal abundances can be utilized to evaluate whether indigenous materials contribute to COI concentrations in an area. For this report, crustal abundance values from Wedepohl (1978) are used for comparison to samples analyzed at the Riverbend site. Table 5 presents average crustal abundances for comparison. Site Geologic Formation Influence on COI Occurrence Tonalite - Tonalite is the dominant lithology underlying the site. Tonalites are a relatively common basement rock in many Proterozoic and accretionary terrains with a composition similar to quartz diorite but with elevated calcium. The composition of typical tonalities is quartz, plagioclase feldspar, amphibole, with accessory sphene, magnetite, and orthoclase. No whole rock petrology for rocks specifically at the site is available at this time. The average crustal abundances as proxy for uncontaminated tonalite at Riverbend are presented in Table 5 for the COI inventory. Some whole rock chemical analyses show abundances of Sb, As, B, Cr, Co, Se, Tl, and V significantly above crustal abundances for average tonalites. Conceivably, weathering of the tonalite could contribute to elevated values for COI because of residual soil formation, but the observed concentrations suggest COI contributions outside of weathering. Soil Development Influence on COI Occurrence Soil formation typically results in the loss of common soluble cations and the accumulation of quartz and clay. Feldspars are hydrolyzed to clays. Increase in the concentrations of COI during the weathering and soil development on the lithology noted above is significant (Table 5). Arsenic values are significantly elevated in boring IDs C-2S, AB-7S, AB-8S proximal to disposal areas, and remain slightly elevated in all soil samples (RBSS CSA Report, Table 6-3). The other COI elements (V, Cr, Co, Se, and Sb) have similar results. The increasing abundance of clay during the natural weathering process can conceivably result in an increase in COI content with M. Filardi RSS MNA Tier I & II October 14, 2015 Page 17 of 24 time. However, the values derived during the CSA suggest the concentrations are being introduced from the ash. Mechanisms for Natural Attenuation at RBSS Active precipitation of COI secondary minerals at the site appears unlikely, with exception of iron and manganese oxides and oxyhydroxides. The very low abundances of the other elements suggest saturation of these elements and precipitation as secondary oxides, arsenates, or carbonates is not attained. The attenuation of chemical contamination by reaction with existing natural materials may play a significant role at the site. The abundant clay content of the soils and host rock lithologies suggests much of the COI concentrations in the ash basin and ash storage may be attenuated by these materials. Harder (1970) and Perry (1972) showed in pioneering studies that boron is adsorbed and incorporated into illite and chloritic clays respectively. The low content of boron in most rivers is attributed to the same processes. Thallium is also documented to be adsorbed to clays, especially potassium and rubidium types, as well as to iron and manganese oxides. Vanadium is also adsorbed and incorporated into clay structures and oxide coatings (Butler, 1953). Krauskopf (1956) suggests vanadium displays a preference for adsorption in decreasing order from Fe oxides > Mn oxides > montmorillonite > organic matter. In addition to the reported abundance of clays at the site, there is significant potential for more clay to be present as an alteration product on the surfaces of the abundant feldspars reported in the soils and rocks. Adsorption to iron oxides and hydroxides has long been demonstrated for As, B, Ba, Cd, Co, Cr, Co, Fe(II), Hg, Mn, Ni, Pb, Sb, Se, SO4-2, V and Zn (Dzombak and Morel, 1990). Soil chemistry results show abundant Fe2O3 and MnO values in soils from the site (Table 6-2, RBSS CSA Report) and a strong potential for adsorption. High Al2O3 (aluminum hydroxide) content is an indicator of clay minerals and also can indicate the presence common soil-forming aluminum oxyhydroxides gibbsite, boehmite and diaspore. Adsorption to most-common gibbsite has been demonstrated for As, B, Cd, Co, Cr, Co, Fe(II), Hg, Mn, Ni, Pb, Se, SO4-2, and Zn (Karamalidis and Dzombak 2010). No amorphous phase content is currently reported from the CSA mineralogical studies; however, high iron, manganese and aluminum content in Method 3051A extractions presage quantification in reporting to follow. When quantified, amorphous Fe-Mn-Al oxide-hydroxides or on amorphous organic materials provide significant potential for natural attenuation of the COI. Amorphous phase content will be determined on soil samples in order to consider adsorption potential. Tier II Discussion & Conclusions Following successful completion of a Tier I demonstration that arsenic, boron, chromium, selenium and thallium are attenuating in groundwater at RBSS, a conceptual model for COI attenuation involving reversible and irreversible interaction with clay minerals, metal oxides, and organic matter is proposed. A Tier II demonstration based on that conceptual model was partially executed. The findings follow: 1. The sampling obtained geomedia representative of the material that the COI plume will traverse. 2. Clay minerals and Fe-Mn-Al oxides were found in all samples. Organic matter is probably not a significant sink for COI at RBSS. 3. Chemical extractions identified that COIs were concentrated in samples exposed to groundwater containing higher concentrations of COIs, validating the attenuation conceptual model. M. Filardi RSS MNA Tier I & II October 14, 2015 Page 18 of 24 4. Chemical extractions were used to determine a probable range of Kd values that suggest attenuation is taking place for arsenic, boron, chromium, selenium and thallium. 5. Additional data collection is necessary to complete the Tier II assessment with respect to specific attenuation mechanisms for each COI, and quantification of the magnitude of that attenuation by specific geomedia to support numerical modeling. Data Gaps and Recommendations Tier II demonstrations address the mechanism and rate of attenuation. The mineralogical and physical characterization of solids with respect to type and adsorbed COI concentrations leads to predictions of COI attenuation expected in the future. Selective extraction (wet chemical analysis of solids), hydrostratigraphy, and broad water quality data collection provide part of the information needed to compare MNA to other remedial alternatives in the Tier II evaluation. The data collected to date has provided critical information on using MNA at RBSS and supports Tier II evaluations. To continue to Tier II and advance to Tier III additional data collection is necessary. The data evaluation and analysis process for MNA is complex and iterative. The following table proposes some specific geochemical analysis approaches to completing Tier II evaluation of mechanism and support Tier III determinations of attenuation capacity of COIs. Quantification of the crystalline and amorphous hydrous metal oxides of aluminum, iron and manganese (HAO, HFO, and HMO respectively) by partial extraction is proceeding with the UNCC effort on Kd quantification and will be very useful when available. Many of the other methods can be performed on archived materials. It was noted in the Tier I evaluation that additional solid-water pair data is needed from areas of higher concentration to improve the fit to these field adsorption isotherms (Field Kd). This will provide opportunity for additional data collection on samples that may be unavailable in archive. Field testing to determine dispersivity would also require redeployment to the site for testing activities. M. Filardi RSS MNA Tier I & II October 14, 2015 Page 19 of 24 Analysis Group Action Method Purpose Wet Chemical Analysis Partial Extractions of Sediments Chemical extractions with differing degrees of action. Chemical analysis of extracts provides information on what solid materials the COI is associated with and the COI concentration on those materials. Total Analysis Complete digestion of sediment. Method 3052. HNO3 + HF digestion provides a total COI analysis of solid matrix to compare with partial extractions. More sensitive for trace metals than oxide analysis. Physical Analysis Field Sedimentology Field logging and classification by geologist. Record used for field selection of samples for preservation and analysis, selection of screen elevations for well installation, and correlation of stratigraphy between borings. Office Petrography Mineralogy by visual examination by geologist. Bulk samples and thin section. Optical and electron microscopy. Knowledge of the aquifer mineralogy and relative percentages of minerals is needed to evaluate observed groundwater chemistry and potential for COI attenuation by reaction with aquifer solids. XRD Reitveld Mineral identification by X-Ray Diffraction. Used to identify clay minerals and confirm/ support visual determinations of mineralogy. EDS Energy Dispersive Spectroscopy Provides information on the location and relative concentration of trace elements on mineral surfaces and polished sections Hydrologic Parameters Tracer tests and break-through curve analysis. Site specific determinations of dispersivity are needed Surface Chemistry Surface area of sediments. Determination of reactive area of the aquifer material. Needed to scale the results of chemical determination. M. Filardi RSS MNA Tier I & II October 14, 2015 Page 20 of 24 REFERENCES AND CITATIONS Butler, J. R., 1953. The geochemistry and mineralogy of rock weathering. I. The Lizard area: Geochimica et Cosmochimica Acta, v. 4, p. 157. Dzombak, D. A. and Morel, F. M. M., 1990. Surface Complexation Modeling: Hydrous Ferric Oxide. John Wylie and Sons, New York. EPRI, 2005. Chemical Constituents in Coal Combustion Product Leachate: Boron, EPRI, Palo Alto, CA. 2005. 1005258. EPRI, 2006. Groundwater Remediation of Inorganic Constituents at Coal Combustion Product Management Sites: Overview of Technologies, Focusing on Permeable Reactive Barriers. EPRI, Palo Alto, CA: 2006. 1012584. Goldberg et al., 2000. Goldberg, S., Lesch, S. M. , and D. L. Suarez. Predicting Boron Adsorption by Soils Using Soil Chemical Parameters in the Constant Capacitance Model. Soil Sci. Soc. Am. J., 64(5): 1356-1363, 2000. Harder, H., 1970. Boron content of sediments as a tool in facies analysis: Sedimentary Geology, v. 4, p. 153. Herbelin and Westall, 1999. A. L. Herbelin and J. C. Westall. FITEQL - A Computer Program for Determination of Chemical Equilibrium Constants from Experimental Data. Report 99-01, Department of Chemistry, Oregon State University, Corvallis, OR 97331. 1999. ITRC, 2010. A Decision Framework for Applying Monitored Natural Attenuation Processes to Metals and Radionuclides in Groundwater. APMR-1. 2010. Interstate Technology & Regulatory Council Karamalidis, A. K., and Dzombak, D. A., 2010. Surface Complexation Modeling: Gibbsite. John Wylie and Sons, New York. Krauskopf, K., 1955. Sedimentary deposits of rare metals: Economic Geology, v. 50, p. 411. Perry, E. A. Jr., 1972. Diagenesis and the validity of the boron paleosalinity technique: American Journal of Science, v. 272, p. 150. Tessier, A., Campbell, P. G. C., and M. Bisson, 1979. Sequential Extraction Procedure for the Speciation of Particulate Trace Metals. Analytical Chem., 51(7)844-850. Thayer, P. A., Kirstein, D. S., and Ingram, R. L., 1970. Stratigraphy, sedimentology, and economic geology of Riverbend Basin, North Carolina: Carolina Geological Society Guidebook, 29p. M. Filardi RSS MNA Tier I & II October 14, 2015 Page 21 of 24 Toride, N., F.J. Leij, and M.Th. van Genuchten, 1999. The CXTFIT Code for Estimating Transport Parameters from Laboratory or Field Tracer Experiments: Version 2. 1. US Department of Agriculture, Agricultural Research Service, US Salinity Laboratory, April, 1999. Research Report No. 137 USEPA, 1999. Understanding Variation in Partition Coefficient, Kd, Values. Volume I: The Kd Model Of Measurement, And Application Of Chemical Reaction Codes. EPA 402-R-99- 004A. United States Environmental Protection Agency, Office of Air and Radiation, August 1999. USEPA, 2007. Monitored Natural Attenuation of Inorganic Contaminants in Groundwater, Vol. 1: Technical Basis for Assessment. 2007. US EPA, EPA/600/R-07/139. USGS, 2009. Harden, S L., M J. Chapman, and D A. Harned. Characterization of Groundwater Quality Based on Regional Geologic Setting in the Piedmont and Blue Ridge Physiographic Provinces, North Carolina. U. S. Geological Survey Scientific Investigations Report 2009–5149. USGS, 2013. Chapman, M. J., C. A. Cravotta III, Z. Szabo, and B. D. Lindsey. Naturally Occurring Contaminants in the Piedmont and Blue Ridge Crystalline-Rock Aquifers and Piedmont Early Mesozoic Basin Siliciclastic-Rock Aquifers, Eastern United States. 1994– 2008. U. S. Geological Survey Scientific Investigations Report 2013–5072. This page intentionally left blank M. Filardi RSS MNA Tier I & II October 12, 2015 Page 22 of 24 Table 3 Solid-water pairs used for Tier I demonstration - RBSS GW Soil Analyte Aluminum Aluminum Antimony Antimony Arsenic Arsenic Barium Barium Boron Boron Beryllium Beryllium Cadmium Cadmium Chromium Chromium Cobalt Cobalt Copper Copper Iron Iron Lead Lead Manganese Manganese Mercury Mercury Molybdenu m Molybdenu m Nickel Nickel Selenium Selenium Strontium Strontium Thallium Thallium Vanadium Vanadium Zinc Zinc mg/kg µg/L mg/kg µg/L mg/kg µg/L mg/kg µg/L mg/kg µg/L mg/kg µg/L mg/kg µg/L mg/kg µg/L mg/kg µg/L mg/kg µg/L mg/kg µg/L mg/kg µg/L mg/kg µg/L mg/kg µg/L mg/kg µg/L mg/kg µg/L mg/kg µg/L mg/kg µg/L mg/kg µg/L mg/kg µg/L mg/kg µg/L AB-1S 8660 2.8 0.14 25.8 21 11.8 0.3 0.028 35.8 0.48 32.7 16.2 37.6 0.8 71900 0.077 563 260 0.033 0.12 14.6 2.9 0.78 11 0.018 92.6 18.5 26 AB-8S 30900 4.6 9.2 300 130 51.7 240 1 0.17 0.13 186 3.1 48.4 0.56 71.2 0.3 57500 1080 140 0.013 70.9 1.8 9.2 1.2 6.7 510 140 0.37 52.7 84 AB-8S 31300 523 130 50 240 1.4 0.17 0.13 177 3.1 70.1 0.56 82.3 0.3 50200 1660 140 0.0087 75.1 1.8 1.2 16.3 510 145 0.37 49.1 84 AS-1S 14000 0.23 2.4 0.95 92.1 95 13.5 1900 0.31 0.031 7.6 0.66 5.1 5.3 14.5 1.5 14400 3.3 0.083 170 1700 0.019 132 4 2.6 8.9 2600 0.065 35.4 0.48 28.7 11 AS-1S 7740 0.23 0.95 82.6 95 6.7 1900 0.3 0.031 3.4 0.66 5.2 5.3 4.2 1.5 7540 1.8 0.083 194 1700 132 3.1 2.6 10 2600 0.065 17.2 0.48 24.7 11 BG-1S 13100 0.23 144 20 0.28 3.8 0.43 12.1 1.9 16.1 0.66 20000 518 240 5.3 1.8 4.7 2.3 14 0.016 59.7 30 14 BG-1S 10900 0.23 189 20 0.29 3.1 0.43 13.9 1.9 16 0.66 19300 6.3 921 240 5.8 1.8 3.1 14 0.016 53.3 31.1 14 BG-1S 9440 0.23 209 20 0.46 4.6 0.43 9.9 1.9 10.3 0.66 13500 817 240 4.7 1.8 4.6 14 0.016 38 33.6 14 BG-2S 11100 0.25 0.14 132 68 0.63 0.041 4.5 0.41 8.5 0.69 4.3 0.37 16000 3.9 0.068 689 360 0.012 0.25 3.9 1.2 8 92 0.031 40.8 38.7 15 BG-2S 15400 0.25 0.14 226 68 1.4 0.041 20.6 0.41 13.8 0.69 20.3 0.37 20600 0.068 1440 360 0.013 0.25 8.5 1.2 16.7 92 0.031 58.8 55.5 15 BG-3S 26700 1.4 276 33 0.84 0.032 6.1 0.33 11 0.54 10.4 1 18100 3.7 0.1 626 240 0.16 7.6 0.69 16.2 67 45.4 0.7 45.8 10 BG-3S 21600 1.4 282 33 0.98 0.032 6.7 0.33 12.1 0.54 93.3 1 23500 3.3 0.1 724 240 0.16 6.9 0.69 22.2 67 50.6 0.7 59.8 10 BG-3S 15900 1.4 159 33 0.63 0.032 5.7 0.33 8 0.54 12.6 1 19400 0.1 542 240 0.16 5.3 0.69 17.3 67 41.6 0.7 43.2 10 C-2S 9940 1600 3.7 0.86 10.2 35 170 0.15 2.1 0.21 9.9 0.51 3.1 42.5 27 5.9 39000 4.8 0.17 20.5 580 0.57 25.7 1.5 3.3 520 0.095 38.1 6.9 94 C-2S 5460 1600 0.86 23.5 35 170 0.086 2.1 0.21 6.4 0.51 14 42.5 7.2 5.9 12800 3.5 0.17 237 580 1.6 25.7 1.5 1.2 520 0.095 40.1 13.3 94 GWA-1S 25200 0.22 164 33 1 92.8 1.1 34.8 0.23 27.9 1.8 31700 0.12 1090 30 0.24 34 1 5.7 35 88.2 0.86 92.4 29 GWA-1S 39600 0.22 288 33 0.96 110 1.1 38.9 0.23 232 1.8 55200 0.12 2940 30 0.24 49.7 1 14.5 35 133 0.86 183 29 GWA-2S 17800 0.14 151 43 16.2 440 0.26 0.032 1.5 1 15 16.8 1.2 24100 0.066 713 56 0.1 9.2 0.45 32.8 180 89.4 1.1 53.2 5.8 GWA-2S 24900 0.14 253 43 21.3 440 0.24 0.032 3.6 1 12.6 25.8 1.2 32300 0.066 1120 56 0.1 12.3 0.45 35.1 180 126 1.1 98.8 5.8 GWA-5S 4390 32.7 17 8 0.32 17.3 4.6 8.2 1.4 16000 575 200 0.13 2 3.8 9.5 34.9 6.5 9.8 GWA-6S 13200 0.17 176 39 16.2 34 0.3 0.24 1.2 0.52 10.5 25.3 28.1 6.2 18200 2.6 523 1100 2.4 4.6 21.8 12 0.041 55.5 56.7 30 GWA-7S 14600 212 21 13 2.3 34.7 0.31 47.9 0.31 23200 813 16 14 1.2 26.5 46 73.8 1 68.4 5.2 GWA-8S 8480 0.18 18.1 120 240 0.25 0.073 131 1.5 87 2 49 5.8 54200 0.63 959 760 0.063 37.2 1.1 0.27 1.9 16 80.1 8.6 32 GWA-10S 14000 57.6 28 170 0.32 0.094 0.038 12.3 1 24.4 10.3 18.6 1.3 30100 32 5.9 1840 240 0.063 3 1.9 2.4 64 59.3 0.39 16.5 13 GWA-20S 6540 12.4 23 30 0.33 0.036 4.4 0.38 3 2.7 0.52 9880 34 0.058 131 150 2.3 7.3 24 0.04 19 7.1 18 GWA-20S 12700 63.3 23 30 0.24 0.036 7.4 0.38 5 3 3.2 0.52 16300 34 0.058 184 150 4.3 7.3 24 0.04 31.1 28.7 18 GWA-23S 3140 140 27.6 38 0.038 1.6 0.54 6.9 3.1 2 0.95 5230 5 0.094 336 81 0.01 1.1 2.3 6.5 9.9 3.8 14 GWA-23S 4560 140 18.4 38 0.038 2.2 0.54 18.6 3.1 2.8 0.95 7710 0.094 210 81 0.01 1.6 2.3 6.5 14.2 7.1 14 M. Filardi RSS MNA Tier I & II October 14, 2015 Page 23 of 24 Table 4 Kd data used for Tier I evaluation - RBSS AB-4D AB-4D AB-2S AB-2S AB-6S AB-6S AB-7S AB-7S GWA-1BRU GWA-1BRUGWA-7D GWA-7D GWA-8D GWA-8D GWA-1S GWA-1S GWA-2S GWA-2S GWA-4S GWA-4S GWA-5S GWA-5S GWA-6S GWA-6S GWA-7S GWA-7S GWA-10S GWA-10S 55-60 FT 55-60 FT 90 FT 90 FT 73-75 FT 73-75 FT 20-25 FT 20-25 FT 78-79 FT 78-79 FT 102-103.5 FT 102-103.5 F19-20 FT 19-20 FT 42-47 FT 42-47 FT 48-52 FT 48-52 FT 20-25 FT 20-25 FT 72-74 FT 72-74 FT 55-60 FT 55-60 FT 22-23 FT 22-23 FT 21-23 FT 21-23 FT Arsenic 904 874 471 430 6 18 19 2044 2148 471 493 306 274 1922 1652 2254 2333 1773 1915 2890 2892 Boron 1.7 2.0 2.0 2.2 3.0 2.3 2.5 3.1 1.9 1.8 Cadmium 1555 1470 111 114 72 79 1555 1470 291 1256 1205 325 321 965 1000 111 114 119 116 46 38 46 46 1075 1261 176 187 Chromium 608 689 Molybdenum 18 16 477 482 317 342 218 273 2384 2408 266 614 671 Selenium 145 124 373 380 573 589 2 1992 2185 1867 1898 558 440 1369 1399 Thallium 541 513 110 118 74 82 112 114 255 319 93 83 256 256 692 696 110 118 132 117 59 54 55 55 1019 1076 196 201 Vanadium 861 782 1798 1800 861 782 13 14 790 961 648 534 1072 1151 Median Kd Min Max Max/min n Arsenic 904 6 2892 474 21 Boron 2.1 1.7 3.1 1.8 10 Cadmium 187 38 1555 41 27 Chromium 648 608 689 1 2 Molybdenum 342 16 2408 147 13 Selenium 573 2 2185 1457 15 Thallium 118 54 1076 20 28 Vanadium 825 13 1800 137 14 M. Filardi RSS MNA Tier I & II October 12, 2015 Page 24 of 24 Table 5 Average trace metal composition for tonalite at the Riverbend Steam Station (ppm) B V Cr Co As Se Sb Tonalite (Wedepohl 1978) 14 99 18 7 1. 4 0. 1 0. 2 Soil (max) 500 400 48 96 8 5 Whole Rock (max) 800 500 43 4 1 4 Page 1 of 1 Summary of Statistically Derived Kd Values – Riverbend Steam Station Variable Num Obs # Missing Minimum Maximum Mean SD SEM MAD/0.675 Skewness Kurtosis CV Arsenic 14 53 251 823.6 519.5 173 46.25 167 0.378 -0.653 0.333 Boron 19 62 1.6 13.1 3.726 2.927 0.672 0.741 2.372 5.658 0.786 Chromium 1 81 1414 1414 1414 N/A N/A 0 N/A N/A N/A Cadmium 41 43 42.4 11934 1510 2467 385.3 374.4 2.726 8.411 1.634 Manganese 28 54 1.6 24011 1287 4889 924 40.1 4.275 18.87 3.798 Molybdenum 18 63 11.7 289.6 61.73 89.64 21.13 5.56 1.915 2.506 1.452 Selenium 22 59 35.1 731.3 195.1 191.3 40.79 63.6 2.066 3.792 0.98 Thallium 48 36 42.6 1104 314.1 302.6 43.68 169.2 1.189 0.416 0.963 Vanadium 4 62 742.8 2192 1211 663.4 331.7 199.2 1.835 3.479 0.548 Note: Prepared by Geochemical, LLC based on UNCC laboratory results.