Developing a FRAMES-Supported Environmental Fate Simulator A Computational

Developing a FRAMES-Supported Environmental Fate Simulator: A Computational System for Estimating Environmental Concentrations of Organic Contaminants Presented to the EPA Exposure Science Community of Practice Feb. 9, 2010 Eric J. Weber US Environmental Protection Agency National Exposure Research Laboratory Ecosystems Research Division Athens, GA 1

General Outline for Today’s Presentation • Address the Primary Questions – What is it? – Why do we need it? – Why now? • Development of the Underlying Process Science • Development and Application of the Model/Software Technology • The Integration of this Knowledge for Conducting Spatially-Explicit Risk Assessments 2

The Source-to-Outcome Continuum Source/Stressor Formation EF&T Models Effect/Outcome Environmental Conc. Exposure Models Biological Event External Dose Target Dose Systems Models BBDR Models PBPK Models One of the primary goals of the Computational Toxicology Research Program is to improve the linkages across the source-to-outcome continuum 3

Why Now? Targeted ORD Research Programs The underlying process science and data is available for simulating chemical transformations -Expo. Cast. TM: Providing an overarching framework for the science required to characterize biologically-relevant exposure in support of the computational toxicology program. -Managing Chemical Risk (Po. BNS): Providing the tools and models for prioritizing chemicals for exposure and effects testing -SP 2 – LTG 2: Developing the data and models for spatially-explicit risk assessments -Hydrolysis -Reductive Transformations Environmental Fate Simulator The modeling/software technology is available -FRAMES -D 4 EM Do. D support through the SERDP program - Physico-chemical properties processor - Reaction pathway simulator 4

Desired Capabilities of the Frames-based Environmental Fate Simulator • Provide access to measured and calculated physico-chemical properties – A growing realization that measured data does not necessarily equate to good data • Provide access to spatially-explicit environmental characterization and source data – A need for the capability to conduct spatially-explicit risk assessments • Provide rate constants for model input and reactivity based binning • Provide dominant transformation pathways and products as a function of environmental conditions • Provide seamless input to chemical exposure models • Conduct uncertainty and sensitivity analyses 5

What is the FRAMES-Supported Environmental Fate Simulator? 1 st Generation: – Physico-chemical properties processor: A tool for accessing computational tools (e. g. , SPARC and EPI Suite) and web accessible data bases (D 4 EM) of measured data to provide the physico-chemical data required for predicting chemical F&T T T – Physico-chemical properties database: A depository for the h calculated and measured data accessed through the Web – Reaction pathway simulator: Based on functional group analysis and knowledge of the environmental system of interest, will provide the transformation products and rates for reductive transformation and hydrolysis 2 nd Generation: – Provide for the seamless parameterization of EF&T models that estimate the environmental concentration (EC) of organic chemicals (e. g. , WASP, EXAMS, PRZM, BASINS, HSPF, MMSOILS, 3 MRA, MULTIMED) 6

Conceptual Design of the Environmental Fate Simulator Environmental Fate Simulation Inputs to EFS User specified Parent Compound Reaction Pathway Simulator Functional Group Analysis Reaction Pathways and Products Outputs of EFS Summary of Relevant Compounds: Parent and Reaction Products Reaction Rate Estimation Environmental Scenario Generic OR Physico-Chemical Properties Processor Calculated Data - SPARC and EPI Suite Physico-Chemical Properties Transformation Rate Constants Measured data (D 4 EM*) User specified Parameritization of the Environmental Scenario Explicit Uncertainty Assessment (D 4 EM) *DFEM – Data for Environmental Modeling 7

Development of the EFS requires: - Knowledge of the current models, data needs and exposure scenarios used by the Program Offices - Knowledge of the processes controlling chemical F&T - The capturing of these processes in mathematical expressions and model code - Software engineering to construct the RPS, provide the linkages to available calculators and data bases of measured data, and to provide for the seamless parameritization of EF&T models 8

An example of a generic scenario is the standard pond scenario used by OPP for pesticide risk assessment A single rain event causes pesticide runoff from a 10 hectare agricultural to a one hectare, 20, 000 cubic meter volume, 2 m deep water body. First Tier Screening Level: GENEEC requires Kd and degradation rate by summing rate constants for aerobic metabolism, abiotic hydrolysis and direct photolysis 9

Developing the Underlying Process Science What do we need to know to predict the reaction pathways and rates for reductive transformations? 1) The functional groups that are susceptible to reductive transformations 2) The molecular parameters describing the “willingness” of chemicals to accept electrons 3) The predominant chemical reductants in natural systems (i. e. , the source of electrons) 4) Pathways for electron transfer 5) Readily measureable indicators of reactivity in natural systems 10

Functional Groups that are Susceptible to Reduction 11

Functional Groups that are Susceptible to Reduction (Cont’d) 12

Process Elucidation in Anaerobic Sediments Conclusions: - Nitroaromatic reduction is facile process in anaerobic sediments - Soluble Fe(II) is a good predictor of reactivity - One-electron reduction potentials are good molecular descriptors for 13 predicting activity

Tools for Calculating Physico-chemical Properties SPARC: SPARC Performs Automated Reasoning in Chemistry SPARC calculates how the X and Y: substituents modify the reactivity of the NO 2 group Prediction of Ionization Constants for 187 Pharmaceuticals A "toolbox" of mechanistic perturbation models calibrated on measured data - Resonance on light absorption spectra - Electrostatic models on ionization equilibrium constants - Solvation models (e. g. , dispersion, induction, H-bonding, dipole-dipole) on vapor pressure, solubility, Henry’s law constants and GC RTs 14

Developing the Modeling/Software Technology Conceptual Multimedia Model The Research Question: How can EPA conduct multi-media, multireceptor, and multi-pathway risk analyses at the 15 national level

FRAMES: Framework for Risk Analysis in Multi-Media Environmental Systems A software system that facilitates the linking and execution of individual models Connections between modules are checked for validity by the system 16

FRAMES: Framework for Risk Analysis of Multi-Media Environmental Systems 17

D 4 EM: Data for Environmental Fate Modeling D 4 EM is a set of reusable components used to automate the retrieval and processing of data for use in executing environmental models Services Provided by D 4 EM: • Data retrieval • Statistical and geoprocessing operations • Data visualization • Model input formatting • Metadata generation 18

Examples of Use Cases: A number of operations performed in concert to accomplish a specific task Examples include: • Downloading data • File formatting • Geo-operations • Logging metadata 19

Transfer Data from D 4 EM Data Store to Modeling System D 4 EM Data Store Unit Definition Processor 20

The Integration of the Process Science and Modeling Technology Allows for Spatially-Explicit Risk Assessments Are public supply wells a source for human exposure to the fungicide pentachloronitrobenzene (PCNB) or its transformation products in Athens-Clarke County? PCNB What information is required to answer this question? • The EF&T processes controlling the reactive transport of PCNB in aquifers • The physico-chemical properties required to simulate these processes • The redox conditions of the aquifers • The location of public supply wells relative to the sources of PCNB and human populations/activities 21

Conducting Spatially-Explicit Risk Assessments Aquifer redox conditions across the US (USGS) Dominant EF&T processes: • Sorption predicted by Koc (Kow values predicted by SPARC and %OC values available in USGS databases) • Nitroaromatic reduction rates predicted by E 1 values (SPARC) and soluble Fe(II) (USGS) Exposure information: • Location of the wells relative to population centers is available The study of public-supply well vulnerability is one of five national priority topics being addressed by the National Water-Quality Assessment (NAWQA) Program 22

Contributors (Laboratory-based Studies) Dalizza Colón (Fed) • Reactivity of iron oxides • QSAR development for reduction of NACs, intermediates, and covalent binding of aromatic amines • Effect of DOM on reduction rates of NACs Rebecca Adams (TAI) • Azo dye reduction in sediments (Disperse Blue 79) Mike Elovitz (NRC) • NAC reduction in sediments (TNT) David Spidle (AI) • Covalent binding of aromatic amines with DOM Kevin Thorn (USGS) • Characterization of with DOM of aromatic amine binding sites in DOM by N 15 NMR John Barnett (EPA) • Analytical support Jean Smolen (NRC), Paul Tratnyek (OGI) • Application of a molecular probe for distinguishing pathways for electron transfer in sediments (abiotic vs. enzymatic) Lisa Hoferkamp (NRC) • Identifying dominant chemical reductants in sediments as a function of redox zonation for NACs Rupert Simon (NRC) • Column studies of sediments (redox zonation) Caroline Stevens (Fed) • Incorporating column kinetic results into a reactive transport model John Kenneke (EPA post doc) • Identifying dominant chemical reductants as a function of redox zonation for halogenated methanes and ethanes in sediments • Identifying molecular descriptors for reduction rates of halogenated methanes and ethanes Said Hilal (Fed), Butch Carreira (UGA) • Development of SPARC calculators for molecular descriptors Judy Zhang (NRC post doc) • Elucidation of pathway for DOM as an electron transfer mediator in sediments • Identification of readily measureable indicators of reactivity for chemical reductants in 23 sediments

Contributors to the Conceptual Design of the EFS Dalizza Colón Susan Richardson Wayne Garrison Caroline Stevens Said Hilal John Washington Jack Jones Jim Weaver Gerry Laniak Gene Whelan Rajbir Parmar Kurt Wolfe 24