Formatted Title

Interactive Dashboard Toolboxes with Coupled Data‑Based Mass‑Flux Models for High Resolution PFAS Conceptual Site Models

Background/Objectives

Environmental engineering practitioners face several unique challenges when dealing with per- and polyfluoroalkyl substances (PFAS). The EPA draft method 1633 has detected 40 target PFAS, but thousands of PFAS are often found at sites where AFFF releases from fire-training areas have occurred. These substances represent a long-term risk as many PFAS precursors can transform into regulated PFAS over time. To identify unmeasured PFAS, TOP/EOF/TOF assays and non‑target analysis are collected while geological, biological, structural, and climate data are used to understand fate and transport. As such, PFAS investigations are complex, and these multiple forms of analysis and target analytes can quickly create extremely large datasets. The sheer volume of data combined with numerous transport, transformation, and exposure processes quickly becomes overwhelming without a robust and reproducible way to evaluate and present the data. This talk will focus on how modern data analytics platforms can build better PFAS high-resolution conceptual site models (HR-CSMs). In the long term, the goal of these platforms is to create robust and reproducible methodologies for PFAS HR-CSM development.

Approach/Activities

Several AFFF or PFAS-impacted sites were evaluated by CDM Smith over the past several years. Several data analysis and visualization tools have been developed through successive trial and error to create PFAS HR-CSMs with multiple advanced analytics tools and visualizations. Data requirements for developing these types of HR‑CSMs are reviewed in the context of field methods used to acquire the data and to support remediation decision-making. Key data requirements include comprehensive soil logs, site specific partitioning data, comprehensive hydrologic data (e.g., surface water and groundwater levels), climatological data, co-contaminant data, target PFAS data from multiple matrices (groundwater, surface water, soil, sediment, porewater), and TOP/EOF/AOF data. Interactive dashboard tools are presented as an alternative to standard report generation for internally and externally presenting this data in a more comprehensive and integrated way. An overview of data analytics tools and their uses include:

• Transformation and Forensic tools
  • Radar plots for PFAS forensic analysis
    • How to distinguish source from background concentrations and source tracking.
  • Fluorine mass balances using target analysis with TOP/EOF data.
    • Filling in the “missing” fluorine and potential for PFAS precursors.
• 3-D Visualization Tools
  • Identification of system heterogeneity in soils and sediments affecting PFAS migration, storage (e.g., sorption or diffusion) and exposure pathways.
  • Evaluation of preferential migration pathways in groundwater, surface water or stormwater and links to receptors.
• Statistical, Geostatistical, and Geospatial Tools
  • Open-source libraries for statistical, geospatial, and geostatistical analysis in Python
    • Mann Kendall/Theil-Sen trend analysis, kriging, variograms, plotly geospatial visualization libraries to understand dynamics (e.g., seasonal variation or expanding plumes) in PFAS migration.
• Simplified Mass Flux Models
  • Rapid PFAS transport risk assessment using publicly available datasets.

Results/Lessons Learned

The foundation for conducting HR-CSMs for PFAS are presented. A comprehensive field program is suggested for sites where HR-CSMs are needed. However, for sites where comprehensive analysis is not feasible due to funding, time, or data constraints, each individual component of the HR-CSM is presented separately. Each data analytics component can be used as a line of evidence for assessing the distribution, fate and transport, or risk of PFAS exposure at an individual site.