Logo image
Back
Field-scale calibration and validation of numerical modeling of per- and polyfluoroalkyl substances (PFAS) leaching during unsaturated zone flushing using lysimeter-derived porewater concentrations
Dissertation   Open access

Field-scale calibration and validation of numerical modeling of per- and polyfluoroalkyl substances (PFAS) leaching during unsaturated zone flushing using lysimeter-derived porewater concentrations

Ayowale Emmanuel Ayodele
Doctor of Philosophy (Ph.D.), Drexel University
May 2025
DOI:
https://doi.org/10.17918/00011013
pdf
Ayodele_Ayowale_20256.99 MBDownloadView
PDFOpen Access (License Unspecified),  Open Access

Abstract

Air-water interface Calibration HYDRUS Numerical modeling PFAS leaching Validation Computer Engineering
Increasing apprehensions of environmental hazards and health issues related to per- and polyfluoroalkyl substances (PFAS) chemical use and persistence in the environment have prompted stringent regulatory requirements and standards. Current PFAS leaching models often use laboratory parameters for calibration and validation, neglecting complicated field-scale PFAS processes such as dynamic water flow and air-water interface (AWI) that govern long-term unsaturated zone leaching. Data from PFAS field investigations and flushing experiments conducted by Schaefer et al. (2019-2023) were utilized for calibrating the HYDRUS numerical code. The calibrated model was validated through an assessment of the role of the AWI and the quantification of its impact on PFAS leaching at various depths and locations. A Wilcoxon Signed-Rank test at 95% confidence level ([alpha] = 0.05) indicates no significant differences in perfluoroheptane sulfonic acid (PFHpS), perfluorooctane sulfonate (PFOS), and perfluorooctanoic acid (PFOA) concentrations in field lysimeters (p = 0.981, 0.255, and 0.490, respectively). The Coefficient of Determination (R²) of 0.86 and Nash-Sutcliffe Efficiency (NSE) of 0.69 suggest a good performance of the model in predicting PFAS leaching amidst complex flow dynamic conditions. The integration of Latin Hypercube Sampling (LHS) and Morris Method ensured a comprehensive approach to parameter sampling, ensuring unbiased coverage and enhancing the robustness of the sensitivity analysis. The results revealed that saturated moisture ([theta]_s), diffusion coefficient (D_w), sorption partitioning coefficient (K_d), air entry pressure ([alpha]), and bulk density ([rho]b) significantly impact output PFAS porewater concentrations. Quantification of AWI impacts on PFAS leaching were performed using five comparative metrics: area under the PFAS concentration vs. time curve (AUC), percentage difference AUC, average leaching rate (ALR), percentage difference ALR, and retardation factor (R_f). The findings indicate that at shallow depths of 15 and 61 cm below ground surface (bgs), AWI significantly enhanced PFAS leaching, whereas at greater depths of 120 and 150 cm bgs, it retarded PFAS leaching. This demonstrates a dual role of AWI, serving as both an enhancer and a retardant of PFAS leaching in the unsaturated zone. The impacts of AWI on PFAS have been shown in this study to be highly time- and depth- dependent, governing both the timing and extent of PFAS leaching. For accurate remediation planning and risk management of PFAS-impacted sites, neglecting these factors may lead to under-or-overestimation of PFAS flux to the underlying groundwater aquifer.

Metrics

1 File views/ downloads
69 Record Views

Details

Logo image