Modeling PFAS transport through the vadose zone – a comparison of model codes for column and field-scale experiments

Predicting transport of per- and polyfluoroalkyl substances (PFAS) through the vadose zone is essential for contamination risk assessment, yet the reliability of available models remains poorly understood. This study compared five available models by exploring their conceptual and technical differences, and assessing the alignment of model capabilities against current theoretical understanding of reactive transport of PFAS. We evaluated predictive performance by forward modeling two column experiments and conducting one virtual field-scale simulation across different PFAS compounds and soil types. While all models agreed well with short-term column experiments (NSE ≥ 0.82), they diverged substantially at field-scale, with mid-point breakthrough times differing by multiple years despite identical parameterization.
Quantification of the air-water interface (AWI) emerged as the primary source of inter-model variability and remains the most disputed aspect in theoretical reactive transport understanding of PFAS transport. Existing approaches compute systematically different AWI values as functions of saturation and soil physical parameters, whilst likely underestimating the interfacial area by neglecting surface roughness of grains and pore-scale complexity.
All examined models employ simplified and empirically-derived solid-phase sorption parameters that do not account for soil-specific behavior, solution chemistry, and soil heterogeneity. Important processes including precursor transformation, competitive sorption, and desorption hysteresis remain largely unimplemented, fundamentally constraining predictive reliability. Hence, comprehensive multi-year field validation datasets across diverse hydrogeological settings are urgently needed to quantify prediction uncertainty and establish robust parameterization strategies.