Per- and polyfluoroalkyl substances (PFAS) in final treated solids (Biosolids) from 190 Michigan wastewater treatment plants.

Trends in concentration, distribution, and variability of per- and polyfluoroalkyl substances (PFAS) in biosolids are characterized using an extensive dataset of 350 samples from 190 wastewater treatment plants (WWTPs) across Michigan. All samples are comprised of final treated solids generated at the end of the wastewater treatment process. Concentrations of both individual and Σ24 PFAS are lognormally distributed, with Σ24 PFAS concentrations ranging from 1-3200 ng/g and averaging 108 ± 277 ng/g dry wt. PFAS with carboxyl and sulfonic functional groups comprise 29% and 71% of Σ24 PFAS concentrations, respectively, on average. Primary sample variability in concentration is associated with long-chain PFAS with higher tendency for partitioning to biosolids. Short-chain carboxylic compounds, most notably PFHxA, are responsible for secondary concentration variability. Usage of FTSA and PFBS replacements to long-chain sulfonic compounds also contributes to variance in biosolids concentrations. Sulfonamide precursor compounds as a collective group are detected at a similar frequency as PFOS and often have higher concentrations. Trends in PFAS enrichment for individual PFAS vary at least 3 orders-of-magnitude and generally increase with compound hydrophobicity; however, partitioning of PFAS onto solids in WWTPs is a complex process not easily described nor constrained using experimentally-derived partitioning coefficients.

Long-duration monitoring and mass balance of PFAS at a wastewater treatment plant following the release of aqueous film-forming foam concentrate.

Approximately 760 liters (200 gallons) of first-generation, PFOS-dominant, Aqueous Film-Forming Foam (AFFF) concentrate entered the sanitary sewer after an accidental release at the Kalamazoo/Battle Creek International Airport and migrated 11.4 km to the Kalamazoo Water Reclamation Plant. Near-daily sampling of influent, effluent, and biosolids generated a high-frequency, long-duration dataset used to understand the transport and fate of accidental PFAS releases to wastewater treatment plants, identify AFFF concentrate composition, and perform a plant-wide PFOS mass balance. Monitored influent concentrations exhibited sharp PFOS declines after 7 days post-spill, yet effluent discharges remained elevated due to return activated sludge (RAS) recirculation, resulting in the exceedance of Michigan's Surface Water Quality Value for 46 days. Mass balance estimates indicate 1.292 kg PFOS entering the plant and 1.368 kg leaving. Effluent discharge and sorption to biosolids account for 55% and 45% of estimated PFOS outputs, respectively. Identification of AFFF formulation and reasonable agreement between computed influent mass and reported spill volume demonstrates effective isolation of the AFFF spill signal and increases confidence in the mass balance estimates. These findings and related considerations provide critical insight for performing PFAS mass balances and developing operational procedures for accidental spills that minimize PFAS releases to the environment.

Per- and Polyfluorinated Alkyl Substances (PFAS) cycling within Michigan: Contaminated sites, landfills and wastewater treatment plants.

Concentrations of Per- and Polyfluorinated Alkyl Substances (PFAS) from public and private sources in Michigan compiled for wastewater treatment plants (WWTPs) (influent, effluent, biosolids), contaminated sites, and landfill leachates reveal complex cycling within the natural and engineered environment. Analysis of 171 contaminated sites in Michigan by source release indicate four dominant PFAS sources - landfills, aqueous film-forming foams (AFFF), metal platers, and automotive/metal stamping - account for 75% of the contamination. Diverse chemical signatures were observed for leachates collected from 19 landfills (mostly type II municipal) with the dominant PFAS ranging from perfluorooctanoic acid (PFOA) and perfluorooctanesulfonic acid (PFOS) to shorter-chained compounds, perfluorohexanoic acid (PFHxA), perfluorobutanoic acid (PFBA), and perfluorobutanesulfonic acid (PFBS). Analysis of PFAS carbon chain length as a function of landfill age shows the transition of C8s in leachate from older landfills to C4s and C6s in younger landfills, consistent with the phasing out and replacement of C8s. PFAS mass flux in leachate for landfills studied range between 5 - 2,000 g/yr and are highest for active landfills, which generate greater leachate volumes and contain fresh PFAS wastes. Detailed study of 10 WWTPs with industrial pretreatment programs indicate numerous chemical transformations across the plants that yield effluent PFAS concentrations as much as 19 times greater than influent, attributed to transformations of unmeasured precursors in the influent to measured, stable PFAS in the effluent. PFOA, PFHxA, perfluoropentanoic acid (PFPeA), PFBA, and PFBS show the greatest increases across the plant ranging from 20% to nearly 2,000%. PFOS concentrations decreased across 6 WWTPs, consistent with a strong tendency to adsorb onto biosolids. Estimated mass of discharge of (mostly unregulated) PFAS from WWTPs to receiving waters range from 40 g/yr to 128 kg/yr.

A Dye Tracer Approach for Quantifying Fluid and Solute Flux Across the Sediment-Water Interface.

We propose a dye tracer method to characterize fluid and solute fluxes across the sediment-water interface. Zones of groundwater discharge within the streambed are first identified, and small volume slugs of 0.5 to 1 mL fluorescein dye are released at known subsurface depths. Fluorescein dye allows for visual identification of interface breakthrough locations and times, and dye concentrations at the point of discharge are recorded over time by a fluorometer to generate high resolution breakthrough curves. Groundwater velocities and dispersivities at the demonstration site are estimated by numerically fitting dye breakthroughs to the classical advection-dispersion equation, although the methodology is not limited to a specific transport model. Breakthroughs across the stream-sediment interface at the demonstration site are nonlinear with tracer release depth, and velocity estimates from breakthrough analysis are significantly more reliable than visual dye (time to first dye expression) and Darcy methods which tend to overestimate and underestimate groundwater velocity, respectively. The use of permanent injection points within the streambed and demonstrated reproducibility of dye breakthroughs allow for study of fluid and solute fluxes under seasonally varying hydrologic conditions. The proposed approach also provides a framework for field study of nonconservative, reactive solutes and allows for the determination of characteristic residence times at various depths in the streambed to better understand chemical and nutrient transformations.

Use of a variable-index fractional-derivative model to capture transient dispersion in heterogeneous media.

Field and numerical experiments of solute transport through heterogeneous porous and fractured media show that the growth of contaminant plumes may not exhibit constant scaling, and may instead transition between diffusive states (i.e., superdiffusion, subdiffusion, and Fickian diffusion) at various transport scales. These transitions are likely attributed to physical properties of the medium, such as spatial variations in medium heterogeneity. We refer to this transitory dispersive behavior as "transient dispersion", and propose a variable-index fractional-derivative model (FDM) to describe the underlying transport dynamics. The new model generalizes the standard constant-index FDM which is limited to stationary heterogeneous media. Numerical methods including an implicit Eulerian method (for spatiotemporal transient dispersion) and a Lagrangian solver (for multiscaling dispersion) are utilized to produce variable-index FDM solutions. The variable-index FDM is then applied to describe transient dispersion observed at two field tracer tests and a set of numerical experiments. Results show that 1) uranine transport at the small-scale Grimsel test site transitions from strong subdispersion to Fickian dispersion, 2) transport of tritium at the regional-scale Macrodispersion Experimental (MADE) site transitions from near-Fickian dispersion to strong superdispersion, and 3) the conservative particle transport through regional-scale discrete fracture network transitions from superdispersion to Fickian dispersion. The variable-index model can efficiently quantify these transitions, with the scale index varying linearly in time or space.

A model-averaging method for assessing groundwater conceptual model uncertainty.

This study evaluates alternative groundwater models with different recharge and geologic components at the northern Yucca Flat area of the Death Valley Regional Flow System (DVRFS), USA. Recharge over the DVRFS has been estimated using five methods, and five geological interpretations are available at the northern Yucca Flat area. Combining the recharge and geological components together with additional modeling components that represent other hydrogeological conditions yields a total of 25 groundwater flow models. As all the models are plausible given available data and information, evaluating model uncertainty becomes inevitable. On the other hand, hydraulic parameters (e.g., hydraulic conductivity) are uncertain in each model, giving rise to parametric uncertainty. Propagation of the uncertainty in the models and model parameters through groundwater modeling causes predictive uncertainty in model predictions (e.g., hydraulic head and flow). Parametric uncertainty within each model is assessed using Monte Carlo simulation, and model uncertainty is evaluated using the model averaging method. Two model-averaging techniques (on the basis of information criteria and GLUE) are discussed. This study shows that contribution of model uncertainty to predictive uncertainty is significantly larger than that of parametric uncertainty. For the recharge and geological components, uncertainty in the geological interpretations has more significant effect on model predictions than uncertainty in the recharge estimates. In addition, weighted residuals vary more for the different geological models than for different recharge models. Most of the calibrated observations are not important for discriminating between the alternative models, because their weighted residuals vary only slightly from one model to another.