Long-term behavior of PFAS in contaminated agricultural soils in Germany

Highlights

• 12-years field-scale leaching data trends from the first documented PFAS contaminated agricultural site in Germany.

• PFAS column leaching tests do not reflect the long-term tailing observed at the field site over 12 years.

• Long-term tailing is potentially caused by transformation of precursor substances into PFOA and PFOS.

• Field data of short chain PFAS show a distinct seasonal patter.

• These case studies demonstrate that it will take decades for PFAS to leach from agricultural topsoils.

Abstract

Per- and polyfluoroalkyl substances (PFAS) contaminated compost materials have been applied over the last few decades to agricultural fields in Germany, resulting in large-scale diffuse PFAS plumes. The leaching behavior of PFAS from the first two identified contaminated agricultural sites in Germany were investigated, one at Brilon-Scharfenberg, North Rhine-Westphalia (BS-NRW), and the other at Rastatt/Mannheim, Baden-Württemberg. The specific objectives of this study were to assess the longevity of the PFAS agricultural sources and compare standardized column percolation tests to long-term leaching of PFAS from contaminated sites. The advection-dispersion model (ADM) was used to compare the leaching behavior of perfluorooctanoic acid (PFOA) and perfluorooctanesulfonic acid (PFOS) from standardized column percolation tests and long-term field leaching data from the BS-NRW site. Column leaching tests conducted with PFOS and PFOA contaminated soil simulated the initial rapid decline but did not predict the long-term behavior (tailing) observed at the field site over 12 years. Trend analyses of the PFAS field data from the BS-NRW showed that concentrations had stabilized and that individual PFAS exhibited distinct seasonal fluctuations; the latter is likely due to the ongoing transformation of precursors and a seasonal influence on production rates of mobile PFAS. Mass balances conducted at both sites indicate that complete removal of these compounds will likely take years to decades to occur, which is expected from the results of the column leaching tests.

Introduction

Per- and polyfluoroalkyl substances (PFAS) are anthropogenic contaminants that comprise over 3,000 individual compounds and are likely widespread in the environment (OECD, 2018). Degradable PFAS (e.g., N-Ethyl perfluorooctane sulfonamide ethanol-based phosphate diesters (diSAmPAP)) are often referred to as precursor substances that are transformed to final transformation products, such as perfluorooctanoic acid (PFOA) and perfluorooctanesulfonic acid (PFOS) (Benskin et al., 2013). These final transformation products are persistent against further breakdown and appear to be widespread within the environment (Chen et al., 2019; Falk et al., 2019; Routti et al., 2015). Primary sources of PFAS in the environment are fire-fighting training grounds, industrial sites, landfills, and wastewater treatment plants (WWTPs) (ITRC, 2018). However, the historical application of biosolids/compost has led to known large-scale diffuse contamination of agricultural land and adjacent bodies of water in the states of Baden-Wuerttemberg, Germany and North Rhine-Westphalia, Germany, and has likely occurred in other regions worldwide (Delschen et al., 2007; Rastatt, 2016). Currently, only a few studies have characterized the extent of PFAS contamination at agricultural sites and their long-term fate in the environment (Lindstrom et al., 2011; Washington et al., 2010; Yoo et al., 2010; Gellrich et al., 2012; Stahl et al., 2018; Söhlmann et al., 2018).

Washington et al. (2010) and Yoo et al. (2010) reported the first PFAS contaminated agricultural fields in the United States of America (USA). In Decatur, Alabama, PFAS contaminated sludges from a WWTP had previously been applied to an agricultural area of approximately 20 km² for over a decade, leading to the pollution of both surface water and groundwater (Lindstrom et al., 2011; Washington et al., 2010). Soil analyses of the polluted agricultural areas showed levels of precursor substances, such as fluorotelomer alcohols (FTOHs), up to 166 μg/kg and final transformation products, such as C6-C14 perfluorocarboxylic acids (PFCAs) and PFOS with perfluorodecanoic acid (PFDA), up to 990 μg/kg (Washington et al., 2010; Yoo et al., 2010). They found that long-chain PFCAs resided predominantly in the topsoil, while C6-C8 migrated to deeper soil layers. Lindstrom et al. (2011) investigated surface and groundwater for PFAS in the vicinity of the agricultural fields in Decatur, Alabama. They detected PFCAs from C4-C9 and perfluorosulfonic acids (PFSAs) between C4-C8, showing the mobility of these compounds in soils, while C10-C14 PFCAs remained in the soil and were not detected in groundwater. This observation is consistent with the findings of Washington et al. (2010) and other studies investigating perfluoroalky acids (PFAAs) mobility in soil (Bräunig et al., 2019; Gellrich et al., 2012).

Two of the first known PFAS contaminated agricultural sites, similar to the one described in Washington et al. (2010), were discovered in Germany in 2006 and 2013, after the application of compost and fertilizers mixed with PFAS-containing waste materials (e.g., paper sludge). During the time of application, no regulatory limits for PFAS existed in the German Fertilizer Ordinance (GFO), which regulates the use of fertilizers, soil conditioners, culture substrates, and plant aids. In 2012, the GFO was updated to include a limit of 100 μg/kg dry weight (∑PFOA+PFOS) to limit the spread of PFAS in the environment (DüMV, 2012). While the use of PFOS and PFOA are regulated for fertilizing materials, Germany still has no binding limits for PFAS in drinking water and groundwater. In 2017, the German Federal Environmental Agency (UBA) released a list with 13 PFAS, recommending regulatory limits for seven of these compounds in drinking water (Table S1 & S2). In 2020, the European Union (EU) included PFAS into their drinking water directive, setting a limit value for the sum of 20 distinct PFAS of 0.1 μg/l and total PFAS of 0.5 μg/l (EU Directive 2184, 2020). The contaminated agricultural sites discovered in Germany triggered laboratory studies, combined with clean-up and monitoring programs, to reduce PFAS loads from contaminated agricultural soils to surface and groundwater. The primary objective of this study was to assess the long-term behavior of PFAS in agricultural soils. Specific objectives of the study were to: (1) evaluate the ability of standardized column percolation tests to predict long-term leaching of PFAS from contaminated agricultural sites; (2) to assess the seasonality of PFAS-leaching behavior in agricultural soil; and (3) quantify the longevity of PFAS agricultural sources.

Field leaching PFAS data of a 10 ha site in Brilon-Scharfenberg, North Rhine-Westphalia (BS-NRW) site, monitored over a period of more than 12 years, were compared to laboratory column studies conducted after the contamination was detected in 2006. Earlier studies showed that data from column leaching tests from various types of material (e.g., contaminated soil, demolition waste, and waste incineration ash) for different compounds (e.g., heavy metals and polyaromatic hydrocarbons) compared well to lysimeters and could be fitted reasonably well with the advection-dispersion model (ADM) (Grathwohl and Susset, 2009). As such, soil column and field leaching data were fitted with the ADM to understand the short- and long-term leaching behavior. Trend analyses were also performed using the Seasonal Kendall test, Mann-Kendall test, and exponential decay functions to estimate rate constants. Data from this site are compared with another case of a 644 ha contamination of agricultural land detected more recently in Rastatt/Mannheim, Baden-Württemberg, Southern Germany, which will further be referred to as the Baden site.