Modelling the transport of per- and polyfluoroalkyl substances (PFAS) from source zones to recipients – challenges and model developments

Per- and polyfluoroalkyl substances (PFAS) constitute an increasing problem for water resources and aquatic ecosystems globally. PFAS are extremely persistent, many are also mobile in water and can transport long distances in the groundwater. Highly contaminated source zones for PFAS exist all over the world in connection to firefighting training areas of rescue services and airports. Modelling tools to predict the subsurface transport of PFAS are important both for understanding the transport, assess risk and design of remediation measures such as in-situ stabilization using sorbents.

PFAS have several properties that distinguish them from many other pollutants. For PFAS reactive transport modelling, several challenges and development needs therefore exist. PFAS are surface-active substances and are attracted to interfaces between air and water, which affects retention in the unsaturated zone. Hundreds of PFAS with different transport properties may be present in typical PFAS-pollution source zones. Many of these substances can be partially degradable (so-called precursors) and break down until a perfluorinated substance is formed. This typically increases mobility, and can be critical for how quickly PFAS leach from the unsaturated zone to groundwater. The different PFAS can also compete for sorption sites, which may increase the mobility of some PFAS and therefore affect both the risks associated with PFAS transport and the efficacy of remediation strategies such as sorbent amendments.

In two recently started research projects, we aim to test and develop practically useful models for subsurface PFAS transport from source zone to recipient. In modelling tools such as GMS/MODFLOW, we have added capability in the transport module to account for competition effects during sorption and we are currently investigating how degradation of precursors coupled to the leaching of PFAS from the unsaturated zone and should best be included in the modelling.