Fate and Transport Modeling for NAPLs
NAPLs, short for non-aqueous phase liquids, are non-soluble organic liquid contaminants commonly found in soil or groundwater caused by industrial releases, accidents, or spills. NAPLs can be categorized into DNAPLs (Dense NAPL) for contaminants denser than water (e.g. coal tar, PCBs, chlorinated solvents) and, therefore, will sink to the bottom of the aquifer; and LNAPLs (Light NAPL), which are lighter than water (e.g. BTEX, gasoline, diesel) and will accumulate on the water table (Figure 1). Because of their toxicity and persisting nature, LNAPLs and DNAPLs can pose a long-term threat to groundwater quality when present in the subsurface in significant concentrations.
Figure 1: DNAPL and LNAPL (Source: Harold et al., 2015)
Evaluating NAPLs contamination and planning remediation strategies often requires fate and transport modeling of the contaminants, i.e., a prediction of the contaminant alteration as they flow through the unsaturated zone and the groundwater aquifer. Numerical modeling of NAPLs is often coupled with groundwater flow modeling but it is not interchangeable with particle tracking modeling because NAPLs do not solely mix with water (although they do have a finite solubility in water) and, therefore, they have different flow characteristics. Additionally, contaminants’ physical and chemical characteristics need to be considered. Because a NAPL can be both dissolved in the groundwater flow, but also exist in a separate immiscible phase, the differing physical and chemical characteristics of the NAPL and groundwater results in a physical separation of the two liquids. This results in differing conditions if you are dealing with a LNAPL versus a DNAPL.
Both LNAPLs and DNAPLs can move through the vadose zone based on the geology of the sediments overlying the groundwater aquifer. The retention of NAPLs in the vadose zone can also result in subsequent leaching of these contaminants into the groundwater table as there is future recharge through the vadose zone to the groundwater table. LNAPLs will start to flow downgradient once they reach the groundwater table (Figure 2a), spreading out as an immiscible layer on top of the water table. Conversely, DNAPLs will tend to continue to move vertically downward beneath the groundwater table until reaching a zone of low permeability, which may be the base of the aquifer or an intermediate low-permeability zone (Figure 2b). The DNAPL may then both flow with the groundwater and/or accumulate in low areas. One of the complications of tracking DNAPLs is that, because they can accumulate in low spots in an aquifer, DNAPL concentrations may only move slowly as native groundwater sweeps across the low area. Therefore, DNAPLs typically do not move with the prevailing groundwater velocity.
Figure 2a: LNAPL Contaminant plumes (Source: Rivett et al., 2006)
Figure 2b: DNAPL Contaminant plumes (Source: Rivett et al., 2006)
Because of the two-phase flow associated with NAPLs in a groundwater aquifer, modeling of these contaminants in a groundwater system can be very difficult. It requires a thorough understanding of not only the characteristics of the contaminants but also the geologic strata, particularly as it relates to DNAPLs, but also to all NAPLs in the vadose zone. Understanding the source location and contaminant characteristics such as solubility, volatility, sorption, degradability, and residual products is crucial for modeling the fate and transport of NAPL contaminants. Geologically, heterogeneous conditions that result in highly variable horizontal and vertical hydraulic conductivities, as well as local variations in longitudinal dispersivity, make it important to understand how to track fluids that are influenced by these factors.
Groundwater levels are used for the calibration of groundwater flow models, therefore, it is also important to have an adequate water level database. In addition, NAPL fate and transport modeling requires observations of contaminant concentrations, both spatially and temporally, to assist with the calibration of the model. Good coverage of observation points that provide geologic, water quality, and water level data can substantially improve model accuracy. A model is only as good as the quality of its input parameters and the availability of observation points for model calibration.
Computation codes for modeling NAPLs are, for example, BIOSCREEN, MODFLOW coupled with MOC3D, MT3D, or RT3D, or FEFLOW. BIOSCREEN provides an analytic solution, while MODFLOW is based on a finite-difference solution, and FEFLOW uses a finite-element solution (see LWS June 30, 2020 blog on “Numerical Modeling Methods in Water Management.”)
A contaminant fate and transport model can be used to predict the contaminant behavior and evaluate the efficiency of proposed remediation plans. The LWS modeling team can help with developing a flow and fate and transport model, including assessing existing databases and recommending the need for additional data development so a representative model can be developed. Ultimately, developing quality input data and calibration targets are vital to obtaining a fully calibrated model that is capable of characterizing NAPL plumes and to simulate potential remediation scenarios. Please do not hesitate to contact LWS with questions regarding fate and transport modeling to characterize contaminant movement, as well as groundwater remediation strategies:
Bruce Lytle, P.E., President of LWS: bruce@lytlewater.com
Chris Fehn, P.E., P.G., Senior Project Engineer: chris@lytlewater.com
Anna Elgqvist, E.I., Senior Engineer: anna@lytlewater.com
References:
Harold F. Hemond, Elizabeth J. Fechner, 2015, The Subsurface Environment. In: Chemical Fate and Transport in the Environment (Third Edition), Academic Press, 219-310
Rivett, M., Drewes, J., Barett, M., Chilton, J., Appleyard, S., Dieter, H.H., Wauchope, D. and Fastner, J. 2006, Chemicals: Health Relevance, Transport and Attenuation. In: Schmoll, O., Howard, G., Chilton, J., Chorus, I (eds). Protecting Groundwater for Health: Managing the Quality of Drinking Water Sources, London, IWA Publishing, pp. 54.
