Dense, Viscous Brine Behavior in Heterogeneous Porous Medium Systems During a Brine- Based DNAPL Remediation Approach

Difficulties of dense nonaqueous phase liquid (DNAPL) remediation arise due to a variety of factors, including subsurface heterogeneities, multiphase behavior, and high DNAPL densities and viscosities. A number of technologies have been developed to address these challenges, including a novel brine-barrier remediation technology (BBRT), which relies upon a brine barrier to control migration of DNAPL once mobilized by other means. The objectives of this study fall into two categories: (1) to evaluate field-scale performance and recovery of a brine barrier and (2) to investigate the impact of subsurface heterogeneity and instabilities on brine barrier establishment and recovery through mathematical modeling. To evaluate performance of a BBRT in a field setting, a pilot study implementing was performed in a test cell at Dover National Test Site in Dover, DE, to evaluate dense, viscous brine behavior and recovery. The test cell allowed for application of a BBRT with the constraint of horizontal barriers due to installation of sheet piles enclosing the cell. In addition, a multilevel sampling panel allowed for extensive monitoring of in situ densities during establishment, maintenance, and recovery stages. It was demonstrated that adequate density could be established and maintained in field setting when horizontal barriers exist. Through downward flushing, we were able to recover approximately 92% of the injected calcium bromide brine. The effects of subsurface heterogeneity and viscosity on brine behavior were investigated utilizing a subset of stochastic functions to represent correlated random permeability fields and to assess their impacts on brine emplacement and brine removal efficiency. For moderately heterogeneous domains, mounding of viscous brines was mitigated by the use of sufficiently low injection rates. However, local heterogeneities were shown to be the dominant factor in subsequent brine removal, emphasizing the importance of subsurface knowledge for accurate modeling and efficient implementation of BBRT.