UVM Theses and Dissertations
Format:
Print
Author:
McKay, Melissa
Dept./Program:
Civil and Environmental Engineering
Year:
2004
Degree:
Ph. D.
Abstract:
Bioremediation of residual Non-Aqueous Phase Liquids (NAPLs) in the subsurface can be a time-consuming process due to the relatively slow mass transfer of NAPL contaminant species from the NAPL phase into the aqueous phase. It has been shown that alcohol flushing assists in raising the solubility of the NAPL, thereby increasing its speed of delivery into the aqueous phase. However, on its own, alcohol flushing does not alleviate the problem of what to do with the contaminant once it has been solubilized. It has been proposed that linking these two remediation techniques can generate an optimal system with the alcohol improving the transfer of NAPL into the aqueous phase where, in conjunction with injection of oxygen and nutrients, it can then be biodegraded readily by the existing microbes. To test this hypothesis, an existing multiphase flow and transport code was reformulated to accommodate the equations necessary to represent the proposed system.
The species being modeled include two NAPL contaminants (NAPL 1 and NAPL 2), oxygen, nitrogen, microbes and alcohol in the aqueous phase. Additionally, the two NAPL contaminant species, oxygen and alcohol can also be modeled in the gaseous phase, with the necessary mass transfer relationships taken into account. Bioremediation was modeled using oxygen and nitrogen limiting Monod reaction kinetics. The alcohol species was assumed to have a toxic effect on the microbes at high concentrations and therefore uses the theory of substrate, or Haldane, inhibition. Additionally, the alcohol is assumed to inhibit the bioremediation of both NAPL contaminant species, with the NAPL 1 species also inhibiting the bioremediation of the NAPL 2 species. Constitutive relationships for the effect of the alcohol on the solubility of the NAPL were also added.
A new LOcollized COllocation Method (LOCOM) was used in conjunction with an existing Hermite collocation based multiphase simulator in the solution of the partial differential equations. This new method drops the degrees of freedom from two in ID, four in 2D and 8 in 3D to one-given any numbers of dimensions. This speed increase was determined to be as much as 3.3 times the original Hermite based method with only a minimal drop in the accuracy of the solution. The new simulator was then used to model an existing bioremediation field site. The existing system was first modeled to assess the current remediation scheme and determine if additional improvements could be made. As a second step, simulation based experiments were run to determine if the addition of alcohol to the remediation system could enhance the removal and degradation of the NAPL phase.
The species being modeled include two NAPL contaminants (NAPL 1 and NAPL 2), oxygen, nitrogen, microbes and alcohol in the aqueous phase. Additionally, the two NAPL contaminant species, oxygen and alcohol can also be modeled in the gaseous phase, with the necessary mass transfer relationships taken into account. Bioremediation was modeled using oxygen and nitrogen limiting Monod reaction kinetics. The alcohol species was assumed to have a toxic effect on the microbes at high concentrations and therefore uses the theory of substrate, or Haldane, inhibition. Additionally, the alcohol is assumed to inhibit the bioremediation of both NAPL contaminant species, with the NAPL 1 species also inhibiting the bioremediation of the NAPL 2 species. Constitutive relationships for the effect of the alcohol on the solubility of the NAPL were also added.
A new LOcollized COllocation Method (LOCOM) was used in conjunction with an existing Hermite collocation based multiphase simulator in the solution of the partial differential equations. This new method drops the degrees of freedom from two in ID, four in 2D and 8 in 3D to one-given any numbers of dimensions. This speed increase was determined to be as much as 3.3 times the original Hermite based method with only a minimal drop in the accuracy of the solution. The new simulator was then used to model an existing bioremediation field site. The existing system was first modeled to assess the current remediation scheme and determine if additional improvements could be made. As a second step, simulation based experiments were run to determine if the addition of alcohol to the remediation system could enhance the removal and degradation of the NAPL phase.