Parameterization of modeling subsurface hydrocarbon contamination and biosurfactant enhanced remediation processes

Li, Zelin (2016) Parameterization of modeling subsurface hydrocarbon contamination and biosurfactant enhanced remediation processes. Masters thesis, Memorial University of Newfoundland.

[English] PDF - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. Download (4MB)

Abstract

Subsurface hydrocarbon contamination caused by accidental spills or operational leakages of petroleum products is a global environmental concern. In order to cost-effectively and eco-friendly recover the contaminated sites, biosurfactant enhanced aquifer remediation (BSEAR) technologies have become a popular subject in both research and practice. However, the inherent uncertainties and complexities of the subsurface systems make it challenging in numerical simulation of the hydrocarbon transport and fate as well as remediation processes. Efforts in developing more efficient and robust parameterization approaches for such modeling purpose, therefore, are highly desired. This research aims to help fill the gap by developing a novel hybrid stochastic – design of experiment aided parameterization (HSDP) method for modeling BSEAR processes. The method was developed and tested based on an integrated physical and numerical modeling system comprised of a set of intermediate scale flow cells (ISFCs) and a numerical simulator named BioF&T 3D. Generally, the HSDP method was performed by: 1) building the design of experiment (DOE) models based on screened parameters and defined responses, which could reflect the goodness of fit between observed and simulated data; 2) identifying the and interactions among parameters and their significance; 3) optimizing the DOE predicted responses; 4) introducing stochastic data within reduced intervals based on the optimized parameters; 5) running Monte Carlo simulation to find the optimal responses with the corresponding combinations of parameters. The flow cell tests proved that the HSDP method could improve both efficiency and robustness of modeling parameterization and significantly reduce the computational demand without compromising the effectiveness in quantifying parameter interactions and uncertainties. Furthermore, a specific lab synthetized surfactin was applied in this study. The effect of dissolution enhancement was observed from parallel flow cell experiments especially during the first 12 hours following the initial hydrocarbon release. The HSDP method was demonstrated to be capable of advancing BioF&T 3D, which lacks the capacity of simulating surfactant. By incorporating the HSDP method, the BSEAR processes were effectively simulated with a satisfactory overall goodness of fit (R² = 0.76, 0.81, 0.83, and 0.81 for benzene, toluene, ethylbenzene, and xylene, respectively). The enhanced dissolution effect was also reflected in the modeling parameterization by increasing the first 12 hours hydrocarbon loading ratio (12LR) compared to non-biosurfactant processes. This research developed a new parameterization method HSDP, which is capable of revealing interactions of parameters, as well as quantifying their uncertainties, in a robust and efficient manner. Also, using this method, this study initiated the attempts to advance simpler numerical models in simulating complicated BSEAR processes, which is particularly attractive for the potential applications in practice.

Actions (login required)

View Item