The California Multimedia Risk Assessment Protocol for Alternative Fuels: Application to Biodiesel and Study of Multiphase Flow of Biodiesel in Porous Media

The research reported in this dissertation describes both the overall multimedia risk assessment (MMRA) for biodiesel and the particular experimental testing for risk of mobility of the fuels in the subsurface.;The MMRA is performed in three tiers, as detailed in the following chapters. The Tier I multimedia risk assessment identified key knowledge gaps regarding aquatic toxicity, biodegradation, and subsurface fate and transport. Tier II experiments were designed and executed to address the knowledge and emphasized Soy and Animal Fat Biodiesel relative to ULSD. Additional studies for air quality were pursued using newer diesel engines and comparing ULSD emissions to those of biodiesel. The biodegradation experiments were performed using aerobic respirometry in microcosms. The aquatic toxicity experiments were performed for 6 species (three estuarine and three freshwater). The subsurface transport experiments were performed using 2D infiltration columns for determining lens formation and redistribution.;The results of the Tier II experiments suggested that both soy and animal fat biodiesel were more readily biodegradable than ULSD under aerobic conditions. The experimental results for toxicity exhibited somewhat increased toxicity to several tested species compared to ULSD. The antioxidant-additized blends increased toxicity for a smaller group of tested species compared to unadditized blends. The subsurface infiltration and redistribution experiments showed that overall soy and animal fat 20% blends resulted in very similar fuel fate and transport in the subsurface, including similar formation of fuel "lens"es on the water table. The neat soy biodiesel also showed very similar lens distribution to ULSD. The neat animal fat biodiesel showed increased residual in the vadose zone and smaller lens geometry than ULSD.;Due to the complexity of laboratory experiments and the qualitative nature of the (photographic) data, simplified numerical simulations of multiphase flow were coded in TMVOC to replicate the conditions seen in the laboratory experiments using physical properties for biodiesel and biodiesel components taken from the literature. Single species, pure biodiesel infiltration experiments were simulated for each soy and animal fat biodiesel and compared to behavior of a simplified ULSD using literature composite properties and other data found with similar carbon chain length, density, and viscosity. The numerical experiments were conducted with homogeneous permeability, and a capillary pressure-saturation relationship for the porous media (medium sand) that was the same for all cases, and I effectively neglected differences in the infiltration due to interfacial tension differences between the biodiesels and ULSD.;The results of the numerical simulations showed a very similar infiltration time to lens formation to the laboratory experiments. The laboratory experiments showed more pore to pore effects not able to be resolved in the macroscale averaged numerical solution. The extent of spreading and thickness of the lenses appear to be consistent between the laboratory simulations and the numerical simulations of the same scale. Due to lower viscosity in the ULSD, it was able to spread slightly further and to make a slightly larger lens in a similar amount of time. In addition, ULSD has a slightly lower density, but it was not low enough to counteract the effects of the lower viscosity.;Based on these numerical model simulations, very similar results can be visualized with the use of literature data when comparing laboratory and numerical simulations. The benefit of the numerical simulations is the ability to control the conditions for consistency between trials. On the other hand, the benefit of the laboratory experiments allows for visualization of small scale effects not able to be seen in the numerical model due to macroscale averaging and to small scale heterogeneities in the pore space. From a multimedia risk perspective, numerical models do provide a way to evaluate the mobility of fuels or other chemicals in the subsurface environment in order to make recommendations regarding relative risk. (Abstract shortened by UMI.).