No abstract

Rapid CO2 hydrate formation was investigated with the objective of producing a negatively buoyant CO2-seawater mixture under high-pressure and low-temperature conditions, simulating direct CO2 injection at intermediate ocean depths of 1.0-1.3 km. A coflow reactor was developed to maximize CO2 hydrate production by injecting water droplets (e.g., approximately 267 microm average diameter) from a capillary tube into liquid CO2. The droplets were injected in the mixing zone of the reactor where CO2 hydrate formed at the surface of the water droplets. The water-encased hydrate particles aggregated in the liquid CO2, producing a paste-like composite containing CO2 hydrate, liquid CO2, and water phases. This composite was extruded into ambient water from the coflow reactor as a coherent cylindrical mass, approximately 6 mm in diameter, which broke into pieces 5-10 cm long. Both modeling and experiments demonstrated that conversion from liquid CO2 to CO2 hydrate increased with water flow rate, ambient pressure, and residence time and decreased with CO2 flow rate. Increased mixing intensity, as expressed by the Reynolds number, enhanced the mass transfer and increased the conversion of liquid CO2 into CO2 hydrate. Using a plume model, we show that hydrate composite particles (for a CO2 loading of 1000 kg/s and 0.25 hydrate conversion) will dissolve and sink through a total depth of 350 m. This suggests significantly better CO2 dispersal and potentially reduced environmental impacts than would be possible by simply discharging positively buoyant liquid CO2 droplets. Further studies are needed to address hydrate conversion efficiency, scale-up criteria, sequestration longevity, and impact on the ocean biota before in-situ production of sinking CO2 hydrate composite can be applied to oceanic CO2 storage and sequestration.

The working lifetime of permeable reactive barriers (PRBs) using Fe 0 as the reactive media is limited by precipitation of secondary minerals, due to reaction of groundwater with Fe 0 . Since PRBs are emplaced at sites with widely differing groundwater chemistry, the suite of minerals that precipitate, as well as the rate of their formation, can vary widely. Using plausible phases obtained from field PRBs, the study shows that chemical equilibrium modeling can correctly predict the amounts of precipitates formed, based on the thermodynamic properties of Fe 0 and groundwater constituents. These predictions were compared to the results from the solid phase analysis from a field column experiment and from a field-installed PRB at Y-12 Plant, Oak Ridge, TN. Using the column chemical data molar distributions of the precipitates along the flow path were modeled. The maximum precipitation at the Fe 0 -sand interface at the influent end was predicted, where pore water showed high saturation index (SI) with respect to calcite and iron (oxyhydr)oxide. In the absence of flow information, the field sampling data were used to construct an SI-pH diagram, from which the extent of reaction with Fe 0 , the potential for precipitate buildup, and relative residence time for the pore water were identified. Kinetic and heterogeneous flow effects were also discussed. To illustrate the application of chemical equilibrium modeling to the design and planning phase of PRBs, groundwater data from four PRB sites were analyzed. The analysis shows that up to 0.63 cm 3 /L solid could form in pore water using an average Fe 0 dissolution rate, leading to severe clogging of Fe 0 medium over a 10-yr period of operation.