One of the methods by which anthropogenic CO2 can be sequestered from the atmosphere is by direct ocean injection. Estimates show that over 300 GtC could theoretically be stored in the deep ocean without increasing the pH by more than 0.18 units, which is comparable to levels of observed natural variability. At deep ocean pressures and temperature, carbon dioxide hydrate particles have been observed to form in the laboratory and in the field. They can be up to 1 percent denser than the ambient seawater (for pure hydrates). Currently a partially reacted hydrate particle extruded from an injector in the field is observed to sink further than 100m on its own, and is predicted to descend to about 300m before they dissolve. Particle plumes consisting of releases of these particles at a rate of 100kg/s of CO2 per second (roughly the output of one power plant) have been numerically predicted to sink to about 1000m below the release point. However, large currents and density stratification in the ocean may affect the plume dynamics of these sinking hydrate particles: the former may waft the plume away from the particles that drive it downward; the latter may cause the plume to form intrusion layers midway down. A possible method by which these ambient effects may be avoided is by building a shroud around the release. The shroud would act like an inverted gas-lift pump, that conveys the sinking particles to a desired depth and uses the particle buoyancy to induce a downward draft to promote dilution.