Carbon dioxide hydrate particles for ocean carbon sequestration

This paper presents strategies for producing negatively buoyant CO2 hydrate composite particles for ocean carbon sequestration. Our study is based on recent field observations showing that a continuous-jet hydrate reactor located at an ocean depth of ~1500 m produced curved negatively buoyant cylindrical particles with diameters ~ 2.5 cm and lengths up to ~ 1 m. Accordingly we performed new laboratory experiments to determine the drag coefficient of such particles and, based on the measured drag coefficient and the initial settling velocity observed in the field, have concluded that the reactor efficiency (percentage of liquid CO2 converted to hydrate) in the field was ~ 16%. Using the dissolution rates observed in the field, we conclude that such particles would ultimately sink to depth below discharge of ~ 115 m. We have also predicted the sinking depth of particles potentially produced from various scaled-up reactors and have shown that, for example, a 10 cm diameter particle produced with a hydrate conversion of 50 % could reach the ocean bottom before completely dissolving. In a real sequestration scenario, we are interested in following large groups of hydrate particles released continuously. We have previously shown that increasing particle size and hydrate conversion efficiency enhances the sinking of hydrate particle plumes produced by the continuous release of CO2 in a quiescent ambient, but that a sufficiently strong current will cause the entrained particles to separate from the plume and settle discretely. In the latter case, particles of different sizes and hydrate conversions (hence different settling velocities) will follow different settling trajectories as they dissolve. This particle fractionation, if employed deliberately, spreads the discharged CO2 in the down current and vertical directions, enhancing mixing, while turbulent diffusion helps spread the CO2 in the third direction. A numerical model that incorporates these processes is used to predict the downstream concentrations and changes in pH from such particle plumes in a ‘strong’ current. An extension of this model simulates hydrate particles that are released continuously from a moving ship. Because of the ship speed, such particles would never form a coherent plume, but the combination of particle fractionation and advection due to the ship motion produces excellent dilution of the discharged CO2. © 2008 Elsevier Ltd. All rights reserved