To mitigate the increasing atmospheric CO2 level and reduce the subsequent impacts, storage of CO2 is one option that might become necessary. Ocean storage and storage of CO2 under the sea bottom with potential leakage into the ocean both require knowledge of the behaviour of CO2 drops in seawater. An analysis of the dissolution and velocity of a single CO2 drop rising from 800 m depth is made. The drop is simulated by a numerical model and theory is compared to data from an ocean experiment performed in Monterey bay, California. Single CO2 drops were there released at 800 m and imaged under hydrate forming conditions. The slip velocity of a hydrate-covered drop has been suggested to correspond to the drag of a spherical rigid particle. However, a clear discrepancy between this theory and the ocean observations is shown. Some possible explanations for this deviation are found. An existing parameterisation for slip velocity accounting for change of shape fits the observed drop velocity reasonably well. That deformation must be taken into account in the calculations of drop terminal velocities is confirmed by similar studies. Hydrate is expected to reduce the mass transfer. Common theory with a reduction factor of 2 matches the observed dissolution rate. This is compatible with laboratory results. A main reason for studying mass transfer and drop velocity is to get increased knowledge about the distribution of dissolved CO2 in the water column. The effects of varying initial drop size, release depth and theories of dynamics and dissolution on the vertical distribution of dissolved CO2 are studied. When deformation is not included in the calculations of drop terminal velocities, an overestimation of the vertical spread of dissolved CO2 might be made. Releasing CO2 near the critical depth leads to a narrower vertical range of dissolved CO2, making release depth an important factor influencing the vertical spread in the ocean. To study statistical probability distributions of drops, a database with information about the mass loss of numerous drops having different initial drop sizes was generated. A simple Matlab program was then developed to extract data from the database and provide different probability distributions. Normal and lognormal distributions were studied with varying standard deviations. Comparing the two showed that the distributions of dissolved CO2 in the water column were equal when a small standard deviation was used. With a larger standard deviation the vertical spread was greater with the lognormal distribution than with the normal probability distribution. The vertical spread is especially sensitive to the presence of large drops in the drop distribution.

%B Master Thesis No. 52 %I Geophysical Institute, University of Bergen %8 2004 %@ 82-8088-415-7 (electronic version) %9 Master Thesis %M 52 %> https://www.nersc.no/sites/www.nersc.no/files/Masteroppgave-gangstÃ¸.pdf