Modeling the release of CO2 in the deep ocean

Abstract

The idea of capturing and disposing of carbon dioxide (CO2) from the flue gas of fossil fuel-fired power plants has recently received attention as a possible mitigation strategy to counteract potential global warming due to the "greenhouse effect." One specific scheme is to concentrate the CO2 in the flue gas to over 90 mol %, compress and dehydrate the CO2 to supercritical conditions, and then transport it through a pipeline for deep ocean disposal. In Golomb et al. (1989), this scheme was studied, with emphasis on the CO 2 capture aspects. In this follow-on study, we concentrate on the mechanisms of releasing the CO 2 in the deep ocean.

Golomb et al. only considered the release of individual liquid CO 2 droplets in the region below 500 m. In this study, we consider all depths in both the liquid and vapor regions, and we model the entire plume in addition to individual droplets or bubbles. The key design variables in the model that can be controlled are: (1) release depth, (2) number of diffuser ports, N, and (3) initial bubble or droplet radius, ro. The results show that we can lower the height of the plume by increasing the number of diffuser ports and/or decreasing the initial bubble or droplet radius. Figure S-1 summarizes the results for a release depth of 500 m. With reasonable values for N and r. of 10 and 1 cm respectively, we can keep the plume height under 100 m. Since our goal is to dissolve all the CO2 before it reaches the well-mixed surface layer at approximately 100 m, we can release our C02 at depths as shallow as 200 m. However, the residence time of the sequestered CO2 in the ocean is also a function of depth. For releases of CO2 less than 500 m deep, we can estimate a residence time of less than 50 years, and for a release from about 1000 m, a residence time from 200 to 300 years. These residence times may be increased by releasing in areas of downwelling or by forming solid CO 2-hydrates which will sink to the ocean floor. For depths greater than 500 m, CO2-hydrates may form but we have ignored them due to lack of data.

We estimate that the local CO2 concentration will increase about 0.2 kg/m 3 . Added to the background concentration of 0.1 kg/m 3 , the resulting total concentration will be about 0.3 kg/m 3 , much less than saturation levels of about 40 kg/m 3 . Similarly, SO2 and NOx concentration increases will be about 1 .10 - 3 kg/m3 and 2 10- 4 kg/m 3 , respectively. Given an ambient current of 10 cm/s, horizontal dispersion will dilute these concentration increases by a factor of two at a distance of about 4 km downstream.

In implementing a CO2 capture and sequester scheme based on an air separation/ flue gas recycle power plant, the price of electricity would double. The reasons for this doubling are: (1) 44% due to derating of the power plant because of the parasitic power required to capture C02, mainly for air separation and CO compression, (2) 42% due to capital charges and operation and maintenance costs (excluding fuel) of the power plant modifications, including air separation and CO2 compression, and (3) 14% due to capital charges and operation and maintenance costs of a 160 km pipeline for deep ocean disposal. These numbers assume that no additional control measures are required to mitigate potential environmental problems are associated with deep ocean disposal of CO02.