Laboratory Experiments on Porous Media Mass Transport: Implications for Carbon Sequestration

Robert Ecke

Short Abstract
Carbon dioxide in the atmosphere has continued to increase throughout the 20th century and continues today. The consensus of the scientific community is that this CO2 increase has contributed substantively to the observed increase in global mean temperature over the last 80 years [1]. Further emissions in CO2 from the burning of fossil fuels will accelerate climate change, affecting regional climate in as yet unpredictable ways. To combat the ever-increasing emissions of CO2, a mitigating strategy is the long-term storage of CO2 in geologic reservoirs [2]. Such reservoirs consist of porous media of varying permeability (average pore size) and porosity (fluid fraction), and an accurate estimate of the rate of absorption of CO2 into such reservoirs is necessary to quantitatively evaluate the efficacy of this sequestration strategy. Roughly speaking, the approach consists of pumping high-pressure (above its critical pressure, i.e., supercritical) CO2 into the porous media at depth. The CO2 is less dense than the surrounding salt water and most of it rises until confined by impermeable cap rock. As more fluid is pumped into the reservoir it spreads laterally under the cap rock and slowly diffuses into the water (the saturation concentration of CO2 into water is about 3%). If diffusion were the only transport mechanism, the dissolution of most of the injected CO2 would take thousands of years. Fortuitously, CO2-saturated brine is heavier than water and can transport mass by the process of convection as well as by diffusion. To determine the storage potential for carbon sequestration strategies involving porous media, accurate determination of mass transport efficiency is required. We have made accurate measurements of mass transport in geometries similar to those relevant for sequestration, namely a gravitationally stable two-layer system where the diffusion interface between the two phases is unstable. The two fluids are water and propylene glycol (PPG) with water on top. In one case the porous media is modeled using the Hele- Shaw geometry with an adjustable gap width to vary the permeability (the porosity in this 2D case is 1). The enhanced mass transport efficiency Nu as a function of dimensionless forcing Rayleigh number Ra is determined with enhancements of up to 250 for Ra = 80,000. Geologic reservoirs are, however, three dimensional and typical Ra in nature are less than about 5000. Therefore, we also measured mass transport in a fully 3D cylindrical geometry for 150 < Ra < 5000 using the same fluids. We discovered a transition from a high mass-transport state to a low mass transport state that typically occurs between 4 and 6 convective times. The implications of these mass transport measurements for carbon sequestration are evaluated.