Abstract
The Nordland shale forms the caprock of the Utsira sands of the Sleipner reservoir currently used for carbon dioxide sequestration. The long-term exposure of shale rocks to supercritical carbon dioxide (scCO2), or scCO2-brine mixtures, may lead to structural and chemical changes in shale that lead to increases in permeability of inter-layers and caprocks, that may mean changes to plume migration behaviour and/or loss of seal efficiency of caprocks. A detailed study has been made of the initial pore structure of Nordland shale and the changes following accelerated treatment with scCO2. Gas sorption scanning curves have suggested that the void space of the original shale consisted of a Network (denoted 1) of micropores and smaller mesopores that is thermodynamically independent of a Network (2) of larger mesopores and macropores. This work introduces a new iodononane pre-adsorption technique to map the macroscopic (>microns) spatial distribution of micropores (<2 nm) and smaller mesopores in shales using CXT. CXT imaging of shale samples with iodononane pre-adsorbed in Network 1, or with entrapped mercury confined to only Network 2, suggested that both small- and large-sized pore networks were pervasive through the shale and associated with the continuous illite matrix phase. The feldspar and quartz grains did not form part of either network, though inter-particle macropores were found surrounding these mineral grains from CXT imaging of mercury entrapped there. Kinetic gas uptake experiments conducted on samples before and after filling Network 1 with iodononane suggested that the smaller mesopores were, despite their small size and thermodynamic independence from the macropores, still critical to mass transport, with the diffusion flux being funnelled through them. Shale surface areas obtained using the homotattic patch adsorption model were found more physically realistic than those determined via the ISO BET method since multi-linear regression of only the logarithm of the former, together with that of the Network 1 pore volume, predicted the gas-phase mass transport coefficient following treatment. This work demonstrated the need for the novel characterisation methods and data analysis presented here to properly understand the structure-transport relationship in shales exposed to scCO2.
NSTL