Drying and salt precipitation in geological formations can have serious consequences for upstream operations in terms of injectivity and productivity. Here we investigate the consequences of supercritical CO 2 injection in sandstones. The reported findings are directly relevant for CO 2 sequestration and acid-gas injection operations, but might also be of interest to a broader community dealing with drying and capillary phenomena. By injecting dry supercritical CO 2 into brine-saturated sandstone, we investigate the drying process and the associated precipitation of salts in a capillary-pressure-dominated flow regime. Precipitation patterns were recorded during the drying process by means of CT scanning. The experimental results and numerical simulations show that under a critical flow rate salt precipitates with an inhomogeneous spatial distribution because of brine and solutes being transported in counter-current flow upstream where salt eventually precipitates. A substantial impairment of the absolute permeability has been found, but despite high local salt accumulation, the effective CO 2 permeability increased during all experiments. This phenomenon is a result of the observed microscopic precipitation pattern and eventually the resulting K() relationship. The findings in this paper are related to unimodal sandstone. In a companion paper (Ott et al., 2014) we present data on the distinctly different consequences of salt precipitation in dual-or multi-porosity rocks.

Virus transport in porous media is affected by the water flow regime. During transient, variably saturated flow, fluctuating flow regimes can enhance virus detachment from both solid–water interfaces (SWIs) and air–water interfaces (AWIs). The objective of this study was to simulate the influence of drainage and imbibition events on the remobilization of attached viruses. Three different modeling approaches were examined. In the first approach, all attachment and detachment coefficients were assumed to be constant, but the values of the detachment coefficients were increased drastically for the duration of transient, unsaturated flow. The second and third modeling approaches involved extensions of the model of Cheng and Saiers, who assumed enhanced detachment of viruses to be proportional to the time rate of change in the water content. Their model was extended to include separate terms for virus attachment–detachment on SWIs and AWIs. In our second approach, we assumed kinetic sorption onto the AWI, with the desorption rate being described as a function of temporal changes in the air content. This approach did not explicitly account for the specific air–water interfacial area. Thus, in our third approach we explicitly included the presence and variation of air–water interfaces and assumed AWI attachment–detachment to be an equilibrium sorption process. The available air–water interfacial area was assumed to be a function of fluid saturation. The models were used to simulate a series of saturated–unsaturated virus transport experiments reported in the literature for conditions of both drainage and imbibition. The most promising results were obtained with the third approach, which explicitly accounts for adsorption to air–water interfaces and assumes equilibrium sorption on the available air–water interfacial area.

No abstract

A major challenge of CO 2 injection into saline aquifers is the risk of formation clogging due to salt precipitation. Capillary-driven flow of brine can provide a continuous transport of dissolved salt toward the dry zone around the injection well where it ultimately precipitates due to evaporation. In this study, core flooding experiments were performed in homogeneous coarse-textured cores and in layered cores consisting of a coarse-textured layer overlying a fine-textured layer. CO 2 was injected through a well in the upper part of the cores, and the bottom parts functioned as brine sources. Impairment in injectivity was found due to accumulation of precipitated salt caused by capillary-driven flow from the brine sources to the upper dryer region. Compared to flow domains without a brine source, we found that capillary-driven upward flow at first prevents complete clogging because the porous medium remains wet, but eventually leads to a more severe clogging of the entire domain. The results show that after sufficient dry-out, a coarse-textured injection layer can draw brine from an underlying fine-textured layer by capillary forces. A connected fine-textured layer can therefore contribute to salt precipitation and clogging of the injection layer.