Pore-scale multiphase flow modeling and imaging of CO2 exsolution in Sandstone

doi:10.1016/j.petrol.2016.10.011

This project addressed the scaling of geochemical reactions to core and field scales, and the interrelationship between reaction rates and flow in porous media. We targeted reactive transport problems relevant to the Hanford site - specifically the reaction of highly caustic, radioactive waste solutions with subsurface sediments, and the immobilization of 90Sr and 129I through mineral incorporation and passive flow blockage, respectively. We addressed the correlation of results for pore-scale fluid-soil interaction with field-scale fluid flow, with the specific goals of (i) predicting attenuation of radionuclide concentration; (ii) estimating changes in flow rates through changes of soil permeabilities; and (iii)more » estimating effective reaction rates. In supplemental work, we also simulated reactive transport systems relevant to geologic carbon sequestration. As a whole, this research generated a better understanding of reactive transport in porous media, and resulted in more accurate methods for reaction rate upscaling and improved prediction of permeability evolution. These scientific advancements will ultimately lead to better tools for management and remediation of DOE’s legacy waste problems. We established three key issues of reactive flow upscaling, and organized this project in three corresponding thrust areas. 1) Reactive flow experiments. The combination of mineral dissolution and precipitation alters pore network structure and the subsequent flow velocities, thereby creating a complex interaction between reaction and transport. To examine this phenomenon, we conducted controlled laboratory experimentation using reactive flow-through columns. Results and Key Findings: Four reactive column experiments (S1, S3, S4, S5) have been completed in which simulated tank waste leachage (STWL) was reacted with pure quartz sand, with and without Aluminum. The STWL is a caustic solution that dissolves quartz. Because Al is a necessary element in the formation of secondary mineral precipitates (cancrinite), conducting experiments under conditions with and without Al allowed us to experimentally separate the conditions that lead to quartz dissolution from the conditions that lead to quartz dissolution plus cancrinite precipitation. Consistent with our expectations, in the experiments without Al, there was a substantial reduction in volume of the solid matrix. With Al there was a net increase in the volume of the solid matrix. The rate and extent of reaction was found to increase with temperature. These results demonstrate a successful effort to identify conditions that lead to increases and conditions that lead to decreases in solid matrix volume due to reactions of caustic tank wastes with quartz sands. In addition, we have begun to work with slightly larger, intermediate-scale columns packed with Hanford natural sediments and quartz. Similar dissolution and precipitation were observed in these colums. The measurements are being interpreted with reactive transport modeling using STOMP; preliminary observations are reported here. 2) Multi-Scale Imaging and Analysis. Mineral dissolution and precipitation rates within a porous medium will be different in different pores due to natural heterogeneity and the heterogeneity that is created from the reactions themselves. We used a combination of X-ray computed microtomography, backscattered electron and energy dispersive X-ray spectroscopy combined with computational image analysis to quantify pore structure, mineral distribution, structure changes and fluid-air and fluid-grain interfaces. Results and Key Findings: Three of the columns from the reactive flow experiments at PNNL (S1, S3, S4) were imaged using 3D X-ray computed microtomography (XCMT) at BNL and analyzed using 3DMA-rock at SUNY Stony Brook. The imaging results support the mass balance findings reported by Dr. Um’s group, regarding the substantial dissolution of quartz in column S1. An important observation is that of grain movement accompanying dissolution in the unconsolidated media. The resultant movement changes the anticipated findings for pore and throat size distributions. For column S3, with cancrinite precipitation accompanying quartz dissolution, the precitiation halts much of the grain movement and more systematic distributions are obtained. Column S4, which was sealed with caustic solution acted as a control sample to study reactive effects during periods when columns S1 and S3 were sealed between flow experiments. No significant changes are observed in S4 with time. At Princeton, the imaging and analysis work focused on the effects of mineral precipitation and advancing our understanding of the impacts of these reactions on reactive transport in subsurface sediments. These findings are described in detail below, and have been published in L.E. Crandell, C.A. Peters, W. Um, K.W. Jones, W.B. Lindquist, 2012. “Changes in the pore network structure of Hanford sediment after reaction with caustic tank wastes.” Journal of Contaminant Hydrology 131 (2012) 89–99. 3) Multi-Scale Modeling and Up-Scaling. Using an array of modeling approaches, we examined pore-scale variations in physical and mineralogical properties, flow velocities, and (for unsaturated conditions) wetting fluid/grain surface areas, and permeability evolution. Results and Key Findings: To predict the column permeability and estimate the impact of mineral precipitation, pore network models were informed using the pore and throat-size distributions from the imaging analyses. As a comparison, supplemental analyses were performed on Viking sandstone specimens from the Alberta sedimentary basin. In another part of this study we sought to understand how carbonate rocks in contact with CO2-rich brines change due to the precipitation or dissolution of fast-reacting minerals such as calcite and dolomite. Using a newly developed reactive-transport pore-network model we were able to identify the conditions that lead to significant permeability changes. These findings are presented below and are compiled in a publication that is under review: J.P. Nogues, J.P. Fitts, M.A. Celia, C.A. Peters. “Permeability evolution due to dissolution and precipitation of carbonates using reactive transport modeling in pore networks”, Submitted: Water Resources Research, 2013.

Title: Pore-scale multiphase flow modeling and imaging of CO2 exsolution in Sandstone

Citation Formats