Evaluation of Advanced Reactive Surface Area Estimates for Improved Prediction of Mineral Reaction Rates in Porous Media

Thursday, 17 December 2015: 15:25
3018 (Moscone West)
Lauren E Beckingham1, Elizabeth H. Mitnick2, Shuo Zhang3, Marco Voltolini1, Li Yang4, Carl I Steefel1, Alexander Swift5, David R Cole5, Julie Sheets6, Timothy J Kneafsey1, Gautier Landrot7, Lawrence M Anovitz8, Saeko Mito9, Ziqiu Xue9, Jonathan Blair Ajo Franklin1 and Don DePaolo1, (1)Lawrence Berkeley National Laboratory, Berkeley, CA, United States, (2)University of California Berkeley, Berkeley, CA, United States, (3)UC Berkeley, Berkeley, CA, United States, (4)Lawrence Berkeley National Lab, Berkeley, CA, United States, (5)Ohio State University Main Campus, Columbus, OH, United States, (6)The Ohio State University, Columbus, OH, United States, (7)Kasetsart University, Bangkok, Thailand, (8)ORNL U Tennessee, Oak Ridge, TN, United States, (9)RITE Research Institute of Innovative Technology for the Earth, Kyoto, Japan
CO2 sequestration in deep sedimentary formations is a promising means of reducing atmospheric CO2 emissions but the rate and extent of mineral trapping remains difficult to predict. Reactive transport models provide predictions of mineral trapping based on laboratory mineral reaction rates, which have been shown to have large discrepancies with field rates. This, in part, may be due to poor quantification of mineral reactive surface area in natural porous media. Common estimates of mineral reactive surface area are ad hoc and typically based on grain size, adjusted several orders of magnitude to account for surface roughness and reactivity. This results in orders of magnitude discrepancies in estimated surface areas that directly translate into orders of magnitude discrepancies in model predictions. Additionally, natural systems can be highly heterogeneous and contain abundant nano- and micro-porosity, which can limit connected porosity and access to mineral surfaces. In this study, mineral-specific accessible surface areas are computed for a sample from the reservoir formation at the Nagaoka pilot CO2 injection site (Japan). Accessible mineral surface areas are determined from a multi-scale image analysis including X-ray microCT, SEM QEMSCAN, XRD, SANS, and SEM-FIB. Powder and flow-through column laboratory experiments are performed and the evolution of solutes in the aqueous phase is tracked. Continuum-scale reactive transport models are used to evaluate the impact of reactive surface area on predictions of experimental reaction rates. Evaluated reactive surface areas include geometric and specific surface areas (eg. BET) in addition to their reactive-site weighted counterparts. The most accurate predictions of observed powder mineral dissolution rates were obtained through use of grain-size specific surface areas computed from a BET-based correlation. Effectively, this surface area reflects the grain-fluid contact area, or accessible surface area, in the powder dissolution experiment. In the model of the flow-through column experiment, the accessible mineral surface area, computed from the multi-scale image analysis, is evaluated in addition to the traditional surface area estimates.