Drainage and Imbibition CO2/Brine Relative Permeability Curves at Reservoir Conditions for Carbonate Formations

Abstract

Proper modeling of the multiphase flow of supercritical CO 2 in deep saline aquifers for CO 2 sequestration (both cycles of drainage during injection and imbibition during CO 2 migration) is critical in being able to understand and predict both the short and long term fate of the injected CO 2 over extended time periods (hundreds to thousands of years). Current numerical models require the use of accurate two-phase CO 2 /brine relative permeability data at representative in-situ conditions in order to be able to accurately conduct these calculations. However, there are virtually no published data in the literature on the high temperature and pressure displacement character of CO 2 /brine systems in actual reservoir rocks, except for the data published by the authors in the June 2008 issue of the SPE Reservoir Evaluation and Engineering Journal. That data set, although it included a few carbonate cases, contained mostly measurements on clastic rocks. This paper presents a new set of nine relative permeability measurements (both drainage and imbibition) for carbonate rocks (limestone and dolomite) of higher permeability values than those in the initial work (which are thus more likely to be representative for candidates for CO 2 sequestration in deep saline aquifers). The new data set to be presented includes also pre and post-test CAT-scan imaging of selected samples to illustrate potential effects of CO 2 contact on potentially soluble carbonate matrices. The paper compares the new data set of measurements for carbonate rocks with the limited set of data available for carbonates from the previous work, and attempts to determine if specific relative permeability and residual saturation trends can be defined based on other rock characteristics that are easier to measure in routine core analyses, to allow extension of the data set to other carbonate facies elsewhere which have not been tested.

Review of CO2 storage efficiency in deep saline aquifers

“…However, water imbibition back in the pore space in the wake of a CO 2 plume reduces the CO 2 saturation, reaching a minimum value at irreducible CO 2 saturation, S CO 2 irr , such that S CO 2 irr ≤ E pore ≤ (1 − S swirr ). Bennion and Bachu (2010) and Bachu (2013) report a wide variability in irreducible saturations: 0.2108 ≤ S wirr ≤ 0.852, and 0.045 ≤ S CO 2 irr ≤ 0.415 based on laboratory measurements of relative permeability characteristics of CO 2 /brine systems for 29 sandstone and carbonate rock samples. This shows that storage efficiency at the pore scale varies in a wide range (by a factor approaching 20), and, as it will be seen later, correspondingly affects storage efficiency at larger scales.…”

Subsurface geochemical fate and effects of impurities contained in a CO2 stream injected into a deep saline aquifer: What is known

“…The displacement of mobile water from the pore space in the CO 2 -plume region will continue until the water content is reduced to irreducible saturation (Sw irr ). Values of Sw irr can vary greatly between rocks; Bennion and Bachu, (2010) and Bachu (2013) report values ranging from 20% to 85% for different rock samples from the Alberta basin, Canada. Water will also be mobilized from this zone by vaporization into the CO 2 phase, although this flux may be partially compensated with a return flow through the water-wet surfaces (Pruess and Müller, 2009).…”

Stochastic and global sensitivity analyses of uncertain parameters affecting the safety of geological carbon storage in saline aquifers of the Michigan Basin

“…! "# data of carbonates/dolomites (Bennion and Bachu, 2008;Bennion and Bachu, 2010), exponential regression (in black), and the half amplitude of confidence interval (in grey). simulations of the hypothetical system based on the Michigan Basin for scenarios S1, S2, and S3 (Table 3).…”

“…"# and !,!! data corresponding to carbonate/dolomite formations are retrieved from the works of Bennion and Bachu (Bennion and Bachu, 2008;Bennion and Bachu, 2010). With these data, a non-linear exponential regression between !…”

“…!"! data obtained from Bennion and Bachu (Bennion and Bachu, 2008;Bennion and Bachu, 2010) together with the exponential regression and the half amplitude confidence intervals.…”