Main content area

CO2-induced dissolution of low permeability carbonates. Part I: Characterization and experiments

Smith, Megan M., Sholokhova, Yelena, Hao, Yue, Carroll, Susan A.
Advances in water resources 2013 v.62 pp. 370-387
X-radiation, calcite, calcium, carbon dioxide, carbon sequestration, carbonates, dolomite, dolomitic limestone, equations, image analysis, magnesium, mass transfer, models, oils, pH, permeability, porosity, scanning electron microscopy, thermodynamics, water resources, Canada
The effect of elevated dissolved CO2 concentrations on compositionally and structurally distinct carbonate sample cores from the Weyburn–Midale CO2-enhanced oil recovery and storage site (Canada) was measured from analysis of 3-D sample characterization and fluid chemistry data from core-flood experiments. Experimental conditions (60°C; 24.8MPa confining pressure) and brine composition were chosen to mimic in situ reservoir conditions. Mineralogy and pore space distributions within the eight individual cores were characterized with X-ray computed microtomography and scanning electron microscopy both before and after exposure to brine with 0.5⩽pCO2⩽3MPa, while solution chemistry and differential fluid pressures were monitored during experiments.Our experimental study aimed to quantify the relationship between fluid flow, heterogeneity, and reaction specific to carbon storage at the Weyburn–Midale field by integrating characterization imaging, pressure data, and solution chemistry. Through the use of non-invasive microtomographic imaging, a variety of dissolution behaviors were observed, with variable effects on the evolution of solution chemistry and permeability as a result of heterogeneity within these two relatively low permeability carbonate samples. Similar-sized, evenly distributed pores, and steadily advancing dissolution fronts suggested that uniform flow velocities were maintained throughout the duration of the higher permeability “Marly” dolostone core experiments. The development of unstable dissolution fronts and fast pathways occurred in the “Vuggy” sample experiments when fluid velocities varied widely within the sample (as a result of increased pore structure heterogeneity). The overall effect of fast pathway development was to increase bulk permeability values by several orders of magnitude, allowing CO2-acidified fluids to travel through the cores largely unmodified by carbonate mineral reaction, as indicated by a lack of change in later-time solution pH levels at the core outlet. Given the impact of heterogeneity within low permeability cores, effort should be taken to incorporate smaller-scale heterogeneity into predictive models and such an averaging approach (utilizing the data and observations discussed here) is the topic of our companion manuscript (see Hao et al., 2013).Solution chemistry results indicated that steady-state carbonate mass transfer conditions were attained in the Marly dolostone experiments and during the earlier (pre-pressure breakthrough) portions of the Vuggy limestone experiments. Steady-state calcium and magnesium concentrations coincided with outlet solutions that were calculated to be at or very near to equilibrium with respect to both calcite and dolomite, relative to available thermodynamic data and considering experimental data scatter. Carbonate mass transfer data were evaluated against a variety of proposed carbonate dissolution mechanisms, including both pH- and pCO2-dependent expressions as well as a simplified pH-independent formulation. Based on this analysis, the calcite reaction rate coefficient was estimated to be ∼17 times faster than that for dolomite dissolution under our experimental conditions. This ratio is consistent with the use of rate equations that depend on carbonate mineral saturation without specifying additional dependence on solution pH or CO2 levels, and may be a result of the narrow experimental pH range. In addition, solution chemistry data were combined with time-dependent pressure data to constrain the exponent in a power-law expression describing the relationship between evolving porosity and permeability within the Vuggy limestones. This relationship as well as proposed carbonate kinetic expressions are further evaluated in our companion paper (see Hao et al., 2013).