10:00 AM - 10:15 AM
[HSC06-05] Experimental study of CO2–brine–rock interactions in mudstone from the Kazusa Group, Japan
As a major source of CO2 emissions, the Tokyo Bay coastal area requires robust CO2 mitigation strategies. The Pleistocene Kazusa Group underlying this region is a promising target for geological CO2 storage (GCS). In GCS, CO2–brine–rock interactions can lead to mineral dissolution and precipitation. Geochemical reactions with CO2 may alter the seal capacity of caprock mudstone, necessitating a study of its reactivity for safe storage. Despite numerical simulations, there are no experimental studies on the geochemical reactivity of the Kazusa Group. In this study, we determined the mineral composition of Kazusa Group mudstone and experimentally assessed its geochemical reactions with CO2.
We used mudstone samples from the Otadai and Kiwada Formations of the Kazusa Group in the Boso Peninsula, central Japan. The mudstone from the Otadai Formation represents a typical caprock in the Kazusa Group, while the mudstone from the Kiwada Formation, interbedded between two tephra layers, is particularly rich in volcanic glass. Given the variable reactivity of volcanic glass and plagioclase with CO2 by elemental composition, we used an electron probe microanalyzer (EPMA) for measurement. Volcanic glass content was determined from EPMA element mapping. Other mineral composition was identified using X-ray diffraction (XRD) and combined with volcanic glass to sum to 100%.
Batch reaction experiments were conducted with powdered and plug samples. The powdered samples, roughly ground using an agate mortar to avoid breaking the grains, weighed 3 g, while the plug samples, molded to 4.75 mm in diameter and 9 mm in length, weighed 0.26 g. We combined 30 mL of ultra-pure water with each rock sample in a vessel, heated it to 80 °C, and pressurized it to 10 MPa with CO2. Durations were 7, 14, and 28 days for powdered samples and 14 days for plug samples. Post-experiment samples were examined using XRD for powdered samples and a scanning electron microscope for plug samples.
Pre-experiment analysis revealed that the samples contain plagioclase (29–31%), quartz (19–28%), volcanic glass (9–25%), chlorite (5–9%), calcite (6–8%), K-feldspar (4–8%), mica (3–5%), and other minor minerals. Plagioclase exhibited a wide range of compositions between albite and anorthite. Volcanic glass was rhyolitic, containing 70–80% SiO2. Calcite appeared as calcareous fossils, dispersed within the matrix.
The results showed similar changes for both Otadai and Kiwada Formation samples. In the plug samples, complete calcite dissolution and slight anorthite surface dissolution were observed, without changes in other minerals. XRD results of powdered samples revealed a reduction in calcite content but no other notable changes in mineral content or crystallinity. Slight precipitation of layered double hydroxide occurred in the powdered samples, but no carbonate minerals precipitated.
The Kazusa Group mudstone contains calcite that is highly reactive with CO2. Although rapid dissolution was confirmed, the amount of calcite is small, and scattered fossil dissolution is unlikely to markedly modify pore structure, indicating minor impact on seal capacity. Plagioclase and chlorite, which are relatively reactive following calcite, showed no significant dissolution. Slight dissolution of anorthite, the more reactive plagioclase end member, was observed in the plug samples, but no changes were seen in albite, and the total plagioclase content remained unchanged in powdered samples, suggesting the impact of the plagioclase reaction is limited. Volcanic glass was rhyolitic with lower reactivity, and no dissolution was observed. Additionally, no carbonate mineral precipitated. In conclusion, Kazusa Group mudstone is unlikely to undergo significant property changes due to rapid geochemical reactions with CO2 in the short-term. For more long-term stability predictions, uncertainties can be reduced through further study, such as longer experiments or geochemical modeling.
We used mudstone samples from the Otadai and Kiwada Formations of the Kazusa Group in the Boso Peninsula, central Japan. The mudstone from the Otadai Formation represents a typical caprock in the Kazusa Group, while the mudstone from the Kiwada Formation, interbedded between two tephra layers, is particularly rich in volcanic glass. Given the variable reactivity of volcanic glass and plagioclase with CO2 by elemental composition, we used an electron probe microanalyzer (EPMA) for measurement. Volcanic glass content was determined from EPMA element mapping. Other mineral composition was identified using X-ray diffraction (XRD) and combined with volcanic glass to sum to 100%.
Batch reaction experiments were conducted with powdered and plug samples. The powdered samples, roughly ground using an agate mortar to avoid breaking the grains, weighed 3 g, while the plug samples, molded to 4.75 mm in diameter and 9 mm in length, weighed 0.26 g. We combined 30 mL of ultra-pure water with each rock sample in a vessel, heated it to 80 °C, and pressurized it to 10 MPa with CO2. Durations were 7, 14, and 28 days for powdered samples and 14 days for plug samples. Post-experiment samples were examined using XRD for powdered samples and a scanning electron microscope for plug samples.
Pre-experiment analysis revealed that the samples contain plagioclase (29–31%), quartz (19–28%), volcanic glass (9–25%), chlorite (5–9%), calcite (6–8%), K-feldspar (4–8%), mica (3–5%), and other minor minerals. Plagioclase exhibited a wide range of compositions between albite and anorthite. Volcanic glass was rhyolitic, containing 70–80% SiO2. Calcite appeared as calcareous fossils, dispersed within the matrix.
The results showed similar changes for both Otadai and Kiwada Formation samples. In the plug samples, complete calcite dissolution and slight anorthite surface dissolution were observed, without changes in other minerals. XRD results of powdered samples revealed a reduction in calcite content but no other notable changes in mineral content or crystallinity. Slight precipitation of layered double hydroxide occurred in the powdered samples, but no carbonate minerals precipitated.
The Kazusa Group mudstone contains calcite that is highly reactive with CO2. Although rapid dissolution was confirmed, the amount of calcite is small, and scattered fossil dissolution is unlikely to markedly modify pore structure, indicating minor impact on seal capacity. Plagioclase and chlorite, which are relatively reactive following calcite, showed no significant dissolution. Slight dissolution of anorthite, the more reactive plagioclase end member, was observed in the plug samples, but no changes were seen in albite, and the total plagioclase content remained unchanged in powdered samples, suggesting the impact of the plagioclase reaction is limited. Volcanic glass was rhyolitic with lower reactivity, and no dissolution was observed. Additionally, no carbonate mineral precipitated. In conclusion, Kazusa Group mudstone is unlikely to undergo significant property changes due to rapid geochemical reactions with CO2 in the short-term. For more long-term stability predictions, uncertainties can be reduced through further study, such as longer experiments or geochemical modeling.