5:15 PM - 7:15 PM
[SCG54-P17] Fluorescence imaging of the water penetration process in a heterogeneous rock including fractures and intrusions
Keywords:water penetration process, fluorescence imaging, heterogeneous rock
Introduction
The water penetration process into a rock mass is important to understand the relationship between fault slip and water. In addition to understanding the seismic process, knowledge of the relationship between rock medium and water is also important for underground utilization, such as geological disposal of radioactive waste and CO2 geological storage. Granite is the primary rock that composes the upper crust. Various previous studies mainly handled granite samples that uniformly include mineral grains. In this study, we take another step and examine the water penetration process in a heterogeneous sample.
Methods
We prepare a block of granitic gneiss rock (200 x 110 x 320 mm) that naturally mixed granite and gneiss part altered from granite. We extract five cores from the block sample, each 40 mm in diameter and 80 mm in height. Because of the large inner fracture in the block, some samples are divided into two. For comparison, we prepare standard rock samples that include uniformly distributed mineral grains: Inada granite, Ogino tuff, and Shimomura sandstone.
We use Methyl Methacrylate mixed with fluorescent substance and red dye (fluorescent MMA liquid) instead of water to visualize the liquid penetrating process. We prepare a black observation box and put the fluorescent liquid into a petri dish up to 5 mm from the bottom. Then, a rock sample is put into the liquid and left to take a photo every 30 minutes under ultraviolet light. Finally, fluorescent MMA liquid hardens in a high-temperature oven. After we hardened the fluorescent part, we made three thin sections inside the sample.
Results and discussion
The liquid penetrated into all samples. The penetration heights are 43.7 mm, 25.9 mm, and 20.5 mm for each of Inada granite, Ogino tuff, and Shimomura sandstone. The time-lapse images estimate penetration rates as 0.05 mm/sec for a granite and 0.08 mm/sec for a tuff. Regarding granitic gneiss, the liquid sometimes penetrates entire samples, while the liquid does not penetrate other samples. The granitic gneiss sample that penetration stopped in the middle of the height shows fast penetration in the granite matrix, and little penetration in the gneiss matrix. All samples show fast penetration at first and gradually approaching a constant rate.
Here, the liquid penetrates the samples without any external force, and the capillary phenomenon plays a role. The strength of the capillary phenomenon depends on both the void size and its continuity. Since granite and tuff include thinner pore sizes than sandstone, the higher penetration height of these rocks is consistent with general knowledge of capillary phenomenon. Past studies indicate liquid passes through the mineral boundary and sometimes liquid intrusion into the plagioclase grains. We observe the thin section images but do not find the liquid intrusion into the plagioclase. A thin section contains a relatively long fracture about 15 mm long that splits the grains. Around the fracture, granite is prominent on one side, and gneiss is prominent on the other side, but liquid only penetrates into the fracture plane and does not show the penetration between the grain boundaries.
Considering the observations, granite usually penetrates more liquid than gneiss, and a large fracture prevents liquid from penetrating grain boundaries.
The difference in the penetration process depending on the heterogeneity indicates the possibility that utilization of structural difference will help capture the liquid inside the bedrock.
The water penetration process into a rock mass is important to understand the relationship between fault slip and water. In addition to understanding the seismic process, knowledge of the relationship between rock medium and water is also important for underground utilization, such as geological disposal of radioactive waste and CO2 geological storage. Granite is the primary rock that composes the upper crust. Various previous studies mainly handled granite samples that uniformly include mineral grains. In this study, we take another step and examine the water penetration process in a heterogeneous sample.
Methods
We prepare a block of granitic gneiss rock (200 x 110 x 320 mm) that naturally mixed granite and gneiss part altered from granite. We extract five cores from the block sample, each 40 mm in diameter and 80 mm in height. Because of the large inner fracture in the block, some samples are divided into two. For comparison, we prepare standard rock samples that include uniformly distributed mineral grains: Inada granite, Ogino tuff, and Shimomura sandstone.
We use Methyl Methacrylate mixed with fluorescent substance and red dye (fluorescent MMA liquid) instead of water to visualize the liquid penetrating process. We prepare a black observation box and put the fluorescent liquid into a petri dish up to 5 mm from the bottom. Then, a rock sample is put into the liquid and left to take a photo every 30 minutes under ultraviolet light. Finally, fluorescent MMA liquid hardens in a high-temperature oven. After we hardened the fluorescent part, we made three thin sections inside the sample.
Results and discussion
The liquid penetrated into all samples. The penetration heights are 43.7 mm, 25.9 mm, and 20.5 mm for each of Inada granite, Ogino tuff, and Shimomura sandstone. The time-lapse images estimate penetration rates as 0.05 mm/sec for a granite and 0.08 mm/sec for a tuff. Regarding granitic gneiss, the liquid sometimes penetrates entire samples, while the liquid does not penetrate other samples. The granitic gneiss sample that penetration stopped in the middle of the height shows fast penetration in the granite matrix, and little penetration in the gneiss matrix. All samples show fast penetration at first and gradually approaching a constant rate.
Here, the liquid penetrates the samples without any external force, and the capillary phenomenon plays a role. The strength of the capillary phenomenon depends on both the void size and its continuity. Since granite and tuff include thinner pore sizes than sandstone, the higher penetration height of these rocks is consistent with general knowledge of capillary phenomenon. Past studies indicate liquid passes through the mineral boundary and sometimes liquid intrusion into the plagioclase grains. We observe the thin section images but do not find the liquid intrusion into the plagioclase. A thin section contains a relatively long fracture about 15 mm long that splits the grains. Around the fracture, granite is prominent on one side, and gneiss is prominent on the other side, but liquid only penetrates into the fracture plane and does not show the penetration between the grain boundaries.
Considering the observations, granite usually penetrates more liquid than gneiss, and a large fracture prevents liquid from penetrating grain boundaries.
The difference in the penetration process depending on the heterogeneity indicates the possibility that utilization of structural difference will help capture the liquid inside the bedrock.