(Poster session/Short oral-presentation) Earth & Resources

[2105-19-04] [Student presentation: Doctoral course] Modeling permeability alteration in simulated seawater flow-through experiments

○Achmad Hafidz¹, Naoki Kinoshita¹, Sho Ogata², Hideaki Yasuhara³ (1. Ehime University, 2. Osaka University, 3. Kyoto University)

Chairperson: Naoki Kinoshita (Ehime University)

Keywords:permeability, simulated seawater, parameter

A numerical model was developed to predict permeability change due to the interaction between fractured granite rock and simulated seawater. The accuracy of the proposed model was checked by comparing our simulation results with the experimental results taken from simulated seawater flow-through experiments at two temperature conditions, 20^oC (E-25) and 90^oC (E-27). The results showed that the proposed model was able to replicate the permeability change of E-25 very well. However, using the same input parameters except for temperature, the proposed model was unable to predict the permeability change of E-25 very well because the permeability decrement in the simulation was considerably greater than that of the experiment. At the beginning of the simulation, the decrease in permeability was dominantly controlled by pressure dissolution. However, in the middle of the simulation, the decrease in permeability was controlled by pressure dissolution and dawsonite secondary mineral precipitation. This dawsonite precipitation occurred because of the increase in Al element concentration supplied from pressure dissolution with a decrease in fracture aperture. Thus, to improve the prediction of permeability change E-27, slight input parameter modification was made by changing the a parameter, which was related to the relationship between average mechanical aperture and contact area ratio. The results showed that no significant reduction in permeability occurred since the contact area ratio increased rapidly with a slight decrease in fracture aperture due to the change in a parameter. Therefore, all input parameters used in the simulation must be accurate.