HR: 0800h
AN: OS41C-0495    [Abstracts]
TI: Failure of Marine Sediments due to Gas Hydrate Dissociation
AU: * Germanovich, L
EM: leonid@ce.gatech.edu
AF: School of Civil and Environmental Engineering, Georgia Institute of Technology, Atlanta, GA 30332-0355 United States
AU: Xu, W
EM: wenyue.xu@eas.gatech.edu
AF: School of Earth and Atmospheric Sciences, Georgia Institute of Technology, Atlanta, GA 30332-0340 United States
AB: Methane gas hydrate (MGH) dissociation in the pore space of marine sediments may be caused by various natural and human-induced processes including sea level decrease, tectonic uplift of continental margins, global warming, and petroleum operations. While these processes generally have different spatial and temporal scales, they result in MGH dissociation, and the released gas and water tend to expand. This may change the pore pressure in the sediments, affecting their mechanical state and failure processes. If the pressure does not change, the hydrate dissociation may still affect the sediment properties by perturbing particle cementation and by introducing phase interfaces (e.g., capillary menisci). In this work, the pressure change has been calculated by coupling the dissociation rate with fluid flow in the sediments based on thermodynamic considerations. The common seafloor failure, submarine landslides, can reach a length of $\sim$100 km, with a length-to-thickness ratio as large as $\sim$1000. It is often assumed that the Storegga Slides were caused by earthquakes that instantaneously created a shallow discontinuity ($\sim$100 m below the seafloor) along the entire slide length of $\sim$100 km. Instead, {\it Puzrin and Germanovich} [2004] reasoned that the MGH dissociation may have resulted in an initial flaw at the scale of only $\sim$1 km. They explained the landslide evolution in submarine slopes by the mechanism of catastrophic shear band propagation of this flaw. Our modeling suggests that the sediment de-cementation and the excess pore pressure due to MGH dissociation may indeed have determined the scale of $\sim$1 km of this initial defect. Our calculations also suggest that dissociation-affected submarine landslides may be common for shallow sea water depths of $<$ 1 km and involve thin sediment layers (usually $\sim$100 m or less). However, the MGH dissociation may also occur underneath a massive and horizontally extended MGH layer, which could serve as a seal or cap-rock. In this case, the excess pressure can be as high as tens of MPa if the sediment permeability is much lower than $\sim$10$^{-16}$ m$^{2}$. Therefore, an excess pore pressure, sufficient for sediment fracturing, may occur at water depths of $>$1 km. In particular, we argue that the seafloor collapse structure at the Blake Ridge site [{\it Dillon et al.}, 2001] may be explained by this mechanism. Alternatively, the excess pore pressure may be sufficient to initiate vertical hydraulic fractures above the dissociation area [{\it Zuhlsdorff and Spieb}, 2004]. We hypothesize that the MGH dissociation occurring in the pore space and MGH lenses [{\it Suess et al.}, 1999] supplies the growing fissures with the fracturing fluid. In this scenario, the phase transition in the pore space is due to the pressure decrease (rather than increase), which further enhances the dissociation and fluid supply. Our calculations suggest that once the dissociation-driven fractures reach a size of up to tens of meters, they separate from the originating MGH region and propagate towards the seafloor. Similar to the models of deep magma transport, this propagation is essentially due to buoyancy. We argue that such periodic episodes of hydraulic fracturing may explain methane ejections, mud volcanoes, and pockmarks observed on the seafloor. Furthermore, the fractures are likely to migrate along closely spaced trajectories. Eventually, quasi-vertical elongated regions of disturbed sediment form. Their horizontal dimensions are unlikely to exceed $\sim$100 m, which is consistent with the quasi-vertical channels discovered in seismic studies.
DE: 8010 Fractures and faults
DE: 8045 Role of fluids
DE: 5104 Fracture and flow
DE: 3022 Marine sediments--processes and transport
SC: Ocean Sciences [OS]
MN: 2004 AGU Fall Meeting