In the preceding section, we have determined the AVO behavior of two micromechanical models of hydrate deposition. We can now link the modeled internal hydrate structure with real seismic data from the Blake Outer Ridge, offshore Florida and Georgia. Since we assumed a very basic rock model in the rock-physics-based seismic modeling, we do not try to match the data quantitatively, but attempt to extract qualitatively parameters such as permeability and strength from the seismically-inferred rock structure.
Figure 11 shows seismic data from a hydrate deposit at the Blake Outer Ridge. The upper panel is an actual CMP gather containing the reflections of the seafloor, the BSR, and a ``flat'' reflector, probably the transition from gas to brine sediment Ecker and Lumley (1994). Above the BSR, there is a ``quiet zone'' where no reflections are visible. This might be caused by the presence of disseminated methane hydrate. It is obvious from the seismogram that there is no strong reflection visible from the top of the hydrate, while there seems to be a clear reflection from the bottom of the gas-saturated sediment layer. The lower panel compares the amplitude picked along the BSR with the theoretically calculated trend using Zoeppritz equations. The amplitudes display a significant increase of amplitudes with increasing offset. Performing a detailed velocity and AVO analysis, Ecker and Lumley 1994 showed that this behavior is compatible with hydrate overlaying gas-saturated sediment. The P-wave velocity above the BSR is significantly larger than that below the BSR, while the S-wave velocity above the BSR is smaller than that below the BSR.
We have seen in the previous section that hydrate cementation results in a decrease of amplitudes with increasing offset. Even though this model can explain the sharp reflection at the BSR and the muting of lithology contrast above the BSR, it cannot explain the real AVO data.
Hydrate deposition within the pore space, however, seems to qualitatively reproduce the observed AVO effect (Figure 10). In this model the hydrate is located away from the grain contacts and thus only weakly affects the stiffness of the sediment frame. The qualitative fit of the in-situ seismic hydrate response with this rock physics model means that: (1) the hydrated sediment is mechanically weak because the grains are not cemented, and (2) the permeability is small (as compared to the cementation case) because the large pore-space conduits are blocked by the hydrate. The latter explains why free gas is trapped underneath the BSR at the Blake Outer Ridge. The fact that the hydrate in this region is underlain by gas-saturated sediment was shown by Ecker and Lumley 1994. Furthermore, significant amounts of hydrate appear to be necessary in the sediment in order to obtain a strong BSR reflection. The data also indicates the absence of a strong reflection from the top of the hydrated sediment. Based on the synthetic modeling of the previous sections, we conclude that the high concentration of hydrate in the sediment just above the BSR gradually decreases with decreasing depth. This means the hydrate is accumulated immediately above the BSR and dissolves further away from the BSR.
 |
Although the hydrate deposition within the pore space can reproduce the data well in a qualitative way, it cannot do so quantitatively. In order to achieve this, we need to assume a more complex grain substance than pure quartz for the rock matrix. Since there is a high amount of clay present in the sediments in the region of the Blake Outer Ridge, its effect on the elastic moduli will have to be determined. However, the use of the simple model already lets us extract information about crucial parameters such as permeability and strength.