Conclusions

In this Chapter, I provided a theoretical tool for estimating the amount of hydrate and gas from seismic interval velocity. I developed rock-physics models that link the elastic wave velocity in high-porosity marine sediment to density, porosity, effective pressure, mineralogy, and water, gas and hydrate saturation. Three micromechanical models of hydrate deposition in the pore space were examined: (A) hydrate is part of the pore fluid, (B) hydrate becomes part of the solid frame, and (C) hydrate cements grain contacts. Using the interval velocities obtained from the stacking velocity analysis described in Chapter 2, I calculated lateral maps of hydrate and gas saturation. The calculations were based both on an averaging approach and a trace-by-trace approach, resulting in an upper and lower bound for the possible hydrate saturation. I find considerably large lateral variations in hydrate saturation. Model A results in a maximum hydrate saturation between 21% and 26% and model B between 15% and 20%. In case of hydrate cementing the sediments, only 1% of hydrate is required in the pore space to increase the velocity corresponding to that observed in the seismic data. The estimated gas saturation is approximately 1%-2%. Subsequently, I evaluated the robustness of these hydrate and gas saturation estimates with respect to the velocity errors that were determined in Chapter 2. The analysis suggests that the maximum errors introduced into the hydrate saturation estimates can be up to $\pm$ 14%, while the gas saturation shows maximum uncertainties of $\pm$ 2%. The uncertainty estimates are upper bounds on possible errors. The comparison with VSP data (see Chapter 2, section 2.3.2) and other seismic velocity investigations Katzman et al. (1994); Korenaga et al. (1997); Wood et al. (1994) suggests that the error in interval velocity is smaller than those upper bounds which would shrink the uncertainties on the hydrate saturation. Using additional velocity and porosity information from well-logs 994 and 995, I evaluated the validity of the proposed technique and the used models. This investigation suggests that the technique is quantitatively accurate.