Methane hydrate structures are increasingly recognized as both a potential energy resource and a resource that might have a strong ``greenhouse'' effect on future global climate Kvenvolden (1993). Associated with the base of the stability zone of these hydrate structures are so-called bottom simulating reflectors (BSRs) that parallel the seafloor at a subbottom depth of several hundred meters. Knowledge of the elastic and mechanical properties of the hydrate structures and the BSRs is essential to an estimation of the total volume of hydrates in place and its recoverability (permeability and formation strength). The key to extracting these parameters is a combination of seismic analysis and rock physics, which will provide a link between seismic velocity and attenuation and internal rock structure.
Several attempts to obtain the properties of hydrated sediment directly from cores have been thwarted by the instability of methane hydrate at normal temperature and pressure Dallimore and Collett (1995); Kvenvolden and Barnard (1983); MacKay et al. (1994). Consequently, the elastic properties have been inferred from seismic reflection data Ecker and Lumley (1994); Hyndman and Spence (1992); Katzman et al. (1994); Singh et al. (1993). A method of linking these data to the desired properties of hydrated sediment is synthetic forward modeling. Typically, a rock physics hypothesis is used in these models to estimate velocities in hydrate-bearing sediments from those in the sediment frame, and in pure hydrate. The latter two are normally available from ocean-drilling cores and from laboratory measurements Pearson et al. (1986); Stoll and Bryan (1979); Whalley (1980). Dvorkin and Nur 1993 give an example of such a synthetic velocity modeling by assuming that hydrate acts as cement at grain contacts. However, little effort has been done so far to correlate rock physics theory directly with seismic hydrate responses.
This paper is a first attempt to interpret seismic AVO data of hydrate structures by performance of rock-physics-based seismic modeling. We explore the effect of two different hydrate models both on the elastic properties of hydrated sediments and on the seismic responses that can actually be measured. The subsequent link of the modeled internal structure with real seismic data emphasizes one goal of rock physics. This goal is not purely a matching of data by the use of an appropriate theory. Rather it is the at least qualitatively extraction of parameters such as permeability and strength from seismically inferred rock structure.
In this paper, we discuss two mechanical hydrate models, their elastic properties and seismic amplitudes, and the resulting interpretation of AVO data from a bottom simulating reflector (BSR) offshore Florida and Georgia. We examined two micromechanical hydrate models: (1) the hydrate cements the grain contact and strongly reinforces the sediment; and (2) the hydrate is located in away from the grain contacts and only weakly affects the stiffness of the sediment's frame. The hydrate-bearing sediment was brine saturated, while the underlaying sediments contained either brine or free gas. The dry rock properties of model (1) were calculated using the cementation theory by Dvorkin et al. , while the parameters of model (2) were based on the contact theory of Hertz and Mindlin 1949. Using Gassmann's equation, we determined the saturated elastic rock properties. Synthetic seismograms were generated using the 1-D, elastic Thompson-Haskell reflectivity method. In order to obtain the amplitude behavior at the BSR without interfering converted waves or critical angle effects, we calculated the AVO responses from Zoeppritz equations. A subsequent qualitative comparison of the synthetic results with data from the Blake Outer Ridge, offshore Florida and Georgia, resulted in an interpretation of the in-situ settings in this region.