The goal of this project was to establish rock physics models for use in generating synthetic seismic signatures of methane hydrate reservoirs. Ultimately, the intent was to improve seismic detection and quantification of offshore and onshore methane hydrate accumulations.
Stanford University, Stanford, CA 94305
Gas hydrate reservoir characterization is, in principle, no different from traditional hydrocarbon reservoir characterization. The seismic response of the subsurface is determined by the spatial distribution of the elastic properties (properties of the subsurface that deform as seismic waves pass through it) and attenuation. By mapping changes in the elastic properties, scientists can identify geologic features, including hydrocarbon reservoirs.
To accurately transform elastic-property images into more easily interpreted images of lithology, porosity, and the pore-filling phase, quantitative knowledge is needed to relate the rock’s elastic properties to its bulk properties and conditions. Specifically, to quantitatively characterize a natural gas hydrate reservoir, scientists must be able to relate the elastic properties of the sediment to the volume of gas hydrate present and, if possible, to the permeability of that reservoir. One way of achieving this goal is through rock physics modeling.
The ultimate goal of rock physics modeling is to determine the gas hydrate saturation in the pore space from seismic data. A strong relationship has been documented previously between the P-wave impedance and the amount of hydrate in the pore space. Therefore, impedance inversion is an appropriate technique for gas hydrate reservoir characterization.
Unfortunately, there are a number of factors that affect the elastic properties of sediment containing gas hydrate. Some of them, such as the bulk modulus and density of the pore fluid, and the differential pressure, are relatively easy to constrain. Other factors, such as porosity, sediment mineralogy, and gas hydrate saturation, are impossible to determine uniquely from the acoustic impedance.
However, modeling end-member cases can help bracket the results. Once a rock physics gas hydrate model has been established and validated by data, it can be used in a predictive mode to assess the seismic signature of methane hydrate away from well control in a “what-if” mode.
The use of first-principle-based rock physics modeling is crucial for gas hydrate reservoir characterization because only within a physics-based framework can one systematically perturb reservoir properties to estimate the elastic response with the ultimate goal of characterizing the reservoir from field elastic data. The challenge then becomes up-scaling these rock physics relationships so that they are applicable to larger-scale seismic reservoir characterization studies.
Previously, several approaches have been used to estimate methane hydrate concentration in-situ from seismic data. These studies have largely utilized empirical data and/or models that have incorporated variables not appropriate for broader and more general application. However, an approach grounded in a physically consistent micromechanical model could be employed successfully in a variety of depositional environments to detect and quantify methane hydrate reservoirs, and the development of such a model is the main goal of this study.
Once a robust and broadly applicable rock-physics model is developed, further validation of the model utilizing high-quality well log data from various locations worldwide will be necessary. In order to make this model usable with seismic data, issues such as seismic resolution and how to deal with realistic, often un-calibrated, seismic impedance must be well understood. Finally, issues of how to up-scale and apply rock physics in a variety of depositional settings, where the methane hydrate reservoir has a range of elastic properties and may overlie free gas, must be resolved.
This study utilized a combination of log-scale rock physics models, upscaling them for use with seismic impedance, to create a synthetic-seismic catalogue over methane hydrate reservoirs. This allows scientists to more accurately detect and characterize methane hydrate reservoirs in both permafrost and marine sediments worldwide. Some of the synthetic modeling tools are interactive and could potentially be used by the methane hydrate research community via the Internet to understand and quantitatively interpret seismic observations.
The project team developed an approach to natural methane hydrate quantification in which the user generates synthetic seismic traces and compares them to real seismic data. If the synthetic matches the observed seismogram, then the reservoir properties and conditions used in synthetic modeling might be the same as the actual, in-situ reservoir conditions. This methodology is based on rock physics equations that link (a) the porosity and mineralogy of the host sediment, pressure, and hydrate saturation, and (b) the resulting elastic-wave velocity and density. The project team developed such rock physics equations to provide this link. One of relationships that appears essentially universal across various methane hydrate provinces is a model for unconsolidated sediment, where the hydrate acts as part of the mineral frame. This rock physics transform, combined with simple earth models, produces synthetic seismic reflections of gas hydrate that can be matched to real data and then perturbed to guide exploration and hydrate reservoir characterization. One result of such seismic forward modeling is a catalogue of seismic reflections of methane hydrate which can serve as a field guide to hydrate identification from real seismic data.
The forward-modeling approach developed and advocated here results in non-unique solutions, because different combinations of rock properties may produce the same reflections. To constrain the spectrum of answers, the earth model used in the modeling has to be geologically-plausible, including ranges of porosity and clay content in the layers which are permissible within the hydrate stability window defined by the pore pressure and temperature. Such constraints are fairly straightforward to impose because most hydrate reservoirs of potential practical significance are high-porosity unconsolidated sands encased in unconsolidated shale. The configurations and spatial distributions of course may vary. The physics-based forward-modeling approach offered here is one way of addressing this natural diversity.
The project has been completed, and the final report is provided below under "Additional Information".
In addition to the information provided here, a full listing of project related publications and presentations as well as a listing of funded students can be found in the Methane Hydrate Program Bibliography [PDF].
Final Project Report [PDF-2.58MB] - June, 2008
Fire in the Ice Article - Summer 2008 [PDF] - Pg. 13 article by Dr. Jack Dvorkin, Stanford University
September 2007 Project Review [PDF-1.40MB]
E&P Paper [PDF-294KB]
Technical Status Assessment [PDF-512KB]
Cordon, I., Dvorkin, J., and Mavko, G., 2006, Seismic reflections of gas hydrate from perturbational forward modeling, Geophysics, 71, F165-F171.
Bandyopadhyay, K., Jack Dvorkin, J., and Gary Mavko, G., 2007, Anisotropy in gas hydrate (in preparation for Geophysics, presented at the 2007 SRB meeting at Stanford).
Dvorkin, J., and Uden, R., 2004, Seismic wave attenuation in a methane hydrate reservoir, The Leading Edge, 23, 730-734.
Dvorkin, J., and R. Uden, 2006, The challenge of scale in seismic mapping of hydrate and solutions, The Leading Edge, Volume 25, n. 5, p. 637-642.