Estimating the composition of gas hydrate using 3D seismic data from Penghu Canyon, offshore Taiwan

Abstract

Direct measurements of gas composition by drilling at a few hundred meters below seafloor can be costly, and a remote sensing method may be preferable. The hydrate occurrence is seismically shown by a bottom-simulating reflection (BSR) which is generally indicative of the base of the hydrate stability zone. With a good temperature profile from the seafloor to the depth of the BSR, a near-correct hydrate phase diagram can be calculated, which can be directly related to the hydrate composition. However, in the areas with high topographic anomalies of seafloor, the temperature profile is usually poorly defined, with scattered data. Here we used a remote method to reduce such scattering. We derived gas composition of hydrate in stability zone and reduced the scattering by considering depth-dependent geothermal conductivity and topographic corrections. Using 3D seismic data at the Penghu canyon, offshore SW Taiwan, we corrected for topographic focusing through 3D numerical thermal modeling. A temperature profile was fitted with a depth-dependent geothermal gradient, considering the increasing thermal conductivity with depth. Using a pore-water salinity of 2%, we constructed a gas hydrate phase model composed of 99% methane and 1% ethane to derive a temperature depth profile consistent with the seafloor temperature from in-situ measurements, and geochemical analyses of the pore fluids. The high methane content suggests predominantly biogenic source. The derived regional geothermal gradient is 40°C km-1. This method can be applied to other comparable marine environment to better constrain the composition of gas hydrate from BSR in a seismic data, in absence of direct sampling.

More information

Identifiers

Catalogue record

Export record

ASCII CitationAtomBibTeXData Cite XMLDublin CoreDublin CoreEP3 XMLEndNoteHTML CitationHTML CitationHTML ListJSONMETSMODSMPEG-21 DIDLOpenURL ContextObjectOpenURL ContextObject in SpanRDF+N-TriplesRDF+N3RDF+XMLRIOXX2 XMLReferReference ManagerSimple Metadata

Altmetrics