In a development well in the Deimos Field, deep water Gulf of Mexico, a large 3D VSP survey was acquired for the purpose of imaging sedimentary structure near a complex salt body where the surface seismic was poorly illuminated. The illumination gaps in the surface seismic image are caused by complex salt geometries which distort the seismic wavefronts. Feasibility modeling was conducted in advance to determine the proper 3D VSP survey parameters. The 3D VSP successfully imaged sedimentary structure near the complex salt body. A subsequent development well target was confirmed with the 3D VSP. This target is within an area that is imaged by surface seismic and established a tie with the 3D VSP. These positive results from the 3D VSP can be used to target future development wells.
As the focus of future exploration and production in deep water Gulf of Mexico moves to subsalt plays, understanding of subsalt amplitude maps have become increasingly important. Also important is to know the acquisition geometry that would best illuminate the complex subsurface, especially subsalt. Forward modeling of seismic data with a realistic velocity model is an answer to this problem. However, full 3D modeling of a large size target is prohibitively high with current state of the art computing capabilities. An elegant and economically affordable alternative presented here provides a qualitative understanding of the subsurface illumination using certain acquisition geometry. The modeling is carried out for a single frequency.
Wagner, Don (BP Upstream Technology) | Sen, Vikram (BP Upstream Technology) | Foster, Paul (BP Upstream Technology) | Albertin, Albertin (BP Upstream Technology) | Read, Randol (BP Upstream Technology) | Burch, Tom (BP Trinidad East Business Unit)
Shallow gas seeps pose a serious problem in seismic imaging but are ubiquitous and can be found in many of the active exploration basins today. Notable among these are Trinidad, the Caspian and the Gulf of Mexico. The abnormally low seismic velocities associated with gas charged zones leads to poor imaging with conventional P-wave data. Expensive yet robust solutions are available in the form of converted wave imaging but often careful processing and proper use of a-priori velocity information can render useful images with conventional towed streamer P-wave data. Our paper presents a case study where we used 3D pre-stack depth migration, residual moveout analysis and tomographic inversion to accurately build a velocity model in an area affected by shallow gas seeps.
Gas-bearing reservoirs southeast of Trinidad are often difficult torepresent geometrically in spite of excellent quality 3-D seismic dataavailable in the area. Developmental challenges of this field include: (1)multiple pay sands (about 20); (2) fault-induced compartmentalization; (3)broad low-relief hydrocarbon accumulations; and (4) complex velocityvariations. Variations in facies and burial rates in addition to shallow andlocalized gas-bearing sands give rise to a complex subsurface velocity field.The localized gas-bearing sands manifest seismically as pronounced "gas sags".Accurate depth representation of proven and non-proven resources wasaccomplished using a minimum number of seismic surfaces (the five besthorizons), all mappable faults, a robust velocity model (for depth conversion),and well data to build a depth model of all of the sands and shales in thesubsurface. Below the shallow gas-bearing sands, initial efforts were notadequate to completely remove the gas sags, which impact over 25% of the studyarea. To accomplish final depth reconstruction, a number of seismic traverseswere interpreted through wells which pass through the shallow gas-bearingsands. These lines were depth converted and used to remove the gas sag wherethe velocity model was inadequate due to incorrect imaging of the seismic data.This was an iterative process which made use of geospatial data (isopachs, welldata, fault displacements, etc.) and was tested against all geophysical data(seismic interpretation, amplitudes, and velocity data). The result was a modelwhich not only added 0.5 TCF of proven resources, but also explained thetrapping mechanism responsible for gas accumulations in multiple reservoircompartments.
The Parang field, part of exploration and production license 120, lies 50miles southeast of the coast of Trinidad (Fig. 1). The field is located in theColumbus basin, which formed between the Caribbean and South American plates.The Miocene to Pleistocene reservoir sands and intercalated shales weredeposited in a variety of depositional environments along the ancestral Orinocodelta. Sedimentary rocks deposited on basin floor to slope fans comprise thelower part of the stratigraphic section. The sands and shales in the middle andupper part of the section were deposited in a paralic setting ranging fromprograding shore facies to a fluvial and transitional barrier complex. Thesource rock is assumed to be early Miocene and/or late Cretaceous in age.