Search

We applied full waveform inversion to a seismic dataset from offshore Gabon. This dataset features complex geology such as diapirism, shallow gas pockets and dewatering faults. We show the results on a large production-size swath from an advanced full waveform inversion method based on the L-BFGS optimization algorithm. The updated velocity model highlights an uplift in resolution and a consistency with the geological features observed in the depth migrated stack. The benefits include identifying features and faults in the overburden and improved top-salt characterization. These result in better imaging in both the post- and pre-salt regions, as well as a velocity model than can be used to aid the geological interpretation.

Numerous evaluations, conducted on various sites, converge towards a shared conclusion: the geophysical value of an acquisition is primarily driven by the density of traces acquired in the field. To confirm this principle also applies in North Africa, ENI and CGG agreed on a technical collaboration. The objective of the collaboration was to demonstrate, on a crew operating in South of Tunisia, that, without additional equipment, using modern high productivity schemes it was possible to efficiently increase the conventional Trace Density by a factor of 10 to 50 with a significant increase in geophysical value all along the processing, imaging and interpretation steps.

In this paper we introduce an AVO QC workflow which was initially designed for ultra-dense land datasets. Its purpose is to monitor the AVO behavior at key steps of a processing sequence on the full area of interest and not only limited to a well neighborhood. The workflow, based on data reduction, consists in the creation of a reliable AVO model that can then be compared to the seismic through different measurable indicators. After some basic interpretation on the whole processed area, using maps and statistics, the processing step is assessed in terms of ‘AVO-friendliness’. If validated, the sequence can be continued and if not, it must be revised. This AVO QC workflow proves the AVO-friendliness of a process. The robustness of the workflow makes it potentially applicable to all types of 3D datasets.

A case history in West Texas, in the Permian Basin, of a non-exclusive seismic imaging project spanning acquisition, processing and reservoir interpretation. Located in Howard and Glasscock counties, the Hobo non-exclusive seismic survey aims at imaging and quantitatively interpreting the Spraberry and Wolfcamp formations.

We demonstrate a recently developed method for computing tomography model uncertainties and mapping them into the migrated domain. The method generates after the final tomography a series of equi-probable velocity model perturbations within a standard deviation confidence level. This allows computing standard deviation-like attributes for velocity and anisotropy parameters and for key horizons. An application to West of Shetland dataset highlights the interest of the estimated uncertainties.

This paper describes the construction of a high-resolution anisotropic FWI model that is used to help image a karstified carbonate layer and also to help define the underlying clastic reservoirs. The karst zones show extreme velocity rugosity, especially as compared to the surrounding sediments. Conventional reflection tomography cannot resolve the anomalous velocities, and the resulting seismic images below these karstified carbonates are distorted and poor for interpretation. This paper presents the strategies for determining the appropriate input data for FWI and how to build a multi-layer horizon-constrained anisotropic model and the uplift that results from using Q-factors .

Traditionally, the regularization step is performed independently for each time-lapse vintage. This disregards any geometrical limitations imposed by different surveys. Here we recast the regularization process as a minimization problem with model-space constraints. These constraints couple geometrical relations between surveys to improve repeatability. We also demonstrate how to solve the minimization using a practical and pragmatic approach.

In this paper a comparison of deterministic and geostatistical inversion approaches to reservoir characterization of a complex gas saturated clastic reservoir with thin coal beds will be discussed. This type of reservoir is characterized by high acoustic contrasts which cause challenges for inversion techniques. To overcome these challenges in deterministic inversion a methodology was applied to update the low frequency model in an iterative manner. This improves the accuracy of elastic properties prediction and, therefore resulting in a more reliable reservoir model, which was verified with data from 28 newly drilled wells.

Water column statics caused by tidal variation and water velocity change during seismic surveys is one major source of noise in marine 4D projects. Correction of this statics effect is a key step in any marine 4D processing. Applying water column statics correction requires a good knowledge of the distance or surface take-off angle when waves travel through the water column, which conventional methods such as ray tracing are not able to obtain accurately when the subsurface velocity is complex. We propose a new method to apply water column statics correction through progressive sparse TauP inversion. This method does not need prior inputs of subsurface velocity and reflector dips, as required for ray-tracing methods, and benefits from the progressive sparse TauP inversion engine that can properly handle spatially aliased marine seismic data and mitigate energy leakage in the TauP domain. We demonstrate the effectiveness of this new method using synthetic ocean bottom seismometer (OBS) data derived from a SEAM velocity model and using real OBS data from 4D surveys over the Atlantis field in the Green Canyon area of the Gulf of Mexico (GOM).