Gerald Kidd
Paradigm
Turbidite systems offshore West Africa are important economic reservoirs for many exploration companies.
While the basic geology of turbidite deposition is straightforward, mapping and evaluating its intricacies ranges from challenging to overwhelming for even the experienced interpreter.
Complying with stringent business guidelines, satisfying exploration managers’ requirements, and analyzing vast amounts of data add to the interpreter’s challenge.
 |
| 3D seismic geomorphology of basement showing accommodation model with sediment pathways. |
One approach combines efficiency and accuracy with state-of-the-art technologies to address these challenges. Using a holistic approach to conduct an initial analysis of the data, evaluate technology options, and then leverage specific workflows, procedures, and parameters gives maximum results with minimum work and minimum elapsed time.
In turbidite regimes, the traditional line-by-line interpretation is unsatisfactory. Objectives frequently are realized only after mapping. The time required by this approach leaves little or no time to investigate deeper and shallower sections which often contain important information.
A common alternative approach to line-by-line interpretation is to increment the line-by-line process, then infill with interpolation and smoothing. This generally produces poor results in complex stratigraphic settings such as turbidites due to the complex discontinuous nature of the reflections and numerous unconformities.
Significant improvements in efficiency, quality, and accuracy can be obtained by using state-of the-art volume interpretation technologies to extract seismic correlations in three dimensions to automatically map and create specialized 3D volume images. Integrating multiple volume extraction methods yields immediate and better results than line-by-line interpretation, and provides a valuable “heads-up” view of the geology which can guide manual interpretation.
Data complexity may require tedious manual interpretation; by performing volume reconnaissance before hand, manual interpretation time can be reduced. The following case study describes this approach.
Study area, background geology
The study area lies near the center of the Tano-Ivoirian basin in easternmost offshore Côte d’Ivoire, West Africa, between the continental shelf break and the lower slope in water depths ranging from approximately 300 m (984 ft) in the north to more than 2,100 m (6,890 ft) in the South. Turonian and Santonian slope fan turbidities pinch-out around the flanks of Albian tilted fault blocks while younger Campanian to present-day large slope canyon complexes cut through and lie off the flanks of the old structural highs.
 |
| Left, formation sculpting showing amplitude distribution. Right, composite of geobody extractions showing major sediment pathways. |
The Bassam Canyon is to the west and the Assinie Canyon is to the east of the study area. The objectives are within the Assinie Canyon system and consist of a deeper Santonian slope fan play and overlying Campanian canyon-fill channel complexes.
Volume interpretation requires the data scaling to be checked and re-scaled if necessary, to improve the dynamic range of the data, enhance the effectiveness of the methods, and improve operational performance. Begin with a full volume 16- or 32-bit QC analysis prior to re-scaling. The QC check and rescaling should be performed by the interpreter according to the data and the objective. In this study, all volumes (amplitude, inversion attributes, and coherency) were re-scaled.
Take time to become familiar with the overall geology and then design strategies to interpret critical features. Identify and evaluate key sequence boundaries, critical faults, and major turbidite systems. Go beyond simple identification and begin to analyze overall 3D geometries.
Although the main focus is on the objective interval, interpretation strategies should aim to integrate the overlying and underlying geology and to understand the structural and stratigraphic framework. In this case study, the following efficient and easily executed steps were used:
- Volume rendering to page through the data
- Reconnaissance sub-volume detections to define and map the extent and thickness of major turbidite systems
- Trace shape based horizon tracking to investigate the extensiveness of the most continuous reflections.
Interpretation challenge
Results of the 3D data overview show the overall turbidite system consists of two Campanian sub-systems, an upper and a lower, each with stacked individual turbidic flows. They are separated by unconformities, overlain by a highly faulted Maastrichtian section and underlain by additional significant unconformities in the Turonian and Cenomanian above the Albanian unconformity. The challenge was to map both turbidite systems and to understand their relationship and differences with respect to the overlying and underlying section and basement.
The deposition of reservoirs and key characteristics are controlled by an accommodation model which must be determined. The most significant control is the topography of the Albanian unconformity and, to a lesser degree, the topography of its conformable overlying section which has various degrees of erosion.
 |
| Seismic trace shape stratigraphic SeisFacies classification of the upper sequence. Left, map view semi-transparent 3D volume rendering. Right, inline dip and bottom right, crossline strike stratigraphic sections. |
In this area, the Albanian unconformity is a major unconformity with significant truncation of older dipping strata and a wide variation of younger depositional features to produce a complex variable phase and amplitude acoustic boundary. Acoustically, the Top Albanian reflection generally is strong, absorbing a considerable amount of seismic energy, with some areas of lower amplitude where the overlying reflections geometries are variable, somewhat masking the basement reflection.
Considering this, a 3D trace shape horizon tracker was applied with carefully selected parameters to analyze the data volume about it, and then propagate automatically along this surface to produce a surface. The procedure completes in a few seconds and subsequent QC of the results was satisfactory, yielding accurate maps of the surface. Both a depth structure map and a 3D volume sculpting showing the amplitude distribution and accommodation model with sediment pathways immediately above basement were generated quickly. The topography of the Albanian unconformity reveals its depositional controls on the overlying sediment extended into the Campanian.
Mapping key sequence boundaries was the most difficult and time consuming task due to the complexity of the unconformities. A key that helped interpret the sequence boundaries was to generate enhanced 3D profiles from an acoustic impedance volume that clearly show the lithologies and sequence boundaries. Due to impedance and phase variations along the boundaries, a grid was manually interpreted and partially filled using a high correlation coefficient 3D volume trace shape horizon propagator.
The final surface was infilled with interpolation. In the section overlying the turbidite systems, 3D wave shape propagation technology was used exclusively to map a highly faulted Maastrichtian surface which is discussed later. Nine detailed surfaces were mapped in less than three weeks.
Sequence analysis
The mapped sequences define the major lithologic and reservoir intervals between the Top Albian to the Maastrichtian. For the sake of thoroughness, each sequence was investigated for potential exploration opportunities using two volume related technologies: formation sculpting and cross-plot/multi-body detections.
Formation sculpting extracts all seismic data between two boundaries, then applies opacity filters to image the amplitude distributions within. The formation sculpting results also serve as input to the cross-plot/multi-body detection technology where all amplitude constrained geobodies are extracted. Surfaces were automatically generated for the 20 largest geobodies.
The areal distribution of the geobodies was studied. Composite images of upper and lower turbidite systems were created along with a grand composite for all geobodies within all sequences. The distributions define the sediment pathways and geobody concentrations and are consistent with the accommodation model, defining two types of stratigraphic traps and point to the lower risk areas of potential hydrocarbon traps.
3D trace shape stratigraphic SeisFacies classification
A set of inversion attribute volumes and sequence boundaries served as input to a seismic facies classification technology to determine which attributes contributed to an optimal trace shape classification and to remove redundancy. The output is a volume within a specific sequence, evaluated in 3D semi-transparent geomorphologic space, and integrated with opaque stratigraphic profiles.
Results show evidence of lithologic facies changes in an Upper Campanian sequence that supports stratigraphic trapping mechanisms updip of sand prone facies. Trace shape analysis provides seismic differentiation that is invisible to amplitude analysis.
 |
| Automatic Fault Extraction (AFE) discriminating faults against stratigraphic discontinuities (red arrows) in a turbidite system. |
To further understand the fault architecture, a horizon through the faulted Maastrichtian interval was created using trace shape horizon propogation with little effort. Numerous minor Upper Maastrichtian faults are seen in a thick claystone interval with a few faults intersecting the Campanian turbidites. Conventional fault mapping is not feasible because of the many insignificant faults. However, their overall structural architecture should be studied and integrated into the interpretation. Therefore, volume-based automatic fault extraction (AFE) combined with multi-directional lighting technology was used to study the significance of faulting and impact on potential reservoirs.
First, a coherency volume was made to “highlight” the discontinuities and used as input to the AFE application which has algorithms that generate fault lineaments and selectively process them into faults. Key AFE parameters defined the proximity of lineaments to be included as the same fault.
Deviation criteria define which lineaments to include as the same fault based on a specified angle between them, and the minimum lineament length to be included in the fault extraction. AFE searches and sorts lineaments within the volume and automatically outputs named faults.
In this case study, over 700 faults were generated and studied by Rose diagrams and by azimuth sorting. The AFE application effectively discriminated between fault lineaments and stratigraphic discontinuities. This is clearly evident on a seismic profile showing a thick complex turbidite reservoir interval consisting of multiple stratigraphic discontinuities intersected by a single automatically extracted fault. Accurate fault extraction reduces quality control and editing time.
A four-color light editor was used to differentiate the azimuth of fault scarps to clearly reveal structural characteristics by offset and trend. The overall structural analysis that the Maastrichtian section underwent a combination of compactional faulting with a slight component of mass movement. This explains why numerous faults are restricted to a shallow interval.
Stacked turbidite systems are complex and time consuming to interpret. Understanding the relationships of younger and deeper sections is required for complete and accurate interpretation. Multiple volume interpretation provides an efficient workflow to understand critical regional and details information on the accommodation model, basin filling, sediment pathways, the stratigraphic architecture of the turbidite systems, its internal geometries, its relationship to the structural elements, and identifies two types of stratigraphic traps in a timely manner. Although more detailed work remains, the results of integrating volume interpretation methods provide a firm understanding of the basic geology in a complex environment in a short period of time.
Acknowledgments
The author gratefully acknowledges Vanco Energy Co. for permission to show this seismic data and for providing the regional geologic context. I also thank Laura Evins and Iris Diaz of Paradigm for their technical contributions, and Marianne Smith and Huw James, also of Paradigm for editing the paper.
