Thierry Coléou

Unsupervised seismic facies classification A review and comparison of techniques and implementation

The use of automatic seismic facies classification techniques has been steadily increasing within E&P interpretation workflows over the past 10 years. It is not yet considered a standard procedure but, with the knowledge of the advantages (and limitations) of the different seismic classification methods, its role in the interpretation process as a successful hydrocarbon prediction tool is anticipated to grow. This paper reviews and compares the unsupervised classification methods presently used in seismic facies analysis: K-means clustering, principal component analysis (PCA), projection pursuit, and neural networks (vector quantization and Kohonen self-organizing maps). The term “unsupervised” covers all classification techniques relying only on input data and not biased by the desired output. These methods are described, compared, and illustrated by case studies taken from deep offshore Louisiana, west and south Texas, onshore California, and offshore Indonesia. Seismic data volumes are huge and consist of highly redundant data. It is now established that their analysis can be greatly optimized by applying efficient data reduction algorithms that preserve essential features of the seismic character. The objective of the facies classification process is to describe enough variability of the seismic data to reveal details of the underlying geologic features. The classification process should do this while preserving a synthesis for the seismic signal changes. In addition to comparing the classification methods, this paper presents relevant information pertaining to data input, data analysis, and data output. Seismic data, from the statistical point of view, have characteristics like continuity, redundancy, and noise that affect the behavior of the seismic classification techniques. At first glance, these classification techniques may appear similar. However, their ability to efficiently describe changes in seismic character and their interpretability can vary significantly. These different techniques are compared in this paper for their suitability to describe seismic data in a meaningful manner directly …

Petrophysical Seismic Inversion

As part of a two years collaborative RD the geomodels are not simply conditioned to seismic attributes by seismic-guided mapping but reproduce prestack seismic measurements (Bornard et al., 2005). A key component of the inversion methodology is a Petro-Elastic Model (PEM) that links the reservoir properties stored in the geomodel (e.g., porosity, rock types and fluid saturations) to the elastic response. The new technique is illustrated using a direct porosity inversion case study involving a large North Sea reservoir.

Multivariate Geostatistical filtering of time-lapse seismic data for an improved 4-D signature.

4-D technology is moving into an accelerated phase with several successful projects to date. Acquisition and processing imprints are considered negligible with standard 3-D processing but become critical when computing time-lapse seismic differences. We introduce a spatial co-filtering geostatistical technique into the 4-D processing sequence to remove acquisition imprints and other uncorrelated noise that increases the repeatability and optimises the 4-D signature. Examples from two fields from the North-Sea, Draugen (Norske Shell) and Forties (BP) are presented.

On the Use of Geostatistical Filtering Techniques In Seismic Processing

When redundancy of seismic data exists factorial cokriging enables the estimation of (1) a common part, based on the common spatial behavior, and (2) the differences relative to the common part of the input data. Coléou (2002) first introduced the automatic implementation of factorial co-kriging (AFACK) as a filtering technique for the time-lapse (4D) processing sequence. It was specifically designed to optimize the critical time-lapse information such as the repeatability and the 4D seismic signature. However, over the last two years we have been developing new applications of this technique in very different processing environments. For example, applications providing data reduction, such as stacking or AVO and EI analysis, have a direct interest in the common part of consecutive offset cubes. Furthermore, we have successfully applied AFACK to more specialized problems such as the merging of OBC and streamer data or the decomposition of wide-azimuth data for fracture characterization.

Petrophysical Seismic Inversion Applied to the Troll Field.

Summary We present an application of petrophysical seismic inversion, a method driven by petro-elastic models, updating a fine-scale geological model in depth to make it fully compatible with pre-stack seismic measurements on a part of the Troll Field Central province, in the North Sea. The results are being evaluated for infill drilling and for a future 4D inversion to determine the remaining oil in the thin oil leg.

Petrophysical seismic inversion applied to the troll field

We present an application of petrophysical seismic inversion, a method driven by petro-elastic models, updating a fine-scale geological model in depth to make it fully compatible with pre-stack seismic measurements on a part of the Troll Field Central province, in the North Sea. The results are being evaluated for infill drilling and for a future 4D inversion to determine the remaining oil in the thin oil leg.

Tomographic Velocities - Challenges and Applications

The resolution of seismic imaging has long been characterized by a mid-frequency gap between the long vertical wavelength components that can be inferred from travel-time tomography and the short vertical wavelength components that can be inferred from seismic migration. The progress in tomography and in broadband acquisition now allows this mid-wavelength gap to be filled and even results in the overlapping between the resolution obtained from velocity model building and from seismic migration. Recent progress in tomographic approaches now provides vertical resolution up to 6 Hz, with in addition, a precise localization of the velocity contrasts. With a case study we discuss their benefits for reservoir characterization, where the missing frequencies are traditionally coming from a model interpolated/extrapolated from well log information. If the seismic inversion for reservoir characterization is not sensitive to the low-to-mid frequencies, this low-to-mid frequency modelling is critical for the estimation of absolute rock properties, and replacing a model extrapolated from sparse well information by actual measurements is of paramount importance.

Petrophysical Seismic Inversion Over an Offshore Carbonate Field

The main objective of this project was to evaluate the ability to derive petrophysical properties like porosity from pre-stack seismic data in a carbonate environment. We apply a direct petrophysical inversion technique to an offshore carbonate reservoir. Starting from an initial geological model in depth and a number of carefully conditioned seismic angle stacks, we derive a detailed 3D model of the porosity and hydrocarbon saturation matching the observed seismic data. We use a wellcalibrated Petro-Elastic Model (PEM) to link the petrophysical properties to the seismic velocities. We compare inversion results obtained using the Xu-Payne and T-matrix PEMs which both account for carbonate pore geometry, lithology, porosity and fluid content but have different elastic sensitivity to fluid saturations. The inverted results provide detailed images of the spatial variations of porosity and fluid content across the reservoir interval. Predicting absolute saturation values is more difficult, as saturation estimation is strongly dependent on the choice of PEM.

Petrophysical Seismic Inversion over an Offshore Carbonate Field

The main objective of this project was to evaluate the ability to derive petrophysical properties like porosity from seismic data in a carbonate environment. A special attention has been given to the possibility of characterizing the geometry of the pore space directly from the pre-stack seismic data. We apply a direct petrophysical inversion technique to a carbonate reservoir offshore Brazil. Starting from an initial geological model in depth and a number of carefully conditioned seismic angle stacks, we derive a detailed 3-D model of the porosity and hydrocarbon saturation matching the observed seismic data. We use a well-calibrated Petro-Elastic Model (PEM) to link the petrophysical properties to the seismic velocities. We compare inversion results obtained using the Xu-Payne and T-matrix PEMs which both account for carbonate pore geometry, lithology, porosity and fluid content but have different elastic sensitivity to fluid saturations. The inverted results provide detailed images of the spatial variations of porosity and fluid content across the reservoir interval. Obtaining estimates of absolute saturations values is more difficult, as saturation estimation is strongly dependent on the choice of PEM.

Static and Dynamic Geo-model and Flow Model Update Using 4D Petrophysical Seismic Inversion: an Application over Troll West

On Troll West area, the Sognefjord producing formation is thick (160 m) with very good reservoir properties and with a thin Oil leg ( straightforward overburden, leading to good quality seismic data. After more than a decade of Oil production, more than 1 million meters of horizontal drains have been drilled and the fifth monitor survey has been shot in 2009. This huge amount of data makes it a-priori particularly suitable for seismic monitoring and should lead to quantitatively reliable follow-up of the thin Oil leg thickness and production forecast. The following paper describes results from a R&D-collaboration between Statoil and CGGVeritas.