William L. Abriel

Multiple-point simulation integrating wells, three-dimensional seismic data, and geology

There are two significant challenges in building a reservoir model integrating all available information. One challenge is that wells and seismic data measure the reservoir at different scales of resolution. The other challenge lies in how to account for conceptual geological knowledge with resolution at multiple scales.In this paper, we present a case study of integrating well data, seismic data, and conceptual geologic models. The well and seismic data are of good quality, but conventional well-seismic data calibration indicates that the seismic data are unable to fully differentiate sand from shale. The reason for this poor well-seismic data calibration is that well log and seismic data measure the reservoir at different scales. Well logs are able to differentiate sand from shale, whereas seismic data are better at detecting larger scale depositional geometries.A new workflow is presented to deal with this problem. First, principal component analysis clustering is used to identify characteristic patterns of certain depositional facies, from which sandy and shaly channels are interpreted. Next, multiple-point geostatistical simulation is performed to build a depositional-facies model, which integrates both hard and soft data but also incorporates realistic depositional-facies geometries provided by our geological knowledge of this reservoir. Finally, different lithofacies (sand and shale) indicators and corresponding petrophysical properties are simulated honoring the limited well data.The results show that not only are the geological features better reproduced, but also is the uncertainty about the reservoir significantly reduced because of a better integration of corresponding three-dimensional seismic data.

Over/under towed-streamer acquisition A method to extend seismic bandwidth to both higher and lower frequencies

Seismic resolution of towed marine-streamer data is affected by free-surface reflections. They strongly modulate the spectrum reducing energy at the so-called notch frequencies, fnu2009=u2009iv/2zu2009{iu2009=u20090,1,…,∞}, in which v is the water velocity and z is the source or receiver depth. As a result there is a very strong loss of useful low-frequency energy and usually a similar loss at higher frequencies. The ghost effect occurs both at the source and at the receiver. At the source, the upgoing part of the source wavefield is reflected off the free surface with inverted polarity before its delayed journey into the Earth. At the receiver, the upgoing part of the receiver wavefield is reflected off the free surface with inverted polarity before it travels down to the receiver again. The over/under acquisition method allows separation of the up- and downgoing wavefields at the source (or receiver) using a vertical pair of sources (or receivers) to determine wave direction. In this article, we present the results of two ...

Reservoir modeling Integrating various data at appropriate scales

Modern reservoir characterization workflows attempt to integrate all available, reliable, and appropriate sources of data into 3D geocellular or numerical earth model(s). These models provide various properties such as lithofacies, porosity, permeability, hydrocarbon/water saturation at each grid cell. Reservoir simulation is then performed on the geocellular model to predict reservoir performance and production history. Geocellular models allow geoscientists to integrate various data from many different sources to calculate oil and gas volumetrics, perform efficient well planning, forecast reservoir performance, and optimize reservoir depletion schemes. In addition, within the same model, the stochastic simulation process can generate multiple equiprobable realizations, all of which honor the input data. The differences among these realizations provide a quantitative envelope of the uncertainty due to the limited reservoir information and data for the subsurface.

Interactive seismic mapping of net producible gas sand in the Gulf of Mexico

In the Garden Banks area of offshore Louisiana several gas sands have been drilled and found productive. However, the sands are laterally variable in thickness and effectiveness. An improved understanding of the spatial distribution of net producible gas sand is highly desirable for reservoir management. The bright reflections from the top and the base of each sand were tracked automatically on an interactive interpretation system. This yielded time structure maps and hence isochron maps for each gross sand interval. The horizon Seiscrop™ sections diplaying amplitudes over the sand interfaces were then summed, adjusted for tuning effects, and smoothed to yield estimates of net gas/gross sand ratio over the area under study. By combining these with the corresponding isochron maps and an appropriate gas sand interval velocity, we obtained net gas sand isopach maps which tie acceptably with well data. Integration of these provided total reservoir volumes. ™Trademark of Geophysical Service Inc.

3-D poststack phase-shift migration in transversely isotropic media

It is now well established that anisotropy can adversely affect migration results when conventional (isotropic) methods are used. A number of extensions to handle anisotropy have been reported in the literature. These algorithms have usually been applied to transversely isotropic (TI) media, which are useful for describing shales and finely laminated beds. Phase-shift time migration is a popular technique for areas with relatively flat-lying geology, but it has yet to be applied to layered, TI media. In this paper the authors describe a 3-D poststack phase-shift method for TI P-waves, and show imaging results from a Gulf of Mexico data set. After a triple Fourier transform over the two horizontal coordinates and time, the data are downward continued using a depth-varying dispersion relation for TI P-waves. This requires four, independent time (or depth) functions, each corresponding to one of the four elastic stiffness constants needed to describe a TI medium. In the implementation, they use the vertical P- and S-velocities, the P-wave stacking velocity, and the horizontal P-velocity. Obtaining these functions can be difficult without well information. The authors show how discrepancies between P-wave stacking, vertical check-shot, and migration analysis velocities can help estimate the amount of anisotropy in themorexa0» area. In the example, increasing the horizontal P-velocity over the vertical by only 10% was enough to improve the image at reservoir depths, illustrating the importance of including anisotropy for better reservoir characterization.«xa0less

Detection of hydrocarbons using non-bright-spot seismic techniques

Interpretation technology for the direct detection of hydrocarbons with stacked seismic reflections is focused mainly on bright spot systems in which the acoustic impedance of the reservoir is lower than the overburden and underburden. Interpretation of polarity reversal and dim spot systems, in which the acoustic impedance of the reservoir is higher than the overburden or underburden, requires the interpreter to use a different and more complicated set of rules to validate these as direct hydrocarbon indicators.

The shapes of Gulf Coast salt intrusions related to seismic imaging

In the Gulf of Mexico, it is important to know the salt‐sediment interface as accurately as possible because of its effect on depth imaging. 3-D depth migrations are rapidly changing our understanding of subsalt structure, since they reveal that 3-D time‐migrated subsalt images are confusing or totally misleading (Figures 1–2). With common velocity contrast ratios of 1.5–2.5 at the salt boundary, Snell’s law implies large refraction angles. Complex 3-D shapes cause data to refract over long lateral distances at sharp angles. Examples of many carefully studied salt intrusions show little hope of invoking 2-D symmetry of the salt for acceptable depth imaging.

The polarity of zero-phase wavelets

Everyone agrees that zero phase is the data condition that makes seismic interpretation easier and more accurate. This becomes increasingly important as an interpretation progresses from structure to stratigraphy to hydrocarbon identification to reservoir analysis. Hence data processing today makes

SEG Advanced Modeling (SEAM) today and tomorrow

This presentation focuses on what we felt are the three most important foundational concepts of the SEG Advanced Modeling Consortium (SEAM). In the our view, these three issues form the basis not only of SEAM, but to a considerable degree the geophysical road ahead. The first is on the emerging techniques for generating models of the Earth’s subsurface with realistic geologic styles. The second discusses the actual algorithms for simulation of seismic acquisition over such models. The third is directed toward the kind of existing and emerging mathematical, physical, and computational technologies that enable extraction of the kind of information required to improve the entire exploration industry’s success rate. We argue that solution of the first leads naturally to application of the second, followed by successful verification of the validity of the techniques of the third. Finally, we suggest reasons for why the technological infrastructure that should emerge from a successful integration of the three basic topics can have a significant impact on more successful exploration with substantially reduced risk.

Determination of transversely isotropic velocity parameters at the Pluto Discovery, Gulf of Mexico

The collection of both a checkshot and multi-level walkaway vertical seismic profile (VSP) in the initial exploration well at the Pluto Discovery in the Mississippi Canyon area of the Gulf of Mexico provided a unique opportunity to assess and model the velocity anisotropy present in the area. Pre-well depth estimates based on a 3D pre-stack depth migration were found to give horizon depths that were five percent greater than those actually encountered in the well. A transversely isotropic (TI) velocity model was chosen to image the velocity anisotropy, which, for P-wave propagation, depends essentially on three independent parameters the interval stacking velocity and two unitless parameters, delta and eta. The sediment velocity model for the Pluto Discovery area was divided into five layers conformable with the water bottom surface. Constant delta and eta parameters were determined for each layer. These two parameters combined imply a horizontal P-wave velocity that is eight to sixteen percent faster than the vertical P-wave velocity in the top 14,000 feet of sediments. Travel time estimates from the ray tracing program agreed with the checkshot and walkaway VSP travel times to within one percent, when anisotropy was taken into account, as opposed to six percent for isotropic ray tracing. Based on this information, a 3D poststack depth migration using the TI velocity model showed, versus an isotropic velocity model, correct positioning of events in depth and better resolution of high dipping events. These are outcomes that we would expect from a TI velocity model migration. A checkshot and offset VSP were collected in the second exploration well at the Pluto Discovery. The checkshot data showed a similar vertical velocity structure to that indicated in the first well. The receiver depth range used for the offset VSP included greater depths than in the first well, so the anisotropic nature of the area could be explored deeper. For at least this sedimentary sub-basin, there appears to be little lateral change in the velocity or TI parameters. The TI velocity model at the Pluto Discovery is useful in correcting the depth implied from the pre-stack depth migration and in computing the correct travel paths of the seismic energy. Additionally, a post-stack depth migration using the TI velocity model resulted in sharper images from the sedimentary section. INTRODUCTION It has been known for some time that seismic waves in the earth can travel with different velocities depending on the direction of travel. In most localities this difference in seismic velocity with direction is small. However, in some areas the difference is large enough to cause distorted seismic images and inaccurate depth predictions. The collection of both a checkshot and multi-level walkaway VSP at the Pluto Discovery in the Mississippi Canyon area of the Gulf of Mexico provided an opportunity to assess and model the velocity anisotropy present in the area. Pre-well depth estimates based on a 3D pre-stack depth migration were found to give horizon depths that were five percent greater than those actually encountered in the well. A checkshot could easily correct this problem; however, a broader assessment and interpretation of the velocity anisotropy were desired. A walkaway VSP acquired in the sidetrack well provided the data necessary to measure the vertical versus horizontal velocity in the area. A transversely isotropic velocity model was chosen whose parameters (described below) were taken from Thomsen (1986), and Tsvankin and Thomsen (1994). The Pluto Discovery is in the Mississippi Canyon area of the Gulf of Mexico, approximately 150 miles southeast of New Orleans. The discovery lies in about 2,800 feet of water. The partners in the drilling of the well were British Petroleum, Chevron, and BHP Petroleum. The first well was drilled to 22,389 feet. A checkshot survey was recorded at approximately 250 foot intervals throughout the length of the original well. A sidetrack well was oriented northward from the original hole to investigate additional targets. A walkaway VSP survey was recorded at two different levels in the sidetrack well. An isotropic earth model is easy to imagine. The velocity of the seismic wave at any point within the earth section depends not on its direction, but rather on its spatial location. This holds true even for a vertically gradient earth. For a TI earth, the velocity is considered constant within a horizontal plane, but that velocity is different from the vertical velocity. A TI earth can arise in three main ways. First, a truly anisotropic rock has velocities that differ parallel and perpendicular to bedding planes. A shale could be seen to have this characteristic. Secondly, a series of isotropic rock layers can give rise to an anisotropic earth. 1998 SEG Expanded Abstracts TI velocity parameters at Pluto, GOM Lastly, stress fields within the earth may cause velocities to vary with direction. It is the first two of these mechanisms, anisotropic rocks and/or layered isotropic rocks, that are thought to give rise to the anisotropy found in the Gulf of Mexico. The horizontal velocity is generally greater than the vertical velocity, and theoretically must be if the TI anisotropy arises from layers of isotropic rocks. The parameters necessary for completely characterizing a TI earth include the interval stacking velocity, delta, and eta. The delta parameter is the percentage difference between the squares of the interval stacking velocity and interval checkshot velocity. The delta parameter has been thought of as the vertical anisotropy and corrects the seismic stacking velocity to the checkshot velocity. Eta is a measure of the difference between the squares of the horizontal and stacking velocities. VELOCITY COMPARISONS There are several different relevant velocity functions useful in seismic exploration. Checkshot velocities provide the definitive tie between seismic and well data. Stacking and migration velocities are used in processing seismic data and are normally picked to provide the best seismic image. Thus they are not necessarily tied to the petrophysical parameters of the geologic section. Figure 1 shows that there is a distinct difference between these three velocity functions of checkshot, stacking, and migration for the Pluto area. The stacking velocity function is the average of the velocity functions from the velocity control locations near the well. The migration velocity function is the one used in the prestack depth migration of the seismic data set. Obviously if either the stacking or migration velocity functions were to be used for time-depth conversions, the estimated depths would be too deep. The pre-stack depth migration for Pluto proved to have depth estimates that were 5% too deep at TD.