Craig S. Calvert

Characterization of stratigraphic architecture and its impact on fluid flow in a fluvial-dominated deltaic reservoir analog: Upper Cretaceous Ferron Sandstone Member, Utah

Fluviodeltaic stratigraphic architecture and its impact on fluid flow have been characterized using a high-resolution, three-dimensional, reservoir-scale model of an outcrop analog from the Upper Cretaceous Ferron Sandstone Member of central Utah. The model contains two parasequence sets (delta complexes), each with five or six parasequences, separated by an interval of coastal plain strata. Each parasequence contains one or two laterally offset teardrop-shaped delta lobes that are 6 to 12 km (4–7 mi) long, 3 to 9 km (2–6 mi) wide, 5 to 29 m (16–95 ft) thick, and have aspect ratios (width/length) of 0.4 to 0.8. Delta lobes have a wide range of azimuthal orientations (120) around an overall east-northeastward progradation direction. In plan view, delta lobes in successive parasequences exhibit large (as much as 91) clockwise and counterclockwise rotations in progradation direction, which are attributed to autogenic lobe switching. In cross-sectional view, parasequence stacking is strongly progradational, but a small component of aggradation or downstepping between parasequences reflects relative sea level fluctuations. We use flow simulations to characterize the impact of this heterogeneity on production in terms of the sweep efficiency, which is controlled by (1) the continuity, orientation, and permeability of channel-fill sand bodies; (2) the vertical permeability of distal delta-front heteroliths; (3) the direction of sweep relative to the orientation of channel-fill and delta-lobe sand bodies; and (4) well spacing. Distributary channel-fill sand bodies terminate at the apex of genetically related delta lobes and provide limited sand body connectivity. In contrast, fluvial channel-fill sand bodies cut into, and connect, multiple delta-lobe sand bodies. Low, but non-zero, vertical permeability within distal delta-front heteroliths also provides connectivity between successive delta-lobe sand bodies.

A comparative study of reservoir modeling techniques and their impact on predicted performance of fluvial-dominated deltaic reservoirs

Multiple techniques are available to construct three-dimensional reservoir models. This study uses comparative analysis to test the impact of applying four commonly used stochastic modeling techniques to capture geologic heterogeneity and fluid-flow behavior in fluvial-dominated deltaic reservoirs of complex facies architecture: (1) sequential indicator simulation; (2) object-based modeling; (3) multiple-point statistics (MPS); and (4) spectral component geologic modeling. A reference for comparison is provided by a high-resolution model of an outcrop analog that captures facies architecture at the scale of parasequences, delta lobes, and facies-association belts. A sparse, pseudosubsurface data set extracted from the reference model is used to condition models constructed using each stochastic reservoir modeling technique. Models constructed using all four algorithms fail to match the facies-association proportions of the reference model because they are conditioned to well data that sample a small, unrepresentative volume of the reservoir. Simulated sweep efficiency is determined by the degree to which the modeling algorithms reproduce two aspects of facies architecture that control sand-body connectivity: (1) the abundance, continuity, and orientation of channelized fluvial sand bodies; and (2) the lateral continuity of barriers to vertical flow associated with flooding surfaces. The MPS algorithm performs best in this regard. However, the static and dynamic performance of the models (as measured against facies-association proportions, facies architecture, and recovery factor of the reference model) is more dependent on the quality and quantity of conditioning data and on the interpreted geologic scenario(s) implicit in the models than on the choice of modeling technique.

Spectral Component Geologic Modeling: An Improved Method for Integrating Seismic Data into Geologic Models

Geologic modeling and reservoir simulation provide information critical to successful commercialization of discovered and undeveloped reserves, as well as to the effective management of producing reservoirs. Often, uncertainty associated with the reservoirs characterization and with the geologic model can be reduced by integrating interpretations from multiple data sources, most commonly from the well and the seismic data. These data types provide very different, yet complementary information to the geologic model. The well data provide information over a broad range of scales, but these data are only sparsely sampled in the reservoir. The seismic data are more densely sampled, but provide only a limited scale of information. The integration of these two data types should properly account for their scale differences. Traditional geostatistical simulation methods do not properly account for the fact that the seismic data and the well data represent very different scales of information. Spectral Component Geologic Modeling (SCGM) properly accounts for the difference in scale between the seismic and the well data. A spectral component is an interpreted map or volume of a rock property. All spectral components contain information over a limited frequency bandwidth. A tentative geologic model is built by mathematically combining spectral components; each component providing information to the model that is limited with respect to its frequency content. The tentative geologic model is then constrained to honor the desired spatial continuity of the modeled property, the statistical distribution (histogram) of that property, and the properties measured at the well locations. Spectral simulation is used to honor the desired spatial continuity of the modeled property. With this method, the exact and non-stationary continuity information contained within the combined spectral components can be preserved in the constrained geologic model. As a result, depositional features such as sinuous channel sands, which are observed in the spectral components, are preserved in the model. An application of SCGM is presented. Results clearly show that the SCGM model honors both the spatial distribution and the spatial continuity of the modeled property, as they are represented in the spectral components. Introduction For most developments, there is economic pressure to keep the well count at a minimum throughout the entire life of the field. Well data alone may be insufficient to adequately describe the reservoir and to constrain the assignment of properties in the geologic model. 3D seismic interpretations in the form of maps and volumes are often integrated into the geologic model to help constrain these assignments. However, there are issues associated with utilizing seismic data for this purpose, possibly the most important being that of the differences in scale between the seismic and the well data. Traditional geostatistical simulation methods, such as simple kriging with varying local means and collocated cokriging (Goovaerts), do not properly account for the fact that the seismic data represent a very different scale of information than do the well data. Block cokriging and simulated annealing methods attempt to account for this difference by assuming that the calibrated seismic data represent vertically averaged rock properties over a specified thickness of the reservoir (Behrens et al., Deutsch et al., Doyen et al., and Behrens and Tran). However, particularly for thick reservoir intervals, the seismic data does not contain the scale of information that is represented by a vertical average, and a simple vertical average does not capture the seismic-scale heterogeneity that is present in that reservoir interval. To properly integrate seismic-scale information into the geologic model, one must consider the scale of the seismic data in terms of its frequency bandwidth. Gilbert and Andrieux propose a method whereby a geologic model of porosity is constructed by summing two independently modeled frequency components, a low-frequency component corresponding to frequencies consistent with the seismicimpedance spectrum, and a high-frequency component corresponding to frequencies above the seismic spectrum. Both components are modeled with Sequential Gaussian Simulation, and both are conditioned to well data that have SPE 91054 Spectral Component Geologic Modeling: An Improved Method for Integrating Seismic Data into Geologic Models Glen Bishop, Craig Calvert, Lincoln Foreman, Tingting Yao, and Kamal Bhuyan, ExxonMobil Upstream Research Company

Seismically Integrated Geologic Modeling: Guntong Field, Malay Basin

Exxon Production Research Company, Esso Production Malaysia Inc., and Petronas Research & Scientific Services participated in a joint research project on seismically integrated reservoir modeling involving the Guntong Field, Malay Basin. The goal of this effort was to develop and test a process for interpreting reservoir properties from 3-D seismic data and for integrating these data into the building of 3-D geologic models that would be suitable for use in flow simulation studies. The project produced a 3-D geologic model for three reservoir intervals (IR010, IR023, IR025) and three predominantly non-reservoir intervals. Each reservoir interval was subdivided into facies that were determined by integrating core; well log, and seismic interpretations. Predictions of porosity and lithology used in building the geologic model were made using seismic attributes calculated from acoustic impedance data. The strong DHI in the reservoir intervals assisted in interpreting lithofacies. The geologic model was built following a sequential process that produced a lithology model, a porosity model, and a permeability model. Each model was generated using geostatistical simulation techniques that integrated seismically interpreted facies, reservoir properties, and reservoir property continuity into the process of cell estimation. Two sets of geologic models were built to assess the contribution of seismically derived reservoir properties to the accuracy of the geologic models. Flow simulation results show a significant improvement in history match using the models that integrated all seismic information.

Seismically Integrated Geologic Modeling: Guntong Field

Exxon Production Research Company, Esso Production Malaysia Inc., and Petronas Research & Scientific Services participated in a joint research project on seismically integrated reservoir modeling involving the Guntong Field, Malay Basin. The goal of this effort was to develop and test a process for interpreting reservoir properties from 3-D seismic data and for integrating these data into the building of 3-D geologic models that would be suitable for use in flow simulation studies. The project produced a 3-D geologic model for three reservoir intervals (IR010, IR023, IR025) and three predominantly non-reservoir intervals. Each reservoir interval was subdivided into facies that were determined by integrating core, well log, and seismic interpretations. The reservoir properties of each facies were modeled independently, producing a geologic model that preserved the unique distribution of reservoir properties that characterized each facies. Predictions of porosity and lithology used in building the geologic model were made using seismic attributes calculated from acoustic impedance data. The strong DHI in the reservoir intervals assisted in interpreting sandy versus shaly lithofacies. The geologic model was built following a sequential process that produced a lithology model, a porosity model, and a permeability model. Each model was generated using geostatistical simulation techniques that integrated seismically interpreted facies, reservoir properties, and reservoir property continuity into the process of cell estimation. Two sets of geologic models were built to assess the contribution of seismically derived reservoir properties to the accuracy of the geologic models. The set of seismically integrated models was built using all of the information that could be interpreted from the seismic data. The set of seismically guided models was built excluding only the seismically derived reservoir properties. Flow simulation results show a significant improvement in history match using the models that integrated all the seismic information.