Calibration of Seismic Attributes for Reservoir Characterization (original) (raw)
Related papers
Use and abuse of seismic data in reservoir characterisation
Marine and Petroleum Geology, 2001
All seismic data contain a mixture of signal and noise. In detailed reservoir characterisation, it is commonly dif®cult to distinguish between real features and seismic artefacts. This is especially a problem when interpreting seismic attribute maps. Such maps are widely used tools during reservoir description, but serious pitfalls exist, which may lead to erroneous interpretations and fatal development plans for oil ®elds. A recent interpretation of seismic attribute maps from a seismic survey collected across Gullveig, an oil ®eld located in the northern North Sea, has been used to illustrate how small faults can be recognised and mapped from such maps. We applied available well data (including core data) and two seismic surveys from the same area, and present convincing evidence that the vast majority of linear features seen on the seismic attribute maps are, in fact, seismic artefacts and not faults. The data from Gullveig are furthermore supplemented by observations from the nearby Gullfaks Field and discussions on the topic of seismic noise. We use these observations and discussions to stress the importance of using all available data to guide and quality control the structural interpretation of attribute map features before utilising such interpretations as input to reservoir modelling or well planning. q
Using drilling data to characterize reservoirs
Second International Meeting for Applied Geoscience & Energy
In this presentation, we will discuss some key insights that can be obtained from drilling data to help better characterize and understand the reservoir. These include changes in lithology, inferring rock strength, identifying, and quantifying localized depletion caused by offset producers, fracture detection both natural and induced, and how completions interact with various geohazards. We will also explain how this information can be collected and ultimately used to make decisions on items like optimal drilling target, stacked pay development, and even well spacing. Specifically, we will look at four applications of how the drilling data can be used. 1) Using rig drilling data to estimate rock strength. 2) The use of rig data, gamma ray logs and geosteering to predict natural fractures. 3) The use of rig drilling data and mud gas logs to identify localized fracture depletion. 4) Understanding hydraulic fracture growth in stacked plays and around geologic features Several case studies will be presented that demonstrate the accuracy and applicability of this data, with a focus on using the data to make actionable decisions, as well as some of the data limitations. At the conclusion of this talk, the importance of incorporating drilling data from every well to improve reservoir knowledge will become evident. Even more so when considering this data is readily available, on every well, and is obtained at a very low or no cost, and with no associated operational risks.
Prediction of Reservoir Variables Based on Seismic Data and Well Observations
Journal of the American Statistical Association, 2002
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
SPE 164830 How to integrate basin-scale information into reservoir models
Objectives and scope of the Study In this paper a new approach is presented to consistently integrate basin-scale information into reservoir models. The impact of the quantitative integration of boundary conditions derived from basin-scale modeling on the facies distribution at the reservoir scale is evaluated. To this purpose, a new workflow was defined based on a geostatistical approach. The aim was that of integrating the typical dataset for reservoir geological modeling, comprising well and seismic data, with a potentially new kind of data obtained from 3-D process-based stratigraphic modeling and related to the distribution of the hydrocarbon bearing volumes. Quantitative coherence between the small scale reservoir volume and the large-scale geological setting defined by the basin model was imposed. Synthetic case studies were set up to verify the effectiveness of the method. Applications The entire process was applied to a fluvio-deltaic environment to integrate the basin-derived information, such as (1) the overall reservoir/non reservoir volumes, (2) the 3D distribution of channelized volumes and (3) related flow directions, to the reservoir model. Eventually, the uncertainty reduction in the description of the final facies distribution at the reservoir scale was evaluated. Results, Observations and Conclusions The developed approach proved very efficient to estimate the lithological fraction of the hydrocarbon bearing rocks (i.e. sands in a shaley/clayey environment). The lithological fraction is of crucial importance during the appraisal phase of a reservoir when relevant decisions have to be taken but few wells are drilled and, as a consequence, a limited amount of data is available to perform a reliable volumetric estimate. Furthermore, the prediction of the 3D facies architecture (such as the channel pattern in a fluvial depositional environment) can effectively assist in the well planning strategy. Besides, the overall uncertainty affecting a reservoir model can be assessed; this uncertainty is both a function of the initial environmental parameters for basin modeling and of the adopted methodological approach for basin-to-reservoir data integration. Therefore, an accurate inference of the basin parameters is needed to achieve a reliable prediction of both the channel location and the sand/shales volumes fractions. Significance of subject matter Reservoir modeling can significantly benefit from the integration of quantitative basin-scale information. In particular, the numerical modeling of the stratigraphic sequence can be used to steer the reconstruction of the reservoir internal geometry and to reduce the uncertainty in the distribution of the hydrocarbon-bearing lithologies. Furthermore, this approach provides a rigorous assessment of the information content of all the available data and thus it might be very useful to guide further data acquisition campaigns.
ECLIPSE, FrontSim, MultiWave Array and RFT (Repeat Formation Tester) are marks of Schlumberger.
2024
Accurate reservoir characterization is critical for optimizing hydrocarbon exploration and production. This study explores the integration of seismic inversion and geostatistical modeling, leveraging seismic attributes and well-log data to enhance lithofacies estimation and reservoir property prediction. The research addresses the challenges of combining multiple data sources to improve the spatial resolution and accuracy of reservoir models. The workflow begins with the acquisition and preprocessing of seismic and well-log data, followed by seismic inversion to derive high-resolution subsurface properties. Geostatistical modeling is then employed to integrate seismic attributes with well-log data, providing a robust framework for predicting lithofacies distribution and reservoir properties. The study evaluates the effectiveness of this integrated approach through a detailed analysis of seismic attribute interpretation, lithofacies classification, and reservoir property distribution. Validation of the models against existing methods demonstrates significant improvements in accuracy and resolution, highlighting the potential of this approach for complex reservoir environments. Key findings reveal that the integration of seismic attributes with well-log data not only enhances the reliability of lithofacies models but also provides a more detailed understanding of reservoir heterogeneity. This research contributes to the advancement of reservoir characterization techniques by offering a practical and scalable solution for improved hydrocarbon recovery. The study concludes with recommendations for applying this approach to diverse geological settings and identifies avenues for future research in the integration of advanced geostatistical methods and machine learning techniques.
Heliyon (Earth Sciences), 2020
This study presents a correct time and depth correlations with enhanced velocity analysis, based on two reservoir horizons mapped across two wells (Osl-1 and Osl-2). It involves the use of high-resolution images to delineate the complex geological structures associated with Reservoir A-horizon (R-A h) and Reservoir B-horizon (R-B h) based on 3-D seismic sections and wireline logs. It focuses on showcasing magnified images of the well to seismic tie (W-S T), to enhance appropriate times and depths posting to aid correct determination of the pay thicknesses (P t), drainage areas (A d) and the mapping of other probable areas within the hydrocarbon field. The idea is to magnify the points of interested at very close intervals (2 feet) to enable the mapping of the actual positions and times of events within the reservoirs. The aim is to enhance better results and confidence in the interpretation, as such, reduce the uncertainty regarding hydrocarbon viability and volume estimation. R-A h is tracking below 9550 feet and 2.460 s in Osl-1. It is below 9510 feet at 2.450 s in Osl-2. Similarly, R-B h is tracking below 10550 feet at 2.655 s in Osl-1 and below 10520 feet at 2.650 s in Osl-2. R-A h is about 70 feet (21.34 m) thick across Ols-1 and Osl-2 while R-B h is 70 feet (21.34 m) thick and 100 feet (30.48 m) in Osl-1 and Osl-2 respectively. In total, A d is 172 acres (69.6 Â 104 m 2) for R-A h and 206 acres (83.4 Â 104 m 2) for R-B h while the P t is 140 feet (42.67 m) for R-A h and 170 feet (51.82 m) for R-B h. Possible wellbore positions to aid future developmental activities could be within the southeast , southwest and northwest directions of Osl-1 and Osl-2. The field is viable with regards to hydrocarbon availability, and the use of high-resolution images has aided accurate evaluation of P t and A d , hence, increased the confidence in the results of the interpretation.
ASEG Extended Abstracts, 2016
Seismic facies classification has been used to reduce risk in laterally heterogeneous reservoir prediction. Studies were focused on the Barrolka/Coolah/Durham Downs Trend in the southwest Queensland sector of the Cooper Basin. The primary reservoir targets are the fluvial channel sediments of the Toolachee Formation. Historically, well success rates across the area have been low, with highly variable reservoir development and connectivity identified as the limiting factors influencing well performance. Conventional seismic attribute analysis has typically yielded inconclusive results, often associated to the presence of thick coals that dominate the seismic response. However, recent drilling campaigns utilised seismic waveform classification mapping, which resulted in an increase in technical success rate of wells. This study aims to investigate the concepts behind the success of the waveform classification method and to determine alternate techniques to further delineate reservoir presence. Key outcomes from rock physics studies indicate subtle variances in the seismic wave shape could be attributed to changing reservoir thickness underlying coal formations. Cross correlation of the wave shape against well results confirmed the concept of dimming seismic amplitude response to be related to increased reservoir thickness. In an attempt to capture the lateral extent of these variances, three adjacent 3D seismic volumes, covering majority of the complex, were subject to a variety of attribute analysis methods. Unsupervised waveform classification was found to be the most efficient and effective method for capturing the dimming seismic reflector, and thus defining the channel trends. The strong correlation between waveform class and reservoir thickness measured in wells enabled the generation of risk-segment maps for the reservoir units. Observed changes in wave shape on full stack seismic data have been related to lithological variations based on rock physics studies. These learnings have been used to select attribute extraction techniques that can best highlight the wave shape variations. Using seismic facies mapping to enhance reservoir prediction capabilities has reduced reservoir risk and improved the technical success rate of drill targets. The output maps from this study have been used to locate future opportunities in the region.