Search

Abstract The reservoirs of the Lower Permian He8 formation are typically tight sandstones, which have geologic features such as rapid lateral changes, strong heterogeneities and complex pore structures. Rock physics characteristics are similar between tight sandstones and mudstones, mean while fluid AVO characteristics are unclear. These factors make reservoir prediction and hydrocarbon detection become difficult. The keys to successful exploration are the effective elimination of non-reservoirs, including mudstones and ineffective tight sandstones, and the successful discovery of favorable areas. Rock physics analysis indicates that tight sandstone can be identified by V P /V S , while gas-bearing can be distinguished by fluid compressibility. In order to improve inversion accuracy, a new method of favorable area prediction is proposed. Firstly, predict sandstone boundary by V P /V S inversion under the control of sedimentary setting. Secondly, distinguish effective reservoirs and gas-bearing areas with porosity inversion and fluid compressibility inversion under the control of sandstone boundary. Finally, confirm favorable areas according to V P /V S , porosity and fluid compressibility. The tight sandstones of the He8 formation can be divided into three types. Favorable area Ⅰ is the first choice of exploration, target followed by the favorable area Ⅱ, whereas the unfavorable area Ⅲ is not chosen as exploration target. The new method has achieved excellent results in the favorable area prediction of tight sandstone in the Kangning area and has significant implications for late exploration. Highlights • Propose a new method for favorable areas prediction of tight sandstone. • Pre-stack parameter inversion with facies control improves inversion accuracy. • Determine the favorable area of He8 formation in the Kangning area. [ABSTRACT FROM AUTHOR]

In conducting studies to make reservoir management decisions, it is important to efficiently interpret results of reservoir simulations. An understanding of how and why predicted reservoir performance changes with model parameters guides evaluation of production strategies as well as exploration of the impact of uncertainty in reservoir description. The aim of this work is to demonstrate the use of dimensionless scaling groups to interpret and qualitatively predict simulation results of multiphase flow in subsurface reservoirs with a large number of wells. Dimensionless scaling groups which quantify the balance between the forces causing fluid flow were computed between well pairs to rationalize simulation results. The data required to partition the model according to injector-producer pairs and estimate the scaling groups were obtained within minutes using simplified, single-phase numerical experiments. We show that scaling groups can be used to classify multiphase flow behaviours observed over the field into a small set of flow regimes characterized by the combination of their dominant forces. Changes in fluid distribution and reservoir performance with the model parameters can be analyzed in terms of changes in the force balance, and qualitatively predicted using the scaling groups. Predictions made using scaling groups may guide, and thereby reduce the use of, time-consuming multiphase flow simulations to optimize field development plans, to improve the calibration of reservoir models to production data and interpreted subsurface heterogeneity, and to assess the impact of reservoir uncertainties on production. [ABSTRACT FROM AUTHOR]

Due to a lack of borehole data, reservoir and seal rock presence in frontier basins is typically inferred from seismic reflection data. However, analysis of the geometry and kinematic development of polygonal fault systems (PFS), which almost exclusively form within very fine-grained strata, provides another, largely untested method to infer reservoir and seal rock presence. We here use very high-quality 3D seismic reflection data from offshore Uruguay and a range of quantitative GIS-based techniques to document the planform and cross-sectional characteristics of a basin-scale (>6400 km 2 ) PFS, and to investigate the role that stratigraphic variations in the Tertiary deep-water host-rock have on its geometrical variability and kinematic development. We demonstrate that a series of likely sandstone-rich deep-water deposits occur at the base of and within the main PFS tier. The geometric characteristics and throw distribution on individual polygonal faults suggest these sandstone-rich deep-water deposits represent a mechanical barrier to fault propagation, thus influencing fault height and areal density and, in some cases, strike. We argue that in largely unexplored, deep-to ultra-deep water basins, such as those characterizing offshore Uruguay, the distribution and geometric attributes of PFS can be used to delineate sandstone-rich reservoir bodies. Furthermore, these characteristics may help exploration geoscientists better understand seal heterogeneity and quality in data-poor basins. [ABSTRACT FROM AUTHOR]

Accurate prediction of reservoir porosity and permeability is essential for prospecting hydrocarbon reserves and petroleum production capacity. We propose an innovative reservoir porosity and permeability prediction method through identifying sedimentary-diagenetic facies, determining the porosity-permeability trends using core measurement data, extrapolating the spatial distribution of the sedimentary-diagenetic facies using log data through the Bayes discriminant analysis and predicting the reservoir porosity and permeability. The essence of the method was illustrated and its effectiveness was demonstrated using the Eocene beach-bar sandstones in the Dongying Depression, Bohai Bay Basin, eastern China. The Eocene beach bar sandstones are classified into fine sandstone, siltstone, and argillaceous siltstone based on grain sizes, sorting, and matrix contents. The major diagenetic processes that influence the porosity and permeability of the beach-bar sandstones are compaction, carbonate cementation and feldspar dissolution. Seven sedimentary-diagenetic facies were identified in the beach-bar sandstones based on the lithological types, and their corresponding diagenesis and influence extent on reservoir properties. The variation ranges of porosity and permeability (log K) of these sedimentary-diagenetic facies are typically less than 6% and 1.2, respectively. Both the porosity and permeability have well defined functional relationships with depth. The sedimentary-diagenetic facies can be identified effectively from logging data through the use of the Bayes discriminant analysis and corresponding cross-plots. The porosity and permeability (log K) of the well evaluated in the study were predicted with errors in the range of ±3% and ±0.6, respectively, based on the distribution and the fitting equation of the trend lines for the seven sedimentary-diagenetic facies. The predicted porosities and permeability of sedimentary-diagenetic facies match the measured porosities and permeability well. [ABSTRACT FROM AUTHOR]

Currently, conventional forecasting methods of well-to-seismic integration are unable to identify turbidite channel sandstones due to scarcity of well data in deepwater areas, small geophysical differences between sandstones and mudstones of turbidite channels and strong sandstones heterogeneity. The reservoir prediction of deep-water turbidite channels is still a difficult issue in deep-water research. On the basis of previous studies, we propose a new technology named “reservoir prediction of deep-water turbidite sandstones with seismic lithofacies control” in view of the characteristics of deep-water turbidite sandstones. This new technology improves the reservoir prediction of complex sedimentary systems after classifying seismic lithofacies and connecting lithofacies with rock-physics. Furthermore, it can accomplish the genetic classification statistics of rock-physics, improve conversion accuracy of seismic elastic parameters/reservoir parameters and achieve the quantitative reservoir prediction under the double control of seismic geomorphology and seismic lithofacies. The C block of Lower Congo Basin is characterized by few well data, complex lithology but high resolution-seismic. We use the technology to predict the reservoirs of this area and have achieved excellent results. This has great significance for the later exploration. [ABSTRACT FROM AUTHOR]

• The paleo-tectonic stress was verified to be an important factor for porosity reduction in tight sandstones. • The structural diagenesis of tight sandstone of Ahe Formation was quantitatively characterized. • The favorable reservoir areas were predicted based on structural diagenesis and stress field simulation. Structural diagenesis is an important factor controlling the reservoir characteristics of the Jurassic Ahe Formation (J 1 a) of the Dibei gas reservoir in the northern structural belt of the Kuqa depression, which is mainly manifested as the compaction-attributed porosity reduction by the paleo-tectonic stress and structural fracture formation. Based on the analysis of the reservoir and tectonic stress characteristics, the control effects of the paleo-tectonic stress on porosity reduction and structural fracture characteristics of the Ahe Formation reservoir are quantitatively investigated. With numerical simulation based on the finite element method, the high-quality reservoir area of the Jurassic Ahe Formation of the Dibei gas reservoir is predicted. The results show that paleo-tectonic stress is an important element affecting reservoir physical properties and structural fracture characteristics of the Ahe Formation. The maximum paleo effective compressive stress is negatively linearly correlated with the average reservoir porosity, while it is positively exponentially correlated with the porosity reduction attributed to tectonic compaction. The direction of the maximum paleo effective compressive stress controls the dominant striking direction of structural fractures, and the stress magnitude determines the development degree of structural fractures. A clear positive exponential correlation is found between the maximum paleo effective compressive stress and both the areal density and plane porosity of fractures. The Well DX1-YN2-DB104 district and Well TZ4 district present high gas production, and the circular zone in the central-southern part of the study area has paleo-tectonic stress and reservoir characteristics similar to those of the above two districts. Thus, it is a potential favorable area for high gas production and should be considered a key exploration orientation of the Dibei gas reservoir. [ABSTRACT FROM AUTHOR]

One of the most indispensable works in the oil and gas exploration and exploitation is reservoir properties forecast. We develop a reservoir properties prediction method based on the mixture of relevance vector regression (RVR) experts. For reservoir properties prediction, an individual machine learning model is probably insufficient with limited training data. The mixture of experts can decompose a complicated problem of reservoir properties prediction into several relatively simple sub-problems by incorporating multiple learning models, where each model processes specific parts of the data. In the proposed method, RVR has been chosen as the expert because it results in sparser regressors and determines hyperparameters automatically. It can also project data into a high-dimensional space through kernel functions to resolve the non-linear problem and avoid the curse of dimensionality. At first, a mixture of RVR experts model is trained on the well data samples. The input features are elastic properties and the output is a reservoir property. Then, the learning model is applied to test dataset and the corresponding reservoir properties are generated. The proposed method is applied to two field data, in which the learning model is obtained by training on the well data and is tested on the well and seismic data, respectively. Compared with an individual RVR expert, some metrics, such as mean absolute derivation (MAD), root mean square error (RMSE), coefficient of determination (R 2) and Akaike information criterion (AIC) etc., are effectively improved. The successful implementation of the method demonstrates its feasibility, certifying the superiority of the new method in the aspect of likelihood of fit and accuracy again. • Mixture of experts strategy is helpful to divide and conquer reservoir data. • RVR can determines hyperparameters automatically and avoid the dimension Curse. • MRVRE improves the accuracy of the reservoir property prediction. [ABSTRACT FROM AUTHOR]

The reservoir fluid mobility, which can reflect the pore connectivity, is an important seismic attribute in reservoir development. The current method to obtain reservoir fluid mobility is based on rock physical experiments. However, the current method can only obtain the reservoir fluid mobility at the well location, rather than for a large range within the exploration area. To accurately extract the reservoir fluid mobility of a large portion of an exploration area, we propose a new method of calculating the reservoir fluid mobility using seismic data. First, the algorithm obtains the approximate reservoir fluid mobility by substituting the reflection coefficient of the dominant frequency content at the low-frequency end of the seismic data for its instantaneous spectrum. Then, the synchrosqueezing generalized S-transform and the Lucy-Richardson algorithm (SSGST-LR) are used to calculate the instantaneous spectrum of the reservoir fluid mobility. Rearrangement of the generalized S-transform's results can improve the temporal and spatial resolutions of the time-frequency transform. Moreover, the interference between the different frequency signals can be eliminated more effectively by the Lucy-Richardson algorithm. Finally, by performing fluid replacement in the well, we demonstrate the influence of fluid type on fluid mobility. We used synthetic data and real data to verify the superiority of our method. This example shows that our proposed method can accurately extract the reservoir fluid mobility and predict the distribution of high-quality reservoirs in an exploration area. • The synchrosqueezing generalized S-transform is proposed to improve the resolution of time-frequency transform. • The synchrosqueezing generalized S-transform is introduced into the calculation of the reservoir fluid mobility. • The fluid types in the well were replaced, and the influence of the fluid types on fluid mobility is discussed. [ABSTRACT FROM AUTHOR]