- Home
- Conferences
- Conference Proceedings
- Conferences
79th EAGE Conference and Exhibition 2017
- Conference date: June 12-15, 2017
- Location: Paris, France
- Published: 12 June 2017
1 - 20 of 1073 results
-
-
A New Method for 2D NMR Log Data Inversion in Tight Sandstone Reservoir
More LessSummaryTo improve the inversion effects of 2D NMR log data of the tight sandstone reservoir, this paper applies the two parameters regularization inversion method to the data. The principle of the method is elaborated in detail. The results of applying the method to the simulated 2D NMR log echo data indicate that two parameters regularization inversion method, which can overcome ill-posed nature of 2D NMR log data inversion and obtain a high inversion precision, is much suitable for inversing 2D NMR log data. The inversion results of the 2D NMR log data of tight sandstone reservoir using the two parameters regularization algorithm can be uesed to identify reservoir fluids.
-
-
-
Estimating Permeability in a Carbonate Reservoir of Campos Basin Combining Well Logs and Multiple Linear Regressions
Authors P. Almeida, A.A. Carrasquilla and A. CarrasquillaSummaryFrom the data set of two wells drilled in a carbonate reservoir in Campos Basin - Southeastern Brazil, this article made the estimation of permeability using geophysical well logs in conjunction with the multiple linear regression technique. The main difficulty to characterize this type of reservoirs consists in that the carbonates have a broad variance in their physical attributes. For this understanding, it was then projected to separate the reservoir into zones according to their different petrophysical features. So, based on gamma rays, resistivity, density, neutronic and sonic logs, certain parameters of the reservoir such as irreducible water saturation, top and bottom of the reservoir, oil/water contact, porosity and oil-producing areas were determined. In addition, geological attributes were applied to validate qualitatively the study, when the reservoir was divided into zones in conformity with the environment energy at the time of deposition of the deposits. In the final, the estimate showed good fit with the permeability measurements in the laboratory through of this simple and fast method.
-
-
-
Integrated Visualization of Subsurface Appraisal and Field Development Plan - A Key Input to Support Post Discovery Capital Intensive Investment Decisions
Authors H. Stigliano, I. Yemez and V. SinghSummaryAfter a field discovery, E&P companies spend billions of dollars every year for subsurface appraisal. Most often, the appraisal decision is difficult and tortuous, lacks transparency and objectivity, and the level of appraisal is less than optimal which affect the full cycle project economics. This paper is an effort to manage uncertainty through appraisal by quantifying its impact on the field development plan and overall project profitability. An integrated approach of selecting appraisal well location and their sequencing using standard discounted cashflow analysis was adopted to calculate commercial value of a project in terms of an economic indicator considering different uncertain variables, their dependencies and correlation between them. To quantify the specific appraisal activity benefits, in terms of uncertainty reduction and its economic reward to the field development plan for an offshore clastic reservoir discovery, the Value of Information approach along with decision tree analysis was used and a comparison between different evaluated scenarios was made for an optimal appraisal selection. This method has helped prioritizing the appraisal locations including their sequencing, ensuring that appraisal plan adds value and allows management to understand the post discovery risks/opportunities for future investments.
-
-
-
Faster reservoir simulation allows multi-million cells models: are fine grids requiring fine reservoir geologist?
Authors F. Bigoni, A. Francesconi, V. Tarantini and R. MarinoSummaryIn the last years, the better performance of computational centers and the more powerful software allow the building of 3D reservoir models with higher number of active cells and the use of the fine geological static model to make dynamic simulations.
In the past, reservoir geologists experienced some frustration when their detailed 3D fine grid geological description resulted almost obliterated by the upscaling processes. Given the new context, geologists can drop their discomfort for the upscaling smoothing effect and focus on a reservoir description “fit for grid”, which can result much more detailed and closer to the real complexity of the rock.
However, this process is not trivial calling for a much more detailed and sharp geological description. In fact, the upscaling process in some way could mask geological wrong assumptions because they resulted smoothed and mitigated by the upscaling process.
-
-
-
A Semi-Empirical Model for Interpreting Rock Strain Sensitivity in 4D Seismic Data
Authors C. MacBeth, A. Kudarova, P.J. Hatchell and J. BrainSummarySeveral models have been recently proposed to connect observations of velocity change with strain deformation in and around reservoirs undergoing production and recovery. In this work we show that a simple compliance-based model combined with the original conceptual understanding of Hatchell and Bourne (2005) can adequately explain the magnitude of R factor values currently observed from calibrated field data in a variety of settings. The model is also used to determine an expression for the gradient of overburden time-shift variation with incidence angle. This gradient is predicted to be low but may vary according to the ratio of tangential to normal compliance at the intergranular contacts. This factor could perhaps be used as an additional parameter to assess the post-production state of the overburden.
-
-
-
Monitoring the Ormen Lange Field with 4D Gravity and Seafloor Subsidence
Authors M. Vatshelle, M. Glegola, M. Lien, T. Noble and H. RuizSummaryThe primary development drilling at the Ormen Lange field at the Norwegian continental shelf concluded in 2014. Reservoir monitoring is employed to reduce subsurface uncertainties and assist with late field-life development decisions. Geophysical methods used include 4D seismic, seafloor geodesy and field-wide 4D gravity and subsidence surveys. This abstract presents the results on both seafloor subsidence and 4D gravity from the surveys performed in 2012 and 2014. These results are provided at a lower cost and significantly faster turnaround compared to 4D seismic.
While 4D seismic does not provide sufficient sensitivity to seafloor subsidence at Ormen Lange because of oceanographic variations, 4D gravity and subsidence surveys provide a clear picture, that is key for understanding and monitoring reservoir compaction.
4D gravity results, in turn, provide a valuable input to understanding mass changes in the reservoir. This input has less lateral resolution than that from 4D seismic, but as gravity is sensitive to mass changes, it provides immediate insight into the energy balance of the field and can be quickly integrated into history matching workflows. The integration of gravity and time-lapse seismic provides a reduction of uncertainties on aquifer influx and strength, with increased confidence through the consistency of independent methods.
-
-
-
Time-Lapse Seismic Signal in the Overburden of Producing Reservoir - Separating Possible Causes
Authors D. Rappin, Y.-M. Leroy and J. FioreSummary4D time-lapse seismic is a standard technology nowadays to describe reservoir property changes during field production, especially pressure variations and fluid movements. Seismic information does not directly provide these physical properties, but seismic velocity changes or impedance changes can be obtained throughout inversion processes.
In this paper, we present several examples of 4D signature encountered in the overburden of different producing reservoirs, in various geological contexts.
The impact in the overburden of producing reservoir is not straightforward. There are many potential causes separate or combined to explain the observed 4D time-shift. Geomechanical effects became obvious in many cases. Effective displacements, subsidence, fault activation and fracturing, can also have significant local contributions. In addition, fluid evolution in small sandy levels located in the overburden can contribute to the observed 4D time-shift.
Several cases and modeling are analyzed with different complementary workflows in order to separate and determine the weight of contribution of the different causes.
-
-
-
A Practical Tool for Simultaneous Analysis of 4D Seismic Data, PEM and Simulation Model
Authors A. Briceño, C. MacBeth and M.D. MangriotisSummaryOne of the main focuses of the geoscience industry is to constrain reservoir models to 3D and 4D seismic data using quantitative workflows that are suitable for model updating and history matching. Seismic history matching (SHM) closes the loop and minimizes the misfit between the observed 4D seismic and that predicted by the reservoir model. A key problem in formulating this misfit function is a lack of understanding as to how uncertainties in the seismic data, petroelastic model and simulation model interact. This can lead to lengthy and time consuming workflows. This study presents a simple and interactive way of visualizing these uncertainties whilst optimizing the SHM. The approach is applied initially to a synthetic example, and then to two different field datasets from the UKCS and Norwegian Sea. The results demonstrate that qualitative updates can be successfully applied to the simulation model in the presence of uncertainty in the PEM and noise in the 4D seismic data.
-
-
-
Wavefield Reconstruction Using a Whole Space Green’s Function Framework
By P. TerenghiSummaryA whole-space Green’s function framework for the representation of the acoustic wavefield is obtained from the Helmholtz wave equation. Under this framework, the wavefield is represented by a distribution of point-sources in space and time, each corresponding to an elementary spherical wavefront.
Intuition suggests the number of equivalent point sources actually required to represent the wavefield of interest might be small compared to the total number of speculative locations. Consequently, in spite of the problem’s innate ill conditioned nature, a satisfying solution can still be obtained by means of sparsity promotion. When a solution is reached, the wavefield can be reconstructed at new or existing locations by utilizing the framework in a forward modelling sense.
-
-
-
F-xy Domain 2D Mathematical Morphological Filtering for Seismic Data Interpolation
More LessSummarySeismic data interpolation is a currently popular research subject in modern reflection seismology. Compressed sensing (CS) based and rank-reduction based methods have great performances in recovering randomly missing traces, but their assumption or precondition, that traces miss randomly, is too hard to satisfy in real data. In this abstract, we propose a new morphology based method, which utilizes 2D mathematical morphological filtering (MMF) to recover missed energy of each frequency component. Because the morphological calculation is based on logical operation and set theory, which is different from the traditional mathematical transforms, this proposed f-xy domain 2D MMF method has strong anti-aliasing capability. Application of it on synthetic and field seismic data demonstrates a successful performance.
-
-
-
Simultaneous Reconstruction and Denoising of 5-D Seismic Data Using Damped Sparse Representation Based Projection Onto Convex Sets
More LessSummaryIrregular landform and obstacles result in a irregular field data acquisition. However, the existing seismic data processing methods and techniques are almost based on the assumption and precondition of regularly input data. Besides, the observed seismic data always contains random noise because of impacts from acquisition equipments or the acquisition environment. Thus methods for reconstruction with the presence of noise are necessary. The simultaneous reconstruction and denoising problem can be effectively solved under the theory of compressed sensing (CS), and the projection onto convex sets (POCS) is one of the effective methods to solve the CS problem. In this abstract, we propose a damped sparse representation (DSR) based POCS method. By a introduced damping operator, the DSR based POCS method can obtain a more accurate estimation of signal, namely, a better result of simultaneous reconstruction and denoising. The feasibility of the proposed method is validated via both 5-D synthetic and field data examples.
-
-
-
A Geometry Dependent Lookup Table Approach to Improve Performance of the OMP Seismic Interpolation Method
Authors Y. Hollander, I. Degani, O. Yilmaz and R. LevySummaryWe present an approach to improve runtime of the \emph{orthogonal matching pursuit} (OMP) method for multidimensional seismic interpolation. OMP is an expansion of \emph{Anti-leakage Fourier Tranfrom} (ALFT) which seeks to further minimize spectral leakage of the computed spatial spectrum. Both methods suffer from low performance efficiency due to the need for iterative computations of forward and backward Fourier transforms. An approach to reduce the necessity of these repeating transform computations was recently presented for the ALFT method. Here we generalize this improvement to OMP, and show a substantial runtime improvement of a factor 4–5 compared to the old approach.
-
-
-
Mapping of Moveout in Orthorhombic Media
By A. StovasSummaryI define the mapping operator for moveout in tilted 3D medium. The proposed operator is applied for orthorhombic medium with rotation by using 3 Euler angles. It allows to define the kinematic properties in tilted orthorhorhombic medium, and can be used to map any of moveout approximation originally defined for vertical orthorhombic medium.
-
-
-
Multifocusing Imaging Comparisons Study Based On Cuckoo Search Algorithm
Authors D.Y. Chang, C. Zhang, T. Hu and W. FengSummaryConventional NMO method causes severe stretching distortion. Moreover, for rugged topography data, conventional NMO and stack are based on the elevation static correction, which can be inaccurate because the assumption of vertical ray path is no longer valid. MF method has great advantages because it not only brings no stretching distortion but also can be directly applied without elevation static correction. MF-based stack method can improve SNR on stack section because this method can stack image traces from different CMP gathers. However, the determination of three MF parameters is the most key as well as difficult issue in MF method. In this article, we introduced a new metaheuristic global optimization algorithm, Cuckoo Search (CS), to calculate three parameters. On physical model with rugged topography data, we tested two MF methods, relief surface MF method and implicit spherical MF method. Both methods show better time correction result without elevation static correction than NMO result, while relief surface MF has better time correction than implicit spherical MF because the latter one calculate corrections without considering sources’ and receivers’ elevation.
-
-
-
Automatic Microseismic Event Detection by Multi-scale Morphology Characteristic Function
Authors Y.Y. Jiang, R.Q. Wang, X.Q. Chen, Y. An and M. ZhangSummaryMicroseismic events detection is a key issue for field real-time microseismic data processing because of the low signal-to-noise ratios. Surface-acquired microseismic events are commonly unpredictable and appear as weak signals, which presents a significant challenge in microseismic data analysis and event detection. A new implementation of the multi-scale morphology characteristic function is proposed for detecting microseismic events automatically. First, the microseismic data are decomposed into different scales by multi-scale morphology decomposition. Then the multi-scale morphology characteristic function is calculated by waveform correlation approach. We have tested the technique on a surface passive seismic monitoring dataset of the microseismic events induced by hydraulic fracturing, and it is proved to be effective by simulation and real data processing. The proposed method has the advantage of suppressing the effect of Gaussian noises and is applicable to the detection of low SNR signal.
-
-
-
Normalized Set of Global Effective Parameters for Pure-Mode and Converted Waves in Horizontally-layered Triclinic Media
More LessSummaryConsidering reciprocity where the traveltime is an even function of the offset or horizontal-slowness, the fourth-order normal moveout (NMO) series are governed by the normal-incidence time and eight effective parameters: three second-order and five fourth-order. Local effective parameters are related to the individual layers, while the global effective parameters are related to the overburden multi-layer model. Local and global parameters are related by forward and inverse Dix-type transforms. The NMO formulae are different in the slowness and offset domains, but the eight parameters are the same in both cases. We suggest a new set of intuitive normalized effective parameters, classified into two “azimuthally isotropic” and six “azimuthally anisotropic” parameters. We provide feasible ranges for the normalized parameters, thus allowing their used for controlled inversion.
-
-
-
Scalable numerical modelling patterns for the cloud
Authors J.W.D. Hobro and A. SharmaSummaryThe ready availability of cloud computing resources presents an opportunity for rapid turnaround in large numerical modelling tasks. This opens up new possibilities for interactive modelling in immersive applications. However, it is only feasible for numerical modelling algorithms that scale well over large computing clusters. This is relatively straightforward for algorithms that are embarrassingly parallel, but achieving linear scaling for algorithms such as coupled numerical modelling problems is much more challenging. We explore the improvements that can be achieved in scalability for this type of algorithm by moving away from a sequential programming approach as conventionally used with the Message Passing Interface (MPI), which encourages large-scale synchronisation across a parallel system. Instead we propose an approach based on the actor model that removes all unnecessary synchronisation during inter-node communication. It avoids synchronisation by introducing flexibility in the order of computation. It also spreads communication evenly in time and therefore significantly reduces network contention. We use a theoretical model to examine the scalability characteristics of the new system, which gives improvements of more than an order of magnitude in scalability compared with the sequential approach.
-
-
-
Finite-difference Seismic Modeling on CPUs/GPUs Using Matrix-vector Products
Authors F. Wittkamp, T. Bohlen and T. SteinwegSummaryModern seismic imaging methods such as reverse time migration (RTM) or full-waveform inversion (FWI) require large high-performance computing (HPC) systems to provide enough computational power to solve a large number of forward problems based on the wave equation. These wavefield simulations are conventionally performed by explicit time-domain finite-difference (FD) methods on regular numerical grids, where the parallelization is often based on a fixed and rather inflexible decomposition of the computational domain. However, such parallelization cannot exploit the computing capacities of modern and especially future exascale HPC architectures, which are expected to become more and more hierarchical and non-uniform. For this purpose, we developed a matrix-vector formulation of the explicit time-domain FD method solving the 3D elastic wave equation. To implement the matrix-vector formalism, we chose the open-source framework LAMA, which allows the development of hardware-independent code. We found that the implementation of such a matrix-vector based 3D elastic forward solver is straightforward. In a strong and weak scaling benchmark, we subsequently explored the scaling behavior of our implementation. The overall scaling performance shows the large potential of our method, which can be improved even further by tuning on the application and framework level.
-
-
-
Optimizing Fully Anisotropic Elastic Propagation On 2nd Generation Intel Xeon Phi Processors
Authors A. Farres, A. Duran, C. Rosas, M. Hanzich, C. Yount and S. FernándezSummaryThis work shows several optimization strategies evaluated and applied to an elastic wave propagation engine, based on a Fully Staggered Grid, running on the latest Intel Xeon Phi processors, the second generation of the product (code-named Knights Landing). Our fully optimized code shows a speed-up of about 4x when compared with the same algorithm optimized for the previous generation processor.
-
-
-
Approximating Q Propagations for Elastic Modeling on GPUs
More LessSummaryPropagating wavefields using the explicit finite difference method is the kernel of reverse time migration (RTM) and high-end velocity algorithms in seismic applications. In recent decades there has been a significant increase of interest in the seismic exploration community to invert the image of the subsurface in larger regions and higher resolutions in the elastic media, which brings tremendous computing challenges. As a result, the optimizing methods for improving the performance of the wavefield propagation are in great demands. This work manages to boost the performance of the wavefield propagation in 3D elastic scenarios by approximating the Q propagation and using the multi-GPU platform. We first extend the constant-Q formulation from the 2D viscoelastic case to the 3D viscoelastic case. Different optimization techniques on GPUs are then described for an efficient modeling kernel. We also propose a set of schemes to reduce the computation to further improve the performance. The experimental results show that we can achieve significant performance speedups of 60 to 200 times with 4 GPUs over the CPU-based solution as a function of Q.
-