- Home
- Conferences
- Conference Proceedings
- Conferences
73rd EAGE Conference and Exhibition - Workshops 2011
- Conference date: 23 May 2011 - 27 May 2011
- Location: Vienna, Austria
- ISBN: 978-90-73834-13-2
- Published: 27 May 2011
61 - 80 of 129 results
-
-
Seismic Wave Extrapolation in Isotropic and Anisotropic Media using Lowrank Symbol Approximation
More LessWe consider the problem of constructing a wave extrapolation operator in a variable and possibly anisotropic medium. Our construction involves Fourier transforms in space combined with the help of a lowrank approximation of the space-wavenumber wave-propagator matrix. A lowrank approximation implies selecting a small set of representative spatial locations and a small set of representative wavenumbers. We present a mathematical derivation of this method, a description of the lowrank approximation algorithm, and numerical examples which confirm the validity of the proposed approach. Wave extrapolation using lowrank approximation can be applied to seismic imaging by reverse-time migration in 3D heterogeneous isotropic or anisotropic media.
-
-
-
3-D Parallel Frequency-domain Visco-acoustic Wave Modelling based on a Hybrid Direct/Iterative Solver
More LessWe present a parallel domain decomposition method based on a hybrid direct-iterative solver for 3D frequency-domain modelling of visco-acoustic waves. Frequency-domain seismic modelling reduces to the solution of a large and sparse system of linear equations for each frequency. The hybrid directiterative approach aims to overcome the memory requirement and the limited scalability of direct-solver approaches, on the one hand, and to iteratively solve better-conditioned system than in global iterative approaches on the other hand. The domain decomposition is based upon the algebraic Schur complement method. The reduced Schur complement system is solved with the global minimum residual method (GMRES) and is preconditioned by an algebraic additive Schwarz preconditioner. The MUMPS direct solver is used to factorize local impedance matrices defined on each subdomain. Simulations in the salt models for frequencies up to 12.5 Hz show that the number of iterations increases linearly with frequencies when the grid interval is matched to the frequency and the size of the subdomains is kept constant over frequencies. This makes the time complexity of the hybrid approach similar to that of global iterative solvers. Possible improvements of the method for multi-source simulation involve the use of block iterative solver and two levels of parallelism.
-
-
-
2D/3D Seismic Modelling via a Massively Parallel Structured Approximate direct Helmholtz Solver
Authors Shen Wang, Jianlin Xia and Maarten V. De Hoop and Xiaoye LiWe consider the discretization and the solution of the 2D/3D inhomogeneous acoustic wave equation in the frequency domain, which is known as the Helmholtz equation, including variable density, transverse isotropy (VTI & TTI), and attenuation scenarios. In particular, we are concerned with solving this equation on a large physical domain, for a large number of different forcing terms in the context of the 3D seismic modeling. The advantage of seismic modeling in the frequency domain lies on that for a single frequency all solutions share a common Helmholtz operator. We resort to a parsimonious mixed grid finite difference scheme for discretizing the Helmholtz operator equipped with the Perfect Matched Layer (PML) boundaries, yielding a pattern-symmetric but non-Hermitian matrix. We make use of 2D/3D nested dissection based domain decomposition, and introduce an approximate direct solver by developing a new parallel Hierarchically Semi-Separable (HSS) matrices compression, factorization and solution approach. We cast our massive parallelization in the framework of the parallel multifrontal method. The assembly tree is partitioned into local trees, which are stored and eliminated locally at each process, and a global tree, whose elimination arouses massive communications among processes. The entire Helmholtz solver is a parallel hybrid between the multifrontal and HSS structures. The computational complexity associated with the factorization is almost linear in the size, say N, of the matrix, assuming r is the maximum numerical rank of all off-diagonal blocks in the multifrontal procedure. We benchmark our Helmholtz solver with the state-of-the-art MUMPS solver, and show that our solver is at least one order of magnitude faster than MUMPS for the same problem and on the same computing platform. We demonstrate the efficiency and accuracy of our solver via displaying various 2D (BP2004 and BP2007 TTI models) and 3D (SEAM model) numerical examples.
-
-
-
A Hybrid Technique for 3-D Modeling of High Frequency Teleseismic Body Waves in the Earth
More LessIn the last decade, the deployment of dense regional arrays such as the USArray transportable array has considerably improved our capacity to image the interior of the Earth.
-
-
-
Application of the Discontinuous Galerkin Finite-element Method to Seismic Modelling and Imaging
Authors V. Etienne, S. Operto and J. VirieuxWe present a discontinuous Galerkin finite-element method (DG-FEM) suitable to seismic modelling and seismic imaging in large scale 3D elastic media. The method makes use of unstructured tetrahedral meshes locally refined to the medium properties (h-adaptivity) and of interpolation orders that can change from one element to another according to an adequate criterion (p-adaptivity). These two features allow us to reduce significantly the numerical cost of the simulations. While the efficiency of DG-FEM has been largely demonstrated with high interpolation orders, we favour the use of low orders more appropriate to the applications we are interested in. In particular, we address the issues of seismic modelling or seismic imaging in case of complex geological structures requiring a fine medium discretisation.
-
-
-
A Stable TTI Reverse Time Migration
More LessWe propose a stable TTI acoustic wave equation system by revisiting the anisotropic elastic equations. Based on a VTI system of equations which is equivalent to its elastic counterpart, we introduce self-adjoint differential operators in rotated coordinates to stabilize the TTI acoustic wave equations. Compared with conventional formulations, the new equation system does not add numerical complexity and can be solved by an extremely high order central finite-difference scheme. We show examples that our method provides stable and high quality TTI reverse time migration images.
-
-
-
3D Forward Modeling for Full Waveform Inversion
Authors Denes Vigh and Jerry Kapoor and Kun JiaoFinite Difference (FD) modeling has been used to simulate seismic acquisition and to find out whether imaging artifacts are interfering with real images. 3D application of the modeling got accelerated when Reverse Time Migration (RTM) became an industry wide accepted migration algorithm. At the same time the Survey evaluation and design (SED) made use of FD modeling instead of ray-based methods to prove or reevaluate particular shooting geometries for illumination and imaging. RTM imaging evolved from the isotropic to anisotropic and very quickly reached the tilted anisotropic stage. RTM continues to moving forward to orthorhombic to add complexity to the migration to produce more accurate images. In the mean time Full waveform inversion (FWI) has been a major player to update and refine velocity fields which also uses FD modeling techniques to emulate the acquired data at each stage of the iterations. The FWI has to be inline with the latest RTM propagators if one wants to use them consistently in the depth imaging. The example show how the FD modeling is used to closely model data for FWI in the different acquisition environments.
-
-
-
Experiments on Elastic Wave Modelling in Isotropic and Anisotropic Media
Authors S. Jin, Yiqing Ren and and Shengwen JinRecent advances in seismic data processing with multi-component land and ocean-bottom-node (OBN) data have shown some contributions from share waves, converted waves as well as anisotropy. To get better understanding of the elastic wave propagation in isotropic and anisotropic media, we compared the wavefields in acoustic and elastic media with and without anisotropy. The preliminary experiments on three synthetic models, i.e., wave/solid interface, free surface with topography, and near-surface low velocity layer, demonstrate the complicated wave propagations in elastic media.
-
-
-
Iterative Krylov Solution Methods for Geophysical Electromagnetic Simulations on Throughput-oriented Graphical Processing Units
Authors Michael Commer, Gregory A. Newman and Filipe R. N. C. MaiaModern graphics processing units (GPUs) are designed for efficiently manipulating computer graphics. Their highly parallel architecture makes them also suitable for compute-intensive scientific applications. To provide access to the multithreaded computational resources and associated memory bandwidth of GPUs, graphics hardware manufacturers have introduced new application programming interfaces enabling numerical calculations in a fashion similar to parallel computing paradigms.
-
-
-
Implementation of the Full Waveform Inversion on GPU Cluster
Authors Henri Calandra, Bertrand Denel, Laurent Choy and Pierre Eric BernardFor more than 30 years, seismic reflection has been the main technology used in the Oil & Gas industry. Physics is well known and is based on solving approximations of the wave equation. In addition, geophysicists are now able to find numerical implementations which are the best suited to the computer hardware. Improving velocity models is still very challenging. It is the key to get an accurate seismic depth image. Mainly based on asymptotic approximations, seismic velocity analysis tools have some limitations when dealing with complex velocity models and complex wave propagation paths. Recent progress in high performance computing gives the ability to revive the Full Waveform Inversion formulation for refining velocity models. This technology, known since the mid 80's, is highly challenging due to the non linearity and the non uniqueness of solution. It is based on the minimization of an objective function measuring the difference between computed and observed data. The minimization process is very CPU intensive. Indeed, this is an iterative method which requires the computation of the derivative of the objective function with respect to the model parameters at each iteration by cross correlating the back propagated residual wavefield with the corresponding forward source propagated wavefield.
-
-
-
Shale Resource Development: From Exploration to Rejuvenation
By Usman AhmedNo two shale reservoirs are the same; however, deployment of a sound life cycle approach to shale gas development can lead to incremental return on investment. In the exploration phase some of the critical issues to address include the use of seismic attributes to identify potential high TOC (total organic content) and levels of Vitrinite reflectance. Acoustic impedance has been correlated to levels of TOC. During the appraisal phase, typically the operator focuses on reservoir assessment, trial investigation and sometimes consolidating land positions. Initial reservoir assessment and trial investigation is very important to define the resource base and exploitation strategy and as such logging, coring and testing becomes important. Two items that define shale gas development phase are horizontal wells and hydraulic fracturing to maximize reservoir contact. Geo-mechanical modeling plays a key role in defining the azimuth, orientation and length of laterals plus the size and stages of hydraulic fracture treatment. Optimization of the fracture treatments are then aided by real time micro-seismic monitoring. Addressing all these technical issues will ultimately have to make economic sense. During the production phase where full scale production is ongoing, some of the key challenges include corrosion / scaling wellbore and equipment, lift optimization (in case there is some fluid production), perforation plugging to name a few. Hence, some of the key solutions to address here include microbial control, scale and corrosion inhibition, friction reduction, H2S scavengers, lift system monitoring and optimization. The production phase comes to a point where the field needs to be rejuvenated (the rejuvenation phase) to enhance the recovery. Some of the key challenges include refinement of reservoir assessment and deployment of techniques to exploit additional reserves. Potential deployment solutions include intervention, remediation, re-fracturing and refine well placement (this may very well include reducing well spacing). Specific challenges and associated solutions will be different for different shale plays at all of the five life cycle phases. Hence, clearly understanding and recognizing these specific challenges and associated solutions are keys to incremental return on investment. Usman Ahmed is Vice President, Reservoir Themes & Solutions and Chief Reservoir Engineer. In this capacity Usman leads BHI’s reservoir-driven, cross-BHI solutions in unconventional reservoirs, mature oil and gas fields, performance completions, intelligent reservoirs and carbonates. Usman has more than 30 years of practical petroleum engineering experience from prior roles with Schlumberger, TerraTek and his own reservoir and production engineering consulting firm. He holds a B. Sc. and M. Sc. (both in Petroleum Engineering) from Texas A&M University. Usman has contributed to the industry through more than fifty publications, textbooks and patents and has held many technical and professional leadership roles.
-
-
-
Geology and Shale Gas Potential of European Sedimentary Basinsan Overview
Authors Hans-Martin Schulz and Brian HorsfieldMany parts of Europe contain prime targets for shale gas exploration. But compared to North America, Europe has a much more complex and compartimentalized setting of geological units. 510 Tcf shale-gas resources were estimated for Western Europe in 1997 by H.-H. Rogner, but the latest estimation by IHS CERA offers a much more optimistic number of 1000 Tcf (”Europe May Match North America in Unconventional Gas Promise”). Palaeozoic and Mesozoic black shales that are attractive for shale gas exploration occur in many European basins where conventional production is declining, an underutilized gathering infrastructure exists and markets are accessible.
-
-
-
Night-time Hunting for Furtive Animals: Data Availability Challenges in International Exploration for Partially Understood Shale Resource Plays
More LessThe recent success of shale resource plays in North America has given rise to an interest in these plays around the world. Europe has seen an almost feverish rush towards securing acreage and large tracts of land have been contracted out for shale exploration in Poland, France, Germany and elsewhere. Unfgfortunately, the exploration results from these licenses are not equally fast forthcoming and, thus far, have not always been very positive. Recently, critical voices have suggested that development of shale resources in Europe may happen at a significantly lower pace than has been the case in the US for a combination of geological, environmental, economical and regulatory reasons.
-
-
-
Production Analysis of Gas Shale Wells: Different Solutions for Different Data Scenarios
More LessProduction analysis is required to evaluate well productivity, forecast production ahead and calculate economics for budgetary reasons. Depending on the amount of data on hands, production analysis may be performed using classic statistical techniques, such as decline curve analysis where production trends are found from historical data and extrapolated over an extended period of time, typically with terminal declines. This form of forecasts is frequent when wells are not well characterized. Extensions of these forms of decline can be the linear flow forms, as it is well established in the literature that cumulative production is a linear function of square root of time. However this is only valid over a certain period of time, until boundary dominated flow prevails. Having more information of the actual reservoir allows to create numerical models which capture gas flow behaviors in nano Darcy rock. Time dependant desorption for instance, transient flow in the matrix cells, or permeability dependency to pressure may be modeled. It is then possible to history match clean up periods and initial gas flow, to better forecast long term performance. In cases where the hydraulic fracture network is characterized with microseismic, better solutions with numerical models are also pertinent. We will demonstrate in this discussion this various techniques through case examples.
-
-
-
New Method for an Easy Use of Stochastic Process-Based Models Such as Flumy to Reproduce a Fluvial Meandering Reservoir
Authors Isabelle Cojan, Jacques Rivoirard and Fabien Ors and Didier RenardFLUMY is a comprehensive modeling tool dedicated to meandering fluvial systems. The simulation is based on the migration of the centerline, based on hydraulic equations. The model reproduces the growth of meander loops, including chute and neck cut-off. Associated deposits correspond to point bars built up in relation to the migration, sand and mud plugs filling the abandoned channel in association to neck cut-off. Aggradation is controlled by the frequency and intensity of overbank floods, resulting in coarse sandy deposits at the bottom of the channel, deposition of silty levees in the vicinity of the channel and shale alluvium further away in the floodplain. Levee breaching results in crevasse splays and may generate avulsion, creating a new channel path. To be operational, process-based models such as FLUMY for fluvial meandering reservoirs make use of a limited number of key parameters, say a dozen. However choice of the parameters is most often a difficult issue as many interactions exist between these: sand body extension is largely related to migration rate and frequency of avulsion, the latter being influenced by the aggradation rate controlling the floodplain topography and giving specific sand ratio. In addition, most of the process parameters cannot be directly inferred from data or analogues.
-
-
-
Recent Advances in Geostatistical Tools in Order to Capture Realistic Geological Features
Authors Biver Pierre and Sergey NaumovIn practical operational studies, facies stochastic modelling is often performed with traditional geostatistical algorithms such as Sequential Indicator Simulation or Truncated Gaussian Simulation. In this presentation, we will focus on recent improvements of geostatistical techniques used to improve geological realism, and we will stress on the key points for their use in operational context. As a basis of discussion for the workshop, we will briefly address: the use and improvements of Multiple Point Statistics; the generalization of Truncated Gaussian Simulation with efficient Pluri- Gaussian Truncation; and the Boolean models with flexible objects.
-
-
-
The Importance of Re-Creating Realistic Reservoir Architecture in 3-D Sub-Surface Models: Lessons from Outcrop Studies
Authors Paul Davies, Huw Williams, Simon Pattison and Andrea MoscarielloThe objective of this study is to define the best approach to recreate realistic reservoir architecture in sub-surface models of wave-dominated shoreface/deltaic deposits. To achieve this goal, a fullydeterministic 3-D model based on 10 measured sections and fully continuous photo-panoramas was first created which exactly matches the sand and shale architecture in a well-exposed series of outcrops in Utah, USA. This model would be used as the benchmark against which all test model results could then be judged. The exceptionally well-exposed, near-horizontal Campanian strata of the Book Cliffs in eastern Utah were chosen as the basis for this study as they provide ideal outcrop analogues which have previously been used to develop, test and refine many sedimentary and stratigraphic models, including the principles and concepts of sequence stratigraphy. To approximate the sparsity of real sub-surface data, only three input well data points were used in all the models. Three different stochastic modelling techniques have been tested to try and capture the sediment body continuity and architecture observed in the outcrop. In addition, a deterministic model was built using only the same three log sections, supplemented only by geological knowledge about the palaeogeography, but strictly following a set of correlation rules and guidelines derived from the outcrop.
-
-
-
Three-Dimensional Numerical Modelling of Clinoforms within Deltaic and Shoreface Reservoirs
Key factors influencing fluid flow and reservoir behaviour include facies architecture and heterogeneity distribution conditioned to stratal surfaces. Within shallow-marine reservoirs, clinoforms are one common type of stratal surface. Clinoforms are palaeoseaward-dipping surfaces whose geometry preserves the depositional morphology of the delta-front or shoreface slope, and whose position reflects shoreline progradation history. Clinoform surfaces control aspects of facies architecture within parasequences, such as facies interfingering, which strongly affects the permeability architecture, due to the association of facies types with major permeability contrasts (Howell et al., 2008; Sech et al., 2009; Jackson et al., 2009). Clinoform surfaces can also act as barriers or baffles to flow (Figure 1), where there is calcite cementation, mica concentration, intense bioturbation, or mudstone and siltstone deposition along them, which further modify permeability architecture (Howell et al., 2008; Jackson et al., 2009). Under certain displacement conditions, it is therefore important to include clinoforms in reservoir models. However, standard reservoir modelling techniques are not well suited to capturing clinoform surfaces, particularly if they are numerous, below seismic resolution and/or difficult to correlate between wells. We present a new numerical algorithm that generates multiple clinoform surfaces within a volume defined by two bounding surfaces, for example a delta-lobe deposit or shoreface parasequence. The geometry and spacing of the clinoform surfaces is specified by the user, and the surfaces can be conditioned to well data where this is available.
-
-
-
Rule-Based Modelling of Modern-Day and Past Carbonate Shoals Environments
Authors Claude-Alain Hasler, Erwin W. Adams and Brigitte M. VlaswinkelCarbonate reservoirs tend to be particularly heterogeneous because, besides physical sedimentary processes, biological and chemical inputs play an important role in shaping the initial depositional architecture. In addition to these processes, diagenetic overprints, which can either follow or cross cut the primary depositional architecture, introduce another level of complexity by altering the primary depositional porosity and permeability. This combination of multiple processes interacting in both a temporal and spatial sense creates carbonate reservoir characteristics with complex architectures and heterogeneities on various scales, i.e. geobodies. As a result, pattern-based modelling is not always applicable, especially because of the difficulties arising in the identification and definition of discrete geobodies. In order to better predict and model geobodies, research developments are steered toward process-oriented methods for interwell modelling. This extended abstract presents a process-oriented approach using cellular automata applied to interwell-modelling.
-
-
-
Mps Facies Modelling – A Promising Technique Tested in Practice
Authors Christian Höcker and Adriaan JanszenSince the synthesis by Caers & Zhang (2002), facies modelling with Multiple-Point Statistics (MPS) has been recognised as a method that can substantially change subsurface modelling. Main advantages are that MPS can generate much more realistic and coherent facies patterns than 2-point statistic methods, that it can be driven with fewer and simpler parameters than genetic or object-based modelling – all this while being easily conditioned to external data. Despite these promising characteristics and implementation in standard platforms for subsurface modelling the technique has not yet been widely used in the oil and gas industry. With performance bottlenecks gradually being removed, the hurdles to successful MPS simulations may be the lack of suitable training images, in particular for modelling of 3D features, and training or guidance to address the somewhat different conceptualisation of MPS modelling runs. We have applied MPS modelling to very different depositional environments with demanding requirements, ranging from carbonate platforms through glacial settings and channelised deposits in meander belts and turbiditic settings. Below a number of topics are discussed which involve the handling of non-stationarity through the auxiliary or trend properties. The shown results were obtained using Impala MPS in JewelSuite.
-