Andrey Bakulin

Followers

Following

Public Views

UT Austin

Schlumberger-Doll Research

Felix Herrmann

University of British Columbia

Alok Soni

A. Berkhout

ADRIANO GOMES

Interests

Uploads

Papers by Andrey Bakulin

Improving Repeatability of Land Seismic Data Using Virtual Source Approach Based on Multidimensional Deconvolution

77th EAGE Conference and Exhibition 2015, Jun 1, 2015

Download

Effective models of stratified media containing porous biot layers

Journal of Mathematical Sciences, Sep 1, 1999

Periodic stratified media in which either two porous Blot layer, or an elastic and a porous layer... more Periodic stratified media in which either two porous Blot layer, or an elastic and a porous layers, or a fluid and a porous layer alternate are considered. The effective models of these media are constructed and investigated. In the case of alternating porous layers, the effective model is a generalized transversely isotropic Biot medium. In this medium, the density of the fluid phase and the mean density acquire tensor character. It is shown that the effective model of a porous-fluid medium is, on the one hand, a generalized transversely isotropic Blot medium of special type and, on the other hand, a generalization of the effective model of a stratified elastic-fluid medium. Bibliography: 14 titles.

Download

Accelerated deep learning-based estimation of wavefront dips and curvatures and their application to 3D prestack data enhancement

Second International Meeting for Applied Geoscience & Energy

We present a novel workflow for the accelerated signal enhancement of massive 3D prestack seismic... more We present a novel workflow for the accelerated signal enhancement of massive 3D prestack seismic data utilizing a Local Wavefront Attributes Deep Neural Network. It is based on automatic local wavefront attributes estimation using a specially trained convolutional deep neural network. The general workflow is adaptive to a particular 3D prestack seismic volume. It requires performing a conventional semblance-based estimation of wavefront dips and curvatures for only about 1% of the whole amount of data. The verification of the proposed approach is done on challenging real datasets, both marine and land. Deep learning allows achieving a significant speed-up compared to the conventional method while preserving an acceptable quality of the results.

Download

Drillbit vibrations enable sonic logs prediction in lateral boreholes using machine learning

First International Meeting for Applied Geoscience & Energy Expanded Abstracts, 2021

The development of unconventional resources requires accurate delineation of formation stiffness ... more The development of unconventional resources requires accurate delineation of formation stiffness along extended horizontal borehole segments. This paper studies the feasibility of prediction of rock stiffness using drilling parameters recorded at the rig, downhole measurements while drilling, and near-bit vibrations. We used a machine learning-assisted workflow to model the wireline logs and assess the importance of various input data. The use of nearbit vibrations reduces the prediction error (of sonic data?) from ~5.3% achieved relying on to 1.8%, which is sufficient for drilling/completion applications. The machine learning algorithm also quantifies the quality of the wireline logs used for training: shear and density logs have ~20% of noisy data while the compressional velocity ~35%. Prediction of wireline shear and compressional sonic data using downhole vibrations measured while drilling has vast potential to reduce logging cost for borehole completion industrywide.

Download

Time-lapse seismic cross-equalization using temporal convolutional networks

First International Meeting for Applied Geoscience & Energy Expanded Abstracts, 2021

Monitoring of geological reservoirs using 4D seismic faces many challenges. The repeatability bet... more Monitoring of geological reservoirs using 4D seismic faces many challenges. The repeatability between different surveys needs to be optimal in which changes are only present in the target zone. Ideal conditions require having the same acquisition parameters for each survey and no nearsurface variations, like those caused by seasonal changes. In practice, data processing and matching techniques are required to improve the repeatability of the data. This study proposes a deep learning approach for post-stack trace-bytrace matching to reduce the remaining 4D noise. We utilize the sequential nature of seismic data to train a temporal convolutional network (TCN), which learns to map the monitor traces to the base data in the overburden region. The goal is to suppress 4D noise while maintaining time-lapse signal caused by the reservoir changes we wish to monitor. We validate the method on synthetic time-lapse zero-offset data and show improvements in repeatability. We also perform an initial test on 4D land data to show the potential for application to real datasets.

Download

Direct estimation of local wavefront attributes using deep learning

First International Meeting for Applied Geoscience & Energy Expanded Abstracts, 2021

Wavefront attributes, such as local dips and curvatures of seismic events, are used in different ... more Wavefront attributes, such as local dips and curvatures of seismic events, are used in different seismic data processing methods, from prestack data enhancement to migration to tomography. The attributes' estimation for prestack data is a time-consuming and computationally expensive process. We propose a new approach based on U-Net convolutional neural network that directly map prestack seismic data to the local wavefront attributes. Using a 3D real data example, we demonstrate that this deep-learning-based approach can reduce the computational time by two orders of magnitude compared to a classical coherency-based optimization technique while preserving a reasonable quality of results.

Download

Prestack data enhancement with phase corrections in time‐frequency domain guided by local multidimensional stacking

Geophysical Prospecting, 2020

ABSTRACTWe present a new approach to enhancing weak prestack reflection signals without sacrifici... more ABSTRACTWe present a new approach to enhancing weak prestack reflection signals without sacrificing higher frequencies. As a first step, we employ known multidimensional local stacking to obtain an approximate ‘model of the signal’. Guided by phase spectra from this model, we can detect very weak signals and make them visible and coherent by ‘repairing’ corrupted phase of original data. Both presented approaches – phase substitution and phase sign corrections – show good performance on complex synthetic and field data suffering from severe near‐surface scattering where conventional processing methods are rendered ineffective. The methods are mathematically formulated as a special case of time‐frequency masking (common in speech processing) combined with the signal model from local stacking. This powerful combination opens the avenue for a completely new family of approaches for multi‐channel seismic processing that can address seismic processing of land data with nodes and single se...

Download

Gulf of Mexico Case Study of Localized Anisotropic Tomography with Checkshot

72nd EAGE Conference and Exhibition incorporating SPE EUROPEC 2010, 2010

Reliably deriving parameters for anisotropic depth models requires use of borehole information. L... more Reliably deriving parameters for anisotropic depth models requires use of borehole information. Localized tomographic inversion attempts to streamline and automate this process by directly incorporating the available well data into conventional reflection tomography. We present a case study from Gulf of Mexico where we conduct local VTI anisotropic tomography using a joint dataset consisting of seismic and checkshot data. Because this area has low structural dip, the results can be compared with more traditional manual 1D layer-stripping inversion. Tomographic inversion for three VTI parameters produces a smooth velocity model that both fits the checkshot traveltimes and flattens all seismic gathers. To regularize tomographic inversion, we apply smoothing operators that are oriented along predominant dips of seismic event and have large lateral extent. The anisotropic profiles derived by tomography and 1D inversion have similar low-frequency components, but differ in finer details. Borehole data require careful conditioning before joint inversion because of potential difference in water velocity between seismic and well surveys. The workflow we present can be applied to calibrating anisotropic parameters in the more general case of 3D models with structural dip and borehole data from deviated wells.

Interaction of Low-Frequency Tube Waves with Poroelastic Reservoirs Containing Perforations

68th EAGE Conference and Exhibition incorporating SPE EUROPEC 2006, 2006

Tube or Stoneley wave is known to strongly interact at low frequencies with poroelastic formation... more Tube or Stoneley wave is known to strongly interact at low frequencies with poroelastic formations provided that flow is not restricted at the borehole-formation interface. Increased permeability leads to increased attenuation and decreased velocity of the tube wave. In this study we focus on reflection of low-frequency tube waves from various finite-size poroelastic structures. First, we examine a model of a thin reservoir and demonstrate good applicability of the approximate 1D effective wavenumber approach to describe interaction of tube waves with porous formations. We confirm that higher permeability leads to higher reflection coefficient. Then we analyze model of an idealized (disk-shaped) perforation inside a poroelastic layer and show that it has higher reflectivity compared to washout zone of the same geometry but with no-flow conditions at the face.

Download

Improving pre-stack land data using smart supergrouping

SEG Technical Program Expanded Abstracts 2015, 2015

3D land seismic data acquired in arid environments is often challenging for data processing and i... more 3D land seismic data acquired in arid environments is often challenging for data processing and interpretation, due to low signal-to-noise ratio and the presence of various types of noise. Traditionally, large source and receiver arrays have been utilized for noise suppression and signal enhancement. A trend in modern seismic data acquisition is to reduce the size of the source and receiver arrays, aiming to record broadband signals for imaging and inversion purposes. For many processing steps and velocity model building, achieving good pre-stack signal-to-noise ratio may be more important. We propose a simple but effective supergrouping technique that significantly enhances prestack data quality. We demonstrate our approach on two 3D onshore datasets from Saudi Arabia.

Download

Time‐lapse traveltime shifts above compacting reservoirs: 3D solutions for prestack data

SEG Technical Program Expanded Abstracts 2007, 2007

Time-lapse (4D) traveltime shifts of reflection events recorded above hydrocarbon reservoirs can ... more Time-lapse (4D) traveltime shifts of reflection events recorded above hydrocarbon reservoirs can be used to monitor production-related compaction and porepressure changes. Existing methodology, however, is largely limited to zero-offset rays and cannot be applied to time shifts measured on prestack seismic data. Here, we obtain traveltime shifts by employing first-order perturbation theory that accounts for the stress-induced anisotropic velocity field, as well as for the deformation of reflectors. The resulting closed-form expression can be efficiently used for 3D numerical modeling of traveltime shifts and, ultimately, for reconstructing the heterogeneous stress distribution around compacting reservoirs. The analytic results are applied to a 2D model that includes a rectangular resevoir embedded in an initially homogeneous and isotropic medium. The computed velocity changes around the reservoir are caused primarily by the deviatoric stresses and produce an anisotropic medium with substantial values of the Thomsen parameters ǫ and δ and variable orientation of the symmetry axis. The offset dependence of traveltime shifts should play a crucial role in estimating the anisotropy parameters and the compaction-related deviatoric stress components.

Download

Reflection and transmission of tube waves in cased boreholes with layers and perforations

SEG Technical Program Expanded Abstracts 2007, 2007

At low frequencies tube or Stoneley waves represent a dominant arrival propagating along borehole... more At low frequencies tube or Stoneley waves represent a dominant arrival propagating along boreholes. They can be excited by the source in a well or by external source due to conversion from other wave types. Tube wave experiences reflection at the bed boundaries, borehole diameter changes and fractures or permeable zones. It was proven in previous studies that 1D effective wavenumber approach provides simple and accurate low-frequency description of tubewave propagation in open boreholes surrounded by radially homogeneous formation. Tube waves become even more dominant in cased boreholes, but casing further modifies wave propagation and reflection/transmission phenomena. In this study we apply 1D effective wavenumber approach to radially inhomogeneous media and demonstrate that it still provides excellent description of low-frequency tubewave propagation. In particular, we focus on three models representative of cased boreholes: reflection from geological interfaces behind casing, reflection from corroded casing section and reflection from idealized diskshaped perforation in cased hole. In all three cases frequency-dependent reflection coefficient obtained by 1D effective method and by finite-difference computations show excellent agreement.

Download

Tube-wave seismic imaging

ABSTRACT The detailed analysis of cross well seismic data for a gas reservoir in Texas revealed t... more ABSTRACT The detailed analysis of cross well seismic data for a gas reservoir in Texas revealed two newly detected seismic wave effects, recorded approximately 2000 feet above the reservoir. A tube-wave (150) is initiated in a source well (110) by a source (111), travels in the source well (110), is coupled to a geological feature (140), propagates (151) through the geological feature (140), is coupled back to a tube-wave (152) at a receiver well (120), and is and received by receiver(s) (121) in either the same (110) or a different receiving well (120). The tube-wave has been shown to be extremely sensitive to changes in reservoir characteristics. Tube-waves appear to couple most effectively to reservoirs where the well casing is perforated, allowing direct fluid contact from the interior of a well case to the reservoir.

Modeling of Borehole Modes Using the Spectral Method

69th EAGE Conference and Exhibition incorporating SPE EUROPEC 2007, 2007

We present algorithm and code that solves the dispersion equation for cylindrically layered media... more We present algorithm and code that solves the dispersion equation for cylindrically layered media consisting of arbitrary number of solid elastic and fluid layers. The algorithm is based on the spectral method which discretises the underlying wave equations with the help of spectral differentiation matrices and solves the corresponding equations as an generalized eigenvalue problem. For a given frequency the eigenvalues correspond to the wavenumbers of different modes. The advantage of this technique is that it is easy to implement especially for cases where traditional root-finding methods are strongly limited or hard to realize, i.e. for attenuative, anisotropic and poroelastic media. We illustrate the application of the new approach using models of a free solid bar and a fluid-filled cylinder. These dispersion curves are in good agreement with analytical results, which confirms the accuracy of the new method.

Download

About the relation between the effective model of layered fluid-solid medium and transversely isotropic biot model

Proceedings, 1996

Download

Computing borehole modes with spectral method

SEG Technical Program Expanded Abstracts 2007, 2007

We present algorithm and code that solves dispersion equation for cylindrically layered media con... more We present algorithm and code that solves dispersion equation for cylindrically layered media consisting of arbitrary number of solid elastic and fluid layers. The algorithm is based on the spectral method which discretises the underlying wave equations with the help of spectral differentiation matrices and solves the corresponding equations as an generalized eigenvalue problem. For a given frequency the eigenvalues correspond to the wavenumbers of different modes. The advantage of this technique is that it is easy to implement especially for cases where traditional root-finding methods are strongly limited or hard to realize, i.e. for attenuative, anisotropic and poroelastic media. We illustrate the application of the new approach using models of a free solid bar and a fluid-filled cylinder. The computed dispersion curves are in good agreement with analytical results, which confirms the accuracy of the new method.

Download

$Research paper thumbnail of Estimation of fracture parameters of monoclinic media from reflection seismic data$

Estimation of fracture parameters of monoclinic media from reflection seismic data

SEG Technical Program Expanded Abstracts 2000, 2000

Physical modeling and theoretical studies of seismic refections from a fault zone

SEG Technical Program Expanded Abstracts 2004, 2004

We simulate a fault/fracture zone in a physical modeling experiment as array of thin, air‐filled,... more

Buried Sources and Receivers in a Karsted Desert Environment

Proceedings, 2012

To evaluate repeatability of emerging seismic technologies for future 4D seismic reservoir monito... more To evaluate repeatability of emerging seismic technologies for future 4D seismic reservoir monitoring studies on Middle East carbonate reservoirs, a seismic field acquisition test was conducted over an onshore field in Saudi Arabia. The effects of near-surface complexity (in the form of sand and karsts) as well as large surface temperature variations are illustrated and quantified by 4D attribute analysis using permanent piezoelectric seismic sources. Even though measured repeatability does not reach values observed in nondesert environments, we show that burying receivers dramatically improves the wavelet amplitude stability. As the complex near-surface scattering layer appears to be thicker than initially expected with the presence of karsts down to a depth of 40 m, we conclude that deeper burial of sources and receivers below the most complex part of the near surface may potentially let us use lower-fold seismic data for reservoir monitoring in complex near-surface desert environments.

Download

Comparing virtual versus real crosswell surveys

SEG Technical Program Expanded Abstracts 2008, 2008

The virtual source method (VSM) is a useful tool for imaging below complex overburden and monitor... more The virtual source method (VSM) is a useful tool for imaging below complex overburden and monitoring in the presence of time-varying overburden. This concept, when extended to crosswell geometry produces data comparable to real crosswell data. Using a field data example we demonstrate that virtual crosswell data is kinematically comparable to real crosswell data, but the virtual crosswell method possesses flexibilities, which are difficult to achieve in a real crosswell survey. Some of these flexibilities include the ability of the virtual source to radiate either horizontally or vertically and the possibility for the virtual source to radiate only P-or only S-waves. It is also possible to create virtual crosswell data that contain only the direct arrivals or only the reflections. These features of the virtual crosswell method should make it useful for crosswell tomography, imaging and reservoir monitoring for moderate interwell distances.

Download

Improving Repeatability of Land Seismic Data Using Virtual Source Approach Based on Multidimensional Deconvolution

77th EAGE Conference and Exhibition 2015, Jun 1, 2015

Download

Effective models of stratified media containing porous biot layers

Journal of Mathematical Sciences, Sep 1, 1999

Download

Accelerated deep learning-based estimation of wavefront dips and curvatures and their application to 3D prestack data enhancement

Second International Meeting for Applied Geoscience & Energy

Download

Drillbit vibrations enable sonic logs prediction in lateral boreholes using machine learning

First International Meeting for Applied Geoscience & Energy Expanded Abstracts, 2021