scholarly journals Seismic velocity modeling in the digital transformation era: a review of the role of machine learning

2021 ◽  
Author(s):  
Abdullah AlAli ◽  
Fatai Anifowose
Keyword(s):  
Machine Learning ◽  
Seismic Velocity ◽  
Waveform Inversion ◽  
Petroleum Industry ◽  
Digital Transformation ◽  
Velocity Modeling ◽  
Full Waveform ◽  
Subsurface Layers ◽  
Push Forward ◽  
Complete Automation

AbstractSeismic velocity modeling is a crucial step in seismic processing that enables the use of velocity information from both seismic and wells to map the depth and thickness of subsurface layers interpreted from seismic images. The velocity can be obtained in the form of normal moveout (NMO) velocity or by an inversion (optimization) process such as in full-waveform inversion (FWI). These methods have several limitations. These limitations include enormous time consumption in the case of NMO due to manual and heavy human involvement in the picking. As an optimization problem, it incurs high cost and suffers from nonlinearity issues. Researchers have proposed various machine learning (ML) techniques including unsupervised, supervised, and semi-supervised learning methods to model the velocity more efficiently. The focus of the studies is mostly to automate the NMO velocity picking, improve the convergence in FWI, and apply FWI using ML directly from the data. In the purview of the digital transformation roadmap of the petroleum industry, this paper presents a chronologic review of these studies, appraises the progress made so far, and concludes with a set of recommendations to overcome the prevailing challenges through the implementation of more advanced ML methodologies. We hope that this work will benefit experts, young professionals, and ML enthusiasts to help push forward their research efforts to achieving complete automation of the NMO velocity and further enhancing the performance of ML applications used in the FWI framework.

Near-surface characterization using traveltime and full-waveform inversion with vertical and horizontal component seismic data

2019 ◽  
Vol 7 (1) ◽  
pp. T141-T154 ◽  
Author(s):  
Md. Iftekhar Alam
Keyword(s):  
Surface Characterization ◽  
Seismic Waves ◽  
Seismic Velocity ◽  
Waveform Inversion ◽  
Horizontal Component ◽  
Full Waveform Inversion ◽  
Near Surface ◽  
Velocity Modeling ◽  
Full Waveform ◽  
Attenuation Models

Seismic imaging of the shallow subsurface (approximately 5 m) can be very challenging when reflections are absent and the data are dominated by ground roll. I analyzed the transmission coda to produce fine-scale, interpretable vertical and horizontal component seismic velocity ([Formula: see text] and [Formula: see text]) models using full-waveform inversion (FWI). Application of FWI is tested through imaging two buried targets. The first target is a pair of well-documented utility pipes with known diameters (0.8 m) and burial depths (approximately 1.5 m). The second target is a poorly documented former location of the pipe(s), which is now a backfilled void. Data are acquired along a 23 m 2D profile using a static array with single-component vertical and horizontal geophones. Our results indicate considerable velocity updates in the [Formula: see text] and [Formula: see text] models across the pipes and backfill. The pipes appear as negative velocity updates in the final inverted [Formula: see text] and [Formula: see text] models, whereas the backfilled area represents negative and positive velocity updates in the [Formula: see text] and [Formula: see text] models, respectively. Variations of the polarities in the inverted models ([Formula: see text] and [Formula: see text]) across the backfill can be indicative of the medium, which respond differently to the vertical and horizontal component seismic waves. The attenuation models show a general decreasing trend with increasing depth. Therefore, simultaneous applications of vertical ([Formula: see text]) and horizontal ([Formula: see text]) component seismic velocity modeling can be an effective tool to understand the subsurface medium in near-surface characterization.

scholarly journals Adding Prior Information in FWI through Relative Entropy

Entropy ◽  
2021 ◽  
Vol 23 (5) ◽  
pp. 599
Author(s):  
Danilo Cruz ◽  
João de Araújo ◽  
Carlos da Costa ◽  
Carlos da Silva
Keyword(s):  
Inverse Problem ◽  
Relative Entropy ◽  
Global Minimum ◽  
Prior Information ◽  
Waveform Inversion ◽  
Petroleum Industry ◽  
Synthetic Data ◽  
Full Waveform Inversion ◽  
Full Waveform

Full waveform inversion is an advantageous technique for obtaining high-resolution subsurface information. In the petroleum industry, mainly in reservoir characterisation, it is common to use information from wells as previous information to decrease the ambiguity of the obtained results. For this, we propose adding a relative entropy term to the formalism of the full waveform inversion. In this context, entropy will be just a nomenclature for regularisation and will have the role of helping the converge to the global minimum. The application of entropy in inverse problems usually involves formulating the problem, so that it is possible to use statistical concepts. To avoid this step, we propose a deterministic application to the full waveform inversion. We will discuss some aspects of relative entropy and show three different ways of using them to add prior information through entropy in the inverse problem. We use a dynamic weighting scheme to add prior information through entropy. The idea is that the prior information can help to find the path of the global minimum at the beginning of the inversion process. In all cases, the prior information can be incorporated very quickly into the full waveform inversion and lead the inversion to the desired solution. When we include the logarithmic weighting that constitutes entropy to the inverse problem, we will suppress the low-intensity ripples and sharpen the point events. Thus, the addition of entropy relative to full waveform inversion can provide a result with better resolution. In regions where salt is present in the BP 2004 model, we obtained a significant improvement by adding prior information through the relative entropy for synthetic data. We will show that the prior information added through entropy in full-waveform inversion formalism will prove to be a way to avoid local minimums.

scholarly journals ML-descent: An optimization algorithm for full-waveform inversion using machine learning

Geophysics ◽  
2020 ◽  
Vol 85 (6) ◽  
pp. R477-R492 ◽  
Author(s):  
Bingbing Sun ◽  
Tariq Alkhalifah
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Inverse Problem ◽  
Optimization Algorithm ◽  
Waveform Inversion ◽  
Slight Modification ◽  
Descent Method ◽  
Quadratic Functions ◽  
Full Waveform Inversion ◽  
Full Waveform

Full-waveform inversion (FWI) is a nonlinear optimization problem, and a typical optimization algorithm such as the nonlinear conjugate gradient or limited-memory Broyden-Fletcher-Goldfarb-Shanno (LBFGS) would iteratively update the model mainly along the gradient-descent direction of the misfit function or a slight modification of it. Based on the concept of meta-learning, rather than using a hand-designed optimization algorithm, we have trained the machine (represented by a neural network) to learn an optimization algorithm, entitled the “ML-descent,” and apply it in FWI. Using a recurrent neural network (RNN), we use the gradient of the misfit function as the input, and the hidden states in the RNN incorporate the history information of the gradient similar to an LBFGS algorithm. However, unlike the fixed form of the LBFGS algorithm, the machine-learning (ML) version evolves in response to the gradient. The loss function for training is formulated as a weighted summation of the L2 norm of the data residuals in the original inverse problem. As with any well-defined nonlinear inverse problem, the optimization can be locally approximated by a linear convex problem; thus, to accelerate the training, we train the neural network by minimizing randomly generated quadratic functions instead of performing time-consuming FWIs. To further improve the accuracy and robustness, we use a variational autoencoder that projects and represents the model in latent space. We use the Marmousi and the overthrust examples to demonstrate that the ML-descent method shows faster convergence and outperforms conventional optimization algorithms. The energy in the deeper part of the models can be recovered by the ML-descent even when the pseudoinverse of the Hessian is not incorporated in the FWI update.

Near-surface full-waveform inversion in a transmission surface-consistent scheme

Geophysics ◽  
2020 ◽  
pp. 1-57
Author(s):  
Daniele Colombo ◽  
Ernesto Sandoval ◽  
Diego Rovetta ◽  
Apostolos Kontakis
Keyword(s):  
High Resolution ◽  
Seismic Data ◽  
Waveform Inversion ◽  
Surface Velocity ◽  
Full Waveform Inversion ◽  
Velocity Mapping ◽  
Signal To Noise ◽  
Near Surface ◽  
Velocity Modeling ◽  
Full Waveform

Land seismic velocity modeling is a difficult task largely related to the description of the near surface complexities. Full waveform inversion is the method of choice for achieving high-resolution velocity mapping but its application to land seismic data faces difficulties related to complex physics, unknown and spatially varying source signatures, and low signal-to-noise ratio in the data. Large parameter variations occur in the near surface at various scales causing severe kinematic and dynamic distortions of the recorded wavefield. Some of the parameters can be incorporated in the inversion model while others, due to sub-resolution dimensions or unmodeled physics, need to be corrected through data preconditioning; a topic not well described for land data full waveform inversion applications. We have developed novel algorithms and workflows for surface-consistent data preconditioning utilizing the transmitted portion of the wavefield, signal-to-noise enhancement by generation of CMP-based virtual super shot gathers, and robust 1.5D Laplace-Fourier full waveform inversion. Our surface-consistent scheme solves residual kinematic corrections and amplitude anomalies via scalar compensation or deconvolution of the near surface response. Signal-to-noise enhancement is obtained through the statistical evaluation of volumetric prestack responses at the CMP position, or virtual super (shot) gathers. These are inverted via a novel 1.5D acoustic Laplace-Fourier full waveform inversion scheme using the Helmholtz wave equation and Hankel domain forward modeling. Inversion is performed with nonlinear conjugate gradients. The method is applied to a complex structure-controlled wadi area exhibiting faults, dissolution, collapse, and subsidence where the high resolution FWI velocity modeling helps clarifying the geological interpretation. The developed algorithms and automated workflows provide an effective solution for massive full waveform inversion of land seismic data that can be embedded in typical near surface velocity analysis procedures.

Direct updating of rock physics properties using elastic full-waveform inversion

Geophysics ◽  
2021 ◽  
pp. 1-64
Author(s):  
Qi Hu ◽  
Scott Keating ◽  
Kristopher A. Innanen ◽  
Huaizhen Chen
Keyword(s):  
Reservoir Characterization ◽  
Seismic Velocity ◽  
Water Saturation ◽  
Quantitative Estimation ◽  
Waveform Inversion ◽  
Rock Physics ◽  
Full Waveform Inversion ◽  
Full Waveform ◽  
Physics Model ◽  
Rock Physics Model

Quantitative estimation of rock physics properties is an important part of reservoir characterization. Most current seismic workflows in this field are based on amplitude variation with offset. Building on recent work on high resolution multi-parameter inversion for reservoir characterization, we construct a rock-physics parameterized elastic full-waveform inversion (EFWI) scheme. Within a suitably-formed multi-parameter EFWI, in this case a 2D frequency-domain isotropic-elastic FWI with a truncated Gauss-Newton optimization, any rock physics model with a well-defined mapping between its parameters and seismic velocity/density can be examined. We select a three-parameter porosity, clay content, and water saturation (PCS) parameterization, and link them to elastic properties using three representative rock physics models: the Han empirical model, the Voigt-Reuss-Hill boundary model, and the Kuster and Toksöz inclusion model. Numerical examples suggest that conditioning issues, which make a sequential inversion (in which velocities and density are first determined through EFWI, followed by PCS parameters) unstable, are avoided in this direct approach. Significant variability in inversion fidelity is visible from one rock physics model to another. However, the response of the inversion to the range of possible numerical optimization and frequency selections, as well as acquisition geometries, varies widely. Water saturation tends to be the most difficult property to recover in all situations examined. This can be explained with radiation pattern analysis, where very low relative scattering amplitudes from saturation perturbations are observed. An investigation performed with a Bayesian approach illustrates that the introduction of prior information may increase the inversion sensitivity to water saturation

Central-difference time-lapse 4D seismic full waveform inversion

Geophysics ◽  
2020 ◽  
pp. 1-42
Author(s):  
Wei Zhou ◽  
David Lumley
Keyword(s):  
High Resolution ◽  
Seismic Velocity ◽  
Waveform Inversion ◽  
Computational Cost ◽  
Time Lapse ◽  
Full Waveform Inversion ◽  
Central Difference ◽  
Data Set ◽  
Velocity Inversion ◽  
Full Waveform

Repeated seismic surveys contain valuable information regarding time-lapse (4D) changes in the subsurface. Full waveform inversion (FWI) of seismic data can provide high-resolution estimates of 4D change. We propose a new time-domain 2D acoustic time-lapse FWI method based on the central-difference scheme with higher-order mathematical accuracy and reasonable computational cost. The method is rigorously tested on the SEAM 4D time-lapse model and OBN data set. High-resolution 4D velocity estimates are obtained, which show strong ~25% velocity increases in a 75 m-thick gas layer, as well as weaker (5%) changes due to geomechanical effects, the latter of which are poorly recovered by the conventional parallel 4D FWI method. We also perform the bootstrap 4D FWI method and the result is contaminated by strong artifacts in the underburden, whereas the proposed central-difference method has fewer underburden artifacts allowing more reliable interpretations. In this realistic case study, acoustic FWI erroneously overfits the elastic scattered waves, and cannot fit the strong elastic 4D coda waves at all. The results show that the proposed central-difference 4D FWI method within the acoustic approximation may be a practical solution for time-lapse seismic velocity inversion.

Sign in / Sign up

Export Citation Format

Share Document