Elastic full-waveform inversion and parametrization analysis applied to walk-away vertical seismic profile data for unconventional (heavy oil) reservoir characterization

Pan, Wugen; Innanen, Kristopher A.; Geng, Yushui

doi:10.1093/gji/ggy087

Cited by 66 publications

(8 citation statements)

References 69 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For the inversion problems of complicated structure reservoirs, the inversion methods based on numerical solutions of wave equation are being studied and applied in depth. For example, recently, Pan et al [35] successfully applied elastic full-waveform inversion to a field dataset for reservoir characterization. However, at present, the published papers on the thin-bed inversion based on numerical solutions of wave equation are very rare.…”

Section: Discussionmentioning

confidence: 99%

P-Wave Reflection Approximation of a Thin Bed and Its Application

et al. 2020

View full text Add to dashboard Cite

“Thin-bed” reservoirs have become important targets of seismic exploration and exploitation. However, traditional amplitude versus offset/amplitude versus angle (AVO/AVA) technologies, for example, those based on Zoeppritz equations and their approximations for a single interface, are not sufficiently accurate for thin-bed stratigraphy. Analytic solutions of thin-bed reflectivity may become practical for thin-bed AVO analysis and inversion. Therefore, a linear analytic approximation of thin-bed P-wave reflectivity is developed under small-incidence and thin-bed assumptions. Numerical simulations show that the amplitude approximation errors are usually smaller than 10% for incidence angles less than 20 degrees, and the thin-bed thicknesses are less than one-tenth of the P-wave wavelength. Based on the least-squares approach, the inversion strategy is proposed using the approximate formula. A synthetic data test shows that the proposed inversion method can produce more accurate thin-bed properties than that based on the Zoeppritz equations, which reveals the potential of the inversion method based on the linear analytic approximate formula in the fine characterization of thin reservoirs.

show abstract

Section: Discussionmentioning

confidence: 99%

P-Wave Reflection Approximation of a Thin Bed and Its Application

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Full waveform inversion (FWI) (Lailly, 1983;Tarantola, 1984) has been widely used in exploration and global seismology for high-resolution parameter estimation (Virieux and Operto, 2009;Tromp, 2020). With the success of the application of mono-parameter FWI to field data (Sirgue et al, 2010;Warner et al, 2013;Operto et al, 2015;Shen et al, 2018), it becomes more and more attractive to study multiple parameters (Operto et al, 2013), to account for the effect of attenuation (Kamei and Pratt, 2013;Fabien-Ouellet et al, 2017;da Silva et al, 2019;Kamath et al, 2021), density (Yang et al, 2016a;Operto and Miniussi, 2018), anisotropy (Prieux et al, 2011;Alkhalifah and Plessix, 2014), or elastic parameters (Brossier et al, 2009;Köhn et al, 2012;Vigh et al, 2014;Pan et al, 2018;Trinh et al, 2019;Wang et al, 2021). FWI formulates seismic inversion into a PDE-constrained optimization problem.…”

Section: Introductionmentioning

confidence: 99%

Parsimonious truncated Newton method for time-domain full-waveform inversion based on the Fourier-domain full-scattered-field approximation

2022

View full text Add to dashboard Cite

In order to exploit Hessian information in Full Waveform Inversion (FWI), the matrix-free truncated Newton method can be used. In such a method, Hessian-vector product computation is one of the major concerns due to the huge memory requirements and demanding computational cost. Using the adjoint-state method, the Hessian-vector product can be estimated by zero-lag cross-correlation of the first-order/second-order incident wavefields and the second-order/first-order adjoint wavefields. Different from the implementation in frequency-domain FWI, Hessian-vector product construction in the time domain becomes much more challenging as it is not affordable to store the entire time-dependent wavefields. The widely used wavefield recomputation strategy leads to computationally intensive tasks. We present an efficient alternative approach to computing the Hessian-vector product for time-domain FWI. In our method, discrete Fourier transform is applied to extract frequency-domain components of involved wavefields, which are used to compute wavefield cross-correlation in the frequency domain. This makes it possible to avoid reconstructing the first-order and second-order incident wavefields. In addition, a full-scattered-field approximation is proposed to efficiently simplify the second-order incident and adjoint wavefields computation, which enables us to refrain from repeatedly solving the first-order incident and adjoint equations for the second-order incident and adjoint wavefields (re)computation. With the proposed method, the computational time can be reduced by 70% and 80% in viscous media for Gauss-Newton and full-Newton Hessian-vector product construction, respectively. The effectiveness of our method is also verified in the frame of a 2D multi-parameter inversion, in which the proposed method almost reaches the same iterative convergence of the conventional time-domain implementation.

show abstract

“…Key ingredients are an efficient forward modeling method to simulate synthetic data, and a local differential approach, in which the gradient and direction are calculated to update the model. When updating several parameters simultaneously, which is a multiparameter full waveform inversion, error in one parameter tends to produce updates in others, a phenomenon referred to as inter-parameter trade-off or cross-talk (e.g., Kamei and Pratt, 2013;Alkhalifah and Plessix, 2014;Innanen, 2014;Pan et al, 2018;Keating and Innanen, 2019). The degree of trade-off or cross-talk between parameters can depend sensitively on the specific parameterization; even though, for instance, λ, µ and ρ have the same information content as V P , V S , and ρ, updating in one can produce markedly different convergence properties and final models that doing so in the other.…”

Section: Introductionmentioning

confidence: 99%

Numerical analysis of a deep learning formulation of elastic full waveform inversion with high order total variation regularization in different parameterization

Zhang,

Sun,

Innanen

et al. 2021

Preprint

View full text Add to dashboard Cite

We have formulated elastic seismic full waveform inversion (FWI) within a deep learning environment. Training this network with a single seismic data set is equivalent to carrying out elastic FWI. There are three main motivations for this effort. The first is an interest in developing an inversion algorithm which is more or less equivalent to standard elastic FWI but which is ready-made for a range of cloud computational architectures. The second is an interest in algorithms which can later (i.e., not in this paper) be coupled with a more involved training component, involving multiple datasets. The third is a general interest in developing the idea of theory-guiding within machine learning solutions for large geophysical problems, wherein the number of degrees of freedom within a network, and the 1

show abstract

Elastic full-waveform inversion and parametrization analysis applied to walk-away vertical seismic profile data for unconventional (heavy oil) reservoir characterization

Cited by 66 publications

References 69 publications

P-Wave Reflection Approximation of a Thin Bed and Its Application

P-Wave Reflection Approximation of a Thin Bed and Its Application

Parsimonious truncated Newton method for time-domain full-waveform inversion based on the Fourier-domain full-scattered-field approximation

Numerical analysis of a deep learning formulation of elastic full waveform inversion with high order total variation regularization in different parameterization

Contact Info

Product

Resources

About