Full Waveform Inversion: A Multiscale Approach to Tackle Gas Cloud Problem
Acoustic and elastic wave equations are computationally efficient and accurate simulation of complex wave propagation in heterogeneous environments. Both acoustic and elastic FWI with Q-absorption on synthetic gas cloud model have been used in this study.
- PDF / 607,086 Bytes
- 11 Pages / 439.37 x 666.142 pts Page_size
- 77 Downloads / 203 Views
Abstract Acoustic and elastic wave equations are computationally efficient and accurate simulation of complex wave propagation in heterogeneous environments. Both acoustic and elastic FWI with Q-absorption on synthetic gas cloud model have been used in this study. Inversion results and misfit analysis show the effectiveness of both equations. In the analysis, the elastic FWI results have not significantly changed, even with wrong assumption of Q-values within the gas cloud. A detailed Q model is not needed for a successful elastic FWI, as long as appropriates passive long wavelength background Q model is in cooperated. Keywords Inversion
Acoustic Elastic Q-attenuation
1 Introduction Getting subsurface properties is very critical and challenging especially in complex environment such as low-velocity overburden area. The presence of gas cloud has long been recognized as a significant problem in the seismic data around the world (North Sea, offshore Southeast Asia). In SE Asia, major hydrocarbon-bearing fields are affected by shallow gas clouds, and therefore, data quality often suffers from serious wipeouts due to shallow gas or gas leaking from a deep reservoir. Gas clouds are easily identified by the low P-wave velocities, whereas their signature is much weaker in the Vs model (Fig. 1). Joint inversion for velocity and Q with viscoacoustic FWI has attracted a lot of interests from the industry; however, it remains a very challenging topic as the joint inversion for both Vp and Q is an ill-posed problem, as they are coupled. Based on wave equation, the first imaging through FWI was proposed in the 1980s by [12] using least-squares approach in time domain and later by [10] in frequency domain. S. Prajapati (&) Centre for Seismic Imaging, Department of Petroleum and Geosciences, Universiti Teknologi PETRONAS, 32610 Seri Iskandar, Perak, Malaysia e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2017 M. Awang et al. (eds.), ICIPEG 2016, DOI 10.1007/978-981-10-3650-7_40
469
470
S. Prajapati
Fig. 1 Gas masking effect: a dense shallow gas cloud masking with poor imaging and b poor imaging with gas leakage due to poor sealing of hydrocarbon reservoir
The basic idea of FWI is to minimize the misfit between observed and calculated dataset after each optimization during inversion. FWI technique is very sensitive to the initial velocity model especially when real data are considered. In the absence or poor starting velocity model, FWI will trend to converse to one of the many local minima due to the wrapping around the nature of phase in the frequency domain or cyclic-skipping problem in the time domain [3, 11]. Earlier, some studies have been done on low velocity or gas cloud using first arrival tomography [13] and using l2 norm [2]. Kohn et al. [6] discuss the elastic isotropic FWI and conclude that P- and S-velocities and densities are better than using impedances. In the present study, I approached this problem in 2D format using acoustic and elastic wave equations with and without considering
Data Loading...
