Source-Independent Acoustic Logging Stratigraphic Imaging: Full Waveform Inversion Based on Time-Domain Waveform Convolution

Application of 2D full-waveform inversion for characterizing potential gas reservoirs

Summary Full Waveform Inversion (FWI) is a mathematically and computationally challenging inverse problem of finding a quantitative rock-property description of the subsurface to match observed seismogram data. The inversion employs a forward model, which relates the subsurface to the observed seismograms. FWI and other seismic applications require High Performance Computing (HPC) to simulate the dynamics of such complex models. Not a long ago, companies, research institutes, and universities used to acquire clusters of computers to maintain on-premise. Recently, cloud computing has become an alternative posing a challenge to the end-users, who have to decide whether they should execute their applications: on their local clusters or burst them to a remote cloud provider. In this paper, we present a decision support method to choose the correct environment considering trade-offs, such as resource costs, performance, and availability on such heterogeneous execution platforms. We evaluated the system using our FWI application and preliminary results indicate that users of HPC applications can benefit from such a cloud advisory system to reduce costs, turnaround times, and even boost local platforms.

This study aims to investigate the feasibility of CO2-EOR monitoring by full waveform inversion (FWI) of time-lapse VSP data in an onshore CO2-EOR site in Abu Dhabi. CO2-EOR monitoring using conventional time-lapse surface seismic in onshore oil fields in Abu Dhabi is often technically challenging for two main reasons. The first is that elastic property change in response to pore fluid substitution is relatively small because the elastic modulus of the reservoir rock frame is far larger than that of the pore fluids. The second is the low repeatability of time-lapse survey data due to high amplitude surface-related noise which varies temporally. However, seismic monitoring with FWI of time-lapse borehole seismic data may offer a solution for these issues. FWI is capable of detecting small velocity changes such as those associated with pore fluid substitution. Furthermore, borehole seismic surveys may provide more highly repeatable, higher quality data compared to surface seismic surveys because borehole seismic data is less affected by surface-related noise. This study consists of two parts, a field data analysis and a synthetic study. In the field data analysis, we studied the resolution and repeatability of FWI results at field-data quality, including the presence of actual noise using time-lapse VSP data. VSP data was acquired at the very early stage of EOR and there was no CO2 injection in the time between the two time-lapse VSP surveys. As a result, a high-resolution P-wave velocity model, consistent with a sonic log, was obtained. The P-wave velocity model also revealed excellent repeatability between the two survey data sets. In the synthetic study, time-lapse FWI was performed using synthetic VSP data representing pre- and post- CO2 injection periods. The results of the synthetic study showed that even in the presence of realistic 4D noise, which was estimated in the field data analysis, FWI successfully delineated the distribution of velocity changes caused by CO2 injection when the cross-sectional area of the injection-induced velocity changes were larger than the resolution of the FWI results. With these results, we demonstrated that FWI using time-lapse VSP data was applicable for CO2-EOR monitoring in the field as long as the criteria were met. This conclusion encourages the application of FWI using time-lapse VSP data for CO2-EOR monitoring in onshore Abu Dhabi.

PreviousNext No AccessSEG Technical Program Expanded Abstracts 2013Full waveform inversion – Dealing with limitations of 3D onshore seismic dataAuthors: Ahmed Al-YaqoobiMike WarnerAhmed Al-YaqoobiImperial College LondonSearch for more papers by this author and Mike WarnerImperial College LondonSearch for more papers by this authorhttps://doi.org/10.1190/segam2013-0799.1 SectionsSupplemental MaterialAboutPDF/ePub ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InRedditEmail Abstract Full-waveform inversion is a promising technique to produce high-resolution, interpretable velocity images of the subsurface. We present one of the first results from application of full waveform inversion to a vibrator seismic land data. The results verify that full waveform inversion can be successfully applied to land seismic data after certain preconditioning procedures and with a good a priori velocity model. Updates of the shallow part of the velocity model will have an impact on better recovering of the deeper part of the migration image. Permalink: https://doi.org/10.1190/segam2013-0799.1FiguresReferencesRelatedDetailsCited ByReferences16 June 2017 SEG Technical Program Expanded Abstracts 2013ISSN (print):1052-3812 ISSN (online):1949-4645Copyright: 2013 Pages: 5258 Publisher:Society of Exploration Geophysicists HistoryPublished: 19 Aug 2013 CITATION INFORMATION Ahmed Al-Yaqoobi and Mike Warner, (2013), "Full waveform inversion – Dealing with limitations of 3D onshore seismic data," SEG Technical Program Expanded Abstracts : 934-938. https://doi.org/10.1190/segam2013-0799.1 Plain-Language Summary PDF DownloadLoading ...

Seismic surveys are routinely used for building precise structural images of the coal bearing formations in Australia but coal production related hazards such as weak strata and zones with an increased gas content remain to be fully resolved by seismic measurements. One way of investigating these issues is through the application of Full Waveform Inversion (FWI) methods. To utilise the full power of these methods a high quality seismic dataset is needed. Such conditions are often met by 2D and 3D reflection seismic data acquired over coal seams in Bowen and Sydney basins. FWI can be used for a high resolution estimate of P- and S-wave velocities and the density that can also be translated into geotechnical parameters of interest to mining operations. In this study, we evaluate the applicability of FWI methods for estimating elastic parameters (P and S wave velocities and density) from the inversion of a synthetic seismic dataset that was recorded over the surface of a 2D earth model that represents subsurface geology in Goonyella coal mine in Queensland, Australia. We generated elastic synthetic shot records using finite difference algorithm and inverted these data back for model parameters to assess the potential of the FWI algorithm. Using only a short array of surface receivers (cheaper option), we show that the application of FWI method can still improve the original earth models towards the true solutions. We were also able to reconstruct the elastic boundaries for a major part of the subsurface models within the seismic bandwidth. Interpretation of the estimated parameters for coal mining objectives is then straightforward

Building reliable starting model remains one of the most topical issue for successful application of full waveform inversion (FWI). In this study, we assess the stereotomography as a tool to build reliable starting model for frequency-domain FWI from long offset (i.e., wide-aperture) data. Stereotomography is a slope tomography method based on the use of traveltimes and slopes of locally-coherent events in the data cube. We assessed a tomographic workflow based on stereotomography and frequency-domain FWI on the 2D acoustic synthetic Val- hall case study. We first computed an acoustic full-wavefield dataset using a finite-difference time-domain modeling engine for a wide-aperture survey with a maximum offset of 24 km. The source bandwidth is between 10 and 45 Hz. Compared to a conventional application of stereotomography, we investigate in this study the benefits provided by the joint inversion of refraction and reflection traveltimes from long-offset data. The use of refraction traveltimes is expected to stabilize and improve the reconstruction of the shallow part of the model. In a similar way as for frequency-domain FWI, we design a multiscale approach which proceeds hierarchically from the wide-aperture to the short-aperture angles to mitigate the non linearity of the inversion. The starting models for FWI were built by stereotomography using three sets of picked events. For the first data set, the picking is limited to reflection travel- times for a maximum offset of 4 km. For the second data set, both refracted and reflected events were picked using an acquisition with a maximum offsets of ± 16 km. In a third test, we extended the maximum offsets to ± 24 km offset for first- arrival travel time picking. We highlight the improvements of the FWI results obtained from the starting stereotomographic model built from the long-offset data set. The improvements due to the use of long offsets data are observed at the reservoir level below the gas layers but also in the upper part of the model where the joint use of refraction and reflection travel- times is helpful to improve the ray illumination. ©2010 Society of Exploration Geophysicists

Summary While there are very few applications of land Full Waveform Inversion (FWI) compared to marine, modern on-shore wide azimuth, dense, broadband acquisition designs offer an outstanding opportunity for FWI application. Several successful examples of acoustic land FWI application to such surveys have been published. The first applications showed the potential of the approach as an early stage velocity model building tool. Later publications demonstrated that acoustic FWI can be used as an efficient model building tool for high resolution velocity models (similar in quality to marine FWI) by investing more effort in dedicated pre-processing and using a two stage workflow based on data selection (Sedova et al., 2017). We illustrate a new case study based on this workflow. We also demonstrate the impact of optimal transport (OT) FWI on the resulting velocity model, which aims to mitigate cycle skipping. We demonstrate how it corrects several localized failures with conventional FWI and constructs a geologically consistent velocity model.

Summary Onshore, the application of full waveform inversion (FWI) for imaging is challenged by the surface waves. Commonly, after their removal, it is possible to approximate the background velocities by inverting the diving/transmitted waves with low-frequency, wide-angle seismic data. However, the surface waves provide information on the near surface that is difficult to obtain from the body waves. A straight forward application of an elastic FWI on land data solely focuses on interpreting the surface waves with more pronounced energy footprints. In this study, we tested an approach based on simultaneous inversion of the surface and diving/transmitted waves, balanced through a time-space weighting scheme. The long-offset data also allows for inversion of the higher modes of the surface waves. Additionally, a novel pre-processing of the acquired seismic data aims at the elimination of the interfering secondary surface waves originating from human activities during the acquisition. Even with a relatively crude initial model, the inverted earth parameters obtained with the elastic FWI show a spectacularly good match with the well logs up to a depth of 1200 m. This study emphasizes the significance and the complementary role of both surface and diving waves in retrieving elastic earth parameters.

Full Waveform Inversion (FWI) has proven to be a powerful tool to quantify the Earth's subsurface. In geological settings, such as gas clouds, gas sand, where attenuation is important, the application of FWI is still very challenging. We have developed a viscoelastic FWI in the time domain. In this paper, we investigate the need to properly account for attenuation when inverting long offset seismic data by comparing the results of elastic FWI applied to viscoelastic data and fully viscoelastic FWI. We carried experiments for short and long offset geometry of acquisition. The effect of attenuation could be divided into two parts: during wave propagation and during reflection. We find the presence of attenuation has a significant effect on wide-angle reflection data, both for reflection and propagation, but it has little or no effect on near-offset reflection data, suggesting that the elastic approximation is only sufficient when inverting pre-critical reflections and the attenuation should be taken into account while inverting wide-angle reflection data.

This study aims at designing optimal crosshole seismic approach for CO2EOR (enhance oil recovery by CO2 injection) monitoring in Abu Dhabi. It is not easy to perform successful reservoir monitoring for carbonate reservoirs in the Middle East because rock properties may not change much associated with fluid replacement. We investigate advantage of newly available elastic full waveform inversion (FWI) for detection of CO2 flooded zones. To investigate advantage, applicability and future problem of time-lapse FWI for CO2EOR monitoring, we conducted rock physics study, elastic seismic simulation, and test application of FWI to synthetic data. We first made realistic velocity models based on Gassmann's equation and actual well logs obtained before and after CO2 injection at carbonate reservoirs in Abu Dhabi. The in-situ light oil at the target reservoirs was replaced by CO2 (Secondary mode CO2EOR). We prepared models with different width and thickness of CO2 flooded zones and analyzed detectability of seismic waveform change. We also conducted test application of elastic full waveform inversion to synthetic data for delineation of CO2 flooded zones. Results of the rock physics study show 1.6% decrease in P-wave velocity, little change in S-wave velocity and density associated with secondary mode CO2EOR. The small property changes are caused by the rigid rock frame and small density difference in light oil and supercritical CO2 at the reservoir condition. Next we performed synthetic seismic analysis based on the rock property models, available crosshole seismic source and realistic well distances. Because of the small property change, seismic traveltime difference and waveform change are quite small when well distance is 116ft. When well distance is 700 ft or greater, guided waves, or trapped energy inside low velocity layers can be observed; relatively large waveform change can be observed in the guided waves due to CO2 injection. Although guided waves have not been used for reservoir monitoring, they are sensitive to small Vp and thus can be an indicator of the CO2 flooding. The elastic FWI result for a model with well distance of 1160 ft demonstrated good representation of the differential Vp with vertical resolution of 20-30 ft and some horizontal resolution. Although we need further investigation on repeatability and signal transmittability using actual field data, our synthetic study indicates possibility of identifying layers with CO2 using guided waves and elastic FWI.

Velocity model building and depth imaging are known to be complicated endeavors in numerous onshore areas in the Middle East. Despite recent advancements toward denser, wide-azimuth, and low-frequency seismic acquisitions, the application of full-waveform inversion (FWI) remains challenging. The main difficulties arise from complex near-surface geology that creates wave mode conversions (P to S waves but also strong multiples) and very energetic and complex ground roll overlaying reflected waves and dominating low frequencies, with the diving waves being less energetic and noisier. This paper presents recent progress in land FWI toward higher-resolution velocity models, with successful applications to case studies from the Sultanate of Oman. The first example shows an application of land FWI imaging. It shows that we are able to invert jointly diving and reflected waves in acoustic FWI. With proper input data conditioning and initial model preparation, we successfully ran acoustic FWI up to 45 Hz, resulting in a high-resolution velocity model and derived FWI image that significantly improved subsurface imaging and geologic understanding. The inverted high-frequency velocity could also be more reliably used for subsequent reservoir characterization steps. Given the important role of a good-quality near-surface model for stabilizing high-frequency FWI updates, the second example highlights the creation of a detailed shallow velocity model. We propose using elastic FWI of surface waves for characterization of the near surface in complex geologic settings. Here, we demonstrate the effectiveness of this workflow to enhance imaging of the shallow subsurface. Compared to conventional methods such as surface-wave inversion, elastic FWI of surface waves has the advantage of 3D full-wavefield modeling with no requirement for picking. Our workflow also benefits from the ultra-low frequencies recovered by interferometry to obtain a more stable deeper update.

Summary Full Waveform Inversion is a powerful tool for quantitative estimation of subsurface parameters (P-wave, S-wave velocities, density, attenuation, anisotropy parameters). This methods has been applied successfully to 2D acoustic and elastic reconstructions, as well as to 3D acoustic reconstructions. Most of the applications of Full Waveform Inversion have been devoted to mono-parameter reconstructions of wave velocities. Multi-parameter Full Waveform Inversion aims at reconstructing simultaneously different class of subsurface parameters. This is a very challenging task: the similarity of the sensitivity of the data to different classes of parameters is the source of trade-off (or cross-talk) which renders the Full Waveform Inversion problem even more undetermined than in the mono-parameter context. This can related to the similarity of the diffraction patterns of different classes parameters for a given propagation regime. In order to overcome this difficulty, the role of the Hessian operator should be crucial. The off-diagonal blocks of this operator accounts for the trade-off between parameters. Incorporating the inverse Hessian operator within the Full Waveform Inversion scheme may help to alleviate this difficulty. On this basis, we provide in this study a very simple example for which we can compute exactly the Hessian operator we use to illustrate these issues.

Summary Acoustic full waveform inversion (FWI) was applied to the cross-well seismic monitoring data acquired in a carbon capture and storage (CCS) test site in Japan in order to monitor CO2 migration. Thorough parameter tests, related to frequency range and trace selection, were conducted using synthetic data of realistic velocity models created based on the real well-log data. These tests revealed the importance of low frequency data in situations where CO2 injection causes a P-wave velocity decrease and resulting high velocity contrast in the reservoir. Carefully optimized pre-processing includes angle-based trace selection, and eliminating non-acoustic waves using an F-K filter and exponential damping. As a result of these optimizations, a high-resolution P-wave velocity model was obtained from the FWI analysis. The high similarity of the field data and synthetic gathers, which were estimated from the final FWI, confirmed the validity of the results. Data elasticity is a remaining challenge, and we anticipate that the application of elastic FWI may improve the detection of 4D responses.

Full-waveform inversion (FWI) has become the centerpiece of velocity model building (VMB) in seismic data processing in recent years. It has proven capable of significantly improving the velocity model and, thus, the migration image for different acquisition types and geologic settings, including complex environments such as salt. With the advent of FWI imaging, the scope of FWI applications has extended further from VMB into the imaging landscape. However, current FWI applications in the industry prevalently employ the acoustic approximation. One common problem of acoustic FWI (A-FWI) is the apparent salt halos at the salt-sediment interface in the resulting FWI velocity and FWI image, the presence of which hinders direct interpretation and imaging focusing around salt bodies. With synthetic and field data examples, we demonstrate that this salt halo is caused mainly by the large mismatch between the elastic recorded data and the acoustic modeled data, particularly at middle to long offsets. To overcome limitations imposed by acoustic assumptions, we developed an elastic FWI (E-FWI) algorithm that combines an elastic modeling engine with the time-lag cost function, which we call elastic time-lag FWI (E-TLFWI). With a more accurate modeling engine, E-TLFWI significantly reduces the salt halo observed in its acoustic counterpart. However, we also observe that the images migrated using the A-FWI and E-FWI velocity models remain similar overall, with some slight improvements around and beneath salt boundaries, particularly near steep salt flanks, as a result of the reduced salt halo. By contrast, FWI images derived from E-TLFWI show considerable benefits over those from acoustic time-lag FWI, such as improved event focusing, better structural continuity, and higher signal-to-noise ratio. The sharpened salt boundaries and enhanced quality of the FWI images reveal the significant value of E-FWI and provide the justification for its greatly increased cost.

Summary Full Waveform Inversion (FWI) is a powerful tool to quantify the Earth's subsurface structure. However, most of the FWI applications have been limited to acoustic media. In geological settings, such as gas clouds, gas sand, melt lens, where the attenuation becomes important, one must use a viscoelastic FWI. Here we present the theory and application of a viscoelastic FWI in the time domain. First, we carried out sensitivity analyses for back-scattered, reflected and transmitted waves. We find that the presence of attenuation has a significant effect on post-critical reflections, but it has little or no effect on near-offset reflection data, suggesting that the inversion of port-critical reflections can help to reduce the cross talks between attenuation and velocity contrast or propagation effects. We have first tested the method on synthetic data and then applied to 15 km long offset data acquired by CGG Offshore central Sumatra, Indonesia. Apart from the recovery of attenuation parameters, the viscoelastic inversion provides sharper P-wave velocity image as compared to the elastic FWI.