Search

Elastic full-waveform inversion (EFWI) and high-frequency EFWI Imaging have gained increasing popularity, primarily with ocean bottom node (OBN) data, in areas where strong velocity contrasts and associated elastic effects are present, such as the Gulf of America (GOA). However, when only vintage streamer data are available in such regions, either wide or narrow azimuth streamer (WAZ/NAZ) data, seismic processing remains largely limited to an acoustic framework, based on the hypothesis that elastic effects are less significant for streamer data with limited offsets. Through a subsalt imaging case study in the GOA, we demonstrate that even with WAZ data, processing within an elastic framework, comprising both EFWI and elastic reverse-time migration (ERTM), is needed to properly utilize the postcritical information of the input streamer data, which leads to imaging improvements in complex salt environments. To better understand the improvements observed from the elastic solution in field data, a synthetic study was conducted to illustrate the potential impact of strong elastic effects arising from the highly dipping salt-sediment interface in the overburden. The synthetic study shows that elastic effects can be recorded from near to far offsets within a typical streamer offset range. Therefore, improvements from the elastic solution can be observed across all offsets in the gather domain, which is consistent with the observations from field data. Moreover, high-frequency EFWI Imaging further unlocks the value of vintage WAZ data, providing images with better signal-to-noise ratio (S/N) and subsalt illumination compared to ERTM, particularly in challenging subsalt areas.

Time-lapse (4D) seismic monitoring generally benefits from a dedicated co-processing where important inconsistencies between vintages are attenuated. We propose a joint formulation for least-squares migration (LSM) that further mitigates effects such as low repeatability and diverse data quality, by constraining the least-squares inversion with information shared among baseline and monitor data sets. It can be applied in both post-stack and pre-stack domains. We show results for two pairs of baseline/monitor data sets with different levels of repeatability in the Brazilian pre-salt. In both cases, joint 4D LSM gave an overall reduction of 4D noise.

Marine sediments are often electrically anisotropic at a scale that affects controlled source electromagnetic (CSEM) data. When the sedimentary sequence is steeply dipping, it is important to model the data using resistivity fields that follow the sedimentary bedding, which is called tilted transverse isotropy (TTI). In this paper, we introduce our approach to the inversion of CSEM data in TTI media, which builds on our previous work using seismic-guided inversion. We illustrate this with synthetic examples representative of the geological settings encountered in Sabah and show how results impact data analysis in the area.

In onshore southern Mexico, thick Mesozoic intervals, extensive salt, and over-pressured shale pose significant challenges for seismic imaging. Leveraging newly acquired data with longer offsets, higher fold, and improved low-frequency signal, we adopt a time-lag full-waveform inversion (TLFWI) driven workflow to resolve the complex overburden and the Mesozoic targets. Moreover, we employ FWI Imaging, which exploits the full wavefield for imaging, to enable clearer salt-flank delineation, enhanced lateral resolution of Mesozoic fault blocks, and more balanced amplitudes compared to reverse time migration (RTM) images. Subsequently, FWI Images are far less sensitive to acquisition irregularities imposed by onshore logistical constraints, such as variable charge strengths and large gaps in the shot coverage due to urbanization.

In recent years, full-waveform inversion (FWI) techniques have evolved significantly - from acoustic FWI (AFWI) that works surprisingly well in salt environments to the latest elastic FWI (EFWI) and high-frequency EFWI Imaging that provides unprecedented clarity in subsalt imaging - largely driven by rapid advancements in algorithms and exponential increases in computing capacity. However, for salt and subsalt velocity updates in the Gulf of Mexico (GoM), poor low-frequency signal-to-noise ratio (S/N), lack of diving wave penetration, and uneven subsalt illumination still impede reliable subsalt imaging, even with state-of-the-art FWI algorithms. With a recently acquired long-offset, low frequency (LOLF) sparse ocean bottom node (OBN) data set, we successfully applied elastic Time-lag FWI (TLFWI) to invert subsalt velocity and generate a high-fidelity FWI Image. The survey utilized a Tuned Pulse Source (TPS) as the energy source, providing usable low-frequency signals down to ~1 Hz. The combination of cutting-edge algorithm, innovative source design, and acquisition results in significant uplift in subsalt imaging in the structurally complex Garden Banks and Keathley Canyon (GB-KC) areas. Some key enhancements include redefined salt bodies, better delineation of over-pressure zones and mobile shales, and improved resolution of thin salt feeders and welds.

Simultaneous source acquisition significantly enhances ocean bottom node (OBN) survey efficiency, but it also introduces blending noise contamination in continuous recordings. Deblending, which separates primary source signals from blended data, is a crucial step in preprocessing for both 3D and 4D imaging using conventional algorithms such as reverse time migration (RTM). Effective deblending normally relies on a good dither scheme and dense wavefield sampling. In this study, however, we performed deblending for a sparse OBN survey optimized for full-waveform inversion (FWI). The survey featured low-frequency sources, a high blending fold, a sparse shot grid, short dither intervals, and strong neighboring survey interference, all of which posed challenges for deblending. Despite these challenges, our hybrid deblending approach can effectively extract primary signals, including those at low frequencies, leading to a significantly higher signal-to-noise ratio (S/N) in time domain data and migrated depth images.

A comprehensive assessment of geomechanical risks is essential for the success of geological carbon storage. Although the relationship between injection-induced pore pressure changes and rock failure is broadly understood in poromechanics, the challenge of characterizing and managing geomechanical hazards associated with subsurface CO₂ injection in saline aquifers persists, affecting its viability as a global strategy for achieving a net-zero carbon economy. Here, we investigate the geomechanical response of the Luna saline aquifer in the North Sea to CO₂ injection using coupled compositional flow and geomechanics simulations based on a seismic-driven high fidelity static model. We find a surface uplift of 13 mm following a constant injection rate of 1.5×10⁶ m³/d (i.e., 1 Mt/y) in the Luna aquifer for 20 years, which is considered reasonable and safe for offshore surface facilities. We show that the upper limits of shear stress level (SSL) and tensile stress level (TSL) are 0.5 and 0.78, respectively, in both the reservoir and caprock. Since SSL and TSL values below 0.8 are considered safe, our results suggest that gas injection induced reservoir and caprock failure is unlikely in the study area, providing confidence that caprock mechanical leakage poses a low risk for long-term CO₂ storage.

Structural distortions are frequently encountered in Abu Dhabi area with significant velocity heterogeneity and complex geological features. Due to the unique challenges, FWI often struggled to accurately resolve small-scale distortions. The hard seabed and ultra shallow water environments lead to significant noise that destabilize diving wave driven FWI. These challenges are further exacerbated by the lack of low-frequency signals. In this paper, we propose a workflow that incorporates MWI in the near surface and utilizes reflections and diving waves simultaneously with TLFWI to overcome the modelling challenges. The workflow enhances inversion stability and improves FWI convergence for deeper sections. The results highlight the enhancement on the continuity and focusing of migrated images in the area, effectively resolving structural distortions caused by complex karst formations and deeper channels.

In this case study, we present results from the first ocean bottom node (OBN) based 4D imaging at the Liza field, offshore Guyana. With two nominally repeated OBN surveys acquired within less than a year, we initially employed a parallel 3D processing flow to generate fast track 4D images within 12 weeks of receiving all the monitor data, facilitating an early interpretation of the 4D response. Subsequently, the full-track flow utilized all necessary 4D co-processing technologies to address non-repeatability issues from their physical origins. The final products offered high-resolution 4D signals with minimal noise, which was crucial for understanding reservoir changes during this brief period of production and gas/water injection. The improvements achieved by transitioning from a fast-track parallel flow to a full-track co-processing flow demonstrate that, even with highly repeatable time-lapse OBN surveys, appropriate 4D co-processing solutions are still essential for maximizing the value of OBN data for reservoir monitoring.