Search

Recent advancements in seismic technologies such as full-waveform inversion (FWI), least-squares Kirchhoff pre-stack depth migration (LS-KirPSDM), and reverse time migration (RTM) hold immense potential for mineral exploration but remain underutilized. These techniques were applied to sparse 3D seismic data from the Oak Dam IOCG deposit in South Australia, delivering transformative insights. The 30Hz FWI model effectively highlighted a regional fault in the cover sequence, aligning with electromagnetic inversion data, while also identifying a low-velocity zone corresponding to the target deposit. This finding was later validated through drilling, demonstrating the accuracy of the FWI model. Furthermore, RTM and LS-KirPSDM provided a significantly enhanced image of the Hardy Hill Fault (HHF), enabling a refined geological model and guiding optimized drilling strategies. Given the deposit s location beneath 700 meters of cover, the use of these advanced seismic methods presented a cost-effective alternative to extensive drilling while yielding equivalent insights.

Up/Down Deconvolution (UDD) is valuable in the field of time-lapse (4D) seismic imaging in a marine environment, as it recovers Earth s reflectivity free from changeable factors such as water column variability and source variability. UDD is very sensitive to the accuracy of the recorded water wave, however. In shallow marine environments such as the North Sea, the high amplitude of the water wave, due to the proximity of the source to the receiver, means it can overload the dynamic range of the recording equipment causing a clipped waveform. Clipping of the water wave creates severe artifacts on the reflectivity produced by UDD. Existing solutions to correct for clipping of the water wave are not robust enough under the scrutiny of 4D imaging when using UDD. Here, we describe a new wavefield filtering method for correcting clipped energy by predicting replacement values from coincident geophone data. In our example from the Valhall life-of-field 4D seismic dataset, the new method outperforms the common industry practice of wavelet replacement. Results suggest this method is robust enough to recover errors of the water wave recording, and it can be used to achieve consistency between up and downgoing wavefields.

Distributed Acoustic Sensing on seafloor fibre optic cables (Surface DAS) provides a cost-effective alternative to Ocean Bottom Seismic. Measuring the change of a fibres optical response over time yields a record of strain along the cable, which can be converted into a seismic signal. While DAS is routinely used for PP primary imaging, the remainder of the wavefield is usually discarded. Here we apply three distinct imaging techniques to a Central North Sea Surface DAS dataset, presenting PP, PS, and multiple migrations. The results showcase the imaging potential of Surface DAS data beyond PP reflections, including in the cross-cable direction.

Carbon sequestration monitoring demands high-resolution imaging to ensure conformance and containment of the injected CO2. This requires good imaging of the reservoir, CO2 plume, CO2 migration pathways, and depositional features, as well as detailed mapping of the overburden, to detect small volumes of CO2 in case of leakage. In 2023, a regional, high-quality, ocean bottom node survey, covering the Sleipner storage site was acquired, comprising long-offsets, multi-azimuth illumination and multi-component recordings. We applied Full Waveform Inversion (FWI) up to 70 Hz, the highest frequency inverted for at the Sleipner plume, to achieve a high-fidelity reconstruction of the CO2 plume structure, migration pathways, and depositional features influencing CO2 migration. The inversion also provided a high-resolution image of the near surface, filling illumination gaps and offering improved definition of vertical migration pathways for CO2 compared to Reverse Time Migration (RTM). Our results demonstrate FWI as valuable tool for conformance and containment assurance, that can also provide valuable insights for CO2 migration forecasting.

Traditional imaging of the Frigg-Gamma structure in the Yggdrasil area of the Norwegian North Sea is compromised by the presence of large vertical columns of calcite cementation. These cemented pipes have a highly contrasted seismic velocity regime compared with the surrounding sediments, causing strong wavefield scattering along their rugose boundary. Consequently, the Frigg-Gamma structure, which sits below these pipes, is in a seismically obscured area (SOA). We address the imaging challenges in this area caused by these, and other, complex overburden features with the use of elastic full-waveform inversion (FWI) using a shallow-water multi-component ocean bottom node data set. The superior modelling of the full wavefield through the elastic propagation in FWI creates a highfrequency velocity model (up to 64 Hz) that improves the imaging of primary data with reverse time migration (RTM). Yet, this primary-only image is still limited by the wavefield scattering caused by the cemented pipes. Our FWI Image reflectivity volume, derived directly from the elastic FWI, benefits from the full-wavefield contribution and can surpass the RTM volume, resulting in a significantly improved image of the SOA.

In recent years distributed acoustic sensing (DAS) using a vertical seismic profile (VSP) geometry has gained significant interest as a technology for addressing the challenges of subsurface imaging. This study presents the integration of data from 16 DAS-VSP wells for simultaneous multi-well processing using a single imaging routine. Conducted in shallow water offshore United Arab Emirates, the study addresses noise and data irregularities from acquisition obstacles. A comprehensive denoise workflow, including a novel 3D sparse tau-p guided de-noise method, was implemented to address noise challenges associated with the lower S/N often associated with DAS-VSP data. Furthermore, the acquisition involved continuous recording with crosstalk from up to four source vessels being separated using a proprietary deblending workflow. Significant surface infrastructure required shot regularization using a two-pass approach to mitigate the impact of these acquisition holes on imaging quality. A wave-equation-based joint primary and multiple imaging (JPMI) algorithm resulted in a subsurface reflectivity down to 7 km. This approach provided valuable near-surface characterization, drilling hazard identification, and deeper reservoir analysis.

Elastic full-waveform inversion (FWI) is becoming more frequently used in compressional velocity (Vp) model building. However, the inversion of shear-waves in elastic FWI is more challenging. Here, we propose a methodology to update a high-resolution shear-wave velocity (Vs) model with elastic FWI, driven largely by the converted waves recorded in multi-component ocean-bottom seismic data. We demonstrate this with a field example from the Central North Sea. Using an accurate Vp model from a 25 Hz elastic FWI update, which utilised both raw compressional hydrophone and vertical geophone components, we subsequently run a 25 Hz elastic FWI update of the Vs model using the raw horizontal geophone components. The Vs update is geologically consistent, showing a good match with the underlying seismic data and the well data. We also show that the inversion result is robust with respect to the starting model. In addition, the extracted Vp/Vs ratio from the elastic FWI highlights geological features such as gas pockets and injectites. A comparison is made to a full-bandwidth amplitude-versusangle inversion of fully processed and imaged PP-data, highlighting benefits in some areas with the new method, as well as comparable features and anomalies elsewhere.

Land Full-Waveform Inversion holds great potential confirmed by recent results, nevertheless, securing well acquired data is imperative for optimal outcomes considering the complexities of near surface geology. Enhanced land FWI performance arises from nodal acquisitions incorporating blending and compressive sensing, offering continuous recording, longer offsets and denser samples even if decimated. Creating passive data via interferometry from continuous recording leads to very lowfrequency FWI updates, subsequently decreasing cycle skipping risks at higher frequencies. Employing very long offsets in FWI ensures exhaustive layered updates involving crucial structures and faults, thus making high-frequency FWI possible. In a 2023 3D land survey exploring the northeastern portion of the Sultanate of Oman, we illustrate how land FWI capitalizes on this recent acquisition, providing even superior FWI imaging results compared to PreSDM in the most structurally complex area, beneath the sand dunes.

The West Africa Atlantic margin is characterized by complex geology and their associated imaging challenges. In this case study, the basin developed from the Late Jurassic to the Early Cretaceous period, showing an example of continental rifting and break-up with strong volcanism affecting the overburden. This geological context has generated complex gravity-driven collapse structures, such as mass transport deposits, showing significant lateral acoustic velocity changes. These changes generate seismic imaging distortions at the lower and upper Cretaceous sand reservoirs beneath. In this study, we describe a seismic imaging velocity model building workflow which captured the details of the mass transport complexes, allowing for restoration of the geometries of the Cenomanian and Albian reservoirs. In a very deep-water environment and with limited offsets and angle range, the restricted illumination from diving waves required our velocity model building to rely only on reflections. We show how we were able to leverage the information from reflections using Dynamic Resolution Time-Lag Full-Waveform inversion (DR-TLFWI; Wang et al., 2023) to solve the challenges of this area.