EAGE 2024 to be held in Oslo on June 10-14 will be the final major conference representation by PGS before the merger with TGS.
The technical content to be presented captures three of the major industry shifts in recent years:
- A far greater focus on the shift to lower carbon energy supplies.
- The growth of HPC and associated AI-driven innovations.
- The eternal focus is on operations with lower costs and a lower environmental footprint.
Hyperlinks to PDF abstracts are provided below, and we welcome any queries on Booth # 4000 in Hall D at EAGE 2024 or via marketing@pgs.com
Multi-Parameter Seismic Imaging and Subsurface Characterization
Simultaneous inversion of velocity and reflectivity, otherwise also known as multi-parameter FWI (MP-FWI) has found global application since PGS introduced it several years ago. Our unique formulation of the underlying physics has notably created intriguing opportunities for robust quantitative interpretation (QI). Used more ‘conventionally’, “Applications of multi-parameter FWI in complex geological settings in the North Sea” by Øystein Korsmo provides a diverse set of case study applications for MP-FWI in various geological settings. Alternatively, “Frontier exploration insights using simultaneous inversion of velocity and reflectivity: a case study, Offshore Canada” by Nizar Chemingui illustrates opportunities for frontier exploration derisking, including the improved delineation of facies morphology, compaction trends, and possible fluid types within vast turbiditic fan complexes.
Although FWI has reached great prominence in recent years for subsalt and presalt imaging improvements, PGS has also been developing solutions for ultra-high-resolution (UHR) 3D data acquired for offshore wind farm planning and management. Nizar Chemingui will showcase such developments with “Incorporating multiples for accelerated construction of ground models for emerging energy applications” in Workshop 6 on Sunday, June 9. The focus of this workshop is on the value of multiples in seismic imaging—a natural feature of FWI and MP-FWI. In the same workshop, Øystein Korsmo will more closely follow the workshop theme with “Exploring multiples in full wavefield imaging and velocity estimations across varied geological landscapes”.
Higher Frequencies, Lower Frequencies, Longer Offsets regards the application of UHR 3D seismic data, as the streamers are typically towed with a focus on the imaging of kilohertz frequencies, the low-frequency content is naturally weaker than for traditional GeoStreamer surveys. Fortunately, “Improving the results of shallow marine surveys imaging with reconstructed low-frequency seismic data” by Ramzi Djebbi describes a novel approach to enhance the application of FWI to shallow marine seismic imaging through reconstructed low-frequency seismic data.
Longer offsets are highly beneficial to FWI model building with deep targets, especially in complex settings, and “Streamer-tail optimization for FWI” by Tashi Tshering describes a new survey design with strategic placement of streamer-tails to efficiently improve both model building and imaging quality.
The relative sea state is also critical to the success of kilohertz UHR deghosting; a topic addressed in “Deghosting of UHR seismic data via dynamic ghost tracking” by Maiza Bekara showcased with real data from a recent survey. Two other PGS presentations describe recent experiences with the imaging and interpretation of UHR 3D seismic data. “Elevating 3D Ultra High-Resolution Processing and Imaging for Wind Farm Site Characterization” by Luca Limonta also presents data from the Irish Sea, and Allan McKay will present “Squaring the Circle of Resolution, Efficiency and Interpretation Ambiguity for Buried Boulder Hazard Assessment” in Workshop 2 on Monday, June 10.
“High-resolution FWI through dual azimuth seismic data integration in Halten Dønna Terraces area: A comparative study” by Amir Asnaashari presents a comparative study on high-resolution Full Waveform Inversion (FWI) using dual-azimuth seismic data integration and longer offsets in the Norwegian Sea. The examples again showcase the new standards being set by FWI applications for seismic imaging in challenging environments.
High-Performance Computing: The Future of Everything and No Longer Monolithic
Such advances do not come for free, and FWI has largely become an industry default only recently through the availability of substantial high-performance computing (HPC) resources. In parallel, vast on-premise supercomputers are no longer required to realize such ambitions, and PGS has been leading from the front in developing flexible and powerful cloud-based seismic imaging and exploitation platforms. Such platforms will become even more critical as the adoption of artificial intelligence (AI) solutions requires HPC solutions on an even greater scale in the future.
In “Seismic inversion on Federated Kubernetes Clusters” by Malcolm Griffiths outlines a cloud-native, fully fault-tolerant approach for executing HPC seismic imaging algorithms using Kubernetes clusters. At a more fundamental level, “Microservices: An Agile Infrastructure for an Agile World” by M’Lynn Jenkins addresses the transition of current monolithic HPC infrastructure to a more agile and robust microservice architecture. This shift aims to replace monolithic, legacy systems with more scalable, modern solutions, optimizing both back-end processes and user interfaces.
Cloud-based seismic data management and delivery will, of course, be the platform for everything going forward, perhaps illustrated no better than by the vendor-collaborative cloud-based ecosystem named Versal. “Versal – how cross-organizational collaboration helps solve industry challenges” by Espen Grimstad provides insight into the development and implementation of the Versal ecosystem—a platform built for the industry by the industry—including the ongoing integration with OSDU.
Machine Learning: From Denoise to Solving the Ultimate Subsurface Challenges
The application of machine learning (ML) to seismic processes is steadily growing. “Machine learning air-leak detection during active acquisition for seismic air gun arrays” by Bagher Farmani implements a solution for real-time detection of air leaks in marine seismic air gun arrays, enhancing operational efficiency by reducing downtime. “Shortening turnaround time for high-resolution velocity model building with deep learning ” by Sean Crawley introduces a novel deep learning approach using Fourier Neural Operators (FNOs) for rapid generation of high-resolution seismic velocity models from migrated data. This method significantly reduces the computational load and iterations required in FWI, as demonstrated with data from offshore Brazil. It’s a grand ambition to finally make real progress.
Rounding out the PGS contributions to better seismic imaging, “Improved multiple attenuation through 5D near-offset extrapolation for convolutional 3D SRME modeling and optimized adaptive subtraction” by Dragica Tomova outlines the implementation of a new 5D near-offset extrapolation technique to enhance multiple attenuation in multi-azimuth (MAZ) seismic data from the Norwegian Sea.
Marine Seismic Acquisition: Fit for More Applications, Lower Cost, Lower Environmental Footprint
PGS continues to push the envelope in marine seismic acquisition, improving efficiency and extending the applicability of towed streamer seismic to both wind farm planning and carbon capture and storage (CCS).
Martin Widmaier will present “Navigating the future: Insights from recent CCS 3D streamer seismic surveys” in Workshop 1 on Sunday, June 9. Multi-source and multisensor streamer configurations facilitate high-resolution imaging across a range of depths and provide the optimum baseline surveys for 4D seismic monitoring that will be crucial for ongoing CCS monitoring and evaluation.
Simultaneous (i.e., blended) shooting is becoming standard in marine acquisition for both towed streamer and ocean bottom node (OBN) surveys but has not been universally perfected. “Revisiting the generation of dither times in the context of blended marine seismic acquisition” by Maiza Bekara proposes an enhanced method for generating dither times with a statistically flatter distribution and wider spread of differences, thereby offering a more controlled and efficient way to meet tighter deblending constraints in survey specifications.
As the merger of PGS and TGS draws closer, the integration of OBN seismic into projects of all scale will increase. “Drop & Pop tri-component ArchiNode” by Didier Lecerf in Workshop 14 on Sunday, June 9, is a timely reminder that innovation in the OBN space is something to look forward to.
Pushing the Envelope on Platforms to Grow CCS
As industry appreciation for the scale of CCS projects necessary to meet net-zero ambitions grows, so do efforts to acquire new 3D datasets fit for both traditional exploration and CCS and developing workflows to extract value from legacy 3D datasets with strategic reprocessing. Part of the challenge includes the identification of CCS storage containers from legacy 3D seismic data, and the other part obviously involves the development of new seismic programs and methods that are both efficient and have reduced environmental impact.
“Subsurface CO2 storage potential of the Central Åsta Graben – Norwegian North Sea” by Eric Mueller evaluates the potential of the Central Åsta Graben in the Norwegian North Sea for large-scale CCS. This study highlights the many geological considerations necessary to identify long-term CCS candidates.
“Rock physics-guided seismic survey optimization for CO2 injection monitoring - model building” by Tilman Kluver introduces the CLEAN4D approach to optimizing seismic survey techniques for CO2 injection monitoring. By reducing the need for strict geometric repeatability in data acquisition, CLEAN4D aims to decrease both the cost and environmental impact of seismic surveys while enhancing data quality.
Three PGS presentations address legacy data reinvigoration for CCS pursuits. “Carbon storage assessment and near field exploration: a dual-purposes reprocessing case study in the UK Southern North Sea” by Eric Mueller describes the reprocessing of 26 seismic surveys covering approximately 12,000 sq km to improve seismic imaging for both CCS site screening in the Triassic Bunter Formation and hydrocarbon exploration in the Permian Rotliegend Group. “Opportunistic 4D using a regional non-repeated 4D monitor, an Ærfugl case study” by Didier Lecerf makes opportunistic use of a multi-client dataset for monitoring gas field production in the Ærfugl field, Norwegian Sea, focusing on the challenges and outcomes of employing a non-repeated 4D monitor setup.
Despite discrepancies in source setup and streamer geometry between the legacy surveys (2005 and 2017) and the 2022 MultiClient survey, the successful methodology prompted the expansion of the monitoring area to cover the entire Ærfugl field, highlighting the importance of continuous collaboration between geophysicists and interpreters in validating processing and imaging steps. Finally, “Shallow water GoM data rejuvenation for CCS prospecting” by Rachel Collings rejuvenated shallow water Gulf of Mexico data for (CCS prospecting, utilizing legacy ocean bottom cable (OBC) data reprocessed with modern imaging technologies.