Abstract
In the 1920s, the first oil discovery in Seminole, Oklahoma, was made at a depth of approximately 4000 ft using seismic reflection data. This breakthrough was achieved through the collaboration of Karcher's Geophysical Research Corporation and Amerada Petroleum Corporation. Today, J. Clarence Karcher's legacy lives on through SEG's prestigious annual award, presented to outstanding young geophysicists. Over the years, there has been a notable increase in the average depth of oil wells drilled. In 1949, the average depth was 3600 ft. By 2008, it was 6000 ft. As the consumption of oil and gas continues to grow globally, exploring deeper is a steady trend for hydrocarbon discovery. Currently the world's deepest oil well in Russia extends 49,000 ft into the earth's surface. Deep exploration below existing production, complex overburden, or at the limits of geophysical resolution is critical for existing and emerging ventures. To meet these challenges, researchers and geoscientists are actively working to acquire better data and develop innovative methods to improve imaging. Recent advances in hardware and equipment, data acquisition, algorithm development, computing, and digital technology allow for offshore and onshore exploration success and extended capability for exploration and production at higher efficiency in challenging deep environments.
Various geologic settings present unique challenges when it comes to deep exploration. When exploring on land, the near-surface challenges and enhanced signal-to-noise ratio are continuously improved to accurately image deep targets. For deep marine exploration, significant advancements in both data acquisition methods and algorithms over the past two decades have had a profound impact. One notable development in marine exploration is the utilization of ocean-bottom nodes (OBNs) for data acquisition. OBNs offer advantages such as wide azimuth coverage, improved signal-to-noise ratios, multicomponent data, and 4D monitoring capabilities. The first 2D OBN case studies were conducted in the North Sea during the 1990s. In 2004, a milestone was achieved with the completion of the first comprehensive 3D OBN survey in the Gulf of Mexico (GoM). Cutting-edge technology now empowers exploration to push beyond its existing limitations. The combination of OBN and full-waveform inversion (FWI) now routinely resolves deep structures in the GoM. Machine learning has also emerged as a valuable tool in deep exploration, aiding the processing, imaging, and interpretation of seismic data, providing valuable insights with improved overall efficiency. The introduction of distributed acoustic sensing (DAS) represents a significant leap in seismic resolution of vertical seismic profiles (VSPs), enhancing efficiency in data collection and analysis. Additionally, the adoption of advanced downhole well-logging technologies has improved formation evaluation and further reduced uncertainty in deep exploration projects.
For this special section, we showcase six technical papers on recent deep exploration methods and applications in hydrocarbon exploration at the leading edge of solving future challenges. To begin the special section, Vigh et al. demonstrate how OBN acquisition and FWI technology yield high-quality inversion results. The authors elaborate on the simultaneous source optimal acquisition design and deblending techniques. In terms of amplitude analysis, the authors suggest substituting Kirchhoff or reverse time migration surface offset gathers with FWI for amplitude variation with offset analysis, allowing for improved amplitude continuity across offsets.
Mohamed et al. present an effective machine learning approach, which was applied to generate a robust initial model for seismic inversion. Its success was demonstrated through two compelling case studies, confirming its feasibility. The results from the multiple linear regression approach show uplift over the conventional methods.
Aldawood et al. show a successful dual-well DAS VSP walkaway acquisition in a deep desert environment in the Middle East. The data obtained from DAS VSP were effectively processed, achieving a high vertical resolution of 15 m. The authors also highlight the advantages of DAS VSP compared to conventional VSP methods in terms of cost and data quality. This information can help readers assess its suitability and identify potential challenges or advantages of DAS for their specific geologic settings.
Bakulin et al. describe an approach to address “speckle noise” caused by short-wavelength heterogeneity in the near surface, which could severely distort the wavefields and diminish imaging of the deeper section. The authors present promising prestack results comparing conventional processing with nonlinear beamforming and demonstrate despeckled results through seismic time-frequency masking.
Zhao et al. present an interesting case study on onshore velocity model building, which was conducted using joint tomography to enhance deep seismic imaging. The field data yielded promising results in the Junggar Basin, a foreland setting where imaging is particularly challenged. The study offers valuable insights for exploration in foothills areas, with comprehensive details on the geologic background and challenges encountered.
Oktariena et al. successfully utilize dipole sonic imager anisotropy information to reconstruct the velocity log from the deviated well measurement and verticalize the log to represent subsurface conditions in the deepwater Sadewa Field. By modifying an existing method, the authors were able to relax the constraints of the initial guess, allowing for more accurate predictions.
The editors of this special section express great gratitude to all the authors and reviewers for their contributions and efforts toward this special section.
