subscribe
 

3D seismic data: taking a smarter approach to interpretation

21st February 2013


The demand for computational tools to underpin the 3D seismic interpretation process has never been more apparent. Jeremy Cresswell reports.

There is a huge and growing mountain of seismic data - new and archived material requiring reprocessing - out there. To interpret it means that the demand for computational tools to underpin the 3D seismic interpretation process throughout the E&P (exploration and production) workflow has never been more apparent, according to UK company Foster Findlay Associates (FFA), which has just teamed up with Dutch IT outfit Bull to deliver a smarter approach.

FFA says that, through use of its high performance computational software, a data driven approach to 3D seismic interpretation can provide the step change in productivity required for geoscientists to 'fully exploit their data and deliver more accurate interpretations of the subsurface, improved understanding of uncertainty and better quantification of risk'.

Indeed, according to FFA's MD, Jonathan Henderson, the partnership with Bull is an important component in the UK firm's development strategy which has always been about delivering innovative and sophisticated tools designed to unlock the innermost secrets of seismic data.

Henderson said in a statement issued early June ahead of the 2009 European Association of Geoscientists and Engineers gathering: "We are excited to be working with Bull, who bring a strong innovative approach to their systems design, which incorporates computational power and scalability with advanced interactive visualisation features. Our development programme is directly addressing the challenges of extracting meaningful information from seismic data at increasing volumes and levels of complexity. These challenges can be significantly advanced by using automated HPC techniques in real or near real time.'

Basically, the two companies will work together to promote the use within the E&P industry of interactive methods for 3D seismic analysis using HPC (high-productivity combinatorial) technology, in particular the use of GPU (graphics processing unit) accelerators.

Under the arrangement forged between the companies, these are fitted into Bull server nodes, from entry-level tower chassis systems with GPU and graphics cards, suitable for seismic interpretation workgroups, to very large hybrid clusters located in data centre, which Bull is already delivering to government and energy customers.

Mostly Aberdeen-based FFA's domain expertise and involvement with a large number of major oil and gas companies' petrotechnical users combined with Bull's expertise in hybrid systems is expected to provide a fast-track for innovation and improved turnaround times within 3D interpretation and modelling workflows.

FFA specialises in the development and application of automated interpretation and modelling tools that are capable of plugging into established 3D interpretation and modelling workflows. That means they can make use of more scalable and powerful technologies of the kind built into Bull's 'extreme' computing systems.

According to the duo, these outperform the traditional commodity platforms currently in use within the oil industry and are expected to deliver significant productivity advantages to those who adopt them."

Crossing the Atlantic to Houston, Spectrum has reported that it has achieved impressive results in processing a 28 000km US Gulf of Mexico seismic survey using Wave Equation Migration (WEM) techniques to reveal subsalt geology.

This is an important step forward as subsalt offers considerable future potential in terms of major new hydrocarbons discoveries, with ness the Jack find by Chevron and strong of massive subsalt (pre-salt) finds offshore Brazil by Petrobras and others.

Spectrum said that the 2D survey was processed at the company's newly refurbished Houston data processing and imaging facility using both Kirchhoff and wave equation depth migration processing.

The company's data interpreters first processed the seismic using Kirchhoff eikonal pre-stack depth migration (PSDM), which produced reliable data and revealed detail about the complex geology in the Gulf of Mexico. They then considered applying wave equation migration and it is this has paid off by revealing clearer subsalt definition.

This is an important breakthrough as subsalt is notoriously difficult to image through and data is often of questionable quality.

WEM embraces multiple arrivals of seismic waves in the subsurface, so overcoming the limitations of standard Kirchhoff PSDM which takes into account only the first arrival of seismic waves.

According to Spectrum, compared to normal Kirchhoff migrations, the wave equation method increases the focusing of seismic images in areas with complex velocity models and excels in imaging below geology that is traditionally difficult to penetrate with seismic signals, such as salt formations.

Kirchhoff and WEM processing are both valuable tools because they help reveal different geological features. However, the workflow is quite different because the input for WEM consists of shot gathers.

Using both techniques Spectrum says it has has been able to learn important information on the subsalt structure from this eastern Gulf of Mexico data.

Following these results, the company also tested the WEM technique on its 'Big Wave' Phase 1 multi-client project and concluded that it showed clear potential to improve the imaging of the Florida escarpment and the base of salt definition.

Remaining in Houston and staying on the subsalt/pre-salt theme, in April, ION Geophysical said it had successfully completed the first phase of its BrasilSPAN seismic data programme and delivered the processed data, seismic images, and an integrated interpretation.

According to ION, BrasilSPAN as the first comprehensive, presalt geophysical study of Brazil's prolific offshore hydrocarbon basins in which several multi-billion bbl discoveries have been announced in recent years.

BrasilSPAN comprises 12000km of seismic data that images the entire crustal section in Brazil's Santos, Campos, and Espirito Santos hydrocarbon basins. ION said it worked closely with regional experts to design the survey and tie the acquired seismic data lines to critical wells in the basins of interest.

The program was acquired with long offsets (a relatively successful technique for imaging subsalt and through basalt), long listen times, and imaged with proprietary reverse time migration (RTM) technology from ION's GX Technology seismic imaging subsidiary to provide improved resolution in the presalt and salt flanks of these prolific hydrocarbon basins.

BrasilSPAN was planned in partnership with ION's AfricaSPAN customers in order to address questions of great interest to global exploration teams, including how the conjugate ties between West Africa and Brazil affect the exploration potential along the continental margins on both sides of the present-day Atlantic Ocean. Not only have huge subsalt hydrocarbons discoveries been made offshore Brazil, BP has located several subsalt fields on block 31 offshore Angola.

The regional, conjugate, and interpretative studies that are part of the BrasilSPAN program are said to assist in an understanding of the mechanism for the break-up of the continents, help develop and de-risk exploration play concepts, and provide a framework to characterise the hydrocarbon potential along the offshore margins of both West Africa and Brazil.

Heading further westward and completing this brief tour of the latest in data interpretation, Nvidia Corporation of Santa Clara, California and the Chinese geophysical services provider GeoStar have unveiled a new hardware and software solution that draws upon the processing power of NVIDIA Tesla GPUs to dramatically accelerate the performance of complex seismic data with GeoStar's seismic software suite.

According to GeoStar general manager, Liu Qin, said the computation of large datasets can be achieved in smaller, more power and efficient GPU-based systems as compared to CPU-only based clusters.







Subscribe

Subscribe



Newsbrief

twitter facebook linkedin © Setform Limited