To grapple with computing-intensive imaging applications, high-performance computing (HPC) becomes essential for exploration and production companies.

The typical size of a seismic dataset is about 56 gigabytes.  When a few metrics are changed in the dataset, the size can quickly scale to one terabyte or more.

As hydrocarbons get increasingly more difficult to find and existing resources dwindle, companies need to explore and interpret larger datasets to bring greater accuracy and precision to their costly oilfield investment decisions.

Using a spreadsheet to understand these voluminous datasets can be a challenge, but analyzing them in a large-format visual environment becomes relatively easy, explained Bill Bartling, senior director, marketing strategy, Silicon Graphics Inc. (SGI) during a joint SGI-Schlumberger presentation in Houston on Jan. 22, 2004.

To process or interpret datasets so that images can be drawn on the screen and made to be interactive, companies would need to place data in the active memory or random access memory of the computer.

Since data-access speeds from typical desktop computer disks can be very slow, the images will be jerky and incomplete because of insufficient numbers of image frames being pulled up at a time.

HPC refers to the system of supercomputers, parallel processing algorithms and software programs that divide data into pieces so that each piece can be executed simultaneously and therefore quickly by separate processors.

E&P companies rely on HPC for data interpretation for complex imaging applications like 3D and 4D seismic and for ever-more powerful modeling techniques.

Favorable economics driving adoption.  E&P companies are rapidly moving from 2D seismic to 3D and 4D (time-lapse seismic) visualization technology.

"There are about 100 active time-lapse seismic programs around the globe," said Judson Jacobs, associate director, CERA during CERAWeek 2004.

Time-lapse seismic is being employed for mature brownfield assets that can still sustain long production lives as well as larger, greenfield assets.

A full 100 percent of the super-majors and national oil companies have adopted 3D seismic visualization technology, according to Bartling.

Approximately 80 percent of larger independents have adopted this technology. 

The 3D visualization technologies are being used much more for the well planning or exploration aspect than the reservoir management or production aspect of the business within E&P companies, he commented.  

What's driving the adoption of increasingly more sophisticated reservoir characterization and visualization technology among E&P companies?

Favorable economics resulting from the use of these visualization technologies is the obvious driver.

Improved reservoir recovery is one such factor, said Jacobs.

"Companies using time-lapse seismic technologies are benefiting from 1.0-percent to 3.0-percent increases in reservoir recovery," according to him.  "That's substantial."

Further, companies are drilling fewer dry holes as a result of applying this technology.

One company interviewed by CERA suggested that it drilled four fewer dry holes in the last five years as a consequence of applying time-lapse seismic technology.

"If an average well costs $10 million to drill, we can start to see how profitable it is to use this technology," said Jacobs.

Bartling explained how new seismic technologies are allowing more to be done in less time.

"When we contrast the current seismic interpretation work processes to the ones 10 or 12 years ago, we realize there has been about a 100,000-fold productivity improvement for geologists and geophysicists using the new technologies," he asserted.

"For example, when I started working on seismic interpretation, I waded through 100 megabytes of 2D line data drawn up on paper, and the papers usually ended up in a big stack in one corner of my office.

"It took me a year to process about 10 percent of this data and create a map of one or two geological surfaces and get some understanding of the geology that we were examining.

"Today, we process 100 percent of 100 gigabytes of data in about a month," he stated.
 
Economics of immersive imaging. Bartling offered examples of two companies that benefited from using immersive 3D-seismic technology in a visualization center, which brings together experts from related fields in a common environment to jointly view seismic and make decisions.

Unlike regular 3D imaging, immersive 3D-seismic technology allows the images to move as the viewer moves.

"The difference [between the two technologies] is almost like watching a movie of the Grand Canyon and actually going there," explained Bartling.

One Latin American client using an immersive 3D visualization center was able to streamline its workflow processes from months to weeks, drill 14 fewer wells in a particular field, and as a result, save approximately $15 million per well or a total of $210 million in a single operation.

In the case of Statoil, the visualization center helped the company shave off drilling costs of about one day of rig time per well, the equivalent of about $20 million to $40 million.

By optimizing well placement in the field, the company was able to benefit from about $375 million worth of produced oil.

Advances in support technologies. A combination of advances in support technologies has made possible the superior reservoir visualization technology available in the market today.

Also, the availability of some support technologies has helped lower total cost of ownership of fancier visualization technology and therefore, hastened adoption among operator companies.

Support technologies include:

• Improved downhole sensor technologies. Fiber-optic sensor arrays are sufficiently small and narrow so that they fit into narrow boreholes and capture seismic information within the well, which can be sent back to visualization centers for interpretation. 
  Rugged coatings for fiber optics developed by a materials research program at the University of Newcastle, England, enable these fibers, which have seismic sensors attached to them, to withstand high borehole temperatures. 
• Miniaturized communication devices. Tiny wireless communication devices that can be fitted in narrow well tubing facilitate communication of frequent readings taken by measurement devices to land-based, centralized monitoring and decision-making centers.
• More powerful and cheaper supercomputers. Supercomputers are getting cheaper and doubling in capacity every 18 months, thus enabling more powerful reservoir characterization software applications and modeling tools.
• Desktop geological and geophysical (G&G) applications using Linux.  G&G software applications that traditionally ran on proprietary operating systems and therefore, expensive hardware and supercomputers are being re-coded in Linux so that they can be used on desktops. 
  Linux is a Unix-like operating system that's more or less open source and freely distributable and runs on many hardware platforms such as PCs and Macintoshes.
  Schlumberger released a Linux-based beta version of Geoframe, a primary component of its GeoQuest software suite in May 2003.
  The original Geoframe software works on the much more expensive to run Solaris operating system developed by Sun Microsystems.
• Cheaper and more widely available broadband and network bandwidth. High-speed broadband is becoming increasingly cheaper and commonplace allowing larger amounts of data transfer to and from remote locations to take place.
• Distributed computing, storage and graphics architecture. Broadband capacity is not expected to keep pace with the need to process larger and larger datasets within the E&P industry. 
  

To overcome broadband bandwidth limitations and constraints in storage and computing capacity, companies can use distributed architecture infrastructure, which still provide consolidated image views and management capability.

Distributed architecture infrastructure such as grid computing, storage area networks and visual area networks linked together with various connectivity technologies help E&P companies, which often operate in geographically dispersed locations, enjoy substantial computing and storage capacity and short processing times for image transfer by reducing data movement.

At the same time, these companies are able to avoid multiple, exorbitant IT infrastructure investments for every location.

Scalable options for imaging. In the near future, E&P companies will have a variety of scalable imaging technology options available to them, which they could choose depending on their functionality and cost-benefit requirements.

At present, many large E&P companies and oilfield service companies have built visualization centers or theater rooms that use 4D and 3D seismic and immersive-imaging technologies.

Oilfield service companies have built similar centers that are also leased to operators who don't own such facilities themselves.

Companies such as SGI, Schlumberger and Halliburton are creating portable versions of these visualization centers that are more cost-effective for E&P companies and can be easily and quickly transferred to different locations.

During its presentation in Houston, SGI demonstrated what it called the SGI Mobile Innovation Center, a portable visualization center with large-screen imaging panels, high-performance computing clusters, and broadband connectivity all set up in a large truck, which could be moved from field to field.

In addition, small E&P companies will soon be able to use 3D seismic visualization capabilities on their desktops on a time-fee basis through application service providers (ASPs).

"The E&P company will simply log on to the application remotely, use the needed computational and technical applications for a period of time, pay a fee and then log off," said Bartling.

ASP imaging and modeling applications developed by the larger oilfield service providers may be available as early as the end of 2004, according to Helen O'Conner, director, real-time systems, Landmark Graphics.