PGS hyperBeam refers to a combination of the beam migration technology developed in the late 1990s by John Sherwood (formerly Applied Geophysical Services or AGS) and the holoSeis immersive visualization technology developed by PGS. Launched in 2009, PGS hyperBeam brings processing and interpretation together in a near real-time interactive system, reducing the cycle time for velocity model building from months to minutes. The implications for depth imaging of seismic data are significant.

A traditional prestack depth migration project will use any one of countless migration algorithms, with the main criteria being dictated by the available time and project budget. Increasingly complex geological regimes require increasingly sophisticated, and thus increasingly expensive, depth migration solutions. An unfortunate paradox with depth imaging algorithms, notably “wave equation” algorithms, is that the input velocity model of the subsurface must be exceedingly accurate (preferably within 1% or thereabouts of the true model), but the seismic data is typically very poor, and so the velocity model building process is painstakingly arduous and difficult. It is not uncommon for subsalt depth migration projects in the Gulf of Mexico (GoM) to take more than six months to complete. Wide-azimuth seismic acquisition is becoming the de facto solution in the GoM, necessitating wave equation depth migration processing, and these projects typically take more than one year to complete.

Targets were believed to be flanks of “shale diapir,” but the beam image revealed eroded recumbent fold. The crest of the “new” structure was drilled as a successful discovery. (Images courtesy of PGS)

Originally conceived as a vehicle to enable the small AGS to compete with better-equipped major service companies in the lucrative GoM depth imaging market, John Sherwood’s AGS beam migration implementation is not only dramatically faster than other depth imaging solutions, but the unique two-step design of the AGS beam algorithm has actually yielded results in many locations that vastly surpass alternative solutions. The key is the “dipscan” process that scans a vast array of 3-D kinematic and dynamic data attributes, and then only uses those data components that will usefully contribute to the final seismic image. Multiple and other noise removal is included — no assumptions are made about the acquisition geometry and sampling, dips in excess of 90º are handled, and residual moveout corrections are inherent in the workflow.

The engineering development between the beam and holoSeis technologies has yielded PGS hyperBeam, a platform that enables near-real-time velocity model building (VMB), and near-real-time migration and screening of multiple velocity model scenarios to be tested. A small PGS hyperBeam machine with only 30 PC compute nodes can turn around 115 sq miles (300 sq km) in less than four minutes. The tomographic routine used for VMB also runs on the same server hardware, and everything is scalable. The holoSeis visualization platform enables true integration of all tools into one environment, sitting on any user’s desktop. There is no restriction on data size, and several vast and independent 3-D volumes and attributes can be manipulated and rendered in real time. Thus, any interpreter can test literally hundreds of depth imaging scenarios in the time it would historically take a conventional depth imaging team to deliver only one depth imaging scenario.

An obvious implication is the reduction in drilling risk and the fast ranking and analysis of drilling prospects. The flexibility offered in VMB with PGS hyperBeam is critical when dealing with complex areas such as the GoM. The beam migration algorithm handles multiple arrivals, unlike Kirchhoff migration solutions. Combined with the powerful dipscan process, beam migration yields the best platform for complex geology imaging. An expected workflow is the combination of PGS hyperBeam with subsequent full wave equation depth imaging, yielding optimum seismic images for appraisal and production in cycle times previously unimaginable.

Many examples now exist of “no seismic” areas yielding quite incredibly detailed structural and stratigraphic images when the AGS beam migration solution is available. Figure 1 is an example of a dramatic success story wherein the geological model was completely revised in only two days, and the revised drilling target at the crest of an eroded recumbent fold (i.e., anticlinal trap) was successfully drilled.

Engineering has provided a new playing field for seismic depth imaging, so the exploration challenge truly passes to the interpreter rather than the processor. It will be intriguing to see how the industry adapts accordingly.