When Repsol-YPF launched its Kaleidoscope project four years ago, its purpose was to study new hardware alternatives to realize the promise of reverse time migration (RTM) for oil and gas exploration.

What the company perhaps didn't envision was the other potential uses of the computer. Repsol is now partnering with the US government to develop a better understanding of how earthquakes occur. The project, being conducted offshore Costa Rica, is intended to help scientists visualize and monitor the folds in the tectonic plates in the region.

"We're really trying to understand zones around the world that produce large earthquakes," said Nathan L. Bangs, senior research scientist, Institute for Geophysics at the University of Texas-Austin. "This is a scientific pursuit to understand the process of subduction zones, which are the settings that produce large earthquakes."

In addition to the University of Texas, the University of California-Santa Cruz and the Spanish Institution for Research are also involved in the project.

Costa Rica is a natural choice for the study. Bangs said the area regularly produces fairly large earthquakes in the range of 6.5 to 7.2 magnitude on the Richter scale. Already it has been the subject of much study.

"There have also been some 2-D profiles," he said. But we really need 3-D seismic to look at some of the structural details and to map out the plate interface zone that slips during the earthquake."

Because of the nature of the area,it is possible for the team to study the subduction zone at relatively shallow depths. Bangs said the Cocos plate, an oceanic crust, is subducting and extending beneath the Caribbean plate. A large 3-D survey was collected earlier this year, and Repsol is processing that data, first in a conventional time domain and then in depth. The depth imaging will be conducted using the RTM algorithm Repsol developed for the Kaleidoscope project.

"Usually people think that RTM is only good for sub-salt," said Francisco Ortigosa, geophysics director for Repsol-YPF. "In other areas of the world where we explore, we've found that RTM is very suitable for this kind of environment where you have thrusting over thrusting and complex structures.

"We think we have technology that is very suitable for this environment."

Time processing should be complete by November, with a final depth migration ready in May 2012. "It's going to be a challenging project, but we are pretty sure we can get a very high-fidelity image of the subsurface in this part of the world," Ortigosa said.

For Bangs and his colleagues, this will be a tremendous help. "These are very complicated structural settings," he said. "This is a subduction zone that has a collision between two plates, and the collision process produces a tremendous amount of deformation of the overlying plates."

He added that the goals are to see the plate interface more clearly and also to study the effect of the tectonic movement on fluids. Fluids are squeezed out of the rocks, and the pressures they produce are important for controlling the strength of the fault zone, he said.

For Ortigosa, the partnership is a good example of cooperation between industry and academia. "We are concentrated on our business, of course," he said. "But it's a nice opportunity to collaborate with academia in a mutually beneficial project for both parties. It's part of Rep-sol's commitment to be socially responsible."