Geoscientists seem to be rediscovering an age-old fascination with rock physics.

Of course it's easy enough to joke that folks like geologists and petrophysicists are so interested in subsurface rocks that they forget that what they're really after is the fluid residing within those rocks.
Nevertheless, rocks do account for a vast percentage of what's under our feet, so it probably makes sense to spend some time learning more about them. And at this year's annual meeting of the Society of Exploration Geophysicists, that focus seemed evident in a number of ways.

First of all, Dr. Gary Mavko from Stanford University gave a talk during a session titled "Geophysics and the Road Ahead." Mavko and his students are responsible for many of the breakthroughs in correlating seismic with rock physics information, and he discussed the discipline's on-again, off-again romance with the study of rock physics.

The underlying benefits are hard to ignore. An understanding of rock physics can answer such basic questions as where the oil is, how much is there and whether or not it will flow. "Rock physics provides a quantitative analysis of the reservoir and helps to manage risk," Mavko said.

But it appears that other, sexier technologies have shifted the focus away from rock physics over the years. While Mavko said that the mention of the subject in publications has grown "exponentially" in recent years, technologies such as amplitude variations with offset (AVO) and time-lapse seismic have eclipsed rock physics, even though both could benefit from a better understanding of the reservoir and source rocks.

The problem with seismic data, Mavko said, is that while seismic waves are sensitive to fluids, the information provided is not unique. For instance, gas, oil and water mixed at a fine scale has a low velocity, while the same makeup of fluids in separate states has a high velocity. Fully saturated rocks demonstrate low attenuation, so do completely dry rocks. And rock physics, he said, is the factor behind compressional waves "getting clobbered" in a gas field.

What would help constrain the seismic information is a better geologic understanding of the reservoir, he said. For instance, hydrocarbons are drastically susceptible to temperature changes in the reservoir. Quantifying such things as temperature, porosity, facies distribution, compaction, etc., would help untangle the seismic signatures.

"We need to recognize that there are powerful ways to figure out how the seismic is likely to change in different depths or depositions," he said.

Imaging permeability

One of these ways is through computational rock physics, which can extract the correct geometry of a rock matrix and compute permeability and rock properties from the pore and grain size of a small sample.
This is something that the graduate students at Stanford have been studying. Stanford, through a cooperative venture with Rock Solid Images, is taking the theory out of the laboratory and into the oil patch through a project intended to develop, test and commercialize technology for permeability estimation in reservoir rock from micro-images of drill cuttings, thin sections and core CT scans. If the 3-D pore space geometry is not directly available from the CT scanning, it is reconstructed from 2-D images such as thin sections and reflected micro-images from flattened rock surfaces.

Rock Solid is seeking partners in this venture. For an investment of US $45,000, each partner receives a "digital rock lab" (DRL) consisting of a microscope, digital camera, laptop or desktop computer, and software for computing porosity and permeability from thin sections and other appropriately prepared specimens. The company hopes to get at least six subscribers into the program.

Phase I will consist of sample collection and preparation, image digitizing and storage, porosity and permeability computation, statistical data analysis and reporting, and distribution of rock lab equipment. Phase II will extend the method to the estimation of capillary pressure, relative permeability and electrical resisitivity, and possibly mechanical or acoustic properties. Ultimate deliverables include software to convert 2-D pore images to 3-D pore volumes, optimized computation workflows for converting images to permeability, software to compute porosity and permeability, complete datasets on more than 100 samples, final reports and recommendations, and a complete DRL with documentation.

Ultimately it's hoped that the DRL could be used to determine permeability while drilling and for logging- and measurement-while-drilling enhancement as well as for reservoir characterization and core analysis.
For more information on the DRL project, visit www.rocksolidimages.com.