|
Pore to core subsurface flow analysis
Zeiss enables non-destructive virtual visibility into core porosity and permeability with multiscale in situ digital core analysis Steve Sutherlin Petroleum News
Carl Zeiss Microscopy LLC can offer porosity data using micro computerized tomography or X-ray microscopy, K.D. Derr, of Zeiss said while presenting at a May 31 technical breakout session at the state Geologic Materials Center in Anchorage. The session focused on the potential for new investigative technologies and machine learning systems to better assist geoscientists and resource companies to meet the challenges of interpreting Alaska geology.
“We make a distinction between micro CT and X-ray microscopy because with micro CT you’ve got a fixed source and a fixed detector, and your resolution depends upon how close you can get your samples to the source,” Derr said. “With an X-ray microscope you have the same type set up, but you have a series of objectives that you have to use (for interior modeling).”
Derr said despite many advances in the science, significant challenges remain, particularly in the areas of process and scale.
“Fluid flow in porous media is dominated by processes occurring at the scale of the microscopic torturous pathways through which the fluid flows and in which the fluids are hosted,” she said. “The last 20 years have seen a step-change in our ability to characterize and examine such flow at the scale at which these physical processes occur, with pore scale imaging and modeling being transformed from a primarily academic pursuit used to examine fundamental processes associated with transport and displacement, to a fully fledged industrial service industry routinely used by the oil industry to predict flow and transport properties of subsurface samples.”
Machine learning has enhanced the ability of computers to predict porosity and subsurface flow.
“You start out with a reservoir model, you do the wireline logging, you pull your core samples, then you do a bunch of measurements and you use that measurement to make a dynamic reservoir model,” she said. “You predict what the production is going to look like but it’s not perfect.
“So, a series of history matching has been created to take the actual production information and to feed that back into the dynamic reservoir model,” she said. “It’s very complex why the predictions aren’t spot on, that’s one of the reasons we think we could get those predictions much more accurate if we had more porosity data.”
Analysis The digital information is analyzed in two different ways, Derr said.
“One is multiscale simulation where we can take that data statically and run digital rock physics on it to get a sense of what the porosity data is,” she said. “The second example is ... we can actually use the data from our X-ray microscope to inform areas where we will pull samples to do in situ experiments, and this is that where we’re actually able to dynamically image rock under reservoir conditions using the X-ray microscope.”
The resulting porosity data varies in an XY and Z direction, which allows the entire sample to be seen non-destructively in three dimensions, rather than just a surface representation.
“We take core samples from different regions of interest and we look at the porosity within that sample; we were actually able to use the objectives within the X-ray microscope to sub sample within that sample,” Derr said. “We can look at those individual areas and look at what the porosity lights and start to model what the flow characteristics are, and we can take that information and plug it back into the core samples and use that to upscale to the entire core.
“This is our vision, and we partner with a lot of other companies to make that happen,” she said. “Zeiss is an optics company, so we’re very clear that our focus is making people able to see what is in their samples, and we play well with all of the companies because our expertise is not data analysis.
“You can plug that data back into the model,” she said. “It’s an iterative process; we can continue to model your samples and take plugs wherever you see a difference in the structure.”
Virtual simulation Using multiple data sources and algorithms, it is possible to create a visual simulation of the interior of the core. It’s like being able to take a virtual thin section without touching the core.
“You can use this information to run digital rock physics ... virtually slicing through the samples,” Derr said.
“One thing that’s really unique about our X-ray microscope is that our scintillators (responsible for converting electrons collected from the sample into photons) are paired to the X-ray energy ranges that are used so we have very high contrast definition between our samples,” she said. “Because we’re able to move our source and our detector away from the samples we’re able to not only have absorption contrast honor images but we also have propagation phase contrast to give you an enhanced (view) so you’re able to separate the porosity much more readily.
“You can look at one of the high porosity sections where the permeability in X and Z is very different than it is in Y,” she said. “This is really important information when you’re trying to understand what’s going on with in your rocks.”
The information can be plugged back into multi-scale data sets.
The samples can be viewed not only in 3D, but in what the company calls 4D imaging.
“This is an in-situ measurement where we take rocks and put them back into reservoir conditions,” Derr said.
The samples are vacuum saturated with brine before the injection of oil (drainage) and chase brine (imbibition) to establish a residual state.
“We can start to understand how the porosity, permeability and the wetability in those rocks work.”
| 