Digital rock physics comes of age
Digital rock physics is redefining how subsurface technologies are implemented by providing a complete picture of what occurs to fluids flowing through rock in real time.
In 2015, Dr Ryan Armstrong, along with his colleague Dr Peyman Mostaghimi, established the Multiscale Transport in Porous Systems (MUTRIS) research group. Armstrong’s particular focus within the group is ‘digital rock physics’, a technology that offers new ways to look at how fluids flow through porous rocks.
According to Armstrong, this technology is an area of increasing interest in industry because it has a wide range of practical applications including flow in groundwater systems, recovery of oil and gas, and carbon sequestration.
“Traditionally these systems have been studied by treating rock as a ‘black box’ by injecting fluids and measuring associated pressures and flow rates,” says Armstrong.
“With digital rock physics we redefine this approach by visualising how fluids flow through rock in real time in addition to studying the chemical, physical and hydrodynamic processes that directly influence the efficacy of subsurface engineering technologies.”
Armstrong explains that fluids flow through constrictions in rock that are only a fraction of the diameter of a human hair and that their research has been able to directly resolve these constrictions and the fluid interfaces that form. This means it provides an unprecedented way to explore the subsurface.
With most of the top oil and gas companies exploring dynamic imaging and digital rock technology through their own research divisions or in-house experts, Armstrong says the technology has now passed the discovery phase and is pushing more into development and deployment.
Petroleum engineers, says Armstrong, are always dealing with a very sparse data set so any intel they can gather is crucial. “Even the most mature reservoirs in the world have an ultimate recovery at around 40% of the actual oil in place, so that means that 60% of the oil is left behind trapped in the rocks,” he says.
With digital rock physics we can test many variables and use data analytics to develop the most effective reservoir management scheme.
Dr Ryan Armstrong, Senior Lecturer, School of Minerals and Energy Resources Engineering
“If we understand the physics better, we can engineer clever ways to recover that oil. One way, for example, might be to inject surfactants and chemicals to help mobilise the oil. With digital rock physics we can test many variables and use data analytics to develop the most effective reservoir management scheme. Even if we can increase oil recovery by 5%, that would be highly significant.”
Although the capabilities of micro-CT imaging and the resolutions required to push this technology forward have been around for about 30 years, it wasn’t until 5-10 years ago that computational processing came up to speed to process dynamic data in a reasonable timeframe.
“During my PhD, some of the numerical simulations took up to a month. Now graphical processing units are able to do these computations in a matter of hours. That was the breakthrough we needed to make this a practical tool,” says Armstrong who has recently returned from the Argonne National Lab in Chicago where he conducted dynamic flow experiments through their multi-billion-dollar synchrotron.
“We use synchrotron X-ray computed microtomography, which is a 3D imaging technology that visualises the internal structure of rocks, to get a really high X-ray flux,” Armstrong continues.
“Essentially, this means we can collect 3D images at micrometer resolution in a matter of seconds and can directly visualise water and oil moving through reservoir rocks in real time.”
The next step in the process is to combine the synchrotron data with high-performance computing, such as Raijin (National Computing Infrastructure in Canberra) and Titan (Oak Ridge National Laboratory in Tennessee), to study the processes that influence subsurface fluid flow.
Although applications are mostly aimed at the oil and gas industry, Armstrong says the technology can be applied to any type of porous media. “The simulation and modelling techniques could be used to understand hydrogen fuel cells and carbon sequestration, for example. It also has applications in mining engineering, membrane science technologies and groundwater hydrology as well.”
The future of this technology is wide open, but one thing Armstrong says will have a big impact is artificial intelligence and machine learning. “There’s a lot of potential, in particular with dynamic imaging, to use machine learning to logically categorise these terabytes of data and use it to make future decisions. This is an exciting development and I think it will accelerate the discoveries and progress that we can make.”