ScienceWise - May/Jun 2008

Getting oil from a stone

Article Illustration
An X-ray slice through a rock sample encased for flooding and subsequent measurement.
Article Illustration
A model of a block of sandstone was used to predict the elasticity of the material. Predictions closely match actual measurements.
Article Illustration
Dr Christoph Arns examines a visualisation of fluid flow through Castlegate sandstone based on a solution of the Navier-Stokes equation using lattice Boltzmann techniques. The visualization was provided by Ajay Limaye (ANU Vislab), using his software package Drishti.

How a super high resolution CAT scanner can improve models of flow in porous rocks

How does oil, water, gas, or nuclear waste flow through porous rocks? It’s far from merely being an academic question. In the petroleum industry, for example, billions of dollars are spent each year making laboratory measurements on rock cores in an effort to understand how oil is stored and might be retrieved. And the manner in which contaminants move through earth materials is critical to the design of nuclear waste facilities that will need to be functional over many lifetimes.

So how to do you understand a complex disordered materials like porous rock? You begin by taking what you think is a representative sample from a rock core and then you measure it in as many different ways as you can. You pump a fluid through the sample to measure its resistance to flow. You apply an electric field across your sample to measure its electrical conductivity. You push against it with a microscopic force to calculate its elasticity. You collect data on its pore space using nuclear magnetic resonance (NMR) spectroscopy. And you image the sample using X-rays to create a representation of its microstructure. Every experimental measurement gives you an additional piece of information on the form and function of the material but the challenge is then integrating what you know and scaling the results up to create useful knowledge applicable in the real world. “Not many people work in the area of drawing the various threads together,” says Dr Christoph Arns. “Most researchers have expertise surrounding one form of analysis which provides useful insights in one area. Often what they discover is very powerful in describing a material at one scale but may have limited value when applied at a larger scale.

“My area of research is based on integrating information from a variety of different characterisations by making numerical measurements. And I’m in a very lucky position because I work with an outstanding NMR researcher and have excellent co-researchers working with our micro X-ray CT facility. These two characterisation technologies are some of the most effective ways of understanding complex materials.”

Dr Christoph Arns is a Senior Research Fellow in the Department of Applied Maths (RSPhysSE). He’s a physicist turned petroleum engineer turned computational physicist. “Petroleum engineering is all about developing ways to improve the extraction of hydrocarbons from oil-bearing rocks,” explains Dr Arns. “That almost inevitably leads you towards ways of better understanding complex materials because this is what oil bearing rocks are.”

Dr Arns has been working for many years on the computational modelling of physical properties of complex materials. He has developed an innovative suite of software that can derive a range of important morphological and physical properties directly from 3D digitised tomographic images created with an advanced micro X-ray computed tomography facility. These include conductivity, diffusivity, fluid permeability, dispersion of a neutral tracer, elastic properties and NMR relaxation/diffusion response. “The predictions we have made from our computer simulations are in good to excellent agreement with independent laboratory measurements across a wide range of pore volume fractions,” says Dr Arns. “This includes agreement in permeability measurements across four orders of magnitude in permeability.”

“In general, we have demonstrated that we can accurately predict properties of complex materials by combining micro-tomographic images with numerical calculations. We’ve recently applied this methodology to a range of petroleum reservoir rocks with excellent results and used the detailed pore-scale information available from the images for an analysis of pore-pore diffusion coupling, and tracer dispersion and local flux. This is significant, as it allowed us to find morphological parameters on the basis of X-ray-CT images, which correlate well with macroscopic physical properties.

“Existing methodologies for estimating the macro properties of disordered materials are limited to overly simplistic representations of microstructure,” observes Dr Arns. “For example, hydrocarbon recovery from reservoir rocks and contaminant dispersion in soils are currently estimated on the basis of laboratory-scale measurements and applied to the field scale. However, large anomalies are observed between what’s predicted and what’s actually measured. This limits our capacity to make accurate predictions of macroscopic properties from morphological information.

“A detailed understanding of correlations of physical properties and morphological descriptors under static structure will also help us, if we consider a dynamical change of the structure itself. “We expect that numerical modelling of reactive flow on the pore scale should lead to improved estimates of permeability changes caused by a range of subsurface processes. This will significantly lower the risk of enormously costly mistakes in the development of oil fields.

“Of similar significance is the potential impact of pore-scale understanding of contaminant migration and reactive flow for groundwater remediation strategies. We’ll be able to generate more accurate numerical modelling where field data cannot be obtained, for example in modelling the risk of storing hazardous materials. Therefore our research has both a significant economic and environmental dimension to it.”

The research being undertaken by Dr Arns is strongly interdisciplinary in its nature, and depends on inputs and interactions with a broad range of scientists.

“NMR spectroscopy is one very important technique for understanding pore volumes and narrow constrictions in porous materials including length scales not accessible by Xray-CT imaging,” explains Dr Arns. “Professor Paul Callaghan at the Victoria University of Wellington, New Zealand, is a world leader in this discipline, and he has been a long term collaborator, providing quality NMR data for this work.

“The other main technology that is contributing to this research is the micro X-ray CT facility. Running a facility such as this requires a range of skills, and I rely on my colleagues to supply me with segmented images and their topological partitions of morphologically interesting samples. Adrian Sheppard, Arther Sakellariou, Tim Senden, Robert Sok and Mark Knackstedt are all involved in this.

"Applied Maths is a great place for this type of research," says Dr Arns. "It's a supportive and friendly place, open to new ideas and always ready to interact with others when it means adding value to the science."

How a super high resolution CAT scanner can improve models of flow in porous rocks
In search of a new drug for antibiotic resistant bacteria
Creating the world’s most powerful telescope
Big bold gestures
Royal jelly is yielding clues to human disorders
Possibly Related ANU Research Articles
How a super high resolution CAT scanner can improve models of flow in porous rocks
Successfully transferring technology from the lab to industry
Analysing rock cores takes more than fancy X-ray equipment
Fiddly little details
How Mathematical Modellers are Helping Improve the World’s Water Distribution
Applications from artificial bone to modeling of oil bearing rocks

Updated:  11 December 2013/ Responsible Officer:  Director, RSPE/ Page Contact:  Physics Webmaster