Zachary Diermyer

Figure 1: Illustration of a production-level shale analysis workflow from left to right showing micrometer-resolution digital rock imaging to GPU accelerated mDPD simulations of fluid flow in realistic nanopores (Figure 2 of Xia et al. [6], reprinted with permission). (a) An example shale rock sample. Image colors indicate organic and inorganic solids, as well as pore space. (b) Digitized representation of shale pore space. Potential regions of interest (ROI) to be used for computational modeling are highlighted with various colors. (c) An ROI is chosen and a subsection of the region is used in the flow simulation. The orange cube shows the resulting simulation domain size. (d) A pressure gradient is applied to the chosen pore section to study fluid transport through the pore space. In this simulation, water is used to flush hydrocarbon liquids out of the pore space, modeling the physical petroleum recovery process. Fluid transport in physical nanoporous material can be predicted by computational models using this method. Deviations from continuum based theoretical predictions can be quantified by comparing them to the recreated, heterogeneous pore model seen on the right.

The dynamics of fluids confined within nanometer-sized structures (nanoconfined fluids) deviate significantly from the predictions of classical fluid mechanics. Nanofluidics is a rapidly expanding field of research with the goal of understanding and controlling those dynamics. Nanofluidics has gained popularity because of the growing influence of nanotechnology. However, the physical behavior of nanoconfined fluids is still debated. If these behaviors become well understood, it would lead to major advances in science and engineering, such as improvements in petroleum recovery and permanent geological CO2 storage. Larger petroleum recovery rates could further lead to increased energy efficiency and less pollution during the recovery process.

Subsurface shale reservoirs containing significant amounts of petroleum (Marcellus, Permian, etc.) are composed of nanometer-sized pores, making them a prime target of nanofluidics research. This is a considerably difficult challenge because of the heterogeneous pore structure in reservoir rocks like shale. It is difficult, if not impossible, to identify rock samples with low variability in pore physical and chemical properties for experimentally characterizing the flow and transport of fluids within nanoporous rock [2]. Because the flow phenomena are influenced by the highly coupled factors of pore size, pore geometry, and fluid–solid interaction at the nanoscale, they remain far from being well understood [3]. If the degree to which each of these factors individually influences fluid transport within natural porous media can be understood, it will enhance our understanding of flow phenomena in the natural environment and could lead to increased recovery during the petroleum extraction process.

To decouple the many factors that complicate nanofluidics research of porous rock, researchers have studied well-defined channel flow systems, where flow can be analyzed by laboratory experimentation [5-6] and numerical simulations [1]. For example, one focus of nanofluidics research has been nanotubes because of their inherently simple geometry. These studies have found differing results based on the chemical composition of the nanotube (e.g., carbon- or silicon-based tubes). This is an active area of study, and no clear consensus has been reached on how structure and chemical properties affect nanoconfined flows.

To understand the controversies and resolve variations reported in the literature, different types of mesoporous materials and nanochannel devices have been fabricated by the Multi-scale Fluid-Solid Interactions in Architected and Natural Materials (MUSE) Energy Frontier Research Center (EFRC). For example, flexible high-pressure flow devices have been built to accurately measure the relationships between volumetric flow rate and pressure drop for different fluids flowing through various nanoporous media.

Additionally, ongoing efforts within MUSE have been devoted to computational models that can reveal the underlying mechanisms and provide fundamental insights into the observed phenomena in experiments. A GPU-accelerated package based on a many-body dissipative particle dynamics (mDPD) model has been developed to simulate the flow of fluids confined in nanopores [5]. The mDPD model manifests a unique, particle-based, multiscale (from nano- to micrometer-scale) capability that can model fluid–fluid and fluid–solid interfaces in pores from sub-continuum and continuum scales. This has resulted in an advanced combination of experimental and modeling capabilities to understand the behaviors of flow under confinement (Figure 1). Simulations of nanoporous flow were conducted on the Summit supercomputer at Oak Ridge National Laboratory and used over one thousand GPUs simultaneously. These simulations would not be possible without the Department Of Energy’s top-notch supercomputing resources.

In the future, a variety of model systems will be experimentally investigated to determine the applicability of the computational findings. Molecular dynamics simulations will be used to calculate the fundamental dimensions of slip length or surface adhesion as a function of rock–fluid wettability. Pore scale simulations (dissipative particle dynamics) will be used to understand flow paths and tortuosity, which will help to delineate the effects of pore size and pore geometry in multiphase flow.

More Information

[1] Rao, Q., Xia, Y., Li, J., Deo, M., and Li, Z., Flow reduction of hydrocarbon liquid in silica nanochannel: Insight from many-body dissipative particle dynamics simulations, Journal of Molecular Liquids, 2021. 344(117673). DOI: 10.1016/j.molliq.2021.117673

[2] Goral, J., Panja, P., Deo, M., Andrew, M., Linden, S., Schwarz, J.O., and Wiegmann, A., Confinement Effect on Porosity and Permeability of Shales, Scientific Reports, 2020. 10. DOI: 10.1038/s41598-019-56885-y

[3] Goral, J., Walton, I., Andrew, M., and Deo, M., Pore system characterization of organic-rich shales using nanoscale resolution 3D imaging, Fuel, 2019. DOI: 10.1016/j.fuel.2019.116049 

[4] Xia, Y., Blumers, A., Li, Z., Luo, L., Tang, Y.H., Kane, J., Goral, J., Huang, H., Deo, M., and Andrew, M., A GPU-accelerated package for simulation of flow in nanoporous source rocks with many-body dissipative particle dynamics, Computer Physics Communications, 2020. 247(106874). DOI: 10.1016/j.cpc.2019.106874

Other Information

[5] Majumder, M., Chopra, N., Andrews, R., and Hinds, B.J., Enhanced flow in carbon nanotubes. Nature, 2005. 438(7064): p. 44-44. DOI: 10.1038/43844a

[6] Mattia, D., Leese, H., and Lee, K.P., Carbon nanotube membranes: From flow enhancement to permeability. Journal of Membrane Science, 2015. 475: p. 266-272. DOI: 10.1007/s10404-008-0293-5


The research is supported by EFRC-MUSE, an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Basic Energy Sciences under Award No. DE- SC0019285.

About the author(s):

Zachary Diermyer is a Mechanical Engineering PhD student in the Multiscale/Multiphysics Theory and Simulation (MuLTS) group at the University at Buffalo in Buffalo, New York. He was given the opportunity to work as an intern at the Idaho National Laboratory (INL) from January until July of 2022. With the guidance of his mentor, Yidong Xia, he was able to create his own two-fluid mDPD simulations and has recently submitted his work for publication. While performing this work, Zach was encouraged to present his research to the MUSE PIs, since the INL project was part of this EFRC. Since then, Zach has been an active MUSE member and he hopes to continue his work with the group. Zach’s interests are physics, fluid mechanics, numerical/statistical methods and simulation techniques, machine learning, and high-performance computing. ORCID ID # 0000-0002-0599-9381.

Newsletter Articles

Research Highlights