Concept visualization of the phenomenon of viscous fingering: displacement of a more viscous fluid (ochre) by a less viscous fluid (blue).
For the team at the USC Geosystems Engineering and Multiphysics Laboratory (GEM Lab), understanding how fluids move through complex geological systems is the key to solving critical energy and environmental challenges.
Working within USC Viterbi’s Mork Family Department of Chemical Engineering & Materials Science, the lab studies the processes that govern subsurface systems, where fluid flow, transport, rock properties, geomechanics and chemical reactions interact across multiple scales. The ability to predict and control these interactions determines the effectiveness of current technologies for groundwater remediation, geothermal energy production, geologic carbon sequestration, underground hydrogen storage, and a wide range of chemical and pharmaceutical processes.
In a numerical study published in Proceedings of the Royal Society A, Birendra Jha, director of GEM Lab and associate professor of chemical engineering and materials science, and doctoral researcher Faeze Ghazvini, report a new way to influence how fluids spread and mix through porous rock.
Using high-resolution three-dimensional simulations, they demonstrate that fracture orientation can be used to control the development of viscous fingering, a hydrodynamic instability that has long limited the efficiency of fluid transport in porous materials.
Viscosity refers to a substance’s resistance to motion under an applied force. Viscous fingering – also known as the Saffman–Taylor instability – occurs when a less viscous fluid pushes a more viscous one; Instead of advancing as a relatively uniform front, the less viscous fluid breaks into branching “fingers” that move ahead of the surrounding flow. These instabilities influence how fluids spread, mix and react in porous materials.
Using one challenge to solve another
“Understanding fractured rock and the impact on the flow of fluids has been recognized as an important problem for a long time,” said Jha.
This study marks a turning point, bringing together two longstanding challenges in subsurface engineering. The first is that fractures create highly conductive pathways that can redirect fluid flow and make subsurface systems difficult to predict. The second challenge is viscous fingering. As the instability develops, some fingers grow faster than others. In many cases, one finger becomes dominant, forming a narrow channel that captures much of the flow and bypasses large portions of the surrounding rock. This channeling reduces mixing and limits the effectiveness of fluid displacement.
The researchers found that fractures can alter this behavior. By varying the angle of a highly conductive fracture relative to the dominant flow direction, they showed that fractures aligned more closely with the flow can draw multiple fingers into the fracture, where mixing increases. Instead of reinforcing channel formation, the fracture disrupts it. Downstream, the fluid emerges as a broader, better-mixed plume rather than a concentrated channel.
“We addressed one challenge with the other challenge,” Jha said. “What we found is we can use the presence of fractures in the rock to actually suppress some of the negative impacts of viscous fingering.”
Models and measurement
The study also highlights the importance of studying these systems in three dimensions. Many previous investigations have relied on two-dimensional models because fully three-dimensional simulations are computationally expensive. GEM Lab found that simplified models can miss important features of instability-driven flow.
Dominant channels often developed outside the plane captured by two-dimensional models, affecting predictions of breakthrough behavior, fluid recovery and mixing efficiency. These differences become increasingly important in fractured systems, where flow paths can evolve in complex ways.
Beyond demonstrating a new way to control viscous fingering, the researchers also developed a new way to measure it. The team introduced a statistical framework based on the evolving probability distribution of fluid–fluid interface length. The method provides a quantitative indicator of when a system transitions from distributed fingering to dominant channel formation. The framework offers researchers a new way to compare flow conditions, evaluate control strategies and study instability-driven transport in both natural and engineered systems.
Extension and interconnection
For Jha, the study represents the beginning of a broader research direction. “We haven’t seen anyone recognizing this interaction between fractures and fluid patterns,” he said. “We believe that this research opens a door where people can explore this further.”
The current work examines a single fracture, but natural subsurface environments often contain extensive fracture networks. Jha’s group is now testing these findings experimentally and exploring how the resulting datasets can be used to train artificial intelligence models capable of predicting unstable flow more efficiently. “The combination of the experiments and the model can be used to scale up some of these conclusions to field scale,” said Jha.
The study not only represents a novel method for greater control over fluid transport in porous materials. It also indicates a way of thinking about engineering problems at large, recognizing an initial challenge as integral to an effective solution. In this case, complex geological structure is not simply a source of uncertainty that engineers must account for; it also provides a clue for how to use that structure to navigate that uncertainty.
Published on June 22nd, 2026
Last updated on June 22nd, 2026