With several colleagues and students, civil engineering associate professor David Mays is pioneering a new approach to clean up contaminated groundwater. According to the National Ground Water Association, groundwater—the water occupying the space between soil grains and fractured rocks in the Earth’s crust—provides drinking water to 44% of Americans plus more than 50 billion gallons per day for agricultural irrigation. When groundwater becomes contaminated, however, cleanup is no easy task.
“Groundwater remediation is a challenge for several reasons,” Mays notes, “For one thing, it is hard to manage subsurface resources, simply because they are out of sight. We share this challenge with geotechnical and petroleum engineers. Second, groundwater remediation works through a complex system of linked hydrological, microbiological, and geochemical processes that we call hydrobiogeochemistry. And third, because groundwater moves slowly, there is essentially no turbulence, which is really frustrating for anyone wanting to mix treatment chemicals into subsurface contaminants. So the cleanup problem is important, invisible, complex, and slow.”
Over the last several decades, researchers from the Environmental and Molecular Sciences Laboratory at Pacific Northwest National Laboratory have developed cutting-edge, sophisticated computer model simulations to understand the hydrobiogeochemistry of groundwater remediation.
Mays explains, “These models account for groundwater flow, geochemical reactions, and microbiological processes, which boils down to solving staggeringly large systems of equations on their Cascade supercomputer. And then, what is really impressive, the team from PNNL can validate the simulations with gene expression data taken from a field site. It’s great stuff.”
While PNNL has been working to address the complexity of groundwater remediation, Mays and colleagues have been working to improve mixing in groundwater aquifers by applying new ideas from chaos theory. According to the fluid mechanics research literature, chaotic advection—where flows have sensitive dependence on initial conditions—provide the best possible mixing in the absence of turbulence. “It sounds like rocket science,” Mays comments, “but actually chaos theory can be quite simple. For us, it boils down to stretching and folding the plume of injected treatment chemical, kind of like a saltwater taffy machine.” This work has been supported by NSF grants awarded in 2011 and 2014, and is illustrated in a short animation.
The goal now is to incorporate chaotic advection into PNNL’s existing computer simulation of hydrobiogeochemistry. Mays explains, “Fortunately, this can be done by a fairly straightforward modification of the hydraulic boundary conditions that does not require changing the overall model architecture. And this has been fun. When I started at CU Denver in 2005, I never imagined that I would ever be doing research with a supercomputer.” Work is in progress, but preliminary results have been presented at the American Geophysical Union’s Fall Meeting in New Orleans, Louisiana in December 2017, and most recently at the Hydrologic Sciences and Water Resources Engineering Seminar at CU Boulder in January 2018.