Modeling Reactive Flows in Porous Media
Code: PFLOTRAN
Allocation: 10 million hours
Geoscience researchers are actively participating in efforts to reduce emissions of greenhouse gases and maintain a safe and sustainable environment. A team led by Peter Lightner of Los Alamos National Laboratory is pursuing two such efforts—to reduce atmospheric levels of the greenhouse gas carbon dioxide and to better explain the behavior of contaminants in groundwater—using an allocation of 10 million processor hours on ORNL’s petascale Jaguar supercomputer.
The ability to predict the transport of wastes and contaminants in complex underground mineral environments is one of the most important and daunting challenges facing the geosciences. Subsurface flow systems have coupled processes, the interplay of fluid flow, transport of substances, heat transfer, and chemical interactions between rocks and fluids. These systems also feature interactions on a vast range of space and timescales not compatible with space and time resolution in earlier models. Recent advances in pore scale modeling suggest that the best mathematical approach for following the behavior of hazardous materials as they interact with the mineral bed (a phenomenon called reactive transport) is to treat the system as a spatially continuous, multiple component fluid, in which the different substances present can react.
To model this activity, Lichtner developed PFLOTRAN (pore reactive flow and transport code), a program which models the flow of substances, their interactions with one another, and their reactivity with the mineral bed at the pore scale. In the current petascale projects, Lichtner’s team will model uranium migration at the Hanford 300 Area in Washington state and will perform two 1,000-year simulations of the carbon sequestered from 400 coal-burning power plants within the Illinois Basin.
