Lawrence Berkeley National Laboratory

Fluorinated gases are critical to a number of high-technology industries including semiconductor manufacturing and related processes such as flat panel display, LED, and photovoltaic manufacturing. These fluorinated gases have very high global warming potentials (GWPs) and, unfortunately are chemically very stable, in some cases lasting for thousands of years in the atmosphere. For environmental reasons, it is imperative that the release of these gases be minimized, but existing methods to destroy them requires a significant amount of energy. The goal of this project was to develop a more energy efficient method of destroying these high GWP gases using computational fluid dynamics, bringing together resources available at the national labs, including high-performance computing hardware and modern computational methods, in order to explore an industrially important problem with key environmental and economic impact. This exploratory project has demonstrated that HPC can indeed provide a unique view into a complex engineering problem - one that is simply inaccessible by any other means. The tools allow one to explore the details of the chemical process, and construct ``what-if'' scenarios to address shortcomings of the physical device in its current operating modes. However, as with many sources of completely new information, many new questions are illuminated with this capability. It will take considerable time, effort and experience before we learn to probe the solutions, pose operating scenarios and confirm modifications to the device that lead to real engineering advances.