What can you do with several million processing hours on a supercomputer? Oh, just change the course of energy storage and vehicle development. That’s the hope, at least, behind a new project awarded 24 million supercomputing hours this week by the Department of Energy.
Working with researchers from several national labs, as well as Vanderbilt University and IBM, Jack Wells of Oak Ridge National Lab will lead an effort to prove the practicality of a rechargeable lithium-air battery that could store up to ten times the energy of a lithium-ion battery (the chemistry of choice for most electric cars now in the works from major automakers) of the same weight and power an electric car for up to 500 miles. The millions of hours of processing power on the supercomputers could accelerate research, bumping it up to “weeks and months” rather than the “years and decades.”
I spoke with Winfried Wilcke, Senior Manager of Nanoscale Science & Technology, and Program Director of Silicon Valley Projects for IBM’s Almaden Research Center (he’s also one of the co-investigators on Wells’ team), about IBM’s plans for rechargeable lithium-air battery tech last summer — not long after the company launched an ambitious project to commercialize the batteries.
Wilcke explained that the technology plugs into areas of IBM’s expertise. First, nanostructures, which are seen as one of the keys for a safe lithium air battery. They’re needed to exclude moisture from the lithium metal electrode (an explosive mix) while letting oxygen in. The second area Wilcke described as a strong fit for IBM in advancing this technology is supercomputing, which can help model things like how potential catalysts affect the reactions and performance of the battery.
IBM and the national labs’ award for this lithium-air project comes as part of 1.6 billion supercomputing processor hours handed out Tuesday for 69 projects through what’s called the Innovative and Novel Computational Impact on Theory and Experiment program (fortunately, it has a catchy acronym: INCITE).
This is the seventh annual round of awards under the program. Other research areas awarded supercomputing time this year (mostly through national labs and universities) include nano-scale solar cells, cellulosic ethanol and next-gen internal combustion engines for alternative fuel vehicles. General Electric (s GE) will also get 19 million processor hours to simulate turbulence issues affecting the efficiency of advanced energy and propulsion systems, such as wind turbines and hybrid diesel engines.
Looking at the number of processor hours awarded for each of these projects — we’re talking tens of millions — offers a reminder of just how much remains to be explored with these technologies. As the research summary for the lithium-air INCITE project notes, the “exciting proof-of-principle work” that has been completed for lithium-air tech “still presents very big scientific challenges before one can be confident that practical propulsion batteries can be based on the Li/Air system.”
To that end, Wells, Wilcke and their collaborators, plan to focus on reactions occurring at the interface between the electrode and electrolyte, cathode surface properties, the role of catalyst selection and the mechanisms for how lithium-air cells discharge and recharge, among other things. Supercomputers at the Argonne National Lab and Oak Ridge National Lab will be used to simulate reactions and model the performance of different materials for the batteries.
As Stacey has reported over on GigaOM, many of the parts that make up a supercomputer — from processors to networking cables —increasingly are the same as those used in everyday corporate computing. But a few million hours of processing time on these machines can still accelerate research, according to the DOE, which in its press release claims the INCITE program will allow awardees to “conduct cutting-edge research in just weeks or months rather than the years or decades needed previously.”
If IBM’s Wilcke is right that lithium-air represents “the only system that has a chance to be as good as gasoline,” and if Dalhousie University’s Jeff Dahn is right that while rechargeable lithium air isn’t worth betting the farm on, “but it has to be explored” — then supercomputing might be just the thing to take us one step closer to next-next gen batteries. Or provide enough answers to show it won’t be practical, after all.