Tuesday, February 23, 2010

Traveling Wave Reactor Works Well -- In Models

Recent developments in supercomputing have enabled the TerraPower scientists to simulate the traveling wave concept and establish its feasibility, they say.

Machiels agrees. "The modeling capability that John Gilleland's team has achieved has allowed a lot of progress. They have fantastic computing capabilities," he said. The team's supercomputer cluster has more than 1,000 times the computational strength of a desktop computer, TerraPower says.

The team draws on support from MIT, DOE's Argonne National Laboratory and other scientific centers, and future testing will require more DOE support. But at this point, the project is a private research venture.

It recalls the famous Tuxedo Park laboratory established by the millionaire investor and amateur scientist Alfred Lee Loomis at his mansion outside New York City in 1926. Its scientists went on to provide critical research in the development of radar and the atomic bomb in World War II.

"This is a type of work that requires a deep, deep pocket," said Machiels. "The fact that this is being funded now by a private firm is good, but very unusual." TerraPower is backed by Nathan Myhrvold, Microsoft's former chief technology officer, who now is CEO of Intellectual Ventures. _NYT
Ultra-sophisticated computer modeling supports the theory behind TerraPower's traveling wave reactor approach to use nuclear waste as fuel. Of course, TerraPower's main backer -- Nathan Myhrvold -- was formerly the chief technology officer for Microsoft, so he should know something about computers and computer modeling.  It is good to remember that computer modeling has been an extremely potent tool for many areas of science and engineering.  We should not allow the perversion of modeling by alarmist climatologists to taint the entire enterprise.
This reactor (pdf) works something like a cigarette. A chain reaction is launched in one end of a closed cylinder of spent uranium fuel, creating a slow-moving "deflagration," a wave of nuclear fission reactions that keeps breeding neutrons as it makes way through the container, keeping the self-sustaining reaction going.

And it goes and goes, perhaps for 100 years, said former Bechtel Corp. physicist John Gilleland. He heads TerraPower LLC, a private research team based outside Seattle that is pursuing the traveling wave reactor design.

"We believe we've developed a new type of nuclear reactor that can represent a nearly infinite supply of low-cost energy, carbon-free energy for the world," Gilleland said in a presentation.

...In the traveling wave reactor, the fuel, initially, is likely to be the vast U.S. stores of depleted uranium, which don't themselves pose a proliferation risk. Plutonium is formed in the reaction process but undergoes transmutation into other elements and is essentially consumed. Depleted uranium is a heavy, lead-like residue from making or enriching uranium fuel. Lacking the volatile isotope U-235 that is used in conventional nuclear power plant fuel and nuclear weapons, depleted uranium is currently used for conventional anti-tank ammunition and in the keels of sailboats. _NYT
When computer modeling works, it is because the researchers involved are honest enough to test the models against physical observations -- rather than the other way around, as in the climate science of carbon hysteria.

Labels: ,

2 Comments:

Blogger J. Paige said...

Can u238 be consumed in a LFTR? Does one hundred years of neutron flux makes containing the cigarette problematic ? Can the traveling wave rxr follow electrical loads? Does it fail to safe mode? What degree of containment required? I would be interested in a fair-minded comparison to LFTR .

JPS

5:42 AM  
Blogger al fin said...

The answer to your first question may lie within this repository of documents dealing with LFTRs. But I think that breeders have to be customised for the particular fertile material -- whether U238 or Thorium etc.

2nd question: Probably not.

3rd question: Probably not.

4th question: Yes, if the model fails, the computer model simply shuts down. ;-)

5th question: Probably similar to conventional fission reactors.

Yes, I would also like to see such a comparison. Neither reactor is a reality, and both are undergoing intensive design study and modification.

6:18 AM  

Post a Comment

Subscribe to Post Comments [Atom]

<< Home

Newer Posts Older Posts