Anatomy of a Fusion press release
Listening to the nonsense mindlessly parroted by media outlets, you’d think we were a few months away from limitless fusion power. I’d like to believe that. I’d like to believe my country was still capable of making important technical breakthroughs. Just as I’d like to believe my country’s foreign policy isn’t run by simpering baboons who don’t know a Cossack from a cassock.
Unfortunately, I went to college, and I once worked in a government lab, so I know better. The actual news behind the recent fusion press release is no cause for rejoicing. NIF isn’t be something we should be lionizing in the press; it’s a national embarrassment. The people who made the press release certainly know this. Unfortunately, the people who read the press release couldn’t be bothered to understand anything about it.
What is NIF? NIF, or “national ignition facility,” is a bunch of big lasers, designed to implode a pellet of deuterium and tritium well enough to achieve fusion “ignition.” Ignition means, the reaction is self sustaining; the fusionables burn after you light them on fire, so to speak; just like the regular-world definition of ignition. LLNL had hoped to have achieved ignition by now, but they haven’t. NIF construction also took about four times what it was supposed to cost, and was turned on five years behind schedule. The original management team (including perennial presidential candidate Bill Richardson) was excoriated in the GAO report. This is a terrible record, even for a government program. But it gets worse…
In 2012, Congress reviewed the project, wondering why the project had not yet achieved ignition. They’re in deep shit with their paymasters. It’s now 2014, and they still haven’t achieved ignition. Hence this press release, just in time for the 2014 fiscal year.
What did they announce in this press release? Well, they claimed a kinda-sorta “scientific break even.” Supposedly, they got 5×10^15 neutrons out, which somehow is higher than the energy they managed to impart on the actual target. This is, to say the least, deceptive. The energy obtained in this reaction was allegedly as high as 14KJ. The energy put into the reaction was 1.8MJ. Their “break even” assertion is that the 1.8MJ laser flash managed to put slightly under 14KJ on target. An article in science magazine mentioned this back in 2013, when the first press release hit; mentioning that the 1.8MJ pulse is about the same as a two-ton truck going 100mph, and the imparted result was something like a baseball going 50mph. The author didn’t mention that the energy required to generate the two-ton truck laser pulse required a few hundred two ton trucks worth of energy input, because the laser system uses old school flash-tube pumped YAG technology, but that’s a factor also. All this totals to around away from true break even. Unfortunately, LLNL made another press release a week or two ago, essentially repeating the September 2013 press release, and nobody bothered checking with a grown up.
NIF as an engineering project actually does what it promised; it consistently delivers 2MJ UV laser pulses. Laser inertial confinement is an old idea at LLNL, and it was an ambitious engineering job to get to this point. As an engineering project, I have to admit it’s pretty cool; the high energy pulse control, the giant KDP crystals and the deformable mirror optics is all good stuff. It ought to be, considering it cost 4X what it was supposed to, but the engineering of the device is blameless. It’s the physics that failed.
While the history of NIF is marred by some lousy management, we have a pretty good idea of why it has failed to do what it is supposed to do. The computer models used to calculate the physics of a laser pulse hitting a target: they don’t work. They’re making a prediction which is at least a factor of 5-10 off from what is observed, in terms of energy absorption. Worse; the material used to compress the fuel seems to have unpleasant characteristics which tend to snuff out the fusion reaction, slowing the neutrons and generally behaving differently than it is supposed to (as documented in this review PDF) and nobody understands that either.
So, the actual fuel doesn’t work like it’s supposed to, and neither does the “match” which is supposed to light it on fire. This should not be a great surprise. Computer simulation works fairly well when you understand all the physics, and when the equations are well behaved. It doesn’t work at all when you leave out important physics, or when the equations are badly behaved. I’d expect something like a highly focused megajoule nanosecond UV laser pulse on exotic materials to be something which falls into the “not so well behaved” class of physical models. I’m guessing the recommendations in the review article are being followed, and the NIF folks are attempting to adjust their models using some indirect observations and calibrations. I think they’re also taking a punt by fiddling with beam pulse shape. It appears the “good” result of September 2013 happened via the latter technique.
One of the saddest things about this debacle is the JASON committee recognized problems with the computer models back in July of 2005 (see page 44-48 of this report). The same group has also pointed out that, one of the alleged benefits of NIF, that of stockpile stewardship, is a mirage, as tiny little explosions don’t have a whole lot to do with great big explosions. That would be true, even if the thing worked.
Sadder still, we didn’t need the JASON committee to tell us this sucker was going to fail. The history of inertial confinement fusion at LLNL provided ample evidence that NIF would fail. Consider the last laser inertial confinement project at Livermore: the NOVA project. It, too, claimed that it would achieve ignition, and using a much smaller laser. They claimed you’d only need tens of kilojoules to achieve ignition. Its performance as far as break-even goes was approximately the same as what NIF has done so far (I think it released 10^13 neutrons, which is about what you’d expect from a laser which generates two orders of magnitude less power). It did so with a much lower energy laser pulse. They blamed all kinds of stuff for its failure, but at the end of the day, the physics of the laser/plasma and “shell/fuel” interaction is poorly understood. This was somewhat known back in the 1980s with the results of the Halite-Centurion experiments. These test results, classified and done with nuclear weapons, seemed to indicate that the implosion approach might need as much as 100MJ laser pulses to achieve the goals of NIF.
Following “current events,” while the thing was being built would also have been instructive. Our media class seem to enjoy making fun of the corruption among the ruling classes of Slavic countries. I figure it is a fair cop; those places are pretty blatant in their corruption, at least looking at things from over here, but the media also almost always overlook corruption in America. This is inexcusable, as this is their job. I can laugh at the Rooskie without any assistance from the tittering pustules in the New York Times. Their only excuse for taking up space is to unearth corruption in America. They don’t do this very well. Here’s a few little pieces of information about NIF from way back in 1999-2003; mostly ignored by the media, yesterday and today.
- The review committee responsible for signing off on this project was widely viewed as stacked in LLNL’s favor. Having witnessed a few DoE national lab committees; this is fairly common, and a really bad idea if you think the idea of “peer review” is useful. There are also rumors that a certain DoE bigwig (one who got his job via bribery; common enough for political appointees these days) who engineered this, hoped to profit from NIF, as he owned a laser optics company. I have been unable to verify it with DoE old timers, but if it’s true, well, I’m not a reporter, but someone should have noticed …
- The associate director of NIF in 1999, E. Michael Campbell, had to resign, as it was revealed that he never got a claimed Ph.D. in physics from Princeton. He had a soft landing; he got a job at General Atomics afterwords, and eventually got a Ph.D. from some lesser school. I hear he was a decent physicist; you don’t need a Ph.D. to be good at this kind of work. But this is shady business. It was a scandal at the time, but somehow nobody thought to investigate whether or not this fraud’s main project was also a fraud.
- A NIF whistleblower, Lee Scott Hall, was brutally stabbed to death in 1999. Hall was one of the NIF designers. I didn’t know him, but I know people who knew him. He definitely wasn’t into any shady business, and by all accounts was a straight shooter and a decent guy. This … really looks like a man was stabbed to death for pointing out that NIF couldn’t work (or any number of other things). Sure, nobody knows who killed him, but this is at least as shady as anything I’ve heard of happening in Russia. This sort of thing has happened quite a few times to whistleblowers at the Department of Energy.
- Two NIF whistleblowers who didn’t die, Les Mikolsy and Luciana C. Messina, were ignored, then fired in 2003. They were involved in the code that controlled the giant contraption, and pointed out some serious problems with it, including problems with safety implications, and the management of the project. Not a peep from anyone about their assertions or the lawsuit they filed against LLNL (it was dismissed; yay objective and completely non-corrupt court system!).
One of the greatest problems with the United States of America right now is a complete lack of accountability among our ruling classes. People still get fired; never the managers who break things. Maybe patience is necessary in this instance, and that inertial confinement fusion is “coming real soon now.” Maybe Ed Moses should be frog marched to the firing squad, or made to peel potatoes until he pays the taxpayer back the billions his gizmo cost. One thing is certain though: NIF, and all who are involved with it (heroic whistle blowers excluded), are presently national embarrassments. They have failed, and they should be roundly mocked for it. This isn’t a matter of, “gee, we should wait a little longer,” this is a matter of, “they promised it would work factors of 100 or 1000 better than it does several years ago, and they have failed repeatedly.”
Wanna know why human progress has slowed? Talented people wasting their lives on crap like NIF, which only exists because of corrupt empire builder bureaucrats in white laboratory jackets.