Fusion issues - when does an exaggeration become deception? [And cringe, but I cite Elon Musk in support!]

I'm a bit surprised, but Sabine Hossenfelder seems to not want to give any encouragement to fusion power skepticism after the "net energy gain" breakthrough announcement from Lawrence Liverpool this week.  

And look, I know that I criticise amateur "armchair experts" on matters like climate change and vaccines, so I feel I am at great risk of being called a hypocrite when I now put my own version of amateur assessment on this topic.  

But, but:  I reckon anyone just has to read a bit more widely to understand that the problem is not just getting fusion to work - it's getting it to ever work in a way that makes economic sense for power generation.  I reckon that it's that aspect which no one is asking the pro-fusion researchers to properly discuss and justify. (Sure, the timeframe question comes up - more on that below - but I reckon there is plenty of reason to doubt that it will ever be economically viable.)

I mean (ugh, I know I shouldn't do this appeal to gut reaction, because it feels so much like the same tactics climate "skeptics" use) but look at this photo:

Does this look even vaguely like an easily deployable system for power generation?   It's the National Ignition Facility at Lawrence Livermore where they made the breakthrough, and of course, it just an experimental set up and it was never meant to be something that would generate useful power.  But still, a picture gives an idea of the complexity of this type of fusion set up, so I'm still running with it. 

And when you read about the set up, it almost seems that the question should be "how come it took so long to even get to the net energy gain"?  

‘Nif is the world’s largest, most energetic laser,’ she explains. ‘It’s 192 separate lasers, each one of which is close to the most energetic in the world. And it’s housed in a building that’s three American football fields wide and 10 storeys tall, which is needed for all the amplifying objects. In fact, it’s the world’s largest optical instrument.’ When it fires, the facility’s beams are amplified by 3070 sheets of phosphate glass doped with neodymium, each weighing 42kg and set at Brewster’s angle, which reduces reflective loss. ‘The idea is we take all of that energy, which comes to about 1.9 megajoules, and focus it down on a target the size of a small ball bearing, about 2mm in diameter.’

As for how long they have been trying to get it to make net energy (and only considering the laser power going in, not the energy needed to make the lasers) Science magazine explains:

The $3.5 billion NIF began its “ignition” campaign in 2010.  ...That self-sustaining burn is what defines ignition, and after more than a decade of effort NIF scientists declared they had achieved that milestone after a shot in August 2021 produced 70% of the input laser energy. But NIF’s funder, DOE’s National Nuclear Security Administration, set NIF’s goal as an energy gain greater than one—the threshold it passed last week.

So, $3.5 billion and 12 years to get a single event in which the energy of the reaction was about "the equivalent of about three sticks of dynamite."  A small energy return on investment, if ever there were one.

The Science article does go on to explain a possible future direction for laser fusion (my bold):

The NIF scheme has another inefficiency, Betti says. It relies on “indirect drive,” in which the laser blasts the gold can to generate the x-rays that actually spark fusion. Only about 1% of the laser energy gets into the fuel, he says. He favors “direct drive,” an approach pursued by his lab, where laser beams fire directly onto a fuel capsule and deposit 5% of their energy. But DOE has never funded a program to develop inertial fusion for power generation. In 2020, the agency’s Fusion Energy Sciences Advisory Committee recommended it should, in a report co-authored by Betti and White. “We need a new paradigm,” Betti says, but “there is no clear path how to do it.”

Now that NIF has cracked the nut, researchers hope laser fusion will gain credibility and more funding may flow. 

[Betti, by the way, is from another research lab.]

About that funding - as everyone who has read anything about this knows, a lot more money is going into tokamak fusion research, in the form of the gigantic and hugely expensive ITER plant being built in France:

However, the leading tokamak device, the ITER reactor under construction in France, is anything but simple. It is vastly over budget, long overdue, and will not reach breakeven until the late 2030s at the earliest. With NIF’s new success, proponents of such laser-based “inertial fusion energy” will be pushing for funding to see whether they can compete with the tokamaks.

As for the cost - it seems a matter of much dispute as to how to actually cost it, which is a little odd, but the range (shared by many nations) seems to be from $22 billion to $65 billion.

All that money for possible breakeven by the late 2030's.

Also, the article I first linked to in this post is from Chemistry World, which explains one of the fundamental issues on the economic development of fusion power - the development of suitable materials needed around a fusion reactor:

The greatest problem faced in fusion isn’t achieving the incredible temperatures required – it’s the materials science required to maintain that environment long-term. It’s why Jet couldn’t go past a few seconds, explains Rimini. ‘Jet is based on fairly old copper coils for the magnetic fields, and the tokamak walls are not actively water-cooled, so the high fusion period is only designed to run for 10–15 seconds at most.’

UKAEA has built a new materials research facility at Culham Science Centre to tackle such problems. One of the staff searching for solutions is Greg Bailey, a computational nuclear physicist. ‘The copper magnets get too hot,’ he says. ‘So, in the future, we’re using superconducting magnets. And hopefully we’ll learn more.’ These material changes have already happened in the past. ‘Jet actually changed the material of its walls,’ Bailey says. ‘Initially we’d made the walls out of carbon, because that made life easier for the experiments. It should have been perfect, but, actually, it was terrible! We were getting a lot of tritium retention – we were losing our fuel into the wall, the hydrogen was drifting inside. So we had to change it.’

The design challenges discovered and solved by Jet are already being fed into Iter, explains Bailey. ‘What does a material for a reactor need to be? Resistant to damage [from radiation], it needs to be able to take the temperatures and extreme environments, and maintain its mechanical properties during its lifetime. So, in terms of a fusion reactor, the vast majority is probably going to be steel. The really interesting bits come inside the vacuum vessels, your housing, because they’re going to be facing extremes. They need armour, obviously.’

This has resulted in plans for Iter to be covered by 440 ‘blanket’ modules, weighing up to 4.6 tonnes, which cover the steel of the tokamak’s structure. Neutrons discharged during the reaction the enter the blanket can be slowed, and their kinetic energy transferred to a coolant system for another form of power. It’s hoped the blanket can also be used to solve another issue for reactors: their feedstock.

‘There’s plenty of deuterium on Earth,’ Bailey says, ‘but deuterium fusion produces much lower energy neutrons; it’s not really a viable source to make a power plant. And tritium is not naturally occurring.’ To obtain their tritium, the team plans to use lithium with an enhanced level of lithium-6, which can break apart under neutron irradiation to produce tritium. Although this is naturally occurring, the problem is that lithium is already in high demand for its use in lithium-ion batteries. ‘Frankly, when lithium comes into our reactor, we’re going to destroy it,’ Bailey says. ‘The fuel is not the problem; it’s how you produce it.’

This is where the blanket could come in, explains Bailey. ‘A lot of designs right now are mixing lithium with lead, or lithium with ceramic and some beryllium in there. The idea is that you get deuterium and tritium, the fusion reactor turns on, and neutrons produced in the fusion reactions smash into the blanket and tritium breeding reactions can occur. We can then extract that tritium to refuel the reactor. And, obviously, the neutron radiation into the blanket will cause a huge amount of heating.’ It’s still not perfected yet, but Bailey is confident the experiments done at Culham will show the way, potentially in collaboration with the private sector; fusion is already attracting major investors, including Amazon’s multibillionaire founder Jeff Bezos. ‘If we want to do fusion on an industrial scale we need to start building that supply chain now,’ says Rimini. ‘We need to start evolving the industry.’

Obvious questions I have:  how long will the "blanket" modules last?   How long will a fusion power plant need to be down while they are replaced?   At 4.6 tonnes each, and presumably all getting radioactive at the same rate - it's going to be a huge maintenance job, and it's something they are only now trying to work out. 

There's a complicated 2017 paper here about the materials science challenges for testing and developing suitable materials:

This paper presents a preliminary evaluation of the materials challenges presented by the
conceptual design [1] for a Fusion Nuclear Science Facility (FNSF) to bridge the development gap between ITER and a demonstration power plant (DEMO). Here the FNSF specifically denotes the concept that has been studied in the recent Fusion Energy System Studies (FESS) supported by the US Department of Energy, also called the FESS–FNSF, which is examining a  conventional aspect ratio tokamak. The FNSF is an experimental machine designed to establish the reliable performance of the critical fusion system technologies required in DEMO and power plants. The FNSF horizontal maintenance system [2] allows for periodic removal, examination, and replacement of full power core sectors.

As far as I can tell, this Facility does not exist yet, and won't for some time.   This presentation from 2014 seems to indicate that it wouldn't really get going until ITER is up and running - in the 2030's - and the 2017 paper says this:

A minimum 20-year timeframe will be required to accommodate the development of the advanced materials to commercialization and code qualification, development of blanket fabrication technologies, evaluation in non-nuclear integrated test programs, and 14 MeV neutron testing in DONES/A-FNS/IFMIF to validate irradiation performance.

So piecing this together, we're getting the "best hope" for tokamak fusion not likely getting to break even until the late 2030's, during which decade a materials research stage which will take a minimum of 20 years will have started.   

Does this sound like commercialisation of fusion power within 20 years?   No it doesn't - sounds more like 40 to 50 - if it is possible at all.  Because isn't this complicated materials science issue likely to be a key one in the question of whether fusion will ever be economically viable?  And we won't even know the answer to that for another 20 to 30 years.

AND YET:   this morning on Radio National, we heard Kim Budil, the director of Lawrence Livermore National Laboratory (home of the "breakthrough") say this at the 12.40 mark:

"So not 50 years away anymore, I would say probably 2 decades of concerted effort and it's plausible we have power plants in development"

To her credit, Patricia Karvelis, sounds skeptical "Wow - really - in 2 decades?"

And Ma says "I think so"

I'm sorry, but ever allowing for the qualifiers of "probably" and "plausible", I reckon that that answer is so practically unrealistic as to be deceptive.  

I'll come back and add a bit more to this post later...

Update:   I had a look at Youtube videos about it, and quickly found one in which a former Secretary of Energy (and nuclear physicist) makes an outlandish claim the he "think[s] we can demonstrate and maybe initially deploy some power plants on the grid within the next decade or so".  [!]

Gee, if anyone invests money in the company he's on the board of, based on this type of spruiking, I reckon it would come close to fraud:

 

More realistically (much, much more realistically) we have an actual former fusion scientist who thinks it's worth pursuing, but he explains in this video from a year ago the huge engineering issues yet to be overcome.  He says there is no way we will have fusion by 2040, and everything I have listed above indicates that is correct: 

 

Finally, and I didn't see this coming or realise it until now, but I'm on the side of Elon Musk!  Here's a short clip in which he says that sure, fusion will be achievable, but it's just not going to be economically viable as a power source, citing the tritium issue mainly.   [I can't embed it, as it's a Youtube short.]   

How embarrassing is that, given that he seems to have driven himself nuts by blowing many billions on Twitter?  Quite - but hey, if the facts are actually on his side on this issue, so be it.