Helion Energy to demonstrate net electricity production by 2024

[Skipjack]

AAAND as if to confirm my earlier post, Helion is planning to do high Q D-T experiments with Polaris...
These are to the best of my understanding to see whether they can make Helium3 breeders also produce usable net energy by burning the Tritium.
It is another option but they have not decided yet. It is all about economics and D-T comes with a lot of issues, even when you do not have to breed Tritium.
I think it is a clever thing to do. Plus, if they can do high Q D-T with their machine, it will further cement their leadership in the field, killing 2 birds with one stone. If they can pull all that of (and I am very confident they can), then all larger D-T fusion designs would be dead in the water. It would only leave more compact designs like Zap and some other small, alternative D-T designs. Otherwise, only those pursuing advanced fuels will have a chance to compete.

Helion_DT.jpg (44.3 KiB) Viewed 2011 times

[mvanwink5]

SJ wrote:

Funny thing, btw. There is no reason why Helion's power plants could not be stood "upright" rather than lying on their side. That opens an interesting design option because then they can use a LiPb waterfall like Zap is doing.

Interesting point.

[RERT]

Apropos fuel mix: DT is highly reactive at relatively low temperature; D3He needs higher temperature to become highly reactive.

Might it work to use a small quantity of T as a 'fuse' to raise plasma temperature into the D3He regime?

That is, start with some fuel mix D-T-He3 and expect an overall higher fusion yield from D3He?

T3He side reactions would be rare and not especially problematic.

[charliem]

I've been reviewing what I could find about Helion's tech for the last couple weeks, patents, papers, presentations, videos, etc.

John Slough patents regarding Helion's prototype reactor design have changed remarkably little for the last 10+ years, at least in general architecture, overall size, methods, and many parameters. The main alterations seem to be fuel selection, and temperature.

D-3He is a relatively recent switch, before it was D-T. Designs from 3+ years back include blankets around the reaction chamber to collect neutrons, and traditional thermal electricity generation in addition to direct conversion.

So, yeah, it looks like they gave a good amount of time to study the possibility of using D-T.

There are a number of interesting figures and data points in theses sources, although most are projections not experimental results, and many have probably been superseded already:

• Some time back they estimated their energy production prototype could reach densities up to 10^24 ions/m3.

• Trenta's magnetic fields can compress FRCs up to 400 times initial formation density.

• With D-T, their prototype would need to invest ~27 MJ to start the process, and harvest ~59 MJ, producing 50 MWe (net) working at 2 Hz.

• The final FRC is remarkably elongated, one of the sources mentions a diameter of only 0.014 m and length of ~2 m.

• Ten years ago, while "selling" Venti to their investors, they wanted to design it to reach a B of up to 20 T (obviously they didn't, so I suppose they had to compromise ?).

• ...

I have not found jet any mention to experimental data regarding confinement time although one of the formulas from a 2018 Dr Kirtley's presentation, applied to some parameters from other sources, suggest a Tau only slightly under 1 ms for the B field and temperature they say are going to try with Polaris.

Interesting reading.

I hope Helion gives more numbers in the future. I'm a big fan of hard numbers.

[Skipjack]

D-3He is a relatively recent switch, before it was D-T. Designs from 3+ years back include blankets around the reaction chamber to collect neutrons, and traditional thermal electricity generation in addition to direct conversion.

The He3 is in the name of the company. They were always looking towards advanced fuels but D-T would have/could have been a first step.
The D-T in more recent times was mainly because ARPA-E required it and they needed that funding source in the beginning.
ARPA- E wanted high gain, low pulse rate. Helion is aiming for low gain high pulse rate.

So, yeah, it looks like they gave a good amount of time to study the possibility of using D-T.

Yep!

There are a number of interesting figures and data points in theses sources, although most are projections not experimental results, and many have probably been superseded already:

Their scaling laws have held up pretty well, actually. If you look at older presentations by John Slough, the equations are pretty much unchanged.

• Some time back they estimated their energy production prototype could reach densities up to 10^24 ions/m3.

This is something that I believe they have changed relatively recently, but I might be wrong. They will still have orders of magnitude higher densities than a Tokamak, but they will likely keep it below the 10^24 ions/m3 but more moderate around or slightly below 10^23 or so.

With high Beta, you can scale between temperature and density linearly. If you want to favor D-D reactions over D-He3 reactions (e.g. in a lean He3 burning machine that breeds its own He3), they would use higher densities and lower temperatures. I believe they might be doing the same with D-T.
For D-He3 experiments they will dial up the temperature and use lower densities. There are a few other advantages to going with higher temperatures. E.g. FRC stability increases with higher temperature. The Te:Ti ratio decreases with temperature. This reduces losses.

• With D-T, their prototype would need to invest ~27 MJ to start the process, and harvest ~59 MJ, producing 50 MWe (net) working at 2 Hz.

I believe their (original) D-T design was aiming for only 25 MWe net production to keep the neutron wall loads manageable or they would have needed a bigger machine. D-He3 machines are actually better in this regard!

• The final FRC is remarkably elongated, one of the sources mentions a diameter of only 0.014 m and length of ~2 m.

Yes, the elongation factor Epsilon helps with tilt instabilities.

• Ten years ago, while "selling" Venti to their investors, they wanted to design it to reach a B of up to 20 T (obviously they didn't, so I suppose they had to compromise ?).

Venti was built for ARPA-E Alpha. That program had two phases. Helion completed Phase 1. They got 10^19 keV s /m3 out of Venti, actually surpassing expectations. That kind of triple product is mighty impressive for what was essentially a bench top system. Phase 2 was going to increase the B to 20 Tesla, but by then their investors decided that they should go straight to Trenta instead. The goal was D-He3 not D-T. So they viewed Venti as a bit of a distraction.

I have not found jet any mention to experimental data regarding confinement time although one of the formulas from a 2018 Dr Kirtley's presentation, applied to some parameters from other sources, suggest a Tau only slightly under 1 ms for the B field and temperature they say are going to try with Polaris.

Venti was tiny and so its confinement time was limited to about 40 microseconds. Trenta did about 0.5 milliseconds from the numbers that I have seen.
Goal for power plants is ~1 ms if I am not mistaken.

[charliem]

I've been able to find a number of patents under the name John Slough and MSNW, describing the machine Helion is developing, the first from 2010, three years before Helion Energy was incorporated.

I also found multiple peer reviewed papers authored by John (and colleagues or students). One of them, also from 2010, describes in fine detail the results obtained with the I.P.A. experiment/machine.

I think Helion considers IPA a predecessor of theirs (3rd generation, maybe ?). Given they are different iterations of the same general design, those numbers suggest orders of magnitud for many parameters for newer machines, but not much more.

I have not been able to find any paper or presentation with the same level of detail about Grande, Venti or Trenta, than that paper offered about IPA.

Any idea whether Helion has published any ? It would be nice if we could compare results by generation.


EDIT: Corrected a mistake in the 1st paragraph, the older patent I found containing Helion's machine general design is US2010093981A2, from 2010, not 2013.

[Skipjack]

I've been able to find a number of patents under the name John Slough and MSNW, describing the machine Helion is developing, the first from 2013, three years before Helion Energy was incorporated.

Yes, Professor John Slough is a co-founder of Helion. John Slough was the owner of MSNW LLC. Helion was spun off from MSNW to commercialize the lessons learned from IPA, IPA-C and previous machines.

I have not been able to find any paper or presentation with the same level of detail about Grande, Venti or Trenta, than that paper offered about IPA.

Any idea whether Helion has published any ? It would be nice if we could compare results by generation.

Because MSNW was a research company with close ties to UW. Most of the work there was financed by grants. Helion is a company for the commercialization of the fusion engine design. They are mostly funded by private investors, not by grants. So they do not have to publish (much). They have published a couple of papers since then though. They also have a new paper in peer review right now, I hear.

[jrvz]

I suppose you've seen this patent?
John Thomas Slough, David Edwin Kirtley, and Christopher James Pihl. Advanced D-3He fuel cycle for a pulsed fusion reactor. European Patent EP 3103119 B1, March 2021.

[Skipjack]

I suppose you've seen this patent?
John Thomas Slough, David Edwin Kirtley, and Christopher James Pihl. Advanced D-3He fuel cycle for a pulsed fusion reactor. European Patent EP 3103119 B1, March 2021.

I have seen their patents, yes.
This is for the D-He3 cycle.
There is another patent for a D-T but also fuel agnostic system that this patent is based on.
https://patents.justia.com/patent/20180025792

[charliem]

I saw it too. It was originally filed in in February 2014, just a year after the creation of Helion, and granted in 2015.

Since starting Helion, Dr. Kirtley has filed a bunch of patents related with electrical and other subsystems derived from his work on fusion, and also a number of non related articles (mostly, aerospace stuff), but nothing too detailed about Helion's results (that I could find).

Let's hope the paper Skipjack just mentioned they have in peer review don't take too long.

[charliem]

Glad to hear confirmation that Helion is going to try D-T in Polaris because, after playing with their own equations for a few days, I've been unable to find a combination of parameters that result in net electricity production using D-3He (in that machine), or anywhere near.

Of course, I'm no fusion scientist so I may well be wrong, but the published formulas are not that hard. Simply put, 20T and 20-30 keV seem to be insufficient.

On the other hand, with D-T the numbers do fit (even letting all the neutrons go to waste). Of course there's the problem of shielding, we are talking megajoules in neutrons per shot.

Fun fact (extracted from Helion's power points): J. Slough and D. Kirtley wanted to make Venti powerful enough to reach those 20 T (peak), and later on had the same goal for Trenta. In the end they had to do with just 8 and 10 T, respectively. Not enough funding, maybe?

[Skipjack]

Glad to hear confirmation that Helion is going to try D-T in Polaris because, after playing with their own equations for a few days, I've been unable to find a combination of parameters that result in net electricity production using D-3He (in that machine), or anywhere near.

I would like to see your math there because net electricity from D-He3 is still the goal for Polaris.

Of course, I'm no fusion scientist so I may well be wrong, but the published formulas are not that hard. Simply put, 20T and 20-30 keV seem to be insufficient.

Which formulas did you use? Did you watch David Kirtley's talk at Princeton?
https://mediacentral.princeton.edu/medi ... 1_9p8c7d85

Fun fact (extracted from Helion's power points): J. Slough and D. Kirtley wanted to make Venti powerful enough to reach those 20 T (peak), and later on had the same goal for Trenta. In the end they had to do with just 8 and 10 T, respectively. Not enough funding, maybe?

See my response to your earlier comment about Venti.

Venti was built for ARPA-E Alpha. That program had two phases. Helion completed Phase 1. They got 10^19 keV s /m3 out of Venti, actually surpassing expectations. That kind of triple product is mighty impressive for what was essentially a bench top system. Phase 2 was going to increase the B to 20 Tesla, but by then their investors decided that they should go straight to Trenta instead. The goal was D-He3 not D-T. So they viewed Venti as a bit of a distraction.

[Carl White]

Helion is (unsurprisingly) pleased with the U.S. Nuclear Regulatory Commission's recent decision to regulate fusion energy systems based on existing nuclear materials licensing.

https://twitter.com/Helion_Energy/statu ... 1827345408

[charliem]

Thank you for the inside data, Skipjack, for people like me there are never enough numbers

Let me say that I am not taking Improbable Matter's side any time soon, I'm just a fan of physics, fusion in particular, someone that takes pleasure in trying to understand as much as I can about these matters. I'm rooting for Helion's success.

Yes, I did watch Kirtley's presentation at Princeton. The formulas I've used come mainly from one of the slides he presented there, the one with equations for Pfus, Pbrem, Psync and Ptransport. Those are incomplete, so I filled the gaps with some material from earlier works by J. Slough, D. Kirtley, and others.

I feel a bit confused about some parameters, I fear I may have mistaken one for another. For example, B. It is used multiple times, but there is more than one B inside the compression chamber, at least we have B coil, B external and B vacuum.

Let me illustrate my doubts with, for instance, estimating triple products:

• nT seems easy to calculate (ηei=electron to ion temp ratio): n*T = β*B^2/[2*μ0*kb*(1+ηei)]
• For confinement time Helion patents mention this old empiric formula: τ_N = 3.3E-15 * ε^0.5 * x_s^0.8 * r_s^2.1 * n^0.6

Skipjack offered that Helion reached a triple product of 1E+22 eV*s/m3 with Venti. As far as I know, Venti did B=8T max, β around 1, and accepting ηei=1/8 (given that plasma temp only reached 2 keV, it was probably less, but that would not change the result much). From those figures we obtain: n*T = 1.5E+30. Dividing triple product by n*T gives τ_N = 0.08 ms.

We know the real time was 40 microseconds, so there's a factor of 2 difference between calculated and experimental data. It would be nice to find out where is the mistake in my calculations.

By the way, doing the same for Trenta, from B=10T, β=1, Ti=9 keV, ηei=1/10, and τ_N = 0.5 ms, I obtain an ion density of 2.5E+22 m^(-3), and a triple product of 1.1E+23 eV*s/m3. How does this compare with the real deal ?

[charliem]

Did you remember to add back the 90%+ of the energy put in that is recovered?

Yes, I did. It is not that, but I have an idea.

One of the numbers Skipjack gave in one of his last posts had me thinking.

The problem with not having enough hard data is that I had to make assumptions about some parameters, and at least one of them seems to be off.

Polaris needs to generate at least 11% of the power it takes to create, accelerate, and compress its fuel, to counter the losses. The input energy needed to create a FRC is independent of its lifetime, but the fusion output IS, and if that FRC dies too soon ...

Skipjack said Helion is aiming for a confinement time of 1 ms. That's incompatible with keeping the reaction chamber size I saw in previous Helion's patents, and that I assumed they were going to use for Polaris. With the ion density resulting from B=20T and T=20keV, plus those dimensions, 1 ms is outside the possible (according the scaling law I have).

Also, there is something Kirtley said in one of his videos, that they found the compression stage has to be 25% wider in order to prevent the plasma impinging on its walls.

So, Polaris is going to have a bigger reaction chamber (and FRC) than I thought, and consequently have a longer ion confinement time. I don't know how much bigger, but tripling my assumed volume should do the trick, that would extend the achievable confinement time to up to 2 ms.

Well, isn't it nice. There is a chance

Edit: grammar, spelling and clarity