I get that, 100% props for the experience. And I totally respect it. The GP question was "what makes fusion different", and the only thing that is different really is the economics of how much it costs vs how much energy you get out.
Energy out is, by definition, the amount of heat differential you can generate. Cost is, again by definition, the total operating cost of the heat source.
So this is where the author and I see things differently, the author wrote: ... the "levelized cost of electricity", dominated by the capital cost of the plant, will still be much higher than other sources of electricity.
I agree with that statement, the difference with fusion is that the amount of energy produced after accounting for the capital cost of the plant will be a million times greater. And as I related in a later comment if you compare things on a dollars/BTU level the fusion plant will produce extremely cheap BTUs. Much cheaper than even the cheapest natural gas plant.
The key here is that the cost to build such a plant is much higher (and it is), but the energy produced by that plant is way higher. That is the ratio that makes fusion different.
But the LCOE is a normalized (usually to $/kwh) price that accounts for the amount of energy produced for a given capex, plus the operating costs, plus the cost of capital, plus the lifetime of the plant. It tries to bake that all in.
Your position is that LCOE will be much lower, because (as I understand you) the plant cost will scale much better than, say, 100MW natural gas plants. I totally accept that my assertion about LCOE might be wrong because it only costs 2x as much money to build a fusion plant that's 100x bigger.
The future of whether or not fusion becomes the next big thing will be watching the LCOE for fusion plants vs everything else.
It is interesting to compare fusion plants to fission plants in this regard. Fusion fuel extraction is much cheaper, fusion waste byproducts are minimal, plant failure risk and mitigation is much much cheaper (no fallout, no long live nucleotides etc), and the energy cycle produces 10 - 20x as much energy as fission.
Edit: And when things get going you can get around Carnot Efficiency by converting the high speed particles directly[1]. This experiment was built at LLNL as well and shown to actually give > 50% conversion efficiency.
You don't have to worry as much about major accidents as in fission, but the risk of minor accidents that aren't a public safety risk but are a risk of ruining the plant is arguably much larger. A fusion reactor puts a large, very complicated thing, with many non-redundant parts, into a hot zone where hands on repair is impossible.
The lesson of TMI is that even if the public is unharmed, an accident that destroys a multibillion dollar investment is ruinous for a utility.
Absolutely agree. We don't know a LOT about how these things would go together once we get past making energy. If the probability of destroying your plant was such that the lifetime expectancy was shorter than the 20 - 30 years a typical plant is expected to operate then it would definitely raise the effective cost.
Also, fusion plants have inherent diseconomy of scale, from the square-cube law. There is a limit to the energy flux through the surface of the reactor, and because of that the volumetric power density of the reactor is inversely proportional to its linear size.
I will admit that I'm not an engineer of any relevant area here. So I offer this only as a matter of general judgment & experience, which you are welcome to say "does not apply." (But please say why it doesn't.)
It's always been my impression that building things on a huge, factory-wide scale is a radically different problem than building a demo, or even reasoning about it. The latter is what you are all doing.
As long as "energy crisis" has been a Thing, I've been hearing about some uber-cool new thing a scientist did, which has the potential to transform everything. Usually that's the last you ever hear of it. I always wonder "whatever happened to X?"
I figure that "X" might look good in the lab, but no one can build a factory around it. The people who build factories (or power plants) are much less starry-eyed than we all are.
That's what we mean by "skin in the game (SITG)." Don't tell me what stocks to buy -- tell me what's in your portfolio.
So if I'm deciding whom to believe in an area where I know next to nothing, I'm much more likely to listen to someone who has, or had, SITG. Obviously if no one's ever built a fusion plant before, then exactly no one has had SITG. That makes it a tough one.
Don't believe me!! I have a solid grasp on why waste-heat-to-power doesn't work (despite free "fuel") and why geothermal doesn't work (despite free fuel). I pray that there is something different about fusion that means it will work... and make no claim to know that it won't. I don't know anything about fusion plants or their economics.
Communication is a difficult thing, especially written communication. I work proactively to get better at it when the opportunities to do so arise.
In your initial response, and in this one, what I "hear" is that what you heard was somehow either dismissive or disrespectful of the experience the OP brought to the table. I am interested in understanding what it was about my words that gave you that impression. I do understand that reality is very complicated and building systems as ones $dayjob in a competitive environment gives you a depth of knowledge that is unmatched.
What I heard in the OP comment was a question, "What makes fusion different?" and in this case the context was why might it be a step change versus other things (some of which the author was involved in) have not gotten traction. And the specific driver was how to compete with costs from extractive technologies like natural gas that are quite mature.
I shared my point of view and reasoning from the perspective that I expected the ratio of the plant construction and operation cost relative to the energy output to be much higher than pretty much any other technology so it's LCOE will be lower.
Do I know that to be the case? No I don't. Can I reason to that position from first principles? I think so, and calling me out on my reasoning as pfdeitz has done a couple of times in this thread is awesome (the most relevant for this thread being that so far we only know how to make fusion plants with a much lower energy/m3 ratio than say nuclear plants)
To your comment about 'whom to believe' I would say I'm not asking you to believe me (and I hope that wasn't a take away from my responses). I was sharing how I thought fusion was different, others have other takes on this, and nobody knows for sure because well it doesn't exist yet. I am curious though, if you feel okay sharing it, if there is anxiety with your belief system.
When I am presented with information and I don't know how much confidence to put into it, I often do similar things which is to look at the source, and to consider how the source arrived at that belief. Ideally I like to hear how they got to that belief which probably comes from reading a bunch of scientific papers that are basically one long essay that use the form: "This is what we believe, this is why we believe it, and this is the evidence we have to support our beliefs." But even in a paper I still hold that belief "lightly." If you see me passing such things on it will usually be of the form, "here is a reference to a paper or article I've read that posits this." As opposed to some definitive statement of truth. For me, I'm not invested in being "right".
I can easily say that most of the best conversations I've had, started with "Chuck you are completely wrong about that and here's why ..." I love conversations like that because it means I'm likely to learn something new, and every time we re-examine what we "know" we often come away with a better understanding of the limits of what we "know" and what we "think" or hypothesize.
Throughout this conversation it has been my intent to be clear that these are things I "think" based on what are, to me, comparable but "solved" problems. Still there are a lot of unknown unknowns such as it is and everything I think could easily be invalidated by one of them. I am always on the lookout for this sort of information. It helps keep me grounded.
Always working on improving communications and critical thinking skills and always open to feedback.
Sorry if I was curt, Chuck. I reread OP's post and your later dialog. It's interesting that you mention something I thought of but I hadn't seen before: the capacity of the grid to handle a very large power source.
I saw "3 Gw" as a comfortable max right now, maybe even that was from you; I don't know how accurate that is, but let's go with it for the moment. Let's say that you can't connect a power source larger than 3 Gw currently, although maybe raising that number is straightforward.
So it seems almost beyond dispute that the capital cost of building a fusion plant will be immense -- far, far larger than any other type, and that would be true for any type of fuel that hasn't been used before, purely because it's Plant #1. Plant #100 always costs way less.
But fusion isn't just any type of new fuel; it's radically different and unknown as to its properties. Will the neutron bombardment break it down in a couple years? Won't there be something that needs regular maintenance? No one knows how this thing will behave in 24x7 use, and yet the cost will be huge.
As OP said, even with free fuel, the capital cost of a geothermal or waste heat plant makes it noncompetitive with natural gas, and your argument is that fusion will produce so much electricity that the inequality will change.
So it costs huge sums to build the thing, and that can only be worthwhile if it produces huge amounts of power. More than 3 Gw. Furthermore, any downtime is horrifically expensive: you have to keep amortizing those construction costs.
So we probably have to upgrade the grid to handle these things and we have to over-engineer the thing grossly to avoid that expensive downtime.
One can argue that these are just part of the learning curve for any new technology. That might be true, but it probably means that it'll be 100 years before fusion is a big win.
No worries, I didn't think you were being curt, I just wanted to be sure I was understanding what you were trying to tell me. My goal is to add light not heat to the discussion (thermo pun).
We definitely approach this sort of thing differently. To use your first example:
"So it seems almost beyond dispute that the capital cost of building a fusion plant will be immense -- far, far larger than any other type, and that would be true for any type of fuel that hasn't been used before, purely because it's Plant #1. Plant #100 always costs way less."
I completely agree that it may turn out that building fusion plants are prohibitively expensive. That isn't "disputing" the assertion it is simply accepting that it may or may not turn out to be the case.
The way I approach it however is this; Let's assume there are a bunch of possible futures. Some of them are like "Fusion plants cost 10x more than even Nuclear plants" and that future is not interesting from discussion point of view because the market won't allow that sort of thing to survive. So let's consider what would indicate that they would be financially feasible.
So looking at that one question, we have a couple of interesting examples. There is ITER which is consuming billions of dollars.
Not all of the expenses there would be present in building the next one as you note. Still, even if they were half as much the plant would be hugely expensive.
Then we have the Wendelstein 7-x stellerator[1],
the HIT-SI system[2] at UW, the Lockheed Martin entry[3], and the MIT ARC project[4].
So in one future one of these alternate implementation strategies "wins" and we get fusion plants that are both functional and affordable.
So if we're going to talk about something we can talk about all the possibilities.
So we can be in complete agreement that fusion may never be feasible while still talking about what would need to be true so that it was[5].
And I'll wrap up with this, you wrote:
"As OP said, even with free fuel, the capital cost of a geothermal or waste heat plant makes it noncompetitive with natural gas, and your argument is that fusion will produce so much electricity that the inequality will change.
So it costs huge sums to build the thing, and that can only be worthwhile if it produces huge amounts of power. More than 3 Gw. Furthermore, any downtime is horrifically expensive: you have to keep amortizing those construction costs."
That wasn't what I was trying to communicate though. IF the cost of production for a fusion plant is comparable to the cost of production of a fission plant, and the fusion plant can produce twice as much energy, then based on the LCOE of Nuclear plants the LCOE of fusion energy would be a market leader. So what I do is look for things that might inform the question, "What's it going to cost to build a fusion plant?" and try to glean any insights I can from what I find.
I also agree that if fusion plants cost much more than nuclear plants to produce the same amount or less energy, then they will not be successful in the market place.
Everyone I've talked to who is trying to build commercial fusion are targeting for much less expensive than nuclear but success is never guaranteed.
[5] Yes, some people will say "Well its a waste of time to talk about something that you don't know if it can even be done." And yet, for me, it is a pragmatic use of time because unexpected things happen all the time. A case I lived through was the GM "EV-1" electric vehicle where many engineers I knew wanted to stop talking about electric cars because GM has "proven" they weren't feasible and no one would buy them because they would take to long to recharge and no one is willing to wait that long at the recharging station. That was all true during those discussions. But Tesla showed what you could do if you built one differently, and now everyone seems to think electric vehicles are the future of all cars. So for me, it isn't worthless to think about these things.
> "Then we have the Wendelstein 7-x stellerator[1], the HIT-SI system[2] at UW, the Lockheed Martin entry[3], and the MIT ARC project[4]."
Well, we're at the limits of my "competence." I guess it comes down to "what does a plant cost to build?" and I have absolutely no clue.
I don't think that "sunny optimism" is always the best attitude in everything, like it's been for us in computers the last 50 years, but you don't seem guilty of that. So: thanks & good luck!
are you suggesting that fusion plants will be petawatt sized? because that's really not the case. i think the lcoe of fusion is predicted to be around the same as natural gas, with the DEMO reactor costing twice as much.
There are a lot of unknowns around fusion plants (with the current big one being is it even feasible to build one). The last time I was looking at grid infrastructure questions the largest power plants on US grids stayed below 3GW because, as I was informed but cannot cite/verify, that was the limit on infrastructure carrying capacity of the western grid.
One of the unknowns is what is the minimum cost to build a plant capable of sustained nuclear fusion. Once you know that, you will also know what its net energy output is as well. Next you will want to know what is smallest practical increment of net energy output and what is the marginal cost of that increment.
Then, as we've been discussing here, what is the sweet spot with respect to total plant cost versus kW capacity (it's LCOE). I have chosen to use as a base estimate the cost of building a nuclear plant (the discussion of nuclear LCOE is here[1], here[2], and here[3]). I base that assumption that if we're going with heat conversion then the primary differences between fusion and fission plants will be reactor construction and fuel efficiency.
If we take it as a given that the LCOE for nuclear is approximately 10 cents per kWh, if our Fusion plants cost the same and can generate twice the power that is 5 cents per kWh. Given the difference in energy production between fusion and fission I'd like to believe it would be closer to 10x more power for the same cost or an LCOE for 1 cent per kWh.
The optimism here stems from an assumption that plant costs scale with size so smaller more powerful plants win in two ways, more power and less cost.
And yes, nobody knows if we can even build these things yet. So it really is all just speculation at this point.
You have it backwards. Fusion will have much lower volumetric power density than fission, for fundamental reasons (one can circulate coolant through the core of a fission reactor, but not a fusion reactor). Fusion reactors will likely be much larger, and therefore much more expensive, than fission reactors.
Look at the power density of ITER (0.05 MW/m^3), of ARC (0.5 MW/m^3), and a PWR reactor vessel (20 MW/m^3).
Energy out is, by definition, the amount of heat differential you can generate. Cost is, again by definition, the total operating cost of the heat source.
So this is where the author and I see things differently, the author wrote: ... the "levelized cost of electricity", dominated by the capital cost of the plant, will still be much higher than other sources of electricity.
I agree with that statement, the difference with fusion is that the amount of energy produced after accounting for the capital cost of the plant will be a million times greater. And as I related in a later comment if you compare things on a dollars/BTU level the fusion plant will produce extremely cheap BTUs. Much cheaper than even the cheapest natural gas plant.
The key here is that the cost to build such a plant is much higher (and it is), but the energy produced by that plant is way higher. That is the ratio that makes fusion different.