Hacker News Reader

Comment by Veedrac | Hacker News Reader

This article is good in some ways and mistaken in others. Particularly, as a criticism of the ITER/DEMO path to fusion, it is correct. That was never a viable path to commercial fusion.

The mistakes made, or else those allowed for but not predicted, listed without particular order,

* Not all fusion requires turbines. Eg. Helion's designs don't.

* Most if not all renewables require either energy storage, or much better energy transport, which must be factored into their price.

* The size of commercial magnetic confinement plants is not well represented by DEMO, given new high temperature superconductors.

* (Personal opinion) The cost of fission plants is mostly a consequence of regulation, and this dominates the financial aspect.

That said, the idea that fusion has extremely tough cost competition from solar and wind, and that it will get increasingly tougher over time, is (and has turned out to be) absolutely true.

View on HN

> Most if not all renewables require either energy storage, or much better energy transport, which must be factored into their price.

This was the big one for me. Many people underestimate the complexity of energy transport over long distances. At some scales, this is still a materials science research topic, but even when it has been commercialised it's exorbitantly expensive over the international distances that would be required.

As for storage, despite the medium scale battery deployments we've seen in the last ~5 years, and pumped hydro, I'm remaining unconvinced that storage is something we'll rely on for industrial-scale grid usage in a post-fossil fuel world. These solutions need to be able to power a country, with all its industry, for ~days while wind and solar aren't running at necessary levels.

>> Most if not all renewables require either energy storage, or much better energy transport, which must be factored into their price.

>This was the big one for me. Many people underestimate the complexity of energy transport over long distances. At some scales, this is still a materials science research topic, but even when it has been commercialised it's exorbitantly expensive over the international distances that would be required.

There's plenty of long distance electricity links that operate commercially, and they are cash positive simply due to price differentials between regions. To give you some examples, both Sweden and Norway are connected to Denmark/Germany and essentially power countries like Italy or France when the demand is there. Essentially much of the European energy grid is already integrated and it is profitable to build extra links simply because of price differentials due to weather. So if you have different evidence, please provide it.

> As for storage, despite the medium scale battery deployments we've seen in the last ~5 years, and pumped hydro, I'm remaining unconvinced that storage is something we'll rely on for industrial-scale grid usage in a post-fossil fuel world. These solutions need to be able to power a country, with all its industry, for ~days while wind and solar aren't running at necessary levels.

Again if your grid is large enough you really don't need to power the whole country, but even so unlike nuclear and fusion, storage solutions are on an exponential cost reduction curve (however you are correct that they are currently still expensive).

I would content that a lot of these long distance projects are 'profitable' of sorts because of the choices made in the respective energy markets.

If everybody would simply have the rule of every per X00000 people you build one reactor in a geographic location near them. Then you would simply not need the waste majority of all long and medium distance energy transport.

You need some connection to be able to support temporary maintenance but for the most part each region is pretty independent and relatively simply.

That reduces your grid complexity a great deal and doesn't require country or even continent wide coordination with lots and lots of highly variable inputs, complex energy price markets and so on.

> * (Personal opinion) The cost of fission plants is mostly a consequence of regulation, and this dominates the financial aspect.

Do you have any evidence to actually back that up? This is being brought up every time we talk about nuclear and every time someone asks for evidence there's nothing. All the serious analyses of the cost, actually show this is not the case. I mean this very article explains why, nuclear reactors are thermal plants based on a steam engine. The article directly mentions even if the reactor is free they are more expensive than wind or solar, in other words just the steam engine is already too expensive.

Well first of all I would argue all in cost for a whole countries power needs, based on current fission is still competitive. If you build a large number of plants you can get a very good stable energy supply.

With renewables you need a great deal more transportation of power, you need a gigantic deal of storage of power to be truly independent.

I would still go with nuclear even if its more expensive in terms of dispatch energy cost. Because as we have seen, you can have pretty long period where both wind and solar are at 5% capacity, and you better have a backup that can last weeks if you have no fossil backup.

All that said however, I think just saying 'regulation'. Doesn't capture the problem. I think the more accurate description of the problem is 'innovation'. After a golden period of the 50/60 innovation slowed down dramatically in the 70s and came to essentially a dead stop and since then we have barley improved other then some marginal gains on old designs.

Not only way most state research money removed all over the world, the current state of the art was embedded in regulation and even if somebody would be willing to put up the capital to research a new reactor making it commercial was essentially impossible. Specially in the US, one of the few places where state independent nuclear might have had some chance, regulation were essentially defined in a way that only water cooled reactors could ever be legal.

In Europe all nations other then France turned on nuclear pretty hard and while some plants were built, actual innovative designs were not ready to be commercial and mood turned against nuclear and most projects were canceled.

France did build a lot of nuclear but overall they were not really innovative, the just scaled the production of existing technology. This is great of course and gave France green energy. However even there as time went on nuclear was seen as a legacy, just waiting to be finally replaced.

So nuclear today is still low mostly temperature, and this requires specially gas turbines. The kind of turbines that are no longer used in modern gas and coal plants. As coal and gas used better turbines and scaled their production nuclear was unable to take advantage of those better steam turbines. And because no high temperature nuclear power existed, there also wasn't much effort put into turbines required for such high temperature reactors.

So today building a PWR you have a couple of big issues. The 'Pressure' in PWR requires you to build a hugely expensive reactor with gigantic heat exchanges that need to be contained in a huge secure building that needs to be able to contain a hydrogen explosion (and that doesn't actually work as you could see at Fukushima). Second, you need a hugely expensive very large, not very efficient steam turbine that today is only used for nuclear, and thus is even more expensive then historical was.

Third, because of the scale of these projects, in addition to the nuclear bit, they are simple gigantic public civil engineering projects that suffer from all the same problems that huge civil engineering projects suffer from.

Now had we innovated and moved to what is now called 'GenIV' reactors many of these things could be solved. First I would say, the term 'GenIV' is really a bad term because some reactors we now call 'GenIV' have been in development of some kind since the very beginning and have no iterative evolution with the Gen1-Gen3. The 'Gen' here refers to commercialization not research.

So how would these reactors potentially solve many of the issues. First, you can massively decrease the size of both your reactor, heat exchanger and your steam turbine. All of this results in a massive improvement in capital cost and therefore also the scale civil engineering project. In addition the increase of safety of these of these plants, should have a significant effect on the required operation cost. The more efficient fuel burn of most of these reactors, also leads to improved operation because refueling is required far less often and can mostly be done online, improving the utilization.

So if you combine those advantages, I would consider all in these plants would make a hell of a lot more sense. And the reason they don't exist is partly regulatory but but those regulations and maybe more importantly the regulatory environment are reflection of the the overall turn against nuclear most of Western society has made. Some of the regulation we have today are over the top or unnecessary, but changing these would not instantly impact the price of the next PWR we build.

A nuclear regulatory agency and a country need to actually WANT nuclear innovation. As an example of such a process, look at what Canada did once they sold CANDU. They actually said, we want to innovated and be part of the next generation of nuclear and they changed their regulatory process and environment to reflect that.

https://nuclearsafety.gc.ca/eng/reactors/power-plants/pre-li...

Something like this could have happened already 40-50 years ago and in a much larger market like the US. But that was not the direction the government went in. In the US, there is clear pre-and-post Nuclear Regulatory Commission (1974) where before new reactors happened all the time, essentially no progress happened. That is not just the fault of the NRC, but also the broader changes in the state and society reflected in those changes.

In my opinion we should go back to building nuclear, for a country like mine, Switzerland, I do not see the extra cost of a nuclear plant as the big problem. We should be able to have energy independence with years worth of energy stored. Renewables are not great here, neither solar nor wind is really well suited to the situation. And if you really want to go 100% the storage requirements are very scary. The capital cost of nuclear can be paid off over literally the next 80-100 years while the plant is in operation.

I think the example of France shows that we could have had green energy for the last 40 years, and that Frances big mistake was to go away from nuclear (they actually started depending on gas). France was one of the few shining lights, and built amazing projects like Superphénix (https://en.wikipedia.org/wiki/Superph%C3%A9nix) but instead of building more, they were killed politically.

I predict in 100 years, nuclear fission will be what powers most of society. All in, from first principle, a very high energy density fuel will simply win the race in the long term. Its only a matter of building the right reactor, and turbine (or figure out aneutornic fusion).

So in summary you don't have evidence, but just make it up. Again the article argues (with evidence), that a fission plant would be more expensive than current wind/solar even if the reactor would be free. Do you have evidence to the contrary, hand-waving about future small reactors having the potential to be cheaper (also you argued it's regulation which is the driver of cost, how can then multiple small reactors be cheaper? Would they not be subject to the same regulations?

> I think the example of France shows that we could have had green energy for the last 40 years, and that Frances big mistake was to go away from nuclear (they actually started depending on gas). France was one of the few shining lights, and built amazing projects like Superphénix (https://en.wikipedia.org/wiki/Superph%C3%A9nix) but instead of building more, they were killed politically.

The reason they moved away from nuclear was economical not political. France is probably the most pro-nuclear nation in the west all major parties are strongly pro nuclear, even the french greens are ambivalent. The argument does not make sense.

I wouldn't think of wind and solar as competition to fusion. Fusion power would have a different role in future electricity grids. Both would be complementing each other, rather than being a replacement.

Solar and wind are comparatively cheap and quick to deploy and are great solutions for decentralised power generation. Wherever reliable continuous power is required, however, other methods of power generation have to step in. This could also still change of course, depending on how storage technology progresses. But the cheap super-battery is probably just as ambitious as fusion power is today.

They really are competitors, in the sense that intermittent sources, and fixed cost dominated sources, are both inflexible, and compete with each other on that basis. Cost optimized solutions for the grid typically contain a lot of one and little or none of the other.

The interesting point this author makes is that nobody is building a future electricity grid. There is nobody who can make such high-level choices about the mix of energy resources - choosing to allocate such-and-such here, thus-and-so there, to get an optimal blend - in the way that all of us armchair engineers tend to think about the problem.

What is actually happening, according to this author, is that bankers are evaluating specific energy projects, individually, and choosing to fund the ones which look like they will offer the best return on investment. That's it: there is no system. The future grid will consist of whatever happened to get built along the way, and our economy will adapt to the emergent characteristics of that "system".

There is no need for storage technology to "progress". It just needs to be built out.

It would be stupid to build out storage before you have built enough renewable generating capacity to charge it from: when renewables are not producing enough, you fill in with gas turbines, as now. As renewable generating capacity rises, and storage is built to take up the surplus, you rarely need to run the turbines. The turbines will eventually be converted to burn ammonia or hydrogen cracked from ammonia, when that becomes cheaper.

There is never a time when you have enough storage for all possible scenarios. Instead, you plan to fill in with other methods when storage runs low. Storage is a cost-cutting measure. Importing ammonia from tropical solar farms is always on the table.

> I wouldn't think of wind and solar as competition to fusion. Fusion power would have a different role in future electricity grids. Both would be complementing each other, rather than being a replacement.

How would that work? Solar is very cheap. Nuclear is several times more expensive. Ergo, if the sun is shining at all the economically rational decision is to buy as much solar as you can.

Same goes for wind.

This means renewables will cut into nuclear, a lot.

The whole "base load" ideal only works assuming nuclear is dirt cheap, but inflexible, and is supplemented with expensive but flexible sources for the times it's too slow moving. That would be the ideal scenario.

The unfortunate thing is that nuclear isn't cheap, and renewables aren't expensive.

The less time you are able to find buyers for your nukes' output, the more expensive that power becomes, in a vicious cycle. Soon no one will buy any.

When the nukes are mothballed, every kWh they ever produced becomes suddenly more expensive.

That renewables are not expensive is not unfortunate.

> That renewables are not expensive is not unfortunate.

I mean unfortunate for the intended business model of nuclear

I think we are not that far away from the "cheap super-battery": https://www.nextbigfuture.com/2022/10/catl-will-mass-produce... Does this qualify as "super-battery", you tell me, but it's pretty good and cheap and could easily scale for storage of solar and wind.

We don't need a "cheap super-battery". Most utility storage will not be chemical batteries.

Reply on news.ycombinator.com