Hacker News Reader

Comment by anovikov | Hacker News Reader

Why doesn't anyone see the elephant in the room: time for fusion has run out, renewables have won. There is no way in the world fusion will ever be cheaper than current costs of solar + LiFePo battery power. And as well no chance commercially proven fusion will exist even on a minimal scale before electricity needs are fully covered by renewables. Continuing fusion research is important to keep the scientific process going, it will surely result in a lot of byproduct discoveries in physics, material science, and a lot more, but the lie about fusion-derived electricity as being a viable and/or necessary thing for humankind, should stop.

View on HN

That is a failure of imagination, like the guy who said let's close the patent office in the 1800s because everything had been invented. We have no idea what such a surplus of energy would make possible that we don't even have the imagination for today.

If we want such a surplus of energy, at this point it's pretty clear Fusion isn't the way to get it in even the medium term, and maybe even not remotely affordably in the long term. We've already sunk 4 or 5 Manhattan projects' worth of resources into Fusion with essentially nothing to show for it. How many more Manhattans is it going to take?

I'm not against pursuing some of the modern, safety first fission technologies. They may well have a place. Maybe Fusion will eventually become viable, but I think it's very clear now that our technological reach is way longer than our grasp. We poured way too many resources into it long before we actually had any reasonable chance of success. We'd be better off dropping the big super expensive projects that aren't going anywhere, and continue with the many smaller basic technology development projects until it becomes a lot clearer if any of those is going to pan out.

Failure of imagination indeed. There's much more to energy than the regular usage cycle of the average electrical supply grid.

What if creating easily-fusable isotopes is the most efficient way to store "renewable" energy? Fusion is the most concentrated release of power humans are able to muster. What's to say such concentrated power will never be required for any purpose?

That's totally right, fusion might have it's uses in the future. For example, the goal it was originally intended to be used when first tokamaks were created: cheap production of large amounts of tritium, enabling more compact thermonuclear bombs producing much less fallout with very flexible yields. Just not electricity production.

Where will surplus of energy come from? Even if fusion reactors will become mass-produced and cheap, well, it's easy to find a baseline there: they will never be cheaper than a coal boiler producing same steam energy, that's easy to see. Which means, because the rest of the system - turbines, generators, heat exchangers, condensers, steam separators, water treatment facilities, transformers, safety systems etc. - remains the same - a fusion power plant is the same thing as a coal power plant running on free coal.

If you take cost of coal-produced electricity and subtract the cost of coal that went into it, it's still more than current costs of solar with storage.

Renewables have currently very difficult problems to solve with storage.

As far as I know the only sufficiently scalable energy storage method is pumped hydro. Storing enough energy to deal with the intermittency of renewables at the scale required involves essentially digging a very large number of very large reservoirs. This is a huge civil engineering undertaking with massive environmental issues of its own. I've not seen anyone in any government really talking about investing in that much storage.

Batteries do not seem to be realistic for the near and medium term (and also come with massive environmental problems), pumped air works where you have appropriate geology, everything else is essentially a meme.

Renewables are great as long as you ignore the storage problem, just as coal is great as long as you ignore the waste problem.

> everything else is essentially a meme.

I don't think this is fair: even if something like hydrogen storage isn't practical or economic now, it might be if generation cost falls 3x or 10x. Which is plausible for renewables in the medium to long term.

> Renewables have currently very difficult problems to solve with storage.

And fusion does nothing to solve these problems.

Even in an optimistic scenario, what you'd end up with fusion is a very expensive power plant that can generate large amounts of energy. Running that as a peaker plant to complement renewables is unlikely to be economically viable.

So your only bet really is to have fusion running as baseload power. However there won't be baseload power in the future, because we'll have times where renewables will provide more than 100% of the energy needed.

I really don't see how fusion fits into any of this. SUre, intermittency problems of renewables need to be solved. But Fusion ain't the solution. Look at hydrogen-fired power plants, advanced geothermal, or heat storage, those are promising options.

Fusion doesn't have those problems. With fusion, you generate electricity 24h/day. Storage is needed for solar and wind for obvious reasons.

That's under the assumption you'll have a fusion only grid. But fusion won't exist in a vacuum. It will exist within the energy system in 20+ years. That will in all likelyhood be dominated by wind and solar, yet with another 20+ years of improvement.

Battery storage works just fine for intraday use.

And you don't even need batteries for industrial heat, which is a very substantial fraction of total energy demand. You can heat up a box of rocks/bricks/sand/graphite with resistive wires and run water through pipes to get the heat out again.

And you don't need to dig out massive reservoirs for pumped hydro in much of the world. You just need a decent-size mountain range somewhere convenient to your electricity grid. Much of the time you can repurpose conventional hydroelectric dams for the job.

Another alternative to storage is interruptible demand. If you're making hydrogen with electrolysis (another potential alternative energy storage medium) you can just turn the electrolysers off when electricity prices are high. There are lots of studies and while the more time you're using your electrolyser the better, the economics of running them intermittently are likely to be quite reasonable.

Finally, it's worth pointing out that most people in the developed world seem likely they will have several day's worth of home electricity usage parked in their driveways by 2035 or so. Tapping just a small amount of that will make a huge dent in dunkelflautes.

As a Brit with solar panels, in June I get 15x the kWh output that I get in December. Which is the month with the highest demand for natural gas, for heating.

So I think we need more than just intraday storage.

Yes, solar power has a huge amount of seasonal variation in the UK and the rest of northern Europe.

And obviously, batteries on their own won't cut it to deal with this seasonal variation.

But northern Europe, particularly the UK, has a lot less solar and a lot more wind in its grids than Spain, California, or Australia, for that very reason.

Beyond that, there are quite a few other current or near-term technologies you can use to deal with the seasonality of supply and demand. First, insulating its houses properly to reduce the winter demand peak. Second, switching to heat pumps is another. Third, expanding biogas production and using it specifically for the winter peak.

In the long term, producing hydrogen in the summer and storing it for the winter seems like the most likely solution for 100 or near-100% renewable grids in northern Europe at this point.

If you want to use wind as baseload in the UK, you need enough storage to go through a wind drought which lasts for weeks.

On top of that, the UK currently largely uses piped gas for space and water heating so unfortunately moving to heat pumps and EVs makes the problem even worse since both increase demand for electricity further (even if they net reduce energy demand/use).

I don't think it's unreasonable to conclude that nuclear is likely to be a significantly cheaper solution when factoring in total system costs, especially if you can build a whole series of nuclear plants and benefit from economies of scale.

Another solution is to simply install 15x solar panels and switch off some when there is too much power. Because a panel currently goes for 11 cents per watt and there's almost no more expenses involved with overbuilding - no inverters, transformers or grid connection as you just switch off them on DC panel output point if there's too much power - it comes dirt cheap.

You can do that kind of thing in the US, but land use is a real constraint in the UK.

The politics of land use in the UK are also diabolical.

You could put them in the Sahara and run a cable https://en.wikipedia.org/wiki/Xlinks_Morocco%E2%80%93UK_Powe...

Battery storage is very cheap already with LiFePo batteries being at about $70/kwh with 80% cycle life of 3000 cycles. That's ~2.6 cents per kwh of energy stored. With invertors, grid connections, buildings etc. (that don't wear out nearly as quickly), it will capitalise to maybe about 6 cents per kwh for multiple hour storage. Because only a small fraction of energy needs to go through storage, it will make a really small impact on overall electricity cost.

That's false. Renewables have not won. See Germany to understand why renewables can not replace stable energy sources. They burn lignite now and buy French nuclear energy to offset nuclear.

Renewables have two major issues that need to be solved: one is recycling, and the other one is storage. Nuclear is the only real "renewable" energy source worth pursuing until fusion is achieved.

Well, in EU overall, we went to only 1/3 of electricity being produced with fossil fuels. I think it pretty much proves that fossil fuels aren't necessary. We are going to be down to under 10% in less than 10 years, using the already existing technologies and with no assumption of further cost declines: just with incremental buildout.

Well maybe but on the other hand we have

>Helion announces world’s first fusion energy purchase agreement with Microsoft New facility aims to deliver at least 50 MW and begin producing electricity by 2028 https://www.helionenergy.com/articles/helion-announces-world...

I mean that may be BS, but renewables have not exactly stopped fossil fuel use yet so why not try both approaches?

> than current costs of solar + LiFePo battery power

As far as I know there's no way yet to set that up at scale, meaning to be able to provide power to tens of thousands to hundreds of thousands of homes (and more) based on power stored inside of batteries. All I've seen are some gimmicks set up/installed in some Western suburbia houses.

Pumped storage seems to fit the need pretty well, in that it’s trivially easy to build and there are hundreds of thousands of potential sites in the US alone.

"trivially easy to build and there are hundreds of thousands of potential sites in the US alone."

Yeah, but plenty of people don't live in the US and their geography is much less suitable for pumped storage than the American (or Norwegian) one.

For example, Finland is very flat, so are the Baltic states and most of Poland. So is Bangladesh, which has three times as many people as all the above together, and a nascent industry that needs energy.

In Czechia, we have rolling hills that are suitable for some pumped storage, but we built up almost all the good locations. In one case, an entire top of a mountain was cut off in order to make a reservoir:

https://cs.wikipedia.org/wiki/Dlouh%C3%A9_str%C3%A1n%C4%9B_(...

I have a vague memory that someone suggested turning a significant fraction of what is now Czechia into a single huge hydro lake — probably one of the many mad things suggested in the inter-war period — but I can't remember the name now to search for it.

At least we could have some mean naval battles. Our landlocked history is severely lacking those.

Where the geography allows, pumped storage should be expanded.

There was an interesting proposal for a massive pumped storage Strathdearn[0]

It has the capacity of about 6800 gigwatt hours (283 gigawatt days) which is about 1000 times the existing scottish current pumped storage in Cruachen.

The article claims to cover the existing daily power requirements of UK and be able to export to europe.

I am not a civil engineer or geologist so it might be too big, not practical and too expensive. lots of smaller schemes might be better, but an interesting read

[0] https://scottishscientist.wordpress.com/2015/04/15/worlds-bi...

What hogwash. People crave security more than they do anything else and it's completely toxic.

Space travel can use infinite energy.

Fusion works underground where many renewables do not.

Infinite energy can be used for climate control of the whole planet.

Fusion plants can fuel megafactories even in war conditions.

Fusion can enable more energy expensive technologies like hydrogen batteries, to be trivially affordable.

I'm starting to find the Western malaise that has come around this decade, to be offensive.

It's like the knees have given out and people are just begging for a comfortable death. Come on! We're better than this!

Where does this idea that fusion == infinite energy come from? Nothing could be further from the truth, all fusion power plant designs have a very fixed amount of power they can generate (limited by the size of the reactor), and the costs to build and maintain one are gigantic, in ways that are unlikely to go down in the next century or so. Scaling them up (building bigger reactors) is extremely hard because of fundamental geometric reasons, and scaling horizontally (many many power plants) is hard because they require teams of experts to run and monitor.

And pretty much the whole plant, with all the expensive magnets, supports, and reactor chamber, has a lifespan of maybe 10-15 years before the constant neutron bombardment makes the materials too brittle to keep supporting their weight. You then need robots to come in and dismantle the highly radioactive components and store them securely for a few 1-200 years while they cool off.

Sure, it's nice that fusion requires relatively little fuel (though even that is not that great, since it requires an extremely rare substance, tritium), but that doesn't mean it's in any way going to scale up infinitely, or even a lot, or even as much as nuclear fission.

I said infinite energy, because that is exactly the phrase required to break people out of the limited mould that traps minds into erroneous thinking.

Neutrons are a problem in JET/Tokamaks. They are less of a problem in other fusion designs.

I feel like I'm the last person on this planet who actually enjoys being uncertain, and uncertain enough to see the big picture.

Most of the problems you have described are getting solved piece by piece in small scale projects, and then the evidence collected from their successes will be combined into a more effective solution that you are willing to discuss.

I'm so tired of the constant emotional vortex sucking the intellect out of the West. Get off my back.

I think it all depends who you believe about how batteries will develop and the consequences of manufacturing them.

On the other hand from what I understand fusion isn't all that clean itself.

The problem is that it's impossible to separate the pr from reality; only time will do that.

I don't think we'll ever have enough energy, we'll always find ways to use more, on earth or elsewhere.

I am pretty sure we'll run into the problem of too much heat long before we need more power.

Fusion will become usable only elsewhere, not on Earth.

On Earth, fusion does not make sense unless it would provide more energy than the energy received from the Sun.

However, if an amount of energy comparable to that received from the Sun would be produced by fusion, after being used, that energy would become heat, causing an even more dramatic climate change than what we are facing now.

Fusion energy could be very useful, but only on spaceships or on planets/satellites/asteroids that are far from the Sun, and never on Earth.

The same is true for nuclear fission energy. In the short term, it can be useful to replace power plants that use fossil fuels with nuclear fission reactors, but the total amount of energy produced by nuclear reactions can never become so great as what can be obtained by capturing directly or indirectly solar energy, without causing climate changes.

The most important research direction should be for efficient methods of long-term high-capacity energy storage, e.g. in synthetic hydrocarbons or in flow batteries, and not in nuclear energy.

Elephant in the room is renewables can't provide the same security/constant power as nuclear/fusion. Even more, I'm not sure it's even possible to cover all our rising needs with renewables only in combo with batteries and hydropower, especially when accounting that not all areas are good for renewables. And there are different challenges too like space needed to provide that amount of energy+storage, managing when that equipment should be replaced and so on. I'm genuinely happy France managed to perform needed maintainance on their nuclear plants so that (in theory) they will import much less from Germany in the winter, or potentially even export and also happy they succeeded to push nuclear as net zero technology

The sun is on 100% of the time.

The only reason that renewables can't provide us with reliable power is because of our human political problems: we can't reliably cable up the always-sunny areas of the planet to the areas where people live, because the governments in the always-sunny areas are unstable.

Well, it will also cost an estimated £16 billion, to build two high capacity electricity transmission lines from Morocco to the UK. You can build a lot of capacity locally for that price.

Or a tiny portion of a railway! Do you have a citation for that 16 billion?

It's the Xlinks project.

This is where I got the 16bn figure: <https://www.investmentmonitor.ai/sectors/energy/uk-morocco-s...>. This source says £20 billion, but half is for the cable, and the other half the solar farm and other expenses: <https://sifted.eu/articles/xlinks-morocco-uk-electricity-cab...>.

Reporting is a bit inconsistent. My point is, transmitting electricity extremely long distances has its own downsides (construction costs & power losses), even if we assume Morocco gets a fair deal out of this.

And this project is "only" 7% of UK electricity supply. What of the rest? What of other countries? Saying "export" and "import" doesn't solve all the problems of intermittency (unfortunately).

Where's always sunny 24hr/day all year around? Or how much storage do we need to cover the night use? How much will it cost to build all this + transmission cables for all this power without big energy loss?

> Where's always sunny 24hr/day all year around?

The aggregate combination of the Sahara, the Gobi, and the Great Australian desert, Arizona, the Namib, the Atacama Desert.

> How much will it cost to build all this + transmission cables for all this power without big energy loss?

For a 1Ω resistance over the entire grid, assuming it was made of aluminium, 40,000 km long, material cost is about $239 billion.

(But on this scale the number is essentially imaginary, as the real answer is "how much do governments value the opportunity cost of the other things they could get people to do other than make a factory to make aluminium").

> renewables have won.

As evidenced by the fact that they have completely replaced fossil fuels and already sustain multiple grids in flat lands 100% now.

Yeah, for grid energy applications in our lifetimes it's hard to see fusion beating a grid made up of solar, wind, lithium-ion and other battery chemistries, topped up with some chemical storage (eg hydrogen) for places with cold winters.

However, in the long term some kind of nuclear propulsion is the only realistic way to open up the outer solar system, and I'd prefer our descendents had a safer option than "nuclear pulse propulsion" - the physically plausible but very 1950s idea of pushing a spacecraft along with nuclear detonations...

"current costs of solar + LiFePo battery power"

At what scale? Can you even build enough batteries to supply, say, entire Europe for 1 hour?

> At what scale? Can you even build enough batteries to supply, say, entire Europe for 1 hour?

Personally? No.

As a species? Sure, why not.

Looks like China makes most of them, but looking at this graph, the world total battery production last year was 5-EU-hours:

https://www.iea.org/data-and-statistics/charts/lithium-ion-b...

https://www.wolframalpha.com/input?i=2%2C753%2C320GWh%2F1yea...

With significant production increases forecast. Those things will probably last in the order of ten years before needing replacement/refurbishment/recycling, which would mean around 50-EU-hours steady-state capacity today

Worldwide is more interesting than just the EU. Current production is global production per year of storage sufficient for storing 30m of global electricity demand, with the same 10-year assumption that's a steady-state level of 5 hours; assuming the forecast growth is correct, the annual production in 2030 rises to 6.79 TWh/year, which is storage for 2.33 hours of global electricity demand made in that year or 23.3 hours with the same 10-year-steady-state assumption.

Interesting, thank you.

The graph that you linked to is interesting in more than one way. If/when China decides to take Taiwan by force, the democratic world is going to be screwed.

We've already had a massive gas supply shock in Europe when Russia attacked Ukraine. We were able to overcome it, but it required kowtowing to disgusting regimes such as Qatar and Azerbaijan.

Sure, batteries last longer than gas. An immediate rupture in supply wouldn't be catastrophic, unless the Chinese vendors can somehow turn off/disable batteries remotely. But just replacing the ones that fail would probably overwhelm our current manufacturing capability.

We've slept on an eco-bio-financial cushion too long.

Money and price is a really poor information vessel for this kind of decision making, as they are disconnected from the actual physical flows involved.

Renewables require a lot of minerals compared to any centralized energy production facility, as they are really diffuse, and also they compete with the surface we need to grow our food and live. This is absolutely not present in current prices.

> they compete with the surface we need to grow our food and live

They can, but they don't need to. Rooftops, extra shade for car parks (or walkways), deserts, gaps between directions of the same road[0]… in principal, if we were solving this together at a global scale rather than a bunch of local competing interests, we can supply well over 100% of current demand just by PV without needing any farm land to be used in the process.

[0] https://www.google.com/maps/@39.740634,-119.0682473,2221m/da...

Reply on news.ycombinator.com