Fusion Power is 5 years away! (Really?)

[Rockphed]

Okay, storage and density make sense. I'm frequently shocked by how wasteful we are with dams, like shipping electricity to an aluminum plant hundreds of miles away instead of putting them right on top of each other.

Where Dams go is pretty hard to change. While you lose some power in transmission, I'm not sure how much it actually is. Per this site, electricity in the US costs between 4.5 and 22 cents per kilowatt-hour. Assuming you aren't running your aluminum foundry in one of the high rate states, you are probably paying about 7.5 cents per kWh. It takes about 17,000 kWh to make a ton of aluminum, so if it costs more than $170 per cent saved in per kWh rate to ship the aluminum ore to the new smelter site and ship the aluminum from the smelter site, then it is better to not ship it. Even if you would save on costs that way, you might have other reasons for where you site your smelter, such as ease of transport, proximity to labor or major customers, or simply that you already own the land and have buildings built.

[UtopiaNext]

They expect to have it running in 5 years.

I haven't heard anything lately from Lockheed-Martin's Compact Fusion project, which, I think, stated in 2013 that they planned to have a working prototype in five years.

From what I understand, this is going to be a running experiment that is intended to "bridge" between the fusion experiments that have gone on so far and an actual productive fusion reactor. IIRC by 2035 they want to be ready for the next step, so everything before that is going to be trying to achieve the 50 mw in / 500 mw out figure they stated.

My general impression is that they don't really know exactly how this will pan out but figure that the effort will at least be useful. Odds are, I think, whatever they figure out should be added to the other fusion projects present and planned, so if we're lucky we might be able to actually make useful fusion reactors by 2035?

[Sigako]

If I learned anything about nuclear fusion in my life, it's a knowledge that fusion reactor is always 5 years away. It's called half-wait period (or something like that, I heard the term in Russian).

[monomer]

One of the questions I had years back is why hydrogen fuel cells never took off. Getting the hydrogen is extremely easy, you just split the water at a dam using the energy from the dam and then ship the hydrogen. So much energy leaks from the lines it wouldn't be much less efficient then what we do now.

I think mostly new energy just isn't interesting to the countries that can afford to explore it, and too expensive to the ones that can't.

Power transmission over long distances in the United Sates has losses of up to around 5% . Distribution adds up to another 5% (though you wouldn't be too worried about that if you are a major customer with dedicated high-voltage line), so in all you are generally looking at around 10% total losses.

Conversion of water to hydrogen via electrolysis is up to 80% efficient at an industrial plant, and around 70% at a smaller plant, so you are already starting at a disadvantage just from making the hydrogen.

For the end-user, in a smaller application like a car, low temperature proton-exchange membrane fuel cells are roughly 60% efficient at converting the hydrogen back to electricity, while larger / hotter industrial fuel cells are up to about 85% efficient with heat reclamation.

So all in all, you have about 10% loss from electricity distribution, while just the conversion from electricity>>hydrogen>>electricity you have at least a 32% loss, which doesn't even take into account the energy required for compression and transportation of the hydrogen.

[factotum]

So all in all, you have about 10% loss from electricity distribution, while just the conversion from electricity>>hydrogen>>electricity you have at least a 32% loss, which doesn't even take into account the energy required for compression and transportation of the hydrogen.

Efficiency isn't a major deciding factor for this sort of usage, though. For instance, the internal combustion engine in a regular car is maybe 25-30% efficient--burning that fuel in a car is by far and away the worst thing you could be doing with it, since burning it in a power plant would be a far more efficient use. However, a tankful of fuel has such a high energy density that it's still advantageous to do that compared to using batteries and an electric motor--even the longest range electric vehicles can't hope to match the range between top-ups that an ICE car can do. If you're using fuel cells in a mobile application it's for the advantages of range and (relatively) quick and easy refuelling, not because of the efficiency.

[Chronos]

Practical fusion has never been claimed to be five years away. The usual claim is 30, with steady and continuing funding and support (which it's never gotten). Without funding and support, the only possible prediction is and always has been "never".

Impractical fusion is here now, and has been for a very long time. The guy who invented television also made a device that fits on a tabletop and which produces fusion. You could build a fusion reactor in any reasonably well-stocked workshop.

Cold fusion is a hoax, and is irrelevant anyway. Even if what Pons and Fleischmann claimed were true, their device wasn't practical, either. And even if there is some real way to produce cold fusion, there's no reason whatsoever to believe it would be any more practical than hot fusion.

[snowblizz]

Where Dams go is pretty hard to change.
Even if you would save on costs that way, you might have other reasons for where you site your smelter, such as ease of transport, proximity to labor or major customers, or simply that you already own the land and have buildings built.

Not to mention the places good for dams are unlikely to be ideal for large industrial sites and transportation networks. Basically you can build the transmission lines almost over any terrain. But try building the road or rail need to ship bulky bauxite ore. A dam can run with a lean staff compared to an aluminium smelting-plant.

With the low cost of "free" electricity the other costs of the equation matter a lot in how it shakes out. So e.g. it makes sense to ship ore to Iceland to smelt it.

Sometimes other whacky considerations like environment and such actually matter too.

One of the questions I had years back is why hydrogen fuel cells never took off. Getting the hydrogen is extremely easy, you just split the water at a dam using the energy from the dam and then ship the hydrogen. So much energy leaks from the lines it wouldn't be much less efficient then what we do now.

I think mostly new energy just isn't interesting to the countries that can afford to explore it, and too expensive to the ones that can't.

I think the main problem stems from needing a completely new secondary infrastructure to support hydrogen fuelcells. And as others have noted it is much more inconvenient to store. Or put a bit hyperbolically do we want every car, every truck and every gas-station be a potential Hindenburg? In comparison take that power you didn't use to create motive bombs and put it into electric cars where you can actually even refuse existing infrastructure, and probably 90-95% of what you need already exists.
Basically, all vehicles need to be replaced already. But do we also need to completely replace all logistics and transportation infrastructure too? Hydrogen fuel-cells suffer from massive bootstrap problem the alternatives doesn't quite have.

[Rockphed]

I think the main problem stems from needing a completely new secondary infrastructure to support hydrogen fuelcells. And as others have noted it is much more inconvenient to store. Or put a bit hyperbolically do we want every car, every truck and every gas-station be a potential Hindenburg? In comparison take that power you didn't use to create motive bombs and put it into electric cars where you can actually even refuse existing infrastructure, and probably 90-95% of what you need already exists.
Basically, all vehicles need to be replaced already. But do we also need to completely replace all logistics and transportation infrastructure too? Hydrogen fuel-cells suffer from massive bootstrap problem the alternatives doesn't quite have.

I suspect that even running a bioreactor to produce gasoline or diesel under artificial light would work better than using fuel cells due to hydrogen's really hard storage requirements.

One fuel technology that hasn't happened and I wonder why is cars using liquified natural gas. I hear about cities using it for their garbage trucks, but haven't seen anything about cars running it. Is it just a different enough engine and tank design that cars can't, or is it that nobody sells the stuff?

[Lord Torath]

Cold fusion is a hoax, and is irrelevant anyway. Even if what Pons and Fleischmann claimed were true, their device wasn't practical, either. And even if there is some real way to produce cold fusion, there's no reason whatsoever to believe it would be any more practical than hot fusion.

There's another type of Cold Fusion, but it's not useful for generating power: Minute Phsysics: Cold Fusion

Not a hoax. Just not useful for generating power, as you only get out about 1/2 the power you put in, assuming 100% transfer/collection efficiency.

[snowblizz]

I suspect that even running a bioreactor to produce gasoline or diesel under artificial light would work better than using fuel cells due to hydrogen's really hard storage requirements.

One fuel technology that hasn't happened and I wonder why is cars using liquified natural gas. I hear about cities using it for their garbage trucks, but haven't seen anything about cars running it. Is it just a different enough engine and tank design that cars can't, or is it that nobody sells the stuff?

A bit column A, bit B I think. Infrastructure's a female dog. They couple years ago commissioned a cruiseliner using LNG and it required a new everything (including crew). Including a pipeline to a refueling plant. LNG is I think easier than hydrogen to store but not by a lot. They wanted a refuelling infrastructure on the other end of it's trip too but that got a firm NIMBY response. You still need the new infrastructure but just switched to a less flexible fossil fuel. Without looking it up I think simply put LNG works better at scale. Like why coal powered ship and train works but not really a car.

[factotum]

You can get cars converted to run on LPG. I believe there are a few cold start issues with such an engine, but the major problem is obviously that LPG is just another fraction distilled from crude oil and therefore isn't a solution to the problem of running out of fossil fuels, so nobody's really pushed hard to replace existing petrol and diesel infrastructure with it.

[Mastikator]

I was thinking how much loss of hydrogen if we powered the earth with 0-energy input fusion compared to how quickly we lose hydrogen to space. Also there was a book where people had settled mars and Earth had decided to stop allowing them to take water for fuel and to terraform mars.

Even if we somehow replaced literally all energy sources with fusion that uses oceanic water the sea level would continue to rise for the next thousand years or so. The CO2 in the atmosphere is incredibly high and the global temperature is slowly catching up.

If you really want to lower the sea level then you have to drop a gigantic ice cube into the ocean. Thus solving the problem once and for all!

[Telok]

One fuel technology that hasn't happened and I wonder why is cars using liquified natural gas. I hear about cities using it for their garbage trucks, but haven't seen anything about cars running it. Is it just a different enough engine and tank design that cars can't, or is it that nobody sells the stuff?

The first and last LNG car I saw/used was almost 30 years ago on a US military base. It was a rental car in Oklahoma City. I recall that it turned out there was exactly one LNG fuel station in the entire state, on the military base.

I suppose it could have been worth it if they had a problem with lots of people running off with rental cars. At the time it just seemed weird.

[Chronos]

You can get a car that runs on liquid propane, but I'm not sure why you would. It's only slightly better than gasoline on the greenhouse gas front, but significantly worse on all of the other pollutants, costs more, and requires a more difficult fuel tank that takes longer to refill, with fewer filling stations available. You might be able to get tax credits or the like for an "alternative fuel vehicle", depending on jurisdiction, but if so, that's just a sign of poorly-worded laws.

Liquefied methane (natural gas) requires an even more difficult fuel tank and takes similarly long to refill, but since many houses already have natural gas piped in, you can install a home fueling station. It also has nonzero but significantly less greenhouse gas emissions (as long as you don't have any leaks; methane itself is a much worse GHG than carbon dioxide), and also less emissions of all of the other pollutants. Currently it's mostly only practical for large centralized fleets like city garbage trucks or city buses, but (if you value the clean air), it is practical for those.

[snowblizz]

Liquefied methane (natural gas) requires an even more difficult fuel tank and takes similarly long to refill, but since many houses already have natural gas piped in, you can install a home fueling station. It also has nonzero but significantly less greenhouse gas emissions (as long as you don't have any leaks; methane itself is a much worse GHG than carbon dioxide), and also less emissions of all of the other pollutants. Currently it's mostly only practical for large centralized fleets like city garbage trucks or city buses, but (if you value the clean air), it is practical for those.

The irony is that the large centralised fleet is one the applications electric vehicles are best at. Even less emissions at point of use, and you can then limited the infrastructure to a centralised plant that just makes electricity.

[tyckspoon]

Liquefied methane (natural gas) requires an even more difficult fuel tank and takes similarly long to refill, but since many houses already have natural gas piped in, you can install a home fueling station. It also has nonzero but significantly less greenhouse gas emissions (as long as you don't have any leaks; methane itself is a much worse GHG than carbon dioxide), and also less emissions of all of the other pollutants. Currently it's mostly only practical for large centralized fleets like city garbage trucks or city buses, but (if you value the clean air), it is practical for those.

The best argument for using LNG is probably just that it provides a market for the stuff, which means gas and oil producers have a reason to collect it, refine it if needed, and transport it.. instead of just burning it off in the field. A lot of natural gas is pumped as a byproduct of drilling and pumping for oil, and if there's no monetary value to it it often just gets flared off on-site as a waste product. If it goes into vehicles, power plants, or even home usage, then it can be usefully burnt to provide energy and have its combustion products run through filters and converters to clean them up some.

[Xyril]

Is this really better than harnessing the (fusion) power of the Sun? I'm not against the research per se, but this has to be a lot of money, and I wonder if investing in photovoltaic research would achieve similar results. Although I guess that anyone could invest in solar panels, while this kind of mega-project requires supranational collaboration.

Define "better."

The first fundamental limit of solar energy is power flux. We can keep making strides in improving efficiency by improving conversion efficiency per area of solar panel, and panel utilization per area of space, and in that regard, there's still a lot of gains to be made, but at some point the only way we can meaningfully scale up is by thinking beyond terrestrial solar panels.


I'm actually in favor of this, since it synergizes well with developing space travel, but from an investment standpoint, developing and deploying space-based solar solutions is probably on the same order of magnitude as developing practical fusion power.


Right now, practical solar panels are about 20% efficient. We can probably get a bit higher by further optimizing the engineering solutions (i.e., finding ways to further reduce parasitic losses without drastically increasing production costs), but the fundamental source of inefficiency comes from the initial capture of light. The first step in a solar panel is the absorption of the photon--when light hits the semiconductor, excites an electron, and hopefully we're able to direct that electron somewhere that it can do useful work before the whole thing returns to an unexcited state. The efficiency of this process--what percentage of electrons get absorbed and excite something, rather than passing through without interacting, and what percentage of those excited electrons do something useful before returning to a low energy state--is largely a material property of semiconductors.

We can tweak these properties by doping--adding impurities--which is why we can get so much interesting behavior out of devices that are 99% silicon, but in terms of tweaking them to be more efficient light absorbers, we've pretty much hit a wall with silicon. This is important because silicon's pretty much the cheapest bulk semiconductor material we have available. They've actually gotten much higher efficiencies using other semiconductors--I want to say indium-gallium was one of the more promising ones--but these materials are currently more expensive to process and to integrate into working cells.

I absolutely believe there is a way forward in terms of improving solar, but the point is that we're at the point where we have to characterize and experiment with new materials, and eventually find a way to integrate these new materials with manufacturing processes, and to hopefully scale up without drastically increasing costs. In other words, we're getting very close to having the same situation as with fusion--where the fundamental science is novel enough that we really can't predict how much money and effort we'll have to sink in to get results, or that we'll even get significant progress within a certain time frame.

And even if we do, we don't know for sure what our energy needs will be in a few decades. It's possible that solar cells at 30-40% efficiency will meet all of our needs, without having to cram the Earth so full of them that we start disrupting the biologically or meteorological processes that also depend on solar energy. However, it's also possible that our power needs will grow so much that only a compact solution--something like fusion, fission--will be viable. What if, for example, we manage to exhaust our supply of ground water? Desalination is often the thing that turns a green boat into one that has to rely on gas generators for long voyages.

[snowblizz]

And even if we do, we don't know for sure what our energy needs will be in a few decades. It's possible that solar cells at 30-40% efficiency will meet all of our needs, without having to cram the Earth so full of them that we start disrupting the biologically or meteorological processes that also depend on solar energy.

I'm going out on the limb and say it never will. Because large parts of the planet does not get enough sunlight all the time. And at some point we hit limits on storing and transferring solar power from places that do.

And there are other issues, am told in some places in Germany where they got a lot of solarpower in towns on the roofs a local cloudformation can disrupt the grid as it shields a significant portion of local power. We get all of these kinds of issues with only one or even few numbers of means of producing. Every time I consider a solar and wind only world e.g. I wonder what we will do a sunny day in early February when it's still. The sun may shine but it only produces a tiny fraction of energy compared to other parts of the year. There's a stagnant highpressure so no wind either. And it'll be -20 C because sunny means cold in February.

That's when you need something else.

However, it's also possible that our power needs will grow so much that only a compact solution--something like fusion, fission--will be viable.

We are likely going to be growing our use of power to match our ability to produce it for a long time still. Fusion and fission provides something that wind and solar can't easily do, which is built in redundancy where you ramp up production for peaks. There's nothing that powergrids hate as much as power that fluctuates. The problem lies in providing constant effect for the grid, how exactly the input is made doens't matter as much as that the input can be controlled or predicted.

The optimum solution probably lies somewhere in setup where fusion plants run at 30-40% capacity and the rest is covered by solar and wind (and others) but where fusion can cover 100% of the need if it has to. If we run fusion on a rare isotope, even though there migth be quite a lot of it, makes sense not to "waste" it and use as much "free" power as we can.