What are disadvantages of nuclear fusion?

[Javster]

Finite doesn't automaticly mean we can spend it all.

Sure hydrogen is finite, but there is so much of it, for practical purposes in this case, it might aswel be infinite

Same as crude oil: technically it's renewable, but we use far too much for the extremely slow generation of it to keep up.

[Nuke]

As for running out of hydrogen, even tritium, not an issue. heat pollution will be an larger problem.

i would just have centralized hot water near fusion plants. that heat still winds up in the environment, but at least you are putting it to good use before it gets there, and the people using it wont need to use so much electricity to heat water.

[Aghanim]

A small trivia: Apparently in the new X-Men, the plane that the characters use was powered by a small tokamak

[phoenix_ca]

A small trivia: Apparently in the new X-Men, the plane that the characters use was powered by a small tokamak

There's no such thing as a "small tokamak".

[Red Iron Crown]

There's no such thing as a "small tokamak".

In fairness, there's also no such thing as "X-Men".

[Seret]

i would just have centralized hot water near fusion plants. that heat still winds up in the environment, but at least you are putting it to good use before it gets there, and the people using it wont need to use so much electricity to heat water.

I'm a big fan of district heating from CHP, but the companies that own the power plants are often less keen. The problem is that to provide useful hot water for district heating they have to run the cool side of their condensers at a higher temperature than they would like. This reduces the electricity generating efficiency of the plant. Since electricity is a more lucrative commodity than heat this hits their bottom line. To counter that the government would have to intervene in the market and make sure they were getting an inflated price for the heat they provide. The bottom line is you tend to only see large-scale use of district heating schemes in countries that are happy to take that kind of interventionist central planning approach (Scandinavia, former Soviet countries, etc).

[Nuke]

in a fusion reactor you dont need water that is as readily available in a nuclear reactor, since the reactor can be shut down quickly if there is a shortage. so its probibly acceptable to source cold intake water from the normal utility. the hot end of the heat exchanger (where the heat exchanger is the sink side of a brayton cycle) goes out for public consumption as hot water/heat. by the time the water goes through the consumers, the water treatment plant, and back to the reactor, it is cold again. the hot water would mostly go to industrial centers built around the power plants for basic applications like heating and manufacturing.

[Aghanim]

Once through cooling? That will work, especially if we use direct conversion reactor instead of tokamak.

[m4v]

That's a little optimistic. Electronics grade silicon pays back it's embodied energy in about 8-10 years. It doesn't really matter if it's poly- or mono-crystalline, the former has lower embodied energy but is also less efficient so it pretty much evens out.

You are however correct that lifetimes are very long and reliability is high. Failures of panels over 20-30 year timescales have been shown to be very low (around 5%) and your 40-year life isn't unreasonable. The weak part of a PV array is the power electronics, these last about 5-10 years, but these represent only a tiny fraction of the array's embodied energy anyway.

Is not optimistic, solar panels don't use electronic grade silicon, they don't need to be %99.9999999 pure, is also a fact (see 7th slide). And that's only for solar panels installed in Europe, ie, not the place of the planet with the best irradiance.

IMO the weakest part of PV is the energy storage systems, batteries are really expensive and barely last 5 years.

Edited June 1, 2014 by m4v

[NERVAfan]

IMO the weakest part of PV is the energy storage systems, batteries are really expensive and barely last 5 years.

Yep. You could probably cover thousands of square miles of Nevada desert with solar panels and power California and Salt Lake City and Tucson etc., but that's a LOT of batteries.

Maybe Tesla Motors will make batteries cheaper eventually... IIRC they're going to push mass production of lithium-ion batteries with a "Gigafactory".

[WinkAllKerb'']

notice that some energy source are not really weather friendly, un/intended disaster friendly, recycling friendly, rentability, linked job and so on family, ... relativise with time shake a little bit and then => a x b x c x ... x t = ?

Edited June 1, 2014 by WinkAllKerb''
i miss existing caracters for this equation GL.

[Nuke]

some wavelengths of light can penetrate cloud cover better than others. couldn't you optimize your panels to operate in those spectra? this may require materials other than silicon though, sort of like how different color leds use different semiconductors to produce different wavelengths of light.

[phoenix_ca]

some wavelengths of light can penetrate cloud cover better than others. couldn't you optimize your panels to operate in those spectra? this may require materials other than silicon though, sort of like how different color leds use different semiconductors to produce different wavelengths of light.

Yes, but that doesn't mean you'll get useful levels of energy or a bandgap (in fact using photovoltaics by its nature limits you to certain wavelengths because you are limited to wavelengths that will cause an electron to jump bands/orbitals) that's large enough to be of practical use. We already have photovoltaic detectors for various wavelengths of photons, e.g. infrared cameras, x-ray detectors/telescopes, gamma ray detectors/telescopes, radio telescopes and detectors, and so on and so on.

Thin-film silicon photovoltaic cells already effectively use infrared and visible light (well, large parts of them anyway).

Edited June 2, 2014 by phoenix_ca

[m4v]

some wavelengths of light can penetrate cloud cover better than others. couldn't you optimize your panels to operate in those spectra? this may require materials other than silicon though, sort of like how different color leds use different semiconductors to produce different wavelengths of light.

Visible light is the spectrum that reaches the surface with most irradiance, peaks around the green, silicon solar panels doesn't quite match that curve (they work best in red/infrarred) but work with visible light mostly. So there's room for improvement, but since silicon panels are the most cheap and easy to manufacture, plus non-toxic, I don't think that will change for a while.

[Nuke]

yea some of the things used in leds are rather nasty, things like arsenic and germanium. one common hack is to use leds as light detectors by reverse biasing them, they will allow current to pass when exposed to whatever wavelength they produce when powered normally. detection is completely different from energy production though.

[Seret]

Is not optimistic, solar panels don't use electronic grade silicon, they don't need to be %99.9999999 pure, is also a fact (see 7th slide). And that's only for solar panels installed in Europe, ie, not the place of the planet with the best irradiance.

They are electronics grade silicon, but I stand corrected about the payback time. Fraunhofer Institute is a credible source.

IMO the weakest part of PV is the energy storage systems, batteries are really expensive and barely last 5 years.

Very few PV systems use batteries. The vast majority of PV is grid-tied, they use the grid as "storage", exporting when there's an excess and importing to make up any shortfall. There's demand for practical energy storage systems for on-grid, but the bottom line is that none of the available technologies are cheaper than just using the grid. Some people do store their excess PV as heat, diverting it to an immersion heater for DHW. It's really only folks that are completely off-grid that use batteries.

[Aghanim]

The small problem with grid tied systems is that they could create excess of electricity in the grid, and driving the prices crazy, without anywhere to store it... However there are solutions to that, like hydrogen generator, molten salt batteries, pumped hydroelectricity, but in the end, we desperately need some sort of energy storage to smooth out our intermittent energy source. Probably if every countries are connected together, that will solve the problem

[PakledHostage]

Very few PV systems use batteries. The vast majority of PV is grid-tied, they use the grid as "storage", exporting when there's an excess and importing to make up any shortfall. There's demand for practical energy storage systems for on-grid, but the bottom line is that none of the available technologies are cheaper than just using the grid.

That may yet change. I recently heard an interesting story about organic flow batteries on the CBC's "Quirks and Quarks" science magazine radio program. It seems that a research group at Harvard University has made significant progress in using quinones in flow batteries. And since increasing the storage capacity of flow batteries is simply a matter of increasing the size of the electrolyte storage tanks, we may yet see a day where such batteries are an integral part of windmills and large solar arrays. Indeed, the state of California has mandated that utility companies build energy storage capability into their networks. This will inevitably fuel the market for batteries and drive additional research into battery technologies.

[Red Iron Crown]

Here's an interesting TED talk from 2012 about large-scale energy storage. The speaker strikes me as a bit arrogant, but that doesn't mean his idea is bad. I wonder if he got anywhere with it?

[Seret]

Not really a problem Aghanim. For a start nobody has anywhere near the level of installed PV where that could become a problem. Second of all grids can manage variability of supply just as easily as they can manage variability of demand. If was a large amount of solar power other large generators could be taken offline. If it was due to embedded generation that the grid couldn't control then the individual inverters would simply drop offline once the frequency went out of range.

[ThinkOutsideTheHangar]

This year they final got a fusion reaction that gave more energy than it took! But to power the laser........

[phoenix_ca]

This year they final got a fusion reaction that gave more energy than it took! But to power the laser........

Source or it didn't happen.

[andrewas]

http://www.nature.com/nature/journal/v506/n7488/full/nature13008.html

The catch is that, while the reaction emitted more energy than it absorbed from the laser, the reactor as a whole is still way below breakeven .

[magnemoe]

Not really a problem Aghanim. For a start nobody has anywhere near the level of installed PV where that could become a problem. Second of all grids can manage variability of supply just as easily as they can manage variability of demand. If was a large amount of solar power other large generators could be taken offline. If it was due to embedded generation that the grid couldn't control then the individual inverters would simply drop offline once the frequency went out of range.

It also depend a lot of the other energy sources, hydro can be throttled fast, gas turbines pretty fast, coal and nuclear is slow to ramp up and down.

Solar has a bonus in hot areas, lots of the peak use is air condition who depend on temperature so solar input follows the peaks well.

[phoenix_ca]

nuclear is slow to ramp up and down.

Uhhhh...nuclear power might be slow to get up to full power, but reactors are designed to be shut-down fast. Like seconds, and at most minutes.