A recent Time magazine had fusion power highlighted on its cover, with the thought, “It might actually work this time”. The concept is that fusion power, if you can make it work, will provide essentially unlimited power, and that will be totally greenhouse gas free. The usual gaseous product is helium, which has no vibrational spectrum at all because it does not form molecules, at least none that have any lifetime that we know of. The so-called miracle here is that this fusion power will come from small companies, and not from government funding to academics. As an example of the biggest government-funded effort, ITER is being constructed in the south of France at a cost of $20 billion, and will take at least another twelve years to construct, then some unknown amount of time to “debug”.
The problems are reasonably obvious. If you can make fusion work, you get/require temperatures in the order of a hundred million degrees centigrade. There is no solid material above about 4,000 degrees centigrade, so how do you contain this? The short answer is by magnetic fields. Unfortunately, the material becomes a plasma after a few thousand degrees, and plasmas are weird. They are extremely difficult to control, nevertheless, in principle plasmas can be controlled with magnetic fields. It is also possible to arrange for the plasmas to generate their own containment fields, but this is not generally done. Of course to get the necessary magnetic field strength you consume an awful lot of electricity, and a lot more to get the material up to reaction temperature, so there is also the problem of getting more energy out than you put in. There have been fusion reactions in the lab, but these have always, so far, consumed more electricity than they could get out. Of course, they were not designed to make electricity, but rather they were designed to uncover and solve the problems in getting fusion to work usefully.
The reason you get so much more energy out than from any other technology lies in the strength of binding. One of the strongest chemical bonds is the hydrogen molecule, which has a binding energy of about 4.3 electron volts (or about 438 kilojoules per mole). The binding energy of deuterium (a proton and a neutron) is about 2.3 MeV, i.e., about a million times more. 4He (the usual form of helium, two protons and two neutrons) is a bit over 28 MeV. So, if you react two deuterium nuclei to form helium, you get about 23.4 MeV. The easiest reaction to get working would be to react deuterium with tritium (a proton bound to two neutrons) and this reacts to give helium plus a neutron. The problem with this is that neutrons will fly off, hit the walls of the reactor, and react with them, making them radioactive. You have to replace your reactor walls every six months, which makes this an added expense. To add to your troubles, tritium is not that stable, and you have to make it somewhere. So, what to do?
One solution that a company called General Fusion (www.generalfusion.com) has come up with is to compress the plasma in a vortex of metal that includes lithium, and lithium captures neutrons and makes tritium. I must confess that as an outsider who is somewhat ignorant of the problems, I like the thinking here. Another reaction is being tried by a company called Tri Alpha https://en.wikipedia.org/wiki/Tri_Alpha_Energy,_Inc.
(they do not seem to have a website that rates on Google!) They appear to have chosen a different reaction still: firing a proton (a hydrogen nucleus) at boron 11 (to get carbon 12, which is a very stable nucleus). All you need to make this go is a billion degrees! That could also be misleading. Heat is random kinetic energy, and what you try to do by heating is to get some parts go fast enough so that when they kit, the kinetic energy is enough to get over the barrier to reaction. This could be done “cold” if the protons were accelerated fast enough, in which case you have directed kinetic energy.
So, how long is this going to take? Who knows? Since at present the problems being solved are still scientific ones, not in the immediate future, because once these are sorted, there will still be engineering problems, including the one of how to get power from the heat. In my futuristic novel “Troubles” I guessed fusion would be made to work about 2050, and my proposed method of recovering energy was by the so-called magnetohydrodynamic effect. There was a power station built in the Soviet Union that worked by taking energy from a plasma, the plasma being made from coal. I gather that while it worked, it generated about 60% of the available energy, which is much better than standard coal-fired plants, and it was limited by the fact that the plasmas collapse in the region of about 1500 degrees Centigrade. However, their plasmas were unlikely to exceed 4000 degrees, and energy recovery is dependent on the temperature range (4000 to about 1800 degrees) and what is thrown out is lost. (The plant also probably failed because coal will also contain silicates, and these would produce a slag. No slag is possible from making helium.) From over a hundred million degrees the losses due to the second law of thermodynamics applied to plasma collapse are negligible. As an aside, my guess was the use of the deuterium – deuterium reaction, to keep neutrons to a modest amount. (There will be some 3He + n.). Unfortunately, I probably will not live long enough to see whether that guess turns out to be right.