Affected continent is Asia, the world is running out of coal and natural gas. A major mega structure project is underway, its goal is to provide electricity across the continent with a single fusion power plant buried beneath the seabed. Once this power plant is commissioned in year 2100 C.E. it will supply over 75% of the entire energy consumption in Asia and that's a lot of energy. Since it is inaccessible to nearly all and will be heavily guarded by advanced navy forces and a network of deep space satellites to provide 24hr surveillance, the power plant only required a small team to operate.

I'm envisioning a possible future setting when an entire continent shares a single power source, would such technology be possible by 2100? If so how would Asia shares the responsibility to keep it running?

The pink dotted lines are where the fusion reactors is buried.

Pink dotted lines denote fusion reactors

  • 3
    $\begingroup$ Not a good idea (unless you're using that big fusion reactor in the sky :-)): what happens when that solar plant has to shut down for maintenance? Having single points of failure is NOT good design! $\endgroup$ – jamesqf Jan 18 '17 at 5:12
  • $\begingroup$ @jamesqf: whole-hearted agreement, my first thought was in the logistics in transmitting energy from a single source across the entire continent! Politically speaking as well, countries coming together to work on a project like this would be extremely unlikely, seeing how you are essentially giving neighbours unlimited access to the core of your economy. $\endgroup$ – Xenocacia Jan 18 '17 at 5:35
  • 1
    $\begingroup$ Yes. It's call the Sun and, even better, it doesn't need engineering nor guards. And it the entire Asia agrees to share the grid/resources, building a distributed network of solar/wind farms would guarantee that there will be sun/wind somewhere to distribute the energy from. If you want a central location for a mega-structure, place there the "hyper-battery" - some chemistries become economical on large scales (e.g. molten NaS - high efficiency, deep cycle and good life cycle) $\endgroup$ – Adrian Colomitchi Jan 18 '17 at 6:45
  • $\begingroup$ My version of the fusion reactor is at least a thousand times longer than LHC in Europe, it runs underneath many countries and oceans in Asia... not a small compact building housing a puny donut! $\endgroup$ – user6760 Jan 18 '17 at 6:58
  • 2
    $\begingroup$ good luck with plate tectonics... $\endgroup$ – Burki Jan 18 '17 at 8:00

What will be Asia's energy consumption in 2100?

The US currently uses 300 GJ per capita per annum. China is at 76 and India is at 24 GJ per capita, currently. Lets say that with rising wealth but improved electric efficiency, Asia uses 100 GJ per capita.

According to the UN via the Washington Post, Asia is set for about 4.5 billion people in 2100. Multiplying those numbers, we need 4.5e20 J of electicity for Asia in 2100, or 14 TW. 75% of that is about 11 TW of electrical power.

Can a fusion generator generate 11 TW?


For massive scale fusion power, the tokamak is probably the only way to go. There are a great variety of potential reactor designs, but the tokamak scales the best...by far. The premise is that you create a toroidal magnetic containment field and then heat fusion fuel inside of it to 10 keV or more. Fusion ensues, and once the temperature gets hot enough, the heat production exceeds the input requirements. You then extract the heat in some manner and generate electricity. A significant portion of this electricity would be needed to generate the magnetic confinement fields. Whatever is left over can power the grid.

Because of the magnetic confinement power requirements, a tokamak works best at the largest scale. If one is ever made, I would expect it to be on the 100 GW scale or maybe even more. It is not inconceivable that you could make TW tokamaks, and set up a group of them in one spot to generate the necessary power.

Can Tokamaks be built by 2100?

This is the question that you as the author will have to answer. There are three primary problems with tokamaks thus far.

First, D-D or D-T fusion produces very high energy neutrons. These neutrons will rapidly activate and embrittle the materials of the tokamak. There is no practical way to stop the neutrons, so either you will have to use a low neutron reaction (called aneutronic; some combination of H-1, He-3, Li-6; or H-1 and B-11). The other option is to develop materials that can sufficiently withstand the embrittlement for the decades of the life of the plant.

Second, the scale of the necessary magnetic fields is very large. In order to generate more energy than is required to operate the field, we would need better magnets than we are currently using. The big technological advance that would solve this problem would be better superconductors. With some advanced superconducting electromagnets, it would become much easier to generate the necessary magnetic fields to contain the very hot plasma.

Third, the understanding of magnetoplasmadynamics is currently insufficient. There are instabilities brought on by the fact that a plasma will have an induced current when exposed to a magnetic field. There is potential for Mega-Amp currents to develop inside the plasma, which will then create its own magnetic field and potentially result in a loss of containment. Solving this problem requires a lot of primary research in to plasma physics.

So can we have a tokamak by 2100? I'm optimistic that the answer is yes. The problems are significant, but reasonably well understood. Dedicated, persistent funding should be able to unlock the necessary physics in a few decades. A little engineering determination later, we could have a working reactor with positive net energy.

How would Asia share the responsibility?

The first thing to note here, is why would you ever make a single reactor? Lets say we have average consumption of 11 TW, and peak of 15 TW. It would be best to have 5 installations of 4 1 TW reactors each, each in different locations. Single reactors are single points of failure. With so many reactors spread out, each facility could take one reactor down for maintenance and still meet the peak power requirements. You mention the seabed...I'm going to ignore that, that is not a good idea.

Since splitting the reactors is basically the only way to do it, they would likely be split between countries who would then sell power to each other. Russia, China, and India are obvious destinations for a powerplant; Iran and Indonesia would probably be next. You can pretty much divvy up the plants between countries any way you wish. If you insist on a single plant, I will tell you that a single plant powering Asia is extremely unrealistic.

| improve this answer | |
  • $\begingroup$ I've upload an image showing the scale of the mega structure, kindly review ur answer if necessary. $\endgroup$ – user6760 Jan 18 '17 at 7:45
  • $\begingroup$ The magnetic field is NOT the problem. Every test plant I'm aware of uses superconducting electromagnets (one reason why they are so god damn expensive). The real problem is stability (like you correctly said) and the heat loss because current reactors are to small, see Lawson criterion. $\endgroup$ – Christoph Jan 18 '17 at 8:02
  • $\begingroup$ @Christoph I'd like to see some articles to that effect. My understanding is that generating a magnetic containment field strong enough is not a problem, but efficiency of the magnetic field generators is. This is the efficiency term of the Lawson criteria; again my understanding is that the efficiency term is the problem, not the radiation or conduction loss terms. $\endgroup$ – kingledion Jan 18 '17 at 14:23
  • $\begingroup$ @kingledion The problem is not the energy efficiency of the magnets (they are superconducting -> almost no loss). The problem is that we can't create strong enough magnetic fields (we would need other superconductors which allow for stronger fields without losing superconductivity e.g. NbTi loses it at ~20 Tesla) to get a fusion rate which exceeds the radiation loss. $\endgroup$ – Christoph Jan 18 '17 at 14:35
  • 1
    $\begingroup$ @Christoph Ah, I was not aware that a magnet could lose its superconducting properties due to a magnetic field/flux increase. I thought only a temperature increase would do that. $\endgroup$ – kingledion Jan 18 '17 at 14:47

Technically yes but I don't think it would make any sense from an engineering/political/military point of view.


Any machine has to undergo some maintenance and can never be trusted to run 100% of the time without any problem. Common sense would make you have a least 2 reactors for redundancy. On top of that, sending energy over large distances is costly unless you assume that supraconductivity has been mastered. You roughly loose 10% of you energy in heat dissipation to the resistance of cables per 1000km.


The country where the fusion plant is based has huge leverage over neighbouring countries. Do you think that China would ever accept a Japan based fusion plant or vice versa?


Again the single node argument. If an enemy country decides to invade Asia, destroying that single fusion plant would probably ensure victory. Even if it s heavily defended it still is an inferior solution compared to having multiple similarly defended smaller redundant sites.

Also why do you limit yourself to a continent, you might as well run the entire planet on a single fusion power plant if the arguments above are somewhat irrelevant in your world.

| improve this answer | |
  • $\begingroup$ I've uploaded an image showing the scale of my mega structure, kindly review ur answer if necessary. $\endgroup$ – user6760 Jan 18 '17 at 7:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.