Question pertaining to the Power Yields of Nuclear Fusion Reactors

  1. Aug 27, 2014 #1
    I'm only beginning to scratch the surface of higher level Physics and, despite my usual haunt being the EE area, I enjoy reading the responses and discussion in this section of the forum.

    I was talkin' to a friend about Fusion power and he asked me why it wasn't used instead of Fission. Now, I know just enough about fusion to speak intelligently; but I could not give him an answer that made sense to either of us. I know, or, at least, I've read, that fusion takes much more power than it gives in return. Is this correct? Why?
    Which method of producing Fusion has the highest yield, i.e. highest Ein/Eout ratio?

    Please forgive me if this question is annoyingly basic for you folks. I searched on Google but I couldn't find anything that gave a coherent answer; perhaps I wasn't holding my tongue right.
  2. jcsd
  3. Aug 27, 2014 #2


    User Avatar
    Gold Member

  4. Aug 27, 2014 #3


    User Avatar
    Science Advisor
    Gold Member

    The basic problem for fusion power is that a working system has yet to developed. Historically, about as far back as I can remember (1960's ?) it has been 20 years off. It still seems to be 20 years off.
  5. Aug 27, 2014 #4


    User Avatar

    Staff: Mentor

    Here's a good (long) PF thread that should give you some insights into the current state of fusion power generation & the issues involved in making it practical:

  6. Aug 28, 2014 #5
    Its not enough for fusion power to produce more energy than you put into it. A fusion power plant has to be economically competitive, it has to be reliable, it has to have a high duty cycle (it has to be able to operate all full power for long periods), etc. I think that we the fusion community have to some extent trivialized the difficulty of producing fusion by focusing on the energy multiplication. Yes its an important part, but fusion will never be realizable as an energy source if it doesn't also satisfy the above criteria.

    Why we don't have fusion yet is easy to answer.... fusion is a hard problem. I could go into specifics, but this statement is self-evident when you look at the physical quantities. Our goal is to heat a plasma up to 100 million degrees. To do so we use 1-10 Tesla magnetic fields (1 Tesla is a incredibly strong magnetic field), we drive millions of amps of current in our plasmas, and we us tens of millions of watts of heating power. The magnetic fields and currents exerts huge stresses on the reactor structure, in some instances stresses can be a large as billions of pascals. Not to mention all the large heat fluxes, neutron fluxes, and corrosive effects of the plasma that plasma facing materials must withstand.
  7. Aug 28, 2014 #6
  8. Aug 28, 2014 #7

    This was what I was saying in my discussion with a friend that lead to this any rate; that's how this came about. I was talking about the complications it produces and my friend was giving me all this green, mother earth mumbo-jumbo (I though he might fly away to his mother ship with the way he was thrusting his arms around, but, alas, no such thing took place; would've been cool if it had). He finally got around to stating "I just don't see why we don't have fusion yet! It's the only viable way to go!". Now, I'm never one to say, or think, I know it all; so I gave him the benefit of the doubt and came to PF. Glad I did. Y'all have affirmed my previous surface knowledge and deepened it as well.

    Just FYI: My friend is not interested in physics or engineering any more than he thinks it makes him look smart; sad thing is, it does, which bothers me. You have to take what he says about anything with a grain of sugar. Just wanted to make sure my facts weren't askew.
  9. Aug 29, 2014 #8


    User Avatar

    Staff: Mentor

    When I took fusion engineering as an undergraduate, controlled fusion was estimated at 10 years away. By the time I left graduate school it was more like 20 years away. It has proven to be rather elusive.

    ITER is the next hope for some achievement of a more sustained reaction.

    For fusion to be economical, it needs to generate way more energy than input, and the fusion community has been struggling to get to breakeven.

    The easiest reaction is d+t, but that involves a 14.1 MeV neutron which takes 80% of the reaction, which then leaves the plasma and goes through the first wall into whatever blanket is present - or otherwise into the structure and magnets. Fast neutrons damage material microstructure, so folks are trying to find the most damage resistant materials.

    Ideally, one could devise an aneutronic reaction, but that too is problematic because it requires higher plasma temperatures, and the easiest aneutronic reaction requires He-3, which is rather exotic and very rare on earth.

    the_wolfman has cited the issues with the magnetic fields. Superconducting magnets have been a challenge, and sitting next to a high energy neutron field is problematic. Magnetic confinement is problematic from the standpoint of plasma stability, but also the energy losses from nuclei and electrons in a magnetic field, as well as losses due to brehmsstrahlug, recombination, etc.

    So a viable fusion reactor must produce an amount of energy that is useful (the electrical energy sold to the market) while covering the energy input to heat and maintain the plasma, as well as covering the losses.

    Some of the heat generated in the fusion reactor could be collected in a conventional power cycle, but that adds additional complication to the design and capital cost.

    When someone mentions that fusion is the same process used in stars, that is just misleading because stars use p+p fusion or CNO, both of which occur under conditions that terrestrial systems cannot accomplish. We simply cannot make materials strong enough to sustain the pressures and plasma densities found in stars.
  10. Sep 3, 2014 #9


    User Avatar

    And we wouldn't just need to make materials strong enough to sustain the pressures/temperatures/plasma densities inside most stars, we would need to outdo most stars. The plasma conditions inside the core of the sun wouldn't produce nearly enough power density to be a viable reactor (if I remember correctly, the core of the sun only has a power output of a couple hundred watts per cubic meter), so to make a 1GW reactor (a large, but not unreasonable amount of electrical production), we would need 3-4GW of thermal energy (since the generators aren't 100% efficient), which would require on the order of 10 million cubic meters of solar core at ~300W/m3. That's a cube of solar core material that is ~200m on a side (which is obviously impractical).

    Basically, any fusion power plant will need to outdo the energy production density of the sun's core by 4-6 orders of magnitude to be viable.
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted
Similar Discussions: Question pertaining to the Power Yields of Nuclear Fusion Reactors
  1. Nuclear power reactors (Replies: 4)