Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Results of steam bubbles in the reactor

  1. Oct 8, 2014 #1
    When I was in the Navy, we were told of the immense importance of maintaining pressure inside the reactor (all the reactors in the Navy use a hydrogen moderator via water). My teacher discussed that one of the results of loss of pressure could be the formation of steam bubbles collecting on the fuel rods. The reason why this is dangerous is because the molecules on the surface of the rods had fewer H2O molecules to collide with and would therefore radiate heat via infra-red waves. But I don't recall learning why it was that molecules should act this way. Quantum physics wasn't really stressed that much in the Navy's curriculum.

    What is happening on the quantum level that makes molecules radiate infra-red energy when that energy is not able to be transferred via molecular collision?
     
  2. jcsd
  3. Oct 8, 2014 #2

    Astronuc

    User Avatar

    Staff: Mentor

    The concern about boiling in a pressurized water reactor (PWR) is primarily about departure from nucleate boiling (DNB). There is sometimes nucleate boiling in hotter or highly rated PWRs, depending on local peaking factors, coolant temperature and pressure. Boiling reduces moderation in that region of the core. Nucleate boiling improves heat transfer, but if there is too much steam, then the heat transfer is reduced. Following DNB, a layer of steam may form on the surface of the cladding surface. Steam has poor thermal conductivity compared to liquid water at the same temperature, and for a given heat flux, the surface temperature and fuel temperature will increase.

    The comment about infrared is probably a reference to radiative heat transfer, which is rather weak compared to conduction and convective heat transfer. Hot objects (e.g., radiators used to heat rooms) will radiate in the infrared, and eventually in the visible.

    The other aspect of boiling is corrosion of the underlying metal surface, which is a concern if the corrosion rate accelerates greatly, and the metal transforms to an oxide. Too much corrosion/oxidation may lead to a breach of the cladding and loss of fission products to the coolant.
     
  4. Oct 8, 2014 #3
    Thank you, sir.
     
  5. Oct 8, 2014 #4

    QuantumPion

    User Avatar
    Science Advisor
    Gold Member

    To add to what Astronuc posted, DNB is of safety concern because it is a runaway, positive-feedback reaction which can lead to fuel damage/melting. If the fuel temperature becomes too high, the coolant will start to boil. The boiling reduces the heat transfer capability of the coolant, thus leading to even higher fuel temperature, which leads to even more boiling, etc, until the fuel becomes "dry" (only steam). At this point the cladding will burn and/or fuel will will melt.
     
  6. Oct 8, 2014 #5
    Could this also be caused by turbulence in the coolant flow? Doesn't fluid turbulence result in pressure differentials?
     
  7. Oct 8, 2014 #6

    QuantumPion

    User Avatar
    Science Advisor
    Gold Member

    No, in fact turbulence is desirable and intentionally induced using vanes to keep the coolant mixed, which helps prevent DNB.

    Think about cooking a pot of stew. If you don't stir it, you will form a steam bubble on the bottom which may burst and explode hot stew in your face. If you do stir it, the stew remains at a uniform temperature and doesn't explode.
     
  8. Oct 8, 2014 #7
    Ahhh, I see.
     
  9. Oct 15, 2014 #8

    Doug Huffman

    User Avatar
    Gold Member

    Hi ya David. Sorry to be late to the thread.

    After dry out, convection heat transfer fails and the only local mechanism is radiation heat transfer, which you may recall is proportional to the fourth root of the absolute temperature difference - not any where as efficient as conduction or convection.

    But before dry out there are a number of wet regimes characterized by less and less moisture available for Hx.

    While fully wet, nucleate boiling increases the mass flux and heat flux at the channel surface. The 'bubbles' cause considerable turbulence (mass flux) and the energy (heat flux) of bubble formation is the heat/thermal energy of that differential volume. Nucleate boiling is a beneficial but transient state, with burn-out on the right hand of the curve. The operating envelope avoids nucleate boiling for the unpredictability of DNB departure for nucleate boiling.

    Do you recall this curve. It's similar to what I used to teach (1980 - 1993) in this area of this subject.
    https://upload.wikimedia.org/wikipedia/en/thumb/4/49/Boiling_Curve.jpg/450px-Boiling_Curve.jpg

    My text for this subject came from Boiling crisis and critical heat flux (AEC critical review series) 1972 by Long Sun (L. S.) Tong.
    http://www.amazon.co.uk/Boiling-crisis-critical-heat-review/dp/B0006C7S84

    There is another version of the curve, I think called "dryout" that I'll see if I can find.

    For our on-lookers, the ultimate problem is the hydrogen catastrophe at about 1000°C (as I recall) when our cladding burns in the oxygen released.
     
  10. Oct 16, 2014 #9
    That's right. The fourth root. Now I remember. Thank you, Doug.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Results of steam bubbles in the reactor
  1. Reactor Design (Replies: 49)

  2. Nuclear reactor (Replies: 6)

  3. Gallium reactors (Replies: 3)

  4. Nuclear reactor (Replies: 16)

  5. Nuclear Reactor (Replies: 1)

Loading...