Wednesday, March 16, 2011

Spent fuel pools at Unit 4

Water levels a concern for Unit 4 spent fuel pools

One of the big issues for those following the news right now has been the status of the spent fuel cooling ponds at Fukushima Daiichi Unit 4. Unit 4 was undergoing regular maintenance prior to the earthquake; no fuel was in the reactor. Thus, until now it has not been as serious of a concern as Units 1, 2, and 3.

Reports from the Japan Industrial Atomic Forum (JAIF) indicate low water levels at the spent fuel pool, with damage suspected to the fuel rods. An explosion and fire have also been reported at Unit 4. While some suspect the explosion may have been hydrogen-related, the spent fuel pool was significantly cooler than the reactors at Units 1 and 3, with temperatures reported to be around 85 C (185 F) over the past two days - thus, the explosion remains difficult to understand. Several fires have also broken out over the last two days in the spent fuel storage area - sources indicate that these have been machinery oil fires, which have since been contained. Reports of low water levels at the Unit 4 spent fuel storage pool appear to have begun at least 24 hours ago (as of 19:00 hours on March 15).

It is difficult to determine the current status of the fuel pool, given conflicting accounts. U.S. NRC chairman Gregory Jazcko has indicated that it is his staff's belief that, "there is no water in the spent fuel pool." This was immediately contradicted by a TEPCO spokesperson who indicated "the condition is stable", however no further updates are available at the TEPCO website. However, water levels at the spent fuel pools have been a reported concern for some time, according to both World Nuclear News and NEI.

The Nuclear Industrial Safety Association - Japan's version of the NRC (and part of the Ministry of Economy, Trade, and Industry - METI) - has most recently reported the following:
<Unit 4>
  • It was confirmed that a part of wall in the operation area of Unit 4 was damaged. (06:14 March 15th) 
  • The fire at Unit 4 occurred. (09:38 March 15th) TEPCO reported that the fire was extinguished spontaneously (11:00 March 15th) 
  • The temperature of water in the Spent Fuel Storage Pool at Unit 4 had increased. (84 ℃ at 04:08 March 14th)
  • The fire occurred at Unit 4. (5:45 March 15th) TEPCO reported that no fire could be confirmed on the ground.(06:15 March 16th)
  • The water injection was stopped. (14:00 March 16th)

CNN and NHK are currently reporting that JSDF helicopters are currently attempting to drop water onto the Unit 4 building in order to add coolant, as well as water cannons (essentially, fire engines) on standby.

Spent fuel pool basics

Unlike the fuel immediately after shutdown like in Units 1 and 3, spent fuel sitting in the cooling pool is considerably cooler, although it still generates enough heat to require water cooling. Spent fuel is generally about 10-13 feet in height; the water covering this fuel is generally at least another 20 feet in height above the fuel itself. This water serves a dual purpose - it both keeps the fuel cool and provides a good measure of radiological shielding; exposure rates immediately above the spent fuel pool are such that workers can generally be in the area without special protective equipment - generally less than 2 millirem per hour (about a fifth of the dose of a chest x-ray).

A spent fuel storage pool (Image courtesy of IEEE spectrum)

The spent fuel storage pool is kept at atmospheric pressure, unlike the reactor. Under normal conditions, pumps will circulate cooling water in order to keep the rods cool, although under emergency conditions, natural circulation is expected to take over (i.e., where warmer water expands and grows less dense, thus rising to the top, while cool water, which is denser, sinks - thus providing a natural "circulation" for heat removal around the rod.) These spent fuel pools tend to be quite robust - a simple failure of pumps or piping would not be able to drain the water level from these pools, which are made of thick concrete and steel.
Generally, the only way significant water level changes could be expected in the spent fuel pool are from evaporation of the water or if a large crack were to develop in the pool itself. Readings over the last two days indicate that the temperatures of the pools were elevated, at around 85 C (185 F) - while certainly warm, well below boiling. (However, one can expect an increased rate of evaporation at this temperature). The IAEA confirms these temperatures for the prior two days, however no data was available for today. If accounts of diminished water levels are to be believed, this would imply that there may be structural damage to the building; this much is unknown at this time.

Reports indicate that there may be fuel damage due to dropping water levels - this would likely be in the form of cladding failures, which have released radioactive fission gases similar to the process in Units 1, 2, and 3. The major concern for workers right now is in personal safety - without the protective layer of water, there is significantly less radiological shielding, making it far more difficult to operate around the spent fuel pool. This would appear to be the reason for using water cannons and helicopter drops in order to supply emergency water, in order to bring radiation levels back down (by providing an adequate layer of water for shielding). 

For those keeping score at home, the radiological source from these spent fuel rods would be deep-penetrating gamma radiation; the danger to workers is that the exposed would be that in the absence of the several feet of water shielding, the spent fuel rods would essentially act as a "gamma flashlight" pointing out of the fuel pool.

Depending upon the age of the fuel rods themselves (which determines the level of decay heat), it seems less likely that any kind of actual fuel melting would occur itself. Rather, the chief issue appears to be in the radiological release from ruptured fuel cladding as well as the higher levels of radioactivity, which make it extremely difficult for workers to get close to the pool in order to refill it with coolant - hence the use of helicopters and water cannons from the ground. 

I will continue to update as I learn more - unfortunately, the news on this appears to be very scarce, and mixed at best.

Update: JAIF confirms helicopters dropping seawater on Units 3 and 4 due to low coolant levels. Private sources have also begun to confirm that water levels are very low at Unit 4.


  1. " the spent fuel rods would essentially act as a "gamma flashlight" pointing out of the fuel pool."

    The problem with the flash light analogy is the ability of gammarays to penetrate.

    I = I(o)e^-(µ/ρ)ρt

    What is the gamma energy range and what would be the half thickness of say, lead?

  2. I should say:

    "What is the gamma energy range of a recently recycled MOX fuel rod (2 months) and what would be the half thickness of say, lead?"

  3. If you want to do a full calculation you would need these for water and I assume lead b/c the helicopter has a lead plate installed?

    You would also need to know the initially loading of the core, the operating history, the time after shutdown and the amount of water covering the pool.

    Since we don't have that, do a back of the envelop calculation:
    Assume that the water level has dropped by a factor of 4 from 16ft to 4th covering the rods. Then your gamma intensity would be I_2 / I_1 = exp(4) = 54.6 times more intense.

    The analogy was mine that I gave to Steve. If those fuel rods did indeed become uncovered this would be an intense gamma source pointing essentially out of the pool.

    A calculation similar was done by a professor here at NCSU for our research reactor in the event of a loss of coolant event:
    In here, dose maps are given if the core was uncovered. Of course this is less fuel. And PWR fuel to boot.

  4. Cyrus is right - one thing to add is that the attenuation coefficient for gammas of air compared to say, water, is fairly negligible. i.e., going back to the NIST pages Cyrus linked to, we can look at the mass attenuation curves for air and water.

    What we see is that mu/rho for both is pretty similar. Here's the rub - the density of water is about 1 gram per cubic centimeter, give or take a little (i.e., for the water being warmer). The density of air at sea level is about 1 kg/m^3, or about 0.01 g/cm^3. In other words, water is about 100 times more dense than air, meaning the attenuation factor has dropped by a factor of 100.

    Basically, this density difference profoundly affects the attenuation - Wikipedia has a list of halving distances for radiation. The halving distance (the distance at which half the gammas would be attenuated) is 18 cm for water, and 15000 cm for dry air - so you can see immediately that air is a far, far less effective radiation shield than even ordinary water.

    Basically, that would imply that for every meter of water lost, the radiation levels increase by a factor of 10. (Or a factor of 10 increase for about every 3 feet the water level drops.)

    If we lost only a foot of water, we'd already be at 20 mrem/hr - and from the sounds of it, the pool may have lost several feet. This can easily get significant.

  5. Although you asked about lead - we're dealing with a half-thickness there of 1.0 cm. Assuming the gammas we are dealing with are on the order of tens of keV to perhaps 10 MeV, we can reference the attenuation curve for lead.

    For simplicity, let's say we're dealing with gammas on the range of 100 keV to 10 MeV. This is actually pretty reasonable.

    mu/rho ranges on the order of 1 to 0.05 for this energy range. The density of lead is 11.3 gm/cm^3.

    So, the "worst" case, for 10 MeV gammas (which are far less common, of course), indicates a halving distance in lead of 1.22 cm. For 200 keV gammas, it would be about 0.06 cm. So, weighting for the higher end of the gammas, we come to around the answer Wikipedia gives - I think 1 cm halving distance in lead is probably a reasonable number for this case.

  6. So lead is 18 times more efficient by thickness than water and if we want the equivilent of 20 ft of water then we need over 1 foot of lead. That is a lot of wieght to move around.

    I was hoping to get a feel for how "bright" the flashlight is without getting out the old textbooks.

    Thank you for your help!

  7. "TEPCO officials say that although one side of the concrete wall of the fuel pool structure has collapsed, the steel liner of the pool remains intact, based on aerial photos of the reactor taken on March 17. The pool still has water providing some cooling for the fuel; however, helicopters dropped water on the reactor four times during the morning (Japan time) on March 17. Water also was sprayed at reactor 4 using high-pressure water cannons."


    This makes it sound like one side wall is simply steel with no concrete.

  8. I would think that from this PDF (takes a while to load):
    That the water is being contained by the steel. I don't think the concrete has anywhere to go. The structure is recessed so even if the earthquake cracked the spent fuel pool its still around. The part where the crack is could be radioactively hot. If it points towards the primary containment then w/e. If it points to the stairs who knows?

  9. "Once the pool water level is below the top of the fuel, the gamma radiation level would climb to 10,000 rems/hr at the edge of the pool and 100’s of rems/hr in regions of the spent-fuel building out of direct sight of the fuel because of scattering of the gamma rays by air and the building structure. At the lower radiation level, lethal doses would be incurred within about an hour. Given such dose rates, the NRC staff assumed that further ad hoc interventions would not be possible."

    I have been wondering why we are seeing so much Gamma ray activity at ground level when we think of the pool as a flashlight. I had not considered scattering of gamma ray radiation.

  10. We see photons reflecting (scattering) off of the surfaces that we shine a real flashlight on. This would be identical behavior whether it be gamma or visible light part of the spectrum.

  11. I think they are talking about Mie scattering by air. I am imagining a flashlight in the fog.

  12. Colin, not mie; compton scattering(inelastic scattering against bound electrons).

    See the nice little plot on wikipedia for the differential cross section of compton scattering at different scattering angles and at different gamma energies:

    Photoelectric effect(dominant at low energy), Compton scattering(dominant at medium energy) and pair-production(dominant at high energy) are the three major effects by which gamma rays interact with matter. These effects allow gamma radiation to be shielded and conversely also to be absorbed in the human body and do damage.

    Alphas interact so strongly with matter that they are trivial to shield, where as neutrinos can happily go through a lightyear of depleted uranium without being any worse for wear.

    Gammas are an unhappy trade-off; they interact weakly enough that it's difficult to provide any sort of compact, lightweight shielding, while also interacting strongly enough to be dangerous.