Thursday, November 24, 2011

Accidental Fallout from LANL?

Although contrary to today’s festive spirit, let's think for a bit about possible future releases of radioactive material from LANL, as a result of earthquake and/or wildfire. How much radioactive fallout would be experienced by local residents? Could this radioactive fallout constitute any sort of health hazard?  Could a retired physicist like myself, with no particular knowledge of environmental science, have anything sensible to say about such questions? Well, let’s see!    

 A release of Pu-239 from PF-4, or the still to be built CMRR-NF, could occur sometime in the future. In the event of a major earthquake (>7 on the Richter scale) followed by fire, there might be a release from either of these facilities of Pu-239 in dust form, or in the form of fumes from burning Pu-metal. Since many tons of Pu-239 is expected to be stored at PF-4, and/or at the CMRR-NF, one might imagine that a release of >1 ton Of Pu-239 could be possible. [1 ton, or 1000 kg of Pu-239 corresponds to 73,000 PE Ci.]

 Such a possibility was said last week, at a meeting in Santa Fe, NM, by the Defense Nuclear Facility Safety Board Chairman, Dr. Peter Winokur, to be “not far-fetched.” (The DNFSB is a group of scientists and engineers tasked by the US Congress to provide advice on safety practices, at US nuclear weapons facilities, to the US President and to the DOE Secretary.)

 A wildfire burning over Area-G might have serious consequences. Currently, there are 10,000 barrels of TRU-waste stored above ground at Area-G, and these are only poorly protected against the effects of  wildfire. Each barrel contains ~10 PE Ci of actinide residue, mainly Pu. If these barrels were to burst during a wildfire, ~100,000 PE Ci could go up into the air. (At the DNFSB meeting in Santa Fe last week, DOE/NNSA manager Anderson testified that the las Conchas wildfire  "was very scary" and that "if the wind hadn't shifted when it did, LANL might have been consumed by fire.")

 Thus, there are at least two different scenarios that might result in the release of ~100,000 PE Ci into the air over LANL.

 If a bad accident at LANL, accompanied by a hot fire, were to release 100,000 Ci of Pu-239 into the atmosphere, and if the ensuing fallout exhibited circular symmetry with an activity decreasing exponentially with distance r from the source, over a characteristic distance R (that is, if r=R, then the concentration at r would be  1/e = 0.37 of its value at the source, according to the law C(r) = C(0) * exp( -r/R), where C(r) is the activity per unit area, at r, in units of Ci per square meter, or Ci/m**2) then the activity of Pu-239 deposited on the ground at a distance r, and for R = 1.0 mile, would be:

                        Table 1

         r (miles)              C(r)  (Ci/m**2)
         --------            --------------------

          0.01                     30.7
          0.10                     0.307
          1.00                     0.00113                     

          5.00                     8.28 * 10**(-7)

          10.0                     1.39 * 10**(-9)

          20.0                     1.59 * 10**(-14)

If, instead, R = 5.0 mile, then:

                       Table 2

         r (miles)             C(r) (Ci/m**2)
         --------            --------------------

          0.01                     30.7
          0.10                     0.307
          1.00                     0.00307                     

          5.00                     4.52 * 10**(-5)

          10.0                     4.16 * 10**(-6)

          20.0                     1.41 * 10**(-7)

 For a small fire, one would expect fallout to be concentrated over and around the facility which was burning, with the density of fallout and, therefore, of activity, both areal and volumetric, decreasing with distance from the facility; i.e., as described by the numbers in Table 1.  For a hotter and more extensive fire, Table 2 might instead pertain. But, if the fire were both very hot and very extensive, such that the hot gases being generated rose up into the stratosphere, then the fallout would be distributed over larger distances than those indicated in the Tables, and the fallout pattern would  elongate along the direction of the prevailing winds. In such a case, both Tables would need to be modified. (The Fukushima-Daiichi disaster exhibited a pattern of fallout that extended in the northwest direction over a range ~10 times that of its width. Since the damaged nuclear reactors were located on the eastern seacoast, strong prevailing sea breezes blew the fallout away from the coast and towards the northwest. In May, 2000, smoke from the Cerro Grande wildfire extended from its source in the Jemez Mts., through Los Alamos, and towards the northeast, over the town of Española, and over the pueblos of  Santa Clara and Ohkay Owingeh, again blown by the prevailing winds. The length of the smoke plume was ~3 times its width, and it extended all the way into Oklahoma.)

 But, to continue:

 Let's estimate the amount of Pu-239 that would be deposited in the lungs of a local resident over the ~1 year's time that it would take for all of the fallout to come to earth. Let's assume that the amount of Pu-239 suspended in the air, during that 1 year's time, remains at a constant value and is distributed evenly in a vertical air column up to an altitude of 1000 m. (One could object to this number since, as we will see, it plays a critical role in our calculation; however, as a rough value I think that it may not be so bad.) The volumetric activity of Pu-239, as a function of r, would then be given by the numerical values in the above Tables, reduced by a factor of 1000 (and in units of Ci/m**3). Then, since the volume of air exchanged per breath by the human lungs is ~0.5 liter, or 0.0005 m**3, and the number of breaths taken per year is ~2 x 10**6, the volume of air exchanged in 1 year would be ~1,000 m**3. If we assume that all of the inhaled Pu-239 is deposited in the lungs, and remains in the lungs, then the activity of the Pu-239 concentrated in the lungs in 1 year, as a function of r, is given by the numerical results appearing in the above Tables (but, now in units of Ci).

 A LANL compendium of radiation effects published in June, 2000 (Los Alamos Radiation Monitoring Notebook, by J. T. Voss) noted that Pu-239 deposited in the lungs of dogs was fatal, within a year, if its activity exceeded ~5 x 10**(-7) Ci/gm (Ci per gm-weight of lung tissue.) Therefore, for adult humans with an average lung mass of 900 gm, inhaled Pu-239 might be lethal for activities > 4 x 10**(-4) Ci (corresponding to a dose > 4000 Rem.) Since single doses < 10 Rem are usually considered to be marginally safe, one could say that for adult humans inhaled Pu-239 would be marginally safe if the activity of all the inhaled material was < 10**(-6) Ci.

 Returning to the Tables, we see that, for R = 1 mile (Table 1), and at a distance of 5 miles from a Pu-239 release of 100,000 Ci into the atmosphere, the amount of Pu-239 concentrated in the lungs of a local resident, within 1 year of the release, does not exceed a marginally safe value. However, at a distance of just 1 mile from the release point the amount accumulated would be lethal. Therefore, for this case, a zone of exclusion extending out to at ~5 miles from the release point would be necessary.

 Similarly, for R = 5 mile (Table 2), and at a distance of 10 miles from a Pu-239 release of 100,000 Ci into the atmosphere, the amount of Pu-239 concentrated in the lungs of each local resident, within 1 year of the release, is marginally safe; but, at a distance of 5 miles the amount accumulated would be unsafe and possibly lethal. In this case, the zone of exclusion would have to extend to ~10 miles.

 Shocking stuff? Well, maybe.

 Clearly, many assumptions have been made in these rough "calculations", some more critical than others. The form of the distribution assumed is very important, as is the value of R chosen, if the distribution were to be exponential. The height of the air column through which the fallout occurs is, obviously, very important; i.e., the actual volumetric density of Pu-239 at ground level is critical. Moreover, there is uncertainty in the amount of Pu-239 which would be inhaled by persons in the vicinity of the release point and the maximum dose of inhaled Pu-239 which can be tolerated by humans is also not very well known.

  But, in spite of these uncertainties, I claim that my rough "calculations" suggest the need for a careful study of these important questions by independent qualified experts. But, perhaps such studies have already been performed? Then I wonder what their results have shown and why they aren’t already available as public information?  Maybe they’re just too shocking?

Friday, November 18, 2011

DNFSB Criticizes LANL's Risky Practices

17 Nov 2011/ Defense Nuclear Facilities Safety Board (DNFSB, or the Board) public meeting held at the Santa Fe, NM Convention Center / Thursday, 17 Nov 2011/ 1PM - 5:30PM; 7PM - 9PM.

The Board was created in 1989 to advise the President and the DOE Secretary on safety issues at the DOE's nuclear weapons laboratories. The Board has ~100 full-time staff and an annual budget of ~$22 million.

Today, safety practices at Los Alamos National Laboratory (LANL) will be addressed by the Board. Of particular interest to the Board are:

 1) effects of the latest information about seismic activity beneath the Pajarito Plateau as it concerns the ongoing planning for the CMRR-NF construction project at LANL;

 2) effects of lessons learned from the Cerro Grande and Las Conchas wildfires on planning for the CMRR-NF construction project at LANL, as well as on the Area G cleanup, and upgrades to the Radioactive Liquid Waste Treatment Facility;

 3) effects of lessons learned from the Fukushima Daiichi natural disaster on planning for the CMRR-NF construction project, and on other projects at LANL.

The Board Chair, Peter Winokur, and four other Board members listen to short formal presentations from two panels of LANL/NNSA managers (6 per panel), and ask follow-up questions. Approximately 50 members of the general public are also in attendance. However, no questions from the general public are allowed and several uniformed security personnel are present to keep order. Nevertheless, the general public has been invited to present comments, 5 minutes per person, during a 45 min period at the end of the meeting. The Board Chair also announces that the meeting record will be held open until 19 Dec, 2011 and that anyone may add remarks to the record. Proceedings of the meeting are being recorded by means of redundent audio and video, a stenographic record is being created, and several photographers are at work in the meeting hall.

Panel member Donald Cook (DOE/NNSA manager) remarks that the latest seismic studies show that earthquakes which might occur beneath the Pajarito Plateau could be so large as to exceed the design standards for LANL's PF4 (an existing plutonium facility) and that the PF4 building structure might experience multiple failures during such an earthquake. However, he asserts that even assuming a maximum release of plutonium (powder, or fumes from burning metal) into the atmosphere, the resulting biohazard would be 10,000 times less serious than current natural occurring biohazards. He also claims that planned upgrades to the facility will reduce future possible releases of plutonium to below levels specified as safe by the DOE.

Cook also claims that lessons learned from Cerro Grande wildfire helped to mitigate effects at LANL of the Las Conchas wildfire. Similarly, he expects that lessons learned from the Las Conchas wildfire will mitigate the effects of future disastrous wildfires at LANL; ditto [somehow] for the Fukushima Daiichi earthquake and flood.

LANL Dir. Charles McMillan talks about "the broad scope of what we have done at the Lab." He says that "safety is our highest priority at the Lab", and that "our goal is to encourage reporting safety issues before they become serious." He asserts that "one of the reasons that we are here today is because of the Lab's self-reporting of possible problems at PF4." He says that, however, "PF4 is so well-designed that, in the event of an earthquake he would feel safer in PF4 than in his own home."

Board Chair Winokur remarks that, in the event of earthquake inspired structural failures and fires at PF4, “the levels of radiation in and around the destroyed facility might rise to 100,000s of Rem.  It is doubtful that anyone could be safe under those circumstances.”

DNFSB LANL site-representative Brett Broderick points out that PF4 was designed and constructed in the 1970s using the best seismic info then available. However, in 2007 new seismic data showed that ground motions during a strong earthquake could be (1 1/2)x greater in the horizontal direction and 2x greater in the vertical direction, than previously thought. Projected radiation releases during such extreme events were shown to be 100x greater than had been thought possible earlier. He said that, nevertheless, approximately 1/2 of the revealed vulnerabilities at PF4 had already been corrected. Remaining vulnerabilities were primarily in the ventilation systems and in the fire suppression systems.

At this point, Board Chair Winokur notes that the US Atomic Energy Act required the DOE to protect the public against hazardous conditions at DOE nuclear weapons facilities. In this regard, he believed that DOE was designing its facilities so that worst case accidents would not result in radiation doses to members of the general public of greater than 25 Rem. However, recently, he has become aware that the DOE has actually been designing its faciities such that doses to individuals could be as high as 2500 Rem.

DOE/NNSA manager Cook immediately objects, saying that the "risk" to the general public is being appropriately computed by DOE as the product of the probability of an accident with the "consequence" of that accident.

Board member John Mansfield objects strenuously, saying that he does not believe that the DOE knows how to correctly calculate risk. [Evidently, a controversy has been brewing about this interesting topic between the Board and the DOE.]

Board Chair Winokur continues: "The risk to the public of an accident occurring once in 2000 yr, with radiation released corresponding to a dose of 2000 Rem is very different than the risk to the public of an accident which occurs once a year and leads to a dose of 1 Rem each time!"

[Now, I think it worthwhile to point out a few facts about radiation dose; i.e., according to experts who have studied the effects of radiation dose on the human body:

Normally, exposure to radiation from environmental sources results in an accumulated annual dose of less than 1 Rem, for each individual; such a low dose is always inconsequential. Also, doses accumulated gradually (over the course of a year, say) are always less problematic than doses accumulated all at once. Although sudden doses of less than 50 Rem usually do not lead to observable physical effects, sudden doses of ~ 100 Rem or more are often physically damaging, and doses of 1000 Rem or greater are usually fatal. The DOE sets 5 Rem as the maximum allowed annual dose for a worker in the nuclear weapons industry.]

[Although Winokur did not say so explicitly, it seems clear that although a release of 2000 Rem might be a rare event, any person would experience it as a deadly event, if it happened to occur in his life-time, or at the end of his life-time; but otherwise not! Which is to say that it is essential not to be confused by the rarity of an event when estimating the risk of that event to an individual.]

Panel member Kevin Smith (DOE/NNSA manager) remarks that the Las Conchas wildfire was very different than the Cerro Grande wildfire in that the more recent wildfire moved much more rapidly (explosively, in fact) and consumed much more of the forest (~10X more.)

Panel member Anderson (DOE/NNSA manager) said that the Las Conchas wildfire "was very scary" and that "if the wind hadn't shifted when it did, LANL might have been consumed by fire."

Board Chair Winokur asked Anderson if, during the Las Conchas wildfire, he had been worried about LANL's Area G; i.e., since Area G stores aboveground and essentially unprotected against the ravages of wildfire many [~10,000] large metal drums, each drum filled with TRU waste [~10 PE Ci per drum]. Anderson said that he had not been particularly worried about Area G.

Board member Joseph Bader asked about accidents so severe that they exceeded the "design basis" threat level. DOE/NNSA's Dr. Stanford answered that "such extreme events will be practiced during upcoming exercises."

Board Chair Winokur remarked that, if one tried to "imagine accidents at LANL beyond the design basis, a large earthquake accompanied by a wildfire would qualify, and did not seem to him to be far-fetched."