From: jfloyd@wam.umd.edu (Jason Edward Floyd)
Subject: Re: Yucca Explosion
Date: 11 Mar 1995 14:52:46 GMT
Organization: University of Maryland College Park
Lines: 437
Message-ID: <3jsdfu$odh@cville-srv.wam.umd.edu>

Here is the Los Alamos internal review committee report on Bowman and
Venneri's findings:

COMMENTS ON "NUCLEAR EXCURSIONS" AND "CRITICALITY ISSUES" 
LAUR-95-0851

Gregory H. Canavan, Stirling A. Colgate, O'Dean P. Judd
Albert G. Petschek, Thomas F. Stratton
Los Alamos National Laboratory

        Technical reviews of papers on criticality and energy release
from underground storage of fissile material concluded the probability
of each of the steps required is vanishingly small and the probability
of occurrence of all of them is essentially zero.  Even if they could
occur, any release would be too small and slow to produce significant
consequences in the repository or on the surface.

        The Laboratory provided technical reviews of papers by Drs.
Bowman and Venneri. The first, entitled "Nuclear Excursions and
Eruptions from Plutonium and Other Fissile Material Stored
Underground"[1] ("Nuclear Excursions") was reviewed in December, 1994,
and a written response was submitted to the authors through Laboratory
management. The second, entitled "Criticality Issues for Thermally
Fissile Material in Geologic Storage"[2] ("Criticality Issues"), which
was a response to the issues raised in the December review, was
reviewed in February, 1995. This review summarizes the assessment of
both. Very recently, the authors released a third paper, entitled
"Underground Autocatalytic Criticality from Plutonium and Other Fissile
Material."[3] ("Underground Autocatalytic Criticality"). However, it is
largely a compilation, without correction, of materials from the first
two; thus, our comments apply to it as well.
        The papers primarily discuss the underground emplacement of
glassy logs containing weapons plutonium, and purport to demonstrate
that after on the order of 10,000 years, geologic action will increase
their reactivity to the point where criticality, auto-catalytic action,
and explosive energy release are probable. The significant difference
between the papers is that the first ascribes the increase in
reactivity to the dilution of plutonium in a dry silicon dioxide
medium, while the second two ascribe the increase of reactivity to the
concentration of plutonium in a wet silicon dioxide medium.
        The review concluded that the discussion in the papers does not
describe a credible sequence of geologic events leading to super
criticality and explosive energy release. The probability of each of
the necessary steps-increase in nreactivity to criticality,
auto-catalysis, and explosive energy release-is vanishingly small, and
the probability of occurrence of all three is essentially zero.
Moreover, even if these steps could occur, any energy release would be
too small and slow to produce any significant consequences either in
the repository or on the surface. Indeed, any surface effects would
occur on times of tens of thousands of years, which are so long as to be
outside the time scale of any credible scientific prediction.

        Emplacement, dispersal, and criticality. The geological
situations discussed in "Nuclear Excursions" were too unrealistic to
provide a useful framework for analysis or to validate the proposed
scenario. That was pointed out in the review, but those situations were
still used in "Criticality Issues." "Nuclear Excursions" postulates the
emplacement of fissile materials in geologic formations of pure silicon
dioxide, which is a weak neutron absorber, is not a common geologic
material, and has not been proposed as a repository material. Other
elements present in all geologic formations absorb neutrons much more
strongly than pure silicon dioxide, which reduces the reactivity of the
mixture. Although the papers mention minor soil constituents with very
large absorption cross sections, their calculations ignore them. The
papers offer unsupported estimates that including them would increase
the critical mass by 50%. When they are properly included, it may not
be possible to achieve criticality for the assumed conditions even with
pure Pu-239. It is not possible to be more quantitative in our response
without further analysis of weapons Pu and spent fuel in realistic
media, which is not performed in these reports. That must be done in a
more careful subsequent project.
        The papers perform most of their calculations for pure Pu-239.
The weapons plutonium of interest has a significant fraction of Pu-240,
a strong absorber that further reduces reactivity. Even for the maximum
loadings postulated in "Nuclear Excursions," weapons plutonium could
never disperse to a condition of criticality in real, dry repository
materials. It is argued that the Pu-240 would decay, leaving the more
reactive Pu-239, but that would happen over several times the 6,500
year half life of Pu-240. Even then the Pu-240 would be replaced by its
daughter U-236, which is a weaker but still noticeable absorber,
degrading the thermally fissile mixture. 
        The assumption of significant dispersion of plutonium into the
surrounding geologic medium is without justification. Geologic
processes would take millions of years, by which time plutonium would
have decayed to uranium-235, which is less reactive than Pu-239. We
have not discovered a credible process that would produce more rapid
dispersal. Anthropogenic measures are unlikely and are routinely
accounted for in repository analyses. "Criticality Issues" argues that
water flowing down through the repository would dissolve the glass log
in 1,000 years and leave a fragile powder, but its calculation
overestimates the amount of rainfall on-and water within-the repository
by factors of 1,000, so the correct time scale for dispersal is about a
million years.[4] Moreover, the temperature gradients driving the
process are overestimated by an order of magnitude, and the leaching
process could leave a residue as strong as the original log.
        Autocatalysis. The papers' assumptions about the behavior of the
fissile mixture near criticality are not credible. Based on their
improper interpretation of published equations of state, "Nuclear
Excursions" and "Underground Autocatalytic Criticality" assumed the
rock in which the fissile material is placed is rigid and would prevent
the expansion of the material. Rock is compressible, and even at depths
of several kilometers, lithostatic stresses are small and anisotropic,
so that confining stresses are small. Even if the mixed material became
critical, it would slowly heat and expand, which would decrease its
reactivity below critical. Then its neutron flux would drop, and it
would cool.[5] Thus, these dry mixtures have the negative temperature
coefficients characteristic of most fissile assemblies, as discussed in
detail in the open meetings of the review, and would not be
autocatalytic for material motion over geologic time scales. 
        "Criticality Issues" again argued that fissile material could
diffuse to criticality, although it shifted its argument to SiO2 with
high amounts of water, which have higher reactivity.[6] However, the
physics for such media is essentially the same as that for dry rock.[7]
There are two parts to the argument, depending on whether the mixture
approaches criticality from the under moderated or over moderated side.
From the under moderated side, as the mixture reached criticality, it
would heat slightly.  That would expel some water, which would reduce
its reactivity, after which it would cool.[8] This is closely related
to the stabilization of dry media by a negative temperature coefficient.
        From the over moderated side, as the mixture gradually passed
through criticality, it would heat slightly-though not enough to expel
significant water-which would cause it to expand. That would reduce its
reactivity, after which it would cool.[9] Thus, over moderated, heavily
hydrated mixtures generally also have negative temperature
coefficients.[10] Thus, there is nothing new in the papers on wet media,
which just repeat the stability errors made in "Nuclear Excursions" in
a different context. 
        A key feature not addressed in the papers reviewed is
importance of the evolution in time of the criticality and temperature
of the mixtures. For those of interest, the time scale for the increase
of reactivity is very long-tens to hundreds of thousands of years.
Thus, the excess levels of criticality and hence the time scales for the
release of energy are correspondingly long-thousands to tens or
hundreds of thousands of seconds. And the temperature increases are
fractions of a degree. The slowness of those processes dominate the
faster time-dependent processes postulated but not analyzed in the
reports. 
        There are some scientifically interesting interactions between
the negative temperature coefficient of such mixtures from expansion
and the potentially small positive coefficient from absorption and
Pu-239 resonance broadening, but those effects are delicate and
comparable even at very high levels of hydration.  Unfortunately, they
cannot be evaluated from the calculations in "Criticality Issues," which
were apparently all performed for cold soil, pure SiO2, and pure
Pu-239. All three of those restrictions would have to be removed to
provide an assessment beyond that in "The Myth of Nuclear Explosions at
Waste Disposal Sites," which predicts overall stability.[11] 
        Energy release. Even if dispersion and criticality are assumed,
the conclusion that an explosion would occur is incorrect. "Nuclear
Excursions" postulates "auto-catalytic" behavior in which the release
of energy leads to greater criticality, but the discussion above shows
that in dry repository material, the release of energy instead reduces
criticality and shuts the reaction off. "Criticality Issues" postulates
auto-catalytic behavior in hydrated mixtures, but the discussion of the
previous section shows that to the extent that the phenomenon has been
quantified by earlier work, the release of energy reduces criticality
there, too. Temperature increases appear to be limited to at most
fractions of a degree for plausible dispersal times. 
        The postulated mechanisms for explosion are not credible. The
essential feature of explosive process is the rate at which energy is
released. The papers do not calculate it; they do not even estimate it.
They simply assume it. For the largest realistic rates the most that
appears possible is heating and evaporation of some water before a
smooth shut down. There is no credible mechanism for releasing energy
on a time scale short enough for even a steam explosion. A nuclear
explosion must make the transition from critical to highly
supercritical in a fraction of a second. A credible means to force such
a transition in a repository has not been found.[12] Thus, the
assertion that an explosion would occur is incorrect.  
        Even if dispersion, criticality, and energy release are assumed,
which appear virtually impossible on the basis of the arguments above,
there would be no serious consequences elsewhere in the repository or
on the surface. Even if an explosion could occur, careful calculations
indicate that the energy released would be on the order of a few
percent of that from the natural decay of the Pu over the same time
scale. Detailed hydrodynamic calculations indicate that the containment
volumes from such explosions would be very small compared to the
nominal spacing between storage elements; thus, there could not be any
coupling between storage elements or any possibility of greater energy
releases through synergisms.[13] 
        Relation with other work. That the critical mass may be reduced
by dilution by moderating material, as discussed in the paper, is well
understood by the nuclear community. Fermi used it to full advantage
when he assembled the first pile under the grandstand at Stagg
Stadium.[14] Fermi also used the advantages of heterogeneity in
minimizing resonance losses in natural uranium, although that is
irrelevant to the discussions of Pu reactivity here.
        The National Academy of Science report does not suggest
emplacement of weapons plutonium in the manner discussed by "Nuclear
Excursions," although it did comment on the advantages of higher
fissile loadings. The Academy was alert to the potential for
criticality and qualified its recommendations by stating that further
analysis and discussion were needed before deciding on the best and
safest geologic disposition of weapons and reactor spent fuel.
        Summary. We should always be alert to unintended consequences
and open to discussions that illuminate potential dangers in nuclear
waste storage.  "Nuclear Excursions" argued that there were serious
dangers in proposed repository concepts, but review found the paper's
major assumptions flawed and its major conclusions incorrect for
fundamental, technical reasons, which were stated in detail and in
writing. "Criticality Issues" did not respond to those criticisms;
instead, it introduced a new scenario, in which it made the same
technical errors in a new context. Those errors were combined for
publication in "Underground Autocatalytic Criticality." We find no
technical merit in these papers. However, they treat technical matters
and apparently contain no classified material; thus, in accord with the
Laboratory's policy of open and unrestricted research and discussion on
unclassified matters, the authors should be free to submit their paper
for publication in a peer reviewed journal. 
        We do not find any value in these two papers that would justify
their publication, and do not see how to produce such a paper from
them. They contain fundamental errors in concept and execution. They
show no grasp of such elementary concepts as the time scale for the
approach to criticality, the rate of energy release, and the crucial
role of the negative temperature coefficient of the systems treated.
Moreover, they show no appreciation of these points even after they
were pointed out clearly in the review by those who do did understand
them. That is compounded by the shifting scenarios on which the papers
are based and the alarmist estimates of potential effects, which have
become less credible and more shrill throughout the review process.
        The authors have shown little interest in technical suggestions
or inclination to respond to them; thus, it would not appear to be
useful to continue this one-sided discussion. However, it would be
irresponsible for the Laboratory to disseminate untested opinions in
this visible and controversial area. Thus, if this program is
continued, and these individuals remain associated with it, the
laboratory would be well served by establishing a permanent red team,
funded by this program and composed of members from the cognizant
technical divisions, with the responsibility of independently checking
the calculations done by those in the program.

References 

1. C. Bowman and F. Venneri, "Nuclear Excursions and Eruptions from
Plutonium and Other Fissile Material Stored Underground," draft, 22
November 1994.

2. C. Bowman and F. Venneri, "Criticality Issues for Thermally Fissile
Material in Geologic Storage," draft, February 1995.

3. C. Bowman and F. Venneri, "Underground Autocatalytic Criticality from
Plutonium and Other Fissile Material," Los Alamos National Laboratory
document LA-UR-94-4022 (draft), March 1995. 

4. T. Kunkle, "Bowman and Venneri, Part Deux," Los Alamos memo EES-5, 
DDEES/CEP:95-006, 15 February 1995.

5. G. Canavan, "Time Dependence of Neutron Density and Material
Temperature in Thermally Fissile Mixtures," Los Alamos National
Laboratory report , March 1995.

6. W. Stratton, "Neutron Transport Reactivity Calculations of Plutonium
Mixed with and Reflected by Silicon Dioxide," Los Alamos memo, 14
December 1994, Fig. 7 and Table V.

7. W. Stratton, "Pu/SiO2 Overmodulation and Autocatalysis," Los Alamos
memo, 10 January 1994.

8. K. Despain, "keff vs. Assembly Radius from MCNP Calculations-Weapons
Grade Pu in Tonopah Springs Rhyolite," Los Alamos memo for blue team,
December 1994.

9. G. Canavan, "Time Dependence of Neutron Density and Material Temperature in 
Thermally Fissile Mixtures," op. cit.

10. W. Stratton, "The Myth of Nuclear Explosions at Waste Disposal
Sites," LA-9360, October 1983. 

11. W. Stratton, "The Myth of Nuclear Explosions at Waste Disposal
Sites," op. cit.

12. J. Kammerdiener, ""Bowman-Venneri paper on 'Nuclear Excursions',"
Los Alamos X-2 memo to Jas Mercer Smith, Blue Team Leader, 23 December
1994.

13. J. Mercer-Smith, "Super criticality Blue Team Preliminary Report,"
Los Alamos memo, December 1994.

14. E. Fermi, "Elementary Theory of the Chain-reacting Pile, Science, 10
January 1947.