Analysis
of information released by DOE on Possible Use of
“Reactor
Grade” Plutonium in Nuclear Weapons
Karanam
.L. Ramakumar
Karanam.ramakumar@gmail.com
Abstract:
Information released by Department of Energy (DOE) on Possible Use of “Reactor
Grade” Plutonium in Nuclear Weapons has been critically analysed. The analysis
is based on published scientific literature and the information released by the
Government Agencies. The primary aim of the analysis is to estimate the Pu-240
content in the plutonium used in the devise exploded in 1962. From the
available literature, it was possible to conclude that Pu used in the 1962
device contained about less than 20% Pu-240 (in the range 15 to 18%).
Possibility of using non-weapons grade plutonium in the nuclear weapons is also
explored.
Keywords:
nuclear weapons; plutonium; reactor grade; fuel grade; Calder Hall nuclear
reactor; DOE database on nuclear explosions
Introduction
The
US government
had announced in 1994[1] that a nuclear weapons test using 'reactor grade'
plutonium was carried out at the Nevada Test Site in 1962. The information was
declassified in July 1977 but additional information was provided in 1994. The
yield of the blast was less than 20 kilotons. According to DOE, extensive
nuclear test database and predictive capabilities maintained by US and the low
yield test revealed that weapons can be constructed with reactor-grade
plutonium. The plutonium was provided by the United Kingdom under the 1958
United States/United Kingdom Mutual Defence Agreement. Various documents on
this subject are available from different sources on the Internet. Based on the
downloaded documents an analysis has been carried out to review the information
released by the DOE. It may be mentioned that the analysis is based on
scientific publications and the information released by Government Agencies.
The primary aim of the analysis is to estimate the Pu-240 content in the
plutonium used in the devise exploded in 1962. Possibility of using non-weapons
grade plutonium in the nuclear weapons is also explored.
Analysis
of information provided by DOE
Details
of nuclear test with “reactor grade” plutonium:
About 1149 nuclear tests/detonations
were carried out by the USA between 1945 and 1992[2]. Ninety-six of them were
carried out in 1962. Two of these 96 tests used plutonium supplied by UK. The
details are given below:
Test No.
|
Test
|
Date (mm/dd/yyyy)
|
Sponsor
|
Location
|
Hole
|
215
|
Pampas Accidental
release of radioactivity detected offsite
|
03/01/1962
|
LANL/UK
|
NTS
|
U3al
|
299
|
Tendrac
|
12/07/1962
|
LANL/UK
|
NTS
|
U3ba
|
Test No.
|
Time
(GMT)
|
Latitude
(degrees)
|
Longitude
(degrees)
|
Surface Elevation
(meters)
|
Type
|
Purpose
|
Yield Range
|
215
|
19:10:00.09
|
37.041
|
-116.030
|
1196
|
Shaft
|
Joint US-UK
|
9.5 kt
|
299
|
19:00:00.10
|
37.052
|
-116.030
|
1202
|
Shaft
|
Joint US-UK
|
Low
|
Nomenclature
on different plutonium grades
The
DOE fact sheet merely mentioned that reactor grade Plutonium was used in 1962
test. The exact isotopic composition of the plutonium remains classified. It
may be noted that until 1976, DOE had designated three different grades of plutonium[3].
These are:
- Super weapons grade, less than 3% Pu-240
- Weapons grade, less than 7% Pu-240
- Reactor grade, 7% or more Pu-240
- Super weapons grade, less than 3% Pu-240
- Weapons grade, less than 7% Pu-240
- Fuel grade, ≥ 7% and < 19% Pu-240
- Reactor grade, ≥ 19% Pu-240
As
the first de-classification on using reactor grade Pu in nuclear weapons
testing by DOE was in 1977, some lingering uncertainty with regard to “reactor
grade” nomenclature existed. Added to this, which definition or designation,
that of the old or new scheme applies to the 1962 reactor grade plutonium
weapon test, has not been officially disclosed.
There
was, however, consensus that UK plutonium came from Calder Hall nuclear power
reactors during late 1960s as at that time only Calder hall and Chapelcross
nuclear power reactors were operating. Primary purpose of these reactors was
for production of weapons grade plutonium and intermittently for electricity
generation. Brief particulars of Calder Hall reactor[4]
relevant to the present context are:
Table-1. Relevant data
for Calder Hall reactor
Electrical output
(gross)
|
46 MWe
|
Thermal output
(gross)
|
182 MWt
|
Efficiency
|
23 %
|
Material
|
Natural uranium
metal
|
Mass of uranium per
reactor
|
120 tonnes
|
Different interpretations were put forward to identify the grade of plutonium used in the test. As the production and isotopic composition of plutonium depends on burnup, it would be interesting to deduce the burnup of the Calder Hall fuel during the period of interest. From the available literature it is seen that in case of MAGNOX type reactors, the cross-over from fuel-grade (Pu-240 ≤ 18%) to reactor-grade plutonium (Pu-240 > 18%) occurs at a burnup of around 3500 MWd/t [5]. Thus, estimating burnup of the MAGNOX fuel which was reprocessed for separating Pu for shipment to the USA is desirable to possibly identify the quality of plutonium used in the 1962 nuclear weapon test.
Burnup
of the fuel could be estimated from the data given in the Table-1.
Assuming the reactor was continuously operating for 300 days,
Burnup
(MWD/T) = (182 (MWt) x 300 (days))/120 (tonnes) = 455 MWD/T
However,
for electricity production the reactors need to be operated at higher burnups.
This would also result in increased Pu-240 content in the fuel. Hinton[6]
mentions that at its peak, the station generated 196 MW — four times as much
power as it did when it opened. This is very close to the figure given by McCrickard[7].
Then the thermal power is estimated to be 784 MWt and assuming that the
reactors operated again continuously for 300 days, the burn up would be 1950
MWD/T. For producing plutonium with > 18% Pu-240, at least some of the
fuel elements in these reactors should have been irradiated to more than 3500
MWD/T burnup and the reactors continuously operated for extended periods of
time. For this purpose, literature [8-11] was studied. Following observations could be made:
(a) Due
to some systematic faults in the fuel, Calder Hall and Chapelcross reactors
operated at somewhat lower burnups than 2000 MWD/t during initial periods of
operation[8].
(b) Later, MAGNOX civil
reactor fuel elements and also some Calder Hall and Chapelcross test fuel
elements fabricated with modified specifications to address the causes of
failure were subjected to test irradiation in Calder Hall and Chapelcross
reactors and had undergone higher irradiations (channel average burnup of more
than 3000 MWD/t) resulting in higher Pu-240 content (> 18%)[9].
(c) The
IAEA publication [10] gives average and maximum core burnup of Calder Hall fuel
as 2700 and 3900 MWD/t respectively for the year 1962. As the test was carried
out in July 1962, it is reasonable to assume that the fuel might have been
discharged during January 1962 for reprocessing to separate plutonium for
shipment to the USA. It is therefore necessary to estimate the average and
maximum core burnup at the time of discharge in January 1962 for deducing
Pu-240 content.
(d) Stewart
in his publication[11] gives details of irradiation history of fuel elements in
Calder Hall reactors. Some of the fuel elements loaded in 1958 were still undergoing
irradiation trials as of August 1963 and these fuel elements did see burnup
exceeding 3000 MWD/Te as of August 1963. Suppose some of these fuel elements
were selectively removed for reprocessing to separate plutonium for shipment to
the US for the 1962 test. As the 1962 test was carried out in July 1962, it is
reasonable to assume that about six months was needed to complete the process
of the discharge the fuel elements, reprocessing followed by shipment of the
separated plutonium to the US for assembling the devise. It would of interest
to deduce burnup of these elements as on January 1962.
What could be the Pu-240 content in the plutonium used in the 1962
device?
As mentioned earlier, the DOE fact sheet
merely mentioned that reactor grade Plutonium was used in 1962 test. At that
time according to DOE, plutonium with more than 7% Pu-240 was designated as
reactor grade. Subsequently DOE modified
its nomenclature and defined plutonium with more than 19% Pu-240 as reactor
grade. Plutonium with Pu-240 content in the range between 7 and 19% was
designated as fuel grade. It would be of interest to deduce Pu-240 content in
the plutonium used in 1962 test to know if it was fuel grade (Pu-240 content
< 19%) or reactor grade (Pu-240 content > 19%). One may not get the
accurate figure but a reasonable accurate band may be enough for discussion.
Isotopic composition of plutonium produced in a nuclear reactor, among
other things depends on the burnup. Very low burnups are needed to produce
weapon-grade plutonium with Pu-240 content less than 7%. Higher burnups result
in more and more Pu-240 production. A knowledge of burnup the fuel was
irradiated to would be useful in this regard.
The IAEA publication [5] gives average and maximum
core burnup of Calder Hall fuel as 2700 and 3900 MWD/t respectively for the
year 1962. As the test was carried out in July
1962, it is reasonable to assume that the fuel might have been discharged
during January 1962 for reprocessing to separate plutonium for shipment to the
USA. It is therefore necessary to estimate the average and maximum core burnup
at the time of discharge in January 1962 for deducing Pu-240 content. The
maximum core burnup at the end of 1961 was shown to be about 3000 MWD/te.
Further, It should be mentioned that average burnup of a channel and an
individual fuel pin within the channel could be much higher than the core
values.
Papers published by Steward in 1963 and 1964 listed channel average and
individual pin burnup values. Some relevant 1963 values given in Table 5 of the
publication are reproduced in the Table below:
High irradiation experience with
natural uranium MAGNOX fuel in Calder Hall (CR) reactors
Description of fuel element
|
Reactor
|
Date loaded
|
Irradiation position at August 1963
|
||
Number
of channels
|
Irradiation
MWD/te
|
||||
Mean channel
|
Peak fuel element
|
||||
Calder Hall Mk. 1A fuel element with coarse- grained cans
|
CR.1
|
September 1958
|
8
|
3440
|
4500
|
CR.3
|
June 1958
|
205
16
32
|
2600
4180
1900
|
3400
5150
2500
|
|
These values are as on August 1963.
Thus, it is seen that channel average and individual pin burnup values are
indeed much higher than the core values.
As
the test was carried out in July 1962, it is reasonable to assume that some of
the pins might have been discharged during January 1962 for reprocessing to
separate plutonium for shipment to the USA. It is therefore necessary to
estimate the channel average burnup and peak fuel element burnup at the time of
discharge in January 1962. For deducing Pu-240 content.
Hardy and Lawton[12] listed the fuel rating values for the inner zone configuration for Calder hall reactor core. Maximum fuel rating was 3.48 MW/te and the average value for the channel when computed is 2.66 MW/te. At 3.48 MW/te fuel rating the burnup of the element for one year irradiation is calculated as 1270 MWD/te. Calculating back the peak fuel element burnup at the time of discharge in January 1962 comes to about 3000 MWD/te. At this value Pu-240 content in the plutonium may not be more than 18%.
It should also be mentioned that
even though the maximum fuel rating of 3.48 MWD/te was mentioned, it seems the
fuel did not actually see this level of fuel rating. Otherwise the peak fuel
element burnup should have been more than 6000 MWD/te instead of 5150 MWD/te as
given in the Table 5 of Stewart. Noting that the fuel elements were loaded in
June, 1958, the peak fuel element burnup of 5150 MWD/te would reach after 5
years at 3.48 MWD/te fuel rating for 300 days of operation every year or 2.8
MWD/te fuel rating for the whole year of 365 days. Either way, at the time of
discharge in January 1962, the burnup could have been about 3400 MWD/te.
On the other hand, if it is assumed
that full channel was discharged for reprocessing in January 1962, then the
channel average burnup in January 1962 would be about 2850 MWD/te.
Thus, we have three values of burnup
at the beginning of 1962: 2850, 3000 and 3400 MWD/te. Pu-240 content in
plutonium at these burnup values range between 15 and 18%.
Thus, it may be assumed that Pu-240 content in the plutonium
used in the 1962 US nuclear weapon test might be in the range between 15 and
18%.
Reactor-
or fuel- grade plutonium in DOE test
Thus,
it is seen from the published literature that the highest burnup of Calder hall
fuel during the period under consideration (1956–1962) ranged between 2850
MWD/t and 3400 MWD/t with Pu-240 content between 15 and 18%. It was reactor
grade plutonium as per pre-1976 definition or fuel grade Plutonium as per
post-1976 definition of Plutonium grades. Clinching proof to this estimate
comes from none other than another US DOE Publication. The DOE’s publication of 1996
[13] clearly mentions that “under the Mutual
Defence Agreement with the United Kingdom from 1959 to 1980, the United States
acquired a total of 5.4 MT of plutonium (5360 kilograms) in exchange for 6.7
kilograms of tritium and 7.5 MT of highly enriched uranium”. The report further
says that this plutonium was of “primarily fuel grade” plutonium. That means
Pu-240 content in the material used was between 7% and 19%. In all probability,
the 0.4T of plutonium received from other countries was reactor grade. One may
refer to the Table 7 on page 44 of the report. It would be very clear why it
had to be reactor grade. The quality of plutonium could be undoubtedly reactor
grade and also the period of Pu receipt is also way beyond 1962.
Surprisingly there was no reference to this statement in the DOE publication of
1996 by any analyst/interpreter.
A
note about using plutonium with more than 7% Pu-240 in nuclear weapons
Typically, nuclear
weapons are designed so that a pulse of neutrons will start the nuclear chain
reaction at the optimum moment for maximum yield; background neutrons from
plutonium-240 can set off the reaction prematurely, and with reactor-grade
plutonium the probability of such "pre-initiation" is large. Pre-initiation [14] can substantially
reduce the explosive yield, since the weapon may blow itself apart and thereby
cut short the chain reaction that releases the energy. If pre-initiation occurs
just at the moment when the material first becomes compressed enough to sustain
a chain reaction, the explosive yield would be of the order of one or a few
kilotons. Historically this yield is referred to as the "fizzle
yield". The term “Fizzle” had become synonymous to failure of a nuclear
weapons test. Treating neutron kinetics in a typical explosion device as purely
deterministic in nature, it has been shown[15] that the energy yield of a
nuclear explosive decreases by one and two orders of magnitude if the 240Pu
content increases from 5 (nearly weapons-grade plutonium) to 15 and 25%,
respectively. Thus in typical nuclear weapon with an expected yield of 20 kT
with 5% Pu-240 device, the yield would reduce to 2 KT and 0.2 KT for 15%
240-Pu and 25% 240-Pu devices respectively. These are “fizzle yields”
according to conventional nomenclature but are by no means fizzle. During 1958,
the USA conducted series of nuclear weapons tests at different locations at Pacific Proving Grounds
[16]. In the nuclear test
with code name Juniper, it was indicated that a minimum threshold yield of 0.2
KT for a boosted primary is required for a successful test. The test device had
a diameter of ~37 cm, and a length of ~38 cm. It weighed ~76 Kg and resulted in
65 kT yield. With optimum design configuration supported by theoretical
calculations, even with up to 25% Pu-240 containing plutonium device, it is
possible to realise this yield.
Observations on the
information released by DOE
According to DOE
1. It was a successful test
2. The yield was less than 20 kilotons
3. The United States maintains an extensive nuclear test data
base and predictive capabilities. This information, combined with the results
of this low yield test, reveals that weapons can be constructed with
reactor-grade plutonium.
The information
released by the DOE is cryptic and many questions arise.
What were the criteria
followed by the DOE to declare a test is successful or not?
Was the process of
testing a success? Was the design of the weapon a success? Was the DOE’s
predictive capability a success?
The reason why the DOE
did not give the yield figure is also very intriguing. There are many instances
where the DOE did publish the actual yield figures including very low values
(less than 10 kiloton). DOE had no hesitation to give these figures for weapons
grade plutonium. Bur for “reactor-grade” plutonium despite the fact that
“Reactor-grade plutonium is significantly more radioactive which complicates
the design, manufacture and stockpiling of weapons”, DOE felt otherwise.
Another question that comes to mind is how low was this figure? Finally, one is
tempted to interpret the last point again in different ways. Did the DOE indeed
predict very low yield in this test? Or were the results of very low yield, formed
the basis for reassessing the predictive capabilities? Or did this test prompt
the DOE to revisit the classification terminology for plutonium grades to
include fuel-grade in between weapons-grade and reactor-grade? Could it be also
the reason why the DOE declared that the plutonium received from other
countries including United Kingdom was primarily fuel-grade? One never knows.
But the analysis given above indicates that the plutonium used in the test
could be fuel-grade with Pu-240 content in the range 12 to 16%. One may give
any name to the grade of plutonium used in the 1962 test. It is immaterial
whether it was reactor grade or fuel grade. What is pertinent is the
Pu-240 content. Treating neutron kinetics in a typical explosion device as
purely deterministic in nature, it has been shown [15] that the energy yield of
a nuclear explosive decreases by one and two orders of magnitude if the 240Pu
content increases from 5 (nearly weapons-grade plutonium) to 15 and 25%,
respectively. Thus, in typical nuclear weapon with an expected yield of 20 kT
with a 5% Pu-240 device, the yield would reduce to 2 KT and 0,2 KT for a 15%
240-Pu and 25% 240-Pu devices respectively. These are “fizzle yields” according
to conventional nomenclature but are by no means fizzle. One should have no
doubt about nuclear weapons built with reactor grade plutonium. 1962 test was
the first test with RG plutonium and there is no way to say with confidence
that the Pu-240 content was between 20-23%. Safe statement could be Pu-240 was
more than 12% as at that time this plutonium was indeed reactor grade.
Sahin’s work shows how
the yield of a weapon could change depending on the Pu-240 content. This is
where knowledge of actual yield value becomes important and crucial. On both
the counts the test would be deemed to be successful.
Conclusions
The
purpose of this analysis in not intended to address the use of reactor- or
fuel- grade plutonium in nuclear weapons. In the current context of the spectre
of nuclear terrorism and the possibility of radiation dispersion devices, it is
always safe to secure any plutonium irrespective of its isotopic content. No
purpose is served by proving or disproving of the use of “reactor grade” or
“fuel grade” plutonium in nuclear weapons. Plutonium being plutonium it has to
be treated with respect. The IAEA is absolutely right in declaring any
plutonium (except that with more than 80% Pu-238 content) to be brought under
safeguards and securing. It is also mandatory on the part of the countries
having Plutonium containing more than 80% Pu-238 to secure to prevent any
malicious and terrorist activities.
References
1.
DOE Fact Sheet: https://www.osti.gov/opennet/forms.jsp?formurl=document/press/pc29.html#ZZ0
(US Department of Energy, 1994)
2.
United States Nuclear Tests: July 1945
through September 1992, DOE/NV--209-REV 15, DOE (2000) and https://www.nnss.gov/docs/docs_LibraryPublications/DOE_NV-209_Rev16.pdf
4.
Description of the Magnox Type of Gas
Cooled Reactor (MAGNOX), S. E. Jensen and E. Nonbol, Riso National Laboratory,
Roskilde, Denmark NKS-2 (1998) https://inis.iaea.org/collection/NCLCollectionStore/_Public/30/052/30052480.pdf
5.
David Albright, Frans Berkhout and William
Walker, Plutonium and Highly Enriched Uranium 1996: World Inventories,
Capabilities and Policies, Stockholm International Peace Research Institute
Oxford university press, New York (1997)
6.
Lord Christopher Hinton, Calder Hall Nuclear Power Station, http://www.engineering-timelines.com/scripts/engineeringItem.asp?id=778
7.
J. McCrickard, The Development of Calder Hall and Chapelcross as
Base Load Nuclear Power Stations, in Proceedings of the Conference on Operating
Experience With Power Reactors, Vol. I, IAEA, Vienna, 4-8 June 1963 pp.
407-423 https://inis.iaea.org/collection/NCLCollectionStore/_Public/44/064/44064225.pdf
8.
A Johnson, “Magnox Fuel Cycles,” Operating Experience with Power
Reactors, Volume II, International Atomic Energy Agency, Vienna, 1963
9.
The Development of Uranium-Magnox Fuel Elements for an Average
Irradiation Life of 3000 MWD/te, H. K. Hardy et al., The Journal of the British
Nuclear Energy Society, Volume 2, 1963 (January 1963)
10. Operating experience
with nuclear power stations in member states until end 1970 https://inis.iaea.org/collection/NCLCollectionStore/_Public/02/014/2014309.pdf.
11. J. C. C. Stewart,
C.B.E., B.Sc., Development and Manufacture of Magnox Fuel, Proc. Instn. Mech.
Engrs., 1963-64 Vol 178 Pt I No 9 (227-240)
12. Hardy HK, Lawton H
(1958) the assessment and testing of fuel elements, Second United Nations
international Conference on the Peaceful Uses of Atomic Energy, a/conf.15/p/3o6
https://www.osti.gov/servlets/purl/4283557
13. Plutonium: The First 50
Years, DOE/DP-0137, DOE (1996) http://fissilematerials.org/library/doe96.pdf
15. Siimer Sahin, Reply to
"Remarks on the Plutonium-240 Induced Pre-Ignition Problem in a Nuclear
Device", Nuclear Technology, 54, 431-432 (1981)
No comments:
Post a Comment