Click for next page ( 33


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 32
Evaluating Pu Disposition Forms Against the Standard BASIS OF COMPARISONS We believe the appropriate spent-fuel "item" for comparison with final weapons plutonium forms from disposition is the LWR heel assem- bly the array of fuel rods and spacers that is the smallest item that could be removed intact from a spent-fuel storage pool or shipping cask. Fuel assemblies for boiling-water reactors (BWRs) are typically 4 m long and .15 m on a side, with a mass of about 250 kg, and for pressurized- water reactors (PWRs) they are typically 4 m long and .25 m on a side, with a mass of about 670 kg. For purposes of comparison with plutonium disposition forms at a nominal 10 years after their creation, we will take "typical" spent fuel to be 30 years old, measured from the time of its discharge from the reactor. (That is, for a comparison in 2020 with a dispositioned plutonium form produced In 2010, we will use spent fuel that was discharged in 1990. For the United States, this would be about In the middle of the age distribution of the spent fuel that will exist in 2020.) We assume that typical spent fuel was irradiated In the reactor to 33,000 megawatt-days per metric ton of initially contained heavy metal (MWd/ MTHM), which is typical of fuel discharged around 1990.~8 The pluto- 18A sense of the distribution of characteristics of the spent-fuel assemblies currently in inventory can be obtained from U.S. data, which show about 107,000 commercial LWR assemblies discharged between 1968 and 1994, some 60,000 of them from boiling-water reactors and about 45,000 from pressurized-water reactors. About 80 percent of the boiling- water reactor assemblies stored as of 1994 had experienced irradiation between 30,000 and 32

OCR for page 32
EVALUATING PU DISPOSITION FORMS AGAINST THE STANDARD 33 nium content would be about 2.5 kg for a typical BWR assembly and about 6 kg for a typical PWR assembly. The gamma-ray dose rate at the surface of a PWR assembly would be about 6,500 rem/hour; one meter from the surface of the assembly at its midpoint the dose rate would be about 800 rem/hour; for a BWR assembly the dose at one meter would be about 500 rem/hour.~9 Spent fuel from plutonium disposition using the once-through MOX option with LWRs would be identical In physical dimensions and mass to the "reference" LWR spent fuel of the corresponding type. It would differ significantly in plutonium content (containing typically 2 to 3 percent by weight plutonium, compared to around 1 percent in the reference spent fuel), and the contained plutonium would differ somewhat In isotopic composition from that in the reference spent fuel. By assumption the spent MOX fuel would differ somewhat from the reference spent fuel in burnup and age since discharge (and therefore In radiation field), inasmuch as we assume the MOX fuel will be irradiated to 40,000 MWd/MTHM, and we compare such spent fuel at age 10 years in 2020 with the middle of the age distribution of spent LEU fuel extant In the United States at that time, aged about 30 years after irradiation to 33,000 MWd/MTHM. The MOX option for plutonium disposition using CANDU reactors is considered here In two variants. . . The standard CANDU option is based on the fuel-assembly con- figuration In general use In commercial CANDU reactors, wherein the fuel assemblies measure about 0.5 m by 0.1 m, with a mass of 24 kg and irradiation to 9700 MWd/MTHM. At typical initial Pu loadings In CANDU MOX, the plutonium content In this spent fuel would be about 1.4 percent. The second option is the CANFLEX approach, which entails com- bining 40 slightly modified CANDU hlel assemblies (each con- taining 43 fuel pins as opposed to 37 in the standard CANDU . 40,000 MWd/MTHM. About 85 percent of the pressurized-water reactor assemblies stored then had been irradiated to between 35,000 and 50,000 MWd/MTHM. See Energy Informa- tion Agency, Spent Nuclear Fuel Discharges from U.S. Reactors: 1994, Report SC/CNEAF/ 96-01, Washington, DC: Energy Information Agency, February 1996, and Carl Walter, "Uniform Descriptive Data," Lawrence Livermore National Laboratory unpublished report, May 1995. 10See CISAC (1995, pp. 270-273, and references cited therein). Because the gamma-ray doses from spent fuel and from dispositioned plutonium forms protected by fission prod- ucts are dominated by 30-year half-life cesium-137 for the period between 5 years and 100 years from the discharge of the fission products from a reactor, knowing the dose rate at one time enables a straightforward calculation of what it would be at other times based on this half life.

OCR for page 32
34 SPENT-FUEL STANDARD FOR DISPOSITION OF EXCESS WEAPON PLUTONIUM design) into a welded rack, which holds 20 tubes in two layers of 10, each tube containing 2 fuel assemblies. This item would have a mass of about 1000 kg. The CANFLEX variant also differs from the standard CANDU option in being irradiated to 25,000 MWd/ MTHM, yielding a plutonium content of 1.8 percent. This approach was proposed In a 1995 study for DOE by Ontario Hydro, the operator of most of the Canadian CANDU reactors, but no further work on the concept has been done since that time.20 Although the CANFLEX approach is not yet a well developed option, we treat it here as a way of taking into account In a preliminary way the technical possibilities for increasing the proliferation resistance of stan- dard CANDU MOX. As indicated above, the can-~n-can~ster approach to the immobiliza- tion track In DOE's two-track ("hybrid") program for disposition of excess weapons plutonium arose as a way to me ze the complications and delays that DOE feared would be imposed by the superposition of the pluton~um-disposition mission on DOE's pre-exishng program to immo- bilize defense high-level wastes at the Savannah River weapons produc- tion complex. The characteristics of the lead~ng-candidate can-~n-canister configuration as developed In the U.S. program as of December 1999 are summarized in Table 3. As discussed further below, this design is intended to be more robust against physical attack than the design dis- cussed ~ this Panel's Interim Report.2i We note that the relevant question In addressing the compliance of the can-~n-canister approach with the spent-fuel standard is whether the plutonium final form In the can-~n-can~ster approach where we take the relevant "item" for evaluation to be the 3 m ~ 0.6 m, 2,500 kg canister is approximately as proliferation resistant as the above-described typical LWR spent-fuel assemblies, not whether it is as proliferation resistant as the homogeneous plutonium-and-fission-product-bearing glass logs pre- viously considered for the immobilization track. 20The CANLEX approach incorporates a modification of the storage system now in use at the Darlington and Pickering CANDU plants, in which 48 tubes are stacked In a structure of four layers. Each tube, of thin stainless steel, contains two assemblies. The tubes are open, but covered with a wire mesh that allows IAE inspectors to verify that the IAEA seals are intact. 21Leonard W. Gray, Gregg Hovis, Robert Jones, and Michael Smith, The Can-in-Canister- Then and Now, Lawrence Livermore National Laboratory Report PIP-99-151, 28 October 1999; Leonard Gray and Thomas H. Gould, Immobilization Team Comments on Interim Report of NAS Panel Review of Spent-Fuel Standard for Disposition of Excess Weapons Plutonium, Lawrence Livermore National Laboratory Report PIP-99-152, 28 October 1999.

OCR for page 32
EVALUATING PU DISPOSITION FORMS AGAINST THE STANDARD TABLE 3 Reference U.S. Can-In-Canister Configuration 35 Term Description Puck Composition: DOE's Synroc ceramic incorporates plutonium into a crystal lattice, with pyrochlore, brannerite, zirconolite and actinide oxide as the principal plutonium-bearing phases and rutile as the primary non-plutonium bearing phase Size: 6.7 cm diameter by 2.5 cm height Mass: 0.5 kg, with 10.5 % Pu by weight, hence 52.5 g Pu Stainless steel cylinders holding 20 pucks each. Size: 7.6 cm diameter by 51 cm height Mass: 2.3 kg empty, 13.6 kg full, 1.05 kg contained Pu Magazine Perforated stainless steel tubes (the final design has not yet been selected) into which 4 cans are loaded on top of each other with glass surrounding each can in the magazine. Size: 8.3 cm diameter by 2.4 m height Mass: 6.4 kg empty, 60.8 kg full, 4.2 kg contained Pu Rack Welded stainless steel frame inside canister consisting of 7 vertical rods (2 cm diameter), 4 scalloped horizontal plates (0.6 cm thick) and a bottom plate into which 7 loaded magazines lock in place in an equally-spaced arrangement. Mass: 47-52 kg empty (range due to variation in base-plate designs), 472.6-477.6 kg full, 29.4 kg contained Pu Canister Stainless steel cylinder that contains rack. Loaded magazines are robotically inserted into rack inside canister. After loading of magazines, radioactive waste glass is poured into the void spaces in the canister. Size: 0.6 m diameter, 3 m height Mass: 500 kg empty, ~2500 kg full, 1523-1568 kg contained glass (depending on height of fill), 29.4 kg contained Pu . Requirements of disposition campaign Disposition of 50 MT of Pu would require 1701 canisters Disposition of 17 MT of Pu would require 578 canisters Disposition of 13 MT would require 442 canisters Sources: Plutonium Immobilization Project, Lawrence Livermore National Laboratory, "Plutonium Immobilization Project Baseline Formulation," UCRL-ID-133089, February 1999; L. Gray and T. Gould, Immobilization Team Comments on Interim Report of NAS Panel Review of Spent-Fuel Standardfor Disposition of Excess Weapons Plutonium, Lawrence Livermore National Laboratory Report PIP-99-152, 28 October 1999; Letter report to Allison Macfarlane from L. Gray, dated December 1999.

OCR for page 32
36 SPENT-FUEL STANDARD FOR DISPOSITION OF EXCESS WEAPON PLUTONIUM COMPARISON MATRICES AND DISCUSSION The matrix in Table 4 provides capsule descriptions and, where pos- sible, quantification of the proliferation-resistance characteristics of the reference spent fuel and of three pluton~um-disposition final forms spent LWR MOX, spent CANDU MOX (conventional variant), and the current lead-candidate can-~n-can~ster configuration. The characteristics are orga- n~zed according to the classes of barriers to proliferation as presented in Table 1. In what follows, we elaborate on these comparisons barrier by barrier, characterizing each option's degree of match or mismatch with spent fuel, with respect to each barrier, as better than comparable, compa- rable, worse Man comparable, or much worse than comparable. (By "better than comparable" we mean that the relevant barriers to proliferation are higher than for the reference spent fuel, and by "worse than comparable" we mean that the barriers are lower than for the reference spent fuel.) These evaluations are then aggregated In another matrix (Table 5) as a basis for our overall judgments about compliance with the spent-fuel standard. Barriers to acquisition: mass and bulk of items The mass and bulk of the canister deemed barriers of moderate importance against acquisition of the plutonium from its storage site in the proliferant-state and subnational-group threat categories (Table 1)- are significantly larger than the corresponding characteristics of the refer- ence (and LWR MOX) spent-fuel assemblies. In mass, the ratio is a factor of 4 to 10 (2500 kg vs. 250-670 kg). The "item" mass In the case of a standard CANDU fuel assembly, however, is a factor of 10 lower than the lightest reference fuel assembly and 100-fold lower than that of the can- in-canister configuration. The "item" size in the CANFLEX variant of CANDU is about 1,000 kg. (An important question, discussed separately below for both the can-~n-can~ster and CANFLEX configurations under the heading of "resistance to energetic attack," is whether these structures are robust enough that dismantling them on site to reduce the mass and radiation field of what needed to be carried away would be impractical (for subnational groups) or more trouble than it is worth (for any proliferator). With respect to this barrier, then, we judge the can-in-canister approach as better than comparable to typical spent LWR fuel, the LWR-MOX and CANFLEX-MOX options as comparable, and the standard CANDU-MOX option as much worse than comparable.

OCR for page 32
37 in - _. Cal o ._ o in o CD cn o o . - _1 cn o in _1 = o - . I U) - a so a) cn - V) o ~ or ._ 0 oo of 0 0 - o g - - 0 - of 0 - .s O , sat ~ =< SIX =~ -~ =m ~ ~~ ax .~. ~ o (d X 'm ~ ~ O ~ ~` ~ = O ~ ~ o ~ O ~ ~ ~ ~ au ~ O ~, ~ O ,~ O ~ ~ O ~ S ,~, ~ ~ ~ ~ ~ ~ ,~; 6-Y o ;' ~ ~ S C ~ ~ ~, . .= ~0 .. 04-D ,,= .m _. `. ~ X ~ g ~ O O ~ ~ O ~, . - ~4 .~0 O c-_ ~ S nl ~ ~ ~ ~ ~0 ._ U) o ~9 ~n O ~ O 1 ~ O _ O ~4 `4 0 _ O `,y Eg ~ U, U) ~ ~ ~ ~ O 3 ~ c o ~ U 1 o ._ 4 - ~4 .= ~ _ ~ ~ 3 t: CD ~ ~ ~ C I Tf b4 ii, - - o o - ~L U]

OCR for page 32
38 di ~ ~ ._' o ED O a . In a no _ 0 Pi a a .~ u, 0 In - P" a, u, u u ~ .s ~ u ~ =0 m ~ O = . m~ ~ ,_1 ~ :^ c ~ ~ ~ ~ ~ C=, o: | ,` c ~ ~ In ~ 0 ~ 0 by ~ ~` ma .3 E c c ~ ; ~ ~ ~ E ~0 =g 0 ~ ~ ~ 0 ~ ~ ' uo up :< c so ~ ' ,:4 c .= o ~ t~ o o {~~ v) : o o o ~ - 4 - u] ~ ~ ~ o 8 ~ o ~ ~ ~u '- O ~ ~ ~ ~ ~ ~ I 10

OCR for page 32
39 ~ an s.= ~ it o oc ~ u ~ s:~= p-- .N ~ ~ ~ ~ A so ~ o ~ O ~ ~ ~ O F in ~ E ._ ' . - ~.o ~ ~ o L' .~ o ~ or o ,~ ~ Pa ~ CO ~ ~ A ~ o U ~ C ~ ~ ~ ~ C O Q ~ ~ ~ ~ 5 in QJ to o :L - _' Ct O ~ BOX CO to in O CL o :` =.= ~ o ~ u) o ~ s~ ~ o ~ ~ >\ co - ~ ~ ~ - .~m ~ ~ ~o ~ w-- o ~ ~ 'E ~, ~ ~ ~ c ~ o C) ~ U) ..o ._ b4 _ _ ~ o bO I ~ C~ :e .~ o CO 4 - _ U) o oo ~ ~, c U) - . - .`n ~ ~ ~ o b4 _ b4 cL Qi CD ~ _ o ~ ~ I_4 CL >4 ~o u] a, _~ 1 .s 1 - l

OCR for page 32
40 dot h ~ o c) .s - ~ ~ A; us ~ o in o - - 04 o in o in o A - - ~ ~ o .~- ~ K~ o ~ to en ~ v ~ + + Go ~ to cO to up V ~ [~! s Go ~ to ~ co 1 1 1 1 pa to m O' ~ '~ ~ o ~ O d. + + o Co ~ o ~ t~ n 'n ~ c~ ~ ~ ~ ' o ~ di 1 1 ~ ~ + + oo ~ o ~ co ~ ~ d~ 1 1 1 1 ~ p~ ~ ~ ~uO l o o - o o ct .m ~ = = ~ ~ x ~ .- ~ ~ - ~it ~ o ~ o .s ~ ~ ~ }~33 ~ ~ ~ 0 a~ ~9'd545- ~ ~ ~ ~ ~ ~ ~ '- ~ ~ ~ ~ ~ t _1 U] . - - .> ~4 - o o - di co o~ o v oo c~ . . ~n o o u) _1 o u, o .= ~d

OCR for page 32
41 ~ - a; v) A En o o - o v) E - ~ En o ~ V'3 I o V) _I E ~ _ o to ~ to Cal ~ hi Cd be _ ~ ~ - o U] - o ~4 o 04 I= ~4 0, 11 o BY a, o o 11 o to UD 11 ~ o u 11 L4 To cn ~ ~ Ct .~ U ~ U o E-~ s~. oo .= ~ ~ ~L, 0 ~o ,~ ._ O ~ U) a, UD .~- ou ~ ~ .' X O X L~ ~ OX g~ U~ ~ OX Cd ^ ~ E~ - o D D o ~ D D t; ~ t;` o o O D J CJ ~ ~ D u ~ _ ~ _ _, _ _ ~ ~ ~ ~ t5 ~ (V ~ ~ ~ ~ ~ O ~ ~ ~ 0- ~ ~ ~ U o ~ ~ t~) ,9 3 ~ ~ ~ ~ ~ ~ ~ ~V ~ o ~ o ~ ~ ~ ~ o o a; _ ~ ~ "Q v 3 v v D v v 3 D v O $ $ E ~ ~ ~ ~ E o o ~ o o ~ ~ ~ ~ o ~ .-o U, ~ ~ ~ ~ ~ CD _ .~ ~ ~ ~ ~ o O ~ O O _ ~ ~ ~ .~ o ~ .0 ~ ~ ~ ~ E r e ~ ~ ~ O , ~ ~ ~

OCR for page 32
42 SPENT-FUEL STANDARD FOR DISPOSITION OF EXCESS WEAPON PLUTONIUM Barriers to acquisition: quantity of material to be acquired The quantity of material that must be acquired to obtain a given quan- tity of plutonium a matter of high importance in the proliferant-state and subnational-group threat categories is about the same in the can-in- canister configuration as in the reference LWR spent-fuel assemblies, both of which have a plutonium concentration near 1 percent by weight. The plutonium concentration is about 30 percent higher in the CANDU MOX configurations and 2-4 times higher in the LWR MOX. For the can-in- canister, theft of a single 2500-kg item would bring 28 kg of plutonium. One would need to steal 4 to 10 of the correspondingly less massive (but similarly bulky) reference LWR spent-fuel assemblies to obtain a similar amount; and one would need to steal nearly 100 of the far less massive (and far less bulky) standard CANDU assemblies to get such a quantity. (We note that, at the 50-70 percent recovery factors that might be assumed for the processing efforts of subnational groups and proliferant states, the 28 kg of Pu in one can-in-canister item or 5-10 reference LWR spent-fuel assemblies would become 14-20 kg.) With respect to this barrier, we judge the can-in-canister and CANDU-MOX options to be comparable to typical LWR spent fuel, and we judge the LWR-MOX option as worse than compa- rable. Barriers to acquisition: hazard to acquirers from radiation The radiation dose at one meter from the midplane of a 10-year-old canister a barrier of moderate importance against acquisition of the plu- tonium from its storage site in the proliferant-state and subnational-group threat categorieswould be about 500 rem per hour.22 This is compa- rable to the 500-800 rem/hr range for 30-year-old spent LWR fuel. Ten- year-old spent LWR MOX would have a radiation field 2-3 times stronger (1000-1500 rem/hr) than the can-in-canister configuration, ten-year old CANDU MOX in the standard configuration a field about ten times weaker (~50 rem/hr). The corresponding field for the CANFLEX variant of CANDU is estimated as 700 rem/hr. The conclusion that the radiation barriers of the can-in-canister and reference LWR spent-fuel assembly are comparable would not hold, how- ever, if the radiation barrier associated with the can-in-canister configura- tion were significantly lower than the range just indicated as a result of 22This is calculated at 2018 assuming fill with 33 kilocuries of Cs-137 per canister in 2008 (Leonard Gray, private communication, December 1999). 6

OCR for page 32
50 SPENT-FUEL STANDARD FOR DISPOSITION OF EXCESS WEAPON PLUTONIUM Barriers to separation: technical difficulty of dissolution and chemical separation In the case of spent LWR fuel, the steps following sawing and chop- ping begin with immersing the pieces in hot nitric acid in order to dis- solve the uranium-oxide ceramic and the plutonium and fission products it contains. Then plutonium and uranium together are separated from the fission products during a first solvent-extraction step. Both the dissolu- tion and the first solvent extraction must be carried out behind shielding to protect the operators from the high radiation field emanating from the fission products. Once these "hot" steps are complete, "cold" operations requiring little or no shielding begin: first the uranium and plutonium are separated from each other with a further solvent-extraction step, and then the plutonium product is "cleaned up" with ion-exchange tech- niques. None of these dissolution and separation steps would be made either appreciably easier or appreciably more difficult by the higher plu- tonium concentration in LWR-MOX and CANDU-MOX spent fuel. In the case of the can-in-canister configuration, we assume in order to avoid double counting of difficulties that mechanical disassembly opera- tion has succeeded in completely separating the ceramic pucks from the rest of the item, so that the input to the dissolution operation is pucks alone. Because the radiation field from the pucks would be far less intense than that associated with the high concentrations of fission products in the glass that has now been removed (and in the comparison spent fuel), the dissolution step could proceed with little shielding. It would begin with grinding the pucks to a powder to ease dissolution. Dissolution itself would not be as straightforward as for spent fuel, however, because the ceramic used in the pucks does not dissolve in hot nitric acid. A proliferator would need to seek out alternative processes (for which task there is considerable relevant information available in the open literature, but no cookbook-style recipe as for spent fuel). Once the ceramic is in solution, the final step would be chemical separation of plutonium from the other constituents of the ceramic, including uranium (which is a con- stituent of the pucks intended to address long-term criticality issues in the repository). The separation step may be accomplished via an ion- exchange process, although this will be complicated by the presence of uranium. In our judgment, the reduction in technical difficulty of dissolution and separation associated with the reduced shielding requirements for the can-~n-canister case, compared to those for typical spent fuel, is sub- stantially offset by the greater difficulty the proliferators would face in mastering the chemistry for these steps. With respect to technical difficulty of dissolution arid separation,

OCR for page 32
EVALUATING PU DISPOSITION FORMS AGAINST THE STANDARD 51 therefore, we conclude that the can-in-canister approach is comparable to typical spent fuel. More obviously, the LWR-MOX and CANDU-MOX options are also comparable in this respect to ordinary LWR fuel. Barriers to separation: quantity of material to be processed Distinct from the question of the quantity of material that must be acquired in the first place to obtain a given quantity of plutonium, which was discussed above under "barriers to acquisition", is the question of the quantity of material that must be passed through the various processing steps. Leaving aside, so as to avoid double counting, the possibility that much of the mass of the can-in-canister configuration could be removed by energetic attack at the site of a theft, the comparison of the mass of material per kilogram of contained plutonium entering the mechanical- disassembly step would be the same as the comparison given above for the masses of material that must be initially acquired that is, the can-in- canister and standard CANDU-MOX cases are comparable to typical spent fuel, and the LWR-MOX case and CANFLEX CANDU-MOX cases are worse than comparable. For the can-in-canister case, the quantity of material needing to be handled goes down by about a factor of 10 following mechanical dis- assembly, the large mass of the radioactive glass and steel structure having been removed at that step. The throughput of material per quantity of contained plutonium is, correspondingly, about 10 times smaller at the dissolution and chemical-separation steps than for the case of typical spent fuel, and 3 times smaller than for LWR-MOX. This deviation is only moderately offset by the smaller anticipated recovery factor of plutonium from the can-in-canister ceramic than from spent fuel: a well run spent- fuel reprocessing operation can recover 85 to 90 percent of the contained plutonium, compared to a Livermore Lab estimate of just over 70 percent as the upper limit for recovery from the can-in-canister ceramic (based on getting 80 percent of the contained plutonium into solution and then losing 10 percent of that in the separation, purification, and conversion- to-metal steps).29 With respect to quantity of material to be processed, we judge the standard CANDU-MOX option to be comparable to typical spent fuel and the CANFLEX CANDU-MOX, LWR-MOX, and can-in-canister options to be worse than comparable. .` 29See Gray and Gould, October 1999, cited at Note 9.

OCR for page 32
52 SPENT-FUEL STANDARD FOR DISPOSITION OF EXCESS WEAPON PLUTONIUM Barriers to separation: hazards to operators The radiation, criticality, and toxic hazards to the operators of the processes used to separate plutonium from the dispositioned form considered separately from the radiation hazards to those acquiring the plutonium-contain~ng items from storage or transport were rated in Table 1 and the accompanying discussion as being of low importance against the threat of host-nation breakout and moderate importance against the proliferant-state and subnational-group threats. As also ~ndi- cated earlier, the toxic component of these hazards is not sufficient to constitute a significant barrier, and we do not consider it further here. Radiation and criticality would be the proliferators' main concerns in the category of hazards of separation. . . . . . .. . .` As noted just above, In the reference case of reprocessing ordinary spent fuel, a high radiation field from the fission products accompanies the plutonium through the dissolution and first solvent-extraction steps. The need for shielding against this field complicates the technical work, as considered above in our discussion of the "technical difficulty of sepa- ration" barrier, and it poses a risk of health-damag~ng or even fatal doses of radiation to the operators In the event of mistakes or in the event of a need for "hands on" repairs during processing. Inasmuch as this hazard is dominated by the fission products rather than by the plutonium, ura- n~um, and other heavy isotopes present, its variations among different spent-fuel forms depend mainly on the fuel's burnup and its age since discharge. The fission products in spent-fuel from the LWR-MOX option, which we assume will have a burnup of 40,000 MWd/MTHM and which we are evaluating at 10 years past discharge, would be generating about twice the radiation field of our designated "typical" spent fuel with its burnup of 33,000 MWd/MTHM and an age of 30 years since discharge. The fission products from CANDU-MOX fuel irradiated to 9700 MWd/ MTHM and aged 10 years would generate a field about half as intense as that from our "typical" LWR fuel;30 and those from the CANFLEX CANDU-MOX option (where the assumed burnup is 25,000 MWd/ MTHM) would be about 20 percent more intense than that from the "typi- cal" LWR fuel. In all these cases, the concentrations of the fission prod- ucts in solution would be similar, as the processing geometries involved presumably would be, so no significant differences in the radiation fields are to be expected from these factors. 30Prior to mechanical disassembly and dissolution, the radiation field from CANDU- MOX is smaller than that of "typical" LWR fuel by a larger factor than this (in the range of 10 to 15), where the additional shortfall comes from geometry in the form of the CANDU fuel assemblies being much smaller.

OCR for page 32
EVALUATING PU DISPOSITION FORMS AGAINST THE STANDARD 53 In the case of the can-in-canister option, the high radiation hazard arising from the fission products persists in the course of processing only up to the point where the plutonium-containing ceramic pucks are extracted from the surrounding fission-product-containing glass. The subsequent dissolution and separation steps entail much smaller shielding require- ments and lower radiation hazards in the event of mistakes or hands-on maintenance needs because the fission products are gone. This part of the "hazards" barrier, then, is significantly lower for the can-in-canister option than for the reference spent fuel and for all of the MOX options. As for criticality, this hazard arises mainly from the possibility that plutonium in solution will find itself in a combination of concentration, geometry, and diluent properties that allow formation of a critical mass.31 The high flux of neutrons from criticality would constitute a potentially deadly radiation hazard, and the energy release could be enough to dam- age or destroy the processing equipment. The main variable among the disposition options, with respect to this hazard, is the plutonium con- centration in solution. In order to avoid discussing the chemistry of disso- lution and separation in excessive detail, we assume here that the con- centration of plutonium in solution is proportional to the plutonium concentration in the items dissolved. If this were so, then the relevant concentration for the CANDU options would be 30 percent higher than in the case of our designated "typical" LWR fuel, and the concentration for the LWR-MOX option would be about 3 times higher. In the case of the can-in-canister option, since only the plutonium-containing pucks and not the initially surrounding glass must be dissolved, the concentration of plutonium in solution would be about 10 times that for typical spent fuel. Thus, the criticality hazard experienced by the operators of pluto- nium recovery processes in the can-in-canister case would be much better than comparable to that associated with reprocessing typical spent fuel, while the radiation hazard would be much worse than comparable. If these two hazards are given similar weight, then these differences sub- stantially offset each other in the overall evaluation of comparability with respect to the barrier of "hazards to the operators," and can-in-canister approach comes out "comparable" in this respect. The LWR-MOX option is better than comparable with respect to criticality and comparable with respect to radiation, and we rate it better than comparable overall. The CANFLEX CANDU-MOX option comparable in both criticality and radia- tion, hence comparable overall; and the standard CANDU option is com- . 31There is also a possibility of this occurring after the plutonium has been extracted from solution that is, when it is in the form of plutonium oxide or plutonium metal but the criticality hazard once this stage is reached does not differ among the disposition options from which the plutonium was obtained and so will not be discussed further here. 6

OCR for page 32
54 SPENT-FUEL STANDARD FOR DISPOSITION OF EXCESS WEAPON PLUTONIUM parable in criticality and worse than comparable in radiation, hence rated worse than comparable overall. In summary, with respect to the barrier of hazards experienced by the operators of plutonium-recovery processes, we judge the can-in-canister and CANFLEX CANDU-MOX options to be comparable to typical spent fuel, the LWR-MOX option to be better than comparable, and the standard CANDU-MOX option to be worse than comparable. Barriers to separation: signatures aiding detection Comparable amounts of radioactive and chemical effluents are re- leased to the environment as a byproduct of the mechanical/chemical (PUREX) process ordinarily used to extract plutonium from either LWR spent-fuel form. The higher concentration of plutonium in LWR-MOX compared to typical spent fuel implies that a smaller processing facility could be used to produce plutonium at the same rate in the case of MOX. On the other hand, dissolution of higher concentrations of plutonium is more difficult, requiring a secondary dissolution step, which in turn gen- erates additional chemical effluents. Based on these considerations, we judge that LWR-MOX is comparable to typical spent fuel with respect to signatures aiding detection, as are the standard and CANFLEX CANDU- MOX options. ~ . The signatures comparison is more complex in the case of the can-in- canister approach. The radioactive and chemical signatures available for detecting the separation of the ceramic pucks from the glass and the sub- sequent processing of the former to extract the contained plutonium are different from those available in extracting plutonium from spent fuel using the PUREX process. The three main differences are: the absence, in the case of the can-in-canister approach, of detectable fission products such as the noble gas Kr-85, released during reprocessing of spent fuel; the need to use processes other than PUREX to separate and dissolve the ceramic and extract the contained plutonium; and the higher concentra- tion of plutonium in the ceramic pucks compared to that in typical spent fuel, which, all else equal, will reduce the scale and/or duration, and hence the detectability, of extraction operations. Taking these differences into account, we judge that the detectability of processing to extract plutonium in the case of the can-in-canister option would never be better, and in most cases would be worse, than for typical spent fuel across the range of diversion scenarios involving the host state, proliferant states, and subnational groups. The detectability of processing in the case of the can-in-canister option can be increased, however, either

OCR for page 32
EVALUATING PU DISPOSITION FORMS AGAINST THE STANDARD 55 directly (by the addition of various infrared active chemical "taggants" to the canister that would be released when it is disassembled) and/or indi- rectly (by adding constituents to the ceramic that would complicate the chemical processing and thus increase the scale and duration of emis- sions). Our preliminary survey of such signature enhancement techniques has indicated a wide variety of possibilities. While these require further analysis with regard to technical feasibility and cost, we judge this to be a promising option for increasing detectability. We also note that detection of a chemical taggant in the vicinity of a storage site could alert authori- ties that a canister had been disrupted on site; this detection possibility, if coupled with adequate response times of security forces, could reduce somewhat the concerns about vulnerability of the can-in-canister configu- ration to such on-site attack. With respect to signatures aiding detection of plutonium sepa- ration, we judge that the LWR-MOX and CANDU-MOX options are comparable to typical spent fuel. We judge the can-in- canister option worse than comparable on this criterion, although there is a high likelihood that it could be made comparable through the use of additives to increase detectability, and pos- sibly it could be made better than comparable in this way. Barriers to utilization: deviation of isotopic composition . The "reactor-grade" plutonium in typical spent fuel differs from "weapons-grade" plutonium in having a higher neutron background (deriv- ing mainly from the higher Pu-240 and Pu-242 content), a higher heat- generation rate (deriving mainly from higher Pu-238 content), and a higher surface gamma-ray dose (deriving mainly from higher Pu-241 and Am-241 content). As indicated earlier, these differences present addi- tional difficulties to the designers, producers, and handlers of nuclear weapons made with reactor-grade plutonium, compared to those encoun- tered when weapons-grade plutonium is used. The barriers posed by these difficulties for the utilization in weapons of plutonium of isotopic composition differing from weapons grade are not insurmountable; for reasons explained above, we have rated them as "moderate" in impor- tance against the host-nation and proliferant-state threats and "low" in importance against the subnational-group threat. As seen in Table 4, the plutonium in the LWR-MOX option has about the same Pu-240 + Pu-242 content as plutonium in the reference spent fuel, a Pu-238 content ranging from less than half as great to about the same, and a Pu-241 + Am-241 content ranging from 50 percent greater to twice as high; on balance, we rate the isotopic barrier of the LWR-MOX

OCR for page 32
56 SPENT-FUEL STANDARD FOR DISPOSITION OF EXCESS WEAPON PLUTONIUM option as comparable to that of typical spent fuel. The standard CANDU- MOX option is likewise comparable to typical spent fuel in this respect, with Pu-240 + Pu-242 concentration in the same range, Pu-238 content considerably smaller, and Pu-241 + Am-241 content about 50 percent lower. The CANFLEX-CANDU option has Pu 240 + Pu-242 content about 50 percent higher than that in typical spent fuel, considerably smaller Pu-238 content, and Pu-241 ~ Am-241 content about the same; we rate this as better than comparable to typical spent fuel. In the case of the can-in- canister option, the isotopic composition of the contained plutonium is unaltered from that of the weapons plutonium provided to the process; we rate this as much worse than comparable to typical spent fuel. In summary, with respect to the isotopic barrier to utilization of the plutonium in nuclear weapons, we judge the LWR-MOX and standard CANDU-MOX options to be comparable to typi- cal spent fuel, the CANFLEX-CANDU option to be better than comparable, and the can-in-canister option to be much worse than comparable. .` Overall judgments on comparability and compliance Table 5 summarizes our judgments on the comparability, with typical commercial spent LWR fuel, of the four disposition forms LWR-MOX, standard CANDU-MOX, CANFLEX CANDU-MOX, and the reference can-in-canister configurationin respect to all of the proliferation barriers considered here. The relative-importance ratings of these barriers against the three classes of threats are indicated (from Table 1) in the first column. In the remaining columns, each disposition form is rated on each barrier as being "comparable" to typical spent fuel, "worse than comparable", "much worse than comparable", or "better than comparable". (Again, 'worse' means a lower barrier to proliferation than provided by the refer- ence spent fuel, and 'better' means a higher barrier to proliferation.) We found no instance in which a rating of "much better than comparable" was warranted. Our judgments on compliance with the spent-fuel standard are then based on "summing," in a manner of speaking, the departures from com- parability for each disposition option. Thus: The LWR-MOX option is worse on Pu concentration and quantity of material to be processed; better on signatures aiding detection of acquisition and hazards to operators; and comparable on the remaining seven barriers. The importance ratings of the two bar- riers on which LWR-MOX is worse than reference spent LWR fuel

OCR for page 32
EVALUATING PU DISPOSITION FORMS AGAINST THE STANDARD 57 are only slightly higher than the importance ratings of the two barriers on which it is better. (The average of these differences is about half a level, that is, half of the difference between "low" and "medium" or between "medium" and "high.") The magnitude of the net deviation from reference LWR spent fuel in importance- weighted barriers to acquisition, separation, and utilization of the contained plutonium is small, well within the spent-fuel standard's requirement that the plutonium be "roughly as difficult to acquire, process, and use in nuclear weapons" as that in the reference spent fuel. Accordingly, we judge the LWR-MOX option to be compli- ant with the spent-fuel standard. The standard CANDU-MOX option is much worse on item mass/ bulk; worse on radiation hazard to acquirers, signatures aiding detection of acquisition, and hazards to operators; and compa- rable on the remaining seven barriers. The "worse" and "much worse" performances are uncompensated by any "better" perfor- mances, and they occur in barriers with importance ratings aver- aging near "moderate" and extending to "moderate to high". The magnitude of this net deviation from reference spent LWR fuel in importance-weighted barriers to acquisition, separation, and utili- zation of the contained plutonium is too large, in our judgment, to meet the spent-fuel standard's requirement of "roughly as diffi- cult..." Accordingly, we judge the standard CANDU-MOX option to be noncompliant with the spent-fuel standard. The CANFLEX CANDU-MOX option is worse on difficulty of on- site reduction of mass & radiation; worse on quantity of material to be processed; better on isotopic composition; and comparable on the remaining eight barriers. The importance ratings of the two barriers on which this option is worse than reference LWR spent fuel are significantly higher than the importance rating of the bar- rier on which it is better, which means that the one "better" perfor- mance does not fully compensate even the less important of the two "worse" performances; the more important of the "worse" performances, which is completely uncompensated, entails "high" importance ratings against two of the three classes of threat. The magnitude of the net deviation from reference LWR spent fuel in importance-weighted barriers to acquisition, separation, and utili- zation is not as great as in the case of the standard CANFLEX CANDU-MOX option, but it is enough to render this case a close call. We judge the compliance of the CANFLEX CANDU-MOX option with the spent-fuel standard to be marginal. The reference can-in-canister option is much worse on isotopic com- position; worse on quantity of material to be processed; better on . .`

OCR for page 32
58 SPENT-FUEL STANDARD FOR DISPOSITION OF EXCESS WEAPON PLUTONIUM item mass/bulk and technical difficulty of disassembly; and com- parable on five other barriers. It could be from much worse to comparable on difficulty of on-site reduction of mass and radia- tion (this evaluation depending on the outcome of a recommended program of analysis and testing), and from worse to better on signatures aiding detection of separation (this evaluation depend- ing on the outcome of investigation of the potential of additives to enhance these signatures). Taking into account the importance ratings of the barriers involved, the two "better" performances could be deemed to compensate for enough of the "worse" on quantity of material to be processed and the "much worse" on isotopic composition to permit a judgment of compliance with the spent-fuel standard if analysis and testing showed performance to be comparable with respect to difficulty of on-site reduction of mass & radiation and if it proved possible, using additives, to make the signatures aiding detection of separation at least comparable to those for typical LWR spent fuel. Accordingly, we judge the compliance of the reference can-in-canister option with the spent- fuel standard to be contingent on the outcome of efforts to clarify this option's resistance against on-site attack and to improve its signatures aiding detection of separation activities. Alternatives to the Reference Can-in-Canister Configuration In the event that the current can-in-canister configuration were to be found noncompliant with the spent-fuel standard, there are other modifi- cations to this approach to plutonium immobilization that could and should be considered as potential remedies. The possibilities, beyond the use of additives to increase detectability of processing as already dis- cussed above, include the following: . . Further modification of theframe-and-can arrangement to increase resis- tance to cutting and/or explosive attack. Considerable efforts along these lines have already been made, but further advances may be possible. Addition of materials to the puck composition to increase the difficulty of chemical extraction of the plutonium. Some of the materials that pose the greatest difficulties in this respect are already part of the cur- rent composition, and have been taken into account in our evalua- tion, but further ingenuity in complicating the mix may be possible. Reduction in the concentration of the plutonium in the ceramic. This is technically easy but would, obviously, increase the amount of

OCR for page 32
EVALUATING PU DISPOSITION FORMS AGAINST THE STANDARD . 59 plutonium-bearing ceramic material to be produced and the quan- tity of radioactive glass and number of canisters to be provided for disposition of a given quantity of plutonium. Addition of cesium directly to the ceramic. This measure would extend the main radiation barrier into the chemical dissolution and sepa- ration steps, corresponding to the situation with spent fuel. This would complicate (and increase the cost of) the puck-production process, but the gain in proliferation resistance might be worth it. Replacement of pucks with pellets. Instead of pucks, the plutonium- containing ceramic could be formed into pellets or marbles, which could be placed in a wire mesh for loading into the canister or added to the molten glass as it is being poured. These modifica- tions would require surmounting certain difficulties, but the result- ing more homogeneous distribution of ceramic in the glass would make it more difficult for thieves to separate the plutonium- containing material at the site of the theft. All of these possibilities have been considered in at least a preliminary way by DOE and its contractors. All of them pose difficulties as well as offering potential for increased proliferation resistance. To investigate these trade-offs would have been beyond our mandate, and we have not done so. But we believe all these approaches would be worth revisiting in the event that the current configuration is ultimately judged noncompliant with the spent-fuel standard. Some combination of them and perhaps others not mentioned her~might suffice to bring the can-in-canister option into compliance.32 32Some of these approaches would undoubtedly increase costs, but we would reiterate in this connection the emphatically stated view of the previous CISAC plutonium reports that security is primary in this matter and cost secondary (unless and until costs become high enough to prevent taking the steps that security requires).

OCR for page 32