ML19323B684

Preliminary Proposal - Emergency Reactor Off-Gas Decontamination Sys. Ltr to Util Encl

ML19323B684
Person / Time
Site:	Crane
Issue date:	06/22/1979
From:	Merriman J UNION CARBIDE CORP.
To:
Shared Package
ML19323B642	List: ML19323B645 ML19323B647 ML19323B649 ML19323B651 ML19323B654 ML19323B657 ML19323B662 ML19323B667 ML19323B670 ML19323B674 ML19323B677 ML19323B680 ML19323B684
References
RTR-NUREG-0662, RTR-NUREG-662 K-ET-244, NUDOCS 8005140071
Download: ML19323B684 (4)

Text

{{#Wiki_filter:-

*V.

., ' (?lij.yihj-{A 't-- 3,;, ! s, -80051% @-.. WM w; q.;;,i~ y 4a '.K/ET-244/i AS_ - '7. g - r. v4 ':.. t. ..p " : ? i',i .,1 -~... PRELIMIKARY PROPOSAL - DERGENCY REACIDR OFF-GAS DECONTAMINATION SYSTEM i) '1 Prepared by the Staffs of the Enrichment Technology Division and i Process Engineering Division Jtne 22,1979 I' t, c l ) Oak Ridge Gaseous Diffusion Plant Union Carbide Corporation-Nucle ~ar Division f Oak Ridge, Tennessee r l' l (Sponsor: J. R. Merriman) 3 f

it M u gg @.@ S. Y gg ?A y, n a w:_ ppP %(g. ; EMERGENCY REAGDR OFF-GAS n DECONTAMINATION SYSTEM Irradiation of nuclear fuel results in the production of substantial Fortunately, the bulk of the quantities of radioactive by-products. fission jiroduct inventory is retained within the ceramic matrix of the s. - reactor fuel. A significant fraction of the noble gases and halogens,. however, migrates through the fuel pellets and accu::ulates in the fuel-Under nornal operation, small anounts of these fission products. rod voids. then enter the primary reactor coolant system via fuel claMing failures. ~ Xenon-133 and Xenon-135 constitute the major source of this modest coolant q. Table I gives the volatile radionuclide core l activity in the usual case. inventory for a representative 1000 We PhR nuclear power plant operated at full power for 500 days and shows the corresponding anounts of these conponents expected in the fuel plentra and primary coolant. During the nornn1 operation of a PhR, essentially all of the noble gases released { from the fuel concentrate in the primary coolant voltm control system and All are subsequently routed to an appropriate decontamination process. ancillary equipment leaks and secondary coolant releases are comonly left l untreated and are lost to the emironment because under nomal reactor. l operation they represent a relatively small amount of noble gas activity, l well below federal release limits. The boiling water reactor operates on a direct steam generation cycle and the contaminated coolant passes through the reactor turbine. Entrained noble gases and other noncondensables are then renoved from the turbine condenser by a steam jet air ejector. Better than 99 percent of all noble gas released from the core of a BhR non ally escapes along this pathway. Consequently, the only BhR noble gas control equipnent is located on the downstream side of the condenser air ejector. Other environneval emissions are again minor and nostly ignored. Delay tanks and charcoal beds are commonly employed to limit noble gas radioactive releases from various power reactors under nomal operating conditions. This nethod achieves decontamination through natural decay. Retention of the reactor off-gas for 45 days results in the decay of the major fraction of all short-lived isotopes and redtres the reactor radio-active emissions by a factor of near 100. Krypton-85, however, has a 10.8-year half-life and therefore novas through such a system virtually unaffected. Reactor radwaste systems are designed to effectively accomnodate only the arount of activity associated vith nomal core releases, which corresponds to 1 percent failed fuel. On this basis, the cooling circuit vaste gas retention equipment is designed to handle a nominal 1 scfm of contaminated gas in the case of a PhR and 5 to 50 scfm for a BhR. Comercial light water reactors have been in operation for a sufficient period of time to denonstrate the effectiveness of various containment systems designed to minimize releases of radioactive by-pmducts during nomal reactor operation. Overall, the U.S. cersnercial power reactor operating record has been outstanding. Actual releases to the environment have only been a fraction of the permissible levels set by federal reg-ulations. Recently, however, the Three Mile Island reactor incident s -~-_~--% - - ~.. -. _ _. ~

' ' $Y. w,; #WTOD;.~'~ Epm ,.~. 3' y. '. QWijf-q,.. s y Table I f ,F p. IN VOLATILE RADIO.WCLIDE IhVENIDRY[5] A 1000 Mie h'JCLEAR POWER PLANT ~ Reactor Fuel Rod Primary Co ,a

Voids, Coolant, 1

Radionuclide Half-Life 10 curie 106 curie curie r'. 1 Iodines: I-131 8.05d 74.9 0.76 465 b I-132 2.3 h 114.0 0.14 186 h I-133

21. h 171.0 0.64 766 I-134 52.

'm 206.0 0.12 117 f. I-135 6.7 h 158.0 0.34 420 ( i-Tryptons: 'I Kr-85 10.8 y 0.66 0.067 334 Kr-85m 4.4 h 33.5 0.95 439 Kr-87

76. m 64.4 0.076 261 L.

Kr-88 2.8 h 93.0 0.149 775 n i' Xenons: I Xe-133 5.3 d 164.0 4.17 52,290 1 Xe-133m 2.3 d 4.0 0.019 692 Xe-135 9.2 h 43.6 0.084 1,488 'j; Xe-135m 15.6 m 46.4 0.016 42 Y: + 3040 MVt pig operating at full power for 500 days, a. b. I percent failed fuel. o O o /

4glw L4,g pg;p s e -4 ,c . p- . ; F ;G +w f deronstrated the inability of the nomal reactor environmental systems to % satisfactorily handle overload situations precipitated by a significant..,. failed fuel condition. In the TMI case, nearly all of the. core inventory?' 54 V of noble gas was released into the primary and secondary reactor coolant circuits. A large fraction of this burden was subsequently dumped into the reactor containment shell, containment sump pump housing, and turbine y building. - Fortunately, the 1MI reactor had been in service for only 90 L days pricr to the accident and the krypton-85 inventory was less than 10 percent of the level given in Table I, which is based on 500-day operation.. All of the short-lived isotopes, however, were present at near Table values since these radionuclides achieve rapid core equilibrium. The reactor gaseous radwaste system, while effective during routine reactor operation, e9 was not available because this equipment is part of the reactor primary coolant cleanup system and is not generally accessible. As already noted, however, the radwaste equipnent would not have offered much relief anyway because of its low capacity relative to the quantity of gas requiring .n decontamination. Some of the noble gas was inadvertently vented to the environnent while a sizable fraction was isolated within the reactor containment shell. The reactor presently contains 35,000 curies of krypton-85. In 1966, the Chk Ridge Gaseous Diffusion Plant (ORGDP) worked on a mobile processing system designed to be transported to the site of a hypothetical reactor within 24 hours following a fuel failure accident to recover the noble gases and other volatile radionuclides that were released to the containmnt vessel. The process was based on selective absorption using to be located in a series of trailer tmcks 11,1 }recoven equipment was fluorocarbon solvents. As envisioned, all f th The design capacity of the nobile unit was 1000 to 1500 scfm and was based upon processing the entire reactor containnent volum (3 x 106 cubic feet) within one week with an overall krypton decontamina+ ion factor of 100. This work did not proceed beyond the conceptual design stage, however, and the program emphasis was shifted to application of.the process to routine cleanup of reactor off-gas during normal operation and to fuel reprocessing plant off-gas treatrent. Tne fluorocarbon process developmnt program has since been essentially co pleted as part of the fuel recycle program at the Oak Ridge National Laboratory (ORNL). Considerable improvenents and simplifications have been rade to the original 1966 version of the process. Perfomance and reli-ability have been demonstrated on an engineering scale over'nany years of pilot plant operation, including tests with krypton-85, xenon-133, and iodine-131. A third generation pilot plant is currently in operation at Oak Ridge. Mass transfer data and rigorous process nodels have been generated that permit confident design of an energency off gas decontamination process applicable to essentially all types of reactors. In view of the current interest in providing additional reactor safety features, it is believed the nobile Kr-Xe removal system deserves re-examination. This proposal specifically requests funds to study, design, and ultimately constmct an energency nobile or fixed reactor decontamination system. Further process development is not required for this application k

~:~~. ___-_ _ _ _ _ _. ' ;-r:0;'m,q; :*gy ~ ~ - ,g h.qs.pi u_. ~? .w s. ,i... .m..,.,. . ~ x n,:.. and the design studies can be initiated imediately. An integreted system q,. (. is proposed that will allow total off-gas cleanup, depending upon 'the

1.

particular accident, including removal of elemental and organic iodine, trTtiated water, semi-volatile species, and particulates. 'Ihe resultin f~ 'i'

~

system operation and performance will be verified and it will then be made available as directed by DOE. 1'b. 3. T*,, 4- ~ y w,. fy e, eY t' '.3 A ,t., l t d r b e# 4k en / .W 6

1. m

[,'* ~. 16 fp~ e 4 e 7 / 4 9 [' = o e YY W3

• l o

Js.,

===we

- - -,;, u ~ , t q,L. g}g Q_ *. fu ja, hQvyj 6 5% 'e b - 9. a g.. n-9 7^ 1. MCKGROUND

; s The fluorocarbon-based selective abso n tion process for krypton and xenon renoval'has been under development since 1967.

At that time, a test 7., 4 facility was built at the ORGDP to estab: 2pJ. general process feasibility; ' i and to collect noble gas absorption data This initial equipment, ' while somewhat limited in operating capability, functioned exceptionally -; well, and valuable overall perfomance information was attained. In 1970; the testing program was oriented to demonstrate specific application of[24), the process to achieve cleanup of the off-gas from light water reactors b Pilot plant tests denonstrated the basic feasibility of the process and showed that better than 99.9% of the krypton and xenon could be removed from various reactor off-gases, su:.h as air, nitrogen, argon, helium, and hydrogen, and process concentration factors in excess of 1000 could be achieved [7,8]e process was subsequently offered commercially for LWR Th application but was not really required to meet nomal reactor release i limits thct were in effect at that time. Based on the denonstrated operability, flexibility, perfomance, and tolerance for impurities that the fluorocarbon process exhibited for typical t off-gas components, efforts were initiated in' 1971 to adapt the process to krypton decontamination of IFFBR fuel reprocessing plant off-gas. In concert with.the overall IJfFBR fuel recycle program at ORNL, a nore sophisticated facility was proposed for the expanded program. A new pilot plant was designed in 1972 and put into operation at ORGDP late in 1974[26,27 Tnis test equip ent was built to be especially flexible, J with the necessary analytical capabilities for overall, as well as detailed, [. component analysis. Allied-Gulf Nuclear Services conducted a study of the application of advanced s fission gas retention systems to the Separation Facility at the Earnwell Nuclear Fuel Plant. Flow sheets incorporating the fluorocarbon-based krypton absorption process in an overall off-gas decontamination package were subsequently developed and rodifications to the Barnwell plant necessary to imolement the flow sheet options were deteminedll83. Nuclear Fuel Services performed a similar study to identify problems associated with ( providing enhanced fission product containment at the West Valley repro-cessing plantl19J. These two studies confimed the overall utility and ( applicability of the proposed selective absorption process to the decontamination of conmercial thR fuel reprocessing plant off-gas but also pointed out that, in the opinion of the two companies involved, sufficient envircraental justification could not be found to warrant installation of noble gas renoval equipment. In 1975, it became apparent that unresolved waste management problems had become detrimental to the continuing development of comercial-LhR fuel cycle facilities. After an evaluation of available waste management alternatives, an ad hoc national task force concluded that effective and reliable off-gas decontamination equipment os, in fact, also needed for the renoval of krypton-85 from IXR facility off-gas (1}. Since there-were only minor differences between IIIFBR and LWR fluomcarbon process I

~ ' 7 h.hMh:. 1 y* f L' y y 7 } _ f.4{ applications, a joint UFER/lXR fluorocarbon process development progiani. ~ R was fomulated among ORGDP, ORNL, and the Savannah River Laboratory (SRL),, ' t [ to efficiently meet the needs of both reactor fuel cycles. During this, e i same time, it became obvious that the fluorocarbon-based process. being;, ~- - i developed for. krypton-85 removal could also be used for effective, simultaneous removal of carbon-14, as carbon dioxide, and various nitrogen L oxides. -Available data also indicated that the same process is an ~ 9 J effective means for renoving iodine, methyl iodide, and water. Con L. sequently, the scope of the fluorocarbon process development effort was. t broadened to include further definition of this general capability of, the process for application to DEBR and IER reprocessing plantslZ83. i p kl Early in 1977, considerable interest was also expressed in the use of. q alternate fuel cycles, such as the one based on thorium, and various optional reprocessing schemes that would reduce the world-wide opportunity 1$ for plutonium diversion. Consequently, in order to keep the program N responsive to the needs of the expanded fuel cycle effort, the fluomcarbon I process developnent program was stru tured to the newly fomed Advanced s-am at ORNL and the Alternate Fuel Cycle Technologies fi Fuel Recycle Pro Program at SRLL2 ]. Radon-222 was identified as a potential off-gas 4> problem for the thorium fuel cycle, and therefore, work was then scheduled to verify that the fluorocarbon process would 'also be an effective means 'l i for renoving this nuclide. g h-All nuclear fuel reprocessing work funded by DOE is currently being managed ~ The ORGDP by ORNL under the Consolidated Fuel Reprocessing Program. fluorocarbon development program is now an integral part of this overall c 3(. The objective of the ORGDP fuel recycle work is to complete the effort. process development activities and design a denonstration off-gas decon-f. tanination facility applicable to several types of reprocessing plants. 3 In order to accomplish this task, the fluorocarbon program was divided ld E, into 4 major work areas: (1) process development, (2) process application, (3) solvent chemistry, and (4) reliability analysis. Tne process develop-rent work is generating the process technology required to conpletely define the f1torocarbon-based process and all associated peripheral equip-

1. y Process application studies are providing the engineering nodels required for process optimization and denonstration plant design [9,16,31),

ment. p 7his work is identifying flow sheet options and pointing out the relative effects and importance of individual process elements and operating 7 The solvent chemistry effort conditions on the overall system function. g is establishing and confirming component solubilities (where.these' data are not available), phase relationships, component interactions, and corrosion characteristics of the fluorocarbon system. The process reli-m ability studies are assessing the component and system reliability and 1 reconmending necessary flow sheet redundancy and backup support systems to ensure a high pmcess on-line efficiencyl33]. 9 r e h" .s 7= L.e, 3 e a

n: h:n g C,. &p.g x v N ~ c.p w

• ~

is 8 y

.s M&gg y

- o

w..

1.[. '/.[L 4 II. PROCESS MSIE

, W

. n ,t 3-Several methods can be considered for reducing noble gas emissions f i I m various nuclear facilities. In general, these operations can be categorized. into four groups: (1) delay tanks and selective adsorption techniques-using such sorbent raterials as charcoal or rolecular sieves, (2) cryogenic-2 distillation, (3) selective absorption in liquid solvents, or (4) special-techniques such as trapping as clathrates or separation by selective permeation. Suitable chemical processes are not available because krypton, ~ and xenon are essentially' inert. Delay tanks and adsorption processes are .m not effective whenever Kr-85 control is demanded as previously discussed. u Cryogenic distillation is a well developed commercial process for air separation in the absence of radiation effects and feed gas impurities. hhen processing radioactive noble gas, however, ozone formation poses a ~ serious threat to the continuous safe operation of the cryogenic equipment. E Also, xenon plugging has proven to be a comon problem with this type of process when the arount of xenon in the feed gas is large relative to that of krypton, such as is characteristic of the noble gas generated in a reactor. De special techniques are not generally effective for treating dilute gas streams. Additionally, the processes that fall into this classification are not cell developed at this time. i l The basis of the selective absorption process is, of course, solubility ditTerences which exist between the different gas mixture constituents and the solvent. In this regard, fluorocarbon solvents have been identified f, as a valuable and somewhat unique group of solvents unusually suited for separai.ing a nu2er of industrially important components from various gas mixtures [14,15), Ihe solvent preferred for noble gas absorption is i dichlorodifluoronethane,CC12F2, comonly referred to as refrigerant-12. This particular solvent was first selected by Steinberg, primarily because of its capacity, separation factors, and thermal and radiation stability, as well as overall process safety and economic features [21,22j, ne physical properties of refrigerant-12 are well know}iO)The basic thenno-dynamic properties are detailed by htHamess, et al , and in a duPont technical bu11etinl43. A substantial amount of equilibrium data now exists for krypton and xenon in refrigerant-12 solution. The initial work was perforned at the Brookhaven National Laboratory (BNL) by Steinberg. Later work was reported from the University of Tokyo (LTT)]by Yamanoto and j Takeda[343. The nost recent data are given by Shaffer[20 at ORNL. f. bbrrinnn[173 reviewed the available data and utilized several techniques f based on Hildebrand's regular solution theory to estimate krypton and xenon equilibrium coefficients. All available data and hbrriman's regular i solution estimates are presented in Figure 1. All investigators show f good agreement. Toth, et al[32], recently conpleted a laboratory study to define the general behavior of other nuclear fuel reprocessing plant off-gas components such as carbon dioxide, various nitrogen oxides, water, iodine, and methyl iodide. Their work shows that carbon dioxide is even nere soluble in. refrigerant-12 than xenon, and they present detailed solubility data talen over the tenperature range of -40 to +20*C and with solute concentrations-ww. y i ~

.m_ -,, y, y 9 ~ : o .o o w's.= o.nio.ve.as : e tul 200 Kr ~ O _ c, f . O * /g w - 100 5,7 80 e 60 O / O a 0 Xe 0 hn O 40 O O K, 20 atm 8 T.. s LEGEND 4 10 / 8 O BNL DATA O UT DATA O ORNL DATA 6 A SOLUTE VAPOR PRESSURE O - MERRIMAN ^ 4 k b 2 5.3 5.4 5.5 5.6 5.7 5.8 in T (OK) Figure 1 DISTRIBUTION COEFFICIENTS FOR KRYPTON AND XENON IN CC1 F22 .s. A A

..y,g gy.$jg .- p , = ~ - w,;.s;NhM fiQf '}<; 10 . ;,g.m =.,,

.,

:n.a..

~ sn E~ of 3 to 661. Elenental iodine was found to be fairly soluble-in A refrigerant-12, while methyl iodide was totally miscible. Figure 2shows) the iodine solubility data of Toth. Beyond this work, some estimated s 'iiolubility relationships have also' been given{l73 Table II sumarizes the solubilities of the various components of. 'J' .c ,J r interest in CC1 F. 22 . e ,g' L ~- I a p Q 1 .r y n; w e b i s. ,e i l l l h f l 6 3 { a 6 \\ e +t s-e o r *

~ y7p;., e e,p y.,~. rs.v.. wyg.n.,x.,qp.49pg.s,yy3p.,q,a7:.e917>.7,p ,..a > c .:...4 1 ......u......., lui i 10. u" n o 1 LOG Il ] = 2.244 - 1544 (T-) o g 10 2 _N 2 a h 104 w" N O ~ g tt u. w J REF.TOTH AND FULLER, ORNL o E ~ t;- j. 10-5 l l l l l l l l l l l j 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4.0 4.1 4.2 4.3 4.4 4.5 ,y l h x 10, oK~l ,. /, 3 '.4 - 0,; i 1 i . j Figure 2 _n.- e 4 A; IN LIQUID CCL F22 . SOLUBILITY OF I2 j i 9. t I p' p 8 3 N Jub, '-).,+' . qf 'i f,[, ,u ,. =. s s e

r: n[ 12 \\ e ? y gig @ ls y ': m. ;..y;;.;.yc%.g <4; a ', ~ .W:?,.py.M?pyd ] .. ~. M.u...A,n,,D 1 .P ** Table II . :?,b.A)\\;yfQ:yi

n :

. ! -- c, Ce,.19 ~ w;-s@- w Y.', St.MiUtY OF CDMPONENT SOLUBILITIES IN CCI F - ec. 22 c :,,, - Q;: 0,9:, ,:,5. y; ye .,.1 g, v. l.- . ' [$ %'. ? k l [.]. '. _ Solubility 4 rM " Corponent mole fraction x 10 Reference - ?d,.! :'{fy 4 .c ~ ' i 'T/i]{^ 9 2 '15 H 5.5 ji., 2 u g.s. u*- .E N 23 17, 31 2 r W,. - 17, 31 0 40 g~..,. 2 t CO 41 17 I Kr 131 17, 20, 31. 0 290 17 3 Xe 465 17, 20, 31 (; CO 700 17, 32 2 ..t 3 I 0.71 32 2 I>- 1. Based on a solution tenperature of -25*F (-31*C) and conponent c ,E partial pressure of 1 atn except as noted. lw H , 'c

e.,

2. 75*F (24 C) 3. Saturated solution. C .N k I I e i 4 9 Jp rg'.t,,... J rW-5 e 4 w a Ma. M t.G.Sai%ei milhipeumi ime i ep iii.i i i-ig - _3 m, [g

L - l N. f [k ~ ' N M UW ' E 13 t h -f $ Qfkh ~

• y jx 4

7 [- III. PROCESS DESCRIPTION {j 7 a, ff'. Figure 3 is a schematic of the selective absorption process as it was " 4, originally conceived. The process serves to remove volatile radioactive ~, contaminants fmm various waste gas streams and subsequently concentrate - p J-the contaminants for long-term radioactive waste storage. Absorption, fb intemediate stripping, and final stripping steps must be performed in k order to accomplish these, two process objectives. Each separation step t, h, exploits prevailing gas-liquid solubility differences that are established - F between the solvent and the various feed gas constituents. The main f separation of radioactive components from the bulk gas is effected in the - p absorber. The intermediate or fractional stripper serves to remove the ?- coabsorbed carrier gas from the solvent, thereby enriching the remaining k dissolved gas in the mre soluble components. The final stripper removes A all remaining gas from the pIncess solvent for collection and regenerates 7~ the solvent for recycle back to the absorber. The absorption section consists E of only a packed colunn, reboiler, and overhead condenser. In addition, D the intermediate stripper also includes a flash drum. Support equipment 0.. items for the basic process include a feed gas heat exchanger, process gas y compressor, solvent pump,* solvent cooler, storage tanks, and several b refrigeration compressors. If the feed gas contains significant 4 quantities of high boiling components, i.e., those components that have S a vapor pressure less than refrigerant-12, a solvent purification still is L available as an in-line option to prevent these mterials from building up in the recirculating solvent. A solvent recovery system is.necessary to remove solvent vapor from the process off-gas. 1w p" Figure 4 shows a photograph of the second generation ORGDP selective g absorption pilot plant. This particular plant was put into operation in 1974 and operated for approximately four years. Detailed engineering f{ drawings and descriptions of the facility are given elsewhere[28). The plant is designed on the basis of handling a nominal 15 scfm of contaminated 4(il gas at absogtion pressures from 100 to 600 psig and temperatures from ninus 45 to plus 25*F. The nominal process solvent flow rate is 1.5 gpm. i f Approximately 900 to 1000 pounds of solvent is required to charge the g pilot plant to normal operating levels. A In the course of the pilot plant operation and analysis, a soluble gas peak 9 was observed in the fractionator columa owing.to gross internal condensation 7 S of the upflowing stripping vapor [31). Further definition of the internal peaking phenomenon showed that when the internal condensation zone was N@ - raised in the column, the magnitude of the soluble gas concentration peak increased dramatically. It became apparent that if sufficient stripping stages were provided below the condensation zone, the final stripping L section of the three column process could be eliminated with the product i~ e y

• 1hese equipment items are for a typical plant. An emergency unit might

+ i well be different in some respects. For example, an emergency unit would f probably use a direct cooling scheme, e.g., via liquid nitrogen, instead' of a mechanical system. g.;g u 2-

~ f .c- , m ;c;: z y-o I w. o-aE9 o Im-wom 20 mm zmF mo o z-or 4z uWO Om

L rz m -t f

JL oe i A 0< m 1 0 m SOLVENT STILL 2r o i mp 5 l V Jk 2d g m r 3 1 g o o o e o n -n oC m r 4m i FINAL STRIPPER ?g-x -i 1m (j v JL _n n 3 o z< JL 3 mz' V n 5T. r A d y FRACT STRIPPER j 3 z si JL >a V Jk Q r m 1"" Jk e Co o m JL o V lI] 9 <o 4 m y -c m2 m

L z -i 2

ABSORBER

• o k

i V >z .D j( d M ' 7.1 o -i \\ a r.n m Om mo o y o< ar %) <m ya ' N~ m2 8 4 m -i 't O j o m z 4LJ z m -l / y ? m> o3 E ^* DO az m> .a >p O to i H m m 4 m O O P m a s'Q,,, f( '( 7:. y . J.n+c - ' ;., - ~. ,U$,4.',a7lf.5*ll ' w . :3.g.. p. n,,_,_

y,q. p q q..p ;m y g g...}.. t %:.ssyhqsp.. j wu ..g.a.; ..,. s,g. 't*.g=gj/y((?. I '..

• t*'

'. V{fy t'.' ~ ,s,, 1.-' g2.y. - > _ ~,. e .s n. a ,1 .p,. <,s 15 ya. .. !i r4 ' L.f,y;4 %,.. .g ..I,.* p y., g f.

g.,

..'.,y v. Y

~. 4 g(jy

\\

• i'.

r :. g ~T 7.-

i a R.2 2-

~. h '.. 1 gn. f

i. c Mi f.

I y, W-j ~ '.estn I;, __ ^: ^^ M ' ~3 l[ -r

s. s

~ 5-4l \\ t k.- = ~. - ?- gt t.* %. 4 ~ Eg0 s , n. N'*". l E-Ff:miiCL ) s; r .J 4 Qwf%as -.,c.s im Mun_.- - - v u ~ ?:t'G Q. e. ~ ~ %2E: ~ * \\ ,f. W~%%h.Qi }.

• w-

...f ~?fiqi Q -W, q: M.- k,$

• "r' 1

~ l ga '8m%@PT '4' ' yi 5 i" f L;L67 d f ~ '.p [.g,... L.h l v g f V j .e .w~- j $NY? d L [bp. L f$h sj@@[(f '

• f ml r

O n [ %e .f. -y b QQ M.w.s.7 ',,,. l,. a I .y.,, + %m.?w. . M ,,.. :. 3453.' 8 Ei o , a < e tc W t. 'p'" -J $r, Fy . ag@'- f .!? w n_m g-- _ ~ _ =. u. t yf ,. ~..... y - W a,.n:1..: _ g y k t % % 9 t M n e % W m J. j'. ..-i Q. D -,rr 7.O J 3 W '-- ' ^^ saa o I v. f $ 0 ". h. W 8 $ Y Y. W N ) 1 s, j.u-.m n 4

s e

r .N %W;;p::2--ing: - a rikf~ ,g tw'- di a r m + n ,x m Ne m s... Ed '&&; x.-2\\ ? pm w;. me >U Q~ ~% ech. - w.3e ? 5-f.-_.. T \\ ~ ~ W <<==:1F..S.,,.!.!,iS.W+;w.... +%WB' .Y .-a. ne:3 n- -. -.

• .g O

. f(i'.c!. !/[- r+J f /, l

t..

fg p .i i y .g s ,t.. y

.{ p, 2 ^ - 7 --- ----- 7 g ?M 9chk?b,,

• v-

~. '. n.c. 2;i.. f;g%y M 'UD ~ i' T^ - 16 y l ' ', p.'s.qqg. s 7 i

c 3,

,, s m ~ being collected as a side stream. Furthemore, it also ieemed"feasibl$ ~to. c place the intemediate section directly below the absorber and operate the ' t'. entire assembly at a conmon pressure. Subsequently, a single colum was ~ designed,that combined the three functional' steps of absorption, fractional n stripping, and final stripping into a continuous contactor [30). ' Figure'.5' y~' gives a schematic of this piece of equipment. Decontaminated off-gas flows <l, from the top of the combination colum and regenerated solvent from the O bottam, while the fission product gases are collected as a side stream. - gi 1he combination colu:m re42 ires substantially less ecuipment and control instnrnentation than the conventional flow sheet, and because of its Rev - greater simplicity, it offers numerous operational and economic advantages. Iy 'f ~ Because of its potential and design uncertainties, a combination colum was recently built and installed at the ORGDP for evaluation [29]. Figure 6 N is a photograph of the column taken during construction. The coltrm is approximately 24 feet tall and has the same flow capacity and perfomance capability as the three-columa developmnt facility. The absorber and m intemediate sections are again 3 inches in diamter, while the final stripper is 6 inches. High effi_iency, wire mesh column packing is used in all sections of the colum. Pilot plant test results c1carly show that scale-up on the columa area is direct for this particular packing as long i,, l as wil designed gas and liquid distributors are employed [9].

Hence, I

larger flow equipment can be designed with confidence using the same type I of columa material. l [I The combination columa has been undergoing perfomance evaluations for almost 1 year. These tests not only established the overall feasibility of the concept, but showed conclusively that the combination column could perform L. nearly as well as the separate three columns. On the basis of a one-to-one

r

' i. conparison of the two options, the combination columa was recently selected as the preferred version of the process for the reprocessing plant appli-4 cation and, similarly, would probably be recommended for the emergency unit. /* ,a F

s h

_m e._,,,.

~' ,_p;en e p L. g:;r:: f;.:- & y w:. x ows. *e o. st/049 3 53 s ',,.4,' ' ' ! I :- ,j it;. tul d I. - C: v, y DECONTAMINATED U VENT G AS {,j {.. A - ~ n s. .h, . + 'f, i .-~. ~ t 4 l~i; ? CONTAMINATED / FEED GAS Ut [ ~ N/ TC R j-3 n 'M VOLATILE SOLUBLE COMPONENTS Kr, Xe, CO2 .h f. r LEGEND h f 4 FC - FLOW CONTROLLER TC -TEMPERATURE CONTROLLER j i 7 LC -LEVEL CONTROLLER j l V APC -OlFFERENTIAL PRESSURE E CONTROLLER l5 A PC -PRESSURE CONTROLLER l J,' CC - COMPOSITION CONTROLLER t - REFRIGER ATION SUPPLY ([ n] R 1 a N -<] Figure 5 l SCHEMATIC OF THE COMBINATION COLUMN s ' g,,, g g, t .~ s

- wy .g, - +. - ~,. _ _ ~ _n, ,p..y qQ y,y,.g_. _ _,g si-

q sr

4. x,

,4 g..a.?. pr g, %

• *.a,.

.~

g

. g w ./ .......g:q,. ; 18 e,<aso no. ntehe ieei ' Y*

O.p* U rus

,,_.y .g .9, .e - ? '. _ _.W&.x \\f) k5, .r. ,, g..ggg -. ' "- ' + -:, : - ,e ~ ~ @r.. / .~ w.,< s'V w e-- r 5, m h,Y ^ 6 l i" S.W 3 t n 9 6 w,_, ~ ~"- M-=- - M st*- W:.eewe.-w 22R. r:d_y%.., WTF f..y ,y 7.;. w. u - ..g:;my w W.c O, - l x 1 p o a, +

T-g d...m y.

x r ::r,16{ ( . _ - -...Q, j I]n..fW s ?r. C;. ~' WMp_ > -uJW - - _ 3N$3F_ " ' p"<S' E W N .fi@~*Mfii58a@: ,j --m 3 -( s 9 +g+u w &.. --- a= S*:=. 26%W'g ~i&W %m. + . f.hE"$qamn.:,.,>., - '$Y$. ~ '[0 '$$Q~ Rf,E---- 1;

=,

w:h a, ( - l...Rs.;.& .........:.....- w eM i1 2.Q P w-f s /, w%+, ;, _- F,wg:d - w h - .E1LT l ii r rp 4 3, eu N M,..;;,.. g 5.if Ti --- u. _; m. MJ'J,3. i=*=.-n%,',,GME {J, U,4 - w i.y

w-ss

~ i l u a. w 2 H 2 h,.:

• 4Q d

typ.-- ,,egTr.- - 9.4;.., r a o.:. [ - w-. -6 ^ i p;;- +++ e 1

t gh3_.E k-.e-i..L; k &@-%

i W W ra rf' gy. f. A %. W Q ~ '*-* tk Lid. W '2 ,: ~J.*:b fm'r

-. ~~

--..,::LT, A sy i J i s. F- --p.,f, : m% ~ ig;p _. -!c qufr:. +- il p u i i .t F.4 e .w .. $ *. : q- - e 2

s...s. sc: 3 4

s. 1.. ^-- a it* < ~g 21 @ f .,e : ' m ' 9. n%5' Ken. Q a.-g_ s e - : 3 -- -' ' ' -4 y,g?9 '. -h.. u sM.f. - .a. '.-.,sy y - -r '. Q. ~

u. Q. J' 1g,.5)~~4.:.

,a y 'c J n JWT 4 1;r * + - ~ N s 2g: s A.- 3..; '.. l N L A . ~. h.; h~ g .g. r m - r

e. +.. W b

i* in j f.lfhW.S [.' ' ? .l t.fs, _ 5 C.~.s 2 :: 3 3 y-d e* m 2 1 t.-. %p>[ Y ~' ' ~~~*_- R -$'h% ? Q b. j'QW y. w m,. +2 Q. l ,I ) ..+-%. 4 ar d _f. s -,c.. n. %..:2 _A -c. -w. rt;' e-=. ~- { q.&p ~:. ~ ~~.,, l *. ; - - t k;MM~ 7 4 p D g' .s' W._ - .3. $ h. e-p @~ M i -Tm Q. % 5hhkbh hb'$y ' q ~' ~%s. a3. ,x w

w. Wu..

,Et".w.g: :.~f. x,,., . : : rP...,u.x.; a .b:M ;.1 .n,. .,ij :'c.u o, ^ 62 .., !:.- 3 ,g ;%:... a fu q

3. --.es. -

s.p ,4 1 u. g p=,' +-;d..1 .3r-q [ --.x. r n . :.y -y 9 c#1*.g *5'"i*-S ~ t.ca r.g gu +c ,, p d e.- -m e-<. *W.j 9 s.. -q q*:N * ' _&lJ,.:3 ,Q st. Q N. [.%'?: p ;~^,., n. %-w.c;tzendpm-?"JR*T. $>,a y %G, ' J t q:,n .1 -

+.-.-,e ; r x

f n.-;:_:.; -" - a.-n. WW, s., --

- t-_ % ;E7J L g.:c & - E -Q :): c-.j- % S f a hu2 ;&r-:Sy44&

n 1.. ~ ,s 4 f h W* h;., g$ k W h Y ~ Y $ m$ $ h'!-n4,C"i. P C - :ic.G e. ? 4 e -n-drew Wi$ .A'$ / t:4;.jsiriR% w g-m&.rI: r.:.- . O.7.B::ls-.*f".o s ~ D".a G- 'I c' pi%M:.f.%m:,mt 4,Srw ?

M;r* *

b.s.W ;-d. w g

~;..a mo +~ ~~' c.m%:.z',.:>. .Fy: mcM's.4*LWk-??TSfW@*- W;m r:T%%::;~.w.SW>:r,..mm'1- -v -.~ms.- -r . '. ~ ~. . s. ,z M. e -'w ". %M5

6. -a : r

• m r. s J: w#.,2,,.m ~ --

q c

s y :- w.m+r%.w. +.:i;w.. 'c, c~w.2r~,w-M..ap- :.. -. : g- ;t,?.,.,,. * -g - - cJ

.,r. -~m m

gw%{dS M M :- g W,L; w y_$:e.g% m.: M hg g,. --~ r, g-. -mE'.'urg ",2

s - : '.

-m.

p. ~.

4;;%hgTch::T54w.g e# ,~- -.L i l Figure 6 OVERALL VIEW OF THE COMBINATION COLUMN DURING CONSTRUCTION ,,.m. I f

.,e 1 _ y "PQSf@qpjQt QQ 19 - 9:g6 ? F 3 g,ii '~. IV. PROCESS DESIGN CRITERIA ./2.>,, .r .r-The decontamination process throughput, fiss' ion product recovery, and k radioactive product purity have to be specified before equipment design can be initiated. Usually, an economic balance would be established i between the cost of removing the noble gas from the reactor following an ? , accident and the cost of the reactor downtime. In this case, however, ~ 3 there may be overriding safety and social ramifications which dictate that the decontamination process proceed posthaste. The process deconta-

g -

mination factor is contingent upon the particular processing scheme but rust at least be high enough to meet applicable federal release regulations.. To illustrate some of the trade-offs, if the decontamination process off-g. gas is recycled back to the reactor containment, lower single pass decon-e7 tamination factors are pemissible. Also, if the concentrated radioactive it product is relatively pure, fewer gas cylinders will be generated for long-term radwaste storage. I Figure 7 presents decay curves for krypton-85 and xenon-133 in the contain-ment shell of a 1000 We reactor following the total release of the reactor 3 core inventory after full power operation for.500 days. The total activity 7 in the containment shell will be.near 2 x 108 curies after 1 day of cooling. Most of this radiation will be due to xenon-133. After 42 days, krypton- ? 85 will become the dominant source at 6.6 x 105 curies. The arount of krypton-85, unlike the generation patterns of the short-lived isotopes, c increases substantia 11 After 1100 days of operation, for instance,1.1 x 10g with fuel burnup. curies will be present[2]. This radiation source { e will remain at virtually the same level for any years thereafter if the sr containment volum is not purged. s= In the absence of a suitable decontamination ' system, the containment gases would have to be held for about two months to give the short-lived isotopes sufficibnt time to decay. The krypton might then be slowly vented if both s atrospheric conditions and public relations were favorable. At a discharge i rate of 103 curies per day, over 600 days would be required in this case to vent the reactor inventory. The krypton-85 release, however, would be in violation of EPA standard 40 CFR Part 190 which states that the total quantity E' of radioactive materials entering the general environment from the entire fuel cycle mst contain less than 50,000 curies of krypton-85 per gigawatt-year of e electrical energy [6]. Thus, slow venting does not seem to be an acceptable altemative. Use of an effective noble gas removal process in conjunction with the decay phenomenon will allow rapid renoval of the containment activity following F an accident. A central sizing issue is the combination of flow rate and single-pass decontamination factor required to reduce the curie loading Figures 8 and 9 within the reactor to a specified level in a given time. show krypton-85 and xenon-133 activity levels as a function of elapsed time These after release to the reactor containment and various processing rates. calculations assume that the radioactivity will be removed simitaneously,by radioactive decay and a noble gas recovery facility. The decay mechanism,

5 e

. m am g' *_. ' a'. '"'*I.*'g 4' W'r - ( o ^l* ,2 .- n y h r .,m- +. :. c...r.e

A.,,.

s 1 . g ~.,%'.~' .. W:q..n' 20 ~

y;;

p - f-- ..y :e; e t- ' o...

o. mir n..' e " -"-

e sus .4 z *.. to i a n ti 4 + t',o ~>

r y;

n.,

8 10 4n, 3, b Xe 133 7 10 J. t B ,4.. .-o r L 6 10 Kr-85 o ~. i, 5 ~ 10 10 I I I I l' i 8 O 10 20 30 40 50 60 70 TIME AFTER ACCIDENT. DAYS Figure 7 .y ' 7'r DECAY OF NOBLE GASES FOLLOWING REACTOR ACCIDENT e

ymy y y (, u -. -l- }.' fM ' ' * ::.N- .y. , + t owg, no. nie-se-s see ',

IuJ, e.

a a a a s 8 10 ,e NOTES: 6

1. CONTAINMENT VOLUME 3 a 10 CUBIC FEET

^

2. 90% SINGLE PASS REMOVAL EFFICIENCY 1 TOTAL RECYCLE

,o'r .r: u 5 5 106 m DECAY ONLY ~ .-z w s I z L E \\ O U Z Eo s i u E 1DO SCFM 4 10 250 500 f 750 1250 g 2000k Ii i I 3 10 O 10 20 30 40 50 60 70 PROCESSING TIME. DAYS Figure 8 REMOVAL OF Kr-85 FROM REACTOR CONTAINMENT AS A FUNCTION OF PROCESSING TIME AND RATE

~ yl{':9 - 4' ~ f{.,T '5"V,'?{}; if ^ w .,...v - 22 02/. -4'.; ? a " l' t , ; *.. ?.N. ':, W,,, :,*s l '.. e* ows. No. asse-te.s aeo tal ,,,.% s s 10' a a a a e .[. NOTES: 8

1. CONT AINMENT VOLUME 3 a 10 cubic FEET

2. 99% SINGLE PASS REMOVAL EFFICIENCY

3. TOTAL RECYCLE

~ 8 10 2 3 t a E107 = 9=z w 1 E s-z Ou E t> 6 f 10 u< A A DEC AY ONLY 5 ~ 0 100 SCFM l l 2000 250 t 1250 500 750 10 il 1 1 I I I 4 0 10 20 30 40 50 60 70 PROCESSING TIME. DAYS Figure 9 ~ REMOVAL OF Xe-133 FROM REACTOR CONTAINMENT AS A FUNCTION OF PROCESSING TIME AND RATE 1 l l"---

. n g q gjn r ' ~ ' 23 /-.'. 4<. * - % p. He of course, will have little effect on the renoval of krypton-85. -~ recovery facility, in this case, would recycle the processed gas back to the containrent vessel following 90 percent recovery of the krypton and 99 percent of the xenon. At a processing rate of 250 ft / min, about 39 days 3 would be' required to get both the krypton-85 and xenon-133 containment activities down to 104 curies. Increasing the processing rate to 500 ft / min reduces this time to 19 days for krypton and 26 days for xenon, 3 3 In-while 1250 ft / min requires only 8 days and 13 days, respectively. creasing the single-pass removal efficiency of the recovery process will also reduce the processing time and/or will allow the use of a smaller thruughput system. He fluorocarbon process can typically remove 99 percent of the feed gas krypton and 99.9 percent of the xenon. I A preliminary schematic of the proposed processing system is shown in Figure 10. Le contaminated gas, withdrawn from the reactor containment, turbine, and/or auxiliary buildings, will first be filtered to remove possible particulates, compressed to approximately 125 psig, and then cooled to near -30'F. The bulk of the reactor airborne tritiated water and iodine will freeze on the heat exchanger surfaces and thereby become. imobili:ed. Trace amounts of water remaining in the feed gas will be ad-sorbed on 3A molecular sieves. Next, the reactor gas will be fed into the absorption section of the combination column and contacted countercurrently. The decontaminated gas 1 caving the top of the with downflowing solvent. column, containing 5 to 10 percent refrigerant-12, will be passed through Alter-a turboexpander and 13X molecular sieve bed for solvent recovery. natively, a low tenperature condenser mght be used to effect the same separation. The process off-gas can be recycled or vented at this point, depending upon the noble gas concentration. Initially, all gas will be

recycled, ne solvent containing the dissolved gases will subsequently flow into the internediate and final stripper sections of the column.

If the Regenerated solvent will be pumped back to the top of the column. solvent contains trace quantities of water and iodine, a 4A molecular sieve and/or silver impregnated zeolite will be used to further decontami-nate the solvent prior to recycle. At a total pressure of 125 psig, the reboiler will operate at around 104*F. A solvent heat exchanger will be used to cool this stream down to -30*F. Exact equipment arrangements vould, of course, be determined as a part of this project. The arount of noble gas present in the reactor will depend upon the fuel e>posure and the reactor specific mrating power level. Figure 11 presents calculated PWR krypton and xenon 1 eration rates for a power level of 30.0 W/MIU[2]. Most of the krypt:n and xenon will be stable and isotopic decay will have little effect on the bulk concentrations. After 500 days of operation,180 standard cubic feet' (4900 liters) of krypton and 1400 If all of standard cubic feet (40,000 liters) of xepon will be present. the core gases are released into a 3 x 10 cubic foot reactor containment shell, the resulting volume will contain 60 ppm krypton and 470 ppm xenon. The noble gas product will be pulled as a combination column side stream. Solvent removal will be achieved by adsorption on 13X molecular sieves. The product flow rate will be less than 0.1 percent of the process feed Provisions will have to be made for. ong-term storage of about l rate.

owo me nio.:e.easi tui \\ 13X MOLECULAR SIEVE i g TURBOEXPANDER A e COMDINATION PRODUCT COMPRESSOR COLUMN ) [' 85 .r T l 13X MOLECULAR m { d SIEVE \\ N \\ 3A REACTOR Kr.Xe PRODUCT. BUILDING GAS RECYCLE MOLECULAR STOR AG E SIEVE l l HEPA FILTER 7 GAS GAS GAS COOLER COOLER COMPRESSOR 7 SOLVENT h ) RECYCLE 4A MOLECULAR SIEVE SOLVENT SOLVENT ', ~ COOLER PUMP ~ ? Figure 10 ily. SCHEMATIC OF THE FLUOROCARBON REACTOR DECONTAMINATION PROCESS ..: - e ..,;gf.g'

~ - - . ~. .V- + = w---- i . f,, '. a q,. s, ;;.,(hy _.;*-.,. ;. a;... y t- + ,,, %, y. ; > o.1 ~;--. (j

• own, eso. st/s to sets tut l

) 10,000 g g g g g l B.000 6.000 ~ 4S# m il. - s.. ~ 2,000

1..

.-5 4 as. 1,000 800 6 600 h VOLUME OF ~ l G AS. SCF Kr ^f 200 2 ~ 1-V NOTES: 100

1. PWR WITH 3.3% U 235 80

2. FUEL LOAD 100 MTU 1 SPECIFIC POWER 30.0 MW/MTU so 40

~7 20 10 I I I I f f 0 200 400 600 800 1000 1200 1400 R E ACTOR OPE RATION, DAYS Figure 11 OUANTITY OF NOBLE GAS GENERATED IN A PWR AS A FUNCTION OF REACTOR OPERATING TIME I p.. 'I e g W m

2: 1600 ft of radioactive prcduct gas. A generally appliedble and acceptable. 3 method for management of the gaseous waste is encapsulation in high-pressure cylinders and ~ storage in a suitable repository [1,33. In this case, the noble gas will be held in standard 1A high-pressure gas cylinders, 9 inches in diareter by 52 inches high. This type of cylinder has an internal voltrae of 1.54 cubic feet (43.S liters) and is normally filled to a pressure of around 2000 psig. Xenon is fairly coxpressible at 150*F, with a com-pressibility factor (z = W/nRT) reaching a minirum of 0.50 at 2000 psig., Figure 12 gives the ntrber of.ylinders required to contain the noble gas generated by a PhR after full power operation for 100, 500, and 1000 days as a function of cylinder pressure. For exarple, 5 cylinders will be re-quired if the reactor has been in operation for 500 days prior to the core release and the gas is stored at a pressure of 2000 psig and temperature of 150 F. The amount of radioactivity centained in the cylinders will decrease rapidly as the xenon and short-lived krypton isotopes decay, until the LTypton-85 source is left. 7~nen, a total of 6.6 x 103 curies will be present in the exarple case. Nearly 1 inch of lead would be required to reduce the external Kr-85 dose rate to 10 n: rem / year at 1 meter from the surface. Correspondingly, the krypton-85 activity will result in a heat. load of near 700 Btu /hr/ cylinder of the noble gas mixture. Biological shielding and heat dissipation will necessi';ge the use of water-filled storage and shipping casks. Bloreke and Perona AJ present the details of such a cash in Figure 13. The cask is 5 feet in diancter and is fabricated from 1-inch-thick, type 304 stainless steel. 'Ihe water provides shielding and serves as a heat transfer raedium. External fins enhance heat dissipation. The design meets applicable inpact, puncture, and fire resis-tance specificatiom; for shipping capsules for curium oxides. A 200 psig rupture disc is provided as a safety raeasure in addition to 16 fusible plugs, In addition, a which are designed to allow steam to escape in case of fire. vapor s' pace is provided sufficiently large to hold the contents of a leaky cylinder without exceeding the cask pressure limits. Each cask can accom-nadate about 5 cylinders of krypton-xenon mixture. A loaded asserbly would weigh about 7 tons. l .) i

1

u a:. m.-

.~;r:0 .3.y.y.f ' 27 ' t '.. 9.?..- ji g, T-IV o.a. no. m/s.n44esi I ~ '

' (

luI I 1.000 g g .g l I I ~ 800 - y V 600'- 400 .t t NOTES:

1. ST ANDARD 1A CAS CYLINDERS 3

~-

2. CYLINDER VOLUME 1.54 FT

/ 200 1

3. CYLINDER TEMPERATURE 15CFF

((

4. PWR GENER ATION SCHEDULE

\\

5. TOTAL CAPTURE OF CORE GAS

'f ~ 100 ~ 80 ~ ( 60 r NUMBER OF GAS CYLIND E RS s I-20 ~ a i * ,YY ~ 10 ~ .g ~ 6 500 4 2 100 DAY OPERATION 1 l l I I I O 500 1000 1500 2000 2500 3000 CYLINDER PRESSURE, psia Figure 12 NUMBER OF GAS CYLINDERS REQUIRED TO STORE PWR NOBLE GAS AS A FUNCTION OF REACTOR OPERATING TIME AND CYLINDER STORAGE PRESSURE s.

,) l

e..s.

_s

• D f

l O w e.No.Mfe 79 teef tui I PUNCTURE SHIELD LlDGASKET A r--m LIFTING EYE O -ps s s x .s s s s s s A TIE-DOWN EYE o g g g-f"us s s s s s s _s [N -1 (__ k{___ ]..

== -? FUSIBLE ALLOY VENTS

=

g WATE ri s eg - eg - rg, LEVEL \\ ~ g ,I /N / N COOLING FINS s ~ t. N .; t. .i, j \\ s .7 , 9, ' ', Nll l g,. PRESSURIZED GAS CYLINDERS N ~3 'l3 D \\ s , Oh ,s - RUPTURE DISC

• i lj

',).i i

N-f,h i% b 6 FT i.q y, c;. s .') h f!h) STAINLESS STEEL SHER sa s W y a s g t i,5 pf, s BOTTOM FINS \\ ii Q.

4 s

\\ '[,' g ~ e hv 1 CASK FEET N r - - h: \\ --1 s -/ & M\\ s s 'a / /// \\\\\\ \\ ,-,s) t s s s s s s s s s s < s s s s s s r _J d

. n

\\ TAKEN FROM ORNL TM-2677 1 5FT c r _I.f-eQ Wq [ - 47,.-'.Y O Figure 13 CONCEPTUAL DESIGN OF SHIPPING CASK FOR CYLINDERS OF COMPRESSED <t I  ?*f.Mi f FISSION-PRODUCT GASES

,4

- :m - ;, s.) 4,g a~

c a. 29 V. h'O?I PROPOSAL A fluorocarbon-based process has been developed that is capable of effi-ciently reporing knpton and xenon from various nuclear off-gases. The process can also remve carbon-14, as carbon dioxide, and iodine. Several process applications have been made during the past several years. Exten-sive pilot plant testing has been performed and rigorous design models have been developed. Sufficient process performance and operational infomation now exists to allow detailed design of an emergency reactor decontamination process with high confidence of reliable, efricient operation. The program outlined here encompasses four main phases: conceptual studies, detailed design, system construction, and system validation testing. 7here are logical "go-no-go" opportunities after the first two. The conceptual study phase is considered particularly important. In addition to the normal conceptual design work related to the kI,fpton-xenon unit itself and the associated trade-off studies (size, decontamination factor, location, etc.), this phase will provide an opportinity to place the post-accident kapton-xenon pmblem in clearer perspective and secure a better understanding of the desired solution. As regards the approach to this project, it is anticipated that: 1. UCC-ND will provide: Criteria and technical requirements for an energency reactor decon-a. tamination system. b. Technical management for the design, procurement, and erection of the prototype system, Operating contractor procurement services, quality assurance, and c. quality control services during system fabrication and erection. d. Required environmental and safety assessments and analyses. Preliminary testing and certification of prototype system c. perfomance. f. Definition of the operating philosophy and applicable operating procedures and guidelines for ultimate energency operations. g. Requirements definition for the continued perfomance testing, n.aintenance, staffing, and operator training that will be needed in the actual deployment of the system. h. Preparation of the actual system for its ultimate deployment.

i. Systems engineering analysis.

1 . :ll.,. ; ' ~ 7 :WT ;< y gpyp j .c. w .. y.Q x .a.J;p @ )),.4%: = g "+.w t .. s _.$; 2

• 1 2.

An architett-engineer will: - g

3..

s a. Produce a conceptual design and preliminazy cost estimate in. [, f cooperation with UCC-ND. ~ ~ ' b. Assume the major evaluation role in aissessing the required inter- ~ ~ ( faces with the reactor types that might be serviced by the system.' 7 Prepare the Title II design drawings and design documentaticm. c. d. In cooperation with UCC-ND, prepare a procurement package for the conmercial fabrication of the system. In cooperation with UT-ND, prepare a cost estimate for the system. e. [ f. Assist UCC-ND in the preparation of required quality assurance assessment and planning. 3. 7he Oak Ridge CPAF construction contractor will provide any required erection services and procurerent related to those seIvicen, as l appropriate. 4. A vendor, selected by carpetitive bidding, will fabricate the system s. in transportable modules. During the conceptual design phase, two alternative concepts will be y evaluated; one will be based on the mobile or " fire wagon" approach, whN the other will be a stationaIy unit that could become an integral part n, each reactor's emergency equipment. One of these concepts will then be selected for detailed design. The single-pass reroval capability of the plant assumed in the baseline design will be 90% of the kryp+wn and 991 j of the xenon. Corresponding iodine and tritium renovals will be in excess f :. of 99.99i.. A prototype will then be fabricated and installed at the ORGDP. Tne design and construction phases will proceed under the scrutiny of an j approved quality assurance prograr. and will conform to the codes and standards specified by the customer. Overall system reliability and availability will be emphasized. Once the prototype has been fabricated, l operability and perfomance of the unit will be thoroughly established. Appropriate reactor interfaces will be determined. OptiImmn or preferred operating conditions and forr:a1 procedures will be provided. 7he require-rents for maintenance, operator staffing, and necessary staff training will be specified. i /' S h L _ - - - - _ - ~ _

o ~ w-eq3 ,.. 5,_. y ~- ,..s ..g, 31 VI. SOiEDULE AND COST F Tnis program will produce an emrgency reactor decontamination system, ready for deployment. This objective will be accoglished in four phases. The first phase will define the system criteria and requirements, pmvide the preliminary safety and environmntal assessnents and analyses as required, and pmduce a conceptual design and cost (stimate. H e second phase will produce the definitive de. sign doctrnents, the procurement package (s), and, the. final safety analysis (FSAR). The third phase will include the vendor fabrication and subsequent erection of the system at a test site at the ORGDP. The fourth phase will include operational testing, perfonrance certification, a A the necessary actions to make the system ready for deployment. The schedule, as depicted in Figure 14, is quite tight in spite of a 4-1/2-year space. So e cogression might be possible by overlapping phases se that pmcurement could proceed during the latter phases of design; hoeever, this would cogromise the concept of placing a clearly defined contract with a fixed-price vendor. Such an action might reduce the total tire to 4 years. Other reductions might be made if certain standard requiremnts are waived. Tne costs for the system are based on extrapolations of earlier estimates produced by UCC-ND and others[18,19]. They are believed to be reasonably conservative; however, they could be altered dramatically by unanticipated custoner or regulatory requiremnts. It should be noted that the schedule does not include gaps or intervals for review and approvals. It is anticipated that these actions will occur on a continuing, concurrent basis during the course of this important project. Tne estirates anticipate that (1) the project will be perforned under the norral DOE cost accotnting niles (not full cost recovery); (2) the pre-lininary environrental assess:ent will validate the assugtion that an environmntal irpact statemnt is not required; (3) the program does not include the cost of the facilities required for the long-tenn support of the system (a reference design and description will be provided); (4) a generic assessmnt of reactor interfaces will be adequate (rather than a detailed reactor-by-reactor design definition); (5) the system will not be required to beco e a part of the reactor containnent; (6) no program-supported facilities will be provided for the disposal of radioactive raterials resulting from the erergency operation of the system; and (7) close, r atually supportive cooperation, of the type nonrally maintained in the DDE system, can be carried out between the customer, the operating con-tractor, the architect-engineer, the system fabricator, and the system As str-arized in Table III, the definition phase is estirated erector. to cost $500,000; the total UCC-ND and A-E engineering package $2 million; the fabrication and erection phase 55 million; the verification phase 5500,000; and total cost $8 nillion (all in FY 1979, second quarter dollars). l Tne total cost, escalated to the year of expenditure, will increase to nearly $11 million.

ui [f li

l

?S gE f nlA,'yQ:;z:p9'"., lJ tr o Y s:n d I' sY'.. 4 8 f l I 9 1 i l! I l g ll l g ll! 1 g! I g g l l 1 1 l W o 3 8 9 lg l l l !l g l l l l l ! i l j l l I g l gl I 1 ill -Y F u' 2 8 9 ll I I l l l ! l g l l g j l! lI g l gl f 1 i l l i W U p g U 1 8 1 i ll Ig ll l g j l I 1 g l !l g l l ! l I l l g l i 9 E L W U U 4 D 1 E I e O r S u U g K 0 8 i S 1 i l l I l g ,j l lg g l l I l l I l g l l ! I I g l g g t 9 F AT B U W U q S t s n t W n P e y e o n ) l s e n p n n a e t n n a n i a n l D n e n n g ei dP mA t i y a a t e a F l t k i m o n r r l m a s c se t ae o v / i a i d n y a ( t c u t t n A l P s a si d s y t nt n q n s a a ' es d oi a e== E g n n i n t n A o l ? ti y P =n t n i i t g C cf b o s e r e t i n r o m e c c li r ue d rD n K a i C n n e a e i n o t( a i v a g n f r c1 d SA r n d a i i r E

• a e n

s t T e E n n n g e - rt n e ns S t a n a n t t n o V g I c t n i r i d A D E n n ri i m r n n I e oa t d n i gt o e t ns f m f M c i m' iS C a e g A 1 mr e e D m e i i n t Q 1 g y ht n m y o n e o r nr r an e o e t i n a l e t t u go E e d o t e t o D R a l s c c R i f r r l rE t A n t y s o A s d P a n R p l e s it s f p i i i S a r S e) n w o S e a pt y i S S c a F r T P P r td e e h A i O D n On, y a o t l n r i e e u e en s t n o e f a a t l t n g a m t e n e ene n i d a o a d i em e = r r a e p r n r r e r r r ri n o l n c i r n t ry r a a u S e a i a o l a a a af i i b a i t a a a ao a t p p l c p r p f p p p p ped t a r c p r pl p t e e a E n a a e r m e e e ed a a t d b e e g e ep e s e r e o r r r r ne c n i a r r n p re r e P P F. A C P n P P C P P P PUP i E D F E P i O PD P c r r v - o r t 1 e j b . R C D E D A 8 C 3 MA a yA B C D E F C F A .. C D E F. T n 1 V 1 1 1 I 1 1 I 'I l l

1 / g l y p ;1 m g y. v g e g e n,gg.7,,q p 4 9 c 9,q ;'e r M.g: +.,;uf. rqqp,,;.;g;4q

g. 19 g g u 4

4 4,, c

• l ' 4

o a

• }M

.J .1 y 1 4 Tabic III i AKflCIPATED PROGIW1 COSTS BY ACTIVITY AND YEAR ($1000) FY-80 FY-81 IT-82 FY-83 IY-84 'IUrAl. 500 I. Project Definition (Criteria, Conceputal 500 Design) II. Engineering Design, Documentation, Titic II 300 1000 400 250 50 2000 a., 5000 0 2000 3000 III. Fabrication and Erection 250 250 500 IV. Testing and Deploymen. Preparation Total $, FY-79 (2nd Quarter) 800 '1000 2400 3500 300 8000 . j Escalated to year of expenditure 900 1200 3200 5000 500 10800 e e am* y be. p

• s

e.,, -.

. q 7 s,, ',.."q g. I + 5; G. n e,.g .,'.. _,, a, w (, y. .~ 8 .% 1 ,S"- )'l ' ~ ; w &R s? . D'. ' /,,Cj,CQ',11, p 's', ,, e' : <..y. n p, 4g; c. -,g,

~~> 3-r, .,_'_.J. [: m n W g :-

, m; w; 34

' C$. ^ ,o n .~ REFEREZES 4 ~ 1. Alternatives for Mm=ging W stes for Reactors and Post-Fission Cperations in 19 LVR Fuel CydZe, USDOE Report ERDA-76-43, Springfield, Virginia. 2. Bell, M. J., ORIGER - Tne CRRL Isotope Generation and Depletion Code, h.._ USDOE Report ORNL-4628, Oak Ridge, Tennessee (1973). f. 3. Blomeke, J. O., and Pe%na, J. J., M1nagement of noble Gas Fission f_, P Product Wastes from Reprocessing Spent Puel, USDOE Report ORNUD4-2677, Oak Ridge, Tennessee (1969). 4. E. I. duPont de Nenours and Company, Inc., Tnermdynamic Properties of Freon-12 Refrigerant, Technical Bulletin T-12, Wilmington, Delaware s (1956). 5. Environ ;cntal An: lysis of the Uranium Fuel Cycle, Przrt II kcLear Pouer Reactors, USEPA Report EPA-520/9-73-003-C, Washington, D. C. l (1974). 6. Environmental Protection Agency (40 CFR Part 190), Federal Register, i l1 Vol. 40, No.104, 23420 (May 29,1975). L 7. Griffith, G., "99% Cleanup of Nuclear Gaseous Wastes," Pouer Engineering [ R:rch, 62 (1973). 8. Ibgg, R. M., "New Radwaste Retention System," AcIecr Engineering International February, 98 (1972). D. Kanak, B. E., Analysis of a Gas Absorption Column with Soluble Carrier x_ Cas and FoZatile ScZvent, USD3E Report K-2007, Oak Ridge, Tennessee (1979).

10. htHarness, R. C., Eiseman, B. J., and Martin, J.

J., " Freon-12," { Refrigeration Engineering, 32 (September 1955). l

11. hbtriman, J. R., A Mobile Processing Unit for Frypton-Xenon Absomtion, USDOE Report K-L-2397, Oak Ridge, Tennessee (1967).

12. hbrriman, J. R., Pashley, J. H., and Smiley, S. H., Engineering Develop-ment of an Absorption Process for the Concentration and Collection of Irypton and Xenon, USD3E Report K-1725, Oak Ridge, Tennessee (1967).

Ebrriman, J. R., Stephenson, M. J., Dunthom, D. I., and Pashkey, J. H., 13. BS Removal of Kr from Reprocessing Plant Off-Gas by Selective Absorption, USDOE Report K-L-6201, Oak Ridge, Tennessee (1972).

14. Fbrriman, J. R., Pashley, J. H., Stephenson, M. J., and Dunthorn, D. I.,

Process for the Separation of Components from Gas M*:tures, U. S. Patent 3,762,133 (1973).

m, 7 m 4.e 35 hkrrinan, J. R., Pashley, J. H., Stephenson, M. J., and Dunthorn, D. I., 15. Removal of Purified Heliwn or Eydrogen from Gas Mi=tures, U. S. Patent, 3,785,120 (1974). bbryinan, J. R., Analysis of a Mtitico :ponent Gas Absorption System 16. t.ith Carrier Gas combsorption, USDOE Report KY-G-300, Paducah, Kentucky (1975). A Critical Review, krriran, J. R., 7ne Solubility of Gases in CCZ Fg: USDOE Report KY-G-400, PaAnh, Kentucky (1977)g 17. // Murbach, E. W., et al, Fission Product Gas Retention Process and Equip-e 18. ment Design Study, USIOE Report ORhl-TM-4560, Oak Ridge, Tennessee (1974). North, E. D., and Booth, R. L., Fission Product Gas Retention Study 19. FincI Report, USD3E Report ORhl-TM-4409, Oak Ridge, Tennessee (1973). Shaffer' J. H., Shockley. W. E., and Greene, C. E., Tne Solubility of 20. Frgpton and. tenon at Infinite Miution in Mchlorodifluoromethane, USDOE Report ORN1-TM-6652, Oak Ridge, Tennessee (1978). Steinberg, M., and Manowitz, B., "Recoveiy of Fission Product Noble 21. Gases," Ind. Ingr. Chem., 51, 1 (1959). Steinberg, N., ne Recovery of Fission Product Xe and Er by Absorption 22. Processes, USIDE Report BNL-542, Brookhaven National Laboratory, Long Island, Ns York (1959). . Stephenson, M. J., Wrrimn, J. R., and Dunthorn, D. I., E:perimental 23. Investigation of the Remval of Kr and Xe from Contaminated G:s Streams Phase I Completion by Selective Abscrption in Fluorocarbon Solvents: Report, USDJE Report K-1780, Oak Ridge, Tennessee (1970). Stephenson, M. J., Merriman, J. R., and Dunthom, D. I., Application 24. of the Selective Absc:ption Process to the Removal of Kr and Xe from Reactor off-C s, USIOS Report K-L-6288, Oak Ridge, Tennessee (1972). Stephenson, M. J., Wrriran, J. R., Dunthorn, D. I., and Pashley, J. H., 25. E:perimental Dc-cnstration of the Selective Absorption Process for Tr-Ie Rempal, USDOE Report K-L-6294, Oak Ridge, Tennessee (1972). Stephenson, M. J., Danthom, D. I., Reed, W. D., and Pashley, J. H., 26. Absorption Process for Removing Frgpton from the Off-Gcs of an IMBR Puel Reprocessir.g Plant, USDDE Report K-L-6338, Oak Ridge, Tennessee (1974). Stephenson, M. J., and Eby, R. S., Development of the FASTER Process for 27. Re oving Frgpton-SS, Carbon-14, and Other Conte:n*nante from the off-Gas FucI Repmcessing Picr.cs, USD]E Report K-GD-1398, Oak Ridge, Tenness,ee (1976). l

) .. i. cr, y m. g ,f. = Y, Stephenson, M. J., Eby, R. S., and Ibffstetler, V. C.,~ ORCDP ScIsotive 28. Absorption Pilot Plant for Decontar~ nation of Puet Reprocessing Pignt ~ Off-Cas, USIDE Report K-1876, Oak Ridge, Tennessee (1977). Stephenson, M. J., Eby, R. S., and Kand., B. E., Reprocessing PZant 29. Off-Gas Decor.tc:iration by Selective Absorption, USDOE Report ( K-GD-1800, Chk Ridge, Tennessee (1978). 4 30. Stephe,nson, M. J., and Eby, R. S., Gas Absorption Process, U. S. Patent 4,129,425 (1978) f Stephenson, N. J., Analysis of a Fractional Gas Stripper, USDOE 31. Report K-1895, Oak Ridge, Tennessee (1978). Toth, L. M., Bell, J. T., and Fuller, D. W., Chemical and Physical 32. An Inter k Behavior of Sc e Conta--irants in the R-12 Off-Gas Process: Report, USDOE Report ORNL ~IM-6484, Oak Ridge, Tennessee (1978).

• g}33. Wood, D. E., Availability Ar. tysis of the Freon Absorption System for

x.ln.

.Wating Effluerts from Reprocessors, USDOE Report ORNL-TN-5797, Oak Ridge, Tennessee (1977). ~ /

34. Ya= roto, Y., and Takeda, H., " Solubility of Kr in Some Organic Solvents," J. Fac. Er.g. U. Tokyo, A-7, 44 (1970).

\\ i \\ 4 k 1

n- / ~ [ ~ 'sf l (,_-) - -i March 28, 1980 i (,[ o ,- y 0[ Public Utilities r " 100 Interpace Parkway ~ ~ ..Passippany,PU 07054 QiAttenticn: Mr. Hermn Dieckamp =.e-.

gDmr Mr. Dieckanp

i l ~ j I offer my hunble solution to the rmova1 of the so called 40,000 curies of krypton gas frm the %rm Mile Island vessel. I feel certain Qthis can be Imoved inexpensively with the least axunt of objecticris by citizens or enviromentalists and above all the most safe' yet suggested..- m t - ~) no use of large ballocns capable of travel to' the stratosphere 1 and large erotch to transport a contaiment vessel capable of carrying an appreciable amount of the gas under pressure.

2) After the balloon reaches a high enotyh altittrie, relief valves on the ccritaiment vessel can be relmsed by radio control, and later the balloon can be destroyed by explosives.' - '~~

.\\\\ his can further solve the apparent transportation problen if an effort is mde to transport the gas by land. le.(citizens objectire to , or accidental discharge). I needn't explain the further reprecussions or costs connected with release of the gas in the atmsphere directly above the Three Mile Island plant. ~ Perhaps this could be the cnly safe solution and certainly should be considered. Sincerely, Albert B. Snizik l gg," cr~KtFn?~dfroTE Det' tW oc: Riifds d W M h Robert Arnold - Executive Vice President, N. R. C. l

~_ SAW7 \\ J. A. Van Vechten [ R. J. Gambino x' J. J. Cuomo 4 \\' Encapsulation of Radioactive Noble Gas Waste in Amorphous Alloy Public demandfor the containment and safe storage ofradianctive waste materials has caused the U.S. Government to reqmre that, begmnmg in January I983 most of the "Kr. which untilnow has been vented on the atmosphere durine the reprocessmg of nuclearfissionfuel rods. well hase to be captured and retainedfor several decades. The cost ofaccom-prishing this with present compressed-gas technology is enough to increase the cost ofnuclear generated electricity by an estimated 0.37c. However. materials developedfor amorphous magnetic bubble memorv devices have been found to be capable of storing large quantities of Kr (30 atomic percent) with great stability up to temperatures above 1070 K. The cost of"Kr storage in the magnetic bubble memorv material appears to be less than IVe of thatfor present com-pressed. gas technology. s -.. Introduction The problem of safe disposal of radioactive wastes from Corp. and which has developed several methods [1] to nuclear fission power plants is a major obstacle to the capture "Kr. The National Engineering Laboratory re-continued and expanded use of fission reactors. Perhaps processes only U.S. Navy nuclear fuels: there are no the most difficult radioactive fission product to capture commercial reprocessing plants at present. and contain is an isotope of the noble gas Kr. "Kr. which has a half.hfe of 10.7 years and emits p. particles at Figure I shows the increase in atmospheric "Kr meak energies up to 0.67 MeV and y-rays at 0.5 MeV [l. 2). sured at various geographic locations up to 1968, at which Unhke most other hssion products it is neither solid time there were about 56 million curies (56 mci)or about (above 128 K)in its elemental form nor can it be reacted 10" atoms of "Kr in the atmosphere worldwide 91. Al-to a stable sohd compound. Although heavier than air. It most all "Kr is introduced by man; of this only 57,is due mites thoroughly in the atmosphere;if released even in a to nuclear weapons testing. If the rate of expansion of deep mine shaft, it would quickly diffuse into the atmo-nuclear power along with the concomitant increases in at-sphere. It also diffuses rapidly through water and earth. It mospheric "Kr experienced up to 1%8 had continued, is produced in about 0.37c of all "U fission events. This~ there would now be about 0.6 GCi or about 10" atoms of is about 6c* of the Kr and 0.87c of the noble gas produced "Kr in the atmosphere [4]. (The medical consequences of by fiuion of "U. (The other major noble gas produced is this dose are argued [4]to be slight.) The actual amount is Xe.) Almost all processors of nuclear fuel around the much less due to slowed progress in bringing on nuclear world have allowed these radioactive gases to escape to fission power as a replacement for fossil fuels.The rate of the atmosphere. (It should be noted that essentially all the release has also been limited by the fact that spent fuel Kr is released in reprocessing: less than IPc is released from power reactors is not being reprocessed at present. from the reactor (3).) One exception is the Chemical Pro-Spent fuel is stored on-site in deep pools, an unsatisfac-cessing Plant at the Idaho National Engineering Labora-tory procedure for long-term storage. If nuclear fission tory. Idaho Falls, which is operated by Allied Chemical power were to provide the projected fraction of our en-Copyright 1979 by International Business Machines Corporat;on. Copying is permitted without payment of royalty provided that (1) each reproduction is done without alteration and (2) the Joumal reference and IBM copyright notice are included on the first page.. The title and abstract may be und without funher permission in corr.puter. based and other information. service systems. Perminion 278 to rrpullish other excerpts 3nould be obtained from the Cditor. 4 8 4 % 6 % D(HtFN FT AL last I. A Fs DFV Ft.nr e vnt. *s..%O. s e se u in -._. a - __ _

m - cw ..w. m_ e. = +. t, s ergy needs and if sirnple venting were to continue. the ,o_ atmospheric burden would level out at well over I GCi. It too so-might also be noted that I GCi of"Kr produces 4 MW of e2' power, which might be put to some practical use if it o,3 "~ [8 could be safely handled; admittedly, this is an almost neg. [3 0 3o-

• p
• ligible amount compared to the total power that would be g

produced by the reactors. 2, I,, " ~I o ] To give perspective to the quantities involved, let us o to- .~ o ge, note that the fission of'"U produces 200 MeV of thermal / o / o ,4.,,,,,,,,,,,, energy directly and, depending on design, approximately 1,33 ss 37 se si as ss et se another 200 MeV of thermal energy by emitting neutrons that produce other fissionable isotopes, principally '"Pu Ys= and '"U, by transmutation. Thus, the complete fission of ngure : Atmospheric "Kr as a runeiton or time un to 1968. one gram of '"U in a typical reactor would produce about pata taken from Ref. f 31 - 5.2 x 10' watt years of heat. As nuclear power plants are about 32% efficient in convening heat to electricity, this one gram of'"U would provide about 1.7 kW of electricity for a year. A typical nuclear power plant generates 1 GW the amount produced by 67 standard nuclear power of electricity. To run such a plant continuously for a year plants, each producing 1 GW of electricity [2].) requires the complete fissioning of 0.6 Mg of '"U. In a typical fueling cycle 3% of the initial charge and 1% of Where the Kr has been captured, the only technology the spent fuelis '"U, so that fifty times as much material available for storing it is to compress it into cylinders [2]; rnust be processed as is fissioned. At this rate of produc-133 cylinders 23 cm in diameter would be required to con-tion, the alternative of storing spent-fuel bundles on-site tain the noble gas released each year at each fuel repro-is untenable. Thus a typical plant would require 30 Mg of cessing plant. There are several problems with this fuel to be reprocessed each year of continuous operation. method of containment. Rubidium, the decay product of Of this mass, about 390 g would be "Kr, about 5 x 10" "Kr. causes a deterioration of ferrous alloys: so there is atoms or 2.8 x 10'Ci. If we project to the year 2000 and doubt about the long-term integrity of the cylinders. assume that each of 3 x 10' Americans is to be provided There is also the danger that the cylinders might burst due electric energy totally supplied by nuclear fission at the to some accident in handling and transport or due to cor-present average consumption rate of 2 kW,i.e.,600 GW rosion-and radiation. induced damage ove: long periods for the nation, then 600 standard 1 GW plants would be of time. Because the radioactive gas is present in large required for the U.S. alone. These would produce 2.3 Mg quantity and under pressure, such an accident could eas-or 1.7 x 10'Ci of"Kr annually. If nuclear power were to ily be fatal to those nearby unless some means of second-provide only a fraction of this energy need c ? 'he aver-ary confinement of the gas is provided. He cost of meet-age electric consumption were to decrease, the "Kr re-ing federally imposed safety standards with the com-Icase would be correspondingly reduced. World produc-pressed gas technology is rather high [2]. The estimated tion of "Kr would be at least three times this figure. cost of a facility to contain on a 40-year cycle the com-pressed gas produced by a single reprocessing plant is U.S. Federal regulations to take effect January 1983 [5] 5208.5 million. For a 30-year loan at an 11.5% interest will limit the amount of "Kr that may be vented to 5 x rate, this would require an annual payment of more than 10'Ci/GW of electricity generated for one year, for fuel 524 million. The cost of compressing the gas, of purchas-irradiated in 1983 or thereafter. [ Editor's note: The global ing and transporting the cylinders, and of salaries and en-body dose rate per capita from the release of all of the ergy would be additional. The warehouse cost alone "Kr generated in continuous operation of a 1-GW (elec-would run to more than $200 million per year for the U.S. tricity) reactor is -2 x 10'* mrem / year (rem = roentgen by the year 2000. In other terms. this would add 50.00006 equivalent man). This dose rate is about 2 x 10-' times to the cost of generating a LWh of electricity, which the av erage background dose rate: see Reference [3].] Re-would he an increase of about 0.3%. processing with unrestricted venting would result in a re- ! case rate about seven times higher than this. The fuel Proposed alternate methods of storage have included I reprocessing plants would be responsible for keeping the incorporation into zeolite lattice pores by high temper- "Kr release down to this level. (A standard reprocessing sture pressure diffusion and by incorporation into crystal-plant handles about 2 Gg of spent fuel per year, which is line [2a] and amorphous [2b] metals. The zcolite method 279 l 88W # mEh D>$ t toF

• WoL. 25 + ho )
• st AY 89M 3 A. V AN 4 tCH f f % F T Ai.

.,r l l -% Tma t*=al Fig. 2. A low. pressure discharge is established in a sput. tering gas between two electrodes. one of w hich is known T"8" as the target and the other as the substrate electrode.De fc,+ {x,+ h,+ spurtering gas is normally chosen to be one of the noble [.' M'M) ~ Y*. h@M$ JM.Ig ' gases. He, Ne. Ar, Kr. or Xe. to avoid chemical reactions i f-f N.' @1 with the target and substrate materials. In practice. Aris +y "4 i. nsually chosen on the basis of cost and sputtering rate. / ' d ; h k M u,p ne Kr and Xe sputter as rapidly in most applications but ?

'g.,g;.

'.N,g L are more expensive. The discharge converts the noble gas ( g g,j g ~. } to a positive ionization state e.g., Kr*. These ions are - 4 g,i - 4 g,+ - 'f y -' --% g accelerated toward the target electrode, which is biased e f l s,,%, negative with respect to the plasms by the targetpoten. I h=* I tial. The plasma is in turn biased from ground by a small -g so.p. plasma porcarial. When the noble gas ions reach the tar. Schematic representation of the bias sputtering pro-get surface they penetrate several atomic layers, produc-re 2 ing a process known as a collision cascade in which the energy of an incident ion is transferred to many atoms of the target material. Several of these atoms are sub-l sequently emitted from the target surface in a manner { similar to the " break" at the start of a game of billiards. f sufers from the fact that if water gets to the material. it The target atoms are generally neutral and travel by vir- '{ reacts and rapidly releases the gas. In crysta!!ine metals the Kr forms small bubbles. At high concentration the tue of their kinetic energy through the intervening space pressure in these bubbles is sufficient to cause mechanical between the target and the substrate, perhaps suffering a failure, a phenomenon known as blistering,in which the few collisions with the sputtering gas on the way. For nor-i mal choices of substrate temperature and materials, virtu-I gas escapes. Furthermore, these bubbles tend to collect ally all of the target atoms reaching the substrate stick at grain boundaries and microcracks along which they dif-there. As normally practiced, this results in the growth of fuse at significant rates even at room temperature. More-a polycrystalline film on the substrate. However. Nowick i l over, due to the power produced by the decay of the "Kr. and Mader (8) discovered that when two or more ele-the containtnent material will be self. heated to a temper-ments are deposited simultaneously and the radii of their ature dependent on the size of the individual container; i atoms are sufficiently different, the resultant films are not the larger the container, the larger the masimal temper-polycrystalline but amorphous. (This me.ms that they are ature and the more severe the thermal diffusion and deg-f e[, radation. For most storage schemes the volumes of con-microscopically disordered but macroscopically homoge-neous as contrasted to the polycrystalline films, which l tainment material required are substantial. Each repro-are microscopically ordered but macroscopically dis- !I cessing plant would require [2] the following volumes per ordered.) It is also possible to obtain amorphous films f year for the various proposed methods: compressed gas with atoms all the same size if one deposits faster than a cylinders,6.5 m*; zeolite. 7.3 m*; Ni,1.3 m*; Al, I.6 m'; glass, >l90 m*. critical rate, this rate being a function of substrate tem. perature [9). Storage in bias-sputtered amorphous meta!!ic alloy in bias sputtering, a substrare hint is also applied be-In the course of development of amorphous materials for tween the plasma and the substrate. This has the effect of i magnetic bubble memory devices [6), we have come upon accelerating rioble gas ions toward the surface of the l a method for the storage of Kr. Xe and other noble gases. growing film as well as toward the ' target. The ion bom-whether or not radioactive, which seems capable of con-bardment of the film during growth has a number of useful j taining the radioactive waste from one of these reprocess-effects. In the first place, it introduces anisutropies in the ing plants in just 0.2 m' of material, and of retaining it properties of the film. In the development of amorphous stably up to temperatures as high as 1070 K. We estimate magnetic bubble materials, it was necessary to use this the cost of storing the "Kr by this method as well under effect to induce a perpendicular easy axis of magnet-i I", that of storage in the compressed. gas cylinders,l.c.. ization. In the second place, it allows one to climinate less th:.n 50.24 rnillion per reprocessing plant. many types of irnpurities that are not as well bound as host atoms. This is done by inducing a collision cascade, The containment materials in question are formed by in the substrate that is not sufficiently violent to remove 280 bias-sputter depositiori [7). This process is illustrated in host atoms. A thir.d effect. which was discovered by s s o urnirs er st. rass s. itrs no rt e>r. soi. s. so i. u ss iv = r g4 _._______h"'-

y 590 K. The authors associate these peaks with mecha. chemical nature of minor impurity constituents and can nisms having activation energies of 1.31. l.74. 2.21. and exist over a broad range of composition. The amorphous 2.78 eV. respectively, a!!oys in question will contain about 30 at% Kr or Xe. but as noted above, only 6% of the total Kr released at Rantanen er al. [13a]also studied the thermal re-emis-the reprocessing plant would be radioactive "Kr. Let us sion spectra of Kr from polycrystalline Ni. They reported assume that the Xe is separated out by distillation so that activation energies of 1.18. l.36,1.50, and 1.7I eV for only Kr is stored. This would seem to be economically this case. They also pointed out that these activation desirable, although one could also easily store the Xe by energies are probably associated with interstitial migra-expanding the size of the sputtering unit. Eventually, tion (1.03-1.09 eV). vacancy formation (I.35 eV). l.8 at% Rb wi!! be contained in the storage material. This vacancy migration (1.55 eV), and surface diffusion would be enough to affect many crysta: fine hosts sub. (1.68 eV). stantially but would have a negligible effect on a drphs-amorphous host. Such host materials are also less suscep. The above results for Krin polycrysta!!me Ni should be tible to radiation damage because the currents produced compared with the thermal release of Kr from amorphous by ionizing radiation do not persist as long and because GdCo and GdCoMo alloy films by Frisch and Reuter[14]; the resultant atomic diffusion does not have as much ef. see Fig. 3(b). The method used to study the amorphous feet on a structure that is already disordered. film was similar to that of Rantanen et rd.. except that the heating rate was 10 K/ min and a high sensitivity mass The selection of the most practical composition from spectrometer was used. Extensive rneasurements have which to form the encapsulating host material requires been made on a large number of these bias-sputtered the consideration of four factors: gas-incorporation ca-a~.r..ns GdCo and GdCoMo alloy films. All of the pacity, thermal stability, chemical stability, and cost. Let thermal re-emission spectra for uno.tidued films have the us start with the amorphous magnetic bubble memory charzcter shown in Fig 3(b). Oxidation lowers the tem-material. GdCoMo. for which the incorporation of large perature at which Kr release occurs [14b]. In the quantities of noble gas was first discovered. This material amorphous alloy films no detectable rate of noble gas evo-can incorporate more than 50 at% Ar and more than lution was observed until the film began to crystallize 30 at% Kr and Xe when the three bias voltages of the [14a]. At the crystallization temperature the gas was system are adjusted properly. This large noble gas incor-evoh ed very rapidly. In this case the kinetics of gas liber-poration capacity occurs because the rare carth element ation are determined by the kinetics of the crystallization. Gd has an atomic radius much larger than the first-series which is a nucleation and-growth process. An activation transition element Co. The second-series transition ele-energy of 4 eV has been estimated for the migration of Kr ment Mo is interrnediate in size and serves to further dis-in amorphous GdCo alloy [14a). This implies that the order the drphs structure so that these mixtures will con-mean time to diffuse one atomic site would be about 10" dense in an amorphous phase over a wide range of com-years at 570 K; at 1070 K, the Kr would diffuse about positions and will have a relatively large nur:ioer of 10 nm in the 40 years required for the radioactivity to interstitial spaces large enough to accommodate a Kr or decay to 37c ofits original value. Xe atom. However, the GdCoMo composition of the { magnetic bubble memory would not be an attractive i A further benefit of an amorphous structure for a mate-choice from the point of view of cost. Because the rare-I rial to contain "Kr is that the disorder improves the abil-earth elements (which in fact are not that rare) are all very ity of the material to tolerate radiation damage and impu-similar in their chemical behavior, they are expensive in rities. Even if the containment material were pure to be-their pure elemental form. A typical price for pure Gd sin with, it would not remain so because the "Kr would be $500'kg. If one instead purchases the rare earth transmutes to Rb by radioactive decay. The stability of a elements in an unseparated form called mischmetal or crystalline hust material would be adversely affected by RMM [15]. the price is much less, typically SI(Fkg. and the simultaneous effects of irradiation, which generally the chemical behavior as it affects Kr storage in enhances atomic diffusion and of the incorporation of the amorphous alloys is no worse. One can also replace Co daughter isotope, which is chemica!!y incompatible with with Fe without affecting the containment properties sig-the crystal lattice of the proposed host materials. This nific:.ntly. With respect to thermal stability, it has been would cause embrittlement of a crystalline host material shown that GdCoMo and GdCoCr ter; try a!!oys are and would accelerate mechanical failure by such mechs-much more stable than binary a!!oys like GdCo or even nisms as blistering. However, those amorphous alloys ternary alloys containing Au or Co. e.g., GdCoAu or which are stabilized by atomic size mismatch and a highly GdCoCu. For example 15 to 20 ats Mo increases the 282 disordered drphs structure are much less sensitive to the crysta!!ization temperature from 770 K for GdCo to more 3 g 44% 4f t Mit % ET 4L law L AFs uLVFinP

• VOt. :)
• sn 3
• M 4Y tv"v

Wareunoisd subursie rad I , Fe,3Cr,,her OOO OO O O O, r OO >7l,-' O O O' OO HeaatamateN*F of subursar.eds OOO OO v- % ~Y'- J OOO OO A OOO y_j e^;:- t OO 7 OOO l OOO OO OOO OO Krypes OOO 6'-' OO 2" R ~~ \\., \\ OOO / h"7 M tb) tc) Figure 4 proposed sputtering apparatus for incorporation of Kr into the amorphous alloy on a production scale. (a) Modular spunering unit (b) target bar assembly, and (c) sputtenng chamber. than 1070 K for the ternary alloys. From the point of view trapped in a cryogenic distillation tower [1]. The Kr and f ofchemical stability, the rare earth concentrations should Xe would be trapped at the end of the distClation sequence be kept low because these materials oxidize (as well as in cold traps or on charcoal cooled to 77 K with liquid cost more than the other constituentsl. Chromium, on the

nitrogen, other hand, significantly improves the oxidation resis-tince and should be added at a concentration consistent The liquified noble gas is maintained at 77 K and trans.

l with its cost. Therefore, an appropriate composition for ferred to the sputtering station for incorporation into the j the containment application would be (in atomic percent): amorphous a!!oy: see Fig. 4. The vapor pressure of the Kr RMM 20% Fe 60%, and Cr 20rc. The 2.1 Mg, or about at this temperature is about 10'Pa (10atm). which is 0.2 m'. of this composition that would be required to enough to bierd through valving into the sputtering cham-st:re the Kr retrieved at each 2-G;/ year reprocessing ber but low enough that the danger of excessive leaks plint would cost about $10 thousand. Of course, this ma-would be easily managed. The gas pressure in the sputter. terial could be recovered and recycled every century or ing chamber is about 10Pa (10atm). [ Compare this so as the level of Kr radioactivity from each charge de-situation with that of the compressed gas cylinders, which creases. handle the gas at a pressure of about 10' Pa (10' atm).] I 1 The process The rate at which material may be deposited by bias-At the reprocessing plant, spent fuel elements containing sputter deposition varies from I pm/h for very simple UO, ceramic pe!! cts encased in metal are dissolved in ni. diode systems to 30 pmh for systems that use electron-tric acid. At this point the Kr and Xe are released and injection or mvetic. field confmement of the plasma. We bubble out of solution together with several other volatile feel that the rnost practical arrangement would be modu-species. The various volatile species can be separated and far and would consist of a hexagonal array of water-283 late 8 k t % DEv Ft.or.

• vol 23 +.*=o )
• W AY f f59

$ A. VAN % FCH rFJs if AL

cooled substrate rods surrounded by bar. shaped target For Anal storage one might wish to pot the entire top electrodes. With this arrangement a continuous deposi. Aange, rod and target remains assembly in cement and tion rate of 10 m/h would be practical. In order to de-wrap it in lead. However, we feel that the amorphous al-posit the 0.2 m' of material per year required to contain lay is so stable a method of storage that th: material could the Kr tetricved at each 2-Gg/ year reprocessing plant. the be released for several practical applications (ranging volume deposition rate will have to be 2.3 x lo-*m'ih. so from nuc! car batteries to fire detectors. cold-cathode sta-that 2.3 m' of deposition area are needed. This can be ac. bilizers. thickness monitors, and simple sources of heat) commodated with a system of 232 rods 2 cm in diameter rather than simply putting it away in a deep salt mine. and 30 cm long arrayed honeycomb fashion in a cy-lindrical vacuum chamber 1,5 m in diameter and 0.5 m Conclusion high. Such sputtering systems sell commercially for about $80 thousand (16]. The materials developed for the amorphous magnetic bubble memory system have been shown to provide a i About 200 kW/m' input power would be required to very stable medium for the long. term /high-temperature sputter at the proposed rate of 10 m/h (17]. Therefore. storage of the noble gases Kr and Xe. The radioactive the sputtering station would consume about 460 kW of isotope "Kr. produced in *U fission reactors. is difficult electrical power in order to capture the Kr retrieved at a and expensive to contain by other means. Compared to 2.Gs/ year fuel reprocessing plant. At 50.04/kWh the cost the present technology of compressed-gas cylinder stor. of this power would be 5160 thousand per year. Perhaps age which is estimated to cost 524 million per year per another $10 thousand per year of electricity would be reprocessing plant for warehouse amortization alone, our consumed running the vacuum, cooling and control sys-process would cost approximately 5180 thousand for cap-I *5* ital equipment, which would be a nortized at less than 540 thousand per year, plus 510 thousand per year for materi-l Due to the inherent simplicity of the sputtering process als and 5170 thousand per year for electricity. In our eco-itself, this could easily be automated or remotely con-nomic analysis we have not considered the cost of the trolled. The cost of special control equipment for the ra-building to contain the process; but since the process runs dioactive environment automated operation should not at high vacuum instead of at high pressure and since the exceed $100 thousand. However, the deposited rods product is quite stable to high temperatures, we feel the would ha. to be removed and replaced periodica!!y.This cost of this building should be minimal. In the high. pres-could be accomplished by valving off the source of"Kr sure cylinder technology the cost of the building is a ma-l and of the cooling water. breaking the vacuum of the sys-jor part of the total expense. With our process the radio-tem, and pu!!ing the top 11ange of the vacuum chamber active materialis present only in small quantities before it I with all the rods and the remains of the target electrodes is incorporated into the solid, and because of the stability attached to it out of the body of the vacuum chamberand of that solid, can be dispersed in practical applications afterwards. removing it from the sputtering station. Operators could then attach a new top flange with substrate. rod assembly and target electrodes to the vacuum and cooling systems. References and notes This should be done about once a month afterabout 7 mm C. L Bendnsen. G F. que, and B. R. Meeler. Q j genie Rare Gas Recovery m Nuc! car Fuel Reprocessing, of material has been deposited on the rods. i Chem. Eng. 78. 24 (1971). i

2. (a) D. A. Knecht. "An Evaluation of Afethods for Immobi-The configuration of the target electrodes shown. Fig.

liting Krypton-85."(Jaka NationalEngineering Laborators in 4 indicates that these consist of Fe,,Cr,, bars with misch-g,,,,o scr.ii25. Allied Chemical Co. Idaho Falls. !D. Jul'y 1977. and personal communications. tb) G. L. Tingey er al., rnetal plugs inserted into drilled holes. This configuration Pat'fe N0'thurst L.sburernry Quarterly Procress Reporr j ts recommended for easy handling of the mischmetal. Nos. BNWL4179. 2:45, 2290. PNL-23771. 4. and.3. and 1 which is hard and brittle. With this configuration one PNL43731 and -2. Pacinc Northwest Laboratories. Rich-land. WA 99352. could also arrange to coat the deposited layer of

3. Communication to the Editor by E. J. hfoniz. htassachusetts

($ amorphous metal with crystallm.e stainless steel.in order Institute of Technology. Dept. of Physics. Cambridge. St A m39, y,no,,y g, 3979, to provide further protection from corrosion and abrasion.

4. E. Csongor. ** Atmospheric "Kr 5feasurements between and to contain the bett-particles emitted by the Kr. This 1966 and 1958 in Debrecen. Hungary." Acta Physica 23.109 would be done by continuing to sputter after the Kr (1970) source had been turned off and the mischmetal plugs 3,.. Environmental Radiation Protection S:andards for Nuclear Power Orerations." rederal Regiurr 42. No. 9. tirte 40. part nearly consumed and the bias voltage would be increased

{0.g357 nu ry 13 77 to 250 V m order to increase the fraction of Fe and Cr in 284 the deposited mitture. .Arnorphous ste:allie Fi!ms for Bubbte Domain Appli. cations." IBM J. Res. Derefop. 17, 66 (1973). 34 v g% % Friet F% f r 44, mM J AES DMi r>P = %OL 23

• No 3
• M 4Y 8 /9

1 I. l f t e

7. D D. Dove. R. J. Cambino. J. J. Cuomo, and R. J. Kobfiska.

Allied Tupir s. Houston. TX. May 1975. p. 560; (b) M. A. "SicaJy.Stase Mass Dal.nce Approach to Substrate Biased Frisch and W. Reuter. " Studies on the Thermal Release of ef Spunering "/. Var. Jef. TerAnnt. 13 % 5(1976). Kr from af Sputter Deposited Amorphous Gd.Co Films by a .l lL A. S. Nomct and S. R. Mader. "A Hard. Sphere Model to UHV Mass Spectrometne Technique." submitted 101. Vac. L Simulate Arioy Thin Films." !8tM l. Res. Develop. 9. J58 Sci. Technol.: also pnvate communications with authors, igr455. IBM Thomas J. Watson Research Center. Yorktown 9 g. l.. Chopra. TAin film rArnumena. McGraw-Hill Book Hei hts. NY. C Co. Inc.. New York.1969. p. IFT.

15. Mischmetalis the material from which cigarette lighter flints 10, J. J. Cuomo and R. J. Gambino " Incorporation of Rare are inade, and is available from Moly Corp.. Inc., White Gnes in Spu'8e'ed Amorphous Metal Films.** /. Voe. Sci.

Plains. NY. Tre Aa,4.14.152 (19771

16. For example. Ultek. Santa Clars. CA: Varian Associates,

11. G. M. MCCr*cken. "The Behavior of Surfaces under lon Palo Alto. CA; and Sloan. Santa Barbara. CA.

Domb.ardment." Reports an Progress in Physics 38. 317

17. N, Hosokawa.T. Tsukada. and T. Misumi "Self. Sputtering s1975).

Phenomena in. High Rate Coasial Cylindrical Magnetron f s2. G. S. Cargillill." Dense Random PacLing of Hard Spheres Sputtering "1. Voc. Sci. TerAnal. 14, 143 11977). es a Siructural Model for Noncrystalhne Metallic Solids."J. Appl. Phes. al. **.249 (1970).

13. ::s R. O. Ransanen. A. l Moeri, and E. E. Donaldson." Ion Burial and Thermal Release of Noble Gases at Nickel Sur.

8,. faces."J. Var. Sci. TerAnol. 7,18 (1970). (b> J. F. Truhlar. Receis ed Nurember 6.1978; revised January 8.1979 l E. E. Donaldson. and D. E. Horne. " Trapping and Thermal Relee e of Noble Cases at a NicLel Surface."1. Appl. Phys. l. s 43, 2139 (1970).

14. (s) M. A. Frisch and W. Reuter " Thermal Stability of Thin Films by Modulating Beam Mass Spectrometry." Pro.

The authors are located at the IBM Thomas J. Watson a erJonys of the 23rd Conference on Slass Spectrometry and Researth Center. Yorktower Heigluts. New York 10398. (St4)SM-Spo f r i t., I. I l 6 ? 9 285 IM8 s a f s Dt s t t UP

• b ot. :)
• NO. )
• kt O 109 I^

^~ d

1 - r.cew r; m - m e_ .... _... :, a ?+ l ~ ' The release of the contaminated water from the reactor poses i the same type of long-term risk to public health. In fact, it is more of a j i threat to public health because the Susquehanna River provides the drinking i l water for southeast Pennsylvania and northeast Maryland residents. It is a I major tributary to the Ctesapeake Bay -- one of the U.S.'s most fragile and productive ecosystems -- thus, further radiation contamination can result by the incorporation of long-lived radionuclides in the food chain. Should the Chesapeake be contaminated by the TMI radioactive wastes the economic and environmental reprecussions would be devestating. The federal government has consistently maintained that TMI radiation releases are not harmful to the public. It has not been able to determine, however, what it causing the increased incidence of spontaneous abortions, stillbirths, and illnesses among TMI residents. Radiation may not be the only reason for this increase but it is unlikely that it has not at least contributed to it. Because releasing the wastes will create the potential for additional health problems among a larger population and contaminate the environment, the Environmental Policy Center proposes that (1) the NRC adopt an alternative to releasing the radiation into the environment, such as entombment; (2) the Environmental Protection Agency increase its on-site and l l off-site monitoring capability; (3) the Pennsylvania and Maryland Health Departments monitor vegetables, fruit, and dairy products grown down stream from TMI for strontium: (4) independent monitoring systems be implemented: (5) the NRC, EPA, state, and independent monitoring data be analyzed by f independent reseachers; and (6) the cost / benefit analyses include the long-l term health costs created by TMI. _ _ _ _ - _ _ _ _ _ _ _ _ _}}