ML19323B677

Development of Cryogenic Krypton-Separation Sys for Off-Gas of Reprocessing Plants, Presented at 15th DOE Nuclear Air Cleaning Conference

ML19323B677
Person / Time
Site:	Crane
Issue date:	04/24/1980
From:	Aamon R, Burkhardt H, Hutter E GERMANY, FEDERAL REPUBLIC OF
To:
Shared Package
ML19323B642	List: ML19323B645 ML19323B647 ML19323B649 ML19323B651 ML19323B654 ML19323B657 ML19323B662 ML19323B667 ML19323B670 ML19323B674 ML19323B677 ML19323B680 ML19323B684
References
NUDOCS 8005140062
Download: ML19323B677 (12)

Text

"

15th DOE NUCLEAR air CI~EINING CONFERENCE NVZTF?m 7e j

~

~

V. sfy?Qg% ;s.

w.

go05i40. %.s....3 M I

~

a..s i

DEVELOPMENT OF A CRYOGENIC KRYPTON-SEPARATION SYSTEM"' "~M 1C p

'e M

-" P FOR THE OFFGAS OF REPROCESSING PLANTS

. _ d-: :

j

y. y '

R.

v. Ammon, H.-G. Burkhardt, E. Hutter and G. Neffe c

Kernforschungszentrum Karlsruhe GmbH, Federal Republic of Germany J

s f.

c' f i E

Abstract l

4

.r i

The concept of cryogenic rare gas separation from reprocessing r

offgas pursued at present at the Kernforschungszentrum Karlsruhe (KfK) is discussed and compared with other offgas purification flowsheets.

The KfK concept includes: separation of O and residual NO by cata-lytic reduction with H adsorptive reten ion of H 0, CO 2

2, NH - etc then Kr-Xe mixtdres..

and cryogenic distillakica of first N -Kr-Xe, y

Some features pertinent to this flowsheet which were studied experi-mentally either on a laboratory or semiworks scale, are described including the following: desublimation of Xe in the first column, coadsorption of Kr at and its selective desorption from the adsorp-

[

tion bed, purification of the Xe product, poisoning of the reduction j It is shown that all of these

/

catalyst and methanization of CO2 features, despite being capable of causing malfunctions of the pro- /

cess, can be controlled by proper process design and operational d

conditions in a way not to impair with a good Kr decontamination factor.

I.

Introduction Plans for retention of the fission gas Kr from the offgas of f

the future large German reprocessing plant as well as of the existing small plant WAK (Wiederaufarbeitungsa_nlage K_arlsruhe) prefer a cryo-9 genic distillation scheme. The reason is the long experience with h

such a proca g in the industrial production of rare gases in air 1

liquefaction hl Two essential differences between the rare gas-oxygen mixtures gi in air liquefaction plants and reprocessing offgases demand for modifications of the process:

- the mole ratio Kr/Xe in fission gas (0,103) which is almost reci-d procal to the ratio in air (13,1 ) ; the resulting higher Xe-concen-l trations enhance the problem of Xe freeze-out;

- the radioactivity of Kr-85 resulting in the radiolytic ozone forma-a tion from 02*

Depending on the risks one places on these differences the existing (or planned. Kr retention units show different flowsheets J

1) 2). At the I.N.E.L.

(U.S.A.)

(3) and at the C.E.T.

(France) d N( fig.

) the complete offgas including 0 is liquefied in the first d

2 column. Here the risk of ozone formation is accepted, but freeze-out l }j of Xe is decreased because of improved solubility in O compared to ld N. Further enrichment of Kr is achieved either by bat h distillation j'!

of the three-component mixture 0 -Kr-Xe in the second column (I.N.E.L.)

2with H between the two columns and Ii or by catalytic reduction of O2 2

!91

,l 640 a

s

_.., ;.,.; e ;p.,4,

~-

_,}

.c.,W',

(h subsequent dictillotion of Kr-Xa (C.E.A.).

3. 7' 4.

1 g

-. e..

r i

_~~c op. s,.

s.

4 M S h.

jf

{"'

r -- - - - 9 D

! o,.n.ao a 'r---

a/

i. - _ _... _ ;

j, sc.-%amoa

.I en g',

, /s, 7

cw >

r.

g h,

g, '

m,

,g

=

in on.sc.n w

p.

1*:

/. \\,

.k[

4.

=

\\

i g1

'3

. 5. s10 *

• /.

,1 C.

bar bo bar 4

g F-

.u N

g.

K Aesorpe n s.

(

Cth

,a

~

m,, / ' /

4

_ ion,,

a con T.,

A,.

$e r---

_-f,,

LL h

J

~

rsunc 2 a _e, w., p'" *5 __ _ _ _ J Fiessure cependence f.ceof bewor) c' tne tve-phase oreo g-y J

50id-laQuid O the CAc5e d'Ogrorn Ny-Kr Xe and Course of bottrn f

p'Oduct Concentrotons durng stort-up ODeroton of cobrnst t

'j'

---*" "' a:>o'o"o' *'

T

/

7' us_ u, y

CO,go ng D'

]

FIGURE 1 withKr recyce Dfierent flowsheets for the retenton of Krypton

[

trorn the offgos of LWR.reprocessrig plants i

i e

Both risks are circumvented by proposals of Linde { Germany) (5)

(U.S.A.) for the Japanese Tokai Mura plant ( ): both 02 and by Airco and Xe are separated from the offgas before the cryogenic part by catalytic reduction and by adsorption, respectively. Workers at KFA prefer a separation of Xe by freeze-out instead of adsorption (7 ).y)After separation of Xe the cryogenic distillation is JUlich (German simplified to the two-component system N ~Kf*

2 In the present concept of KfK(8) similar to the plans of the (Belgium)(9) O., is also removed by catalytic reduction, but C.E.N.

l adsorption of Xe is omitted. Thus the three-component system N,-Kr-Xe is driven off at must be distilled in the first column. Whereas N 2 Kr and Xe are collected in the bottom of the still. After the head, discontinuous transfer to the second column Kr and Xe are separated I

into pure products and filled into steel cylinders. The cryogenic Part of this concept together with a preceding molecular sieve bed it was installed on a cold" 'semiscale basis (gas through-as drying ug/h at S.T.P.). These units which bear the names KRETA Put 30-50 m and ADAMO (Adsorption (jirypton-Entf ernungs-T_ief temperatur-Anlage) have been in operation now for 4000 hours0.0463 days <br />1.111 hours <br />0.00661 weeks <br />0.00152 months <br />. The

)

an _Molekularsleben) 641 2

ta n;

};

-e -

y L! $,,.

docign of thace unito hne b;en doccribed olecwh;ro(10).. Some Cf the gj results gained are reported in the following sections.

h II. Features of the KfK Flowsheet and Results 4

~-

M i 1.

Desublimation of Xenon 3Ji 1.1 Solid-Liquid Equilibrium

>s The phase diagram of the three-component system N -Kr-Xe has a 2

f region of limited solubility on the N -Xe side. Solid Xe is in 2

equilibrium with N -Kr-Xe liquid (fig. 2). This two-phase area 2

decreases on increasing the pressure and thus the temperature. One r

bM could therefore prevent Xe from freezing out completely by operating YU the column at 17-20 bar (11-13). Because of safety aspects in plants f

containing large inventories of radioactivity we chose an intermediate M

pressure of 5 bar which is also familiar practice in air liquefaction.

a fd Fig. 2 also shows the problem of Xe freeze-out during the start-i up period of the column: if the fission gas mixture is fed into the column cooled down with liquid nitrogen (at 95 K) one would pass 9

straight through the critical area (dot-dashed line). This situation 6

can be avoided, if the column is filled with a N -Kr mixture before 2

the fission gas feed is opened. In this case the stationary bottom j

concentration is approached first on a straight line

. hen, after opening the bottom product transfer line to the second column, on a H

curved line (dashed line). The critical area is not intersected.

Alternatively, one can also raise the Kr/Xe ratio to or near that of 11 the air by means of an additional Kr feed. In the second Kr separation H

unit at Karlsruhe being in the planning state now for the offgas of 7

1 the WAK plant a Kr recycling loop from the top of the second cplumn to the first column possibly will be installed (see fig. 1)(10. In

',j this case one would be on the safe side in all parts of the column r

(dotted line in fig. 2).

dw We have tested all three versions in the KRETA campaigns and 3

confirmed the expectations: whereas in the first version we had serious malfunctions of the column because of Xe freeze-out, we have 2

'k attained Xe concentrations up to 80 % in the bottom product in the

-5 second version without having detected any precipitation of solid Xe.

3 2

DuringunperturbedoperationofghecolumntheKrdecontamin-

.h (DF) at the head was 2 10 the upper limit set by the ation factor j

sensitivity of the He ionization detector of our gaschromatographic analysis (100 ppb) and the normal Kr feed concentration of 160 ppm by vol. W1 { 4 1.2 Gas-Solid Ecuilibrium 9 More essential than the solid-liquid equilibrium is in our . d' column the gas-solid equilibrium. Due to the constructica of the 'H-column which is a sieve plate column containing 37 plates with the 'i feed point above the 13th plate (overall height: 8 m) and due to 3 the temperature profile within Ehe column (gas temperature at the J feed point: 135 K) we observe desublimation of Xe at the first plate above the feed point (temperature: 95 K) under certain conditions.' J3 There aslid Xe is not dissolved completely by liquid N2, but accumu-

• ?w

.,j 642 9W C i

'15th DOE NUCLEAR AIR CLEANRU CONFERENCE .g-until the cicvc holes c.ro pluggcd. According to tha cublima-latoc: tion pressure curve of Xe (fig. 3) the limiting Xe-concentration in 95 K should be only 80 ppm. In practice we observe.a the feed gas at limiting Xe concentration which is higher by almost an order of magni-We conclude therefore that the difference is either dissolved tude. arried down the column mechanically in solid form. by liquid N2 r The aim of our present work is to increase this concentration by in-creasing the temperature profile in the vicinity of the feed point and by altering the construction of the column interior in that area. What are the consequences of this result for the rare gas separation from a real offgas ? The offgas most pertinent to our development at themomgnt is the offgas of the WAX. This offgas consists of 80 - 120 m /h at S.T.P. of essentially air containing a i / maximum of 1600 ppm of Xe at the peak of a dissolution. If continuous-it should result in serious plugging te ly fed to a column like ours, after a short time. However, a dissolution of fuel in WAK is ecsen-a. tially completed after two hours followed by a day until the next (fig. 4) and dissolution starts. We simulated such a feed gas in KRETA showed that such a rare gas concentration peak does not lead to any even if another rare gas peak simulating a disso-signs of plugging, lution is fed to the column tsa hours af ter the end of the first run. f l$30 TR ri D,= o'?d*L A h 10 '. Kr Xe ~ .L F= 45 Nm}h I- $10' E mood 1 l T l hl W Si 4 13 : w A f Krypton -[ i e0 ppm-. l l 1 1 10 e5 so 13 no s: it.c 0 60 120 180 240 Time (mm) RGURE 3 F8GURE 4 Kr-and Xe-concentmtons in a s:mulated offgos of e Subbmotion pressure of Krypton and Xenon c:ssoluton ct WAK i, l 643 i wi -

15th x_..y p.< s >a, l In the case of a lcrga reprocaccing plcnt wh ro raro gna cin-centrations probably will be higher and on a more continuous level, this concept would have to be changed, however, unless we can' avoid plugging by desubliming Xe by means of modifying the column structure. ~ L:

2. Ozone Formation

,,n ^h The primary reason for the separation of 02 in our concept was

f to prevent excessive ozone formation. However, even during normal operation, not to speak of malfunction of the reduction catalyst traces of 02 will reach the cryogenic column. Ozone formation rates t

9 therefore have to be known. In addition, use of liquid O2 as solvent processalternativeinmanyrespectsgeniccolumnisaninteresting s for the rare gases in the first cryo 3). If one considers this 1 alternative, the safety risks of ozone formation have to be assessed. a .M We have therefore carried out a calculation of the ozone concentra-tj tion to be expected in the first column considering three process j, alternatives (table I): je ji

1) The first case is the alternative presently investigated by us in 3

the KRETA pilot plant; it is assumed that 10 ppm 02 pass the re-duction unit. p [

2) Thesecong)casewasproposedfortheoffgaspurificationsystemat the WAK Il called project AZUR. Here the ratio Kr/Xe is raised p

to that of air by a Kr recycling loop; O again is reduced down to jl' 2 ] 10 ppm. j

3) In the third case, 02 which is present in the offgas in air con-W 1

centration (20 %) is not reduced; it is partially (50 %) liquefied s j thus acting as solvent for the rare gases; Kr is not recycled. I culated by $ and Kr concentration profiles along the column were cal-The O tridiagonalmatrix method (15). Stationary column operation A h has been assumed which is attained 700 hours after startup. 'd The ozone formation rate was calculated according to the ekuationshowninfig. 5, where an average G-factor of 10 was assumed k ( 6), where E is the o elect where the Kr concen-2 takes inko ace ron fraction, 0 i;i Ebd where the radiation dose rate is the product of the tration C unt the abundancy of Kr-85 in overall Kr (6 %) l ! specific activity of Kr-85 and its average B-energy. The dimensions h of the column considered are typical of the KRETA pilot plant y (liquid volume of bottom product: 15 1, and of each of 10 theoretical y plates: 1,2 1). 1 The calculated ozone concentrations in the bottom product are given in table I. Surprisingly the highest value is not obtained in the case of O peration (case 3), but with 10 ppm 0 in the feed a gas (case 1),2whereas in the Kr recycle mode (case 27 it is lowest. 7 }d } The explanation is found in the different transfer rates of the bottom product; whenever this rate is high, the residence t!me of 0 d and Kr f.n the column is.short and thus the ozone formation rate is 2 T low. This is particularly true"for cases 2 and 3. In case 2 the effect of high Kr concentration is even overcompensated. 1 Generally, the calculated ozone concentrations are very low and presumably beyond a critical concentration (17,18). The calculation 4 644 y 3 3.- - - =

~ 15th D6E NUCLEAR AIR CLEANING CONFERENCE ' ~ ~ ?~ i =. 1 2 3 Flowsheet KRETA" Kr-Recyde 0 -Uguef chon 2 Kr 500 WI.-ppm 5 W.-% 500Wl.-ppm Feed Compos. ton Xe 5000 Vol.-ppm 5000Wl.-ppm 5000Wl-ppm IN Nwr M 0 10 WL-ppm 10 Vol-ppm 20 WI.- % 2 Transfer Rote of Bottom Product [t/h) 0.56 3.98 8.t1 D (Cs] l} 2.21. 21.2 0.06 ( 1.10 10 108 10 3,06 10' 5 5 t Ozone-Concetraton[Hd-ppm] i 170 8.2 53 IBottom) Tcble I I Ozone formation in the first cryogerac column n different p.m flowsheets d[0) G feV") Co,(MouMd) CKr[Mol/Mol) dose rote [eVl"s"} 3 dt NL (Mol") df0} 1010[eV ) Co, 0D6 C r 8.5110'fCel") 3.710"[C[ s") 2,t.910'[eV) 3 K dt 6D2 1023[ggaj d 0,. 7.8 10 Eo, C r (M0l'l^ '5 ] ^ j K dt 3 FIGURE 5 Ozone formaton rate n cryog,enic rare gas mixtures l l ( 645 I t I

I 15th DOE NUCLEAR'Al'RFLE N .' ;pf ' r.~ chow 3 that ozona formntion 10 not c eariou3 problem in cryog:nic Kr J s.eparation, since it can be controlled by proper design of the pro ; t' cess. It must be kept in mind, however, that further build-up of the. ozone level may take place in the second column, especially in the Xe f product, if an ozone and oxygen reduction step between the columns is' not considered (4, 17, 19). Table I indicates another inherent safety aspect: the inventory of Kr and thus of the activity. It varies by almost three orders of 3 magnitude and is lowest in the 0 Peration mode and highest in the 2 Kr recycle mode. 3. Points of Krypton Leakage Inherent to the Process The head of the first column is the most important, but not the only point, where Kr is released to the atmosphere. Two additional points are discussed in the following sections. 3.1 Adsorption Unit A conventional adsorber unit is provided in our pilot plant to dry the feed gas and, lateron, to retain all offgas components which, if present, would freeze out in the cryogenic part, i.e. CO2, NOx and NH3. Some Kr is also adsorbed during this step. Although this I coadsorption of Kr is weak at room temperature (20), it would be l sufficient to spoil the good column decontamination factor, if it l would be released to the atmosphere on regeneration of the adsorber. [ Therefore a purging step is included in the operation mode of the ad-j sorber bed, where the coadsorbed Kr is selectively desorbed and re-4 cycled. This concept led to the design of three adsorber lines each g consisting of two beds. The first one is presently filled with 70 kg g of silicagel, the second one with 100 kg of molecular sieve type 10A. Breakthrough (at c/cE = O,5) of Kr (160 ppm by vol.) and Xe j (1600 ppm) in 31 m /h (S.T.P.) at 5 bar and room temperature takes i place after 4.2 and 26.0 minutes, respectively, through the two beds of a line. Desorption curves fog /h Kr labelled with Kr-85, and Xe using a purge gas stream of 5 m ( S.. T. P. ) at 1 bar and room tempe-l rature are shown in fig.

6. The bulk zcount of Kr has been desorbed f

already after 30 minutes, after 2 hours only 1 %o of the Kr original-j ly adsorbed remains on the beds. If one would stop the purging step i at this point, this residual Kr would be released on heating the bed in the regenerating step. It can be shown that a release of 14 %o in-stead of 1 %o residual Kr would still limit the release to the same amountwhichisreleasedatgheheadofthefirstcolumnatacolumn decontamination factor of 10 / 3.2 Xenon Product The bottom product of the first column after transfer to the i second column (packed bed) is separated by distillation into Kr (head i product) and Xe (bottom product). This process step is only sensible, if storage volume of radioactive product can be saved. This is pos-sible only, if the inactive Xe can be purified from Kr-85 to such an extent that it can be released to the atmosphere (or utilized com, mercially). If we postulate again that the amount of Kr-85 released with the Xe product should not exceed the amount released at the head 646

15th DOE RUCLEQL;J m;J CSEANING CONFERENCE 2

o is c

xc. m y rso so i nr xe T= i =1 h3 7 s t } t iso - / isoo 5 t I I v me a. j \\ a, so soc 30 M 9c 12 0

• s0 E

t. H 3" FIGURE 6 oesorpton cf Kr and xe fromtne saco endmsecuce m 2 bees n tro::t plant 13M0 um 5 Nm /h ts et tbar cnd 25'c 5 s. l 3 head of the first column at a DF = 10 then the Kr content of the Xe must be limited to 100 ppm by vol. ~ This postulate could be fulfilled almost routinely. At the moment, the Kr and Xe products are both collected in steel cylinders by means of a membrane compressor. Lateron, transfer of the products ~ into the cylinders by way of cryogenic pumping will be applied. and NO 4. Catalytic Reduction of O2 x 4.1 Poisoning of Noble Metal Catalysts On the basis of laboratory experiments a Ru cagalyst on Al O23 carrier of relatively low specific surface area (9 m./g) has been selected as catalyst for the reduction of O2 and residual NO with H2-x From all nobel metals investigated, this catalyst is most resistant to poisoning by iodine and organic phosphorus compounds which are both trace components of reprocessing offgases. The poisoning effect of 12 was studied with respect to activity.and selectivity on a series of Pt, Pd and Ru catalysts. Some of the results are shown in Fig. 7 and 8. The activity of Pt and Pd is definitely < reduced by. increased loading with I2 As can be expected, the activity of cata-lysts of high specific surface is more strongly affected than of 647 i 2. ._w-

,w-15th DOE NUCLEAR AIR CLEANING CONFERENCE ST? ~ i , ' ~ ~ * :.fs a; e p. h l 'l j L c!Fmm220m2 ist:0 g 96 neo-a% =.-% l l Y w 8 etamm.%tzeti s 9 ,I m, I y-I h,'xunam-%een.i2g m=-% j i ,l E i : f %:f 'a 1 yl/ 6 ,ooo. q \\ a:- -s,.co j / / ( p _ MI-g t? 2.0 IC 20 30 40 1 -Loaarg e cessyst hig) Ir-Loo 884 of Colo'ru8 [*99) I s 2 FIGURE 7 f12VSEl. Ptusaing effect of 1 on Pt-end Pd-cotolysts wrth resoed Poesoring effect of 12 on Ru-cctctysts with respect to 2 to cetety and selectivity of NO-reduction. se!ectmty of NO-reduction Inlet ecncentrction 0.754-%Or, C54-%tO,2,75d-%H. Inlet concentrotens 0.75Vol-% 0, a5W-% NO 2.75W-%H - 2 2 2 Space velocity : 5000hi400 *C space velocsty: 5000h",400*C. f. I 4 ? j~, i types of lower specific surface area. The selectivity which is not good at the beginning (about 40 - 50 % NH formed, the rest N ), is 2 not af fected significantly, however (21),3 On the contrary, the activity of Ru catalysts is not influenced even by very high I2 loadings: under the experimental conditions, NO is reduced throughout to a residual level of = 1 ppm. A poisoning effect can only be observed in the selectivity: formation of NH3 which is very low in the unpoisoned state (22), increases more or less loading (fig. 8). Again, the influence of strongly on increasing I2 f rmation never reaches the specific surface area is obvious. The NH3 ? the amount produced at Pt and Pd. Thus, on the basis of these data, further work will Se carried out with the Ru catalyst of low specific surface area. It can be shown that in a real offgas the time of operation of such a catalyst bed shouldnotbelimitedbelowagoutoneyearbyiodinepoisoning,ifan iodine filter with a DF of 10 or better is installet before the cata-P 1 lyst. 4.2 Form'ation of Methane - It is well known that Ru is an active catalyst for the reduc-j tion of CO to CO and CH (23). CO is an essential component of the 2 4 2 a N~ ; 648 I i e

15th DOE NUCLEAR AIR CLEANING CONFERENCE offgoc if cir ic the ccrrier 900. In addition, CO 2 10 form;d, if anic substances are burned on an oxidation catalyst, which is con-from the fuel elements predominantly as }COe tional practice in air liquefactio

2. CO and CH4, if formed the reduction catalyst, will not be retained at the adsorption bed atand will reach the cryogenic column. According to their boiling points (81.7 K and 111.6 K at 1 bar, respectively) they will attain a certain equilibrium concentration there. For two reasons CH forma-4 tion must be avoided:

- it is a safety hazard in the presence of O ( r0 2 3 - retention of C-14 will be inhibited. The reduction of CO2 by H2 is governed thermodynamically by the equilibria given in table II (23). Whereas CO is formed preferably at higher temperatures (" water gas" reaction), CH4 is formed preferably at lower temperatures ("Fischer-Tropsch" reaction). Tcb'e U : Reductiori of CO by H2 2 I CO2.H2 d CO. H O 2 Kp U I 0.017 at 250 *C E CO.3H ed CH + H O 2 4 2 2.62 at 350 'C Kp = 3.04 at 250 *C 10 CO2.LH2 *=* CH.2H O s 2 0.18 ct 350'C Ig 0.3 ct 250 *C I II The overall equilibrium constant is the product of both Kp - and Kp shifting reaction III to the right with increasing tempegature. The experimental results show that at temperatures above 400 C substan-tial amounts of CO, but little CH4 is formed, whereas at lower tempe-l ratures high yields of CH4 are obtained depending strongly on the H - 2 concentration. Thus one requirement for the minimization of CO -re-l i 2 duction is to keep the overstoichiometric H am unt with respect to 2 649 I IU Z lf - w.. M g

p y;c tha rcduction of 02 cnd NO co low cc pocciblo. %~' x 1 As a matter of fact, the presence of 0., suppresses the'fdrma-l tion of CH drastically, the overall atmosphere still being reducing. g 3 If O is added to a gas mixture containing only CO and H in N, the } form $ tion of CH immediately drops down below the $evel ok detebtion 3 (20 ppm by gaschromatography). Equilibrium III (table II) is shifted to the lef t by a decrease of the H - and an increase of the H 0-con-2 2 's, centration. [ It can be concluded from these results that in a realistic 'offgas-formation of CH4 at the reduction catalyst can be controlled to such e J l an extent that it poses no serious problems from a safety stand 5oint. 1 I Acknowledeements ,f We acknowledge gratefully the assistance of the following co-workers and colleagues carrying out the experiments and calculations: i I W.

Bumiller, G.

Franz, E. HauB, G. Kimmig, G. Knittel and K. Schulz. We also thank E.

Henrich, C.H.

Leichsenring, R.-D. Penzhorn and W. WeinlEnder for stimulating discussious. t i References I 1) Ullmanns Encyklop5 die der techn. Chemie,

3. Aufl.

(W. Foerst, ,g Hrsg.), Band 6, p. 208, Urban und Schwarzenberg, Munchen-Berlin 1955. k.l. 2) R.

v. Ammon and H. Beaujean, in "Chemie der Nuklearen Entsorgung" (F. Baumg5rtner, Hrsg.), Thiemig-Taschenbuch Bd. 66, 1978.

i Y'I 3) C.L. Bendixsen and F.O. German, ICP-1057 (1975), l 4) A. Chesnb, J.P.

Goumondy, P. Miquel and A.

Leseur, CEC-Seminar on, ,j Radioactive Effluents from Nuclear Fudl Reprocessing Plants, f}j. Karlsruhe 1977,

p. 447 5)

R. Glatthaar, Kerntechnik 18, 431 (1976). l 6) T. Kon and S. Motoyama, Techn. Committee on Removal, Storage and l 11 Disposal of Gaseous Radionuclides from Airborne Effluents, IAEA Vienna 1976. 7) J. Bohnenstingl, M. Heidendael, M.

Laser, S. Mastera and E. Merz,I j

IAEA-SM-207/20 (1976). l f 3 8) R.

v. Ammon, W.

WeinlMnder, E.

Hutter, G.

Neffe and C.H. Leichsen-f ring, KfK-Nachrichten 7, 63 (1975), l i 9) L.H. Baetsl6 and J. Broothaerts, CEC-Seminar on Radioactive l l l Effluents from Nuclear Fuel Reprocessing Plants, Karlsruhe 1977, j p. 421 I . f 10) R.

v. Ammon, E.

Hutter, C.H. Leichsenring, G.

Neffe and W. Wein-j 1Hnder, in KfK-2262 (1976), p. 144; Deutsches Atomforum, Reaktor-g tagung 1976, p. 339. 11) R.

v. Ammon, W.

Bumiller, B".

Huttar and G. Neffe, KfK--2570' (197 8 ) p. 242. 1 12) S.

Mastera, J.

Bohnenstingl, M. Laser and E. Merz, Brennstoff- } WHrme-Kraft 29, 214 (1977). 4 1I ESO 4 .a ?" w n m a - - - m,..,.u-....~ . (ggy-y

15th DOE NUCLEAR AIR CLEANIN3 CONFERENCE 13) F* ^"O** *' "* Y* Y *" ( 'E* 9 14a)H. Gutowsky, W. Haas and A. Patzelt, Deutsches Atomforum, Reaktor- ,ho tagung 1978, p. 441- '3 14b)H. Beaujean, U. Tillessen, G. Engelhardt and G. Israel, CEC-seminar on Radioactive Effluents from Nuclear Fuel Reprocessing '~ plants, Karlsruhe 1977,

p. 551, 15)

W. Pfeifer and G. Neffe, unpublished.

h 16)

J.F. Riley, ORNL-3176 (1961), p. 33. 17) C.L. Bendixsen, F.O. German and R.R. Hammer, ICP-1O23 (1973).

18) E. Karwat and G.

Klein, Linde Berichte aus Technik und Wissen-schaft Nr. 4 (1950), p. 3. 19) G.E. Schmauch, ASME 74-WA/NE-2 (1974). 20) 5. Kitani and J.

Takada, J. Nucl. Sc. Techn.

2, 51 (1965). 3 21) R. v.

Ammon, K.

Strauch, W. WeinlMnder and W.

Wurster, KfK-2437 (1977) 22) M. Shelef, Catal. Rev. 11, 1 (1975).

23) Ullmanns EncyklopMdie der techn. Chemie,

4. Aufl., Band 14 p. 329, Verlag Chemie, Weinheim-New York 1977.

n n m 651 M-n= = 1 = _-}}