ORNL-TM-13142.txt

 

S S

s
S 2

ST
ST an
s

o

 

ot

e
SRS
Tansaaidn

LI
aa S

By

e

e
G

ir

T

S

TR RRRARN N
S
GnEnaR
S .

B

2

o

i
S

i

o

i

Tt

o

o
ok

o

S

i

o

S

Tt

S

Sl
Sl
i

o Sy
L
T

et

et

iny
i

SRS o
Shmtas

e
A

o

e

SEadunis

e .
s e
i

i Tl
o Shis
S T
A
e
Sh e

2 = e
e 2

Y

o

R

e

s
ek

%

o
T

e
L

5
R

e

i
b

i
= .:?.n:r:.\;&;fi‘

S

G

3
P
SR
B

ahy
e
e

-

it
S
S

e

&

bl

i
Chi

s

i S s
B SEn

T

TR

e

e 5
S Sl

L il
s
R

Sl
SR

o B
U s
el

R

s
= L
SR
e

S

e
s

e

o

L

G

o

 

N OF THIS DOCUMENT IS UNLIMITED|

i

s
L

ORNL/TM-13142

RECEIVED
APR 0 2 13%
oS

 

A Descriptive Model of the Molten Salt
Reactor Experiment After Shutdown:
Review of FY 1995 Progress

Sy
o
‘;/;??fii; 2

D. F. Williams
G. D. Del Cul
L. M. Toth

2

.
e

 
 

 

This report has been reproduced directly from the best available copy.

Available to DOE and DOE contractors from the Office of Scientific and Techni-

cal Information, P.O. Box 62, Oak Ridge, TN 37831; prices available from (615)
676-8401, FTS 626-8401.

Available to the public from the National Technical Information Service, U.S.
Department of Commerce, 5285 Port Royal Rd., Springfield, VA 22161.

 

 

 

This report was prepared as an account of work sponsored by an agency of
the United States Government. Neither the United States Government nor any
agency thereof, nor any of their employees, makes any warranty, express or
implied, or assumes any legal liability or responsibility for the accuracy, com-
pleteness, or usefulness of any information, apparatus, product, or process dis-
closed, or represents that its use would not infringe privately owned rights.
Reference herein to any specific commercial product, process, or service by
trade name, trademark, manutacturer, or otherwise, does not necessarily consti-
tute or imply its endorsement, recommendation, or favoring by the United States
Government or any agency thereof. The views and opinions of authors
expressed herein do not necessarily state or reflect those of the United States
Government or any agency thereof.

 

 

 
 

DISCLAIMER

Portions of this document may be illegible
in electronic image products. Images are
produced from the best available original
document.

 
 

ORNL/TM-13142

Chemical Technology Division

A DESCRIPTIVE MODEL OF THE MOLTEN SALT REACTOR EXPERIMENT
AFTER SHUTDOWN: REVIEW OF FY 1995 PROGRESS

D. F. Williams
G. D. Del Cul
L. M. Toth

Date Published — January 1996

Prepared by the
OAK RIDGE NATIONAL LABORATORY
Oak Ridge, Tennessee 37831-6285
managed by
LOCKHEED MARTIN ENERGY RESEARCH CORP.
for the
U.S. DEPARTMENT OF ENERGY
under contract DE-AC05-960R22464
CONTENTS

LIST OF TABLES ....ooooneoomeeeeseeeemmesssseemmnssseeessmsssssssasessesssmassesessomssssmmsessssmmessssessesssmmmnns v
LIST OF FIGURES ........ovvveoeessseeemssesssseesssmmeessssssesssessessssssmmesssssesmmenssssmmmssssesssssmsecsssmose vii
EXECUTIVE SUMMARY ..couneeeeevreesesssseeemsessssscessssssssssssssmssssssssssssssstsssnsssssesmanssessees ix
1. INTRODUGCTION ....covvoeeeesreeesssssmesesssssemmssesseessasessesmasssessssssesssemssssssemsessseemmessesmessee 1
2. MSRE FUEL INVENTORY .....coouerereeeeemsssssesessmmsssesssssmmssssssesesssssssmmasssseesansessssmmes 2
3, CHEMICAL INTERACTIONS DURING MELTING OF THE FUEL SALT .......... 15
4. MODELING OF THE ACB «....oovveeeeeseeeeeeeeessssessmmsssssssenmsssssssseasessssssessmssssseasosssssans 17

4.1 RADIATION MODELING ...oovccneveereeseremmssessemmesssseemsssssonmssssssesmasssesmssssesnss 17

4.2 HEAT-TRANSFER MODELING ...cccuu.nccoommmeeseeeemmmesssssmmmessssemmmesssssmmmsssssnsnes 19
5. MSRE SALT RADIOLYSIS ........oveeeeomeeeeeummeasnnassommmsasnessssessssssssssssssssssssmsssssssssssssees 21

5.1 RADIOLYSIS EXPERIMENTS ....cooouneeeeeemmserseseessssssssmmssssssssssmssssssssssssssens 22

5.2 ABSORBED DOSE AND GAS GENERATION ESTIMATES .....c..occumenrceenns 28
6. THERMAL FLUORINATION TESTS .....covuummmenereemmmssssssessmssseemmsmssessssssnsssssssesssssss 34
ACKNOWLEDGMENTS ...coouneeemeaeeeseesssmsseesmmmmessssssssmessssmesssssssmmsssssssasssssommssssssssssses 37
REFERENCES ......cooveeeesesseeseesssssssssemmmmessssessesmesssssssssmmssssssmsmmessssinssmssssssssmssssssssssesseses 39
Appendix A. ORIGEN-S RUN INPUT FILE..........oooouueesseeemmesssesmeesssecessmssseesssssssnaes 41

Appendix B. CALCULATION OF RADIOLYTIC YIELD FROM
IRRADIATION OF MSRE FUEL SALT IN THE HFIR COOLING

POOL.....oovvvmmmnnns e seee st R AR SR eRRR eS8 RS RRRRA S 008545 AR08 EE R 47
Appendix C. ESTIMATION OF CROSS-TRANSFER OF FISSION
PRODUCTS AND PLUTONIUM TO THE FLUSH SALT ..........ccouee.... 53
Appendix D. HEAT-TRANSFER ANALYSIS OF IRRADIATION
SPECIMENS ...covecunmmmaosssssecsssessesssssmmmsssssssssessssssssmensosssssssssmmmmnsosssssssssmonss 57
iii

 
LIST OF TABLES

Table Page
1. Comparison of measured and projected fission product activity at shutdown ........ 5
2. Distribution of major fission product decay energies between salt-seeking

and metallic element CIASSES ........covreivnccrnsacccssesnsserassicsssassssssessasaessssnessessnsnsssessones 5
3. Primary inventory Of Stored MSRE Salts ........ccccorvecsnnessessesenrnaessassesnsrsssnssseasassens 7
4. Secondary inventory of stored MSRE Salts .........cocecriericrensecccresncscseeraesnssesasssesses 7
5. Detailed inventory of stored MSRE salts (1995 basis) .......cccceeuerneercnecraecnseeresecssens 8
6. Inventory of radioactive isotope activity (Ci) and elemental mass (g) ........ccveeu.... 10
7. Results from analysis of MSRE off-gas samples taken in 1994 ..............cccceveue.. 14
8. Estimate of material removed from MSRE salt beds.........cccoecirnmrreccssrennaestonsnnes 14
9. Electrode potentials of major fuel-salt constituents................... weeresetenrsne e s s nrens 15

10. Summary of radiolysis experiments on MSRE fuel salts ........ccccvvceereccncnscaannnns 24

11. Source spectra for MSRE fuel-salt irradiation .......ececevcesveecssncsssacssscrssssssacsnacsaanes 26

12. Drain tank gamma spectrum (2583 kg salt basis, 25 years after discharge).......... 28

13. Distribution of source-term power by radiation category (total fuel and

FIUSD-SAIE DASIS) ...eveeerrrererrnesssassssansssanssssnassssnssssssasssnasssrasesssassssnnsssssasssssssassssasssorsones 30

 
 

 

 
LIST OF FIGURES

Figure Page
1. Primary elements of the MSRE fuel-salt Storage SyStem. .........cceceeverneresrernereessnssens 3
2. Schematic depiction of uranium dEPOSIt ASSAY. ......ccccereererressarsesnssresessassessassaossasess 18
3. Fluorine generation curves for 1986 and 1995 irradiation experiments. ............... 23
4. Comparison of source spectra for MSRE fuel-salt irradiation. .......cccceevceenaennennas 25
5. Distribution of source-term power by radiation category. ............. sresasvensanensasrersase 29
6. Projected fluorine generation in the absence of annealing and induction

EETECLS. oo eieeiecceecnecncesstenancintesnncnssnesneesnnsenssansressnsesnnssnsasnsenssansessanssnnsassnnesnanaesssnse 31
7. Projected accumulation of radiolytic fluorine in the absence of annealing

SINCE 1971 coieciinicnnninissnntnssstsisscesnsissassssssssasssasssasssssassssasssessesssssssansssassssanasnes 31
8. Projected accumulation of radiolytic fluorine in the absence of annealing

between 1994 and PreviOUS YEATS. ..cc.ciicreccsessssersesssesasessasssassssessssssasssssssasssssasaneas 32

vii
 

 
EXECUTIVE SUMMARY

Laboratory experiments, field measurements, and coordinated analysis efforts have
helped the ORNL technical staff gain a better understanding of the status and behavior of
the Molten Salt Reactor Experiment (MSRE) after its shutdown on December 12, 1969.
Laboratory experiments showed that conventional (i.e., “thermal”) fluorination of the
UF, in MSRE fuel salt by molecular fluorine does not occur under static (i.e., nonflow)
conditions at temperatures below 300°C. However, further studies are required to rule
out the possibility of conventional fluorination of the fuel salt at temperatures below the
230°C annealing treatment limit. A separate investigation has identified and quantified
the stoichiometry and thermochemistry of the reactions between F,/UF, mixtures and
activated carbon. This work séeks to explain the chemistry in the auxiliary charcoal bed
(ACB) and is documented in a separate report.

Field measurements at the MSRE have identified material that has evolved from the fuel
salt and now resides in the off-gas system. The following items are of particular

importance:

* Analysis of radiatiori and temperature measurements provide
independent and consistent estimates of ~ 2.6 kg of fuel-salt uranium
deposited in the top of the ACB.

« Off-gas samples drawn just upstream of the ACB indicate that the off-
gas piping and tank plenums contain more than 1.8 kg of uranium and
more than 47 mol of fuel-salt fluorine. Based upon the off-gas analysis
and the ACB uranium assay, it is projected that an additional 68 mol of
fuel-salt fluorine is deposith in the ACB.

 
 

« Therefore, the total inventory of material removed from the fuel salt is

projected to be greater than 4.4 kg of uranium and more than 115 mol of
F,. This represents a removal of more than 12% of the 37.6 kg of fuel-

salt uranium and an addition of 230 equiv of reductant to the remaining
fuel. Under these net reducing conditions, significant amounts of
uranium metal can form during melting of the fuel if the salt redox
chemistry is not adjusted.

Revised source-term and radiation-transport calculations were conducted and support
improved estimates of the decay energy deposited in the fuel salt and the generation and
accumulation of fluorine by radiolysis. Based upon a one-dimensional transport
calculation, more than 88% of the gamma decay energy is deposited in the fuel salt. The
remaining 12% that escapes corresponds to an exposure at the inner tank wall of about
600 R/h. The upper bound on the yield for salt radiolysis indicates that less than 650 mol
of F, has accumulated since the cooling of the salt in 1971. Best-estimate yield values
put the figure at 300 mol of radiolytic fluorine. Projections also show that the current
measure of liberated fluorine (115 mol) could not have been generated recently.
Accofding to these estimates, generation of fluorine must have occurred prior to 1989,

and probably started much earlier than this.

 
 

1. INTRODUCTION

During FY 1995 considerable progress was made toward gaining a better understanding
of the chemistry and transport processes that continue to govern the behavior of the
Molten Salt Reactor Experiment (MSRE). As measurements in the MSRE proceed,
laboratory studies continue, and better analyses are available, our understanding of the
state of the MSRE and the best path toward remediation improves. Because of the
immediate concern about the deposit in the auxiliary charcoal bed (ACB), laboratory
studies in the past year focused on carbon-fluorine chemistry. This work is documented
in a separate report []. Secondary efforts were directed toward investigation of gas

generation from MSRE salts by both radiolytic and nonradiolytic pathways.

In addition to the laboratory studies, field measurements at the MSRE provided the basis
for estimating the inventory of uranium and fluorine in the ACB. Analysis of both
temperature and radiation measurements providéd independent and consistent estimates
of about 2.6 kg of uranium deposited in the top of the ACB. Further analysis efforts
included a refinement in the estimates of the fuel-salt source term, the deposited decay

energy, and the projected rate of radiolytic gas generation.

This report also provides the background material necessary to explain new developments
and to review areas of particular interest. The detailed history of the MSRE is
extensively documented and is cited where appropriate. This work is also intended to

update and complement the more recent MSRE assessment reports [2-4].
 

2. MSRE FUEL INVENTORY

The inventory of the stored MSRE fuel by element, isotope, and location is the starting
point for most analyses, and a number of studies /2—6] have reported inventory values.
There are two important reasons to revisit this subject: (a) the recently discovered

transport of material within the MSRE has not been accounted for in these reports, and
(b) previous reports contain inconsistencies that need to be reconciled. The goal of this

section is to report an inventory based upon the best and most current estimates.

After MSRE reactor operations ended on December 12, 1969, the entire fuel-loop

inventory was emptied into the two fuel-salt drain tanks in the drain tank cell, as shown in

Fig. 1. Flush salt was then circulated through the fuel loop to remove any heel or

deposits and then returned to the flush-salt drain tank in the drain tank cell. The salts

solidified upon cooling below 434°C and were maintained between 230 and 340°C for

1 year before being allowed to cool to ambient conditions in 1971/7-9]. With the

exception of the heel of flush salt left in the fuel loop and the heels of fluorinated salt in -
the fuel storage tank and salt still, these three tanks in the drain tank cell contain virtually

all of ihe radioactive fuel salt [6, 10-11]).

Consideration of the inventory after shutdown (i.e., “discharge” inventory) is the natural
starting point. Adjustments are made to this baseline to account for the decay and
transport of species after shutdown. Inconsistencies in the reported inventory derive from
differing assumptions, different bases, and the inherent uncertainty in measurements.
Even though most of the discrepancies are rather minor, it is important to adopt a logical
basis for resolving these differences. Estimation of the discharge inventory is based upon

a variety of measured parameters: (a) the isotopic distribution of uranium and plutonium

in the fuel salt, (b) the fission product loading of the fuel salt, and (c) the weight of salt in

 

 
"WId)ISAS ades03s Jres-[any FYSIA Y3 jo spudwdpd Arewnrg ‘131

ANYL 2=dNVL L =ANVL
HSN14 13nd  Nivdqg 13nd  NIvyag 13nd

 

    

 

n "iviol
6% 148

1vs 0314101108

1130 Q38 TVOJIVHIY

 

 

 

  

(1 606) (0 z92)

 

        

   

ANI3H
30vds SVO

   

 

(0 961) SLS £LS
dNA1oH

JNNTOA

    

        

INIOd
J1dNVS

        

         

   

           

        

         
   

       
    

(1 90¢)
o | 48
MIVLIS IN3IA TVOIUVYHI Sdvil
Jd34OLINOW AYVIIXNY 30viyL TIILEVd Zvs
ONv 038 WOIYVH) JviL
vd3iH Ol e
££S aNY 393nd TVOOUVHI QILVAILDV FIEES)
ANINg 0l

°H EEE)
SLIvsS qitos EEER

LS~-VS6 9MG INYO

 
 

the reactor loop and drain tanks (both at discharge and during operation). Each of these
inventory clements has a different level of certainty. Probably the most accurate
measurement is that of the uranium and plutonium isotopic concentrations. Uranium and
plutonium isotopic concentrations were carefully measured throughout the MSRE
operating cycle and provided the most sensitive and accurate determination of power
output, burnup, and total uranium and plutonium masses [11]. It was not possible to
measure the remaining activation and fission products so completely and accurately. The
entire fission product inventory can be estimated only by modeling the generation and

decay of isotopes.

The primary objective of fisSion product measurements was to aid in modeling the
transport behavior of the elements in the molten salt [12]. A fairly complete picture now
exists for the partition of fission products between the salt and the surroundings: (a) the
first four periodic groups (IA, IIA, ITIA, and IVA) and the rare earths are salt-seeking
elements and remain homogeneously distributed in the fuel salt; (b) the noble gas fission
products are removed to the off-gas; and (c) the noble metals class (Nb, Mo, Tc, Ru, Rh, ‘
Pd, Ag, Sb, and Te/I) dissolves 'in the salt to a minor extent, and probably exists as a
separate phase that deposits on surfaces. The good agreement between the final MSRE
fission product measurements for salt seekers and the projected inventory as calculated by
Bell /5] is shown in Table 1. A correction factor of ~10%, due to differences in the basis
for calculation of measured and projected activities, brings the values in Table 1 into
agreement within the limits of experimental precision. It is impossible to know the fate
of the noble metal isotopes, but it is certainly reasonable to assume that most of them
were flushed into either the drain tanks or the flush tank. The noble metal fission

products are relatively short-lived and comprise a significant fraction of the decay energy

only during the first few years after shutdown, as shown in Table 2.

 
Table 1. Comparison of measured and projected fission product

 

 

activity at shutdown
Measured
. inventory  Ratio of measured to Ratio of measured to

Isotope Half-life (Ciya projected activity?  projected activity¢

Salt-seeking elements
Sr-89 51d 93,900 0.58 0.532
Y-91 58.5d 166,200 0.91 1.017
Zr-95 64d 149,700 0.75 0.848
Cs-137 30y 9,520 0.85 0.793
Ce-144 285d 118,200 0.93 1.058

‘Metallic elements

Nb-95 35d 8,540 0.05 0.054
Ru-103 39d 6,860 0.09 0.11
Ru-106 1.02y 568 0.08 0.051
Te-129m 34d 2,920 0.11 0.278

 

4 Based upon 12-5-69 sample reported in ORNL-4865 (complete citation in note “c™) and a circulating
loop inventory of 4350 kg.

bSource: Bell, M. 1., Calculated Radioactivity of the MSRE Fuel Salt, ORNL/TM-2970, Oak Ridge
National Laboratory, May 1970.

CSource: Compere, E. L., et al., Fission Product Behavior in the Molten Salt Reactor Experiment,
ORNL-4865, Oak Ridge National Laboratory, October 1975.

Table 2. Distribution of major fission product decay energies between lsalt-seeking
and metallic element classes |

Fission product decay energy (W)

 

1 year after 5 years after 25 years after
Half-life shutdown shutdown shutdown

Salt-seeking elements

Sr-89 50.6d 7.2

Sr/Y-90 285y 21.3 82.5 50.0

Y-91 58.54d 8.7

Zr-95 64d 194 -

Cs/Ba-137 Ny 52.7 48.3 30.0

Ce/Pr-144 285d 417.8 11.9

Pm-147 262y 10.5 3.7 —
Subtotal 607.6 146.4 80.0

Metallic elements |

Nb-95 35d 394

Ru/Rh-106 1.02y _36.2 24 .
Subtotal 75.6 24 0.0
Total 683.2 148.8 : 80.0

 
 

Perhaps the least accurate element in establishing the salt inventory is simply the total salt
weight. The d;ain tank load cells that were originally intended to provide accurate salt
weights were found to be too inaccurate for independent determinations [6]. To obtain a
value for the inventory of the fuel and the flush-salt weights, considerable material
balance work is required. Accounting for the numerous additions, withdrawals, and
flushes of the fuel loop—-in addition to the effect of the heels remaining in the drain
tanks—requires good judgment and extensive prOéess knowledge. The best values
available are bounding estimates provided by the MSRE staff members who are most
knowledgeable about the history of operation [6, 11]. These values (Tables 3 and 4), in
conjunction with the measured uranium and plutonium isotopics [17] and the fission
product/activation projections of Bell /5], form the best basis for establishing a discharge
inventory. The inventory of major salt constituents calculated on this basis is

summarized in Table 5.

The major uncertainty in Table 5 is the distribution of plutonium and fission products
between the fuel and flush salts. Measurements of uranium concentration in the flush salt
cannot be used to directly infer fission product or plutonium concentrations, because of
the removal of uranium from the flush salt after the initial phase of operation with 235U.
However, the steady increase of uranium measured in the flush salt after each circulation
in the flow loop did establish that ~20 kg of fuel salt was transferred to the flush salt
during each flush operation {11]. The present inventory of uranium in the flush salt
(~1.3% of the total) represents the cross-transfer from two flush operations conducted
during the final phase of operations with 23U. The fission product and plutonium cross-
transfers also had contributions from the seven flushes during 225U operation. In contrast
to the relatively constant amount of uranium transferred per flush operation, the
magnitude of the fission product/plutonium cross-transfers grew from near zero to the
maximum value associated with 20 kg of spent fuel salt. It is assumed that the fuel-salt

fission product and plutonium inventories grew in direct proportion to burnup during

6

 

_—— i
Table 3. Primary inventory of stored MSRE salts

 

Maximum weight ~ Minimum weight?  Salt density

 

Major components?® (kg) (kg) (g/mL at 26°C)
Fuel salt 2.48
Fuel Drain Tank -1 2583 2479
Fuel Drain Tank -2 2263 2171
Subtotal 4846 4650
Flush salt 2.22
Fuel Flush Tank 4274 4265

 

@Sources: MSRE Fuel and Flush Salt Storage, Request for Nuclear Safety Review and Approval, NSR
0039WMO00013A (approved 12/28/93; expires 12/31/95); Thoma, R. E., Chemical Aspects of MSRE
Operations, ORNL-4658, Oak Ridge National Laboratory, December 1971, pp. 5865, 99-112.

b These minimum weights are most consistent with the process history.

/

Table 4. Secondary inventory of stored MSRE salts

 

 

Minor components3 | Fue:l-s(ailctg ;;vcight Flush-s(all(l;)weight
Fuel storage tank 1750
Distillation experiment 30b
Reactor flow-loop heel : 20 ]
Drain tank cell piping ' 12b.c
Processing cell piping gb.c
Release to drain tank cell 0.1

 

 

ASources: MSRE Fuel and Flush Salt Storage, Request for Nuclear Safety Review and Approval, NSR
0039WMO0C13A (approved 12/28/93; expires 12/31/95); F. . Peretz, ORNL, personal communication,
September 6, 1995. - ' N

bThese salts have been fluorinated and have low uranium concentration (<100 ppm).
CThis value also includes the contribution of unspecified flush or fresh salt to the inventory.

 
Table 5. Detailed mventory of stored MSRE salts (1995 basis)?

 

 

Total weight
Fuel salt Flush salt (kg)
Bulk composition mol % (wt %)
LiF 64.5 (42.6) 65.9 (51.3)
BeF, 30.4 (35.8) 33.9 (47.8)
ZiF, 4.9 (20.5) 0.18 (0.89)
Major elements
U, kg 37.1 0.5 37.6
Pu, %P, 98.2 1.8 0.737
Fission products, %% 98.3 1.7 2.71
Rare earths 1.47
IA, TIA 0.275
Zr 0.626
Other metals /1 0.334
Fissile element isotopes, wt % ¢
B2y 160 ppm? 75 ppm®
23y 83.92 39.4
B4y 7.48 3.6
235y 2.56 17.4
6y 0.104 0.245
238y5 5.94 394
239py 90.1 94.7
240py 9.52 4.8
other Pu 035 0.50

 

@Source: Thoma, R. E., Chemical Aspects of MSRE Operations, ORNL-4658, Oak Ridge National
Laboratory, December 1971 pp. 58-65, 99-112 .

bDistributions based upon estimates in Appendix C.
CFlush salt values are the average of two analyses.

4 stimate obtained from Bell, M. J., Calculated Radioactivity of the MSRE Fuel Salt, ORNUI'M-2970
Oak Ridge National Laboratory, May 1970.

€Flush salt 232U 233U ratio assumed to be that of the fuel salt.

 

 
35U operations. During 23U operations, breeding of plutonium was negligible, and the

change in plutonium concentration in the fuel salt was dominated by depletion due to
fission/transmutation and replenishment by PuF, refueling operations. Appendix C
provides the details that support the transfer of ~2% of the fission products and plutonium
to the flush salt.

The previous projections of the MSRE spent-fuel activity had either a very short focus
(< 5 years) or were concerned with projections far into the future—the intermediate time
period between S and 100 years has not received detailed attention. Because of this gap
in the literature, additional ORIGEN-S runs were performed at complementary time
intervals /13]. The discharge inventory and decay calculations are summarized in

Table 6. The ORIGEN-S input file is included in Appendix A.

The final inventory item that must be considered is the transport of material out of the salt
beds. Except for the generation of fluorine by radiolysis of MSRE salt, no other
mechanism for producing mobile species was known before 1994. Annual reheats of the
drain tanks were intended to recombine the radiolytic fluorine before it was released from
the salt, thereby preserving the salt chemistry and eliminating any substantial release of
F,. A completely new and unexpected pathway for volatilizing MSRE constituents was
discovered during sampling of the off-gas syst;:m upstream of the ACB in 1994. Off-gas
samples (Table 7) indicated the presence of a considerable volume of F, and UF, in
addition to small amounts of HF , MoFg, and CF,. The presence of such a large amount
of F, and UF in the off-gas suggested that the ACB be inspected as a possible sink for
these reactive gases. Radiation and temperature measurements confirmed that a
significant quantity of uranium was deposited in the upper portion of the charcoal bed.
Careful analysis of this data led to an estimate of 2.6 kg of uranium immobilized on the

carbon (Sect. 4). The quantity of fluorine held in the ACB was inferred from the F,/UF,

mole ratio in the off-gas sample (Table 7).

9

 
 

 

€1 00L°97 PPE  weTl
1'6 008°L6 Yor'| 6Z1
L€ 98¢ 0v6'E P60T  wLTI
9'€ 8LE 00L'TE yp'6 Lzt
€0 01 Szl vy 8'EL £21 L8T P8S gzl
@ 1) @1¢) @i z 3L T8
€0 01 oy 1S 81 70¢ 208 L¥9 AeLe czTl
3z90) 3 qs I
3 L0£) @ L0g) 3 ¢80 z pd 9
T8 9 196 osL'E 0£6's § OF 901
LIT 006'€L wog  ugol
(3 g9) (3 ¢5) (8 L05) T 't S
T8 9T 196 0SL'E ow'L £70'1 901
| LT1 000'¥YL PT6E £01
@ ¢p) (B ¢p) @iy 3 ny ¥
S0 S0 S0 S0 S0 $0 $0 S0 S0 £¢ave 66
(3862) (8860) (3860 3 9], €
@1€n @10 @L11) z o Ty
0'S oSty P9E  wWG6
I'E ove's . 000°€Ll PSE 6
@) 1 qN 1§72
¥'1 0£8°C 000°00Z PP9 6
£0 €0 £0 €0 £0 €0 €0 €0 €0 £93 ST €6
(3199) (3 919) @zL9) 3 17 oy
. _ ¥'0 ozv'e 000°€81 PS8BS 16
061°% 0,99 0SS'L 0£S's 00601  00EZT  O000'El  009°€l 009'€1 PLT 06
@z @z @1y z X 6€
061'l 0499 0sS'L 0£5'8 00601  O00E'ZT  O000'El  009'¢l 000'¥1 £¢87 06
080'1 000791 P9°0S 68
@ 11) (3 1°89) (3or1) 3 1§ 8¢
3 8p'8) (B gy's) (381'8) 3 Qy LE
S1IMpoLg U0ISSY,]
Koot Koe Ast Aot Aot Ag  A¢ K1 odmysiq oJj[eH SSeN Wowd[g  Z

 

 

(8) ssew1 [ejUdWIAA pue (1)) A)Ande Ad0)0s] IAIJIROIPR. JO £I0JUIAU] "9 QR

(Saumd a1e syun JAYIo [[e ‘sasoyuared up SSeUI [RIUA]T)

10

 

 
 

019'¢ 00¥'¥C  00S°LT  001'1€  O00I'TY  00I'8S  O009'€8  000'L0T  000°008°1
: - (12) &ntanov yompoad uoyssty
@11Lo @r1Le) “ @1
(8) ssow sonpoud vosssy g
(2 06°0) (8 c98°0) B0 I PO 9
'y €6 61 $'s8 6L1 e ¥Z€ 9LE L96¥ gCT
I'g Ly oL LSt S'€ Lt $'TE Tse £33 73
't Sl 61 €€ [A 4 Ly TS ¢S Lggl sl
@B st'g) B 6t'9) (3Lg9) 3 ng £9
89 LT1 1z 9zl 9¢1 1821 ¥rl 9l Lyl £06 ISt
(3 s01) @orn) @z I ug 29
4 0S0'T Py  wsyl
1'0 0 PHS 8vl
Vel €0S 681 089'C 0£6'6 008'91  009'8T  00T'LE £297 Lyl
@500 @zon 3 wy 19
(89LS) @9L5) (3 9¢¢) X PN 09
o 602 €21 6TL 0 weL  whpl
SLl 064'1 018's 001°ZS 000'821 weLl 1
(B 181) (@y81) @oLn) 3 14 6S
¢Ll 06+'1 018'8 001'ZS 000'LT1 PS8z ¥l
oLl 000°01+ PSZE b1
(8 79¢) (8 79¢) (3 oth) 3 D 8¢
@oLD) @oL1) @oLl) X e LS
050'1 062'S ov6's 099'9 06€'8 0z’ 0L8'6 00€°01 005°01 w9y Wil
(Bo61) @ Lc1) (3 08) T ey 96
ot1'l 009'S 0679 090'L 068'8 0866 00401  006°01 00T'11 Lo LET
z0 01 07 6€ ¢S L90T el
@sgzn @y @e621) 3 5D S
@z91) @z9o1) @6y 3 I £S
K001 Aog Agz Loz Lo1 K¢ Ag A1 3Breyosiq IJBH SSEN WoUR[d  Z

 

(SaLmd 2re SIUN JIYIO [[8 ‘sasarpuared Ul sseW [RIUSWIS[Y)

(PonuyL0) °9 Qe

 

11

 
 

 

b8Y 1€6 6L6 9Z0'1 6Z1°1 91°1 SST'T SIT'T  690'1
(10) £nnanov kvoap-apmnsy |
@ze6rn @ 6v's) (3 L81°0)
(8) ssowu Kpaap-aprun1 oy

L't 80 L0 <0 €0 1'0 0 £00EL 622

99 ZEX 6€1 opl 191 991 o) 651 €51 £61 82T
@ L1) (3 06'€) _ (3 L81°0) T 4L 06
LT 8'0 L'0 S0 £0 1’0 0 POl &6 v 68
L' 80 L0 <0 €0 10 0 P8I 7T

99 €l 6€1 opl 191 991 91 . 091 €St P99t T ey 88
L' 8'0 Lo S0 €0 10 0 w6y 1 ! 8

99 €l 6€1 9p1 191 991 91 Wr €SI $9'CS e W 98
L' 80 Lo S0 €0 10 0 sw zg e w £8

99 €1 6€1 9p1 191 991 91 091 €51 sw S| 917
LT 80 Lo S0 €0 1'0 0 sy €T
€Ty 818 1'68 9't6 €01 901 901 01 §'L6 $ Cp e od 8
LT 8'0 L0 S0 £0 1'0 _ 0 w9'GH €12

99 Z€l 6£1 op1 191 991 91 091 st y10T1 e g £€8

99 €l 6€1 opl 191 991 91 091 €ST Y901 e
LT 80 L0 <0 €0 1'0 0 ysze 60T
B ¢Sy) @65T) 3 qd 8
L'€T oLy 0s ¢'ZS 8'LS 9'65 €65 p'LS g wgoE 807 1L 18

S4a1y43nop KD22p-Iprunoy
K001 Aot Agz Loz L01 k¢ Ag A1  oBreyosiq SJ-JleH SSBN WeuR[  Z

 

- (SaLmDd 3w SIUN I3YI0 [f8 ‘sosoypuared Ul SSBW [RIUSWIT)

(Panuyucd) ‘9 3qe,

12

 
*K10}udAUI AZIBYISIP §,J]od WOL] PIALIP
ore suondofod 1910 NIV "0L6T AN *K1ojeioqe [euonieN 98pR Y80 ‘0L6Z-NL/INIO /10§ 1214 TUSW 241 fo nanovoppy paiomaip) uy pauodar se
T1°d [ " £q pareno[ed 95043 Uey) IOMO] % S—¢ aIe pue sisAeue JId0I0ST WOI} PIALIdp e 5« Z 1doox9) sonpea L10judAur umiuoinid pue wnueln 4

4

 

SL 187 $08 £88 SO1'T 092'1 €EC'T 6I%'T 08t'1
(12) Kn3ano0 NY I 1010
(8 0g£'8€) (3 0vE'8e) (8 ove'se)
(3) ssow Y 1010
0 TS ST PEIT T w) 96
e T S'1Z ol A ¥iL 0's v'T 960 Legy 1T wy S6
17L TIT oLT prE 85¢ 0lL 8L 798 $06 Aypl 14T
Y | €SI €Sl €<l €Sl €Sl £l €51 €Sl Kovs'o Ot
L'y Ly Ly L'ty L'ty L'y Uiy L'y Ly LOTI'YT 65T
IS0 68°0 760 960 10'1 80'1 or't 60’1 00’1 L1188 86T
(3zeL) @ LeL) | (3 evL) | 3 nd 6
KeALry 8T
KLAVET 9€T
A8AY0L  SET
Ll LI Ll LI A Ll L 'L ¥iLl Aeasye  vET
20€ 20€ 70€ 20¢ 20€ 20¢€ 20€ 70¢ T0¢ A¢H6ST  €€T
9 621 SET Al LST SOl 891 TLl €LI LoL  Z€T
(8 009°LE) (8009'LE) _ (3 009°LE) 3 n 26
p$2d0j1osy umpp.msupa |

 

Aoor Log Asz Kot Lor LS K¢ A1 oBrmyosiq op-JieH SSE oWl  Z
p(SOLIND Are SIUN JIYIO 18 (SISAYIUATRd UI SSBUI [RIUIW[H)
(panunuod) 9 3qe[,

13
 

Table 7. Results from analysis of MSRE off-gas samples taken in 1994°

 

 

Partial Pressure

Component (mm Hg)

F, 350

Inerts 305

UF 692

Mok, 10

CF, 5

HF 0.74 (1000 ppm)

N-F compounds Trace

 

4Source: Toth, L. M. ORNL, personal communication, Feb. 14, 1995.

bsaturation pressure of UF at the sample temperature of 21°C is 79 mm Hg.
The off-gas assay and charcoal bed analysis provide a basis for estimating the amount of
- material that has migrated out of the MSRE salts. Because the source of these volatile products
was far upstream of the sample point and at a lower temperature, it is possible that the amount
of UF, and F, in the off-gas piping and tank plenums is greater than that predicted on the basis
of a homogeneous vapor space with no deposition of material by reaction, condensation, or
sorption. The most unbiased approach is to proceed with the limiting case of a homogcneous
atmosphere, as shown in Table 8. These assumptions lead to a projection that more than 4.4 kg

of uranium and 115 mol of fluqrine have been removed from the fuel salt.

Table 8. Estimate of material removed from MSRE salt beds?

 

 

 

Uranium Fuel-salt fluorine
kg mol (mol F,)?
Off-gas volume® > 1.8 >7.7 > 46.7
ACB deposit? .26 11.2 > 68.0
Total removed > 4.4 >189 > 115
Remaining f,‘glltg;‘;} <332 <1425 > 115 deficient

 

 

9Basis: 2029-L off-gas volume, 20°C average temperature, 738 mm-Hg-pressure.
bincludes removal of fluorine as UE,oF,.

€ Assumes off-gas is at 1994 sample conditions shown in Table 7.
d Assumes 5.07:1 F,/UF ¢ ratio of Table 7 applies to the ACB deposit.

14

 
3. CHEMICAL INTERACTIONS DURING MELTING OF THE FUEL SALT

The safe and effective removal of salt from the tanks must account for the chemical condition
of the stored salt, and a preliminary discussion of this issue is needed to summarize our
present understanding and to plan future work. Except for the radiolytically driven reactions
described in Sect. 5, the stored salts are believed to exist as an otherwise stable one-phase
solid. However, the salt is also in a net reducing condition because of the more than 115 mol
of fluorine that was generated by radiolysis and removed from the solid. This represents a
net 230 equiv of reductant present in the form of isolated metal sites (Li° and Be®). The
maintenance of the salt, when molten, in such a highly reduced state was one of the chief
concerns of the original MSRE staff because of the likelihood of catastrophic phase
segregation under these conditions. The redox chemistry was closely monitored during
operation of the MSRE to ensure that highly reducing conditions did not develop. The
present reducing potential of the stored salt is latent in the solid form, but once the salt is
melted the reducing potential of these sites can be realized, and the metal species will react
according to their redox potentials, Li > Be > U ~ Zr, as shown in Table 9.

Table 9. Electrode potentials of major fuel-salt constituents?
E®° = reduction potential (V)

 

 

 

Half-cell reaction 450°C 725°C
Li* + e — Li(s) -2.770 -2.559
Be?t + 2¢- — Be(s) . -1.958 -2.460
U + 3¢ —» U(s) -1.606 -1.433
U* + 4 - Uy ~1.522 -1.336
Zi*t + de= —  Zx(s) -1.542 -1.335
U + e —» UM ~1.268 -1.045

 

@Potentials referenced to HF/H,, F~ in 0.67LiF-0.33BeF,, Source: Baes, C.F., “The Chemistry and
Thermodynamics of Molten Salt Reactor Fuels,” Nucl.Metal .15, 617-44 (1969).

15
 

The following reactions are a consequence of this reduction series:

2Li° + BeF, — 2LiF + Be° 1)
Li° + UF, — UF; + LiF (2)
Be® + 2UF, — 2UF; + BeF, 3
4UF; o 3UF, + U° 4)
3Be® + 2UF; — 2U° + 3BeF, ©)
2Be® + ZiF, — Zr° + 2BeF, ()]

It is expected that the reduction of beryllium by lithium is kinetically favored and that the
cascade of subsequent reduction steps eventually converts all of the uranium to U3+ and some
fraction of the uranium and zirconium to the metallic state. Using the lower bound of

230 equiv of reductant, R, and the estimate of 142 mol of uranium in the fuel salt, a projection

of the chemistry of uranium in the molten salt begins with the stoichiometry of the initial

reduction step:
R + UF, - UF; + RF (7)
[142 142 142 142]

The close proximity of uranium and zirconium on the redox scale makes it difficult to predict
the subsequent reduction to these metals, and the possibility of alloy formation between the

two further complicates the picture due to the lowering of the U activity. In the event that the
excess of R (88 equiv) reduces uranium preferentially, a mass of 6.8 kg of uranium metal will
be formed by reactions (4-5). Even if uranium metal is not formed or alloyed with zirconium,
the solubility of UF, is limited in the MSRE sélt and a considerable portion of the uranium may
precipitate upon melting [14]. These results su ggcst that adjustment of the redox chemistry of

the salt prior to or during melting (e.g., fluorination or hydrofluorination) will be required.

16

 
4. MODELING OF THE ACB

The basic approach of assaying the uranium deposit by measuring the temperature and
radiation fields surrounding the charcoal bed is depicted in Fig. 2. It is possible to
estimate the amount of uranium present based on its action as a local heat source and the
extended radiation field it produces. In this particular case, the radiation-modeling
calculations require a good approximation of the actual source geometry—accurate
single-point spectroscopic measurements can be used with confidence only after the
extent of the source has been defined. Even though the heat-transfer calculations are
relatively insensitive to the source geometry, it is likely that the radiation measurements

will yield a more accurate estimate because of their greater accuracy and specificity.
4.1 RADIATION MODELING

The estimate of the amount of uranium deposited in the charcoal bed is based upon the
following [15]:
(a) aknown isotopic concentration of 22U in secular equilibrium with
its daughters;

(b) a mapping of the radiation dose in the charcoal bed cell by
thermoluminescent dosimetry (TLD), followed by analysis to infer a
source geometry; and |

(c) measurement of the 2.6-MeV gamma-ray intensity from the 208T1
daughter, followed by shielding analysis to convert this to a source
activity.

The 22U content of the deposit is based upon the measurement of uranium activity in an
alpha-monitor filter sample from the MSRE vent house. The measured value of 135 ppm
232(J compares with the projected value of 160 ppm 232U (Table 5) . Secular equilibrium
between 232U and its daughters is inferred from the stable radiation field surrounding the

charcoal bed.

17
 

(4/wes) ¥3i¥M NI 4 10 31vY 3504

 

 

 

 

DZ 8L 9 ¥ ¥ O 9 9 ¥ T O
T T T T T T T T T T %08
- 191 1Nds ~ s6
FUNLVYIANIL TIYMIQIS e
= QIYVYINI rl o6
FUNIVYIJNIL TIVAIAIS = = = =
| NOILVIOYY = = mem \. 58
— \-J 08
_ J s
. L
‘\ oL g
9
-
~ ~ I 9z
[ J -y
=3
-~ |
= » gS
\ 7,
- \ \ - os
/
~
[ - of
i Lt 11 oz
0L 69 89 (9 99 €3 ¥y f£9 T9 19 09

(4.) 3uNLlvy3dN3L

499Q-vrE 9MC INHO

*Kesse J1sodap winjuean jo uopdidap d>1BWIYIS T 314

 

Try———

. v aw N
Saonle AT
3 L .

 

 

 

’
-
v
¥
Wt

+
.
-

   

 

I‘l-flN- 3

Q34 i [
IYOIUVYHO
ANYEIXNY

 

 

 

 

 

 

 

 

 

 

v6/81/41)

r6/r/14

¥6/02/¢L

T

. AOOTONOYHD
T3ATT H3IYM

 

 

 

(19sA WAl

 

30 NOWYD0Y)

ue

INIVA

3ISNOH_INIA
SN

 

 

annod
sy

 

 

 

s

B~

 

 

 

.
e

 

882
922
Tuu
Tm«
ore
82z
912
jroz
1z61
logs
{891
191
trr
Yzs)
ozt
1801
C 4 86
e
7

BY
9%
14
Zl

 

 

 

X 7
o~8¥8 13 ML ‘v 8L
3076 THIIHS

 

Pt—

 

 

 

 

 

 

 

T....:.. "3 oL

SNOILYHIING
Jejeuinip W

Cu)
IvIS
H1d3a

18

 

 
The mapping of the radiation field within the charcoal bed cell was performed at radial
distances of 8.3 and 15.4 in. from the bed centerline and at axial positions ranging from
24 to 108 in. below the top of the shield block. Both point-kernel (MARMER) and
MonteCarlo (MORSE) codes gave consistent predicted dose-rate profiles for the assumed
source dist:ibutions. Note that it is not possible to deconvolute the measured dose profile
to identify a unique source distribution; instead, some judgment must be used to constrain
the choices for the form of the source distribution; At present, it is most reasonable to
assume a uniform cylindrical deposit of uranium, even though the measured profile

(Fig. 2) is not exactly symmetrical (as expected for a uniform source). The best fit far a
uniform cylindrical source was for a deposit extending 12 in. below the top of the

charcoal bed.

Careful measurement of the 2.6-MeV gamha—ray intensity emerging from the empty
shield plug atop the charcoal bed cell was coupled with the appropriate shielding
parameters in the program MICROSHIELD to provide the estimate of 7.74 Ci of 22U in

the source. This corresponds to a total uranium mass of 2.6 kg in the deposit.
4.2 HEAT-TRANSFER MODELING

Two separate types of heat-transfer calculations were performed in order to estimate

(a) the strength of the heat source contained within the charcoal bed and (b) the centerline
temperature in the bed, given the source distribution assumed in Sect. 4.1 (12-in.-long
uniform source)/16]. The heat source-strength is most readily estimated by surnming the
convective and radiative heat flux over the outer boundary surface enclosing the bed
(6-in. schedule 10 stainless steel pipe). Both of these fluxes are functions of the
experimentally measured wall and boundary temperatures (Fig. 2). Integration of these

fluxes over the surface of the pipe—using the McAdams correlation [17] for natural

19
 

 

convection to air and a pipe emissivity of 0.7 —results in a heat source of 2.36 W. The
locally deposited energy (i.c., “thermal power™) for the same isotopic mix as considered
in the previous section is 0.932 W/kg U and is consistent with a deposit of 2.5 kg of

uranium—almost the same value as derived in the previous section.

The temperature within the charcoal bed is a concern because of the potential for further
reaction of the carbon-fluorine compounds formed by the reaction of UF, and F, with
activated carbon. Under certain conditions it is possible to initiate exothermic
decomposition of these C;F compounds by heating them to a temperature above that at
which they were formed [18].

The projection of the maximum temperature (i.e., centerline temperature) within the bed is
a more complex task than estimating the overall source strength. Because of the system
geometry and nonlinear boundary conditions, a numerical solution of the governing
differential equations is required. An added complication is that the most important heat-
transfer parameter, the effective bed thermal conductivity, &, _; cannot be éstimated with
any real precision //9]. Because of this uncertainty the solution involved parameter fitting
for k.. For a 2.36-W source, ;cl value for &, of 0.064 Bru/(h - ft-°F) provides the best fit
to the measured wall and centerline temperatures. None of the plausibie alternative source
strengths and heat-transfer parameters that were examined produced a temperature
difference between the centerline and the cell that deviated far from the measured value of
12°F . Only for the condition of filling the cell with vermiculite was the centerline
temperature pi'ojcctcd to increase appreciably. Projected temperature differences between

the centerline of the bed and the surrounding cell for this case ranged from 20 to 55°F.

20

 
S. MSRE SALT RADIOLYSIS

The liberation of fluorine gas by radiolysis of the lighter constituents of the MSRE salt
was first recognized in 1962 /20] and has been studied intermittently since that time. A
simplified picture of the process assumes the formation of radical species by homolytic
cleavage of the salt, followed by the formation and liberation of molecular fluorine and

the deposition of a resident active metal center in the salt lattice:

LiF Li- + E

+hy = - FT @
BeF, Be: + 2F.

The net production of fluorine is governed not only by the forward steps shown here but
| also by a temperature-dependent back-reaction (i.e., recombination) of the metal and
fluorine that restores the original salt. Various studies have identified minimum
“annealing” temperatures, where the radiolysis and recombination rates are equal, that
range between 70 and 150°C 21, 22]. Only recently has it been recognized that the

room-temperature fluorination of UF, in the MSRE fuel salt may also occur under the

storage conditions:
UF, + 2 — UF 1 ®)

The following sections update and summarize the ‘evidence regarding radiolysis of the

MSRE salt. At present the experimental evidence is restricted to fluorine generation—

only field measurements at the MSRE have confirmed the radiolytic generation of UF.
Prediction of the generation rateé for F, and UF requires both a reaction model and an
estimate of the decay energy deposited in the fuel salt. In Sect. 5.1 the experimental

evidence supporting a simple model for the radiolytic generation of fluorine is updated

and reviewed. Section 5.2 examines the deposition of decay energy in the salt beds and

21
 

 

couples the resulting dose estimates with yield values to project the potential gcneraition

and accumulation of fluorine at the MSRE.

5.1 RADIOLYSIS EXPERIMENTS

The 1986 experiment of Toth and Felker [2]] explored the behavior of fuel-salt simulant
in the limit of high radiation doses and sought to establish the asymptotic limit of
radiolytic damage. The results from this work are reconsidered here because they
illustrate some important points and because the radiolytic yield for this experiment was
recently calculated and should be reported. In this study a 30-g powdered salt sample
was exposed to the intense gamma flux from spent-fuel elements recently discharged
from the High Flux Isotobe Reactor (HFIR). Radiolysis was followed by measuring the
pressure rise due to fluorine generation as a function of time. In Fig. 3 the basic pressure
vs time data have been transformed into the standard format of amount of radiolytic
product vs absorbed dose (Appendix B). The resulting sigmoidal curve displays three
distinct regions: (a) an induction period that extends to 17 Wh/g when no fluorine is
released, followed by (b) a linear generation region whose slope corresponds to a
radiolytic yield of Gg,=0.012 tholecules of F, per 100 eV of absorbed energy, and
eventually (c) an (appatent) asymptotic damage limit that occurs at about 150 Wh/g, or
2% damage (i.e., metal center concentration). The first two regions have been identified
in previous studies /22] and are typical of many radiolytic processes. The existence of a
damage limit results from the accumulation of active metal sites to the extent that the rate
of recombination counterbalances radiolysis. These three parameters—induction period,
yield, and damage limitfform the basis for making projections about the generation of

fluorine from MSRE salts.

22

 
 

The 1995 results displayed in Fig. 3 were derived from experimental conditions that are -
believed to be the same as in the 1986 trial (Appendix B); however, in the 1995
experiments UF, generation was the primary focus. No UF was found in the gas space
above the sample, even after heating to 200°C, and the fluorine generation rate is far
below that (:xpected. The difference in the particle size of irradiated samples is thought to
be the cause of this discrepancy: the 1986 material was a 50-100 mesh powder, whereas
the 1995 sample consisted of loose chunks of 0.5-1.0 cm. A heat-transfer analysis of the
1995 sample conditions, contained in Appendix D, indicates that it is likely that the large
chunks of salt experienced considerable heating and thus promoted the recombination of

fluorine. Future experiments will resolve these issues.

 

 

0.35 [ T T T ] T T T ! 1 T T T ]
03 [ --e--1986 e’
b .'
—e— 1995 ’
0.25 | ”

 

o
N

0.15 }

Fluorine Accumulation
(mmol F2 / g salt)

0.1

0.05 |

 

 

_llllllllllllllllllllllllllllll

 

200

Dose (Wh/g salt)

Fig. 3. Fluorine generation curves for 1986 and 1995 irradiation experiments.

23

 
 

Despite the unexpected results of the 1995 trial, a more consistent picture appears if one
examines all of the radiolysis litérann'c for MSRE salts. Table 10 summarizes the results
for experiments that used a variety of radiation sources to generate fluorine from fuel salt
and fuel-salt simulants. The consensus of these studies is that the expected yield from
radiolysis at room temperature is about 0.02 molecules per 100 eV of deposited energy
and that a value of 0.045 represents a likely upper limit.

It is not yet clear how radiolysis varies with the energy spectrum of gamma radiation, but

it appears that the MSRE spectrum is comparable to or less energetic than the sources

used in radiolysis experiments. Figure 4 and Table 11 show that the source spectrum of

Table 10. Summary of radiolysis experiments on MSRE fuel salts

 

 

F, yield
(molecules per
Date/ID Radiation Salt form Induction period 100 eV)
19634 Post-irradiation Plllg Erratic 0.005-0.031
(MTR-47-5)  decay energy <1id 0.02
19635 %0Co v, Plug 25d 0.045
(Savageetal) 0.72 MR/h 1.3 Wh/g
19630 Van de Graaf p,  Farticles, None evident  0.02
(Baker, Jenks) 1000 MR/h ~900 ym
1964 - Soft x-rays, <50 pum "Notreported  0.005-0.04
(Rainey etal) 0.13 MR/ ~700 pm 0.0006-0.004
1986°¢ HFIR-pool v, 10 days 0.012
(Toth, Felker) 20 MR+ oW 17 Whyg
19904 238 P No F, detected
(Toth, Felker) Pu 9‘ lug after 1 year

 

@Source: Blankenship, F. F., et al., in Reactor Chemistry Division Annual Progress Report for the Period
Ending Jan. 31, 1963, ORNL-3417, Qak Ridge National Laboratory, pp. 17--30.

bSource: Reactor Chemistry Division Annual Progress Report for the Period Ending Jan. 31, 1964,
ORNL-3591, Oak Ridge National Laboratory, pp. 16-37, May 1964.

CSource: Toth, L. M., and Felker, L. K., “Fluorine Generation by Gamma Radiolysis of a Fluoride Salt
Mixtare,” Radiat. Eff. Def. Solids 112, 201~10 (1990).

dSource: Toth, L. M., unpublished data, 1990.

 
MSRE fuel salt is comparable to that of the HFIR cooling pool and is, on average, less
energetic than the radiation field from a %°Co source. Beta and gamma radiation appear
to be equally effective for salt radiolysis, and the preliminary indication is that alpha
particles do not radiolyze the salt to any significant extent [23]. At this point we adopt

the conservative assumption of equal effectiveness for all forms of radiation.

 

   
   
 
 
    

 

 

 

 

 

 

100 @ -
]
G

A 10 ¢ 3
>" s -
80 ]
8 -
a — & - HFIR-pool source spectra: )
g 1 d of cooling y
5 1 F —— HFIR-pool source spectra: _' -
v - 5 d of cooling 3 :
© L | ---©--- HFIR-pool source spectra: -
& - 21 d of cooling B -
S : i

g ----&--- MSRE source spectra: :
S 0.1 L 25 years after discharge 4
R 5
| i %Co : ]
1 energy i% ]
- ¢ 4 -
0.01 L 1 etz a1 sl L i i 1 a1 1 J: b1 1 1 1%

0.01 0.1 1 10

Q = Gamma Energy (MeV)

Fig. 4. Comparison of source spectra for MSRE fuel-salt irradiation.

25

 
'$661 ‘61 AIng “20uspuodsanio) FEwIA] INYO ‘SYUDL WD 194 FYSI Ut uotsoda(q (81au7 Jo saypuiysg pasiaay
‘SUBIEM *d "d €661 39q010 ‘SaqUIFSSY 1o 1adS Y[JH 40f SUODIMIW)) [ ZUdS1Q UONBINI[E) QAN ‘620-€6-HIAH-D “JoUSL] *H 'S :$904m0S,,

 

26

 

 

[S/ASIN] 4101 X9'T ‘[S/T]1,{01 X €'y = 90IM0Ss [I10], [S/ASIA] (101 X 6 ‘[S/1] 0T XS'T = 92mOS [RIOL,
000 000 6 o1 000 000 ¢'6 Il 81
000 000 TL 8 000 000 L 8 L1
000 000 Y $9 000 000 S 9 91
0070 000 Sy ¢ 100 000 S'e v S1
000 000 St v VTl L1°0 cL't £ vl
00°¢ 6£0 SL'T £ 88°0 S1°0 §TT 4 ¢l
00°0 000 A ST 90'1C 65y SLT z z1
00 10°0 £8°1 C - 689 ore STl 9 | I
8C0 L00 S6v'1 99°1 897 a0cl ¢80 1 01
00 ST°0 Sor'1 £e’l ¥6'9¢C C8L1 CLS O L0 6
0L0 870 60 I v’y ¢S’y SLEQ Sv'0 8
cL'18 091y Lo 80 99°¢ 656 ¢t £0 L
0T°¢ LS'1 S0 9°0 99°C org - SC1'o c1°0 9
8¢°1 09°'1 ¢e0 ¥o 120 Yo'y 6800 I'0 S
96°1 08T ¢C0 £0 960 69°¢ SLSO0 00 v
06°C 689 S1°0 0 68°0 £0'6 SLe00 Sv0°0 £
YA 6901 SL00 1’0 0 899 §200 €00 c
98°C y6'tt £00 §0'0 v'0 ¥8°91 10°0 <00 1

dnox8 w1 dnoi3 ur (ASIN) (ASN) dnoi3 ur dnoi8 ur  (A9N) £810ud (ASN) dnoi8

A310u2 95 vwwwed ¢, A310ud o8eroay punoqiaddn  ASuo g  vwwed g oferoay  punoq laddny  £Sisug

 

a8 reyosIp 10158 s1edk ¢7 :enoads TYSIN afireyosip 101ye p | :enoads jood Surjood YIIH

 

pUOL)BIPBLN Jjes-]onj TUSI 10§ 81323ds 32anog *IT dqeL
 

Far less confidence can be placed on projections about the induction period for fluorine
release. One can imagine that 2 number of factors, such as impurity levels and salt
morphology, make it difficult to predict induction times a priori, and the lack of
convergence in the reported values in Table 10 bears this out. The induction period of
1.3 Wh/g reported by Savage [22, 24] was the basis for assuming that annual annealing
treatments of the stored salts would be sufficient to preclude fluorine release. The recent
measurements of the MSRE off-gas and.the history of increasing radiation levels in the
MSRE off-gas system indicate either that this induction period is not correct for the

stored MSRE fuel salt or that the annealing heat treatments were not effective.

27

 
5.2 ABSORBED DOSE AND GAS GENERATION ESTIMATES

The radioactive source terms reported in Table 6 provide the basis for estimating the
energy deposited in the stored fuel salt. It is clear that all of the alpha and beta decay
energy will be deposited in the salt and that only a fraction of the gamma energy will be
absorbed. Estimates of the “leakage” of gamma energy from the 2583 kg of fuel salt in
fuel-salt drain tank number 1 were obtained using the transport code XSDRN2.7 and by
assuming an equivalent spherical tank geometry [13]. These calculations showed that
more than 88% of the gamma energy is deposited within the salt bed and that the
deposition is very uniform except for a narrow depletion zone at the wall. The spectrum
and intensity of the gamma flux at the inner wall of the tank are summarized in Table 12

and correspond to an exposure rate of about 600 R/h [25].

Table 12. Drain tank gamma spectrum (2583 kg salt basis, 25 years after discharge)

 

 

Fraction of Fraction of Fraction of
Energy Upper Lower gammasingroup gammasin group gammas in group
group bound bound at tank center at midradius at inner wall

no. MeV) (MeV) (%) (%) (%)

18 10.00 8.00 0.0002
17 8.00 6.50 0.0002
16 6.50 5.00 0.0001 0.0001 0.0002
15 5.00 400 0.0001 | 0.0001 0.0002
14 4.00 3.00 0.0002 0.0002 0.0002
13 3.00 2.50 0.481 0.476 0.545

12 250 - 2.00 0.0614 -~ 0.0586 0.0796
11 2.00 1.66 0.0528 0.0507 0.0673
10 1.66 1.33 0.114 0.111 0.136

9 1.33 1.00 0.211 0.208 0.246

8 1.00 0.80 0.291 0.289 0.328

7 0.80 0.60 30.0 30.1 31.2

6 0.60 0.40 13.5 13.5 16.8

5 0.40 0.30 9.81 9.79 11.9

4 0.30 0.20 17.7 17.7 19.7

3 0.20 0.10 23.8 23.8 18.2

2 0.10 0.05 3.66 3.67 0.8

1 0.05 0.01 0.353 0.355 0.0023

Total flux [gamma/(cm? - s)] 2.358 x 10° 2.347 x 10° 7.331 x 108

 

28

 
The decay-power history is displayed according to radiation category in Fig. 5 and
Table 13. Since 1970 beta-gamma decay has been the dominant source, with the beta
source being roughly twice the size of the gamma source. At present the alpha source is
only a third of the total, but this fraction is slowly and steadily increasing and will

eventually become the dominant source.

 

 

 

 

 

 

o= ' C r'j'll l ' ' l"'ll l ' ' "l'l-
10° -
,S B
a0 i
=
& i
N
Q/"\
52 10° L
o -
A -
g :
= :
I 3 |
10 A\
- | N
i 9 3 1 llljj! § 1 i IlLlll i 1 IlllJ_}
0.1 1 10 100

Time After Discharge (years)

Fig. 5. Distribution of source-term power by radiation category.

29
 

Table 13. Distribution of source-term power by radiation category
(total fuel and flush-salt basis)

 

 

Years after Total power Alpha source Gamma source  Beta source
discharge (W) W) (W) W)
0 5699.5 46.5 |
03 2269.8 46.8 - 967.0 1256.0
1 733.2 47.1 160.7 525.4
3 276.1 419 54.5 173.7
5 201.3 43.1 - 43.6 109.6
10 168.2 47.0 36.8 84.4
20 138.3 440 29.2 65.2
25 126.4 42.6 26.0 57.8
25.5 125.3 42.5 25.7 57.1
30 115.6 412 23.2 51.2
100 41.8 27.1 492 9.78

 

If we adopt the simplifying assumption that all radiation is equally effective in
radiolyzing the fuel salt and neglect the (uncertain) effects of annealing treatments and
induction periods, then projections about the potential for fluorine generation and
accumulation are straightforward. Under these conditions the fluorine generation rate is
proportional to the upper decay-power curve in Fig. 5, and the accumulation of fluorine is
proportional to the area under this curve. The fluorine generation and accumulation
curves for both the best-estimate yield value of 0.02 molecules of F, per 100 eV and the
upper bound of 0.045 molecules of F, per 100 eV are presented in Figs. 6-8. The
fluorine flow rates in Fig. 6, which range from ~200 cm?/h 1 year after discharge to

20 cm3/h in 1995, can produce pressure rises in file 2000-L off-gas volume, which range
from 13 psifyear (1971) to 1.3 psifyear (1995). The area under the flow curve, starting in
1971 when the salt was cooled below its annealing temperature, represents the amount of
F, accumulated to date and is displayed in Fig. 7. Depending on the assumed radiolytic
yield, between 300 and 650 mol of F, could have been generated by 1994. This

represents an inventory of two to five times the amount of fluorine that has been

30
A - Gm = 0.02 molecules Fz /100 eV, or
A Flow(cm®/h) = 0.167 » Decay Power (W)
—a&— G, = 0.045 molecules F, /100 eV, or
A Flow(cm’/h) = 0.376 « Decay Power (W)
102 |~ -
= -
-g E i ‘\\
g = .
c E [
§ o |
L'DN 2 ]
L
. A
10 E’ S . ‘_‘]
| 2 1 i 1.0 .2 291 1 A 1 A3 1. 1?
1 10 100
Time After Discharge (years)

Fig. 6. Projected fluorine generation in the absence of annealing and

 

 

     

 

 

 

 

 

 

 

 

 

 

 

induction effects.
700 T 1 NI T

: Trer. L .
=~ 600 ry=¢ P ‘ ~

[ (Deca wer « dt ‘ i
2 . F2 yro ) G = 0.045
é : 1971 L’ F2 ]
7 500 | ,,' h
al L’ ]
[+] 8 ’ ‘
3 5 , ]
= o 400 ’ ,
22 :
s E ]
“% 300} -
8 -
5 b -
2 3 p
g 200 | -
< 100 ]

o r i . | s a1 g L+ 3 9 1
0 5 10 15 20 25 30

Period After 1970 Discharge (years)

Fig. 7. Projected accumulation of radiolytic fluorine in the absence of
annealing since 1971.

31
 

 

 

  

 

 

 

 

400 T T T T 1 M T )
. ]
i 1994 s ]
350 ., _ G S -
o Y =0, (Decay Power » dt) ’
o 9
o - ’
% :E; 300 N T:cf. "
'§ e 250 - )
[ C S
Gt o | . ’
S > 200 F _e
€ 8 : G,=0045 ¢ _jo2
E § [ - ) o F2 :
2 & 150 % ]
§ & | S ]
g & s SR ]
g 100 : : ]
50 + ]
0 L1 nv:l_n_ a1 * 1 o4 4 0 1 4 4 4 1 :

1992 1988 1984 1980 1976 1972

T,., _» Reference Year

Fig. 8. Projected accumulation of radiolytic fluorine in the absence of
annealing between 1994 and previous years.

identified in the off-gas system. The various factors that may account for the extra F,
identified in this estimate include (a) fluorine deposition by corrosion or reaction,

(b) partial recombination of fluorine during annealing treatments, and (c) the condensation
or deposition of UF, that is as yet unaccounted for. Because the off-gas analysis of Table 7
indicates the presence of volatile fluoride corrosion products at very low levels, it appears
that consumption of fluorine by corrosion upstream of the ACB is a minor factor. Itis also
possible that fluorine was deposited in the ACB in a larger proportion than that inferred
from the 1994 off-gas analysis {1]. There is no detailed information to support conjectures
about the effectiveness of annealing treatments or to estimate the holdup of fluorine in the
form of condensed UF,. At this point we can only say that the generation of 300-650 mol

of F, is consistent with what is known about the present condition of the MSRE.

32
Because of the uncertainty about annealing treatments, induction times, and the UF
formation mechanism, it is impossible to construct a precise time line for fluorine
generation and accumulation. The fact that the radiation field surrounding the ACB
deposit is stable indicates that most of the uranium deposit is at least 5 years old. Some
questions about the genesis of UF, and F, can be answered by looking at fluorine
accumulation from the reverse perspective, starting in 1994 and integrating to an earlier
time. Figure 8 presents the accumulation term from this perspective and is useful because
it allows one to see that the 115 mol of F, identified in the MSRE could not have been
generated recently. According to these projections fluorine generation began prior to

1989 and it probably started much earlier than this time.

33
6. THERMAL FLUORINATION TESTS

During the review of UF, generation at the MSRE, it was suggested that the annual
annealing of the salt to 150-230°C may have provided a nonradiolytic pathway for UF
generation. The consensus in the literature is that significant conversion of pure UF, to
UF, by F, under static (i.e., nonflow) conditions does not occur below 350°C /26].
However, thin samples of UF, powder exposed to a flowing gas stream exhibit the onset
of fluorination at temperatures between 220 and 230°C f27]. This 120°C difference
between onset of fluorination for static and flow systems has been explained on the basis

of alternate reaction paths. For the flow tests with thin samples, it is believed that the

UF formed does not have the opportunity to react with UF, to form fluorides of
intermediate oxidation state (e.g., UF;). These “intermediate” fluorides have been shown
to be less reactive than pure UF, [27]. During static tests on a thick sample UF has the

opportunity to react with the resident UF, and form the inhibitory intermediates.

At this point it is not clear which of these onset limits applies to the situation of dilute

UF, and F, trapped in the stored fuel salt. Furthermore, it appears that low-temperature

fluorination of UF, in the MSRE salt requires a more oxidizing species than F,, such as

atomic fluorine. In fact, atomic fluorine has been shown to be an effective fluorinating

agent for UQ, at room temperature /28], and it is expected that the formation of UF; in

the fuel salt is caused by the atomic fluorine generated by radiolysis.

In to order to see if low-temperature treatments of MSRE salt with molecular fluorine
could also be responsible for the generation of UF,, the following static (nonflow)
experiments were conducted. Ten grams of salt simulant, held in a nickel crucible and
contained by a passivated 50_0-cm3 quartz chamber, was held for 6 h at 200, 250, 300, and
400°C in 0.5 atm of fluorine. Infrared spectra of the gas sample after each temperature

trial showed no evidence of UF or change in salt weight, and only minor fluorine losses

34

 

L.
were noted. After the 400°C treatment, trace activity was measured on the chamber

walls. A separate trial with pure UF, in the same apparatus was then performed to verify

the previous observation at this temperature. A measurable amount of UF, appeared in

the gas phase during this run. However, a week-long test with salt simulant in an all-

Monel system at 300°C confirmed the finding of no measurable conversion to UF at

temperatures below 400°C.

35

 
ACKNOWLEDGMENTS

J. P. Renier and I. Remec of the Computational Physics and Engineering Division of
ORNL carried out the ORIGEN-S and XSDRN code calculations. Their expert advice
and guidance are greatly appreciated.

37

 
 

10.
11.

12.

13.

14,

15.

REFERENCES

Del Cul, G. D., Some Investigations on the Chemistry of Activated Charcoal with
Uranium Hexafluoride and Fluorine, to be published as ORNL/TM-13052, Oak
Ridge National Laboratory.

Notz, K. J., Extended Storage-in-Place of MSRE Fuel Salt and Flush Salt,
ORNL/TM-9756, Oak Ridge National Laboratory, September 1985.

Notz, K. J., Feasibility Evaluation of Extended Storage-in-Place of MSRE Fuel Salt
and Flush Salt, ORNL/CF-85/71, Oak Ridge National Laboratory, March 1985.

Notz, K. )., Decommissioning of the Molten Salt Reactor Experiment—A Technical
Evaluation, ORNL/RAP-17, Oak Ridge National Laboratory, January 1988.

Bell, M. J., Calculated Radioactivity of the MSRE Fuel Salt, ORNL/TM-2970, Oak
Ridge National Laboratory, May 1970. |

MSRE Fuel and Flush Salt Storage, Request for Nuclear Safety Review and
Approval, NSR 0039WMO00013A (approved 12/28/93; expires 12/31/95).

Rosenthal, M. W., et al., Molten Salt Reactor Program Semiannual Progress Report
Jor the Period Ending Feb. 28, 1970, ORNL-4548, Oak Ridge National Laboratory,
August 1970, pp. 3, 24-25.

Rosenthal, M. W, et al., Molten Salt Reactor Program Semiannual Progress Report
Jor the Period Ending Aug. 28, 1970, ORNL-4622, Oak Ridge National Laboratory,
January 1971, p. 1.

Rosenthal, M. W, et al., Molten Salt Reactor Program Semiannual Progress Report
for the Period Ending Feb. 28, 1971, ORNL-4676, Oak Ridge National Laboratory,
August 1971, p. 1.

F. J. Peretz, ORNL, personal communication, September 6, 1995.

Thoma, R. E., Chemical Aspects of MSRE Operations, ORNL-4658, Oak Ridge
National Laboratory, December 1971, pp. 58-65, 99-112.

Compere, E. L., et al., Fission Product Behavior in the Molten Salt Reactor
Experiment, ORNL-4865, Oak Ridge National Laboratory, October 1975.

Williams, D. F., Revised Estimates of Energy Deposition in MSRE Fuel Drain Tanks,
ORNL Internal Correspondence, July 19, 1995.

Jennings, W., et al., “Solubility of Uranium Trifluoride in MSRE Fuel Salt Mixture,”
pPp- 50-52 in Reactor Chemistry Division Annual Progress Report for the Period
Ending Jan. 31, 1964, ORNL-3591, Oak Ridge National Laboratory.

Ramey, D. W., Technique for Estimating the Uranium Content in the MSRE
Auxiliary Charcoal Bed, ORNL Internal Correspondence, Sept. 12, 1994,

39
 

16.

17.
18.

19.

21.

23.
24.

25.
26.

27

28.

Sulfredge, C. D., and Morris, D. G., Summary of Thermal Calculations for Filling
the MSRE Charcoal Bed Cell with CO, or Vermiculite, ORNL Internal
Correspondence, Mar. 14, 1995. |

McAdams, W. H., Heat Transmission, McGraw-Hill, New York, 1954, pp. 172-176.

Nakajima, T., and Watanabe, N., Graphite Fluorides and Carbon-Fluorine
Compounds, CRC Press, Boca Raton Florida, 1990.

Williams, D. F., “Charcoal Bed Conductivity,” Internal Memorandum to B. D.
Patton and L. M. Toth, June 21, 1995.

Blankenship, F. F., et al., in Reactor Chemistry Division Annual Progress Report for
the Period Ending Jan. 31, 1963, ORNL-3417, Oak Ridge National Laboratory,
pp. 17-30. .

Toth, L. M., and Felker, L. K., “Fluorine Generation by Gamma Radiolysis of a
Fluoride Salt Mixture,” Radiat. Eff. Def. Solids 112, 201-10 (1990).

Haubenreich, P. N., Fluorine Production and Recombination in Frozen MSR Salts
after Reactor Operation, ORNL/TM-3144, Oak Ridge National Laboratory,

Sept. 30, 1970.

Toth, L. M., unpublished data, 1990.

Savage, H. C,, et al., “Gamma Irradiation of a Simulated MSRE Fuel Salt in the
Solid Phase,” pp. 27-31 in Reactor Chemistry Division Annual Progress Report for
the Period Ending Jan. 31, 1964, ORNL-3591, Oak Ridge National Laboratory.
Etherington, H. (ed.), Nuclear Engineering Handbook, Sect. 7-7, 1958.

Brater, D. C., and Smiley, S. H., “Preparation of Uranium Hexafluoride,” in
Il’;ggress in Nuclear Energy, Ser. 3, Vol. 2—Process Chemistry, F. R. Bruce (ed.),

Labagon, V. Y., and Johnson, K. D. B., “Kinetic Studies of the Fluorination of
Uranium Tetrafluoride by Fluorine,” J. Inorg. Nucl. Chem. 10, 74-85 (1959).

Beattie, W. H., and Salopek, M. A., “The Reaction Between Dissociated Fluorine
and Oxides of Uranium,”/J. Fluorine Chem. 30, 59-72 (1985).

40
 

Appendix A
ORIGEN-S RUN INPUT FILE
 

$SSSSSSSSSS
S$SSSS5S8S8S5S8S
SS ss
Ss
§s
S§5S555558888
§555588558s88
§s
ss
3 ss
SSS8SSSSSSSSS
S55S5588858

ccccceeceee
ccececeecececced
cC cc
cc
cc
cc
cc
ccC
cc
cc cc
cccceecceecee
cceceececeee

aaaaa2aaa
aaaaaasaaaa
aa aa
aa aa
aa aa
aaaaaaaaaaaas
aaaaaaasaaaas
aa aa
aa aa
aa aa
aa aa
aa aa

it '
tLeeeeeetett
teeereeetettt

HRR AR R AR R R AR R R SRR R R R R R R R R R R R WA AR i i R Wi e A R felr ek
HRRR i R R R R AR R AR R i dr e e AR AR A A i R e R R AR W e R AR R A AR R R A W e e e A e e e A de e i ek
A e A W A W e R e R R A e e A R A A AR R R R R A VR R SR A R W e e R A AR A Ak e Al el ek

fieiehen
seirdrirk
dririeden
drirdrir
ke

program verification information

code system:

scale version:

4.2

riededrd
ik
fedrdrird
hhhw
Yedrdrdrd

AR R A IR i v e dr e deve st de ke s vedk dede e ve v e v W e e e e s s e s s e e e e v s s s e dede s s e et de S et e e S e e e e e Wk e e
ARARA SRR AR FERRRRAR R R R R AR R R drdr de b i s ded W e W i e e e dr i e i de s v s e el s e e o e e e e e e e i

program: origns
creation date: 05/31/95
library: /scale/exe
prodact;'on code: origens
version: 2.5
jobname: i7r
date of execution: 06/02/95
time of execution: 13:01:05

AEREARER TR TR RARRTTRRR R R RRRR TR RTRTRRAAREEET TR R i R R Rkt ey

43

 
 

 

note:

....................

esitsannrenccnscavesnrnannn ...colums 1-72...... Gtessmassnsecscacsesnanna

0ss Al

* input echo * (with break between col. 1-72 and 73-80)

7 ET

DECAY CASE
338 211127 A62 A3318 ET

3588 0 1T

5438 AB 1 A11 1 E
5638 A2 10 A5 1 A0 O A3 116 ALS A3 E

ST** 0 ¢

T

MSRE DECAY FROM 12/12/1969 10 12/31/95
ENTIRE MSRE CORE (U-233 FUEL)
60 .3 1351020 25 25.5 30 100

only comments are permitted after colum 72.

--------------------

420980

601450
621520

61** F 0.000001
6538
TGRAM-ATOMS  GRAMS CURIES WATTS-ALL WATTS-GAMMA
32 1 01 1 0 1 32 32 62
37 10 1 10 1 2 32 62
32 1 01 1 0 1 32 32 62
818820261 E
828 0 0 00 002 2 2 2
83**  1.E+7 8.E+6 6.5+6 5.E+6 4 .E+b
I.E+ 2.5E+6 2.E+6 1.66E+6 1.33e+6
1.E+6 8.E+5 6.E+5 4.E+5 3.E+5
2.E+5 1.E45 S.E+d 1.E+6
B4** 2. E+7 6.434E+6 3.E+6 1.85E+6 1.4E+6
9.E+5 4. E+5 1.E+5 1.7E+4 3.E+3
5.56+2 1.E+2 3.E#+ 1.E+1 3.04999E+0
1.77E+0 1.29999+0 1.12999+0 1.E+0 8.E-1
4,E-1 3.25E-1 2.256-1 9.999985€-2 5.E-2
3.E-2 Q.999998E-3 1.E-S
73sS 370870 380B80 380890 380900 390890 390910 400900 400910
400920 400930 400940 400950 400960 410950 420950 420970
421000 430990 441010 441020 441030 441040 441060 451030
461050 461060 461070 511250
521271 521280 521291 521300 531290 551370 561370 S61380
571390 581400 581410 581420 581440 591410 601430 601440
601460 601480 601500 611470 6114B1 621470 621500 621510
631520 631530 631540 631550
922320 922330 922340 922350 922360 922380 942390 942400
942410 952410
410930 440990 471070 521250 531270 541290 621480 631510
641540 641550 20040 822060 822070 822080 832090 882260 902290
902300 902320 932370 -
390900 410951 451031 451060 501231 521251 521270 521290
551340 561371 591440
812080 822120 832120 842120 B42160 862200 882240 902280
942380 962420
641620 651621

30070 40000 400000 90000
74** B 480 5.64 5.573 9.893E+1 6.660E+1 7.458 4.440 9.730E+1 1.100E+2
1.190€+2 1.180E+2 9.306 1.140E+2 4.397 2.440E+1 3.250E+1 3.160E+1

2.850E+1 2.9BOE+1 1.950E+1 1.560E+1 2.292 7.880 2.238 5.070E+1
1.820E+1 4.610 5.640 6.176E-1
4.175E-1 9.460 8.859E-1 2.090E+1 1.490E+1 1.287E+2 5.560 7.440E+1

1.700E+2 1.890E+2 1.439E+1 1.730E+2 3.979E+1 1.700E+2 1.670E+2 1.130E+2
1.070E+2 8.410E+71 4.630E+1 1.900E+1 4.011E+1 4.912E-2 2.390E+1 2.710E+1

5.585 1.460E+1 3.110E-2 5.45 1.303£-1 7.629E-1
7.846 3.232E+4 2.911E+3 9,B60E+2 6.800E+1 2.370E+3 6.223€+2
7.501E+1 8.742 2.793E-1

 

comment or title end
00000000000000000C000
2.50E-2 1.09€-2 2.27E-3 2.50E-6 3.64E-6 1.04E-2 1.24E-2 4.67E-3
4.26E-3 1.95E-5 1.69E-3
1.85E-7 1.10E-4 1.04E-5 5.46E-16 4.25E-10 1.66E-7 9.60E-4 1.87E-1
5.B4E-2 7.40E-3
1.98E-8 3.16E-7
528901 319130 533050 3233452
75¢¢ 3 3 3 333333333333 333333
33333333
3 3333333333333 33333333
33333333
2 2 222122222
3 333 3333332222222 222
33333333 333
2 222222222
33
1 4 4 4 17

MSRE DECAY FROM 12/12/1969 TO 12/31/95 TIMESTEP 7
MSRE DECAY FROM 12/12/1969 TO 12/31/95 TIMESTEP 8
MSRE DECAY FROM 12/12/1969 TO 12/31/95 TIMESTEP 9
MSRE DECAY FROM 12/12/1969 10 12/31/95 TIMESTEP 10
5688 0 0A0 7 ET
563 0 O A0 8 ET
5688 C 0D A0 9 ET
568$ 0 0 A0 10 ET

5688 FO T

when job “fails", make sure no fido INPUt...ccvveccnnenee Getebeetemeenaransanvanssres is out here!
0% array 12 entries read
1114

dbl. prec. machine word applied has, at least, a 16 significant figure accuracy.
short-lived split test fraction, gxn = 9.1188E-04
half-norm of matrix used, axn = 7.0000E+00
4-place-accuracy-retention ratio, ratiot = 6.4516E-13
3% array 33 entries read

ot

45

 
 

library information...

cross-section data taken from position number 1 of library on unit 21,

pass 1

pass O i

*scale-system control mockile sas2 library*

used a time-dependent neutron spectrum, for each of the above passes
pass 0 applies start-up fuel densiities
pass n applies mid time densities of nth library interval

first library updated was...

WIRR AR R AR R R LRI RN AR R AR AR R R AR ded A s e e dritde dr e e s Wt i el e e s de el e et el e dr e e et

* *
* prelim lwr origen-s binary working library--id = 1143 *
* made from modified card-image origen-s libraries of scale 4.2 *
* data from the light element, actinide, and fission product libraries *
* decay data, including gamma and total energy, are from endf/b-vi *
* ‘ *
* neutron flux spectrum factors and cross sections were produced from *
* the "presas2" case updating all nuclides on the scale "burnup" library *
* *
* fission product yields are from endf/b-v *
* *
* photon libraries use an 18-energy-group structure *
* the photon data are from the master photon data base, *
* produced to include bremsstrahiung from uo2 matrix »
* *®
* see information above this box (if present) for later updates *
* *

SeRedededede dodede Rl vl e A dr e R Wk R W S R RN R W e R e S R e R e e R R e e e e e A R e e e e e e de e e i e dede e
* *

BRR R AR dArk R R Rk A ki R A R e A AR A A R R R R R R AR e SR e e W e de e R et s e e e ke e

.other identification and sizes of library.
data set name: /scale/datalib/origen/binrylib/prlimlur
4/ 471995 date library was produced
1697 total number of muclides in library
689 number of light-element nuclides
129 number of actinide nuclides
879  number of fission product nuclides

7935 number of nonzero off-diagonal matrix elements

R R R R R R E W Rk Sk A AR R RA AR AR TR R TR R AR R AR AT RRERR AT TR R AR TR R TR vl R ik Sk d ki d g

 
 

Appendix B :

CALCULATION OF RADIOLYTIC YIELD FROM IRRADIATION OF MSRE
FUEL SALT IN THE HFIR COOLING POOL
The pressure-time data from irradiation experiments must be transformed into an amount

of radiolytic product (moles of F, per gram of salt) and an accumulated dose (absorbed

energy per gram of salt). Because of the rapid decay and replacement of spent fuel in the

HFIR cooling pool, the accumulated dose must be calculated from the irregular gamma-

flux profile shown in Fig. B.1. The accumulated exposure is simply the area under this

curve.

100

Gamma Flux (MR/h)

 

 

 

 

----- 1986 experiment
— 1995 experiment

 

 

 

 

 

20 40 60

Irradiation Time (d)

Fig. B.1. Gamma-flux profile during HFIR-pool irradiation experiments.

49
 

Because the irradiation specimens constitute a thin absorbing medium (i.e., constant flux
throughout the sample), the dose can be calculated by referencing the energy deposition

in salt to that in air by using the ratio of their mass absorption coefficients (L/p):

Dose[eV/g] = (0.873 (Ej-di‘ll) [‘—'afl—] Exposure[R] 6.25x 1013 [Ex\_;_{ig_] , Of

WP)air R

Dose [eV/g] = 5.456x 1013 (Wp)sah
Exposure [R] . - (Wp)air

Mass absorption coefficients are both energy and composition dependent, but for the
lighter elements there is little variation over a broad range of energies [Templin, L. J.,
Reactor Physics Constants, ANL-5800, Argonne National Laboratory, July 1963, p. 652]. |
This ratio should be near unity. In the absence of a detailed analysis of absorption
coefficients, we use the ratio of 1.1455 determined by Savage in his %Co irradiation of
MSRE salt [Savage, H. C,, et al., “Gamma Irradiation of a Simulated MSRE Fuel Salt in
the Solid Phase,” pp. 27-31 in Reactor Chemistry Division Annual Progress Report for
the Period Ending Jan. 31, 1964, ORNL-3591, Oak Ridge National Laboratory]:

Dose[eV/g] _

= 13 = 13
Exposure [R] = >-4°6%x107 (1.1455) = 6.25x10

The conversion of accumulated fluorine pressure to a number of moles is simply a gas-

law calculation for the void volume of the specimen:

 

n = % _ P (psi) (75.4/1000 IIZ; — 6.02 x 102 mglc%lla
0.08205 L-atm (393 gy 19./ PSt mo.
mol-K atm

 

I - 1.202x1020 molecule
P psi

50
 

The radiolytic yield (G Fz) can now be calculated from the slope of the pressure-exposure

curve as follows:

Example: 1986 Experiment
dP/dE = 1.93 x 10 psi/R at an average exposure rate of 20 MR/h

193x 109 [P%] 1.202 x 10%° [%]

 

dP n
G,=dn -EP . , | ,
: dD % m 6.25 x 1013 [e—w&] 30 g salt]
R
= 12x10% molecule
eV

_ molecule F,
Or, = 0012 1655 absorbed gamma energy

51
Appendix C

ESTIMATION OF CROSS-TRANSFER OF FISSION PRODUCTS AND
PLUTONIUM TO THE FLUSH SALT
. =1
ey iy @ aeer : _ €89 . . —
re1o oY1 Jo %8130 ‘nd89'€r= {8 epL+ 30LS + | £89/(907) = @nuing %) M 3065 | + 01 X8EY
1[es [ong 3 095y €'89 _
gv e {3epL+30LS + [(a194d ¢gz-n Sunnp parq nd 8 066) ¢ T dnung ) w = pawRjsuen nd Jo junoure
01 '~
810130 %L1 = | 001/(k6E) = Tumnag) S | ° 0T XY =

1=1
N = pauajsuexn sjonpoid uorssty Jo 9,
6

SajeuIn)sT JIJSUBIY-SS0))

wes Py 3Y 09sv) 001
psuen 8y 07 | (dnwng %)

"90U19JA1 SIY JO £ “d uo udAId suyl own Sunerado Y YPIM JUANSISUOD 10U ST I IK) “bodas v..aE_ofi. JO €9 *d U0 PoYNUAPI §I JOJSURI-SSOID STILL
"0L61 KB *Asoreioqe [euorieN 93pry ¥eQ ‘0L6T-WL/INYIO “IvS 1ond FYSHW a1 Jo 1anavoipoy paiomap) “f "W TI9d :994n08
" TII-66 'S9-8S "dd ‘IL61 Jaquiadaq ‘Ai0reioge] [euoneN d3pry NeQ ‘8S9y-TINNO ‘swonpiadQ JYSW Jo $102dsy [oopuay) “g ~J ‘BWOY], :29.4108,,

 

 

8 epL (%001) 918'V01  69/EI/TL: 69/E1/8  €€-07-dd : S-61-dd 6

8 oLs (%L'L8) ¥¥6'16 69/21/8 : 89/v1/8 7-61-dd: 1-S1-4d  (uomesadonger) 8

3 06$ (%€°89) 08S°1L 80/8T/V: L9/I1/6  TL-¥I-dd: €-€1-dd L
Pmod resur | (%ETY) T9E DY L9/8/6 * LI/LT/9 [-€1-dd : €-C1-dd 9
Ymoss reour| (%8'€€) S8E'SE L9/O1/9 : LOTV/S 1-71-dd : 09-11-dd oS
[ImoIs reour| (%8'€E) SBE'SE  LY/OL/S : 99/T1/T1 65-11-dd : 2-01-dd v
yimous resur (%1TT) 11911 99/VT/11 : 99/¥/11 8-6-dd : L1-8-dd €
yImoIs Jeoury (%€°6) 00L'6~ 99/2/11 : 99/0€/6 S1-8-dd : ¥-8-dd z
pmoI3 seour (%S'L) €28L 99/3/6 : S9/81/T1 1-8-dd: 0l-y-dd  (VonwdoNgg) 1
pAIOIIAAL 1 [RIO], [(%) UMW pSolep o[dures  (ysniy JoIge pue 210j3q) '0U 19J5UBL-$SO1D)

qdnumg pSordures jfes-ysnpg

 

MBS Ysnyj 0} [oNj UI0J SIJSUBII-SSOID JO 30T °I°D) 3qe

35
Appendix D |
HEAT-TRANSFER ANALYSIS OF IRRADIATION SPECIMENS

 
As indicated in Table D.1, three different types of MSRE salt irradiation specimens have
been used to investigate radiolysis: (a) solid plugs, (b) powder beds, and (c) large
granular chunks.

Table D.1. Characteristics of selected irradiation specimens

 

 

F, yield
Saltbedsizz ~ Exposurerate (molecules per

Experiment Sample form (in.) (MR/h]) 100 eV)
Savage—1963  Solidplug  0.78 ID x 3.2 long 0.72 0.045

80Co source ‘

Toth—1986  Powderbed (079D x3.1long  15-25 0.012

HFIR pOOl v <0.02-cm diam

Toth—1995  Large granules (079D x3.1long  20-80 <0.001

HFIR pooly  ~ 1-cm-diam

All of these situations correspond to heat transfer in the presence of a uniform heat source
in the sample (due to gamma heating). However, there is an important distinction
between the cases of a plug and powder bed and that of a bed of large granules. For the
plug or powder bed, the entire bed can be treated as a continuum with an effective
thermal conductivity that governs the temperature profile _in the bed. The situation for
granular material is quite different, because the continuum approach is not valid and the

potential exists to support large temperature differences across an individual particle.

Single-Particle Calculations

The temperature rise experienced in an irradiated sphere (radius = a, diameter = d) can be
estimated based upon the source energy per unit volume, Q, and the resulting surface heat

flux, g/A:

q/A = Q » volume /surface area = Q » (nd3/6)/(rd?) = Qd /6

59

 
 

From Appendix B we have the following:

Q [cal/(cm?3 +5)] = 1.65 x 103 « E [MR/h], or
g/A [cal/(cm?+s)] = 2.75x 10« E[MR/h]  d[cm]

For conduction-limited transfer to the surrounding gas at T., [Carslaw, H. S., and Jaeger,
J. C., Conduction of Heat in Solids, Clarendon Press, Oxford, 1959, p. 232], the following

calculation governs the external temperature difference:
ATy = (T(,;, — T.) = Qd?% (12¢kgas) = q/A*d/ (2kgas) (D.1)
The temperature within the particle is given by:
Ty — Tima) = Qd2/ Qdkea) + (1 — 12/a2), and
AT, = centerline hcating. = Te=0) — Ter=a) = Qd%/ (24%ksan)

Because the salt is very conductive [ksa) > 0.0058 cal/(cm - s - °C), Source: Rosenthal,
M. W, et al., Molten Salt Reactor Program Semiannual Progress Report for the Period
Ending Aug. 31 1969, ORNL-4449, Oak Ridge National Laboratory, February 1970,

p. 92], the centerline heating is usually small compared with the temperature drop across

the particle surface.

The conduction-limited regime applies to very small particles—for larger particles, free
convection must be considered. Free convection from a sphere in the boundary layer
regime between creeping flow and turbulence (10 <Ra < 10%) is well correlated by the
semiempirical expression [Rosenhow W. M., and Hartnett, J. P. (eds.), Handbook of Heat
Transfer, McGraw-Hill Co., New York, 1973, p. 6-15]:

60

 
Nu =0.49 (Ra)l# (D.2)

Ranz and Marshall [Transport Phenomena, 1960, p. 413; Chem. Eng. Prog. 48,
141-173, (1952)] propose a more general expression for nonturbulent free convection,
which extends to the creeping-flow condition of the smallest particle (with conduction as

an asymptote for Gr =Ra =0):
Nu = 2 + 0.6 (PD!R « (Gn)'# (D.3)

where

_ _ (gA)d _ . — Dre
Nu = Nusseltno. = T T (T.-T) Ra = Rayleigh no. = Pr«Gr

T, -T.)d
Gr =Grashof no. = gP(Tu-T)& Pr=Prandtlno. =c, u/k

and at T.. = 300 K:

¢p = heat capacity [cal/(g-°C)] = 1.24 for He, 0.24 for air;

B = viscosity [g/(cm+s)] = 2x 104 for both He and air;

v = kinematic viscosity [cm?/s] = wp = 1.12 for He, 0.154 for air;

p = ideal gas density [g/cm®] = 1/22,400 « MW = 1.8 x 10 for He, 1.3 x 1073 for air;
k = thermal conductivity [cal/(cm-s+°C)] = 36 x 10 for He, 6x 10 for air;

B = coeff. of thermal expansion = 1/T.. for ideal gas;

g = accel. of gravity =980 cm/s%
Tw = wall or surface temperature;
d = particle diameter [cm] =1 and 0.02.

It should be noted that the approach of adding conduction and convection contributions
used in Eq. (D.3) is an approximation and is in conflict with the idea that there is a

critical Raylei gh number for the onset of natural convection. This approach probably

61

 
 

overestimates the influence of natural convection at low Rayleigh numbers and therefore
produces low estimates of the temperature difference required to support a particular heat

flux.

Solving for AT (= Ty, ~ T..) in Eq.(D.2) gives:

AT [°C] = ¢; * E*# [MR/h] * d [cm] D.4)
(forHe: c¢; =1.27, forair: ¢) =2.34)

Solving for AT (= Ty — T..) in Eq. (D.3) gives:

AT + c, AT = ¢ ®.5)

 

T ev2

@ \# A)-d
c, = 0.3Pr1’3( £ ) c; = (——q/Zi:

A bound on the radiant transfer of heat is included for comparison and to ensure that it is

not the controlling mode of transport:

14
AT =(q—’:- + T,;‘) -T D.6)

Here the emissivity assumed to be 1, and © is the Stefan-Boltzmann constant [1.36 x 1012

cal/(cm? - s-K*)]. Predictions based upon the preceding development are summarized in
Table D.2.

62
| Table D.2. Single-particle heat-transfer predictions
J Temperature difference (°C)

 

Conduction  Free convection Radiant Rayleigh no.

s Cover d E Egs.
gas (cm) (MR/h) Eq.(D.1) Eq.(D4) Eq.(D.5) Eq.(D.6) (D4)/(D.5)

 

He 1 20 7.6 — 5.1 32.0 25/9
1 50 19.1 — 11.8 67.3 52/21
1 80 30.5 — 18.0 94.7 76/ 33
0.02 20 0.003 — 0.003 0.7 <<1
0.02 50 0.008 — 0.008 1.9 <<1
0.02 80 0.012 — 0.012 3.0 <<1
Air 1 20 45.8 25.7 15.8 320 2820/1735
1 50 115 53.5 34.6 67.3  5870/3800
1 80 183 77.9 51.6 94.7 8540/ 5665
0.02 20 0.018 — 0.018 0.7 <<1
0.02 50 0.045 — 0.045 1.9 <<1
0.02 80 0.072 — 0072 3.0 <<1

 

- Initially the irradiation specimens have a helium cover; however, as radiolysis proceeds,
fluorine is generated, helium is withdrawn, and the transport properties of the ambient gas
become more like those of air. So the estimates for a helium cover represent the
conditions at the beginning of irradiation, and the estimates for an air cover bound those
conditions at the end of irradiation. These estimates support the assertion that the large
granular particles can develop a significant temperature gradient at the particle surface. A
temperature rise of 30°C puts the particle temperature at >70°C, which is above the
lowest estimate of the salt annealing temperature [Toth, L. M., and Felker, L. K.,
“Fluorine Generation by Gamma Radiolysis of a Fluoride Salt Mixture,” Radiat. Eff. Def.
Solids 112, 201-10 (1990)]. Thus it is likely that the decreased fluorine yield seen in the
1995 experiment is due to heating of the large particles to temperatures that promote

significant recombination.

63

 

 
 

Solid and Particulate Bed Calculations

The preceding single-particle calculations must be reconciled with our existing
knowledge and best estimates of heat transfer in packed and solid beds. It is useful to
examine some bounding predictions for the temperature distribution in a packed or solid
bed that contains a uniformly distributed energy source. The temperature in an infinite
cylinder of radius @, with uniform heat generation Q, and imposed surface temperature

Tr=a = Teo is given by:

AT = Tgy—Te = Qd2/ (16°kea) * (1 - 12/ 22)

Therefore, the centerline heating is as follows:

AT = (Tg=0) — Tg=a)) = Qd%/ (16%Keap)

This one-dimensional approximation overestimates the temperature rise in the bed
because it omits heat loss from the cylinder ends. In these projections the particulate bed
energy source is reduced by a factor of two from the previous single-particle values in
order to account for the bed voids. The value for the particulate bed conductivity is
derived from previous estimates [Wllhams D. F., “Charcoal Bed Conductivity,” Internal
Memorandum to B. D. Patton and L. M. Toth, June 21, 1995]. The estimates in

Table D.3 suggest that the 1985 experiment may have also experienced a slightly
enhanced recombination rate due to sample heating. Note that a large fraction of the bed

is cooler than this centerline value, as indicated by the preceding equations.
Table D.3. Projected centerline heating in irradiated samples

 

 

 

Sample Exposure Q Bed conductivity AT,
Trial form (MR/h) (ca]/cm3 *S) [cal/(cmes. °C)] | (°C)
Sa\l'ggg— Plug 0.72 0.0012 >5.8x103 - <01
__ Powder bed |
Toh—  (C002om 1525 00120021  ~4x104 ~10-20
particles)
65
46.
47,

48.
49,
50.

51.
52.

54

55
56
57
58
59-60

 

BN =t ik et el ek ek ek ek ek
eV NAMBEWN -

i
PPN, LD=

ORNL/TM-13142

INTERNAL DISTRIBUTION
S. N. Burman 22. B.D. Patton
C. H. Byers 23. F. J. Peretz
R. M. Canon 24. D. W. Ramey
J. M. Cosgrove 25. 1. Remec
A. G. Croff 26. J. P. Renier
S. Dai 27. 1. C. Rudolph -
G. D. Del Cul 28. J. E. Rushton
M. F. P. DeLozier 29. G. W. Sherrill
J. R. Devore 30. R. M. Szozda
R. L. Faulkner 31. L. M. Toth
L. K. Felker 32. L. D. Trowbridge
M. K. Ford 33. K. L. Walker
M. A. Green 34-38. D.F. Williams
E. M. Hodges 39. ER Document Management
J. S. Ivey Center
J. R. Keiser 40. Central Research Library
B. L. Kimmel 41. Y-12 Technical Library
L. E. McNeese Document Reference Center
G. T. Mei 42. ORNL Patent Section
D. E. Mueller 43-44. ORNL Laboratory Records
S. H. Park 45. ORNL Laboratory Records RC
EXTERNAL DISTRIBUTION

J. Caja, 9052 Highbridge Drive, Knoxville, TN 37922

E. C. Crume, Jr., Pragmatics Inc., 708 S. lllinois Avenue, Suite E105, Oak
Ridge, TN 37830
J. R. Engel, 118 E. Morningside Drive, Oak Ridge, TN 37830

P. A. Haas, 1205 Chickering Way Lane, Knoxville, TN 37923

M. R. Jugan, Environmental Restoration, Department of Energy, Oak Ridge
Operations, P.O. Box 2001, Oak Ridge, TN 37831
J. J. Laidler, Director, Chemical Technology Division, Argonne National
Laboratory, 9700 S. Cass Avenue — Bldg. 205, Argonne, IL 60439

N. W. Lingle, Environmental Restoration, Department of Energy, Oak Ridge
Operations, P.O. Box 2001, Oak Ridge, TN 37831

53. K. J. Notz, Henley Point-River Road, Kingston, TN 37763

Idaho Falls, ID 83415-5219

. A. L. Olson, Idaho National Engineering Laboratory, P.O. Box 1625, MS-5219,

. L. P. Pugh, 2024 Cedar Lane, Kingston, TN 37763

. R. E. Thoma, 119 Underwood Road, Oak Ridge, TN 37830

. R. G. Wymer, 188-A Outer Drive, Oak Ridge, TN 37830

. E. L. Youngblood, 198 N. Purdue Avenue, Oak Ridge, TN 37830

. Office of Scientific and Technical Information, P.O. Box 62, Oak Ridge, TN 37831
 

L4