ORNL-TM-0500.txt

  
 
 
  
   
   
  
 
  
  
   

 

'__5.___;OAK RIDGE NATIONAL I.ABORATORY
. . "jf_-_operuted by
UNION: CARBlDE CORPORATION

- - for the o |
U S ATOM!C ENERGY COMMlSSION

 

   
   
 

  

- e
- G

 

    

L= '35

 

 

~ NOTICE

‘This report contains patentable, preliminary,

‘unverified, or erroneous information. For

one or more of these reasons the author or

| issuing installation and responsible office

- 1-71 have limited its distribution to Governmental
1.7 ] agencies and their contractors as authorized

| by AEC Manual Chapter 3202-062. A formal
report will be published at a later date when
the data is complete enough to warrant publi—

. cation. - : :

 

 

 

 

 

        

: It is subject e
present a final .
eprinted or otherwise: gi:‘v‘:n’e::f::m ;l':
e ORNL pofent brcnch Legal cnd fnfor- e

 

s -'lnformafion ‘is nof. s

o be: -abstracted,

- semination without the npprovn! of ihr
o mcfion Controi Depurtment.

    

ot i i

 

kR L i s
 

@ e e ey

LY LKA g g g

ey ke 5 D

 

 

 

 

 

 

 

 

'L;dii. »ioflce,

Tbis npon was propund os an account of Gov-rnmom sponsoroci work. Noirher the Unmd S!u:u, :

nor the Comminlon, nor any person acting on behalf of the Commission:

A, Mokes any warranty or representation, ‘expressed or implied, with respect to !ho accuracy,"r'fir

completeness, ‘or viefulness of the information contained in this report, or that the use of

_ any information, cppurows, molhod or proc“s dm:lond in this rnpofl may net infnnge

-privataly owned rights; or - — -

B. Assumes any liabilities with respecf |c the use of, or for Jcmogas uwlhng ‘from fl\a use of

any informetion, epparatus, nflhod or process disclosed in this report..

As used in the above, *person ueflng on behalf of the Commission® includes any employes or | ..
“contractor of the Commission, or smployes of such contractor, to the extent that such employse - |-
“or contractor of the Commission, or employes of such contractor prepares, disseminates, or -

provides cccess to, any information wsucnt to’ hls ‘employment of contract with fhe Comm:sslon, .

or his Cmploymum with such conircflor. o

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
 

 

“

w}

ORNL-T-500
Copy

Contract No. W-7405-eng-26

REACTOR CHEMISTRY DIVISION

RADIATION CHEMISTRY OF MSR SYSTEM

Written by: W. R. Grimes

The experiments described in this document required
the cooperative effort of many people in the Reactor,
Metals and Ceramics, Operations, Analytical Chemistry,
and Reactor Chemistry Divisions of the Oak Ridge National
Laboratory. D. B. Trauger, A. R. Olsen, E. M. King,

A. Taboada, and F. F. Blankenship have been in respon-
sible charge of important segments of this work. While
he accepts responsibility for accuracy of the reported
material , the author makes no claim to credit for design
and operation of the experimental program.

DATE ISSUED

‘MAR 13 1963

QAK RIDGE NATIONAL LABORATORY
Oak Ridge, Tennessee
- operated by
UNION CARBIDE CORPORATION
I - for the
U. S. ATOMIC ENERGY COMMISSION

 
 

 

 

 

INTRODUCTION . . .

EXPERIMENT ORNL-MTR-47-4 . . . . . .

Irradiation of Specimens . . .

Post Irradiation Examination .

Analysis of Cover Gas . . .

Gross Examination of Capsules

Examination of the Metal .

Examingtion of Salt . . .

Condition of the Graphite

Conclusions .

- - - . - -

EXPERIMENT ORNL-MTR-47-5 . . « . . .

Behavior Under Irradiation . .

Behavior

Behavior

Behavior

Behavior

Behavior

Conclusions .

with Reactor at Full Power. .

with Reactor Shut Down.

at Intermediate Reactor Power

Through Startup and Shutdown.

of Capsules After Termination

11

. 11

19
22
22
27
27
28

. 33

. 34

37

41

. 4l

. 43
 

 

 

 

)

-9

RADTATION CHEMISTRY OF MSR SYSTEM

INTRODUCTION

Compatibility of molten fluoride mixtures, near in composition to that
proposed for MSRE, with grfiphite and with INOR-8 has been demonstrated
convincingly in many out-of-pile tests over a period of several years. A
few in-pile capsule studies of graphite-fluoride-INOR systems with fluoride
melts of a rather different composition have been performed in past years;
examination of thosé capsules showed no adverse effect of radiation. How-
ever, in-pilé testing of this combination of materials under conditions
similar to those expected in MSRE has been attempted only recently.
Accordingly, no completely realistic experiments have been reported.

An assembly ORNL-MTR-3 containing four capsules, was irradiated in
MTR during the summer of 1961 to provide assurance as to the non-wetta-
bility of grasphite by molten LiF-BeF,-ZrF,-UF,-ThF,; though such assurance
was obtained, this experiment yielded some surprising results.ls2,3,% 1In
brief, the following anomalous behévior Qas observed.

1. The cover gas within the sealed capsules contained appreciable
quantities of CF,.

2. The cover gaé contained the expected quantity éf Kr from all
capsules; the Xe was present at the expected }evel in two, but
its concenfiration-fias less by 100-fold in the other two capsules.

3. While the-INORfB,-Eothas_capsule walls and as small paftialiy

| - immersed teét-speéimens, was eésentially undafiaged, small coupons
of molybdenufi'were-thinned to roughly half the original thickness

with no observable corrosion deposit.

 
 

 

 

 

4. No evidence of wetting of the graphite by the salt was observed. | Ei’

However, the irradiated salt was deep black in colof; it con-

tained loose, roughly spherical beads of condensed salt of

various colors from clear through blue to black.

The difference in behavior of Kr and Xe seemed guite inexplicable;
the behavior.of molybdenum, while not of direct conseqfience to the MSRE,
was disturbing. The black color of the salt was removed by annealing at
temperatures below the liquidus and was shown to be due, largely if not .
entirely, to radiation-induced discoloration of crysfalline LiF and
2LiF+BeF,. This color, and the occurrence of the salt beads (wvhich were
clearly due to condensation on the relgtively cool capsule walls of mate-
rial distilled from the very hot pool of salt) were judged to be trivial.
The appearance of CF, in the cover gas, however, remained a most disturbing
observation. Thermodynamic data for reactions of fluérides with carbon
such as
4UF,; + C = CF, + 4UF;3

suggest that the equilibrium pressure of CF, over such a system should not
exceed 10-8 atmospheres. Moreover, out-of-pile controls with identical
material , geometry and thermal histories (insofar as possible to obtain
same with external heat sources) showed no evidence (> 1 part per million)
of CF,. This gas, accordingly, clearly arose from some previously unknown
radiation-induced reaction. Such generation might, of course, have serious
conseqnences. If CF, were génerated at an appreciable rate in the MSRE
core it might well be removed in the gas stripping section of the MSRE
pump and be lost to the system. If so, the loss of carbon from the moder-

ator graphite would be trivial, but an appreciable loss of fluoride ion g-}
 

 

L 4]

-

from the.melt; with the resultant appearance of Ut3 or other reduced
species in the fuel, would, at the least, cause frequent shutdowns to
reoxidize the fuel; Additional in-pile experiments were, therefore, con-
ducted in an attempt to establish the guantity of CF, to be expected under
power levels, temperatures and graphite-~fuel geometries more representative
of the MSRE. Two assemblies of capsules, designated ORNL-MTR-47-4 and
ORNL—MTF—47-5, have been irradiated in the MTR for this program. While

the examination of these experiments has not been completed, much informsa-
tion has been obtained. The following is a brief statement of progress to
date in this experimental program and a review of the conclusions which

can presently be drawn from the informagtion.

EXPERIMENT ORNL-MTR-47-4
Trradiation of Specimens
The irradiation assembly, designated as 47-4 in the following, con-

tained six INOR-8 capsules. These were immersed in a common pool of
molten sodium which served to transfer £he heat generated during fission
through & helium filled annular gap‘to_water circulating in an external
jacket. The four large cefisules, as shown in Fig. 1, were l-inch diameter
X 2.25-ifiches length-andeonteieea-a core of CGB grephite (l/2—inch diam-
eter x ;-inch leegth)_efibmerged about 0.3-inch (at temperature) in about
25 grams of fuel. The two Smalier capsules, see Fig. 2, contained 0.5-
inch diameter cylindricai crucibleslof CGB graphite containing about 10

grams of fluoride melt. The large capsules included a thermowell which

-permitted measurement'ofitemperature within the submerged graphite speci-

men; detailed analysis suggests that this measured'tempereture should

approximate closely the temperature at the salt-graphite interface.

 
 

 

 

UNCLASSIFIED
ORNL-LR-DWG 67714R

Cr=Al THERMOCOUPLE\fl /NICKEL THERMOCOUPLE
WELL

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

NICKEL POSITIONING LUGS (2)~
NICKEL FILL L|NE>k < - > )’(MCKEL VENT LINE
| INOR-8 CAP
’ A 3
T
HELIUM COVER GAS / L
(3.5 cm?3) - f \ :
; N [Y=—INOR-8 CAN
N
PUNCTURE AREA  / N_A) A
FOR GAS SAMPLING \-_——_—'_: \
N oA \
- N
\ MOLTEN SALT FUEL
E L \ (25¢g)
N
_ N
E L N
N
- - N Cr-Al
A N THERMOCOUPLE
E - — N
b B
—CGB GRAPHITE
12.4 cm? INTERFACE
INOR—8 CENTERING PIN
0 'z {
INCH O NICKEL POSITIONING LUG

Fig. 1. Assembly 47-4 submerged graphite capsules.
 

-

£}

HELIUM COVER GAS

(2.4 cm3)

HELIUM GAS GAP- —

UNCLASSIFIED
ORNL-LR-DWG 67715

  
  

NICKEL POSITIONING LUG

PUNCTURE AREA FOR
N GAS SAMPLING

 

 

MOLTEN-SALT FUEL
(10 g)

 

—CGB GRAPHITE
CRUCIBLE

NN N N N N WA

 

-<— INOR- 8 CAN

 

 

 

 

 

 

O--— NICKEL POSITIONING LUG

 

Fig. 2. Assembly 47-4 graphite crucible capsules.

 
 

 

 

 

 

No such provision for temperature measurement could be incorporated into
the smaller capéules.

The CGB graphite used in all capsules of 47-4 had a surface area of
0.71 m?/g as determined by the BET method. Other properties of this

material follow:

Permeability of a 1.5-in.-0D, 6.56 x 10~% cm® of He (STP)
0.5-in.-ID, 1.5-in.-long specimen per second

Density (Beckman sir pycnometer) 2.00 g/cm?

Bulk density 1.838 g/cm?

Bulk volume accessible to air 8.7%

Total void volume as percentage of 19.5%
bulk volume

Spectrographic analysis of the graphite revealed only the usual low
levels of trace elements; some of these may have been introduced during
machining and handling operations.

The UF,; used in preparing the salt mixtures was fully enriched in all
cases. The four large (submerged graphite) capsules and one of the small
capsules each contained fluoride mixtures of LiF-BelF,-ZrF,-Th¥,-UF,; for
which analyzed samples indicated the composition 71.0-22.6-4.7-1.0-0.7
mole %; analysis for Ni, Cr, and Fe in the material showed 25, 24, and 97
parts per million. The other small capsule was loaded with a very similar
mixture but with the uranium content raised to 1.4 mole %; the Ni, Cr, Fe
analyses were 15, 40, and 123 ppm. The fused salt mixtures were prepared
by mixing the proper quantities of pure fluorides, and then, at 750 to
800°C, sparging for 2 hours with H,, 8 hours with a 5:1 mixture of H, and
HF', and.48 hours with Hj.

The large capsules were filled by transferring the molten fluoride
mixture under an atmosphere of He; the level to which they were filled was

controlled by blowing excess liquid back through a dip-line adjusted to
 

)

“‘

0

9

the proper level. Tfi;‘ievel of salt was mofiifored with a television x-ray
system. Final closure was made in a helium-filled glove box where the ends
of the fill-tubes were crimped and then welded shut. This technique should
have affected neither the composition nor the amount of cover gas within
Ithe capsule.

The smaller capsules were filled, in the glovebox, with ingots of
solid fuel mixture. They were closed by inert-gas arc-welding of the end
cap. This welding operation, which used argon as the electrode cover gas,
raised the capsule temperature gpprecigbly and permitted mixing of some
argon with the helium within the capsule before thg seal fias effected.
Accordingly, neither the amount of gas nor the He:A ratio within the
sealed capsule at time of closure can be certified. |

The amount of fuel chosen for each capsule should have yielded a
vapor volume of 3.5 cm? in the large, and 2 cm? in the sfiall, capsules at
tempergture. The filled and segled capsules were x-rayed and examined to
insure that the molten salt had flowed to its proper position in the
assembly.

Assembly 47-4 wes irradiated thfough three MTR cycles in the period
March 15 to June 4, 1962. The'tempeiafure history of the six éapsules (as

read from the chromel-alumel thermocofiples'in the four large capsules and

as calculated for the two small capsules) is shown in Tsble 1. The maximum

measured temperature among’the four instrumented capsules was 1400 * 50CF.

The‘mean measured temperatures of the other three instrumented capsules

were approximately 1380, 1370, and 1310°F, respectively. The correspond-

ing calculated INOR-8 Capsule.walIQto-salt interface temperatures were

1130, 1125, 1115, and 1110°F. The temperature of the surface of the INOR-8

 
 

 

10

Table 1

Temperature History of Fuel Salt in Capsules from 47-4

 

Time et Temperature (hr)

 

“

 

 

Temperature Interval Submerged Graphite Graphite®
(°c) Core be  Crucible

(24, 36, 45, 12, 6°) AN

Steady-state operation
0-100 390.6 390.6
100-700 44 .0 ' 12.1
700-750 1203.7 3.8
800-850 11.4
850-900 1456.9
Total 1892.3 1892.3
Nonsteady-state operation 51.7 - 51.7
Total irradiation 1553.4 1553.4

 

 

 

&Calculated temperature history based on temperatures measured in submerged
graphite capsules.

bGraphite crucible capsule containing fuel with 0.7 mole % UF,.

Ceapsule identification number.
 

-}

¥

-)

11

end cap in contact with the fuel salt was calculated to be the same as
that.of the graphite-to-salt interface temperature. The area of the end
cap wetted by the salt was abofit 12% of the total 5.6 in.? of metal sur-
face in contact with the salt.

The accumulated time during transient operation ineludes only times
for temperature changes of more éhan about 30°C. Of the 121 such temper-
ature changes recorded, 60 included decreases to or below the solidus tem-
perature of the fuel salt. It is estimated that the fuel freezes within
five minutes after shutdown of the MTR, and cooling to below 200°F should
occur within half an hour.

Exposure data for the capsules in 47-4, and those for out-of-pile
controls which were given as similar a thermal history as was practicable,
are summarized in Tables 2 and 3.

The temperatures were controlled from Capsule 24, and the changes in
the retractor position required to hold a constant temperature were not

unusual. The thermocouple reading for capsule 45 appeared to drift down-

- ward by about 35°C during the 12-week exposure, but this was the only

symptom of deviant temperature recdrdings.
Post Irradiation Examination
| The assembly was rembved.from.the_MTR_and partially disassembled at
that loéation for shipment £6 ORNL. The assembly was completely dismantled
in ORNL'hqt cells énd the capsules were recovered for complete examination.

A1l dismentling operations went smoothly, and no evidence of failure of

. any‘capsule was observed.

Analeis of Cover Gas

The six irradiated and two unirradiated control capsules were

 
 

 

et -t e s —p < A G B

 

Table 2. Exposure Data for Large Capsules in Assembly 47-4

 

 

 

 

 

 

 

Thermal-Neutron Fast-Neutron Temperature (°C)
Weight of Urenium Content Flwx® Based (>3 Mev) Flux Power  Calculated
Capsule Fuel on Co80 Based on Co”® Graphite- INOR-8-to- Density Burnug
(g) Mole & Activation Activation to-Salt Selt (w/cm®) (% U23°)
ole g (neutrons/cm?-sec)  (neutrons/ecm?®-sec) Interface Interface?
x 1012 x 1012
3¢ 25.532 0.7 0.993 750
ge 26,303 0.7 1.023 750
12 25,17 0.7 0.979 2.10 2.1 6804 610 67 5.5
24 25.37% 0.7 0.987 2.70 3.2 7602 605 83 7.0
36 24.886 0.7 0.968 2.7 3.3 yal 595 85 7.0
45 25.598 0.7 0.996 3.85 3.8 7od 600 117 9.7
aAverage external neutrqn flux,
bEstimated temperatures.
CUnirradiated controls.
d'I‘hermocouple readings prior to termination of final irradiation cycle.
Table 3. Exposure Data for Small Capsules in Assembly 47-4
: Thermal-Neutron Fast-Neutron
Weight of  Urenium Content Flux® Based (>3 Mev) Flux TemperatuTe . Power  Calculated
Capsule Fuel e ————— on Co%° * Based om Co”8 Re iona Density Burnug'
(g) Mole % Activation Activation (§b) (w/em®) (% U233)
: : g (neutrons/cm?-sec) (neutrons/cm?- sec) :
x 1013 x 1012
lg ' 9.381 0.7 0.365 . 750
3 6.805 0.7 0.265 750
5 9.829 1.47 0.737 . 895 ' .
4 10.101 1.47 0.758 4.79 5.2 895¢ 260 11.4
6 9.915 0.7 0.386 1.31 1.3 7.5¢ 43 1.3

 

afiyerage external neutron flux.
bUnirradiated controle.
thermocouple readings prior to termination of final irradiation cycle.

A

 
 

 

 

 

 

 

")

R

13

punctured and the covér gas was recovered for analysis in the interval
August 8 to October 23, 1962. These operations, conducted in the hot cells
of Bldg. 4501 at ORNL, used a screw-driven puncturing tool sealed with a
bellows and by & neoprene O-ring which butted on a flat previously ma-
chined into each capsule (see Figures 1 and 2). Gas escaped into a
collection system whose volume was calibrated and which had, in each case,
been evacuated and checked for leaks. The initial collections were per-
formed with a mercury-in-glass Toepler pump and glass éample bulbs in a
glass and metal system; when this proved inadequate, as described below,
it was réplaced by an all-metal gas collection system without a pump.

This metal system was conditioned with elemental fluorine according to
wéll-established profiedures before use.

The first observations were made on the small capsule (No. 6) which
had suffered the least burnup of uranium and on its out-of-pile control
(No. 5). These capsules were handled without difficulty with the glass-
metal system., After collection of the gas, the void volumes in the cap?
sules were determined by pressure drop on expansion of helium from a known
volume. Data obtained, including analyses of the gases by the mass spec-
trometer, are shown in Table 4.-:The puncturing device leaked slightly,
présumably atlthe O;ring séal,.during opening of irradiated Capsule 6.

After correction of the analyses for the quantity of air (47%), the total

volume of He + A agreed very well with that from the out-of-pile control.

The Xe and Kr were.r6coverediat'verj_nearly the expected ratioc, and the

absolute volume bf_these_gasesragreed reasonably well with the quantity

(0.35 em? total) calculated from the burnup shown in Table 3. The only
radiocactive materigls recovered on Toepler pumping from Capsule 6 were Kr

and Xe.

 
 

 

 

14

Table 4

Cover Gas Analysis for Capsules 5 and 6 of Assembly 47-4

 

Capsule 5 : Capsule 6

 

 

Gas collected: 1.3 cm? Gas Collected: 3.2 cm’

Gas volume in capsule: 2.7 cm’® Corrected velue*: 1.7 cm’
, Gas volume in capsule: 2.5 cm?

 

 

 

 

 

Gas Quantity Quantity
Volume Volume
% (cm3) g% (cm3)
He 15 0.2 23 0.3
1.2 1.2
A 79 1.0 51 0.86
0 0.7 0.01 - -
CO + N» 5.3 0.07 - -
Kr 0 0 4.2 0.0
0.4
Xe 0 0 19.0 0.32 '

 

% Corrected for leakage of air (as indicated by mass-spectrometric analysis)
at time of sampling.

&)

"
 

 

 

=}

»)

15

Cover gas recovery and analysis from all other capsules revealed very

different behavior.

Capsule No. 24, the first of the large capsules to be opened, showed
the cover gas to be quite‘reactive even toward the Hg in the Toepler pump,
and truly quantitative analysis of the gas was precluded by the inadequacy
of the gas handling system. 1In spite of the considerable but unknown loss
of gas by reaction, a large quantity (84 em? at STP) of gas was collected.
The product of reaction with the Hg was shown to consist entirely of Hg,F;.
Unlike the gas from Capsule 6 this material was quite radioactive; Te
activity.(presumably was volatile TeFg) wae primarily responsible. Since
the measured free volume in Capsule 24 at ambient temperature was shown to
be 4.6 cm?®, a pressure of at least 18 atmospheres existed within the cap-
sule at the time of puncturing.

Based on mass spectrometry of two 0.1-cm? samples, the gas from Cap-
sule 24 was 5% He, 0.1% Xe, 0.4% Kr, 4% (Co + N,), 6% CO,, and 17% CF,.
The remainder was O, and SiF, in roughly equal amounts, matching the pro-
ducts of the reaction of F, with the glass

810, + 2F; —» SiF, T + 0, T.
The krypton yield listed above is in fair agreement with the calculated
value, but the xenon yield is far too low. Later samples obtained by long

exposuré of the capsule to an evacuated 3-liter container, had 0.3 cem’® of

- xenon and less than one-tenth that'amount of krypton.

All other capsules in this series were opened into an essentially all-

metal, preconditioned:gas systenm, which1permitted much more representative

‘samples of the gas to be obtained. Table 5, for example, shows data ob-

tained from Capsule 36, the near duplicate in irradiation conditions of

Capsule 24.

 
 

B L 1T T TS, ] 10 1 1 bS8 om0 1 11 W 0 80 A AR 1.0 O

 

16

Tfible 5

Analyses of Cover Gas from Irradiated Capsule 36

 

 

Volume of gas space in capsule at room temperatuie: 4.6 cm3
Volume of gas space gt operating temperature: 2.6 cm’®
Volume of gas removed from capsule: 188 cm?
Gas Quentity (vol %) Volume (em?)
Fa 83.5 156

CF,, 10.0 | 19

0 2.5 4.7

Xe . 0 0

Kr 0.3 0.56

He 2.8 5.3

CO + Ne 0.1 0.19

002 007 lo3

")

 
 

 

 

n)

ol

17

By contrast, Capsule 3A, which was an unirradiated control for
Capsules 24 and 36 and which had been thermally cycled in a manner similar
to their in-pile history, yielded 4.5 cm® (STP) of 99.7% He, 0.05% of
CO + N, and 0;2% COp. The lack of A and the totai pressure of almost
precisely 1 atmosphere was expected because of the speed and simplicity
of the final closure of capsules of this type; this analysis undoubtedly
represents the gas composition of Capsules 24 and 36 before irradiation.

The pertinent data for all irradiatéd capsules in 47-4 are shown as
Table 6. Inspection of this information shows that all capsules except
No. 6 (previously described) yielded large quantities of F, and consider-
able quantities of CF,. The total gas yield from Capsule 45 indicated a
pressure of 60 atmospheres at room temperature before opening; had this
gas been present at the operating pressure and in the considerably
smaller volume then available it must have exceeded 250 atmospheres. It
is almost certain that the capsule could not have withstood such a
pressure even had the gas been inert.

The large volume of gas drastically reduced the concentration of Kr
and other minor constituehts, and the resctive nature of the samples
required use of-special maés spectrometers whose precision was not of the
highest Quality. Both'these*factors_ddversely.affect the precision and
accuracy of the analyses.i It appears, however, that most, if not all, the
Kr was recovered'from alléémples while the Xe is recovered poorly, if at
all, from allrcapsules excépf Nb.'é. Though no positive proof is avail-
able, it seems highiy probable that Xenon récovery fails because of forma-
tion of .XeF;* (or pérhaps of 'ot’hér fluorides or oxyfluorides) which ma.y—be

strongly adsorbed on the fuel or graphite within the system.

 
 

 

 

Quantity and Composition of Gases from 47-4 Capsules

 

Small Capsules

 

Large Capsules

 

Capsule No. 6 36 45 12
Fuel Content (g) 9.915 24..886 24,598 25.174
Uranium Content (g) 0.386 0.968 0.957 0.979
Burnup (%) 1.3 7.0 9.7 5.5
Sampling Date 8/6 9/7 10/16 10/23
Gas Collected (em?® STP) 1.6 188 270 153
Gas Composition (%)
P, 0 84.1 88.7 48.1°
CF[' 4-3 9-3 8-1. I*l-l-l
Xe 9.0 0 0 0
Kr 4.1 0.4 0.8 0.2
He + A 2.6 2.7 2.95 4.0
Other - 3.5 0017 6-6

 

@Minimum value ;s some gas lost in reaction with container.
Corrected from originel analysis which showed SiF, + 0,.

"

)

 
 

 

")

19

While it is possible to see that the production of F, (or F, + CF,;)
generally increases with quantity of uranium burned, no quantitative cor-
relation of yields of either or both together with any of the experimental
parameters can be shown. Moreover, the datea does not permit any decisions
as to effect of cooling time on yield of these materials. It is worthy of
note that the Fp + CF,; produced in the most extreme case (Capsule 45)
répresents 3% of the fluorine contained in the salt or 1.6 times the
quantity of fluorine in the UF, in that sample. It is clear that these
data indicate very large‘losées of F, from the salt.

Gross Exafiination of Capsules

Visual examination 6f the capsules after removal from the 47-4
assembly showed them to be sound and in good condition. No dimensional
éhanges were apparent.

All capsules have been sectioned to permif direct observation of the
metal, graphite, and salt. Figs. 3 and 4 show typical views (of Capsules
24 and 45) obtained after rough polishing of the sawed specimens. Several
features in these views are noteworthy. The fuel salt is black in color,
but seems to'shqw_no unusualrtendengy to érack or shatter during the saw-
ing process. N§ evidence0fwétting of the graphité by the salt is
apparent. Nonwetting of tbe”metal by the salt is 6ccasionally obéerved;
this behavior, which isjUnfisual,,seéms-to be due to a layer of "scum" on
the salt surface near the_capéule_fiall.'

One bilt of curious béhavior, not matched,by the but-of-pile controls,
is the oécurrence of'smoothééuffaced voids, which have appérently been
gas bubbles during high temperature operation, in the bottom hemi spher-

ical section of all the large capsules.

 
 

o
N

 

 

UNCLASSIFIED

11213

 

 

 

after exposure.

2

-4

MTR~47

2

Capsule 24

Fig- 3.

 

 

 

 

 
-

UNCL ASSIFIED
- HCO-1866

[
P
-
‘.
i
]
b

 

'Fig; 4. Capsule 45 3 MTR-4’7-4, after exposure.

 

 

 

 

 
 

 

 

 

 

22

A

Examination of the Metal | ' ki;
Fluorine at elevated temperatures is an extremely corrosive material. |

Fig. 5, for example, affords comparison of an untreated specimen of INOR-8

with one after exposure for 22 hours at 1100°F to flofiing helium containing

1% (by volume) of F,. An evaluation of the corrosion of the capsule metal

should permit definitive statements as to whether F, was produced and was

present during the high temperature operation'of Assembly 47-4. Accord-

ingly, a very careful metallographic examination has been made of sections

)

from Capsule 24, and selected specimens from other capsules are under study.
Especial attention has been paid to metal surfaces exposed in the gas phase
as well as at the gas-salt interface and fiithin the molten salt. Typical
photomicrographs from Capsule 24 are shown as Figs} 6 and 7.

There is no evidence of attack on the INOR-8 at any portion of Cap-
sule 24; while definitive data are not yet available for other cafisules,
their appearance makes it highly unlikely that evidence of attack will be

observed. (No attack would be expected in an Out-of-pile capsule test of

¥

this duration at this general temperature level.) Further, there is evi-

dence (see Fig. 7) that wall thickness has not been reduced, so the

(w

unlikely possibility of uniform corrosion followed by leaching of the
corrosion film by the molten salt can also be excluded. It can, apparently,
be stated with certainty that none of these capsule walls was exposed to
fluorine at high temperature for apprecigble periods of time.

Examination of Salt

 

Small samples of salt from the irradiated capsules have been examined
carefully with the optical microscope to determine the nature of the crys-

talline material. These materials are in general, and as is frequently the i j
 

 

 

 

 

X004 . X004 o
q o] ) | | = o w| | O
e L CIHONI GEO'Q ————e——reerrrerrerrrema ) e S3HINI S£0'0 ————————msmnane ] m o
w3 O f
=3 8 0
<% ? -
o> —
z . o
= St %
23
' o...wl_ -m
+
. 0w o
g 8
S o
o
3 ¥
‘ 8 |
- 5
. m* S
o QL HH
9 -
o
. o By
. . et O
- 83
- n | . ,
0 i
: w g
o
m Q
T h
L N
2 . .
wy
| 35
, d“of ~
o g
¥
. _ , o3
&5
, o m
. & . ' ¢ % { €

 
 

 

i
i
1

ok

ETCHED

Fig. 6.
of capsule 24.

 

 

 

 

CONTROL

Longitudinal section

24

UNCLASSIFIED
R-12958

 

 

 

 

ETCHED  |RRADIATED

of INOR-8 from hemispherical portion

 

ah

 

=~
-

 
 

 

 

 

 

 

25

 

UNCL.ASSIFIED
R-11624

 

 

INCHES

0.02

 

 

 

 

Fig. 7. Transverse section of INOR-8 control material and specimen

from hemispherical portion of capsule 24, MTR-47-4. Reduced 8%.

 

 
 

 

 

 

26

O

case for rapidly cooled specimens, too fine grained to be positively iden-
tified. However, a sampie from near the graphite at the center of Capsule
24 has been shown to contain the normally expected phases, though some are
discolored and blackened; these phases included 7LiF-6Th¥, which was shown
to contain UF, (green in color) in solid solution. It seems certain,
therefore, that the uranium was not markedly reduced at the time the fuel
was frozen. |

An attempt was made, with Capsule 35;.to remove the salt by melting
under an inert atmosphere td provide representative samples for chemical
analysis and, especially, for a determination of the reducing power of the
fluorine-deficient salt. This melting operation was not successful since
only about one-half of the salt was readily removed in this way. That
portion which melted and flowed from the capsule was green in color and
obviously contained some tetravalent uranium. Since the salt was highly
reduced--2% of the fluorine, approximately equivalent to that as UF;, had

been lost--melting the material would be expected to result in deposition

¥)

of some metal; this metal may have resulted in a high-melting heel of un-

certain composition.

n

A variety of chemical analyses have been performed on materials from
these capsules. No evidence of unusual corrosion behavior is observed in
any of the analyses.

Salt specimens have been examined by gamma ray spectrometry to eval-
fiate gross behavior of fission product elements.» In general, the ruthenium
activity appears to be concentrated at near the bottom of the capsules;
this may indicate its existence largely in the metallic state so that a

gravity separation could occur. Cerium as well as zirconium-niobium { J

+
 

 

 

27

activity appears, as expected, to be fairly evenly distributed in the fuel.
Some cesium, with some Zr-Nb, activity is found on metal surfaces exposed
to the gas phase.

Condition of the Graphite

Graphite specimens from all the capsules‘have been recovered and
examined visually with low power microscopes. No evidence of damage can
be observed. The surface of the graphite appears identical to unirradiated
specimens, the specimen in all cases appears structurally sound, and the
saw cuts (see Figs. 3 and 4) appear normal. An autoradiograph of the
graphite f}om Capsule 36 indicates quite uniform activity over the specimen
with no visible evidence of fuel penetration. No quantitative measurements
of physical and structural properties have been completed.

Conclusions

Of the six capsules in Assembly 47-4, five were found to contain
large quantities'of F, and CF,; while the sixth, which had received a con-
siderably smaller radiation dose, showed a small quantity of CF, but no F,.
This latter capsule alone yielded the expected quantity of xenon while all
others retained this element nearly quantitatively. The quantity of F~
released from the fuel melt génerally increased with increase in uranium
fissioned (or any of the consequences of this) but no quantitative correla-
tion was found. Neither -the. quantity of CF, nor the ratio CF,:F, showed an
obvious correlation with fisSiOn"rate; burnup, or time of cooling before
examination. | |

The appearanée of tfie o§ened'capsules, and"all completed examination
of cdmponents'from these, seem to preclude the possibility that the F, was

present for any appreciable period during the high temperature operation of

 
 

 

 

28

this assembly. Accordingly, the generation of F, from the salt at low tem-
peratures during the long cooling period must have occurred.

Considerable energy is, of course, available from fission product
decay to produce radiolytic reacfiions in the frozen salt, and the quantity
of energy increases with burnup of uranium in the fuel salt. Capsules as
small as those can absorb only a sfiall fraction of the gamma energy re-
leased by fission product decay. On the other hand, a substantial fraction
(perhaps as much as one-half) of the beta energy would be absorbed. A .
yield (G value) of less than 0.04 fluorine atoms per 100 electron volts
absorbed would suffice to produce the fluorine ohserved in any capsule.
Production of F, with such a low G value seems & reasonsble hypothesis
though no observations of fluorine production from salts on bombardment by
B irradiation have been published. Attempts to demonstrate F; generation
by eiectron bombardment of fuel mixtures at ambient temperatures are under
vay.

EXPERIMENT ORNL-MTR-47-5

Assembly ORNL-MTR-47-5 was designed, after the observations of CF, in
the gas space in capsules from 47-3 referred to sbove, to include six cap-
sules of INOR-8; four of these capsules contained graphite specimens such
that large variation in salt-graphite interface area and in the ratio of
graphite area:metal area in contact with salt was provided. The other two
capsules, whose behavior is described below, were generally similar to the
large capsules of Assembly 47-4. However, these capsules included addi-
tional vapor space above the molten salt, and each was provided with inlet
and outlet gas lines extending to manifold systems beyond the reactor. It

was possible, accordingly, to sample the cover gas above the melt during

)

(n

a
 

 

 

29

reactor operation as well as during and after reactor shutdown.

The two purged capsuies, referred to as A and B in the following and
illustrated in Fig. 8, are 0.050-in. thick INOR-8 vertical cylinders, l-in.
diameter x 2.25-in. long‘with hemispherical end caps. Each capsule contains
a core of CBG graphite (see preceding éection) l/2-in. diameter x l-in.
long submerged to a depth of approximately 0.3 in. in sabout 25 grams of
salt. The top cap contains two l/4-in. diameter purge lines, é l/8-in.
diameter thermocouple well, and a projéction to keep the graphite sub-
merged. The two capsules differ only in the composition (primarily in the
uranium concentration) of the fuel salt. Table 7 shofis the nominal compo-
sition of the salt mixtures in each capsule.

Preparation of the salt, and handling and filling operétions were very
similar to those described for Assembly 47-4 above.

The six capsules were, as was the case in previous experiments, sus-
pended in a tank filled with sodium which served to transfer the thermal
energy from the capsules to the tank wall. Heat was removed from the
sodium tank through a variable annulus fiiled with helium to cooling
water flowing in an external jacket. As in past experiments, no auxiliary
heating was provided. The capsule wall afitained temperatures of about
600°C during high level power'operation.

The twin manifolds which,comprise the pressure monitoring and gas
collection systems were modefately compiex. - A schematic diagram of one of
the assemblies is shéwn.as Fié.ié. .The equipment was éarefully calibrated
to insufe that thesémplesdrafin as desired into 200 cc metal sampiing
bulbs includéd a large and known fraction of the gas contained in the

capsule.

 
 

 

 

 

30

UNCLASSIFIED
ORNL—-LR—DWG 71570OR

INOR—-8 THERMOCOUPLE WELL _*

   
 
 

   
 
  
  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

INOR-8 INOR-8
COVER GAS COVER GAS
PURGE LINE PURGE LINE
| ) \\ _
) \
=3 N\
INOR-8 CAP \
\,—.-_
FOR N / (3 cm”)
GAS SAMPLING 7 7
Y
N . N—A
A— ; =
A ) T
INTERFACE AR A _;'I:; ::/ MOLTEN-SALT FUEL
NT EA - Kl HINRSKY—-- Y -
ssad) 2RSSt  es e
cm AR TR
Bl A1
Bk ~p
INOR-8 CAN—" [}~ »’0’0’& 1/
/ RS HPSRRS 7
'''' AL ey ==~
A - RN XY - - - -
/ R s /
CHROMEL-ALUMEL [}~ — KX Jegied— Y Sen srapiTe
THERMOCOUPLE—="FN (%] —— 1/ INTERFACE AREA
»,020:0’«! }:0}:: Y (12.5 cm?)
qd RN
- - RIS =1
LRI

INOR—8 CENTERING PIN

)
== &
& 2%
~—rt :.r/ z’:
S
>

 

NICKEL POSITIONING LUG

Fig. 8. Assembly 47-5 purged capsules.
 

31

Table 7. Fuel Composition

 

Composition, mol %

 

 

Fuel | Capsule A Capsule B
LiF 6'7.36 67.19
BeF, 27.73 27.96
ZrF, 4,26 .51
UF, - 0.66 0.34
| Specific gravity at 1200°F 2.13
Liquidus 84.2°F

Thermsl conductivity at 1200°F  3.21 Btu/hr-ft2.0F/ft

Specific heat at 1200°F 0.455 Btu/1bme°F

 

 
 

 

32

UNCLASSIFIED
ORNL-LR-DWG 77838

 

 

——
m— X XXX XK X

i
—P--XXXXX:

 

 

—b
- XXX X i

 

|-
|

 

 

COUPLING

    

—— MOLTEN SALT

e

 

SUBMERGED GRAPHITE

 

 

 

 

 

 

 

 

SAMPLE BOTTLES
CARBON TRAPS

xxxoooc CAPILLARY TUBING
PR PRESSURE RECORDER

Fig. 9. ORNL-MTR-47-5 gas collection system.

-
XX XXX XXX DRIER HELIUM

TO MTR
OF F-GAS

 

 

 

 

o
 

 

 

 

®

!

33

Unfortunately, this experimental assembly was designed and con-
structed before the caPSUIes-of Assembiy 47-4 were opened. Accordingly,
the entire complex was designed to handle CF, but not F,. In the short
time between recognitidn of F, in the previous capsules and the insertion
of 47-5 into the MTR, a number of materials changes were made to improve
the ability to handle fluorine. For example, valves seated with fluorocar-
bon plastic were substituted for’less desirable types in the assembly, and
sample bulbs of preconditioned nickel were substituted for the original

ones of stainless steel. However, some stainless steel tubing, which could

not have been replaced without a complete dismantling and reconstruction

operation requiring several months, was retained in the assembly. Short
lengths of this tubing, which were welded to the INOR-S purge tubes within
the sodium tank, attained the temperature of the sodium bath. Stainless
steel contains CF, quite well, but it is not at all resistant to fluorine
at elevated temperatures. Acco?dingly, evolution of F, with the assembly
at high temperature might have been obscured by reaction of F, with the
stainless steel. Observations below suggest very strongly that no such
effect occurred.
Beha#ior Ufider-Irradiation

Assembly MTR447-5 was carried'through five MTR qycles during the four
and one-half months ending January 23, 1963. Opefating conditions.for this
assembly were varied as desired throughout the'test to include conditions

anticipated-in'different regions of the MSRE. Temperature and power level

~ in Capsules A and B were'Variedfindependently over g considerable interval

- but wete, insofar aS~possible3‘fiéldrconstant during periods represented by

accumulagtion of the gas samples. Fuel temperaturés during reactor

 
 

 

 

 

 

34

operation varied from about 190°F to 1500°F with power levels ranging from
3 watts/ce to 80 watts/cc within the fuel.

Gas samples were taken by isolating the capsule for a known and pre-
determined interval (6 to 96 hours) and then purging the accumulated gases
into the sample bulb. Pressure was carefully monitored (to the nearest
0.1 psi) while the capsule was isolated. A total of 59 gas samples were
taken from the two capsules during the fifst'four MTR cycles under five
distinct sets of conditions within the capsules. These sets of conditions
(and numbers of samples) are:

1. Reactor at full power (40 Mw) with fuel molten, (34),

2. Reactor at intermediate power (5 to 20 Mw) with fuel frozen, (8),

3. Reactor shutdown with the fuel at about 90°F, (9),

4. Periods spanning fuel melting at reactor startup, (4), and

5. Periods spamning freezing of fuel following reactor shutdown (4).

In addition, after termination of the experiment on January 23, the
assembly was removed from MTR, the purge lines were capped with valves,
and the assembly was quickly shipped to ORNL. Pressure monitoring of Cap-
sules A and B at ambient temperature in the hot cell is under way, and
samples of the evolving gas have been analyzed.

Behavior with Reactor at Full Power

Pressure Monitoring of the capsules with the reactor at 40 Mw and the
fuel molten has shown no case of perceptible pressure rise. Moreover, none
of the 32 gas samples taken during such conditions (over power levels rang-
ing from 3 to 80 w/cc) have shown evidences of F,. [The presence of heated
stainless steel in the lines might, as stated earlier, conceivably be
responsible.] Analyses of gases collected from Cepsules A and B during

the first MTR cycle are shown in Tables 8 and 9. These analyses, which are

typical of all that have been collected under these conditions, show
 

 

 

 

 

 

 

 

 

 

Table 8. Analysis of Gas Samples from Capsule A, MTR-47-5 During First MTR Cycle
Fuel Composition Fuel Weight 24.822 grams
(mole 7 Graphite Weight  5.696 grams
LiF  68.0
UF, 0.7
Accumu-
lation | _ Power Gas Analysis
Time Temp. . Level Percent (by Volume) Parts per Million
(hr.) (°F) (W/ee) He N, +CO O» H,0 A CO» H, Kr Xe CF,
29 1050 13 99,2 0.45 0.00 0.04 0.0L 0.11 0.15 24 95 <l
R 1050 13 99.5 0.31 0.00 0.10 0.00 0.06 0.03 3 18 <1
50 1248 33 96.7 2.54 0.03 0.20 0.04 0.40 0.00 77 439 <l
24 1250 33 98.0 1.34 0.00 0.22 0.02 0.29 0.05 39 248 <1
24, 1380 65 98.9 0.40 0.00 0.10 0.18 0.34 7L 405 0
96 1400 57 98.5 0.60 0.00 0.05 0.0L 0.20 0.44 222 1762 <2
16.3% 90 - 99.1 0.63 0.00 0.07 0.0L 0.09 0.11 14 22 7
*Unscheduled Seram: 4 hrs. 17 min. at 33 watts/ce, 12.5 hrs. at zero power.

Ge

 

 
 

Table 9. Analysis of (Gas Samples from Capsule B, MTR-47-5 During First MIR Cycle

 

 

 

 

 

Fuel Composition Fuel Weight 25.240 grams
(mole 7 Graphite Welght 5.587 grams
LiF 68.0
BeF, 27.0
ZI'F4 4.65
UF, 0.35
Accurmu- Gas Anslysis
lation Power .
Time Temp. Level ‘ Percent (by Volume) | Parts per Million
(hr.) (°F) (W/ee) He N,+C0 0,  H0 A €0, H, Kr Xe CF,
29 900 7 96.9 2.18 0.31 0.08 0.03 0.40 0.08 6 28 <1l
R 900 7 97.8 1.70 0.37 0.09 0.02 0.04 0.02
50 1100 17 94.1 4.65 0.70 0.20 0.06 0.26 0.00 16 96 0
24 1090 17 95.0 3.9 0.61 0.19 0.05 0.25 0.00 6 - 56 0
24 11200 34.6 62.5 29.26 5.95 0.27 0.35 0.28 0.14 10 140 0
96 1230 30 96.7 2.54 0.03 0.20 0.04 0.40 0.00 77 439 <1l
16.3% 90 . - 97.1  2.26 0.27 0.02 0.03 0.19 0.00 4 36 0

 

*Unscheduled secram: 4 hrs. 17 min. at 17 watts/cc, 12.5 hrs.

at zero power.

9¢

 
 

 

£y

)

*

&

37

considerably more Xe than Kr (the proper ratio is sbout 6:1) and very
little if any CF;.

Qualitative spectral analysis of gamma rays penetrating the nickel
sample bottles have shown Xel33? to contribute the only detectible peak
under these collection conditions.

Behavior with Reactor Shut Down

Capsules A and B operate, by virtue of their differing uranium con-
tents, at power levels which differ by about a two-fold factor (see Tables
g8 and 9). At the time of the first long MTR shutdown Capsule A had been
fissioning at a rate corresponding to about 15.5 days at 35 w/cc.

Shutdown of the MIR deprived the assembly of heat; the capsules
cooléd to below the fuel solidus in a few minutes and to 90°F within per-
haps 90 minutes. Six and one-half hours after shutdown a perceptible
pressure increase was observed in Capsule A while some 36 hours elapsed
before the increase was observed in Capsule B. Rate of pressure increase
decreased with time for both capsules but'showed no evidence of reaching
equilibrium during the 4-day shutdown. On subséquent shutdowns, Capsule A
always showed the pressure increase with its onset 2 to 8 hours after shut-
down; Capsule B showed pressure increases only after the first and fourth
shutdowns. In all‘cases_CaPSule'B showed smaller rates of rise than
Capsule ‘A.

Gas samples taken during shutdowns when pressure rises were apparent

have shown much more radidactivify_than those taken when the assembly was

132 and.

at high temperature; spectral ahalysis of the gamma rays show Te
T132 a5 primarily responsible with Xe*33 as & minor contributor.

Handling, transporting, and analyzing of these samples were

 
 

 

38

complicated by the considerable radioactivity; long waiting periods and
some dilution of the samples have been required. Accordingly, F, has been
observed in only one of the samples for which pressure rise data and the.
increased radioactivity strongly suggest it should be present. In that
sample, however, 5.8% of F, and no detectible CF, was observed. No CF,
has been observed in any sample accumulated during & reactor shutdown.

Tt is of interest, since this information strongly suggests the
evolution of F, at low temperature by radiolysis of the solid fuel, to
compare the shape of the pressure rise curves with the shape of the decay
energy release curve for B-emitting fission products.

Calorimetric data’ on fission-product decey energy release were used
since energy absorbed, rather than total energy released, was of prime
coficern. No correction was made for gamma energy absorbed from the in-pile
environment. By interpolation between the l-week and l-month curves in
ANL-4790, the values of P4/Py for 15-1/2 dasys of pile operation were
obtained and plotted vs cooling time. This curve was graphically inte-
grated to obtain the curve of total energy absorbed as a function of
cooling time. Multiplication of the ordinate by P, (capsule power during
irradiation) yields

t

[ P_at

o B
the total energy absorbed to a given cooling time, t. In order to chéck
whether the shape of the pressure vs time curve corresponded to the shépe
of the energy vs time curve, a series of energy release curves were pre-

pared by multiplying the original ordinates by appropriate factors,

leaving the time scale unchanged. These scaling curves, the original curve

, O
 

)

&

39

represented on different energy scales, were then compared with the
pressure rise data plotted on the same time scale. Two particular calcu-
lated lines fit the pressure rise data very well when the origins of the
energy curves were shifted, as shown in Fig. 10. The energy scale used

is such that 1 scale unit = 31,500 joules = 1.97 x 1017 Mev. The ratio of
the ordinate of the Capsule B line to that of the Capsule A line is 0.57
at any time, corresponding well to the estimated ratio of Po's for the

two capsules (0.5 %= 0.1).

The excellent fit of the pressure points to the energy lines may be
partly due to the elements of arbitrariness in selecting energy scales and
the positions of the origins. However, a reasonable case may be madg for
the main assumptions in the treatment given. First, reasonable G values
are obtained (see below), justifying the choice of energy scale. Second,
the observed induction period before pressure rise suggests that F, formed
in the early part of the radiolysis is consumed by easily accessible
reduced materials in the environment and/or is contained in the crystal
up to a ceftain capacity. These effects would be represented graphically
by moving the energy curve origin downward below that of the pressure rise
curve. Third, a time delay effect may be operative, as F, formed in the
fuel crystals at a certain rate diffuses slowly through the crystal and
appears later, but at the earlier rate. Graphically this effect would
require moving the energy curve origin to the right. Since the best fit
wa.s obtainedby moving_the'energy curve origins both downward and to the
right, it is suggested thét both capacity (or consumption) and time delay

effects are operative.

 
 

 

 

UNCLASSIFIED
ORNL-LR-DWG 77160

 

     
  
 

 

 

 

 

30 = o CAPSULE NO. 4 PRESSURE 14
o CAPSULE NO. 3 PRESSURE
— 14
25 |—
g e
w 20 — .
0 o
@ =
u W —10
E 45 — ? 0
3 - 2
b‘:’ g 2
a =18 w
10— & 5
(vl w
5 | &
uz.: -6 O
0.5 — o
o
10 of
A/ 1, &
O — ] 2 L
0 80 90 100
| | 0
0 10 20 30 40 50 60 70 80 90 100
| | | | | | | o

 

 

0 10 20 30 40 50 60 70 80 90 100
COOLING TIME (hr) '

Fig. 10. Comparison of observed pressure rise in capsules from MIR-47-5 with energy‘.
released in fission product decay.

 
 

 

 

&)

4

»)

4

41

Behavior at Intermediate Reactor Power Levels
The design of Assembly 47-5 allowed considerable variation of capsul.e
temperature and neutron flux (fission power). However, the adjustment
fange would not permit maintenance of the fuel below the melting point
with the MTR at 40 Mw. It was of interest to observe whether intermediate
MTR power levels (5 to 20 Mw) with frozen fuel would produce Fp. Accord-
ingly, such experiments were performed for short time intervals énd with
the assembly set at the fully retracted position with minimum gas gap.
Gas pressure within the capsule was monitored and gas samples were taken
for analysis. Data obtained from these experiments are shown in Table 10.
Examination of this data indicates that F, generation may have

occurred in Capsule A with the MTR at 5 Mw and the salt temperature at
about 200°F and, perhaps, in Capsule B with the MIR at 20 Mw and the salt
at 410°F. The fact that the solid salt can withstand such energy releasés
at these relatively low temperatures seemé quite reassuring. ©Since ¥,
generation at 95°F (see Section above) occurs with much less energy avail-
able, it would appear that the "recombination" phenomenon is most sensitive
to temperature.
Behavior Through Startuprand Shutdown

| The last entries in Tables 8 and 9 show data for samples taken near

the end of the first MTR cycle for a 1l6.4-hour period which included an

unscheduled MTR scram so that the sample collected represents 4 hours 17

minutes at power and 12}5_hours at zero power. The gas in Capsule B

seems quite normal. The gas sample from Capsule A is the only one in

which a real concentration (7 ppm) of CF,; has been observed; it may be

observed as well that the Xe:Kr ratio appears to be low. However, the

 
 

42

Table 10. Data from 47-5 with MIR at Intermediate Power Levels

 

 

Capsule MIR Power Estimated Capsule

Capsule Na Bath®

 

 

Number Mw Power Density Temp. Temp. Resultéb
w/ce OF OF
A 20 10 610 378 No pressure rise
B 20 6 510 378 No pressure rise
A 10 5 312 200 &P =0, 0% F,
B 10 3 285 200 AP =0, 0% F,
A 15 - 7.5 380 260 AP =0, 0% Fp
B 15 4.5 350 260 AP = 0, O% F
A 5 2.5 199 122 AP = +0.43 psia’
B 5 1.5 180 122 AP =0
A 20 9 480 303 AP = -0.1 psia
B 20 5 410 303 AP = +0.2 psia

 

 

a'I'he sodium bath temperature is the highest temperature occurring in the
stainless steel tubing leading out from the capsules.
negligible attack of stainless steel by 1% F, at temperatures below 40C°F.

hr of constant temperature operation.

Tests indicate

bPressure readings, probably accurate to 0.1 psi, were taken during 2 to 3

cThe pressure rise occurred over a period of 65 min; for the subsequent 2
hr of the 5-Mw operation there was no further pressure rise.
 

&

4

43

radiocactivity of the gas sample (through the Ni bottle) was low and its

activity pesks were not unusual.

Behavior of Capsules After Termination of MTR Exposure

Assembly 4145 was removed from the MIR and was detached from the
manifold sys?em; the purge lines from Capsules A and B were closed off by
valves. After transfer to ORNL and removal of the assembly to a hot cell
the capsule lines were attached to manifolds. Capsule pressures were
measured, samples were taken, and the very radioactive gas was pumped -

from the capsules into chemical traps. The evacuated capsules were iso-

lated in sections of the manifold contalning pressure gauges and were

allowed to stand at ambient temperature (70°F). Generation of gas was
immediately apparent in both capsules; pressure increases from Capsule A
were larger than those from Cepsule B. Rate of generation decreased

appreciably during the first 72 hours but seemed nearly linear thereafter.

-Analysis of the gas shows it to contain high concentrations of F, with

considerable concentrations of 0, (but little N,) suggesting that air
in-leakage is small but that reaction of ¥, with oxide films on the
apparatus is still appreciasble. Small quantities of COz, COF,, and OF;

are also observed, but virtually no CF; is found. The capsules are still

“under dbservation, and a series of experiments to ascertain the effect of

- temperature on the rate of Fz'generation is beginning.

The'quantity'of F, generated from Capsule A since the start of exper-
iments at ORNL is estimated (February 12) to exceed 160 cm® (STP). This

quantity plus that préviousl&*rembved in samples, purges, etc. during

| reactor'shutdowns,'now“approaches'the quantities observed in the most

highly irradiated capsules from MTR-4.

 
 

 

 

 

 

L4

None of the capsules from MTR-5 have yet been opened for.examination;.
at present, it is expected that another 40 to 60 days will elapse before -
examination begins. Meanwhile, F, generation from Capsules A and B will
be examined as a function of temperature.

Conclusions

All data from Assembly MTR-47-5 indicate that neither F, nor apprec-
iable concentrations of CF, are generated when fission occurs in the molten
salt. It can be stated with certainty that F, is released from the irrad-
iated salt on standing at temperatures below 100°F; the analyses show,
however, that little, if any, CF, is released from such samples. Genera-
tion of F, is apparent at these low temperatures only after an appreciabie
time. This time delay does not seem to depend entirely on extent of
irradiation; in Capsule B it clearly exceeded 100 hours on both the
second and third MTR shutdown but was only 36 hours in the first shutdown.

No auxiliary evidence, such as pressure rises, Xe:Kr ratios, or radio-
activity of the sampled gas, suggests difficulty when the salt is at power.
All of these phenomena, however, are observed when the irradiated salt is
maintained at low temperatures.

Behavior when the salt is frozen but fission is occurring is less
clearly defined since few data are available. It appears, however, that
no gas generation occurs when the salt is above 400°F. Out-of~pile studies
indicate that F, deliberately added to specimens of reduced (fluorine
deficient) fuel reacts rapidly at temperatures well below the melting
point of the salt. It seems very likely that "recombination" processes
are sufficiently rapid at temperatures above 350°C to prevent any loss of

F, from the salt.

y

*
 

Yy

45

The virtual nonappearance of CF, at any of the conditions experi-
enced in MTR-5 is most encouraging. It may suggest that the CF, found in

previous capsules from MTR-3 and MTR-4 arose in the following manner: F,

“was generated in those capsules during some or all of the MTR shutdown

periods; this F, was not removed from the sealed capsules and on subse-
quent heatup by reactor power it reacted indiscriminantly with the fuel,
the metal, and the graphite. The CF,; formed was partially retained and
was observed at termination of the test.

The complex events taking place in the experiments summarized here
are not understood in detail. It cannot, for example, be demonstrated
with certainty that the graphite is without influence on the phenomena
observed. It is, however, apparent that the generation of F; is a low
temperature process occurring only under circumstances of relatively
little importance to the MSRE and that generation of CF,; occurs to a
negligible extent, if at all, at MSRE temperatures and power levels.

Additional effort will be required to evaluate possible difficulties
in the frozen flange and the freeze plug concepts. Such studies are
under way. It seems clear, however, that the phénomena of CF,; and ¥,

generation should not have important effect on the operation of the MSRE.

‘References
1. "MSRP Semiann. Prog. Rep. Feb. 28, 1961," ORNL-3122, p. 108.

2. "MSRP Semiamn. Prog. Rep.Feb..28, 1962," ORNL-3282, p. 97.

3. "Reactor Chem. Div. Am. Prog. Rep. Jan. 31, 1962," ORNL-3262, p. 19.

4. "MSRP Semiann. Prog. Rep. Aug. 31, 1962," ORNL-3369, p. 105.

5. é. Untermeyer and J. T. Weills, Heat Generafion in Irradiated Uranium,
ANL-4790 (Feb. 25, 1952).

 
 

 
 

 

P

ovoo~Toumdwnn e

i
o

12.
13.
14,
15.
16.

18.
19.
20.
21,
22.
23.
24 .
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.

36. .

37.
38.
39.
40.

41,

42.

mw?zwspupuzfiwwmdmwwummbw

HunhEHQQ

G.
J.
E.

. B.

R.
A.
H.
L.
J.
R.
P.

. K.

BI

. R,

G.
H.
H.
N.
C.
P.

* P.

J.

. M.

S.

. W

B.
N.
G.
B.
J
K.

R,
. L.
. R.
. R.

. Affel

Alexander
Beall
Bettis

. Billington

Blankenship

. Blumberg

Bohlmann
Borkowski
Boyd
Briggs
Bruce
Conlin
Cook
Crovley
Ditto
Engel
Epler
Ergen
Gallaher
Grimes
Grindell
Guymon
Harley
Haubenreich
Hise
Holz
Jarvis
Kedl
King
Kirslis
Krewson

-Lindauer

Lyon
MacPherson
MeDonald
McDuffie
MeGlothlan
Mixon
Moore
Olsen

Payne

47

Internal Distribution

 

43.
Z14‘.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
7.
58.
59.
60.
6l.
62.
€3.
64-65.
66-67.
68-70.
7L.
72-"73.

74=75.
76.

77-82.

83.

84.
85-98.

J. L. Redford

M. Richardson

R. C. Robertson

J. BE. Savolainen

D. Scott

C. H. Secoy

J. H. Shaffer

M. J. Skinner

A. N. Smith

P. G. Smith

I. Spiewak

J. A. Swartout

A. Taboada

J. R. Tallackson

R. E. Thoma

D. B. Trauger

W. C. Ulrich

G. M. Watson

A. M, Weinberg

B. H. Webster

J. C. White

Central Research Library
Document Reference Section
Laboratory Records
Laboratory Records (LRD-RC)
Reactor Division Library

External

D. F. Cope, Reactor Division
AEC, ORO

A. W. Larson, Reactor
Division, AEC, ORO

H. M. Roth, Division of Research
and Development, AEC, ORO

W. L. Smalley, Reactor Division,
AEC, ORO

J. Wett, AEC, Washington

DTTE