ORNL-4658.txt

ORNL-4658
UC-80 — Reactor Technology

CHEMICAL ASPECTS OF MSRE OPERATIONS

Roy E. Thoma

OAK RIDGE NATIONAL LABORATORY
operated by

UNION CARBIDE CORPORATION
for the

U.S. ATOMIC ENERGY COMMISSION

DISTRIBUTION OF THIS DOCUMENT IS UNLIMITED

Printed in the United States of America. Available from
National Technical Information Service
U.S. Department of Commerce
5285 Port Royal Road, Springfield, Virginia 22151
Price: Printed Copy $3.00; Microfiche $0.95

This report was prepared as an account of work sponsored by the United
States Government. Neither the United States nor the United States Atomic
Energy Commission, nor any of their employees, nor any of their contractors,
subcontractors, or their employees, makes any warranty, express or implied, or
assumes any legal liability or responsibility for the accuracy, completeness or
usefulness of any information, apparatus, product or process disclosed, or
represents that its use would not infringe privately owned rights.

DISCLAIMER

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, completeness, or
usefulness of any information, apparatus, product, or process
disclosed, 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, manufacturer, or
otherwise does not necessarily constitute 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.
Contract No. W-7405-eng-26

MOLTEN-SALT REACTOR PROGRAM

ORNL-4658
UC-80 — Reactor Technology

CHEMICAL ASPECTS OF MSRE OPERATIONS

Roy E. Thoma

REACTOR CHEMISTRY DIVISION

NOTICE
This report was prepared as an account of work
sponsored by the United States Government, Neither
the United States nor the United States Atomic Energy
Commission, nor any of their employees, nor any of
their contractors, subcontractors, or 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 disclosed, or represents that its use
would not infringe privately owned rights,

DECEMBER 1971

OAK RIDGE NATIONAL LABORATORY
Oak Ridge, Tennessee 37830
operated by
UNION CARBIDE CORPORATION
for the
U.S. ATOMIC ENERGY COMMISSION

DISTRIBUTION OF THIS BUC

¥
.

CONTENTS
ABSTRACT
EXECUTIVE SUMMARY
ACKNOWLEDGMENTS
KEY WORD INDEX
I INTRODUCTION

2 CHEMICAL BEHAVIOR IN THE FUEL AND COOLANT SALT SYSTEMS DURING
PRENUCLEAR OPERATIONS

2 1 Preoperational Procedures
2 2 Flush Salt
2 3 Coolant Salt

24 Fuel Salt
2 4 1 On Site Preparation
2 4 2 Uramum Assay
2 4 3 Structural Metal Impurities
2 4 4 Oxide Contaminants
24 5 Analysis of Helium Cover Gas
24 6 Lithium Analysis
2 4 7 Examination of Salts after Zero-Power Experiment
2 4 8 Appraisal of Chemical Surveillance in Prepower Tests

3 CHEMICAL COMPOSITION OF THE FUEL SALT DURING NUCLEAR OPERATIONS
31 Introduction
3 2 Component Analysis
33 Oxide Analysis
3 4 Uranium Concentration
3 5 Structural Metal Impurities
3 6 Chemical Effects of Reprocessing

3 7 Materal Balances for 2°*U and 2?2 U Operations
371 Recovery of 23°Uand ?3%U
3 7 2 Inventores for Stored Salts
37 3 Salt Loss from Leakage

111

10
14

15
15
16
19
26
30
31
31
33

36
36
50
50
52
53
55

57
57
58
59
10

11

12

1V

CHEMICAL COMPOSITION OF THE FLUSH SALT DURING NUCLEAR OPERATIONS
4 1 Role of Flush Salt Analysis in the Determination of Salt Residue Masses
4 2 Transfer of Uranum and Plutonium to Flush Salt in >*3 U Operations
4 3 Flush Salt Loss to Off-Gas Holdup Tank

CHEMICAL BEHAVIOR OF THE COOLANT SALT
51 Composition Analysis

5 2 Corrosion Behavior

CORROSION IN THE FUEL CIRCUIT
6 1 Modes of Corrosion
6 2 Corrosion in Prenuclear Operations
6 3 Corroson in Power Operations
6 4 Additions of Reductants and Oxidants to the Fuel Salt
6 5 Effect of Uranium Trifluoride on the °° Nb Concentration of the Fuel Salt

DETERMINATION OF REACTOR POWER
7 1 Power Estimates with 23°U Fuel from Heat Balance and Other Methods
7 2 Power Output of the MSBR Based on the Isotopic Composition of Plutonium
7 3 Isotopic Composition of Uranum during 233U Qperations

7 4 Isotopic Composition of Uranium during 22 °U Operations

PHYSICAL PROPERTIES
8 1 General Properties
8 2 Denstty of Fuel and Coolant Salts
8 3 Crystallization of the MSRE Fuel

INTERACTIONS OF FUEL SALT WITH MODERATOR GRAPHITE AND
SURVEILLANCE SAMPLE MATERIALS

CHEMICAL SURVEILLANCE OF AUXILIARY FLUID SYSTEMS

101 Water Systems
1011 Cooling Tower Water
10 1 2 Treated Water Supply
10 1 3 Vapor Condensing System

10 2 Hehum Cover Gas
10 3 Reactor Cell Air
10 4 Oil Lubncation Systems

TRANSPORT OF MATERIALS FROM SALT TO COVER GAS SYSTEMS
11 1 Fission Products

11 2 Restrictions in the Off Gas System

11 3 Tritwum Transport in the MSRE

IMPLICATIONS OF THE MSRE CHEMISTRY FOR FUTURE MOLTEN SALT REACTORS

61
61
64
64

65
66
69

71
71
74
74
79
94

99
99
101
108
110

112
112
112
116

118

121

121
122
122
123

127
128
128

131
131
132
138

139
CHEMICAL ASPECTS OF MSRE OPERATIONS
R E Thoma

ABSTRACT

In this report are tabulated all resuits of laboratory
analyses performed in surveillance of MSRE salt, water,
cover gas, and o1l systems Excepted are analytical data
pertaining to fission product and trittum distribution
and transport The report recapitulates conclusions
derived from chemical analyses performed from the
1964 preoperational test period until termination of
power operations 1n 1969, modified by the results of
postoperational examination of the reactor compo-
nents Surveillance results were evaluated with respect
to their significance as indicators of the performance of
the MSRE and as indicators of the need and potential
for development of specific in-line methods of analysis
for molten-salt power reactors

As judged from chemical data, the MSRE was highly
successful as a materials demonstration The flowing
salts did not wet their containment systems, fuel salt
netther wetted nor penetrated the graphite moderator
surface The cumulative generalized corrosion within
the fuel circurt resulted in the removal of chromium
from the alloy to an average depth of 04 muil, while
that in the coolant system was undetectably low The
results of postoperational examinations, although cor-
roborative of predicted corrosion, also indicated finite
but shight intergranular attack In operations which
successively employed 235U, 223U, and 23°Pu as
sources of power 1n the reactor, the circulated fluids
remained chemically stable, free of radiation damage,
and free of contamination The average full-power
output of the reactor, as computed from experimental
results of 1sotopic dilution mass spectrometric analysis
of fissile species and subsequently confirmed by cap-
ture-to-absorption ratio measurements, was shown to be
7 4 MW(t)

EXECUTIVE SUMMARY

The Molten Salt Reactor Experiment (MSRE) was
conducted durnng the period from 1965 to 1969 as the
first extensive demonstration of the operability of
molten-salt reactors Durnng this period, continuous
survetllance of the chemical behavior in the circulating
fluid salt, water, cover gas, and o1l systems was
maintained through a program of laboratory analysis
The principal function of the program was to ensure
that the Experiment would proceed with the freedom

from chemucal problems that was anticipated from the
results of prior supporting research and development
programs The results of chemical analyses were also
used for assistance in developing operational plans for
nuclear engineering experiments with the reactor In
these experiments, the reactor was employed to a
hmited extent as an experimental chemical facility to
obtain chemical data that were not otherwise available

Al of the results of laboratory analyses (except the
mass of data on fission product and tritium distribution
and transport) performed 1n surveillance of the MSRE
salt, water, cover gas, and o1l systems are summarized 1n
the current report In this archive record are recapitu
lated various conclusions derived from the laboratory
analyses performed from 1965 to 1969 as modified by
the results of postoperational examination of the
reactor components

Surveillance results were evaluated with respect to
their significance as indicators of the need and potential
for development of specific in-line methods of analyses
for molten-salt power reactors Examination of metal-
graphite assemblages removed periodically from a flow
channel 1n the graphite moderator confirmed inferences
from chemucal data that materials compatibility was
excellent

This report 1s divided into chapters that pertain to the
chemical behavior of flush, coolant, carrier, and fuel
salts 1n the prenuclear operational period, the chenical
composttion of the fuel and flush salts during nuclear
operations, the results of corrosion surveillance, power
estimates from chemical and 1sotopic dilution data, and
the results obtained from analysis of samples from
auxiliary fluid systems

Experience with the MSRE throughout the shake-
down periods preceding power operations confirmed
that the molten fluoride salt mixtures were intrinsically
noncorrosive to Hastelloy N and that effective proce-
dures were employed to prevent serious contamination
of the salt circuits during this period

Upon mitiatton of power operations, orifices in the
fuel off-gas system became restricted Investigation
showed that the cause was organic material o1l that
seeped nto the fuel pump Improved filters successfully
alleviated the plugging problem, but the continued
passage of hydrocarbons through the pump constituted
a chemucal factor that introduced a degree of un-
certainty to some interpretations of chemical behavior
in the MSRE
In operations which successively employed 23°U,
233y, and 2??Pu as sources of power in the reactor,
the circulated fuel salt remained chemaically stable, free
of radiation damage, and essentially free of contami-
nation.

While chemical data alone were useful as statistical
indicators of trends in the concentration of fissile
species 1n the fuel salt over extended penods of power
operation, they were of secondary importance 1n
day-to-day operations because on-site reactivity balance
measurements proved to be some ten times more
sensitive to changes 1n the concentration of fissile
material than the mdividual chemical results The
combined results of chemical and mass spectrometric
analysis, however, furnished information that was
uniquely suited to use in establishing numerous abolute
values and were applicable for determination of trends
m performance, in computation of inventory, and in
establishing the distnbution of uranmwum between the
fuel- and flush-salt systems

After three years of operations with 2352380 fyel
salt, uranium was removed from the carnier salt in
preparation for tests of **>*U as a fuel for molten-sait
reactors Uranmium was removed from the carrier salt by
fluorination, the uranwm hexafluoride product was
absorbed on NaF beds Recovery of the uramum from
the NaF absorber beds yielded less uramum than
expected A painstaking investigation was made which
led to a refinement in the material balances of
fissionable species in the reactor from the outset of
operations From these 1t was deduced that the dis-
parity was caused by retention of 0.8 kg of 233U (2 48
kg ZU) 1n the chemical reprocessing facility at the
MSRE.

Fuel salt was prevented from becoming increasingly
oxidizing as burnup of fissionable material proceeded
by the addition of small amounts of beryllium metal to
the salt flowing through the pump bowl. The results of
these experiments showed that the disposition of > Nb
between the containment materials and the salt could
be used as an indicator of the freedom from our
development of a potentially oxidizing condition 1n the
fuel salt.

Disposal of gaseous tritium emanating from the
MSRE posed no radiological hazard, consequently, no
program for completely defining 1ts distribution was
instituted at the outset of operations After recognition
of the importance of tritium control 1n large molten-salt
reactors, studies of tritium in the MSRE were actively
pursued. Results of these studies are described in other
MSRP reports and are not treated extensively here.

As judged from chemical data, the MSRE was highly
successful as a materials demonstration. The flowing
salts did not wet their containment systems, fuel salt
neither wetted nor penetrated the graphite moderator
surface. Chemical analyses showed corrosion with the
2L1F-BeF, coolant system to be neghgible. This has
been borne out by subsequent examination of the salt
side of the tubes from the air-cooled radiator and of the
coolant side of the primary heat exchanger. Sumilar but
more numerous analyses suggested that corrosion with-
i the fuel system was slight (but observable). The
cumulative generalized corrosion within the fuel circuit
resulted 1n the removal of chromium (the most chemi-
cally active constituent of the alloy) from an average
depth of 0.4 mul, some ten tumes less than was
anticipated from the preoperational laboratory mea-
surements of self-diffusion coefficients of chromium n
Hastelloy N. It 1s inferred that the major fraction of this
corrosion resulted from interactions of atmospheric
oxygen retained in the graphite moderator after periods
of reactor maintenance.

Postoperational examination by metallographic tech-
niques confirmed the low generalized corrosion but
disclosed a grain-boundary effect near surfaces exposed
to the fuel which resulted 1n cracks to a depth of one
grain 1n strained specimens. This hitherto unobserved
phenomenon 1s currently being mvestigated.

A salient conclusion from the chemical studies de-
scribed 1n the current report 1s that development of
automated 1n-line methods for determination of redox
potential ([U**]/[ZU]) of fuel salts, for dynamic
assessment of corrosion rates, and for measurement of
the presence of oxides at low (<50 ppm) concen-
trations 1n flowing salt will be required for operation of
larger reactors

Operation of the MSRE served to demonstrate the
practicality of the molten-salt reactor concept, its
safety, reliabihty, and tractability to simple mainte-
nance methods, These operations confirmed that the
molten fluonide salts are immune to radiation damage,
equally serviceable with varnous fissie species as energy
sources, and tolerant of the buildup of fission and
corrosion products The MSRE thus fulfilled its role
and demonstrated chemical compatibility of matenals,
stmple refueling and reprocessing of salts, and the
potential need for automated in-line analysis as part of
the operational controls system in molten-salt reactors.

ACKNOWLEDGMENTS

The data recorded in this document were obtamed
through the efforts of a large number of people in
several Divisions of the Oak Ridge National Laboratory:
Analytical Chemistry, Chemistry, Chemical Tech-
nology, Metals and Ceramics, Operations, Reactor, and
the Solid State Divsions. We are especially indebted to
the staff of the Analytical Chemistry Division, for only
through the enthusiastic commitments of these
chemists to the needs of the Molten-Salt Reactor
Program was it possible to obtain the excellent quality
of chemical information pertaining to MSRE operations
that we now have. The efforts of this group were led by
R. F. Apple, R. E. Eby, J. M. Dale, C. E. Lamb, A. S.
Meyer, Jr., and W. F. Vaughan, with the administrative
support of L. T. Corbin, I. C. White, and M. T. Kelley.
We were assisted as well by E. M. King and his
associates in hot-cell examinations of specimens from
the MSRE.

Counsel and advice was provided throughout the
period of reactor operations by members of the MSR
program staff, =~ M. W. Rosenthal,  R. B. Briggs,
P. N. Haubenreich, J.R. Engel, R.H. Guymon, and
B. E. Prince, and by the chemists who have been my
associates in the program. Of this group W. R. Grimes,
E. G. Bohlmann, F. F. Blankenship, and H. F. McDuftfie
supplied special assistance and provided significant
contributions to the conclusions offered here,

KEY-WORD INDEX

*analytical chemistry + *burnup + *chemical properties
+ *chemistry + *corrosion products + *fluorides +
*fuels + *fused salts + *MSRE + *power measurement
+ *primary salt + *reactors + *sampling + *secondary
salts + *single-fluid reactors + *surveillance + *uranium
fluorides + *uranium-233 + *uranium-235 + beryllium
fluoride + coolants + cover gas + fuel preparation +
graphite + Hastelloy N + inventories + lithium fluoride
+ MSRP + nickel alloys + oxides + oxidation + oxygen +
phase equilibria + plutonium fluorides + primary system
+ reactor vessel + zirconium fluoride

1. INTRODUCTION

During the last two decades a great number of reactor
concepts have been proposed to fill the foreseeable
need for electric power toward the end of the century
and to conserve supplies of fissionable materials. Of
these concepts, only a few remain of potential signifi-
cance to the nuclear economy. Foremost among this
group are the Liquid-Metal Fast Breeder (LMFBR), the
Gas-Cooled Fast Breeder (GCFBR), the Light Water
Breeder (LWBR), and the Molten-Salt Breeder (MSBR).

The initial efforts to develop the molten-salt system
began more than 20 years ago at the Oak Ridge
National Laboratory. A detailed examination of the
program which followed is described in a series of
papers published early in 1970.! By 1964 development
of MSR’s had culminated in the construction and
operation of the Molten-Salt Reactor Experiment
(MSRE) as a demonstration of the practicality of these
reactors. It was designed to employ, as nearly as was
feasible, the same materials that were proposed for use
in molten-salt breeder reactors. Thus the MSRE was
constructed to circulate uranium fuel as UF, dissolved
in a molten fluoride mixture within a Hastelloy N
circuit. The fuel mixture was pumped at a rate of 1200
gpm through a graphite core matrix contained in a
cylindrical core vessel (Fig. 1.1). Dry, deoxygenated
helium was supplied at 5 psig to the pump bowl. A flow
of this gas carried xenon and krypton out of the pump
bowl to charcoal beds.

When the reactor was operated at full power, fuel
entered the graphite core at 632°C (1170°F) and was
heated to 654°C (1210°F). The salt was then dis-
charged through the shell side of a tube and shell heat
exchanger, returning through a fuel inlet to the reactor
vessel. A coolant salt circulated through the heat
exchanger, through the air-cooled radiator to the
coolant pump, and back to the heat exchanger to
complete the circuit. At full power, the temperature of
the coolant salt varied from 546°C (1015°F) to 579°C
(1075°F) in this circuit. Design parameters of the
MSRE are summarized in comparison with those for
larger molten-salt reactors in Table 1.1. Here it is noted
that two fuel salt compositions were employed in the
MSRE. The reactor was operated initially with a 235U
fuel charge; the uranium from this charge was recovered
and replaced with 233U for the latter period of reactor
operations. Nuclear characteristics of the MSRE with its
2351 fuel charge are listed in Table 1.2.

From the inception of operations with the MSRE in
1965, the performance of the MSRE was positive
indication of the technical feasibility of molten-salt
reactors. The MSRE has shown that a molten-fluoride
reactor can be operated at temperatures above 1200°F
without attack on either the metal or graphite parts of
the system; that reactor equipment in the radioactive
parts of the plant can be repaired or replaced; and that
xenon can be stripped continuously from the fuel.

Operations with the MSRE were terminated as plan-
ned late in 1969. The reactor was operated for a
cumulative period of 13,172 equivalent full power
hours during the period of nuclear operations from
June 1, 1965, to December 12, 1969. A chronological
ORNL-LR-DWG 6i09TRIA

FLEXIBLE CONDUIT TO
CONTROL ROD DRIVES

GRAPHITE SAMPLE ACCESS PORT

COOLING AIR LINES

ACCESS PORT COOLING JACKETS
ha

FUEL OUTLET REACTOR ACCESS PORT

SMALL GRAPHITE SAMPLES
- HOLD-DOWN ROD

OUTLET STRAINER

CORE ROD THIMBLES
LARGE GRAPHITE SAMPLES
CORE CENTERING GRID

FLOW DISTRIBUTOR -
VOLUTE

GRAPHITE - MODERATOR \
STRINGER

FUEL INLET /

_ T~- CORE WALL COOLING ANNULUS
REACTOR CORE CAN —

REACTOR VESSEL —1

ANTI-SWIRL VANES -

VESSEL DRAIN LINE - MODERATOR

SUPPORT GRID

Fig. 1.1. MSRE reactox vessel.
Table 1.1. Comparnson of MSRE, MSBE, and MSBR design data’

% MSRE MSBE MSBR
Reactor power, MW( 1) 73 150 2250
Peak graphrie damage flux 3x 1013 5% 1014 3x 1014

(E, > 50 keV),

neutrons cm 2 sec ™!

Peak power density, W/cc
Primary salt 30 760 500

Core including graphite 66 114 65
Peak neutron heating in 02 26 17
graphite, W/cc
Peak gamma heating 1n 07 63 47
graphite, W/ce
Primary salt
Volume fraction 1n core 0225 015 013
Composttion, mole %
TLiF 65 (64 5)°
BeF, 29 2(302) 715 717
ThE,4 0(0) 16 16
235 238yp, 0 83 (0) 12 12
233UF, 0 (0 14) 05 03
Zr¥, 502 None None
~ Liquidus, °C 434 500 500
Liquidus, °F 813 932 932
Density, Ib/ft3 at 1100°F 141 211° 210
Viscosity, Ib 171 hr ™! at 1100°F 19 29¢ 29
- Heat capacity, Btu/1b°F 047 032 032
Thermal conductivity, 83 075 075
Btu hr™? £t oE !
Volumetric heat capacity, 66 66 66
Btu ft ™3 °F !
Temperature, °F
Inlet reactor vessel 1170 1050 1050
Outlet reactor vessel 1210 1300 1300
Cuculating primary salt volume, ft> 70 266 1720
Inventory fisstle, kg 76 (32)b 396° 1470
Power density primary salt 4 20 46

cairculating average, W/cc

Y rom J R McWherter, Molten Salt Breeder Experiment Design Bases, ORNL-TM 3177, p 3
(November 1970)
Figures n parentheses refer to the second fuel loading, contamning 223 UF,
€206 at 1300°F, 212 at 1050°F
916 4 at 1300°F, 34 2 at 1050°F
€2334 1taal
Table 1.2. Nuclear charactenstics of MSRE with 235U fuel

Thermal neutron fluxes,a neutrons cm "2 sec”!

Maximum 379 x 1013
Average in graphite moderated regions 148 x 1013
Average n circulating fuel 474X 1012
Reactivity coeffiaentsb
Temperature, (°F) ! 77% 107
235y concentration 0253
Fuel salt density 023
Graphite densty 053
Prompt neutron lifetime, sec 28x%x 107

At operating fuel concentration, 7 4 MW

bAt mafial cmfical concentration Where umits are shown,
coefficients for variable x are of the form (1/k)/(8k/ax), other
coeffictents are of the form (x/k)/(8k/ax)

history of reactor operations 1s summarized in Fig 12
Detailed accounts of these operations are described 1n
the Molten-Salt Reactor Program semmiannual progress
reports

In operation, the MSRE employed three salt mix-
tures fuel, coolant, and flush salt (that was used to
scavenge mpurities from the fuel circuit and from
surfaces of the graphite moderator before and after the
fuel containment system was opened)

Fuel and coolant salts for use in the MSRE were
selected on the basis of considerations which are
discussed 1n detail 1n an earlier report > The fuel salt
consisted essentially of a carrier muxture into which
suitable amounts of fissile material could be dissolved
to produce fuel salt The carrier was selected to be a
maxture of 7LiF BeF,-ZrF, such as to provide the
optimum physical and chemical properties of the fuel
salt Some of the criteria included 1n optimuzing the
carrier composition were liqudus temperature, vis-
costty, and zircomum (included to ensure that UO,
could not be precipitated from the molten fluoride
solution) concentration The phase diagram of the
LiF-BeF,-ZrF, system®** shown in Fig 1 3 illustrates
the options available for choice of carrier salt The salt
compostiron selected on the basis of the considerations
mentioned was ' LiF BeF,-Z1F, (65-30-5 mole %)

An additional measure was adopted 1n the selection of
salt composition of the fuel to minimize the possibility
that troublesome deposits containing fissionable ma-
terial might segregate from the molten-fluoride fuel
solution This was the choice to constitute the uranium
fuel charge of about two-thirds 2°®U and one-third
235U, 1t was based on chemical considerations, and
arose from uncertainty as to the probable value of the
oxidation-reduction potential that would prevail n the

fuel salt 1n normal operations From Long and Blanken-
ship’s* results, it was concluded that the dispropor-
tionation of UF, would not proceed to the extent that
the amount of metallic uranmmim produced would
precipitate or cause problems by alloying with the fuel
salt containment system If, for some reason, however,
the [U*}/[ZU] concentration ratio were to rise above
50%, formation of uranmum alloys and carbides was
foreseen as possible This difficulty was recognized 1n
the 1nexactness of our information concerning the value
of the average total cation-anion balance that would
result from one fission event in the reactor environ-
ment If the tendency was toward a slight excess of
cations, the potential for reduction of U* — U3" and
disproportionation was increased Increasing the total
inventory of uramum would reduce proportionately the
rate of development of unfavorable {U**]/[ZU] con-
centration ratios For these reasons, the choice was
made to specify that the concentration of uranium in
the fuel salt would be 09 mole %, even though <0 3
mole % of highly enniched uranmium would have been
sufficient to make the MSRE cntical Such measures, 1t
was shown, were overly conservative in affording
protection which 1t 1s recognized now was not essential
Notwithstanding, they, like the meticulous operational
methods which were employed, comprse the margins
of safety which were appropriate to the experiment
Under simlar considerations, the coolant salt was
chosen as a muxture corresponding to the fuel composi-
tion but contamning neither fissile material nor zur-
conum Salt of the composiion ’LiF-BeF, (66-34
mole %) was used both as the coolant and as the flush
salt

Samples of the MSRE fuel mixture and (less fre-
quently) the coolant mixture were analyzed routinely
durning all periods when salts were circulated m the
reactor On each occasion of 1ts use, the flush salt also
was analyzed The concentrations of the salt constit-
uents, oxide contaminants, and fission product species
were momtored on a continung basis Chemical
analyses were performed regularly with samples re-
moved from the circulating salts in order to evaluate the
utility of a contmuous surveillance program as well as
to fix goals for in-line analytical controls for future
molten-salt reactors The MSRE provided the imtial
experience 1n these respects, although a molten-salt
reactor, the Aircraft Reactor Experniment,® previously
demonstrated the operability of molten-salt reactors,
the scheduled period of its operation was brief and did
not include a program of chemical analysis For the
MSRE, however, we sought to demonstrate through a
long period of operation the stability of such reactors,
SALT IN
FUEL LOOP

POWER

DYNAMICS TESTS

INVESTIGATE
OFFGAS PLUGGING

REPLACE VALVES
AND FILTERS

RAISE POWER
REPAIR SAMPLER
ATTAIN FULL POWER

CHECK CONTAINMENT

g gy T ST

FULL - POWER RUN

-— MAIN BLOWER FAILURE

REPLACE MAIN BLOWER
MELT SALT FROM GAS LINES
REPLACE CORE SAMPLES
TEST CONTAINMENT

RUN WITH ONE BLOWER
INSTALL SECOND BLOWER

-

ROD OUT OFFGAS LINE
CHECK CONTAINMENT

30—dey RUN
AT FULL POWER

} REPLACE AIR LINE
DISCONNECTS

SUSTAINED OPERATION
AT HIGH POWER

REPLACE CORE SAMPLES
TEST CONTAINMENT

} REPAIR SAMPLER

SALT IN

FUEL LOOP POWER

J

.

erchan et P
g S e g

L.

e
gt

ORNL-DWG 69— 7293R2

XENON STRIPPING
EXPERIMENTS

MAINTENANCE

} INSPECTION AND
REPLACE CORE SAMPLES

TEST AND MODIFY
FLUORINE DiSPOSAL
SYSTEM

PROCESS FLUSH SALT
PROCESS FUEL SALT

LOAD URANIUM-233
REMOVE LOADING DEVICE

233, 7ERO - POWER

' PHYSICS EXPERIMENTS

INVESTIGATE FUEL
SALT BEHAVIOR
CLEAR OFFGAS LINES

REPAIR SAMPLER AND
CONTROL ROD DRIVE
233 DYNAMICS TESTS

INVESTIGATE GAS
N FUEL LOOP

HIGH-POWER OPERATION
T0 MEASURE 233y 4 /o,

REPLACE CORE SAMPLES
REPAIR ROD DRIVES
CLEAR OFFGAS LLINES

INVESTIGATE COVER GAS,
XENON, AND FISSION
PRODUCT BEHAVIOR

ADD PLUTONIUM
IRRADIATE ENCAPSULATED U

MAP F.P DEPOSITION WITH
GAMMA SPECTROMETER
MEASURE TRITIUM,
SAMPLE FUEL

REMOVE CORE ARRAY
PUT REACTOR IN STANDBY

0 2 4 6 8 0

Fuel S5 POWER (Mw) ruee NS POWER (Mw)
FLusH 1 Flusd 1

Fig. 1.2. Chronological outline of MSRE operations.
PRIMARY PHASE AREAS:

® uF

LigBefgZrfg

© Li,BeE,

@ LigZrig

® LizZrF,

®) LizZr,F
Bef,

©
® zrf, 80Q

LigZryfg .

~- 520\
£-507-, \‘ré

ORNL-DWG €6—7321R3

TEMPERATURES IN °C
COMPOSITION IN moie %

2-LIQUIDS

</
£-360

Fig. 1.3. The system LiF-BeF,-Z1F,.

and their capability to operate free from corrosion
problems for long pertods of time at high temperatures.
It was also our purpose to examine the possibility of
adapting those analytical chemical methods which
proved to be most successful with the MSRE to the
development of in-line methods for MSBR analysis and
control. Accordingly, our primary goals were to deter-
mine continually whether the concentration of uranium
in the fuel salt coincided with that expected from
reactor physics calculations and on-site neutronic
measurements, whether chemical evidence would indi-
cate that the salts remained free of oxide as a
contaminant, and to establish rates of corrosion. The
frequency of sampling was set at the beginning of
operations essentially to coincide with estimates of

reasonable maximum capability of the analytical labora-
tory, and was at first about one sample per shift,
decreasing in frequency as it became evident that no
serious problems were developing in relation to cor-
rosion and reactivity anomalies to one per day, three
per week, and in the final stages of MSRE operations to
once a week.

The results of chemical analyses have served as a good
measurement of the generalized corrosion and have
provided a convincing demonstration of the compati-
bility of molten-fluoride fuels and coolants with Hastel-
loy N. Analyses of the changes in isotopic composition
of the plutonium and uranium in the fuel salt samples
established with good precision the average power
output of the reactor. Studies of the relation of ?5Nb
disposition within the fuel system of the MSRE showed
that the distibution of the isotope within the system
reflects the oxidation-reduction potential of the sait 1n
such a manner that behavior of ?*Nb 1n molten-salt
breeder reactors may be exploited as an in-line indicator
of potential corrosion

Results of the chemical analyses performed in support
of MSRE operations comprise the basic data from
which various inferences can be made pertaining to the
performance of the MSRE For example, those relating
to corrosion phenomena, power output, fuel stability,
fission product chemistry, and distribution of tritrum
within the system have been judged on the basis of
results obtained in a program of chemical momtoring of
specimens removed from the MSRE They comprise a
body of information which forms the basis for ranking
priorities 1n future research and development programs
For these reasons, this report has been prepared to
mclude all of the available results of chemical analysis
which were obtained 1n connection with the operation
of the MSRE Except for the correlations of the
disposition of **Nb 1n the fuel salt n relation to the
oxidation-reduction potential of the salt, fission prod
uct chemustry 1s omutted Studies of the behavior of
fisston products i the MSRE have been carried out by
F F Blankenship, S S Kuslis, E G Bohlmann, and
E L Compere The results of their studies are described
1n a report® which summarizes this aspect of the opera-
tional expertence with the MSRE

References

f—

Nucl Appl 8(2) (February 1970)
2 W R Grimes, Nucl Appl 8,137 (1970)

3 R E Thoma et al ,J Nucl Mater 27, 166 (1968)

4 G Long and F F Blankenship, Reactor Chem
Dw Ann Progr Rep Jan 31, 1965, ORNL-3789, p
65

5 E S Bettis and W K Ergen, “Aircraft Reactor
Experiment,” chap 16 in Fluid Fuel Reactors ed by
J A Lane, H G MacPherson, and F Maslan, Addison-
Wesley, Reading, Pa , 1958

6 F F Blankenship, E G Bohlmann, S S Kirsls,
and E L Compere, Fission Product Behavior n the
MSRE, ORNL-4684 (in preparation)

2 CHEMICAL BEHAVIOR IN THE FUEL
AND COOLANT SALT SYSTEMS
DURING PRENUCLEAR OPERATIONS

2 1 Preoperational Procedures

Three types of fluoride muxtures were prepared for
use 1n the MSRE (1) a salt mixture of the approximate
composttion " LiF BeF, ZrF, (65 30 5 mole %), for use
as the fuel carrier (2) another of the composition
"14F BeF, (66 34 mole %), for use as the coolant and
flush salts, and (3) "LiF 23%UF, and "LiF 235 UF,
(73 27 mole %) muxtures, for use mn constituting the
mmtial fuel charge as well as for subsequent enrichment
of the fuel as required A detailed description of the
methods employed 1s reported elsewhere *

Late in 1964, the MSRE fuel and coolant salt systems
were heated to ~650°C and purged of moisture, the
fuel and coolant drain tanks were then charged with salt
of the composition "11F BeF, (66 34 mole %) 2611 kg
in the coolant system and 4187 kg in the fuel system
The coolant loop was filled on January 9, 1965, and on
January 12, 1965, flush salt was circulated in the fuel
loop for the first time

After the operability of the equipment and the
cleanness of the system had been demonstrated, fuel
carrter salt (lacking enrtched uranium) was charged nto
fuel dramn tank No 2 (FD-2) beginming on April 21,
1965 The contents of 35 shupping contamners (4560 kg
of salt) were melted and transferred to the drain tank
To this was added 236 kg of “LiF UF, eutectic
contamng 147 kg of 23%U (depleted 1 2?°U) Batch
contamers were heated above the hiquidus temperature
of the salt mixture in auxihary furnaces and connected
by a small diameter Inconel tube to the drain tank The
connecting tube extended to the bottom of the batch
contamner so that the salt mixture could be transferred
as a hquid by controlhing the differential gas pressure n
the two containers All fill operations were accom
plished in a routine manner without causing detectable
beryllium contamination of the reactor facilities

The wmutial fuel loading of the MSRE required
approximately 75 ft®> of fused fluorides having the
composttion ' LiF BeF, ZrF, UF,; (65029250083
mole %), fissionable 22°U comprsed about one third of
the urantum nventory To provide for an orderly
approach to cntical operation of the reactor and to
facilitate fuel preparation, the fuel was produced from
the three salt mixtures described above. The enriched
fuel concentrate mixture, m which all 233U was
combined with "LiF as UF, 93% enriched in 235U, to
form the bmary eutectic mixture (27 mole % UF, ), was
prepared in six small batches (15 kg of #3°U each) for
nuclear safety and for planned incremental additions to
the reactor fuel system. The balance of the uranium
required for the fuel was provided as a chemically
wdentical muxture with UF; depleted of 2°°U. The
third component mxture, the barren fuel solvent,
conststed of the remamning constituents of the reactor
fuel and had the chemucal composition ” LiF-BeF,-ZrF,
(64.7-30 1-5.2 mole %). The reactor fuel for the
zero-power experiments was produced subsequently by
adding small increments of "LiF-235UF, into this
carrier salt m the drain tanks and finally into the pump
bowl The composition of the salt was fixed at this
point by the amount of 2**UF, required for criticahty
to be sustained with one control rod completely
mserted.

The addition of the enriched fuel concentrate mixture
to the MSRE to within 1 kg of 233U of cnticality was
accomplished during the latter part of May 1965. The
first major addition of enriched fuel concentrate con-
sisted of the transfer of about 44.17 kg of 23%U from
three containers directly into the fuel drain tank Three
subsequent additions of 2?3 U to the reactor drain tank
increased 1ts 235U inventory to 59.35, 64.42, and
finally 68.76 kg The transfer of less than batch-size
quantities of 225U was made by inserting the salt
transfer hne to a predetermuned depth in the batch
contamer.

It was anticipated that the composition of the fuel at
this point would be 7 LiF-BeF,-ZrF,-UF, (65 0-29.17-
5.0-0.83 mole %), the composition calculated from the
weights of the carrier and enniching salts added to the
reactor was changed steadily throughout the zero-power
experiments as capsules of the enriching salt were added
to the pump bowl.

Complete results of all analyses performed with
MSRE salts during the precritical and zero-power
experiments are summarized in the following sections
of this chapter.

2.2 Flush Salt

Part of the LiF-BeF, mixture which was furnished for
use m the MSRE was employed to flush the fuel system
imtially and later, on occasions before and after
maintenance periods 1if 1t were likely that atmospheric
contaminants could have entered the system. Samples
of each of the batches of "LiF-BeF, mixtures for use as

10

flush and coolant salts were analyzed 1n the facilities of
the ORNL Analytical Chemistry Division prior to use in
the MSRE. Analytical results for the 61 batches of salt
used as flush salt and primary coolant are listed in Table
2 1. Although the nominal composition was ’ LiF-BeF,
(63-34 mole %), the average composition as determined
by chemical analysis was " LiF-BeF, (63.56 = 0.005—
36.44 * 0.005 mole %). Examinations of the material
balance of the production operations indicate that the
disparity 1n composition was due to an analytical bias.
Efforts to deterrmine whether such a bias actually
existed were made by the analytical chemists, but were
unfruitful.

Before salt was charged into the reactor, the dramn
tanks and salt loop systems were carefully dried by
evacuating and purging with dry helum. A detailed
description of the procedures employed 1s given else-
where.? Thereafter, late in 1964, approxmmately 4187
kg of flush salt was charged into the fuel drain tank; it
was then transferred among the tanks in the drain tank
cell. These operations served to calibrate the weighing
devices, to check elevations and volumes, and to
establish the operating requirements of the freeze
valves. Throughout the prenuclear test period, speci
mens of the circulating flush salt were obtamned for
chemical and spectrochemical analysis. The results of
spectrochemical analysis are listed in Table 2.2, chemi-
cal analyses are listed in Table 2.3. Spectrochemical
data were of principal value to ensure that during the
early stages of operation the salt charge was free from
any unexpected cationic contaminants in the contain-
ment system. These analyses were very useful 1n that
they provided the assurance required, but with ex-
tended experience with the reactor they became of less
significance and were discontinued.

The data shown in Table 2.3 suggest that the
operations connected with transfers of the flush salt
among the storage tanks in the drain tank cells 1n
November and December of 1964 did not introduce
appreciable amounts of structural metal contaminants
mmto the salt, the analytical results for 1ron suggest,
rather, that since its apparent concentration in the salt
samples removed from the drain tank was lower by
about one standard deviation than the average concen-
tration in the salt before use and in use, 1t was present
in the flush salt mitially as metallic particulates which
were precipitated from the salt during this period. Little
or no change 1s evident 1n the chromium concentration
before circulation of the flush salt, the data in Table 2.3
can be interpreted, however, to indicate that after
completion of the flushing operations in January—
March 1965, the chromium analyses represent an
11

Table 2.1. Chemical analyses of LiF-BeF, (66-34 mole %) produced
for use as MSRE flush and coolant salts

7

Batch Net wt. Li Analyses
No. of salt Assay Li Be F Cr Ni
(kg) (wt %) (ppm)

116 117.3 99.992 13.76 9.61 77.3 18 14 204
117 119.3 99.992 13.77 9.75 77.5 32 14 171
121 119.8 99.992 13.29 9.51 77.2 62 6 125
125 117.8 99.991 13.20 9.59 77.3 16 5 35
126 119.8 99.992 13.27 9.69 76.9 15 32 103
127 119.6 99.992 12.98 9.86 76.8 7 12 123
128 119.6 99.992 13.06 9.81 76.9 9 45 112
129 117.1 99.992 13.20 9.53 77.2 <5 <5 97
131%* 117.2 99.991 13.30 9.67 77.0 14 31 104
132 117.7 99.992 13.00 9.78 77.1 <5 18 76
133 124.8 99.992 12.41 9.30 76.3 8 79 i6l
134 126.1 39.991 12.80 9.84 77.3 6 72 151
135 120.4 99.991 12.90 9.82 77.5 7 21 123
136 120.0 99.991 12,20 9.97 77.2 8 11 180
137 119.4 99.991 12.76 9.49 77.8 ) 10 72
138 119.2 99.991 i2.70 9.90 77.0 15 23 85
139 105.5 99.991 12,91 9.78 76.2 6 66 119
140 125.3 99.991 12.70 g.81 77.1 21 91 130
141 117.7 99.991 12.70 9.86 77.4 37 36 138
142 111.0 99.991 12.90 10.00 77.2 23 5 130
143 124.4 99,993 12.70 9.97 77.3 19 34 79
l44% 118.1 99.994 12.90 9.98 77.2 14 56 110
145 119.8 99.994 12,80 9.96 77.0 15 <5 142
146 119.8 99.994 13.00 10.00 77.4 23 134 135
147 118.8 99.994 12.90 10.00 77.1 18 107 117
148 119.1 99.993 12.90 10.10 77.3 24 12 119
149 117.5 99.993 12.80 9.89 77.4 i3 105 166
150 118.7 99.993 13.20 9.70 76.7 18 34 86
151 124.6 99.993 12.90 G.82 77.4 13 12 143
152 112.3 99.994 12.50 9.62 77.4 9 27 172
153 125.6 99.994 13,00 9.84 76.7 17 31 132
154 95.9 99.994 11.10 9.55 77.2 26 47 106
155 128.1 99.994 13.90 8.92 76.2 27 91 117
156 119.1 99.993 12.70 9.55 77.3 17 28 92
157 118.8 99.993 13.20 9,80 77.0 11 29 23
158 119.4 99.993 13.20 9.80 77.1 17 29 124
159 119.2 99.993 13.10 9.74 76.9 13 108 163
160% 119.6 99.994 13.20 9.77 77.0 13 11 99
161 118.9 99.994 13.20 9.80 77.2 8 i8 75

*
Production Excess.
Table 2.2. Spectrochemical analysis of MSRE flush salt
during mtial use

Sample Impunity concentration?

No Al B Bi Ca Cu Mg Mn Na Pb S1 Sn Zr

FD-2-20
FD-2-3
FD-2-4
FD-2-5
FP-1¢
FP-2
FP-3
FP4
FP-§
FP-6
FP-7
FP-8
FP-9
FP-10
FP-11
FP-12
FST-14
FST-2
FST-3
FD2-7
FD2 8
FD2-9

o W
WmW W

o0
mmmmwo OO0 N

AN
A
mwWmOOOOOOCOO0OO0000O0O0n

oloRoNeNoNoRoNaloReoNoNoNaNONONO NGRS Ne!

>A
>A
>A

Wm W SRR R W
o s e o e e~ e e - e e 2

4Ly and Be omitted A = 100 to 1500 ppm, B = 10 to 100
ppm,C=1i0 10 ppm

bED designates samples removed from the fuel drain tank

CFP designates samples removed from the fuel pump bowl

dEsT designates samples removed from the fuel storage tank

12

average final value of 60 ppm If 1t 1s assumed that the
increase resulted from scavenging of moisture or oxides
from the system, one would anticipate a corresponding
increase of 12 ppm of oxygen in the salt (see Sect 6 1)

Results of oxygen analyses of the flush salt exhibited
perplexing vanance The larger values are believed to
reflect adsorption of moisture inadvertently introduced
through the handling procedures after samples were
removed from the reactor If the oxygen concentrations
were representative of either contamination of the salt
by water or oxidation of the container alloy by
atmospheric oxygen, the occurrence of chemical attack
on the Hastelloy N walls of the circuit should have been
reflected 1n increased concentrations of chromium in
the salt For example, if an increase of oxide concentra-
tion m the order of 200 ppm were to have been caused
by such reactions, an equvalent increase 1in chromium
concentration of 60 ppm should have been observed

To test whether a large bias might have been mnvolved
i the fluorination assay of oxygen concentration (the
KBrF, method), large (50 g) samples of the salt were
obtained and analyzed by a newly developed method
which, like the punfication procedures, depends on the
removal of water by sparging the molten salt with an
H,-HF stream For the three samples that were ana
lyzed 1n this manner, the average concentration found
was 75 ppm

After the flush salt was drained from the reactor on
completion of the tests in which 1t was used, the flush

Table 2.3. Composition of MSRE flush salt in prenuclear operations

Weight percent

Parts per milhion

Date Sample U

11 Be Zr

= Ie Cr Ni Pu o4

Book  Analytical

Charge salt®? 1268 962
11/30/64 DC-1 1312 968
12/8/64 DT-1
12/11/64 DT?2
12/15/64  FD-2-3
1/12/65 FD-2-5
1/12/65 FP-1

1/13/65 FP-2 1365 983
1/14/65 FP3
1/16/65 P4 1355 935
1/18/65 | ]
1/20/65 FP-6 1350 996
1/23/65 FP-7 1365 946
2/3/65 FP-8
2/3/65 I'pP-9 1335 998
2/11/65 FP-10 949

2/23/65 FP-11
3/4/65 FP-12

Av 1347 968

76 51 99 01 137 17 34

7708 99 94 45 45 7 432

59 16 24 419

48 18 <3 390

44 22 49 555

140 24 10

35

80 52 104 62 i <10 33 56
46¢

79 34 102 27 212 62 30 74
72¢

80 07 103 56 125 <10 <20 180

77 85 101 00 180 54 <20 150
106¢

75 80 9917 210 57 <20 142
75 05 128 60 <20 15604

No analyses performed
144

7796 10166 118 38 22

4K BrF4 method unless otherwise noted
bAverage for 61 batches

CHF purge method
dSample stored 48 hr m capped plastic container before analysis
salt was reprocessed > The salt was transferred by gas
pressure to the fuel storage tank in the fuel reprocessing
system and sparged with a mixture of H, and HF It
was concluded from measurements of the HF concen-
tration 1 the off-gas stream that the reprocessing
operations were effective in removing 115 ppm of oxide
from the salt > The oxide concentration should have
been reduced by this process to the same concentration
which resulted from preparation and purification of the
salt, ~55 ppm®* The results of chemucal analysis
indicated that the chromium concentration of the flush
salt increased by 43 ppm 1in use, corresponding to an
increase of 180 g of chromium, and equivalent to 55 5 g
of oxide, or to an increase of 13 ppm in oxide
concentration The H,-HF analytical data indicated
that the average concentration of oxide 1n the {lush salt
during 1ts use was 75 ppm, or that the oxide concentra-
tion of the salt increased by 20 ppm mn use This
increase, while shightly greater than would have been
anticipated from the above discussion, may be ration-
alized by the following considerations An indetermui-
nate fraction of the oxide removed in reprocessing was
introduced from the oxide scales present on the
surfaces of the chemical reprocessing facility containers,
the fuel storage tank, and the walls of the fuel circunt
Prior purge of the system with dry 1nert gas was
performed before the facility was used for reprocessing
the flush salt, but the system was not previously flushed
with a scavenger salt It might therefore have been
anticipated that the oxide concentration of the salt i
the reprocessing system, after its initial use, would yield
a hagher concentration of oxide than that removed from
the salt in the fuel circulation system Thus, if 1t 1s
assumed that in the reprocessing experiment ~0 4 kg of
oxide scale was removed f{rom the system, good
agreement exists among the analytical data, except for
measurement of oxide concentration by the KBrF,
method

One potential use of the flush salt that had been
anticipated was 1o remove any oxides that might
deposit 1n the heat exchanger 1f the fuel salt were to
become seriously contaminated Concurrent with the
beginning of MSRE operations, chemical data ap-
peared® which showed that the solubihity of the least
soluble oxide i the LiF-BeF,-ZrF, carner salt was
some three times that in LiF-Bel, mixtures It was thus
necessary to regard the flush salt as an agent which
might have limited capacity to scavenge oxide impu-
rities from the fuel circuit unless, if saturated, 1t were
reprocessed Its principal function after imitial use was
to remove residues contaiung uranium fluoride from
the fuel circuit after 1t was drained Concurrently, the

13

capacity of the flush salt to contamn dissolved oxides
increased with each use in the reactor, because the fuel
restdues carried into the flush salt added about 2 5 kg
of zircontum to the salt each time 1t was used It would
probably have questionable value 1f 1t were expedient to
dissolve precipitated oxide deposits from the heat
exchanger This did not preclude the use of LiF-BeF,
flush salt for such application, because at 650°C the
solubility of the saturating oxide phase, BeO, 1n the salt
was known to be ~200 ppm, corresponding to a
capacity of the flush salt to dissolve and retain 1 135 kg
of ZrO, 1n solution at this temperature Successive
repurtfication and removal of oxides that were picked
up 1n the salt would permit it to be used to clear the
heat exchanger of an oxide plug if necessary No such
requirement developed A preferable alternative was
also available if needed The increased solubility which
small amounts of zircomum fluonde provide, increasing
the oxide solubility to 800 ppm for a few mole percent
of ZrF,, showed that if 1t were expedient to do so, the
composition of the flush salt could have been changed
simply by the addition of "LiF ZrF,; so as to enhance
its solvent capacity for precipitated oxides

Operation with the flush salt also shed lLight on
another aspect of the use of molten fluorides mn a
arculating system, namely, the appearance of salt (or
salt constituents) in the cover gas Analysis of the
hehhum cover gas by Million and Pappas (see Sect 2 4 5)
indicated the presence of fluorides m the fuel off gas,
but this probably represented salt droplets rather than
any decomposition product

The principal difference in the design of the fuel and
coolant pumps 13 that the fuel pump contains a spray
nng which functions to remove fission product gases
from the fuel The major salt flow into the pump tank
1s through a bypass that s taken from the volute
discharge line into a toroidal spray ring in the upper
part of the pump bowl From there the salt sprays out
through two rows of holes and mmpinges on the sait
surface in the tank to prowvide gas-liquid contact for gas
stripping ¢ This difference 1n design of the two pumps
seems to account for the tact that it was occasionally
necessary to remove solids from filters, valves, and hines
n the fuel system, but rarely necessary with the coolant
system Material recovered from the coolant off.gas
system showed only traces of salt constituents, but the
fuel off-gas solids generally contained minute beads of
frozen sait

One experiment conducted dunng the precritical
operation pertod consisted of krypton stripping tests
For these tests, a special mnsert was installed into the
side of the off-gas flange nearest to the fuel pump bowl
14

PHOTO 4863 -71

oy

Fig. 2.1. Fluonde glass beads removed from MSRE off-gas line after precritical experiment PC-1.

Later, the flange was opened to remove the insert,
inspection showed that small amounts of small glassy
beads were deposited between the flange faces

Toward the end of the first set of experiments with
flush salt, fuel pressure control became erratic (see
Chap 11) After the salt was drained from the circuit in
March 1965, the gas pressure control valve far down-
stream at the vent house was found to be partially
plugged The valve body was rinsed with acetone, which
was found to contain beads (1 to 5 y n diameter) of a
glassy material

After a week of carner salt circulation 1n May 1965,
the small control valve again began to plug This time, it
was removed and cut open for examination A deposit
on the stem was found to be about 20% amorphous
carbon, and the remainder was 1- to 5-u beads having
the composition of the flush salt A photograph of
these beads, obtamned with the petrographic mucro-
scope, 1s shown in Fig 2 1 The glass beads were found
to have a refractive index of 1315 As computed by
interpolation of the refractive indices of the crystalline
components LiF and BeF,, its composition would be

LiF-BeF; (54-45 mole %) Chemical analysis showed,
however, that 1ts actual composition was LiF-BeF,
(66-34 mole %) The anomalous refractive index of the
glass results from tts glassy character, that s, the hiquid
phase at LiF-BeF, (66-34 mole %) 1s of significantly
lower density than the crystalline solids of this COMmpo-
sition  The beads were glass rather than crystalline,
mplying rapid cooling of the molten-salt mist The
carbon was presumably soot from oil that had been
thermally decomposed in the pump bowl

2 3 Coolant Salt

The coolant salt mixtures for use in the MSRE were
prepared as described elsewhere ' Twenty-four batches
of the salt muxture were allocated for use as the
coolant The composition and purity of these mixtures
are listed 1n Table 24 A charge of 2610 kg of
"LiF-BeF, (66-34 mole %) was delivered i molten
form to the coolant drain tank in October 1964 The
coolant system was preheated and purged of moisture
in a similar manner to the fuel system, but was not
Table 2.4, Fluoride production for MSRE coolant salt mixture
LiF-BeF, (66-34 mole %)

Chenncal analyses

7
Batch of salt L1 Weight percent Parts per
No assay million
(kg)

| ] Be F Cr Nt Fe
101 118.1 99991 1368 943 769 8 32 133
102 116 2 99.991 14.23 9,14 772 5 8 175
103 121.2 99991 1381 932 767 10 58 182
104 1195 99990 13.68 982 772 17 11 208
105 113.2 99.990 1283 957 771 27 45 216
106 1190 99.990 1343 966 772 11 68 309
i07 1150 99990 1312 958 769 10 18 94
108 1145 99990 12.81 965 773 T 13 246
109 1199 99990 1373 927 770 8 5 62
110 1279 99.990 1315 957 766 19 12 94
111 121.2 99990 1399 921 765 13 8 142
112 1200 99991 13.81 947 770 11 48 218
i13 117.6 99.991 13.80 9.53 77.0 <5 2§ 212
114 1209 99991 1314 979 770 9 22 196
115 120.8 99.991 1374 954 774 14 5 131
118 125.0 99991 1306 991 773 9 20 134
119 114.5 99,991 12.41 10.16 769 10 18 182
120 125.5 99.991 13.65 9.52 771 8§ 35 127
122 117.8 99991 1326 9.48 773 39 70 126
123 117.9 99991 1337 948 769 48 10 113
124 118.5 99.991 1335 945 771 86 14 154
130 119.9 99,990 13,10 955 774 40 13 188

preflushed with molten salt. The system was filled and
drained 18 times durning the time the MSRE was
operated. The chemical behavior of the coolant salt was
not noticeably different during prenuclear operations
from those that followed, and will be described 1n detail
later in this report.

2.4 Fuel Salt

2.41 Onssite preparation. After circulation in the
reactor in March 1965, the flush salt was dramed to fuel
drain tank FD-1. The fuel salt was then constituted
within the reactor. The "LiF-BeF,-ZrF, fuel carrier
mixture was charged into fuel drain tank FD-2, starting
April 21, 1965. On-site records show that 4558.1 kg
(the contents of 35 shipping containers) of this carrier
salt was melted and transferred to the dramn tank; to
this was added 236.2 kg of "LiF-*3®UF,, contamning
147 6 kg of *3®U (depleted in 23°U) In our current
review of the significance of the results of chemical and
mass spectrochemical analyses, 1t becomes evident that
these values are in munor error. The "LiF-BeF,-ZrF,-
238UF, as described above was circulated through the
fuel curcuit for some 250 hr during precritical test PC-2.

15

Each of the two fuel dran tanks on the MSRE
incorporates two pneumatic weigh cells for estimating
the inventory of the salt in these tanks. Calibrations of
the weigh cells in the drain tanks were made carefully
during the 1mtial stages of operations. The precision of
these measurements was found to be about £0.8% (£28
kg).” Continued experience with this equipment mdi-
cated long-term drifts i the readings. It was necessary,
therefore, to determine inventories on the basis of
analytical chemical data, and on computations involving
salt densifies and volumes at certamn reference points
(the circulating loops filled to the pump bowls or the
tanks filled to the level probe points) to eliminate the
effects of extraneous forces on the indicated weight.?

During the imtial fill operations, "LiF-BeF, flush salt
was admutted to all parts of the fuel circutt system. On
draining the reactor, flush salt remained on the pump
rotor and in the freeze valves, as well as residue 1n the
heat exchanger. At this point, some 35 kg of salt was
unaccounted for by the weigh cell data and was
presumed to have remamned in these locations. The
"LiF-BeF,-ZrF,-2*3UF, salt mixture was introduced
mto the same circuitry at the beginning of the PC-2 test
and was, therefore, diluted slightly by flush salt. Later
in this report, 1t is shown that the increased concentra-
tion of uramum in the flush salt after use indicates that
the salt residue left in the fuel circust system after 1t 1s
drained 1s about 20 kg. Estumation of the nominal
concentration of uranium 1n the fuel during precritical
test PC-2 thus requires the assumption that ~20 kg of
flush salt was added 1mtially to the fuel. As noted 1n
Sect. 2.4.2, however, the best agreement between
expected and observed concentration 1s found when it
18 assumed that a negligible amount of flush salt residue
remained in the system at completion of PC-2. This
assumption 1s untenable. An alternate assumption, that
less carrier salt was delivered to the reactor than was
credited 1n the loading operation, 1s more credible as an
explanation of the apparent anomaly.

Chemucal analyses were conducted durning the precriti-
cal experiments for the purpose of establishing analyt-
ical base lines for use in the full-power operating period.
Prior to the zero-power experiment, analyses of all salt
samples were conducted with conventional equipment
i the facilities of the ORNL Analytical Chenustry
Division. Thereafter, the fission product activity made
it necessary to perform all analyses in the ORNL
High-Radiation-Level Analytical Laboratory (HRLAL).
The facilities of this laboratory have been described
elsewhere.” In order to provide continuity which would
relate the results of prenuclear operations with those
that followed, specimens of the fuel salt obtained
during the precritical test period were analyzed concur-
rently in both facilities

Mass spectrometric analyses were performed routinely
throughout #33:238( operations Samples of the fuel
salt were analyzed before, duning, and at the comple-
tion of the ?*°U loading operation The results
reflected the 1sotopic difution of 2*®U precisely and
were used in tests for corroboration of the agreement
between estimated and analytical values of the concen
tratton of uranmum in the fuel salt and to refine
estimates of the amount of depleted uranium that was
charged 1nto the fuel system

The composition of the fuel salt was changed steadily
throughout the zero power experiments (run No 3) as
capsules of the ennching salt were added to the pump
bowl Compositional analysis during this perniod served,
therefore, to permit evaluation of the fuel composition
dynamucally rather than as a statistical base for ref-
erence during the power run

Most of the uramum required for enrichment was
loaded i four charging operations to one of the fuel
dramn tanks, FD 2 After each addition the salt was
transferred to the second drain tank and back again to
ensure thorough mixing The mixed salt was loaded 1nto
the reactor system after each charging operation, and
count rate data were taken at several salt levels in the
core and with the reactor vessel full These data were
compared with the barren-salt data to momtor the
neutron multiphcation and to estabhsh the size of the
next addition '® Extrapolation of inverse count-rate
plots with the reactor vessel full showed that the
loading after the fourth addition was within 0 8 kg of
235U of the critical loading when the salt was not
circulating and the control rods were withdrawn to
their upper hmts The remainder of the *3°U was
added directly to the circulating loop with enriching
capsules These were inserted into the fuel-pump bowl
via the sampler-enncher to increase the loading by 85 g
of uranium at a time Count rates were measured after
each capsule with the fuel pump off and the control
rods withdrawn The reactor became critical after the
eighth capsule with the pump off, two rods fully
withdrawn, and one rod poisoning 0 03 of 1ts available
worth

These zero-power experiments (run 3) were com-
pleted on July 4, 1965, after 764 hr of circulation of
the salt The fuel was then drained and mixed with the
salt remaining in the drain tanks After the fuel loop
was drained, 1t was filled with flush salt, which was
circulated 1 3 hr, sampled, and drained Analysis of the
flush salt led to the conclusion that 0 77 kg of uranium
remained n the fuel circurt on completion of run No 3

16

Final preparations were made for operation of the
reactor at power In December 1965, low power exper1
ments were begun As soon as containment testing was
fimished, the instruments, controls, and equipment were
given the check outs required prior to startup The fuel
system was then heated, and flush salt was circulated
for three days Samples of the flush salt were taken
Then fuel was charged to the loop for tests in the
reactor at various low power levels, 100 and 500 kW
and finally 1 MW The experiments that were con-
ducted during this time extended over a one month
period, providing the single period of operation when
the composition of the fuel salt remained nominally
constant

Samples of the fuel salt were obtained regularly
during this period in a continuing effort to appraise the
utility of the chemical surveillance program and in
order to verify that the reactor system was in suitable
condition for the mnitiation of power operations Good
agreement between anticipated and observed results was
found The results of chemical analyses indicated that
the system had remained free of corrosion during all
preliminary operations, and it seemed that the analyses
would serve as a reliable measure of fuel stability and
corrosion

2 42 Uranium assay Chemical analyses and on site
nuclear measurements were accumulated regularly when
fuel salt was circulated in the reactor in order to
evaluate as many aspects of the behavior of the molten
radioactive fuel salt as feasible in contact with the
Hastelioy N containment system and with the graphite
moderator A number of important characteristics of
the reactor soon became evident For example, by the
time that statistically significant populations of analyti
cal chemical data were produced from the HRLAL it
had become dapparent that on site neutronic measure
ments of the reactivity balance were sufficiently sensi
tive to detect changes of as little as 0 1% i the
concentration of fissile material circulating in the fuel
system ' This represented at least a factor of 10
greater sensitivity than was indicated by the statistically
significant results of chemical analysis In addition,
reactivity balance measurements were made dynami
cally and, essentially, on a continuous basis While
chemical data alone were useful as statistical indicators
of trends in the concentration of uranwum over ex
tended periods, the combined results of chemical and
mass spectrometric analyses furnished information that
was umquely suited for use in estabhishing numerous
fundamental values for use in subsequent evaluations of
MSRE performance, and 1n accounting for the changes
in the distribution of urammum between the fuel and
flush salt systems
In order to compare the concentration of urantum in
the fuel salt as measured analytically with expected
values, we have evaluated the results of chemical
analysis, mass spectrometric analysis, the weights of the
fluonides as measured in the production facility, and
on-site measurements of the weights of the salt con-
tainers delivered to the MSRE From the weights of the
enriching salts credited to the reactor and from the
isotopic composttion a uranium material balance was
computed Progress reports for this period of MSRE
operations indicate that satisfactory agreement between
nomunal and analytical values was found As the results
of mass spectrometric analyses were accumulated,
however, comparisons of the nominal 1sotopic composi-
tion of uranium during initial loading operations at the
MSRE with the results of mass spectrometric analyses
disclosed that closest agreement 1s reached when the
assumption 1s made that the initial "LiF-**®UF, charge
contained 145 6 kg of uramum rather than 147 6 kg as
credited (see below) This lesser amount of uramum,
carried by 234 5 kg of "LiF-?33UF, enriching salt and

17

added to the amount of "LiF-BeF,-ZtF, carrier salt
credited to the MSRE (4558 1 kg), would produce a
salt mixture in which the concentration of uramum 1s
3038 wt %

The average concentration observed was 3 044 wt %
In the early stages of analysis, a bias of —0 8% was
discovered 1n the amperometric data'? which, when
used to correct the analytical results, causes the average
observed concentration in precritical run No 2 to be
3068 wt % Thus the nomnal concentration of
uranium 1n the fuel salt ts inexplicably lower than the
averages of the observed values if the amount of carrier
salt credited to the MSRE (4558 1 kg) 1s assumed to be
accurate A computation of the material balance for the
MSRE fuel salt in prepower tests, using this value, 1s
summarnzed in Table 2 5, which shows that the analyti-
cal values tend to indicate slightly greater average
concentrations than nominal If, however, we assume
that ~40 kg of the carrier salt was not charged into the
fuel circuit, and the fuel charge during the precritical
experiment was constituted from 145 60 kg of uranium

Table 2.5, Material batance for MSRE fuel salt in prepower tests

Inventory (kg)

Uranmam concentration {wt %)

U 7L1F-UF4 Totat Nomuinal Analytical

Run PC-2 (av) 145 60 2345 4792 6 3038 3 068
Run 3, FP-3-1,2

Added +47 49 765

Total 193 09 3110 4869 1 3965 3998
Run 3, FP 3-3

Added +2143 34 53

Total 214 52 345 53 4903 6 4 375 4 390
Run 3, FP-3-4

Added +4 67 +7 52

Total 219 19 35305 4911 1 4 463 4 446

Loop charge 195 23 4374 3 4463

Left in drain tank 2396 536 8 4 463
Run 3-F

Added +7 83 +12 74

% charge 227 02 45237 4611

Loop charge 203 06 4386 9 4 629

Loop residue -0 77 165

To drain tank 202 29 4370 4 4 629

Already in drain tank +23 96 +536 8 4 462

FD-10-3 226 25 4907 2 4611 4 648
Run 4 I loop charge 208 15 4514 6 4611

Flush residue +18 0

Loop charge 208 15 4532 6 4592 46420028
Run 5-1, drain tank 226 25 49325 2 4593
Runs 5 7 (av of 14 samples) 4629+ 0026
Run 4 (av of 22 samples) 4642+ 0028
Runs 4 7 (av of 36 samples) 4638 + 0025

and 4754.5 kg of salt, the nominal concentration of
uranium should be 3.062 wt %, as compared with an
analytical value of 3.068 wt %. Comparably improved
agreement 1s found between the nominal and observed
concentrations of uranmum for all other prepower
measurements when the assumption 1s made that the
above quantittes of uranium and salt were charged to
the MSRE fuel system. A uranium material balance for
the prepower period was computed on the basis of
these values, and 1s given i Table 2.6. We deduce,
therefore, that the salt charge which was used for the
precritical test, PC-2, consisted of 145.6 kg ***Um a
total salt charge of 4754.5 kg. The results of the
chemical analyses performed with samples of the fuel
salt during this test are hsted in Table 2.7 Most of the
uranium required for enrichment was loaded mn four
charging operations to one of the fuel drain tanks,
FD-2. Samples of the fuel salt were obtamned frequently
during this period 1n a continuing attempt to appraise
the utiity of the chemucal surveiilance program, and to

18

verify that the reactor system was in suitable condition
for the imitiation of power operation.

As the concentration of uranium 1 the fuel salt was
mncreased in preparation for operating the reactor at tull
power, fuel-salt samples were obtamned for 1sotopic and
compositional analysis after groups of three or four
capsules of enriching salt were added. Changes in the
1sotopic composition of uranium in the fuel salt during
mitial loading operations are summarized in Table 2.8.
Compositional analyses were obtained over a period
when the incremental change in the concentration of
urantum 1n the fuel salt was large 1n comparison to that
in the normal operating mode of the reactor. They
serve, therefore, to permut evaluation of the fuel
composition dynamically rather than as a statistical
base for reference duning the power run. The results of
these analyses are listed 1n Tables 2.9 and 2.10.

As noted previously, the salt mixture used in the
precritical test, PC-2, was formed from purnfied salt
mixtures of the nominal compositions ” LiF-BeF,-ZrF,

Table 2.6. Material balance for MSRE fuel salt in prepower tests

Inventory (kg)

Uranium concentration (wi %)

U 7L1F-UF4 Total Nominal Analy tical

Run PC-2 (av) 145 60 234.5 4752.6 3.064 3.068
Run 3, FP-3-1,2

Added +47.49 76.5

Totat 193.09 311.0 4829.1 3.998 3.998
Run 3, FP-3-3

Added +21 43 34.53

Total 214.52 345 5 4863.6 4410 4.390
Run 3, FP-34

Added +4.67 +7,52

Total 219 19 353.05 4871 1 4 500 4.446

Loop charge 195 23 4338.6 4500

Left in dramn tank 2396 532.5 4 500
Run 3-F

Added +7 83 +12.74

Z charge 227.02 4883 8 4.648

Loop charge 203.06 43513 4.667 4.665

Loop residue -0.77 -165

To dram tank 202 29 43348 4.667

Already m dram tank +23 96 +532 5 4 500

FD-10-3 226.25 4867.3 4 648 4.648
Run 4-1, loop charge 208 15 44779 4 648

Flush residue +18 0

Loop charge 208.15 44959 4.630 4,642 + 0.028
Run 5-1, drain tank 226 25 4885 3 4,631
Runs 57 (av of 14 samples) 4,629 = 0.026
Run 4 (av of 22 samples) 4,642 + 0.028
Runs 4—7 (av of 36 samples) 4,638 + 0.025

19

Table 2.7. Summary of MSRE salt analyses, expeniment No. 2, fuel sait
Nominal composition of salt LiF-BeF ,-Z1F 4-238 UL 4 (64 86-29 54-5 07 0 53 mole %)

Weight percent

Parts per million

Date Sample L U

Nominal

z Fe Cr Ni O

Analytical

5/11/65 FP-2-13¢ 1030 699 1132 3062
5/12/65  FP-2-14¢ 1020 655 1149 3062
5/12/65 kP 2-15 1014 709 1142 3062
5/13/65 TP-2-16 1050 660 1209 3062
5/13/65  ¥P-2-17 1030 654 1108 3062
5/13/65  FTP-2-18 1050 654 1181 3062
5/14/65 FP2-19 1020 616 1141 3062
5/15/65  FP-2-20 1053 664 1136 3062
5/16/65  FP 2-21 1030 638 1108 3062
5/17/65  ¥P-2-22 1020 621 1117 3062
5/18/65 kP 2-23 030 616 1123 3062
5/18/65  FP-2-24 1078 684 1118 3062
5/19/65 PP 2-25 1052 650 1136 3062
5/19/65  FP-2-26 1060 667 1127 3062
5/20/65  FP-2-27 1046 644 1128 3062
5/20/65  FP 2-28 1055 657 1105 3062
5/21/65  FP 2-29 1051 730 1111 3062
5/21/65  ¥P-2-30 1061 659 1166 3062

3 043 67 50 98 66 74 26 34 20455
3041 7190 10318 149 <50 36

3023 95 19 18 4715
3 044 71 60 10384 271 <50 33
3053 67 40 98 38 75 21 21 155
3053 7340 105 31 152 <50 43
3 061 7140 102 22 130 26 47
3023 68 20 99 76 120 23 21 660
3039 71 00 101 81 133 25 49
3052 7110 10174 221 28 55
3033 70 50 101 23 159 28 42
3014 68 60 100 42 76 23 <5 215
3014 67 40 98 40 110 24 17
3038 7140 102 98 113 24 <20
3063 67 10 98 35 27 23 14 70
2 998 72 30 103 47 123 34 <20
3053 67 70 9968 203 28 20 405
3025 719 10370 175 27 <20

95 23 165
+21 42 44

2 Analyses performed under supervision of W F Vaughan
bAnalyzed by KBrF4 method
“Analysis performed under supervision of C E Lamb

(64 7-30 1-5 2 mole %) and "LiF-UF, (73-27 mole %)
The results of chemical analyses of the individual
batches of these salts are listed in Tables 2 11 and 2 12
Results of the analyses of the carrier salt indicate that
its composition was ’ LiF-BeF,-ZrF, (62 25-32 47-5 28
mole %), or that the Li/Be ratio 1s skewed as noted
previously 1n connection with the coolant and flush
salt Whale the results from chemical analyses show
similar bias 1n determination of lithum and berylhum
when compared with nomunal values, they are 1n good
agreement when the analyses of the salts are compared
with those for the starting materials Reexamination of
the material balance in the preparation of these salts
indicated that dewiations from the nominal composi-
tions are much less than is indicated by the analytical
data '3 In order to estimate how closely the composi-
tion of the fuel-salt muxture matched the design
composition, 1t is necessary to compare the analytical
results with a nominal composition computed from the
analyses of production plant samples The composition
of the salt mixture used 1n the precritical run compares
as shown in Table 2 13 with the nonmunal composition
which would result from (1) combining the carrer salt
and L1F-2*#UF, of the design compositions and (2)

combining compositions as indicated from chemical
analysis

In the foregomng discussion, we have estimated the
quantities of carrier salt and LiF-UF,; mixtures used to
constitute the MSRE fuel for power tests The chemical
compostition of the salt circulated in the fuel circuit at
various stages during the zero power and in all
prenuclear operations 1s compared with the results of
chemical analyses in Tables 2 14 and 2 15 The values
listed 1n these tables comncide within the precision limits
of the analytical chemical results, the mass spec
trometric methods, and the estimates of the physical
properties of these fluids We have therefore adopted
the nominal composttions hsted in Table 215 as
reference compositions for appraisal of results through-
out the remainder of MSRE operations

2.4.3 Structural metal impurities. One of the features
which originally suggested the potential feasibiity of
molten-salt reactors was the recognition that if the
physical and mechanical properties of any one of
several mickel-based alloys, such as those composed
principally of Ni-Mo-Cr-Fe, were suitable for reactor
construction, the alloys would almost certainly be
Table 2.8. Isotopic composition of uranium in the MSRE fuel salt
during initial loading operations

— Loop Inv. 234, 235, 236, 238, 234, 235, 236, 238,
(kg) (kg) (kg) (wt %)
LiF-2 50, Loading 145.60 - 0.323 - 145.277
LiF-“77UF, to 5/25/65 47.49 0.452 44,143 0.187 2.708
Sample No. FP3-2 193.090 0.452 44,466 0.187 147.985 0.234 23.029 0.097 76.640%
23 0.234 22,501 0.099 77.086=
LiF-?3%F, to 5/30/65  +26.100 0.249  +24.261 +0.102  +1.488
219.190 0.701 68,727 0.289 149.473 0.320 31.355 0.132 68,193
195.240 0.625 61,218 0.257  133.140
Sample No. FP3-7 +0.820 +0.008 +0.762  +0.003 +0,047
196.060 0.633 61.980 0.260 133.187 0.323 31.613 0.133 67.932
0.247 23.133 0.103 76.,437%
Sample No. FP3-12 41,094 +0.010 +1,017  +0.004 +0.063
197.154 0.643 62,997 0.264 133,250 0.326 31.953 0.134 67.587
0.335 3L.784 0.138 67.743
Sample No. FP3-17 +1,789 +0.017 +1.663  +0.007 +0.102
198.943 0.660 64,660 0.271 133.352 0.332 32.502 0.136 67.030
0.340 32.606 0.136 66,918
Sample No. FP3-22 +1.787 +0.017 +1.661  +0.007 +0.102
200.730 0.677 66,321 0.278 133.454 0.337 33,041 0.138 66.484
0.352 33.122 0.141 66.386
Sample No. FP3-27 +1.030 +0.010 40,957 +0.004  +0.059
201.760 0.687 67.278 0.282 133.513 0.340 33.346 0.140 66.174
0.355 33.340 0.141 66.165
Sample No. FP3-32 +1.274 +0.012 +1.184  +0.005 +0.073
203.034 0.699 68,462 0.287 133.586 0.344 33,720 0.141 65.795
0.355 33,719 0.139 85.787
Run 3-Residue -0.77 -0.003 -0.241  -0.001 -0.525
202.264 0.696 68.221 0.286 133.061
23,950 0.077 7.510 0.032 16.332
Run 4-I 226,214 0.773 75.730 0.318 149.393  0.342 33.477 0.141 66.041
0.354 33.553 0.145 65,9488
0.350 33.250  0.140 66.260%

2Bold face type indicates nominal values,

b .
—Results of mass spectrometric analyses.

c
—fnomalous results; possible misidentification of sample.

. i

d
—Average of 3 analyses,

e
—Average of 4 analyses,

0T
21

Table 2.9. Summary of MSRE salt analyses, experiment No. 3, fuel salt

Weight percent

Parts per million

2354 in fuel

Date Sample L
L e & £ Fe C Ni o crouit(ke)
Nominal Analytical?

5/27/65 FP-3-1 10.45 6.4 10.90 4.01 3.984 66.8 98.62 115 28 38 335 39.3

5/27/65 FP-3-2 10.7 6.14 11.19 4.01 3.822 68.63 10049 161 30 44 3350 39.3

5/30/65 FP-3-3 10.7 6.51 11.08 4.42 4.355 69.23 101.88 323 24 53 3915 56.9

5/31/65 FP-3-4 10.8 6.57 11.10 451 4.411 68.4 101.33 371 36 48 60.93
5/31/65 FP-3-5%  10.35 6.71 11.04 4.51 4.450 67.6 100.15 162 28 30 490 60.93
6/2/65 FP-3-6 10.75  6.16 11.60 4.52 4,457 69.69 102.66 112 31 39 330 61.46
6/3/65 FP-3-7 10.6 6.62 1118 4.52 4,416 70.9 103.72 121 31 44 61.46
6/4/65 FP-3-8 10.65 6.39 11.57 4.53 4,448 70.10 103,16 130 28 49 520 61.62
6/6/65 FP-3-9 16.6 6.29 11.43 4.53 4,448 69.35 102,13 113 34 6l 320 61.87
6/7/65 FP-3-10  10.75 6.75 11.39 4.54 4.464 70.12 10345 122 34 25 875 62.05
6/8/65 FP-3-11 10.65 6.46 11.11 4.54 4.444 7090 103.56 123 35 62 3715 62.22
6/9/65 FP-3-12 1055 6.70 11.69 4.55 4,461 68.21 101.61 238 42 68 1745 62.47
6/10/65 FP-3-13 1070 6.56 11.06 4.56 4.494 70.08 102.89 150 33 53 2465 62.89
6/11/65 FP-3-14 1050 6.38 11.32 4.57 4.439 69.72 10236 141 38 60 1760 63.23
6/12/65 FP-3-15 10.70 6.55 11.27 4.57 4.454 69.25 10222 116 38 39 1115 63.31
6/13/65 FP-3-16 10.30 6.36 11.23 4.58 4.462 68.15 100,90 185 48 34 555 63.81
6/15/65 FP-3-17 1050 6.28 11.11 4.59 4.523 68.94 101.56 151 44 44 855 64.14
6/16/65 FP-3-18 10.80 6.46 11.62 4.59 4.463 6948 102.82 143 42 68 64.30
6/17/65 FP-3-19 10.00 6.53 11.13 4.60 4.566 67.12 99.54 139 36 71 605 64.71
6/18/65 FP-3-20 10.50 6.68 10.75 4,60 4.499 68.57 10L03 105 49 61 423 65.12
6/19/65 FP-3-21 1045 6.31 10.69 4.61 4.504 68.30 100.32 97 31 74 705 65.46
6/20/65 FP-3-22 10.50 6.54 11.24 4.62 4.535 68.90 101.82 134 35 73 820 65.80
6/21/65 FP-3-23 990 6.67 10.99 4.63 4.568 68.09 10047 122 35 84 1570 66.13
6/22/65 ¥pP-3.24  10.30 6.50 11.17 4.63 4.582 68.99 101.54 77 37 47 840 66.13
6/23/65 FP-3-25 10.60 6.80 11.12 4.63 4.587 67.99 101.10 135 38 52 580 66.21
6/24/65 FpP-3-26 10.45 6.51 1148 4.63 4.557 68.20 101.21 135 34 90 440 66.37
6/25/65 FP-3-27 10.40 6.44 10.27 4.64 4,633 68.68 10042 167 53 28 310 66.80
6/26/65 FP-3-28 10.65 6.26 10.85 4.65 4.610 67.43 99.80 105 38 34 475 67.05
6/27/65 FP-3-29 1040 6.52 10.18 4.66 4.640 68.54 100.28 110 36 44 520 67.56
6/28/65 FP-3-30  10.35 6.55 10.54 4.66 4.651 68.25 10034 121 37 40 935 67.90
6/29/65 FP-3-31 10.10 6.56 10.84 4.67 4.642 67.83 99.97 128 39 49 326 67.98
6/30/65 FP-3-32  10.10 697 11.20 4.67 4.605 71.66 104,55 149 42 93 1235 67.98
7/2/65 FP-3-33 1040 6.47 1124 4.67 4.563 68.20 100.87 170 40 88 965 67.98
7/3/65 FP-3-34¢ 4.67 67.98
7/4165 FP-3-35 1040 6.85 11.15 4.67 4.597 67.79 100,79 158 48 74 1070 67.98

Av(34) 148 37 55

2Corrected for 2*3U enrichment.
bCriticality.
€50-g sample obtained for A. S. Mever.

compatibie with molten-fluoride mixtures. This recogni-
tion developed from the fact that numerous metal
fluorides were more stable than the fluorides of these
alloy constituents and could be used for fuel and
coolant mixtures. By the time conceptual plans for the
MSRE were formulated, an alloy of this type was
developed at ORNL specifically for application in
molten-salt reactor systems, and is now designated as
Hastelloy N. Its average composition is 70Ni-16Mo-
7Cr-7Fe-0.05C wt %. The selection of salt constituents
with respect to their compatibility with such alloys has

been reviewed by Grimes'* and does not require
further elaboration here.

Values for the standard free energies of formation of
the structural metal fluorides, CrF,, FeF,, and Nik,,
signify that, as impurities in the reactor salts, iron and
nickel should be present in metallic form.

In the molten-salt reactor fuel salt, the fluid contains
small amounts of UF; generated in the equilibrium
reaction Cr? + 2UF;= 2UF; + CrF,. In power
operations with the MSRE the concentration of UF,
was adjusted occasionally by in-situ generation of U?*
22

Table 2.10. Comparison of nominal and analytical values of uranium concentration
in the MSRE fuel circuit — zero-power experiment

Sample No. Reactor Inventory Loop Inventory Additions Concentration U
U Salt U Salt U Salt Nominal Observed®
(kg) (wt %)

Run 3-I 145.60 4754.5 145.60  234.5 3.044 3.068b
FP 3-1 193.09 4831.0 47.49 76.5 3.997 4.016
FP 3-2 3.822 3.853
FP 3-3 214.52 4865.5 21.43 34.53 4,409 4.390
FP 3-4 219.19 4873.0 4,67 7.52 4,498

195,24  4340.4 5.498 4,446
FP 3-5 195.24 4340.4 4,498 4.486
FP 3-6 195.70 4341.1 0.4581 0.7394 4.508 4.493
FP 3-7 196.06 4341.7 0.3622 0.5847 4.516 4,452
FP 3-8 196.24 4342.0 0.1813 0.2926 4,520 4,484
FP 3-9 196.51  4342.5 0.2687 0.4337 4,525 4,484
FP 3-10 196.70 4342.7 0.1853 0.298% 4,529 4,500
FP 3-11 196.88 4343.0 0.1847 0.2981 4.533 4.480
FP 3-12 197.15 4343.5 0.2742 0.4426 4,539 4.497
FP 3~13 197.60 4344 .2 0.4527 0.7306 4.549 4.530
FP 3-14 197.97 4344.,8 0.3632 0.5862 4,557 4,475
FP 3-15 198.06 4345.0 0.0893 0.1441 4,558 4,490
FP 3-16 198.59 4345.8 0.5323 0.8589 4,570 4,498
FP 3-17 198.94 4346 .4 0.3516 0.5674 4,577 4,559
FP 3-18 199.12 4346.7 0.1723 0.2782 4,581 4.499
FP 3-19 199.56 4347 .4 0.442% 0.7136 4,590 4.602
FP 3-20 200.01 4348.1 0.4494 0.7253 4.600 4,535
FP 3-21 200.37 4348.7 0.3648 0.5887 4.608 4,540
FP 3-22 200.73 4349.3 0.3581 0.5779 4,615 4,571
FP 3-23 201.09 4349.8 0.3558 0.5742 4.623 4,604
FP 3-24 201.09 4349.8 4,623 4.618
FP 3-25 201.17 4350.0 0.0883 0.1426 4,625 4.623
FP 3-26 201.35 4350.3 0.1756 0.2833 4,628 4,593
FP 3-27 201.80 4351.0 0.4494 0.7252 4.638 4.669
FP 3-28 202.16  4351.6 0.3599 0.5760 4.646 4,646
FP 3-29 202.61 4352.3 0.4538 0.7324 4,655 4.676
FP 3-30 202.98 4352.9 0.3675 0.5932 4,663 4.688
FP 3-31 203.07 4353.0 0.0918 0.1482 4,665 4,679
FP 3-32 263.07 4353.0 4,665 4,642
FP 3-33 203.07 4353.0 4,665 4,599
FP 3-34 203.07 4353,0 4.665
FP 3-35 203.07 4353.0 4,665 4.634
Run 3-F 203.07 4353.0 4,665
Run 3-F circuit residue -0.77 -0.77 16,50 4,665

202.30  4353,0°
Drain tank charge 23.95 532.6
Run 4-1 226.25 4885.6 4,631

3Corrected for -0.8% bias.
Average for Precritical Experiment, PC-2.

cNeglects probability that mass of fuel and flush residues are not identical.
Table 2.11. Chemical analyses for MSRE fuel carrier salt

23

TLiF-BeF,-ZrF4 (64.7-30.1-5.2 mole %)

Batch Net wt. 7Li Analyses

of salt Assay Li Be Zr F Cr Ni Fe

(kg) (wt %) (ppm)

F-162 130.8 99.994 10.6 7.43 11.5 70.0 23 17 38
F-163 131.6 99.994 10.6 7.36 12,0 69.9 21 21 54
F-164 133.8 99.996 11.0 7.43  11.6  69.9 14 24 145
F-165 132.0 99.996 10.8 7.30 12.1 70.1 29 24 39
F-166 137.0 99.996 11.3  6.98 1l1.5 70.0 23 8 124
F-167 125.2 99.996 10.9 7.23 12.1 70.1 19 <3 26
F-168 133.1 99.996 10.8 7.30 12.0 70.0 13 3 33
F-169 132.6 99.995 11.1 7.38 12,0 69.6 15 16 36
F-170 134.5 99,995 10.5 7.37 11.8 70.3 20 7 46
F-171 132.9 99.995 10.69 6.98 11.90 70.3 17 17 91
F-172 134.2 99.995 10.61 7.33 11.72 70.3 26 16 89
F-173 133.1 99.995 10.50 7.29 12.12 70.0 18 24 79
F-174 133.8 99.995 10.66 7.57 11.99 70.0 31 17 141
F-175 133.6 99.995 11.17 7.05 11.97 70.37 19 <5 120
F-176 133.8 99.995 11.65 7.27 11.84 69.65 13 <5 80
F-177 133.8 99.995 10.58 6.95 11.92 69.77 29 26 90
F-178 131.1 99.995 11.03 7.37 12.09 69.91 28 7 84
F-179 131.1 99.995 16.79 7.47 12.24 69.75 33 11 106
F-180 133.4 99.995 11.19 7.49 11.72  70.45 29 40 198
F-181 134.8 99.995 10.84 6.90 12.48 69.1 25 29 141
F-182 134.2 99.995 10.73 6.89 12.44 69.3 27 29 240
F-183 133.0 99.994 10.99 7.21 12.35 69.80 22 8 53
F-184 132.8 99.995 10.90 7.16 12.28 69.27 23 <5 61
F-185 133.2 99.995 10.92 7.30 11.99 70.24 22 <5 58
F-186 133.5 99.995 10.99 6.86 12.10 69.59 19 <5 53
F-187 133.4 99.995 10.87 7.13 12.25 69.71 27 9 136
F-188 133.8 99.995 10.86 7.14 11.84 69.86 24 19 113
F-189 134.2 99.995 10.87 7.32 11.93 70.74 22 6 63
F-190 134.1 99.995 10.89 7.26 11.78 70.62 22 8 88
F-191 132.0 99.995 10.35 7.23 12.0 71.2 17 13 69
F-192 134.3 99.995 10.31 7.19 11.5 72.0 28 28 150
F-193 133.4 99.995 9.75 7.40 11.4 71.4 13 10 220
F-194 137.3 99.995 10.82 7.33 12.1 70.3 21 12 137
F-195 134.1 99.994 10.56 7.34 12.2 70.5 19 16 173
F-196 131.1 99.994 10.48 7.34 11.4 71.5 24 11 151
F-197 132.6 99.99%4 10.88 7.39 11.18 70.6 21 10 130
F-198A 146.4 99.994 4.68. U-62.26 33.60 8 29 66
F-198B 136.6 99.994 4,67 U-62,.82 33.82 10 <5 33

mn the salt or by addition of FeF, as an oxidant
Generally, an effort was made to hmut the [U>*]/[ZU]
ratio to ~0 0l as a means of retarding the corrosion
reaction

The reaction Ct® + 2UF, = 2UF; + C1F, has an
equilibrium constant with a small degree of temperature
dependence Since the equilibrium constant increases
with temperature, operation of a loop with a significant
temperature gradient causes the chromium concenira-
tion of the alloy surface that 1s at high temperature to
decrease, transferring metal to the cooler regions Since
chromium comprises a small fraction of Hastelloy N
and 1ts availability to the salt 1s diffusion limited,
attainment of equilibrium by this reaction 1s very slow

In prenuclear operations with nonuraniferous salts,
the salt loops were operated in essentially 1sothermal
conditions Under these conditions, corrosion of the
circuit walls could arise only from contaminants

In all the activities associated with the startup of the
MSRE, stringent measures were taken to remove all
traces of moisture from the newly constructed system
and to prevent subsequent introduction of impurities

Table 2.12. Chemical analyses for TLiR-? 35UF4 (73-27 mole %)
fuel concentrate for the MSRE

Chemical analysis

Batch

No Weight percent

Parts per mllion
L: U F Cr N1 Fe

E 201 479 6153 330 17 <5 26
E 202 4 80 6202 332 27 89 74
E 203 492 62 26 334 29 47 11
E 204 4381 6121 333 24 9 69
E 205 468 6159 330 21 14 22
E 206 470 6151 327 <5 21 <5

24

During October and November 1964, the fuel system
was readied for use, 1t was purged of moisture, heated
to 650°C, and tested for leak tightness The purge gas
was high-punty hehum To ensure that it was free of
oxygen and moisture 1n use, the gas flowed through a
series of dryers and preheaters and through hot tita
nium sponge before 1t entered the reactor

From the results of analyses of the individual salt
batches produced for use in the MSRE (see Tables 2 11
and 2 12) the average concentrations of Cr, Fe, and N1
in the blended salt charge (" LiF-BeF,-ZrF4-2*® UF, ) at
the beginning of the precritical experiment PC-2 should
have been 21, 99, and 14 ppm respectively The salt
samples removed from the fuel pump during this test
were supplied to the ORNL General Analytical Labora
tory (GAL) as well as to the HRLAL so the overlapping
results would show whether shight biases might result as
new equipment and necessary modifications in methods
were employed As determined by the GAL, the average
concentrations of structural metal impurities in the
PC-2 fuel circuit salt were Cr 19, Fe 99, N1 23 ppm,
while from the HRLAL, the average concenirations
were found to be Cr 37, Fe 163, N1 34 ppm (Table
27) Following run PC-2, "LiF ??3UF, enriching salt
was added to the carrier salt to constitute the final
235U fuel mixture The average concentrations of the
impurities on completion of this experiment as deter-
mined by the HRLAL were Cr 37, Fe 148, N1 55
ppm (Table 2 7)

With extended experience, we learned that one
standard deviation for Cr analysis was ~7 ppm, there
fore, the difference in chromium concentration as
measured 1n the two laboratories can be regarded as a
bias correction — that 15, HRLAL results could be
expected to average about 18 ppm higher than those
from the GAL

Table 2.13. Composition of the MSRE fuel salt 1n the precritical examnation

L1 Be Zr 237,934y Total
Carrier (norminal composition), kg 51691 309 62 541 36 45200
7L1F—238UF4 (nomunal composition), kg 1155 145 60 234 5
Total fuel salt (nominal), kg 528 46 309 62 541 36 145 60 47545
Weight percent 1112 651 1139 3062
Weight percent analytical (18 samples) 1042 6 60 11 35 3 068
Carrier (analytical composition), kg 487 75 327 61 539 06
"L1F-238UF, (analytical composition), kg 10 96 146 34 2345
Total fuel salt (analytical), kg 498 71 32761 539 06 146 34
Weight percent 1049 6 89 11 34 3077
Weight percent analytical (18 samples) 1042 6 60 1135 3 068

25

Table 2.14. Summary of MSRE salt analyses, experiment No. 4, fuel salt

Weight percent

Parts per mitlion

Date Sample ya
Ia Be VA b2 Te Cr N1 QO
Nominal Analytical
12/21/65  FP-4-15 800
12/21/65  FP-4-16€
12/22/65  FP-4-17 655
12/24/65  FP-4-18 10 25 668 1124 4651 6744 10026 144 43 44 859
12/22/65 FP-4-19¢
12/23/65  FP 4-20¢
12/24/65  FP-4-21 10 47 640 1082 4671 66 79 99 15 121 60 52
12/25/65  FP-4-22 10 54 654 1095 4 664 64 68 97 37 116 44 84
12/26/65  FP-4-23 1027 674 1086 4 642 65 06 97 57 99 46 35
12/27/65  FP-4.24 10 65 665 1096 4 655 67 66 10058 116 48 45
12/31/65  FP 4-25 10 60 637 1141 4 646 65 44 98 47 26 35 42
1/1/66 FP-4 26 10 60 663 1120 4 642 66 90 99 97 89 48 47
1/2/66 FP-4-27 10 55 653 1107 4618 67 68 10045 222 50 41
1/3/66 FP-4-28 10 60 642 1110 4 663 66 19 9897 211 49 34
1/5/66 'P-4-29 1070 671 1154 4 654 69 75 103 35 11T 39 31
1/7/66 FP-4-30 10 63 681 1119 4 661 67 32 100 58 83 49 27
1/10/66 FP-4-31 10 30 663 1126 4632 68 51 10133 190 37 41
1/12/66 FP-4-32 10 55 671 11 80 4625 6766 101 34 173 43 33
1/14/66 P-4 33 11200 675 1107 4596 66 25 99 97 55 50 39
t/17/66 FP-4-34 11355 649 1113 4 601 6820 10182 164 58 16
1/19/66 FP-4-35 1136/ 668 1086 4721 69 35 103 01 74 54 <5
1/21/66 FP-4-36 11255 632 1120 4632 6676 10016 125 47 <5
1/22/66 I P-4-37 11300 633 1108 4622 69 35 10268 189 53 80
1/24/66 FP-4-38 10 70 654 1146 4 608 67 25 10056 311 51 25
1/26/66 P4-39

@Values corrected to 33 696 at % >>°U
bur purge method

“No sample obtamed

dKBrF4 method

€For amperometric analysis

T Erroneously high because of weak batteries in automatic pipet

On completion of the fueling operation in July 1965,
the reactor was dramned and flushed Radiation levels
were Jow enough to permit mamntenance and mstalla-
tion work to begin mmmediately in all areas, n
preparation for high-power operation In August 1965,
the assembly of graphite and Hastelloy N surveillance
specimens, which had been 1n the core from the
beginning of salt operation, was removed While the
reactor vessel was open, mnspection revealed that pieces
were broken from the horizontal graphite bar that
supported the sample array The pieces were recovered
for examination, and a new sample assembly, designed
for exposure at high power and suspended from above,
was installed The fuel-pump rotary element was re-
moved 1n a final rehearsal of remote mamtenance and
to permut inspection of the pump nternals It was
remnstalled after inspection showed the pump to be in
very good condition

Tests had shown that the heats of Hastelloy N used 1n
the reactor vessel had poor high-temperature rupture
life and ductiity in the as-welded condition Since the
vessel closure weld had not been heat treated, the entire
vessel was heated to 760°C (1400°F) for 100 hr, using
the installed heaters, to improve these properties

The experience which was developed with the MSRE
throughout the shakedown peniod preceding power
operations confirmed that the molten-fluoride salt
mixtures were intrinsically noncorrosive to Hastelloy N
and that effective procedures were employed to prevent
serious contarmnation of the salt circuits during this
period

By December, maintenance and modifications were
completed, and flush salt was circulated through the
fuel system for a period of three days On December
20, 1965, the fuel salt was loaded into the system, and
the zero-power experiment began On completion of
26

Table 2.15. Chemical composition of the MSRE fuel salt in prenuclear operations

1 Be Zr U b3
Carrier salt, kg 516 91 309 62 541 36 45200
TLF-238UTF,, ke 1155 145 60 234 5
TLiF-235UF, (bulk charge), kg 584 7359 118 55
Total fuel salt, kg 534 30 309 62 541 36 21919 4873 05
Loop charge (89 07%), kg
Carrner salt 460 41 27578 482 19
LiF-238yF, 10 29 129 69
LiF 235UF, 520 65 55
Additions to loop 6 32 7 83 12 736
Total 43530
Loop charge, wt % 1108 6 34 1108 4 665
FP 3-32-35 (3), wt % 10 30 676 1120 4 624
Fuel salt residue in loop, kg 183 105 183 017 165
Salt to drain tank, kg 480 39 274 73 480 36 202 30 4336 5
Dramn tank charge, kg 58 40 3384 5908 2395 5326
53879 308 57 539 44 226 25 4869 1
Drain tank charge, wt % 1107 6 34 11 08 4 647
Charge to loop for run 4 (92%), kg 495 69 283 88 496 29 208 15 4479 6
Flush salt residue m loop, kg 251 1 66 180
Total fuel salt, kg 498 20 285 54 496 29 208 15 4497 6
Total fuel salt, wi % 11 08 6 35 1104 4628
Analytical (69 samples), wt % 10 785 + 0 064 6571 +0076 11197 20260 4641+ 0026

this period the concentrations of the structural metal
mmpurities had average values of Cr 48 £ 7, Fe 131 %
65, N1 40 = 20 ppm Here, for the first time, there was
evidence that an increase in the concentration of
chrommum in the fuel-salt mixture had taken place As
will be noted later, such mcreases followed a regular
pattern, occurring almost exclusively after periods of
maintenance when the reactor was opened to the
ambient atmosphere for some period of time It now
seems that circulation of the flush salt for periods of
only a few days was insufficient to negate the effect of
the exposure completely

2 4.4 Oxide contaminants. Among the summary
reports on various aspects of MSRE operations, no
separate review of analytical chemical developments is
included Except for a few analyses, for example, for
oxide and for lithium, the methods in the ORNL Master
Analytical Manual were employed routinely and found
to be satisfactory In response to the need for improved
methods for deternunation of the concentration of
oxides in the MSRE salts, a rehiable and accurate
method was devised and applied concurrent with early
MSRE operations This development was an important
aspect of MSRE chemical development and 1s therefore
described 1n detail here

Long before operations with the MSRE were started
1t was demonstrated 1n laboratory tests that precipita-
tion of oxide as the saturating phase in the molten fuel,
flush, or coolant salt was very unhkely However,
assurance that the concentration of oxide as a con-
tarminant of these salts was not increasing could not be
obtained from imtial measurements using conventional
methods '® Many of the results of early analyses of
MSRE flush and carrier salts were anomalous, as shown
in Tables 2 2,2 7, and 29 At face value, the analytical
data shown 1n these tables might suggest that the
concentration of oxide in the samples examined ranged
sporadically from 100 to 4000 ppm If real, the higher
values would represent the presence of more than 0 5%
Z10, 1n the specumens Each of the 52 specimens
obtained for analysis durning precrnitical run No 2 and
duning the zero-power expermment was subjected to
petrographic exanmunation With the carner and LiF-
BeF, salt, the sensitivity for detection of well-formed
crystals of ZrQ, by the petrographic methods 1s well
below 100 ppm, neither crystalline ZrO, nor oxy-
fluorides were found to be present in any of the
specimens examined

The sampling procedures employed at the MSRE
effectively protected salt specimens from contact with
moisture up to the pomnt where they were transferred
into a transport container, subsequent handling proce-
dures, however, were not so stringent and are presumed
to have allowed moisture to contact the surfaces of the
fluoride specimens on occasion Beryllium fluoride and
LiF-BeF, glassy phases are hygroscopic Crystallization
of each of the MSRE salt mixtures can produce minute
quantities of these phases, typical sampling conditions
effect rapid cooling of the samples and are therefore
even more conducive toward the production of the
hygroscopic phases than equilibrium crystallization of
the salt muxtures These hygroscopic phases, once
moistened, cannot be redried completely by ordinary
desiccants at room temperature The anomalously high
values for oxide analyses during the early stages of
MSRE operations are thus indicative of exposure of the
salt samples to moisture-laden atmospheres

An effort was nitiated to develop improved methods
for oxide analysis that would be adaptable for use n
the HRLAL After a study of several methods which
might have application under conditions such as those
which prevailed in the HRLAL and where no provision
for atmospheric control m the hot cells was made, the
analytical chemists'® concluded that a hydrofluorina-
tion-transpiration method based on the reaction O? +
2HFE(g) = H,0(g) + 2F  was potentially the most
useful Their development of the new method 1s
described 1n the following paragraphs

In princaple, the amount of water evolved from
purging a molten-sait sample with an H,-HF gas
mixture would serve as a measure of the quantity of
oxide 1n the molten-salt sample Since the water evolved
and the HF consumed are both proportional to the
amount of oxide present, either compound would serve
as an indicator species Water momitoring methods were
selected because of their greater convenience and
reliability than those for HF

The apphication of this method to the analysis of
radioactive samples required the development of (1) a
sampling technique which minimized atmospheric con-
tamination, (2) the incorporation of 4 water-measure-
ment techmque which was convenient for hot-cell
operations, and (3) the fabrication of a compact
apparatus for conserving the himited space available in
the hot cells It was necessary to adapt the sampling
techniques from methods already developed to obtain
samples for wet analysis By using a copper ladle of
nearly the same dunensions as the enriching capsules,
approximately 50-g samples were obtamed which could
be transported in the existing transport container
Exposure to the atmosphere was minumized by remelt-
ing and hydrofluonnating the entire sample n the

27

sample ladle The ladle was sealed 1n a nickel-Monel
hydrofluorinator, with a delivery tube spring-loaded
against the surface of the frozen salt Before the salt
was melted, the apparatus was purged with a hydro-
fluorinating gas mixture to remove water from the inner
surface of the hydrofluorinator and from the exposed
surface of the salt As the sample was melted, the
delivery tube was driven to the bottom of the ladle by
spring action for efficient purging of the sait

In all preliminary tests the water in the effluent purge
gas was measured by Karl Fischer titration While the
Karl Fischer reagent was shown to be remarkably stable
to radiation, the titration would have been excessively
difficult to perform routinely 1n the hot cell As an
alternative, an electrolytic mosture monitor was
adapted for these measurements Since such equipment
1s subgect to interference and damage from HF,' 7 1t was
pracfical to include a sodwm fluoride column 1n the
effluent gas traimn In operation at about 90°C 1t
removed HF from the etfluent gas without significant
holdup of water

A schematic flow diagram of the appatatus mstalled
in the HRLAL 1s shown in Fig 22 The apparatus 15
pictured m Fig 23 A modular design was selected to
facilitate any changes found necessary during com
ponent testing and to permit necessary repairs in the
hot cell Except for the hydrofluorination furnace, all
hot cell components were contained within a cubical
compartment, 16 1in on a side

Samples of the flush and fuel salt taken during the
December startup of the reactor were analyzed for
oxide Table 2 16 summanzes the results Figure 24 152
recording of the data output from analysis of the first
fuel-salt sample taken afier the fuel was loaded into the
reactor The results of the analyses by the hydro-
fluorination method were in good agreement with those
by the KBrF, procedure The KBrF, values paralleled
the trends shown by the hydrofluorination method but

Table 2.16. Oxide concentrations of flush and fuel salt
from the MSRE

Tume of salt crculation Oxide concentraion

Sample

(hr) (ppm)
Flush salt 247 46
29 1 72
47 6 106
Fuel salt 34 120
238 105
322 30
525 65

28

ORNL-DWG 65-8855A

MOISTURE
/ MONITOR

Pb SHIELD

/ He SPLITTER
> (¢ Dyl -
U T
PURIFIER / N\ LHe BACK FLUSH
/ 'f? | NoF - HF TRAP
N N\
N\
" N
CELL WALL § \
D

HYDROFLUORINATOR

Fig. 2.2, Schematic flow diagram of the apparatus for the determination of oxide in MSRE fuel by hydrofluorination.

averaged shightly higher. This bias was not unexpected,
since the pulverized samples required for the KBrF,
method were easily contaminated by atmospheric mois-
ture. Subsequently, 50-g fuel-salt samples were taken at
various power levels from zero to full power. With an
in-cell radiation monitor, the initial sample read 30 R at
1 ft. This activity increased to 1000 R at 1 ft at the
full-power level. Results of the oxide analyses are given
in Table 2.17.

The oxide in the coolant salt sample, 25 ppm, is, for
practical considerations, identical with a value of 38
ppm obtained for a coolant salt sample taken on
January 25, 1966, and analyzed in the laboratory after
three weeks’ storage. The fuelsalt analyses are in
reasonable agreement with the samples analyzed in the
development laboratory before the reactor was oper-
ated at power. The oxide concentration in these
nonradioactive samples seemed to decrease gradually
from 106 to 65 ppm. Between the FP-6 and FP-7 series
the sampler-enricher station was opened for mainte-
nance, and thus the apparent increase in oxide concen-
tration (ca. 15 ppm) may represent contamination of

Table 2.17. Oxide concentrations of coolant
and fuel salt from the MSRE

QOxide concentration

Sample Code (ppm)
Coolant salt CP-4-4 25
Fuel salt FP-6-1 49

FP-64 53
FP-6-12 50
FP-6-18 47
FP-7-5 66
FP-7-9 59
FP-7-13 66
FP-7-16 56

the samples by residual moisture in the sampling
system; however, the number of determinations was not
at the time sufficient to establish the precision of the
method at these low concentration levels.

In an attempt to determine whether radiolytic fluo-
rine removes oxide from the fuel samples, sample
FP-7-9 was removed from the transport container and
29

nd
&
®
©
T
o
T
a

cell apparatus.

. Hot

3

.2,

Fig
30

ORNL- DWG 66-402BA

40 H, FLOW, 200 cc/min
e HF, 0.2 atm
E FUEL SALT, 48.7¢q
> OXIDE FOUND, 120 ppm
3 30 - n,,300°C
o
|.:I->| —»Ha-HF,300°C ~SALT MELTED
i 20 \ r‘\
e
2
o
T 10 \

v | ¥
0 __’__/ —
0 20 40 60 80 100

TIME (min)

Fig. 2.4. Determination of oxide in MSR fuel sample FP4-11 by hydrofluorination.

stored in a desiccator for 24 hr prior to analysis. Since
the oxide content, 59 ppm, is comparable with that of
the remaining samples for which analyses were started 6
to 10 hr after sampling, no significant loss of oxygen is
indicated. A more direct method of establishing the
validity of these results by measuring the recovery of a
standard addition of oxide was used in subsequent
operations. A tin capsule containing a known amount
of SnQ, was heated to 550°C in the hydrofluorinator
as hydrogen passed through the system. The SnO, was
reduced to the metal, and the water that was formed
passed on to the electrolysis cell. Two standard samples
of Sn0O, were analyzed after a four-month interim, and
oxide recoveries of 96.1 and 95.6% were obtained.

The slight negative bias was attributed to momentary
interruptions in the flow of the hydrofluorinator
effluent gas through the water electrolysis cell. Diffi-
culty with cell plugging was encountered throughout
the period of development of the oxide method. As an
attempt to eliminate the negative bias and also to
provide a replacement cell for the remote oxide
apparatus, it was deemed necessary to find a method of
regenerating the electrolysis cell that would permit a
steady gas flow at relatively low flow rates.

The water electrolysis cell contains partially hydrated
P, O; in the form of a thin viscous film in contact with
two spirally wound 5-mil rhodium electrode wires. The
wires are retained on the inside of an inert plastic tube
forming a 20-mil capillary through which the sample
passes. The 2-ft-long tubing element is coiled in a helix
inside of a %-in.-diam pipe and potted in plastic for
permanence.

During the course of the investigation of the cell, it
was found that a wet gas stream in itself did not cause

the electrolysis cell to plug. It was also necessary for
curtent to be flowing through the cell for flow
interruptions to occur. This indicated that the hydrogen
and oxygen evolving from the electrodes create bubbles
in the partially hydrated P, Qs film, which then grow in
size sufficiently to bridge the capillary and form an
obstructing film.

After many unsuccessful approaches, an acceptable
solution to the problem was obtained by means of a
special regenerating technique which employed dilute
acetone solutions of H3PO, as the regenerating solu-
tions. This provided a desiccant coating sufficient to
absorb the water in the gas stream and gave a minimal
amount of flow interruptions during electrolysis. The
cells were thereafter successfully regenerated in this
manner and yielded oxide recoveries of 99.6 + 1.3%
from standard SnQ, samples.

2.4.5 Analysis of helium cover gas. As noted in Sect.
2.4.4, a reliable means for the determination of oxide in
the fuel salt did not exist when the initial experiments
with the MSRE began. The results of the oxide analyses
were perplexing, and they were almost certainly invalid
as indicators of contamination by moisture, since the
concentrations of structural metal contaminants in the
salt remained essentially constant. In order to obtain
some direct assurance that negligible inleakage of
moisture was occurring in the fuel system, temporary
measures were devised and applied. The method con-
sisted in bleeding off a small flow of the helium gas in
the pump bowl and passing it through a monitoring
system in the high-bay area.

Mean values for the concentration of HF in the
helium cover gas were obtained during the PC-2 and
zero-power experiments by adaptation of a continuous
internal electrolysis analyzer for gaseous fluorides
Through the cooperation of ORGDP personnel, data
were obtained regularly throughout prenuclear opera-
tions The results showed that HF was evoived from the
salt at a maximum mean value of 150 to 200 ug/hiter ' ®
The validity of this value 1s somewhatl questionable
because of the probable positive bias that particulate
fluorides would contribute The results are, therefore,
conservative to an indeterminate degree, but correspond
to the introduction of no more than 1 ppm of oxide
into the salt per day through the reaction

2H, O + Zrk, = 210, + 4HF |

an amount which would escape detection by other
methods

2 4 6 Lithium analysis In Sect 2 4 2 we noted that
chemical analyses of the fuel salt routinely indicated
that Be/Li ratios were significantly greater than for the
nomunal composition of the salt A careful examination
of the resulis for all the salt mixtures prepared for use
m the MSRE shows that if 1t 18 assumed that the
analytical results are correct for berylhum in the
coolant and flush salts, for beryllium and zirconium in
the carrier salt, and for urammum in the LiF-UF, salt
muxtures, and if a correction factor of 5 to 6% 1s added
to each of the results of hithium analyses, the average
results of composition analyses nearly coincide with
nomunal values

When this disparity was first noted, a reexamination
of the production records was made but did not
disclose any dispartty in the material balances that
would account for the anomaly in Be/Li results
Evidence from the purification-plant mventory data
indicated that the composition of the salt delivered into
the reactor system was of the design composition, 66 0
*+ 025 mole % LiF, 34 0 £ 0 25 mole % BeF, A sample
of the MSRE coolant, taken from one of the batches
loaded into the reactor, showed a hquid-sohd phase
transttion temperature of 457 6°C, thus 1s within 0 1°C
of the temperature of the equilibrium hquid-solid
reactions which can occur in LiF-BeF, muxtures richer
than ~33 mole % LiF Chemical analyses of this
material indicated 1ts composition to be LiF-BeF,
(63 63-36 37 mole %) The thermal data indicated,
however, that the matenal contained at least 65 5 mole
% LiF, and thus confirmed the composition indicated
by the weights of the maternals used in 1ts preparation

Application of the same correction factor to the
results of hthium analyses for either LiF-UF, or
LiF-BeF, mixtures effects equal improvement of the

31

match of analytical results with nominal values, we
must conclude therefore that a negative bias of 5 to 6%
existed m the lithium results Lithium was determined
in the MSRE salt samples by flame spectrophotometric
methods The results are compared with calibrated
standards, and regarded by the analytical chemisis to
match satisfactorily with these standards However,
view of the comncidence of values which results when a
correction of ~5% 1s applied to the results, we must
infer that a negative bias of that magnitude affected the
MSRE results

The lithtum used to produce the salt charges for the
MSRE was selected from stockpiled LiOH in which the
®L: had been depleted to 001% or less Assays were
available on each batch of LiOH, these analyses served
as criterta for selection of the material used 12-2° The
assays of the batches of LiOH which were to be used
for the MSRE ranged from 00072 to 0 0085% °La,
with the average of the batches which were to be used
for the flush salt and the fuel carrier salt as 0 0074%
After each converston of hydroxide to the fluoride,’
the hthium 1n each product batch was again assayed
before the LiF was used to make up the coolant, flush,
or carrier salt The assays of the batches of LiF used to
make up the fuel salt ranged from 0 004 to 0 006% and
averaged 0 0049% 21

In view of the relation of tritium production in the
MSRE to the 1sotopic composttion of lithrum, Hauben-
reich?? recently completed a full review of the possible
sources of triium and of the analyses on which
production estimates were based In connection with
that review, S Cantor obtained new analyses for ®Lt in
unused LiF-BeF,-ZrF, carrier salt The results comn
cided with those obtatned imtially for LiOH, but were,
for reasons still unknown, substantially hugher than the
~50 ppm which was previously used as the basis for
estimation of production rates 23

247 Examination of salts after 7ero-power experi-
ment. At termination of the zero-power experiment,
the fuel salt was drained for temporary storage n fuel
drain tank FD-2 Flush salt was circulated through the
fuel system for a 24-hr period and drained for storage
Four specimens of fuel salt were obtained from the fuel
dramn tank at intervals during the first 300-hr penod of
storage This practice was not continued after power
generation began because of the hazards which would
be mcurred by attempts to provide temporary access
for samphing radioactive salt One purpose of the effort
to sample salt from the dratn tank was to confirm that
the composition of the salt in the tank conformed to
that expected from mixing the newly constituted salt in
the fuel circuit with that of lower uranium concen-
tration m the dramn tank Any mismatch noted at this
time would serve as a base-hne correction to data
obtamed later during power operations A second
purpose was (o investigate the effectiveness of quiescent
storage of the salt in reducing the amounts of sus-
pended particles of the structural metals won and
nickel In retrospect, this period afforded the oppor-
tunity for securing a wide vanety of base-line data, an
opportunity that was exploited only to a modest
extent, 1 that only four salt samples were removed for
analysis Their compositions, as determined from chem-
ical analysis, are hsted in Table 2 18

The data shown in Table 2 18 indicate that approxi-
mately the same bias noted previously n the analysis of
carrier constituents continued to be evident and that
the disparity between the analytical data and the
nominal concentration of uranium was ~0 3%, which 1s
within the precision normally found for uramium
analyses

A comparison of the values for 1ron and nickel in the
samples drawn from the drain tank and those previously

32

obtained from the pump bowl shows no significant
difference  Ewidence confirmung this expectation was
provided later by analysis of two samples removed from
the pump bowl before the reactor was operated at
power Samples were withdrawn from the reactor n
enricher ladles and were transferred under an inert
atmosphere to the graphite crucible of an electro-
chemical cell assembly for electrochemical studies by D
L Manning The cell assembly and electrodes developed
for these electrochemical studies are described else
where 2% Average total concentrations of iron and
nickel in the melt, as determmed by conventional
methods, were about 125 and 45 ppm respectively Iron
and nickel, as cationic species in the molten fuel, are
electroreducible in the melt and can thus give voltam
metric reduction waves By voltammetry and by stand-
ard addition techniques, Mannmg found?®? that the salt
contained ~10 ppm of wron as Fe?" and that if nickel
were present as a cationic species, 1ts concentration was
below the limit of detection by voltammetry (<1 ppm)

Table 2.18. Composition of MSRE fuel-salt samples obtained from fuel dratn tank FD-2

Sample Composition {wt %) Composition (mole %)
No Li Be Zr U Lit BeF, ZiF, UF4
FD-2-10 1013 6 54 1110 4 580
FD-2-11 10 18 6 54 11 27 4 638
FD-2-12 10 55 6 68 11 95 4614
FD-2-13 10 55 622 1095 4613
Average 10 35 641 11 32 4611 63 38 30 46 532 0836
Nomunal 65 19 29 00 501 0 808
Sample Parts per mllion
No Cr Fe N1

FD-2-10

A? 37 141 83

B? <30 53 <30
FD-2-11

A 37 97 72

B <30 63 <30
FD-2-12

A 26 91 31

B <70 50 7
FD-2-13

A 38 147 132

B <70 50 6
Average

A4 35 119 80

A(3) 33 110 62
Run 3

A 3748 154 £55 48 £ 19

ZAnalysis by wet chemical methods
bAnalysm by spectrochemmcal methods
33

PHOTO 80844

INCHES

Fig. 2.5. Flush sait residue from volute of the pump bowl.

Since all previous transits of the salt were conducive
to retaining suspended metal in the salt, we felt that
possibly the finely divided metal (some threefold more
dense than the fluoride mixtures) would gradually setile
in drain tanks on conclusion of the zero-power experi-
ment. In an attempt to test this possibility, samples of
the static fuel salt were obtained from the fuel storage
tanks throughout a two-week period following the
zero-power experiment. The concentrations of the
structural metal impurities were not found to change
during this storage period. We concluded, therefore,
that thermal convection in the storage tank was
sufficient to prevent settling of fine metallic particles in
this tank and that the concentrations of these metals 1n
the salt might be expected to reman essentially
constant throughout the remainder of MSRE opera-
tions.

On removal of the fuel pump rotor for examina-
tion,?® approximately I kg of flush salt was found to
have been retained in the rotor flange area (Fig. 2.6)
and 1n the volute of the pump bowl (Fig. 2.5).
Additional small fragments of salt were found in various
adjacent locations: in the grooves of the shield-block
O-ring and in line 903, through which pump-bowl gas
was vented to an HF analyzer that was installed
temporarily in the prenuclear operational period to
analyze the composition of the off-gas. All salt speci-
mens were submitted for microscopic and/or chemical
analysis and were found to be entirely free of oxides or
oxyfluorides. Results of the chemical analyses per-
formed with these samples are listed in Table 2.19. A
minor oil leak allowed oil to enter the shield-block

Table 2.19. Chemical analysis of fuel pump salt samples

Chemical composition (wt %)

Sample locatton
L1 Be Zy U I

Labyrinth flange 1336 9.381 0.15 0.0177 7805
Pump volute 1370 9.71 0021 0.0195 77.68
Upper O-ring groove  9.81 6.9 10.87 4.294  62.57
Lower O-ring groove 12.1 0.910 0.295 72.76
Pump suction No chemical analysts run. Contained
dispersion of fine nongraphitic carbon.
1025 671 11.11 4.602  68.87
13.12 968 O 0 77.08

Fuel salt
Flush salt

section of the pump. Thermal decomposition products
of the pump oil were in contact with the salt specimens
in these areas and were found as partial films on some
of the salt (Fig. 2.5) or dispersed through fragments of
other specimens. In all locations where crystallized salt
deposits were found, it was evident that the molten salt
had not adhered to the metal surfaces. This was
particularly evident in the volute deposit, where the
molten specimen cooled rapidly and maintained the
surface angle tangency characteristic of sessile drops
which are free of contaminant oxides.

2.4.8 Appraisal of chemical surveillance in prepower
tests. From a chemical standpoint, operations of the
MSRE during the precritical and zero-power experi-
ments were performed in a manner that maintained the
purity of the salt charges during all transfer, fill, and
circulation operations. The results of the chemical
analyses obtained with samples from the fuel and
34

o o g nr o LA

PHOTO 81120

Fig. 2.6. Flush salt in rotor flange area.

coolant systems reinforced the longstanding conclusion,
drawn in laboratory and engineering-scale experiments,
that pure molten-fluoride mixtures are completely
compatible with nickel-based alloys.

The utility of routine analysis of salt samples removed
from the circulating systems came to be realized as of
questionable value with respect to several of the
constituents after brief experience with the MSRE. It
became quite clear, for example, that only easily
recognized operational perturbations were likely to
cause the average concentrations of carrier or coolant
salt constituents to vary beyond standard deviations in
the analyses and that without major refinement in the
analytical methods for determining the concentration
of uranium in fluoride mixtures these analyses would
not provide operational control. Nonetheless, the
potential value for applications of statistically signifi-
cant results justified, in our judgment, a continued and
active program of chemical analysis. Furthermore, by
the time the zero-power experiment was completed, the
value of chromium analysis as the single quantitative
indicator of corrosion or of corrosion-free performance

was solidly established. Analysis of salt samples for this
impurity involves wet chemical dissolution techniques,
which account for some 90% of the cost of a complete
analysis. Because of this factor, there was little reason
to omit analysis of more than one component, for such
omission would preclude determination of overall com-
position.

Several aspects of the relation of analysis to MSRE
operations proved to be discomfiting by the completion
of the zero-power experiment. Annoyingly, chemical
analysis of the flush-salt samples removed from the
system indicated that the amounts of salt residues
remaining in the system after drain operations were
somewhat more than could be accounted for by the
on-site estimates of the volumes of probable void space
where it was deduced that salt could reside. Ultimately,
the application of neutron activation and mass spec-
trometric analysis to this problem confirmed that
estimates made from chemical data were the most
nearly correct. Until this problem was resolved, con-
tinued exchange of fuel and flush salt by drain-fill
operations suggested a negative bias in the analysis of
uranium, Methods were developed for slight refine-
ments in the uranium analysis, but not surprisingly,
these did little to resolve the fundamental problem.
Weigh cell data, in which some reliance was placed
originally, proved to be of little value with continued
experience, and accordingly, increasing reliance on
chemical analyses was developed.

For a lengthy period, there was no absolute method
for establishing the level of contamination of the salts
by small amounts of oxide ion. This assay was
performed in a partly satisfactory manner during the
precritical and zero-power experiments, but the method
was not adapted for application to highly radioactive
materials until well into the power period.

The lack of dynamic methods for obtaining various
analyses of the gas and vapor above the surface of the
fuel salt proved to be a severe limitation in the
interpretation of the chemical behavior which occurred
in the fuel system. In the absence of on-site gas analysis,
either by mass spectrometry or gamma-ray spec-
trometry (which was developed later for postopera-
tional examinations), interpretation of several aspects
of molten-salt reactor behavior remained less than
completely resolved — in particular, of greater signifi-
cance in power operations, the experimental investi-
gation of factors controlling the transport of noble
metal fission products.

In general, the surveillance program was, to this stage,
a success principally because of the careful procedures
which were employed to ensure its trouble-free opera-
tion and because of the intrinsic stability and compati-
bility of the materials used in its construction.

References

1. J. H. Shaffer, Preparation and Handling of Salt
Mixtures for the Molten-Salt Reactor Experiment,
ORNL4614 (January 1971).

2. MSR Program Semiannu. Progr. Rep. Feb. 28,
1965, ORNL-3812, p. 5.

3. R. B. Lindauer, Processing of the MSRE Flush and
Fuel Salts, ORNL-TM-2578 (August 1969); G. Long
and F. F. Blankenship, The Stability of Uranium
Trifluoride, ORNL-TM-2065 (November 1969),

4. R. B. Lindauer, private communication.

5. C. F. Baes, “The Chemistry and Thermodynamics
of Molten-Salt Reactor Fluoride Solutions,” ITAFA
Symposium on Thermodynamics with Emphasis on
Nuclear Materials and Atomic Transport in Solids,
Vienna, Austria, 1965.

35

6. A more detailed description of the hydraulics in
the system is given in the report by J. R. Engel, P. N.
Haubenreich, and A. Houtzeel, Spray, Mist, Bubbles
and Foam in the Molten-Salt Reactor Experiment,
ORNL-TM-3027 (June 1970).

7. MSR Program Semiannu. Progr. Rep. Feb. 28,
1965, ORNL-3812, p. 10.

8. MSR Program Semiannu. Progr. Rep. Feb. 28,
1965, ORNL-3812, p. 10.

9. L. T. Corbin, W. R, Winsbro, C. E. Lamb, and M.
T. Kelley, “Design and Construction of ORNL High-
Radiation-Level Analytical Laboratory,” /1th Hot Lab
Conf. Proceedings, ANS, November 1963; “High-
Radiation-Level Analytical Laboratories,” Anal. Chem.
37(13), 83A (1965).

10. B. E. Prince et al., Zero-Power Physics Experi-
ments on the MSRE, ORNL-4233 (February 1968).

11. J. R. Engel and B. E. Prince, The Reuactivity
Balance in the MSRE, ORNL-TM-1796 (March 1967).

12. MSR Program Semiannu. Progr. Rep. Feb. 28,
1966, ORNL-3936, p. 122.

13. J. H. Shaffer, internal correspondence, Oct. 9,
1964,

14. W. R. Grimes, Nucl. Appl. Technol. 8, 137
(1970).

15. G. Goldberg, A. S. Meyer, Jr., and J. C. White,
Anal. Chem. 32,314 (1960).

16. A. S. Meyer, R. F. Apple, and J. M. Dale, ORNL
Analytical Chemistry Division.

17. W. S. Pappas, Anal. Chem. 38, 615 (1966).

I8. Results summarized in a memorandum from J. G.
Million (ORGDP) to R. E. Thoma, July 14, 1965.

19. P. N. Haubenreich, “Selection of Lithium for
MSRE Fuel, Flush and Coolant Salt,” internal corre-
spondence, Apr. 10, 1962.

20. H. F. McDuffie, “Selection of "Li Batches for
MSRE Fuel, Flush and Coolant Salt,” internal corre-
spondence, May 22, 1962.

21. J. H. Shaffer, personal communication.

22. P. N. Haubenreich, A Review of Production and
Observed Distribution of Tritium in MSRE in Light of
Recent Findings, ORNL-CF-71-8-34 (in press).

23. P. N. Haubenreich, Tritium in the MSRE: Calcu-
lated Production Rates and Observed Amounts, ORNL-
CF-70-2-7 (Feb. 4, 1970).

24. D. L. Manning, internal correspondence, Jan. 19,
1966.

25. C. H. Gabbard, Inspection of Molten-Salt Reactor
Experiment (MSRE) Fuel Pump after Zero Power
Experiments (Run 3), ORNL-CF-66-8-5 (Aug. 4, 1966).
3 CHEMICAL COMPOSITION OF THE FUEL SALT
DURING NUCLEAR OPERATIONS

3 1 Introduction

The MSRE was not the first reactor to use a
molten fluoride fuel mixture The ARE (Aircraft Re-
actor Experiment) was operated as the 1nitial demon
stration of a molten-salt reactor in November 1954,
using a fuel solution of NaF-ZiF,;-2?°UF, (53-41-6
mole %) ! The time during which the molten fluoride
fuel was circulated was brief, ~1000 hr, and did not
include a program of chemical surveillance In contrast,
the MSRE was intended to provide the experience and
operational data on which design and plans for molien-
salt reactors of 30-year operating lifetimes could be
based Plans for its operation, therefore, included a
chemical surveillance program that was comprehensive
n scope Moreover, chemical information and quantita-
tive data pertaiming to a number of phenomena that
could only be obtained from operating reactors were
lacking, even though a thorough program of laboratory,
engineering-scale, and in-pile tests had established the
principal parameters required to design and operate the
MSRE Among the most important of these are the
chemical factors which control the distribution of
fission products within the fuel and off-gas systems, and
the redox chemistry typical of dynamuc fuel systems in
which fuel burnup and replenishment proceed concur-
rently It was also important to establish the corrosion
resistance of Hastelloy N under realistic and typical
operating conditions, to examine the capability of a
molten-salt reactor to operate under circumstances
where contaminants might enter the salt occasionally or
chronically, and to study the possible effects that
operation of numerous chemical components of the
reactor might have on fuel and coolant chemustry It
was the mtent of these efforts to evaluate, on a
continiing basis, the adequacy of the surveilance
program to assess the performance and safety of the
MSRE as the prototype of a molten-salt breeder
reactor

Imtial activities of the program of chemucal surveil-
lance served to establish that the behavior of the fuel
and coolant salts would conform with that expected
from laboratory and engineering tests We sought
information that would indicate whether the fuel and
coolant salts remained chemically stable and noncorro-
sive, to what extent the consumption of fissile matenal
deviated from that expected from nuclear considera-
trons, and whether or not urammum losses could be
detected by chemical examination of salt samples In

36

the beginning stages of MSRE operation, samples were
taken on one-day intervals to assess the chemical
stabiity of the fuel, to establish whether or not the
concentration of oxide contamunants remained below
the saturation lumts (~700 ppm for the average
operating temperature of 650°C), and to correlate
concomitant corrosion, if 1t occurred, with possible
introduction of contaminants into the system Finally,
we anticipated that through measurement of 1sotopic
changes 1n the uranium composition of the fuel it might
be possible to obtain accurate comparisons of the
amounts of uranium burned with those expected from
nuclear considerations

After a significant amount of uranium had undergone
fission, procedures were imtiated to monitor the rela-
trve concentration of trivalent uranium 1n the fuel salt
Studies of the chemical effects of uranwum fission 1n
molten-fluonde muxtures had led to an early estimate
that the fission reaction would be mildly oxidative and
would produce ~0 8 equivalent of oxidation per gram
atomic weight of urantum consumed This estimate was
based on an appraisal of the yield and chemucal stability
of the state of the fission products Since 1t was possible
to evaluate only the equilibrium states of all the various
fission product species — for example, the rare earths,
iodine, tellurium, and the “noble” metals, including
Mo, Nb, Ru, Ta, Re, and T¢ — nonequilibrium behavior
in highly radioactive environments of the MSRE would
not have been surprising and would conceivably intro-
duce uncertainties into interpretation of fission product
behavior 1n the MSRE Previous experience with fission
product behavior had been explored almost exclusively
through in-pile capsule tests, although some appraisal of
the ARE fission product distribution was made 2 A
vigorous 1nvestigation of the fate of fission products in
the MSRE was therefore imnitiated and pursued through-
out the entire period of reactor operations The results
of those studies are summarized elsewhere 3 Concur-
rently, laboratory studies were developed 1n support of
the on-site efforts to establish the relationships of
fission product distributions in the MSRE with chemi-
cal parameters During the course of MSRE operations
it became 1ncreasingly evident that munor variations in
the oxidation-reduction potential of the fuel salt almost
certainly had a real effect on some of the fission
product distributions within the system and that
accurate means for the determination of the concentra-
tion of trivalent uranum 1n MSR’s would be needed
Active work was devoted thereafter to the development
of means for analysis of U3+/ZU

Based on the considerations discussed above, a pro-
gram of chemical surveillance was formulated and
Table 3.1. Composition of the MSRE fuel salt

(Weight %) (ppm)_ 3+

Date Sample Equiv. Li Be Zr F Z Fe Cr Ni U in Au (kg) L Eq. L Eq. [Uu” /Iu]

Full N Obs Circula- Oxida~  Reduc~ %

P޴?r om. : tion tion tion Nom.
2/13/6¢ Run 5-1 4.631 209,950 0 0 3.633 0.41
2/14/66 FP 5-1 4 10.65 6.53 11.45 4,631 4,625 68.18 10Ll.44 68 51 15 0 0 3.633 0.41
2/16/66  Run 5-F
475766 Run 6-1 a
476766 FP 6-1 Concentration of oxide: 49 ppm
4/6/66 - Fuel circuit dralned because of restriction in valve No. 561.
4/8/66 - Fuel circuit filled
4/9/66 FP 6-2 11.43 6.22 10,80 4,631  4.622 68.97 102,07 77 57 63 209.950 0 0 3.633 0.41
4/13/66 FP 6-3 Sample for oxide analysis: analysis unsuccessful
4/14/66  FP 6-4 Concentration of oxide: 55 ppm®
4/15/66 FP 6-5 15 10.45 6.37 11.12 4.630 4.605 69.85 102.40 234 65 35 209.944 -0.006 -0.005 3.628 0.41
4/16/66  FP 6-6 15 10.55 6.46 11.32 4.630 4.625 68.56 101,52 101 46 49
4/17/66 FP 6-7 10.32  6.41 11.42 4,630 4,647 69,67 102.47 71 46 58
4/19/66 FP 6-8 30 10.43  6.49 11.29 4,630 4.655 68,90 101,77 89 45 58 209.938 -0.012 0.039 3.594 0.41
4/20/66 FP 6-9 39 10.48 6,66 11.52 4,630 4,684 68.92 102.26 168 58 55 209.934 -0.016 0.050 3.583 0.40
4f25/66  FP 6-10 55 10.30 6.76 11.54 4.630 4,595 68.81 102.00 129 42 36 209.928 -0.022 0.070 3.563 0.40
5/9/66 FP 6-11 115 10.30 6.42 11.28 4.630 4.612 70.09 102.70 95 59 125 209.904 =0.046 0.147 3.486 0.40
5/10/66  FP 6-12 138 Concentration of oxide: 50 ppm®
5/11/66 FP ©-13 147 10.50 6,41 11.06 4,629 4.628 66.79 99.38 101 52 54 209,891 -0.059 0.188 3,445 0.39
5/13/66 FP 6-14 166 10.50 6.45 11.33 4.629  4.617 68.18 101.08 107 48 74 209.884 ~0.066 0.212 3,421 .39
5/14/66 FP 6-15 190 10.55 b6.86 11.35 4,629  4.601 68.18 101.54 94 45 52 209.874 -0.076 0.243 3.390 0.38
5/18/66 FP 6-16 254 10,50 6.58 11.31 4.628 4.629 67.88 100,90 84 45 36 209. 848 -0,102 0.326 3.307 0,37
5/23/66 FP 6-17 322 11.44 6,64 11.05 4.628  4.652 68.26 102,04 122 49 54 209.822 -0.128 0.411 3.222 0.36
5/25/66  FP 6-18 Concentration of oxide: 47 ppm?
5/26/66 FP 6-19 400 10,40 6.88 11.72 4.627 4,667 69.10 102.77 99 39 39 209.790 -0.160 0.512 3.121 0.35
5/28/66 Run 6-F 400 4,627 209, 790
6/12/66 Run 7-1 400 4.627 209.806
6/12/66 FP 7-1 400 10.60 6.78 11.16 4.627  4.647 69.26 102,45 108 51 70 209, 806
6/15/66 FP 7-2 Sample for oxide analysis: sample unsatisfactory for analysis
6/17/66 FP 7-3 487 10,55 6.56 11.44 4,626 4.656 68.24 101.45 148 50 51 209,756 -0.194 0.623 3.010 0.34
6/20/66 FP 7-4 526 10, 50 6.40 11.61 4.626  4.640 67.82 100,97 110 52 38 209.740 -0.210 G.673 2.960 0.34
6/22/66 FP 7-5 Concentration of oxide: ppm?
6/24/66 FP 7-6 658 10.60 6,63 11.35 4.625 4,614 69,22 102.41 115 o6l 66 209.687 -0.263 0.843 2.790 0.32
6/26/66 FP 7-7 706 10.55 6.65 11.13 4.624  4.641 69.60 102.57 78 46 54 209.668 -0.282 0.904 2.729 0.31
7/1/66 FP 7-8 725 10.60 6,59 11.67 4.624  4.663 67.87 101.39 94 49 52 209.661 -0.289 0.927 2.796 0.31
7/4/66 FP 7-9 Concentration of oxide: ppm*2
7/6/66 FP 7-1C 846 10,63  6.91 11,21 4.623 4.609 68.20 101.5¢6 36 49 <20 209.612 -0.338 1.084 2.549 0.29
7/10/66 FP 7-11 944 10,45 7.00 11.22 4,622 4.630 69,48 102.78 58 39 36 209,573 -0.377 1.210 2.423 .27
7/13/66 Fp 7-12 1012 10,50 6.65 11.04 4,622 4,640 68.44 101,27 84 46 45 209.546 -0.404 1.300 2.333 0.26
7/15/66  FP 7-13 Concentration of oxide: ppm?
7/18/66 FP 7-14 1032 10.55 6.50 11.26 4,622 4.660 68.79 101.76 53 44 63 209.538 -0.412 1.321 2.312 0.26
7/20/66  FP 7-15 1047 16.55 6.71 11.60 4.621 4.638 67.59 101,08 83 44 28 209.532 -0.418 1.341 2.292 0.26
7/22/66  FP 7-16 Concentration of oxide: ppm@
7/25/66  Run 7-F 1047 4,621 209.532 -0,418 1.341 2.292 0.26
9/25/66  Run 8-1 1047 4.603 208,711
10/8/66 FP 8-5 1047 11.10 6.23 11,21 4,603 4,643 71,21 104,39 139 108 208.711 -0.434 1.341 2.292 0.26
10/11/66 FP 8-6 1100 10,95 6.37 10.97 4,603 4,624 70.24 103,15 133 61 72 208.706 ~0.439 1,408 2,225 0.25

LE
Table 3.1 (continued)

(ppm)

{(Weight %) "

Date Sample Equiv Li Be Zr U F L Fe Cr Ni LU in AU (kg) L Eq. £ Eq. ([u” /ru]

Full Nom Oos Circula- Oxida- Reduc- z

nger ! ’ tion tion tion Nom.
10/13/66 FP 8-7 Concentration of oxide: 44 ppm?
10/17/66 FP 8-8 1166 10.95 6.88 11.23 4,603  4.638 69.94 103.64 219 65 35 208,680 ~-0.465 1.491 2.142 0.24
10/19/66 FP 8-9 1200 11.00 6.60 10,79 4.602 4.626 72.27 105.29 124 66 42 208,666 -0.479 1.536 2.097 0.24
10/21/66 Fp 8-10 1228 11,00 6.43 11.41 4.602 4.630 67.55 101.02 89 73 119 208.655 -G, 490 1.572 2.061 6.23
10/24/66 FP 8-11 Sample for oxide analysis: analysis unsuccessful
10/26/66 FP 8-12 1357 11.15 6.62 11.20 4,601  4.650 69.01 102.63 106 51 26 208.603 -0.542 1,738 1.895 0.22
10/28/66 FP 8-13 1380 14,25 6.49 11.29 4.601  4.623 68,82 105.47 84 63 60 208,594 -0.551 1.767 1.866 0.21
10/31/66 FP 8-14 1401 13,85 6.65 11.17 4.601  4.620 69.08 105,37 82 63 25 208.586 ~-0.559 1.793 1.840 0.21
10/31/66 FRun 8-F 1401 4.601 208.586
11/7/66  Run 9-1 4.582 207,754
11/7/66 FP 9-1 1401 10.93 6.55 11,70 4,582 4.618 68.39 102.19 140 66 50 207.754 -0.559 1.793 1.840 0.21
11/9/66  FP 9-2 1449 Concentration of oxide: 44 ppm?
11/11/66 FP 9-3 1480 10,95 6.60 10,97 4,582 4,621 69.06 102.20 140 65 58 207.723 ~0.591 1.895 1.738 0.20
11/14/66 FP 9-4 1528 [u3t/zu] = 0.10% 207,703  -0.610  1.956 1.677  0.19
11/16/66 FP 9-3 1544 10.95 6.64 10.96 4,581 4.618 67.96 101,13 145 56 26 207,697 -0.616 1.976 1.657 0.19
11/17/66 FP 9-6 Concentration of oxide: ppm
11/19/66 TFP 9-7 1662 - 6.71 10.95 4.580 4,557 68.79 176 57 74 207.650 -0.663 2.126 1,507 0.17
11/20/66 Run 9-F 1662 4.580 207.650 -0.663 2.126 1.507 0.17
12/12/66 Run 10-1I 1662 4.561 206.815
12/12/66 FP 10-3 1662 Sample for determination of organics in offgas; sample faulty
12/12/66 FP 10-4 Sample for determination of organics in offgas by Cu0; 600 ppm HF
12/14/66 FP 10-5 1662 11.10 6.47 11.00 4,561 4.608 66.65 99.83 181 58 93 206,815 -0.663 2.126 1.507 0.17
12/16/66 FP 10-6 1662 11,05 6.25 11.13 4,561  4.612 67.57 100.61 168 62 55 206,815 -0.663 2.126 1.507 0.17
12/19/66 FP 10-7 1677 11.20 6.82 11,27 4.561 4.605 66.98 100.88 174 53 65 206.809 -0.669 2,146 1.487 0.17
12/22/66 FP 10-8 1746 11.20 6.69 11.13 4,561  4.604 67.21 100.83 149 56 62 206.781 -0.697 2.235 1.398 0.16
12/24/66  FP 10-9 1794 11.05 6.51 10.8% 4.560 4,575 70.20 103.22 167 62 49 206.762 -0.716 2.296 1.337 0.15
12/27/66  FP 10-10 50 g sample for oxide analysis; no results available
12/27/66  FV 10-11G Freeze valve capsule-gas sample
12/28/66 FP 10-12 1813 11.10 6.66 11.24 4,560 4.577 67.72 101.29 91 63 48 206.754 -0.724 2.322 1.311 0.15
12/30/66 FP 10-13 1846 11.30 6.84 11.09 4.560 4,628 65.82 99.68 162 61 82 206.741 -0.737 2.364 1.269 0.15
1/1/67 FP 10-14 1870 Be addition: 3 g as powder 206.732 -0.746 2.393 1.906 0.22
1/2/67 FP 10-15 Special sample (50 g}; not subjected to chemical analysis
1/3/67 FP 10-16 1920 Be addition; 1 g as powder 206,712 -0.766 2.457 2.064 0.24
1/3/67 FP 10-17 1920 11.15 6.47 10.88 4.559 4.633 69.03 102.16 102 64 166 206.712 -0.766 2.457 2,064 0.24
1/4/67 FP 10-18 1948 Be addition; 1.63 g as rod 206,701 -0.777 2,492 2.391 0.27
1/6/67 Fp 10-19 1968 11.15 6.64 11.03 4.559 4,621 67.75 101.19 145 62 62 206.693 ~0,785 2.518 2.365 ¢.27
1/9/67 FP 10-20 2161 11.08  6.40 10.82 4,557 4,622 69.30 102,22 164 57 57 206.615 -0.863 2.768 2.115 0.24
1/11/67 FP 10-21 50 g sample for oxide determination
1/11/67 FV 10-22G Freeze valve capsule~gas sample
1/13/67 FP 10-23 2247 Be addition: 10.65 g as rod 206,581 -0.897 2,877 4.369 0.50
1/13/67 FP 10-24 2247 11,23  6.55 11.21 4.556  4.606 68.73 102.36 151 56 87 206.581 -0.897 2.8717 4.369 0.50
1/15/67  FP 10-25 2288 50 g sample for UM/IU analysis; UX/IU = 0.66%C 206.565  ~0.913  2.928 4.318  0.50
1/16/67 Run 10-F 2288 4.556 206,565 -0.913 2.928 4,318 0.50
1/28/67 Run 11-~I 2288 4.556 206,585 -0.913 2.928 4,318 0.50
1/28/67 Fp 11-1 2288 11,18 6.28 10.90 4,556 4.603 67.46 100.42 131 66 54 206,585 -0.913 2.928 4,318 0.50
1/30/67 FP 11-2 2308 11.10 6.27 10.65 4.556 4,599 69,16 101,78 112 75 63 206.577 -0.921 2.954 4.292 0.49
2/1/67 FP 11-3 2331 16.42 6.41 11.08 4,556 4.606 66.72 99.03 150 61 64 206,568 -0.930 2.983 4,263 0.49

Q

8¢
Table 3.1 (continued)

(Weight %) (epm) 3+
Date Sample Equiv. Li Be Zr u ¥ L Fe Cr Ni U in AU (kg) L Eq. L Eq. [U” [EU]
Full T o6 Circula- Oxida-  Reduc- A
Paver ome s tion tion tion No.

2/3/67 FP 11-4 2372 11.10 6.33 11.10 4.556 4,592 67.87 100.99 145 62 22 206.551 ~-0.947 3.037 4.209 0.48
2/6/67 FP 11-5 2468 50 g sample for U'/IU analysis; UI*/IU = 0.60%¢ 206.513 -0.985 3.159 4,037 0.47
2/8/67 FP 11-6 2516 11.25 6.31 10.97 4,554  4.555 67.44 100.52 131 67 33 206.494 ~1.004 3.220 4.026 0.46
2/10/67 FL 11-7 2541 11.38  6.70 11.27 4.554 4,558 69.92 103.83 172 62 50 206.484 -1.014 3.252 3.994 0.46
2/13/67 FP 11-8 2614 11.50 6.62 10.81 4.553  4.569 68.89 102.39 312 54 107 206,455 -1.043 3.345 3.901 0.45

FV 11-9G Freeze valve capsule-gas sample
2/15/67 FP 11-10 2663 Be metal addition; 11.66 g as rod 206,435 ~-1.063 3.409 6.424 0.74
2/171/67 FP 11-11 2779 10.67 6.57 10.87 4,552 4,551 69,94 102,60 165 73 43 206.389 -1.109 3.557 6.276 0.72
2/21/67 FP 1i-12 2811 10.93 6.27 10.88 4,552 4,567 69.96 102.60 76 75 34 206.376 -1.122 3.598 6.235 0.72
2/22/67  FP 1i-13 2835 50 g sample for U¥t/IU analysis; UX/IU = 0.69%C 206.366  -1.132  3.631 6.202  0.71
2/24/67 FP 11-14 2884 10.98 6.35 11.26 4,551 4,525 67.83 100.95 98 78 75 206.347 -1,151 3.691 6.142 0.71

FP 11-14 2884 4,551  4.539 206.347 -1.151 3.691 6.142 0.71
2/28/67 FP 11-15 2950 11.11 6.49 10.74 4,551 4,552 68.89 101.78 71 62 37 206,321 -1.177 3.775 6.055 0.70
2/28/67 FV 11-16G Freeze valve capsule-gas sample
3/1/67 FP 11-17 2994 11.33 6.47 11.21 4,550 4,553 70.25 103.81 67 58 47 206,303 -1,195 3.833 6.000 0.69
3/2/67 FP 11-19 3014 10.73  6.67 11.17 4.550 4,576 68.68 101.83 120 56 43 206.295 =1,203 3.858 5.975 0.6%
3/3/67 FP 11-19 3038 10.47 6.57 11.11 4.550 4.589 67.12 99. 86 117 68 42 206.285 -1.213 3.890 5.943 0.68
3/6/67 FP 11-20 3109 10.51 6.72 10.98 4.549  4.561 67.60 100.37 122 59 49 206,257 -1.241 3.980 5.853 0.67
3/8/67 FP 11-21 3126 10.50 6.36 10.89 4,549  4.576 66,83 99.16 168 63 46 206.250 -1,248 4,003 5.830 0.67
3/9/67 FP 11-22 3126 10.55 6.57 10.92 4.549 4,572 66.05 98.66 104 62 63 206.250 -1.248 4,003 5.830 0.67
3/10/67 FP 11-23 3126 10,53 6.49 10.95 4,549 4,583 69.70 102.25 136 63 55 206.250 ~-1.248 4,003 5.830 0.67
3/13/67 FP 11-24 3169 10.55 6.51 10.90 4,549 4,547 68,50 101.01 173 67 63 206,233 -1,265 4,057 5.776 G.66
3/16/67 FP 11-25 50 g sample for oxide analysis; analysis unsuccessful
3/20/67 FP 11-26 335% 10.43 6.49 10,91 4,547 4,570 67.17  99.57 118 52 53 206,157 -1.341 4.301 5.532 0.64
3/21/67 FP 11-27 3380 10.52 6.85 10.85 4,547 4,577 66,04 98,84 72 63 80 206,149 -1.349 4,327 5.506 0.63
3/21/67 FP 11-28 Concentration of oxide: 58 ppm
3/22/67 FP 11-29 3407 10.48  6.46 10.92 4,546 4,584 64,62 97.06 126 64 56 206.138 -1.360 4,362 5.471 0.63
3/23/67 FP 11-30 3460 10.48 6.39 11.02 4.546 4,597 64,51 96,99 115 66 71 206,117 -1.381 4,429 5.404 0.62
3/27/67 FP 11-31 3524 10.53 6.58 11.06 4,546 4.559 67,06 99.79 80 64 50 206.092 -1,406 4,509 5.324 0.61
3/28/67  FP 11-32 13548 50 g sample for UM/IU analysis; U3*t/IU = 0.045%C 206.082  -1.416  4.541 5.292  0.61
3/29/67 FP 11-33 3571 10,53 6.43 11,14 4,545 4,567 66.38 99,04 142 72 72 206.073 -1.425 4,570 5.263 0.61
3/31/67 FP 11-34 3619 10.55 6.33 11.37 4,545 4,582 68,99 101.82 146 64 64 206.054 =1.444 4,631 5.202 0.60
4/3/67 FP 11-35 3690 10.53  6.35 11,12 4.544 4,566 67.24 99.81 194 73 64 206,025 =1.473 4,724 5.109 0.59
4/4/67 FV 11-36G Freeze valve sample; capsule penetrated by leaching solution
4/5/67 FP 11-37 3739 10.55 6.33 10.75 4,544 4,541 65.93 96,10 79 80 49 206.006 =1.492 4.785 5.048 0.58
476/67 FP 11-38 3763 50 g sample for U/IU analysis; U3t/LU = b 205.996  -1.502  4.817 5.016  0.58
4/7/67 FP 11-39 3856 11.57 6,44 10.92 4.543  4.536 66,55 100,02 182 69 52 205.959 -1.539 4.936 4.897 0.56
4/10/67  FP 11-40 3856 Be addition: 8.40 g as rod 205.959  -1.539  4.936 6.761  0.78
4/10/67 FP 11-41 3856 10.42 6.37 10.77 4.543 4,579 68.58 100.72 135 56 58 205.959 -1.539 4,936 6.761 0.78
4/11/67 FV 11-42G Freeze valve sample - wt gain: O g Pump off 40 min. Sampler port 2" above salt surface
4112767 FP 11-43 3904 50 ¢ sample for /Ly analysis; sampler machined improperly; no results obtained
414767 FP 11-44 3937 10.50 6.60 11.01 4.542 4.561 69.88 102.55 140 59 44 205.927 -1.571 5.038 6.659 0.77
4/17/67 FP 11-459 4007 10.58 6.50 10.65 4,541 4,548 67.13 99.41 88 54 41 205.899 -1.599 5.128 6.569 0.77
4/18/67 FV 11-46G Freeze valve sample. Pump on. Sampler port above salt surface
&/21/67 FP 11-47 4167 10.95 6.48 10.96 4,558 4.604 66.65 99.64 169 71 75 206.678 -1.639 5.257 6.440 0.74
L/24/67 FP 11-48 4172 10.45 6.52 10,85 4,558 4,578 67.23 99.63 210 49 58 206,652 -1.665 5.340 6.357 0.73
4/25/67 FP 11-49 4196 50 g sample for UX/ZU analysis; »>2000u moles HF 206.642 ~1.675 5.372 6.325 0.73

6¢
Table 3.1 (continued)

(Weight %) (ppm) 3+
Date Sample Equiv Li Be Zr U F ) Fe Cr Ni LU in aU (kg) I Eq. Z Eq. {U” /3y]
Full o Ohe Circula- Oxida~ Reduc- A
nger ’ ’ tion tion tion Nom.
4/26/67 FP 11-50 Tandem graphite specimens - wire showed 30 X activity at salt-gas interface
4/28/67 FP 11-51 4274 11.20 6.45 10.97 4,557 4,571 69.61 102,80 114 6l 61 206.611 -1.706 5.471 6,226 0.71
5/1/67 FP 11-52 4341 11.33 6.45 10.79 4.556 4,566 69.27 102.41 80 60 25 206.584 -1.733 5.558 6.139 0.70
5/3/67 FV 11-53G 4388 Freeze valve sample. Pump on, level low (51%) 206,566 ~-1.751 5,616 6.081 0.70
5/5/67 FP 11-54 4436 16.93 6.63 10.95 4.556 4.551 69.91 158 61 63 206,547 -1.770 5.677 6.020 0.69
5/8/67 FP 11-55 4507 50 g sample for use in hot cell experiments 206.518 -1.799 3.770 5.927 0.68
5/9/67 FP 11-56 50 g sample for oxide analysis - poor results obtained; estimate: 50-150 ppm
5/10/67 FP 11-57 Sampler was found to be empty
5/10/67 FP 11-58 4513 10.48 6.66 11.27 4,556 4,607 70.85 103.97 131 81 42 206.516 -1.801 5.777 5.920 0.68
5/10/67 Run 11-F 4513 4.556 206.5106 -~1,801 5.777 5.920 0.68
6/19/67 Run 12-1 4513 4,536 205.679
6/19/67 FP 12-5 45,13 11.2 6.74 10.94 4,536 4.550 66,32 99,75 123 52 60 205.679 -1.801 5.777 5.920 0.68
6/21/67  FP 126 4513 50 g sample for U3*/ZU analysis U3t/Lu = 0.71%C 205.679  -1.801  5.777 5.920  0.68
6/21/67 ¥V 12-7G Freeze valve sample for Run 12 baseline data
6/21/67 ¥P 12-8 Be addition: 7.933 g as rod 205.679 -1.801 5.777 7.676 0.89
6/23/67 FP 12-9 4558 Be addition: 9.840 g as rod 205.661 -1.819 5.834 9.807 1.13
6/26/67 FP 12-10 4602 11.6 6.91 10.57 4,536 4,525 67.27 110.87 134 71 72 205,643 -1.837 5.892 9.749 1.12
6/29/67  FP 12-11 4674 50 g sample for U3T/SU analysis U3t/iu = 1.3%¢ 205.615  -1.865  5.981 9.660  1.10
6/30/67 FP 12-12 4706 11.6 6.54 10.91 4,535 &4.545 66.66 100,76 113 64 62 205.602 -1.878 6.023 9.618 1.11
7/3/67 FP 12-13 4739 Be addition: 8330 g as rod 205.589 -1.891 6.065 11.424 1.32
/57167 FP 12-14 4787 11.5 6.50 11.22 4.534 4,557 67.95 101.73 145 82 47 205.569 -1.911 6.129 11. 360 1.31
7/6/67 FP 12-15 4811 Be addition: 11.677 as rod 205.560 -1.920 6.158 13.922 1.61
7/7/67 FP 1216 4836 11.4 6.40 10.62 4.534 4,567 68.27 101.26 269 110 68 205.550 -1,930 6.190 13.890 1.60
7/10/67 FP 12-17 4885 11.3 6.40 10.66 4.533 4,532 66.92 99.81 216 144 53 205.530 -1.950 6.254 13,826 1.60
7/11/67 FP 12-18 4909 Concentration of oxide 57 ppm
7/11/67 FP 12-19 4917 11.5 6.19 11.00 4.533 4,522 65.05 98.26 100 102 62 205,518 -1,962 6.292 13.788 1.59
7/12/67 FP 12-20 4933 10.6 6.36 10.76 4.533 4,557 65.76 98.04 81 64 L4 205.511 -1.969 6.315 13,765 1.58
7/13/67  FP 12-21 4981 50 ¢ sample for UX/LU analysis: U3H/IU = 1.0%C 205.492  -1.988  6.376  13.704 1,58
7/13/67 FP 12-22 4981 10.5 6.52 10,43 4,532 4,566 66.18 98,19 247 90 50 205,492 -1.988 6.376 13.704 1.58
7/14/67 FP 12-23 4998 10.6 6.68 10.58 4,532 4,526 65.26 97.64 154 78 76 205,485 ~1.995 6.398 13.682 1.58
7/15/67 FP 12-24 5022 11.38 6.36 10.67 4.532 4,567 66.46 99.44 176 67 56 205.476 2,004 6,427 13.653 1.58
7/16/67 FP 12-25 5046 16.7 6.46 10.53 4.532 4,496 66,10 98.29 208 68 62 205.466 -2.014 6.459 13.621 1.57
7/17/67 FV 12-26G Freeze valve capsule-gas sample
7/17/67 FP 12-27 5070 10.7 6.61 10.78 4,532 4,550 67.50 100.14 195 75 72 205.457 -2.023 6.488 13.592 1.57
7/18/67 FP 12-28 5097 16.7 6.44 10,66 4,531 4.569 69.00 101.37 150 68 52 205. 446 -2.034 6.523 13.557 1.56
7/19/67 FP 12-29 5121 16.7 6.41 10.87 4,531 4,529 67.11 9G.62 177 84 78 205.436 -2.044 6.555 13.525 1.56
7/19/67 FP 12-30 U Addn. caps. no. 99
7/19/67 FP 12-31 U addn. caps. No. 100
7/20/67 FP 12-32 U addn. caps. no. 101
7/20/67 FP 12-33 U addn. caps. no. 102
7/10/67 FP 12-34 U addn. caps. no. 103
7/21/67 FP 12-35 U addn. caps. no. 104 4,543 205.983 =2.044 6.555 13.525 1.56
7/21/67 FP 12-~36 5169 10.5 6.68 10.96 4,543 4.554 66.40 99.09 156 84 129 205.964 -2.063 6.616 13.464 1.55
7/21/67 FP 12-37 U addn. caps. no. 105
7122767 FP 12-38 U addn, caps. no. 106
7/22/67 FP 12-39 U addn, caps. no. 107
7/22/67 FP 12-40 U addn. caps. no. 108 4,551 206,328 -2.063 6.616 13.464 1.55

oy
Table 3.1 {continued)

(Weight %) (ppm) 3+
Date Sample Equiv. Li Be Zr U F I Fe Cr Ni. iU in AU (kg) I Eq. L Eq. [U” /EU]
Full o ohs Circula- Oxida- Reduc- A
Power * ’ tion tion tion Nom.
7/23/67 FP 12-41 5229 10.6 6.52 10.68 4.550 4,562 68.00 100.36 110 64 192 206.304 -2.087 6.693 13.387 1.54
7/23/67  FP 12-42 U 6.52 capsule No. 109
7/23/61 FP 12-43 U addn. caps. no. 110
FP 12-44 U addn. caps. no. 111
FP 12-45 U addn. caps. no, 112
FP 12-456 U addn, caps. no. 113 4.560 206, 760 ~2.087 6.693 13,387 1.54
7/25/67 FP 12-47 5266 10.6 6.50 10.50 4.560 4,586 66.34 98.33 120 72 70 206,745 ~2,102 6.741 13.339 1.53
FP 12-438 U addn. caps. no. 114
FP 12-49 U addn, caps. no. 115
FP 12-50 U addn. caps. no. 116 4.566 207.020 -2.102 6.741 13.339 1,53
7/26/67 FP 12-51 5296 11.22 6.60 10.32 4.566 4,588 66,32 99.05 94 72 39 207.008 ~-2.114 6.780 13.300 1,52
7/28/67 FP 12-52 5366 10.7  6.47 10.7F 4.565 4,594 65.98 98.45 119 72 92 206.980 -2.142 6.870 13.210 1.51
1731767 FP 12-53 5433 10.5 6.39 10.66 4,565 4.503 62,48 97.53 182 72 60 206.954 -2.168 6.953 13,127 1.50
8/1/67 FP 12-54 5433 50 g sample for isotopic analysis
8/2/67 FP 12-55 5463 4,564 206,942 -2.180 6.991 13.089 1.50
10.75 6.44 11,12 4.564 4,577 65.72 98.61 156 72 300 206.942
10.80 6.42 10.95 4,564 4.575 64.57 97.32 136 58 720 206.942
8/3/67 FP 12-56 5492 Be addition: 9.71 g as rod 206.930 -2.192 7.030 15.205 1.74
8/3/67 FP 12-57 5496 4,564 206.928 ~2.194 7.036 15.199 1.74
8/4/67 FP 12-58 5500 11.33 6,59 11.08 4.564 4.587 65.14 98.73 156 54 170 206,927 -2.195 7.040 15,195  1.74
11.3¢ 6.74 11.28 4.564 4.549 66.53 170.40 160 64 424 206.927
B/4/67 FP 12-59 5500 11.00 6.58 10.77 4.564 4.600 66.62 99,57 138 74 66 206.927 ~-2,195 7.040 15.195 1.74
8/5/67 Run 12-F 5500 Sampler cable severed
9/15/67 Run 13-I 5500 4,543 205.984 -2.195 7.040 15,195 1,74
9/15/67 FP 13-4 5568 10.98 6.90 11.21 4,543 4,552 65,88 99.55 141 82 82 205.957 -2,222 7.126 15.109 1.74
9/15/67 FP 13-5 5568 50 g sample for U3+/IU; u3*/zy = 1.60%¢ 205.957 -2,222 7.126 15,109 1.74
9/18/67 FP 13-6 5666 10.95 6.42 10,78 4.542 4,587 67.01 99.77 118 66 76 205.918 -2.261 7.25% 14,984 1.72
9/18/67 Run 13-F 4,542 205,918
9/20/67 Run 14-I 4.542 205.924 ~-2,261 7.251 14,984 1.72
3/20/67 FP 1l4~1 Sample capsule suspended in empty pump bowl for 10 minutes
9/21/67 FP 14-2 No sample obtained; sampler found to have no sample port.
9/22/67 FP 14-3 50 g sample for isotopic analyses
9/25/67 FP l4-4 5757 10,70 6.32 11.52 4.541 4.577 66,16 99.30 104 76 46 205.887 -2.298 7.370 14,865 1.71
9/26/67 FP 14=5 50 g sample for isotopic analyses
9727767 FP 146 50 g sample for isotopic amalyses
9/28/67 FP 14-7 5828 10.70 6.24 10.98 4.540 4.572 66.19 98.71 102 79 63 205.859 -2.326 7.460 14.775 1.70
10/2/67 FP 14-8 5923 10.80 6.13 10.98 4.540 4,584 66,54 99.12 98 74 68 205.821 ~2.364 7.582 14,653 1.69
10/3/67 FP 14-9 5933 Sample for oxide analyses; no results obtained
10/5/67 FP 14-10 5995 10.90 6.74 11.40 4.539 4,587 65,48 99.13 150 80 76 205.792 -2.393 7.675 14.560 1.68
10/9/67 FP 14-11 6086 10.80 6.20 11.06 4.538 4,579 65.18 97.78 128 88 66 205,756 -2.429 7.790 14,445 1.66
10/12/67 ¥P 14-12 6155 10.90 6.88 11.01 4,538 4,579 66.82 100,22 176 60 60 205.728 ~-2.,457 7.880 14,355 1.65
10/16/67 FP 14-13 6249 10.50 6.36 11.08 4.537 4,575 65.73 98.27 146 65 57 205.691 -2.494 8.000 14.235 1.64
10/20/67 ¥P lé-14 6344 10.65 6.52 11.25 4,536 4,480 66,88 99.81 110 80 112 205.653 -2.532 8.120 14.115 1.63
10/23/67 FP 14-15 50 g sample for hot cell experiments
10/24/67 FP 14-16 6440 10.80 6.30 11.28 4.535 4,556 65.75 98.72 158 76 105 205.615 -2.570 8.242 13.993 1.61
10/26/67 FP 14-17 6440 10.80 6.26 10.81 4.535 4,557 66.14 98.59 124 80 55 205.615 -2.570 B.242 13,993 1.61
10/30/67 FP 14-18 6506 10.53 .00 11.00 4.535 4.553 65.24 97.35 112 70 62 205.594 ~2.591 8.310 13.925 1.61

184
Table 3.1 (continued)

(Weight %) (ppm) 3+
Date Sample  Equiv. Li Be Zr U F L Fe Cr Ni ZU in AU (kg) I Eq. L Eq. [U”/1U]

Full Circula- Oxida~- Reduc- A

Pgief Nom, Obs. tion tion tion Hom.
11/3/67 FP 14-19 6604 10,73 5.92 11.25 4,534 4.553 66.68 99.13 108 70 51 205.549 -2.636 8.454 13.781 1.59
11/3/67 FV 14-208 Freeze valve capsule - salt sample
11/6/67 FP 14-21 6678 10.60 5.94 10.69 4.533 4,557 65.88 97.69 128 76 61 205.520 ~2.665 8.547 13.688 1.58
11/7/67 FP 14-22 6702 10.60 5.86 11.32 4.533 4,529 66.74 99.07 106 72 74 205.510 -2.675 8.580 13.655 1.58
11/8/67 FP 14-23 6726 10.43 6.04 11.29 4.532 4.534 66.89 99.21 94 68 56 205.501 -2.684 8.608 13.627  1.57
11/9/67 FP 14-24 6751 10.63 6,18 11.34 4,532 4.567 66.42 99,17 169 72 72 205,491 -2.694 8.640 13.595  1.57
11/13/67 FP 14-25 6848 10.40 6.12 10.92 4.531 4,548 66.19 98,20 124 72 62 205.452 ~2.733 8.765 13,470 1.55
11/14/67 FP 14-26 6872 10.63 6.10 11.00 4.531 4,572 66.34 98.66 103 66 60 205,442 -2.743 8.797 13.438 1.55
11/27/67 FP 14-27 6913 10.20 6.19 11.12 4,531 4,553 66.53 98.62 126 72 69 205,426 -2.759 8.848 13,387 1.54
11/30/67 FP 14-28 7020 10,53 6.19 11,18 4.530 4,589 66.79 99.31 135 72 60 205,383 -2.802 8.986 13.249 1.53
12/4/67 FP 14-29 7142 10.40 6.34 10.88 4.529 4,564 66.42 98,63 121 10 12 205.335 ~2.850 9.140 13.095 1.51
12/5/67 FV 14-30S Freeze valve capsule - salt sample
12/6/67 FP 14-31 50 g sample for use in hot cell experiments
12/11/67 FP 14~32 7313 10.40 6.45 10,94 4.527 4,550 66.52 98.88 120 70 58 205,266 -2.919 9.362 12,873 1.49
12/12/67  FP 14-33 10 g sample exposed to atmosphere in Line 928, 1.5'ft. above pump bowl
12/13/67 FP 14-34 10 g sample exposed to atmosphere in Area 1C for 10 min.
12/14/67 FP 14-35 50 g sample for isotopic analysis
12/15/67 FP 14-36 50 g sample for isotopic analysis
12/18/67 FP 14-37 50 g sample for isotopic analysis
12/19/67 FP 14-38 7464 10.80 6.38 10.56 4.526 4,561 66.06 08.39 174 82 96 205.206 -2.979 9.554 12.681 1.47
12/20/67 FP 14-39 Concentration of oxide: 46 ppm
12/26/67 FP 14-20 7597 10.30 6.28 10.87 4,524 4.531 66.31 98.32 113 79 69 205.103 -3.032 9.724 12,511 1.45
1/2/68 FP 14-41 7732 10.30 6.34 11.12 4.524 4,536 66,23 98,55 134 84 59 205.099 ~3.086 9,897 12.338  1.43
1/8/68 FP 14-42 7799 10.50 6.10 11.16 4.523 4,531 66.41 98.73 148 87 73 205.072 -3.113 9.984 12,251 1.42
1/11/68 FP 14-43 7849 10.10 6.46 10.96 4.523 4,524 66.36 98,33 118 T4 70 205.052 -3.133 9.984 12.251 1.42
1/15/68 FP 14-44 7917 9.80 6.42 10,45 4.522 4,466 66.51 97.67 120 9c 68 205.025 -3.160 10,13 12,105 1,40
1/16/68  FP 14-45 7939 50 g sample for USt/LU analysis; 205.016  -3,169  10.16 12.075  1.40
1/17/68 FP 14-46 7953 10.20 6.50 11.16 4,522 4.516 67.06 99.47 152 84 72 205.011 -3.174 10.18 12.055 1.39
1/18/68 FP 14-47 7970 10.50 6.40 11.01 4,522 4,536 66.78 99.25 136 81 53 205.005 ~3.180 10.20 12.035 1.39
1/22/68 FP 14-48 8038 10.25 6,42 ig.92 4.521 4,557 66.96 99,13 111 0 51 204.977 -3.208 10.29 11,945 1.38
1/25/68 FP 14~49 8040 10.28 6.56 10.85 4.521 4.538 66.92 99.17 108 74 56 204,976 -3.209 10.29 11.945 1.38
1/29/68 FP 14-50 8040 10.20 6,38 11,38 4,521 4.541 66.24 98,76 121 79 70 204.970 -3.209 10.29 11.945 1.38
2/1/68 FP 14-51 8057 10.3 6,70 10.96 4.521 4.54]1 66.43 98.96 132 81 73 204,969 -3.216 16.31 11.925 1.38
2/5/68 FP 14-52 8129 10.15 6.56 10.90 4.520 4,541 66.57 98.74 110 68 48 204,941 -3.244 10.40 11,835 1.37
2/6/68 FP 14-53 Concentration of oxide: 58 ppm
2/8/68 FP 14-54 8186 9.88 6.03 10.38  4.520 4. 446 66.58 97.35 105 80 50 204.918 -3.267 10.48 11.755 1.36
2/9/68 FP 14-55 Ni rod suspended in Ni basket - 2 hours
2/12/68 FP 14-56 8236 16.28 6.36 11.08 4.519 4.533 66.53 98.84 117 90 68 204,898 ~-3.287 10.54 11.695 1.35
2/13/68  FP 14-57 50 g sample for UX*/IU analysis; UH/SU = ~0.35%
2/15/68 FP 14-58 8305 10,30 6.40 106.93 4.515 4,529 66.70 98.89 94 82 58 204.870 -3.315 10.63 11,605 1.34
2/19/68 FP 14-59 8401 10.23 6.58 10.7¢ 4.518 4,527 65.93 98.09 101 82 42 204.832 -3.353 10.75 11.485 1,33
2/20/68 FP 14-60 50 g sample for F. P. experiments - S. S. Kirslis; 10 g obtained
2/22/68 FP 14-61 8456 - 6.21 10.96 4.517 - 66,63 - 74 204,810 -3.375 10.82 11.415 1.32
2/26/68 FP 14-62 50 g sample for use in hot cell experiments
2727768 FV 14-638 Freeze valve capsule - salt sample 80
2/28/68 FP 14-64 8602 10.30 6.24 10.80 4.516 4,552 66.04 97.94 144 80 56 204.752 =-3.433 11.01 11.225 1.30
3/4/68 FP 14-65 8697 10.2G 6.40 11.08 4.515 4.510 66.58 98.80 63 83 36 204,714 ~3.471 11.13 11.105 1.29

cr
Table 3.1 (continued)

| .

(Weight %) {ppm)
Date Sample  Fquiv.  Li Be Zr U F ¥ Fe  Cr  Ni  Pu_ LU in AU (kg) £ Eq. L Eq. ¢[U3F/zu)
Full Nom Obs Circula- Oxida-  Reduc- y

Power ’ ’ tion tion tion Nori.
3/5/68 FV 14~-665 Freeze valve capsule - salt sample
3/6/68 FV 14-67G Freeze valve capsule - gas sample
3/7/68 FP 14-68 8742 10,09 6.24 1G.85 4,515 4.474 66.41 98,10 118 78 72 204,696 -3.489 11.19 11.045 1.28
3/8/68 FP 14-69 50 g sample for use in hot cell experiment
3/11/68 FP 14-70 Malfunction of sampler; no sample obtained 204.588 -3,594 11.59 10.645 1.23
3/28/69 FP 14-71 50 g sample capsule; no sample obtained
3/28/68 Run l4-F 9005 -3.615
8/18/68  FST-19 50 g sample for UF3 analysis (after transfer to FST)
8/19/68 FST-20 Isotopic analysis for ETA measurement
8/19/68 FST-21 Isotopic analysis for ETA measurement
8/19/68  FST-22 Isotopic analysis for ETA measurement
8/19/68 FST-23 Isotopic analysis for ETA measurement
8/19/68 FST-24 Isotopic analysis for ETA measurement
8/21/68 FST~25 6.92 1054 4,306 131 170 36
8/21/68  FST-26 6.80 10.46 4,293 132 164 35
8/29/68 FST=27 26 ppm 400 420 840 at finish of Fy
9/2/68  FST-28 8 ppm 430 420 1540 filtered; after  hr I,
9/3/68 FST-29 3 ppm 400 420 530 filtered; after 33 hr Hy
9/4/68 FST-30 11.42 4 ppm 360 440 530 not filtered

380 460 180
9/4/68 FST-31 50 g sample for oxide analysis
9/6/68  FST-32 11.52 2 ppm 110 100 10
9/7/68 FST-33 no results; filter end empty
9/14/68  FP 15-5 10 g sample for analysis of total concentration of reductants
9/14/68 FP 15-6 10.40 6.86 11.11 0.513 0.516 64,38 93.30 140 42 46 113 21.887 - - - 0
9/15/68 TP 15-7 Be addition: 10.08 g as powder - - 2.24 0
Scrapings of metal deposit from Be cage 12,0% 0.11% 8.03% -

9/15/68 FP 15-8 10 g sample for total reducing power; lost by explesion in 1ab
%/17/68 FP 15-9 10.70 6.74 11,40 0.663 0.648 69,32 98.84 131 50 75 116  28.373 +6.486
9/19/68 FP 15-10 10.45 6.44 11.28 0,768 0.764 67.18 96.13 125 29 60 102 32,862 +4. 489
9/20/68 FP 15-11 Capsule No. 30 Enrichment No, 1
10/1/68 FP 15-12 ~ 6.18 11.07 0.770 0.764 66.68 - 109 31 70 120 32,952 +0.090
10/2/68 FP 15-13 Capsule No, 28 Enrichment No, 2
10/2/68 FP 15-14 Capsule No. 25 enrichment No., 3
10/2/68 FP 15-15 Capsule No. 26 Enrichment No. 4
10/3/68 FP 15-16 Capsule No. 24 Enrichment No. 5
10/3/68  FP 15-17 Capsule No. 23 Enrichment No, 6
10/5/68 FP 15-18 10.63 6.62 10.94 0.780 0.822 67.61 99.66 159 36 82 106 33,430 +0.478
10/6/68 FP 15-19 Capsule No. 20 Enrichment No, 7
10/6/68 FP 15-20 Capsule No. 19 Enrichment No. 8
10/6/68 FP 15-]1 Capsule No. 27 Enrichment No, 9
1G/7/68 FP 15-22 Capsule No. 29 Enrichment No. 10
10/8/68 FP 12-34 Capsule No. 16 Earichment No. 11
10/9/68  FP 15-24 Capsule No. 21 Enrichment No, 12

134
Table 3.1 (continued)

(Weieght %) (ppm) 34
Date Sample  Equiv. Li Be Zr U F Z Fe Cr Ni Pu ZU in AU (kg) I Eq. I Eq. [U7/LZu]
Full Circula- Oxida- Reduc- i
Poggr Nom. Obs. tion tion tion Nok
16/ /68 FP 15~25 10 g sample for analysis of total concentration of reductants: % = 0.6%, [UST]/[LU] equivalent: 0.097%
16/10/68 FP 15-26 11.60 6.72 10.95 0.794 C.804 71.36 102,24 140 50 51 34.009 +0.579
16/12/68 FP 15-27 Capsule No. 22 Enrichment No. 13 34.009
10/12/68 FV 15-28% Freeze valve capsule - salt sample; no sample obtained
10/12/68 FV 15-29G Freeze valve capsule - gas sample
10/13/68 FP 15-30 Be addition; 8.34 g as rod - - - 0
Scale from Be rod 36.5% 50.3% 13.2%
Scale from Ni cage 78.6% 4.0% 17.4%
16/15/68 FP 15-318 Capsule No. 35 Enrichment Neo. 14 34.189 +0,090
10/15/68 FV 15-328 11.75 6.44 11.17 0.795 0.785 - - 134 59 - 34,189 -
10/17/68 FP 15-33 11.55 6.36 11.16 0.795 0.797 68,36 98,23 158 56 61 133 34,189 -
106/18/68 FP 15-34 Capsule No. 42 Enrichment No. 15
10/19/68 FP 15-35 Capsule No. 18 Enrichment No. 16
10/19/68 FP 15-36 Capsule No. 17 Enrichment No. 17
10/20/68 FP 15-37 Capsule No. 36 Enrichment No. 18
10/20/68 Fuel circuit drained
10/23/68 Fuel circuit filled
10/23/68 FP 15-38 11.45 6.76 11.45 0.804 0.797 66.72 97.21 154 68 43 146 34,443 +0,571 - - 0
10/26/68 FP 15-3%@ 0.804 0.795
10/28/68 FP 15-40 12.37 - - 0.804 0.820 - - 676 139 74
10/28/68 FP 15-41 Capsule No. 41 Enrichment No,19 34.533 +0.090 - - ¢
16/29/68 FV 15-428 - 6.13 11,09 0.804 0.779 - - 132 60 143 135 34.533 - = - ¢
10/29/68 FV 15-43G Freeze valve capsule - gas sample
10/30/68 FV 15-44 Capsule No. 38 Enrichment No. 20
10/30/68 FP 15-45 Capsule No. 34 Enrichment No. 21
10/31/68 FP 15-46 Capsule No. 45 Enrichment No, 22
10/31/68 FP 15-47 Capsule No. 39 Enrichment Ne. 23
11/4/68 FP 15-48 Capsule No. 32 Enrichment No. 24
11/4/68 FP 15-49 Capsule No. 31 Enrichment No. 25
11/5/68 FP 15-50 Capsule No, 37 Enrichment No. 26
11/6/68 Fv 15~518 - - 11.18 0.821 0.798 35.144 -
11/6/68 FV 15-52G Freeze valve capsule -~ gas sample
11/6/68 FP 15-53 Cu capsule containing a magnet; exposed to salt 5 min.
11/9/68 FP 15-54 Capsule No, 44 Enrichment No. 27 35.234 +0.0%0 0
11/11/68 FP 15-55 11.30 6.53 10.99 0.823 0.813 68.00 97.68 181 80 140 35.234 +0.070 - - 0
11/11/68 FP 15-56 Capsule Wo. 32, from PF 15-48, for laboratory tests
11/12/68 FV 15-57S Freeze valve capsule ~ salt sample
filtrate 0.823 0.736 2510 1290 340
residue 0.823 0.794 183 76 76
11/12/68 FV 15-58G Freeze valve capsule - gas sample
11/13/68 FP 15-59 Segmented magnets - suspended in salt 1 hr; pump off
11/15/68 FP 15-60 11.25 6.64 11,15 0.823 0.828 68.50 98.41 176 62 31 148
11/15/68 FP 15-61 Segmented magnets - suspended in salt 1 hr; pump on
11/15/68 FP 15-62 Be addition: 9.38 g as rod 0
Scale from Ni cage cap 4.72% 3.26% 4.71%
Scale from Ni cage body 2.74% 4.53% 0.10%
11/16/68 FP 15-63 11.50 6.50 11.04 0.823 0.818 69.52 99.42 143 62 46 153

I 4

4%
Table 3.1 (continued)

(Weight 7) (ppm) +
Date Sample Equiv. Li Be 7T U F n Fe Cr Ni LU in AU (kg) I Eq. L Eq. [U3 JiU]
Full Nom obs. Circula- Oxida~  Reduc- %
Power hx ‘ tion tion tion Nom.
11/19/68 FP 15-64 Magnet No. 4 (segmented) immersed 4 times
11/20/68 FP 15-65 11.80 6.99 11.23 0.823 0,843 66.91 97.90 98 63 52 171
11/20/68 FP 15-66 Segmented magnets, separated by Be metal spacers. Be addition: 1.0 g.
11/21/68 FP 15-67 Magnet No. 6; immersed 5 min.
11/22/68 FP 15-68 11.68 6.73 11.02 0.823 0.816 68.70 98.99 148 62 26 168
11/25/68 FV 15-69G Freeze valve capsule - gas sample
11/26/68 FP 15-70 50 g sample for oxide analysis; specimen not usable
11/27/68 FV 15-71S Freeze valve capsule - galt sample
11/28/68 Run 15-Ff
12/11/68 Run 16-1
12/12/68 FP 16-1 Magnet capsule
12/13/68 TFP 16-2 Magnet capsulel2/13/68 FV 16-3S
12/13/68 FV 16-38
12/16/68 FV 16-4S ~ 6.85 10.90 0.823 0.807 67.88 - 152 84 30
12/16/68 Run 16-F
1/12/69 Run 17-1 t]
1/12/69 P 17-1 G 11.35 6.39 10,94 0.823 0.780 70.63 100,09 116 52 57 140 35.234 4] G 0 0
1/14/69 Fv 17-28 2 - 7.02 11.08 0.823 0.812 66,18 - 122 68 53 35,233 -0.0005
1/16/69 FP 17-3 84 50 g sample for oxide determination - overheated - not usable 35.207
1/21/69 FP 17-4 11.58 6.82 11,32 0.822 0.812 67.89 98.46 112 62 56 144 35.207 -0.0266
1/21/69 Fp 17-5 50 g sample for oxide determination: 61 ppm
1/22/69 FV 17-68 Freeze valve capsule ~ galt sample
1/22/69  FV 17-75 124 Freeze valve capsule - salt sample 35.195 -0.039 0.13 -0.13 -0.09
1/22/69 FV 17-8 124 Be addition: 8.57 g as rod 1.77 1.17
1/27/69 FP 17-9 148 11,35 6.88 11.08 0.822 0.823 66.64 96.97 149 58 64 158 35.187 ~0.047 0.15 1,75 1.16
1/28/69 ¥V 17-105 172 6.68 10.3% 0.822 0.598¢ 162 128 270 35.180 -0.053 0.17 1.73 1.15
1/30/69 FP 17-11 220 cr® rod exposed to fuel for 6.5 hr, 4.73 g dissolved (0.18 eq.) 35.165 -0.070 0.23 1.85 1.23
2/6/69 FP 17-12 374 11,50 6.58 10.71  0.820 0.815 70.09  99.74 148 78 60 145 35.116 -0.118 0.38 1.70 1.13
2/8/69 FP 17-13-16 50 g samples for mass spectrometric analysis
2/10/69 FV 17-17G Freeze valve capsule - gas sample
2/12/69 FP 17-18 467 11.60 6,90 11.14 0.819 0.820 67.63 98.13 163 70 54 156  35.087 -0.148 0.48 1.60 1.07
2/19/69 FP 17-19 543 11,40 7.00 11.12 0,819 0.817 68.50 98.88 125 70 56 156  35.063 -0.172 0.56 1.52 1.01
2/26/69 FP 17-20 698 11.30 6.92 10.86 0.818 0.817 69.03 98.97 143 63 54 145 35,014 -0.221 0.72 1.36 0.91
2/26/69 FP 17-21 Dumbell shaped N1 rod exposed to fuel galt for 30 seconds
2/31/69 Fv 17-225 818 Freeze valve capsule - salt sample 34.976 -0.258 G.84 1.24 0.83
3/5/69 FP 17-23 839 11.60 6.75 10.60  0.817 0.814 132 70 58 132 34,969 =0.265 0.86 1,22 0.81
3/13/69 P 17-24 921 11.50 6.64 10.62 0.816 0.816 136 76 40 144 34,943 -0.291 0.95 1.13 0.75
3/17/69  FV 17-25G Freeze valve capsule - gas sample
3/17/69 FP 17-26 1006 50 g sample for determination of U3*/ZU concentration 34.916 -0.318 1.04 1.04 0.69
3/19/69 FP 17-27 1048 11.50 6.72 11.03 0.815 0.806 98 70 77 141 34,903 -0.331 1.08 1.00 0.67
3/26/69 FP 17-28 1147 11.93 6.82 11.56 0.814 0.805 147 71 33 159 34.872 -.0362 1.18 0.90 0.60
3/26/69  FV 17-295 1171 Freeze valve capsule-salt sample 34,864 -0.370 1.21 0.87 0.58
4/1/69 FP 17-30 1292 il.40 7.01 11.09 0.813 0,817 68,60 98.96 148 69 34 148  34.826 -0.,408 1.33 0,75 0.50
4/2/69 FV 17-318 10,460 7.24 7,69  0.813 0.972
4/3/69 FV 17-325 1340 Freeze valve capsule - salt sample 34.811 -0.423 1.38 0.70 0.47
474769 FV 17-33G Freeze valve capsule - gas sample
4/7-8/69 FP 17-34~40 50 g samples for mass spectrometric analysis
4/9/69 Run 17-F 1538 34,748 ~0.486 1.58 0.50 0. 34
4/14/69 Run 18~I 1538

Sy
Table 3.1 (continued)

(Weight Z) (ppm) 3+
Date Sample  Equiv. Li Be Zr U F I Fe Cr Ni Pu IU in AU (kg) I Eq. L Eq. 4[U”/LU]
Full m o Circula- Oxida- Reduc- %
?Oflgr om. S- tion tion tion Nom
4/14/69 FP 18-1 1538 11.30 6,50 10,65 0.812 0.800 66.50 95.78 151 86 33 157 34,787 =0.486 1.58 0.50 0.34
4/14/69  FV 18-25 1538 11.36 6.70 14.16 0.812 0.897 Nb: 42% 34,787 -0.486 1.58 0.50 0.34
4/18/69 FP 18-3 1564 Zr rod exposed to fuel for hr; 20.24 g dissolved (0.89 eq.) 34.779 -0.494 2.61 1.36 0.91
4/19/69  FV 18-4S5 1565 10.50 6.62 13,82 0,812 0.861 34.779 -0.494 2,61 1.36 0.91
4/18/69  FDE-A
4/22/69  FDE~B 9.50 5.14 17.68 0.179% 2.73% 0.037%
4/23/69  FDE-C 5.49 5.67 13.39 0.510% 1.61% 0.227%
4/23/69 FP 18-5 1718 11.35 6.53 11,41 0.811 0.805 69.10 99.23 100 77 50 156 34,730 -0.543 2.77 1.20 0.81
4/23/69  FV 18-6%5 6.28 6.19 9.3 0,8if 0,755
4725769 FP 18-7 1766 Zr Rods exposed tc fuel: 24,04 g dissolved (1.05 eq.) 34,715 -0.558 2.25 1.77 1.18
4/26/69 FP 18-8 50 g sample for isotopic analyses
4/29/69 FP 18-9 50 g sample for isotopic analyses
4/29/69 FP 18-10 1855 11.50 6.38 11,18 0.810 0.826 69,40 99.32 119 79 44 139 34.687 ~-0.586 1,91 2,11 1.41
4/30/69  FP 18-11 Empty Ni cage, exposed to salt for 10 hours
5/2/69 FP 18-12S 9.61
5/5/69 FP 18-13 1979 10.50 6.21 10.95 0.809 6.781 69.80 98.28 119 65 50 147 34,648 ~0,625 2.04 1.98 1.33
5/6/69 FV 18-14G Freeze valve capsule ~ gas sample
5/6/69 FV 18-15G Freeze valve capsule - gas sample
5/8/69 FP 18-16 2044 50 g sample for U3*/LU analysis. USt/IU = 0.4% 34.627 ~0.646 2.10 1.92 1.28
5/8/69 FP 18-17 30 g FeFy added to pump bowl. (0.64 oxidation equiv}.
5/9/69 FP 18-18 Interfacial tension measurement 2,74 1,28 0.86
5/9/69 FV 18-195 7.16 10.61 0.672
5/12/69  FP 18-20 Cu cage exposed to fuel salt for 10 hr.
5/13/60 FV 18-21G Freeze valve capsule - gas sample
5/14/6% FP 18-22 2224 11,23 6.89 11.32 0.807 0.804 70.10 100.38 107 65 33 154  34.570 -0.703 2.93 1.09 0.73
5/15/69  FP 18-23 2248 Be addition: 5.68 g as rod (1.26 eq) 34.563 -0.710 2.95 2,33 1.57
5/17/69 FP 18-24 Interfacial tension measurement
5/17/69 BV 18-25G Freeze valve capsule - gas sample
5/17/69 FV 18-26G Freeze valve capsule - gas sample
5/19/69 FP 18-26 Fission product deposition sampler (ligquid phase test), 1 hr. exposure
5/20/69 FP 18-27 2344 10,50 7.10 16.82 0.807 0.799 68.20 97.46 149 79 33 133 34.436 =0.740 3.05 2.23 1.51
5/20/69 FP 18-28 2344 Interfacial tension measurement - Be® present in sampler (3.17 g:0.704 eq) 34.532 =0.740 3.05
5/21/69 FV 18-29G Freeze valve capsule - gas sample
5/21/69 FP 18-30 50 g sample for isotoplc analyses (48.1 g obtained)
5/22/69 FP 18-31 50 g sample for isotopic analyses (41.7 g obtained)
5/22/69 FP 18-32 50 g sample for isotopic analyses (36.5 g obtained)
5/23/69 FP 18-33 50 g sample for isotopic analyses (44.1 g obtained)
5/23/69 FP 18-34 30 g sample for isotopic analyses (40.4 g obtained)
5/23/6% FP 18-35 50 g sample for isotopic analyses (37.4 g obtained)
5/26/69 FP 18-36 50 g sample for isotopic analyses (42.6 g obtained)
5/26/6% FP 18-37 30 g sample for isotopic analyses (42.0 g obtained)
5/271j69 FP 18-38 50 g sample for isotopic analyses (41.0 g obtained)
5/27/69 ¥FP 18-39 50 g sample for isotopic analyses (39.7 g obtained)
5/27/69 FP 15-40 50 g sample for isotopic analyses (44.5 g obtained)
5f27/69 FP 18-41 50 g sample for isotopic analyses (41.3 g obtained)
5/28/69 FV 18-42G Freeze valve capsule - gas sample
5/28/69 FP 18-43 2464 10.40 5.92 10,40 0.806 0.791 70.60 98.16 199 72 38 153 34,495 -0,778 3.17

9t
Table 3.1 (continued)

| .

(WEiEht z) (me) I+
Date Sample  Equiv. Li Be Zr U F L Fe Cr Ni pu U in AU (kg) I Eq. L Eq. {U” /fIU]
Full Yom Obs Circula- Oxida~  Reduc~- A
Pog;r : : tion tion tion Hom.
5/29/69 FV 18-44C Freeze valve capsule - gas sample
6/1/69 FV 18-45G Sample taken 30 min. after shutdown
6/1/69 FY 18-45G Sample taken 6 hr after shutdown
6/6/69 Run 18-F 2547 0.805 34.468 -0.805 3.26
8/11/69 Run 19-1 2547 0.802 3.26 0
8/16/69 FP 19-8 2547 10.50 6,12 10.90 0.802 0. 850 78.5 106.82 186 78 92 igég 34. 360 -0.805
8/18/69 FV 19-95 Freeze valve capsule - salt sample 66 Nb: 34%
8/19/69 FP 19-10 Fuel enrichment - 162,047 g salt = 100.145 g U
8/19/69 FP 19-11 Fuel enrichment - 155.387 g salt = 96.G29 g U
8/21/69 FV 19-12 2547 Fuel enrichment - 155.383 g salt = 96.027 g U IU = 292,201 34,652 ~0.805 o
8/21/69 FV 19-13G Freeze valve capsule - gas sample e
8/21/69  FV 19-14G Freeze valve capsule - gas sample 8,
8/21/69 FV 19-15G Freeze valve capsule - gas sample 2
8/21/69 FV 19-16G Freeze valve capsule ~ gas sample 2
8/27/69 FP 19-17 2629 11.0 5.69 10.94 0.809 (0.811 72,2 100.68 221 83 48 l;;: 34.626 -0.830 g
8/28/69 FP 19-18 2646 11.4 5.85 11.01 0.809 0.797 68.0 97.09 197 75 42 124 34,621 -0.836 8
9/4/69 FV 19-19G Freeze valve capsule - gas sample n
9/4/69 ¥V 12-20¢ Freeze valve capsule - gas sample 8
9/5/69 FP 19-21 2727 10,90 7,20 11,20 0.808 0.795 6%9.80 99.92 177 89 15 98 34,595 -0,861 e
9/9/69 FP 19-22 2793 10.95 6.92 11.30 0.807 0,799 68.50 98,50 147 87 20 104 34,575 -0.882 2
9/10/69 FV 19-23G Freeze valve capsule - gas sample +
9/10/69 FV 19=245 Freeze valve capsule -~ salt sample 17 0,81 0.55
9/12/69 FP 19-25 PuF4 addition: 31.6 g PuF3 (Zr: 0.62 g; = 0.03 eq.)
9/19/69 FP 19-26 PuF3 addition: 35.6 PuF3 0.84 0.57
9/19/69 FP 19-27 2793 10.70 7.12 10.80 0.807 0.792 70.10 99.55 183 96 50 144 34,575 -0.882 0.84 0.57
9/23/69 FV 19-28G Freeze valve capsule - gas sample
9/23/69 FV 19-29G Freeze valve capsule - gas sample
9/24/69 FP 19-30 10.70 7.38 10,90 0,807 0.781 66.30 96.10 227 97 48 142
9/24/69 FP 19-31 PuF3 addition: 39.2 g PuFj
9/25/69  FP 19-32 PuF3 addition: 40.8 g PuF3
9/25/69 FP 19-33 PuFq addition: 42,2 g PuFj
9/26/69 FP 19-34 PuFq addition: 39.2 g PuFj 165
9/29/69 FP 19-35 2968 10,75 7.07 11,10 0.806 0,775 69,00 98,73 214 97 52 ¢ 34,519 -0.938 0.18)% 0.71 0.48
9/29/69  FV 19-368 Freeze valve capsule - salt sample 97 &g%n
9/30/69 FV 19-37G 2992 Freeze valve capsule - gas sample 34.512 -0.945 (0.21) 0.68 0.46
9/30/69 FV 19-38G 2992 Freeze valve capsule - gas sample
10/1/69 FP 19-39 Surface tension exp. no Be present
10/2/69  FP 19-40 3040 Surface tension exp. - 2.87 g Be dissolved (0.64 eq.) 34,497 -0.960 (0.25) 1.28 0.87
10/2/69 FP 19-40 Img: 1171 0.54 22.4 <0.01
10/3/69  FV 19-41G Freeze valve capsule - gas sample
10/3/69  FV 19-428 Freeze valve capsule - salt sample Nb: 0%
10/5/69 FP 19-43 3106  10.60 6.61 11.40 0.805 0.767 71.70 183 96 45 165 34,476 =0,981 (0. 32) 1.21 0.82
10/6/69  FV 19-44S Freeze valve capsule - salt sample
10/6/69 FP 19-45 3130 50 g for YI/Iv 34,468 =0,989 (0.35) 1.18 0.79
10/7/69  FV 19-46G Freeze valve capsule - gas sample
10/7/69  FV 19-475 3154 Freeze valve capsule - salt sample

Ly
Table 3.1 {continued)

(Weieht %) (ppm) 3+
Date Sample Equiv. Li Be Zr U F E Fe Cr Ni Pu LU in AU {kg) L Eq. % Eq. [U” /Zu]
Full Circula- Oxida~ Reduc=- A
Poger Nom. Cbs, tion tion tion Nom.
X
10/8/69 FP 19-48 3178 Be addition: 4.91 g as rod (1.09 eq.) 34,461 -0.996 0.37) 2.25 1.52
8§.58 6.95 14,7 0.804 0.849 (65.11) 0.217% 3.53% 0.07%Z 161 34.453 -1.004 (0. 40) 2.22 1.50
10/9/69 FP 19-49 Graphite assembly exposed to pump bowl wvapor
10/9/69 FP 19-50 Graphite assembly exposed to pump bowl salt
16/9/69 FP 19-51 Graphite assembly exposed to pump bowl salt
10/13/69 P 19-52 Copper rod exposed to pump bowl salt
10/13/69 FP 19-53 3321 11.48 6,79 11,60 0,804 (.786 68.50 235 102 40 164 34,411 ~1.049 (0.54) 2.08 1.41
10/14/69 FV 19-54G Freeze valve capsule - gas sample
10/14/69 FV 19-558 Freeze valve capsule - salt sample
10/15/69 FV 19-56G Freeze valve capsule - gas sample
10/17/69 FV 19-57sS Freeze valve capsule - salt sample
10/17/69 FV 19-58S Gas sample takenm 1 hr after reduction of power to 10 kw
10/18/69 FvV 19-59S Gas sample taken 4 hr after reduction of power to 10 kw
10/20/69 FP 19-60 3458 50 g sample for uH/tu analysis; uH/y = 0.07%h 34.365 -1.092 (0.68) 1.94 1.31
10/21/69 ¥FP 19-61 Nb foil exposed to fuel salt 45 min
10/22/69 FV 19-62G Gas sample with intake port at bottom
10/22/69 FP 19-63 3506 11.25 6,78 10,90 0.802 0.802 70.0 99.78 136 95 33 173 34,350 ~1.107 0.73) 1.89 1.29
10/22/69 FV 19-64G
10/23/69 FV 19-656 Freeze valve capsule - gas sample
10/24/69 TFP 19-66 Capsule for FP plating experiment - 10 hr exposure
10/24/69 ¥FP 19-67 Capsule for FP plating experiment - 3 hr exposure
10/27/69 F¥P 19-68 Capsule for FP plating experiment - 10 min exposure
10/28/69 FP 19-69 3627 50 g sample for oxide analysis; sample not removable from transport container. 34,311 -1.146 (0. 86) 1.7¢6 1.20
10/28/69 FV 19-70G Gas sample with intake port at top
10/28/69 FP 19-71 50 g sample for hot cell tests
10/29/69 FP 19-72 50 g sample for hot cell tests
10/29/69 FV 19-73G
10/30/69 FpP 19-74 3698 11,20 7.31 10.70  0.801 0.796 69. 30 99.35 124 101 20 168 34,289 -1.168 (0.93) 1.69 1.15
10/30/69 FP 19-75 50 g sample for U3*/IU analysis: UST/IU = 0.02%h
10/31/69 Fv 19-76S Freeze valve capsule - salt sample
10/31/69 Fv 19-77S Freeze valve capsule - salt sample
11/2/69  FV 19-78G Gas sample obtained 1 hr after drain
Run 19-F 3777 34,267 -1.193 (1.01) 1.61 1.10
Run 20-1
11/26/69 FV 20-15 0.800 0.784 34.274 -1.193 - - ¢
11/26/69 FP 20-2 3777 Surface tension experiment - control sample, 2 hr exposure
11/27/69 FP 20-3 7LiF-233yF, addition U = 97.2 g
11/27/69 FP 20-4 50 g sample for oxide determination: 58 ppm 34.371 -1.201
11/28/69 FP 20-5 Sample for UX/IU analysis by spectrophotometric method
11/28/69 FP 20-6 3825 10.00 7.19 10,90 0.802 0.783 72,80 101,71 199 98 72 117 34, 356 ~1.208 0.15 0.1
11/29/69 FP.20-7 3849 Be addition: 6.974 as rod (1.55 eq.) 34.384 -1.216 ( - ) 1.53 1.04
White Salt 9.92 7.42 11.95 0.803 0.799 -— 0.17% 1,88% 0.04%
Dark Crust 0.10% 6.447 0,047%
11/30/69 FP 20-8 Cu0-Ni assembly exposed to fuel pump vapor 8 hr
12/1/69  FP 20-9G Freeze valve capsule - gas sample
12/1/69  FP 20-10 Ni powder in Ni capsule, exposed to fuel pump vapor 8 hr
12/2/69  Fp 20-11 Electron microscope screen assembly, exposed to fuel pump vapor 2 hr
12/2/69  FP 20-12G Freeze valve capsule -~ gas sample
12/3/69  FP 20-13 Ni rod, exposed to fuel pump vapor & hr

514
Tablie 3.1 (continued)

| | .

(Weight %) (ppm} I+
Date Sample Equiv. Li Be Zr U F z Fe Cr Ni Pu U in AU (kg) L Eq. L Eq. [U” /ZU]
Full Circula=- Oxida- Reduc- %
power Nom. Obs. tion tion tion Nom.,
Hi
12/3/69  ¥P 20-14 50 g sample for U/IU analysis: 0.11%
12/3/69  FP 20-15 Sample for U3*/IU analysis by spectrophotometric method
12/4/69  FP 20-16 CuD-Ni assembly exposed to fuel pump vapor 8 hr
12/4/69 FP 20-17 Cu0-Ni assembly exposed to fuel pump vapor 8 hr
12/4/69  FP 20-18 'LiF-233UF, addition, U:92 g
12/5/69  FP 20-19S8 Freeze valve capsule - salt sample
12/5/69  FP 20-20 Ni bar exposed to fuel pump 8 hr
12/8/69  FP 20-21 Cu0-Ni assembly exposed to fuel pump salt 8 hr
12/9/69 FP 20-22 4093 Be addition; 9.874 g as rod (2.19 eq.) 34,271 ~1.293 (0.25) 3.54 2,41
12/9/69 FP 20-23 Surface tension experiment, with Be®, 4 hr exposure
12/9/69 FP 20-24 Sample for U3'/IU analysis by spectrophotometric method. !
12/10/69 FP 20-25 Ni bar exposed to fuel pump vapor 8 hr
12/10/69 FP 20-26 50 g sample for U'/IU analysis; U/SU = 0.20%
12/10/69 FP 20-27G Freeze valve capsule - gas sample
12/10/69 FP 20-28 Cu0-Ni assembly exposed to fuel pump salt 8 hr
12/11/69 FP 20-29 Cu0-Ni assembly with Pd windows exposed to pump bowl vapor 8 hr
12/11/69 TFP 20-30 4140 Sample for U/IU analysis by spectrophotometric method 34,256 -1.308 (0.30) 3.49 2.37
12/12/69 FP 20-31 4165 10.80 .80 10,90 0,800 0.801 71.60 101.03 265 92 160 34,249 -1.316 (0.33) 3.46 2.35
12/12/69 FP 20-32¢G Obtained 1 hr after fuel drain
12/12/69 Run 20-F 4167 0.800 34,249 -1.316 (0.33) 3.46 2.35

*Analyzed after storage for 30 hr at room temperature,
a MSRP Semiann. Prog. Rept. for P/E Feb. 28, 1967, ORNL-4119, p 156.
b MSRP Semiann. Prog. Rept. for P/E Aug. 31, 1967, ORNL-4191, p. 168.

0

R. E. Thoma and J. M. Dale, ORNL-4396, Feb., 28, 1969, p. 134,

d 0.819 kg U added; sample addition capsules not designated by FP-numbers

Samples removed from the fuel storage tank

e
£f 50 g sample for determination of uranium concentration by fluorination.

& Increase in oxidation equivalents after FP 19-26. ORNL-4548, p 180

h Sample was not usable for U3+/EU analysis. Oxide concentration determined to be 71 ppm.

6
maintained throughout the entire period of reactor
operations with the MSRE. All salt, metal, and gas
specimens removed from or introduced into the fuel
system were assigned sample designation numbers.
Analytical data obtained with these specimens are
summarized in Table 3.1, which will serve as the basis
for the discussion in the remainder of this section.

3.2 Component Analysis

As soon as a statistically significant number of
fuel-salt samples were obtained it became evident that
the average variation in the analytical values for the
components Li, Be, Zr, and F within one standard unit
of deviation (1 o) would preclude the use of the results
of chemical analyses for these components for opera-
tional control of the reactor. Of special interest in this
connection is the component zirconium. It might seem
that if reduction in the concentration of zirconium in
the salt were detectable analytically, such reduction
would signal that the oxide saturation limit was
exceeded. However, assuming that the salt was origi-
nally completely free of oxide and that the solubility
limit is ~700 ppm at 650°C, the fuel salt could
accommodate 0.34 kg of oxide before it became
saturated. If the source of the oxide was from moisture,
saturation by oxide would correspond to complete
reaction of 0.23 kg of HF with the containment vessel,
from which 0.29 kg of Cr would be leached into the
salt, increasing its concentration there by 59 ppm. It is
thus evident that unless highly sensitive techniques were
available for measuring the oxide concentration of the
salt introduced from contamination of the salt by
moisture, the initial indication of such contamination
would be the increase of chromium. The [ ¢ sensitivity
limit for zirconium corresponds to *0.24%, which
corresponds to 1.75 kg of ZrO,. Although the solu-
bility of ZrQ, in the fuel salt is sufficiently temperature
dependent to expect that on saturating the molten-salt
solution with oxide and that further formation of oxide
would result in the deposition of oxide crystals at the
coldest point in the circuit, it has not been established
experimentally that, with rapidly circulated streams,
such crystal growth would actually occur. If not,
significant amounts of zirconium might be carried in
the salt in slurry form and an analysis of samples
removed from the pump bowl would not signify a loss
of zirconium from the salt as one might expect
otherwise. It is thus clear that by the time it would have
been possible to establish from zirconium analysis that
the 10 limit had been exceeded, at least 1.41 kg of
ZrO, would have been precipitated and probably

50

carried by the salt as a slurry. Since, of all the fuel salt
constituents, zirconium is the one whose reduction in
concentration would independently reflect increasing
amounts of oxide impurity, only perfunctory attention
was given to the precision of the analyses of the carrier
constituents. Analysis of these constituents was con-
tinued on a routine basis because of their use in
providing general confirmation of the inference that the
fuel was chemically stable and because the principal
analytical costs were incurred from operations related
to handling and processing the radioactive samples;
laboratory tests which followed dissolution of the salts
for analysis comprised a minor fraction of the overall
charges. Mean values of the MSRE fuel-salt composition
as determined from analysis of samples removed from
the pump bowl are given in Table 3.2.

3.3 Oxide Analysis

Until the chemical equilibria involving H,, HF, and
oxides in LiF-BeF, melts were defined completely,? no
satisfactory analytical method existed for determina-
tion of oxide in MSRE salts. The KBrF; method,>
while occasionally satisfactory, produced results that
reflected the frequent contamination of samples after
they were removed from the reactor by quantities of
water adsorbed in transit and during storage, and were
for the most part generally incredibly large and erratic.

Subsequently, an improved method of analysis was
developed® based on the equilibrium 0% + 2HF(g) =
H,O(g) + 2F ", which occurs when a molten-salt sample
is purged with an HF-H, gas mixture. Although this
hydrofluorination method was regarded as less sensitive
than the inert-gas fusion method” or the KBrF,
method, satisfactory sensitivity was achieved by ~50-g
samples. After development, only this method was used
for analysis of MSRE salt samples. During the period
when the MSRE was in operation it was important to
know the approximate concentration of oxides in the
fuel and coolant salts, but the information was not
considered to be particularly relevant to a need to
develop highly sensitive in-line analytical methods for
reactors in which on-line fuel reprocessing is done. This
arises from the fact that on-line reprocessing methods
are likely to employ chemical processes in which, for
rare earth removal, there is a complete turnaround in
the uranium inventory. If this is done, the turnaround
will be accomplished by fluorination, and the reconsti-
tuted stream will be adjusted in uranium (IV, III)
concentration.

The fuel-salt mixtures which are most suited for use
in breeder or converter systems employ a fertile carrier
Table 3.2. Mean values of the MSRE fuel composition from chemical analysis

Sample No, of Li Be Zr U F LiF BeF7 ZrFy UFy, Cr Fe Ni Nom, Aver
Group Samples wt 7) (mole %) {(ppm) Wt 4 U
Carrier® 36 10.79+0.33 7.25+0.18 11.93+0.31 - 70.1740.62 62.24+0.90 32.48+0.90 5.28+0.16 22+5 102+54 1449

Run 22 8 10.4440.19 6.79+0.31 11.2630.13 3.0440.02  67.50%0.62 62.84+1.08 31.46+1.12 5.16+0.11 0.54+0.01 1948 98+51 23+3 3.038
Run 2 10 10.3940.17 6.44+0.02 11.43+0.34 3.05+0.02 71.65+0.80 63.72+0.43 30.40+0.44 5.33+0.14 0.55+0.01 37413 163+49 34+11 3.038
Run 3 51 10.45+0.21 6.49+0.21 11.22+0.35 3.98+0.71 69.80+1.72 63.46+0.83 30.60+0,84 5.23+0.17 0,71+0,12 3748 154455 48+19

Run 4 22 10.51+0.14 6.55+0.16 11.14+0.30 4.642+0,028 67.17+1.44 63.36+0.57 30.65+0.58 5.15+0.12 0.83+0.01 48+7 131+65 40+20 4.648
Run 5-7 14 10.52%0.28 6.54+0.19 11.32+0.23 4.629+0.026 68.72+0.86 63.36+1.03 30.58+0.91 5.23+0.14 0.82440.013 50+7 10844 54+25  4.627
Run 4~7 47 10.52+0.18 6.57+0.18 11.2440.27 4.638+0.025 67.96+1.36 63.29+0.72 30.70+0.70 5.19+40.13 0.824+0.011 49+7 114+55 46421

Runt 8 8 11.78+1.41 6.53+0.20 11.16+0.19 4.6324+0,011 69,77+¢1.49 65.84+2.49 28.57+2,12 4,8240.35 0.771+0.058 64+7 122445 61+36 4.605
Runt 9 4 10.99+0.10 6.63+0.067 11,15+0.37 4.603+0.031 68.55+0.48 64.17+40.10 30.04+0.15 4.99+0.19 0.794+40.011 61+5 150+17 52420 4,587
Run 10 10 11.14+0.08 6.58+0.19 11.05+0.15 4.609+0.020 67.82+1.38 64,65+0.45 29.6430.48 4.92+0.06 0.791+0,010 60+4 150+30 74¥35  4.567
Run 11aA 31 10.80+0.35 6,46+0,15 10.97+0.18 4.570+0.018 67.81+1.46 64.30+0.90 29.88+0.82 5.02+0.13 0.804+0.018 64+6 131+48 54416 4,557
Run 11B 6 10.89+0.36 6.53t0.09 10.97+0.17 4.579+0.023 68.94+1.66 64.27+1.00 29.96+0.85 4.97+0,14 0.799+0.017 64+11 144+46 54+18 4.565
Run 1l 38 10.82+0.35 6.47+0.15 10.9740.17 4.571+0.019 67.99+1.53 64.30+0.90 29.89+0.81 5.01+0.13 0.803+0.018 64+7 133+47 54416

Run 10-11A 41 10.8940.34 6.49+0.17 11.0040.17 4.580+0.024 67.83+1.43 64.34+0.82 29.81+0.76 4.99+0.13 0.801+0.017 63+6 136+45 59+24

Run 10-11 48 10.88+0.34 6.49+0,.16 10.99+0.17 4.579+0.024 67.95+1.47 64.3740.83 29.84+0.76 4.99+0.13 0.801+0.017 63+7 136+44 58+23

Run 12-4 15 11.004+0.45 6.50+0.18 10.75+0.21 4.544+0.023 66.79+1.10 64.65+0.96 29.70+0.90 4.85+0.13 0.790+0.020 7149 166+54 61+11 4.535
Run 12-B 11 10.85+0.32 6.54+0.11 10.82+0.29 4.570+0.028 66.10+0.90 64.21+0.57 30.08+0.49 4.92+0.14 0.800+0.017 69+8 139+26 226+209 4.571
Run 12 26 10.93+0.40 6.5230.15 10.78+0.24 4.555+0.027 66.50+1.06 64.46+0.83 29.86+0.77 4.88:0.14 0.794+0.021 7040 154446 110+ ?

Run 13-14A 24 10.65+0.20 6.28+0.28 11.104+0.21 4.561+0.024 66.27+0.52 64.51+0.74 29.51+0.80 5.16+0.14 0.81740.017 63+6 123422 68+15

Run 14B 20 10.35+0.28 6.39+0.16 10.914+0.24 4.523+0.030 66.49+0.32 63.55+0.92 30.48+0.86 5,1540.14 0.821+0.012 82+7 121+147 62414

Run 14 44 10.51+0.28 6.33+0.24 11.014+0.24 4,5430.032 66.3740.45 64,07+0.95 29,95+0.95 5.15+0.13 0.819+0.015 77+8 123+43 65+15 4,548
Run 4-14 176 10.7540.50 6.48+0.21 11.04+0.27 4.585+0.046 67.44+1.49 64.09+1.11 30,05+1.00 5.06+0.19 0.809+0.024 64+13 130+45 67+67

Run 16-18 21 11.33+0.39 6.71+0.30 10.994+0.29 0.809+0.011 68.71+1.26 65.08+0.91 29.93+0.93 4.85+0,14 0.137+0.004 7249 135+24 49+12 0.8155
Run 19 11 10.95+0.34 6.83+0.48 11.08+0.27 0.791+0.013 69.40+1.66 63.94+1.67 30.96+1.74 4.97+0.18 0.137+0.005 93+8 186+38 38+13 0.8045
Run 19-20 13 10.86+0.40 6.85+0.45 11.05+0.26 0.791+0,012 69.83+1.86 63.68+1.75 31.21+1.79 4.97+0.17 0.137+0.005 93+8 193+41 41+16

Run 16-20 33 11.14+0.45 6.76+0.37 11.02+0.28 0.802+0.015 69.18+1.60 64.53+1.46 30.43+1.46 4,9040.16 0.13740.004 8C+14 157+43 46+14

Run 2-20 258 10.79+0.50 6.530.25 11.0540.28 3.881+1.462 67.87+1.65 64,10+1.24 30.16+1.08 5.,05+0.19 0.685+0.261 62+17 136+58 62+58

a Analyses from the General Analysis Lab., all others from the HRLAL.

1S
52

salt of the approximate composition ’LiF-BeF,-ThF,
(72-16-12 mole %). It has been deduced recently that in
uraniferous fuel salts constituted from this base, the
oxide tolerance at operating temperatures is in the
range 30 to 70 ppm.® Although there is no need for
incorporation of an oxygen getter, such as ZrF,, in the
salt directly, this low oxide tolerance indicates that
there will be a need for swift, satisfactory, and
preferably in-line methods for determination of Q%
(and UF;) concentrations in the fuel after storage and
during startup and operation. Further, it will be
necessary to incorporate a demonstrated continuous
process for removal of 0% to a satisfactorily low level
from a substantial side stream from the reactor. The
results listed in Table 3.1 show that the concentration

of oxide never exceeded ~10% of saturation and were
thus regarded, along with the corrosion data, as
satisfactory evidence that the fuel system did not
experience contamination during periods when fuel was
circulated in it. Other material pertinent to this
discussion is described in Chap. 6, concerning corrosion
and anomalies in the behavior of UF ;.

3.4 Uranium Concentration

One of the purposes of the Molten-Salt Reactor
Experiment was to examine the applicability of various
techniques for rapid, accurate, reliable means of estab-
lishing the inventory of uranium in the fuel-salt system.
We projected that the experience gained with the MSRE

ORNL-DWG 71-9987R

I
468 o *BASED ON INITIAL LOADING OF
) 226.25 kg U, 4885.3 kg 2 SALT,
. U CONSUMPTION RATE=0.3991042 g/EFPH [7.4 Mw (th)]
466 °
> © ©
. ® o )
52 B o
_— .
z 464 r-X 7 - R
— L
| \.. [ ] ®
w b \d
") a2 - - -
L @ &
Z ® e o
o B 'R [ ® e
Z 460 2 ~~ hd o
= b ® ®
ilj ®
o) s Qoo o ¢
(] 9 .'
£ 458 s >
o s 3o ® 3 ° %
e ® ®
=z ® ®
= ® e ® ®
= o & &g .fi ®
D 4.56 o ® @ & £
= o 8 oq [ | e e b
=z ~ 9 8 o e ¢ ®
o ® D % oF @ @ @
> iy~ L .
5 4.54 @ ~ :"
> NOMINAL VALUE®” @ o S~ '
S o %
E . . : \\o
4,52 —]
- §
Z
Lt @
s ®
O 4,50
O ®
4.48 &
RUN NO. ®
5-7 8.9 10 1 12 13-14 o ,
T T T T
o) 2 4 6 8 (x10%)

equivalent full power hours

Fig. 3.1. Comparison of analytical and computed values of uranium in the MSRE fuel salt (*2%U and 23U tuel charge).
would fix the alternatives to be explored for the
development of larger molten-salt reactors. When it was
noted early in MSRE power operations that the
sensitivity of on-site determination of uranium concen-
tration coefficient in the reactivity balance was some
tenfold greater than the statistical variation in chemical
analysis, further consideration of the applicability of
laboratory analysis of individual samples for operational
control was abandoned.

Other sections of this report discuss uranium burnup
rates (Chap. 7), fissile inventory (Sect. 2.4.2), and
material balances (Sects. 3.6 and 3.7), the results of
which led to the final estimates of the concentrations of
uranium in the fuel salt as listed in Table 3.1. After
final adjustments in the estimated amounts of salts
contained in the reactor system and of the power

53

production rates, comparisons of the analytical data for
uranium concentration were made with the nominal
values for 233U and 232 U fuel, respectively, in Figs. 3.1
and 3.2. It is evident in Fig. 3.1 that the analytical data
contain a small positive bias in reference to nominal
values. The bias is of little consequence, for with the
average uranium inventory a difference in the estimated
carrier salt mass of 1 kg changes the nominal concentra-
tion of uranium by 0.01 wt %.

3.5 Structural Metal Impurities

In the development of molten-salt reactor technology,
continuvally recurring needs for accurate values of the
properties of moving fluids become apparent. Conse-
quences of ever increasing amounts of structural metal

ORNL-DWG 71— 9988R

0.830

0.825

0.820

0.815 €

Q.810

0.805 & ®

0.8C0 2

URANIUM CONCENTRATION {w! %)

0.795

0.790

0.785

0. 780

0.775

2

3 4 (x10%)

equivalent full power hours

Fig. 3.2. Concentration of uranium (*>>U) in the MSRE fuel circuit.
impurities that might be generated in the MSRE and
that might alter the properties of the flowing sait
streams into which they were released were envisioned
at the outset of operations. With the possibility that
impurities would exist both as ionic and metallic
species, it was desirable to examine the interrelation-
ships of their development with respect to corrosion of
the container circuits, fission product transport, cavita-
tion effects, changes in the flow characteristics, and
their impact on MSBR reprocessing capability. Since all
but the initial aspect of this behavior comprises an
ongoing effort, the scope of the present report includes
consideration of structural metal impurities in the
MSRE only with respect to the implication of changes
in their concentration as related to corrosion. Specific
discussion of the chemical significance of the amounts
of impurities and changes in the concentrations is
deferred for consideration under Sects. 5 and 6, in
which corrosion in the fuel and coolant circuits is
discussed, and for subsequent reports where the rela-
tionships of physical and chemical properties of the
MSRE to metallurgical behavior and to fission product
behavior are assessed.

Midway through the experiment, phenomena were
observed that led to the speculation that changes in the
surface-related properties of the impurities were of
possible significance to the changes observed. It was not
experimentally feasible at that time to utilize the
reactor as the experimental tool for studies of the
interfacial behavior in salt-metal systems. By the time
that laboratory studies had advanced to the stage where
a program of studies with the reactor could be
formulated, the schedule for operation of the reactor
called for its termination.

The combined experience from laboratory investiga-
tions and from an extensive engineering test program in
which various fluoride mixtures were circulated in
thermal convection and forced circulation loops led to
the expectation that the concentration of structural
metals in the MSRE salts would increase only slightly
throughout the planned period of reactor operation. As
anticipated, only modest increases occurred.

The MSRE fuel was produced, as described in Sect.
2.4, from carrier salt and fuel concentrate, for which
the chemical analyses indicated average concentrations
of impurities were quite low. Chemical analyses were
performed with samples of fuel salt removed regularly
from the fuel pump bowl, and occasionally from the
fuel drain tank. The average concentrations of the
structural metal impurities found in these samples are
listed in Table 3.2. Individual analyses are listed in
Table 3.1.

54

Table 3.3. Average concentration of structural metal impurities
in the MSRE fuel salt

Number of Impurity concentration (ppm)
Sample group samples Cr e Ni
Fuel concentrate
Run No.
2
3 34 377 14859 55%18
4 69 53+8 122451 52%27
57 14 507 108*44 54125
8 8 647 122*45 61%36
9-10
i1 38 647 133%47 54%16
12 18 72t9 160+52 85%73
13-14
Early stages 24 73+t6 125%£22 68*15
Latter stages 20 827 12124 62%14
Total 44 778 123+23 6515
15
1618 21 7229 135+t24 49*12
19-20 13 93+8 193*41 41*16

The methods employed by the analytical chemists for
determination of structural metal impurities were de-
scribed previously.? Each of the methods employed is
capable of approximately the same precision. Yet, as
shown by the results in Table 3.3, the experimental
precision obtained was much better for chromium than
for either iron or nickel. Corrosion reactions in the
MSRE were expected to produce only CrF, as a
dissolved species in the salts; iron, nickel, and molybde-
num, if present, were expected to persist in the metallic
species. The differences in observed standard deviations
for the three elements suggest that behavior very similar
to that anticipated did result. Inspection of the indi-
vidual results in Table 3.1 shows that the relative
concentrations of iron and nickel found among indi-
vidual samples rarely were constant. This is a somewhat
surprising result since the excellent mixing conditions
which existed in the circulating salt should have been
conducive for allowing these metals to produce equilib-
rium solid solution alloys. The possibility that the metal
particles might have deposited from the salt also was
expected but seemingly did not occur either during
circulation or during storage. In retrospect, these
observations are compatible with the early conclusion
that the pure salt fluids do not wet metal surfaces and
with the fact that the low viscosities of the fluoride
mixtures enable thermal convection currents to main-
tain the fluids in well-mixed condition during storage.
That thermal mixing occurs in these tanks was substan-
tiated by the first composition analysis in PC-2. As
evidenced by the composition of a sample which was
obtamned from No 2 dram tank shortly after 7LiF-
238JF, was added to the carrer salt, virtually com-
plete mixing of " LiF-UF, and carrier salt was achieved
before intertank transfers were made to ensure homoge-
nization Comparably efficient mixing 1s also probably
adequate to prevent settling of the very fine (below the
limit of microscopic detection) metallic particles of iron
and nickel in the drain tanks

A quality control program was imtiated by the
Analytical Chemistry Division as chemical surveillance
of the reactor salt began The results of this survey
showed that comparable accuracy and precision were
achieved i the analysis of structural metal impurities
in the salt samples 19 From the beginning, molybde-
num was not detected in the samples Its absence then
and later was interpreted to signify that such corrosion
as occurred 1n the reactor was essentially diffusion
finmted

3.6 Chemical Effects of Reprocessing

The MSRE ncluded a fuel processing plant for
removal of oxides, if necessary, from the fuel, flush,
and coolant salts The plant was also intended for use 1n
recovering uramum from the fuel and flush salts after
termination of expertments with the MSRE A detailed
description of the methods, operating procedures, and
components 1s given i the MSRE Design and Opera-
tions Report 't It proved to be unnecessary to repurify
any of the salt charges in MSRE maintenance opera-
tions, although flush salt was repurified after its mitial
use and prior to nuclear operations 1n order to remove
from the salt the oxides that 1t had accumulated during
its 1nitial use 1n the fuel system As planned, the
235,2380) fye] salt was fluorinated mn the chemucal
reprocessing plant after completion of experiments with
235y fuel salt Decision to utilize 233U as fuel entailed
not only fluorination of the 22> U salt but subsequent
reduction of the structural metal impunities that enter
the salt as 1t 1s fluonnated Decision to reuse the carrier
to constitute 2**U fuel made 1t exigent to remove
metal particulates from the salt carrier before returning
it to the fuel system Procedures for reduction of the
fluonides were evaluated and tested, and an optimum
filter medium was selected 12 A sintered metal filter
was then designed?3 and incorporated in the return hine
from the processing plant

Flush salt was reprocessed first to remove the
uramum 1t had accumulated 1n use (see Sect 4 2)
Following fluorination, the fuel was treated, as was the
flush salt, to reduce the structural metal fluorides
present in their metallic forms

55

The amounts of uranium which were recovered from
the fuel and flush salts were estimated from the
increases in weights of the NaF absorbers in which UFy
was collected The actual weights of UF, were not
measurable because fluormation of the salts in Hastel-
loy N processing vessels converted small amounts of
molybdenum to 1ts volatile fluondes, which were
absorbed along with UF4 on the sodmum fluoride beds
The approximate amounts of molybdenum absorbed on
these beds were estimated from the proportionate
amounts of 1ron, chromium, and nickel that were
dissolved mto the flush and carrier salts after fluorina-
tion Estimates of the weights of uranium recovered
were then obtammed by subtraction of the probable
amounts of molybdenum absorbed from the total
weights gained by the absorbers In a summary of
experience with the processimng plant, Lindauer has
estimated that the amounts of uranium recovered from
the flush and fuel salts at this time were ~6 5 and
217 85 kg respectively 14 The sodum fluoride absorb-
ers were subsequently dehvered to the Goodyear
Atomic Plant in Portsmouth, Chio, where wet chemical
procedures were used to recover the uranium charge
from the NaF absorbers The results of these efforts and
their significance are discussed 1n Sect 3 7

The reprocessing procedures were notably successful
i decontaminating the uranum charge from fission
products The only activity collected in any measurable
amount on the absorbers with UF, was ?°Nb None of
the filled absorber vessels exceeded maxmmum permis-
sible radiation levels, each vessel could be removed from
the processing plant without protective shielding

At termination of operations of the MSRE with 2°°U
fuel, the concentration of structural metal contami-
nants in the flush and fuel salts was scarcely greater
than when they were fust used Fluormation of the
salts 1 the processing plant mcreased the concentra-
tions of chromium, iron, and nickel contaminants
temporarilly On completion of reprocessing procedures,
therr concentrations were reduced once again (o satis-
factorily low values Changes in these concentrations as
a result of processing the salts are histed 1n Table 3 4

For operation of the reactor with 2**U fuel, a charge
of only about 40 kg of urantum was required as
compared with the some 220 kg of uranmuum that was
used during 23°U operations In order to provide a
sufficient volume of the fluid fuel mixture, the reactor
was loaded with a supplementary charge of fuel carrier
salt amounting to 129 9 kg

Methods for removal of the lanthanide element fission

products from the fuel streams of molten-salt reactors
using reductive extraction techniques were not fully
Table 3.4. Structural metal fluoride concentrations?

Concentration (ppm)
Crt10 Fetd40 Nit1s

Flush salt?
In reactor system
Before fluorination
After fluorination
After 10.8 hr of H, sparging

76 150 52
104
133 210 516

No sample taken

After 604 g of Zr and 9 hr of H, sparging 100¢ 174¢ 50¢

After 1074 g of Zr and 25 hr of H, sparging

No sample taken

After filtration 76¢ 141c 26
Fuel salt?

In reactor system 85 130 60
Before fluorination i70 131 36
After fluorination 420 400 840
After 17.1 hr of H, spaiging 420¢ 430¢ d
After 33.5 hr of Hysparging 420¢ 400¢ 520¢
After 51.1 hr of H, sparging 460¢ 380¢ 180¢
After 5000 g of Zr and 24 hr of H, sparging 1006¢ 110¢ <1¢¢

After 5100 g of Zr and 32 hr of H, sparging

After filtration

No sample taken
34 110 60

From R. B. Lindauer, Processing of the MSRE Flush and Fuel Salts,

ORNL-TM-2578 (August 1969).

ng sparging times are cumulative.

‘Filtered sample.
dContaminated sample.

developed until after the MSRE had operated for some
time. Until then the most promising method for their
separation was the high-temperature low-pressure distil-
lation of fuel carriers after removal of uranium. Tests of
this method with irradiated salt were scheduled to take
place at the MSRE on completion of 23U operations
after removal of uranium from the fuel salt. Approxi-
mately 48 liters of irradiated carrier salt was allocated
for this experiment; when the charge was delivered,
however, it was found that only a total of 12 liters was
transferred. Partly because of this anomalous behavior
and because of its importance in the evaluation of the
reactor fuel inventory, a considerable effort was made
to obtain an accurate measurement of the volume of
salt remaining in the fuel storage tank.

At the beginning of 33U operations, samples of fuel
salt were removed from the MSRE pump bowl and
analyzed regularly by colllometric methods which had
been applied previously and which were checked against
standards on a routine basis. Their precision and
accuracy is regarded to be ~+0.5%. The results of these
analyses were, on the average, greater than the nominal
values by 0.008 wt %, ~1% of the nominal value, and
indicated (disregarding precision limits for purposes of
this calculation) that the net weight of the carrier salt
during this period was 4639 kg rather than 4708 kg, or

that an additional mass of 69.0 kg (1.13 ft*) of salt,
that is, a total of 2.99 ft* of the original carrier salt,
was not returned to the reactor to constitute the 233U
fuel charge. In an attempt to support the conclusion
that some 3.0 ft* of carrier salt was left in the fuel
storage tank, an isotopic dilution experiment was
performed in April 1969, in which two samples of ® LiF
(35 g) were added to this tank. Retrieval of samples
from the tank using a windlass and 10-g sample ladles
was subsequently successful in removing small amounts
of salt (which incidentally were found to be encrusted
with metal residues) only through repeated attempts.
The total weight of ®LiF added was 66.28 g (15.8 g of
Li). Prior to ®LiF addition the carrier salt was found
by analysis to have an average " Li/ZLi concentration of
999905 £ 0.0015 wt %. The samples recovered after
the addition had a concentration of 99.914 + 0.024 wt
%15 which would indicate that the volume of the salt
is 3.0 £ 0.84 ft>. However, samples of the carrier salt
delivered to the still-pot section of the distillation
apparatus were found to have concentrations of °Li/
ZLi = 1.59 and 2.85 wt %, indicating that the ® LiF was
not dispersed homogeneously in the storage tank but
instead had dissolved preferentially in the salt fraction
delivered to the still pot. Homogeneous dispersal would
have resulted in a higher average ®Li concentration in
Table 3,5. Chemical composition of the MSRE carrier salt as indicated by chemical analysis

57

Number Chemical composition (mole %)
Specunen of

Samples LiF BCFQ ZI'F4
Carrier salt, as produced 36 62241090 32482090 5282016
Carrier salt, assunung transfer of 13 93 kg Z:4 62 33 32 58 515
Runs4 14, all samples 176 6461112 3029%10 510019
Run 14 (only) 44 6461 095 3020095 5192013
Runs 17 20, all samples 33 64 53+146 3043 %146 4901016

2Assumption of the removal of 13 93 kg of zirconium from this salt permits comparison after its dilution by six flush salt

residues

the sample recovered from the fuel storage tank, and a
lesser volume of salt would have been computed from
the results of the 1sotopic dilution analysis experiment
Thus, although its lower limit cannot be deduced
unequivocally from the 1sotopic data, the volume of salt
retained 1n the storage tank cannot have been as high as
3 84 ft?

The efforts to obtain samples from the storage tank
provided additional informatton that allows a separate
estimate of the salt volume Sometimes small quantities
of salt were obtained, sometimes none If this 1s
interpreted to mean that the salt level was at the point
in the dished head directly below the samplers, the pool
contained ~3 0 ft> of salt Assuming that the storage
tank contained 3 0 ft* of salt rather than 1 13 ft* used
before, we correct the earlier result 4707 — 1826 (30
ft3) kg + 1130 (1 13 ft?) kg and conclude that the
drain tank contained 4638 kg of carrer salt at the
beginming of 233U power operations Thus the net
weight of the fuel charge at the begmning of ***U
power operations was 4638 kg + 38 3 kg = 4676 kg, and
the uranium concentration was O 819 wt %

Ample thermodynamic evidence existed to indicate
that reprocessing procedures would not change the
average composition of the fuel carrer salt This
expectation was confirmed by comparnson of the
average composition of the fuel salt before and after
reprocessing and by determunatton that the uranwum
concentration m the 233U fuel mixture conformed to
our expectations A discussion of the uranmum assay
duning this period may be found m Sect 73 The
results of chemucal analyses shown 1n Table 3 5 support
the conclusion that the average composition of the fuel
carrier salt remained essentially constant throughout
the entire period of MSRE operations

3.7 Material Balances for 23°U and 2*® U Operations

37.1 Recovery of 235U and 2°%U. A matenal
balance of the fuel and flush salts was maintained on a

continuous basis from the beginning of experimental
operations with the MSRE With refinements 1n physi-
cal property data for the salts, in neutron absorption
cross-section data, and 1n analytical methods, the
quality of the balance mmproved steadily The accuracy
afforded by these improvements was established finally
by recovery operations at ORNL and at the Goodyear
Atomic Corporation, Piketon, Ohio There, the uranium
hexafluoride removed from the 22°U fuel and flush
salts by chemucal processing at ORNL was recovered by
wet chemical methods The amounts of uranium recov-
ered using these methods were 6 420 = 0 006 kg from
absorbers 6, 7, and 8 {from the flush salt) and 214 572
+ 0204 kg from the remaimmng 25 absorbers The
procedures employed 1n the recovery of uranium were
described 1 detail in a report from the Goodyear
Corporation as follows !©

The NaF matertal (25 absorbers) was processed using the
continuous dissolver and resulted in 120 batches of solution, tor
a total of 46,049 liters contamnng 214,456 grams of uranium
These batches were measured and sampled, i duphcate, using
the solution recovery accountability measurmng columns Batch
samples were proportionally composited, in duphcate, repre
senting from 10 to 20 batches of solution All composiie
samples were analyzed for total uramwum and weight percent
U 235 by thermal mass spectrometer In addition, the contin
uous dissolver operation generated 1,588 pounds of filter cake,
contaiming 105 grams of uramum This matenial was also
sampled in duplicate and composited on a weight basis Vent
losses resulting from steam effluent were calculated to be 11
grams of uranmum

The lumts of error used for the measurement of solution and
filter cake resulting from processing the remaiming sodmum
fluonde were 0 18 and *10 percent of the reported value per
composite analyzed The hmit of error for the total uranium
recovered was based on propagation of error applicable to each
composite group measured Also, ncluded in the hmat of error
1s a bias estimate for the volume measurements of £( 25 hiter
per batch, equivalent to 126 grams of uranum

These lumts of error are based on previously determined
precision and accuracy estimates Statistical evaluation of the
analyses obtained was not performed at this time since the
variation between samples indicates the precision will be well
within the estimated values. Analyses of standard samples for
comparison {0 known uranium value were made during the
processing period and showed excellent agreement.

The following precautions and special consideration were taken
while processing the NaF material to minimize measurement
uncertainty:

1. The continuous dissolver and solution recovery systems were
dedicated to processing NaF material exclusively from
mid-March through June, 1970, to minimize rinsing of
equipment as well as possible crossover of other material in
process.

2. The accountability measuring columns were recalibrated on
a weight basis and the recycle time required for representa-
tive sampling was determined by test data.

3. The accountability measuring columns wese rinsed twice, 30
liters per rinse, between each batch measured and sampled.

4. AH containers, plastic, and miscellancous scrap generated
during opening the 25 absorbers were deconiaminated and
the resulting solution processed through the dissolver.

5. Vent losses, even though minimal, were calculated based on
total steam flow to atmosphere and the uranium concentra-
tion determined by analyses of condensate samples through-
out the processing period.

6. System clean-out was initiated after processing the solution
generated from the last three absorbers which were esti-
mated at zero uranium content. This solution, 4,210 liters,
contained an average of 0.033 grams of uranium per liter.
Final rinse of the system averaged less than 0.002 grams
U/liter,

A disparity between the ORNL material balance and
the Goodyear results of 0.83 kg of 235U was evident.
In an attempt to resolve this discrepancy we requested
the Goodyear Corporation to retain the absorber vessels
until investigation of the cause of the disparity was
identified. In a telephone conversation between R. B.
* Lindauer and J. G. Crawford,!7 two pertinent details
were disclosed: the few fines remaining in the absorber
tanks were removed (by wiping the interior surfaces
with moistened tissues) and added to the process
solutions. Further, the interior surfaces of the absorbers
were noted to be bright and shiny, a condition which
seems to preclude the possibility that significant
amounts of uranium were retained on these surfaces.

The overall results strongly suggest that rationaliza-
tion of the disparity in the amount of uranium which
we anticipated would be recovered by the Goodyear
plant and that which was actually recovered would be
possible only through a careful review of MSRE
operations. In this connection, two possibilities were
suggested, by J. R. Engel and R. B. Lindauer, respec-
tively.!8

Engel noted that the particle filter (a 9-ft filter
designed to remove corrosion product solids from the

58

fluorinated salt before its reuse in the reactor) in the
line between the fuel storage tank and the processing
tanks could, after treatment of the flush salt was
completed, have contained an unknown amount of
zirconium metal, delivered to this location as the
processed flush salt was returned to the reactor system.
It is difficult to assign high probability to the events
which could have reduced the uranium from the fuel
charge as, subsequently, it passed through this filter, in
such a way that some 2 kg of uranium remained in the
filter; however, the possibility cannot be excluded and
merits further examination.

Another possible site where uranium may have been
retained, as has been suggested by R. B. Lindauer, is the
high-temperature sodium fluoride absorber bed which is
positioned between the fuel storage tank and the NaF
absorbers. The design temperature for operation of this
absorber is 750°F, based on previous laboratory
studies.!? The laboratory studies indicate that this
absorber would not retain UF, at the operating
temperature; the possibility that temperature gradients
prevailed within the absorber at periods near the end of
fluorination operations which allowed the retention of
some uranium within the absorber should now be
examined,

The possibilities that such errors as might be ascribed
to misestimates of power output of the reactor,
reactivity anomalies, and implications of short-term
trends in the results of chemical analyses, have been
reexamined; we conclude that their possible contribu-
tion to the disparity described is negligible.

The possibility that the retention of uranium in either
site might be established with certainty by application
of newly developed neutron interrogation techniques
using a californium source seemed favorable. Mockup
experiments were devised and tested in the.remote
maintenance practice cell at the MSRE with a neutron
source and an ORR fuel element. These tests showed,
however, that the method could not be used without
considerable modification, including procurement of a
more intense source, and the effort was abandoned.
Although confirmation of the results of the existing
isotopic dilution analyses would be desirable by direct
experiments such as were planned, the data which were
used to monitor the transfers of uranium and pluto-
nium within the reactor system appear to be suf-
ficiently reliable to estimate the amounts of 233U and
238 retained in the reprocessing system and for
computation of final inventory distribution.

3.7.2 Inventories for stored salts. Later in this report,
methods for establishment of the power output of the
MSRE are described (see Sect. 7). By application of the
59

Table 3.6. Inventory of residual uramum and plutonium n the MSRE?

A. Uranium

233y 234y 235(; 236 238y; sy
Fuel circuit inventory, run 20-1, kg 28 782 2 545 0876 0036 2035 34 274
Total inventory, run 20-1, kg 31052 2746 0945 0039 2196 36 978
Drain-tank inventory, run 20-I, kg 2270 0201 0069 0003 0161 2704
Fuel-circuit mventory, run 20-F, kg 28 739 2 563 0876 0 036 2035 34 249
Transfer to flush salt, run 20-F, kg 0411 0 037 0013 0 001 0029 0491
Charged mto drain tank, run 20-F, kg 28 328 2 526 0863 0035 2 006 33758
Drain tank residue, kg 2270 0201 0 069 0003 0161 2 704
Final dramn tank inventory, kg 30 598 2727 0932 0038 2167 36 462
U/sU, wt % 83918 7479 2556 0104 5943

B. Plutonmm
239p, 240p, 238,241,242p, sPu

Luel circuit mventory, run 20-1, g 6258 6181 2 39 690 0
Total mventory, run 201, g 6802 67 19 2 60 749 99
Dramn tank mventory, run 20-1, g 544 538 021 59 99
Fuel-circunt inventory, run 20F, g 6156 6543 2 37 683 4
Transfer to flush salt, un 20-F 2 g 618 6 57 023 68 6
Charged into dramn tank, run 20-F, g 5538 58 86 2 14 614 8
Drain tank residue, g 544 538 021 59 99
Final drain tank inventory, g 608 2 64 24 235 674 79
Pu/ZPu, wt % 90 13 952 035

IWeights are based on comparisons of analytical results and computed values These comparisons mdicate that maximum power
output was 7 4 MW(t) Final estimates assume 4167 EFPH (equivalent full-power hours) at 7 4 MW(t)

Brhis item makes the simphifying assumption that the total amount of plutonium estimated to be transferred to the flush satt was

transferred during the final flush of the fuel circust

1sotopic analyses used for this determination, together
with other analytical methods, it became possible to
estabhish the power output of the reactor with good
preciston, and accordingly to compute final materal
balances for the fuel and flush sait systems. The results
of these computations are listed in Table 3.6

In storage, the fuel salt 1s divided equally between the
two drain tanks. Both fuel and flush salt were frozen in
storage and are maintamed between 232 and 343°C to
minimize the evolution of fluormme from the frozen
salts

An 1nventory of the uranium and plutonmum con-
tamned 1n the drain tanks, based on data Iisted in Table
3.6, together with results of mass spectrometric analy-
ses, 1s shown in Table 3.7. Using the values in this
inventory the composition of the fuel salt was calcu-
lated. Nommnal values are compared with analytical data
in Table 3.8.

3.7.3 Salt loss from leakage. Experiments with the
MSRE were termunated on December 12, 1969. The
fuel and coolant salts were dramned for storage in the

drain tanks. After the fuel was dramned, and while the
freeze valves were being frozen, am increase n the
radiocactivity 1n the reactor containment cell was ob-
served, which indicated that a very small leak had
occurred 1n the prumary containment. After several
hours the activity began to decrease after having driven
the momitor on the recirculating cell atmosphere to a
maximum of 35 mR/hr. The fuel loop and drain tanks
were pressurized without causing an apparent effect on
the activity. A sample of the cell atmosphere indicated
that the activity was caused prmmanly from '?3Xe
Since leakage had apparently stopped, flush salt was
charged into the fuel system and circulated for 17 hr
Before flush salt was dramned from the circuit the
system was pressurized to 20 psig for 2 hr without
showing any signs of leakage It was concluded after
further investigation that the leak was in freeze valve
FV-105 or in the immediate vicinity of this valve To
conclude operations with the fuel circuat, flush salt was
transferred through the fill line to ensure that FV-103
and the adjacent line were filled with salt
Table 3.7. Material balance for the MSRE fuel and flush salts

Uranium concentration

Inventory (kg) (Wt %)

Uranium  Fuel salt  Carrier salt Calculated  Observed

Total uranium charged to the MSRE, 1964 229.020

Correction to 238U inventory4 -2.0

Fuel inventory at termination of run No. 3 227.02 4883.8 4656.8 4.648 4.648
Total uranium burned in 23°U operations, 9057 EFPH at 7.4 MW()?  —3.615 —-3.615 0

Uranium added as fuel replenishment +2.461 +3.074 +0.613

Total uranium removed in samples —0.256¢ ~5.583 —5.327

Uranium returned to circuit with 233U charge -1.935 0 0

Transfer balance, fuel to flush salt —6.420 -12.0 -12.0

Maximum amount of uranium recoverable 217.255 4865.68 4648.42

Maximum amount of uranium recovered (ORNL)4 217.99

Maximum amount of uranium recovered (Goodyear)€ 214.776

ORNL inventory minus Goodyear recovery value ~-2.479

Carrier salt retained in storage tank (3.0 ft3) ~182.6

Supplementary addition of carrier salt +129.9

LiF-233UF,4 (including 1.935 kg 235:238(J residue) 38.298 57.367  +21.00

Fuel salt charge at beginning of run No. 16 38.298 4654.80 4616.50 0.823 0.820
Total uranium burned in 233U operations ~1.316 -1.316 0

Uranium added as fuel replacement +0.389 +0.630 +0.241

Transfer balance, fuel to flush salt —-(0.289 ~0.289 0

‘Fotal uranium removed in samples —-0.025 -3.167 —3.142

Fuel inventory at termination of MSRE operations 37.057 4650.5 4613.6 0.797 0.792f

ISee sect. 2.4.2.

PBurnup rate: —0.3991042 g/EFPH. See B. E. Prince, MSR Program Semiannu. Progr. Rep. Aug. 31, 1969, ORNL-4449, p. 25.
€J. R. Engel, MSRE Book Uranium Inventories at Recovery of 235U Fuel Charge, MSR-68-79 (May 15, 1968).
dR. B. Lindauer, Chem. Technol. Div. Annu. Progr. Rep. May 31, 1969, ORNL4422, p. 40.

€See sect. 3.7.1.

f, Average of four samples obtained during run 20.

Table 3.8. Composition of fuel salt
stored in the MSRE drain tanks

TLiF BeF, ZiF, 233%2ygp, 23%1lipyp,

Mole Percent

Nominal 64.50 30.18 5.194 0.134
Analytical®? 64.53 30.43 4.90 0.137

2.38 % 1073

Weight Percent

Nominal 41.87 3544 21.67 1.033 0.0177
Analyticat? 41.37 35.27 20.19 1.06

“Current calculations do not include corrections for transfer
of carrier solvent residues to flush salt or flush salt residues to
fuel. Disparity between nominal and analytical values for
zirconium will be reduced by introduction of this correction
factor.

bRun 17-20, average of 33 samples.

References

1. E.S. Bettis et al., Nucl. Sci. Eng. 2,841 (1957).

2. ). A. Lane, H. G. MacPherson, and Frank Maslan
(eds.), Fluid Fuel Reactors, p. 590, Addison-Wesley,
Reading, Mass., 1958.

3. F. F. Blankenship, E. G. Bohlmann, S. S. Kirslis,
and E. L. Compere, Fission Product Behavior in the
MSRE, ORNL-4684 (in preparation).

4. A. L. Mathews and C. F. Baes, Jr., Oxide Chem-
istry and Thermodynamics of Molten Lithium Fluo-
ride—Beryllium Fluoride by Equilibration with Gaseous
Water—Hydrogen Fluoride Mixtures, thesis, ORNL-TM-
1129 (May 7, 1965).

5. G. Goldberg, A. S. Meyer, Jr., and J. C. White,
Anal. Chem. 32,314 (1960).

6. MSR Program Semiannu. Progr. Rep. Aug 31,
1965, ORNL-3872, p. 140.
7. MSR Program Semiannu. Progr. Rep. Feb. 28,
1965, ORNL-3812, p. 160.

8. C. E. Bamberger and C. F. Baes, Jr., MSR Program
Semiannu. Progr. Rep. Aug. 31, 1970, ORNL-4622, p.
91.

9. MSR Program Semiannu. Progr. Rep. Feb. 28,
1966, ORNL-3936, p. 168; ibid., Aug. 31, 1966,
ORNL-4037, p. 200;ibid., Feb. 28, 1967, ORNL4119,
p. 165.

10. M. T. Kelley, Statistical Quality Control Report,
January—March 1969, ORNL-CF-69-4-30; April-June
1969, ORNL-CF-69-7-65;July—September 1969, ORNL-
CF-69-10-36; October—December 1969, ORNL-CF-70-
1-37.

11. R. B. Lindauer, MSRE Design and Operations
Report, Part VII, ORNL-TM-907 (May 1965).

12. R. B. Lindauer and C. K. McGlothlan, Design,
Construction, and Testing of a Large Molten-Salt Filter,
ORNL-TM-2478 (March 1969).

13. J. H. Shaffer and L. E. McNeese, Removal of Ni,
Fe, and Cr Fluorides from Simulated MSRE Fuel
Carrier Salt, ORNL-CF-68-4-41 (April 1968).

14. R. B. Lindauer, Processing of the MSRE Flush
and Fuel Salts, ORNL-TM-2578 (August 1969).

15. Analyses performed by J. R. Sites, Analytical
Chemistry Division.

16. Correspondence from W. B. Thompson, Supervi-
sor of Process Engineering, Goodyear Atomic Corpora-
tion, Piketon, Ohio, to the Oak Ridge National Labora-
tory, September 2, 1970.

17. Goodyear Atomic Corporation, Piketon, Ohio.

18. Personal communication.

19. G. L. Cathers, M. R. Bennett, and C. J. Shipman,
MSR Program Semiannu. Progr. Rep. Aug. 31, 1968,
ORNL-4344, p. 321.

4. CHEMICAL COMPOSITION OF THE FLUSH
SALT DURING NUCLEAR OPERATIONS

4.1 Role of Flush Sait Analysis in the
Determination of Salt Residue Masses

The potential utility of a flush salt to scour contam-
inants from the MSRE fuel circuit was recognized early
in the MSRE development program; the sagacity of the
choice to use this salt was confirmed immediately after
preliminary experiments with the reactor — experi-
ments in which the flush salt was repurified after its
first period of circulation through the fuel system.! The
repurification procedures employed at that time
showed that in its first use the salt had served to remove

61

a considerable amount of contaminants from the fuel
and moderator graphite system. Not foreseen, however,
was the fact that it would function later to have a
significant influence on evaluations of reactor perform-
ance. This arose from the fact that the quantities of
residues of fuel and flush salt, intermittently cross-
transferred before and after reactor maintenance, could
not be adduced from on-site inferences. Soon after
experiments began with the circulating fuel salt, it
became evident that only the results of highly accurate
chemical analysis of the flush salt would make it
possible to establish reliable values for the amounts of
fuel and f{lush salt residues left in the reactor fuel
system after it was drained. Accuracy in these values
was also necessary in order to compute changes in
nominal values of the concentration of fissionable
material in the fuel circuit as the reactor experiment
proceeded.

Results of chemical analyses indicated that after the
fuel salt was drained, circulation of the flush salt
removed an average of 20 kg of fuel salt from the
drained circuit. Refinements in the values for the
physical properties of the fuel and flush salts allowed an
estimation from the relative density of the flush and
fuel salts that the average mass of flush salt residues
would be ~17.5 kg. With these values, nominal values
for the composition of the fuel salt were calculated for
the period when experiments were conducted with
235¢ fuel.

The total mass of flush salt originally charged into the
MSRE was about 4190 kg. From averages of the
increments in the changes in concentration of uranium
which developed in the flush salt in use, it was expected
that after its final use with ?°3U fuel the flush salt
would contain uranium at a concentration of 1500
ppm, representing a total pickup of 6.29 kg of uranium.
Analysis of the flush salt after its final use to remove
23517 fuel residue from the fuel circuit showed that the
concentration of uranium was 1488 ppm and, there-
fore, that the amount of uranium which was recover-
able from the salt was 6.23 kg; the actual amount of
uranium recovered was 6.420 kg, With the final amount
of uranium in the flush salt established as 6.420 kg, the
average increments were established as 0.917 kg, and
the average mass of fuel residue as 20.0 kg.

Complete results of the chemical analyses performed
with the flush-salt samples removed from the MSRE are
listed in Table 4.1. The average concentrations of
uranium in the samples removed from the systemn
during 2?3 U operations are shown in Fig. 4.1.

Flush salt was circulated in the MSRE fuel-salt
circuitry 13 times as a part of the 23°U operations and
62

Table 4.1. Summary of MSRE salt analyses, flush salt

Weight % Parts per million
Date Sample U4
Li Be Zr F b3 Fe  Cr Ni ob
Book Analytical

PC-1 13.12 9.68 77.08 5993 45 45 7 4320

FP-3-36 13.70 9.71 0.021 0.0195 77.68 101.17 144 64 75 12050
12/15/65  FP-4-1 35b
12/16/65 Fp-4-2 . 13.65 9.83 <0.0025 0.0210 80.52 104.04 116 <19 33 562
12/16/65 FP4-3. 46°
12/16/65 FP-4-4 13.55 9.35 <0.0025 0.0207 79.34  102.30 212 62 30 740
12/17/65 FP-4-5 72¢
12/17/65 FP-4-6 13.50 9.96 <0.0025 0.0200 80.07 103.58 125 <19 <20 180%
12/17/65 FP-4-7 13.65 9.46 <0.0025 0.0241 77.85 i01.03 i80 54 <20 1502
12/17/65 FP-4-8 106¢
12/17/65 FP-4-9 13.35 9.98 <0.0025 0.0221 75.80 29.20 210 37 <20 1425
12/18/65 FP-4-10 9.49 <0.0025 0.0230 75.05 98.26 128 60 <20 13002
9/28/66 FP-8-1 12.40 9.79 0.02 0.0409 80.70 10299 262 75 31
9/29/66 FP-8-2 12.40 9.66 0.25 0.505 78.14  101.01 197 82 225
9/30/66 FP-8-3 225
9/30/66 FP-8-4 12.43 9.02 0.22 0.458 75.75 98.93 382 53 74
11/2/66 FP-8-15 9.34 <0.01 0.0578 78.74 128 65 <15
11/3/66 FP-8-16 13.50 9.24 <0.10 0.0715 78.01 100.94 125 61 <15
11/4/66 FP-8-17 13.40 8.62 <0.10 0.0556 73.63 95.81 76 52 <15
11/24/66  FP-9-8 13.35 9.59 0.24 0.0840 72.15 95.43 175 53 26
12/11i/66  FP-10-1 13.15 8.90 0.145 0.0828 72.67 9498 169 71 53
12/11/66  FP-10-2 13.10 8.41 0.190 0.6747 73.03 9484 194 75 42
5/10/67 FP-11-59 13.43 7.57 (.345 0.0292 76.10 97.50 222 68 40
5/10/67 P-11-59 0.9304
5/16/67 FP-11-59 0.1080¢
511/67 FP-11-60 13.40 9.36 0.260 0.0268 79.23 102.30 119 74 34
6/16/67 ¥P-12-1 13.40 8.64 <0.2 0.0778 80.22 102.56 108 66 26
6/16/67 Fp-12-2 13.70 9.59 <0.2 (.0793 75.40 98.89 104 70 23
6/17/67 FP-12-3 13.60 9.38 <0.2 0.0826 77.40  100.68 93 68 24
6/17/67 FP-12-4 50-g sample for oxide analysis
9/8/67 FP-13-1 12.70 8.42 <0.2 0.1392 75.14 96.64 106 82 260
9/10/67 FP-13-2 12.80 8.84 0.72 0.1084 76.12 98.72 140 60 250
9/11/67 FP-13-3 12.80 10.04 0.70 0.1069 7772 10142 202 76 232
4/28/68 FP-14-72 12.53 9.66 0.42 0.1507 73.85 9664 150 76 52
4/28/68 FP-14-72 0.1488¢
8/2/68 FST-11 0.1172 104  (before F5)
8/4/68 FS8T-12 24 ppm 176 119  (after Fy)
8/4/68 FST-13 7 ppm 232 112 516 (after F5)
8/5/68 FST-14 6 ppm 174 133 543 (after F5)
8/8/68 FST-15 No sample obtamned
8/8/68 FST-16 0.30 206 136 36
8/8/68 FST-17 50-g FVS capsule after Zr addition 174 100 50
8/10/68 FST-18 No sample obtained
8/14/68 FP-15-1 13.43 9.16 0.546 0.0036 152 76 26

8/15/68 Fp-15-2 50-6 sample for eta experiment — no sample retrieved
8/15/68 FP-15-3 25-g sample for eta experiment

8/16/68 FP-15-4 25-g sample for eta experiment

8/11/69 FV-19-15

8/12/69 FP-19-2 12.43 9.64 0.51 0.0072 79.7 102.30 128 81 23
8/13/69 FP-19-3 50-g sample for oxide determination
8/13/69 FP-19-4 12.50 9.53 0.41 4.0065 79.2 101.68 197 74 25
8/13/69 FP-19-5 11.55 7.55 0.49 0.0075 80.5 100.14 241 80 40
12/13/69  FP-20-33 11.55 8.19 0.430 0.0118 7740  103.59 146 93 49
12/13/69  FP-20-34  50-g sample of flush salt for oxide analysis 7t
12/14/69  FP-20-35 11.50 8.45 0.540 0.0108 7790 1034 155 88 46

aValues corrected to 33.696 at. % 23°U. dRecheck by fluorometric method.

bKBrF4 method emploved unless otherwise noted. €Delayed neutron activation analysis.

¢HF purge method used.
63

. ORNL-DWG 71-9989

CTTTTTITTITTITITTTT]
® FLUOROMETRIC ANALYSIS 8 | _liiss
‘ A FLUOROMETRIC ANALYSIS, CONFIRM
® DELAYED NEUTRON ACTIVATION ANALYSIS
1400 = () EXPERIMENTAL AVERAGE
] e e b e {341
(1272)
1200
(7ss) T T'°7
B @ $
1000
oo T 1
— &
£
O
5
) ? @
800 iobzsol
(808) &
- ®
— 603
: 600 & (616)
l ®
@
“nll‘a?“ unm-426
400 ® (424)
&
by o oo i o feEs o 1T G ---.249
(215) 9 & ¢
200 (g F-26
» O o
P TT T T YT T I TRt an s
Mt ITTTTODOODoDoO Lo NN 0n T
0 o O o a o 000 o o0 oo ad ol 0O oo 4o A
| VRS 1 T W R I SN W SV N A U S TN I TN NN N 5 SN ¥ N S ¥ A I ENO ¥ W ¥ N W UV I ER § SO N

SAMPLE NUMBER

Fig. 4.1. Uranium concentration in MSRE flush salt. Calculated average mcrement = 177 ppm (includes transters of five fuel
residues totaling 11.34 kg from fill line).
three times as part of 233U operations, once on each
occasion before the fuel was constituted in the reactor
system, and thereafter whenever the fuel system was to
be opened for maintenance. If during the maintenance
period, the possibility developed that the fuel circuit
was exposed to ambient cell atmosphere, the circuit was
again washed with the molten flush salt before resump-
tion of experiments with the fuel salt. In use, the flush
salt was circulated for brief periods, usually 24 to 48 hr,
because it was assumed that molten LiF-BeF, flowing
in the metal circuit at 650°C would be extremely
effective as a solvent for contaminants that may have
entered the system during the previous maintenance
period. Gross exposures of the circuit to potentially
contaminating atmospheres were avoided during the
maintenance periods first by flushing the reactor system
with argon, and subsequently by maintaining a nitrogen
atmosphere in the standpipe column through which
access to the reactor core was gained. Neither blanket
gas was specially purified for this application.

It was judged that the flush salt was circulated for
sufficient periods of time to be effective, because, in
use, major increases were not observed in the concen-
tration of contaminants in the salt nor were there major
increases in the concentration of chromium in the fuel
salt on initiation of the subsequent experiments with
fuel salt. After each of these occasions, however, the
chromium concentration of the fuel salt gradually rose
to a higher value than had preceded the maintenance
period. The significance of these changes is discussed in
Sect. 6.

4.2 Transfer of Uranium and Plutonium
to Flush Sait in ?*2 U Operations

Reclamation of 6.420 kg of uranium from the flush
salt indicates that the average increase in uranium
concentration developed from the removal of a single
fuel-salt heel during 2*°U operations was 219 ppm
rather than the average 215 ppm we had observed from
analyses performed during the course of reactor opera-
tions. With the 233U fuel salt, a reduction of the
conceniration increments was anticipated proportionate
to the lower concentration of uranium in the fuel salt,
and should have resulted in an average increase of ~40
ppm per flush operation. The average of the two results
obtained with the flush salt after completion of run No.
20 shows that the final concentration of uranium in the
salt was 113 ppm, or that for each of the three times
flush salt was used to remove fuel salt residues, the
average increase in the concentration was 38 ppm, in
good agreement with the anticipated value.

64

Table 4.2. Results of mass spectrometric analysis
of MSRE flush salt

In weight percent

Uranium
233 234y 235y, 236y 238y

Sample No,

FP-20-33 40.55 3.73 17.67 0.25 37.80

FP-20-35 38.20 3.47 17.06 0.24 41.03
Plutonium
Sample No. 233py 240py 241py 242py
FP-20-33 94,74 4.78 0.44 0.04
FP-20-35 94.63 4.87 0.45 0.05

After termination of experiments with 233U fuel the
reactor was drained and flushed. Two samples of flush
salt were obtained (FP-20-33 and -35) in which the
average concentration of uranium was 113 ppm, an
increase of 42 ppm over that found during the single
previous use in which it removed a fuel-salt heel in
2331 operations (the beginning of run No. 19). The
single flush-salt sample obtained during the startup
operation preceding the introduction of 233U fuel into
the system was reported to have contained 36 ppm of
uranium. The three groups of data are thus consistent —
an indication that the concentration of uranium in-
creased as predicted. They also suggest that the flush
salt contained 35 to 40 ppm of uranium when it was
returned from the chemical reprocessing plant.

Isotopic analyses obtained from the last two samples
of flush salt are shown in Table 4.2. The isotopic
composition of the uranium contained in these samples
indicates that (correcting for the contribution from the
fuel of 0.023 kg and neglecting a correction for 238U
burnout) the flush salt contained ~50 ppm (0.21 kg} of
uranium. This is somewhat more than one would expect
considering the excellent agreement between expected
and observed amounts of uranium previously recovered
from the flush salt, and from analyses obtained during
the uranium recovery experiments with flush salt in
which results indicated the residual concentration was
no higher than 24 ppm, and possibly as low as 4 ppm.

4.3 Flush Salt Loss to Off-Gas Holdup Tank

The results of the chemical and mass spectromeiric
analyses described above provide a basis for the
conclusion that only very small amounts of either fuel
or flush salt could have been lost from the primary
containment or came in contact with the cell atmos-
phere through the leak that developed adjacent to the
freeze valves (see Sect. 3.6.3). If cell air were aspirated
into the fuel salt as it drained from the reactor at
termination of run No. 20, a probable consequence
would have been that ZrO, deposits may have formed
in the region of the break, later to be dissolved in the
flush salt. Possibly, some uranium would have been
deposited as well, along with the ZrO,. Small amounts
of ZrF, are transferred to the flush salt as a part of fuel
residues, yet, as shown in Table 4.1, the concentration
of zirconium in the flush salt after its final use does not
seem to have increased. Second, sizable amounts of
uranium deposited in the area of the break would both
increase the anticipated concentration in the flush salt
and alter the isotopic ratio in favor of an anomaious
increase in 223U fraction. The isotopic analyses shown
in Table 4.2 denote a slight anomaly in isotopic ratios
of uranium, but of opposite character than suggested
above; that is, they indicate that slightly more 235U
and 23%U were present than was expected from the
prior reprocessing operations. In one previous incident,
the possibility developed that flush salt was inadvert-
ently? lost from the salt system.

In Piper’s account® of operations in July 1966, he
notes that the fuel loop was filled with flush salt in
preparation for maintenance work in the reactor cell.
Flush salt was transferred into the overflow tank to
flush out the residual fuel and to check the indicated
level in the pump bowl at the overflow point, and
transfer began when the indicated level was 9.6 in. (In
January the indicated level at the overflow point had
been 9.2 in.) After approximately 0.6 ft*® of salt was
transferred, the fuel pump level suddenly rose off scale.
Rising level in the overflow tank triggered a drain, but
not before salt had entered some of the lines connected
to the top of the pump bowl. This behavior resulted
from an operational error which allowed the salt level in
the flush tank to be lowered too far, and also allowed
the pressurizing gas to enter the fill line and pass up
into the reactor vessel. The gas expanded rapidly as it
rose through the salt, causing salt to flood the pump
bowl. The pump bowl reference line was plugged with
frozen salt, and enough salt was frozen in the sampler
tube to obstruct passage of the latch. A thermocouple
indicated that some salt also entered the off-gas line,
but this line was not plugged. Salt also froze on the
annulus around the fuel-pump shaft, preventing its
rotation. In subsequent maintenance operations electric
heaters were applied to the outside of the lines to melt
out the salt in the bubbler reference line and the
sampler tube. The short flexible portion of the off-gas
line was replaced because of uncertainty over the
possible effects of salt in the convolutions.

65

We noted above that for the increments observed in
the change of concentration in the flush salt the
projected amount of uranium recoverable for the salt
was ~0.3 kg, and that a total of 6.42 kg was recovered.
The projected value was calculated from the final
analytical values, 1488 ppm X [4187 + 7 X 2.6 kg (the
average mass increase from fuel-flush cross transfer)] =
6.257 kg. The fact that 6.42 kg of uranium was
recovered suggests that if the analytical data are free of
bias, only a small volume of flush salt could have
transferred to the off-gas in the overflow accident. For
example, if 50 kg of flush salt were lost by transfer the
maximum amount of uranium which could be re-
covered from the flush salt at a concentration of 1488
ppm would not exceed 6.18 kg. Although it cannot be
contended that the above results unequivocally demon-
strate that the amount of flush salt transferred was
negligible, they provide convincing evidence that the
amount was indeed small.

In retrospect, the flush salt has served as the source of
several kinds of valuable information pertaining to the
performance of the MSRE. It now seems evident that it
could have been used more effectively as a more
valuable source if the periods of its use had been
extended and designated for additional chemical investi-
gations.

References

1. R. B. Lindauer, Processing of the MSRE Flush and
Fuel Salts, ORNL-TM-2578 (August 1969).

2. H. B. Piper, MSR Program Semiannu. Progr. Rep.
Aug. 31, 1966, ORNL-4037, p. 24.

5. CHEMICAL BEHAVIOR
OF THE COOLANT SALT

Of some 7000 kg of LiF-BeF, (66-34 mole %)
prepared for the MSRE, 2610 kg was designated for use
as the coolant salt and the remainder as flush salt. This
coolant was circulated by a 75-hp motor and pump
through a circuit which included a tube and shell heat
exchanger and an air-cooled radiator. Detailed descrip-
tions of this coolant system are given in the MSRFE
Design and Operations Report.' At termination of the
Molten-Salt Reactor Experiment the salt had circulated
in the system for a period of ~26,000 hr. With the
reactor at full power, coolant salt entered the tube
bundle of the heat exchanger (Fig. 5.1) at a tempera-
ture of 546°C (1015°F) and at 77 psig, removing heat
from the fuel salt as it circulated through the shell side
of the exchanger, and flowed on to the air-cooled
===
p——=

sz cEr
)
e

THERMAL-BARRIER PLATE e
==
TUBE SHEET = \\\\\\\\\\\
COOLANT INLET
%‘ =

T-STREAM COOLANT QUTLET

LAN
SEPARATING BAFFLE
FUEL OUTLET

i)

\\\\\@\\

66

ORNL.-LR-DWG 52036R2

FUEL INLET

i/2-13.-CD TUBES

=

CROSS BAFFLES

16.4-1n 0D x 0 2-in. WALL x 8-ft LONG

Fig. 5.1. Primary heat exchanger.

radiator (Fig. 5.2). On reentry to the heat exchanger,
the salt was at a temperature of 579°C (1075°F). The
coolant was circulated under these conditions for
approximately half the period of its use. At zero or low
power the salt was circulated isothermally, at about
650°C (1200°F). The coolant pump operated at con-
stant speed, developing a salt flow of 6286 kg/min (793

gpm).

5.1 Composition Analysis

The salt was circulated for some 1200 hr during the
prenuclear test period. Since flush and coolant salts
were supplied from the same inventory, the coolant salt
was analyzed during this period only for impurities.

At the end of prenuclear testing, concentrations of
the structural metal impurities, chromium, iron, and

nickel, were established to be approximately 30, 90, and
7 ppm respectively — remarkably low considering that
the circuit had not been flushed previously with molten
salt. The results of spectrochemical analyses showed
that no significant amounts of additional impurities
were introduced during this period. As the reactor was
brought to full power early in 1966, coolant salt was
again circulated, sampled regularly, and subjected to
compositional and impurity analysis. The results of
these analyses, together with all other analyses of the
coolant salt, are listed in Table 5.1.

It was stipulated that the chemical analyses include a
determination of zirconium concentration as part of the
procedures for coolant-salt analysis. Although such
analyses were not conceivably applicable as primary
indicators of fuel leakage, their incorporation supplied
base-line data for the contingency that possible analysis
67

ORNL-LR-DWG 55841R2

PENTHOUSE

DRIVE CHAIN .

[/

DOOR DRIVE MOTOR AND
GEAR REDUCER

|
’k
SUPPORTING STEEL ~fl

< WIRE ROPES —

w~
M B =y
DOOR CAM GUIDE 4

3

BLDG 7503, FIRST FLOOR
(ELEV 852 ft-0n)

SHOCK SPRING
B

R .

N
0
. AIR BAFFLES —
] ) i .
| / | a \/\;«\
. AIR INLET DUCT ~~_ " §T~AIR QUTLET DUCT
- / . 3
/" AIR DUCT FLANGE — Al Goane e
L ‘ b ‘ 3 | / {
- ~
. MAIN AIR BYPASS DUCT fi 5 RADIATOR ENCLOSURE

RADIATOR TUBES

AIR FLOW - DAMPER

Fig. 5.2. MSRE radiator coil and enclosure.
Table 5.1. Summary of analytical chemical data for the MSRE coolant circuit

68

Date Total Number Sample Li Be Zr F Cr Fe Ni 0
Hr, in circuit Design (wt %)

1/9/65 0.3 CP1-1 - - - - 20 117 8 420

1/10/65 24 CP1-2 - - - - 25 181 <5 433

1/11765 b4 CP1-3 - - - - 26 184 <5 472

1/13/65 94 CP1l-4 - - - - 27 96 12 580

1/15/65 143 cPl-5 - - - - 27 83 <12 325

1/17/65 186 CPl-6 - - - - 27 93 <12 525

1/20/65 267 Run 1-F

1/22/65 267 Run 2-F

1/22/65 277 CP2-1 - - - - 30 63 5 388

1/27/65 393 Cp2-2 - - - - 29 101 7 270

2/1/65 514 CP2-3 - - - - 28 65 25 263

2/11/65 757 CP2-4

2/18/65 913 cp2-5

2/24/65 1062 CP2-6

3/5/65 1249 cp2-7

3/5/65 1249 Run 2-F

6/26/65 1249 Run 3-1

6/26/65 1250 CP3-1

6/30/65 1342 CP3-2

6/30/65 1364 CP3-3 12.87 9.53 - 33 78 38 695

7/1/65 1370 Run 3-F

1/8/66 137¢ Run 4-I

1/8/66 1371 CP4~1 36 65 <5 -

1/18/66 1375 CP4-2 26 83 <5 130

1/20/66 1423 CP4-3 13.78 8.91 16.50 40 40 40 190

1/20/66 1431 CP4-4 50 g sample for oxide analysis 25b

1/25/66 1544 CP4-5 14.20 8.87 <0.002 76.80 50 50 20 180

1/25/6% 1558 CP4~6 50 g sample for oxide analysis 38P

1/27/66 1593 Run 4-F

2/7/66 1593 Run 5-I

2/7/66 1600 CP5-1 13.86 8.85 0.0015 76.70 35 5 <10 150

2/10/66 1664 CP5-2 13.82 8.55 0.0016 76.60 35 <2 <10 110

2/15/66 1784 CP5~3 12,76 10.51 0.0062 76.04 46 57 64 65

2/18/66 1865 CP5-4 12.60 9.17 0.0175 76.80 35 119 14 <20

2/23/66 1977 CP5-5 12.60 9.15 0.0153 76.70 42 87 28 120

3/2/66 2176 CP5-6 13.01 10.06 0.0050 77.20 20 100 8 260

3/4/66 2219 Cp5-7 13.17 9.49 0.00%9 76.40 24 112 <2 171

3/8/66 2331 CP5-8 12.70 10.22 0.0069 77.20 29 36 <2 71

3/11/66 2406 CP5-9 12.77 9.43 0.0105 76.70 32 44 <2 35

3/15/66 2482 CP5-10 12.97 5.59 0.0016 76.90 33 31 <2 <20

3/20/66 2560 CP5-11 12,92 9,61 0.0032 76.40 36 49 53 <20

3/22/66 7648 CP5-12 12.75 9,74 0.0030 76.40 44 57 8 <20

3/25/66 2722 CP5~-13 13.10 9.98 0.0997 76.20 32 73 10 162

3/25/66 2730 Run 5-F

3/29/66 2730 Run 6-1

3/29/66 2825 CP6~1 13.09 9.73 0.0031 76.10 30 43 41 <20

4/1/66 2907 CP6-2 12.68 9.56 0.0220 76,40 38 28 9 <20

4/20/66 3321 CP6-3 12.82 9.13 - 76.80 45 76 47 210

4/24/66 3431 CP6~4 12,76 9.63 0,0055 76.00 57 68 15 210

5/16/66 3946 CP6~5 12.70 9.79 0.0095 74.80 20 57 6 135

6/8/66 4499 CP6-6 - 9.65 0.0050 74.80 28 37 6 <40

6/8/66 4499 Run 6-F

6/28/66 4499 Run 71

6/28/66 4807 cP7-1 13.64 8.72 0.0032 14.65 32 60 <2 <20

6/29/66 4826 cp7-2 12.16 9.88 0.0070 74.50 31 51 <2 300

7/5/66 4967 CP7-3 12,39 10.24 0.0096 74.80 51 80 <2 145

7/12/66 5134 CP7-4 12.70 9.26 0.0008 75.90 46 24 28 85

7/22/166 5374 Run 7-F

10/2/66 5374 Run 8-1I

10/2/66 5376 CcPg-1 12.90 9.94 0.0008 74.55 17 23 <5 53

10/25/66 5929 cPg-2 12.40 9.35 0.0013 75.70 24 19 <2 <25

10/25/66 Run 8-F

11/8/66 Run 9-1

11/8/66 6264 CP%-1 12,30 9.71 0.0012 17.40 4 30 29 73

11/22/66 6608 CP9-2 12.65 9.55 <0.0005 74.23 33 25 <29 75

11/22/66 Run 9-F

12/7/66 Run 10-1

12/7/66 6962 CPi0-1 i1.90 9.64 ¢.0033 74.80 30 36 <5 75

1/5/67 7655 CP10-2 12.80 9. 40 0.0032 74.50 20 653 34 475

1/5/67 Rur 10-F

2/2/67 Run 11-1

2/2/617 8570 CP11-~1 12.50 9.73 <0.,0020 77.00 18 78 <10 330

2/16/67 8798 CP11-2 11.90 9.57 0.0017 76.50 23 35 <10 530

3/7/67 9258 CP11-3 13.20 9.46 0.0090 76.00 13 105 15 380

3/21/67 9751 CP11-4

4/13/67 10145 CP11-5 12.90 9.70 <0.002 76.00 32 24 <10 -

5/11/67 10836 CP11-6 13.50 9.46 0.0060 76.10 35 35 14 314

5/11/67 Run 11-~F

6/16/67 Run 12-1

6/16/67 10848 CPl2-1 12.70 9.30 0.0074 77.40 39 168 10 -

7/6/67 11346 CP12-2 14.90 9.10 <0.0026 75.10 26 33 <20 88

7/11/67 11460 CP12-3 12.73 9.40 <0.0040 74.30 30 27 26 230

7/11/67 Run F-12
69

Table 5.1 (continued)

Date Total Number Sample Li Be Zr F Cr Fe Ni 0®
Hr. in circuit Design wt %) (ppm)
Run 13,14-I
g;;?égy 12150 CP13-1’ 12.40 9,27 0.0045 73.80 <15 62 zgg igg
9/29/67 12550 CP1l4-1 12,60 9.87 0.0098 72.80 24 <10 20 e
10/25/67 13174 CP14-2 14.30 8.86 0.0011 76.32 24 25 o o
11/27/67 13966 CP14-3 15.20 9.35 0.0100 75.50 24 61 o g
1/18/68 15214 CPl4-4 12.50 10.10 0.0019 76.10 25 2; Pt o
3/20/68 16702 CP14-5 14.60 9.61 0.0078 17.20 <15 3
3/28/68 16894 Run 1l&~F
9/ /[68 Run 15-1
9/17/68 17714 CP15-1
11/27/68 19414 Run 15-F
12/ /68 Run 16-1
12/10/68 19524 CP16-1
12/ /68 Run 16-F
12/6 Run 17-1
;§4§é99 20076 CpP17-1 13.10 9.26 <0.0020 76.60 26 21 <1; zggg
3/6/69 20796 CP17-2 13,40 9.49 0.0021 77.00 25 15
Run 17-F
Run 18-I
4/27/69 22044 Cp18-1 13.00 9.52 ¢.0013 77.60 25 12 ;5 ;gg
5/29/69 22812 CP18-2 13.70 9,22 0.0014 77.50 25 14 <10
22860 Run 18-F
2860 Run 19-1
8/10/69 52370 {rie-1 13.40 9,44 <0.0019 77.30 35 <5 10 1820
9/14/69 22966 Cr19-2¢ 13.40 9.25 0,112 77.60 35 143 2
9/14/69 22966 CcP19-3¢ 13.85 9.13 0.103 78,30 35 128
11/2/69 23084 Run 19-F
11/26/6 23084 Run 20-I
llfiZGifig 23094 CP20-1 14,00 9.55 0.0019 77.20 46 32 11 3200
12/10/69 23430 Cr20-2 13.70 9.31 0.0206 77.20 60 114 10 1700b
12/11/69 23454 CP20-3 Ni bar, exposed to coolant for tritium experiments
12/11/69 23454 CP20-4 N1 bar, exposed to coolant for tritium experiments
12/12/69 23479 CP20-5 Cu0 specimens, exposed to coolant for tritium experiments.
12/16/69 23479 CP20~6 Cul specimens, exposed to coolant for tritium experiments.
12/12/69 Run 20-F

Lnless otherwise noted, oxygen analyses were obtained by the KBrF

a
b Results obtained by HF~Hp transpiration method,

< Samples analyzed bv HRLAL personncl.

of fuel-coolant mixing accidents might arise. Results are
omitted from Table 5.1, because in none of the analyses
of the coolant salt was zirconium detected.

As noted in Sect. 3.3, procedures used for the determi-
nation of the concentrations of oxides in MSRE samples
were unsatisfactory until development of the HF-H,
transpiration method was completed. Not until late in
the program of reactor operations was this method
applied to coolant-salt samples, and then, for the single
sample available, it produced questionable results. The
results of analyses for oxides listed in Table 5.1 were
obtained with the KBrF, method, a method which is
generally satisfactory for nonhygroscopic materials, but
which was applied to samples which are hygroscopic
and had not remained isolated from ambient atmo-
spheres after their removal from the reactor. In view of
the established absence of corrosion in the coolant
circuit, their credibility is dubious.

Analyses of coolant salts were conducted by the
General Analysis Laboratory of the ORNL Analytical
Chemistry Division, while those for the fuel were
obtained from the High-Radiation-Level Analytical
Laboratory of that Division.

4

method.

5.2 Corrosion Behavior

Corrosion of Hastelloy N as a container for molten
LiF-BeF, mixtures may originate from a very limited
number of sources: from impurities in the melt, from
oxide films on the metal, and from mass transfer of
metal constituents in the fluoride. Of these sources,
only the latter, which is caused by the differential
temperature coefficient for solubility of metals in salts,
affords a mode of continuous attack in reactor systems
that are protected from inleakage of contaminants.

Since the procedures adopted for MSRE operations
assured that the circuit would be free from chronic
sources of contaminants, it was anticipated that surveil-
lance of the coolant salt would indicate imperceptibly
low corrosion throughout the experiment.

In the early stages of MSRE operations, samples of
the flush salt were removed from the circuit at a rate of
one per week. However, as our experience developed,
and it was confirmed that chroni¢ sources of oxidizing
impurities were absent, the frequency was decreased,
finally to intervals of a month or longer. In contrast to
the fuel system, the coolant salt system was not
ORNL-DWG 71~9990

&0

50 b

40 e S

20 2 @

0 5 10

15 20 { x10%)

CUMULATIVE CIRCULATION PERIOD (hr)

Fig. 5.3. Concentration of chromium in the MSRE coolant circuit salt.

exposed to the cell atmosphere during the course of
MSRE operations. Under these circumstances, it was
reasonable to expect that corrosion would not occur in
the circuit, and these expectations were borne out in
MSRE operations as indicated by the results of chemi-
cal analyses, and later in the postoperational examina-
tions of the radiator (see below).

The results of the chemical analyses performed with
coolant-salt samples showed that no measurable in-
crease in the concentration of Cr, Fe, and Ni (with
average values of 32, 64, and 20 ppm respectively)
developed after the coolant salt was first charged into
the MSRE. The concentrations of chromium in the
coolant-salt samples, as determined in chemical analy-
ses, are shown in Fig. 5.3. The fact that the chromium
concentration remained unchanged is remarkable, since
it indicates that within the limitation of the analytical
precision (7 ppm) no corrosion, excepting that pos-
sibly resulting from mass transfer, occurred in the
coolant circuit during the entire period of MSRE
operations. The demonstrated compatibility of the
coolant salt with its containment alloy is believed to be
unmatched in any prior experience with either molten
salts or liquid metals as recirculating heat-exchange
media.

The average value of the chromium concentration in
the coolant salt remained at 32 ppm for the entire
period of reactor operation. The apparent trend toward
higher values toward the end of the use period, as is
reflected in Fig. 5.3, tends to discount the credibility of
the view that the system remained free of contami-

nants. However, the results of postoperational examina-
tions support this view completely and indicate thereby
that continued use and analysis of the salt would have
shown that the data were not indicative of a trend, but
rather represented normal scatter. As part of the
postoperational examination of MSRE components,
sections of the radiator tubing were removed from the
inlet and outlet ends and tested by ORNL metallurgists.
Their analysis of the surfaces of the alloy that were
exposed to the coolant salt by electron microprobe
techniques disclosed that no compositional variations
in Cr, Mo, Ni, or Fe existed within 2 u (the allowable
working range on these samples) of the surfaces.?

Comparisons of the chemical composition of sections
of this tubing with machined turnings showed a higher
concentration of carbon at the salt interface after
service, as was suggested from examinations of the
microstructure.> However, changes in the tensile
strength of the alloy were found to be slight and did
not indicate serious embrittlement of the alloy.

The chemical behavior of the coolant salt in the
MSRE is a unique demonstration of the compatibility
of pure fluoride mixtures with nickel-based alloys, and
shows as well that, with normal precautions to ensure
that oxidizing contaminants are prevented from enter-
ing these systems, corrosion will not occur. The
behavior of the coolant salt in the MSRE is a benign
indicator of the potential application of LiF-BeF,
based systems, but has only marginal implication with
respect to the coolant salt of larger-scale molten-salt
reactors in which lower liquidus temperatures are

mandatory for the coolant, and where the capital cost
of the coolant salt will have a pronounced effect on
power economics. These factors seem to preclude the
choice of 7LiF-BeF, mixtures as coolants in such
reactors. The current choice for the MSBR coolant salt
is the NaF-NaBF, eutectic mixture. The highly success-
ful performance of the molten fluoride salt "LiF-BeF,
in the MSRE is, however, of special significance in that
it leads to the recognition that the possibilities of
materials combinations are more flexible than was
estimable prior to operation of the MSRE.

References

1. R. C. Robertson, MSRE Design and Operations
Report, Parr I, ORNL-TM-728 (January 1965).

2. H. E. McCoy and B. McNabb, MSR Program
Semiannu. Progr. Rep. Aug. 31, 1970, ORNL-4622, p.
120.

6. CORROSION IN THE FUEL CIRCUIT
6.1 Modes of Corrosion

Early recognition that numerous inorganic fluorides
are among the most chemically stable materials in
nature supplied the initial impetus for investigation of
the chemical feasibility of molten-salt reactors. It was
noted that the fluorides of commonplace alloy con-
stituents such as iron, nickel, chromium, and molyb-
denum were less stable than the various compounds
which might serve as components of molten-salt reactor
fuels and coolants, and thus good alloys were poten-
tially available as containers for molten fluoride mix-
tures (see Table 6.1). Subsequently, a materials develop-
ment program was initiated culminating in the develop-
ment of the alloy now designated as Hastelloy N
specifically for use in constructing the MSRE; its
approximate composition was Ni-Mo-Cr-Fe (71-17-7-5
wt %). The basis for the selection of this exact
composition is described by Taboada' in a detailed
review of the alloy development program.

Since the major components of the MSRE salts, LiF,
BeF,, ZrF,, and UF,, are much more stable than the
fluorides which can result from the corrosion of
Hastelloy N, that is, MoF;, NiF,, and FeF,, compati-
bility of the molten-salt mixtures and the alloy was
essentially assured.

The most probable mechanisms by which corrosion
might occur in the MSRE were identified and examined
extensively prior to operation of the reactor. The
chemistry of such processes was reviewed recently by

71

Table 6.1. Relative stability? of fluorides
for use in molten-salt reactors

Free energy Absorption cross

Compound of formatoion N}I?eoiit;r;g section? for
at 1000°K ©C) thermal neutrons
{kcal per F atom) {barns)
Structural metal fluorides
CrF, 74 1100 3.1
FeF, ~66.5 930 2.5
NiF, —58 1330 4.6
MoF g =50 17 2.4
Diluent fluorides
CaF, —125 1330 0.43
LiF —125 848 (.033¢
Bak, —124 1280 1.17
StF, —-123 1400 1.16
CeF4 —118 1430 0.7
YF; ~113 1144 1.27
MgF, —-113 1270 0.063
RbF -112 792 0.70
NaF —112 995 0.53
KF - 109 856 1.97
BeF, —104 548 0.010
ZrF, —94 903 0.180
AlF 4 -90 1404 0.23
SnF, ~62 213 0.6
Pbl, —-62 850 0.17
BiF, -50 727 0.032
Active fluorides
ThE, —101 it
UF,4 -95.3 1035
UF, -100.4 1495

9Reference state is the pure crystalline solid; these values are,
accordingly, only very approximately those for solutions in
molten mixtures.

bOf metallic ion.

€Cross section for 7Li.

Grimes.? In the absence of impurities, corrosion pro-
ceeds through two mechanisms, by mass transfer and by
selective oxidation of chromium, the most chemically
active constituent of the container alloy. Both types of
corrosion are limited by self-diffusion of chromium in
the alloy and by temperature, which in the MSRE fuel
reached extremes of 654°C (1210°F) and 632°C
(1170°F).

Preponderantly, oxidation-reduction reactions are re-
sponsible for all the corrosion in molten-salt reactor
fuel systems. The principal reaction which controls the
process is

UF,4(d) + %, Cr(e) = UF5(d) + % CrF, (d) (1)
for which the standard free energy at 1000°K (727°C)
is +15.1 kcal; at this temperature the equilibrium
constant for the reaction is

Ny, Netk, )
K= =5 X 1074, (2)
Nyr, ac;

where NV is mole fraction of a reactant or product and «
is the activity coefficient. For purposes of examining
the way this equilibrium affects corrosion behavior,
assume conditions similar to those existent with 233U
fuel, where Ny, = 1.38 X 1073, Cr** = 5.15 X 1073
(150 ppm), an, = 3 X 10722 The equilibrium
concentration of UF; thus becomes 2.832 X 107° and

Nyg,/Nyp, =2.832 X 107°/1.38 X 107 = 0.205% .

These results illustrate that the reaction tends to come
to equilibrium with a small fraction of the uranium in
the trivalent state. Chemical factors which remove UF,
from the salt system promote the forward reaction (1)
and tend to remove chromium from the container atloy
by selective oxidation.

Fission in molten-salt reactor fuel systems causes a
gradual increase in the oxidation potential in molten-
salt fuels (based on UF,) as uranium is consumed.
From estimates of fission yields* and probable oxida-
tion states of the species produced, the effect of fuel
burnup on corrosion can be evaluated. A recent
appraisal of this balance indicates that when reducing
conditions are maintained and xenon and krypton are
removed rapidly from the fuel, the sum of the electrical
charges on the fission product cations is less than an
average value of +4 per mole of uranium burned, and
~0.76 equivalent of oxidation results from the fission
of one gram atomic weight of uranium.’ It was useful
for this reason to increase the Ny /Nyp, concen-
tration in the fuel salt occasionally as a means of
IINITNIZIng corrosion.

Self-diffusion coefficients of chromium in nickel-base
alloys in the temperature range 600 to 900°C were
determined by Watson and co-workers® by monitoring
the total intake of %! Cr by the alloys exposed to salt
solutions containing this radiotracer and by measuring
the tracer concentration profiles through successive
electropolishing of the specimens.

They found from the loop experiments that the
diffusion coefficient of 5'Cr in Hastelloy N at 650°C
(1200°F), the mean temperature of the MSRE fuel, was
1 X 107'* cm?/sec. In recent examination of metal
surveillance specimens from the MSRE core,” evidence

72

ORNL-DWG 70-4833

7 ==
-1
— & ,// //
\e
8 y
Z / 14 2
S s / MEASURED (0 = 4 x 107 '® cm2/sec}———
l_
< /
S
= 4 7~ 7L CALCULATED (£ = 2 x 10~ ** cm2/sec)
2.0/
O / /
2 !/
= 4 ~
o /
[nag
T /
[GI
)/
0
0 1 2 3 4

DISTANCE FROM SURFACE (mils)

Fig. 6.1, Chromium gradient in Hastelloy N sample exposed
in MSRE core for 22,533 br.

was found that indicated the diffusion coefficient of
chromium in the core specimens may have been as high
as 4 X 107'* ¢m?/sec (Fig. 6.1). The fact that the data
from the MSRE samples were of necessity obtained
from hot-cell operations, whereas no such constraints
were imposed in the original laboratory measurement of
self-diffusion coefficients, lends somewhat greater credi-
bility to Watson’s values. Both values may be accurate,
however, because Watson noted that for Inconel the
self-diffusion coefficients of chromium were strongly
dependent on annealing conditions at low temperatures.
Conditions leading to large grains led to low diffusion
coefficients and vice versa.

Extrapolation of Watson’s data for Hastelloy N (see
Fig. 6.2) indicates that the self-diffusion coefficient for
Cr in Hastelloy N at the lowest temperature of the
MSRE fuel circuit is 8 X 107! ¢m?/sec. The rate of
mass transfer is limited by the rate of diffusion of
chromium into the alloy in the coolest area of the
system. Corrosion from mass transfer was therefore
expected to be of little consequence in the MSRE.*

Fick’s equation, M, = 2Cy(Dt/m)'/? (where C, =
concentration of the diffusing element in the bulk
species, D = diffusion coefficient, # = time) can be used
to predict the quantity of material removed by dif-

*Postoperational examination of sections of Hastelloy N from
the heat exchanger disclosed no evidence of an enrichment of
chromium at the salt-metal interface. This observation has led
to a suggestion by F. F. Blankenship that since the coolest
location in the fuel circuit is in the area of freeze valve 103, it
would be useful in a future examination to determine the
profile of chromium concentration in the alloy at this location.
73

ORNL-LR-DWG 46403
TEMPERATURE, (°C)

850 800 750 700

N L

A COEFFICIENTS BASED ON TOTAL SPECIMEN COUNTS

® COEFFICIENTS BASED ON Cr3' CONCENTRATION —
GRADIENT MEASURED BY COUNTING ELECTROLYTE

© COEFFICIENTS S8ASED ON Cr®' CONCENTRATION

@ GRADIENT MEASURED BY COUNTING SPECIMEN
-2 |8
’g @
3 \\\O\ @ oo
e LN © 0
L A \
Q 4a A\ka\\i‘
2 A
g 8

. T~

T

N A
o] \A
fa

OBTAINED UNDER NON-CORROSIVE CONDITIONS ¢ ° T8
BY ADDITION OF Cr*F, TO NoF-Zrf, l
-14
8.75 9.00 9.25 9.50 9.75 10.00 10.25 10.50

‘O’OOO/TPK_')

Fig. 6.2. Self-diffusion coefficients for >1¢r in INOR-8; loop 1248, (from ref. 3).

Table 6.2. Comparison of predicted and observed amounts of chromium in 2350 fuel salt

Diffusion coefficient = D

Predicted
Time (Grams Parts per million Observed
o) po1x10® p=2x107* p=4x10"* p=1x10* p=2x10""* p=4axio*  rm)
cmy/sec cmz/sec cmz/sec cmz/sec em? [sec cm2/sec

1,000 263 372 525 54 76 108 48
2,000 371 525 743 76 108 152 48
4,000 548 775 1097 112 159 225 48
6,000 643 910 1286 136 186 263 48
8,000 743 1050 1486 152 215 304 62
10,000 337 1184 1674 171 242 343 64
12,000 909 1286 1819 186 263 372 64
14,000 982 1387 1965 201 284 402 70
16,000 1050 1486 2101 215 304 430 72
18,000 1114 1575 2227 228 322 455 82
20,000 1174 1661 2349 240 3490 480 82

fusion under conditions where the surface concen-
tration of the diffusing element is zero. Table 6.2 lists
the amounts of chromium which might have been
expected to appear in the fuel salt during 23°U
operations if all the chromium accessible to the salt
were oxidized to CrF,. It is apparent from these data
that generalized corrosion, as inferred from increases in
chromium concentration of the fuel salt, was only
one-fourth to one-third of that expected from the

diffusion coefficients. By contrast, the apparent corro-
sion rates that were observed during the initial stages of
MSRE operations with 233U fuel and again in the
beginning of run 19 were approximately 5.4 and 3.84
times as rapid as predicted using a diffusion coefficient
of 2 X 107** cm?/sec. Such rates were inconsistent
with previous observations and more rapid than the
maximum allowed by the maximum value observed for
the diffusion coefficient, 4 X 1071% c¢m? /sec. Since the
rates of corrosion on these occasions were unlikely to
be permitted by chromium diffusion, it seems likely
that bulk diffusion involving other constituents of the
alloy was operative.

From the following argument it is concluded that the
anomalously rapid corrosion during these periods and
occasionally during 22>U operations was caused by
contaminants introduced into the fuel system during
periods of maintenance.

6.2 Corrosion in Prenuclear Operations

As noted in Chap. 2, chemical analyses were per-
formed concurrently in the General Analysis Labora-
tory and in the High-Radiation-Level Analytical Labora-
tory with salt samples removed from the MSRE during
the period before nuclear operations began. This pro-
cedure reflected that a slight operational bias existed in
the results of the two laboratories® for some of the
species and enabled us to use those biases as correction
factors for assessment of changes in the salt. With
respect to chromium analyses as a corrosion indicator,
the data corrected by this factor showed that at the
average value of its concentration, 37 + 7 ppm, no
change in the chromium concentration of the fuel
circuit salt had occurred throughout the period ending
with the completion of the zero-power experiments. In

the low-power experiments following the first main-
tenance period, during which time the pump rotor was
removed for examination, the new average value of
chromium in the fuel salt was 48 * ppm. Since the
standard deviations for these two values overlap, it can
be inferred that the flushing operation preceding the
low-power operation (run 4) effectively removed
corrosion-inducing contaminants that might have
entered the system during the maintenance period.

6.3 Corrosion in Power Operations

In Fig. 6.3 the results of all chemical analyses of fuel
salt samples for chromium are summarized for power
runs. The scattered data are described most simply as
increasing linearly with time and correspond io an
increase in chromium concentration in the salt at a rate
of 12 ppm per year. Statistical analyses of the indi-
vidual groups of data, however, indicate that the
average values (least-squares method) are as listed in
Table 3.3.

In operation of the reactor with 233U fuel, the
concentration of chromium in the samples of fuel salt
increased sharply only during two periods, during the
initial stages of run Nos. 15 and 19. One interpretation
of these analytical results is that corrosion rates
exceeded those predicted by diffusion data only after

ORNL-DWG 70-2164

150
140 ¥
130 - b —
120 L -
0.30-
10 - 0.45
100 . 0.25- “( =
- 0.40 1ol
& 90 . e /
= , - 0.20- . ff
2 80 — R P -0.35 . , o ;
= . 1] & ‘2 /e s . . -
8 70 5 * .| =i O 1 |—_ .
T . P B A A ' -0.30 £ . .
I -{/‘— o --:" R | . 0.10 S g
. o e RN L0.25 | / ]
T f' - :: 1.1 o 0.054 /
; { T - -0.20 (
1N 0- .
i RUN: 4-14 15-20 + -
| ] E | 3 CORROSION | | |
L] 1 i fiu | O T ol | GEEE | (mils) B ]
RUNNO. |4 |5 7 8|9i 10 1 12| 113 14 5 |16 T 18 19 20
FLUSH ® [ ] 1 LI II‘ '‘BEIEREL 1 l 1: [ i = l ‘ '
DJFM‘AMJJASOND{JFMAMJJASONDJFMAMJJASGNDJFMAMJJASOND
1965 1966 1967 1968 1969

Fig. 6.3. Corrosion of the MSRE fuel circuit in 2*>U and 233U power operation.

.“
some specific event. A significant increase in the
concentration of chromium in fuel salt samples was
noted only during the initial stages of runs Nos. 4, 8,
12, 15, and 19, with an apparent increase partway
through run No. 14. In the beginning periods of run
Nos. 4, 8, 12, 15, and 19, we find that in each instance
a common set of circumstances existed: the reactor
core vessel had been opened previously, either for
maintenance or to exchange test arrays positioned
within the graphite moderator. Although reasonable
measures were adopted to minimize the amount of
airborne contaminants that might be introduced into
the system during these periods, the possibility that
significant amounts of oxidants were introduced into
the open vessel cannot be excluded. During the first
period of power operation, when a significant tempera-
ture differential was imposed on the circuit for the first
time (run 4), some corrosion was anticipated as the Cr°
+ 2UF, = CrF, + 2UF; equilibrium reaction adjusted
to the temperature profile of the circuit. Under these
conditions the increase in the concentration of chro-
mium in the fuel salt could have resulted from the
establishment of the equilibrium reaction and was not
necessarily a signal of the presence of oxidizing con-
taminants.

The increase in the concentration of chromium in the
fuel salt well after run 14 began (see Fig. 6.3) seems to
be inconsistent with the premise that external con-
taminants were the principal cause of corrosion. How-
ever, in the period preceding run 14 part of the graphite
and metal specimens in the core were removed and
replaced. It seems quite possible, therefore, that the
residual concentration of reductant which was gen-
erated within the fuel salt during run 12 was sufficient
to offset the combined oxidizing effects of whatever
contamination was incurred during shutdown and that
characteristic of the fission reaction® only through the
early part of run 14 and that the subsequent rise in
chromium concentration represents the normal com-
pensating shift in the equilibrium corrosion reaction.
Generally, it is assumed that the contaminant most
likely to be responsible for corrosion is moisture.

The inference that moist air was the corrosion-
inducing contaminant calls into question the efficacy of
the flush salt. As an agent for removal of adsorbed
moisture, molten LiF-BeF, flush salt is extremely
effective, as demonstrated in numerous laboratory
experiments, and unquestionably was an effective mois-
ture scavenger. If an oxidizing contaminant or contam-
inants were capable of diffusion within the graphite or
reacting with species deposited in the surface layers of
the graphite, the probability of its removal by brief

75

circulation of flush salt might be slight; instead it might
be released into the salt gradually after the moderator
was heated to high temperatures. Thus, oxygen (per-
haps as CO), rather than water, seems more likely to be
the cause of the observed corrosion. This conclusion is
supported by the fact that the scale found on the nickel
cages which were used to expose Be® to the salt during
run 15 was comprised preponderantly of iron, whereas
on other occasions the principal structural metal in such
scales was chromium. Results of chemical analyses
showed that the prior fuel reprocessing treatment was
effective in reducing the concentration of chromium in
the salt from 133 to 34 ppm and of iron from 174 to
110 ppm. The effectiveness of the reprocessing opera-
tions in reducing Cr®>" preciudes the likelihood that
significant amounts of Fe?" were delivered to the fuel
circuit from the chemical reprocessing plant. The
reduction of Fe? to Fe® by Be? suggests rather that
Fe?* was generated after the beginning of fuel circula-
tion by an oxidizing contaminant contained in the
closed fuel circuit. As increasing amounts of Be® were
added to the salt mixture, the ratio of metallic Fe/Cr
found on the nickel metal cages was reduced until a
normal balance was reestablished and corrosion ceased
(see Tables 6.3 and 6.4).

Laboratory studies of the stability of FeF, have led
to the conclusion that at equilibrium little or no Fe?"
should exist in the MSRE fuel (see Sect. 2.4.7). The
results of these tests confirmed that, of the structural
metal impurities in the melt, iron and nickel persist
almost entirely as metallic species. The results of other
laboratory studies, particularly those concerned with
the reduction of Fe?" by hydrogen,® suggest that even
in concentrations as large as was found by Manning, the
presence of divalent iron was likely to have resulted
from reoxidation after transfer and remeiting of the salt
samples. Subsequent laboratory tests with the MSRE
fuel were precluded by the generation of fission
products in the salt.

Appraisal of the premise that maintenance operations
might possibly permit the ingress of enough oxygen to
account for the analytical results requires the following
considerations. The cumulative amount of oxidation
introduced into the fuel salt (during both 233U and
2330 operations) based on the increases of chromium
at the beginning of run Nos. 8, 12, 15, and 19 and on
the apparent losses of UF; at the end of runs 7, 18, and
19 amounts to 51.31 equivalents, or 410.5 g of 0*", and
corresponds to a cumulative exposure of ~50 ft° of air.
During 233U operations, the amount of oxygen enter-
ing the salt might have been expected to increase the
concentration of oxide by 83 ppm, in excess of the
76

Table 6.3. MSRE fuel salt analyses, run No, 15

Be® added Weight percent % M/z Fe + Cr+ Ni
Sample
(g) Be Te Cr N1 Fe Cr Ni

Carrier, before Z:® addition 0 0380 0 0460 0 0180 372 452 176
Carner, after Zr® addition 00110 0 0034 <0 0010 714 220 65
FP-15-7 10 08

Scale from Ni cage 120 0106 8 03 597 0353 398
FP-15-30 8 34

Scale from Be® rod Z mg 31D (42 8) (11 2) 365 503 132

Scale from N cage zmg @61 493) (21 2) 786 40 174
FP-15-62 9 38

Loose particles 6 06 383 375 144 425 416 159

Scale from top 629 472 326 471 372 257 371

Scale from cage 7 34 274 453 0 096 372 615 13
Salt average, 11/23/68 659 00140 0 0067 0 005 55 26 19

Table 6.4. Relative fractions of Fe? and Cr® reduced
from MSRE fuel salt in run No. 15

Sample  Equivalents of Corrosion e’ /Cr0 on nickel
No Be? added rate (milsf/year) cage?
FP-15-7 224 088 113
FP-15-30 4 09 054 195
FP-15-62 617 035 061

2Average Fe/Cr in carrier salt was 2 24 at the incepiion of run
No 15

sensitivity hmuts for the analytical method and well
above that observed It must be recalled, however, that
early 1n *3°U power operations the concentration of
07", as measured expenimentally, dechned from 120 to
60 ppm, suggesting that under power operation 0® 1s
partially removed from the salt as a volatide species [t
may be concluded, therefore, that the corrosion ob-
served in the MSRE 1s likely to have been caused as
described above but that the mechanism has, as yet, not
been demonstrated unequivocally

The rationale proposed above has several implications
concerning the behavior of the MSRE dunng 233U
operations During the first 16 hr in which fuel salt was
circulated at the beginming of run No 15, the salt did
not transfer to the overflow tank and behaved as
though it contained a neghgibly small bubble fraction
Thereafter, Be® was introduced, and the bubble frac-
tion began to increase, with further exposures of the
salt to Be® the fraction varied erratically ' Certanly
the Be® reduced the surface tension and thereby
allowed easier transport of gas from graphite to salt,
followed probably by oxidation of the metallic iron
impurity, which acts as an oxidant to the circuit walls

Corrosion would continue until the oxidants were
consumed The model of corrosion proposed here has a
relation to the changes in bubble fraction The corro-
sion data suggest that with respect to its physical and
chemical properties, the fuel did not achieve a reference
state until the beginning of run 17 That 1ts bubble
fraction then was greater than observed m 235U
operafions probably was related principally to 1ts lower
density

The probability that atmospheric oxygen was the
primary causative agent of corrosion mn the MSRE was
discounted by Grimes in the following appraisal of the
observed behavior 1!

We find no evidence whatsoever that air was introduced mto
MSRE dunng tts periods of normal high temperature operation
There 1s, on the other hand, no doubt that air i appreciable
guantities was admutted to this reactor system during shut
downs when the reactor circuit was at temperatures of 300°F or
below Only a small fraction of the oxygen admitted at such
low temperatures should have reacted with the reactor metal
and should have been available to cause subsequent corrosion
Flushing of the reactor circuit with hehum during the reactor
heatup should have removed most of the unreacted air
(oxygen) Moreover, use of the flush salt before admission of
the fuel should have removed some (and perhaps a large
fraction) of the reacted oxygen However, some oxidant was
admutted, and some corrosion from this source probably
occurred in MSRE The evidence suggests that corroston from
admitted ar must have been a small fraction of the minor
amount which took place

Since a detailed deserniption of the corrosion picture in MSRE
at all ttmes 1 not yet available, 1t 1s not possible to define the
precise amount of corrosion due to ingress of oxygen It 1s
easily possible, however, to set upper himits upon the amounts
of oxygen that could have been mvolved These amounts are
small If all the observed corrosion (as evidenced by rise in CrF,
concentratton) i MSRE Runs 1 thru 9 s aitributed to imngress

of oxygen the oxide 1on concentration of the fuel should have
nsen by about 11 ppm If corroston (by the same indicator) in
Runs 10 thru 14 were due to oxygen ingress, and if this oxygen
mgress were also responsible for oxadizing all the UF;3 created
by the Be® added dunng those runs, the oxide ton concentra
tron of the fuel should have risen by about 25 ppm Durnng the
entire sequence of runs with 235UF4, therefore, all the
chemically observed corrosion could be due (it seems very
certamn that it was not) to the mgress of oxygen resulting in less
than 40 ppm of added oxide 1on mn the fuel Our systematic
chemical analyses for 0% 1n the MSRE fuel showed no evidence
of such an mcrease MSRE was provided with no mechanism
intended to remove 02_, and we have been unable to postulate

much less to demonsirate — an imadvertent mechanism for
0% removal Our methods for determmnation of QF might
possibly fail to find this <40 ppm increase It seems much more
plausible, however, that the actual increase in 0% was a small
fraction of this figure and that the observed corroston was due
to oxadation of chromium by UF, through mechanisms similar
to those postulated from many years of testing It 1s true that
we have drawn detailed curves of the chromium behavior ot the
mdividual MSRE runs with 233 UF,, and we have been tempted
to speculate about them However the marked scatter in the
chromium data gives such curves httle or no statistical signifi
cance

Rise i the chromium concentrations in MSRE Runs 15 thru
19 (that s in the runs fueled with 233UF,) was more dramatic
and the mitial mcreases were almost certainly more statistrcally
signtficant Several pieces of evidence suggest that the 233UF4
fuel, which by virtue of reuse of the carrier salt was necessarily
less well characterized and probably less pure than was the
235UF,4 wnrtial fuel charge, was relatively oxidizing This may
have been due fo oxygen mgress and oxidation of the MSRE
metal surfaces duning the long fuel change shutdown between
Run 14 and Run 15 If all the corrosion during Runs 15 thiu
termunal Run 19, again as adduced from increases in chromium
concentration of the fuel and allowmng for oxidation of all
reductant added, s attributed to admitted oxygen, the oxude
ion concentration of the fuel should have mcreased by some 45
ppm, this increase, again, seems conceivable but unhkely

In summary, the corrosion picture shows that the quanfity of
oxide 1on which could have entered the fuel during the whole
MSRE operation (involving 20 separate shutdowns) was less
than 100 ppm It seems virtually certamn that the actual amount
which entered was much less than this The ZrF, present i this
fuel would have prevented precipitation of appreciable UQ,
even if much larger quantities had entered, oxide 1on at the
maximum levels suggesied above would not have precipitated
Z10, Subsequent corrosion of the rteactor circuit through
mgress of oxygen would, however, not have been prevented (or
appreciably affectied) by the presence of ZrF, m the MSRE,
this would be equally true for ZiF,4 1n the MSBR fuel

It has been postulated that one of the principal
reasons for the expectedly low corrosion observed is
that the metal surfaces of the fuel circuit have been
covered with a film of the noble-metal fission products
Nb, Mo, Tc¢, and Ru about 10 A thick Results of
electron microprobe analysis of the metal surveillance
spectmens'? removed from the MSRE m May 1967
lend support to this view in that they did not reveal any

77

change in chromium concentration below a depth of 10
M, the lhimut of measurement, as do the results of
electron mucroscopic nvestigation of the specimens,’ ?
which 1ndicated that the surface was covered to a depth
of several thousand angstroms by those metals Al-
though the postulate of a dynamucaily produced and
autoregenerative noble metal film 1s an attractive
rationale of the very low corrosion observed in 225U
operations, 1t fails when the behavior of the MSRE
during the early stages of run 19 1s considered Here,
rapid corrosion was proceeding concurrently with the
generation and presumably the deposition of the same
noble metal fission products which seemed earlier to
have restricted the corrosion of the fuel circuit walls
Another hypothesis has been suggested recently by
Grimes' > as an alternate rationale of data that indicate
the almost complete absence of corrosion dunng
extended periods of operation, while at other times
rates were more rapid than could be accounted for by
diffusion-controlled mechanmisms It presupposes that
the presence of UF; in the fuel salt does not counter
the corrosion of Hastelloy N effectively once the
concentration of Cr®" in the salt has risen to a level of
~70 to 80 ppm Rather, at concentrations of this order
and n the presence of the intense radioactive flux m
the reactor core, chromium carbide, Cr3C,, 1s formed
at the surfaces of the graphite moderator The forma-
tion of such a phase thus causes the graphite moderator
to act as a sink for chromium and promotes corrosion
of the container alloy at the most rapid rate allowed by
the diffusion coefficient of chrommum If such a
mechanism operates, the difference in the chromium
concentration as observed in the fuel salt samples and
the maximum possible concentration represents the
quantity of chromium which was deposited at the
surface of the moderator graphite On occasions when
the oxidation potential of the salt increased substan-
tially, this carbide coating would then decompose and
release large amounts of Cr?" to the salt, thus causing
rapid changes in the concentration of chromium in the
salt samples submutted for chemical analysis to suggest
that the corrosion rate was mncreasing rapidly This
hypothesis 1s attractive for a variety of reasons (1)1t
implies that throughout the operation of the MSRE the
corrosion rate was invariant and diffusion limited, (2)
the [UF3]/]UF,] remains so low 1n such systems at all
times that attempts to repress corrosion in MSBR fuels
by external adjustment of the [UF;]}/[UF4] are
needless, as well as the previously foreseen need to
develop in-line analytical methods for 1ts determination,
and (3) the hypothesis provides a rationale for the
otherwise mnexplicably low rates of corrosion which
seemed to characterize operations of the MSRE with
2350 fuel.

It seems unlikely that significant quantities of chro-
mium carbides will have formed on the moderator
graphite surfaces for the following reasons:

1. In run No. 15, the principal constituent of the slag
found on the nickel cages used for the introduction
of beryllium was iron (see Sect. 6.7). Presumably,
this iron deposit was formed by reduction of Fe?*
from the melt and represents a condition in the melt
which is so oxidizing that, if present, the relatively
unstable compound Cr3C: (AG, 5p0°k = —23 keal)
should have decomposed completely during this
period. The possible amount of chromium which
should have been released to the fuel salt by such
decomposition, as calculated using diffusion coeffi-
cients from 1 X 107'% to 4 X 107** cm?/sec,
would increase the concentration of chromium in
the fuel salt to values of 240 to 480 ppm (see Table
6.2), that is, to much higher concentrations than
were observed at any time during MSRE operations.

2. The coincidence of apparently stepwise increases in
Cr followed maintenance periods (as noted previ-
ously).

3. The carbides of niobium are more stable than is
Cr3C,; in 223U operations the oxidation potential
of the salt increased gradually during runs by fission
to the point that niobium entered the salt (see Sect.
6.5) and was subsequenily reduced and removed
from the salt by Be addition; under these conditions,
Cr3C, should have become reoxidized, entered the
salt, and increased the chromium concentration of
the samples sharply, where, in fact, no perturbation
of the average concentration of chromium was
noted.

In postoperational examinations with one of the
stringer bars from the reactor core, a search was made
to ascertain whether or not a coating of chromium
carbides adhered to the surface. No evidence of such a
coating was found.

A surprising result of the postoperational tests per-
formed with specimens of the MSRE control rod
thimbles was the observation of anomalous intergran-
ular penetration of the alloy which had been exposed to
the fuel salt. Occasionally, inspection of Hastelloy N
specimens removed from the reactor core showed grain
boundary cracks at the surfaces of specimens subjected
to tensile tests. These cracks appeared in material
strained at room temperature, which is not normal for
Hastelloy N, but they penetrated only a few mils and

78

had no detectable effect of the mechanical properties of
the specimens. The cracks in the control rod thimbles,
however, were deeper than previously observed, at some
locations as deep as 7 mils,’* some ten times the depth
of the generalized attack indicated by changes in the
concentration of chromium in the fuel salt. Curiously,
several fission products, preponderantly tellurium, had
penetrated to depths comparable with those of the
cracks.

It is not possible to ascribe this attack to radiation
effects per se, for in-pile tests with salt-graphite-alloy
assemblages prior to tests with the MSRE showed no
corresponding effects. These unpredicted results have
caused a renewal of the investigation of materials
removed from the MSRE, with the initial objectives to
determine if fission products are indeed the cause of
grain boundary separation.

While the initial results seem to implicate tellurium as
a causative agent, its role is not positively identified.
One mechanism, proposed before tellurium was be-
lieved to have deposited preferentially, relates the
attack to impurities in the following way. Electron
microscopic examinations of Hastelloy N show typi-
cally that an unidentified group of complex phases
comprised of Mo, Cr, C, N, and B, morphologically
similar to face-centered M,C carbides, are deposited
preferentially at intergranular surfaces. It is also ob-
served that the loss of ductility that occurs with
Hastelloy N after irradiation correlates with more
pronounced intergranular fracture. Such behavior very
possibly means that the chemical activity of the
carbide-like phases deposited in these locations is
increased. If so, it would not be surprising to find that
these phases were highly susceptible to oxidative
corrosion. It was noted previously that two unusually
oxidative regimes occurred in the MSRE, at the
beginning of run 15 and again at the beginning of run
19. In an earlier assessment of corrosion of the MSRE
fuel circuit we have deduced that the most probable
oxidant at those periods was atmospheric oxygen.'® We
reach the tentative conclusion, therefore, that the
intergranular attack noted was a result of attack by
oxygen on the carbide-like phases deposited at inter-
granular surfaces. If oxygen were released from the
moderator, as suggested previously, the first metal
which it might contact would be the control-rod
thimbles. It seems likely, therefore, that the inter-
granular penetration observed in the postoperational
tests resulted from the development of unusually
oxidative conditions in the MSRE and is atypical of
normal reactor behavior. The overall results of corro-
sion surveillance in the MSRE, while not totally

unequivocal, appear to indicate that corrosion behavior
in molten-salt reactors can properly be anticipated to be
negligible over the planned use period of these reactors.
Furthermore, the probability of recurrence of the
anomalous penetration in future reactors can be
assessed in tests with the modified alloys which are
currently under development and which will contain a
group of carbides perhaps unlike those found in the
MSRE alloy.

6.4 Additions of Reductants and
Oxidants to the Fuel Salt

As discussed earlier in this report (Sect. 6.3) the fuel
salt, free of moisture and HF, should remove chromium
from Hastelloy N only by the equilibrium reaction

,C1° + UF4(d) = %, CrF,(d) + UF5(d) .

When the above corrosion equilibrium was first estab-
lished in MSRE power operations, the UF; produced in
this reaction, together with that originally added to the
fuel concentrate, should have totaled 1500 g, with the
result that as much as 0.65% of the uranium of the
system could have been trivalent soon after the begin-
ning of power operation. The UF; content of the
MSRE fuel was determined after approximately 11,000
MWhr of operation to be no greater than 0.05%. The
fuel salt was considered to be far more oxidizing than
was necessary and certain to become more so as
additional power was produced unless adjustment was
made in the UF; concentration. A program was
initiated early in 1967 to reduce 1 to 1.5% of the
uranium inventory to the trivalent state.

On the basis that it met chemical criteria and that its
introduction into fuel salt could be accomplished
conveniently, beryllium metal was selected as the most
suitable substance to use for in-situ reduction of U(IV)
in the fuel salt. Results of laboratory tests had shown
that the metal would be sufficiently reductive to
require exposure for conveniently short periods of time
but was not so reductive as to cause concern that it
might reduce U(IV) to the metallic state. The pure
material was available in forms that were easily accom-
modated for use with the sampler-enricher device, and
the metal afforded a maximum of reductive equivalents
per unit mass.

Initially, 4 g of beryllium was introduced into the salt
by melting a mixture of "LiF-BeF, carrier salt and
powdered beryllium in the MSRE pump bowl sampler
cage. Subsequently, three additions were made by
suspending specimens of ¥-in.-diam beryllium rods in

79

the salt in the pump bowl. The capsules used for adding
beryllium were similar in size to those used for sampling
for oxide analysis but were penetrated with numerous
holes to permit reasonable flow of fuel salt. The
beryllium rods were found to react with fuel salt at a
steady rate, dissolving at approximately 1.5 g/hr. A
summary of all the additions of reductants and oxidants
introduced into the fuel salt system via the sampler-
enricher apparatus is given in Table 6.5.

With 235U fuel salt, all dissolutions of beryllium
metal into the fuel salt proceeded smoothly; the bar
stock which was withdrawn after exposure to the fuel
was observed to be smooth and of symmetrically
reduced shape.

No significant effects on reactivity were observed
during or following the beryllium additions, nor did the
results of chemical analyses indicate that the actual
concentration of U(HIT) had increased until after run 12
had begun (see Table 6.6). The additions preceding run
12 had increased the U3* concentration in the total
uranium by approximately 0.6%. Four exposures of
beryllium were made at close intervals during the early
part of that run. Samples taken shortly after the last of
these four exposures (FP-12-16 et seq., Table 3.1)
began to show an unprecedented increase in the
concentration of chromium in the specimens, followed
by a similar decrease during the subsequent sampling
period.

Previous laboratory experience has not disclosed
comparable behavior, and no well-defined mechanism
was available at the time to account satisfactorily for
the observed behavior. Speculation as to the cause,
partially supported by experimental data, included the
following consideration.

On the two occasions when the most rapid rates of
dissolution of the beryllium rods were observed, chro-
mium values for the next several fuel samples, FP-11-10
et seq. and FP-12-16 et seq., rose temporarily above the
lo level and subsequently returned to normal. That the
increase in chromium levels in samples FP-12-16 to -19
was temporary indicates that the high chromium
concentration of fuel samples removed from the pump
bowl was atypical of the salt in the fuel circuit and
implies that surface-active solids were in suspension at
the salt-gas interfaces in the pump bowl.

That atypical distribution of species in this location
does indeed take place was demonstrated earlier by the
analysis of sample capsule support wires that were (1)
submerged below the pump-bowl salt surface, (2)
exposed to the salt-gas interface, and (3) exposed to the
pump-bowl cover gas. The results showed that the
noble-metal fission products, Mo, Nb, and Ru, were
80

Table 6.5. Summary of adjustments of [U°']/[=U] in the MSRE fuel salt

g . Reductant or oxidant Equivalents of [U3+/2U]
Date ample added reductant (%),
Number .
Form Weight (g) added nominal

2/13/66 Runs 5-1 041
1/1/67 FP-10-14 Be powder 3.0 0.67 0.22
1/3/67 FP-10-16 Be powder 1.0 0.89 0.25
1/4/67 FP-10-18 Be rod 1.63 1.25 0.29
1/13/67 FP-10-23 Be rod 10.65 3.61 0.51
2/15/67 FP-11-10 Be rod 11.66 6.20 0.74
4/10/67 FP-11-40 Be rod 8.40 8.06 0.79
6/21/67 FP-12-8 Be rod 7.93 9.82 0.87
6/23/67 FP-12-9 Be rod 9.84 12.01 1.12
713/67 FP-12-13 Be rod 8.33 13.86 1.30
7/6/67 FP-12-15 Be rod 11.68 16.45 1.59
8/3/67 FP-12-56 Be rod 9.71 18.60 1.72
9/15/68 FP-15-7 Be rod 10.08 2.24 0
10/13/68 FP-15-30 Be rod 8.34 4.09 0
11/15/68 FP-15-62 Be red 9.38 6.17 0
11/20/68 FP-15-66 Be washer 1.00 6.39 0
1/22/16 FP-17-8 Be rod 8.57 8.29 1.19
1/30/69 FP-17-11 Crrod 4.13 8.48 1.19
4/15/69 FP-18-3 Zr rod 20.24 9.37 0.65
4/26/69 FP-18-7 Zr rod 24.04 1042 1.19
5/8/69 FP-18-17 FeF, powder 30.00 9.78 0.52
5/15/69 FP-18-23 Be rod 5.68 11.04 1.20
5/20/69 FP-18-28 Be rod 3.17 11.74 1.59
9/12/69 FP-19-25,26 Zr foil 0.62 11.77 0
9/24/69 FP-19-31-4 Zr foil 1.20 11.82 0
10/2/69 FP-19-40 Be rod 2.87 12.46 0.99
10/8/69 FP-19-48 Be rod 4.91 13.55 1.61
10/21/69 FP-19-51 Nb foil 0.018 13.55 1.37
11/29/69 FP-20-7 Be rod 6.974 15.10 1.16
12/9/69 FP-20-22 Be rod 9.894 17.30 2.49
12/9/69 FP-20-22 Be rod 3.019 17.97 2.95

deposited in abnormally high concentrations at the
salt-gas interface. Such behavior suggests that the high
chromium concentrations in the fuel specimens were
caused by the occurrence of chromium in the pump
bowl in nonwetted, surface-active phases in which its
activity was low. A possible mechanism which would
cause such a phenomenon is the reduction of Cr** by
Be® with the concurrent reaction of Cr® with graphite
present on the salt surface to form one or more of the
chromium carbides, for example, CrzC, (AH Of =21
kecal at 298°K). Such phases possess relatively low
stability and could be expected to decompose, once
dispersed in the fuel-circuit salt.

The possibility that surface-active solids were formed
as a consequence of the Be® additions was tested late in
run 12 by obtaining salt specimens at the salt-gas
interface as well as below the surface. First, specimens
were obtained in a three-compartment sample capsule
that was immersed so that the center hole was expected

to be at the interface (Fig. 6.4). Next, a beryllium metal
rod was exposed to the fuel salt for 8 hr with the result
that 9.71 g of beryllium metal was introduced into the
fuel salt. Twelve hours later a second three-compart-
ment capsule was immersed in the pump bowl. Its
appearance after removal from the pump bowl is shown
in Fig. 6.5. Chemical analyses of the fuel-salt specimens
FP-12-55 and -57 did not show significant differences in
chromium; however, the salt-gas interface in FP-12-57
was blackened as compared with FP-12-55 (Fig. 6.6).

An additional purpose of sampling with the three-
compartment capsule was to determine whether foam-
like material was present in the sampler area and would
be collected in the upper compartment. Globules were
noted on the upper part of FP-12-57 (Fig. 6.5),
indicating that conditions in the pump bowl were
substantially different after beryllium was added to the
fuel salt.

.‘
Table 6.6. Concentration of UF; in the MSRE fuel salt?

Uranium Uranium - o Net Equivalents . 3+
Date SaI:Inp te Megawatt-Hours  Consumed  Consumed Nethqxf(xlvaI.cnts ggctlaldBe of Reductant Ne; gqgwatlaentts v @ -
o (ke {moles) of Oxidation ed (g Added Ob Keauctan Calculated  Analytical
0 0 3.13 0.33
11/14/66  FP9-14 12,345 0.632 2.67 2.14 0 3.13 0.99 0.10 0.10
1/1/67 FP10-14 14,950 0.766 3.23 2.58 3 3.80 1.22 0.13
1/3/67 IP10-16 15,050 0.771 3.25 2.60 6 4.46 1.80 0.19
1/4/67 [P10-18 17,100 0.877 3.70 2.96 7.63 4.82 1.86 0.20
1/13/67 FP10-23 17,852 0.915 3.86 3.08 18.28 7.19 4,11 0.43
1/15/67 FP10-25 18,050 0.924 3.90 3.10 18.28 7.19 4.09 0.43 0.66
2/6/67 FP11-5 19,712 1.010 4.26 3.40 18.28 719 3.79 0.39 0.60
2/15/67 FP11-10 21,272 1.090 4.60 3.68 29.94 9.77 6.09 0.64
2/22/67 P11-13 22,649 1.161 4.90 3.90 29.94 9.71 5.87 0.62 0.69
3/28/67 FP11-32 28,342 1.453 6.13 4.90 29.94 9.77 4.87 0.51 0.45
4/10/67 FP11-40 30,900 1.584 6.68 5.34 38.34 11.64 6.30 0.66
6/21/67 FP12-6 36,055 1.663 7.01 561 38.34 11.64 6.03 0.64 0.71
6/21/67 P12-8 36,055 1.663 7.01 561 46.27 13.40 7.79 0.82
6/23/67 FP129 36,416 1.866 7.87 6.30 56.11 15.58 9.28 0.98
6/29/67 FP12-11 37,400 1.932 8.15 6.52 56.11 15.58 9.06 0.95 1.30
7/3/67 FP12-13 37,856 1.940 8.19 6.55 64.24 17.38 10.83 1.14
7/13/67 FP12-15 38,345 1.966 8.30 6.64 76.12 20.02 13.38 1.40
7/13/67 FP12-21 39,500 2.023 8.54 6.83 76.12 20.02 13.19 1.39 1.0
8/3/67 I'P12-56 43,872 2.248 9.49 7.59 85.83 22.18 14.59 1.54
9/15/67 I'P13-5 44,781 2.314 - 9.76 7.81 85.83 22.18 12.42 1.31 1.60
3/26/68 FP14-(I) 72,454 3.743 15.79 12.63 85.83 22.18 9.55 1.01
9/15/63 FP15-7 0 0 10.08 2.23 0 0
10/13/68  ['P15-30 0 0 18.42 4.09 0 0
11/15/68  [P15-62 0 0 27.80 6.17 0 0
11/20/68  FP15-66 0 0 28.80 6.39 0.22 0.15
1/22/69 FP17-8 850 0.044 0.19 0.1 37.37 8.29 1.97 1.31

?These numbers assume that the 235U salt originally was 0.16% rcduced; that the increase in Cr before initial 235y power operations was real, occurred before 11/14/66,

and resulted in reduction of U to U3+; that each fission results in oxidation of 0.8 atom of U3+; and that there have been no other losses of U3+.

18
82

% of

ey

PHOTO 1865-71

PHOTO 1864-71

Wi

R O N Pt
St g o el L U
o bre

[ W

2
©
-

SEE

Pl Bje @

P RCCTR S SN

4 E i
s

w;
o 18
S &
e S T

w
& s
lete k& m‘ya

©rmren e
fvfifiwMM

g Sk,

b
.
I A N

SR T
B P g il
L A ek

L, e
m

Ty
@yw?;‘%‘
S

o

R o

R )

i

s
NS

&

A
A
A

et

37

QIRS
NG

5,

Y

5%

o5

-

.,

Fol7EE

s

2

=
o

o

g3
=
oy
- o
ns
Q
Bt
w
3 o
@
@ &
ER:
a,
[
Q
—
-+
2

.

1c

Fig. 6.5. Three-compartmeni n
MSRE pump bowl after exposure of beryl

salt. August 3, 1967, FP-

Each of the three compartments

12-57.

the capsules shown in F
outside by two holes as shown here.

5 was open to the
The capsules were

4 and 6.

6

igs.

n

imn

kel capsule suspended

) {4

-compartiment n

Fig. 6.4. Three
MSRE pump bowl before exposure of berylt

positioned approximately so that the center compartment was

expected to be at the salt-gas in

to 235U fuel

um

active

terface to collect surface-

12-55.

salt. August 2, 1967, FP

foam 1f 1t were present. No appreciable residue was found 1n the

solids. The upper compartment was to serve as a collector of
upper compartment.
%@@‘&%@m@%flb P &t

ST, R-39267
8767

= =

-

iog

A

S
et

Bl oo O o
(& i g%fmfisg}&w o e Fet e

5 8

Sk
e fimé;;ar’w -

R-39266

(£)

Fig. 6.6. Surface appearance of fuel salt before and after
beryllium exposure to 2350 fuel sait. (@) FP-12-55, (b)
FP-12-57. Salt analyses did not reflect significant differences in
chromium concentration nor were they indicative of the
identity of the material at the blackened surface.

83

Examination of the metal basket that contained the
beryllium rod while it was exposed to the fuel showed
the presence of dendritic crystals along with a small
amount of salt residue (Fig. 6.7). Spectrochemical
analysis of material removed from the basket (Fig. 6.7)
indicated that the material contained 7.8 wt % chro-
mium and less than 10 ppm of iron and nickel.

The evidence obtained did not permit inference as to
the identity of the phases which formed within the
pump bowl as a consequence of the beryllium addi-
tions. It strongly implied that nonwetted flotsam can be
formed and accumulated temporarily in the MSRE
pump bowl.

During this period, fuel salt accumulated in the
overflow tank steadily during operation and remained
there in relative isolation from the fuel stream. At
intervals of about one day, part of the salt (60 Ib) was
returned to the fuel stream. Recognizing that chromium
might be injected into the pump bowl as the salt
returned, we performed an experiment in which salt
samples were obtained from the pump bowl within an
hour after fuel was returned from the overflow tank to
the pump bowl. The purpose of the experiment was to
determine whether material from the overflow tank
contributed appreciably to the perturbations in the
chromium concentration. The results were negative,
possibly, in part, because sampling and salt-transfer
operations were not performed concurrently for safety
reasons.

In view of the mechanism developed recently'® to
account for the transport behavior of the noble metal
fission products in the MSRE, we can infer that the
exposures of beryllium which generated high concen-
trations of chromium in the salt samples simply reduced
the Cr*" in the flowing fuel salt to the metal, which in
turn became a part of the particulate pool of suspended
metals at the surface of the salt in the pump bowl. It
remained there temporarily, gradually oxidizing to the
fluoride as it reacted with the fuel salt.

In their examinations of the behavior of the noble
metal fission products, Kirslis and Blankenship! 7 did
not observe a significant effect of fuel reduction on the
noble-metal concentrations in the fuel. The concentra-
tions frequently rose rather than fell, as expected, after
adding beryllium, They noted that the °°Mo, '°3Ru,
106 Ru, and '32Te showed parallel rises and falls. The
®°Mo results were extraordinarily high for samples
FP-11-8 and FP-11-12. The high values were checked by
reruns on fresh samples. If all the ?°Mo produced by
fission remained uniformly distributed in the fuel, the
calculated concentration would be 1.4 X 10'! dis
84

Fig. 6.7. Residue from berylium addition capsule FP-12-56, August 3, 1967. Capsule was exposed to fuel salt for 8 hr, 9 71 g of
Be® dissolved Residue contaned 7 8 wt % Cr and less than a total of 10 ppm Fe and Na .
min~! g7'. A calculation showed that if all the *®Mo
produced by neutron activation of the *®Mo in the first
0.1-mm thickness of the Hastelloy N reactor contain-
ment vessel diffused instantaneously into the fuel melt,
the increase in ?°Mo concentration would be only
about 10° dis/min per gram of fuel. It thus appeared
that a mechanism was operating which either concen-
trated fission-produced Mo and other noble metals in
the pump bowl or resulted in large temporal and spatial
variations in their concentrations. Dissolved ?°Mo
would undoubtedly be uniformly distributed. If the
noble metals circulated as a suspension of insoluble
metal particles, it was reasoned that they might
concentrate in the pump bowl or vary in concentration
with pump bowl level, cover gas pressure, and other
operating variables. Since clean metal surfaces are not
wetted by the fuel salt, there might also be a tendency
for metal particles to collect around helium bubbles,
which are probably most numerous in the pump bowl.

The °°Nb concentrations varied erratically and did
not parallel the behavior of the other noble metals. This
was ascribed to analytical difficulties. An unavoidable
difficulty was that a large correction for ?3Zr decay
must be made for each salt analysis.

From the beginning of operations of the MSRE with
2337 fuel, the behavior of the fuel salt during and after
exposure to the beryllium rods was markedly different
than previously observed. A detailed description of the
changes in fluid salt and gas behavior during these
periods is described in a separate report.’ ® Introduction
of beryllium metal into the 232U fuel salt apparently
caused a major perturbation of salt-gas interactions and
resulted in an increase in void fraction of the fuel salt.
The exact reasons for development of the greater void
fraction that persisted throughout ?22U operations
have not been completely determined. In addition to
the purely chemical effects which resulted from the
addition of beryllium and other reductants, other
factors such as pump speed, variation in the solubility
of inert gases as a function of hydrostatic pressure,
difference in density of the 233U and #33U fuel salt,
and changes in interfacial tension were recognized to be
related. Interaction of beryllium with the fuel salt
during run 15 caused a major change in only one
physical property of the fuel — its surface tension. This
is the most puzzling aspect of the effect of beryllium on
the fuel salt in run 15 since the effect was unprece-
dented in the fuel system. However, slight changes in
interfacial energies in fluid systems are known to
account for marked changes in hydraulic behavior.
Since the addition of beryllium in the early part of the
233U operations effected both chemical and physical

85

changes concurrently, the exact way in which these
changes altered the void fraction of the salt cannot,
therefore, be evaluated exactly, although some aspects
of the observed phenomenon can be attributed tenta-
tively to the additions.

If, as inferred in Sect. 6.3, atmospheric oxygen was
the causative agent for the increased oxidation potential
of the salt, release of the gas from the moderator
graphite might well have been triggered by a reduction
in the salt-gas interfacial tension, resulting in an increase
in the bubble fraction of the salt.

Photographs of the cage assemblies removed from the
fuel pump bowl after exposure to the fuel salt are
shown in Figs. 6.8—6.23. While none of these cages
appeared to have been wetted by salt during 235U
operations, each of those removed after treatment of
the 233U salt showed evidence that salt had wetted the
nickel cage and had adhered to it. A typical example is
shown in Fig. 6.17. The presence of bubbles is
suggested by the appearance of the upper part of
capsule FP-17-8 (Fig. 6.17), on which structures of
collapsed bubbles seem to be visible.

Of particular significance to the interpretation of the
behavior of the fuel salt from August 1968 to January
1969 are the changes in the relative amounts of iron
and chromium contained in the deposits on the nickel
cages that were used to suspend beryllium into the
pump bowl (Table 6.5). From this and other chemical
evidence we deduced that corrosion in the fuel circuit
was not checked until early in 1969. By that time, the
bubble fraction in excess of that observed in 23%U
operations was reduced to a fraction that could in
general be accounted for by the other factors men-
tioned above. It may be speculated that the coincidence
of the disappearance of reactivity blips with reestab-
lishment of the expected ratio of structural metal
fractions in the residues adhering to the nickel cages
was indicative of the removal of the last fraction of
anomalous excess of bubbles in the fuel salt.

Among the conjectures advanced to account for the
excessive rates of salt transfer during this period was
one based on the recognition that sharp gradients in the
density profile of the fuel salt in the pump bowl
existed; from this it was suggested that voluminous
amounts of foam were developed in the pump bowl.
There was no evidence that such foam had existed in
233U operations, but rather that a salt mist was present
in the pump bowl. Indication that such a mist existed
was obtained by suspension of a cage-rod assembly of
nickel into the pump bowl during run 14. Photographs
of the assembly after removal from the pump bowl
(Fig. 6.10) show that the surfaces of the rod exposed to
86

. 235
-Fig. 6.8. Dendritic crystals and salt on basket of nickel capsule used to expose beryllium to U fuel salt. August 3, 1967,
FP-12-56.. '
gt z‘“&?

% ):\?“(m

B
e

=

iy
2

M
-

i b
& -

SRRy
a

Y
k3

s
u’w;@f
o

———
v ¢
o
g
#

L A g

F

ri

-
ke

-

]

Fig. 6.9. Appearance of nickei basket and rod after suspen-
sion in MSRE fuel salt for 2 hy. FP-14-55.

the vapor just above the salt were coated with con-
densed droplets. Although the vapor space in the
sample shroud was probably more quiescent than that
in the pump, it was reasoned that foam should also
appear here if it were prevalent in the pump bowl.

87

Attempts to produce foam were performed in the
laboratory'® but were unsuccessful. Tests for foam
production consisted of measurements of the fuel salt
level in a 10-in.-diam nickel vessel at intervals during the
reduction of essentially all of the uranium to its
trivalent state and after stepwise oxidation of the nickel
with nickel fluoride. Melt levels were determined by the
abrupt change in electrical resistance noted as the
insulated nickel rods made contact with the salt
mixture., Without exception, the tests showed no
evidence of the developmeni of a foam either by
increasing or decreasing the oxidation potential of the
salt.

In an attempt to identify other possible agents which
might cause development of foams in fluoride melts,
Kohn and Blankenship'?® tested the effect of each of
the additives: carbon dust, graphite powder, beryllium
metal, finely divided nickel metal, and pump oil vapor.
They were unable to promote the development of
stable foams with any of these additives in clean melts.
Foams were formed only by introducing enough water
into the melt, either in the sweep gas or by adding solid
hydrates, to give a definite cloudiness. Even so, the
foam would collapse very quickly after the purge gas
stream was removed.

The phenomena observed in run 15 and later remain
as puzzling because the opportunity to perform realistic
tesis to resolve unanswered questions was lost with the
termination of operations of the reactor. It appears that
development of a void fraction of unprecedented
magnitude took place during a period of rapid corrosion
of the containment system and that contaminants
which entered the core of the reactor while it was
exposed to the atmosphere during maintenance were
the source of the corrosion. These contaminants mark-
edly affected the interfacial energies of salt-metal,
salt-gas, and salt-graphite surfaces but did not generate
foam in the fuel circuit. The conditions which allowed
these events to occur are avoidable and are atypical of
the operating procedures that are projected for future
MSR’s.

Additional investigation of the physical properties of
molten salts will be necessary to establish quantitatively
the relationships of interfacial energies in salt systems
to the retention of and stability of bubbles in flowing
streams. The initial stages of these investigations can
properly be carried out in the laboratory; however, the
results may prove to be wmrelevant unless the effect of
highly radioactive fluxes on the properties is shown to
be inconsequential.
88

PHOTO 1867~ 71

Fig. 6.10. Appearance of nickel rod from FP-14-55 after Fig. 6.11. Nickel cage after second exposure of beryliium to
suspenston in MSRE pump bowl for 2 hr. (¢) Upper end, 233U fuel salt. October 13, 1968, FP-15-30, § 34 g Bef
(b) lower end dissolved from Be rod
=

i

Fig. 6.12. Copper capsule containing several short magnets
before exposure to 233U fuel salt.

89

Fig. 6.13. Metallic particles on copper capsule used to expose
a magnet to 233U fuel salt. November 15, 1968, FP-15-61.
Dendrites on capsule were composed of fine particles of iron.
90

e
L

7%
< M
t8
g —~
=
= O
59
2L
~ oy
~ £ a
} g nchF
o) S )
= : 2%
00} 9 RN
o -8
O . =R
k- - TR
o -9 S 5
T = 223
o o mfl
= lvm
2, 2.3
~ & & 8
o -]
o —
) - 5
= o 2 L
o0 &% o
= e 2 0O
-
£ o3
Gy O=E
< .um
Wy
2 B
[ 4] 6SM
v
my . Q=
Q. 8 = o
= M= mgs
= 8 .=
= 3 E=
el
4”a
¥ E
8 B
-t
o0 g
=3
S

w8
e,

b & B -
w # TS E

5 o% 8 )
o L4 5

S
-t
PR

i ¢

5y

i
Houg

o oy, g it
M)
AP CRT
A

i mfl.
o

PHOTO 1872—71

L

B

| el ot
o e wmw
. =
&2
e
St B

m«mmfi
el B B
E TR Ny ety @

fihwrmfm%& e

o
o

91

9.38¢ Be® dissolved from

+
E

flium exposure in 233y fuel salt. November 15, 1968, FP-15-62

d bery

1F

kel cage from th

ic

Ni

. 6.16.

g

F

rod
92

. PHOTO 94278

Fig. 6.17. Nickel cage after fourth exposure of berylhum to
233U fuel salt. November 15, 1968, FP-17-8, 8 57 g of Be?
dissolved from rod

PHOTO 94273

Fig. 6.18. Top of nickel cage from fourth beryllium exposure
in 233U fuel salt January 22 1969 FP-17-8, 8 57 g Be® dis
solved from rod After leach treatment with Verbocit and nitric
acid solutions

93

1 lPHOTO 1873-711

Fig. 6.19. Chromium rod exposed to 2337 fuel sait.
January 30, 1969, FP-17-11.

e
£y
3

5™
g

PHOTO 94272

£

Fig. 6.20. Cross section of chromium rod showing
surface deposit. January 30, 1969, FP-17-11.

&)

'

3

YE T v AR
] piSE
AR e

3,

ki

G s
: o

g

£

94

§ PHOTO 1874 —71

o

F{%

v
£

o

5

o
Of e
kS %r Y

\
i
s T

¢

XA
5
EL

L

BRI
e

5
£

T
i ”i'g
<

Y
=
=3
-,

o

- iy

s
e ¥,
=)
Tl 5%
¥ B b

%
e
3

5 AT
e

Bye
B
i)
pg®

%
£h

2

=
)

&

s

k)

o

i

T

5
o

)
i L
e

P

o

V.M
e
L o
o
i B0
e
5
§W
5
i,
i
G
=%
B
g
R

F el " 5 4 s
5 T e ol s
T e e o

e
£

Fig. 6.21. Nickel cage after exposure of beryllium rod to
2337 fuel salt. October 8, 1969, FP-19-48; 4.91 g BeY dissolved
from rod.

6.5 Effect of Uranium Trifluoride on the
25 Nb Concentration of the Fuel Salt

Minor adjustments in the concentration of uranium
trifluoride in the MSRE fuel salt were made occasion-
ally during the period when the MSRE was operated
with 233U fuel. Their primary purpose was to offset
the oxidizing effects anticipated to result from the
fission reaction. Within this period the [U*"]/[ZU]
concentration ratio was estimated to have varied within
the range 0.1 to ~1.7% (Fig. 6.24). No evidence was
found that indicated that such variation effected
significant changes in either corrosion rate or fission
product behavior in the fuel salt within the reactor.
This derives from the fact that the ?35U fuel was a
highly buffered system in comparison with the 233U
fuel used later; that is, the total amount of uranium in
the 225U fuel exceeded that contained in the 233U fuel
by sixfold. In contrast, operation of the MSRE with

PHQOTO 4875 — 74

L
3§ f TR
P ol e By
SRS

)

-

5, ,%5)

‘273-3

ANCE

Fig. 6.22. Nickel cage from exposure of beryilium rod to
233y fuel salt. November 29, 1969, FP-20-7: 6.97 g Bed
dissolved from rod.

233U fuel showed pronounced changes 1n the corrosion
rates and fission product chemistry as the concentration
of UF; was altered. Of considerable interest was the
appearance of ?*Nb in the fuel salt, noted for the first
time in initial operations with 233U fuel.?® Ths
observation signaled the potential application of the
disposition of ?*Nb as an in-line redox indicator for
molten-salt reactors.

Operation of the MSRE with 33U fuel thus gave
evidence that pronounced changes in fission product
and corrosion chemistry resulted as variations of the
concentration of UF; in the fuel salt were made. The
233U fuel was experimentally more tractable for study
than the previous charge of 235:238J fuel because of
the much lower uranium inventory carried in the 223U
fuel. A serious disadvantage was realized, however,
when it was discovered that the total amounts of UF;
which were obtainable in samples of the 233U fuel
caused the previously satisfactory method?! used for

e
e
s Sl 6l P BT
2

i

g
E:3
Sty s, T o
S SRS R e .
hmmwmnfimfifiw
AR

E
=

L

FP-20-22; 9.87 g Be" dissolved.

?

95

PHOTO 1876— 74

s

y‘syfi;@
ol 8™

Lo n
A
N

%

k3

3 e

R 8E T F e

w

1969

b

Fig, 6.23. Nickel cage from exposure of beryllium rod to 233y fyel salt. December 9
96

ORNL-DWG 7i- 9991R

1.8
1.6 o
N
o ANALYTICAL RESULTS N
(Hy- HF TRANSPIRATION METHOD)
1.4
) \

N \\
3 N
~ 1.0 o

(u3h/izm

0.4 f\ i

0.2 \\ ~

--------- = 5% %150 222U moles
o |
LR PP 4% bt b4
FREEZE VALVE SAMPLES |
0 f 2 3 4 5 6 7 8  (x10%)

full power hours

Fig. 6.24. [U3+]/[EU] in the MSRE fuel salt, runs 5—14, Assumed maximum power, 7.4 MW(t).

determination of U3 in the fuel salt to be of little value
and required the development of new methods of
analysis.

Adjustments of the uranium trifluoride concentration
of the MSRE fuel salt were made frequently during
2337 operations in order to study the inhibition of
corrosion by control of the [U**]/[ZU] concentration
ratio and to evaluate the possible application of °5Nb
disposition as a redox indicator. The experimental
results obtained during this period of operations pro-
vided concrete evidence that values of the equilibrium
constant for the reaction Cr° + 2UF, = CrF, + 2UF,,
as assigned from standard free energy and activity data,
are reasonably accurate. Together with niobium distri-
bution data they show that after the fuel system had
been opened for maintenance the fuel salt subsequently
appeared to become oxidizing with respect to the
MSRE containment circuit even though it did not seem
to have caused corrosion in the drain tanks. Compari-
sons of the relative fraction of the ?>Nb inventory

which appeared in the fuel salt as the redox potential of
the salt changed indicated that when mildly reducing
conditions were imposed, as for steady-state operation
of the MSRE, niobium very likely became involved in
reaction with the moderator graphite to form niobium
carbide.

During August and September 1968 the 233U fuel
charge was constituted from "LiF-2**UF, and "LiF-
BeF, -ZrF, carrier salt which had previously contained
235,2381F,. Concurrent with the inception of corro-
sion in the fuel circuit, as evidenced by a rapid increase
in the concentration of Cr in the fuel salt (Fig. 6.25),
niobium began to appear in the fuel salt for the first
time?! and persisted there until 28.80 g (6.54 equiva-
lents) of Be® had been added. During this period the
concentration of chromium in the salt rose from 35 to
65 ppm, indicating the removal of 140 g (5.89
equivalents) of chromium from the circuit walls. Thus,
when ?SNb disappeared from the fuel salt after the
final addition of beryllium (sample FP-15-62), a total of

97

ORNL DWG 69 5503

1968

- B - o é o o e o
a Be T ©
o
= & S (ADDED i mo ° o
— i i H I 1
3 J i f ] s 80
& g
C )
XIRE: ;
i &
e 020 — ° "
Lz £ —_
g © = =
= = EE @ o ] a
2« o - .f I — 60 &
o o ) o
(TATE! T d ©
Y o< 9, | =
- 5
= 0 w = 040 =
S 2 o0 &C b s
=0 35 s - 50 ¢
eq h & i z
se | &5 e ©
= Q
55 u —_ e e 40
QD I
z o I 530 _m«"f‘r o {0 g SAMPLES
& 0 b ig e FREEZE VALVE SAMPLES |
o <1 G N 4 i
% é o 1 30
I =
o o
=
g L ! 1 | i ) 20
SEPTEMBER OCTOBER NOVEMBER DECEMBER | JANUARY FEBRUARY

1969

Fig. 6.25. Corrosion of the MSRE fuel circuit in run Nos. 15 and 16, September—December 1968.

12 43 equivalents were mvolved in the reduction of
Fe?* and establishment of the Cr® + 2UF, = 2UF, +
CiF, equilibrium  Samples of the salt obtained during
the brief pertod of the subsequent run (No 16) as well
as the beginning of 233U power operations in run No
17 (FP-16-4 and FP-17-2) showed the presence of 52
and 29% of the ? 3 Nb inventory respectively

In Fig 626, nomnal values of [U3]/[ZU] are
shown for runs 17—20 These operations comprse
nearly the total power operation of the MSRE with
2337 fuel The concentrations of UF; shown m Fig
6 26 are based on the assumption that 0 76 equivalent
of oxidation results from the fission of 1 at wt of
uranium® and that the maximum power achieved by
the MSRE was 74 MW(t) They also assume that the
observed increases 1n the concentration of chromium in
the fuel salt at the beginning of runs 19 and 20 are
indicative of the total loss of U?* from the salt, the
nominal values for [U**]/[ZU] at these instants are
thus shown as zero The equibibrium constant for the
corrosion equibbrium Cr® + 2UF, = 2UF, + CrF,
reaction at 650°C, assuming an activity for Cr® mn the
Hastelloy to be 0 03, 151 271 X 107 Thus, 1n a regime
such as that which prevailed during the 1nitial stages of
run No 19, the rate at which Cr° 1s leached from the
Hastelloy N circuit gradually decreases as the Cr®*
concentration of the fuel salt increases During the
mnitial period of run No 19 the Cr?* concentration of
the circulating fuel salt rose from 72 to 100 ppm At
that point the equhbrium concentration of
[U*1/[ZU] 1n the fuel salt anticipated from free
energy and activity data 1s ~0 5%

The disposition of *>Nb 1n the 22U fuel durnng the
mitial period of run No 19 ndicates that when
[UP*]/[ZU] was less than ~0 5%, Nb became oxidized
and entered the salt, possibly as Nb>" or Nb** Then, as
the corrosion reaction Cr® + 2UF, = 2UF; + CrF,
proceeded to equilibrium, the U**/TU concentration
ratio increased, and at a [U*]/[ZU] value of 0 5%,
?SNb precipitated from the fuel salt Two levels of
nominal U**/ZU concentration are given during run
Nos 17 and 18, the hugher values based on the assump-
tion that corrosion of the fuel circuat during the early
stages of run No 17 may have accounted for a fraction
of [U*]/[ZU] The extent to which this reaction might
contribute to the total concentration of UF; in the fuel
at the beginning of run No 17 1s obscure, because the
MSRE was operated at full power at the inception of
run No 17 Such operation deposits the noble metal
fission products on the surface of the Hastelloy N,
causing the activity of Cr® at the alloy surface to be
effectively reduced The beginning period of run No 19
1s not analogous, for not until the corrosion equilibrium
was established was the reactor operated at full power
for sustained periods

Freeze-valve samples were obtained at the request of
E G Bohlmann and E L Compere during runs Nos
1720 Their analyses of the salt removed from the
pump bowl showed ?>Nb disposition as indicated by
the data points in Fig 626 It 1s evident that, as
[U*]/[ZU] decreased below 0 5% 1n run No 17,%°Nb
was oxidized and distnnbuted to the fuel salt, to be
removed subsequently as this concentration ratio was
exceeded Of the data shown in Fig 626 only the
98

ORNL-DWG 6914557

3.0
N
2.8 — N
2.6 U3+/SU NOMINAL CONCENTRATION
- = U3T/ZU NOMINAL CONCENTRATION, ASSUMING CREDIT
FOR CORROSION AT BEGINNING OF RUN NO. {7
2.4 s 9%Np EXPERIMENTAL
DISPOSITION OF 25Nb
2.2
2.0
1.8
& 16
7 .
m 1.4 N
I y \
i \h\ \\ h\ \\ My \
2 g \\ < N \ \ 60
Y
X \\\ \ \ \
1.0 NN A hd " 50
\\\\ 4 A \
0.8 AY . 40 @
AR 3 \ -
0.6 \ A b A AN . 30 ‘z
S \\\\\\\\\\\w\\m\{m\m\ AMMHTITET g - oos
0.4 ] y / 20 o
\
0.2 AN 10
/] 4 / I/ 4
0 it / b 0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Mwhr (x103)
\ y I\ . I\ y N R—
RUN 17 RUN 18 RUN 19 RUN 20

Fig. 6.26. Effect of U>'/XU on distribution of >°Nb in the MSRE fuel salt.

second data point in run 18 appears to be anomalous. It
seems likely that this result represents incomplete
equilibrium since the sample was obtained just after
[UP*]/[ZU] had been adjusted to a nominal value of
0.7%. In view of the fact that Be?, in contact with the
molten fuel, generates reduction potential gradients
which result in the reduction not only of U*" to U**
but of Cr*" to Cr® as well, one might anticipate a
kinetic factor to be significant in Be® = Be?", U** =
U, Cr** == Cr%, Nb3 = Nb? equilibria in the MSRE
fuel salt.

The results described in Fig. 6.26 indicate that when
the [U>]/[ZU] of the 233U MSRE fuel was poised at
~0.5%, disposition of ?>Nb toward solution in the salt
or deposition within the reactor was at a null point,

and, as indicated by behavior during ??°U operations,
the U>*/°5Nb ratio is the controlling factor. At 0.5%
U?* this ratio is 8.5/1.

Little is known concerning the chemistry of niobium
in the MSRE fuel salt. Preliminary results of laboratory
experiments indicate?? that under mildly oxidative
conditions niobium assumes an oxidation number of
~3.7. The fact that during run 15, when the oxidation
potential of the fuel salt was sufficiently high to permit
Fe?" to exist in the salt in significant concentrations,
nearly all of the 3Nb inventory of the fuel salt was in
solution, whereas in subsequent operations when the
oxidation potential was less, no greater than ~50% of
the ?>Nb was found in the salt. This behavior seems to
indicate that when niobium was deposited on the
moderator graphite it reacted to form niobium carbide
and that under the various redox regimes which were
established in the MSRE during runs 17 to 20 niobium
carbide was not removed from the moderator graphite.
The prevalence of niobium as the carbide is compatible
with the experimental observations by Blankenship et
al.'® and as noted by Cuneco and Robertson,?® who
found that the concentration of *>Nb at all profiles in
three different types of graphite was greater than would
have been anticipated if after deposition the isotope re-
mained as the metallic species.

References

1. A. Taboada, MSR Program Semiannu. Progr. Rep.
July 31, 1964, ORNL-3708, p. 330.

2. W. R. Grimes, Nucl. Sci. Appl. Technol. 8, 138
(1970).

3. M. B. Panish, R. F. Newton, W. R, Grimes, and F.
F. Blankenship, J. Phys. Chem. 62, 980 (1962).

4, J. O. Blomeke and M. F. Todd, “Uranium-235
Fission Product Production as a Function of Thermal
Neutron Flux, [Irradiation Time, and Decay Time,”
USAEC report ORNL-2127, Oak Ridge National Lab-
oratory (November 1958).

5. C. F. Baes, “The Chemistry and Thermodynamics
of Molten Salt Reactor Fluoride Solutions,” Proc.
TAEA Symposium on Thermodynamics with Emphasis
on Nuclear Materials and Atomic Transport in Solids,
Vienna, Austria, July 1965,

6. W. R. Grimes, G. M. Watson, J. H. DeVan, and R.
B. Evans, “Radiotracer Techniques in the Study of
Corrosion by Molten Fluorides,” Proceedings of the
Conference on the Use of Radioisotopes in the Physical
Sciences and Industry, Sept. 6—17, 1960, vol. 1Ii, p.
559, International Atomic Energy Agency, Vienna,
Austria, 1962.

7. H. E. McCoy, An Evaluation of the Molten Salt
Reactor Experiment Hastelloy-N Surveillance Speci-
mens — Fourth Group, ORNL-TM-3063 (1970).

8. MSR Program Semiannu. Progr. Rep. Aug. 31,
1965, ORNL-3872, p. 113.

9. MSR Program Semiannu. Progr. Rep. July 31,
1963, ORNL-3529, p. 130.

10. J. F. Engel, P. N. Haubenreich, and A. Houtzeel,
Spray, Mist, Bubbles, and Foam in the Molten Salt
Reactor Experiment, ORNL-TM-3027 (June 1970).

11. Letter dated February 9, 1971, from D. B.
Trauger to M. Shaw, Assessment of Need for Oxygen
Getter in MSBR Fuel.

12. Furmnished through the courtesy of C. Crouth-
amel, Chemical Engineering Division, Argonne National
Laboratory, Argonne, Il

99

13. W. R. Grimes, unpublished work, 1970.

14. B. McNabb and H. E. McCoy, internal corre-
spondence, March 31, 1971,

15. R. E. Thoma, MSR Program Semiannu. Progr.
Rep. Feb. 28, 1970, ORNL-4548, p. 93.

16. F. F. Blankenship, E. G. Bohlmann, S. S. Kirslis,
and E. L. Compere, Fission Product Behavior in MSRE,
ORNL-4684 (in preparation).

17. S. 8. Kirslis and F. F. Blankenship, MSR Program
Semiannu. Progr. Rep. Feb. 28, 1967, ORNL-4119, p.
131.

18. J. H. Shaffer and W. R. Grimes, MSR Program
Semiannu. Progr. Rep. Feb. 28, 1969, ORNL-4396, p.
135.

19. H. W. Kohn and F. F. Blankenship, ibid., p. 137.

20. J. M. Dale, MSR Program Semiannu. Progr. Rep.
Aug. 31, 1967, ORNL-4191, p. 167.

21. E. L. Compere and E. G. Bohlmann, MSR
Program Semiannu. Progr. Rep. Feb. 28, 1969,
ORNL-4396, p. 139.

22. C. F. Weaver, private communication.

23. D. R. Cuneo and H. E. Robertson, MSR Program
Semiannu. Progr. Rep. Aug. 31, 1968, ORNL-4344, p.
141.

7. DETERMINATION OF REACTOR POWER

7.1 Power Estimates with 23° U Fuel from Heat
Balance and Qther Methods

Estimates of the power developed by the MSRE can
be derived from nuclear and heat balance data. Both
bases were used from the beginning of power operations
with the MSRE. As refinements and corrections wete
introduced into the physical property and nuclear data,
new estimates were made. Consequently, MSR Program
progress reports cite various values for the maximum
average power of the reactor ranging between 7 and 8
MW(t).

The results of the investigation described in this
chapter fixed the maximum power of the MSRE at 7.4
MW(t), a value with which other estimates have more
recently agreed.

The power generation rate of the MSRE was cal-
culated routinely by an on-line computer; power
production was determined from computation of the
heat balance in the fuel and coolant systems. Details of
the methods employed are described elsewhere.!*?
Once heat balances were established, nuclear instrumen-
tation systems were calibrated to correspond to the
nuclear power indicated by the heat balance.
The results of chemical and isotopic analysis of fuel
salt were assessed carefully throughout the period when
the MSRE was operated with ?**U fuel and later when
the fuel charge was comprised of 223U and plutonium,
with the purpose of employing these results to monitor,
if possible, the power generated by the reactor.
Attempts to use chemically determined values of the
concentration of uranium in the fuel-salt samples were
generally unsatisfactory because of the point-to-point
scatter in these data and because of the overwhelming
effect that the use of flush salt in the fuel circuit had on
the uranium concentration base lines. The amounts of
fuel- and flush-salt residues remaining in the fuel circuit
after drains were estimated by comparisons of the
chemical analyses of uranium in the flush salt with
those in the fuel salt during each period of use. Not
until the uranium concentration of the flush salt had
undergone several increments was the precision of the
average mass of fuel-salt residues narrowed to within 5
kg. This was, however, of insufficient precision to
afford a sufficiently accurate base line for computation
of the power generation. It was thus evident that the
results of wet chemical analyses for uranium in the fuel
salt would be of potential use in establishing burnup
rates only after long periods of power generation which
were free from drain-flush-fill interruptions.

Until March 1968, calculations of the heat balance
indicated that the maximum power generated by the
MSRE was 7.2 MW(t). By that time, a discrepancy
between nominal concentration and analytical values
for uranium in the circulating fuel salt began to appear;
the analytical data over an extended period in run No.
10 showed a negative divergence from the nominal
concentration of the uranium in the fuel of about 10%.
During this period, however, computations of the
reactivity balance did not indicate corresponding or
anomalous decreases in the fissile concentration of the
salt. Reactivity balance calculations had previously
indicated that this method of evaluating reactor per-
formance was sensitive to a factor of 10 greater than
chemical analysis with respect to detection of changes
in uranium concentration of the fuel. The chemical data
suggested that the maximum actual power output of
the reactor was ~8.0 MW rather than 7.2 MW. However,
little credibility could be accorded to this conclusion
for the reasons cited above.

By early 1968 a sufficient amount of 23°U was
generated in the fuel salt to suggest that comparison of
the analytical results of mass spectrometric measure-
ments would provide a good measure of the integrated
power. Tests of this comparison® yielded a slope that
was within 1% of the theoretical slope for a power

100

ORNL~DWG 68- 5509

1000
[ ]
pd ./
800
o/
82"
= o
= - ®
4
t5 600 4
>_
) /
= 1@
& . /
2 L
" 400 < S
g 7
&

= %

200

0
0 10 20 30 40 50  (x10°)

INTEGRATED HEAT-BALANCE POWER (Mwhr}
Fig. 7.1. 236y buildup in MSRE vs power production.

generation rate of 7.2 MW. The uncertainty in the
theoretical slope was regarded to be probably less than
10%, and the actual integrated power was probably
within 10% of the value indicated by the heat balances
(see Fig. 7.1).

Although the 23®U production rate seemed to con-
firm estimates that indicated the maximum power
output to be 7.2 MW, suspicion was growing that the
heat balance calculations were in error, for data
collected at different power levels indicated that the
value employed for the specific heat of the LiF-BeF,
coolant salt was not temperature dependent,® whereas a
temperature-dependent relationship was employed in
the heat balance.®

A program of laboratory measurements of the en-
thalpy of solid and liquid Li, BeF4 from 273 to 900°K
was completed at this time, by investigators at the
National Bureau of Standards,® from which it was
shown that the heat capacity was 0.56 cal g 7' (°C)™1. A
fluoride mixture was synthesized for confirmation at
ORNL of the NBS value. Results of these measurements
indicated that the derived heat capacity of the coolant
sali was 0.577 £ 0.008 cal g7'(°C)™",” in good
agreement with the NBS investigations but substantially
higher in the operating temperature range of the MSRE
than the previously used value, and showed essentially
no variation with temperature. The new value of the
specific heat was incorporated into the computerized
heat balance computations prior to the beginning of
operatton with **3U fuel Calculated full-power level
was changed from 72 to 80 MW as a result of the
revision in the value for specific heat and m accord with
the analytical chemical results

Throughout this entire period, changes in the nominal
amounts of uranium 1sotopes in the fuel salt were
computed based on average crosssection data for
thermal and epithermal neutron reactions in the MSRE
spectrum that were current to 1965%? The con-
sumption and production rates were

234y ~2 167 X 107™% g/MWhr

235y 5417 x 1072 g/MWhr

236y +1 083 x 1071 g/MWhr

238y 7017 X 1072 g/MWhr
Total ~5146 X 1072 g/MWhr¢
4Ret 9

7.2 Power Output of the MSBR Based on the Isotopic
Composition of Plutonium

The potential use of plutonium as a fuel for molten-
salt reactors has been assessed periodically for more
than a decade The results of one early study'® showed
that a PuF;-fueled two-region homogeneous fluoride
salt reactor was operable, although its performance was
poor Further development was not pursued for neither
the chemical feasibility nor methods for mmproving its
performance were obvious Although the thermo-
chemical properties of the plutonium fluorides were not
well established at that tume, 1t was clear that the most
soluble fluoride, PuF,, was too strong an oxidant for
use with the available structural alloys The solubility of
PuF;, while sufficient for cnticahty even in the
presence of fission fragments and nonfissionable 1s0-
topes of Pu, was assumed'! to linut the amount of
ThF; which could be added to the fuel salt This
limitation, coupled with the condition that the con-
tinuous use of 23?Pu as a fuel would result 1in poor
neutron economy in comparison with that of ?33U-
fueled reactors, vitiated further efforts to exploit the
plutomum fluornides for application in two region
MSBRs

Recent developments in fuel reprocessing chemustry
and 1n reactor design have established the feasibility of
single-fluid molten-salt breeder reactors One of the
alternative modes of operating such reactors is to
employ plutonium in place of enriched ??3U for the

101

imtial fuel loading and startup of the reactors Assess-
ments of plutomum to start up MSBRs 1n this way have
concluded that plutomium should prove to be a very
satisfactory alternative to enriched uranium, either as
initial fuel for a molten-salt breeder reactor or as mnitial
fuel and subsequent feed material ' 2714

It now appears that 1t will be possible to operate an
LiF-BeF,-ThF,-PuF; single-flnd molten-salt reactor
with lower concentrations of thonum and plutonrum
than earlier considerations required, for example, with
thorium fluoride concentrations of 8 to 12 mole % and
with a plutomum fluoride concentration of approxi-
mately 25% less than required for 233U loading, that 1s,
<02 mole % These conclusions indicated the de-
sirabtlity of demonstration experiments to examine, 1n
as nearly similar application as possible, the behavior of
plutontum m an MSBR The chemical feasibihty of this
application was evaluated!® and found to hold pronuse
for successful application Therefore, when the MSRE
resumed operations with 222U fuel, plans were made to
include plutonium as a constituent of the fuel Using
pre-1965 cross-section data and assuming that the
maximum operating power of the reactor was 8 0 MW,
efforts were 1nitiated to establish a material balance for
plutonium.

During the final period of operation with
fuel, a sufficient amount of plutonium was generated
for 1ts detection n salt samples to be tractable by
standard analytical methods Approximately 600 g of
plutonium was generated in the MSRE fuel as 1t was
operated with 235U fuel by neutron absorptions in
2381, enough to afford a comparison of the analytical
data with anticipated values The results of that
comparison, expressed as a matertal balance for plu-
tonum, showed that the analytical chemical methods,
which were previously satisfactory for determination of
the concentration of plutonium in the fuel salt, weire of
questionable utility for use with the 232U fuel charge
and that 1sotopic dilution methods, using mass spectro-
metric analyses, afforded the most satisfactory means
of analysis They mdicated, in addition, that at the
maximum concentrations in which plutonium occurred
in the MSRE fuel salt, 1t existed as a stable chemical
entity and, by inference, that Pu, O3 1s not precipitated
in the presence of low concentrations (50 to 60 ppm)
of oxide 1on

The net production rates for plutonium generated n
the 235:2381J fuel salt have been estimated’ ¢ to be

235,238y

G239 = G238.( 018584(¢—0 605377’ T

=0 3318757y
G240 = G238.(( 008233~ 0 29166 7° T
0 0648799¢—0 291667 T

+0056647¢—0 333187°7)

where

Gk = mass of k i crculation (g) (k = 238, 239, 240
refers to 238U, 229Pu, and 24 °Pu respectively),

T = time mtegrated power (MWhr)

At termunation of the %3°U expeniment, 589 g of
plutonium should have been generated in the fuel salt
with the reactor operating at a maximum power of 8 0
MW For a fuel charge of 4900 kg, this corresponds to
120 ppm The average concentration of plutonium 1n
the fuel salt, as determuned from the results of analyses
of 18 fuel salt samples obtained during the latter period
of power operations with 235238 fuel, was 118 ppm
The analytical data do not show a significant trend and
are probably not sufficiently precise to use as a basis to
mnfer that a real difference exists between calculated
and analytical values On completion of these power
operations, urantum was removed from the fuel salt by
fluorination ' 7 Tt was anticipated that the plutonium
would remam 1n the carrier salt Five samples of the
carrter salt were removed from the fuel drain after
fluorination and found to have an average concen
tration of 120 ppm of plutonium, representing a total
of 562 g, as compared with an expected value of 125
ppm, that 15, 27 g of plutontum 1s not accounted for by
the results of these analyses

The concentration of plutonium 1n a salt specimen 18
calculated from the relation

Concentration of Pu (ppm) = dpm/g X 3 97 X 107®
(338N X 1470+ 239N X 5402 X 107% + 240y
X200X 1072 +24 N X 410X 107

+2%2NX 35X 107%),

where N 1s the atom fraction of plutonmum as the
1sotope designated

The concentration of plutonium 1n each of the 10 g
samples of fuel salt taken since the beginning of 233U
operations was determined using conversion factors
which assumed that the plutonium in the MSRE
consisted entirely of 2*°Pu and 2#°Pu The average
concentration of plutonum after full loading of the
reactor was achieved was found to be 147 ppm (Table

102

7 1) Ths value indicates the presence of a total of 689
g of plutomum, thus the enriching salt might then have
been expected to contain 100 g of plutonium In an
attempt to substantiate this conclusion a sample of the
"LiF 2?3UF, ennching salt was obtained from a
section of transfer line at the TURF and submutted for
chemical and mass spectrometric analysis The salt
residue which was contained in the line was considered
to be typical of that delivered for use in the MSRE The
results of mass spectrometric analysis showed that this
salt contained plutontum 164 wt % of which was
238Pu Rewvised conversion factors were therefore re
quired for calculation of the total plutonium concen
tration of the TURF salt and 233U fuel salt because of
the high specific activity, 6 46 X 10° dis sec™ pg™",
for 238Pu The plutomum concentration of the TURF
salt appears now to have been 249 ppm using the
revised factor and corresponds to the addition of 16 4 g
of plutonium along with the "LiF 223 UF, eutectic
Current estumates of the nominal concentration of
plutonium are now based on the assumption that this
amount of plutonium was added to the fuel salt and,
therefore, that the MSRE contained ~605 g of plu
tonium at the beginning of 233 U power operations

A plutontum 1nventory of the MSRE fuel was
computed, both before and after loading with 223U
fuel, based on Prince’s estimates of the production and
fission rates for plutonium and on the assumption that
16 4 g of plutomium was contained 1n the enriching salt
These values showed that the plutonium production
rate during *3° U operations was greater than previously
antictpated and, correspondingly, that the relative
changes 1n ?*%Pu and 23°Pu during 233U operations
were also shightly different

Estimates of the variation of 23°Pu and 2#°Pu during
recent power operations require that the quantity and
composition of the final plutonium nventory be known
accurately As noted previously,'® and as shown m Fig
7 2, attempts to determine the concentration of plu
tomum 1n the fuel salt from gross alpha count measure
ments were not very satisfactory because of the high
specific activity of 238Pu An improved estimate of the
plutonium iventory of the system was made from
extrapolations of the observed changes in *>°Pu and
24%Pu 1n the beginning stages of power operation with
233U fuel The mitial 24 °Pu/?3°Pu concentration ratio
was computed to be 00453, with the plutonium of the
reactor at that point as 568 g, approxmmately 2% more
than estimated from previous analyses Current esti
mates of inventory values have been computed for this
revised starting mventory The values obtaining at the
time the samples were taken were based on estimated
103

Table 7.1. Summary of MSRE fuel-salt analyses: plutonium

Net Weight of Pu in Weight % Pu/ZPu Concentration of

Sa;nopie Mwhr Fuel Sait (g) Calculated Analytical Puppm)
' 23%py 240py z 239p, 240p, 23%p, 240p,  Calculated? Analytical
FP1441 67,767 487 18.6 506 96.32 3.68 96.46 3.31 94 a5
FP14-42 62,305 492 18.8 511 95 120
FP1443 62,705 495 19.0 514 95 125
FP14-44 63,251 498 19.3 517 96 128
FP14-46 63,537 500 19.5 520 96 127
FP1447 63,671 501 19.6 521 97 122
FP14-48 64,211 505 19.8 525 97 126
FP14-49 64,232 505 19.8 525 97 125
FP14-50 64,234 505 19.8 525 97 123
FP14-51 64,367 507 20.0 527 97 119
FP14-52 64,994 510 20.3 530 98 111
FP14-54 65,397 513 20.5 534 99 98
FPi4-56 65,809 517 20.8 538 100 119
FP14-58 66,351 521 21.2 542 101 118
FP14-59 66,982 525 21.8 547 102 120
FP14-64 68,720 537 22.7 560 95.94 4.06 96.05 3.67 104 145
FP14-65 69,481 543 233 566 114 97
FP14-68 69,838 546 235 570 115 102
FP14-Final 72,454 563 25.6 589 120 120
FP15-6 72,454 571 27.1 599 128 113
FP15-9 72,454 573 2.5 602 128 112
FP15-10 72,454 575 27.9 604 129 96
FP15-12 72,454 575 27.9 604 129 112
FP15-18 72,454 575 21.9 604 129 100
FP13-33 72454 575 27.9 604 129 134
FP15-38 72,454 576 28.0 605 129 143
FP1542 72,454 576 28.0 605 129 135
FP15-60 72,454 576 28.0 605 129 129
FP15-63 72,454 576 28.0 605 129 141
FP15-65 72,454 576 28.0 605 129 159
FP15-68 72,454 576 28.0 605 95.85 3.99 95.43 4.12 129 157
FP17-1 72,454 576 28.0 605 129 130
FP17-4 73,127 575 28.0 605 129 134
FP17-9 73,830 572 29.4 602 128 148
FP17-12 75,434 569 304 598 128 138
FP17-18 76,183 564 3L.9 597 127 149
FP17-19 76,791 563 325 596 127 147
FP17-20 78,026 560 32.7 594 127 141
FP17-23 79,154 557 34.8 593 127 130
FP17-24 79,814 555 35.5 592 126 142
FP17-27 80,828 552 364 589 126 140
FP17-28 81,610 550 37.2 588 125 134
FP17-30 82,771 547 38.3 586 125 149
FP18-1 84,741 542 40.1 583 124 160
FP18-5 86,454 538 41.8 581 124 162
FPi8-10 87,267 536 42.5 580 124 145
FP18-13 88,265 533 43.5 578 123 154
FP18-22 89,886 529 44.8 575 92.63 7.20 91.81 7.33 122 164
FP18-27 90,898 527 45.7 574 122 144
FP1843 92,142 524 46.9 570 92.30 7.53 91.38 7.70 122 171
FP18-Final 92,985 524 47.3 570 122

@Based on total fuel charge.
104

ORNL-DWG 69-7561 .

i80

170

. lfi . f/ \ \y
| i ffi |V

——

>
o

o
o

g

\\\-

NOMINAL CONCEN’TRATIONZ
\ BEGINNING OF 233U POWER OPERATIONS ~ |
1 | 1
% %l END OF 233y OPERATIONS

Y
[

PLUTONIUM (ppm)

LY

100 f !
o
90 : .
80 5
55 60 65 70 70 75 80 85 90 (x10°) .
Mwhr

Fig. 7.2. Comparison of nominal and analytical values for the concentration of plutonium in the MSRE fuel salt.

Table 7.2. Isotopic composition of plutonium in MSRE fuel salt

Sample No. Description 238p, 23%py 2%0p, 24ipy 242p,
FP14-41 Z g (calculated) 487 18.6 a a
% calculated 96.32 3.68
% analytical 0.011 96.45 3.32 0.22 0.002
FPi4-64 Z g {calculated) 537 22.7
% calculated 95.94 4.06
% analytical® <0.015 96.05 3.67 0.28 0.005
FP15-68 Z g generated in MSRE (calculated) 563 25.6
¥ g added with 7LiF-?33UF, 0.27 12.9 2.44 0.37 0.33
g 0.27 575.90 28.04 0.37 0.33 .-
% calculated 0.04 95.20 4.64 0.06 0.06
% analytical <0.08 95.43 4.12 0.34 0.13

a241py and 2*2py not included; B. E. Prince estimates T g 241,242p, < 1.8 g -
bAnalyses performed by R. E. Eby. .
average values for the rates of change of 2?°Pu and
240Py in the period between samples (Table 7.2).

In the MSRE, fluid fuel was circulated at rates which
were sufficiently rapid with respect to changes in the
isotopic composition of the fissile species that the salt
samples removed from the pump bowl were repre-
sentative of the circulating stream. This characteristic of
molten-salt reactors makes it possible to use the results
of isotopic analyses for a variety of purposes. One
potential application, that of appraising the cumulative
power generated by the MSRE at various periods,
became apparent with the initiation of 2**U opera-
tions, for with 223U fuel the isotopic composition of
the plutonium inventory (produced partly by that
generated in 233U operations as well as from that
added later) would change significantly during power
production and would possibly serve as an accurate
indicator of the power produced.

About 600 g of plutonium was produced during
power operations with 233U fuel. Thereafter, addi-
tional plutonium was introduced into the fuel salt as a
contaminant of 7 LiF-2?? UF, enriching salt and later to
replenish the fissile inventory of the MSRE during
2331 power operations.

105

Additions of plutonium to the fuel salt as PuF; were
accomplished smoothly by use of capsules sealed by 12
zirconium disks. These containers were designed and
loaded by Carr et al,'® a typical capsule is shown in
Fig. 7.3. In contact with the fuel salt the zirconium
dissolved, permitting the PuF; to disperse and to
dissolve in the fuel salt. Photographs of one of the
capsules after use are shown in Fig. 7.4.

Resolution of the analytical problems with plutonium
showed that estimated production rates of plutonium in
the 235U fuel salt, based on pre-1965 cross-section
data, were low. The programs used for reactor physics
calculations were revised using post-1965 data, and as a
consequence, both consumption and production rates
for plutonium and uranium isotopes were revised
significantly.?°

Samples of the MSRE fuel salt were submitted
routinely for determination of the isotopic composition
of the contained fissile species. Comparisons of the
results of plutonium assays with nominal values that
should result from operations at various power levels
from ~7 to 8 MW were made. Within this range, best
agreement between calculated and experimental values
was obtained for a maximum power output of ~7.40

PHOTO 9747

s
5 2
[
¥ f‘»gd‘éd‘r

i

@

Fig. 7.3. 239PuF3 capsule for MSRE refueling.
e
o L
]
ey
=5 5

k)

iy

o
it
o, 4

FP 19-25 TOP SECTION, OPPOSITE SIDE FROM
THAT SHOWN IN OVERALL VIEW,

FP 18-25 BOTTOM SECTION, OPPOSITE SIDE
FROM THAT SHOWN IN OVERALL VIEW. SALT

RESIDUE MORE CLEARLY EVIDENT HERE THAN
IN OVERALL VIEW,

106

PHOTO 1877~ 71

FP 19-25 FIRST CAPSULE
USED TO ADD PuFz TO THE
MSRE FUEL SALT. CAPSULE
WAS PARTIALLY SUBMERGED
IN FUEL SALT FOR FOUR HOURS

FP 19-25 CAPSULE VIEWED FROM BOTTOM,
SHOWING SALT RESIDUE IN CAPSULE.

Fig. 7.4. Capsule sections.

Table 7.3. Isotopic composition of plutonium in the MSRE fuel salt at a power generation rate of 7.40 Mw(th)

f .

Fuel Circuit Inventory

Isotopic Composition

Isotopic Composition

Sample pi:ir — A239py A240p, (Caloulated) (2)? (Calculated) (Analytical)
No. e (e per 1000 Mwhe)  (gper 1000 Mwho) 355 Tl TR0 WURRWIR gy g, WLERWEPU g g
239Pu 240Pu 239Pu 240Pu

Run 17-1 0 0 5415 24,53 5684 9526  4.32 0.0453

FP 17-9 148 537 -3.025 +1,178 539.8  25.16 5674 95.14  4.43 0.0466 9526  4.35 0.0457
FP 17-18 466 2,227 ~2.996 +1.144 534.8 2709 5643 9476  4.80 0.0507 9428  5.16 0.0547
FP 17-19 542 3,656 ~2.949 +1.180 $30.6 2878 S61.8 9444 512 0.0542 9448  5.00 0.0508
FP1720 697 4,492 2,924 +1.105 528.1 2970 5603 94.26  5.30 0.0562 9420 5.5 0.0557
FP 17-23 920 5,862 ~2.894 +1.086 5242 3118 5578 93.97  5.95 0.0595

FP 1727 1047 7,31 ~2.871 +1.072 5205 3254 5555 9370  5.86 0.0625 9358  5.80 0.0620
FP17-28 1145 7,946 ~2.854 +1.063 518.2 3342 5540 9353 6.03 0.0645 9430  5.97 0.0639
FP17-30 1290 8827 ~2.845 +1.054 5157 3435 5515 9351 623 0.0666  93.16  6.18 0.0663
Run 17-F 1536 10,245 ~2.824 +1.039 S11.7 3582 549.9 93.04 651 0.0700

FP 181 1536 10,245 5141 3491 5514 9322 633 0.0679

FP 18-1 1536 10,245 ~2.875 +1.075 Si4l 3491 5514 9322 6.33 0.0679 9292  6.36 0.0684
FP 18-5 1562 11,231 ~2.860 +1.066 S13.5 3511 5511 9318 637 0.0684 9265  6.61 0.0713
FP18-10 1851 12,373 ~2.843 +1,055 507.6 3732 5473 9273 6.82 0.0735 9238 6.84 0.0740
FP18-13 1976 13,873 ~2.826 +1.038 505.0 3826 5457 9254 7.0l 0.0757 9216  7.04 0.0764
FP1822 1976 13,873 ~2.826 +1.038 505.0 38.26 5457 9254 7.0l 0.0757  91.80  7.36 0.0802
FP 1827 2306 15,523 ~2.799 +1.017 498.3 4069 5414 9203  7.52 0.0817 9163  7.49 0.0817
FP 1843 2461 17,281 ~2.773 +0.999 4952 41.81 5395 91.80  7.75 0.0844 9148  17.63 0.0834
Run 18F 2544 18,143 2758 +0.989 493.5 4241 5384 9167  1.87 0.0859

Run 19 2544 18,143 4952 4181 5394 9179 775 0.0844

FP19-17 2625 18,739 ~2.770 +0.993 493.5 4239 5384 9167  7.87 0.0858 9122  7.84 0.0860
FP19-18 2642 19,093 ~2.763 +0.989 4932 42,51 5382 9165  7.90 0.0862 9119  7.87 0.0863
FP19-21 2724 19,449 ~2.755 +0.985 4916 4310 S537.1 9152  8.02 0.0877 9102  8.03 0.0882
FP19-22 2791 19,693 ~2.752 +0,985 490.2  43.58 5363 9142  8.13 0.0889 9090 8.11 0.0892
FP19-24 2791 19,693 2,752 +0.985 490.2 43,58 5363 9142  8.13 0.0889  89.88  8.99 0.1000
FP-19-256 2791 19,693 ~2.760 +0.983 5411 46,74 5906 91.62  7.91 0.0864

FP19-27 2791 19,693 ~2.760 +0.983 5411 46,74 5906 9162 791 0.0864  91.01  8.03 0.0882
FP19-30 2818 19,693 ~2.760 +0.983 540.5 4693 5899 9163  7.96 0.0868  90.89  8.13 0.0895
FP19-31-4 2818 19,693 ~2.750 +0.983 662.6 5451 719.6 92.08  7.58 0.0823

FP19-35 2964 20,961 ~3.820 +1.369 658.6 55.96 7171 91.85  1.80 0.0850 9135 777 0.0851
FP19-43 3102 21,989 ~3.810 +1.357 654.8 5732 7145 9164  8.02 0.0875 9118  7.90 0.0866
FP19-53 3294 23,185 ~3.795 +1.339 649.5 5918 7111 9133 832 0.0911  90.88  8.25 0.0908
FP19-63 3561 24,631 3,775 +1.339 6422 6177 706.4 9091  8.74 0.0962  90.49  8.50 0.0939
FP19-74 3693 26,078 ~3.750 +1.316 639.3 6277 7045 89.09 891 0.0982  90.15  8.78 0.0974
Run19-F 3774 27,069 ~3.720 +1.302 637.1 6354 7031 90.62  9.04 0.0997

Run20-F 3774 27,069 625.8 6181 6900 90.69 896 0.0988

FP 20-6 3820 27,236 ~3.670 +1.309 624.6 6225 6893 90.61  9.03 0.0996  89.89  8.99 0.1000
FP20-31 4159 28,294 ~3.660 +1.293 615.6 6543 6834 90.07  9.57 0.1062  89.36  9.43 0.1055
Run20-F 4159 28,294 ~3.660 +1.293 615.6 6543 6834 90.07  9.57 0.1062

2 Average for period between samples,
b A ssumes 92% fuel charge in circulation,

LO1
108

ORNL-DWG 706755

0.4%
/O

G50 a/

. /v

/a
Q.09 .?: /“
/V./ Ll/' ¢
o

0.08 2

240 /239Pu

yd
Q.07

i

o, *

0.05 / e

-]
e ’\
0.06 NOMINAL VALUES OF 2407239 AT 740 Mw (th) — ]

0.04 [ i | |

@ RUN
RUN 17—————'——-RUN 48—>‘<——RUN 19———-—'*;(7,'

i ! | !

0 1000 2000

3000 4000

equivalent full -power hours

Fig. 7.5. [2*°/239Py] in the MSRE fuel salt circuit during 2> >U operations.

MW(t). A comparison of calculated and observed values
for the isotopic ¢omposition of plutonium which
should result from a maximum power output of 7.40
MW is shown in Table 7.3 and in Fig. 7.5. Agreement
tests indicate that the standard deviation between
calculated and observed values is £0.63% and that the
average positive bias in the experimental data is 0.093%.
On this basis the maximum power output was 7.41 %
0.05 MW(t). The precision of this value seems to be
adequate for related analyses of reactor operations.

It is considered unlikely that further refinements in
cross-section data for the plutonium isotopes will
require any substantial changes in the calculated in-
ventories used for this comparison. However, for the
purpose of estimating uncertainty in the combined
cross-section data and neutronic model used to cal-
culate reaction rates, an equivalent of 2% in the power
output is judged conservative. The results of experi-
ments designed to measure the 233U capture-to-
absorption ratio in the fuel of the MSRE were reported
recently by Ragan.?! From these results he concluded
that the full-power output of the reactor was 7.34 +
0.09 MW, in excellent agreement with our value.
Further corroboration of these estimates was provided
by Gabbard?®? in a recent reevaluation of the accuracy
of the data produced by the MSRE coolant salt flow
transmitters. He found that either the pressure trans-

mitter or the square-root converter component, which
generated 2- to 10-v signals for the computer, was
faulty and caused the device to indicate higher than
actual flows. Using a corrected flow rate of 793 gpm
rather than a nominal flow rate of 850 gpm, as
indicated by the computer, to recompute the heat
balance power, he estimated that the full-power output
of the reactor was 7.65 MW rather than the previous
value of 8.2 MW,

7.3 Isotopic Composition of Uranium during
233U Operations

For operation of the MSRE with 233U fuel, esti-
mation of the power output of the MSRE from
measured changes in isotopic composition of the fissile
material is achieved with considerably greater precision
from analyses of plutonium than from uranium. This
arises from the fact that the rate of change in the
relative fraction of the most abundant isotopes for
plutonium, ?*?Pu and 2*°Pu, is some four times that
for the uranium pair, 233U and 2% U. Analyses of the
isotopic composition of uranium in the fuel circuit
during 233U operations were employed, therefore,
primarily to determine whether they atforded approxi-
mate confirmation of the power estimate as inferred
from the plutonium data (Sect. 7.2). Calculations of the

109

Table 7.4. Isotopic composition of uranium in the MSRE fuel-salt circunif?

Sample

U/ZU (Wt %)

b 234 233
No, EFPH 233U 234U 235U '236U 238U { U]/[ U]
Run 17-] 0 84687 6,948 2,477 0.0808 35.807 0.08204
FP 17-18 466 84.590 7.011 2.489 0.084 5.828 0.08288
84.690 6.990 2.470 0.084 3. 771 0.08253
FP 17-24 920 84.489 7.073 2,501 0.087 5.849 0.08371
84.382 7.058 2487 0.089 5.986 0.08304
FP 17-32 1338 84.440 7.131 2.510 0.090 5.867 0.08445
84,445 7.128 2,487 0.087 5.843 0.08440
Run 17-F 1536 84.363 7.152 2.513 0.091 5.875 0.08477
Run 18-1 1536 84,393 7.136 2.511 0.091 5.870 0.08455
Run 18-2 1536 84.199 7138 2.507 0.089 6.067 0.08477
FP 18-4 1563 84.385 7.141 2.511 0.091 5.871 0.08462
&84.249 7.158 2.527 0.091 5.975 0.08496
FO 18-10 1852 84.326 7.180 2.518 0.092 5.883 0.08514
54.269 7.178 2.507 0.091 5.955 0.08517
FP 18-13 1976 84,298 7.199 2.521 0.092 5.890 0.08539
84.060 7.203 2.517 0.087 6.133 0.08568
FP 18-22 2221 84,241 7.232 2.529 - 0.095 5.902 0.08584
54.167 7.208 2.517 0.098 6.016 0.08563
FP 18-43 2461 84.189 7.265 2.534 0.097 5.912 0.08629
84.041 7.232 2.537 0.093 - 6.097 0.08605
Run 18-F 2544 84.169 7.279 2,536 0.098 5.916 0.08648
Run 19-1 2544 84.185 7.267 2.534 0.097 5.913 0.08632
FP 19 10-12¢ 2544 84,224 7.268 2.526 0.097 5.882 0.08629
FP 19-35 2964 84,377 1.349 2.543 0.100 5.919 0.08709
83,987 7.338 2.537 0.099 6.036 0.08737
FP 1943 3102 84.103 7.348 2,539 0.101 5.909 0.08726
83.994 7.328 2,533 0.099 6.047 0.08724
FP 19-53 3294 84.060 7.375 2.546 0.102 5.917 0.08773
83.912 7.358 2.537 0.101 6.092 0.08768
FP 19-63 3561 84.000 7.413 2,553 0.104 5.931 0.08825
83.927 7.408 2,542 0.102 6,021 0.08826
FP 19-74 3693 83.971 7.430 2.555 0.105 5.937 0.08848
83.801 7418 2.569 0.102 6.102 0.08851
Run 19-F 3774 83,953 7.442 2,557 0.105 5.942 0.08864
Run 20-1 3774 83.973 7.427 2.555 0.105 5.938 0.08844
FP 20-3¢ 3774 83.996 7.426 2.549 0.105 5.923 0.08840
FP 20-6 3820 83.986 7.433 2.550 0.105 5.925 0.08850
83.614 7.435 2577 0.104 6.271 0.08892
FP 20-31 4159 83.911 7.482 2.558 0.108 5.940 0.08916
83.742 7.484 2.577 0.105 6.092 0.08936

Upright type indicates values computed on the basis that the maximum power generated by the MSRE was 7.41 Mw(th). The
following rates (furnished by B. E. Prince) were used: 233U: —4.643 X 1072 g/Mwhr; 234U: +3.6325 X 1073 g/Mwhr; 23°U:
+9.5596 X 1075 g/Mwhr; 236U: +3.0725 X 10~% g/Mwhr; 238U: —2,90 X 10™* g/Mwhr. Results of mass spectrometric analyses

are listed in italicized type.
bEquivalent full-power hours.
“Fuel addition,
110

ORNL—-DWG 70-6756

0.080
P
L.
/'.,o’l//
0.088 -
o/
Afl/
0.086
QD ./
g /'/
§ ./'.o
0.084 > - P
_ NOMINAL VALUES OF 224233 AT 7.41 Mw (th)
/
0.082 !'
; RUN 17 +—RUN 18~——|-*—F?UN 19 %—Rgg %
0.080 E | t | z \ | n
0 1000 2000 3000 4000

equivalent full — power hours

Fig. 7.6. 2347233y in the MSRE fuel circuit during 233y operations.

isotopic composition changes of the uranium in the
MSRE fuel circuit that should have accompanied
operation of the reactor at a maximum power output of
7.41 MW(t) were made and compared with the results
of mass spectrometric analyses. The results of this
comparison are shown in Table 7.4 and in Fig. 7.6; they
indicate that the changes observed in the isotopic
composition of the uranium were in excellent agree-
ment with those of plutonium.

7.4 Isotopic Composition of Uranium during
235 Operations

Once the anomalies in the power production of the
MSRE were resolved by analysis of the mass spectro-
metric data, and assisted by refinements in the cross-
section data, internal consistency developed quickly
among various other analytical results. In particular, use
of the final estimate of power generation at a maximum
rate of 7.4 MW to appraise both isotopic analysis and
chemical analyses for uranium during the 23°U oper-
ational period showed that these results were entirely
consistent with those for uranium and plutonium
during 233U operations. In addition, the revised esti-
mates in the changes in consumption and production
rates for 234U, 235U, 236U, and 238U, together with
improved values for the amounts of uranium transferred
to the flush salt, brought nominal and analytical values
for the concentration of uranium in the fuel into good
agreement as shown in Table 7.5.

Isotopic composition of the uranium in the circu-
lating fuel salt was recalculated on the basis that the

reactor had operated at 7.4 MW and that the con-
sumption and production rates of the uranium isotopes
at this power level were those given in Prince’s revised
estimates.?® A comparison of the resulis obtained is
shown in Table 7.5. It will be noted that the greatest
disparity in nominal and analytical values listed in Table
7.5 is observed for 233U, In Sect. 3.6 it is noted that
further analysis of these data led to the conclusion that,
of the 23#U nominally charged to the MSRE drain
tanks, some 2 kg was probably not delivered. Re-
computation of these results in line with that con-
clusion would bring the nominal and observed values
for the relative fractions of uranium isotopes, as listed
in Table 7.5, into even closer agreement.

References

I. R. H. Guymon, MSRE Design and Operations
Report, Part VIII, Operating Procedures, ORNL-
TM-908, vol. Il (January 1966).

2. G. H. Burger, J. R. Engel, and C. D. Martin,
Computer Manual for MSRE Operators, internal memo-
randum, MSR-67-19 (March 1967).

3. R. C. Steffey, Jr., and J. R. Engel, MSR Program
Semiannu. Progr. Rep. Feb, 29, 1968, ORNL-4254, p.
10.

4. C. H. Gabbard, Specific Heats of MSRE Fuel and
Coolant Salts, internal memorandum, MSR-67-19
(March 1967).

5. R. B. Lindauer, Revisions to MSRE Design Data
Sheets, Issue No. 9, ORNL-CF-64-6-43 (June 1964).
Table 7.5. Isotopic composition of uranium in the MSRE fuel system?

111

Sample
No? — 234 235, 236, 238, - 234, 235, 236, 238,
(kg) {wt %)
Initial d
Loading - 0.325 - 147.272  147.60 - 0.282 99.7785
0.775 75.645 0.320 4.640 81.38 0.952 95.953 0.398 5,702
0.775 75.973 0.320 151.912  228.980 0.339 33,179 0,140 66.342
Run 3-F 0 228,210
Run 4-1I 0 0.711 69.660 0.294 139.288 209,953 0.339 33,179 0.140 66.362
FP 4-14 0 0. 350 53.850 g.140 86. 260
FP 4-28 0 a.350 33,249 0.14L €6.360
FP 4-33 0 0. 344 33.132 0.187 66,387
FP 4-37 0 0.343 33.188 0.138 66.351
FP 6-14 166 0.712 69.594 0.308 139.276 209,889 0.339 33,157 0.147 66.357
6.349 33,534 0.144 65,973
FP 6-19 4060 0.712 69.498 0.327 139.258 209,795 0.339 33,127 0.156 66.378
6. 344 33.445 0.152 66.060
FP 7-8 725 0.710 69.366 0.353 139.235 209.664 0.339 38.084 0.168 66.408
0. 358 33.592 0.176 85.876
FP 7-12 1012 0.709 69,250 0.376 139.213  209.548 0.338 353,047 0.179 66.435
4. 356 35,552 0.188 §5.904
FP 7-15 1047 0.709 69.236 0.379 139.211  209.535 0.338 33,043 0.181 66.438
0.347 35,161 0.177 86,315
FP 8-5 1047 0.707 68.992 0.371 138.657 208,727 0.339 33.053 0.178 66.430
0.348 33.205 0.175 66.214
¥P 10-7 1677 0.700 68.203 0.414 137.494  206.811 0.338 32.978 0.200 66.484
0.347 33.259 0.198 66.198
FP 11-14 2884 0.698 67.732 0.509 137.409 206,348 0.338 32.824 0.247 $6.591
0.343 32.892 0.241 66,524
FP 11-39 3856 0.697 67.340 0.588 137.337  205.962 0.338 32.695 0.286 66.681
0. 349 32.973 0.278 £6.408
FP 11-44 1937 0.696 67.305 0.595 137.331  205.927 0.338 32,684 0,289 66.689
R 0.351 32,962 6.283 66,404
FP 11l-44 4107 0.704 67.997 0.611 137.365  206.677 0.341 32.900 0,29 66.463
0,348 32,987 0.287 66.378
FP 11-48 4172 0.704 67.971 0.617 137.361  206.653 0.341 32.891 0.299 66.469
0.351 33.067 G.292 86.290
FP 12-5 4513 0.700 67.568 0.627 136.784  205.679 0.340 32.851 0.305 66.504
0.354 33.133 0.296 66.217
FP 1229 5121 0.700 67.321 0.677 136.740  205.438 0.341 32.769 0.330 66.560
. 0.25L 32,963 0,320 66, 366
¥P 12-51 5296 0.714 68.777 0.697 136.822 207.010 0,344 33.224 0.337 66.095
0.355 33.475 0.325 65.845
FP 12-58 5500 0.714 68.695 0.715 136.807  206.931 0.345 33.197 0.346 66.112
0,353 33.350 (.333 65,964
FP 14-25 6848 0.708 67.788 0.813 136,147  205.456 0.344 32.994 0.396 66.266
0.357 33.390 6.379 65.874
FP 14-46 7953 0.703 67.337 0.903 136.064  205.007 0.343 32.846 0.441 66.370
0.355 33,171 6.422 66.052
FP 14-68 8742 0.701 67.017 6.968 136.008  204.694 0.342 32.740 0.473 66.445
0.35¢ 32,945 6.450 66.251
FP 14-72 9006 0.701 66.910 0.989 135,988  204.588 0.343 32.705 0.483 66.469
0.350 33.083 0.260 66.367

?Bold faced type indicates values computed on the basis that the maximum power generated by the MSRE was 7.40 Mwth.
rates were used: 234U: - 1.665 x 16—3 o /EFPH, 235U:

- 4.05150 ¢ 1071 g/EFPH,

236

(see ORNL-4449 , p. 25). Results of mass spectrometric analyses are listed in italicized type.

=

Equivalent full power hours.
[

d

Refers to total inventory in fuel system.

€0.819 kg uranium added.
f1.642 kg uranium added.

All following items refer to fuel circuit inventory only.

J. H. Shaffer memorandum to R. E, Thoma, Sept., 28, 1970.

Us + 8.14 x 1072 g/EFPH,

238U:

The following
- 7.3691 g/EFPH
6. T. B. Douglas and Wm. H. Payne, J. Res. NBS, Ser.
A, 73A,479 (1969).

7. J. W. Cooke, L. G. Alexander, and H. W, Hoffman,
MSR Program Semiannu. Progr. Rep. Aug. 31, 1968,
ORNL-4344, p. 100.

8. B. E. Prince, MSR Program Semiannu. Progr. Rep.
Feb. 28, 1967, ORNIL-4119, p. 79; B. E. Prince, J. R.
Engel, and C. H. Gabbard, Reactivity Balance Calcu-
lations and Long-Term Reactivity Behavior with *3°U
in the MSRE, ORNL-4674 (in preparation).

9. B. E. Prince, MSR Program Semiannu. Progr. Rep.
Aug. 31, 1969, ORNL-4449, p. 25.

10. D. B. Grimes, MSR Program Quart. Progr. Rep.
June 30, 1958, ORNL-2551, p. 13.

11. J. A. Lane, H. G. MacPherson, and F. Maslan,
Fluid Fuel Reactors, p. 656, Addison-Wesley, Reading,
Mass., 1958.

12. P. R, Kasten, ORNL Reactor Division, personal
communication, 1968.

13. P. R. Kasten, J. A. Lane, and L. L. Bennett, Fuel
Value Studies of Plutonium and U-233, unpublished
work, 1962,

14. A.M. Perry, unpublished work, 1971.

15. R. E. Thoma, Chemical Feasibility of Fueling
Molten-Salt Reactors with PulF';, ORNL-TM-2256 (June
1968).

16. B. E. Prince, personal communication.

17. R. B. Lindauer, Processing of the MSRE Flush
and Fuel Salts, ORNL-TM-2578 (July 1969).

18. R. E. Thoma, MSR Program Semiannu. Progr.
Rep. Aug. 31, 1969, ORNL-4449, p. 98.

19. W. H. Carr, W. F. Shaffer, and E. L. Nicholson,
MSR Program Semiannu. Progr. Rep. Aug. 31, 1969,
ORNL-4449, p. 245.

20. B. E. Prince, MSR Program Semiannu. Progr.
Rep. Aug. 31, 1969, ORNL-4449, p. 22.

21. G. L. Ragan, MSR Program Semiannu. Progr.
Rep. Aug. 31, 1970, ORNL-4622, p. 31.

22. C. H. Gabbard, internal correspondence, Mar. 22,
1971.

8. PHYSICAL PROPERTIES
8.1 General Properties

Successive refinements in the design or in operational
criteria for molten-salt reactors evoke repeated reap-
praisals of the accuracy and precision of the available
physical property data for reactor materials.

As experience with the MSRE grew, most of the
values for physical properties of the fuel, flush, and

112

coolant salts were reconfirmed, while for some proper-
ties, the values originally employed appeared to be of
questionable accuracy. The properties of greatest signif-
icance in these respects were viscosity, thermal conduc-
tivity, electrical conductivity, phase transition behavior,
heat capacity, heat of fusion, density, expansivity,
compressibility, vapor pressure, surface tension, solu-
bility of the gases helium, krypton, and xenon, isocho-
ric heat capacity, sonic velocity, thermal diffusivity,
kinematic viscosity, and Prandtl number. Of these, the
precision and accuracy are intrinsically variable and
depend on methods of measurement or estimation.
While the accuracy for some classes of properties, for
example, liquidus-solidus temperatures, developed to be
of minor concern, others, such as surface tension and
heat capacity, had neither been measured nor were
measurable with the precision that was desirable for
reactor performance evaluations. In response to MSRE
program requirements, reappraisals and refinements of
physical property data were repeatedly made whenever
it was feasible. The final evaluation led to values which,
for the coolant and flush salt, are summarized in Table
8.1 and, for the fuel salt, are summarized in Table 8.2.

8.2 Density of Fuel and Coolant Salts

Laboratory measurements of the density of molten
fluoride mixtures were performed by various groups of
investigators before the MSRE was operated. The
precision in the results, however, was considered to be
less than desirable for use in on-site calculations of
reactivity balances and for appraisal of temperature
effects. An effort was made, therefore, to make direct
measurements with the reactor that would serve these
purposes and that would afford independent measure-
ment of the salt inventories. An early account gives a
perspective on the success of these efforts: !

The weigh cells on all the salt drain tanks were calibrated with
lead weights shortly after the equipment was installed and
before the tanks and connected piping were heated. Additional
data were obtained with the tanks hot during various salt-charg-
ing and transfer operations. These data served both to calibrate
the weighing systems and to give a measure of the density of the
salts at operating temperature. Throughout the operation, salt
inventories were computed from weigh-cell readings, using scale
factors and tare corrections obtained from the calibration tests.

Cold and hot calibration of the coolant drain-tank weigh cells
gave scale factors differing by less than 0.5%. Fuel drain tank 2
(FD-2) was calibrated hot twice, with flush salt and with fuel
carrier salt; scale factors were within 0.2% of each other but
were about 4% higher than the original, cold calibration. The
reason for this discrepancy has not been established.

The coolant-salt density was measured in the reactor by three
different methods; values ranged from 121.3 to 122.3 Ib/ft? at
1200°F, with an average of 121.9 Ib/ft3. When flush salt, which

113

Table 8.1 Physical properties of ithrum fluoroberyllate, Li;BeF,?

Property Value Fstimated precision
Viscosity n(centipoises) = 0 116 exp [3755/1(°K) ] +7
Thermal conductivity 0010 watt cm ™! °C™! *10%
Electrical conductivity K=154X60x 1073 (ohm cm)! at 500°C +10%
Melting pomnt? 459 1°C 0 2°C
Crystal structure Hexagonal, space group R3,2=13 294, c=891A 001 A
Heat capacity

Liquid Cp=057calg™t °C™! 3%

Sohd Cp=031+361X107*7(°C) cal g™ °C™! +3%
Density

Liquid pP=2214 ~ 42X 107%7TCC) g/em? 1%

= 122 Ib/ft® at 650°C

Sohd 0=21953 g/cm3¢
Expansivity 2 14 X 107%/°C at 600°C *10%
Compressibility Br(°K)=23X 10712 exp [1 0 X 10737(°K)] cm?/dyne Factor 3
Vapor pressure log P(torrs) = 8 0 — 10,000/7(°K) Factor 50 from 500 to 700°C
Surface tension v =260 — 0 127(°C) dynes/cm +30, 10%
Solubihity of He, Kr, Xe T(°C) He Kr Xe 2 Factor 10

500 66 013 003
600 i06 055 017
700 151 17 067
800 201 44 20

X 1078 moles cm™ melt atm ™!
Isochoric heat capacity, C,, Cy c
o r
TCC) calg™ calgmole™ calgatom™ —
oK—l oK—I oK—l Cv
500 0489 16 , 69, 115
600 043, 15 68, 11g
700 0475 15 4 67, 124

Sonic velocity
500°C
600°C
700°C

Thermal diffusivity
500°C
600°C
700°C

Kinematic viscosity
500°C
600°C
700°C

Prandtl number
500°C
600°C
700°C

M= 3420 m/sec
M= 3310 m/sec
M= 3200 m/sec

D =204 % 1073 cm?/sec
D=214% 1073 em?/sec
D=214 X 1073 cm?/sec

v="T744 % 1072 cm?/sec
V=434 X 1072 em?/sec
P=28¢ X 1072 cm?/sec

PI':356
PI=204
PI'=131

9Phystcal properties for the pure compound LiyBeF 4 approximate those for the MSRE flush and coolant salts within
limits of experimental uncertainty Unless otherwise noted, property values correspond to those reported in Physical Properties
of Molten Salt Reactor Fuel Coolant and Flush Salts S Cantor, ed , ORNL-TM-2316 (August 1968)

bg A Romberger and ] Braunstein MSR Program Semiannu Progr Rep Feb 28 1970 ORNL-4548,p 161

€Y H Burnsand E K Gordon, Acre Cryst 20, 135 (1966)
114

Table 8.2. Physical properties of the MSRE fuel salt

Property Value Estimated precision

Viscosity n(centipoises) = 0 116 exp [3755/T(°K)] 7
Thermal conductivity 0010 watt cm ™! °C™1 *10%
Electrical conductivity =_222+681 X 1073T(°C)a +10%
Liguidus temperature 434°C 3°C
Heat capacity

Liquid Cp =057 cal glec! 3%

Solid Cp=031+361X107*T(°C) calg™! °C™ +3%
Density

Liquid p=2575-513x 1074T(°C) 1%

=139 9 1b/ft> at 650°C

Expanswvity 214 X 1074/°C at 600°C +10%
Compressibility Br(°K)=23X 10712 exp [1 0 X 1073 T(°K)] cm?/dyne Factor 3
Vapor pressure log P(torrs) = 8 0 — 10,000/T(°K) Factor 50 from 500 to 700°C
Surface tension ¥ =260 — 0 12T(°C) dynes/cm +30, —10%
Solubility of He, K1, Xe TCC) He Kr Xe 2 Factor 10

Isocheric heat capacity, €,

Sonic velocity

500 66 013 003
600 106 0 55 017
700 151 17 067
800 201 44 20
X 1078 moles cm ™3 melt atm ™!

Cy
o Cp
TCC) calg™ calgmole™ cal g-atom™ R
ole OK—l °K'1 v
500 048 16 , 69, 11,
600 048, 159 68, 11g
700  047; 15 4 67, 12

500°C = 3420 m/sec
600°C u=3310 m/sec
700°C p= 3200 m/sec

Thermal diffusivity

500°C D =209 X 1073 cm?/sec
600°C D =214 X 1073 cm?/sec
700°C D =215 X 1073 cm?/sec

Kinematic viscosiy

500°C v=744 X 1072 cm?/sec
600°C v=4 34 X 1072 cm?/sec
700°C v=128¢ X 1072 cm?/sec

Prandtl number

500°C Pr=35 4
600°C Pr=20 4
700°C Pr=13,

2Applicable over the temperature range 530 to 650°C The value of electrical conductivity given here was estimated by
G D Robbins and 1s based on the assumption that ZrF4 and UF4 behave identically with ThF4, see G D Robbins and
A S Gallanter, MSR Program Serannu Progr Rep Aug 31 1970, ORNL-4548, p 159, 1b1d , ORNL 4622, p 101
18 1dentical with coolant salt, was charged into FD 2, the
amount of salt added between two level probes mdicated a
density of 124 5 1b/ft3 A denuty of 1209 Ib/ft3 was
computed from the pressure requured to hft salt from the dran
tank to the fuel loop Weigh cell indications of flush-salt
density, using the “hot” calibration factor, ranged from 123 2
to 131 5 1b/ft3 at 1200°F, the “cold” calibration factor would
have given 1184 to 126 3 Ib/ft> The data on coolant and
flush salt densities thus tend to support the “cold” calibration
factor for FD-2

The density of the fuel carner salt, LiF-BeF, Zrk4 (6530 5
mole %) was measured as the salt was being charged to FD 2
This measured density, computed from externally measured
weights and the volume between the level probes in FD-2, was
140 6 Ib/ft® at 1200°F Addition of all the uranum added
through run 3 would be expected to increase the density by
about 53 Ib/ft> Four measurements were made after the
uranum was added, using the weigh cells and the level probes
Densities based on the *“hot™ calibration of the weigh cells
ranged from 149 9 to 152 2 Ib/ft®, with an average of 151 0
1b/ft3 at 1200°F With the “cold” scale factor the average was
145 1 [o/ft3, very close to the expected density

Salt densities were computed on several occasions from the
change 1in weigh cell readings as the fuel loop was filled In every
case the computed density was less than given by other means,
suggesting that a full loop volume may not have been
transferred

The temperature coefficients of density for the salts were
computed from the change 1a salt level with loop temperature
Measured values of (Ap/p)}/AT were for the coolant salt, -1 06
X 1074 ¢! (average of three measurements), for the flush,

115 x 10™* (°F)7!, and for the fuelsalt, 109 x 107% and
~115%x 107% ¢F)™ (two measurements)

The bulk of the inventory data accumulated to date is on the
flush salt, because more transfer and fill and dramn operations
have been done with this salt Calculated inventories (using
“hot” scale factors) have ranged from 1 7% below to 2 6%
above the nominal or “book” mventory for no ascribable
reason

With continued experience 1t became evident that
on-site measurements of salt masses were less and less
reliable and would have little consequence in appraising
reactor performance because of 1naccuracies in the
welgh-cell measurements In the absence of such infor-
mation the amounts of salts delivered to the circulating
system or remaimng in the dramn tanks were approxi-
mated from values of the density of the salt mixtures
and the dimensions of the container vessels Efforts
were 1mtiated to appraise and improve, if feasible, the
accuracy and preciston of data pertaimning to the
densities of molten fluoride mixtures

Relatively few measurements of the densities of liquid
salt mixtures were made 1n the development of molten-
salt technology at ORNL prior to MSRE operations
Experimental values for the densities of a number of
Z1F 4-containing nuxtures were made as a part of the
Axrcraft Nuclear Propulsion program 2 Sinular measure-
ments, which employed the buoyancy principle, as did

115

the ORNL program, were made later with LiF-BeF,-
UF,; muxtures by the Mound Laboratory® under con-
tract work with ORNL Most density values for the
nuxtures used i the MSR program, however, were
estimated by the method of muxtures 2 This method
was believed to have an accuracy of approximately
5%

In an effort to obtain more accurate values for the
salt mixtures used m the MSRE, attempts were made to
develop new methods for the laboratory determination
of the densities of molten salt mixtures Employing a
new techmque, Sturm and Thoma# obtained density
measurements for the MSRE fuel and coolant salt
muxtures Measurements of the depth of salt in a
cylindrical container of accurately known dumensions,
which was contained in a controlled atmosphere glove
box, were obtained using an electrical probe attached to
a vernter caliper Contact of the probe with the melt
was indicated by completion of an electrical circuit
through the caliper and the melt Appropnate correc-
tions for thermal expansion at a particular temperature
were applied for the contaner and the probe Although
some shortcomings were evident in the procedure,
prncipally owing to the effects of small amounts of
liquid adhering to the probe, 1t afforded a direct
measurement of the densities of hiquids which were
visually observable during measurement The results of
the measurements were found to be in satisfactory
agreement with on-site measurements of the densities of
the salt stored mn the dran tanks > Laboratory efforts
were therefore discontinued Concurrently, a method
for esttmating densities which assumes additivity of
molar volumes was devised by Cantor © With continued
refinement, the method was developed to the extent
that 1ts accuracy was within 5% A comparison of the
densities of the MSRE fuel and coolant salts, as found
using the several methods described above, 1s given in
Table 8 3

On the basis of reactor physics analysis, 1t was
estimated that the average fraction of the fuel charge
circulated during power runs with the MSRE was 92%
This value, together with a value of the volume of the
crreuit computed from component dimensions, 71 3
ft*, and the mass of the circulated charge as deduced
from chemical and 1sotopic analyses (Table 2 6), indi-
cates that the density of the *2°U fuel at the beginning
of power operations was 139 01 1b/ft®, whereas the
value computed by Cantor’s method 1s 139 9 Ib/ft® In
more recent application of Cantor’s method to other
salt mixtures, excellent agreement has been found
between estimated and observed values Such confirma-
tion, along with 1ts apparent precision for MSRE fuel
116

Table 8.3. Density of MSRE salt mixtures at 650°C (1200°F)
p=a—bTCC)g/cm3

Density parameters p, density
Composition (mole %) Source
a b g/cm3 b/ft3
x107%
LiF-BeF, (66-34) Method of mixtures? 2.24 6 1.85 115.5
Mound Laboratory? 2.158 3.7 1.921 119.9
On-site estimate (see text) 1.89--2.11 118-132
Electrical probe 2.296 4.82 1.983 123.8
Molar volume estimate® 2.214 4.2 1.941 121.8
LiF-BeF,-ZrF 4 (64.7-30.1-5.2) Molar volume estimate® 2471 4.95 2.15 134.2
On-site estimate 2.25 140.6
LiF-BeF,-Z1F 4-UF,
(64.7-29.38-5.1-0.82) Method of mixtures? 2.61 7 2.15 134.2
On-site estimate 2.32--2.44 145-152
Electrical probe 2.848 7.69 2.35 146.6
Molar volume estimate® 2.575 5.13 2.24 1399

4S. I. Cohen and T. N. Jones, 4 Summary of Density Measurements on Molten Fluoride Mixtures and a Correlation for Predicting
Densities of Fluoride Mixtures, ORNL-1702 (July 1954, declassified Nov. 2, 1961).

bMound Laboratory report MLM-1086 (see ref. 3).
€S. Cantor, unpublished work, 1969.

salt, permits the inference that the most accurate values
of MSRE salt densities have been derived by this
method.

An important characteristic that easily tends to pass
unnoticed from attention is that Li,BeF, is unique
among the complex fluorides which have significance in
molten-salt reactor technology, in that it undergoes the
least volume change associated with the melting-freez-
ing transition of any of the compounds encountered.
This value has not been measured directly but can be
estimated by assuming that the linear coefficients of
thermal expansion for Li,BeF, and LiF are similar
enough to be used interchangeably. If this assumption is
valid, the density of Li,BeF,; is 2.064 g/em® at the
melting point, about 6% less than at room temperature.
The density of liquid of the stoichiometric composition
at the melting point, as indicated by the density
expression in Table 8.1 is 2.021 g/cm®. These two
values indicate that, on melting Li, BeF,, the density is
reduced only about 2.07%.

It is important to recognize that the MSRE fuel,
coolant, and flush salts were nearly of the composition
Li, BelF,;, and to some extent the freedom that the
MSRE showed from difficulties with freeze valves and
from distortion of the radiator tubes under off-specifi-
cation cooling conditions”? is due to the very small
volume change this compound undergoes at the melting
point.

8.3 Crystallization of the MSRE Fuel

Laboratory studies of fluoride mixtures were carried
on for some years before the MSRE was operated. In
these studies, the crystallization behavior of the LiF-
BeF,-ZrF4-UF, (65-29.1-5-0.9 mole %) mixture was
established to approximate that of the MSRE fuel salt
when it operated with *3*U fuel. Crystallization was
found to follow the equilibrium crystallization se-
quence described below:

On cooling the liquid mixture to 434°C, crystalline
Li, BeF, is formed. This phase continues to precipitate
on further cooling, and at 431°C, Li, ZrF, begins to
crystallize; the onset of crystallization by the tertiary
phase, LiUF;, begins at 416°C. The liquid portion of
the mixture decreases but is present down to tempera-
tures as low as ~350°C. When completely frozen, the
fuel at equilibrium should be composed of crystalline
Li, BeF,, Li,ZrF4, LiUF;, and BeF, in volume frac-
tions of 0.735, 0.204, 0.032, and 0.029 respectively.
The usual cooling paths for LiF-BeF,-ZrF,-UF,; mix-
tures of compositions similar to the MSRE fuel involve
a glassing of the BeF,-rich liquids that are low melting
and, thus, the last to freeze. Hence, crystals of pure
BeF, are generally not expected; rather, the last liquid
solidifies as glass which incorporates variable amounts
of the phases listed above with BeF, .
117

PHOTO ©67747AR

INCH

POSITION wt % U

2.94
35.49
4.5

3.32
4.28
4.62
5.74

~N D WM e

Fig. 8.1. MSRE fuel ingot resulting from slower cooling rate.
The homogeneity generally characteristic of multi-
component salt mixtures in the liquid state is progress-
ively destroyed as the mixture undergoes gradual
crystallization. The possibility that the MSRE fuel
mixture, LiF-BeF, -ZrF,-UF, (65-29.1-5.0-0.9 mole %),
might, on cooling in the MSRE drain tanks, experience
sufficient solid-phase fractionation to create potentially
hazardous conditions was examined in laboratory-scale
experiments. Some 650 g of simulated fuel salt was
cooled at rates approximating that expected of the
entire drain tank assembly and that expected of the fuel
alone, 3.46 and 0.387°C/hr respectively. In both cases
the radiative cooling geometry was controlled to simu-
late as nearly as possible that expected in the drain
tanks, even though it was realized that the horizontal
AT profile in the cooling radioactive fuel mixture
would probably be substantially different from that
prevailing in the laboratory experiment.

The concentrations of uranium were found to be
identical at the top and bottom fractions of each of the
two ingots, irrespective of cooling rate. A photograph
of the ingot resulting from the slower cooling rate
experiments is shown in Fig. 8.1. Chemical analyses of
the salt specimens from each of the locations designated
in Fig. 8.1 were obtained. For areas 1 to 7, the uranium
concentrations were found to be 2.94, 3.49,4.15,3.32,
4.28, 4.62, and 5.74 wt %. In comparison with the
nominal concentration of uranium in the MSRE fuel,
5.13 wt %, these experiments show a maximum increase
of 23.4% in uranium concentration on very slow static
cooling. A fact which accounts for the small degree of
segregation of the uranium phases in these experiments
is that, at the onset of crystallization of LiUFs,
simultaneous crystallization of the three solid phases
Li, BeF,, Liy ZrF4, and LiUF; takes place. In addition,
the volume of the liquid phase is being reduced steadily,
so sharply, in fact, that in this experiment as well as in
similar previous ones, some of the liquid phase was
apparently occluded among dendritic-like crystals of
the solidified phase, a phenomenon which helps prevent
compositional variation in the mixture.

In the freezing of multicomponent mixtures, maxi-
mum segregation of crystalline phases takes place under
equilibrium cooling conditions. The segregation repre-
sented by the results obtained in the fractionation
experiments described here represents, in a practical
way, the nearest approach to equilibrium cooling that
the MSRE fuel salt may experience in a single crystalli-
zation sequence.

The crystallization behavior described above pertains
exclusively to the fuel-salt mixtures used in the MSRE
and is, of course, quite unlike that for fuel mixtures

118

proposed for use in a molten-salt breeder reactor.
Crystallization equilibria similar to those for the
breeder fuel are described elsewhere by Thoma and
Ricci.®

References

1. P. N. Haubenreich et al., MSR Program Semiannu.
Progr. Rep. Aug. 31, 1965, ORNL-3872, p. 30; see also
Sect. 2.4.2 of this report.

2. S. I. Cohen and T. N. Jones, A Summary of
Density Measurements on Molten Fluoride Mixtures
and a Correlation for Predicting Densities of Fluoride
Mixtures, ORNL-1702 (July 19, 1954, declassified Nov.
2, 1961).

3. Density and Viscosity of Fused Mixtures of
Lithium, Beryllium and Uranium Fluorides, Mound
Laboratory report MLM-1086 (December 1956).

4. B. J. Sturm and R. E. Thoma, Reactor Chem. Div.
Annu. Progr. Rep. Jan. 31, 1965, ORNL-3789, p. 33.

5. P. N. Haubenreich, private communication, Aug.
16, 1965.

6. S. Cantor, Reactor Chem. Div. Annu. Progr. Rep.
Jan. 31, 1962, ORNL-3262, p. 38.

7. T. L. Hudson, C. H. Gabbard, and D. M. Richard-
son, MSR Program Semiannu. Progr. Rep. Aug. 31,
1966, ORNL-4119, p. 41.

8. R. E. Thoma and J. E. Ricci, Fractional Crystalli-
zation in the System LiF-BeF,-ThF,;, ORNL-TM-2596
(July 1969).

9. INTERACTIONS OF FUEL SALT WITH
MODERATOR GRAPHITE AND SURVEILLANCE
SAMPLE MATERIALS

A new method of in-situ analysis for lithium, beryl-
lium, and fluorine was invented by Macklin, Gibbons,
and Handley' in response to our appeal for their
assistance in examination of graphite specimens re-
moved from the MSRE core. The method they devised
employed proton bombardment of target nuclides and
measurement of the yields of neutrons and gamma rays
produced. It was found to be applicable for measure-
ment of the concentration of target nuclides in graphite
at the few-parts-per-million level as a means to deter-
mine the extent to which MSRE fuel salt components
penetrated into the graphite under radiation. This
method has the advantage of applicability to matrix-
dispersed samples and in the presence of considerable
radioactivity from fission products, features which
make it well suited to the examination of MSRE
graphite.
The results obtained by these workers were reported
in detail previously.?*® There remain several puzzling
aspects of their findings, and because of this the results
are reviewed here.

Data were obtained from three samples, (1) a control
sample (Y-S5 from CGB bar 635) exposed to nonradio-
active salt, (2) sample Y-7, removed from the reactor
core after 33,400 MWhr of exposure in May 1967, and
(3) sample X 13, removed March 25, 1968, after 66,637
MWhr of exposure.

Specimens were moved across a beam of 2.06-MeV
protons collimated through a slit of 0.0075 cm width at
the ORNL 3-MV Van de Graaff accelerator. Measure-
ment of the resulting prompt gamma rays from
YE(p,oy)' °O showed that fluorine varied in sample
X-13 from 350 ppm near the surface to 60 ppm at the
center (Fig. 9.1). The observed ratio of fluorine to
lithium was near that characteristic of the MSRE fuel
(Fig. 9.2). However, as shown in Figs. 9.1 and 9.3, the
lithium and fluorine did not show a simple dependence
on depth.

Results of the examination of sample X-13 suggested
that much of the Li and F came from bulk intrusion, a
finding which differed from that resulting from the
previous examination of sample Y-7 (Fig. 9.4), where

119

the Li/F ratio became increasingly higher at greater
distances from the surface. Equally puzzling is that
comparisons with the results of the analysis for 2?°U
by Kirslis and Blankenship®* (see Figs. 9.1 and 9.3)
showed that the relative concentrations of F and 2°%U
became steadily divergent with penetration depth and
thus seemed to rule out bulk salt intrusion as a
mechanism. Further, in the absence of radiation, a
control specimen showed less penetration of both salt
and uranium by a factor of 100. The possibility was
considered that introduction of a fuel aerosol from the
gas phase, as was suggested to rationalize the intrusion
of certain fission products,®* was responsible for fuel
having penetrated the graphite voids as an aerosol. The
cause of this phenomenon is still not resolved.

The uranium profiles shown in Figs. 9.1 and 9.3 were
derived in the investigation by Kirslis and Blankenship,
in which it was found, from delayed neutron activation
analysis, that the uranium concentration profile ranges
from values as high as 100 ppm at the surface to 6.5
ppm at 50 mils depth, but “in most cases, the total
range of variation of the **5U concentration in 50 mils
penetration was less than two orders of magnitude. This
moderate slope indicates a higher mobility for uranium
in graphite than for most fission products [but] ...
represents only 1 g of 23U per 1000 kg of graphite.”*

ORNL-DWG 68-14530

Ak T T
R N O FIRST SURFACE TO CENTER -
- & SECOND SURFACE TO CENTER
500 e — @ 2350 IN SIMILAR SPECIMEN
e ] Pl
s —raA
] \\ | i - ]
1--~A—~CL—1 N }fi ]
e X‘fi\ “\ A A ‘;}3.\\ \ V ]
N /B s | o
Ty 1 P
+od /o\ jfi ! \ k"\
c — e s ] A, O -1 - 4
g. 100 - 17 N ] | By a. ab U F
o | A b Y !
= - A1t
T il v:oqw |
\ ‘I’ ko) _Q‘59|: S
50 ] H ¥
~ W Y
~d
PO N
{ 5&\ e - L | I
M 235U
20 N IR Ly
\
@ \\
0 —— |——1+1 - NS - I e
] i ~L _
i g N
5 i i
1 10 100

DISTANCE FROM SURFACE {mils)

Fig. 9.1. Fluorine concentration as a function of distance from the surface.
120

ORNL-DWG 68—14532

50
Al
o FIRST SURFACE TO CENTER
@ SECOND SURFACE TO CENTER|]
20 g
[
C
o 10 Q
= RATIC FOR
o i ; o 4 FUEL SALT
¥ .
= e | @ ®goo0 | ® )
E 5 @ O [o) : L
) c08
- o
» RATIO FOR LiF
1
1 2 5 i0 20 50 100 200 500 1000
DISTANCE FROM SURFACE ( mils)
Fig. 9.2. Mass concentration ratio, F/Li, vs depth.
ORNL-DWG 68-14534
100 o —— 1000
o FIRST SURFACE TO CENTER
™ & SECOND SURFACE TO CENTER
50 \\\ e 2354 | SIMILAR SPECIMEN 500
<
S B e ;jl \b
20 R AR 200
H )—4\& ! J 9 'CrC\l
~ J‘d 8‘5 / g}&%@g
- RO ~ 7, -
£ ? N Li £
a {0 100 o
e o b N 2
<
3 235U»\ N
5 - 50
Y
H-—i
A\A‘A-A ‘,4‘
o™ \,g‘\
2 Hs - 20
Bapi
1 10
1 10 400 1000

DISTANCE FROM SURFACE (mils)

Fig. 9.3. Lithium concentration as a function of distance from the surface.

Although the results of the proton reaction analysis
described by Macklin and co-workers do not seem to
permit quantitative generalization, they do suggest
several points of significance to further development of
molten-salt reactor technology:

1. The penetration of the graphite moderator by
lithium and fluorine appears to be real, if not
massive. Only samples of CGB graphite were ex-
amined in the experiments performed by Macklin et
al. Whether similar penetrations of more dense or

coated graphites can be expected for improved
MSBR graphites is not estimable from the current
results.

Effects of salt-gas-graphite interfacial properties on
transport of lithium and fluorine through films is
unknown at present.

. The current data were derived from a fuel salt with a
composition, LiF-BeF,-ZrF,-UF, (65-29.2-5.0-0.8
mole %), whose effect on transport conceivably may
differ from that for the MSBR. Whether the trans-
121

QRNL-DWG 68~ 14533

1000 — | —
] sl e
[ H
500 — A b —
Al 1 I
e I \‘T‘\ J_I, Lj_l W‘ —
z ™
200 — N\ | | SN ! ’ =
L | SAMPLE X~13
100 F— J . DN -
[ - — \\, _—
—_ — L S
3\‘\ _ RN M~ ]
—- 50 _ N I
E 1 1IN L ]
8 N
2 — \ 1 . - _
ful |
20 |— - e
SAMPLE Y -7
s\\
10 —a N -
T N ]
[ | _S\ i OO S |
5k - N T . .
I 4:;,; N ~
| :
2 | S — ? I 3 S— — -
| ; g t ‘ w \
| | |
1
1 2 5 0w 20 50 100 200 500

DISTANCE FROM SURFACE (mils}

Fig. 9.4. Comparison of fluorine concentrations in samples
Y-7 and X-13, a smooth line having been drawn through the
data points.

port would be markedly affected by the large
heavy-metal concentration of the MSBR fuel, con-
taining a total of 12 mole % ThF4 + UF,, is of
justifiable curiosity.

The results obtained indicated with 2?°U fuel salt
that there are a number of parameters whose quanti-
tative relationship to the transport of salt species to
moderator graphites should be examined by further
examination of graphite specimens, especially those
removed from the MSRE core after completion of the
2337 experiments. Several notable differences between
2337 and *?°U operations existed, including for 233U
operation a uranium concentration less than 20% of
that with 22°U, a continuously higher void fraction in
the fuel salt, and variable differences in the gas-salt
interfacial tension. The latter two of these factors were
regarded as possibly conducive to enhanced transfer of
materials to the graphite moderator and prompted
further examination of graphite specimens from the
MSRE.

In postoperational examinations of a graphite stringer
from the MSRE core vessel, Kirslis and Blankenship®
obtained spectrochemical and delayed neutron analyses
of the graphite milled from the surface of the stringer.
Analyses by the proton bombardment method were not
performed. The analytical resulis that were obtained
were in partial agreement with the earlier results of
Macklin and co-workers and indicated bulk penetration

of the fuel salt, probably via cracks in the graphite, to a
depth of ~2 mils. In contrast, however, they did not
indicate penetration to greater depths. Although the
results of the two groups of analyses are in good
agreement with respect to penetration of the outermost
layers, Kirslis and Blankenship did not find anomalous
divergence in the relative concentration of any of the
components, nor did they find that uranium pene-
tration had occurred to greater depths. The absence of
detectable amounts of uranium in the graphite at
depths where uranium had been detected with samples
from ?3* U tests tends to support a model of bulk salt
penetration, since the 2?*U concentration of the fuel
was less than 20% of that of the >*3 U fuel.

Applications of new and increasingly sensitive
methods of analysis, such as the proton bombardment
method, in molten-salt research and development
programs are significant to the continued development
of the technology. Continued refinement and adapta-
tion of this new method are highly desirable within the
framework of future research programs for possible
application to MSBR development.

References

1. R. L. Macklin, J. H. Gibbons, and T. H. Handley,
Proton Reaction Analysis for Lithium and Fluorine in
Graphite Using a Slit Scanning Technigue, ORNL-
TM-2238 (July 1968).

2. R. L. Macklin et al.,, MSR Program Semiannu.
Progr. Rep. Feb. 9, 1968, ORNL-4254, p. 119.

3. R. L. Macklin et al., MSR Program Semiannu.
Progr. Rep. Aug. 31, 1969, ORNL-4344, p. 146.

4. S. 8. Kirslis and F. F. Blankenship, MSR Program
Semiannu. Progr. Rep. Aug. 31, 1969, ORNL-4344, p.
115.

5. 8. S. Kirslis and F. F. Blankenship, MSR Program
Semiannu. Progr. Rep. Feb. 28, 1971, ORNL-4676, p.
73.

10. CHEMICAL SURVEILLANCE OF AUXILIARY
FLUID SYSTEMS

10.1 Water Systems

Potable water from the ORNL distribution system
was supplied to the MSRE for a variety of uses as
described in the MSRE Design and Operations Report:!

After passing through a backflow preventer, the water is used
in the liquid waste system, in the vapor-condensing system, for
general cleanup of equipment, as makeup for the cooling tower
water system, and for cooling of the charcoal beds. Two
520-gpm centrifugal pumps are provided for circulating cooling
tower water, which is cooled by a two-fan induced draft cooling
tower. The cooling tower water is used for air compressors, air
conditioners, in the chemical plant, for the lube-oil systems, in
the charcoal beds, and for condensing steam from the drain
tank steam domes. (Process water can also be used for this.)
Cooling tower water is also used in a shell-and-tube heat
exchanger to provide cooling for the treated water system. Two
230-gpm centrifugal pumps circulate treated water in a closed
loop to cool in cell components. Makeup water is supplied by
condensing building steam in a shell and tube heat exchanger
using cooling tower water as the coolant, Treated water is also
used to fill the Nuclear Instrument peneiration. This water is
continuously recirculated through a closed loop by a 5-gpm
pump to maintain uniformity of the water condition through-
out the penetration. In order to minimize corrosion, potassium
tetraborate and potassium nitrite were added to the water
supply and maintained at concentrations of 500 and 1500 ppm,
respectively. Steam condensate is also used io supply water to
the feedwater tanks. This untreated water is used in the drain
tank bayonets to remove decay heat from the reactor fuel afier
the reactor has been drained.

The several water systems at the MSRE were sampled
periodically and analyzed to determine corrosion rates,
buildup of contaminants, loss of corrosion-inhibiting
chemicals, etc. Water samples were submitted regularly
for chemical analysis in the General Analysis Labo-
ratory. At more frequent intervals, on-site tests were
performed to give continuity to the control data and to
reduce dependency on analyses from the laboratory.
Details of sampling procedures, methods of additions of
corrosion inhibitors, and of on-ite testing procedures
are given in the MSRE Design and Operations Report,
Sect. 6-1.1

10.1.1 Cooling tower water. Samples of the cooling
tower water were tested on-site daily to check the
concentration of its chromate ion, which served as a
corrosion inhibitor, pH, and hardness. Samples were
submitted on a weekly basis for laboratory analysis,
where more complete analyses were performed. In use,
the cooling tower water supply was maintained in the
pH range 7 to 8, concentration of chromate ion at 30 to
50 ppm, hardness <250 ppm {as CaCO3), and Fe <0.03
ppm. The results of laboratory analyses with cooling
tower water samples are listed in Table 10.1.

10.1.2 Treated water supply. A supply of treated
water was circulated within a closed system to remove
heat from the thermal shield and other equipment in
the reactor and drain cells. A program of water system
surveillance was incorporated into the procedures em-
ployed at the MSRE to minimize the probability that
operational difficulties would arise from this auxiliary
system. Since standardized methods for control of the
water chemistry of auxiliary systems such as those
which existed in the MSRE were fairly well stand-

122

ardized, it was anticipated that surveillance of the water
supply would require only cursory attention. However,
when power was first generated with the MSRE, it was
found that the chemicals used in the treated water
system had become activated. The source of the activity
was identified as *2K, produced from the potassium
nitrite—potassium tetraborate mixture used to inhibit
corrosion in the system. Extrapolation of the radiation
levels observed at low power indicated that if the
reactor were operated for extended periods at full
power, equilibrium radiation levels of about 400 mR/hr
would prevail in the heat exchanger in the diesel
equipment room and in the water control room.

As the alternative, the potassium-containing inhibitors
were discarded and replaced with an analogous lithium-
based mixture, highly enriched in the 7Li isotope to
minimize tritium production. Lithium nitrite was pre-
pared commercially by ion exchange from potassium
nitrite and lithium hydroxide for this use. After the
4000-gal ireated-water system was diluted with de-
mineralized water to reduce the potassium from 800 to
3 ppm, the desired inhibitor concentration was attained
by adding " Li nitrite, boric acid, and 7 Li hydroxide.

When the reactor was next operated at power, in run
5, activation again occurred, this time caused by the
presence of sodium, ~1 ppm of which was present in
the demineralized water from the ORNL facility.
Condensate was produced at the MSRE with less than
0.1 ppm of sodium and used to dilute the sodium in the
treated-water system to 0.3 ppm. The concentrations of
sodium and ®Li were thereafter reduced to sufficiently
low concentrations that shielding and zero-leakage
containment of the water system were not required.

Criteria were established for the treated water after
replacement of the inhibitor mixture to meet the
following specifications: pH: 7.0 to 9.0, Na* <3 ppm,
NO, = 760 to 860 ppm, B> 57 ppm, K™ <8 ppm, AAl
= 0 ppm, Fe = 0 ppm, tritium < 2.2 X 10° dis min™
cm™3, AZ hardness (as CaCO;) = 0 ppm.

Chemical analyses of treated-water samples were
obtained at frequent intervals in order to maintain these
specifications and as guides for adjustments of the
concentrations of chemicals in the treated-water system
supply. On-site tests were also performed with greater
frequency than laboratory analyses for control of
operations, with supplementary analyses obtained from
the Analytical Chemistry laboratory at approximately
biweekly intervals. The results of the chemical analyses
obtained for treated-water samples obtained from the
circulating system are listed in Table 10.2. Those
obtained from the nuclear instrument penetration are
listed in Table 10.3.

N
The results listed in Tables 10.2 and 10.3 do not
warrant detailed comment or interpretation because
they were used primarily as guides for controlling the
concentrations of the corrosion inhibitors. It is evident
from the results listed that they served satisfactorily for
this purpose, and that the treated-water systems were
essentially free of operational difficulties.

10.1.3 Vapor condensing system. Among the water
supplies which were subjected to routine chemical

123

surveillance was included in the reservoir of the vapor
condensing system. This system was incorporated into
the MSRE for service only under the accident con-
ditions described in the MSRE Design and Operations
Report, Part 1.2 1If an accident occurred in which hot
fuel salt and the water used to cool equipment inside
the cells became mixed, it would be necessary to
contain within the MSRE the steam generated by the
accident. A vapor condensing system, shown sche-

Table 10.1. Results of chemical analysis of MSRE cooling tower water

CTW Date Total CrOh_ pH Te CTW Date Total cro, pH Fe
Sample Hardness Sample Hardpess

No. (ppm) No. (ppm)

8 6/18/65 214 39 8.53 <2, 456 12/2/66 178 12 8.00 < 0.3
9 6/25765 185 13 8.63 <2.0 458 1274766 251 68 8.00 <1.,0
10 7/2/65 284 39 8.70 <2.0 467 12/12/66 197 4 0.3
11 7/9/65 133 141 8.10 - 473 12/18/66 176 14 - <2.0
12 7/23/65 105 21 9,58 <1.0 482 12/27/66 187 28 817 0.3
13 7/30/65 105 30 7.97 <1.0 494 1/8/67 178 16 8.37 0.5
14 8/6/65 185 7z 7.80 <1.0 502 1/15/67 175 18 - 0.3
15 8/23/65 194 - 8.51 <0.3 512 1/30/67 171 9 8.10 0.3
16 9/4/65 248 - 6.90 4.0 516 2/5/67 165 11 8.78 0.3
18 9/17/65 3.0 37 8,77 1.0 524 2/13/67 171 - - 0.3
26 9/24/65 223 25 8.60 - 532 2721767 161 26 8.02 0.3
32 10/8/65 111 44 8.00 0.5 538 2/27/67 175 17 8.10 <0.3
33 10/15/65 30 32 8.10 0.5 546 3/6/67 231 27 8.77 0.3
48 11/10/65 121 14 8.08 0.3 559 3/19/67 223 79 8,36 0.3
55 11/17/65 141 15 8.22 0.2 575 413167 - 23 8.57 0.3
67 11/29/65 161 16 7.0 <0.2 581 418767 342 36 8,60 0.3
79 12/10/65 129 29 7.37 - 588 4/18/67 241, - 8.30 <1.0
83 12/15/65 137 21 8.0 <0.3 596 4/24/67 239 29 8.50 <1.0
97 12/28/65 143 - 7.80 <0.3 603 5/2/67 235 26 8.47 <0.3
105 1/5/66 - - 8.32 <G.3 607 5/7/67 286 66 - 0.5
109 1/9/66 157 20 8.17 0.5 615 5/15/67 193 50 8.30 <0.3
118 1/19/66 151 17 8.30 <0.3 622 5/22/67 143 54 8,07 0.3
123 1/23/66 153 17 8.25 <0.3 630 5/30/67 152 53 8.00 <0.3
142 2/12/66 150 15 8.35 <0.3 634 6/4/67 132 59 7.77 0.5
145 2/13/66 169 13 - <0.5 641 6/12/67 155 44 7.93 <0.3
162 2/28/66 262 42 8.40 <0.5 655 6/26/67 217 108 8,31 0.3
164 3/3/66 128 - 8.20 <0.5 661 7/2/67 - 71 8.33 <0.3
172 3712766 153 17 8,07 <0.3 669 7/10/67 245 58 7.3 €.3
181 3/19/66 151 17 8.38 <0.3 676 7/16/67 260 58 8.0 <0.3
186 3/23/66 186 1% 7.90 <0.3 689 7/30/69 149 - - <0.3
189 3/27/66 156 19 8.20 <0.3 697 8/7/67 152 68 8.54 0.3
209 4/18/66 179 17 8.14 <0.5 704 8/14/67 214 54 8.40 0.35
219 4/27/66 - 17 - <0.3 711 8/21767 176 42 8.51 <0.2
224 5/1/66 178 20 8.4 <0.3 721 8/31/67 199 67 8.22 <0.3
232 5/8/66 194 21 8.42 <0.5 725 9/3/67 155 40 8.44 <0.3
240 5/15/66 - 18 8,32 <0.3 734 9/12/67 212 56 8.53 <0.4
248 5/22/66 200 20 8.28 <0.3 740 9/18/67 248 71 8.50 <0.4
256 5/30/66 198 15 8.22 <0.3 748 9/25/67 248 71 8.38 1.1
262 6/5/66 187 14 8.24 <0.3 756 10/2/67 268 71 8.55 0.62
269 6/12/66 183 21 8.08 <0.3 763 10/9/67 247 56 8,40 <0,4
277 6/19/66 212 20 8,45 <0, 3 769 10/16/67 276 129 8.135 <0.3
285 6/26/66 218 17 8.37 <0.3 776 10/23/67 250 62 6.80 <0.2
292 7/2/66 210 8 8.55 <0.3 783 10/30/67 228 72 8.10 <0.4
300 7/10/66 216 20 8.40 <0.3 790 11/6/67 205 58 8.0 <0.5
307 7/17/66 203 20 8.05 <0.3 795 11/13/67 213 56 8.3 <0.5
315 7/23/66 156 20 8.38 <0.3 800 11/18/67 215 58 8.46 <0.5
317 7/24/66 172 17 8,46 <0.3 810 11/27/67 54 67 8.50 <0.2
325 7/31/66 20 20 - <0.3 824 12/10/68 273 67 8.40 <0.4
332 8/6/66 126 15 7.71 <1.0 844 1/1/68 -~ - 8.24 <0.2
340 8/15/66 129 32 - <0.3 921 3/30/68 168 60 8.12 <0.4
346 8/21/66 130 21 7.70 <0.3 928 4/8/68 151 69 7.48 <2.0
354 8/29/66 121 13 - - 964 6/21/68 160 100 8.98 0.1
360 9/5/66 127 17 8.10 <0.4 992 6/30/68 150 76 8.24 <0.4
366 9/11/66 260 90 - <0.3 1027 8/4/68 193 47 8.44 <0.3
375 9/19/66 145 12 8.20 <0.3 1186 1/8/69 177 60 8.05 <0.1
383 9/25/66 158 11 8.20 <0.3 1241 3/2/69 409 155 8.50 <0.3
392 10/3/66 - 18 8.50 <0.,4 1274 4/6/69 186 36 8.32 <0.1
414 10/24/66 294 30 - <0.5 1359 7/6/69 209 65 7.76 0.12
422 10/30/66 210 17 8.57 - 1385 8/3/69 228 112 7.57 0.08
433 11/10/66 185 14 8.70 <2.,0 1413 8/31/69 275 76 8.06 7.0
438 11/14/66 198 16 8.70 <0.3 1491 10/9/69 270 46 8.59 <0.1
b44 11/20/66 187 13 - <0.3 1503 10/20/69 493 98 8.82 <0.2
452 11/28/66 176 21 8.16 <1.0 1539 12/4/69 225 71 8.32 <0.1

matically in Fig. 10.1, was provided to prevent steam
pressure from rising above the 40 psig allowable
pressure for the cells and to retain the noncondensable
gases. The vapor condensing tank in this sysiem was
kept about two-thirds full of potable water at all times
from 1964 to 1969.

124

Corrosion inhibition in the water reservoir tank was
effected by addition of potassium nitrite and potassium
tetraborate, as in the recirculated water systems. The
water supply in the water reservoir tank was to meet
the same specifications as the circulating systems.

Table 10.2. Chemical analysis of treated-water samples from the MSRE circulating system

Date Treated pH N02 B Al Fe L Hardness Trit/:ium
Water dpm/cm
Sample No. (ppm)

3/2/66 129 9.45 720 60 <1.0 <0.5 -

3/7/66 132 9,40 860 61 <1.0 - -

3/17/66 137 9. 40 760 57 <1.0 <0.3 15

3/19/66 138 8.98 760 57 <0.5 <0.3 20

3/19/66 139 9.40 750 57 <1.0 <0.3 20

3/23/66 141 9.30 740 56 - <0.3 19

4/3/66 145 9.50 710 56 <1.0 <0.5 27

4/3/66 146 8.45 740 57 <1.0 <0.3 25

4/18/66 149 9.33 630 56 <1.0 <0.5 22

4/22/66 152 9.34 670 70 - - -

5/1/66 156 9,28 600 70 <1.0 <0.3 25

5/10/66 158 9.27 590 60 <1,0 <0.5 28

5/16/66 169 9.17 460 67 <1.0 <0.3 25 3
5/16/66 170 - - - - - - 9.04 x 10;
5/27/66 179 - - - - - - 1.72 x 10
5/27/66 180 9.07 7.60 - - - -

6/5/66 187 9.10 720 64 <1.0 <0.3 26 4
7/3/66 208 9.20 690 64 <1.0 <0.3 26 1.96 x 10
9/7/66 224 9.38 520 42 - <0.3

9/19/66 227 9.00 540 L4 <1.0 <0.3 14

10/3/66 233 10.10 500 51 13 <0.3 7

10/5/66 235 10.10 1204 51 15 <0.3 12

10/30/66 244 9.62 560 68 22 <0.3 16 A
11/10/66 245 9.50 740 69 26 <0.2 26 1.4 x 10
12/4/66 250 9.40 1020 67 12 <1.0 18

1/24/67 260 9.20 860 58 49 <0.3 20

2/5/67 263 9.17 810 59 1.0 <0.3 16 A
3/6/67 268 9.07 840 58 31 <0.3 16 8.07 x 19
3/9/67 271 9.14 732 58 5 <0.3 17 1.1 x 10
4/3/67 274 9,07 715 58 27 <0.3 19 7.85 x 104
5/7/67 279 9.17 640 73 4 <0.3 14 1,26 x 109
5/9/67 280 9,27 790 71 <1.0 <0.5 13 1.48 x 103
6/4/67 283 9.08 752 70 0.5 <0.5 18 1.76 x 103
6/14/67 285 9,07 900 72 1.0 <0.3 13 1.37 x 102
6/26/67 286 9.18 840 66 1.0 <0.3 - 1.33 x 107
7/2/67 287 9.17 890 65 <0.4 <0.3 13 1.42 % 103
7/5/67 288 9.18 850 65 0.5 <0.3 12 1,33 x 10°
8/7/67 294 9.20 810 64 0.8 <0.3 8 5.0 x 104
9/3/67 298 8.81 820 69 <0.4 <0.4 17 1.57 x 107
10/2/67 302 9.25 790 63 <0.4 <0.5 21 1.59 x 10°
1/1/68 312 9.20 850 60 <0.2 0.5 10 2.23 x 102
3/30/68 320 9,12 800 60 <0.5 <0.4 9 2.44 x 10°
6/2/68 327 9,22 760 58 <1.0 0.3 10 2.32 x 102
6/30/68 331 9.34 780 55 <0.5 <0. 4 9 2.47 x 105
8/4/68 333 9.23 730 56 0.18 <0.1 10 1.11 x 10°
1/7/69 343 9.08 746 45 <0.2 0.12 10

3/2/69 349 9.14 700 b7 0.6 <0.3 10 2.26 x 10°
4/6/69 352 9.16 800 67 0.4 0.2 10 2.57 x 103
7/6/69 355 8.84 758 71 <0.4 <0.3 10 2.85 x 10°
8/3/69 357 9.10 742 70 0.73 <0.3 - 2.83 x 102
8/31/69 359 9.19 765 70 0.74 <0.3 9 2,82 x 109
10/9/69 361 9.15 730 69 1.0 <0.1 10 4.41 x 109
10/19/69 362 9.18 686 67 <0.2 <0.2 10 3.10 x 10°
11/4/69 363 9.18 723 68 <0.2 <0.2 8 2.98 x 10°
12/4/69 365 9.17 660 69 <0.7 <0.1 9 3.00 x 109

125

Table 16.3. Chemical analyses of treated water from the MSRE instrument shaft

Date Treated pH NO, B8 Al Fe L Hardness Tritium

Water dpm/cm
Sample No.

9/24/65 18 9.71 120 46 2 <0.5 4

10/8/65 20 9.51 1250 60 2 0.5 8

10/15/65 21 9.47 1240 60 2 0.5 9

11/10/65 22 9.60 1110 58 5 0.3 6

11/17/65 23 9.55 1110 59 3 0.2 9

11/29/65 24 9.52 - 59 2 <0,2 10

12/15/65 28 9.50 1170 59 2 <0.3 35

12/28/65 33 9.50 1050 60 2 <0.3 6

1/5/66 36 9.50 1060 58 1 <0.3 -

1/9/66 37 9.53 1100 59 2 0.5 9

1/15/66 39 9.18 1100 91 2 <0.3 11

1/23/66 42 9.15 1280 91 1 <0.3 -

2/16/66 48 9.22 740 61 1 <0.3 12

3/2/66 53 9.25 720 60 1 <0.5 20

3/7/66 56 9.20 860 60 1 - -

3/12/66 58 9.07 830 58 <1 <0.3 16

3/19/66 61 9.15 810 58 1.0 <0.3 11

3/23/66 63 9.12 820 60 - <0.3 13

3/23/66 64 9.25 820 61 1.0 <0.3 16

4/8/66 67 9.16 820 62 1.0 <0.5 13

5/1/66 69 9.10 700 62 <1.0 0.3 14

5/10/66 71 9.08 700 61 <1.0 <0.5 20

7/3/66 79 8.90 600 65 <1.0  <0.3 17 5

7/31/66 87 8.90 800 60 <1.0 <0.3 20

8/29/66 91 8.70 760 62 10 <0.3 18

9/5/66 92 8.75 770 60 11 <0.4 19

10/3/66 97 8.82 780 61 11 - -

10/30/66 103 8.84 670 61 11 <0.3 21 9.

11/1/66 105 8.85 920 - 23 <2.0 27

12/4/66 109 8.82 830 62 11 <1.0 19

12/31/66 115 8.26 750 63 <0.5 <0.3 19

2/5/67 120 8.78 690 60 1.0 <0.3 20

3/6/67 125 8.96 830 72 8 <0.3 18 2. X

3/19/67 128 8.78 878 71 4 <0.3 20 2. X

4/3/67 131 8.76 822 72 3 <0.3 20 3. X

4f/16/67 132 8.75 840 71 <1 <1 18 3. b

5/7/67 134 8.73 690 72 4.0 <0, 3 17 5. X

5/9/67 135 €.70 740 71 <1.0 <0.5 15 5. X

6/4/67 137 8.67 780 68 0.5 0.5 22 7. X

7/2/67 139 8.68 830 70 0.5 <0.3 16

7/24/67 142 8.78 790 70 0.8 0.3 49 1. X

8/7/67 147 8.74 860 69 0.8 <0.3 11 1. X

9/3/67 151 8.81 820 69 <0.4 <0.4 17 5.36 x

10/2/67 155 8.80 890 67 0.5 0.9 21 4, X

11/6/67 160 8.70 895 69 <1 0.7 16 6. x

1/1/68 164 8.81 1060 68 0.6 0.4 16 6. X

3/30/68 173 8.73 1070 67 1.0 <0.4 13 8. X

6/2/68 179 8.77 990 66 <1 0.3 14 7. X

6/30/68 183 8.91 1040 62 <0.5 <0.4 13 8. x

8/4/68 185 8.83 1040 66 <0.5 <0.3 13 7. X

1/7/69 194 8.88 965 63 <0.2 <0.3 12

3/2/69 200 8.96 1640 64 1.2 <0.3 10 7. X

4/6/69 203 8.99 1070 64 1.4 <0.3 9 8. X

7/6/69 206 9.02 1090 61 <0.4 <0.3 7 9. X

8/3/69 208 9.10 1058 62 0.9 0.3 6 8. X

8/31/69 210 9.07 993 62 0.7 0.34 17 9.08 x

9/14/69 211 9.0 1040 62 1.0 <0.3 8

10/9/69 212 9.01 1040 62 1.0 <0.3 8 1.03

11/4/69 214 9.01 1030 63 1.3 <0.3 8 1.02

12/4/69 216 8.99 890 64 0.7 <0.3 8 9.83

SPECIAL EQUIPMENT ROOM

IN REACTOR BLDG

FLOOR ELEV. 852 ft

40 ft FROM REACTOR BLDG.

ORNL-LR-DWG. 87i82R2

v

2-in.-DIAM. VENT LINE
TO FILTERS AND STACK~

GROUND ELEV.
851 ft- 6 in._~

T ELEV. 858 ft

| it

- GAS \
RETENTION |

ELEV. 848 ft i = 1 - LA
10-ft-DIAM. x 66 f1-LONG
{2-in.-DIAM. 3200-13
BURSTING DISK RELIEF LINE t
=== RELIEF VALVE
\ ety — ~ 100§t = 1 el |
{ 4-in. LINE -~ =" _
ELEV. 836 ft - ol —
X X «@ TO STACK s
| L Sok— | VAPOR __ELEV. 830+t
- CONDENSING | AVEN=TE
{0-ft DIAM.x 23- ¢t HIGH SHALE
FROM REACTOR CELL
BUTTERFLY ELEV. B24 ft
VALVES

Fig. 10.1. Diagram of MSRE vapor-condensing system.

1
Samples obtained from the vapor condensing tank were
analyzed at approximately semiannual intervals. The
results of these analyses are listed in Table 10.4. It may
be noted from the last entry in Table 10.4 that the final
analysis of the vapor condensing system water reservoir
was made more than one year before termination of the
reactor experiment. The reason is implicit in the results
listed, which show adequate stability of the corrosion
inhibitor during the entire period water was stored in
the reservoir tank.

10.2 Helium Cover Gas

Helium was used to fill voids in the salt circulation
systems with an inert protective gas. Commercial
helium, supplied from tank trucks, was passed through
a purification system to reduce the oxygen and water
content to below 1 ppm before the gas was admitted to
the reactor. The helium supply was passed continuously
through the fuel pump bowl at ~200 ft3/day (STP) to

127

transport the fission product gases to activated charcoal
absorber beds. The cover gas system was also used to
pressurize the drain tanks to move molten salts into the
fuel and coolant systems. Exhaust gas from these
operations was discharged through the charcoal beds
and filters to the atmosphere via the off-gas stack.

Incoming helium gas was passed through hot titanium
sponge and then through a moisture analyzer to ensure
that the purity of the gas was not degraded by inleakage
of moisture at faulty connections. Samples of the
helium supply were obtained from each new trailer of
the gas delivered to the reactor site and were analyzed
using mass spectrometric methods. The upper limits of
the allowable concentrations of contaminants in the gas
prescribed in design and operations specifications were
set at <1 ppm oxygen and <6 ppm water. Results of
the mass spectrometric analyses performed with
samples of the helium gas supply (Table 10.5) showed
that the supply was consistently of adequate quality for
use in the MSRE.

Table 10.4. Chemical analysis of water samples from the MSRE vapor condensing system

Parts per miilion

Date Sampie pH -
Li Na K Al B Fe NG, 3, hardness
2/7/66 V(CS-1 8.70 0.1 6.3 935 10 82 <0.3 590 68
8/31/66 VCS-3 9.00 1.1 6.3 748 15 70 <0.3 690 97
5/27/67 V(CS-4 8.86 1.1 6.6 800 1.0 63 <0.3 690 101
5/24/68 VCS-5 8.77 6.4 6.6 758 <1.0 61 <0.1 560

Table 16.5. Purity of helium cover gas for the MSRE as measured by mass spectrometric methods

Date Laboratory Weight percent
y No. He H, H,0 N, + CO 0, Ar CO,
10/28/65 6639 99.998 <0.0004 0.0003 0.0010 0.0002 <0.0001 0.0001
1/3/66 6806 99.981 0.0004 <0.0001 0.0187 <0.0001 0.0001 <0.0001
1/5/66 6827 99.977 0.0004 0.0002 0.0222 0.0004 0.0001
3/25/66 7050 99.990 0.0006 0.0003 0.0084 0.0001 <0.0001 <0.0001
6/2/66 7253 99,995 0.0004 0.0001 0.0059 0.0001 0.0001 0.0001
8/9/66 7456 99.996 <0.0001 0.0001 0.0033 0.0001 0.0002 0.0001
10/18/66 7665 99.996 0.0002 <0.0001 0.0035 0.0004 0.0002 0.0001
12/22/66 7865 99.998 0.0003 <0.0001 0.0009 0.0002 0.0001 0.0001
3/8/67 9410 99.998 <0.0002 <0.0002 <0.0001 0.0001
8/8/68 99,997 <0.0001 <0.0007 0.0018 0.0004 0.0002 0.0001
8/15/68 99,997 0.0001 0.0009 0.0017 0.0002 <0.0005 <0.0001
9/16/68 473 99.997 0.0001 0.0006 0.0019 0.0003 0.0002 <0.0001
4/11/69 1192 99,995 0.0004 0.0026 0.0015 0.0002 0.0001 0.0001
6/17/69 1438 99.998 0.0003 0.0002 0.0014 0.0001 0.0001 <0.0001
10/10/69 1812 99.995 0.0005 0.0001 0.0034 0.0007 0.0002 0.0001
10/28/69 1952 99.995 0.0003 <0.0001 0.0016 0.0003 0.0024 0.0001

“Each entry represents the supply from one trailer (39,000 f13) charge as delivered io the MSRE site.
10.3 Reactor Cell Air

Composttion of the reactor cell atmosphere was
adjusted to consist primarily of nitrogen contaiming less
than 5% oxygen Oxygen was maintained at this low
concentration to mummize the possihiity of com-
bustion hazards which mght occur if lubricating o1l
leaked from the fuel salt circulation pump system
Nitrogen was added to the cell as needed to make up
for air inleakage The normal cell pressure was 12 7
psia, with a maximum permussible air inleakage rate of
042 scth Under these conditions the required mitrogen
purge rate was 1 5 scfh Samples of the atmosphere of
the reactor cell were obtamned periodically and sub-
mitted for mass spectrometric analysis in order to
ensure that the specifications were met Results of these
analyses are histed in Table 10 6

Table 10.6. Chemucal analyses of MSRE reactor cell

air samples
Cell awr vol %
Date
sample No 0, CO, H,0

4/19/66 2 152
5/8/55 3 501
5/24/66 4 235
5/30/66 5 126
6/12/66 6 253
6/21/66 7 209
6/30/66 8 229
7/3/66 9 234
7/10/66 10 29
7/17/66 it 2 36
7/25/66 12 188
1/4/67 13 945
1/5/67 14 667
1/9/67 15 47
8/2/67 16 70 0 09
12/14/67 18 36 002
1/4/68 19 263 011 329
1/23/68 21 304
2/5/69 23 292

At ternination of operations with the MSRE, in-
spection of the reactor cell components was initiated,
the prelitminary results of which indicated that the
machinery in the cell remamned relatively free of
atmospheric corrosion The details of these exami-
nations were not complete at the time this report was
made and will appear subsequently in progress reports
of the Molten Salt Reactor Program

128

10.4 Oil Lubrication Systems

Each of the two salt pumps 1n the MSRE was served
by an independent lubrication o1l system, inter-
connected so that either system could serve both pumps
i an emergency These o1l systems supphed both oil for
lubrication of bearings and seals, as well as for cooling
pump radiation shields Details of the o1l circulation
systems are described elsewhere * Specifications of the
oil to be used in the MSRE were recommended by
Grindell 1n 1960 4

The MSRE employed a turbine-grade paraffinic-base
lubricating o1l having a viscosity of 66 SSU at 100°F,
commercially available as Gulfspin-35 No attempis
were made to characterize the lubncating o1l com-
pletely during reactor operations, for in these oper-
ations the principal concern was that iis lubrnicating
properites would not be degraded in use Composition
and physical property analyses were performed rou-
tinely to provide this assurance The composition
analysis conformed with the average specifications for
the product, for Gulfspin-35 analytical specifications
fixed the carbon content of the paraffinic hydrocarbon
at 854 wt % The average formula of the paraffinic
compound conforming to thas fraction 1s C3oHqg

Little was known about the identity of the additives
i the MSRE lubricating o1l, except that nominally they
were thermally stable through the operating temper-
ature range to which the o1l would be subjected in
MSRE wusage Analyses normally requested were for
carbon, sulfur, moisture, total solids, bromine number,
acid number, flash point, viscosity at 100 and 210°F,
mfrared and spectrographic analyses In operation, an
acid number of 0 06 was to be maintained as indicative
of oxidation-free performance, the interfacial tension
was not to exceed 18 dynes/cm at 77°C The results of
these analyses are summanzed in Table 10 7

It was recognized that mmgration of the fission gases
from the pump tank to the region of the shaft lower
seal via the shaft annulus could polymerize o1l as 1t
leaked down the shaft, plugging the drain hne The
pump was designed accordingly with a flow of purge gas
down the shaft annulus to minimize this migration ° It
was concluded that the oil to be used in the MSRE
could withstand radiation doses of as lugh as 107 rads/g
without significant degradation of 1ts properties and
that this amount of radiation would still permt the oil
to flow from the o1l catch basin Although this
conclusion, which proved to be valid, led to the
expectation of satisfactory performance of the MSRE
lubrication o1l 1 service with components that were
exposed to highly radioactive fluxes, early experience
129

Table 10.7. Properties of MSRE lubricating oil as measured in laboratory tests

Sample Viscosity Bromine Interfacial Composition (wt %) Total Flash Acid
Date Desig.a (Centistokes) Number Tension Carbon Sulfur Water Solids Point Number
38°c 99°C (dyne/cm) (ppm) (°C)
8/25/66 1-N 1.95 86,37 1.13 < 0,04 - 163
8/25/66 2=N 1.86 86.43 1.01 <0.,04 - 158
8/25/66 3-N 1.67 85.68 1.67 <0,04 - 158
8/25/66 4-N 1.56 85.52 0.98 <0.04 - 162
8/25/66 5=-N 1.86 85,82 0.92 <0.04 - 163
3/27/66 16~C 1.78
4/3/66 17=C 1.35
4/3/66 18-F 1.40
4/15/66 19-F 1.26
4/15/66 20-C 1.24
4/18/66 21-C 9.60 1.18
4/18/66 22-F 9,22 1.27
4/29/66  23-F 9,32 1.48
4/29/66 24-C 9,89 1.44
5/1/66 25-F 8.65 1.42
5/1/66 26-F 9, 89 1.39
5/8/66 27«F 8,43 1.47
5/8/66 28-C 10.05 1.44
5/24/66 29=C 10.77 1.67
5/24/66 30-C 10.13 1.64
6/3/66 31-F 0.03 0.78
6/3/66 32-C 0.04 0.81
6/3/66 33-N 0.03 0.61
6/21/66 34-C 9,48 1.38
6/21/66  35-F 9,20 1.33
6/26/66 36-C 10.60 1.58
6/26/66 37-F 10.70 1.59
7/3/66 38~F 10.14 1.76
7/3/66 39-C 10,33 1.86
7/10/66 40-F 10.25 1.50
7/10/66 41-C 1.40
7/17/66 42-F 10.13 1.10
7/17/66 43-F 10.71 1.00
7/24/66 44~F 10.51 2.77 1.70 10.63 85.04 0.032 0.20 2120 165 0.83
7/24/66 45-C 10.27 2.07 1.60 10.56 85.28 0.045 0.15 2076 166 0.88
8/12/66 46-TF 10.00 2.60 17.35 .92
8/8/66 47-N 9.94 2.64 1.26 17.46 86,88 0.012 0.16 1872 155 .93
8/16/66 48-C 10.00 2.27 17.55 0.93
8/13/66 49-N 1.30 18.32 86.21 0.029 0.017 1874 156 0.96
10/17/66 50-F 10.16 2.60 16.40 G.33
i0/17/66 51-C 10.06 2.58 16.59 0.30
10/25/66 52-C 17.5 0.26
10/25/66 53-F 17.10 0.29
10/30/66 54-F 17.18 0.79
16/30/66 55-C 16.80 0.60
11/10/66 56~F 16.50 0.57
1i/11/66 57-C 16.75 Q.55
12/18/66 58-C 15,80 1.01
12/18/66 59-F 15.490 0.83
12/27/66 60-C 15.8L 0.90
12/27/66 61~F 16.97 0.85
1/3/67 62-C 16,08 1.32
1/3/67 63-F 16.13 1.30
1/9/67 64-C 15.70 1.30
1/9/67 65-F 16.00 1.30
2/2/67 66-C 9.93 2.64 15.70 1.40
2/2/67 67-F 16.00
2/10/67 68-F 15.60 1.4
2/13/67 69-C 9.94 2.64 1.90 15.40 86.21 0.10 0.06 161 1.21
2/13/67 70-F 10.26 2.67 1.70 16.00 86.60 0.07 0.09 164 1.15
3/2/67 71-C 9.88 2.63 15.70 1.40
3f/2/67 72-F 10.20 ].66 15.80 1.26
3/21/67 75-C 14.88 1.05
3/21/67 76-F 15.07 i1.01
3/27/67 77-C 14,40 1.28
3/27/67 78-F 15.70 1.15
413167 80-C 14.60 1.25
4/3/67 81-F 15.60 1.21
4/10/67  82-C 9.97 2.60 14. 40 1.41
4/10/67 83-F 10.40 2.65 2.06 14.90 85.90 0.025 <0.05 158 1.27
130

Table 10.7 (continued)

Sample Viscosity Bromine Interfacial Composition (wt %} Total Flash Acid
Date Desig.a (Centistokes) Number Tension Carbon  Sulfur  Water Solids Point Number
38°¢C 99°¢C (dyne/cm) (ppm) (°C)

4/17/67  84-C 11,30 1.30
4f/17/67  85-F 14,20 1.30
4/24/67  86~C 14.80 1.31
4/24/67  81-F 16.10 1.24
5/1/67 88-C 15,80 1.35
5/1/67 89-F 14,40 1.40
5/7/67 90-C 15.10 1.39
5/7/67 91-F 16.80 1.30
7/3/67 95-N 15.60 1.30
1/3/67 96-N 16.20 1.36
7/10/67 97-C 6.01 2.57 16.43 1.30
7/10/67 98-F 10.15 2,57 15.50 1.30
7/17/67  99-C 10.01 2.56 15.80 1.31
7/17/67 100-F 10.16 2.57 15.80 1.31
7/24/67 101-C 15,60 1.38
1/24/67 102-F 15.90 1,29
7/31/67 103-C 15.83 1.36
7/31/67 104-F 15.05 1.43
8/7/67 105-C 15.66 1.52
8/7/67  106-F 15,47 i.42
8/13/67 107-C 9.92 2.55 i.96 11.78 83, 3¢ 0.011 0.013 1809 157 1.12
9/18/67 108-F 15.95 1.30
9/18/67 109-C 16.14 1.30
9/25/67 110-F 16.50 1.38
9/25/67 111-C 16.50 1.40
10/2/67 112-F 16,50 1.01
10/2/67 113-C 16.30 . 1.11
10/9/67 1l4-F 16. 4G 0.84
10/9/67 115-C le. 30 1.03
10/16/67 116-F 17,10 1.14
10/16/67 117-C 17.10 1.21
10/23/67 118-F 16.10 1.22
10/23/67 119-C 15.90 1.40
10/30/67 120~-F 15.64 1.38
10/30/67 121-C 15,17 1.46
11/6/67 122-C 15.47 1.35
11/6/67 123-F 14,81 1.23
11/13/67 124~-F 15.84 1.27
11/13/67 125-C 15.94 1.35
11/27/67 126-F 16.19 1.15
11/27/67 127-C 15.62 1.35
12/10/67 128-C 15.61 0.83
12/10/67 129-F 15.04 .85
12/19/67 130-C 14,60 1.31
12/19/67 131-F 15.36 1.23
1/1/68  132-F 16.33 1.36
1/1/68  133-C 15.48 1.51
3/2/69 146-C 13.70 1.11
3/2/69 147-F 16.10 0.86
3/16/69 148-F 15.80 ) 0.96
4/6/69  149-F 15.80 1.08
4/6/69  150-C 13.70 1.30
4/20/69 151-F 15.60 1.00
4/11/69 152 1G.55 2.67 1.69 - 13.00 85.60 0.016 0,090 1780 164 1.06
4/11/69 153 9.90 2.64 0.63 11.00 85.70 0.045 0.080 1640 159 0.91
5/19/69 154~F 9.94 2.58 1.29 14.15 78.00 0.012 0.015 1800 163 1.43
5/19/69 155-C 10.52 2.67 1.55 16.41 76.70 0.050 0.020 1800 167 1.20
9/14/69 156-C 9.93 2,73 0.63 12.78 86.30 0.019 0.002 1730 168 1.10
9/14/69 157-F 16.19 2,73 1.10 15.89 86.30 0.021 0.006 1810 175 0.90
9/14/69 158-F 9.80 2.60 1.11 16.10 85.90 0.021 0.003 1840 180 0.81
10/9/69 159-F 15.45 0.38
16/9/69 160-C 10. 14 1.14
10/19/69 161-F 10.58 3.80 1.35 15.60 85.40 0.018 0.002 1940 174 0.88
12/4/69 167-F 15.90 1.02
12/4/69 168-C 13.40 1.22

a
letter designations with sample numbers denote source of sampl = i
ple, N = unused oil from storage, C = sample from c
system, F = sample from fuel system. 8e P oolant
with the MSRE in power operations showed that o1l
vapors transported to restricted areas could certainly
polymerize (see Sect 11 1) We had no foreknowledge
as to whether polymenzation of the recirculated o1l
mught occur, although we faced this prospect with
trepidation because of the formation of varnishes from
oll vapors in the off-gas valves This fear seems to have
been unwarranted, for throughout the operation of the
MSRE the properties of the o1l apparently remained
unchanged, both as judged on the basis of performance
of the components in which 1t was used as well as from
the results of the analyses histed in Table 10 7

On one occasion, at the time when hydrocarbon
varmshes were plugging the off-gas stream valves, the
lubricating o1l was fractionally distilled and the re
fractive index of each 10% volume cut was measured
The range was found to include refractive indices from
14726 to 14817, as compared with the refractive
index of the undistilled o1l of 1 4738

Surveillance analysis included infrared absorption
tests to determune whether oxidation was occurnng
Occasionally these tests indicated that C-O bands were
present, but these bands were only slightly more intense
than in the reference sample and seemed not to become
more so with continued use of the oil

Among the least well-understood chemical phe-
nomena governing behavior 1n the MSRE and, as well,
among the least tractable for experimental investigation
with the MSRE were the chemical relationships in-
volving o1l from the MSRE fuel pump and the transport
and distribution of fission products and trtium within
the reactor system The properties of the oil, as
determuned through the chemical monmitoring program,
disclosed little that was significant concerning these
matters Unquestionably, the thermal degradation prod-
ucts formed in the pump bowl were of some signif-
1cance 1 the transport behavior of fission products and
trittum  Resolution of their exact effects on reactor
operations, however, will be derived from future nvesti-
gations rather than from further scrutiny of MSRE
behavior

References

1 R H Guymon, MSRE Design and Operations
Report, Part VIII, Operating Procedures, ORNL-TM-
908, vol 1,p 3C-1

2 R C Robertson, MSRE Design and Operations
Report, Part I, ORNL-TM-728, p 444

3 R C Robertson, ibid , p 144

4 A G Grndell, Spectfication of a Lubricant for
MSRE Pumps, ORNL-MSR-60-37

131

5 A G Guondell and P G Smuth, MSR Program
Senuannu Progr Rep July 31, 1964, ORNL-3708, p
155

11 TRANSPORT OF MATERIALS FROM SALT TO
COVER GAS SYSTEMS

11 1 Fission Products

In solid fuel reactors, knowledge of the modes of
fission product distribution 1s of utmost importance to
the assessment of reactor safety and control and for the
development of reprocessing methods Changes 1n state
and distribution of fission product species in sohd fuel
matrices have been examined so intensely and for so
long a period of time that only problems related to
accident conditions remain as challenging to the under-
standing For molten-salt reactors 1t i1s necessary that an
even more comprehensive understanding of fission
product behavior be achieved, for with fission products
intentionally circulated with fuel salt, their disposition,
as controlled by mechanical and chemucal influences, 15
of considerably greater importance than in static fuel
systems In MSRs, some fission products are removed
from the fuel salt continuously and become distributed
to the off-gas system, to the fuel storage tanks, and to
the processing plant They must be contained and their
decay heat removed under all concewable circum-
stances

The probable fates of fission products produced n
molten-salt reactors were deduced from thermodynamic
considerations, from laboratory studies, and from 1n-
pile static capsule tests From the results of these
studies, one may group fission product elements in
categonies according to general chemical properties
noble gases Kir, Xe, noble metals Pd, Rh, Ru, Mo, Ag,
Te, Tc, Nb, Se, As, transition metals Sn, Sb, In, Cd,
Ga, Zr, Zn, lanthamides Y, La, Ce, Pr, Nd, Pm, Sm, Eu,
Gd, Tb, Dy, Ho, Er, active metals Cs, Rb, Ba, Sr,
halides Br, I

Of the elements produced n high yields, only the
lanthanides developed concentrations in the MSRE that
were n the range of chemical detection This group
accounts for 53 8% of the fission products Operation
of the MSRE resulted 1n the fissioning of 2 88 kg of
235U fuel and 132 kg of 233U, a total of 1792
gatoms of uranmum From fission, therefore, 9 64
g atoms of lanthanides, about 15 kg, were produced
ultimately to produce a final concentration of 300 ppm
of these elements in the fuel salt At the very low
concentrations which prevailed in the fuel salt, the
lanthanmides were of little consequence with respect to
general chemical surveillance, and no special effort was
made to monitor their concentrations or distributions
directly One method for their removal from MSR
carrier salts relies on high-temperature distillation Tests
of this separation method with MSRE fuel carrier salt
disclosed that the measured concentrations of the
lanthanides 1n the salt agreed well with computed
values !

The MSRE provided the first opportunity to begin
studies of fission product transport in cwculating salt
systems Although some preliminary information was
derived from the results of experiments with the
Aircraft Reactor Experiment, its scheduled pernod of
operation was so bnef as to obviate any real opportu-
mty for such studies Recogmition that the knowledge
of fission product behavior was 1in a rudimentary
condition required that programs for their investigation
proceed independently from that for general chemical
survelllance Detailed results of the studies of fission
product behavior are therefore excluded from the
present report and will be described separately ?

11.2 Restrictions in the Off-Gas System

In both the fuel and coolant systems, control of the
pressure 1n the cover-gas systems over extended periods
proved to be difficult and unpredictable The cause was
the accumulation of solids and tarry deposits, mostly
originating from lubricating oil, in the off-gas throttling
valve used for fuel-system pressure control and in the
filter just upstream of the coolant-system pressure-
control valve

During the imtial period of circulation of coolant salt
in the reactor, a period of 1200 hr, the off-gas filter
plugged and was replaced twice Examnation of the
filter medium showed that i1t was covered with amor-
phous carbon contamning traces of the constituents of
the coolant salt and Hastelloy N Although the filter
was replaced by another one with some 35 times the
surface area, pressure control again became erratic It
was realized that oil seepage took place down the shaft
of both the fuel pump and the coolant pump Rates of
seepage were determined and found to be varnable,
ranging from indicated rates of zero to 4 cm® of o1l per
day seeping 1nto the pump bowl

The rates did not seem to increase appreciably from
tune to time, nor was their variation clearly related to
any operational practice It became routine to replace
the filter in the coolant salt off-gas system after several
months’ use Each time the filter was replaced, 1t was
found to be covered with o1l residues

132

Pressure control of the fuel system became erratic
toward the end of the first extended operation with salt
in the circuit Durning this peniod, ®3Kr was used for
tests of the stripping efficiency of the noble gas from
the hehum stream After saturation of the salt to near
equiltbrium concentrations, the wmflow of ®°Kr was
stopped and the gas purged from the system As one
part of the tests, the effect of salt level in the pump
bowl was examined with the LiF BeF, (66 34 mole %)
salt at three levels At this time the void fraction of the
salt in the fuel loop was measured for the first time and
found to be 06% At termination of these tests the
off-gas filter was removed and examined It was found
to be free of deposits Examination of the pressure
control valve showed that 1t was partially plugged The
residue was washed out with acetone The hquid
suspension was dark in color, and the residue left after
evaporation of the acetone consisted of spheres of a
glassy matenal, 1 to 5 u 1n diameter Shortly after the
cleaned control valve was reinstalled, 1t became partially
restricted agamn 1t was then replaced with another one
of greater flow-through capacity Inspection of the
original valve revealed that a black deposit partially
covered the tapered stem The deposit was about 20%
amorphous carbon and the remainder was again 1- to
5-u spheres that were found to have the composition of
the flush salt

The program for bringing the MSRE to full-power
operation was interrupted on two occasions by flow
restrictions in the off-gas system Symptoms of plugging
became pronounced in each case after about 12 hr of
operation at 1 MW Restrictions were first noted in the
capilary restrictor in Iine 521, at the same time, check
valve HCV-533 became 1noperative Restriction was also
noted tn the stainless steel line filter and 522 valve
assembly In order to resume operation of the reactor,
the capillary restrictor and HCV-533 check valve were
replaced by nonrestrictive sections, and the filter-valve
assembly was replaced The filter element which was
removed was capable of blocking passage of more than
90% of the particulate matter of 20 7 u in diameter It
was replaced by a filter designed to block particulates
of <50 u1n diameter

Continued difficulty was experienced with restric-
tions 1n the off-gas system of the fuel circutt for some
time during the early stages of power operation with
the MSRE These difficulties led to a practice which
relieved excess pressure 1n the fuel system by venting
gas through HCV-533 to the auxiliary charcoal bed
During this period, evidence of restrictrons in the
charcoal beds began to appear On resuming operation
at 1 MW, restnictions appeared to develop in three
locations at valve 522B, at the entries to charcoal beds
1A and 1B, and m the lines ahead of the auxiliary
charcoal bed Small specimens of the matenals which
were suspected to have caused plugging were removed
from the affected parts of the off-gas system and
subjected to laboratory tests in an attempt to determine

133

the reasons for flow restrictions The results of these
examinations are summarized in Table 11 173

The results shown in Table 11 1 indicate that the
restriction 1n the off-gas system may be attributed to
varnish-like organic materials whose origin was the
Gulfspmn-35 lubnicating o1l The refractive index of the

Table 11.1. Results of examinations of spectmens removed from the MSRE off-gas lines

Specimen description
and orgin

Morphology

Deposit from exiat orifice of
capallary restrictor

Scrapings from spool piece
adjacent to capulary

Depostit from check
valve 533

Scrapings from valve 522
poppet

Scrapings from valve 522
seat

Ol drops from valve 522
body

Scrapings from stainless
steel filter element

Metallic scrapings from
stainless steel filter element

Deposit from HV-621 valve
stem

Isotropic particles,
appearing as
partly coalesced
amber globuies

Sameas 1

Isotropic, amber,
varnish hke
particle, ~50 X
100 1

Isotropic, amber,
varnish hike
matrix contain
g embedded
1sotropic(?)
crystalline
matenal of
lower refractive
index

Same as 4

Isotropic, faintly
colored matenal,
more nearly
scalelike than
glassy 1n ap
pearance

Granular opaque
particles, low

index, transparent,
birefringent crystal-

line matertal
spalled off metal
on MICTOSCOpe
slide

Isotropic, faintly
colored materal,

varnish like 1 ap-
pearance with peb
bly surface

Activity at Predominant
Refractive ¢ lv; Y ta 1sotopes Spectrochemical
index C((;{n/}?f) (from gamma data
scan)
1520 11, 99 ug, Be,
124 ug, 71,
100 pg
1540
Be, 095 ug, L1,
2ug, Zr,<05
ng
~1 540 25 No Zr, Nb, Ce
1 544 to 15 89gr, 1903,
1550 1407 4
1509 No Zr, Nb,
Ce
1524 to 140p, 14074
1526 103Ru,
137(33, no
Ce, Zr, Nb
~1526 1321¢

heaviest 10% volume fraction obtained on vacuum
distillation® of the oil (see Sect. 10.4) was found to be
approximately 1.50. The high refractive index of the
varnish-like materials removed from the off-gas system
suggests that oil vapor and fractionation products
carried into the off-gas were polymerized in transit to
and on the surfaces of the valve parts, and thus plugged
the system. This inference is strengthened further by
the observation that the varnish-like materials were
found to have limited or no solubility at room
temperature in xylene, petroleum ether, carbon tetra-
chloride, or acetone.

Typical examples of the deposits removed from the
MSRE and examined in the laboratory are shown in
Figs. 11.1-11.4. In one attempt to identify the
materials that were causing the restrictions, the reactor
was drained, and for several days after, all helium flow
through the fuel loop was stopped® and the contained
helium was recirculated at 100°F. Three samples of the
recirculated gas were isolated for analysis, but the
results were inconclusive,

The appearance of restrictions in the off-gas system of
the fuel circuit comprised a major operational impedi-
ment. Considerable effort was expended to identify the
causes of these restrictions and to alleviate the con-
comitant problems. In April 1966 a particle trap was
installed in the fuel system off-gas line, just downstream
from the first holdup section. It was replaced in August
1966 with an identical unit. In January 1967 an
improved unit was used to replace the original one and

PHOTC 28500

Fig. 11.1. MSRE valve HCV-621 poppet.

134

remained in service throughout the remainder of opera-
tions with the MSRE. Each of the particle traps
employed stainless steel mesh as the roughing section;
in addition, finer media, both fibrous metal (Feltmetal)
and Fiberfrax, served to block passage of the smallest
particles. It served to alleviate the major part of the
difficulties that had beset previous operations, but
because it was not capable of completely blocking the
passage of volatile hydrocarbons, it was ineffective for
providing complete freedom from the development of
restrictions in the charcoal beds. The system was
operable throughout the remainder of MSRE tests,
although with chronic indications that oil fractionation
products were being transferred downstream to the
charcoal beds.

A particle trap that had been installed in the fuel
system off-gas line in April 1966 was examined, and
samples were removed from the trap for analysis. The
results of the examination of this trap were described in
detail by Scott and Smith.* The chemical results are
summarized below.

Mass spectrometric analysis of material leached from
Yorkmesh matting at the inlet section of the filter
indicated that it had collected a number of fission
products. The eluted material contained 20 wt % Ba, 15
wt % Sr, and 0.2 wt % Y. In the same analysis the salt
constituents Be and Zr were estimated to be 0.01 and
0.05 wt % respectively. Other samples removed from
the metal filter contained small quantities of Cr, Fe,
and Ni. Results of gamma-ray spectrometric analysis
indicated the presence of '37Cs, ®°Sr, !°3Ru or
106Ry, P10 Ag 95Nb, and '*°La. All samples were
analyzed for determination of the presence of Be, and
the concentration present was below the detectable
limit of 0.1%. Attempts to determine the presence of Zr
were complicated by the presence of large quantities of
Sr. The fraction of soluble hydrocarbons was deter-
mined for the materials present on the Yorkmesh mat
and in other samples was found to range from 60 to
80%. Carbon disulfide extracts were made of the
organics and were allowed to evaporate; the residues
were used for infrared analysis. Al samples were
identical and indicated the presence of long-chain
hydrocarbons. There was no evidence of any functional
groups other than those involving carbon and hydrogen,
nor was there any evidence suggesting double or triple
bonds. There was an indication of a possible mild
cross-linkage. It is likely that there is more cross-linkage
of the organic in the gas stream than appeared in these
samples, and the low indication could be due to the
insolubility of the cross-linked organic and the high
operating temperatures of the wire mesh, which would
135

PHOTO 82614

sl o
&»T%%fr

#
e

Fig. 11.2. MSRE sample 8-1. Fragment of materal removed from hine 521 pressure equalizing capillary Photographed in 1414
refractive index ol 206X
136

PHOTO 82607

o
o
e

o

el

Fig. 11.3. MSRE sample 8-3. Fragment of material removed from line 521 pressure equalizing capillary. Photographed in
refraction index oil. 645X . Reduced 20.5%.

1.520
137

™~
0
©
3\
oo
O
[
o
I
o

e

Dol

o B

T

1ter Refractive index

Fragment of translucent birefringent material removed from stainless steel fi

2¢.

MSRE sample 8-

4

11.

Fig.
more than 1

468. 160X.

B
138

cause breakdown of the organic into elemental carbon
and volatiles.

The following conclusions can be drawn from the
results of the examination:

1. Since the spectrographic analysis indicated that the
concentrations of Be and Zr were very small
compared with those of the fission products Ba and
Sr, the amount of entrained salt mist carried to the
filter was negligible.

2. The high activity and the large amount of barium
and strontium in the inlet section of the particle trap
indicate that a large fraction of the solid daughters
of Kr and Xe which decay in the line is carried down
the line with the off-gas stream.

3. The disiribution of activity indicates that entrance
areas of the Yorkmesh mat were unexpectedly
effective in trapping the solid fission products.
Decay heat in this area resulted in temperatures
above 1200°F.

4. The collection of hydrocarbon mist on the York-
mesh mat had probably enhanced the collecting
efficiency of the mesh for solid particles.

With the adoption of methods to minimize the effect
of hydrocarbon transport in the off-gas streams, little
further operational difficulty was encountered with the
MSRE, and no additional chemical studies were made.

The chemical effects of continually passing hydro-
carbon degradation products through the fuel stream
are largely unknown. They were viewed as intractable
to quantitative assessment with the means at hand for
dynamic analysis. In attempts to gain some under-
standing of the possible relation of the hydrocarbons to
various chemical and physical aspects of fuel salt and
fission product behavior, a gas sampling facility’ was
developed and installed at the MSRE. In use, it
provided few data that were of value in appraising the
effect of oil seepage and was adapted to other uses.
Thermal degradation of the pump oil certainly pro-
duced a small finite concentration of hydrogen in the
pump bowl atmosphere. There is no evidence in the
chemical results obtained from salt analyses that indi-
cates a recognizable effect of its presence. Possible
consequences of its effects on fission product chemistry
and transport were recognized some time ago. Their
assessment may be found in the summary report on
fission product behavior.®

11.3 Tritium Transport in the MSRE

Shortly before scheduled termination of operations
with the MSRE, it was realized that the hot metal

containment system should be highly permeable to
hydrogen and that, therefore, the tritium produced in
the fuel salt could be expected to diffuse out of the fuel
system. An investigation was initiated to measure the
amounts of tritium in the air vented to the discharge
stack, that released through the radiator tubes, and the
amount of tritium in fuel salt samples. Equipment was
installed by the analytical chemists in the MSRE
vent-house sample station in order to measure the
concentrations in the various reactor effluent gas
streams. A fraction of the gas stream was passed over a
bed of hot copper oxide in order to convert the tritium
to water. Tests were conducted with the copper oxide
beds either at 340 or 800°C. The higher temperature
allows the oxidation of methane, while the lower
temperature does not. The results of these tests,
conducted by Dale, Apple, and Meyer,” showed that
the maximum amounts of tritium released from the air
stack, from the radiator, and found in fuel salt,
corresponded to the production of 4.6, 0.6, and 22.7
Ci/day respectively. At full power operation the MSRE
produces 60 Ci of tritium per day.

In an evaluation of the results, Briggs'® concluded
that the disparity in the material balance suggested that
the following behavior may have occurred:

1. Either the metal walls have a much greater resistance
to the diffusion of tritium than has been estimated,
or most of the tritium is present in a form other
than T, or HT. Even assuming that graphite is a sink
for T,, one cannot otherwise explain the small
fraction that passes through the radiator tubes and
the piping and vessel walls in the primary system.

2. When TF is taken into consideration but reactions
with the graphite are neglected, the fraction of the
tritium that leaves the fuel pump off-gas system
increases and the amount that passes through the
metal walls decreases. Assuming equilibration of
liquid and gas in the pump bowl and UF,/UF; =
1000, the distribution begins to look like that
measured for the MSRE. One would have to further
assume that much of the tritium is retained on the
carbon beds as TF or organic compounds and that
T, and TF react with materials in the off-gas system
to produce the organic compounds found in the exit
of the carbon beds. These calculations contain
several uncertainties and simplifications that can
considerably affect the results; the results are in-
sensitive to the solubility of TF but are quite
sensitive to the solubility of T, in the salt and to the
value of the equilibrium constant, k. The presence of
hydrogen and organic compounds from decompo-
sition of oil in the pump bowl could be important
139

but is neglected. Agreement between calculations
and measurements may be entirely accidental.

3. When retention or reaction with the graphite in the
core is considered along with the presence of TF, the
graphite becomes a major sink for tritium, mostly as
TF. If UF,/UF; increases from 100 to 1000, almost
all the tritium enters the graphite as TF. Very little
diffuses through the metal walls or through the fuel
pump off-gas system. A distribution about like that
observed for the MSRE would be calculated if one
assumed partial reaction of TF with graphite to
produce hydrocarbons that are released from the
graphite and removed through the fuel pump off-gas
line. Such reactions are believed to be highly
unlikely.

Briggs’ assessment thus shows that the transport
mechanisms for tritium in the MSRE were not clarified
from the data obtained and that inevitably their
characteristics must be established because of their
significance to design of and procedures for operational
control of MSBRs. Termination of operations with the
MSRE precluded further assessment of transport mech-
anisms under realistic conditions. This was not neces-
sarily unfavorable, for the leakage of oil from the fuel
pump shaft seal into the pump bowl was unques-
tionably a complicating factor in attempts to interpret
the results of on-site analysis. The fuel pump was
known to have permitted small amounts, generally
believed to have been ~1 g/day, of lubricating oil to
enter the fuel pump bowl and thus become rapidly, and
perhaps heterogeneously, thermally degraded. It seems
very probable that radiation also must have had some
role in the degradation process. Generation of a
spectrum of hydrogenous degradation products in the
off-gas must complicate theoretical models of transport
mechanisms to a great extent and probably even
accounts for the wide range of analytical values
obtained at the MSRE. An initial model of tritium
behavior was advanced recently by Strehlow.!! As
fundamental to the development of a model of tritium
transport, Strehlow noted that “transport processes and
chemical behavior of hydrogen isotopes are interrelated
parameters in molten-salt reactor design considerations.
The transport of hydrogen is determined by the
solubility and diffusivity in moderator, salt, and metal.
The chemical behavior is a function of the oxidation-
reduction potential in the salt, radiation field, and
possibly reactivity with the graphite moderator.”

Additional tests were conducted at the MSRE near
the end of operations in which CuQ specimens and Ni
rods were exposed to the gas and salt before and after

the addition of beryllium to the salt. The results of
these tests have not yet been completed. After termi-
nation of the MSRE, one of the graphite stringers was
removed for examinations and tests. These tests, con-
ducted by S.S.Kirslis and F.F. Blankenship, were
completed in July 1971. The results indicated that
about 15% of the tritium produced in the MSRE was
sorbed on the moderator graphite. The concentration of
tritium in thin surface samples was as high as 10! dis
min~! g7!, fell to about 10° dis min™' g~! at 60 mils
depth, and then decreased slowly to about 6 X 10® dis
min~! g~! for penetrations to 800 mils. These results
are of particular interest in that they suggest a means
for controlling the distribution of tritium in molten-salt
reactors.

References

1. J. R. Hightower, Jr., and L. E. McNeese, MSR
Program Semiannu. Progr. Rep. Feb. 28, 1970, ORNL-
4548, p. 317.

2. F. F. Blankenship, E. G. Bohlmann, S. S. Kirslis,
and E. L. Compere, Fission Product Behavior in the
MSRE, ORNL-4684 (in preparation).

3. These and other related examinations were made
by T.N.McVay (consultant) and the author in the
High-Radiation-Level Analytical Facility (HRLAF).

4. Performed by L. J. Brady et al., Analytical
Chemistry Division, ORNL..

5. MSR Program Semiannu. Progr. Rep. Aug. 31,
1965, ORNL-3872, p. 26.

6. D. Scott and A. N. Smith, MSR Program Semi-
annu. Progr. Rep. Feb. 28, 1967, ORNL-4119, p. 47.

7. MSR Program Semignnu. Progr. Rep. Feb. 28,
1966, ORNL-3936, p. 67.

8. F. F. Blankenship, E. G. Bohlmann, S. S. Kirslis,
and E. L. Compere, Fission Product Behavior in MSRE,
ORNL-4684 (in preparation).

9. J. M. Dale, R. F. Apple, and A. S. Meyer, MSR
Program Semiannu. Progr. Rep. Feb. 28, 1970, ORNL-
4548, p. 183.

10. R. B. Briggs, Additional Calculations for the
Distribution of Tritium in the MSRE, ORNL-CF-71-7-8
(July 1971).

11. R. A. Strehlow, MSR Program Semiannu. Progr.
Rep. Feb. 28, 1970, ORNL-4548, p. 167.

12. IMPLICATIONS OF THE MSRE CHEMISTRY
FOR FUTURE MOLTEN SALT REACTORS

The Molten-Salt Reactor Experiment was spectac-
ularly successful as a materials demonstration: the
molten fluoride fuel and coolant-salt mixtures flowed
through their containment circuitry for thousands of
hours, causing little or no corrosion in these systems;
the graphite moderator experienced no measurable
dimensional changes in use and remained free of
penetration by the fuel salt. The reactor equipment
operated reliably, and the radioactive liquids and gases
were contained safely. The fuel was completely stable.
Xenon was removed rapidly from the salt. When
necessary, radioactive equipment was repaired and
replaced in reasonable time and without overexposing
maintenance personnel.

That the MSRE was completed so successfully did not
.come about as the result of frequent adjustments in the
chemical properties of the circulating fluids. A major
conclusion that emerges from experience with the
MSRE is that chemical surveillance, as carried out with
individual samples obtained and removed from the
reactor and transferred via necessarily cumbersome
processes to off-site laboratories, is of value principally
as a basis for attempting to control trends in behavior.
Irrespective of the quality of the analytical data
obtained, the information feedback process is generally
too slow to afford frequent control direction. Chemical
surveillance for future MSRs coupled to steam plants,
therefore, must be obtained through on-line monitoring
systems, and such surveillance systems must function
dynamically. In addition, for safety, incorporation of
the associated on-line instrumentation that is required
must not add significantly to the complexity of the
reactor system.

The next molten-salt reactor which is likely to be
built will be constructed either exclusively as a federal
enterprise, for example, the MSBE, involving both a
steam generation system and on-line chemical reprocess-
ing for the removal of rare-earth fission products, or a
larger reactor developed partly under the impetus of
private indusiry. Such a reactor would generate steam
but would not include a chemical reprocessing plant for
the removal of rare-earth fission products. Either of
these reactors would utilize a fuel salt containing the
same constituents as have been proposed for the MSBR
fuel, for which the standard reference design composi-
tion is ”LiF-BeF,-ThF,-UF, (72-16-11.7-0.3 mole %).
Heat removal would be accomplished through a second-
ary salt system. Final choice of the coolant for this
system has not yet been made for the reasons described
below.

It is essential that the tritium produced from fuels
based on ’LiF-BeF, be managed so that essentially
none of the tritium produced in the reactor core can
find its way into the steam plant. Complete removal of

140

tritium from the fuel sait before it passes into the heat
exchanger is highly improbable, and accordingly, man-
agement of the tritium transferred through the heat-
exchanger walls is required.

Fluoroborate mixtures were proposed as suitable for
molten-salt reactor coolants' and have been under
examination for some time for that application. Consid-
eration was given to the possibility that the NaF-NaBF,4
eutectic mixture would be substituted for the LiF-BeF,
coolant salt in the MSRE before operations with the
reactor were terminated. However, the plan was not
implemented for lack of financial support.

A barrier to the selection of fluoroborates as suitable
coolants for molten-salt reactors has been their seeming.
inability to retain hydrogenous species in melts con-
tained in nickel-based alloys. Unless a finite but small,
~50 ppm, concentration of hydrogenous species can be
retained in these melts as a reservoir for TF, their
applicability seems tenuous. As an alternative, a reactor
design which employs Hitec [NaNO;-NaNQ,-KNO;
(6.9-48.5-44.6 mole %)] as the coolant salt, isolated
from possible contact with fuel salt by an LiF-BeF,
buffer salt, has been proposed. It is necessary to
eliminate the possibility that Hitec might come into
contact with the graphite moderator. Evaluations of
this alternative have begun but will require a con-
siderable research and development effort before the
potential applicability of Hitec can be judged.

It is thus evident that the next molten-salt reactor to
be built, of whatever type, will employ both fuel and
coolant salts unlike those which were used in the
MSRE. Even so, the chemistry of molten salts, as
advanced by experience with the MSRE, is sufficiently
similar to be of practical value.

Until specific choice of a coolant is made, it seems
unnecessary to project from MSRE experience the
character of surveillance programs for the coolant
system. It is most likely that the earliest indications of
coolant-fuel mixing would be given by the nuclear
surveillance instrumentation and that, therefore, chemi-
cal surveillance of the separate systems would be of
little operational consequence. The details of coolant-
system surveillance must therefore be relegated to
future planning efforts.

At this point,.no timetable exists for construction of
follow-on reactors to the MSRE. Accordingly, as our
understanding of the chemical behavior of materials in
the MSRE is extended, it can serve to develop detailed
surveillance programs as needed. Fuel surveillance, at a
minimum, must include consideration of the following
chemical factors:
1. Oxide contaminants in the fuel salt. A fuel salt
which does not contain a highly polar cation compo-
nent which interacts chemically with the oxide ion, as
did Zr*" in the ZrF, contained in the MSRE fuel salt,
will almost certainly not tolerate relatively large con-
centrations of dissolved oxide ion without precipita-
tion. In the MSRE, the oxide tolerance of the fuel salt
was estimated to be ~680 ppm at the operating
temperature of the fuel salt, 650°C (1200°F), whereas
in the MSBR reference fuel salt the tolerance is
probably not greater than ~40 ppm at this temperature.
It has been tacitly assumed that all versions of
molten-salt reactors operated in the future would
include processing equipment for maintaining the con-
centration of oxide in the fuel salt at quite low levels
through continuous processing of a bleed stream of salt.

A detailed evaluation of the consequences of contami-
nation of fuel salt by oxides has been made, from which
it was concluded that ‘“there is no necessity for an
oxygen getter, such as ZrF,, in the MSBR fuel,” but
that “swift, satisfactory, and preferably on-line,
methods for determination of 0% and UF; concentra-
tions of the MSBR fuel after storage and during startup
and operation”? must be available.

Preliminary conceptual designs of on-line oxide detec-
tion apparatus have been devised by members of the
Analytical Chemistry Division.®> Increasing efforts will
undoubtedly be devoted to this activity in the future.

2. Corrosion detection. The extent of generalized
corrosion in the MSRE was estimated routinely from
changes in the concentration of chromium found in the
salt samples removed from the reactor. Results of
examinations of surveillance specimens removed occa-
sionally from the reactor core and from postoperational
examinations of the alloy removed from the heat
exchanger confirmed the general validity of these
estimates. The use of Cr?* concentration as a corrosion
indicator thus continues to be uniquely attractive. To
verify corrosion-free operation of a large reactor from
analyses of numerous samples removed from the reactor
would probably be limited by the slow feedback as
mentioned above. An alternate means would be far
more desirable. One generalization which emerges from
MSRE experience is that during periods when the
relative concentration of [U**]/[ZU] in the fuel salt
was 20.5%, the system appeared to be protected against
corrosion. This observation, along with a consideration
of the expense, inconvenience, and irrelevance of
individual results of chemical analyses, suggests that a
more suitable means for establishing that corrosion-free
conditions prevail would be to ensure that at all times
the proper redox potential existed in the salt stream.

141

As part of the MSRE experiment, evidence appeared
to indicate the possibility that the disposition of
niobium in the fuel salt could be exploited as a means
of monitoring corrosion in molten-salt reactors on a
continuous basis. The lack of success we have experi-
enced in attempting to determine the concentration of
trivalent uranium in the fuel salt when the total
concentration of uranium was <0.5 mole % suggests
that there is a clear incentive to develop methods of
detecting corrosion such as by niobium monitoring by
gamma spectrometry, which can provide on-line con-
tinuous monitoring of the chemical potential of the fuel
salt.

As noted in Chap. 6, several aspects of corrosion
chemistry in molten-salt reactor systems were left
unresolved at termination of MSRE operations. The
outstanding example of these was that corrosion attack
during ??°U operations might have been anticipated to
be some threefold greater than was observed. Further,
examinations of metal surveillance specimens removed
from the heat exchanger indicated that mass transfer of
metal from hot to cold zones was so little as to be
undetectable. Temperature differences in the salt cir-
cuits were modest, ~50°F, whereas in future molten-
salt reactors projected differences run to about 200°F.
The absence of base-line data for mass transfer that
results from these examinations obviated the possibility
of projecting mass-transfer coefficients for larger reac-
tors. Generalized corrosion was deduced from chemical
measurements of the chromium concentration in fuel-
salt samples.

Current developments in the attempts to improve the
radiation-induced loss of ductility that Hastelloy N
experiences after use in high radiation fluxes include
the possible inclusion of small amounts of titanium,
hafnium, or zirconium in the alloy.* All of these metals
are more chemically active than chromium, the most
chemically active constituent of the unmodified alloy;
their inclusion as constituents in significant concentra-
tions may affect the chemical methods for monitoring
corrosion on a dynamic basis.

3. Oils and hydrocarbons in the fuel system. The
extent of oil leakage into the fuel system must be
known accurately and ascertained on a regular basis
because of the operational implications of such leakage
and because of the chemical effects induced by the
presence of hydrocarbon degradation products in the
system. Hydrogen, evolved in the thermal degradation
of hydrocarbons, will unquestionably be involved in the
distribution of tritium in the reactor and will control
fission product chemistry, perhaps even favorably if, for
example, it enhances the retention of iodine in the
system as iodide or tellurium as the element. Surveil-
lance of the off-gas by means of a gas spectrometer
seems to be very desirable; it may possibly be an
important requisite for continuous operation. If the gas
analysis by mass spectrometry proves to be adaptable
for determination of a variety of species, it could
provide the most precise source of control data for the
reactor.

4. Miscellaneous. One of the most useful kinds of
analysis in the MSRE was that provided by mass
spectrometric measurements of the isotopic composi-
tion of uranium and plutonium. Occasional removal of
salt samples from reactors built in the future, princi-
pally for mass spectrometric determination of inven-
tory, burnup, etc., could also be used for confirmation
by general analyses that the on-line instrumentation was
functioning correctly.

Little advantage would accrue from any routine
efforts to determine chemically whether there were any
chronic leaks from fuel to coolant or in the reverse
direction, for, in either event, nuclear data would
provide operational control. Similarly, laboratory deter-
mination of the condition of auxiliary fluids, such as
cooling tower water, oil from pump reservoirs, etc.,
should probably not be done at all. Experience with the
MSRE has shown that there will be little difficulty in
maintaining these fluids in condition so they will meet
physical and chemical criteria over long periods of time.
Information pertaining to their condition can be ac-
quired simply and quickly and should be obtained at
suitable predetermined intervals by in-line automated
monitoring equipment.

In summary, it becomes evident that operational
adjustments for future molten-salt reactors should

142

originate from chemical controls provided almost ex-
clusively from on-line instrumentation. It is imperative
that such instrumentation be stringently restricted to
the minimum for safety, because there should be litile
reason to use the information derived for frequent
adjustments of the salt systems. These reduce to (1) a
continuous method for determining the identity of
species in and composition of the gas above the fuel
salt; in all likelihood this will entail the development of
an automated mass spectrometer; (2) a method of
establishing the redox potential of the fuel salt to
ensure corrosion-free operation; for this function, fur-
ther development of a niobium monitor by on-line
gamma spectrometry will be required; (3) verification
that bismuth does not enter the fuel stream via the
incoming stream from the chemical reprocessing plant;
(4) development of a highly sensitive and accurate
method for continuous or frequent on-line determina-
tion of oxide concentration in the fuel salt.

References

1. R. E. Thoma and G. M. Hebert, Coolant Salt for a
Molten Salt Breeder Reactor, US. patent No.
3,448,054, June 3, 1969.

2. W. R. Grimes, correspondence to Milton Shaw,
“Assessment of Need for Oxygen Getter in Molten Salt
Breeder Reactor Fuel,” February 3, 1971.

3. A. S. Meyer and J. M. Dale, personal communica-
tion.

4. H. E. McCoy et al., Nucl. Appl. Technol. 8, 156
(1970).
ORNL-4658
UC-80 — Reactor Technology

INTERNAL DISTRIBUTION

1. J. L. Anderson 41. D. E. Ferguson 82. Dunlap Scott

2. R.F. Apple 42. L. M. Ferris 83. J. L. Scott

3. C.F. Baes 43. A.P.Fraas 84. H. E. Seagren

4. C. E. Bamberger 44. J. H. Frye, Jr. 85. J. H. Shaffer

5. C.J. Barton 45. W. K. Furlong 86. M. J. Skinner

6. J. B. Bates 46. C. H. Gabbard 87. A.N. Smith

7. H. F. Bauman 47. R. B. Gallaher 88. G.P. Smith

8. S. E. Beall 48. L. O. Gilpatrick 89. A. H. Snell

9. M. J. Bell 49, W. R, Grimes 90. Din Sood

10. M. Bender 50. A. G. Grindell 91. R. A, Strehlow

11. E. S. Bettis 51. R. H. Guymon 92. D. A. Sundberg

12. D.S. Billington 52. P. H. Harley 93. J. R. Tallackson

13. F. F. Blankenship 53. P.N. Haubenreich 94. E.H. Taylor

14. E. G. Bohlmann 54. R. F. Hibbs 95--97. R.E. Thoma

15. G. E. Boyd 55. G. H. Jenks 98. L. M. Toth

16. J. Braunstein 56. E.M. King 99. D. B. Trauger

17. M. A. Bredig 57. A.P.Malinauskas 100. G. C. Warlick

18. R. B. Briggs 58. H. E. McCoy 101. C. F. Weaver

19. H. R. Bronstein 59. H. F. McDuffie 102. A. M. Weinberg

20. G. D. Brunton 60. H. A. McLain 103. J. R. Weir
21. S. Cantor 61. L. E. McNeese 104. M. E. Whatley
22. D. W. Cardwell 62. J. R. McWherter 105. J. C. White
23. R.S. Carlsmith 63. A.S. Meyer 106. R. P. Wichner
24, W_ L. Carter 64. R. L. Moore 107, L. V. Wilson
25. E. L. Compere 65. E. L. Nicholson 108. H. C. Young
26. W. H. Cook 66. L. C. Oakes 109. 1. P. Young
27. J.W.Cooke 67. R. B. Parker 110. J. W. Cobble (consultant)
28. L. T.Corbin 68. A.M. Perry 111. P. H. Emmett (consultant)
29. 1. L. Crowley 69. H. B. Piper 112. H. Insley (consultant)
30. F. L. Culler 70. B. E. Prince 113. E. A. Mason (consultant)
31. D. R.Cuneo 71. A.S. Quist 114. R. F. Newton (consultant)
32. J.M. Dale 72. G. L. Ragan 115. J. E. Ricci (consultant)
33. J.H. DeVan 73. J.D. Redman 116. C. H. Secoy (consultant)
34. J. R. Distefano 74. D. M. Richardson 117. H. Steinfink (consultant)
35. S. 1. Ditto 75. G. D. Robbins 118. L. R. Zumwalt (consultant)
36. F. A.Doss 76. R. (. Robertson 119-121. Central Research Library
37. A.S. Dworkin 77—78. M. W. Rosenthal 122. ORNL — Y-12 Technical Library
38. W.P. Eatherly 79. R. G. Ross Document Reference Section
39. 1. R. Engel 80. J.P. Sanders 123—157. Laboratory Records Department
40. R. B. Evans 111 81. H. C. Savage 158. Laboratory Records, ORNL R.C.

143
144

EXTERNAL DISTRIBUTION

159—160. MSBR Program Manager, AEC, Washington, D.C.
161. Merson Booth, AEC, Washington, D.C.
162. Paul Cohen, Westinghouse Elect. Corp., P.O. Box 158, Madison, Pa. 15663
163. D. F. Cope, RDT Site Office, ORNL
164. J. D. Corbett, Jowa State University, Ames, lowa 50010
165. J. W. Crawford, AEC, Washington, D.C.

166. F. E. Dearing, RDT Site Office (ORNL)

167. D. R. deBoisblanc, Ebasco Services, Inc., 2 Rector St., New York 10006
168. C. B. Deering, Black & Veatch, P.O. Box 8405, Kansas City, Mo. 64114
169. A. R. DeGrazia, AEC, Washington, D.C.

170. S. G. English, AEC,Washington, D.C,

171. J. J. Ferritto, Poco Graphite, P.O. Box 2121, Decatur, TX 76234

172. T. A. Flynn, Jr., Ebasco Services, Inc., 2 Rector St., N.Y., N.Y. 10006

173. E. E. Fowler, AEC, Washington, D.C.

174. J. E. Fox, AEC, Washington, D.C.

175. Norton Habermann, RDT, USAEC, Washington, D.C. 20545

176. A. Houtzeel, TNO, 176 Second Ave., Waltham, MA 02154

177. G. M. Kavanagh, AEC, Washington, D.C.

178. H. H. Kellogg, Henry Krumb School of Mines, Columbia U., N.Y., N.Y. 10027

179. M. Klein, AEC, Washington, D.C.

180. S.J. Lanes, AEC, Washington, D.C.

181. Kermit Laughon, AEC, RDT Site Office (ORNL) .
182. J. M. Longo, Esso Research & Eng., P.O. Box 45, Linden, N.J. 07036

183. T.W. Mcintosh, AEC, Washington, D.C. 20545

184. J. Neff, AEC, Washington, D.C.

185. R. E. Pahler, AEC, Washington, D.C.

186. A.J. Pressesky, AEC, Washington, D.C.

187. M. V. Ramaniah, Bhabha Atomic Research Centre, Radiological Labs., Trombay, Bombay-85, AS, India
188. David M. Richman, AEC, Washington, D.C.

189. H. M. Roth, AEC-ORO, Oak Ridge, Tn.

190. R. M. Scroggins, AEC, Washington, D.C.

191. M. Shaw, USAEC, Washington, D.C.

192. T. G. Schlieter, AEC, Washington, D.C.

193. J. M. Simmons, AEC, Washington, D.C.

194. S. A, Szawlewicz, AEC, Washington, D.C.

195. A. R. Van Dyken, AEC, Washington, D.C.

196. N. Srinivasan, Bhabha Atomic Research Centre, Trombay, Bombay 74, India

197. R.C. Steffy, Jr., TVA, 540 Mkt. St., Chattanooga, Tn. 37401

198. A. E.Swanson, Black & Veatch, P.O. Box 8405, 1500 Meadowlake, K.C., Mo.

199. J. A. Swartout, UCC, New York, N.Y. 10017

200. B. L. Tarmy, Esso Research & Engr. Co., P.O. Box 101, Florham Pk., NJ. 07923

201. J. R. Trinko, Ebasco Services, Inc., 2 Rector St., N.Y., N.Y. 10006

202. C. E. Larson, Commissioner, AEC, Washington, D.C.

203. W. W. Grigorieff, Oak Ridge Associated Universities -
204. Leo Brewer, Lawrence Radiation Laboratory

205. R. C. Vogel, Argonne National Laboratory

206. S. Spaepen, Head of Chemical Technology, SCK-CEN, MOL, Belgium -
207. V. K. Moorthy, Metallurgy Div., BARC, Bombay 83, India .
208
209

210.
211.
212.
213.
214.
215.
216.
217.
218.
219.
220.
221.
222--224.
225--226.
227-231.
232.
233.
234-452.

145

Penneman, Los Alamos Scientific Laboratory

Steunenberg, Argonne National Laboratory

Reser, American Ceramic Soc., 4055 N. High St., Columbus, Ohio 43214

Sidney Langer, Gulf General Atomic, San Diego, California

C. W. Keenan, Dept. of Chem., University of Tennessee, Knoxville

Gleb Mamantov, Dept. of Chem., University of Tennessee, Knoxville

Rustum Roy, Materials Res. Lab., Penn State Univ., University Park, Pa.

Minoo D. Karkhanavala, Chem. Div., BARC, Bombay, India

W. Danner, Max-Planck-Institut Fur Plasmaphysik, 8046 Garching Bei, Munchen, West Germany
David Elias, AEC, Washington, D.C.

Ronald Feit, AEC, Washington, D.C.

P. A. Halpine, AEC, Washington, D.C.

W. H. Hannum, AEC, Washington, D.C.

R. Jones, AEC, Washington, D.C.

Director, Division of Reactor Licensing (DRL), AEC, Washington, D.C.

Director, Division of Reactor Standards (DRS), AEC, Washington, D.C.

Executive Secretary, Advisory Committee on Reactor Safeguards, AEC, Washington, D.C.
Laboratory and University Division, AEC, ORO

Patent Oftice, AEC, ORO

Given distribution as shown in TID4500 under Reactor Technology category (25 copies — NTIS)

. RLA.
. R K.
M. K.